Академический Документы
Профессиональный Документы
Культура Документы
Handbook of
Neurochemistry and
Molecular Neurobiology
Development and Aging Changes
in the Nervous System
Volume Editors: J. Regino Perez-Polo and Steffen Rossner
Volume Editors
J. Regino PerezPolo Steffen Rossner
I. H. Kempner Professor and Chair Universitat Leipzig
Dept. Biochemistry & Molecular Biology Medizinische Fakultat
Senior Scientist PaulFlechsigInstitut fur Hirnforschung
Sealy Center for Molecular Medicine Jahnallee 59
Staff Scientist Shriners Hospital for Children 04109 Leipzig
Department of Biochemistry & Molecular Biology Germany
University of Texas Medical Branch Steffen.Rossner@medizin.uni-leipzig.de
301 University Blvd.
Galveston
TX, 77555
USA
regino.perez-polo@utmb.edu
ISBN: 9780387326702
Additionally, the whole set will be available upon completion under ISBN: 9780387354439
The electronic version of the whole set will be available under ISBN: 9780387304267
The print and electronic bundle of the whole set will be available under ISBN: 9780387354781
All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the
publisher (Springer ScienceBusiness Media, LLC., 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in
connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval,
electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is
forbidden.
The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as
such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.
springer.com
Animals share the challenge of maintaining an internal environment that is restricted to fairly low ranges of
temperature, pH, and water content within a well-protected envelope while engaged in continuous
exchanges with the environment in terms of gases, liquids, energy, even as movement of body parts and
the entire organism itself is necessary for survival. This dynamic spectrum of changes is further amplified
during developmental events or more acutely during responses to pernicious environmental factors in due
to trauma and disease. In addition, persistent incidents associated with aging can result in irreversible
changes to the allostasis that characterizes the living condition.
In the nervous system, a very high metabolic turnover, fragile but steep ionic gradients, and morpho-
logical and structural constraints dictated by the necessity for prompt neuronal transmission of electrical
impulses and necessary plasticity result in a highly fragile organ system.
Here we address a small sampling of major constituents of neural function at the cellular and molecular
level that play important roles in development and aging, two endogenous processes that embody features
of allostasis or the dynamic shifts in set points for specific homeostatic mechanisms associated with
development and aging.
The opening chapters discuss the major players in the neurotrophic hypothesis, the neurotrophins.
These growth factors have been shown to play a significant role during development and in the maintenance
of the adult cholinergic system in CNS as well as in the development of the sensory and sympathetic
nervous system. That they are also involved in plasticity events associated with memory and behavior points
to the degenerate nature of signaling molecules that achieve specificity by acting in concert as part of
ensembles of molecules rather than solitary regulators.
It is widely known that oligodendroglia and myelination events are late arrivals in the developmental
scheme of the brain and are also prime targets in early development of ischemic insults. Thus, a chapter on
oligodendroglia and myelination in development and aging serves to introduce these nonneuronal partners
vital to proper neuronal transmission. Molecular participants in stress responses to both acute and chronic
stressors are discussed from different perspectives in following chapters with varying degrees of emphasis
on injury versus normal aging and neurodegenerative disease.
The study of neural responses to stress of various kinds has led to a realization of the importance of
plasticity and the complexity of the mechanism allowing plasticity in the nervous system. The chapters
addressing the topic are followed by an introduction to the amyloid hypothesis, and what may be its central
character the enzyme held mostly responsible for the generation of beta amyloid. This is followed by a
broader discussion of misfolding proteins in the nervous system and its possible interventions to counteract
aging-associated deficits.
Limited in scope but offering a broad sampling, these chapters stress the dynamic features of neuronal
responses to internal (developmental) cues or the more harmful external events (injury and disease) in a
modern perspective.
Contributors
K. Abid M. V. Chao
George and Cynthia Mitchell Center for Molecular Neurobiology Program, Skirball Institute of
Neurodegenerative Diseases, Departments of Biomolecular Medicine, Departments of Cell Biology;
Neurology, Neuroscience and Cell Biology and Physiology and Neuroscience, New York University
Biochemistry and Molecular Biology, University of School of Medicine, 540 First Avenue, New York,
Texas Medical Branch 301 University Blvd, Galveston, New York 10016, USA
TX 77555, USA
J. de Vellis
J. M. Angelastro Department of Neurobiology, Mental Retardation
Department of Pathology and Center for Neurobiology Research Center, Semel Institute for Neuroscience and
and Behavior, Columbia University College of Physicians Human Behavior, David Geffen School of Medicine at
and Surgeons, New York NY 10032, USA UCLA, USA
T. Arendt L. A. Greene
Paul Flechsig Institute for Brain Research, Department Department of Pathology and Center for Neurobiology
of Neuroanatomy, 04109 Leipzig, Germany and Behavior, Columbia University College of Physicians
and Surgeons, New York NY 10032, USA
C. S. Atwood
School of Medicine, University of Wisconsin and X. Hu
William S. Middleton Memorial Veterans Administration, Department of Biolchemistry and Molecular Biology
Madison, Wisconsin 53705, USA and the Sealy Center for Molecular Science, University
of Texas Medical Branch at Galveston, Texas
R. L. Bowen 77555-1072, USA
Voyager Pharmaceutical Corporation, Raleigh, North
Carolina 27615, USA J. A. Joseph
Human Nutrition Research Center on Aging, Tufts
D. A. Butterfield University, 711 Washington St., Boston, MA 02111, USA
Department of Chemistry, University of Kentucky,
Lexington, KY 40506-0055, USA I. Konig
AG Molecular Mechanisms of Plasticity, Department of
V. Calabrese Neurochemistry & Molecular Biology, Leibniz Institute
Section of Biochemistry and Molecular Biology, for Neurobiology, 39118 Magdeburg, Germany
Department of Chemistry, Faculty of Medicine,
University of Catania, Catania, Italy M. R. Kreutz
AG Molecular Mechanisms of Plasticity,
G. Casadesus Department of Neurochemistry and Molecular Biology,
Institute of Pathology, Case Western Reserve University, Leibniz Institute for Neurobiology,
Cleveland, Ohio 44106, USA 39118 Magdeburg, Germany
S. F. Lichtenthaler H. Soreq
Adolf-Butenandt-Institut, Ludwig-Maximilians Department of Biological Chemistry, Institute of
University Munich, 80336 Munich, Germany Life Sciences, The Hebrew University of Jerusalem,
Israel 91904
E. Meshorer
Department of Biological Chemistry, Institute of
C. Soto
Life Sciences, The Hebrew University of Jerusalem,
George and Cynthia Mitchell Center for
Israel 91904
Neurodegenerative Diseases, Departments of
Neurology, Neuroscience and Cell Biology and
M. Mikhaylova
Biochemistry and Molecular Biology, University of Texas
AG Molecular Mechanisms of Plasticity, Department of
Medical Branch, 301 University Blvd, Galveston,
Neurochemistry & Molecular Biology, Leibniz Institute
TX 77555, USA
for Neurobiology, 39118 Magdeburg, Germany
J. Neman C. Spilker
Department of Neurobiology, Mental Retardation AG Molecular Mechanisms of Plasticity,
Research Center, Semel Institute for Neuroscience Department of Neurochemistry & Molecular Biology,
and Human Behavior, David Geffen School of Leibniz Institute for Neurobiology,
Medicine at UCLA, USA 39118 Magdeburg, Germany
D. B. Pereira A. M. G. Stella
Molecular Neurobiology Program, Skirball Institute of Section of Biochemistry and Molecular Biology,
Biomolecular Medicine, Departments of Cell Biology; Department of Chemistry, Faculty of Medicine,
Physiology and Neuroscience, New York University University of Catania, Catania, Italy
School of Medicine, 540 First Avenue, New York,
New York 10016, USA
L. Vergara
J. R. Perez-Polo George and Cynthia Mitchell Center for
Department of Biolchemistry and Molecular Biology Neurodegenerative Diseases, Departments of
and the Sealy Center for Molecular Science, University Neurology, Neuroscience and Cell Biology and
of Texas Medical Branch at Galveston, Texas Biochemistry and Molecular Biology, University of Texas
77555-1072, USA Medical Branch, 301 University Blvd, Galveston, TX
77555, USA
G. Perry
Institute of Pathology, Case Western Reserve University, K. M. Webber
Cleveland, Ohio 44106, USA Institute of Pathology, Case Western Reserve University,
Cleveland, Ohio 44106, USA
D. K. Rassin
Department of Pediatrics, The University of Texas
X. Zhu
Medical Branch at Galveston, Texas 77555-0344, USA
Institute of Pathology, Case Western Reserve University,
Cleveland, Ohio 44106, USA
S. Roner
Paul Flechsig Institute for Brain Research, Department
of Neurochemistry, 04109 Leipzig, Germany W. Zuschratter
AG Molecular Mechanisms of Plasticity,
M. A. Smith Department of Neurochemistry & Molecular Biology,
Institute of Pathology, Case Western Reserve University, Leibniz Institute for Neurobiology,
Cleveland, Ohio 44106, USA 39118 Magdeburg, Germany
Table of Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v
Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . x
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
1 Neurotrophins and Central Nervous
System Development
D. B. Pereira . M. V. Chao
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Abstract: The formation of the vertebrate nervous system is characterized by widespread programed cell
death, which determines cell number and appropriate target innervation during development. The neuro-
trophins, which include nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), neuro-
trophin-3 (NT-3) and NT-4, represent an important family of trophic factors that are essential for survival
of selective populations of neurons during different developmental periods. Neurotrophins exert their cellular
effects through the actions of two different receptors, the tropomyosin-related kinase (Trk) receptor tyrosine
kinase and the p75 neurotrophin receptor, a member of the tumor necrosis factor receptor superfamily. Much
attention has been given to the consequences of neurotrophin action in the peripheral nervous system (PNS);
however, neurotrophins are widely expressed in the brain and spinal cord. This chapter focuses on new views
concerning effects of neurotrophins in central nervous system (CNS) development.
1 Introduction
The neurotrophic hypothesis postulates that during nervous system development, neurons approaching the
same final target vie for limited amounts of target-derived trophic factor (Levi-Montalcini and Angeletti,
1968; Thoenen and Barde, 1980). In this way, the nervous system moulds itself to maintain only the most
competitive and appropriate connections. Competition among neurons for limited amounts of neurotro-
phin molecules produced by target cells accounts for selective cell survival. Several predictions can be
gleaned from this hypothesis. First, the efficacy of neuronal survival depends on the amount of trophic
factors produced during development. Second, specific receptor expression in responsive cell population
dictates neuronal responsiveness.
The neurotrophins are initially synthesized as precursors or pro-neurotrophins that are cleaved to
release the mature, active proteins. The mature proteins form stable, noncovalent dimers and are normally
expressed at very low levels during development. Pro-neurotrophins are cleaved intracellularly by furin or
pro-convertases utilizing a highly conserved dibasic amino acid cleavage site to release carboxy-terminal
mature proteins of approximately 13 kD (Chao and Bothwell, 2002). The mature proteins mediate
neurotrophin actions by selectively binding to members of the Trk family of receptor tyrosine kinases to
regulate neuronal survival, differentiation, and synaptic plasticity. NGF binds most specifically to TrkA;
BDNF and NT-4 to TrkB; and NT-3 to TrkC receptors (> Figure 11). Neurotrophin binding to Trk
receptors results in the activation of the phospholipase Cg (PLCg) pathway, involved in synaptic transmis-
sion and plasticity (Matsumoto et al., 2001; Minichiello et al., 2002), and in the recruitment of a number of
proteins that are involved in the activation of extracellular signal-regulated kinase (ERK), phosphatidyli-
nositol 3-kinase (PI3K) and its downstream target Akt. While ERK1/2 mediate neurotrophin-induced
neuronal survival and differentiation (Huang and Reichardt, 2003), PI3K activity is responsible for
retrograde survival signaling (Kuruvilla et al., 2000).
The p75 receptor can bind to each neurotrophin but has the additional capability of regulating a
Trks affinity for its cognate ligand. In addition, pro-neurotrophins can also interact with p75 independently
(Lee et al., 2001).
Neurotrophin factors fit well with the neurotrophic hypothesis, as many peripheral neuronal subpo-
pulations exhibit a predominant dependence on a specific neurotrophin during the period of naturally
Neurotrophins and central nervous system development 1 3
. Figure 11
Selectivity of neurotrophin binding to Trk receptors and p75. NGF binds most specifically to TrkA, BDNF and NT-
4 to TrkB and NT-3 to TrkC. NT-3 can also potentially interact with TrkB and TrkA receptors. In addition, all
neurotrophins are capable of binding to the p75 receptor with equal affinity
occurring cell death. However, in the CNS, there appears to be a much more complex situation. In the brain
and spinal cord, there is overlapping expression of multiple neurotrophin receptors and their cognate
ligands. This allows for the generation of diverse connectivity that extends well into adulthood. In addition
to promoting axonal and dendritic branching, neurotrophins also possess acute regulatory effects on
neurotransmitter release, synaptic strength and connectivity (Thoenen, 1995; Bonhoeffer, 1996; McAllister
et al., 1999). Many of these effects reflect activity-dependent events based upon competitive remodeling of
axon and dendritic terminals. Other complexities remain, such as the molecular mechanisms underlying
the retrograde signal, a pathway that must efficiently transmit information over long distances, at times over
a meter, as well as anterograde transport of neurotrophins.
The NGF family of trophic factors has provided a remarkable mechanism for controlling a wide gamut
of cellular activities, including target innervation, apoptosis, differentiation, synaptic plasticity, and axon
guidance. Defining the populations in the central nervous system (CNS) that are affected bidirectionally by
neurotrophins provides insights into the mechanisms responsible for regulating cell growth, survival, and
changes in complex circuits. This chapter considers some of the key decision-making events in different
populations of central neurons that lead to these diverse biological outcomes.
The broad distribution of TrkB, TrkC and their ligands throughout the CNS is suggestive of a putative role
for neurotrophins in CNS neuronal survival during development, in similarity to their action in the PNS
(peripheral nervous system). The result of initial studies showing that exogenously applied neurotrophins
can rescue specific populations of central neurons from naturally occurring and induced cell death, seemed
to support this idea. Among the neuronal populations that show increased survival when cultured in the
presence of neurotrophins are embryonic cerebellar granule cells (reviewed in Segal et al., 1992) and
substantia nigra dopaminergic neurons (Hyman et al., 1991), which are responsive to brain-derived
4 1 Neurotrophins and central nervous system development
neurotrophic factor (BDNF), cerebellar Purkinje cells (Cohen-Cory et al., 1991), which are sensitive to
NGF, and septal cholinergic neurons, which are responsive to both (Hartikka and Hefti, 1988; Alderson
et al., 1990). In vivo studies have also shown that exogenously applied NGF and BDNF or NT-3, can rescue
axotomized basal forebrain neurons (Williams et al., 1986) and motor neurons (Oppenheim et al., 1992;
Sendtner et al., 1992; Yan et al., 1992; Koliatsos et al., 1993), respectively. Despite what these pharmacologi-
cal studies have initially suggested, mouse mutants lacking neurotrophins or their receptors, which were
generated in the early-to-mid 1990s, have failed to confirm a significant neuroprotective role for neuro-
trophins in the CNS during embryonic development (reviewed in Snider, 1994).
background (Naumann et al., 2002). On the other hand, the same study reports a significant increase in the
cholinergic neuron number in the p75exonIV mutant mouse (Naumann et al., 2002).
The role of p75 in basal forebrain cholinergic neuron survival remains, therefore, unresolved. Moreover,
the involvement of neurotrophins and Trk signaling in the putative role of p75 in the survival of the
aforementioned neurons was not yet investigated. Clarifying the controversy regarding the two different
strains of p75 mutant mice will require conditional mice, together with crossing these animals with other
neurotrophin mutants. This will be crucial to the understanding of neurotrophin and p75 regulation of
central cholinergic neuron survival.
The TrkB/ mouse, which was the first neurotrophin receptor mutant to be reported, was initially
described to have a decrease of 35% in the number of spinal cord motor neurons (Klein et al., 1993).
This would be in agreement with the rescue of developmentally regulated or axotomy-induced motor
neuron death by BDNF (Oppenheim et al., 1992; Sendtner et al., 1992; Yan et al., 1992; Koliatsos et al.,
1993). However, two different strains of mice lacking BDNF failed to show any reduction in the survival of
several motor neuron populations (Ernfors et al., 1994a; Jones et al., 1994), including spinal cord neurons.
Mice deficient for the other two neurotrophins that can activate TrkB, NT-4 and NT-3, did not exhibit any
reduction in facial motor neurons numbers (Ernfors et al., 1994b; Farinas et al., 1994; Conover et al., 1995;
Liu et al., 1995). BDNF and NT-4 double mutants also have normal facial and lumbar cord motor neurons
pools (Conover et al., 1995; Liu et al., 1995). A reduction in motor neuron survival was only observed in
triple BDNF, NT-4 and NT-3 deficient mice, which show a modest 20% decrease in neuronal numbers in
several nuclei (Liu and Jaenisch, 2000). Finally, reexamination of the TrkB mouse mutants and analysis of
mice lacking both TrkB and TrkC showed no significant difference in the number of facial or spinal motor
neurons between these mutant mice and their wildtype littermates (Silos-Santiago et al., 1997). Therefore,
neurotrophins appear to play a redundant and modest role in the survival of motor neurons during
embryonic development. It is still possible, since not all motor neuron pools were examined in each animal,
that particular sets of motor neurons are dependent on specific neurotrophins for survival, as it is the case
for other trophic factors (Huang and Reichardt, 2001). As an example, the survival of g-motor neurons,
whose nerve fibers are present at the L4 ventral spinal root, is greatly reduced in the NT-3/ mouse
(Kucera et al., 1995).
The most dramatic reduction in neuronal survival in the CNS of neurotrophin or neurotrophin-receptor
mutant mice was observed in the mesencephalic trigeminal nucleus (MTN). The MTN, which is located in
the brainstem, contains the proprioceptive neurons of the trigeminal system that convey information from
6 1 Neurotrophins and central nervous system development
the head region. Both the BDNF/ and the NT-3/ mice show an approximate 50% reduction in the
number of neurons in this area (Ernfors et al., 1994a, b; Farinas et al., 1994; Jones et al., 1994). Double
BDNF/NT-3-deficient mice show a loss of 88% of these neurons and triple BDNF/NT-3/NT-4 mutants are
practically devoid of MTN neurons, with only about 5% of them surviving (Fan et al., 2000). Therefore,
unlike their PNS counterparts, which are completely dependent on NT-3 (reviewed in Snider, 1994; Huang
and Reichardt, 2001), different subsets of these CNS proprioceptive neurons seem to depend on different
neurotrophins for survival.
Most of the other regions of the brain were surprisingly normal in these mutant mice. No gross abnormality
was detected in the cytoarchitecture of the brain areas examined, including the cortex, hippocampus, and
thalamus, in TrkB and TrkC single mutants (Klein et al., 1993, 1994). TrkB/TrkC double mutants also
appeared to have normal brain anatomy (Silos-Santiago et al., 1997), ruling out redundancy as a possibility
for the lack of a phenotype. A more comprehensive examination was performed on the BDNF mutants.
With the exception of the granule cerebellar neurons, which exhibit increased cell death in these animals
(Schwartz et al., 1997), most of the neuronal populations known to be sensitive to this neurotrophin
showed no decrease in survival. This included substantia nigra and VTA dopaminergic neurons, basal
forebrain cholinergic neurons, cerebellar Purkinje neurons, and GABAergic and cholinergic neurons of the
hippocampus and striatum (Jones et al., 1994; Ernfors et al., 1994a). There was, however, a marked decrease
in the expression of certain neuronal markers, such as calbindin, parvalbumin, and neuropeptide Y (NPY;
Jones et al., 1994). This was interpreted as a problem in neuronal differentiation since GABAergic neurons,
where calbindin, parvalbumin, and NPY are expressed, are present in normal numbers in the BDNF-
deficient mice. Importantly, although their brain structure seems normal, BDNF/ brains are smaller and
have less neuropile (Conover et al., 1995). A reduction in the thickness of all cortical layers is also evident
(Jones et al., 1994).
Subsequent reports looking specifically for neuronal apoptosis in the postnatal brain of TrkB and
double TrkB/TrkC mutants uncovered a need for these receptors in the survival of several populations of
CNS postmitotic neurons. TrkB deficient mice show a progressive increase in the number of apoptotic cells
from P10 to P18, in all cortical layers and hippocampal subfields (Alcantara et al., 1997). The dentate gyrus
is the most affected area. In other brain regions, such as the striatum and the reticular thalamic nucleus, the
increase in apoptotic cells in the TrkB mutant is apparent only during the first postnatal weeks (Alcantara
et al., 1997). The majority of dying cells were confirmed to be neurons by the use of specific neuronal
markers. Mice deficient in both TrkB and TrkC also exhibit increased cell death in the hippocampus, after
the first postnatal week (Minichiello and Klein, 1996). Cerebellar granule neurons, but not Purkinje cells,
are also reduced in these animals, in similarity to what was observed in the BDNF mouse mutants
(Minichiello and Klein, 1996; Schwartz et al., 1997).
Conditional mouse mutants lacking TrkB at specific developmental time points and in specific cell types
have been generated using the Cre/loxP recombination system (Minichiello et al., 1999; Xu et al., 2000;
Medina et al., 2004). These mice have been a valuable tool to study the role of TrkB in the CNS of adult
mice. Two of these strains, where depletion of TrkB occurs either embryonically or during the second to
third postnatal weeks, show a severe reduction in the thickness of their visual and somatosensory cortex
(but Medina et al., 2004; see also Minichiello et al., 1999; Xu et al., 2000). Although it is likely that this
compression of the cortical layers is associated with migration and differentiation problems (Xu et al., 2000;
Medina et al., 2004), a decrease in neuronal survival was also observed in 10-week-old animals (Xu et al.,
2000). Interestingly, the survival of different neuronal populations, identified by the expression of specific
transcription factors, were differentially affected by the lack of TrkB (Xu et al., 2000).
Taken together, all the work involving conventional and conditional neurotrophin and Trk receptor
mouse mutants has revealed a disappointing small role for these molecules in the survival of CNS neurons
during embryonic development, as compared with their PNS counterparts. With the exception of the MTN,
Neurotrophins and central nervous system development 1 7
other brain regions seem to be dependent on neurotrophins for neuronal survival only throughout
postnatal development. As we will describe in the following sections, neurotrophins appear to have a
bigger contribution to CNS development by modulating migration and differentiation of CNS neurons.
Neuronal migration is an essential step in brain tissue formation during development, allowing the proper
positioning of neurons before the onset of neuritogenesis and synaptogenesis. To reach their appropriate
position, neurons undergo one or two different types of migration. In radial migration, neurons migrate
from their progenitor regions following the radial disposition of the neural tube (reviewed in Marin and
Rubenstein, 2003). This usually involves the use of radial glia processes, which extend from the ventricular
zone to the pial surface of the brain, as a guide on which neurons migrate until they reach their destination.
Radial glia-independent radial migration can also occur by a process called somal translocation, where
neurons themselves extend a process to the pial surface and then translocate their soma to the proper
position. The other type of migration, termed tangential migration, follows a path that is orthogonal to the
direction of radial migration (reviewed in Marin and Rubenstein, 2003). The physical substrates used by
cells migrating tangentially are other migrating neurons or growing axons. Alternatively, cells can just
disperse in an individual manner not following any cellular substrates.
Neurotrophins have been described to influence both types of neuronal migration by acting as
motogenic factors that stimulate the motility of migrating cells. In the following sections, we will focus
on studies performed on the cerebral cortex and the cerebellum, where the role of neurotrophins in
neuronal migration has been most intensively investigated.
Messenger RNA for the TrkB and TrkC receptors is present in high levels during cortical development, when
neuronal migration occurs (Klein et al., 1990; Lamballe et al., 1994). On the other hand, their ligands,
BDNF and NT-3, have opposite patterns of expression during corticogenesis: NT-3 expression is high
during embryogenesis and decreases with maturation while BDNF expression is low during embryonic
development increasing during adulthood (Maisonpierre et al., 1990; Timmusk et al., 1994). Surprisingly,
in spite of these expression patterns, it is BDNF and not NT-3 that seems to play a role in neuronal
migration in the neocortex. Both BDNF and NT-4 can induce chemotaxis (directed migration) of
embryonic cortical neurons in vitro in a manner dependent on Ca2+ and Trk activation (Behar et al.,
1997). In the same assay, NT-3 can only elicit migration when used at high concentrations (100 ng/ml).
In vivo studies have shown that increasing the concentration of BDNF in the developing embryonic brain,
8 1 Neurotrophins and central nervous system development
. Figure 1-2
Cerebral cortex organization during development and adulthood. Postmitotic neurons, originating from the
ventricular zone (VZ), migrate radially towards the surface of the brain to form the preplate (PP). Most of these
neurons migrate on radial glial processes represented in the diagram by vertical lines. For simplicity, the cell
body of these glial cells was not represented. The PP is then split into the marginal zone (MZ) and the subplate
(SP) by successive migrating neurons that form the cortical plate (CP) in between. Layers II-VI are then formed at
the CP by an inside-out migratory sequence. IZ, intermediate zone
either by intraventricular injection of the neurotrophin or by creating transgenic mice that overexpress
BDNF, produces neuronal heterotopias (collections of cells) in the MZ (Brunstrom et al., 1997; Ringstedt
et al., 1998; Alcantara et al., 2006). Injection of NT-4 has an even greater effect in the production of
heterotopias, as compared with the injection of BDNF, while infusion of NT-3 has no effect (Brunstrom
et al., 1997). This accumulation of cells in the MZ of the developing cortex appears to be the result of
increased neuronal migration and it is at least partially due to a decrease in the expression of reelin
(Brunstrom et al., 1997; Ringstedt et al., 1998). Reelin, which is produced by the CajalRetzius cells of
the MZ, is required for normal lamination of the cortex during embryonic development and is down-
regulated postnatally as maturation occurs (reviewed in Marin and Rubenstein, 2003). This downregulation
is dependent on BDNF since it is delayed in mice lacking this neurotrophin (Ringstedt et al., 1998).
Therefore, BDNF appears to act as a cortical maturation factor that, by being expressed in high amounts
ahead of time, induces cortical lamination defects. Does this imply, however, that endogenous BDNF plays a
role in migration in the neocortex?
Preliminary analysis of the BDNF-deficient mouse brain revealed a reduction in the thickness of the
cortical layers (Jones et al., 1994). A similar phenotype was seen in TrkB conditional mouse mutants where
TrkB expression is reduced during embryogenesis being virtually eliminated following birth (Medina et al.,
2004). These mice show a delay in migration of newly born cortical neurons that assume altered positioning
in the newly formed cortical layers (Medina et al., 2004). However, it is not clear whether the reduction in
TrkB expression during embryogenesis and early postnatal life, achieved in these conditional mutants, is
enough to reveal the full plethora of defects that the lack of this receptor would cause in the context of
neuronal migration. It would be interesting to study in more detail the brain architecture of embryonic and
early postnatal TrkB and BDNF/NT-4 double mutant mice to clarify the role of these neurotrophins in
neuronal migration.
In support of a role for endogenous TrkB ligands in migration to the appropriate cortical layer TrkB
expression is seen in migratory neurons in the CP in vivo, during the period of maximal neuronal migration
(Behar et al., 1997). In addition, a gradient of BDNF transcripts is observed in the neocortex, with the CP
Neurotrophins and central nervous system development 1 9
expressing higher amounts of this neurotrophin as compared to the ventricular zone from where cortical
neurons originate (Behar et al., 1997). This may be the reason why BDNF, and not the more highly
expressed NT-3, is important for neuronal migration. By being present in low amounts in discrete regions,
which allows the formation of a gradient, BDNF may direct the migration of TrkB-expressing neurons
toward regions of higher BDNF concentration.
Tangential migration of GABAergic interneurons into the cortex is also promoted by NT-4 and BDNF.
Exogenous application of these neurotrophins to brain slices increased the number of MGE neurons that
migrate into the cortex and the rate at which they migrate (Polleux et al., 2002). NT-3 was again without
effect on the migration of these interneurons. The role of endogenous neurotrophins in this process was
demonstrated by a reduction in the number of MGE neurons that migrate into the cortex in brain slices of
TrkB deficient mice and in slices treated with the Trk inhibitor K-252a (Polleux et al., 2002). Interestingly,
transactivation of Trk receptors by endocannabinoids has been reported to induce interneuron migration
in vitro (Berghuis et al., 2005).
The Trk signaling pathways involved in the modulation of cortical neuronal migration by neurotrophins
have only recently been studied. The PI3K pathway was suggested to mediate the effects of TrkB ligands
on radial migration as well as interneuron tangential migration, by pharmacological approaches (Polleux
et al., 2002; Yoshizawa et al., 2005). However, both the Shc-binding site, which leads to the activation of
PI3K and ERK, and the PLCg-binding sites on the TrkB receptor need to be mutated to induce a delay in
radial migration similar to that observed in conditional TrkB mouse mutants (Medina et al., 2004). Mice
with single point mutations at either site show normal CNS development (Medina et al., 2004). The PI3K
pathway was also suggested to lead to the activation of P-Rex1, a Rac guanine nucleotide-exchange factor
(GEF) that plays an important role in BDNF-induced migration of cortical neurons (Yoshizawa et al.,
2005). A dominant negative form of P-Rex1 was shown to inhibit the migration of cortical neurons
toward higher concentrations of BDNF in vitro and to give rise to radial migration impairments in vivo
(Yoshizawa et al., 2005).
Other molecules such as the Src-family kinases, which have been involved in Trk signaling and function
(Alema et al., 1985; Kremer et al., 1991; Tsuruda et al., 2004; Rajagopal and Chao, 2006), were recently
shown to be important for neuronal migration. Double mutant mice lacking both Src and Fyn show a defect
in cortical lamination similar to that observed in Reelin/ mice (Kuo et al., 2005). However, it is unknown
whether these molecules play a direct role in neuronal migration in the context of neurotrophin receptor
signaling. Future studies will be necessary to uncover the relative contribution of several signaling pathways
to neurotrophin-induced migration in the cortex.
form the internal germinal layer (IGL; Hatten, 1999). This leads to the establishment of the three layers of
the cerebellum: the outer molecular layer (ML), composed of granule cell axons and Purkinje cell dendrites,
the Purkinje cell layer and the IGL.
Neurotrophins appear to be important for the proper migration of cerebellar granule neurons during
development. Both TrkB and TrkC are expressed, although to various degrees, by the principal neurons of the
developing cerebellum (Klein et al., 1990; Ernfors et al., 1992). NT-3 mRNA levels peak in the IGL at the time
of granule neuron migration and decrease afterwards (Rocamora et al., 1993). Although BDNF mRNA was
not detected in the IGL in the first two postnatal weeks (Rocamora et al., 1993), when granule neurons
migrate radially to form that layer, BDNF protein was detected at P7, both in the IGL and the ML (Borghesani
et al., 2002). While this apparent controversy remains unsolved, it is interesting to note that BDNF protein
levels are low in the EGL, the starting point of granule neuron radial migration, and high in the ML and IGL,
where these neurons migrate to. Additionally, TrkB immunoreactivity in the developing cerebellum is highest
in migratory granule neurons when they are passing through the ML (Gao et al., 1995). Moreover, BDNF is
acutely motogenic for granule cells in culture and when these cells are exposed to a gradient of this
neurotrophin they migrate toward higher concentrations of BDNF (Borghesani et al., 2002).
A role for BDNF and TrkB in the migration of cerebellar granule neurons was confirmed in vivo by
studying the development of the cerebellum from BDNF and TrkB mouse mutants. The BDNF-deficient
mice exhibit an increase in the thickness of their EGL, at P14, while their IGL is sparser and thinner
(Schwartz et al., 1997; Borghesani et al., 2002). The EGL abnormally persists until later in development, in
BDNF mutants, being still present at P17-P21 (Jones et al., 1994; Schwartz et al., 1997). These phenotypes
are believed to be the result of a delay rather than a deficit in migration since at P23 the EGL is already not
apparent in the BDNF-deficient animals (Jones et al., 1994). Indeed, BDNF-deficient granule neurons show
a decreased migratory index: the number of cells that migrate is lower although the speed of migration is
similar to wildtype (Borghesani et al., 2002). Therefore, BDNF appears to be important for the initiation of
cerebellar granule neuron migration (Borghesani et al., 2002).
TrkB conventional and conditional mouse mutants also displayed a delay in granule cell migration, in
similarity to the BDNF-deficient mice (Minichiello and Klein, 1996; Rico et al., 2002). However, TrkB/TrkC
double mutants (TrkB/; TrkC+/ or TrkB+/; TrkC/) exhibit a more pronounced enlargement of
the EGL as compared with the single mutants (Minichiello and Klein, 1996). This suggests that both NT-3
and BDNF may contribute to the proper migration of cerebellar granule neurons. Analysis of NT-3
conditional mutant mice failed to detect a defect in the layering of the cerebellum (Bates et al., 1999).
Nevertheless, this study was performed at P8, when the layering of BDNF-deficient cerebellum is also not
yet different from the wildtype (Schwartz et al., 1997). On the other hand, the NT-3-deficient cerebellum
presents a similar foliation defect to that seen in BDNF mouse mutants, which may or may not be related
with a problem in migration (Schwartz et al., 1997; Bates et al., 1999).
In conclusion, neurotrophins appear to be important for the proper timing of neuronal migration in
both the cerebral cortex and the cerebellum. However, the overlapping effects of different neurotrophins
in these brain regions may be preventing the clarification of the exact role of these molecules in neuronal
migration. A more comprehensive analysis of conventional and conditional single and double neurotrophin
mutants could undoubtedly further clarify the promising role of neurotrophins in neuronal migration.
Soon after neurons reach their appropriate location, or even during migration, they extend neurites that
eventually differentiate into dendrites and axons. This polarization sets the stage for the formation of an
elaborate network of synaptic connections, in which neurite outgrowth, axonal guidance, and synaptogen-
esis take part. Although no evidence was yet presented for a role of neurotrophins in neuronal polarization,
these molecules are established regulators of dendritic growth and complexity while a role in axonal
guidance and synaptogenesis is also emerging (McAllister, 2001; Gillespie, 2003). On the last section of
this chapter, we will focus on the most documented aspect of neurotrophin function on central neuron
differentiation: the regulation of dendritic growth and complexity.
Neurotrophins and central nervous system development 1 11
Neurons in the CNS evolve from immature cells with just a few processes to bearers of highly complex
dendritic arbors, which integrate most of the afferent input received by these neurons. To achieve this level
of intricate dendritic morphology neurons appear to follow an intrinsic growth program, capable of
generating a basic dendritic arborization, and to respond to extrinsic signals that shape their dendritic
tree into a final mature form. Neuronal activity, in response to afferent neuronal connections, has been
regarded as a key factor in the regulation of dendritic morphology by controlling intracellular Ca2+
concentration. Alterations in the concentration of this ion can activate signaling pathways that may locally
induce cytoskeletal reorganization and direct the synthesis of new proteins needed for dendritic growth
(reviewed in Dijkhuizen and Ghosh, 2005b).
Neurotrophins and their receptors are highly expressed in the developing CNS during the time period
when dendritic growth is occurring (reviewed in McAllister et al., 1999). Moreover, neurotrophins are
secreted in response to neuronal activity and therefore appear as plausible candidates to mediate activity-
dependent dendrite plasticity (Ghosh et al., 1994; Blochl and Thoenen, 1995; Goodman et al., 1996;
Wang and Poo, 1997; Hartmann et al., 2001). Indeed, neurotrophins, and in particular BDNF, have
been extensively described as crucial regulators of dendrite number, length and branching (reviewed in
Dijkhuizen and Ghosh, 2005b; McAllister, 2001).
2003). NT-4 was also able to increase the same dendritic parameters on non-pyramidal interneurons in
cortical organotypic slice cultures (Wirth et al., 2003), while NT-3 was without effect (Baker et al., 1998).
However, it is not certain whether the differential effects of these neurotrophins were not due to different
experimental conditions, such as differences in the developmental stage of the culture, duration of
neurotrophin treatment, and brain region analyzed.
Other types of central neurons that show increased dendritic growth in response to neurotrophins
include rat Purkinje cells and mouse parvalbumin-positive interneurons of the olfactory bulb, which are
responsive to BDNF (Hirai and Launey, 2000; Berghuis et al., 2006), rat dopaminergic mesencephalic
neurons, which are affected by NT-4 (DeFazio et al., 2000), and mouse cerebellar granule neurons and
basilar pontine nuclei neurons that are responsive to both neurotrophins (Gao et al., 1995; Rabacchi et al.,
1999). An interesting effect of exogenous BDNF was reported in the Xenopus visual system, where
microsphere-coupled BDNF increases retinal ganglion cell (RGC) dendritic complexity in vivo when
injected in the optic tectum, their target area, while having the opposite effect when injected in the retina,
where RGC dendrites are located (Lom and Cohen-Cory, 1999; Lom et al., 2002). These results indicate that
local effects of neurotrophins in different subcellular domains of central neurons may exert distinct
outcomes on dendrite differentiation.
in 8-week-old mice. These mice also exhibit behavior abnormalities such as increased anxiety and decreased
contextual-dependent memory, in similarity to the BDNF heterozygous mice (Chen et al., 2006). These are
important findings since this SNP was found to decrease hippocampal-dependent memory in humans
(Egan et al., 2003), thus associating dendritic development deficits with a behavior-impairing human
polymorphism.
et al., 2006). Tiam1 is therefore a promising new TrkB effector, suggesting the existence of a simple signaling
pathway with a few intermediaries that closely links TrkB activation and actin cytoskeleton dynamics, which
are crucial for dendritic remodeling.
Acknowledgments
The authors are supported by grants from NIH (NS21072 and HD23315). D.B.P. is supported by a
postdoctoral fellowship from Fundacao Portuguesa para a Ciencia e a Tecnologia.
References
Adcock KH, Metzger F, Kapfhammer JP. 2004. Purkinje cell GABAergic neurons in the marginal zone plays a role in
dendritic tree development in the absence of excitatory the development of cortical organization. Cereb Cortex 16:
neurotransmission and of brain-derived neurotrophic factor 487-499.
in organotypic slice cultures. Neuroscience 127: 137-145. Alder J, Cho NK, Hatten ME. 1996. Embryonic precursor cells
Alcantara S, Frisen J, del Rio JA, Soriano E, Barbacid M, et al. from the rhombic lip are specified to a cerebellar granule
1997. TrkB signaling is required for postnatal survival of neuron identity. Neuron 17: 389-399.
CNS neurons and protects hippocampal and motor neurons Alderson RF, Alterman AL, Barde YA, Lindsay RM. 1990.
from axotomy-induced cell death. J Neurosci 17: 3623-3633. Brain-derived neurotrophic factor increases survival and
Alcantara S, Pozas E, Ibanez CF, Soriano E. 2006. BDNF- differentiated functions of rat septal cholinergic neurons
modulated spatial organization of Cajal-Retzius and in culture. Neuron 5: 297-306.
Neurotrophins and central nervous system development 1 15
Alema S, Casalbore P, Agostini E, Tato F. 1985. Differentiation Caviness VS, Jr. 1982. Neocortical histogenesis in normal and
of PC12 phaeochromocytoma cells induced by v-src onco- reeler mice: A developmental study based upon [3H]thy-
gene. Nature 316: 557-559. midine autoradiography. Brain Res 256: 293-302.
Altman J, Bayer SA. 1985. Embryonic development of the rat Chao MV, Bothwell M. 2002. Neurotrophins: To cleave or not
cerebellum. III. Regional differences in the time of origin, to cleave. Neuron 33: 9-12.
migration, and settling of Purkinje cells. J Comp Neurol Chen KS, Nishimura MC, Armanini MP, Crowley C, Spencer
231: 42-65. SD, et al. 1997. Disruption of a single allele of the nerve
Anderson SA, Marin O, Horn C, Jennings K, Rubenstein JL. growth factor gene results in atrophy of basal forebrain
2001. Distinct cortical migrations from the medial and cholinergic neurons and memory deficits. J Neurosci 17:
lateral ganglionic eminences. Development 128: 353-363. 7288-7296.
Ang ES, Jr., Haydar TF, Gluncic V, Rakic P. 2003. Four-dimen- Chen Y, Wang PY, Ghosh A. 2005. Regulation of cortical
sional migratory coordinates of GABAergic interneurons in dendrite development by Rap1 signaling. Mol Cell
the developing mouse cortex. J Neurosci 23: 5805-5815. Neurosci 28: 215-228.
Angevine JB, Sidman RL. 1961. Autoradiographic study of cell Chen ZY, Jing D, Bath KG, Ieraci A, Khan T, et al. 2006.
migration during histogenesis of cerebral cortex in the Genetic variant BDNF (Val66Met) polymorphism alters
mouse. Nature 192: 766. anxiety-related behavior. Science 314: 140-143.
Baker RE, Dijkhuizen PA, Van Pelt J, Verhaagen J. 1998. Cohen-Cory S, Dreyfus CF, Black IB. 1991. NGF and
Growth of pyramidal, but not non-pyramidal, dendrites excitatory neurotransmitters regulate survival and morpho-
in long-term organotypic explants of neonatal rat neocor- genesis of cultured cerebellar Purkinje cells. J Neurosci 11:
tex chronically exposed to neurotrophin-3. Eur J Neurosci 462-471.
10: 1037-1044. Conover JC, Erickson JT, Katz DM, Bianchi LM, Poueymirou
Bates B, Rios M, Trumpp A, Chen C, Fan G, et al. 1999. WT, et al. 1995. Neuronal deficits, not involving motor
Neurotrophin-3 is required for proper cerebellar develop- neurons, in mice lacking BDNF and/or NT4. Nature 375:
ment. Nat Neurosci 2: 115-117. 235-238.
Bayer SA, Altman J, Russo RJ, Dai XF, Simmons JA. 1991. Cell Crowley C, Spencer SD, Nishimura MC, Chen KS, Pitts-Meek S,
migration in the rat embryonic neocortex. J Comp Neurol et al. 1994. Mice lacking nerve growth factor display perinatal
307: 499-516. loss of sensory and sympathetic neurons yet develop basal
Behar TN, Dugich-Djordjevic MM, Li YX, Ma W, Somogyi R, forebrain cholinergic neurons. Cell 76: 1001-1011.
et al. 1997. Neurotrophins stimulate chemotaxis of embry- DeFazio RA, Pong K, Knusel B, Walsh JP. 2000. Neuro-
onic cortical neurons. Eur J Neurosci 9: 2561-2570. trophin-4/5 promotes dendritic outgrowth and calcium
Berghuis P, Agerman K, Dobszay MB, Minichiello L, Harkany currents in cultured mesencephalic dopamine neurons.
T, et al. 2006. Brain-derived neurotrophic factor selectively Neuroscience 99: 297-304.
regulates dendritogenesis of parvalbumin-containing inter- Dijkhuizen PA, Ghosh A. 2005a. BDNF regulates primary
neurons in the main olfactory bulb through the PLCgamma dendrite formation in cortical neurons via the PI3-kinase
pathway. J Neurobiol 66: 1437-1451. and MAP kinase signaling pathways. J Neurobiol 62:
Berghuis P, Dobszay MB, Wang X, Spano S, Ledda F, et al. 278-288.
2005. Endocannabinoids regulate interneuron migration Dijkhuizen PA, Ghosh A. 2005b. Regulation of dendritic
and morphogenesis by transactivating the TrkB receptor. growth by calcium and neurotrophin signaling. Prog
Proc Natl Acad Sci USA 102: 19115-19120. Brain Res 147: 17-27.
Blochl A, Thoenen H. 1995. Characterization of nerve growth Dreyfus CF. 1989. Effects of nerve growth factor on choliner-
factor (NGF) release from hippocampal neurons: Evidence gic brain neurons. Trends Pharmacol Sci 10: 145-149.
for a constitutive and an unconventional sodium- Egan MF, Kojima M, Callicott JH, Goldberg TE, Kolachana
dependent regulated pathway. Eur J Neurosci 7: 1220-1228. BS, et al. 2003. The BDNF val66met polymorphism affects
Bonhoeffer T. 1996. Neurotrophins and activity-dependent activity-dependent secretion of BDNF and human memory
development of the neocortex. Curr Opin Neurobiol 6: and hippocampal function. Cell 112: 257-269.
119-126. Ernfors P, Lee KF, Jaenisch R. 1994a. Mice lacking brain-
Borghesani PR, Peyrin JM, Klein R, Rubin J, Carter AR, et al. derived neurotrophic factor develop with sensory deficits.
2002. BDNF stimulates migration of cerebellar granule Nature 368: 147-150.
cells. Development 129: 1435-1442. Ernfors P, Lee KF, Kucera J, Jaenisch R. 1994b. Lack of
Brunstrom JE, Gray-Swain MR, Osborne PA, Pearlman AL. neurotrophin-3 leads to deficiencies in the peripheral
1997. Neuronal heterotopias in the developing cerebral nervous system and loss of limb proprioceptive afferents.
cortex produced by neurotrophin-4. Neuron 18: 505-517. Cell 77: 503-512.
16 1 Neurotrophins and central nervous system development
Ernfors P, Merlio JP, Persson H. 1992. Cells Expressing mRNA Horch HW, Kruttgen A, Portbury SD, Katz LC. 1999. Desta-
for Neurotrophins and their Receptors During Embryonic bilization of cortical dendrites and spines by BDNF.
Rat Development. Eur J Neurosci 4: 1140-1158. Neuron 23: 353-364.
Fan G, Copray S, Huang EJ, Jones K, Yan Q, et al. 2000. Huang EJ, Reichardt LF. 2001. Neurotrophins: Roles in neu-
Formation of a full complement of cranial proprioceptors ronal development and function. Annu Rev Neurosci 24:
requires multiple neurotrophins. Dev Dyn 218: 359-370. 677-736.
Farinas I, Jones KR, Backus C, Wang XY, Reichardt LF. 1994. Huang EJ, Reichardt LF. 2003. Trk receptors: Roles in neuro-
Severe sensory and sympathetic deficits in mice lacking nal signal transduction. Annu Rev Biochem 72: 609-642.
neurotrophin-3. Nature 369: 658-661. Hyman C, Hofer M, Barde YA, Juhasz M, Yancopoulos GD,
Gao WQ, Zheng JL, Karihaloo M. 1995. Neurotrophin-4/5 et al. 1991. BDNF is a neurotrophic factor for dopa-
(NT-4/5) and brain-derived neurotrophic factor (BDNF) minergic neurons of the substantia nigra. Nature 350:
act at later stages of cerebellar granule cell differentiation. 230-232.
J Neurosci 15: 2656-2667. Jin X, Hu H, Mathers PH, Agmon A. 2003. Brain-derived
Gascon E, Vutskits L, Zhang H, Barral-Moran MJ, Kiss PJ, neurotrophic factor mediates activity-dependent dendritic
et al. 2005. Sequential activation of p75 and TrkB is involved growth in nonpyramidal neocortical interneurons in devel-
in dendritic development of subventricular zone-derived oping organotypic cultures. J Neurosci 23: 5662-5673.
neuronal progenitors in vitro. Eur J Neurosci 21: 69-80. Jones KR, Farinas I, Backus C, Reichardt LF. 1994. Targeted
Ghosh A, Carnahan J, Greenberg ME. 1994. Requirement for disruption of the BDNF gene perturbs brain and sensory
BDNF in activity-dependent survival of cortical neurons. neuron development but not motor neuron development.
Science 263: 1618-1623. Cell 76: 989-999.
Gillespie LN. 2003. Regulation of axonal growth and guidance Klein R, Martin-Zanca D, Barbacid M, Parada LF. 1990.
by the neurotrophin family of neurotrophic factors. Clin Expression of the tyrosine kinase receptor gene trkB is
Exp Pharmacol Physiol 30: 724-733. confined to the murine embryonic and adult nervous
Goodman LJ, Valverde J, Lim F, Geschwind MD, Federoff HJ, system. Development 109: 845-850.
et al. 1996. Regulated release and polarized localization of Klein R, Silos-Santiago I, Smeyne RJ, Lira SA, Brambilla R,
brain-derived neurotrophic factor in hippocampal neu- et al. 1994. Disruption of the neurotrophin-3 receptor gene
rons. Mol Cell Neurosci 7: 222-238. trkC eliminates la muscle afferents and results in abnormal
Gorski JA, Zeiler SR, Tamowski S, Jones KR. 2003. Brain- movements. Nature 368: 249-251.
derived neurotrophic factor is required for the mainte- Klein R, Smeyne RJ, Wurst W, Long LK, Auerbach BA, et al.
nance of cortical dendrites. J Neurosci 23: 6856-6865. 1993. Targeted disruption of the trkB neurotrophin recep-
Greferath U, Bennie A, Kourakis A, Bartlett PF, Murphy M, tor gene results in nervous system lesions and neonatal
et al. 2000. Enlarged cholinergic forebrain neurons and death. Cell 75: 113-122.
improved spatial learning in p75 knockout mice. Eur J Kohara K, Kitamura A, Adachi N, Nishida M, Itami C, et al.
Neurosci 12: 885-893. 2003. Inhibitory but not excitatory cortical neurons require
Hartikka J, Hefti F. 1988. Development of septal cholinergic presynaptic brain-derived neurotrophic factor for dendritic
neurons in culture: Plating density and glial cells modulate development, as revealed by chimera cell culture. J Neurosci
effects of NGF on survival, fiber growth, and expression of 23: 6123-6131.
transmitter-specific enzymes. J Neurosci 8: 2967-2985. Koliatsos VE, Clatterbuck RE, Winslow JW, Cayouette MH,
Hartmann M, Heumann R, Lessmann V. 2001. Synaptic Price DL. 1993. Evidence that brain-derived neurotrophic
secretion of BDNF after high-frequency stimulation of factor is a trophic factor for motor neurons in vivo. Neuron
glutamatergic synapses. EMBO J 20: 5887-5897. 10: 359-367.
Hatten ME. 1999. Central nervous system neuronal migra- Kremer NE, DArcangelo G, Thomas SM, DeMarco M,
tion. Annu Rev Neurosci 22: 511-539. Brugge JS, et al. 1991. Signal transduction by nerve growth
Hatten ME. 2002. New directions in neuronal migration. factor and fibroblast growth factor in PC12 cells requires a
Science 297: 1660-1663. sequence of src and ras actions. J Cell Biol 115: 809-819.
Hirai H, Launey T. 2000. The regulatory connection between Kucera J, Ernfors P, Walro J, Jaenisch R. 1995. Reduction in
the activity of granule cell NMDA receptors and dendritic the number of spinal motor neurons in neurotrophin-3-
differentiation of cerebellar Purkinje cells. J Neurosci 20: deficient mice. Neuroscience 69: 321-330.
5217-5224. Kumar V, Zhang MX, Swank MW, Kunz J, Wu GY. 2005.
Horch HW, Katz LC. 2002. BDNF release from single cells Regulation of dendritic morphogenesis by Ras-PI3K-Akt-
elicits local dendritic growth in nearby neurons. Nat mTOR and Ras-MAPK signaling pathways. J Neurosci 25:
Neurosci 5: 1177-1184. 11288-11299.
Neurotrophins and central nervous system development 1 17
Kuo G, Arnaud L, Kronstad-OBrien P, Cooper JA. 2005. McAllister AK, Katz LC, Lo DC. 1997. Opposing roles for
Absence of Fyn and Src causes a reeler-like phenotype. endogenous BDNF and NT-3 in regulating cortical den-
J Neurosci 25: 8578-8586. dritic growth. Neuron 18: 767-778.
Kuruvilla R, Ye H, Ginty DD. 2000. Spatially and functionally McAllister AK, Katz LC, Lo DC. 1999. Neurotrophins and
distinct roles of the PI3-K effector pathway during NGF synaptic plasticity. Annu Rev Neurosci 22: 295-318.
signaling in sympathetic neurons. Neuron 27: 499-512. McAllister AK, Lo DC, Katz LC. 1995. Neurotrophins regulate
Lamballe F, Smeyne RJ, Barbacid M. 1994. Developmental dendritic growth in developing visual cortex. Neuron 15:
expression of trkC, the neurotrophin-3 receptor, in the 791-803.
mammalian nervous system. J Neurosci 14: 14-28. Medina DL, Sciarretta C, Calella AM, Von Bohlen Und
Lee KF, Li E, Huber LJ, Landis SC, Sharpe AH, et al. 1992. Halbach O, Unsicker K, et al. 2004. TrkB regulates neo-
Targeted mutation of the gene encoding the low affinity cortex formation through the Shc/PLCgamma-mediated
NGF receptor p75 leads to deficits in the peripheral sensory control of neuronal migration. EMBO J 23: 3803-3814.
nervous system. Cell 69: 737-749. Minichiello L, Calella AM, Medina DL, Bonhoeffer T, Klein R,
Lee R, Kermani P, Teng KK, Hempstead BL. 2001. Regulation et al. 2002. Mechanism of TrkB-mediated hippocampal
of cell survival by secreted proneurotrophins. Science 294: long-term potentiation. Neuron 36: 121-137.
1945-1948. Minichiello L, Klein R. 1996. TrkB and TrkC neurotrophin
Levi-Montalcini R, Angeletti PU. 1968. Nerve growth factor. receptors cooperate in promoting survival of hippocampal
Physiol Rev 48: 534-569. and cerebellar granule neurons. Genes Dev 10: 2849-2858.
Liu X, Ernfors P, Wu H, Jaenisch R. 1995. Sensory but not Minichiello L, Korte M, Wolfer D, Kuhn R, Unsicker K, et al.
motor neuron deficits in mice lacking NT4 and BDNF. 1999. Essential role for TrkB receptors in hippocampus-
Nature 375: 238-241. mediated learning. Neuron 24: 401-414.
Liu X, Jaenisch R. 2000. Severe peripheral sensory neuron loss Miyamoto Y, Yamauchi J, Tanoue A, Wu C, Mobley WC.
and modest motor neuron reduction in mice with com- 2006. TrkB binds and tyrosine-phosphorylates Tiam1, lead-
bined deficiency of brain-derived neurotrophic factor, neu- ing to activation of Rac1 and induction of changes
rotrophin 3 and neurotrophin 4/5. Dev Dyn 218: 94-101. in cellular morphology. Proc Natl Acad Sci USA 103:
Lom B, Cogen J, Sanchez AL, Vu T, Cohen-Cory S. 2002. 10444-10449.
Local and target-derived brain-derived neurotrophic factor Morfini G, DiTella MC, Feiguin F, Carri N, Caceres A. 1994.
exert opposing effects on the dendritic arborization of Neurotrophin-3 enhances neurite outgrowth in cultured
retinal ganglion cells in vivo. J Neurosci 22: 7639-7649. hippocampal pyramidal neurons. J Neurosci Res 39:
Lom B, Cohen-Cory S. 1999. Brain-derived neurotrophic 219-232.
factor differentially regulates retinal ganglion cell dendritic Naumann T, Casademunt E, Hollerbach E, Hofmann J,
and axonal arborization in vivo. J Neurosci 19: 9928-9938. Dechant G, et al. 2002. Complete deletion of the neurotro-
Maisonpierre PC, Belluscio L, Friedman B, Alderson RF, phin receptor p75NTR leads to long-lasting increases in the
Wiegand SJ, et al. 1990. NT-3, BDNF, and NGF in the number of basal forebrain cholinergic neurons. J Neurosci
developing rat nervous system: Parallel as well as reciprocal 22: 2409-2418.
patterns of expression. Neuron 5: 501-509. Nery S, Fishell G, Corbin JG. 2002. The caudal ganglionic
Marin-Padilla M. 1971. Early prenatal ontogenesis of the eminence is a source of distinct cortical and subcortical cell
cerebral cortex (neocortex) of the cat (Felis domestica). populations. Nat Neurosci 5: 1279-1287.
A Golgi study. I. The primordial neocortical organization. Niblock MM, Brunso-Bechtold JK, Riddle DR. 2000. Insulin-
Z Anat Entwicklungsgesch 134: 117-145. like growth factor I stimulates dendritic growth in primary
Marin O, Rubenstein JL. 2003. Cell migration in the forebrain. somatosensory cortex. J Neurosci 20: 4165-4176.
Annu Rev Neurosci 26: 441-483. Oppenheim RW, Yin QW, Prevette D, Yan Q. 1992. Brain-
Matsumoto T, Numakawa T, Adachi N, Yokomaku D, derived neurotrophic factor rescues developing avian
Yamagishi S, et al. 2001. Brain-derived neurotrophic factor motoneurons from cell death. Nature 360: 755-757.
enhances depolarization-evoked glutamate release in Paul CE, Vereker E, Dickson KM, Barker PA. 2004. A pro-
cultured cortical neurons. J Neurochem 79: 522-530. apoptotic fragment of the p75 neurotrophin receptor is
McAllister AK. 2001. Neurotrophins and neuronal differenti- expressed in p75NTRExonIV null mice. J Neurosci 24:
ation in the central nervous system. Cell Mol Life Sci 58: 1917-1923.
1054-1060. Peterson DA, Dickinson-Anson HA, Leppert JT, Lee KF,
McAllister AK, Katz LC, Lo DC. 1996. Neurotrophin regula- Gage FH. 1999. Central neuronal loss and behavioral
tion of cortical dendritic growth requires activity. Neuron impairment in mice lacking neurotrophin receptor p75.
17: 1057-1064. J Comp Neurol 404: 1-20.
18 1 Neurotrophins and central nervous system development
Peterson DA, Leppert JT, Lee KF, Gage FH. 1997. Basal fore- Silos-Santiago I, Fagan AM, Garber M, Fritzsch B, Barbacid M.
brain neuronal loss in mice lacking neurotrophin receptor 1997. Severe sensory deficits but normal CNS development
p75. Science 277: 837-839. in newborn mice lacking TrkB and TrkC tyrosine protein
Polleux F, Whitford KL, Dijkhuizen PA, Vitalis T, Ghosh A. kinase receptors. Eur J Neurosci 9: 2045-2056.
2002. Control of cortical interneuron migration by Smeyne RJ, Klein R, Schnapp A, Long LK, Bryant S, et al.
neurotrophins and PI3-kinase signaling. Development 1994. Severe sensory and sympathetic neuropathies in mice
129: 3147-3160. carrying a disrupted Trk/NGF receptor gene. Nature 368:
Rabacchi SA, Kruk B, Hamilton J, Carney C, Hoffman JR, 246-249.
et al. 1999. BDNF and NT4/5 promote survival and neurite Snider WD. 1994. Functions of the neurotrophins during
outgrowth of pontocerebellar mossy fiber neurons. J Neu- nervous system development: What the knockouts are
robiol 40: 254-269. teaching us. Cell 77: 627-638.
Rajagopal R, Chao MV. 2006. A role for Fyn in Trk receptor Thoenen H. 1995. Neurotrophins and neuronal plasticity.
transactivation by G-protein-coupled receptor signaling. Science 270: 593-598.
Mol Cell Neurosci 33: 36-46. Thoenen H, Barde YA. 1980. Physiology of nerve growth
Rakic P. 1975. Timing of major ontogenetic events in the visual factor. Physiol Rev 60: 1284-1335.
cortex of the rhesus monkey. UCLA Forum Med Sci: 340. Timmusk T, Belluardo N, Persson H, Metsis M. 1994. Devel-
Rakic P, Stensas LJ, Sayre E, Sidman RL. 1974. Computer- opmental regulation of brain-derived neurotrophic factor
aided three-dimensional reconstruction and quantitative messenger RNAs transcribed from different promoters in
analysis of cells from serial electron microscopic montages the rat brain. Neuroscience 60: 287-291.
of foetal monkey brain. Nature 250: 31-34. Tsuruda A, Suzuki S, Maekawa T, Oka S. 2004. Constitutively
Rico B, Xu B, Reichardt LF. 2002. TrkB receptor signaling is active Src facilitates NGF-induced phosphorylation of
required for establishment of GABAergic synapses in the TrkA and causes enhancement of the MAPK signaling in
cerebellum. Nat Neurosci 5: 225-233. SK-N-MC cells. FEBS Lett 560: 215-220.
Ringstedt T, Linnarsson S, Wagner J, Lendahl U, Kokaia Z, Van der Zee CE, Ross GM, Riopelle RJ, Hagg T. 1996. Survival
et al. 1998. BDNF regulates reelin expression and Cajal- of cholinergic forebrain neurons in developing p75NGFR-
Retzius cell development in the cerebral cortex. Neuron 21: deficient mice. Science 274: 1729-1732.
305-315. von Schack D, Casademunt E, Schweigreiter R, Meyer M,
Rocamora N, Garcia-Ladona FJ, Palacios JM, Mengod G. Bibel M, et al. 2001. Complete ablation of the neurotrophin
1993. Differential expression of brain-derived neurotrophic receptor p75NTR causes defects both in the nervous and
factor, neurotrophin-3, and low-affinity nerve growth fac- the vascular system. Nat Neurosci 4: 977-978.
tor receptor during the postnatal development of the rat Wang XH, Poo MM. 1997. Potentiation of developing
cerebellar system. Brain Res Mol Brain Res 17: 1-8. synapses by postsynaptic release of neurotrophin-4. Neu-
Schwartz PM, Borghesani PR, Levy RL, Pomeroy SL, Segal ron 19: 825-835.
RA. 1997. Abnormal cerebellar development and foliation Ward NL, Hagg T. 1999. p75(NGFR) and cholinergic neurons
in BDNF/ mice reveals a role for neurotrophins in CNS in the developing forebrain: A re-examination. Brain Res
patterning. Neuron 19: 269-281. Dev Brain Res 118: 79-91.
Schwyzer L, Mateos JM, Abegg M, Rietschin L, Heeb L, et al. Williams LR, Varon S, Peterson GM, Wictorin K, Fischer W,
2002. Physiological and morphological plasticity induced et al. 1986. Continuous infusion of nerve growth factor
by chronic treatment with NT-3 or NT-4/5 in hippocampal prevents basal forebrain neuronal death after fimbria fornix
slice cultures. Eur J Neurosci 16: 1939-1948. transection. Proc Natl Acad Sci USA 83: 9231-9235.
Segal RA, Takahashi H, McKay RD. 1992. Changes in neuro- Wirth MJ, Brun A, Grabert J, Patz S, Wahle P. 2003. Acceler-
trophin responsiveness during the development of cerebel- ated dendritic development of rat cortical pyramidal cells
lar granule neurons. Neuron 9: 1041-1052. and interneurons after biolistic transfection with BDNF
Sendtner M, Holtmann B, Kolbeck R, Thoenen H, Barde YA. and NT4/5. Development 130: 5827-5838.
1992. Brain-derived neurotrophic factor prevents the death Xu B, Zang K, Ruff NL, Zhang YA, McConnell SK, et al. 2000.
of motoneurons in newborn rats after nerve section. Nature Cortical degeneration in the absence of neurotrophin sig-
360: 757-759. naling: Dendritic retraction and neuronal loss after removal
Shimada A, Mason CA, Morrison ME. 1998. TrkB signaling of the receptor TrkB. Neuron 26: 233-245.
modulates spine density and morphology independent of Yacoubian TA, Lo DC. 2000. Truncated and full-length TrkB
dendrite structure in cultured neonatal Purkinje cells. receptors regulate distinct modes of dendritic growth. Nat
J Neurosci 18: 8559-8570. Neurosci 3: 342-349.
Neurotrophins and central nervous system development 1 19
Yan Q, Elliott J, Snider WD. 1992. Brain-derived neurotrophic Yoshizawa M, Kawauchi T, Sone M, Nishimura YV, Terao M,
factor rescues spinal motor neurons from axotomy- et al. 2005. Involvement of a Rac activator, P-Rex1, in
induced cell death. Nature 360: 753-755. neurotrophin-derived signaling and neuronal migration.
Yeo TT, Chua-Couzens J, Butcher LL, Bredesen DE, Cooper J Neurosci 25: 4406-4419.
JD, et al. 1997. Absence of p75NTR causes increased basal Zagrebelsky M, Holz A, Dechant G, Barde YA, Bonhoeffer T,
forebrain cholinergic neuron size, choline acetyltransferase et al. 2005. The p75 neurotrophin receptor negatively mod-
activity, and target innervation. J Neurosci 17: 7594-7605. ulates dendrite complexity and spine density in hippocam-
York RD, Yao H, Dillon T, Ellig CL, Eckert SP, et al. 1998. Rap1 pal neurons. J Neurosci 25: 9989-9999.
mediates sustained MAP kinase activation induced by
nerve growth factor. Nature 392: 622-626.
2 Nerve Growth Factor Regulated
Gene Expression
L. A. Greene . J. M. Angelastro
2 The Role of Regulated Gene Expression in the NGF MechanismWhat Genes Does NGF
Regulate and Why Do We Want to Know? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5 Closing Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Abstract: This chapter provides a current accounting of the identities of genes and gene products that are
subject to longterm regulation by nerve growth factor (NGF). We provide tabular listings of (a) NGF
responsive genes and proteins that are reported in the literature and (b) transcripts found by comprehensive
serial analysis of gene expression (SAGE) to be differentially expressed in PC12 cells before and after longterm
NGF exposure. Transcriptional changes are key elements in the mechanisms by which NGF affects neuronal
differentiation, function, protection, and repair. The information provided here is thus meant to serve as a
resource for those interested in the molecular mechanisms and consequences of trophic factor actions.
List of Abbreviations: ATF5, activating transcription factor 5; EST, expressed sequence tag; NGF, nerve
growth factor; SAGE, serial analysis of gene expression
The aim of this chapter is to provide a current account of what we know about the identities of the genes and
gene products that are subject to longterm regulation by NGF. Our objective is like that of traditional
handbooks of chemistry and physics, that is, to list information without including extensive interpretive
rhetoric. We hope that the reader will find this information to be a useful resource that will inspire further
experimental and intellectual advances regarding the development, function, and repair of the nervous system.
There are two principal portions of this chapter. One is a tabular listing of the genes and gene products
that have been identified in the literature as being regulated by NGF. The second is a tabular listing of
identified NGF regulated transcripts that we have detected in a largescale SAGE comparison of the
transcriptomes of nave and longterm NGFtreated PC12 cells. In each case, the regulated genes have
been subdivided into categories based on the current information about their functional properties.
Part of the appeal of putting together such a chapter is the opportunity provided by current electronic
technology to easily and swiftly update and extend the information herein and to make such modifications
rapidly accessible to readers. We intend to continue to add to this chapter new findings in the literature as
well as the results of our continuous efforts to fully match the output of our SAGE study with known genes.
We will also continue to take into account new findings about the functional activities of the various genes
listed here. An additional aspect is that, such a format will permit us to correct errors and omissions that we
anticipate have not only occurred, but will be (and we encourage this) pointed out to us by readers. Finally, as
noted earlier, the present chapter includes genes and gene products subject to longterm regulation by NGF.
By long term, we mean to the exclusion of immediate early genes and potentially, to some genes that are only
transiently regulated by NGF. Should the opportunity arise, we will endeavor to add such genes in the future.
Though we now largely take for granted the notion that neurotrophic factors such as NGF act in part by
regulating gene expression, this was not the case always. The capacity of NGF to promote neuron survival as
well as differentiation led to an early debate over whether the factor was merely permissive for differentia-
tion or whether it truly possessed instructive actions. Moreover, the relative stability of NGFregulated
transcripts and proteins and the capacity of NGF to promote neurite outgrowth in explanted ganglia led to
the hypothesis that the factor acted by a mechanism that was independent of transcription (Partlow and
Larrabbee, 1971; Mizel and Bamburg, 1976). The advent of cell systems such as PC12 cells without a
requirement for NGF or without prior exposure to the factor (in contrast to dissected neurons) permitted
the first demonstrations that NGF does regulate genes (McGuire et al., 1978 ; Greene, 1981; Greenberg et al.,
1985) and that such effects play essential roles in neuronal differentiation and function (Burstein and
Greene, 1978; Greene et al., 1983). Moreover, the introduction of means to culture adult sensory neurons,
which respond to NGF, but do not require it, has extended our capacity to uncover effects of the factor on
gene expression (Lindsay, 1988; Lindsay and Harmar, 1989). Additional advances in neuron culture, in
Nerve growth factor regulated gene expression
2 23
introducing NGF in vivo, and in generating animals that are null for key cell death genes and in which
neurons can consequently survive in absence of trophic support, have further widened the opportunities to
identify NGFregulated genes and gene products. Finally, molecular technologies have greatly improved the
sensitivity and efficiency with which NGFdependent changes in gene expression can be monitored.
If NGF acts in part by regulating genes, then the question naturally arises about their identities.
Knowing the identities of NGFregulated genes is important for several reasons. NGF participates in a
number of fundamental activities including promotion of differentiation (itself a multifaceted response),
survival, plasticity, repair, neuroprotection, and neurite outgrowth. Many of the genes regulated by NGF are
key regulators and participants in these major activities and their identification thus provides fundamental
insight into the molecular mechanisms by which they occur. Furthermore, understanding the mechanisms
by which NGF acts, must include a clear description of the genes subject to its regulation. In this context,
knowledge about the genes that NGF regulates not only permits construction of mechanistic pathways, but
also provides candidates for experimental extension of, and intercalation within, these pathways. Con-
fronting ourselves (and our colleagues) with the knowledge that there are large numbers of NGFregulated
genes for which we understand neither function nor potential roles, is not only humbling, but also spurs us
to understand how they contribute to the NGF mechanism of action. Finally, NGF is one of many
neurotrophic factors. If we can identify and understand the functions of the genes that NGF regulates,
we will have an enormous insight into the role of such genes in the actions of other trophic factors as well.
Although the list of NGFregulated genes in the literature is quite substantial, we felt that it was likely to be
far short of the numbers that are actually responsive to NGF. To address this, in 2000 we carried out a
comprehensive comparison, using SAGE technology, of transcripts expressed by rat PC12 pheochromocy-
toma cells before and after 9days of treatment with NGF (Angelastro et al., 2000). PC12 cells proliferate in
serumcontaining medium and do not require NGF for survival. In response to NGF under the same
conditions, PC12 cells exit the cell cycle and over a timecourse of days undergo differentiation into neurite
bearing, electrically excitable cells that resemble sympathetic neurons. Thus, they represent a convenient
system in which to identify NGFpromoted changes in gene expression. SAGE is an unbiased and highly
24
2 Nerve growth factor regulated gene expression
. Table 2-1
NGF regulated changes in gene expression
Up/
Gene name Access. # Down RNA/Prot System References
Antioxidant
Catalase (Cat) NM_012520 Up R PC12 Sampath et al.
(1994)
Cat NM_012520 Up R Striatum in vivo Frim et al. (1994)
Glutathione peroxidase Up R PC12 Sampath et al.
(1994)
GlutathioneStransferase mu Up P PC12 NurEKamal
et al. (2000)
GlutathioneStransferase pi Up P PC12 NurEKamal
et al. (2000)
Hypoxia inducible factor AJ277828 Down R, P PC12 NaranjoSuarez
2 alpha (Hif2a) et al. (2003)
Thioredoxin NM_053800 Up R, P PC12 Bai et al. (2003)
Calcium binding
Calbindin 1 (Calb1) NM_031984 Up P, R PC12 Iacopino et al.
(1992)
Calmodulin Up R PC12 Bai and Weiss
(1991)
Calpactin I heavy chain/ NM_019905 Up P, R PC12 Jacovina et al.
Annexin II (2001)
S100 related protein, clone NM_031114 Up R PC12 Masiakowski and
42C/calpactin I light chain Shooter (1988)
Cell cycle
APC NM_012499 Up P PC12 Dobashi et al.
(1996)
Cdc2/Cdk1 Down P PC12 Buchkovich and
Ziff (1994)
Cyclin dependent kinase NM_199501 Down P PC12 Buchkovich and
2 (Cdk2) Ziff (1994)
Cyclindependent kinase 4 NM_053593 Down P PC12 Yan and Ziff
(CDK4) (1995)
Cdkn1a cyclindependent NM_080782 Up P, R PC12 Yan and Ziff
kinase inhibitor 1A/Waf1 (1997)
Cyclin B1 (Ccnb1) NM_171991 Down P PC12 Yan and Ziff
(1995)
Cyclin D1 (Ccnd1) NM_171992 Up P, R PC12 Yan and Ziff
(1995)
Cyclin D2 (Ccnd2) D16308 Down P PC12 Tamaru et al.
(1994)
Cyclin F (Ccnf) XM_340763 Down P, not R PC12 Movsesyan et al.
(1996)
Cyclin G Up P PC12 Gollapudi and
Neet (1997)
Deleted in colorectal U68725 Up R PC12 Lawlor and
carcinoma (Dcc) Narayanan
(1992)
E2F transcription factor 1 D63165 Down P PC12 Persengiev et al.
(E2f1) (1999)
26
2 Nerve growth factor regulated gene expression
Disease related
Amyloid beta (A4) precursor NM_019288 Up P, R PC12 Villa et al. (2001)
protein (App)
App NM_019288 Up R Developing Mobley et al.
brain (1988)
Prion protein (Prnp) NM_012631 Up R PC12 Wion et al.
(1988)
Prnp NM_012631 Up R Developing Mobley et al.
brain (1988)
Ion channels
Ca channel, voltage NM_147141 Up R PC12 Colston et al.
dependent, N type, alpha 1B (1998)
subunit
Cacna1c Ca channel, voltage NM_012517 Down R PC12 Colston et al.
dependent, Ltype alpha 1C (1998)
Kcnb1 K voltage gated NM_013186 Up P, not R PC12 Sharma et al.
channel, Shabrelated, (1993)
member 1
Cacna1d Ca channel, voltage NM_017298 Down R PC12 Colston et al.
dependent, Ltype alpha 1D (1998)
K small conductance Ca NM_019313 Up P DRG in vitro Boettger et al.
activated channel N1 (Kcnn1) (2002)
Proton gated cation channel NM_173135 Up R Sensory Mamet et al.
DRASIC neurons (2003)
Na channel, voltagegated, NM_030875 Up R DRG neurons in Zur et al. (1995)
type 1, a polypeptide (Scn1a) vitro
Na channel, voltagegated, NM_012647 Up R PC12 Mandel et al.
type 2, a1 polypeptide (1988)
(Scn2a1)
Na channel, voltagegated, NM_012647 Up R DRG neurons in Zur et al. (1995)
type 2, a1 polypeptide vitro
(Scn2a1)
Na channel, voltagegated, NM_013119 Down R Adult DRG in Black et al.
type III, a polypeptide (Scn3a) vitro (1997)
Na channel, voltagegated, NM_133289 Up R PC12 DArcangelo
type 9, a polypeptide (Scn9a) et al. (1993)
Na channel, voltagegated, NM_017247 Up P Adult DRG in Black et al.
type 10, a polypeptide vitro (1997)
(Scn10a)
Na channel, voltagegated, NM_017288 Up R DRG neurons in Zur et al. (1995)
type I, beta polypeptide vitro
(Scn1b)
Lectins
Lectin, galactose binding, NM_031832 Up P, R PC12 Kuklinski et al.
soluble 3 (Lgals3) (2003)
Lgals3 NM_031832 Up P DRG in vitro Pesheva et al.
(2000)
Miscellaneous
ATPase, Na K transporting, NM_012504 Up P PC12 Kurihara et al.
alpha 1 (1994)
Collagenase 1 Up R PC12 Vician et al.
(1997)
CTP: phosphocholine Up R PC12 Carter et al.
cytidylyltransferase beta2 (2003)
28
2 Nerve growth factor regulated gene expression
Neurotransmission
Acetylcholinesterase (Ache) NM_172009 Up R PC12 DeschenesFurry
(stabilized) et al. (2003)
Adenosine A2a receptor NM_053294 Down P, R PC12 Arslan et al.
(Adora2a) (1997)
Agrin (Agrn) NM_175754 Up R PC12 Smith et al.
(1997)
Choline acetyltransferase Up R Basal forebrain in Higgins et al.
(ChAT) vivo (1989)
Cholinergic receptor, NM_080773 Down R Telencephalic Eva et al. (1992)
muscarinic 1 (Chrm1) neurons in vitro
Cholinergic receptor, NM_031016 Up R Striatal neurons Ebstein et al.
muscarinic 2 (Chrm2) in vitro (1993)
Cholinergic receptor, NM_012527 Down R Telencephalic Eva et al. (1992)
muscarinic 3 (Chrm3) neurons in vitro
Cholinergic receptor, XM_345403 Up R Telencephalic Eva et al. (1992)
muscarinic 4 (Chrm4) neurons in vitro
Chrm4 XM_345403 Up R PC12 Lee and Malek
(1998)
30
2 Nerve growth factor regulated gene expression
Substrate interactions
Integrin alpha 1 (Itga1) NM_030994 Up R PC12 Danker et al.
(2001)
Neural cell adhesion molecule NM_017345 Up P, R PC12 Grant et al.
L1 (NcamL1) (1996)
Neural cell adhesion molecule NM_031521 Up P, R PC12 Prentice et al.
1 (Ncam1) (1987)
Matrix metalloproteinase 3 NM_133523 Up R PC12 Machida et al.
(Mmp3)/stromelysin1 (1989)
Thymus cell antigen 1, theta NM_012673 Up P,W PC12 Doherty et al.
(Thy1) (1988)
Transcriptional regulators
Achaetescute complex NM_022384 Down R PC12 Grumolato et al.
homologlike 1 (Ascl1)/ (2003)
MASH1
AP2 Up P PC12 Paggi et al.
(2001)
Activating transcription factor NM_172336 Down P, R PC12 Angelastro et al.
5 (ATF5) (2003)
DNA (cytosine5) NM_053354 Down P, R PC12 Deng and Szyf
methyltransferase 1 dnmt1 (1999)
Estrogen receptor Up P, R PC12 Sohrabji et al.
(1994)
Fak1/fetal Alzheimer antigen/ Up R PC12 Rhodes et al.
falz (2003)
Foslike antigen 1/FRA1 NM_012953 Up P PC12 Cosgaya and
(Fosl1) Aranda (2001)
Foslike antigen 1/FRA NM_012954 Up P PC12 Cosgaya and
2 (Fosl2) Aranda (2001)
LIM homeobox protein 1 NM_145880 Up R DRG in vitro Jameson and
(Lhx1)/Rlim Lillycrop (2001)
Hypoxia inducible factor AJ277828 Down R, P PC12 NaranjoSuarez
2 alpha (Hif2a) et al. (2003)
LIM homeobox protein 3 AF370447 Up R DRG in vitro Jameson and
(Lhx3) Lillycrop (2001)
Math2/Nex1/Neurod6 XM_001059051 Up P PC12 Uittenbogaard
and Chiaramello
(2002)
Cmyc Y00396 Down R, P Human Chen et al.
neuroblastoma (1994)
lines
Nerve growth factor regulated gene expression
2 33
Translational regulators
Eukaryotic translation Z48225 Up P PC12 Kleijn et al.
initiation factor 2B (eIF2B) (1998)
sensitive approach to defining cellular transcriptomes (Velculescu etal., 2000). Put briefly, each transcript
recovered from the cell of interest is converted into a defining SAGE tag, which includes the most 30 CATG
in the transcript followed by the next 11 bases. The relative number of tags representing a given transcript
that are recovered in the analysis is directly proportional to its relative abundance in the cell. If a large
number of SAGE tags are analyzed to compare the same cell in two different states (in this caseNGF
treatment), this provides a very comprehensive view of the changes that occur in gene regulation. In our
study, we analyzed large libraries of approximately 80,000 tags each (76,280 without NGF; 87,004 with
NGF). The results revealed that approximately 4% of the 22,000 transcripts detected in PC12 cells
responded to NGF by undergoing changes in expression of sixfold or greater. One of the challenges of
interpreting SAGE data is to match the various tags with known transcripts. Thus far, we have, by a variety
of informatic means (Angelastro etal., 2000), matched to known transcripts approximately half of the tags
representing genes that are regulated by sixfold or more in response to longterm NGF exposure.
> Table 2-2 lists the NGFresponsive transcripts thus far (up to September 2004) identified from our
SAGE study. The fold changes in expression are listed either as fold increase after NGF treatment (positive
numbers) or as fold decrease in expression after NGF treatment (negative numbers). When no tags were
detected, for the purposes of calculating a fold change, the tag number was set to 1 (to avoid a ratio of
infinity). Thus, some of the fold changes may be underestimated. The fold changes reported reflect
normalization for the differences in tag numbers for the two libraries. With the sixfold cutoff that we
have chosen, all reported changes are significant with a p value of 0.05 or less.
As in the previous table, we have subdivided the regulated genes into functional categories and in some
cases, we have listed the same gene under multiple categories. We have also included relevant accession
numbers. The present table updates and substantially expands the lists we published in Angelastro etal.
(2000) and such additions reflect the enormous increase in genome information available since that time.
We have done our very best to ensure the accuracy of the data presented in the table, but acknowledge that
errors are possible. Readers are encouraged to provide feedback in this respect.
We wish to note that because of the considerable delay between preparation of > Table 2-1 (September,
2004) and publication of this chapter, the Table does not reflect recent advances in annotation of the rat
genome. Nevertheless, the accession numbers given in the Table should permit the interested reader to find
the current annotation for each of the listed transcripts. It is our intention to provide updates for the Table
in future electronic versions of this chapter. Readers may also directly request the lastest version of this table
by email (lag3@columbia.edu).
34
2 Nerve growth factor regulated gene expression
. Table 2-2
NGFPROMOTED longterm changes in gene regulation in PC12 cells as detected by SAGE
() (;) Fold
Tag NGF NGF change Identity Accession #
Energetics
AAGGTTCACTC 19 1 17 Similar to Mtch1 protein (mitochondrial carrier XM_215358
homolog 1)
GCATACGGCGC 18 0 16 Atp5k: ATP synthase, H transporting, NM_080481
mitochondrial F1F0 complex, subunit e
AGCTTGATTAA 16 1 14 Similar to RIKEN cDNA 2410011G03O; XM_216880
0rtholog of human NADH: ubiquinone
oxidoreductase
GTGGCCCACTT 23 2 10 Cox17: cytochrome c oxidase, subunit XVII NM_053540
assembly protein homolog (yeast)
AATAAAAGTTC 15 2 7 Atp5a1: mitochondrial HATP synthase alpha NM_023093
subunit
TATCCAAACAG 7 1 6 Pdha1: pyruvate dehydrogenase E1 XM_343787
alpha 1
CCAAGGAAAAC 51 8 6 Ldha: lactate dehydrogenase A NM_017025
TACCATCTTTC 1 6 7 Similar to integral membrane protein CII3/ XM_2139361
Sdhc: succinate dehydrogenase complex,
subunit C
TACTAGAAAAG 1 7 8 Similar to NADH dehydrogenase (ubiquinone) XM_213940
FeS protein 2
ATCCAAGTCGC 1 8 9 Similar to Cytochrome oxidase biogenesis XM_214182
protein OXA1, mitochondrial precursor (OXA1
like protein) (OXA1Hs)
AGGTCGCTTGG 1 8 9 Similar to xylulokinase homolog; xylulokinase NW_000361
(H. influenzae) homolog
CCCGACTGGGT 5 66 15 Homolog of human LOC56901: ubiquinone NM020142
oxidoreductase MLRQ subunit homolog
Metabolism
GCTGGAATTGA 10 1 9 Farnesyl diphosphate synthase (Fdps) NM_031840
AAGGGTCCCCG 8 0 7 Prpsap1: phosphoribosylpyrophosphate NM_022545
synthetaseassociated protein (39kDa)
TCTGTCCTGCT 8 0 7 Degs: degenerative spermatocyte homolog XM_346454
(Drosophila); transmembrane protein involved
in meiosis in fly; lipid metabolism
ATTTGCTTCTT 8 0 7 Similar to ectonucleotide pyrophosphatase/ XM_236956
phosphodiesterase 5
TTTCAGCAGTG 7 0 6 Hprt: hypoxanthine guanine phosphoribosyl NM_012583
transferase
GGCCCCCAAGT 1 5 6 Ortholog of murine BC004012: cDNA sequence BC004012
BC004012; putative inorganic polyphosphate/
ATPNAD kinase
GCTTTAATGGA 6 31 6 Similar to 2310047E01Rik protein; encodes XM_343416
ortholog of murine Car12: carbonic
anyhydrase 12
GACAATGAAAA 2 12 7 Isocitrate dehydrogenase 3 (NAD), gamma X74125
(Idh3g)
GGAGGACCTCG 2 12 7 Dimethylarginine dimethylaminohydrolase XM_215315
2 (Ddah2)
Nerve growth factor regulated gene expression
2 35
Signaling
AATGTGAGTCA 14 0 12 1433 protein gammasubtype NM_019376
TGCACAGTGCT 14 0 12 S100a4: S100 calciumbinding protein A4 NM_012618
AAATCCTTTCA 12 0 11 Pleiotrophin NM_017066
TGTAGCTCAAT 11 1 11 Guanine nucleotide binding protein, alpha o NM_017327
(Gnao)
ACTCGGAGCCA 10 0 9 Calmodulin 1 (Calm1) (phosphorylase kinase, NM_031969
delta)
AGACACTTCCT 10 1 9 Anxa2: calpactin I heavy chain/ANNEXIN 2 NM_019905
ATTCTGTGCTG 10 1 9 Cd9: CD9 antigen (p24) X76489
GGTGTGCCAGG 10 0 9 Similar to RIKEN cDNA 5730466P16; similar to NM_003876
protein PM1; putative receptor protein
AAGCCTTGCTG 9 0 8 Growth factor receptor bound NM_030846
protein 2 (Grb2)
TTTTGTGATGG 9 0 8 Similar to RIKEN cDNA 2010107K23 encoding XM_346291
protein with MAGE domain; Homolog to
human MAGEH1
TCAGGCATTTT 9 0 8 Similar to Rasrelated protein Rab1B X13905
CCCCCTGGATC 26 3 8 S100a6 calcium binding protein A6 (calcyclin) NM_0533485
CTGCCATCCCT 8 0 7 LOC294900: similar to PC1; Tpd52: tumor XM_215524
protein D52; possible roles Camediated
signaling and proliferation
TAGTCCAGGCT 8 0 7 LOC291675: similar to RIKEN cDNA AI028965
2010001M09encodes protein similar to
human PACAP
TCCTGTGTCCT 8 1 7 Inositol 1, 4, 5triphosphate receptor 3 (Itpr3) NM_013138
CTTCCAAATGT 15 2 7 Rtn4: NogoA NM_031831
AATGCCCCCAG 7 1 6 Presenilin2 (Psen2) NM_031087
CGCGCGCGCGC 7 0 6 Ptpns1: protein tyrosine phosphatase, non NM_013016
receptor type substrate (Bit, SHPS1)
CTGTTAGGTGG 7 0 6 Nicastrin Component of the gammasecretase NM_174864
complex (Ncstn)
36
2 Nerve growth factor regulated gene expression
Cytoskeleton
TGGAACCTTGC 16 1 14 Dncli2: LIC2 dynein light intermediate chain NM_031026
53/55
AGTTGATGCAA 32 2 14 Nfl: neurofilament, light polypeptide NM_031783
AGTTTGCTGAT 43 3 13 Ortholog of murine Tagln2: transgelin 2 BC009076
GAAAAATAGTC 11 0 10 Nfl: neurofilament, light polypeptide NM_031783
TATTTGTTTTG 10 1 9 Prnp: Prion protein, structural NM_012631
AGCTTTCCTGT 10 1 9 Cortactin isoform B (Cttnb) NM_021868
CCACACTGTCT 9 1 9 Ortholog of murine Capza1: capping protein NM_009797
(actin filament) muscle Zline, alpha 1
CAGTATCCCTA 9 1 8 Msn: Moesin NM_030863
GACTGTGCCAA 8 1 7 Pin: dynein, cytoplasmic, light chain 1 NM_053319
GAGGAGGGGGA 8 0 7 Ortholog of murine Tuba6: tubulin, alpha 6 NM_009448
CCCTTCCCTGC 7 0 6 Ctxn: cortexin L15011
GCAATAAATGG 7 0 6 Dbn1: drebrin 1; actin binding protein NM_031024
GCTTAGCCATT 13 2 6 Microtubuleassociated protein 4 XM_345984
CTGCTAGCACC 1 5 6 Similar to Epidermal growth factor receptor XM_341958
pathway substrate 8like protein 2; Eps8l2; PTB
domain/Eps8 actin regulator
ACCTTCTTGGT 2 11 6 Add1: adducin 1, alpha NM_016990
GAAGGAGAACT 0 6 7 Add3: Adducin 3, gamma; Actin capping NM_031552
protein
GCAGTGGGCTC 1 7 8 Coronin relative protein; actin binding protein NM_139115
38
2 Nerve growth factor regulated gene expression
Protein synthesis
CTCAGACAGTG 14 1 12 Similar to 40S ribosomal protein S27 X59375
CTCAAACACCA 1 6 7 LOC363023: similar to mitochondrial ribosomal XM_343354
protein L4 isoform a; Mrpl4
TGTATGAAGCA 1 7 8 Metap2: methionine aminopeptidase NM_022539
2/Initiation factor 2 associated 67 kDa protein
(Amp2); protein synthesis
TTCGTGGTCAA 2 16 9 Eef2: eukaryotic translation elongation factor 2 NM_017245
ATCCTCGCTGA 1 9 10 LOC287126: similar to TCE2; Ortholog of XM_213234
murine Mrps34 mitochondrial ribosomal
protein S34
ACAGAAAGTGG 1 9 10 Similar to mitochondrial ribosomal XM_214751
protein L18
Proteasomal pathway
GTGCTGGACCT 10 1 9 Psme2: protease (prosome, macropain) 28 NM_017257
subunit, beta; activator of proteasome
AGAGGAAGTGG 10 1 9 Similar to Fbox protein FBL2; related to skp2 XM_217519
so putative involvement in proteasomal
targeting
AGACGCCTGTG 10 0 9 Similar to 26S proteasomeassociated pad1 XM_215745
homolog
CCTTACACTTG 9 1 8 Similar to tetratricopeptide repeat domain 3; XM_340973
Ortholog of murine Ttc3; putative ubiquitin
ligase
AAAACACCTTG 7 1 6 Nedd4a: neural precursor cell expressed, XM_343427
developmentally downregulated gene 4; E3
UBIQUITIN PROTEIN LIGASE
AAGTAGCTGGA 7 0 6 Similar to RIKEN cDNA 1300013G12; encoded XM_222627
protein very similar to murine Ubxd2: UBX
domain containing 2
TGATGTCTCTC 0 6 7 Ortholog of human COP1: constitutive NM_022457
photomorphogenic protein; E3 ligase that can
target p53
CAGGGCGAGAT 1 6 7 Similar to RIKEN cDNA 6330414O09; ortholog XM_218405
of Irf2bp1: interferon regulatory factor2
binding protein1; E3 ligase domain
CTCCTCCTGAT 1 6 7 Similar to Ubiquitin carboxylterminal XM_341033
hydrolase 12/USP12 ubiquitin specific
protease 12
GACCTTGGAGT 0 11 12 Ortholog of murine Rnf26: ring finger NM_153762
protein 26; putative ubiquitinprotein ligase
Lysosomal function
CTGATCCCCAT 3 17 6 Lamp1: Lysosomal associated membrane NM_012857
protein 1 (120 kDa); membrane glycoprotein
Nerve growth factor regulated gene expression
2 43
Transporters
TTCTAGCATAT 11 0 10 Atp1b1: ATPase Na/K transporting beta 1 NM_013113
polypeptide
TGCTACCACAC 7 1 6 HesB protein; contains Hesb domain involved NM_181626
in Inorganic ion transport and metabolism
GTGGGCCAAAC 2 10 6 Similar to solute carrier family 39 (zinc XM_342286
transporter), member 1; zinciron regulated
transporterlike gene;
TAGAAAAATGG 1 5 6 Slc16a1: solute carrier family 16, member 1 NM_012716
GTGGGCCAAAC 1 5 6 Ortholog of murine Slc39a1: solute carrier NM_013901
family 39 (zinc transporter), member 1/Zip1
GGGCAGACACA 1 5 6 Timm44: translocator of inner mitochondrial NM_017267
membrane 44 NM_017267
GACATAGCCCA 2 10 6 CHOT1: choline transporter NM_017348 NM_017348
GTGGCCAATCA 0 6 7 LOC291840: amino acid transporter NM_1003705
TTGTATAATAG 1 6 7 Ant2: Adenine nucleotide translocator 2, D12771
fibroblast isoform (ATPADP carrier protein)
CAGTGGGTGGG 5 33 8 LOC287642: galactose transporter NM_199081
ATTCTCTGGAT 1 7 8 Atp2a2: ATPase, Ca transporting, cardiac NM_017290
muscle, slow twitch 2
GGCTTGCTCCT 1 8 9 LOC290673: similar to cationtransporting XM_214310
atpase
CTGGAGCTGGG 1 9 10 Slc25a10: solute carrier family 25 NM_133418
(mitochondrial carrier; dicarboxylate
transporter), member 10
ACAGTGAAGGG 3 26 10 Slc3a2: solute carrier family 3 (activators of NM_019283
dibasic and neutral amino acid transport),
member 2
AACGCTGACCA 1 13 15 LOC288919: similar to arsenic resistance XM_213848
ATPase
TTGGTGAGGTA 1 16 18 Slc7a8: solute carrier family 7 (cationic amino NM_053442
acid transporter, Y system), member B/ LAT4
Neurotransmission/Synapses
CTGGAGGTGTG 22 2 10 Homolog of murine Syn2: synapsin II NM_013681
CCGCTATAACA 8 0 7 Syn2: synapsin II NM_919159
CTAGACACCTG 8 1 7 Scg3: secretogranin III NM_053856
AGTAATTTTAG 7 1 6 Acp1: acid phosphatase 1, soluble; possible NM_021262
role synaptic transmission
GCACACTGTGT 7 0 6 Sh3d2c1: SH3 domain protein 2 C1; implicated NM_031238
in vesicle recycling
ACATTTCAATT 1 5 6 Similar to gammaaminobutyric acid (GABA(A)) XM_216288
receptorassociated proteinlike 1
46
2 Nerve growth factor regulated gene expression
Ion channels
GCTCAGATCCA 0 5 6 Kcnh2: potassium voltagegated channel, NM_053949
subfamily H (eagrelated), member 2
GAGAGCTAACA 1 6 7 Nnat: neuronatin; possibly regulates ion U08290
channels during brain development
GAAATGTCTGA 0 11 13 Similar to hypothetical protein MGC27385; XM_223921
ortholog of murine Kctd6: K channel
tetramerization domain containing 6
TTGGGCTGGTT 0 12 14 Asic4: SPASIC protein: Noninactivating NM022234
protongated ion channel in brain
Antioxidant action
GTGATAGACAA 0 6 7 Peroxiredoxin 6 (Prdx6); antioxidant enzyme NM_053610
AGCTGGGACTT 0 7 8 Sod2: Superoxide dismutase 2, mitochondria X566001
Peroxisomal function
TGCTTGCCACA 0 7 8 Similar to peroxisomal acylCoA thioesterase XM_234398
2B; likely ortholog of mouse peroxisomal acyl
CoA thioesterase 2B
Miscellaneous
TAGAGCGTGCT 11 0 10 Similar to nucleosome assembly protein 1like XM_341967
4; nucleosome assembly and gene expression
GTTTTGCTACC 8 0 7 Similar to NIPSNAP1 protein: XM_341249
4nitrophenylphosphatase domain and non
neuronal SNAP25like protein homolog 1
(C. elegans)
GAACGCACACC 7 0 6 Similar to erthyrocyte band 7 integral XM_216045
membrane protein, protein 7.2B, stomatin
GTCTAGGTCAC 7 0 6 Similar to hepatitis B virus xinteracting XM_215674
protein; HBxinteracting protein; HBixP
ACACGGAGGAG 0 5 6 Haptoglobin (Hp); has serinetype NM_012582
endopeptidase activity
CTCAGCAAAAC 0 5 6 LOC317431: similar to RIKEN cDNA 2610028I09; XM_228840
ortholog of murine Ribc1: RIB43A domain with
coiledcoils 1
GTGCGGTACCT 0 5 6 LOC300222: similar to microspherule protein 1 XM_217048
P78; MCRS1; MSP58; ICP22BP; putative cell
cycle/transcription regulator
GGAAGCTGCAA 1 5 6 LOC361425: similar to hypothetical protein XM_341703
COX4AL ortholog of murine Noc4: neighbor
of Cox4
GTAGCAGCCAG 1 5 6 LOC296731: similar to NEDD8 ultimate buster XM_231280
1; NUB1; binds NEDD8 a ubiquitinlike protein
48
2 Nerve growth factor regulated gene expression
ESTs
GTATTAAATAG 11 1 10 ESTs BQ192965
AGCTAGAGCTG 9 1 8 ESTs BF392924
TAAAGTACTCA 9 1 8 ESTs BQ202150
TGTAATGAGAT 8 0 7 ESTs BE117509
GCTCCAGCTAC 8 1 7 EST AA858996
ATTGTCTTTTC 7 0 6 ESTs AI548927
TGGGCACTGGG 7 0 6 EST AA818499
TGTTCTATAGG 7 1 6 ESTs AI060317
52
2 Nerve growth factor regulated gene expression
Though our aim here is not to interpret or speculate the significance of the various regulated genes, a
few generalizations seem warranted. One is that although the findings in > Tables 2-1 and > 2-2 do overlap
to some extent, they also include many genes that do not. In many cases, this reflects the difference in fold
change required to reach significance in that changes of only greater than sixfold are reported for the SAGE
study. In other cases, due to the absence of 30 sequence, it may not be presently possible to verify the SAGE
tag corresponding to the genes given in > Table 2-1.
Another point of interest is that the SAGE study has uncovered many regulated transcripts that encode
novel proteins of unknown function. Likewise, a number of additional regulated transcripts encode
proteins of at least partially defined function, but that have no obvious or previously known roles in
neuronal differentiation or function. This suggests that we have a very long way to go in our quest to fully
understand how NGF works and that there are many opportunities to fit new proteins into the NGF
mechanism. One example of this is our initial work on ATF5, a transcription factor that is 30fold
downregulated by NGF and that previously had no known role in the nervous system. Functional studies
inspired by our SAGE findings revealed that ATF5 is highly expressed in neuroprogenitor cells but not
mature neurons and glia and that it must be downregulated by trophic factors such as neurotrophins to
permit neuroprogenitor cells to exit the cell cycle and to differentiate (Angelastro etal., 2003; Angelastro
et al., 2005; Mason et al., 2005).-
5 Closing Remarks
In summary, we provide here information about the identities of genes and gene products that are subject
to regulation by NGF. We anticipate that many such genes will be regulated by other neurotrophic factors in
multiple cellular settings. In contrast, others will show both cell and factorspecific regulation. We hope that
Nerve growth factor regulated gene expression
2 53
perusal of the information here will permit the reader to formulate overarching ideas about how NGF and
other trophic factors influence the properties of their target cells and will perhaps inspire them to choose
among the regulated genes their own favorites for further study.
References
Aletta JM, Angeletti R, Liem RK, Purcell C, Shelanski ML, Badoyannis HC, Sharma SC, Sabban EL. 1991. The differen-
et al. 1988. Relationship between the nerve growth factor tial effects of cell density and NGF on the expression of
regulated clone 73 gene product and the 58kilodalton tyrosine hydroxylase and dopamine betahydroxylase in
neuronal intermediate filament protein (peripherin). PC12 cells. Brain Res Mol Brain Res 11: 79-87.
J Neurochem 51: 1317-1320. Bai G, Weiss B. 1991. The increase of calmodulin in PC12 cells
Alheim K, McDowell TL, Symons JA, Duff GW, Bartfai T. induced by NGF is caused by differential expression of mul-
1996. An AP1 site is involved in the NGF induction of IL1 tiple mRNAs for calmodulin. J Cell Physiol 149: 414-421.
alpha in PC12 cells. Neurochem Int 29: 487-496. Bai J, Nakamura H, Kwon YW, Hattori I, Yamaguchi Y, et al.
Allen JM, Martin JB, Heinrich G. 1987. Neuropeptide Y gene 2003. Critical roles of thioredoxin in nerve growth factor
expression in PC12 cells and its regulation by nerve growth mediated signal transduction and neurite outgrowth in
factor: A model for developmental regulation. Brain Res PC12 cells. J Neurosci 23: 503-509.
427: 39-43. Biswas SC, Greene LA. 2002. Nerve growth factor (NGF)
Anastasiadis PZ, Kuhn DM, Blitz J, Imerman BA, Louie MC, downregulates the Bcl2 homology 3 (BH3) domainonly
et al. 1996. Regulation of tyrosine hydroxylase and tetra- protein Bim and suppresses its proapoptotic activity by
hydrobiopterin biosynthetic enzymes in PC12 cells by NGF, phosphorylation. J Biol Chem 277: 49511-49516.
EGF and IFNgamma. Brain Res 713: 125-133. Black JA, Langworthy K, Hinson AW, DibHajj SD, Waxman
Angelastro JM, Klimaschewski L, Tang S, Vitolo OV, SG. 1997. NGF has opposing effects on Na channel III
Weissman TA, et al. 2000. Identification of diverse nerve and SNS gene expression in spinal sensory neurons.
growth factorregulated genes by serial analysis of gene Neuroreport 8: 2331-2335.
expression (SAGE) profiling. Proc Natl Acad Sci USA 97: Boettger MK, Till S, Chen MX, Anand U, Otto WR, et al.
10424-10429. 2002. Calciumactivated potassium channel SK1 and IK1
Angelastro JM, Ignatova TN, Kukekov VG, Steindler DA, like immunoreactivity in injured human sensory neurones
Stengren GB, et al. 2003. Regulated expression of ATF5 is and its regulation by neurotrophic factors. Brain 125:
required for the progression of neural progenitor cells to 252-263.
neurons. J Neurosci 23: 4590-4600. Bournat JC, Allen JM. 2001. Regulation of the Y1 neuropep-
Angelastro JM, Mason JL, Ignatova TN, Kukekov VG, Steng- tide Y receptor gene expression in PC12 cells. Brain Res Mol
ren GB, et al. 2005. Downregulation of activating transcrip- Brain Res 90: 149-164.
tion factor 5 is required for differentiation of neural Bu G, Sun Y, Schwartz AL, Holtzman DM. 1998. Nerve
progenitor cells into astrocytes. J Neurosci 25: 3889-3899. growth factor induces rapid increases in functional cell
Apfel SC, Wright DE, Wiideman AM, Dormia C, Snider WD, surface low density lipoprotein receptorrelated protein. J
et al. 1996. Nerve growth factor regulates the expression of Biol Chem 273: 13359-13365.
brainderived neurotrophic factor mRNA in the peripheral Buchkovich KJ, Ziff EB. 1994. Nerve growth factor regulates
nervous system. Mol Cell Neurosci 7: 134-142. the expression and activity of p33cdk2 and p34cdc2 kinases
Araki T, Nagarajan R, Milbrandt J. 2001. Identification of in PC12 pheochromocytoma cells. Mol Biol Cell 5:
genes induced in peripheral nerve after injury. Expression 1225-1241.
profiling and novel gene discovery. J Biol Chem 276: 34131- Burstein DE, Greene LA. 1978. Evidence for RNA synthesis
34141. dependent and independent pathways in stimulation of
Arslan G, Kontny E, Fredholm BB. 1997. Downregulation of neurite outgrowth by nerve growth factor. Proc Natl Acad
adenosine A2A receptors upon NGFinduced differentia- Sci USA 75: 6059-6063.
tion of PC12 cells. Neuropharmacology 36: 1319-1326. Caillaud T, Opstal WY, Scarceriaux V, Billardon C, Rostene W.
Avila AM, DavilaGarcia MI, Ascarrunz VS, Xiao Y, Kellar KJ. 1995. Treatment of PC12 cells by nerve growth factor,
2003. Differential regulation of nicotinic acetylcholine re- dexamethasone, and forskolin. Effects on cell morphology
ceptors in PC12 cells by nicotine and nerve growth factor. and expression of neurotensin and tyrosine hydroxylase.
Mol Pharmacol 64: 974-986. Mol Neurobiol 10: 105-114.
54
2 Nerve growth factor regulated gene expression
Calissano P, Volonte C, Biocca S, Cattaneo A. 1985. Syn- expression of the alpha1 integrin subunit in PC12 cells. Biol
thesis and content of a DNAbinding protein with Chem 382: 969-972.
lactic dehydrogenase activity are reduced by nerve growth De Leon M, Nahin RL, Mendoza ME, Ruda MA. 1994. SR13/
factor in the neoplastic cell line PC12. Exp Cell Res 161: PMP22 expression in rat nervous system, in PC12 cells,
117-129. and C6 glial cell lines. J Neurosci Res 38: 167-181.
Carter JM, Waite KA, Campenot RB, Vance JE, Vance DE. Deng J, Szyf M. 1998. Multiple isoforms of DNA methyltrans-
2003. Enhanced expression and activation of CTP: Phos- ferase are encoded by the vertebrate cytosine DNA methyl-
phocholine cytidylyltransferase beta2 during neurite out- transferase gene. J Biol Chem 273: 22869-22872.
growth. J Biol Chem 278: 44988-44994. DeschenesFurry J, Belanger G, PerroneBizzozero N, Jasmin BJ.
Cavallaro S, DAgata V, Guardabasso V, Travali S, Stivala F, 2003. Posttranscriptional regulation of acetylcholines-
et al. 1995. Differentiation induces pituitary adenylate terase mRNAs in nerve growth factortreated PC12 cells
cyclaseactivating polypeptide receptor expression in PC by the RNAbinding protein HuD. J Biol Chem 278: 5710-
12 cells. Mol Pharmacol 48: 56-62. 5717.
Cavicchioli L, Flanigan TP, Vantini G, Fusco M, Polato P, et al. Dobashi Y, Bhattacharjee RN, Toyoshima K, Akiyama T. 1996.
1989. NGF amplifies expression of NGF receptor messen- Upregulation of the APC gene product during neuronal
ger RNA in forebrain cholinergic neurons of rats. Eur differentiation of rat pheochromocytoma PC12 cells. Bio-
J Neurosci 1: 258-262. chem Biophys Res Commun 224: 479-483.
Chang A, Toloza E, Bulinski JC. 1986. Changes in the expres- Doherty P, Mann DA, Walsh FS. 1988. Comparison of the
sion of beta and gamma actins during differentiation of effects of NGF, activators of protein kinase C, and a calcium
PC12 cells. J Neurochem 47: 1885-1892. ionophore on the expression of Thy1 and NCAM in PC12
Chao CC, Su LJ, Sun NK, Ju YT, Lih JC, et al. 2003. Involve- cell cultures. J Cell Biol 107: 333-340.
ment of Gas7 in nerve growth factorindependent and Drubin DG, Feinstein SC, Shooter EM, Kirschner MW. 1985.
dependent cell processes in PC12 cells. J Neurosci Res 74: Nerve growth factorinduced neurite outgrowth in PC12
248-254. cells involves the coordinate induction of microtubule as-
Chen J, Liu T, Ross AH. 1994. Downregulation of cmyc sembly and assemblypromoting factors. J Cell Biol 101:
oncogene during NGFinduced differentiation of neuro- 1799-1807.
blastoma cell lines. Chin Med Sci J 9: 152-156. Ebstein RP, Bennett ER, Sokoloff M, Shoham S. 1993. The
Cisneros B, Rendon A, Genty V, Aranda G, Marquez F, et al. effect of nerve growth factor on cholinergic cells in primary
1996. Expression of dystrophin Dp71 during PC12 cell fetal striatal cultures: Characterization by in situ hybridiza-
differentiation. Neurosci Lett 213: 107-110. tion. Brain Res Dev Brain Res 73: 165-172.
Colston JT, Valdes JJ, Chambers JP. 1998. Ca2 channel alpha Eva C, Fusco M, Bono C, Tria MA, Ricci Gamalero S, et al.
1subunit transcripts are differentially expressed in rat 1992. Nerve growth factor modulates the expression of
pheochromocytoma (PC12) cells following nerve growth muscarinic cholinergic receptor messenger RNA in telence-
factor treatment. Int J Dev Neurosci 16: 379-389. phalic neuronal cultures from newborn rat brain. Brain Res
Corcoran J, Maden M. 1999. Nerve growth factor acts via Mol Brain Res 14: 344-351.
retinoic acid synthesis to stimulate neurite outgrowth. FariasEisner R, Vician L, Silver A, Reddy S, Rabbani SA, et al.
Nat Neurosci 2: 307-308. 2000. The urokinase plasminogen activator receptor
Corness J, Stevens B, Fields RD, Hokfelt T. 1998. NGF and LIF (UPAR) is preferentially induced by nerve growth factor
both regulate galanin gene expression in primary DRG in PC12 pheochromocytoma cells and is required for NGF
cultures. Neuroreport 9: 1533-1536. driven differentiation. J Neurosci 20: 230-239.
Cosgaya JM, Aranda A. 2001. Nerve growth factor activates FariasEisner R, Vician L, Reddy S, Basconcillo R, Rabbani SA,
the RARbeta2 promoter by a Rasdependent mechanism. et al. 2001. Expression of the urokinase plasminogen acti-
J Neurochem 76: 661-671. vator receptor is transiently required during priming of
Costello B, Meymandi A, Freeman JA. 1990. Factors influen- PC12 cells in nerve growth factordirected cellular differen-
cing GAP43 gene expression in PC12 pheochromocytoma tiation. J Neurosci Res 63: 341-346.
cells. J Neurosci 10: 1398-1406. Feinstein SC, Dana SL, McConlogue L, Shooter EM, Coffino P.
DArcangelo G, Paradiso K, Shepherd D, Brehm P, Halegoua S, 1985. Nerve growth factor rapidly induces ornithine decar-
et al. 1993. Neuronal growth factor regulation of two boxylase mRNA in PC12 rat pheochromocytoma cells. Proc
different sodium channel types through distinct signal Natl Acad Sci USA 82: 5761-5765.
transduction pathways. J Cell Biol 122: 915-921. Fischer I, RichterLandsberg C, Safaei R. 1991. Regulation of
Danker K, Mechai N, Lucka L, Reutter W, Horstkorte R. 2001. microtubule associated protein 2 (MAP2) expression by
The small Gtpase ras is involved in growth factorregulated nerve growth factor in PC12 cells. Exp Cell Res 194: 195-201.
Nerve growth factor regulated gene expression
2 55
Forander P, Broberger C, Stromberg I. 2001. Glialcellline clusterin gene expression in PC12 cells. Biochem J 339
derived neurotrophic factor induces nerve fibre formation (Pt 3): 759-766.
in primary cultures of adrenal chromaffin cells. Cell Tissue Hashimoto H, Hagihara N, Koga K, Yamamoto K, Shintani N,
Res 305: 43-51. et al. 2000. Synergistic induction of pituitary adenylate
Frim DM, Wullner U, Beal MF, Isacson O. 1994. Implanted cyclaseactivating polypeptide (PACAP) gene expression
NGFproducing fibroblasts induce catalase and modify by nerve growth factor and PACAP in PC12 cells.
ATP levels but do not affect glutamate receptor binding J Neurochem 74: 501-507.
or NMDA receptor expression in the rat striatum. Exp Hayakawa K, Nakashima S, Ito Y, Mizuta K, Miyata H, et al.
Neurol 128: 172-180. 1999. Increased expression of phospholipase D1 mRNA
Furnish EJ, Zhou W, Cunningham CC, Kas JA, Schmidt CE. during cAMP or NGFinduced differentiation in PC12
2001. Gelsolin overexpression enhances neurite outgrowth cells. Neurosci Lett 265: 127-130.
in PC12 cells. FEBS Lett 508: 282-286. Heese K, Nagai Y, Sawada T. 2004. Nerve growth factor (NGF)
Gallo R, Zazzeroni F, Alesse E, Mincione C, Borello U, et al. induces mRNA expression of the new transcription factor
2002. REN: A novel, developmentally regulated gene that protein p48ZnF. Exp Mol Med 36: 130-134.
promotes neural cell differentiation. J Cell Biol 158: Henderson LP, Gdovin MJ, Liu C, Gardner PD, Maue RA.
731-740. 1994. Nerve growth factor increases nicotinic ACh receptor
Ghassemi F, DibHajj SD, Waxman SG. 2001. Beta1 adducin gene expression and current density in wildtype and protein
gene expression in DRG is developmentally regulated and is kinase Adeficient PC12 cells. J Neurosci 14: 1153-1163.
upregulated by glialderived neurotrophic factor and nerve Hens J, Nuydens R, Geerts H, Senden NH, Van de Ven WJ,
growth factor. Brain Res Mol Brain Res 90: 118-124. et al. 1998. Neuronal differentiation is accompanied by
Gibbs TC, Meier KE. 2000. Expression and regulation of NSPC expression. Cell Tissue Res 292: 229-237.
phospholipase D isoforms in mammalian cell lines. J Cell Higgins GA, Koh S, Chen KS, Gage FH. 1989. NGF induction
Physiol 182: 77-87. of NGF receptor gene expression and cholinergic neuronal
Gollapudi L, Neet KE. 1997. Different mechanisms for inhibi- hypertrophy within the basal forebrain of the adult rat.
tion of cell proliferation via cell cycle proteins in PC12 cells Neuron 3: 247-256.
by nerve growth factor and staurosporine. J Neurosci Res Hirayama K, Kapatos G. 1995. Regulation of GTP cyclohydro-
49: 461-474. lase I gene expression and tetrahydrobiopterin content by
Grant NJ, Claudepierre T, Aunis D, Langley K. 1996. Gluco- nerve growth factor in cultures of superior cervical ganglia.
corticoids and nerve growth factor differentially modulate cell Neurochem Int 27: 157-161.
adhesion molecule L1 expression in PC12 cells. J Neurochem Holtzman DM, Lee S, Li Y, ChuaCouzens J, Xia H, et al. 1996.
66: 1400-1408. Expression of neuronalNOS in developing basal forebrain
Greenberg ME, Greene LA, Ziff EB. 1985. Nerve growth factor cholinergic neurons: Regulation by NGF. Neurochem Res
and epidermal growth factor induce rapid transient 21: 861-868.
changes in protooncogene transcription in PC12 cells. Huang XC, Shenoy UV, Richards EM, Sumners C. 1997. Mo-
J Biol Chem 260: 14101-14110. dulation of angiotensin II type 2 receptor mRNA in rat
Greene LA, Rukenstein A. 1981. Regulation of acetylcholines- hypothalamus and brainstem neuronal cultures by growth
terase activity by nerve growth factor. Role of transcription factors. Brain Res Mol Brain Res 47: 229-236.
and dissociation from effects on proliferation and neurite Iacopino AM, Christakos S, Modi P, Altar CA. 1992. Nerve
outgrowth. J Biol Chem 256: 6363-6367. growth factor increases calcium binding protein (calbindin
Greene LA, Liem RK, Shelanski ML. 1983. Regulation of a D28K) in rat olfactory bulb. Brain Res 578: 305-310.
high molecular weight microtubuleassociated protein in Ichikawa M, Kikuchi T, Tateiwa H, Gotoh N, Ohta K, et al.
PC12 cells by nerve growth factor. J Cell Biol 96: 76-83. 2002. Role of PTBlike protein, a neuronal RNAbinding
Greene LA, Bernd P, Black MM, Burstein DE, Connolly JL, protein, during the differentiation of PC12 cells. J Biochem
et al. 1983. Genomic and nongenomic actions of nerve (Tokyo) 131: 861-868.
growth factor in development. Prog Brain Res 58: 347-357. Ikeda T, Kitayama S, Morita K, Dohi T. 2001. Nerve growth
Grumolato L, Louiset E, Alexandre D, AitAli D, Turquier V, factor downregulates the expression of norepinephrine
et al. 2003. PACAP and NGF regulate common and distinct transporter in rat pheochromocytoma (PC12) cells. Brain
traits of the sympathoadrenal lineage: Effects on electrical Res Mol Brain Res 86: 90-100.
properties, gene markers and transcription factors in dif- Jacovina AT, Zhong F, Khazanova E, Lev E, Deora AB, et al.
ferentiating PC12 cells. Eur J Neurosci 17: 71-82. 2001. Neuritogenesis and the nerve growth factorinduced
Gutacker C, Klock G, Diel P, KochBrandt C. 1999. Nerve differentiation of PC12 cells requires annexin IImediated
growth factor and epidermal growth factor stimulate plasmin generation. J Biol Chem 276: 49350-49358.
56
2 Nerve growth factor regulated gene expression
Jameson HL, Lillycrop KA. 2001. Nerve growth factor induces and carboxypeptidase H in rat PC12 cells. Neuroscience 49:
the expression of the LIM homeodomain transcription 443-450.
factor Isl1 with the kinetics of an immediate early gene in Lawlor KG, Narayanan R. 1992. Persistent expression of the
adult rat dorsal root ganglion. Neurosci Lett 309: 130-134. tumor suppressor gene DCC is essential for neuronal dif-
Jongsma Wallin H, Danielsen N, Johnston JM, Gratto KA, ferentiation. Cell Growth Differ 3: 609-616.
Karchewski LA, et al. 2001. Exogenous NT3 and NGF Lee NH, Malek RL. 1998. Nerve growth factor regulation
differentially modulate PACAP expression in adult sensory of m4 muscarinic receptor mRNA stability but not gene
neurons, suggesting distinct roles in injury and inflamma- transcription requires mitogenactivated protein kinase ac-
tion. Eur J Neurosci 14: 267-282. tivity. J Biol Chem 273: 22317-22325.
Joseph R, Tsang W, Dou D, Nelson K, Edvardsen K. 1996. Lee YJ, Zachrisson O, Tonge DA, McNaughton PA. 2002.
Neuronatin mRNA in PC12 cells: Downregulation by nerve Upregulation of bradykinin B2 receptor expression by neu-
growth factor. Brain Res 738: 32-38. rotrophic factors and nerve injury in mouse sensory neu-
Kane MD, Vanden Heuvel JP, Isom GE, Schwarz RD. 1998. rons. Mol Cell Neurosci 19: 186-200.
Differential expression of group I metabotropic glutamate Leonard DG, Ziff EB, Greene LA. 1987. Identification and
receptors (mGluRs) in the rat pheochromocytoma cell line characterization of mRNAs regulated by nerve growth fac-
PC12: Role of nerve growth factor and ras. Neurosci Lett tor in PC12 cells. Mol Cell Biol 7: 3156-3167.
252: 1-4. Levi A, Eldridge JD, Paterson BM. 1985. Molecular cloning of
Kaplan MD, Olschowka JA, OBanion MK. 1997. Cyclooxy- a gene sequence regulated by nerve growth factor. Science
genase1 behaves as a delayed response gene in PC12 cells 229: 393-395.
differentiated by nerve growth factor. J Biol Chem 272: Li XM, Qi J, Juorio AV, Boulton AA. 1997. Reciprocal regula-
18534-18537. tion of the content of aromatic Lamino acid decarboxylase
Kim SJ, Park K, Rudkin BB, Dey BR, Sporn MB, et al. 1994. and tyrosine hydroxylase mRNA by NGF in PC12 cells.
Nerve growth factor induces transcription of transforming J Neurosci Res 47: 449-454.
growth factorbeta 1 through a specific promoter element Lindenbaum MH, Carbonetto S, Mushynski WE. 1987. Nerve
in PC12 cells. J Biol Chem 269: 3739-3744. growth factor enhances the synthesis, phosphorylation, and
Kleijn M, Welsh GI, Scheper GC, Voorma HO, Proud CG, metabolic stability of neurofilament proteins in PC12 cells.
et al. 1998. Nerve and epidermal growth factor induce J Biol Chem 262: 605-610.
protein synthesis and eIF2B activation in PC12 cells. Lindsay RM. 1988. Nerve growth factors (NGF, BDNF) en-
J Biol Chem 273: 5536-5541. hance axonal regeneration but are not required for survival
Kobayashi H, Itoh S, Yanagita T, Yokoo H, Sugano T, et al. of adult sensory neurons. J Neurosci 8: 2394-2405.
2004. Expression of adrenomedullin and proadrenomedul- Lindsay RM, Harmar AJ. 1989. Nerve growth factor regulates
lin Nterminal 20 peptide in PC12 cells after exposure to expression of neuropeptide genes in adult sensory neurons.
nerve growth factor. Neuroscience 125: 973-980. Nature 337: 362-364.
Kojima M, Ikeuchi T, Hatanaka H. 1994. Nerve growth factor Lindsay RM, Shooter EM, Radeke MJ, Misko TP, Dechant G,
induces trkA mRNA expression in cultured basal forebrain et al. 1990. Nerve Growth Factor Regulates Expression of
cholinergic neurons from 17day fetal rats. Neurosci Lett the Nerve Growth Factor Receptor Gene in Adult Sensory
169: 47-50. Neurons. Eur J Neurosci 2: 389-396.
Kuklinski S, Vladimirova V, Waha A, Kamata H, Pesheva P, Liu H, Force T, Bloch KD. 1997. Nerve growth factor decreases
et al. 2003. Expression of galectin3 in neuronally differ- soluble guanylate cyclase in rat pheochromocytoma PC12
entiating PC12 cells is regulated both via Ras/MAPK cells. J Biol Chem 272: 6038-6043.
dependent and independent signalling pathways. Liu XW, Gong LJ, Guo LY, Katagiri Y, Jiang H, et al. 2001. The
J Neurochem 87: 1112-1124. Wilms tumor gene product WT1 mediates the downreg-
Kurihara K, Hosoi K, Ueha T, Nakanishi N, Yamada S. 1994. ulation of the rat epidermal growth factor receptor by nerve
Effects of nerve growth factor and dexamethasone on Na, growth factor in PC12 cells. J Biol Chem 276: 5068-5073.
K()ATPase of cultured PC12h cells. Horm Metab Res 26: Ma Y, Campenot RB, Miller FD. 1992. Concentration
14-18. dependent regulation of neuronal gene expression by
Lah JJ, Burry RW. 1993. Neuronotypic differentiation results nerve growth factor. J Cell Biol 117: 135-141.
in reduced levels and altered distribution of synaptophysin Machida CM, Rodland KD, Matrisian L, Magun BE, Ciment
in PC12 cells. J Neurochem 60: 503-512. G. 1989. NGF induction of the gene encoding the protease
Laslop A, Tschernitz C. 1992. Effects of nerve growth factor on transin accompanies neuronal differentiation in PC12 cells.
the biosynthesis of chromogranin A and B, secretogranin II Neuron 2: 1587-1596.
Nerve growth factor regulated gene expression
2 57
Mahata SK, Mahata M, Wu H, Parmer RJ, OConnor DT. Mizel SB, Bamburg JR. 1976. Studies on the action of nerve
1999. Neurotrophin activation of catecholamine storage growth factor. III. Role of RNA and protein synthesis in the
vesicle protein gene expression: Signaling to chromogranin process of neurite outgrowth. Dev Biol 49: 20-28.
a biosynthesis. Neuroscience 88: 405-424. Mobley WC, Neve RL, Prusiner SB, McKinley MP. 1988.
Mamet J, Lazdunski M, Voilley N. 2003. How nerve growth Nerve growth factor increases mRNA levels for the prion
factor drives physiological and inflammatory expressions of protein and the betaamyloid protein precursor in develop-
acidsensing ion channel 3 in sensory neurons. J Biol Chem ing hamster brain. Proc Natl Acad Sci USA 85: 9811-9815.
278: 48907-48913. Mohiuddin L, Fernandez K, Tomlinson DR, Fernyhough P.
Mandel G, Cooperman SS, Maue RA, Goodman RH, Brehm P. 1995. Nerve growth factor and neurotrophin3 enhance
1988. Selective induction of brain type II Na channels neurite outgrowth and upregulate the levels of messenger
by nerve growth factor. Proc Natl Acad Sci USA 85: RNA for growthassociated protein GAP43 and T alpha 1
924-928. alphatubulin in cultured adult rat sensory neurones.
Masiakowski P, Shooter EM. 1988. Nerve growth factor in- Neurosci Lett 185: 20-23.
duces the genes for two proteins related to a family of Mourey RJ, Vega QC, Campbell JS, Wenderoth MP, Hauschka
calciumbinding proteins in PC12 cells. Proc Natl Acad SD, et al. 1996. A novel cytoplasmic dual specificity protein
Sci USA 85: 1277-1281. tyrosine phosphatase implicated in muscle and neuronal
Mason JL, Angelastro JM, Ignatova TN, Kukekov VG, Lin G, differentiation. J Biol Chem 271: 3795-3802.
et al. 2005. ATF5 regulates the proliferation and differenti- Movsesyan V, Whalin M, Shibutani M, Katagiri Y, Broude E,
ation of oligodendrocytes. Mol Cell Neurosci 29: 372-380. et al. 1996. Downregulation of cyclin F levels during nerve
Mathew TC, Miller FD. 1990. Increased expression of T alpha growth factorinduced differentiation of PC12 cells. Exp
1 alphatubulin mRNA during collateral and NGFinduced Cell Res 227: 203-207.
sprouting of sympathetic neurons. Dev Biol 141: 84-92. Mukouyama Y, Kuroyanagi H, Shirasawa T, Tomoda T, Saffen
Max SR, Rohrer H, Otten U, Thoenen H. 1978. Nerve growth D, et al. 1997. Induction of protein tyrosine phosphatase
factormediated induction of tyrosine hydroxylase in rat epsilon transcripts during NGFinduced neuronal differen-
superior cervical ganglia in vitro. J Biol Chem 253: tiation of PC12D cells and during the development of the
8013-8015. cerebellum. Brain Res Mol Brain Res 50: 230-236.
McGuire JC, Greene LA, Furano AV. 1978. NGF stimulates Nakada N, Hongo S, Ohki T, Maeda A, Takeda M. 2002.
incorporation of fucose or glucosamine into an external Molecular characterization of NDRG4/Bdm1 protein iso-
glycoprotein in cultured rat PC12 pheochromocytoma forms that are differentially regulated during rat brain
cells. Cell 15: 357-365. development. Brain Res Dev Brain Res 135: 45-53.
Mearow KM. 1998. The effects of NGF and sensory nerve NaranjoSuarez S, Castellanos MC, AlvarezTejado M, Vara A,
stimulation on collateral sprouting and gene expression in Landazuri MO, et al. 2003. Downregulation of hypoxia
adult sensory neurons. Exp Neurol 151: 14-25. inducible factor2 in PC12 cells by nerve growth factor
Meisinger C, Hertenstein A, Grothe C. 1996. Fibroblast stimulation. J Biol Chem 278: 31895-31901.
growth factor receptor 1 in the adrenal gland and PC12 NurEKamal, Qureshi MM, Ijaz MK, Galadari SH, Raza H.
cells: Developmental expression and regulation by extrinsic 2000. Protooncogene ras GTPaselinked induction of
molecules. Brain Res Mol Brain Res 36: 70-78. glutathioneStransferase by growth factors in PC12 cells.
Metheny LJ, Skuse GR. 1996. NF1 mRNA isoform expression Int J Oncol 16: 1043-1048.
in PC12 cells: Modulation by extrinsic factors. Exp Cell Res Ohba H, Chiyoda T, Endo E, Yano M, Hayakawa Y, et al. 2004.
228: 44-49. Sox21 is a repressor of neuronal differentiation and is
Miller FD, Naus CC, Durand M, Bloom FE, Milner RJ. 1987. antagonized by YB1. Neurosci Lett 358: 157-160.
Isotypes of alphatubulin are differentially regulated during Oosawa H, Fujii T, Kawashima K. 1999. Nerve growth factor
neuronal maturation. J Cell Biol 105: 3065-3073. increases the synthesis and release of acetylcholine and the
Miller FD, Mathew TC, Toma JG. 1991. Regulation of nerve expression of vesicular acetylcholine transporter in primary
growth factor receptor gene expression by nerve growth cultured rat embryonic septal cells. J Neurosci Res 57:
factor in the developing peripheral nervous system. J Cell 381-387.
Biol 112: 303-312. Osaka H, Sabban EL. 1997. Requirement for cAMP/calcium
Min DS, Ahn BH, Rhie DJ, Yoon SH, Hahn SJ, et al. 2001. response element but not AP1 site in fibroblast growth
Expression and regulation of phospholipase D during neu- factor2elicited activation of tyrosine hydroxylase gene
ronal differentiation of PC12 cells. Neuropharmacology 41: expression in PC12 cells. Brain Res Mol Brain Res 49:
384-391. 222-228.
58
2 Nerve growth factor regulated gene expression
Paggi MG, Bonetto F, Severino A, Baldi A, Battista T, et al. Sampath D, Jackson GR, WerrbachPerez K, PerezPolo JR.
2001. The retinoblastomarelated Rb2/p130 gene is an ef- 1994. Effects of nerve growth factor on glutathione peroxi-
fector downstream of AP2 during neural differentiation. dase and catalase in PC12 cells. J Neurochem 62:
Oncogene 20: 2570-2578. 2476-2479.
Partlow LM, Larrabee MG. 1971. Effects of a nervegrowth Sano K, Kikuchi A, Matsui Y, Teranishi Y, Takai Y. 1989.
factor, embryo age and metabolic inhibitors on growth Tissuespecific expression of a novel GTPbinding protein
of fibres and on synthesis of ribonucleic acid and pro- (smg p25A) mRNA and its increase by nerve growth factor
tein in embryonic sympathetic ganglia. J Neurochem 18: and cyclic AMP in rat pheochromocytoma PC12 cells.
2101-2118. Biochem Biophys Res Commun 158: 377-385.
PeinadoRamon P, Wallen A, Hallbook F. 1998. MAP kinase Sharma E, Lombroso PJ. 1995. A neuronal protein tyrosine
phosphatase1 mRNA is expressed in embryonic sympa- phosphatase induced by nerve growth factor. J Biol Chem
thetic neurons and is upregulated after NGF stimulation. 270: 49-53.
Brain Res Mol Brain Res 56: 256-267. Sharma N, DArcangelo G, Kleinlaus A, Halegoua S,
Persengiev SP, Kondova II, Kilpatrick DL. 1999. E2F4 actively Trimmer JS. 1993. Nerve growth factor regulates the
promotes the initiation and maintenance of nerve growth abundance and distribution of K channels in PC12 cells.
factorinduced cell differentiation. Mol Cell Biol 19: 6048- J Cell Biol 123: 1835-1843.
6056. Sheehy AM, Phung YT, Riemer RK, Black SM. 1997. Growth
Persengiev SP, Li J, Poulin ML, Kilpatrick DL. 2001. E2F2 factor induction of nitric oxide synthase in rat pheochro-
converts reversibly differentiated PC12 cells to an irre- mocytoma cells. Brain Res Mol Brain Res 52: 71-77.
versible, neurotrophindependent state. Oncogene 20: Shibutani M, Lazarovici P, Johnson AC, Katagiri Y, Guroff G.
5124-5131. 1998. Transcriptional downregulation of epidermal growth
Pesheva P, Kuklinski S, Biersack HJ, Probstmeier R. 2000. factor receptors by nerve growth factor treatment of PC12
Nerve growth factormediated expression of galectin3 in cells. J Biol Chem 273: 6878-6884.
mouse dorsal root ganglion neurons. Neurosci Lett 293: Shiozuka K, Watanabe Y, Ikeda T, Hashimoto S, Kawashima H.
37-40. 1995. Cloning and expression of PCPTP1 encoding protein
Planas B, Kolb PE, Raskind MA, Miller MA. 1997. Nerve tyrosine phosphatase. Gene 162: 279-284.
growth factor induces galanin gene expression in the rat Smith MA, Fanger GR, OConnor LT, Bridle P, Maue RA.
basal forebrain: Implications for the treatment of choliner- 1997. Selective regulation of agrin mRNA induction
gic dysfunction. J Comp Neurol 379: 563-570. and alternative splicing in PC12 cells by Rasdependent
Prentice HM, Moore SE, Dickson JG, Doherty P, Walsh FS. actions of nerve growth factor. J Biol Chem 272:
1987. Nerve growth factorinduced changes in neural cell 15675-15681.
adhesion molecule (NCAM) in PC12 cells. Embo J 6: Sobue K, Kanda K. 1989. Alphaactinins, calspectin (brain
1859-1863. spectrin or fodrin), and actin participate in adhesion and
Probst JC, Landgraf R, Behl C. 1997. Expression of the trefoil movement of growth cones. Neuron 3: 311-319.
polypeptide ITF in PC12 cells. Biochem Mol Biol Int 42: Sohrabji F, Greene LA, Miranda RC, ToranAllerand CD.
425-432. 1994. Reciprocal regulation of estrogen and NGF receptors
Rhodes J, Lutka FA, JordanSciutto KL, Bowser R. 2003. by their ligands in PC12 cells. J Neurobiol 25: 974-988.
Altered expression and distribution of FAC1 during NGF Stefanis L, Kholodilov N, Rideout HJ, Burke RE, Greene LA.
induced neurite outgrowth of PC12 cells. Neuroreport 14: 2001. Synuclein1 is selectively upregulated in response to
449-452. nerve growth factor treatment in PC12 cells. J Neurochem
Rogers SW, Mandelzys A, Deneris ES, Cooper E, Heinemann S. 76: 1165-1176.
1992. The expression of nicotinic acetylcholine receptors by Stein R, Orit S, Anderson DJ. 1988. The induction of a neural
PC12 cells treated with NGF. J Neurosci 12: 4611-4623. specific gene, SCG10, by nerve growth factor in PC12 cells
Romano C, Nichols RA, Greengard P, Greene LA. 1987. is transcriptional, protein synthesis dependent, and gluco-
Synapsin I in PC12 cells. I. Characterization of the phos- corticoid inhibitable. Dev Biol 127: 316-325.
phoprotein and effect of chronic NGF treatment. Takahashi T, Yamashita H, Nakamura S, Ishiguro H, Nagatsu T,
J Neurosci 7: 1294-1299. et al. 1999. Effects of nerve growth factor and nicotine on
Rong P, Bennie AM, Epa WR, Barrett GL. 1999. Nerve growth the expression of nicotinic acetylcholine receptor subunits
factor determines survival and death of PC12 cells by regu- in PC12 cells. Neurosci Res 35: 175-181.
lation of the bclx, bax, and caspase3 genes. J Neurochem Takebayashi M, Hayashi T, Su TP. 2002. Nerve growth factor
72: 2294-2300. induced neurite sprouting in PC12 cells involves sigma1
Nerve growth factor regulated gene expression
2 59
receptors: Implications for antidepressants. J Pharmacol kinesin heavy chains during differentiation of human neu-
Exp Ther 303: 1227-1237. roblastoma and PC12 cells. Eur J Neurosci 8: 536-544.
Takekoshi K, Nomura F, Isobe K, Motooka M, Nammoku T, Villa A, Latasa MJ, Pascual A. 2001. Nerve growth factor
et al. 1998. Identification and initial characterization of modulates the expression and secretion of betaamyloid
stathmin by the differential display method in nerve precursor protein through different mechanisms in PC12
growth factortreated PC12 cells. Eur J Endocrinol 138: cells. J Neurochem 77: 1077-1084.
707-712. Weinberger RP, Henke RC, Tolhurst O, Jeffrey PL, Gunning P.
Tamaru T, Okada M, Nakagawa H. 1994. Differential ex- 1993. Induction of neuronspecific tropomyosin mRNAs
pression of D type cyclins during neuronal maturation. by nerve growth factor is dependent on morphological
Neurosci Lett 168: 229-232. differentiation. J Cell Biol 120: 205-215.
Taniguchi T, Morisawa K, Ogawa M, Yamamoto H, Fujimoto S. Winston J, Toma H, Shenoy M, Pasricha PJ. 2001. Nerve
1988. Decrease in the level of poly(ADPribose) synthetase growth factor regulates VR1 mRNA levels in cultures of
during nerve growth factorpromoted neurite outgrowth in adult dorsal root ganglion neurons. Pain 89: 181-186.
rat pheochromocytoma PC12 cells. Biochem Biophys Res Wion D, Le Bert M, Brachet P. 1988. Messenger RNAs of beta
Commun 154: 1034-1040. amyloid precursor protein and prion protein are regulated
Uittenbogaard M, Chiaramello A. 2002. Constitutive over- by nerve growth factor in PC12 cells. Int J Dev Neurosci 6:
expression of the basic helixloophelix Nex1/MATH2 tran- 387-393.
scription factor promotes neuronal differentiation of PC12 Woo CW, Lucarelli E, Thiele CJ. 2004. NGF activation of TrkA
cells and neurite regeneration. J Neurosci Res 67: 235-245. decreases Nmyc expression via MAPK path leading to a
Vedder H, Affolter HU, Otten U. 1993. Nerve growth factor decrease in neuroblastoma cell number. Oncogene 23:
(NGF) regulates tachykinin gene expression and biosynthe- 1522-1530.
sis in rat sensory neurons during early postnatal develop- Wood JN, Lillycrop KA, Dent CL, Ninkina NN, Beech MM,
ment. Neuropeptides 24: 351-357. et al. 1992. Regulation of expression of the neuronal POU
Velculescu VE, Vogelstein B, Kinzler KW. 2000. Analysing protein Oct2 by nerve growth factor. J Biol Chem 267:
uncharted transcriptomes with SAGE. Trends Genet 16: 17787-17791.
423-425. Yan GZ, Ziff EB. 1995. NGF regulates the PC12 cell cycle
Vetter J, Betz H. 1989. Expression of synaptophysin in the rat machinery through specific inhibition of the Cdk kinases
pheochromocytoma cell line PC12. Exp Cell Res 184: and induction of cyclin D1. J Neurosci 15: 6200-6212.
360-366. Yan GZ, Ziff EB. 1997. Nerve growth factor induces trans-
Vician L, Basconcillo R, Herschman HR. 1997. Identification cription of the p21 WAF1/CIP1 and cyclin D1 genes in
of genes preferentially induced by nerve growth factor PC12 cells by activating the Sp1 transcription factor.
versus epidermal growth factor in PC12 pheochromocyto- J Neurosci 17: 6122-6132.
ma cells by means of representational difference analysis. Zubiaur M, Neer EJ. 1993. Nerve growth factor changes G
J Neurosci Res 50: 32-43. protein levels and localization in PC12 cells. J Neurosci Res
Vician L, Silver AL, FariasEisner R, Herschman HR. 2001. 35: 207-217.
NID67, a small putative membrane protein, is preferential- Zur KB, Oh Y, Waxman SG, Black JA. 1995. Differential up
ly induced by NGF in PC12 pheochromocytoma cells. regulation of sodium channel alpha and beta 1subunit
J Neurosci Res 64: 108-120. mRNAs in cultured embryonic DRG neurons follow-
Vignali G, Niclas J, Sprocati MT, Vale RD, Sirtori C, et al. ing exposure to NGF. Brain Res Mol Brain Res 30:
1996. Differential expression of ubiquitous and neuronal 97-105.
3 Myelinating Cells in the Central
Nervous SystemDevelopment,
Aging, and Disease
J. Neman . J. de Vellis
1 Oligodendrocyte Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
1.1 General Developmental Stages/Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
1.2 Embryonic Origins of Oligodendrocyte in the Spinal Cord . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
1.2.1 Motor NeuronOligodendrocyte Precursor (MNOP) Cell Hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . 64
1.2.2 Dorsal Spinal Cord Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
1.2.3 Glial-Restricted Precursor Cell Hypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
1.3 Oligodendrocyte Development in the Forebrain . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
1.4 Oligodendrocyte Development from the SVZ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
1.5 Myelination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2 AxonMyelin Membrane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Abstract: Many neurological diseases are caused by myelin deficiencies, which can be the result of both
genetic and environmental factors. The myelin forming cell, oligodendrocyte (OL), is a major target for the
known causes of white matter diseases. During development, oligodendrocytes pass through a series of cell
phenotypes from undifferentiated stem cells to mature myelin-forming cells. The general idea that there
might be different subclasses of oligodendrocyte derived from different precursor subtypes is an area of
active debate. Can cells that are born of progenitors in the different parts of the embryo, under the influence
of different positional signals and expressing different sets of patterning genes, ever converge on precisely
the same phenotypic endpoint? The following sections will review literature about controversy over
oligodendrocyte origins and degenerative process resulting in demyelination in the postnatal aging brain
and Alzheimers disease.
List of Abbreviations: AD, Alzheimers Disease; AEP, anterior entopeduncular; APP, amyloid precursor
protein; bFGF, basic fibroblast growth factor; CGE, caudal ganglionic eminence; CNS, central nervous
system; Gsh2, genomic screened homeobox 2; LGE, lateral ganglionic eminence; MBP, myelin basic protein;
MGE, medial ganglionic eminence; MNOP, motor neuronoligodendrocyte precursor; NEP, neuroepithe-
lial cells; NFT, neurofibrillary tangles; Nrcam, neuronal-adhesion molecules; NRG, neuregulin; NRP,
neuron-restricted precursor; OL, oligodendrocyte; OLP, oligodendrocyte progenitor; PDGF, platelet-
derived growth factor; PDGFR/, platelet-derived growth factor receptor-/; PLP, proteolipid protein;
PSA-NCAM, polysialylated form of the neural cell adhesion molecule; SHH, sonic hedgehog; SVZ,
subventricular zone; VZ, ventricular zone
1 Oligodendrocyte Development
. Figure 31
Model for oligodendrocyte development. The figure illustrates the characteristics that accompany the sequen-
tial differentiation within the oligodendrocyte lineage beginning from very early stem cells to the mature
oligodendrocyte. Abbreviation for cell marker: SSEA1, stage specific embryonic antigen-1; A2B5, polysialic acid-
NCAM, ganglioside; CNP, 2,3-cyclic nucleotide-3-phosphohydrolase; GD3, ganglioside; GPDH, glycerolpho-
sphate dehydrogenase; GC, galactocerebroside glycolipid; O4, O1, ganglioside; PLP, proteolipid protein; MBP,
myelin basic protein; MOG, myelin oligodendrocyte glycoprotein; Tf, transferrin; GSTpi, glutothione S-trans-
ferases pi; PLP, proteolipid protein; GPDH, glycerol 3-phosphate dehydrogenase
greatly expand the OLP cell population over extended periods of time. Upon removal of bFGF or both
mitogens, the progenitor cells differentiate along the oligodendrocyte lineage.
In addition, insulin-like growth factor 1 has been implicated in OLP proliferation and differentiation.
While transforming, growth factor-beta 1 is also associated with differentiation because it inhibits the
mitogen-induced proliferation of OLP through exit from the cell cycle and thus might have a role in
triggering differentiation (Franklin, 2002).
The morphology of oligodendrocytes varies according to the axons that they myelinate. Those that
ensheath large-diameter axons have a large cell body that lies close to the axon and they may synthesize only
a single internodes worth of myelin, while other oligodendrocytes make as many as 30 internodes small
axons. There are also molecular differences between oligodendrocytes on large- versus small-bore axons
for example, in their gap junction proteins (connexins). It is not known whether these are intrinsic
differences or phenotypic variations of a single, plastic cell type. The general idea that there might be
different subclasses of oligodendrocyte is derived from different precursor subtypes is an area of active
64 3 Myelinating cells in the central nervous systemdevelopment, aging, and disease
debate. Can cells that are born of progenitors in different parts of the embryounder the influence of
different positional signals and expressing different sets of patterning genesever converge on precisely
the same phenotypic endpoint? The following sections will review literature about controversy over
oligodendrocyte origins (Richardson et al., 2006).
The MNOP hypothesis is based on observations that both motor neurons and oligodendrocytes arise in
similar zone of the ventral spinal cord with similar concentrations of Sonic Hedgehog (SHH) required for
the induction of both cell types. In addition, in vitro studies show that the induction of oligodendrocytes is
frequently accompanied by the induction of motor neurons (Pringle et al., 1996; Orentas et al., 1999).
Therefore, it seems logical to have motor neurons and oligodendrocytes developmentally related to each
other since ensheating the axons of these neurons with myelin would increase the conduction velocity.
Interestingly, in annelids and crustaceans, myelin-like membranes are associated with axons required for
rapid escape responses. Therefore, it would be intuitive to ensure that motor neurons and the cells that
ensheath them arise at the same place would be to derive both cells from the same precursor.
In 2002, it was discovered by three separate groups that the Olig gene family of basic HLH transcription
factors is involved in the developmental production of both motor neurons and oligodendrocytes (Lu et al.,
2002; Takebayashi et al., 2002; Zhou and Anderson, 2002). Specifically, Olig1 and Olig2 genes are
expressed in the developing mouse spinal cord within the specific region that appears to give rise to both
oligodendrocytes and motor neurons. Forced expression of Olig1 or Olig2 in neuroepithelial stem cells
induces expression of early markers of the oligodendrocyte lineage. Expression of Olig2 in conjunction with
neurogenin2 appears to be critical for the generation of motor neurons. Moreover, the targeted disruption
of Olig2 prevents oligodendrocyte and motor neuron specification in the spinal cord. Disruption of Olig1,
in contrast, disrupted normal maturation of oligodendrocytes (Noble et al., 2004).
Initially, it was believed that oligodendrocytes in the spinal cord are only derived from ventral sources.
However, recent new evidence shows that oligodendrocytes in the spinal cord are derived from both ventral
and dorsal sources (Cai et al., 2005; Fogarty et al., 2005; Vallstedt et al., 2005) (> Figure 32).
The Nkx6 transcription factors are normally expressed in the ventral part of the embryonic spinal cord
in progenitor domains p3, pMN, p2, and p1, and are activated by SHH signaling from the notochord and
Myelinating cells in the central nervous systemdevelopment, aging, and disease 3 65
. Figure 32
Oligodendrocyte development in the spinal cord. Neurons are formed before glia (astrocytes and oligodendro-
cyte precursors (OLPs)). In general, OLPs are formed before astrocytes and ventral cell types before dorsal. The
figure shows the expression domains of several transcription factors. Dashed bars indicate that the expression
domain boundaries shift during development, in the direction of the small arrowsfor example, expression of
the transcription factor Nkx2.2 expands dorsally, and expression of the developing brain homeobox gene Dbx1
contracts. Approximately 85% of all spinal cord oligodendrocytes are generated from pMN and the remainder
from other progenitor domains. It is not known whether astrocytes are also generated from pMN but, if so, they
are probably produced in small numbers relative to oligodendrocytes. Abbreviations: dP1dP6, dorsal progen-
itor domains; Msx3, a homeobox gene; Olig2, oligodendrocyte lineage gene 2; Pax7, paired box gene 7; pMN
and p0p3, ventral progenitor domains. (Reproduced with permission from Richardson et al. (2006).)
floor plate at the ventral midline. Once activated, they in turn activate Olig2, which as stated before, is
required for the generation of both motor neurons and oligodendrocyte precursors from progenitors in the
ventral progenitor domain pMN. Therefore, in spinal cords of Nkx6-null mice (specifically Nkx6.1 and
Nkx6.2) there is a loss of Olig2 expression in pMN leading to developmental blockage of both motor
neurons and oligodendrocytes.
However, oligodendrocyte precursors that express normal markers of platelet-derived growth factor
receptor-/ (PDGFR/) and Olig2 continue to be produced in the dorsal spinal cord of Nkx6-null mice. The
dorsal precursors coexpress paired box gene 7 (Pax7), confirming their dorsal origin. In wild-type mice,
some oligodendrocyte precursors in the dorsal part of the cord are also found to express Pax7, which
indicates that dorsal production is a normal phenomenon. These precursors were missed in previous
studies presumably because they are generated after their ventrally derived counterparts and mingle with
them unnoticed. There are fewer Pax7-expressing oligodendrocyte precursors in wild-type spinal cord than
in Nkx6 mutant cord, which suggests that ventrally produced oligodendrocyte precursors normally
suppress their dorsal counterparts, perhaps because they compete more effectively for essential proliferation
and/or survival signals such as PDGF alpha. Generation of the ventral precursors starts a couple of days
earlier than that of the dorsal ones (embryonic day (E) 12.5 compared with E15), so that they have plenty
of time to get preestablished (Richardson et al., 2006) (> Figure 33).
66 3 Myelinating cells in the central nervous systemdevelopment, aging, and disease
. Figure 33
Oligodendrocyte developmental sequence in developing spinal cord and forebrain. (a) In the mouse spinal
cord, 85% of oligodendrocyte precursors are generated from pMN in the ventral ventricular zones (1), starting
at about embryonic day (E)12.5. At about E15, generation of a secondary wave of precursors starts in more
dorsal regions by trans-differentiation of radial glia (2). (b) In the telencephalon, the ventral-most precursors in
the medial ganglionic eminence are produced from about E12.5 (1), production of the lateral ganglionic
eminence-derived precursors starts a few days later (2), and production of the cortex-derived precursors occurs
mainly after birth (3). Diagram not to scale. (Reproduced with permission from Richardson et al. (2006).)
Further evidence for dorsally derived oligodendrocytes has been provided by Cre-lox fate-mapping
experiments in transgenic mice under the control of regulatory elements surrounding the Dbx1 (developing
brain) homeobox gene. In these mice, Cre expression mirrors the normal pattern of Dbx1 expression, which
is restricted to neuroepithelial precursors in p1, p0, dP6, and dP5that is, four progenitor domains
centered on the dorsalventral midline. The Dbx1-derived oligodendrocytes comprised 3% of all oligo-
dendrocytes in the spinal cord and were less widely spread than the majoritybeing mainly located in the
lateral white matter radially opposite to their site of origin in the ventricular zone. Some Dbx1-derived,
Olig2-positive cells retained a radial process and transiently coexpressed the radial glial cell marker RC2,
which indicates that they are formed by direct interconversion from radial glia.
In more recent fate-mapping studies with Msx3Cre transgenic mice it appeared that 1015% of all
oligodendrocytes in the cervical spinal cord originate in the dorsal half of the cord. Many of these are
concentrated in the dorsal funiculus where they contribute up to 50% of the oligodendrocytes (Richardson
et al., 2006).
At the same time that studies have been ongoing on origin and relation of motor neurons and oligo-
dendrocytes, a separate line of investigation has raised questions about whether oligodendrocytes are
developmentally more closely related to astrocytes rather than motor neurons.
This latter investigation has led to the idea that during spinal cord development, the progression from
pluripotent neuroepithelial stem cell to differentiated cell types requires prior generation of lineage-
restricted precursor cells. Furthermore, it is thought that neurons and glial cells come from different
lineage-restricted precursor cells, NRP and GRP, respectively. GRP cells, in response to local environmental
signals, can generate either astrocytes or oligodendrocytes, with oligodendrocyte generation involving the
intermediate generation of OLPs (Noble et al., 2004).
Both GRP cells and neuron-restricted precursor (NRP) cells can be directly isolated from the devel-
oping rat spinal cord and grown as purified populations. Freshly isolated cells exhibit the same lineage
restrictions as those cells derived from neuroepithelial stem cells in vitro. Clonal studies have demonstrated
that GRP cells retain their tripotential nature even after weeks of in vitro expansion and several serial
Myelinating cells in the central nervous systemdevelopment, aging, and disease 3 67
reclonings and also exhibit these same restrictions following transplantation in vivo. GRP cells generate
both oligodendrocytes and astrocytes following transplantation into the brain or spinal cord, and do not
generate neurons even when they migrate into such neurogenic zones as the rostral migratory stream and
olfactory bulb. NRP cells, in contrast, generate only neurons (including motor neurons), even upon
transplantation into such CNS regions as the adult spinal cord (Noble et al., 2004).
GRP cells exist throughout the rostra caudal axis and were initially isolated from the developing
rat caudal neural tube around E12 to E14 and can be distinguished from neuroepithelial cells (NEP) by
the acquisition of A2B5 immunoreactivity and their initial lack of NG2 and PDGFRa expression, and
their limited adhesion to laminin and preference for fibronectin. GRP cells therefore represent an early
intermediate progenitor committed to glial lineage between NEP cells and fully differentiated glia and
may be one of the earliest born glial precursors (Liu and Rao, 2004).
NRP cells, in contrast, expressed the polysialylated form of the neural cell adhesion molecule
(PSA-NCAM) and were shown to give rise to multiple different kinds of neurons and not to glia (Noble
et al., 2004).
. Figure 34
Sources of oligodendrocytes in the developing forebrain. (a) Three waves of oligodendrocytes generated from
different lineages in the telencephalic ventricular zone. Cells from the Nkx2.1 population begin to migrate
around E12.5, those from the Gsh2 population around E15.5, and those from the Emx1 population around P0.
Nkx2.1 and Gsh2 expressions overlap in the ventricular zone overlying the medial ganglionic eminence. Also
shown are presumed migratory pathways of oligodendrocyte precursors from each of these areas. (b) Sources
of oligodendrocytes in the corpus callosum (square) at three time points. At P0, most of the cells arise from the
ventral, Nkx2.1 and Gsh2 domains. By P10, cells from the Nkx2.1 lineage have mostly disappeared, replaced by
a mix of Emx1 and Gsh2 cells. In the adult, the Emx1 and Gsh2 lineages make up most of the oligodendrocytes,
but an unlabeled population is also present. (Reproduced with permission from Ventura and Goldman (2006).)
a secondary proliferative population of progenitors that simultaneously generates both neurons and glia.
Progenitor cells specified to become glia migrate radially from the SVZ into the subcortical white matter,
corpus collosum, striatum, and cerebral cortex, where they differentiate into astrocytes and oligodendro-
cytes (Marshall et al., 2003). In addition, oligodendrocytes continue to be produced in the adult brain and
participate in myelin repair (Chari and Blakemore, 2002).
Furthermore, work from James Goldmans laboratory show that in the postnatal forebrain, oligo-
dendrocytes are generated from progenitor cells that reside near the tips of the lateral ventricles (Levison
et al., 1999). Yet, the primary progenitors for new oligodendrocytes born in the adult SVZ have not
been identified.
Recent work from Alvarex-Buylla and colleagues suggest that GFAP-positive type B astrocytes and a
subpopulation of actively dividing type C cells in the SVZ express the transcription factor Olig2. They
also show that type B cells generate a small number of nonmyelinating NG2-positive OLPs and mature
myelinating oligodendrocytes. Olig2-positive, polysialylated neural cell adhesion molecule-positive, PDGFRa,
and beta-tubulin-negative cells originating in the SVZ migrated into corpus callosum, striatum, and fimbria
fornix to differentiate into the NG2-positive nonmyelinating and mature myelinating oligodendrocytes.
Furthermore, primary clonal cultures of type B cells gave rise to oligodendrocytes alone or oligodendrocytes
and neurons. Importantly, the number of oligodendrocytes derived from type B cells in vivo increased fourfold
after a demyelinating lesion in corpus callosum, indicating that SVZ-derived oligodendrocytes participate
in myelin repair in the adult brain (Ligon et al., 2006; Menn et al., 2006) (> Figure 35).
Myelinating cells in the central nervous systemdevelopment, aging, and disease 3 69
. Figure 35
Olig2 contributes to ongoing oligodendrogenesis in progenitor cells of the adult SVZ. The cartoon depicts cells
surrounding the lateral ventricle (V). Type B cells give rise to rapidly cycling type C progenitor cells, which about
95% of the time produce type A neuroblasts that enter the rostral migratory stream and become olfactory
bulb interneurons (OBN). A small subpopulation (5%) of type C cells express Olig2 and give rise to NG2+ OLP
and mature oligodendrocytes. Abbreviation: E, ependymal layer. Reproduced with permission from Ligon
et al. (2006).)
It would be interesting to know what the relationship between the embryonic and postnatal germinal
zones is with respect to oligodendrocyte development. As suggested by Richardson and colleagues, when
neurogenesis comes to an end during late embryogenesis, the forebrain ventricular zone (VZ) regresses
until only a remnant remains at the cortico-striatal boundary, which remains active and continues to
generate new oligodendrocytes after birth and into adulthood. The postnatal VZ and its neighboring SVZ is
derived mainly from the embryonic LGE and lateral cortex, with no contribution from more ventral
regions. Therefore, the most ventral, MGE/AEP-derived progenitors leave no descendants in the postnatal
SVZ. This might contribute to the gradual loss of MGE/AEP-derived oligodendrocytes during postnatal life
(Richardson et al., 2006).
1.5 Myelination
Myelin, a multilamellar membrane, is formed by the spiral wrapping of glial (oligodendrocytes in the CNS
and by Schwann cells in the peripheral nervous system (PNS)) plasma membrane extensions around the
axon. The evolutionary need for the rapid and efficient conduction of action potentials in vertebrate
neurons has resulted in the development of the myelin sheath. Neuron-derived signaling molecules can and
do regulate the proliferation, differentiation, and survival of oligodendrocytes.
Myelin is a lipid-rich membrane (lipids constitute 70% of dry myelin weight) that is highly enriched in
glycosphingolipids and cholesterol. The major glycosphingolipids in myelin are galactosylceramide and its
sulfated derivative sulfatide (20% of lipid dry weight). The two major CNS myelin proteins are MBP and
the proteolipid proteins (PLP/DM20). During the active phase of myelination, each oligodendrocyte must
produce as much as 550 103 mm2 of myelin membrane surface area per day (Pfeiffer et al., 1993).
The timing of myelination is crucial because the ensheathment of axons must not occur before
neurons signal to oligodendrocytes. In turn, signals from oligodendrocytes to neurons are necessary to cluster
70 3 Myelinating cells in the central nervous systemdevelopment, aging, and disease
multiprotein complexes in the axonal membrane into distinct subdomains at the nodes of Ranvierthe
gaps between myelinated segments of neurons. Therefore, the next event after differentiation of OLPs
is the formation of myelin. Myelination is a multistep process requiring precise coordination of several
different signals.
Oligodendrocytes wrap their plasma membrane around axons that have a minimum axon caliber of
2.0 mm diameter. The physiological rationale for this requirement could be that saltatory conduction would
not be significantly enhanced on smaller diameter fibers. The g-ratio is the ratio of the axonal diameter
divided by the diameter of the axons and its myelin sheath. Most myelinated axons have an approximate
g-ratio of 0.6. Therefore, larger axons have thicker myelin, while smaller ones have thinner myelin. The
exception to the constant g-ratio is seen in axons that have been remyelinated and are therefore typically
thinner than expected (Sherman and Brophy, 2005).
The neuregulin (NRG)-ErbB receptor system has been shown to have considerable influence on the
thickness of the myelin sheath. Recent studies show that NRG-1 type III on the axonal surface is required
for the myelination by Schwann cells in the PNS (Taveggia et al., 2005). Transgenic mice with reduced NRG-
1 expression display hypomyelination, whereas over expression of NRG-1 induces increased myelin
thickness (Nave and Salzer, 2006). Myelin-forming Schwann cells thus appear to use NRG-1 signals to
know whether and to what extent axons require myelination. The signaling pathways involved in the CNS
are not known. Whether this signaling system operates in CNS myelination remains an open question of
major importance for human demyelinating diseases (Simons and Trajkovic, 2006).
After oligodendrocytes have established proper contact with the axonal membrane, the process of
myelination begins. One signal that seems to be required to trigger myelination is the electrical activity from
the neuron that leads to secretion of promyelinating factors of adenosine from neurons and/or leukemia
inhibitory factor from astrocytes (Demerens et al., 1996; Stevens et al., 2002; Ishibashi et al., 2006). Once the
coordinated differentiation of the axon and myelination has occurred, the axonal membrane is differen-
tiated into distinct molecular, structural, and functional domains. These domains include the nodes of
Ranvier, the paranodal junction, the juxtaparanodes, and the internodal regions.
2 AxonMyelin Membrane
The Nodes of Ranvier are periodical interruptions in the myelin sheath that are spaced at intervals that are
about 100 times the axonal diameter. These gaps are required for efficient and rapid propagation of action
potentials. The nodes are flanked by paranodal loops that form septate-like junctions with the axonal
membrane. Adjacent to them are the juxtaparanodal domains that lie under the compact myelin sheath.
There are some structural differences between PNS and CNS. In peripheral nerves, the entire myelin unit is
covered by a basal lamina, and the outermost layer of the Schwann cell extends microvilli that cover the
nodes. The perinodal space that lies between the axolemma and the basal lamina contains the microvilli and
is also filled with a filamentous matrix. In addition, in the CNS there are no basal lamina and the nodes are
contacted by perinodal astrocytes (synantocytes) (Butt et al., 2005).
The nodes are characterized by a high density of Na+ channels that are essential for the generation of the
action potential during saltatory conduction. Nodes of Ranvier in the adult CNS and PNS mostly contain
Nav1.6 (Yu et al., 2003). During development, both PNS and CNS nodes express Nav1.2, which is later
replaced by Nav1.6 that may allow neurons to adapt to high-frequency firing (Kaplan et al., 2001). In
addition, there exist two K+ channels at the nodes. Kv3.1 is mainly found in large axons in the CNS and only
in few nodes in the PNS, whereas Kcnq2 is located in all PNS nodes and most CNS nodes (Devaux et al.,
2003). Several transmembrane and cytoskeleton proteins have been identified at the node as well. They
include the neuronal-adhesion molecules (Nrcam), neurofascin-186, and their ligand gliomedin (Eshed
et al., 2005), the cytoskeleton adaptor/scaffolding protein ankyrin G, and the actin-binding protein spectrin
Beta-IV (Garrido et al., 2003). In general, it is believed that interaction with neurofascin, NrCAM, and
gliodemin recruits ankyrin G that contains binding sites for Na+ channels. Recruitment of stabilizing
components such as bIV spectrin further enhances the clustering of Na+ channels by anchoring the complex
more firmly to the axonal cytoskeleton (Simons and Trajkovic, 2006).
Myelinating cells in the central nervous systemdevelopment, aging, and disease 3 71
The paranodal junction attaches the myelin sheath and serves as a fence that limits lateral diffusion of
axolemmal proteins and to separate the electrical activity at the node of Ranvier from the internodal region
under the compact myelin sheath. Both in the PNS and the CNS, the formation of the paranodes is
dependent on axonglia interactions. The interaction of the loops that form septate-like junctions with the
axonal membrane is mediated by neurofascin-155 and contactin (Sherman and Brophy, 2005). The
interaction of neurofascin-155 in glia to a complex of contactin and contactin-associated protein (Caspr)
in axons initiates the assembly of the paranodal complex, which is then stabilized by interactions with a
specialized axonal cytoskeleton consisting of proteins such as the membrane skeleton component protein
4.1B, ankyrin B, and a/bII spectrin (Ogawa et al., 2006).
The juxtaparanode is located in a short zone just beyond the innermost paranodal junction. Caspr2, the
second member of the Caspr family, interacts with heteromultimers of the delayed rectifier K+ channels of
the Shaker family, Kv1.1, Kv1.2, and Kv2. In addition, Kv1.6 is present at this site, predominantly in
small axons. These K+ channels may act as an active damper of reexcitation and to help in maintaining
the internodal resting potential Two other proteins that are found at the juxtaparanodes are transient
axonal glycoprotein-1 (Tag1), a GPI-anchored CAM that is related to contactin118, and connexin 29
(Cx29), which is found at the glial membrane. Two recent studies showed that Caspr2 and Tag1 form a
juxtaparanodal complex, consisting of a glial Tag1 molecule and an axonal Caspr2/Tag1 heterodimer. This
complex is essential for the accumulation of K+ channels in the juxtaparanodes, as targeted disruption of
Caspr2 or Tag1 results in a striking reduction in the juxtaparanodal accumulation of these channels in both
PNS and CNS axons (Poliak and Peles, 2003) (> Figure 36 and > 37).
Several functions of the myelin-producing cells, the oligodendrocytes, make it an integral cell of the CNS
and make it pertinent issue when discussing aging and pathology of Alzheimers Disease (AD). Most studies
on the effects of aging on the nervous system have focused on the changes that occur as a consequence of
AD. The age risk factor is present in nongenetic and genetic forms of AD; with increased amyloid beta
peptides (Ab), deposition in the latter form. Myelin breakdown may be a contributing factor to the
pathology of both aging and AD. In addition, these breakdowns have been observed early in disease in
spite of infarction, Wallerian degeneration, or white matter amyloid angiopathy (Bartzokis et al., 2004). An
attribute to this white matter breakdown maybe due to the accumulation Ab, reduced cholesterol, and
myelin proteins (Roher et al., 2002).
The dependence of the brain on oligodendrocyte-produced cholesterol has implications for CNS
development and its continual functional plasticity. All brain cholesterol is synthesized de novo by
oligodendrocytes. The human brain, which is approximately 2% of the body by weight, contains approxi-
mately 25% of the bodys membrane cholesterol. It is thus not surprising that myelin membrane changes
are found to drive brain lipid changes with age as well as species differences in membrane composition.
Cholesterol enrichment may contribute to the free exchange of cholesterol from oligodendrocytes to
neurons and astrocytes with the aid of apolipoproteins. The low water binding produced by high
cholesterol levels in myelin breakdown may promote the hydrophobic ends of Ab aggregates to preferen-
tially interact with and damage myelin.
As mentioned earlier, myelin is essential for normal brain function. Myelination results in salutatory
conduction that results in increases signal transmission. Once myelin function in salutatory conduction is
compromised, there is a decrease in transmission velocity and refractory period of the axon. These changes
and disruption due to these delays can have impact on synchronization of impulses. The effects of this
desynchronization would be most apparent on brain functions that involve encoding and retrieval of
memories. Aging reduces 50% number of fast-conducting CNS axons in cats. Human stereological studies
estimate that the total length of myelinated axons is reduced by 2745% in old age, primarily through loss
of fibers with small diameter that myelinate later in development and are most susceptible to Ab pathology.
This aging-related myelin breakdown negatively impacts cognitive performance in primates and humans.
The age-related loss of myelin function may also explain the conduction delays observed in aging animals,
72 3 Myelinating cells in the central nervous systemdevelopment, aging, and disease
. Figure 36
Myelinated axons. (a) Myelinating glial cells, oligodendrocytes in the central nervous system (CNS), or Schwann
cells in the PNS form the myelin sheath by enwrapping their membrane several times around the axon. Myelin
covers the axon at intervals (internodes), leaving bare gapsthe nodes of Ranvier. Oligodendrocytes can
myelinate different axons and several internodes per axon, whereas Schwann cells myelinate a single internode
in a single axon. (b) Schematic longitudinal cut of a myelinated fiber around the node of Ranvier showing a
heminode. The node, paranode, juxtaparanode (JXP), and internode are labeled. The node is contacted by
Schwann cell microvilli in the PNS or by processes from perinodal astrocytes in the CNS. Myelinated fibers in the
PNS are covered by a basal lamina. The paranodal loops form a septate-like junction (SpJ) with the axon. The
juxtaparanodal region resides beneath the compact myelin next to the paranode (PN). The internode extends
from the juxtaparanodes and lies under the compact myelin. (c) Schematic cross-section of a myelinated nerve
depicting the inner and outer mesaxons (IMA and OMA, respectively). (d) Drawing of the specializations found
along the internodes. A strand composed of paranodal molecules (Caspr, Contactin; inner line) flanked by
juxtaparanodal proteins (Caspr2, K+ channels and TAG-1; outer lines) extends along the internodal region (the
juxtamesaxon) and below the SchmidtLanterman incisures (the juxtaincisure). In addition, Nf155 as well as
connexins 29 and 32 are found at the glial side, opposite to these axonal strands. (Reproduced with permission
from Poliak and Peles (2003).)
Myelinating cells in the central nervous systemdevelopment, aging, and disease 3 73
. Figure 37
Molecular composition of domains at the Nodes of Ranvier. Components at the nodes include neurofascin-186,
NrCAM, and voltage-gated Na+ and K+ channels (Nav1.2, Nav1.6, Kv3.1, and Kcnq2), which are tethered to a
complex containing ankyrin G and bIV-spectrin. The paranodes contain a complex of Caspr, contactin, and 4.1B
at the axonal membrane, which binds to neurofascin-155 (NF155) on the septate junction of a paranodal loop.
The multiprotein complex in the juxtaparanode contains a cis complex of Caspr2 and TAG-1, which interact with
4.1B and a PDZ domain-containing protein associated with the two shaker-type K+ channels (Kv1.1 and 1.2).
This complex is linked through a trans interaction with TAG-1 to the glial membrane
humans, and patients with AD. Myelin loss may also underlie the reduced myelin staining in postmortem
studies of aging and the aging-related loss of brain volume (Bartzokis, 2004).
Since cholesterol- and lipid-synthesizing enzymes require iron to function, oligodendrocytes have the
highest iron content of all and as much as 70% of brain iron is associated with myelin. Age-related increases
in iron levels may contribute to the increased intracellular oxidation necessary to trigger oligodendrocyte
precursors to differentiate. Inadequate iron levels result in poor myelination and mental deficiencies in
children. Normal ferritin, a spherical protein in which more than of 90% of tissue nonheme iron is stored,
can sequester and store iron and other transition metals. Many normal as well as pathological processes that
have been shown to damage oligodendrocytes can also release iron from ferritin. Oligodendrocytes may be
more vulnerable than other cells to such iron releases since in addition to containing the highest iron stores,
their particular ferritin subunit composition makes iron available with greater ease than in other cells.
Recent evidence suggests that elevated iron levels increase the production of amyloid precursor protein
(APP) and that the soluble Ab can act as an iron chelator. However, iron and other transition metals such as
copper and zinc can also promote Ab oligomerization. Oligomerization makes Ab toxic, thus making
the homeostasis of iron and Ab critically important. Therefore, late-myelinating oligodendrocytes and
their precursors are present at the cortical site of amyloid beta deposits observed in aging and AD
(Xu et al., 2001). In addition, iron and Ab to promote the formation of reactive oxygen species (Butterfield,
2003). This could explain as to why myelin breakdown from older and AD patients has been attributed to
increased levels of lipid peroxidation.
Since myelination markedly reduces neuronal energy expenditure, the loss of axonal myelin would
require an estimated increase of up to 5000-fold in neuronal energy expenditure in order to maintain
neurotransmission levels. Approximately 23% of the oxygen consumed in normal mitochondrial
respiration is obligatorily transformed into free radicals, and with aging, an increasing percentage of oxygen
74 3 Myelinating cells in the central nervous systemdevelopment, aging, and disease
is converted to superoxide. The aging-related loss/dysfunction of myelin would result in a further increase
in the production of damaging free radicals. Both neurons and especially oligodendrocytes are very
susceptible to damage from free radicals. Free-radical (oxidative) damage has been shown to be strongly
aging related and has been implicated in the pathophysiology of AD. An increase in neuronal free-radical
production has also been postulated to contribute to AD tangle-related neuropathology. Oxidation of tau
induces its dimerization and polymerization into insoluble filaments, the precursor to the intraneuronal
neurofibrillary tangles (NFT), the second pathognomonic lesion observed in AD brain. These damaging
oxidative processes are also observed in other neurodegenerative diseases and it is thus not surprising
that many other neurodegenerative disorders manifest NFTs while normal individuals rarely do so
(Bartzokis, 2004).
Oligodendrocytes are markedly heterogeneous based on when in the protracted process of human brain
development they differentiated into myelin-producing cells. Oligodendrocytes that differentiated late in
life ensheath 3050 smaller diameter axons as opposed to 1 oligodendrocyte per myelin segment of large
CNS motor and primary sensory area axons. These late-differentiating cells cannot produce the same
myelin thickness per axon segment as earlier-myelinating oligodendrocytes. The thinner, later-myelinating
sheaths are more susceptible to functional impairment and destruction. In addition, later-differentiating
oligodendrocytes have different lipid properties, may have a slower rate of myelin turnover, and reduced
ability for myelin repair than earlier-differentiating cells. This development-dependent oligodendrocyte
heterogeneity could contribute to reason as to why the neocortical regions of the brain are the most
vulnerable to developing AD lesions that consist of extracellular amyloid neuritic plaques and intraneuronal
NFT. Furthermore, the hippocampal and primary sensory and motor areas develop lesions only later on
in the disease process. Interestingly, it is the late-myelinating neocortical regions that are the most
vulnerable to AD lesions first, whereas the early-myelinating primary sensory and motor areas are affected
last (Braak et al., 2006).
References
Bartzokis G. 2004. Age-related myelin breakdown: A develop- Demerens C, Stankoff B, Logak M, Anglade P, Allinquant B,
mental model of cognitive decline and Alzheimers disease. et al. 1996. Induction of myelination in the central nervous
Neurobiol Aging 25(1): 5-18. system by electrical activity. Proc Natl Acad Sci USA 93(18):
Bartzokis G, Sultzer D, Lu PH, Nuechterlein KH, Mintz J, 9887-9892.
et al. 2004. Heterogeneous age-related breakdown of white Devaux J, Alcaraz G, Grinspan J, Bennett V, Joho R, et al.
matter structural integrity: Implications for cortical dis- 2003. Kv3.1b is a novel component of CNS nodes. J Neu-
connection in aging and Alzheimers disease. Neurobiol rosci 23(11): 4509-4518.
Aging 25(7): 843-851. Eshed Y, Feinberg K, Poliak S, Sabanay H, Sarig-Nadir O, et al.
Braak H, Rub U, Schultz C, Del Tredici K. 2006. Vulnerability 2005. Gliomedin mediates Schwann cell-axon interaction
of cortical neurons to Alzheimers and Parkinsons diseases. and the molecular assembly of the nodes of Ranvier.
J Alzheimers Dis 9(3): 35-44. Neuron 47(2): 215-229.
Butt AM, Hamilton N, Hubbard P, Pugh M, Ibrahim M. 2005. Fogarty M, Richardson WD, Kessaris N. 2005. A subset of
Synantocytes: The fifth element. J Anat 207(6): 695-706. oligodendrocytes generated from radial glia in the dorsal
Butterfield DA. 2003. Amyloid beta-peptide [142]-asso- spinal cord. Development 132(8): 1951-1959.
ciated free radical-induced oxidative stress and neurode- Franklin RJ. 2002. Why does remyelination fail in multiple
generation in Alzheimers disease brain: Mechanisms and sclerosis? Nat Rev Neurosci 3(9): 705-714.
consequences. Curr Med Chem 10(24): 2651-2659. Garrido JJ, Giraud P, Carlier E, Fernandes F, Moussif A, et al.
Cai J, Qi Y, Hu X, Tan M, Liu Z, et al. 2005. Generation of 2003. A targeting motif involved in sodium channel clustering
oligodendrocyte precursor cells from mouse dorsal spinal at the axonal initial segment. Science 300(5628): 2091-2094.
cord independent of Nkx6 regulation and Shh signaling. Ishibashi T, Dakin KA, Stevens B, Lee PR, Kozlov SV, et al.
Neuron 45(1): 41-53. 2006. Astrocytes promote myelination in response to elec-
Chari DM, Blakemore WF. 2002. Efficient recolonisation of trical impulses. Neuron 49(6): 823-832.
progenitor-depleted areas of the CNS by adult oligoden- Ivanova A, Nakahira E, Kagawa T, Oba A, Wada T, et al. 2003.
drocyte progenitor cells. Glia 37(4): 307-313. Evidence for a second wave of oligodendrogenesis in the
Myelinating cells in the central nervous systemdevelopment, aging, and disease 3 75
postnatal cerebral cortex of the mouse. J Neurosci Res 73(5): Richardson WD, Kessaris N, Pringle N. 2006. Oligodendro-
581-592. cyte wars. Nat Rev Neurosci 7(1): 11-18.
Kaplan MR, Cho MH, Ullian EM, Isom LL, Levinson SR, et al. Roher AE, Weiss N, Kokjohn TA, Kuo YM, Kalback W, et al.
2001. Differential control of clustering of the sodium chan- 2002. Increased A beta peptides and reduced choles-
nels Na(v)1.2 and Na(v)1.6 at developing CNS nodes of terol and myelin proteins characterize white matter
Ranvier. Neuron 30(1): 105-119. degeneration in Alzheimers disease. Biochemistry 41(37):
Levison SW, Young GM, Goldman JE. 1999. Cycling cells in 11080-11090.
the adult rat neocortex preferentially generate oligoden- Sherman DL, Brophy PJ. 2005. Mechanisms of axon ensheath-
droglia. J Neurosci Res 57(4): 435-446. ment and myelin growth. Nat Rev Neurosci 6(9): 683-690.
Ligon KL, Fancy SP, Franklin RJ, Rowitch DH. 2006. Olig gene Simons M, Trajkovic K. 2006. Neuron-glia communication in
function in CNS development and disease. Glia 54(1): 1-10. the control of oligodendrocyte function and myelin bio-
Liu Y, Rao MS. 2004. Glial progenitors in the CNS and genesis. J Cell Sci 119(Pt 21): 4381-4389.
possible lineage relationships among them. Biol Cell 96(4): Spassky N, Olivier C, Cobos I, LeBras B, Goujet-Zalc C, et al.
279-290. 2001. The early steps of oligodendrogenesis: Insights from
Lu QR, Sun T, Zhu Z, Ma N, Garcia M, et al. 2002. Common the study of the plp lineage in the brain of chicks and
developmental requirement for Olig function indicates a rodents. Dev Neurosci 23(45): 318-326.
motor neuron/oligodendrocyte connection. Cell 109(1): Stevens B, Porta S, Haak LL, Gallo V, Fields RD. 2002. Aden-
75-86. osine: A neuron-glial transmitter promoting myelination in
Marshall CA, Suzuki SO, Goldman JE. 2003. Gliogenic and the CNS in response to action potentials. Neuron 36(5):
neurogenic progenitors of the subventricular zone: Who 855-868.
are they, where did they come from, and where are they Takebayashi H, Nabeshima Y, Yoshida S, Chisaka O, Ikenaka
going? Glia 43(1): 52-61. K, et al. 2002. The basic helix-loop-helix factor olig2 is
Menn B, Garcia-Verdugo JM, Yaschine C, Gonzalez-Perez O, essential for the development of motoneuron and oligo-
Rowitch D, et al. 2006. Origin of oligodendrocytes in the dendrocyte lineages. Curr Biol 12(13): 1157-1163.
subventricular zone of the adult brain. J Neurosci 26(30): Taveggia C, Zanazzi G, Petrylak A, Yano H, Rosenbluth J, et al.
7907-7918. 2005. Neuregulin-1 type III determines the ensheathment
Nave KA, Salzer JL. 2006. Axonal regulation of myelination by fate of axons. Neuron 47(5): 681-694.
neuregulin 1. Curr Opin Neurobiol 16(5): 492-500. Tekki-Kessaris N, Woodruff R, Hall AC, Gaffield W, Kimura S,
Noble M, Proschel C, Mayer-Proschel M. 2004. Getting a GR(i) et al. 2001. Hedgehog-dependent oligodendrocyte lineage
P on oligodendrocyte development. Dev Biol 265(1): 33-52. specification in the telencephalon. Development 128(13):
Ogawa Y, Schafer DP, Horresh I, Bar V, Hales K, et al. 2006. 2545-2554.
Spectrins and ankyrinB constitute a specialized paranodal Vallstedt A, Klos JM, Ericson J. 2005. Multiple dorsoventral
cytoskeleton. J Neurosci 26(19): 5230-5239. origins of oligodendrocyte generation in the spinal cord
Orentas DM, Hayes JE, Dyer KL, Miller RH. 1999. Sonic and hindbrain. Neuron 45(1): 55-67.
hedgehog signaling is required during the appearance Ventura RE, Goldman JE. 2006. Telencephalic oligodendro-
of spinal cord oligodendrocyte precursors. Development cytes battle it out. Nat Neurosci 9(2): 153-154.
126(11): 2419-2429. Xu J, Chen S, Ahmed SH, Chen H, Ku G, et al. 2001. Amyloid-
Pfeiffer SE, Warrington AE, Bansal R. 1993. The oligodendro- beta peptides are cytotoxic to oligodendrocytes. J Neurosci
cyte and its many cellular processes. Trends Cell Biol 3(6): 21(1): RC118.
191-197. Yu FH, Westenbroek RE, Silos-Santiago I, McCormick KA,
Poliak S, Peles E. 2003. The local differentiation of myelinated Lawson D, et al. 2003. Sodium channel beta4, a new disul-
axons at nodes of Ranvier. Nat Rev Neurosci 4(12): 968-980. fide-linked auxiliary subunit with similarity to beta2.
Pringle NP, Yu WP, Guthrie S, Roelink H, Lumsden A, et al. J Neurosci 23(20): 7577-7585.
1996. Determination of neuroepithelial cell fate: Induction Zhou Q, Anderson DJ. 2002. The bHLH transcription factors
of the oligodendrocyte lineage by ventral midline cells and OLIG2 and OLIG1 couple neuronal and glial subtype spec-
sonic hedgehog. Dev Biol 177(1): 30-42. ification. Cell 109(1): 61-73.
4 SulfurContaining Amino Acids in
the CNS: Homocysteine
D. K. Rassin
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
6 Methodological Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Abstract: The sulfur amino acid metabolic pathway plays an important role in supporting a number of
functions within the CNS. The metabolic intermediate, homocysteine, sits at an important branch point in
this pathway and disturbances related to this compound are related to detrimental outcomes for the CNS.
Homocysteine first came to notice when biological dysfunction related to its excess was reported in
individuals with the inherited metabolic defect homocystinuria. Subsequently, a number of biological
dysfunctions have been associated with this amino acid unrelated to metabolic diseases. Increased homo-
cysteine has been identified as an epidemiologic risk factor for neural tube defects, cardiovascular disease,
cerebrovascular disease, Alzheimers disease and, potentially, several dysfunctions that occur in the new-
born. The mechanisms by which homocysteine may exert detrimental effects on the CNS include oxidative
damage, apoptotic induction, synthesis of the excitotoxic metabolite homocysteic acid, and inhibition of
energy-related enzymes. Methodologic issues, related to its disulfide/sulfhydryl structure and to the small
absolute increases that are associated with increased risk, make it difficult to define homocysteines role in
the CNS. The potential importance of modulating the amount of this amino acid is tempered by the
numerous unanswered questions about the mechanism of its functions.
List of Abbreviations: CNS, central nervous system; GABA, gaminobutyric acid; MTHFR, N5,N10
methylenetetrahydrofolate reductase; NMDA, NmethylDaspartate; PARP, polyADPribose polymerase;
SAH, Sadenosylhomocysteine; SAM, Sadenosylmethionine
1 Introduction
The methionine metabolic pathway supports numerous functions within the central nervous system (CNS)
beyond providing an essential amino acid precursor for protein synthesis. These functions include
neurotransmitter synthesis, methyl group donation, polyamine precursor, osmotic protection, antioxidant
synthesis, and DNA salvage synthesis. In previous editions of this Handbook the inherited genetic defects
related to this pathway have been described (Gaull, 1973), and the general metabolism of methionine,
especially as related to nutrition and development, has been reviewed (Tallan et al., 1983).
In this chapter a general overview of the various constituents of this pathway and their role in the CNS
are presented followed by a discussion of the role of homocysteine, as an agent that may have implications
for potential damage to the brain. In the years since the first edition of this Handbook, there has been an
evolution of our understanding of the role that homocysteine plays in biology. (Homocysteine exists in a
variety of forms, see > Sect. 6, throughout this chapter the compound is presented with this spelling and
discussed as if it represents the total of all forms unless stated otherwise.) Our appreciation of the functions
of this amino acid has changed as investigations have progressed from viewing homocysteine as the
potential toxic agent in a very rare inherited metabolic disease (homocystinuria) to a risk factor in several
common human conditions, such as cardiovascular disease, cerebrovascular disease, and Alzheimers
disease.
Methionine is the essential amino acid precursor for the sulfur amino acid pathway (> Figure 4-1) and
is primarily metabolized via homocysteine either to taurine or is remethylated to methionine (a good
example of the fact that a nutritionally essential amino acid may be synthesized in vivo). Methionine is
supplied in the diet and enters the brain from the peripheral circulation via the large neutral amino acid
transporter (Sershen and Lajtha, 1979). Methionine itself has not been shown to have neurotransmitter
activity in the CNS; however, behavioral effects have been noted after its administration to schizophrenics.
Individuals with schizophrenia consistently worsened after methionine loading (Pollin et al., 1961; Brune
. Figure 4-1
Enzymes of the methionine metabolic pathway. 1 methionine adenosyltransferase, 2 numerous methylating
reactions, 3 Sadenosylhomocysteine hydrolase, 4 cystathionine bsynthase, 5 gcystathionase, 6
cysteine oxidase, 7 cysteinesulfinic acid decarboxylase, 8 hypotaurine oxidase, 9 chemical oxidation,
10 cysteic acid decarboxylase, 11 N5methyltetrahydrofolate: homocysteine methyltransferase, 12 serine
hydroxymethyltransferase, 13 N5,N10methylene tetrahydrofolate reductase. THFA, tetrahydrofolic acid; B6,
pyridoxal50 phosphate cofactor; B12, methylcobalamin cofactor
and Himwich, 1962), an effect that appeared to be associated with an increase in the synthesis of methylated
products of biogenic amines (Spaide et al., 1968). This response reflects the first step in the metabolism of
methionine to Sadenosylmethionine (SAM), catalyzed by the enzyme methionine adenosyltransferase.
SAM serves as the primary methyl donor in the brain (and elsewhere in the body) for biogenic amines,
RNA, DNA, and phosphatidylcholine metabolism and formation. SAM is also the precursor for the
polyamines (notably spermine and spermidine), metabolizing putrescine in the process catalyzed by the
enzyme Sadenosylmethionine decarboxylase. The polyamines have been proposed as regulators of DNA
and RNA metabolism, but of particular interest relative to the CNS is their potential role in the regulation of
neurogenesis (Malaterre et al., 2004). SAM has also been proposed as a nutritional supplement that may
help you feel good by the dietary supplement industry; however, this advice ignores the fact that increased
SAM may increase homocysteine resulting in greater health risks.
80 4 Sulfurcontaining amino acids in the CNS: Homocysteine
The demethylated product of reactions that utilize SAM as a methyl donor is Sadenosylhomocysteine
(SAH). SAH is catabolized to homocysteine and adenosine, catalyzed by the enzyme Sadenosylhomocys-
teine hydrolase. SAdenosylhomocysteine hydrolase appears to favor the reverse direction of synthesis of
SAH (Schatz et al., 1981), which may be one explanation for the low to absent amounts of homocysteine
usually observed in physiologic fluids and tissues.
Homocysteine is the major branch point of the methionine metabolic pathway; approximately half the
homocysteine that reaches this point is remethylated to homocysteine via the folic acid metabolic pathway
and the remainder is metabolized via the transsulfuration pathway (Mudd and Poole, 1975). The transsul-
furation pathway begins with the synthesis of cystathionine from homocysteine and serine catalyzed by the
enzyme cystathionine bsynthase. Cystathionine is an unusual sulfur ether compound that is found
primarily in the brain (Tallan et al., 1958). Although this compound has been suggested to be a neuro-
transmitter (Werman et al., 1966), proof of this function has not been forthcoming. Cystathionine
primarily serves as an intermediary in the transfer of a sulfur moiety from homocysteine to serine. This
conversion is completed by the catabolism of cystathionine to cysteine, catalyzed by the enzyme
gcystathionase. There is little measurable activity of gcystathionase in the brain (Shimizu et al., 1966;
Brown and Gordon, 1971), which probably accounts for the accumulation of cystathionine in this organ.
Cysteine, an amino acid required for protein synthesis, is a precursor for the major intracellular
antioxidant glutathione (gglutamylcysteinylglycine) and is further metabolized to cysteinesulfinic acid
catalyzed by cysteine oxidase. Cysteinesulfinic acid is decarboxylated to hypotaurine in a reaction catalyzed
by cysteinesulfinic acid decarboxylase. Cysteinesulfinic acid may also be oxidized to cysteic acid, which may
be decarboxylated to taurine, and hypotaurine also may be oxidized to taurine. Taurine may function as an
inhibitory neurotransmitter; it is a structural analogue of gaminobutyric acid (GABA) but is more likely to
be an important osmoregulator in the brain (PasantesMorales and Schousboe, 1988). Cysteinesulfinic acid
and cysteic acid are structural analogues of glutamic acid and have similar excitatory properties in the CNS;
however, it is not clear whether or not they are excitatory neurotransmitters in their own right. Taurine
appears to be the end product of the methionine metabolic pathway and is either excreted from the body via
the urine or as one of the bile salt conjugates, for example, taurocholic acid.
The enzymatic activities of gcystathionase and cysteinesulfinic acid decarboxylase are absent or low in
the brain during early development, thus both cysteine and taurine must be supplied from exogenous
sources (Sturman et al., 1970; Pascal et al., 1972; Gaull et al., 1977; Rassin, 1982).
The remethylation of homocysteine to methionine is catalyzed via a remethylation reaction in the CNS
by N5methyl tetrahydrofolate: homocysteine methyltransferase, an enzyme dependent upon a form of the
cofactor of vitamin B12, methylcobalamin. N5Methyltetrahydrofolate (the methyl donor for homocysteine)
is regenerated from the product of the remethylation reaction, tetrahydrofolic acid, in a sequence of two
reactions. The first, catalyzed by serine hydroxymethyltransferase, adds a methylene group from the amino
acid serine to tetrahydrofolic acid, subsequently producing the inhibitory neurotransmitter glycine and N5
N10methylenetetrahydrofolic acid. N5 N10Methylenetetrahydrofolic acid is converted to N5methyltetra-
hydrofolic acid in a reaction catalyzed by N5,N10methylenetetrahydrofolate reductase (MTHFR). Homo-
cysteine may also be remethylated to methionine by a reaction catalyzed by betaine/homocysteine
methyltransferase, utilizing betaine as the methyl donor. However, this reaction only appears to take
place in the liver and the kidney (Sunden et al., 1997).
Thus, methionine metabolic pathway is responsible for the synthesis of several putative neurotransmit-
ters (cystathionine, cysteinesulfinic acid, cysteic acid, hypotaurine, taurine, and glycine), the major intracel-
lular antioxidant (glutathione), the primary CNS methyl donor (SAM), the polyamines, and an important
osmoregulator (taurine), in addition to providing the protein precursors, methionine and cysteine. The
pivotal role of homocysteine in this pathway appears to be important for the health of the CNS.
As noted above, original interest in homocysteine began when the enzymatic defect in cystathionine
bsynthase was identified and shown to be associated with a number of clinical signs and symptoms,
Sulfurcontaining amino acids in the CNS: Homocysteine 4 81
including mental retardation, and biochemical signs, including increased plasma methionine and homo-
cysteine and decreased cysteine (Mudd et al., 2001). Subsequent to this discovery, it was found that
increased plasma homocysteine occurred associated with defects in the enzymes N5methyltetrahydrofolate:
homocysteine methyl transferase, N5,N10methylenetetrohydrofolate reductase, and in cobalamin transport
and metabolism (Rosenblatt and Fenton, 2001). These diseases all are associated with CNS deficits, with the
common biochemical feature being increased homocysteine in the blood and/or urine.
It is interesting to note that two additional metabolic defects in the methionine metabolic pathway have
been identified, one associated with methionine adenosyltransferase and the other with gcystathionase
(Mudd et al., 2001). Neither of these latter defects are associated with increased homocysteine, though
increased plasma methionine is a hallmark of methionine adenosyltransferase deficiency. Increased
cystathionine in the plasma and urine is the primary biochemical sign observed in cystathioninuria.
These two latter defects are not generally associated with functional CNS deficits, though demyelination
has been observed in some patients with methionine adenosyltransferase deficiency (Surtees et al., 1991).
Thus, the primary association with poor cognitive function in these inherited metabolic diseases is the
presence of increased homocysteine concentrations in the blood and/or urine. The possibility that increased
methionine, decreased cysteine, or a general imbalance of the various potential neurotransmitters in this
pathway are responsible for the CNS deficits appear to be unlikely explanations.
In searching for mechanisms to explain the neurologic deficits associated with homocysteine, the first
finding that seemed to be plausible was related to the thromboembolic episodes seen in a large proportion
of patients with cystathionine bsynthaserelated homocystinuria. McCully (1969) suggested that homo-
cysteine was responsible for initiating these events and developed a rabbit model that supported this
hypothesis (McCully and Ragsdale, 1970). Although not all experimental results supported the finding by
McCully and Ragsdale (1970) of increased atherosclerotic plaques in homocysteinetreated rabbits
(Donahue et al., 1974), these investigations stimulated further examination of the role of homocysteine
in cardiovascular disease.
Lead by the study of Wilcken and Wilcken (1976), a number of investigations identified mildly
increased plasma homocysteine as a risk factor for cardiovascular disease (Wilcken et al., 1983; Reis et al.,
1995; Graham et al., 1997). The finding by Mudd et al. (1981) that obligate heterozygotes for cystathionine
bsynthase deficiency were not at increased risk for thromboembolic episodes and the increased interest in
the role of folic acid in prevention of neural tube defects (MRC Vitamin Study Group, 1991; Czeizel et al.,
1992) suggested that the remethylation pathway of homocysteine might be of particular importance in
understanding the pathophysiologic effects of this compound.
The potential role of the remethylation pathway in the observed increased homocysteine concentrations
was further supported by the finding that a common thermolabile variant of MTHFR existed (MTHFR 677
C!T) that was associated with an increased risk for cardiovascular disease (Kang et al., 1991). However, the
association of the variant form of the enzyme itself with vascular disease has been inconsistent (Deloughery
et al., 1996). The variant is associated with increased homocysteine and reduced folate (Deloughery et al.,
1996). In another study that compared patients with acute ischemic stroke, high risk for atherosclerosis
related stroke, and elderly controls, no differences were observed amongst the groups related to either
MTHFR variants or to increased homocysteine (Press et al., 1999). However, this elderly population did
appear to have generally increased plasma homocysteine concentrations.
The combination of findings regarding neural tube defects and the benefits of folic acid, and the possible
role of homocysteine remethylation failure in cardiovascular disease, led to investigations of folate supple-
mentation on homocysteine. The grain supply of the United States was fortified with folic acid starting in
1998 (Mills and England, 2001) and this fortification appears to have reduced the incidence of neural
tube defects by 19% (Honein et al., 2001). Cereal fortification appears to result in a reduction in homocysteine
concentrations (Malinow et al., 1998; Jacques et al., 1999). However, this food fortification has raised the issue
that additional folic acid supplementation may not be effective in reducing homocysteine concentrations and
82 4 Sulfurcontaining amino acids in the CNS: Homocysteine
their further role as a risk factor (Bostom et al., 2001, 2002). In addition, lifestyles (smoking and overweight)
and aging are associated with increased homocysteine (Nurk et al., 2004). Increased homocysteine with age
has been associated with increased risk of hip fractures (McLean et al., 2004). Patients with renal disease may
not clear homocysteine well, so have increased homocysteine (Tamura et al., 1996). Given the results of these
studies, it is not surprising that increased homocysteine has been associated with an increased risk for
cerebrovascular disease, estimated to be between 2.0 to 2.5fold increases in risk (Boushey et al., 1995).
In considering the above studies, it might be concluded that only vascularrelated diseases of the CNS
would be associated with increased homocysteine. However, several studies have indicated that homocys-
teine is a risk factor for cognitive function in the normal aging process (Riggs et al., 1996), in geriatric
patients with various dementias (Nilsson et al., 1996), and in patients with depression (Bell et al., 1992; Fava
et al., 1997).
Intense interest arose in the role of homocysteine in CNS dysfunction with publication of the study by
Clarke and coworkers (1998) in which increased homocysteine was clearly associated with Alzheimers
disease. This study has been considered particularly important because a subset of the patients had
pathologically proven Alzheimers disease, and these patients had the highest plasma concentrations of
homocysteine (DiazArrastia, 1998; Miller, 1999). The study by Clarke and coworkers (1998) also indicated
that the progression of Alzheimers disease was accompanied by an increase in homocysteine concentra-
tions. Patients with Alzheimers disease and also those with vascular dementia have increased plasma
homocysteine and folate deficiency (Quadri et al., 2004). These findings suggested that early increased
homocysteine may signal cognitive decline in the elderly.
Data to date implicate homocysteine as a risk factor in both vascular and nonvascular diseases during
the aging process (Parnetti et al., 1997), and suggests that homocysteine is associated with a cognitive
decline in normal aging populations (Ravaglia et al., 2003; Lewerin et al., 2005). The unanswered questions
related to these studies are whether or not supplementing the aged population with B vitamins and folic
acid (cofactors at a number of steps in the methionine metabolic pathway) (see > Figure 4-1) will delay the
cognitive decline (even if they reduce homocysteine) and what are the exact biological mechanisms that
may explain the effects of homocysteine on the CNS (Vollset and Ueland, 2005). Further, few studies have
actually measured homocysteine in the CNS, none was detected in the brain of homocystinuric patients
(Brenton et al., 1965) or in the brain of a fetus carried by a mother homozygous for homocystinuria (Rassin
et al., 1979). However, methodologic issues, especially protein binding as discussed below, may have
influenced these results.
While most of the data described above reflect homocysteine as a risk factor to the brain during aging or
in disease processes usually found in the geriatric population, there is some evidence for similar risk
associations during early development. Homocysteine plasma concentrations increase with age in child-
hood and this increase appears to be associated with a decrease in folate (van Beynum et al., 2005). The
increased homocysteine was not strongly associated with the MTHFR 677 C!T polymorphism in this
study. Increased homocysteine has been associated with poor pregnancy outcomes (Vollset et al., 2000;
Hague, 2003), neural tube defects (Mills et al., 1995; Knott et al., 2003), and fetal death (Burke et al., 1992).
A number of investigators have associated increased plasma homocysteine in children with ischemic stroke
(Cardo et al., 1999; van Beynum et al., 1999) and venous thrombosis (Koch et al., 1999). Given these risk
factors, it is of particular interest that mild hypoxia in neonatal rats may result in increased homocysteine
when the animals are vitamin deficient (Blaise et al., 2005). This increased homocysteine may be a signal in
the cascade that translates oxidative stress in neonates to cognitive dysfunction.
A number of potential mechanisms have been proposed via which homocysteine may exert its toxic effects
upon the CNS. These mechanisms include the direct toxic effect of homocysteine, interaction with
the NmethylDaspartate (NMDA)glutamate receptor, stimulation of the production of oxidative
products, the production of neurotoxic products, inhibition of functional enzymes, and the induction
of apoptosis or programmed cell death.
Sulfurcontaining amino acids in the CNS: Homocysteine 4 83
initiated by homocysteine are: DNA damage, polyADPribose polymerase (PARP) activation, caspase
activation, and activation of the tumor suppressor protein p53, followed by a decline in mitochondrial
potential and finally nuclear disintegration (Kruman et al., 2000). These investigators also concluded that
homocysteine may sensitize hippocampal neurons to the neurotoxic effects of excitatory amino acids such
as glutamate (Kruman et al., 2000). Thus, there are a number of different possible pathways by which
homocysteine may exert its toxic effects, with oxidation and interaction with the NMDA receptor being
particularly interesting potential mechanisms.
6 Methodological Issues
The measurement of homocysteine has been an issue because of its ability to bind to both proteins and
other sulfurcontaining moieties via disulfide bonds. The proteinbound pool of homocysteine was first
described by Kang and coworkers (1979) and further characterized by Malloy and coworkers (1981a, b).
Homocysteine (in a manner similar to that of both cysteine and glutathione) exists as the sulfhydryl, the
homogeneous disulfidehomocystine, the mixed disulfide (with cysteine or glutathione), or as the protein
homocysteine disulfide (Rassin, 1996) (> Table 4-1). Indeed, the complexity of representing the multiple
forms of homocysteine has been addressed in a paper specific to the problem (Mudd et al., 2000). Most of
. Table 4-1
The forms of homocysteine
Form Name Abbreviation/Brief description
Reduced (thiol, sulfhydryl) Homocysteine HcySH
Oxidized (disulfide) Homocystine HcySSHcy
Mixed disulfide Mixed disulfide RSSHcy
(R cysteine, glutathione, protein)
Protein bound (via disulfide bond) Protein bound ProteinSSHcy
Sum of all but protein bound Free homocysteine (Free) f Hcy
Sum of all forms Total homocysteine (Total) t Hcy
The various forms of homocysteine found in biology. Adapted from Rassin (1996) and Mudd et al. (2000)
the published literature addressing the role of homocysteine as a risk factor has presented the data in terms
of total homocysteine (usually measured by treatment of samples with reducing agents and then determin-
ing the total amount of the sulfhydryl form produced). However, it is not unreasonable to suggest that
in vivo the free homocysteine is most available for transfer into the CNS and that pool, in equilibrium with
that bound to proteins and other sulfhydryls, may determine the degree of damage that results from
increased homocysteine. Also, homocysteine in blood and urine usually is present in the disulfide form,
homocystine, which may alter its transport into tissues. Thus, homocysteine availability may be modified
by its chemical form and by changes in the amounts of plasma proteins (particularly albumin), cysteine
(with which homocysteine competes for protein binding), and other sulfhydrylcontaining compounds
such as glutathione (Malloy et al., 1981a, b).
A second methodological issue relates to the amount of homocysteine that circulates, and the changes
that have been associated with its role as an increased risk factor. In five typical patients with cystathionine
bsynthaserelated homocystinuria, their total plasma homocysteines (the sum of free homocysteine,
proteinbound homocysteine, and mixed disulfide homocysteine) ranged from 101186 mmole/L (Malloy
et al., 1981a). In most of the epidemiologic studies of homocysteine as a risk factor, plasma homocysteine is
in the range of 920 mmole/L, and often a statistically significant result will be obtained when a 10%
reduction or increase is observedin the order of 1 mmol/L. For example a reduction from 11.4 3.4
(mean SD) to 9.7 2.3 mmol/L, an 11.0% change was very statistically significant (p < 0.001) in a study
Sulfurcontaining amino acids in the CNS: Homocysteine 4 85
of the effect of folic acid supplementation in cereal on blood homocysteine (Malinow et al., 1998). The
amounts of homocysteine observed in patients with Alzheimers disease were 13.2 4.0 mmole/L in
controls, 15.3 8.4 mmole/L in clinically diagnosed and 16.3 7.4 mmole/L in histologically confirmed
patients (Clarke et al., 1998). Some investigators have addressed the difficulty of assessing these quantita-
tively small changes by analyzing their data in terms of quartiles (McLean et al., 2004) or setting a possible
abnormal limit, for example, total plasma homocysteine >13 mmole/L (Jacques et al., 1999). From a
population perspective, monitoring plasma total homocysteine may be very useful but determining an
individuals risk factor in the light of these data may be considerably more troublesome; for example, how
would one interpret a blood total homocysteine of 10.0 or 11.0 or 12.0 mmole/Ldo any of these
concentrations represent a specific risk?
7 Conclusion
The sulfur amino acid metabolic pathway plays an important role in supporting a variety of functions
within the CNS. Homocysteine sits at an important branch point in this pathway and disturbances of its
equilibrium appear to be related to detrimental outcomes for the CNS. A number of questions remain to be
answered regarding the role of this amino acid in the CNS.
1. What is the amount of plasma homocysteine that represents safety for the CNS?
2. How should homocysteine be best determined and expressed?
3. Does increased plasma homocysteine reflect increased brain homocysteine?
4. What is the most appropriate milieu relative to the availability of vitamins B6, B12, and folic acid?
5. What is the role of homocysteic acid in the toxicity of homocysteine?
6. Does the MTHFR677C ! T polymorphism determine the homocysteine risk factor?
7. Does homocysteine play a common role in the pathway leading to the degenerative processes observed
in diseases such as Alzheimers, amyotrophic lateral sclerosis, atherosclerosis, and in the natural aging
process?
References
AlChalabi A, Leigh PN. 2000. Recent advances in amyo- Boushey CJ, Beresford SAA, Omenn GS, Motulsky AG. 1995.
trophic lateral sclerosis. Curr Opin Neurol 13: 397-405. A quantitative assessment of plasma homocysteine as a risk
Bell IR, Edman JS, Selhub J, Morrow FD, Marby DW, et al. factor for vascular disease: Probable benefits of increasing
1992. Plasma homocysteine in vascular disease and in folic acid intakes. JAMA 274: 1049-1057.
nonvascular dementia of depressed elderly people. Acta Brenton DP, Cusworth DC, Gaull GE. 1965. Homocystinuria:
Psychiatr Scand 86: 386-390. Biochemical studies of tissues including a comparison with
Blaise S, Alberto JM, Nedelec E, Ayav A, Pourie G, et al. 2005. cystathioninuria. Pediatrics 35: 50-56.
Mild neonatal hypoxia exacerbates the effects of vitamin Brown FC, Gordon PH. 1971. A study of L[14C] cystathionine
deficient diet on homocysteine metabolism in rats. Pediatr metabolism in the brain, kidney, and liver of pyridoxine
Res 57: 777-782. deficient rats. Biochim Biophys Acta 230: 434-445.
Bostom AG, Jacques PF, Liaugaudas G, Rogers G, Rosenberg Brune GG, Himwich HE. 1962. Effects of methionine loading
IH, et al. 2002. Total homocysteine lowering treatment on the behavior of schizophrenic patients. J Nerv Ment Dis
among coronary artery disease patients in the era of folic 134: 447-450.
acidfortified cereal grain flour. Arterioscler Thromb Vasc Burke G, Robinson K, Refsum H, Stuart B, Drumm J, et al.
Biol 22: 488-491. 1992. Intrauterine growth retardation, perinatal death, and
Bostom AG, Selhub J, Jacques PF, Rosenberg IH. 2001. Power maternal homocysteine levels. N Engl J Med 326: 69-70.
shortage: Clinical trials testing the homocysteine hypo- Cardo E, Vilaseca MA, Campistol J, Artuch R, Colome C, et al.
thesis against a background of folic acidfortified cereal 1999. Evaluation of hyperhomocysteinaemia in children
grain flour. Ann Intern Med 135: 133-137. with stroke. Eur J Paediatr 3: 113-117.
86 4 Sulfurcontaining amino acids in the CNS: Homocysteine
Carson NA, Neill DW. 1962. Metabolic abnormalities detected Gerritsen T, Waisman HA. 1964. Homocystinuria: Absence of
in a survey of mentally backward individuals in northern cystathionine in the brain. Science 145: 588.
Ireland. Arch Dis Child 37: 505-513. Gerritsen T, Vaughn JG, Waisman HA. 1962. The identifica-
Chung YH, Hong JJ, Shin CM, Joo KM, Kim MJ, et al. 2003. tion of homocysteine in the urine. Biochem Biophys Res
Immunohistochemical study on the distribution of homo- Commun 9: 493-496.
cysteine in the central nervous system of transgenic mice Graham IM, Daly LE, Refsum HM, Robinson K, Brattstrom
expressing a human Cu/Zn SOD mutation. Brain Res 967: LE, et al. 1997. Plasma homocysteine as a risk factor for
226-234. vascular disease. The European Concerted Action Project.
Clarke R, Smith AD, Jobst KA, Refsum H, Sutton L, et al. JAMA 277: 1775-1781.
1998. Folate, vitamin B12, and serum total homocysteine Hague WM. 2003. Homocysteine and pregnancy. Best Pract
levels in confirmed Alzheimer disease. Arch Neurol 55: Res Clin Obstet Gynaecol 17: 459-469.
1449-1455. Honein MA, Paulozzi LJ, Mathews TJ, Erickson JD, Wong LY.
Curtis DR, Watkins JC. 1960. The excitation and depression 2001. Impact of folic acid fortification of the US food
of spinal neurones by structurally related amino acids. supply on the occurrence of neural tube defects. JAMA
J Neurochem 6: 117-141. 285: 2981-2986.
Czeizel AE, Dudas I. 1992. Prevention of the first occurrence Jacques PF, Selhub J, Bostom AG, Wilson PWF, Rosenberg IH.
of neuraltube defects by periconceptional vitamin supple- 1999. The effect of folic acid fortification on plasma folate
mentation. N Engl J Med 327: 1832-1835. and total homocysteine concentrations. N Engl J Med 340:
Deloughery TG, Evans A, Sadeghi A, McWilliams J, Henner 1449-1454.
WD, et al. 1996. Common mutation in methylenetetra- Kang SS, Wong PW, Becker N. 1979. Proteinbound homocyst
hydrofolate reductase: Correlation with homocysteine (e)ine in normal subjects and in patients with homocysti-
metabolism and lateonset vascular disease. Circulation nuria. Pediatr Res 13: 1141-1143.
94: 3074-3078. Kang SS, Wong PW, Susmano A, Sora J, Norusis M, et al.
DiazArrastia R. 1998. Hyperhomocysteinemia: A new 1991. Thermolabile methylenetetrahydrofolate reductase:
risk factor for Alzheimer disease? Arch Neurol 55: 1407- An inherited risk factor for coronary artery disease. Am
1408. J Hum Genet 48: 536-545.
Donahue S, Sturman JA, Gaull G. 1974. Arteriosclerosis due Kilpatrick IC, Mozley LS. 1986. An initial analysis of the
to homocysteinemia. Am J Pathol 77: 167-174. regional distribution of excitatory sulphurcontaining
Fava M, Borus JS, Alpert JE, Nierenberg AA, Rosenbaum JF, amino acids in the rat brain. Neurosci Lett 72: 189-193.
et al. 1997. Folate, vitamin B12, and homocysteine in major Kim WK, Pae YS. 1996. Involvement of NmethylDaspartate
depressive disorder. Am J Psychiatry 154: 426-428. receptor and free radical in homocysteinemediated toxi-
Freed WJ, Taylor SP, Luchins DJ, Wyatt RJ, Gillin JC. 1980. city on rat cerebellar granule cells in culture. Neurosci Lett
Production of convulsions in mice by the combination 216: 117-120.
of methionine and homocysteine. Psychopharmacol 69: Knott L, Hartridge T, Brown NL, Mansell JP, Sandy JR. 2003.
275-280. Homocysteine oxidation and apoptosis: A potential cause
FritzerSzekeres M, Blom HJ, Boers GH, Szekeres T, Lubec B. of cleft palate. In Vitro Cell Dev Biol Anim 39: 98-105.
1998. Growth promotion by homocysteine but not by Koch HG, Nabel P, Junker R, Auberger K, Schobess R, et al.
homocysteic acid: A role for excessive growth in homo- 1999. The 677T genotype of the common MTHFR thermo-
cystinuria or proliferation in hyperhomocysteinemia? labile variant and fasting homocysteine in childhood
Biochim Biophys Acta 1407: 1-6. venous thrombosis. Eur J Pediatr 158: S113-S116.
Gaull GE. 1973. Abnormal metabolism of sulfurcontaining Kruman II, Culmsee C, Chan SL, Kruman Y, Guo Z, et al.
amino acids associated with brain dysfunction. Handbook 2000. Homocysteine elicits a DNA damage response in
of Neurochemistry, Vol. VII. Lajtha A, editor. New York: neurons that promotes apoptosis and hypersensitivity to
Plenum Press; pp. 169-190. excitotoxicity. J Neurosci 20: 6920-6926.
Gaull GE, Rassin DK, Raiha NCR, Heinonen K. 1977. Milk Lee SJ, Kim KM, Namkoong S, Kim CK, Kang YC, et al. 2005.
protein quantity and quality in lowbirthweight infants III. Nitric oxide inhibition of homocysteineinduced human
Effects of sulfur amino acids in plasma and urine. J Pediatr endothelial cell apoptosis by downregulation of p53
90: 348-355. dependent noxa expression through the formation of
Gaull GE, Tallan HH, Lajtha A, Rassin DK. 1975. Pathogenesis Snitrosohomocysteine. J Biol Chem 280: 5781-5788.
of brain dysfunction in inborn errors of amino acid metabo- Lewerin C, Matousek M, Steen G, Johansson B, Steen B, et al.
lism. Biology of Brain Dysfunction, Vol. 3. Gaull GE, editor. 2005. Significant correlations of plasma homocysteine and
New York: Plenum Publishing Company; pp. 47-143. serum methylmalonic acid with movement and cognitive
Sulfurcontaining amino acids in the CNS: Homocysteine 4 87
performance in elderly subjects but no improvement from Nilsson K, Gustafson L, Faldt R, Andersson A, Brattstrom L,
shortterm vitamin therapy: A placebocontrolled rando- et al. 1996. Hyperhomocysteinaemia a common finding in
mized study. Am J Clin Nutr 81: 1155-1162. a psychogeriatric population. Eur J Clin Invest 26: 853-859.
Lipton SA, Kim WK, Choi YB, Kumar S, DEmilia DM, et al. Nurk E, Tell GS, Vollset SE, Nygard Refsum H, Nilsen RM,
1997. Neurotoxicity associated with dual actions of homo- et al. 2004. Changes in lifestyle and plasma total homocys-
cysteine at the NmethylDaspartate receptor. Proc Natl teine: The Hordaland homocysteine study. Am J Clin Nutr
Acad Sci USA 94: 5923-5928. 79: 812-819.
Malaterre J, Strambi C, Aouane A, Strambi A, Rougon G, et al. Parnetti L, Bottiglieri T, Lowenthal D. 1997. Role of homo-
2004. A novel role for polyamines in adult neurogenesis in cysteine in agerelated vascular and nonvascular diseases.
rodent brain. Eur J Neurosci 20: 317-330. Aging Clin Exp Res 9: 241-257.
Malinow MR, Duell PB, Hess DL, Anderson PH, Kruger WD, PasantesMorales H, Schousboe A. 1988. Volume regulation
et al. 1998. Reduction of plasma homocyst(e)ine levels by in astrocytes: A role for taurine as an osmoeffector. J Neu-
breakfast cereal fortified with folic acid in patients with rosci Res 20: 503-509.
coronary heart disease. N Engl J Med 338: 1009-1015. Pascal TA, Gilliam BM, Gaull GE. 1972. Cystathionase:
Malloy MH, Rassin DK, Gaull GE. 1981a, Plasma cyst(e)ine in Immunochemical evidence for absence from human fetal
homocyst(e)inemia. Am J Clin Nutr 34: 2619-2621. liver. Pediatr Res 6: 773-778.
Malloy MH, Rassin DK, Gaull GE. 1981b, A method for Pollin W, Cardon PV Jr, Kety SS. 1961. Effects of amino acid
measurement of free and bound plasma cyst(e)ine. Anal feedings in schizophrenic patients treated with iproniazid.
Biochem 113: 407-415. Science 133: 104-105.
McCully KS. 1969. Vascular pathology of homocysteinemia: Press RD, Beamer N, Evans A, De Loughery TG, Coull BM.
Implications for the pathogenesis of arteriosclerosis. Am J 1999. Role of a common mutation in the homocysteine
Pathol 56: 111-128. regulatory enzyme methylenetetrahydrofolate reductase in
McCully KS, Ragsdale BD. 1970. Production of arteriosclero- ischemic stroke. Diagn Mol Pathol 8: 54-58.
sis by homocysteinemia. Am J Pathol 61: 1-11. Quadri P, Fragiacomo C, Pezzati R, Zanda E, Forloni G, et al.
McLean RR, Jacques PF, Selhub J, Tucker KL, Samelson EJ, 2004. Homocysteine, folate, and vitamin B12 in mild cog-
et al. 2004. Homocysteine as a predictive factor for hip nitive impairment, Alzheimer disease, and vascular demen-
fracture in older persons. N Engl J Med 350: 2042-2049. tia. Am J Clin Nutr 80: 114-122.
Medical Research Council (MRC) Vitamin Study Group. Rassin DK. 1982. Taurine, cysteinesulfinic acid decarboxylase
1991. Prevention of neural tube defects: Results of and glutamic acid in brain. Taurine in Nutrition and Neu-
the medical research council vitamin study. Lancet 338: rology. Huxtable RJ, PasantesMorales H, editors. New York:
131137. Plenum Publishing Corporation, NY; pp. 257268.
Miller JW. 1999. Homocysteine and Alzheimers disease. Nutr Rassin DK. 1996. Source, metabolism and function of cysteine
Rev 57: 126-129. and glutathione in the central nervous system. Paradigms
Mills JL, England L. 2001. Food fortification to prevent neural of Neural Injury Methods in Neuroscience, Vol. 30. Perez
tube defects: Is it working? JAMA 285: 3022-3023. Polo JR, editor. USA: Academic Press, San Diego, CA;
Mills JL, McPartlin JM, Kirke PN, Lee YJ, Conley MR, et al. pp. 167177.
1995. Homocysteine metabolism in pregnancies compli- Rassin DK, Fleisher LD, Muir A, Desnick RJ, Gaull GE, 1979.
cated by neuraltube defects. Lancet 345: 149-151. Fetal tissue amino acid concentrations in argininosuccinic
Mudd SH, Poole JR. 1975. Labile methyl balances for normal aciduria and in maternal homocystinuria. Clin Chim
humans on various dietary regimens. Metabolism 24: Acta 94: 101-108.
721-735. Ravaglia G, Forti P, Maioli F, Muscari A, Sacchetti L, et al.
Mudd SH, Finkelstein JD, Refsum H, Ueland PM, Malinow 2003. Homocysteine and cognitive function in healthy
MR, et al. 2000. Homocysteine and its disulfide derivatives: elderly community dwellers in Italy. Am J Clin Nutr 77:
A suggested consensus terminology. Arterioscler Thromb 668-673.
Vasc Biol 20: 1704-1706. Reis RP, Azinheira J, Reis HP, Bordalo A, Santos L, et al. 1995.
Mudd SH, Havlik R, Levy HL, McKusick VA, Feinleib M. Homocysteinaemia after methionine overload as a coro-
1981. A study of cardiovascular risk in heterozygotes for nary artery disease risk factor: Importance of age and
homocystinuria. Am J Hum Genet 33: 883-893. homocysteine levels. Coron Artery Dis 6: 851-856.
Mudd SH, Levy HL, Kraus JP. 2001. Disorders of transsulfura- Riggs KM, Spiro A, Tucker K, Rush D. 1996. Relations of
tion. The Metabolic and Molecular Bases of Inherited vitamin B12, vitamin B6, folate, and homocysteine to cog-
Disease. Scriver CR, Beaudet AL, Sly WS, Valle D, Childs nitive performance in the normative aging study. Am J Clin
B, et al. editors. New York: McGrawHill; pp. 2007-2056. Nutr 63: 306-314.
88 4 Sulfurcontaining amino acids in the CNS: Homocysteine
Rosenblatt DS, Fenton WA. 2001. Inherited disorders of folate Vol. 3, 2nd ed. Lajtha A, editor. New York: Plenum Pub-
and cobalamin transport and metabolism. The Metabolic lishing Corporation, NY; pp. 535-558.
and Molecular Bases of Inherited Disease. Scriver CR, Tamura T, Johnston KE, Bergman SM. 1996. Homocysteine
Beaudet AL, Sly WS, Valle D, Childs B, et al. editors. New and folate concentrations in blood from patients treated
York: McGrawHill; pp. 3897-3933. with hemodialysis. J Am Soc Nephrol 7: 2414-2418.
Schatz RA, Wilens TE, Sellinger OZ. 1981. Decreased trans- van Beynum IM, den Heijer M, Thomas CMG, Afman L,
methylation of biogenic amines after in vivo elevation of Oppenraayvan Emmerzaal D, et al. 2005. Total homocys-
SadenosylLhomocysteine. J Neurochem 36: 1739-1748. teine and its predictors in Dutch children. Am J Clin Nutr
Sershen H, Lajtha A. 1979. Inhibition pattern by analogs 81: 1110-1116.
indicates the presence of ten or more transport systems van Beynum IM, Smeitink JA, den Heijer M, te Poele
for amino acids in brain cells. J Neurochem 32: 719-726. Pothoff MT, Blom HJ. 1999. Hyperhomocysteinemia:
Shimizu H, Kakimoto Y, Sano I. 1966. A method of determi- A risk factor for ischemic stroke in children. Circulation
nation of cystathionine and its distribution in human 99: 2070-2072.
brain. J Neurochem 13: 65-73. Vollset SE, Ueland PM. 2005. B vitamins and cognitive func-
Spaide JK, Tanimukai H, Bueno JR, Himwich HE. 1968. tion: Do we need more and larger trials? Am J Clin Nutr 81:
Behavioral and biochemical alterations in schizophrenic 951-952.
patients. Arch Gen Psychiatry 18: 658-665. Vollset SE, Refsum H, Irgens LM, Emblem BM, Tverdal A,
Streck EL, Vieira PS, Wannmacher CMD, DutraFilho CS, et al. 2000. Plasma total homocysteine, pregnancy compli-
Wajner M, et al. 2003. In vitro effect of homocysteine on cations, and adverse pregnancy outcomes: The Hordaland
some parameters of oxidative stress in rat hippocampus. homocysteine study. Am J Clin Nutr 71: 962-968.
Metab Brain Dis 18: 147-154. Waller SJ, Kilpatrick IC, Chan MW, Evans RH. 1991. The
Streck EL, Zugno AI, Tagliari B, Wannmacher CMD, Wajner influence of assay conditions on measurement of excitatory
M, et al. 2002a, Inhibition of Na, KATPase activity by dibasic sulphinic and sulphonic aamino acids in nervous
the metabolites accumulating in homocystinuria. Metab tissue. J Neurosci Methods 36: 167-176.
Brain Dis 17: 83-91. Werman R, Davidoff RA, Aprison MH. 1966. The inhibitory
Streck EL, Matte C, Vieira PS, Rombaldi F, Wannmacher action of cystathionine. Life Sci 5: 1431-1440.
CMD, et al. 2002b. Reduction of Na, KATPase White AR, Huang X, Jobling MF, Barrow CJ, Beyreuther K,
activity in hippocampus of rats subjected to chemically in- et al. 2001. Homocysteine potentiates copper and amyloid
duced hyperhomocysteinemia. Neurochem Res 27: 1593- b peptidemediated toxicity in primary neuronal cultures:
1598. Possible risk factors in the Alzheimerstype neurodegener-
Sturman JA, Gaull G, Raiha NCR. 1970. Absence of cystathio- ative pathways. J Neurochem 76: 1509-1520.
nase in human fetal liver: Is cystine essential? Science 169: Wilcken DE, Wilcken B. 1976. The pathogenesis of coronary
74-76. artery disease. A possible role for methionine metabolism.
Sunden SL, Renduchintala MS, Park EI, Miklasz SD, Garrow J Clin Invest 57: 1079-1082.
TA. 1997. Betainehomocysteine methyltransferase expres- Wilcken DE, Reddy SG, Gupta VJ. 1983. Homocysteinemia,
sion in porcine and human tissues and chromosomal ischemic heart disease, and the carrier state for homocysti-
localization of the human gene. Arch Biochem Biophys nuria. Metab Clin Exp 32: 363-370.
345: 171-174. Wuerthele SE, Yasuda RP, Freed WJ, Hoffer BJ. 1982. The
Surtees R, Leonard J, Austin S. 1991. Association of effect of local application of homocysteine on neuronal
demyelination with deficiency of cerebrospinal fluid S activity in the central nervous system of the rat. Life Sci
adenosylmethionine in inborn errors of methyltransfer 31: 2683-2691.
pathway. Lancet 338: 1550-1554. Wyse ATS, Zugno AI, Streck EL, Matte C, Calcagnotto T, et al.
Tallan HH, Moore S, Stein WH. 1958. Lcystathionine in 2002. Inhibition of Na, KATPase activity in hippocam-
human brain. J Biol Chem 230: 707-716. pus of rats subjected to acute administration of homocys-
Tallan HH, Rassin DK, Sturman JA, Gaull GE. 1983. Methio- teine is prevented by vitamins E and C treatment.
nine metabolism in the brain. Handbook of Neurochemistry, Neurochem Res 27: 1685-1689.
5 Stress Response Signal
Transduction
Xiaoming Hu . J. R. PerezPolo
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
7 Decoy Treatments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96
Abstract: Trauma evokes a common response in all organ systems that includes early energy depletion
followed by dysfunction of ionic gradients and triggering of stress response mechanisms with outcomes
ranging from recovery of function to delayed cell death and further dysfunction. The specific outcome to
trauma in terms of this spectrum is determined to some extent by the specific signaling mechanisms
triggered. In the central nervous system (CNS), stress response signaling includes changes in transcription
of cytokines and chemokines and also inflammatory enzymes. A common regulatory component in all of
these is the nuclear factor kappa B (NF-B) transcription factor. Interventions interfering with NF-B
activation can affect the outcome to CNS trauma.
List of Abbreviations: CNS, central nervous system; COX2, cyclooxygenase2; iNOS, inducible nitric
oxide synthase; ICE, IL1converting enzyme; IL1, interleukin1; IL1R, IL1 receptor; IL1Ra, IL1-
Rantagonist; IL1RAacP, IL1R accessory protein; LPS, lipopolysaccharides; MCAO, middle cerebral arterial
occlusion; MnSOD, manganese superoxide dismutase; RPA, ribonuclease protection assay; RTPCR, real
time polymerase chain reaction; TNF, tumor necrosis factor
1 Introduction
Classical stress response studies dealt with hormonal changes in the face of abrupt environmental stimuli
and their effects on physiological processes: circulation, temperature regulation, motor activity, and so on.
This was followed by inclusion of immune responses as part of the physiological stress responses, typically
involving assessments of cytokine expression changes in the central nervous system (CNS) and associated
growth factors. At about this time, quantitative molecular approaches began to take the place of cellular
assays based on migration and inflammatory events at the organismic level. In addition, the role of cytokine
expression changes in the CNS in response to more traumatic and acute events, typically related to abrupt
physical trauma and ischemia, became more common even as various experimental and animal models of
head trauma, stroke, perinatal ischemia, and spinal cord injury were developed and characterized. Here, we
focus on some of the more common molecular components of stress in the CNS in the acute and chronic
situation. By necessity, this is not a definitive treatment but rather an attempt to integrate some mechan-
isms of responses to trauma in CNS into a cohesive, albeit complex, conceptual framework.
Cellular stress responses are best characterized by an early stage of energy depletion inactivating membrane
bound pumps responsible for the maintenance of transmembrane gradients of an assortment of ions and
signaling molecules. This departure from cellular homeostasis triggers an inflammatory response that, if
sufficient energy stores are available, triggers active secondary cell death with apoptotic features. Cellular
apoptosis processes include chromatin condensation and subsequent margination against the nuclear
envelope, cytoplasmic and nuclear condensation, overall cell shrinkage, cytoplasmic vacuolization, and
convolution of the nuclear and the cytoplasmic membranes followed by the formation of apoptotic bodies.
These apoptotic bodies then undergo phagocytosis. Under conditions in which the initial disruption of
cellular processes is sufficiently robust, the cell death process displays necrotic features, characterized by
overall cell swelling, chromatin clumping, and organelle disruption, with early loss of membrane integrity
(Hockenbery, 1995; Majno and Joris, 1995).
For example, necrotic cell death is typically an early event at the core of the traumatic insult to the CNS
in the case of hypoxia ischemia (HI) (Ray et al., 2003). Delayed cell death with apoptotic features usually
starts between 12 and 24 h later at the core site and spreads over time to sites away from the impacted area
and includes apoptosis of neurons and glia (Northington et al., 2001). In adult animal models, both
apoptosis and necrosis have been reported following focal cerebral ischemia in rats and mice (Charriaut
Marlangue et al., 1996; Dux et al., 1996; Chen et al., 1997; Matsushita et al., 1998; Guegan and Sola, 2000)
and after global cerebral ischemia (Nitatori et al., 1995; Ni et al., 1998), often emphasizing delayed neuronal
death in the selectively vulnerable regions of the hippocampus.
Stress response signal transduction 5 91
Inflammatory responses have also been implicated in chronic glial activation in the aged rodent brain, in
rodent models of adult spinal cord injury and stroke, head injury, as well as hypoxia ischemia (Yang et al.,
1995; Hagberg et al., 1996; Bona et al., 1999; Nesic et al., 2001; Rossner et al., 2001; Chu et al., 2002; Hedtjarn
et al., 2002; Nesic et al., 2002; ToliverKinsky et al., 2002; Cole and PerezPolo, 2004; Hu et al., 2005; Rossner
et al., 2005). Typically, CNS trauma increases glial expression of the proinflammatory cytokines: interleukin
1 (IL1)b, tumor necrosis factora/b (TNFa/b), and IL6 (FosterBarber et al., 2001; Lau and Yu, 2001; Nesic
et al., 2001, 2002; Saliba and Henrot, 2001; Hu et al., 2005).
The bestcharacterized early response inflammatory cytokine is IL1 (Szaflarski et al., 1995). The IL1
family includes the agonists IL1a and IL1b, the endogenous receptor antagonist IL1 receptor antagonist
(IL1Ra), and IL18/IL1g (Shapiro et al., 1998). The role of IL1b in CNS trauma is probably the best
understood (Rothwell et al., 1997; Nesic et al., 2001; Hu et al., 2005). There are reports showing that IL1b
mRNA and protein levels increase as early as 1 h after brain injury and persist during the development
of infarction in various adult animal models (Minami et al., 1992; Zhang et al., 1998a, b; Davies et al.,
1999). Sources of IL1 in CNS are endothelial cells, microglia, and macrophages or monocytes (Buttini
et al., 1994) although neurons, astrocytes, and oligodendrocytes can also secrete IL1b under traumatic
conditions (Sairanen et al., 1997). Levels of IL1a in CNS have been shown to also increase promptly and
robustly after temporary or permanent occlusion of the middle cerebral artery in mice (Hill et al., 1999;
Touzani et al., 1999).
Following synthesis, proIL1b (31 KD) remains primarily in the cytoplasm until cleaved by the IL1b
converting enzyme (ICE, caspase1), producing mature and active IL1b (17 KD). Caspase1 also cleaves
and activates IL18 but not IL1a (Dinarello, 1999). There is evidence showing correlation between
increased IL1b levels and subsequent neurodegeneration. Thus, administration of exogenous IL1b
markedly exacerbates neuronal or glial damage in rodents exposed to focal cerebral ischemia or excitotoxin
administration (Lawrence et al., 1998; Stroemer and Rothwell, 1998). Decreasing effective IL1b levels using
antiIL1b antibodies (Touzani et al., 1999) or inhibitors of ICE (Loddick and Rothwell, 1996; Hara et al.,
1997) decreases neuronal losses after cerebral ischemia. IL1Ra is widely used to study the function of IL1
because it is a selective endogenous receptor antagonist that blocks the actions of IL1a and IL1b
(Dinarello et al., 1998). Injection or overexpression of IL1Ra significantly inhibits cell death after hypoxia
ischemia although the mechanisms involved were not determined (Betz et al., 1995; Stroemer and Rothwell,
1997, 1998; Yang et al., 1998).
The responses to IL1a and IL1b are initiated by binding to IL1 typeI receptor (IL1RI), which then
associates with an IL1 receptor accessory protein (IL1RAcP) (Wesche et al., 1997), leading to signal
transduction (Cao et al., 1996; Muzio et al., 1997). IL1RIs are expressed on both glia and neurons (French
et al., 1999; Pinteaux et al., 2002). Ligand binding of IL1, in situ hybridization, and immunohistochemical
studies have shown the expression of IL1RI receptors in specific brain regions in rats, particularly, the
hippocampus, the dentate gyrus, the choroid plexus, and the cerebellum. A second receptor IL1 typeII
receptor (IL1RII) has a much shorter cytoplasmic domain and is thought to act as a decoy (nonsignaling)
receptor (McMahan et al., 1991). The activation of IL1RI stimulates ubiquitous transcription factors
(NFkB, AP1, AFP) and some key intracellular signaling molecules (JNK, IP3, or PKC kinases; Auron,
1998; ONeill and Greene, 1998). Based on in vitro studies, it has been suggested that activation of the
NFkB pathway through IL1RI binding after ischemia results in gene transcription with deleterious
outcomes (Dunn et al., 2002).
For hypoxia ischemia, it has been shown that IL1 mRNA and protein levels increase as early as 1 h after
the injury and persist during the development of infarction (Minami et al., 1992; Szaflarski et al., 1995;
Hagberg et al., 1996; Zhang et al., 1998a, b; Davies et al., 1999; Qiu et al., 2004). Regional differences in IL
1b distribution patterns have been observed in normal rodent brain (Ilyin et al., 1998; Vitkovic et al., 2000);
92 5 Stress response signal transduction
for example, there are lower levels of IL1b in hippocampus compared with cortex in shamtreated animals
(Hu et al., 2005). This variation might reflect endogenous neuromodulatory activities in an untreated rat.
It is reported that the hippocampus is more sensitive to hypoxia ischemia injury than other brain
structures (CervosNavarro and Diemer, 1991; Grafe et al., 1994; Yue et al., 1997). The time course of IL1b
protein levels correlates well with the time course of cell death level and caspase3 activity observed after
injury. Thus, it would be reasonable to speculate that the cell deathinducing threshold of IL1 is lower in
hippocampus than that in the cortex, which explains the lower level of IL1b after hypoxia ischemia in
hippocampus. This is consistent with there being temporal and spatial differences in IL1b protein
expression after HI, not surprising given similar reports of spatial temporal differences between hippocam-
pus and basal forebrain cell death mechanisms in an adult rodent hypoxia model (Qiu et al., 2001). Thus, it is
likely that there are different mechanisms of stress response signal transduction at the level of IL1 signaling
in different brain regions. This could also explain some of the particular aspects of longterm outcomes in the
clinical context and the reported fragility of hippocampus, for example, over other brain regions.
Using the techniques of in situ hybridization, ribonuclease protection assay (RPA), and realtime
polymerase chain reaction (RTPCR), it has been reported that IL1RI mRNA was expressed in high levels
in cortex (Gayle et al., 1997; Wang et al., 1997) and in low to moderate levels in hippocampus (Wong and
Licinio, 1994). At the cellular level, IL1RI mRNA has been localized in cerebrovascular endothelium,
astrocytic and oligodendroglial (Wong and Licinio, 1994; Hu et al., 2005), as well as neuronal components
(Takao et al., 1990; Wong and Licinio, 1994; Hu et al., 2005).
Significantly protective effects of IL1Ra have been documented in an adult stroke rat model, where cell
death after stroke was significantly reduced when IL1Ra was administered as a single injection at the time
of middle cerebral artery occlusion and hypoxia ischemia (Mulcahy et al., 2003; Hu et al., 2005). The maximal
cortical protection occurred after intracerebroventricular injection of IL1Ra 2 h post middle cerebral
arterial occlusion (MCAO) (Mulcahy et al., 2003). Protective effects of IL1Ra can last for at least 7 days
after injury (Garcia et al., 1995; Liu et al., 1999; Loddick and Rothwell, 1996). Injections of L1Ra 2 h after
hypoxia ischemia injury markedly decreased hippocampal cell death levels and caspase3 activity when
measured 24 h after injury, in support of the hypothesis that IL1 contributes to hypoxia ischemiainduced
cell death.
The transcription factor NFkB consists of homo or heterodimers of subunits, which constitute a family of
related proteins, including p50, p52 (also called p49), p65 (also called RelA), CRel, and RelB (> Figure 51).
All of them share a highly conserved 300 residue NH2terminal domain for DNA binding or dimeriza-
tion (Rel homology domain), which enables them to form dimers and bind to an array of homologous
decanucleotide sequences with varying affinities. NFkB/Rel proteins can be divided into two classes based
on their Cterminal domain. One class includes p65, CRel, and RelB proteins. All of them contain C
terminal transactivation domain. Another class includes p50 and p52, which have no transactivation
domain at the C terminus (Bours et al., 1990, 1992; Ghosh and Baltimore, 1990; Nolan et al., 1991;
Ruben et al., 1991; Schmid et al., 1991; Ryseck et al., 1992). p50 and p52 are first synthesized as precursors
p105 and p100, respectively, both of which have ankyrinlike repeats with close homology to that of IkB
family in their C terminus. After degradation of the Cterminal region, mature p50 and p52 are released.
Dimerization of NFkB subunits produces species with various intrinsic DNAbinding specificities, trans-
activation properties, and subcellular localization (Baeuerle, 1991; Siebenlist et al., 1994). NFkB dimeric
combinations must contain at least one of the transactivating members (p65, CRel, or RelB) to have the
ability to activate DNA transcription (Baeuerle, 1991; Lenardo and Siebenlist, 1994; Siebenlist et al., 1994;
Chen et al., 1998). The most commonly described active NFkB subunit combinations are the p65/p50 and
CRel/p50 heterodimer. Homodimeric combinations of p50 or p52 have no transactivation activity and
often behave in vivo as transcriptional inhibitors at high affinitybinding sites (Franzoso et al., 1992, 1993;
Kang et al., 1992).
Stress response signal transduction 5 93
. Figure 51
Schematic diagram illustrating possible effects of traumainduced IL1 increases on NFkB activation. IL1
stimulates degradation of IkBa and increased nuclear levels of Bcl3. IkBa degradation enables nuclear
translocation of p65/p50. Nuclear Bcl3 binds and dissociates p50 homodimers from cognate NFkBbinding
sites enabling p65/p50 dimers to bind and activate target genes
A mechanism underlying IL1induced cell death involves the activation of the transcription factor
NFkB and the subsequent induction of multiple genes. NFkB is constitutively expressed in CNS at low
levels (Kaltschmidt et al., 1994), but is stimulated above basal levels by stimuli associated with ischemic
insults. In the rodent brain, NFkB activation is stimulated in the forebrain and the hippocampus after
hypoxia ischemia (Koong et al., 1994; Schmidt et al., 1995; Yang et al., 1995; Gabriel et al., 1999; Nurmi
et al., 2004). In most of these studies, the expression and activity of p65/p50 was measured and used as a
representative for NFkB activity, which overlooks changes in other NFkB proteins. For example, NFkB
CRel/p50 selectively binds the Bclx gene promoter CS4 sequence while p65/p50 preferentially binds the
IgGkB promoter sequence (Hu et al., 2005).
NFkB regulates the expression of antiapoptotic genes such as BclxL, Bcl2, Manganese superoxide
dismutase (MnSOD), and inhibitorofapoptosis proteins, and proinflammatory genes such as IL1, TNFa,
matrix metalloproteinase9, cyclooxygenase2 (COX2), and inducible nitric oxide synthase (iNOS), thereby
playing a dual role in neuronal survival (Mattson and Camandola, 2001). Given that the p65/p50 dimer
activates genes coding for proteins with proinflammatory properties, while CRel/p50 dimer activates genes
coding for BclxL protein that prevent cell death (Qiu et al., 2001; Pizzi et al., 2002), we believe that some of
the present confusions in the field as to the proapoptotic or antiapoptotic nature of NFkB activities in in
vivo lesion paradigms result from a lack of data addressing activation of specific promoters by different
94 5 Stress response signal transduction
NFkB dimer species. The activation of different NFkB dimers under different stimuli and at different time
points may be responsible for the differential beneficial or detrimental outcomes in cell survival.
NFkB is expressed in the nervous system and shows a low level of constitutive activity in neurons
(Kaltschmidt et al., 1994). NFkB is induced above constitutive levels in the brain by various stimuli, most
of which are stressful or neurotoxic. For example, rodent NFkB is induced in the forebrain and the
hippocampus following hypoxia ischemia (Koong et al., 1994; Schmidt et al., 1995; Yang et al., 1995; Qiu
et al., 2001), in the cortex after traumatic brain injury (Yang et al., 1995), and in the limbic structures
following seizures (Rong and Baudry, 1996). NFkB is reported to regulate cell death by modulating a
diverse array of genes, known to be important for cell death or survival, such as Bcl2, BclxL, MnSOD,
iNOS, COX2, and some inflammatory cytokines (IL1a/b, TNFa/b) (FosterBarber et al., 2001; Lau and
Yu, 2001; Saliba and Henrot, 2001).
The extent to which NFkB activation contributes to neuropathology versus neuroprotection and
recovery remains unresolved. In a number of experimental models, including cell lines and tissues under
different stimuli, NFkB activation appears to result in both apoptotic and antiapoptotic outcomes
(Abbadie et al., 1993; Barger et al., 1995; Jung et al., 1995; Barger and Mattson, 1996; Grilli et al., 1996;
Grimm et al., 1996; Baichwal and Baeuerle, 1997; Lipton, 1997; Grilli and Memo, 1999). It is possible that
the different compositions of NFkB protein dimers determine the outcomes of their activation. For
example, the p65/p50 dimer may activate genes coding for proteins with proapoptotic properties whereas
CRel/p50 dimer may activate genes coding for proteins that prevent cell death (Qiu et al., 2001; Pizzi et al.,
2002; Hu et al., 2005). Therefore, different stimuli might activate different NFkB dimers, resulting in
beneficial or detrimental outcomes.
There is a generic NFkB DNA sequence consisting of a 10bp consensus sequence that encompasses 128
different decameric sequences, 50 GGGRNNTYCC30 (Gguanine, Rpurine, Nany nucleotide, Y
pyrimidine, Ccytosine, Tthymine). Distinct NFkB subunit combinations exhibit different binding
affinities for the multiple consensus sequence combinations, and this complex variation likely determines
the specificity of NFkB transcriptional regulation (Perkins et al., 1992; Qiu et al., 2001). For example,
p65/p50 heterodimers bind preferentially to the decameric sequence present on the IgGkB promoter
(50 GGGACTTTCC30 ) and the CRel/p50 heterodimer binds preferentially to the decameric sequence 50
GGGGTCTCC30 present on the Bclx promoter (Qiu et al., 2001). Specific combinations of NFkB proteins
can thus distinguish among the various kB sites to selectively regulate gene expression.
Nadjar et al. (2003) documented the activation of NFkB by the visualization of p65 translocation in the
cells of adult rat brain after intraperitoneal or i.c.v. IL1b injection. IL1RIdeficient mice demonstrated
impaired NFkB activation upon the stimulation of IL1b. Huang et al. (2003) documented that decreased
NFkB activity was observed in ischemic ICE null adult mice and suggested that IL1 signaling contributes
to NFkB activation and subsequent ischemic damage.
One of the most important features of NFkB proteins is that the function of NFkB is strictly regulated
by its subcellular localization. Its activation depends on the translocation from the cytoplasm to the
nucleus. The movement of NFkB proteins into the nucleus and binding to DNA is controlled by a family
of inhibitor proteins called IkB (Zabel and Baeuerle, 1990; Haskill et al., 1991; Thompson et al., 1995).
There are several inhibitory IkB proteins (IkBa, IkBb, IkBg, and Bcl3, > Figure 51) that regulate NFkB
activity. These proteins have several homologous amino acid stretches known as ankyrin repeats that
specifically interact with NFkB/Rel proteins (Bours et al., 1992; Grilli et al., 1993). IkBa and Bcl3 are the
two most important proteins in this family. They contribute to the regulation of NFkB activity via different
mechanisms.
IkBa takes effect in the cytoplasm, where they bind to NFkB heterodimers and inhibit their action. In
response to stimuli, including IL1, TNFa, and bacterial lipopolysaccharides (LPS), IkB is phosphorylated
by an IkB kinase, ubiquinated and degraded by proteasomes (Ghosh and Baltimore, 1990; Liu et al., 1993;
Israel, 1995), uncovering masked nuclear localization signals on the NFkB dimers, which are then
translocated to the nucleus, bind to DNA consensus sequences on gene promoters, and activate transcrip-
tion (Ghosh et al., 1998). In particular, it has been shown that IkBa must be phosphorylated on Ser32
and Ser36 and ubiquitinated at Lys21 and Lys22 so that it can be targeted for subsequent degradation by
the ubiquitin and proteasome system (Traenckner et al., 1995; Baldwin, 1996). Interestingly, only the NFkB
Stress response signal transduction 5 95
dimers that contain at least one transactivation member (p65, CRel, or RelB) are effectively regulated by
IkBa. Thus, the p52 and p50 homodimers are not usually retained in the cytoplasm by IkBa. Their activity is
regulated within the nucleus by Bcl3.
Although structurally homologous, Bcl3 and IkBs act differently in regulating NFkB activity. Whereas
the IkBs are primarily cytoplasmic, Bcl3 is predominantly nuclear bound and has a binding specificity for
p52 or p50 homodimer (Bours et al., 1993; Franzoso et al., 1993; Fujita et al., 1993; Nolan et al., 1993;
Bundy and McKeithan, 1997). There are two mechanisms by which Bcl3 can regulate the NFkB activity.
First, Bcl3 can successfully compete with promoter DNAbound p50 or p52 homodimers, removing the
NFkB homodimers away from cognatebinding sites on target promoter DNA sites (Franzoso et al., 1993;
Nolan et al., 1993; Siebenlist et al., 1994; Zhang et al., 1998a, b), allowing p65/p50 or CRel/p65 to bind and
activate gene transcription. Second, Bcl3, together with p52 or p50 homodimer, can bind to DNA kB site
and form a ternary complex.
The tethering of Bcl3 to DNA via p52 or p50 homodimer allows the activation of gene transcription,
whereas Bcl3 and p50/p52 homodimers alone cannot since Bcl3 has only the transactivation domain and
p50/p52 has only the DNAbinding domain (Bours et al., 1993; Fujita et al., 1993; Rocha et al., 2003). The
different regulatory effects of Bcl3 may depend on its phosphorylation states and concentration in the
nucleus (Fujita et al., 1993; Nolan et al., 1993; Bundy and McKeithan, 1997; Cogswell et al., 2000).
NFkB transactivation involves IkBa and Bcl3 proteins. IkBa phosphorylation at Ser32 and Ser36
(Traenckner et al., 1995) is a requirement for its ubiquitination and subsequent degradation by the 26 S
proteasome, an important step in the translocation of NFkB from the cytoplasm to the nucleus. An in vitro
study of IL1bchallenged C6 (astrocytoma) cells showed newly synthesized IkBa to be phosphorylated at
Ser32 (Uehara et al., 1999), which may explain the increase of phosphorylated IkBa 24 h after hypoxia
ischemia (Hu et al., 2005). Increased nuclear Bcl3 levels also stimulate NFkB activation. Western blot
analyses have shown an increase in translocation of Bcl3 from cytoplasm to nucleus 24 h after hypoxia
ischemia in hippocampus and cortex (Hu et al., 2005). A report based on DNA microarray analyses and RT
PCR by Elliott et al. (2002) using cultured SW1353 cells also shows Bcl3 transcription to be stimulated by
IL1b signaling. It is reported that NFkB is stimulated by hypoxia via increased IkBa degradation and NFk
B nuclear translocation. A small number of researches indicated that an independent pathway involving
Bcl3 stimulation also might contribute to the NFkB activation after hypoxia (Gozal et al., 1998; Zhang
et al., 1998a, b; Qiu et al., 2001).
There is evidence that nitric oxide (NO) is an important mediator of ischemic and neurodegenerative
pathology in the CNS (Hara et al., 1996; Mizushima et al., 2002; Brown and BalPrice, 2003). NO is
endogenously produced as a byproduct of arginine metabolism by different nitric oxide synthase (NOS)
isoforms: neuronal NOS (nNOS), endothelial NOS (eNOS), and inducible NOS (iNOS). Both nNOS and
eNOS are constitutively expressed, whereas iNOS is expressed in response to a variety of stimuli. There are
reports documenting the relationship between iNOS induction and ischemic lesions in the brain (Hara
et al., 1996; Mizushima et al., 2002). In addition, the administration of IL1b to cultured brain endothelial cells
and hippocampal neurons induces the expression of iNOS mRNA and increases NO release (Bonmann et al.,
1997; Serou et al., 1999). These observations, together with evidence showing that induction of iNOS is
regulated by NFkB p65/p50 (Teng et al., 2002), also support the hypothesis that activation of p65/p50 by IL1
after CNS trauma may further stimulate iNOS synthesis and NO formation, which in turn triggers cell death.
COX is an important enzyme in the inflammatory process. COX catalyzes the ratelimiting step in the
conversion of arachidonic acid to prostaglandins. There are two isoforms of COX, designated COX1 and
COX2. COX1 is expressed constitutively and appears to be responsible for ongoing physiological function,
whereas COX2 is present only in certain tissues where it is transiently induced by growth factors,
inflammatory cytokines, tumor promoters, and bacterial toxins (Chun et al., 2004). There is evidence to
support an involvement of COX2 in brain damage (Hara et al., 1998; Nagayama et al., 1999; Sugimoto and
Iadecola, 2003). IL1b transcriptional control of COX2 by NFkB p65/p50 has been demonstrated in cell
96 5 Stress response signal transduction
culture (Newton et al., 1997). There is a significant increase in iNOS and COX2 expression in the
hippocampus and the cortex 24 h after hypoxia ischemia (Hu et al., 2003). Treatment with IL1Ra reversed
the HIinduced increases in iNOS and COX2, a novel finding. Interestingly, sequence analyses of the iNOS
and COX2 promoters have shown the presence of DNAbinding consensus sequences specific to the p65/
p50 binding (Hu et al., 2005).
7 Decoy Treatments
Sequencespecific inhibition of NFkB can be accomplished with synthetic doublestranded (ds) phos-
phothiorate oligonucleotides containing a NFkB consensus sequence, which acts in vivo as a decoy cis
element to bind transcription factors and block the activation of cognate genes (Morishita et al., 1997; Qiu
et al., 2004; Tomita et al., 1998, 2000, 2001). The principle of the transcription factor decoy approach is
simply the inhibition of promoter activity due to the binding of endogenous transcription factors to
saturating amounts of exogenous doublestranded oligonucleotides displaying specific DNA sequences that
are present in the promoter region of the gene of interest. For example, an IgGkB decoy, which contains the
NFkB consensus sequence in the IgGkB promoter, has been used to inhibit p65/p50 activity (Blondeau
et al., 2001). However, there is no information about decoys targeted to other NFkB dimers.
The pharmacokinetic profiles of phosphothiorated oligonucleotides of varying base composition in rat
are similar (Agrawal and Zhao, 1998). The stereotaxic injection of fluorescent IgGkB decoys into brain
results in the cellular incorporation of decoys in hippocampi as early as 2 h after injection, with dissipation
taking place by 7 h. The ability to selectively intervene in the binding of different NFkBbinding consensus
sequences to cognate NFkB protein dimers during a fairly short time period should allow for the
determination of the role of NFkB in the regulation of specific genes at specific times after a traumatic
insult to the CNS. It has been proposed that transcriptional injury responses are biphasic and that there are
different consequences to delayed transcriptional activation or to persistent versus transient transcriptional
factor activation of the NFkB transcription factor. The ability to abolish NFkB binding to selective
promoter sites in a transient fashion should provide a useful tool in establishing the role of different
traumainduced pathways in cellular commitment to cell death and survival.
Since IgGkB decoy can specifically inhibit p65/p50binding activity, it is not surprising that it can
inhibit the transcription of inflammatory cytokines, which are the target genes for NFkB p65/p50 (Hu
et al., 2005). These results are consistent with recent reports showing antiapoptotic features to the
stimulation of CRel and proapoptotic features to the stimulation of p65 DNA binding (Qiu et al., 2001;
Pizzi et al., 2002). These results are also consistent with the CRelmediated regulation of Bclx gene
transcription in response to hypoxia ischemia (Qiu et al., 2001, 2004). Genes that were significantly
downregulated by IgGkB decoys were not affected by Bclx decoys, and vice versa.
8 Conclusion
Following injury to the CNS, there is a cascade of molecular and cellular responses that determine outcome
in terms of cell survival and function. One important pathway activated by CNS trauma is via induction of
IL1, which binds to cognate receptors and activates selective gene expression of genes such as COX2 and
iNOS via the transcription factor NFkB. The aim here was to describe the activation mechanisms by which
IL1 contributes to cell death.
We also discuss the effects of selective interventions in (i) IL1 receptor binding or (ii) injuryinduced
p65/p50 NFkB activation. The absence of side effects of IL1Ra and its safety in clinical trials in rheumatoid
arthritis would suggest that this intervention may be applicable to perinatal ischemia. Decoys to NFkB
binding DNA consensus sequences are useful to determine regulatory features affecting individual genes.
Use of decoy treatments to alter physiological and pathological outcomes to trauma requires concerted
pulsed treatments with cocktail of different decoy sequences that take into account the large number of
genes displaying NFkBbinding sites in their promoters.
Stress response signal transduction 5 97
References
Abbadie C, Kabrun N, Bouali F, Smardova J, Stehelin D, p105NFkappa B participates in transactivation through a
et al. 1993. High levels of cRel expression are asso- kappa B site. Mol Cell Biol 12(2): 685-695.
ciated with programmed cell death in the developing Bours V, Franzoso G, Azarenko V, Park S, Kanno T, et al. 1993.
avian embryo and in bone marrow cells in vitro. Cell The oncoprotein Bcl3 directly transactivates through
75(5): 899-912. kappa B motifs via association with DNAbinding p50B
Agrawal S, Zhao Q. 1998. Antisense therapeutics. Curr Opin homodimers. Cell 72(5): 729-739.
Chem Biol 2(4): 519-528 (Review). Brown GC, BalPrice A. 2003. Inflammatory neurodegenera-
Auron PE. 1998. The interleukin 1 receptor: Ligand interac- tion mediated by nitric oxide, glutamate, and mitochon-
tions and signal transduction. Cytokine Growth Factor Rev dria. Mol Neurobiol 27(3): 325-355.
9(34): 221-237. Bundy DL, McKeithan TW. 1997. Diverse effects of BCL3
Baeuerle P. 1991. The inducible transcription activator NF phosphorylation on its modulation of NFkappaB p52 homo-
kappa B; regulation by distinct protein subunits. Biochem dimer binding to DNA. J Biol Chem 272(52): 33132-33139.
Biophys Acta 1072(1): 63-80. Buttini M, Sauter A, Boddeke HW. 1994. Induction of inter-
Baichwal VR, Baeuerle PA. 1997. Activate NFkappa B or die? leukin1 beta mRNA after focal cerebral ischaemia in the
Curr Biol 7(2): R94-R96. rat. Brain Res Mol Brain Res 23(12): 126-134.
Baldwin AS Jr. 1996. The NFkappa B and I kappa B pro- Cao Z, Henzel WJ, Gao X. 1996. IRAK: A kinase asso-
teins: New discoveries and insights. Annu Rev Immunol ciated with the interleukin1 receptor. Science 271(5252):
14: 649-683. 1128-1131.
Barger SW, Mattson MP. 1996. Induction of neuroprotective CervosNavarro J, Diemer NH. 1991. Selective vulnerability in
kappa Bdependent transcription by secreted forms of the brain hypoxia. Crit Rev Neurobiol 6(3): 149-182.
Alzheimers betaamyloid precursor. Brain Res Mol Brain CharriautMarlangue C, Margaill I, Represa A, Popovici T,
Res 40(1): 116-126. Plotkine M, et al. 1996. Apoptosis and necrosis after revers-
Barger SW, Horster D, Furukawa K, Goodman Y, Krieglstein J, ible focal ischemia: An in situ DNA fragmentation analysis.
et al. 1995. Tumor necrosis factors alpha and beta protect J Cereb Blood Flow Metab 16(2): 186-194.
neurons against amyloid betapeptide toxicity: Evidence for Chen FE, Huang DB, Chen YQ, Ghosh G. 1998. Crystal
involvement of a kappa Bbinding factor and attenuation of structure of p50/p65 heterodimer of transcription factor
peroxide and Ca2 accumulation. Proc Natl Acad Sci USA NFkappaB bound to DNA. Nature 391(6665): 410-413.
92(20): 9328-9332. Chen J, Jin K, Chen M, Pei W, Kawaguchi K, et al. 1997. Early
Betz AL, Yang GY, Davidson BL. 1995. Attenuation of stroke detection of DNA strand breaks in the brain after transient
size in rats using an adenoviral vector to induce overexpres- focal ischemia: Implications for the role of DNA damage in
sion of interleukin1 receptor antagonist in brain. J Cereb apoptosis and neuronal cell death. J Neurochem 69(1):
Blood Flow Metab 15(4): 547-551. 232-245.
Blondeau N, Widmann C, Lazdunski M, Heurteaux C. 2001. Chu D, Qiu J, Grafe M, Fabian R, Kent TA, et al. 2002. Delayed
Activation of the nuclear factorkappaB is s key event in cell death signaling in traumatized central nervous system:
brain tolerance. J Neurosci 21: 4668-4677. Hypoxia. Neurochem Res 27: 97-106.
Bona E, Andersson AL, Blomgren K, Gilland E, PukaSundvall Chun KS, Cha HH, Shin JW, Na HK, Park KK, et al. 2004.
M, et al. 1999. Chemokine and inflammatory cell response Nitric oxide induces expression of cyclooxygenase2 in
to hypoxiaischemia in immature rats. Pediatr Res 45(4 Pt 1): mouse skin through activation of NFkappaB. Carcinogen-
500-509. esis 25(3): 445-454.
Bonmann E, Suschek C, Spranger M, KolbBachofen V. 1997. Cogswell PC, Guttridge DC, Funkhouser WK, Baldwin AS Jr.
The dominant role of exogenous or endogenous interleukin1 2000. Selective activation of NFkappa B subunits in
beta on expression and activity of inducible nitric oxide human breast cancer: Potential roles for NFkappa B2/p52
synthase in rat microvascular brain endothelial cells. Neu- and for Bcl3. Oncogene 19(9): 1123-1131.
rosci Lett 230(2): 109-112. Cole K, PerezPolo JR. 2004. Neuronal trauma model: In
Bours V, Villalobos J, Burd PR, Kelly K, Siebenlist U. 1990. search of Thanatos. Int J Dev Neurosci 22: 485-496.
Cloning of a mitogeninducible gene encoding a kappa B Davies CA, Loddick SA, Toulmond S, Stroemer RP, Hunt J,
DNAbinding protein with homology to the rel oncogene et al. 1999. The progression and topographic distribution
and to cellcycle motifs. Nature 348(6296): 76-80. of interleukin1beta expression after permanent middle
Bours V, Burd PR, Brown K, Villalobos J, Park S, et al. 1992. cerebral artery occlusion in the rat. J Cereb Blood Flow
A novel mitogeninducible gene product related to p50/ Metab 19(1): 87-98.
98 5 Stress response signal transduction
Dinarello CA. 1999. Interleukin18. Methods 19(1): 121-132 Ghosh S, May MJ, Kopp EB. 1998. NFkappa B and Rel
(Review). proteins: Evolutionarily conserved mediators of immune
Dinarello CA, Novick D, Puren AJ, Fantuzzi G, Shapiro L, responses. Annu Rev Immunol 16: 225-260.
et al. 1998. Overview of interleukin18: More than an Gozal E, Simakajornboon N, Gozal D. 1998. NFkB induc-
interferongamma inducing factor. J Leukoc Biol 63(6): tion during in vivo hypoxia in dorsocaudal brain stem of
658-664. rateffect of MK801 and LNME. J Appl Physiol 85(1):
Dunn SL, Young EA, Hall MD, McNulty S. 2002. Activation of 312-316.
astrocyte intracellular signaling pathways by interleukin1 Grafe MR. 1994. Developmental changes in the sensitivity of
in rat primary striatal cultures. Glia 37: 31-42. the neonatal rat brain to hypoxic/ischemic injury. Brain
Dux E, Oschlies U, Uto A, Kusumoto M, Hossmann KA. 1996. Research 653(12): 161-166.
Early ultrastructural changes after brief histotoxic hypoxia Grilli M, Memo M. 1999. Nuclear factorkappaB/Rel proteins:
in cultured cortical and hippocampal CA1 neurons. Acta A point of convergence of signalling pathways relevant in
Neuropathol (Berl) 92(6): 541-544. neuronal function and dysfunction. Biochem Pharmacol
Elliott SF, Coon CI, Hays E, Stadheim TA, Vincenti MP. 2002. 57(1): 1-7.
Bcl3 is an interleukin1responsive gene in chondrocytes Grilli M, Chiu JJ, Lenardo MJ. 1993. NFkappa B and Rel:
and synovial fibroblasts that activates transcription of the Participants in a multiform transcriptional regulatory sys-
matrix metalloproteinase 1 gene. Arthritis Rheum 46(12): tem. Int Rev Cytol 143: 1-62.
3230-3239. Grilli M, Pizzi M, Memo M, Spano P. 1996. Neuroprotection
FosterBarber A, Dickens B, Ferriero DM. 2001. Human peri- by aspirin and sodium salicylate through blockade of NF
natal asphyxia: Correlation of neonatal cytokines with MRI kappaB activation. Science 274(5291): 1383-1385.
and outcome. Dev Neurosci 23(3): 213-218. Grimm S, Bauer MK, Baeuerle PA, SchulzeOsthoff K. 1996.
Franzoso G, Bours V, Azarenko V, Park S, TomitaYamaguchi Bcl2 downregulates the activity of transcription factor NF
M, et al. 1993. The oncoprotein Bcl3 can facilitate NF kappaB induced upon apoptosis. J Cell Biol 134(1): 13-23.
kappa Bmediated transactivation by removing inhibiting Guegan C, Sola B. 2000. Early and sequential recruitment of
p50 homodimers from select kappa B sites. EMBO J 12(10): apoptotic effectors after focal permanent ischemia in mice.
3893-3901. Brain Res 856(12): 93-100.
Franzoso G, Bours V, Park S, TomitaYamaguchi M, Kelly K, Hagberg H, Gilland E, Bona E, Hanson LA, HahinZoric M,
et al. 1992. The candidate oncoprotein Bcl3 is an antago- et al. 1996. Enhanced expression of interleukin (IL)1 and
nist of p50/NFkappa Bmediated inhibition. Nature IL6 messenger RNA and bioactive protein after hypoxia
359(6393): 339-342. ischemia in neonatal rats. Pediatr Res 40(4): 603-609.
French RA, Van Hoy RW, Chizzonite R, Zachary JF, Dantzer Hara K, Kong DL, Sharp FR, Weinstein PR. 1998. Effect of
R, et al. 1999. Expression and localization of p80 and p68 selective inhibition of cyclooxygenase 2 on temporary focal
interleukin1 receptor proteins in the brain of adult mice. cerebral ischemia in rats. Neurosci Lett 256(1): 53-56.
J Neuroimmunol 93(12): 194-202. Hara H, Huang PL, Panahian N, Fishman MC, Moskowitz
Fujita T, Nolan GP, Liou HC, Scott ML, Baltimore D. 1993. MA. 1996. Reduced brain edema and infarction volume in
The candidate protooncogene bcl3 encodes a transcrip- mice lacking the neuronal isoform of nitric oxide synthase
tional coactivator that acts through NFkappa B p50 homo- after transient MCA occlusion. J Cereb Blood Flow Metab
dimers. Genes Dev 7(7B): 1354-1363. 16(4): 605-611.
Gabriel C, Justicia C, Camins A, Planas AM. 1999. Vasculitic Hara H, Friedlander RM, Gagliardini V, Ayata C, Fink K, et al.
neuropathy in association with chronic graftversushost 1997. Inhibition of interleukin 1beta converting enzyme
disease. J Neurol Sci 168(1): 68-70. family proteases reduces ischemic and excitotoxic neuronal
Garcia JH, Liu KF, Relton JK. 1995. Interleukin1 receptor damage. Proc Natl Acad Sci USA 94(5): 2007-2012.
antagonist decreases the number of necrotic neurons in rats Haskill S, Beg AA, Tompkins SM, Morris JS, Yurochko AD,
with middle cerebral artery occlusion. Am J Pathol 147(5): et al. 1991. Characterization of an immediateearly gene
1477-1486. induced in adherent monocytes that encodes I kappa Blike
Gayle D, Ilyin SE, PlataSalaman CR. 1997. Interleukin1 activity. Cell 65(7): 1281-1289.
receptor type I mRNA levels in brain regions from male Hedtjarn M, Leverin AL, Eriksson K, Blomgren K, Mallard C,
and female rats. Brain Res Bull 42(6): 463-467. et al. 2002. Interleukin18 involvement in hypoxicischemic
Ghosh S, Baltimore D. 1990. Activation in vitro of NFkappa brain injury. J Neurosci 22(14): 5910-5919.
B by phosphorylation of its inhibitor I kappa B. Nature 344 Hill DB, Devalaraja R, JoshiBarve S, Barve S, McClain CJ. 1999.
(6267): 678-682. Antioxidants attenuate nuclear factorkappa B activation and
Stress response signal transduction 5 99
tumor necrosis factoralpha production in alcoholic hepatitis Lipton SA. 1997. Janus faces of NFkappa B: Neurodestruc-
patient monocytes and rat Kupffer cells, in vitro. Clin Bio- tion versus neuroprotection. Nat Med 3(1): 20-22.
chem 32(7): 563-570. Liu J, Sen R, Rothstein TL. 1993. Abnormal kappa Bbinding
Hockenbery D. 1995. Defining apoptosis. Am J Pathol 146(1): protein in the cytoplasm of a plasmacytoma cell line that
16-19 (Review). lacks nuclear expression of NFkappa B. Mol Immunol
Hu X, NesicTaylor O, Qiu J, Rea HC, Fabian R, et al. 2005. 30(5): 479-489.
Activation of nuclear factorkappaB signaling pathway by Liu XH, Kwon D, Schielke GP, Yang GY, Silverstein FS, et al.
interleukin1 after hypoxia/ischemia in neonatal rat hippo- 1999. Mice deficient in interleukin1 converting enzyme
campus and cortex. J Neurochem 93: 26-37. are resistant to neonatal hypoxicischemic brain damage.
Hu X, Qiu J, Grafe MR, Rea HC, Rassin DK, et al. 2003. Bcl2 J Cereb Blood Flow Metab 19(10): 1099-1108.
family members make different contributions to cell death Loddick SA, Rothwell NJ. 1996. Neuroprotective effects of
in hypoxia and/or hyperoxia in rat cerebral cortex. Int J Dev human recombinant interleukin1 receptor antagonist in
Neurosci 21(7): 371-377. focal cerebral ischaemia in the rat. J Cereb Blood Flow
Huang FP, Wang ZQ, Wu DC, Schielke GP, Sun Y, et al. 2003. Metab 16(5): 932-940.
Early NFkappaB activation is inhibited during focal cerebral Majno G, Joris I. 1995. Apoptosis, oncosis, and necrosis. An
ischemia in interleukin1betaconverting enzyme deficient overview of cell death. Am J Pathol 146(1): 3-15 (Review).
mice. J Neurosci Res 73(5): 698-707. Matsushita K, Matsuyama T, Kitagawa K, Matsumoto M,
Ilyin SE, Gayle D, Flynn MC, PlataSalaman CR. 1998. Inter- Yanagihara T, et al. 1998. Alterations of Bcl2 family pro-
leukin1beta system (ligand, receptor type I, receptor ac- teins precede cytoskeletal proteolysis in the penumbra, but
cessory protein and receptor antagonist), TNFalpha, TGF not in infarct centres following focal cerebral ischemia in
beta1 and neuropeptide Y mRNAs in specific brain regions mice. Neuroscience 83(2): 439-448.
during bacterial LPSinduced anorexia. Brain Res Bull 45: Mattson MP, Camandola S. 2001. NFkappaB in neuronal
507-515. plasticity and neurodegenerative disorders. J Clin Invest
Israel A. 1995. A role for phosphorylation and degradation in 107(3): 247-254 (Review).
the control of NFkappa B activity. Trends Genet 11(6): McMahan CJ, Slack JL, Mosley B, Cosman D, Lupton SD,
203-205. et al. 1991. A novel IL1 receptor, cloned from B cells by
Jung S, Yaron A, Alkalay I, Hatzubai A, Avraham A, et al. 1995. mammalian expression, is expressed in many cell types.
Costimulation requirement for AP1 and NFkappa B tran- EMBO J 10(10): 2821-2832.
scription factor activation in T cells. Ann N Y Acad Sci 766: Minami M, Kuraishi Y, Yabuuchi K, Yamazaki A, Satoh M.
245-252. 1992. Induction of interleukin1 beta mRNA in rat brain after
Kaltschmidt C, Kaltschmidt B, Neumann H, Wekerle H, transient forebrain ischemia. J Neurochem 58(1): 390-392.
Baeuerle PA. 1994. Constitutive NFkappa B activity in Mizushima H, Zhou CJ, Dohi K, Horai R, Asano M, et al.
neurons. Mol Cell Biol 14(6): 3981-3992. 2002. Reduced postischemic apoptosis in the hippocampus
Kang SM, Tran AC, Grilli M, Lenardo MJ. 1992. NFkappa B of mice deficient in interleukin1. J Comp Neurol 448(2):
subunit regulation in nontransformed CD4 T lympho- 203-216.
cytes. Science 256(5062): 1452-1456. Morishita R, Sugimoto T, Aoki M, Kida I, Tomita N, et al.
Koong AC, Chen EY, Giaccia AJ. 1994. Hypoxia causes the 1997. In vivo transfection of cis element decoy against
activation of nuclear factor kappa B through the phosphor- nuclear factorkappaB binding site prevents myocardial
ylation of I kappa B alpha on tyrosine residues. Cancer Res infarction. Nat Med 3(8): 894-899.
54(6): 1425-1430. Mulcahy NJ, Ross J, Rothwell NJ, Loddick SA. 2003. Delayed
Lau LT, Yu AC. 2001. Astrocytes produce and release interleu- administration of interleukin1 receptor antagonist pro-
kin1, interleukin6, tumor necrosis factor alpha and inter- tects against transient cerebral ischaemia in the rat. Br J
ferongamma following traumatic and metabolic injury. Pharmacol 140(3): 471-476.
J Neurotrauma 18(3): 351-359. Muzio M, Ni J, Feng P, Dixit VM. 1997. IRAK (Pelle) family
Lawrence CB, Allan SM, Rothwell NJ. 1998. Interleukin1beta member IRAK2 and MyD88 as proximal mediators of IL1
and the interleukin1 receptor antagonist act in the stria- signaling. Science 278(5343): 1612-1615.
tum to modify excitotoxic brain damage in the rat. Eur J Nadjar A, Combe C, Laye S, Tridon V, Dantzer R, et al. 2003.
Neurosci 10(3): 1188-1195. Nuclear factor kappaB nuclear translocation as a crucial
Lenardo M, Siebenlist U. 1994. Bcl3mediated nuclear regu- marker of brain response to interleukin1. A study in rat
lation of the NFkappa B transactivating factor. Immunol and interleukin1 type I deficient mouse. J Neurochem
Today 15(4): 145-147. 87(4): 1024-1036.
100 5 Stress response signal transduction
Nagayama M, Niwa K, Nagayama T, Ross ME, Iadecola C. actions in murine microglial cells. J Neurochem 83(4):
1999. The cyclooxygenase2 inhibitor NS398 ameliorates 754-763.
ischemic brain injury in wildtype mice but not in mice Pizzi M, Goffi F, Boroni F, Perkins SE, Liou HC, et al. 2002.
with deletion of the inducible nitric oxide synthase gene. Opposing roles for NFkappa B/Rel factors p65 and cRel
J Cereb Blood Flow Metab 19(11): 1213-1219. in the modulation of neuron survival elicited by glutamate
Nesic O, Svrakic NM, Xu GY, McAdoo D, Westlund KN, et al. and interleukin1beta. J Biol Chem 277(23): 20717-20723.
2002. DNA microarray analysis of the contused spinal Qiu J, Grafe MR, Schmura SM, Glasgow JN, Kent TA, et al.
cord: Effect of NMDA receptor inhibition. J Neurosci Res 2001. Differential NFkappa B regulation of bclx gene
68: 406-423. expression in hippocampus and basal forebrain in response
Nesic O, Xu GY, McAdoo D, High KW, Hulsebosch C, et al. to hypoxia. J Neurosci Res 64(3): 223-234.
2001. IL1 receptor antagonist prevents apoptosis and Qiu J, Hu X, Nesic O, Grafe MR, Rassin DK, et al. 2004. Effects
caspase3 activation after spinal cord injury. J Neurotrauma of NFkappaB oligonucleotide decoys on gene expression
18(9): 947-956. in P7 rat hippocampus after hypoxia/ischemia. J Neurosci
Newton R, Kuitert LM, Bergmann M, Adcock IM, Barnes PJ. Res 77(1): 108-118.
1997. Evidence for involvement of NFkappaB in the tran- Ray AD, Roberts AJ, Lee SD, Farkas GA, Michlin C, et al. 2003.
scriptional control of COX2 gene expression by IL1beta. Exercise delays the hypoxic thermal response in rats. J Appl
Biochem Biophys Res Commun 237(1): 28-32. Physiol 95(1): 272-278.
Ni B, Wu X, Su Y, Stephenson D, Smalstig EB, et al. 1998. Rocha S, Martin AM, Meek DW, Perkins ND. 2003. p53
Transient global forebrain ischemia induces a prolonged represses cyclin D1 transcription through down regulation
expression of the caspase3 mRNA in rat hippocampal of Bcl3 and inducing increased association of the p52 NF
CA1 pyramidal neurons. J Cereb Blood Flow Metab kappaB subunit with histone deacetylase 1. Mol Cell Biol
18(3): 248-256. 23(13): 4713-4727.
Nitatori T, Sato N, Waguri S, Karasawa Y, Araki H, et al. 1995. Rong Y, Baudry M. 1996. Seizure activity results in a rapid
Delayed neuronal death in the CA1 pyramidal cell layer of induction of nuclear factorkappa B in adult but not juve-
the gerbil hippocampus following transient ischemia is nile rat limbic structures. J Neurochem 67(2): 662-668.
apoptosis. J Neurosci 15(2): 1001-1011. Rossner S, Apelt J, Schliebs R, PerezPolo JR, Bigl V, 2001.
Nolan GP, Ghosh S, Liou HC, Tempst P, Baltimore D. 1991. Neuronal and glial betasecretase (BACE) protein expres-
DNA binding and I kappa B inhibition of the cloned p65 sion in transgenic Tg2576 mice with amyloid plaque pa-
subunit of NFkappa B, a relrelated polypeptide. Cell thology. J Neurosci Res 64: 437-446.
64(5): 961-969. Rossner S, LangeDohna C, Zeitschel U, PerezPolo JR, 2005.
Nolan GP, Fujita T, Bhatia K, Huppi C, Liou HC, et al. 1993. Alzheimers disease betasecretase BACE1 is not a neuron
The bcl3 protooncogene encodes a nuclear I kappa Blike specific enzyme. J Neurochem 92: 226-234.
molecule that preferentially interacts with NFkappa B p50 Rothwell N, Allan S, Toulmond S. 1997. The role of interleu-
and p52 in a phosphorylationdependent manner. Mol Cell kin 1 in acute neurodegeneration and stroke: Pathophysio-
Biol 13(6): 3557-3566. logical and therapeutic implications. J Clin Invest 100(11):
Northington FJ, Ferriero DM, Flock DL, Martin LJ. 2001. 2648-2652.
Delayed neurodegeneration in neonatal rat thalamus after Ruben SM, Dillon PJ, Schreck R, Henkel T, Chen CH, et al.
hypoxiaischemia is apoptosis. J Neurosci 21(6): 1931-1938. 1991. Isolation of a relrelated human cDNA that poten-
Nurmi A, Lindsberg PJ, Koistinaho M, Zhang W, Juettler E, tially encodes the 65kD subunit of NFkappa B. Science
et al. 2004. Nuclear factorkappaB contributes to infarction 251(5000): 1490-1493.
after permanent focal ischemia. Stroke 35(4): 987-991. Ryseck RP, Bull P, Takamiya M, Bours V, Siebenlist U, et al.
ONeill LA, Greene C. 1998. Signal transduction pathways 1992. RelB, a new Rel family transcription activator that
activated by the IL1 receptor family: Ancient signaling can interact with p50NFkappa B. Mol Cell Biol 12(2):
machinery in mammals, insects, and plants. J Leukoc Biol 674-684.
63(6): 650-657. Sairanen TR, Lindsberg PJ, Brenner M, Siren AL. 1997. Global
Perkins ND, Schmid RM, Duckett CS, Leung K, Rice NR, et al. forebrain ischemia results in differential cellular expression
1992. Distinct combinations of NFkappa B subunits deter- of interleukin1beta (IL1beta) and its receptor at mRNA
mine the specificity of transcriptional activation. Proc Natl and protein level. J Cereb Blood Flow Metab 17(10): 1107-
Acad Sci USA 89(5): 1529-1533. 1120.
Pinteaux E, Parker LC, Rothwell NJ, Luheshi GN. 2002. Expres- Saliba E, Henrot A. 2001. Inflammatory mediators and neo-
sion of interleukin1 receptors and their role in interleukin1 natal brain damage. Biol Neonate 79(34): 224-227.
Stress response signal transduction 5 101
Schmid RM, Perkins ND, Duckett CS, Andrews PC, Nabel GJ. adhesion molecule. Expression in glomerular cells in vitro
1991. Cloning of an NFkappa B subunit which stimulates and in vivo by transcription factor decoy for NFkappaB.
HIV transcription in synergy with p65. Nature 352(6337): Exp Nephrol 9(3): 181-190.
733-736. Tomita N, Morishita R, Tomita S, Yamamoto K, Aoki M, et al.
Schmidt KN, Traenckner EB, Meier B, Baeuerle PA. 1995. 1998. Transcription factor decoy for nuclear factorkappaB
Induction of oxidative stress by okadaic acid is required inhibits tumor necrosis factoralphainduced expression of
for activation of transcription factor NFkappa B. J Biol interleukin6 and intracellular adhesion molecule1 in en-
Chem 270(45): 27136-27142. dothelial cells. J Hypertens 16(7): 993-1000.
Serou MJ, De Coster MA, Bazan NG. 1999. Interleukin1 beta Touzani O, Boutin H, Chuquet J, Rothwell N. 1999. Potential
activates expression of cyclooxygenase2 and inducible nitric mechanisms of interleukin1 involvement in cerebral is-
oxide synthase in primary hippocampal neuronal culture: chaemia. J Neuroimmunol 100(12): 203-215.
Plateletactivating factor as a preferential mediator of cyclo- Traenckner EB, Pahl HL, Henkel T, Schmidt KN, Wilk S, et al.
oxygenase2 expression. J Neurosci Res 58(4): 593-598. 1995. Phosphorylation of human I kappa Balpha on ser-
Shapiro L, Puren AJ, Barton HA, Novick D, Peskind RL, et al. ines 32 and 36 controls I kappa Balpha proteolysis and
1998. Interleukin 18 stimulates HIV type 1 in monocytic NFkappa B activation in response to diverse stimuli.
cells. Proc Natl Acad Sci USA 95(21): 12550-12555. EMBO J 14(12): 2876-2883.
Siebenlist U, Franzoso G, Brown K. 1994. Structure, regula- Uehara T, Matsuno J, Kaneko M, Nishiya T, Fujimuro M, et al.
tion and function of NFkappa B. Annu Rev Cell Biol 10: 1999. Transient nuclear factor kappaB (NFkappaB) activa-
405-455. tion stimulated by interleukin1beta may be partly depen-
Stroemer RP, Rothwell NJ. 1997. Cortical protection by loca- dent on proteasome activity, but not phosphorylation and
lized striatal injection of IL1ra following cerebral ischemia ubiquitination of the IkappaBalpha molecule, in C6 glioma
in the rat. J Cereb Blood Flow Metab 17(6): 597-604. cells. Regulation of NFkappaB linked to chemokine pro-
Stroemer RP, Rothwell NJ. 1998. Exacerbation of ischemic duction. J Biol Chem 274(22): 15875-15882.
brain damage by localized striatal injection of interleukin Vitkovic L, Bockaert J, Jacque C. 2000. Inflammatory cyto-
1beta in the rat. J Cereb Blood Flow Metab 18(8): 833-839. kines: Neuromodulators in normal brain? J Neurochem
Sugimoto K, Iadecola C. 2003. Delayed effect of administra- 74(2): 457-471.
tion of COX2 inhibitor in mice with acute cerebral ische- Wang X, Barone FC, Aiyar NV, Feuerstein GZ. 1997. Interleu-
mia. Brain Res 960(12): 273276. kin1 receptor and receptor antagonist gene expression
Szaflarski J, Burtrum D, Silverstein FS. 1995. Cerebral hypoxia after focal stroke in rats. Stroke 28(1): 155-161.
ischemia stimulates cytokine gene expression in perinatal Wesche H, Korherr C, Kracht M, Falk W, Resch K, et al. 1997.
rats. Stroke 26(6): 1093-1100. The interleukin1 receptor accessory protein (IL1RAcP)
Takao T, Tracey DE, Mitchell WM, De Souza EB. 1990. is essential for IL1induced activation of interleukin1
Interleukin1 receptors in mouse brain: Characterization and receptorassociated kinase (IRAK) and stressactivated pro-
neuronal localization. Endocrinology 127(6): 3070-3078. tein kinases (SAP kinases). J Biol Chem 272(12): 7727-
Teng X, Zhang H, Snead C, Catravas JD. 2002. Molecular 7731.
mechanisms of iNOS induction by IL1 beta and IFN Wong ML, Licinio J. 1994. Localization of interleukin 1 type I
gamma in rat aortic smooth muscle cells. Am J Physiol receptor mRNA in rat brain. Neuroimmunomodulation
Cell Physiol 282(1): C144-C152. 1(2): 110-115.
Thompson JE, Phillips RJ, ErdjumentBromage H, Tempst P, Yang K, Mu XS, Hayes RL. 1995. Increased cortical nuclear
Ghosh S. 1995. I kappa Bbeta regulates the persistent factorkappa B (NFkappa B) DNA binding activity
response in a biphasic activation of NFkappa B. Cell after traumatic brain injury in rats. Neurosci Lett 197(2):
80(4): 573-582. 101-104.
ToliverKinsky T, Rassin D, Perez Polo JR. 2002. NFkappaB Yang GY, Liu XH, Kadoya C, Zhao YJ, Mao Y, et al. 1998.
activity decreases in basal forebrain of young and aged rats Attenuation of ischemic inflammatory response in mouse
after hyperoxia. Neurobiol Aging 23: 899-905. brain using an adenoviral vector to induce overexpression
Tomita N, Morishita R, Tomita S, Gibbons GH, Zhang L, et al. of interleukin1 receptor antagonist. J Cereb Blood Flow
2000. Transcription factor decoy for NFkappaB inhibits Metab 18(8): 840-847.
TNFalphainduced cytokine and adhesion molecule ex- Yue X, Mehmet H, Penrice J, Cooper C, Cady E, et al. 1997.
pression in vivo. Gene Ther 7(15): 1326-1332. Apoptosis and necrosis in the newborn piglet brain
Tomita N, Morishita R, Tomita S, Kaneda Y, Higaki J, et al. following transient cerebral hypoxiaischaemia. Neuro-
2001. Inhibition of TNFalpha induced cytokine and pathol Appl Neurobiol 23(1): 16-25.
102 5 Stress response signal transduction
Zabel U, Baeuerle PA. 1990. Purified human I kappa B Zhang MY, Harhaj EW, Bell L, Sun SC, Miller BA. 1998b.
can rapidly dissociate the complex of the NFkappa B trans- Bcl3 expression and nuclear translocation are induced by
cription factor with its cognate DNA. Cell 61(2): 255-265. granulocytemacrophage colonystimulating factor and
Zhang Z, Chopp M, Goussev A, Powers C. 1998a. Cerebral erythropoietin in proliferating human erythroid precur-
vessels express interleukin 1beta after focal cerebral ische- sors. Blood 92(4): 1225-1234.
mia. Brain Res 784(12): 210-217.
6 Aging and Oxidative Stress
Response in the CNS
V. Calabrese . D. A. Butterfield . A. M. Giuffrida Stella
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
6 Genetics of Human Longevity: Role of Vitagenes in Prolongation of Healthy Life Span . . . . . 133
8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
Abstract: Cellular oxidant/antioxidant balance has become the subject of intense study, particularly
focused on brain aging and neurodegenerative disorders. There is now evidence to suggest that reduction
of cellular expression and activity of antioxidant proteins and the resulting increase of oxidative stress are
fundamental causes for both the aging processes and neurodegenerative diseases. However, to survive
different types of injuries, brain cells have evolved networks of different responses, which detect and control
diverse forms of stress. Efficient functioning of maintenance and repair process seems to be crucial for both
survival and physical quality of life. This is accomplished by a complex network of the socalled longevity
assurance processes, which are composed of several genes termed vitagenes. Among these, heatshock
proteins (Hsps), proteasome, and mitochondrial uncoupling protein systems are highly conserved mechan-
isms responsible for the preservation and repair of the correct conformation of cellular macromolecules,
such as proteins, RNA, and DNA.
Recent studies have shown that the heatshock response contributes to establishing a cytoprotective state
in a wide variety of human diseases, including ischemia and reperfusion damage, inflammation, metabolic
disorders, cancer, infection, trauma, and aging. Among the various Hsps, Hsp32 also known as heme
oxygenase I (HO1), has received considerable attention, as it has been recently demonstrated that HO1
induction, by generating the vasoactive molecule carbon monoxide (CO) and the potent antioxidant
bilirubin, could represent a protective system potentially active against brain oxidative injury. The major
neurodegenerative diseases, Alzheimers disease (AD), Parkinsons disease (PD), amyotrophic lateral sclero-
sis (ALS), Huntingtons disease (HD), and Friedreichs ataxia (FA), are all associated with the presence of
abnormal proteins. Given the broad cytoprotective properties of the heatshock response, there is now strong
interest in discovering and developing pharmacological agents capable of inducing the heatshock response.
These findings have opened up new perspectives in medicine and pharmacology, as molecules inducing this
defense mechanism appear to be possible candidates for novel cytoprotective strategies. Particularly,
modulation of endogenous cellular defense mechanisms such as the heatshock response, and the protea-
somal system, through nutritional antioxidants or pharmacological compounds may represent an innova-
tive approach to therapeutic intervention in diseases causing tissue damage, such as neurodegeneration.
Moreover, by maintaining or recovering the activity of vitagenes, it would be possible to delay the aging
process and decrease the occurrence of agerelated diseases with resulting prolongation of a healthy life span.
List of Abbreviations: AD, Alzheimers disease; AP1, activator protein1; Ab, amyloid betapeptide; CNS,
central nervous system; GSH, reduced glutathione; GSSG, oxidized glutathione; Hsp, heatshock protein;
JAK, janus kinase; JNK, cjun Nterminal kinase; MAPK, mitogenactivated protein kinase; NFkB, nuclear
factor kappaB; NFT, intraneuronal fibrillary tangles; NOS, nitric oxide synthase; PD, Parkinsons disease;
PLA2, phospholipase A2; RNS, reactive nitrogen species; ROS, reactive oxygen species; SAPK, stress
activated protein kinase; STAT, signal transducer and transcription activator; TNF, tumor necrosis factor
1 Introduction
Reduction of cellular expression and activity of antioxidant proteins and the resulting increase of oxidative
stress are fundamental causes in the aging processes and neurodegenerative diseases (Mattson et al., 2002).
Numerous theories have been suggested to explain the aging process (> Table 61). However, current
thoughts generally propose that senescence results from various extrinsic events that lead progressively to
cell damage and death and/or characteristic intrinsic events related to the genomebased theory. These
general theories have been presented in various ways. It is agreed that aging is a combination of several
theories, the free radicals and mitochondrial theories presumably being the most important, while others
may play a definite, although less critical, role. However, it should be emphasized that no single theory is
entirely satisfactory.
Although several lines of evidence suggest that accumulation of oxidative molecular damage is a
primary causal factor in senescence, it is increasingly evident that the mitochondrial genome may play a
key role in aging and neurodegenerative diseases. Mitochondrial dysfunction is characteristic of several
neurodegenerative disorders, and evidence for mitochondria being a site of damage in neurodegenerative
Aging and oxidative stress response in the CNS 6 105
. Table 61
Theories of aging
A. Stochastic (random event)
1. Somatic mutation
2. Error catastrophe
3. Protein glycosylation
B. Developmental
1. Immune
2. Neuroendocrine
C. Genomebased
1. Intrinsic mutagenesis
2. Programmed
D. Free radical and mitochondrial dysfunction
disorders is partially based on decreases in respiratory chain complex activities in Parkinsons disease (PD),
Alzheimers disease (AD), and Huntingtons disease (HD) (Calabrese et al., 2001a). Such defects in
respiratory complex activities, possibly associated with oxidant/antioxidant balance perturbation, are
thought to underlie defects in energy metabolism and induce cellular degeneration. Among these, chaper-
ones are highly conserved proteins responsible for the preservation and repair of the correct conformation
of cellular macromolecules, such as proteins, RNA, and DNA. Chaperonebuffered silent mutations may be
activated during the aging process and lead to the phenotypic exposure of previously hidden features and
contribute to the onset of polygenic diseases, such as agerelated disorders, atherosclerosis, and cancer
(Soti and Csermely, 2002). Recently, the involvement of the heme oxygenase (HO) pathway in antidegen-
erative mechanisms operating in AD has received considerable attention, as it has been demonstrated
that the expression of HO is closely correlated to that of amyloid precursor protein (APP) (Dore, 2002;
Perry et al., 2003). HO induction, which occurs together with the induction of other Hsps during various
physiopathological conditions, by generating the vasoactive molecule carbon monoxide (CO) and
the potent antioxidant bilirubin, represents a protective system potentially active against brain oxidative
injury. HO1 gene is redox regulated and this is supported by the fact that HO1 gene has a heatshock
consensus sequence as well as activator proteins (AP)1, 2, and nuclear factor kappaB (NFkB) binding
sites in its promoter region. In addition, HO1 expression is rapidly upregulated by oxidative and
nitrosative stresses, as well as by glutathione depletion. Given the broad cytoprotective properties of the
heatshock response, there is now strong interest in discovering and developing pharmacological agents
capable of inducing the heatshock response (Butterfield et al., 2002a). In the present chapter, we discuss the
role of free radicals and mitochondria in brain aging and neurodegenerative disorders, then review the role
of NO and CO gases and that of the heatshock system in brain stress tolerance and their relevance to
mechanisms of longevity.
As one of the most prominent current theories of aging, the freeradical theory postulates that free radicals
generated through mitochondrial metabolism can act as a causative factor of abnormal function and cell death.
Various toxins in the environment can injure mitochondrial enzymes, leading to increased generation
of free radicals that over the life span would eventually play a major role in aging (Knight, 2000; Fries, 2002;
Anisimov et al., 2003; Sastre et al., 2003). During the last few years, cellular oxidant/antioxidant balance has
become the subject of intense study, particularly by those interested in brain aging and in neurodegenerative
mechanisms.
Several lines of evidence suggest that accumulation of oxidative molecular damage is a causal factor
in senescence. The direct evidence for this hypothesis is that overexpression of antioxidative genes for
106 6 Aging and oxidative stress response in the CNS
There are several enzymes systems that have been found to repair damage to DNA caused by oxidizing species
(Demple and Harrison, 1994). These include endonucleases, exonucleases, thymine glycol glycolases, and
DNA polymerases. DNA polymerases so far detected in the mammalian brain (a, b, d, and E) undergo age
dependent changes in activity, but it is not known which cell types contain which polymerases, and the ability
of nuclei from different brain regions to repair specific types of oxidative DNA damage is unknown. In
addition, recent evidence indicates that genetic instability, such as telomere loss and somatic and mtDNA
mutations, increases with age (Aviv et al., 2003). Levels of oxidative damage in mtDNA isolated from various
brain regions appear to be at least tenfold higher than those of nDNA (Bohr and Dianov, 1999), although
owing to technical difficulties there has as yet been no definitive study of oxidative damage to mtDNA
(Beckman and Ames, 1998). This increase correlates with the 17fold higher evolutionary mutation rate in
mtDNA compared with nDNA. These higher levels of oxidative damage and mutations in mtDNA recognize
different causal factors, including location of the DNA near the inner mitochondrial membrane sites (where
oxidants are generated), lack of protective histones, mitochondrial polymerase errors, and activations of
genes involved in errorprone DNA repair (Schapira, 1998). The ageassociated accumulation of oxidative
damage to mtDNA correlates with the level of mtDNA deletions found in various tissues composed of
postmitotic cells (Floyd and Hensley, 2002). It is possible that this damage leads to mutations that result in
mitochondrial dysfunction, which has been suggested to be involved in the pathogenesis of neurodegenera-
tive disorders (Calabrese et al., 2003a).
An increase in protein oxidative damage, as indicated by the loss of protein sulfhydryl groups and by a
decline in the activity of enzymes such as glutamine synthetase (GS) and glucose6phosphate dehydroge-
nase (G6PDH), has been demonstrated to occur in the brain during aging (Stadtman, 2001). A number of
experimental evidence indicates that increased rate of freeradical generation and decreased efficiency of the
reparative/degradative mechanisms, such as proteolysis, are the two factors that primarily contribute to age
related elevation in the level of oxidative stress and brain damage. With respect to this, it has been suggested
that decreases in levels of enzymes which ordinarily protect neuronal cells against oxidative stress with age
may be responsible for increased levels of freeradical damage in the brain, or that these enzymes themselves
are susceptible to inactivation by freeradical molecules which increase with age in the brain (Calabrese
et al., 2000a). During aging a number of enzymes accumulate as catalytically inactive or less active forms.
The agerelated changes in catalytic activity are due in part to reactions of proteins with oxygen and/or
nitrogen freeradical species produced during exposure to ionizing radiation or to metal ioncatalyzed
oxidation systems. The levels of oxidized proteins in brain extracts of rats of different ages increase
progressively with age, and in old rats can represent 30%50% of the total cellular protein (Calabrese
et al., 2001a). The agerelated increase in oxidized protein is accompanied by a loss of GS and G6PDH
activities, and by a decrease in the level of cytosolic neutral protease activity, which is responsible for the
degradation of oxidized (denatured) protein. Of particular significance are the results of experiments
showing that similar agerelated changes occur in the gerbil brain and that these changes are accompanied
by a loss of shortterm memory. Chronic treatment of old animals with the freeradical spintrap reagent
Ntertbutylaphenylnitrone (PBN) resulted in normalization of the several biochemical parameters to
those characteristic of the young animals; coincidentally, the shortterm memory index was restored to the
values seen in young animals (Carney et al., 1991). These results provide strong evidence that there is a
linkage between the agedependent accumulation of oxidized proteins and the loss in brain physiological
functions. It has recently been proposed that a primary mechanism leading to neuronal cell death in aging
and common neurodegenerative disorders is interference with proteasome function (Hyun et al., 2003).
Proteasomal dysfunction can involve genetic defects, direct inactivation of proteasome by reactive
oxygen and nitrogen species, or overloading with proteins. The latter can be caused by excessive production
of normal proteins or by the formation of poorly degradable proteins as a result of genetic mutations, faulty
posttranslational modification, or protein modification by freeradical damage. The major neurodegenera-
tive diseases, Alzheimers disease (AD), Parkinsons disease (PD), amyotrophic lateral sclerosis (ALS),
Huntingtons disease (HD) and Friedreichs ataxia (FA), are all associated with the presence of abnormal
proteins (Hyun et al., 2002). The origin of HD and FA involve specific genetic defects that lead to
production of abnormal proteins, whereas AD, ALS, and PD have been described mostly as sporadic,
108 6 Aging and oxidative stress response in the CNS
although familial types of AD, ALS, and PD are recognized. Examples are the rare mutations in synuclein or
parkin that cause familial PD and the 2% of patients with ALS associated with mutation in the gene
encoding Cu/ZnSOD (Halliwell, 2002). Even in the more common sporadic version of PD, ALS, and AD
abnormal proteins are present to a significant extent. Thus, the senile plaques typical of AD contain not
only bamyloid but also a wide range of other proteins. Most of them are oxidized and nitrated. Similar
damage has been described for proteins in Lewy bodies in sporadic PD. Oxidative as well as nitrosative
protein damage are also elevated in ALS. In nondividing cells, such as the great majority of neurons in the
adult brain, the protein content of cells is approximately constant. Since protein synthesis is continuous,
there must be an equilibrium between synthesis and degradation. Cellular protein can be degraded by the
lysosomal system, but a system of equal or greater importance is the proteasome. The 20S proteasome
(according to its sedimentation coefficient) is a cylindrical structure comprised of multiple protein subunits
and containing a narrow channel where proteolysis occurs. The 20S proteasome can degrade a wide range of
proteins including oxidatively damaged proteins, but most or all of the 20S proteasome in the cell is
associated with a 19S cap complex, which binds in an ATPdependent manner and confers specificity for
the degradation of polyubiquitinated proteins (Halliwell, 2001). Several other proteins are known that can
associate with proteasomes and increase (or in some case decrease) rates of protein clearance. Although it is
usually assumed that increased levels of oxidative damage are due to the increased generation of freeradical
species, however, increased levels of oxidative damage can equally ensue a decreased clearance of oxidatively
modified biomolecules. Oxidized protein levels in the central nervous system (CNS) tend to increase with
age (Friguet, 2002), consistent with several reports that proteasome activity decreases with age and in
neurodegenerative disorders (Mc Naught et al., 2001).
The agedependent accumulation of oxidized dysfunctional proteins with reactive carbonyl groups
leads to interand intramolecular crosslinks with protein amino groups, thus altering the efficiency of the
electron transport. Imbalances in the stoichiometry of functional electron transport proteins is proposed
to lead to a leakage in the flow of electrons to the terminal electron acceptor, cytochrome oxidase (Ojaimi
et al., 1999), and increased likelihood of superoxide generation. Studies on isoprostanes, the end product
of lipid peroxidation that can be measured in the CSF and urine in various neurodegenerative disorders,
suggest that lipid peroxidation is an early stage in these disease processes. Similarly, another end product
of lipid peroxidation, the aldehyde 4hydroxytransnonenal (HNE), which is highly neurotoxic, avidly
binds to proteins, and HNEprotein adducts are demonstrable in senile plaques and tangles in AD, tissues
from ALS patients, and Lewy bodies in PD. Protein carbonyls can be generated by direct oxidative damage
to proteins, by binding of cytotoxic aldehyde such as HNE to proteins, and by glycosidation of proteins
(Drake et al., 2003a). The content of protein carbonyls in Alzheimers brain samples is greater than in age
matched controls (Butterfield, 2002), and this provides the clearest indication of greater accumulation of
oxidized proteins in this disease. Brain regions show specific changes in this regard, and carbonyl levels
correlate well with tangles (Butterfield, 2002). Importantly, the accumulation of oxidation products in
particular regions of the brain seems to be related to specific cognitive defects (Forster et al., 1996). In support
of this, administration of the spintrap PBN to gerbils retards both protein oxidation and such neurological
defects (Carney et al., 1991).
inclusions. Cardiolipin, an acidic phospholipid that occurs only in mitochondria, has been shown to
decrease with age (Paradies et al., 2002). This inner membrane lipid is known to have optimal electrical
insulating properties, thereby contributing significantly to the transmembrane potential that drives the
formation of ATP via ATP synthase. Indeed, a decrease in membrane potential in mitochondria from older
animals has been demonstrated (Calabrese et al., 2001a).
It has been proposed that accumulation of mtDNA during life is a major cause of agerelated disease,
and this is because of its high mutagenic propensity. As discussed before, the lack of introns and protective
histones, limited nucleotide excision and recombination DNA repair mechanisms, and location in proxim-
ity of the inner mitochondrial membrane which expose it to an enriched freeradical milieu are all factors
contributing to a tenfold higher mutation rate occurring in the mtDNA than in the nDNA. Moreover, a
large body of evidence indicates that mtDNA mutations increase as a function of age, reaching the highest
levels in the brain and muscle. More than 20 different types of deletions have been documented to
accumulate in aging human tissues. The first report on an agerelated increase in mtDNA deletion was
found in brains from elderly subjects and in PD (Calabrese et al., 2001a). This deletion has been described
to occur between 13bp sequence repeats beginning at nucleotides 8470 and 13447, removing almost a 5kb
region of mtDNA between the ATPase 8 and the ND5 genes. The deletion is thought to occur during
replication of the mtDNA, the absent sequence encoding for six essential polypeptides of the respiratory
chain and 5 tRNAs. It has been associated with several clinical diseases, such as chronic progressive external
ophthalmoplegia and Kearns Sayre syndrome. Several agerelated disorders have been shown to be linked to
higher levels of mtDNA mutations than agematched controls. In the CNS, 17 times higher levels of the
common deletion in the striatum of patients with PD have been demonstrated, compared with age
matched controls. Evidence also exists indicating higher levels of this deletion in patients with AD, which
parallel increased levels in the oxidized nucleotide 8OHdG (Bohr and Dianov, 1999).
A major feature of mtDNA disease in humans is the presence of cells with low cytochrome c oxidase
activity, and evidence exists that indicates that the mechanism for these changes is likely to be clonal
expansion of individual mtDNA deletions within single cells (Schapira, 1998). Complex IVdeficient cells,
which occurred only sporadically earlier than the sixth decade of life, were present regularly after this age,
with the loss of enzyme activity being always confined to single, randomly distributed cells. Similarly,
cytochrome c oxidasenegative neurons have been demonstrated to exist in abundance in the CNS of
patients with mitochondrial disorders (Cottrell et al., 2001). These findings establish the relationship
between ageassociated accumulation of mtDNA mutations and bioenergy dysfunction as a key feature of
the aging process, at least in tissues predominantly composed of postmitotic cells, such as the CNS and
skeletal muscle. Relevant to mitochondrial bioenergetics, in fact, is the finding of a significant decrease in
state 3/state 4 ratio, which has been observed to occur in brain during aging (Calabrese et al., 2001a). Since
this ratio relates to the coupling efficiency between electron flux through the ETC and ATP production, an
increase in state 4 would result in a more reductive state of mitochondrial complexes and, consequently, to
an increase in freeradical species production. A decrease in state 3/state 4 respiration during aging has been
found associated with a significant decrease in cardiolipin content in brain mitochondria (Cottrell and
Turnbull, 2000). This loss could play a critically important role in the agerelated decrements in mitochon-
drial function and appears to be associated with both quantitative and qualitative regionspecific protein
changes, which are parallel to structural changes, such as decrease of the inner membrane surface, smaller as
well as sparser cristae, decreased fluidity, and increased fragility. Modifications in cardiolipin composition is
recognized to accompany functional changes in brain mitochondria, which include all proteins of the inner
mitochondrial membrane that generally require interaction with cardiolipin for optimal catalytic activity
(PorteroOtin et al., 2001; Quiles et al., 2002). Acetylcarnitine fed to old rats increased cardiolipin levels to
that of young rats and also restored protein synthesis in the inner mitochondrial membrane, as well as
cellular oxidant/antioxidant balance (Paradies et al., 1999), suggesting that administration of this com-
pound may improve cellular bioenergetics in aged rats (Hagen et al., 1998a, b). Interestingly, caloric
restriction, a dietary regimen that extends life span in rodents, maintains the levels of 18:2 acyl side chains
and inhibits the cardiolipin composition changes (Selman et al., 2003). In addition, caloric restriction was
shown to retard the agingassociated changes in oxidative damage, mitochondrial oxidant generation, and
antioxidant defenses (Mattson, 2003).
110 6 Aging and oxidative stress response in the CNS
There is now evidence to suggest that reduction of cellular expression and activity of antioxidant proteins
and the resulting increase of oxidative stress are fundamental causes for both aging process and neurode-
generative diseases (Ferri et al., 2003; Genova et al., 2003). However, to survive different types of injuries,
brain cells have evolved networks of different responses, which detect and control diverse forms of stress.
Efficient functioning of maintenance and repair process seems to be crucial for both survival and physical
quality of life. This is accomplished by a complex network of the socalled longevity assurance processes,
which are composed of several genes termed vitagenes. Signaling mechanisms adopted by regulatory
proteins to control gene expression in response to alterations in the intracellular redox status are very
common in prokaryotes. Gene activation by oxidative stress was first described in bacteria where
regulatory proteins such as OxyR were discovered as an activator of antioxidantand stressresponsive
genes. As the cytoprotective mechanism triggered by SoxR in E. coli includes the expression of critical
antioxidant defensive proteins, such as superoxide dismutase, the emerging concept is that an analogous
system might operate in mammalian cells (Calabrese et al., 2001a, 2003a, 2004a). In eukaryotes, typical
examples are the HO and Hsp genes, thioredoxin, detoxificant enzymes (MnSOD, glutathioneStransferase,
NADPH/quinone reductase), cytokines, immunoreceptors and growth factors (Calabrese et al., 2003c;
Colombrita et al., 2003).
NOS2 and, in others, inhibits the expression of NOS2 and consequently of NO (Motterlini et al., 2002a).
NO upregulates HO1 with production of CO (Maines, 1997). We have recently found evidence for a
functional relationship between CO and NO. In endotoxic shock, the salutary action of CO in the rat brain
appears to depend sequentially on the activation of NFkB, which triggers transcription of NOS2 with
production of NO, and subsequently in the upregulation of HO1. In the absence of any of these steps, the
beneficial effect of CO is lost (Scapagnini et al., 2002a). This has been also demonstrated in mice treated for
hepatitis induced by TNFa and Dgalactosamine (Otterbein et al., 2003a). To what extent CO and NO act
interdependently in other physiopathological conditions that are responsive to CO and/or NO is unknown.
Cys 199 being a likely candidate site of posttranslational modification (Rosenberg et al., 1999; Motterlini
et al., 2002a). The expression of these protective genes renders bacteria more resistant to oxidant
damage (Motterlini et al., 2002b). As the cytoprotective mechanism triggered by SoxR in E. coli includes
the expression of critical antioxidant defensive proteins, such as superoxide dismutase (Motterlini
et al., 2003), the emerging concept is that analogous systems might operate in mammalian cells. In
eukaryotes, typical examples are genes such as the HO gene, thioredoxin and detoxificant enzymes
(MnSOD, glutathioneStransferase, NADPH/quinone reductase), cytokines, immunoreceptors, and
growth factors. That the antioxidant protein HO could sense NO, and thus protecting against ROS and
RNS insults, is supported by the following findings: (1) NO and NOrelated species induce HO1 expression
and increase HO activity in human glioblastoma cells, hepatocytes, and aortic vascular cells; (2) cells
pretreated with various NOreleasing molecules acquire increased resistance to H2O2mediated cytotoxicity
when HO is maximally activated; and (3) bilirubin, one of the end products of heme degradation by HO,
protects against the cytotoxic effects caused by strong oxidants such as H2O2 and ONOO (Rosenberg et al.,
1999; Motterlini et al., 2002a). The conception that NO and RNS can be directly involved in the modulation
of HO1 expression in eukaryotes is based on the evidence that different NOreleasing agents can markedly
increase HO1 mRNA and protein, as well as HO activity, in a variety of tissues, including brain cells
(Scapagnini et al., 2002a). In rat glial cells, treatment with lipopolysaccharide (LPS) and interferong
(IFNg) results in a rapid increase in both iNOS expression and nitrite levels followed by enhancement of
HO1 protein (Calabrese et al., 2000b). In the same study, the presence of NOS inhibitors suppressed both
nitrite accumulation and HO1 mRNA expression. Modulation of HO1 mRNA expression by iNOS
derived NO following stimulation with LPS has also been reported in different brain regions, particularly in
the hippocampus and substantia nigra in an in vivo rat model of septic shock (Scapagnini et al., 2002a).
Moreover, the early increase in iNOS protein levels observed in endothelial cells exposed to low oxygen
tension seems to precede the stimulation of HO1 expression and activity, an effect that appears to be finely
regulated by redox reactions involving glutathione (Motterlini et al., 2000a, 2002a). Taken together, these
findings point to the central role of NO as a signaling molecule which, by triggering expression of
cytoprotective genes such as HO1, may lead to adaptation and resistance of brain cells to subsequent,
eventually more severe, nitrosative and oxidative stress insults (Butterfield et al., 2002b). Thus, a direct
interaction of NO groups with selective chemical sites localized in transcription proteins that can be
activated through nitrosative reactions could effectively contribute to the enhancement of both HO1
gene expression and stress tolerance. Recent knowledge concerning the modulation by thiol redox state of
the activity of several transcription factors that recognize specific binding sites within the promoter and
distal enhancer regions of the HO1 gene include Fos/Jun (AP1), NFkB, and the more recently identified
Nrf2 proteins (Balogun et al., 2003; Poon et al., 2004a). Importantly, both AP1 and NFkB contain cysteine
residues whose interaction with oxidant or nitrosant species might be crucial for determining the DNA
binding activity (Rosenberg et al., 1999; Motterlini et al., 2002a). Data in the literature show that NO can
either activate or inhibit these transcription factors, and that in many circumstances activation depends on
the reversibility of the posttranslational modification elicited by the various RNS (Butterfield et al., 2002a;
Poon et al., 2004b).
We have demonstrated in astroglial cell cultures that cytokineinduced nitrosative stress is associated
with an increased synthesis of Hsp70 stress proteins, which was also found after treatment of cells with the
NOgenerating compound sodium nitroprusside (SNP), thus suggesting a role for NO in inducing Hsp70
protein expression (Calabrese et al., 2000b). In vivo experiments performed in our laboratory have also
demonstrated that the redox glutathione status is a critical factor for induction of cytoprotective Hsp70
(Calabrese et al., 2000c, 2002c, 2004a; Balogun et al., 2003).
endogenously produced CO arose from the degradation of hemoglobin released from senescing erythro-
cytes (Siorstrand, 1949). Greater than 75% of CO produced in humans arises from erythrocyte turnover
generated as a byproduct of heme metabolism. In 1969, the source of endogenous CO was discovered, as
Tenhunen and collaborators (1969) described and characterized HO as the enzyme responsible for breaking
down heme in the body, demonstrating that heme catalysis resulted in the subsequent release of CO and free
iron as byproducts (Tenhunen et al., 1969). Since then, supported by a large body of experimental
evidence, CO is proving to be an extraordinary signaling molecule generated by the cell, which is vital in
the regulation of cellular homeostasis. In the brain, CO is emerging as a chemical messenger molecule
which can influence physiological and pathological processes in the central and peripheral nervous systems.
This gaseous molecule is now considered a putative neurotransmitter, owing to its capability to diffuse
freely from one cell to another, thereby influencing intracellular signal transduction mechanisms. However,
unlike conventional neurotransmitters, CO is not stored in synaptic vesicles and is not released by
membrane depolarization and exocytosis. It seems likely that CO is involved in the mechanism of cell
injury (Turcanu et al., 1998). This is evidenced by the fact that CO binds to iron of heme of the enzyme
guanylyl cyclase to activate cGMP (Piantadosi et al., 1997). Indeed, it has been found that CO is responsible
for maintaining endogenous levels of cGMP. This effect is blocked by potent HO inhibitors but not by NO
inhibitors (Maines, 1997). On the basis of endogenous distribution of HO in the CNS, it has been suggested
that CO can influence neurotransmission like NO (Verma et al., 1993). CO appears to be involved as a
retrograde messenger in LTP and also in mediating glutamate action at metabotropic receptors (Graser
et al., 1990). This is evident from the fact that metabotropic receptor activation in the brain regulates the
conductance of specific ions channels via a cGMPdependent mechanism, which is blocked by HO
inhibitors (Glaum and Miller, 1993). Experimental evidence suggests that CO plays a similar role like
NO in the signal transduction mechanism in regulating cell function and celltocell communication
(Maines, 1997). HO resembles NOS in that the electrons for CO synthesis are donated by cytochrome
P450 reductase, which is 60% homologous at the amino acid level to the half carboxyterminal of NOS
(Calabrese et al., 2001a). CO like NO binds to iron in the heme moiety of guanylyl cyclase. However, there
are some differences in function between CO and NO. Thus, NO mainly mediates glutamate effect at
NMDA receptors while CO is primarily responsible for glutamate action at metabotropic receptors. Taken
together, it appears that CO and NO play an important role in the regulation of CNS function, therefore
impairment of CO and NO metabolism results in abnormal brain function (Calabrese et al., 2000a). A
number of evidence suggest a possible role of CO in regulating nitrergic transmission. Endogenous CO has
been suggested to control constitutive NOS activity. Moreover, CO may interfere with NO binding to
guanylyl cyclase, and this in addition to the important role of HO in regulating NO generation, owing to its
function in the control of heme intracellular levels, as part of the normal protein turnover (Calabrese et al.,
2003a). This hypothesis is sustained by recent findings showing that HO inhibition increases NO produc-
tion in mouse macrophages exposed to endotoxin (Turcanu et al., 1998). CO may also act as a signaling
effector molecule, by interacting with targets other than guanylate cyclase. Notably, it has been recently
demonstrated that K(Ca) channels are activated by CO in a cGMPindependent manner (Wang and Wu,
2003) and also that COinduced vascular relaxation results from the inhibition of the synthesis of the
vasoconstrictor endothelin1 (Coceani et al., 1997). Little, however, is known about how CO is sensed on a
biological ground. Interestingly, the photosynthetic bacterium Rhodospirillum rubrum has the ability to
respond to CO through the heme protein CooA that, upon exposure to CO, acquires DNAbinding
transcriptional activity for the CO dehydrogenase gene, thereby encoding for CO dehydrogenase, which
is the key enzyme involved in the oxidative conversion of CO to CO2. Remarkably, heart cytochrome c
oxidase possesses COoxygenase activity, thus metabolizing CO to CO2 (Calabrese et al., 2001a). Whether
this occurs also in brain mitochondria remains to be elucidated. Aside from the CNS, the protective effects
of CO were initially demonstrated in a model of acute lung injury and endotoxic shock, and subsequently in
a mouse cardiac xenotransplantation model (Otterbein et al., 2003a). Mouse heart transplanted to
immunosuppressed rats survive indefinitely. However, if HO1 activity cannot be expressed in the mouse
heart, either as a consequence of absent phenotypical expression of the HO1 gene (mice hmox/) or as a
consequence of HO1 activity being inhibited with a selective inhibitor Tin protoporphyrin (SnPPIX), the
hearts are rejected rapidly. HO1 expression in the transplanted heart is essential to prevent rejection in this
Aging and oxidative stress response in the CNS 6 115
model. Surprisingly, if the donor and recipient were both treated with 250 ppm CO, even a heart that
cannot express HO1 activity still survives indefinitely (Otterbein, 2002). In this scenario CO appears to be
able to substitute for HO1 in suppressing the proinflammatory response, which is the leading cause of graft
rejection. CO emerges as a powerful antiinflammatory promoting agent acting at the level of the macro-
phage cell line, a cell that probably controls the balance of inflammation in many conditions. Macrophages
stimulated with bacterial LPS produce proinflammatory cytokines such as TNFa and an inflammatory
cytokine interleukin10 (IL10) is also produced (Otterbein et al., 2003b). If macrophages overexpress HO1
or are exposed to CO in vitro before stimulation with LPS, the proinflammatory response, and consequently
TNFa, is markedly diminished, whereas the antiinflammatory response, characterized by IL10 production,
is enhanced. At least, three important actions of CO contribute to its antiinflammatory effects: (1) CO
prevents platelet aggregation and the consequent thrombosis (Otterbein et al., 2003a); (2) CO down-
modulates the expression of plasminogen activator inhibitor type 1 (PAI1); and (3) CO prevents apoptosis
in several cell types, including endothelial cells, fibroblasts, hepatocytes, and pancreatic bcells (Otterbein
et al., 2003a). In addition, CO suppresses the proliferative response of smooth muscle cells, which contribute
to neointimal proliferation associated with inflammatory lesions in vivo. Many of the observed effects of CO
have been obtained by exposing cells or animals to gaseous CO and its subsequent inhalation. Interestingly,
the recently discovered COreleasing molecules (CORMS) appear to afford similar protective action,
thereby providing an alternative therapeutic approach for those pathophysiological conditions where CO
administration is warranted (Motterlini et al., 2002b, 2003).
After heat shock, for instance the synthesis of Hsp70 increases to a point where it becomes the most
abundant single protein in a cell. Once synthesized, Hsp70 binds to denatured proteins in an ATP
dependent manner. The Nterminal end contains an ATPbinding domain, whereas the Cterminal region
contains a substratebinding domain. Hsps serve as chaperones that bind to other proteins and regulate
their conformation, regulate protein movement across membranes or through organelles, or regulate the
availability of a receptor or activity of an enzyme.
In the nervous system, Hsps are induced in a variety of pathological conditions, including cerebral
ischemia, neurodegenerative disorders, epilepsy, and trauma. Expression of the gene encoding Hsps has
been found in various cell populations within the nervous system, including neurons, glia, and endothelial
cells (Kelly et al., 2002). Hsps consist of both stressinducible and constitutive family members. Whether
stress proteins are neuroprotective has been the subject of much debate, as it has been speculated that these
proteins might be merely an epiphenomenon unrelated to cell survival. Only recently, however, with the
availability of transgenic animals and gene transfer, it has become possible to overexpress the gene encoding
Hsp70 to test directly the hypothesis that stress proteins protect cells from injury, and it has been
demonstrated that overproduction of Hsp70 leads to protection in several different models of nervous
system injury (Wang and Wu, 2003). Following focal cerebral ischemia, mRNA encoding Hsp70 is
synthesized in most ischemic cells except in areas of very low blood flow, because of limited ATP levels.
Hsp70 proteins are produced mainly in endothelial cells, in the core of infarcts in cells that are most
resistant to ischemia, in glial cells at the edges of infarcts, and in neurons outside the areas of infarction. It
has been suggested that this neuronal expression of Hsp70 outside an infarct can be used to define the
ischemic penumbras, which is the zone of protein denaturation in the ischemic areas (Balogun et al., 2003).
A number of in vitro studies show that both heat shock and Hsp overproduction protect CNS cells against
both necrosis and apoptosis. Mild heat shock protects neurons against glutamatemediated toxicity and
protects astrocytes against injury produced by lethal acidosis (Narasimhan et al., 1996). Transfection of
cultured astrocytes with Hsp70 protects them from ischemia or glucose deprivation (Fink et al., 1997).
Hsp70 has been demonstrated to inhibit caspase3 activation caused by ceramide, and also affects JUN
kinase and p38kinase activation (Mosser et al., 1997). In addition, Hsp70 binds to and modulates the
function of BAG1, the bcl2 binding protein (McLaughlin et al., 2003), thus modulating some type of
apoptosisrelated cell death.
A large body of evidence now suggests a correlation between mechanisms of oxidative and/or nitrosa-
tive stress and Hsp induction. Current opinion holds also the possibility that the heatshock response can
exert its protective effects through inhibition of NFkBsignaling pathway (Calabrese et al., 2001a). We have
demonstrated in astroglial cell cultures that cytokineinduced nitrosative stress is associated with an
increased synthesis of Hsp70. Increase in Hsp70 protein expression was also found after treatment of
cells with the NOgenerating compound SNP, thus suggesting a role for NO in inducing Hsp70 proteins.
The molecular mechanisms regulating the NOinduced activation of the heatshock signal seems to involve
cellular oxidant/antioxidant balance, mainly represented by the glutathione status and the antioxidant
enzymes (Calabrese et al., 2000a, 2003a; Motterlini et al., 2000a).
Ubiquitin. Ubiquitin is one of the smallest Hsps and is expressed throughout brain in response to
ischemia. It is involved in targeting and chaperoning of proteins degraded in proteasomes, which include
NFkB, cyclins, HSFs, hypoxiainducible factor, some apoptosisrelated proteins, tumor necrosis factor, and
erythropoietin receptors (Mayer, 2003).
Hsp27. Hsp 27 is synthesized mainly in astrocytes in response to ischemic situations or to kainic acid
administration. It chaperones cytoskeletal proteins such as intermediate filaments, actin, or glial fibrillary
acidic protein following stress in astrocytes. It also protects against FasApo1, staurosporine, TNF, and
etoposideinduced apoptotic cell death as well as H2O2induced necrosis (Bechtold and Brown, 2003).
Hsp47. Hsp47 is synthesized mainly in microglia following cerebral ischemia and subarachnoid
hemorrhage (Valentim et al., 2003).
Hsp60, glucoseregulated protein 75 (GRP75), and Hsp10. Hsp60, glucoseregulated protein 75 (GRP75),
and Hsp10 chaperone proteins within mitochondria. GRP75 and GRP78, also called oxygenregulated
proteins (ORPs), are produced by low levels of oxygen and glucose. These protect brain cells against
ischemia and seizures in vivo, after viralinduced overexpression (Turner et al., 1999). Hsp60 is encoded in
Aging and oxidative stress response in the CNS 6 117
the nucleus and resides mainly in the mitochondria (Calabrese et al., 2003c, 2004a, 2005). Hsp60 forms
the chaperonin complex, which is implicated in protein folding and assembly within the mitochondria
under normal conditions (Izaki et al., 2001). Most mitochondrial proteins are synthesized in the cytosol
and must be imported into the organelles in an unfolded state (Izaki et al., 2001). During translocation, the
proteins interact with Hsp70. ATPdependent binding and release of Hsp70 provide the major driving force
for complete transport of polypeptides into the matrix. Most imported polypeptides are released from
soluble Hsp70; however, a subset of aggregationsensitive polypeptides must be transferred from Hsp70 to
Hsp60 for folding (Okubo et al., 2000). Owing to the close functional interaction between this chaperonin
system and the Hsp70 system, it is likely that upregulation of Hsp60 may be a fundamental mechanism
targeted by nutritional interventions leading to restoration of mitochondrial respiratory complex function
compromised by oxidative stress (Calabrese et al., 2004a, 2005; Perluigi et al., 2005; Poon et al., 2005;
Sultana et al., 2005).
Hsp32. Hsp32 or HO is the ratelimiting enzyme in the production of bilirubin. There are three
isoforms of HO, HO1 or inducible isoform, HO2 or constitutive isoform, and the recently discovered
HO3 (Scapagnini et al., 2002c).
HO isoenzymes are also seen to colocalize with different enzymes depending on the cell type. In the
kidney, HO1 colocalizes with erythropoietin, whereas in smooth muscle cells HO1 colocalizes with NOS.
In neurons, HO2 colocalizes with NOS, whereas the endothelium exhibits the same isoform that coloca-
lizes with NOS III. The cellular specificity of this pattern of colocalization lends further support to the
concept that CO may serve a function similar to that of NO. Furthermore, the brain expression pattern
shown by HO2 protein and HO1 mRNA overlaps with the distribution of guanylate cyclase, the main CO
functional target (Coceani et al., 1997).
HO3, the third isoform of HO, shares a high homology with HO2, both at the nucleotide (88%) and
protein (81%) levels. Both HO2 and HO3, but not HO1, are endowed with two CysPro residues
considered the core of the hemeresponsive motif (HRM), a domain critical for heme binding but not
for its catalysis (Hon et al., 2000).
Although the biological properties of this isoenzyme still remains to be elucidated, the presence of two
HRM motifs in its amino acidic sequence might suggest a role in cellular heme regulation (Maines, 2000).
Studying the HO3 mRNA sequence (GenBank accession no.: AF058787), we have observed that its 50
portion corresponds to an L1 retrotransposon sequence, a member of a family of retrotransposons recently
found to be involved in evolutionary mechanisms (Kazazian, 2000). On the basis of the close similarity to a
paralogous gene (HO2) and on the preliminary data from our group demonstrating the absence of introns
in the HO3 gene (Scapagnini et al., 2002c), it is possible that this 50 -portion could have originated from the
retrotransposition of the HO2 gene. In addition, this genetic mutation in the rat may represent a species
specific event since no other sequence in the public databases match that of the rat HO3.
Induction of HO1 gene could be used for clinical diagnosis. However, the length of the GT polymor-
phism in the promoter of the gene encoding HO1 that regulates the magnitude of the HO1 response to a
given stress signal can render this approach difficult for those individuals with long GT repeats, which are
associated with low HO1 responsiveness. This polymorphism appears to be of functional significance in
the short repeats, which are associated with high responsiveness and seem to be also associated with lesser
likelihood of restenosis after angioplasty (Otterbein et al., 2003a).
HO1 gene is induced by a variety of factors, including metalloporphyrins and hemin, as well as
ultraviolet A (UVA) irradiation, hydrogen peroxide, prooxidant states, or inflammation (Tyrrell, 1999).
This characteristic inducibility of HO1 gene strictly relies on its configuration: the 6.8kb gene is organized
into four introns and five exons. A promoter sequence is located approximately 28 base pairs upstream from
the transcriptional site of initiation. In addition, different transcriptional enhancer elements such as the
HSE and metal regulatory element reside in the flanking 50 region. Also, inducerresponsive sequences have
been identified in the proximal enhancer located upstream the promoter and, more distally, in two
enhancers located 4 kb and 10 kb upstream the initiation site (HillKapturczak et al., 2003). The molecular
mechanism that confers inducible expression of HO1 in response to numerous and diverse conditions has
remained elusive. One important clue has recently emerged from a detailed analysis of the transcriptional
regulatory mechanisms controlling the mouse and human HO1 genes. The induction of HO1 is regulated
principally by two upstream enhancers, E1 and E2 (Sun et al., 2002). Both enhancer regions contain multiple
stress (or antioxidant)responsive elements (StRE, also called ARE) that also conform to the sequence of the
Maf recognition element (MARE) (Martin et al., 2004) with a consensus sequence (GCnnnGTA) similar to
that of other antioxidant enzymes (Balogun et al., 2003). There is now evidence to suggest that heterodimers
of NFE2related factors 2 (Nrf2) and one or another of the small Maf proteins (i.e., MafK, MafF and MafG)
are directly involved in induction of HO1 through these MAREs (Gong et al., 2002). A possible model,
centered on Nrf2 activity, suggests that the HO1 locus is situated in a chromatin environment that is
permissive for activation. Since the MARE can be bound by various heterodimeric basic leucine zipper
(bZip) factors including NFE2, as well as several other NFE2related factors (Nrf1, Nrf2, and Nrf3), Bach,
Maf, and AP1 families (Sun et al., 2002), random interaction of activators with the HO1 enhancers
would be expected to cause spurious expression. This raises a paradox as to how cells reduce transcriptional
noise from the HO1 locus in the absence of metabolic or environmental stimulation. This problem could
be reconciled by the activity of repressors that prevent nonspecific activation. One possible candidate is the
heme protein Bach1, a transcriptional repressor endowed with DNAbinding activity, which is negatively
regulated upon binding with heme. Bach1heme interaction is mediated by evolutionarily conserved heme
regulatory motifs (HRM), including the cysteineproline dipeptide sequence in Bach1. Hence, a plausible
model accounting for the regulation of HO1 expression by Bach1 and heme is that expression of HO1
gene is regulated through antagonism between transcription activators and the repressor Bach1. While
under normal physiological conditions expression of HO1 is repressed by Bach1/Maf complex, increased
levels of heme displace Bach1 from the enhancers and allow activators, such as heterodimer of Maf with
Nrf2, to promote the transcription of HO1 gene (Sun et al., 2002). To our knowledge, the Bach1/HO1
system is the first example in higher eukaryotes that involves a direct regulation of a transcription factor for
an enzyme gene by its substrate. Thus, regulation of HO1 involves a direct sensing of heme levels by Bach1
(by analogy to lac repressor sensitivity to lactose), generating a simple feedback loop whereby the substrate
effects repressoractivator antagonism.
The promoter region also contains two metalresponsive elements, similar to those found in the
metallothionein1 gene, which respond to heavy metals (cadmium and zinc) only after recruitment of
another fragment located upstream, between 3.5 and 12 kbp (CdRE). In addition, a 163bp fragment
containing two binding sites for HSF1, which mediates HO1 transcription, are located 9.5 kb upstream of
the initiation site (Balogun et al., 2003). The distal enhancer regions are important in regulating HO1 in
inflammation, as they have been demonstrated to be responsive to endotoxin. In the promoter region, there
also resides a 56 bp fragment that responds to the STAT3 acutephase response factor, involved in the
downregulation of HO1 gene induced by glucocorticoid (Raju et al., 1997).
ROS in brain cells (Butterfield et al., 2002b). A compromised GSH system in the brain has been associated
with the oxidative stress occurring in neurological diseases (Butterfield et al., 2002c). Recent data demon-
strate that, besides intracellular functions, GSH has also important extracellular functions in the brain. In
this respect astrocytes appear to play a key role in the GSH metabolism of the brain, since astroglial
GSH export is essential for providing GSH precursors to neurons (Dringen and Hirrlinger, 2003). Of
the different brain cell types studied in vitro only astrocytes release substantial amounts of GSH. In
addition, during oxidative stress astrocytes efficiently export glutathione disulfide (GSSG). The multi-
drugresistance protein 1 participates in both the export of GSH and GSSG from astrocytes (Dringen and
Hirrlinger, 2003). Glutathione exists in either a reduced (GSH) or oxidized (GSSG) form and participates in
redox reactions through the reversible oxidation of its active thiol. In addition, GSH acts as a coenzyme of
numerous enzymes involved in cell defense. In unstressed cells the majority (99%) of this redox regulator is
in the reduced form, and its intracellular concentration is between 0.5 and 10 mM depending on the cell
type (Drake et al., 2003b). Depletion of glutathione has been shown to occur in conditions of moderate or
severe oxidative stress and has been associated with increased susceptibility to cell damage (Calabrese et al.,
2000b; Balogun et al., 2003; Catania et al., 2003). There is now evidence to suggest that a direct link between
a decrease in glutathione levels by oxidant stress and rapid upregulation of HO1 mRNA and protein exist
in a variety of cells, including the rat brain, human fibroblasts, endothelial cells, and rat cardiomyocytes
(Foresti and Motterlini, 1999). This finding is supported by the fact that Nacetylcysteine (a precursor of
glutathione) abolishes oxidative stressmediated induction of HO1 gene (Foresti et al., 1997, 2001). In
addition, increased production of NO and RSNO can also lead to changes in intracellular glutathione.
In astroglial cell cultures, stimulation of iNOS by exposure to LPS and IFNg decreases total glutathione, while
increasing GSSG, and this effect was abolished by pretreatment of glial cells with NOS inhibitors (Calabrese
et al., 2000b). Moreover, elevation of intracellular glutathione prior to exposure of endothelial cells to NO
donors almost completely abolishes activation of the HO pathway, which suggests that thiols can antagonize
the effect of NO and NOrelated species on HO1 induction (Calabrese et al., 2000b). We have recently
demonstrated in endothelial cells subjected to hypoxia that induction of HO1 is associated with a decrease in
the GSH/GSSG ratio and with an increase in RSNO levels resulting from early induction of iNOS (Motterlini
et al., 2000a). This implies that in conditions of low oxygen availability both oxidative and nitrosative reactions
may serve as a trigger for induction of the HO1 gene (Motterlini et al., 2000b). All these evidence corroborate
the notion that generation of ROS and RNS are important signal transduction mechanisms linking HO1
activation to cell stress tolerance (Mancuso et al., 2003; Pocernich et al., 2005).
microglia and macrophages occur long after traumatic injury and are correlated with Wallerian degeneration
and remodeling of surviving tissue (Mautes et al., 2000).
Since the expression of Hsps is closely related to that of APP, Hsps proteins have been studied in the
brain of patients with AD. Significant increases in the levels of HO1 have been observed in AD brains in
association with neurofibrillary tangles (NFTs) (Takeda et al., 2000), and also HO1 mRNA was found
increased in AD neocortex and cerebral vessels (Premkumar et al., 1995). HO1 increase was not only in
association with NFTs but also colocalized with senile plaques and glial fibrillary acidic proteinpositive
astrocytes in AD brains (Schipper et al., 2000). It is conceivable that the dramatic increase in HO1 in AD
may be a direct response to increased free heme associated with neurodegeneration and an attempt to
convert the highly damaging heme into the antioxidants biliverdin and bilirubin (Calabrese et al., 2004b).
Upregulation of HO1 in the substantia nigra of patients with PD has been demonstrated. In these
patients, nigral neurons containing cytoplasmic Lewy bodies exhibited in their proximity maximum HO1
immunoreactivity (Ewing and Maines, 1992). New evidence showed a specific upregulation of HO1 by
oxidative stress in the nigral dopaminergic neurons (Poon et al., 2004a).
Multiple sclerosis (MS) is a common, often disabling disease of the CNS. It has been suggested that
inappropriate stress response within the CNS could influence both the permeability of the bloodbrain
barrier and the expression of Hsps, thereby initiating the MS lesion (Aquino et al., 1977; Maines, 2000).
However, cytokines, immunoglobulins, and complement complexes may elicit a survival response in the
oligodendrocytes, involving the induction of endogenous Hsps and other protective molecules, which
indicates that redox systems and therefore the oxidant/antioxidant balance in these cells are of great
importance in MS (Calabrese et al., 1994, 1998, 2002a, 2003b; Bagasra et al., 1995). The expression of
HO1 is increased in the CNS of mice and rats with experimental allergic encephalomyelitis (EAE), an
animal model of MS (Catania et al., 2003). To investigate the role of HO1 in EAE, SnPPIX was
administered to SJL mice during active disease. SnPPIX (200 mmol/kg) attenuated clinical scores, weight
loss, and some signs of pathology in comparison to vehicle treatment. Glutathione levels were greater in
treated EAE mice than in those receiving vehicle, indicating lower oxidative stress in the former group.
These data suggest that inhibition of HO1 attenuated disease and suppressed freeradical production
(Chakrabarty et al., 2003). On the contrary, in another study, high expression of HO1 in lesions of EAE was
enhanced by hemin treatment, a procedure that resulted in the attenuation of clinical signs of pathology,
whereas tin mesoporphyrin, an inhibitor of HO1, markedly exacerbated EAE (Liu et al., 2001). These
results strongly suggest that endogenous HO1 plays an important protective role in EAE, and that targeted
induction of HO1 overexpression may represent a new therapy for the treatment of MS. We have recently
shown that thiol disruption and nitrosative stress are associated in active MS with induction of Hsp70 and
HO1 in central and peripheral tissues of MS patients and that acetylcarnitine was able to counteract nitrosative
stressmediated damage, an effect associated with enhancement of Hsp stress signaling (Poon et al., 2004a).
All these findings can open up new therapeutic perspectives, as molecules activating these defense mechanisms
appear to be possible candidates for novel neuroprotective strategies (Calabrese et al., 2000c).
with relatively weak endogenous antioxidant defenses such as the CNS and the myocardium. Interestingly,
increased levels of bilirubin have been found in the cerebrospinal fluid in AD, which may reflect the increase
of degraded bilirubin metabolites in the AD brain, derived from the scavenging reaction against chronic
oxidative stress (Kimpara et al., 2000). Similarly, a decreased risk for coronary artery disease is associated
with mildly elevated serum bilirubin, with a protective effect comparable to that of HDL cholesterol (Dore
et al., 2000). The most likely explanation for the potent neuroprotective effect of bilirubin is that a redox
cycle exists between bilirubin and biliverdin, the major oxidation product of bilirubin. In mediating the
antioxidant actions, bilirubin would be transformed to biliverdin, then rapidly converted back to bilirubin
by biliverdin reductase, which in the brain is present in large functional excess, suggesting a mechanism to
amplify the antioxidant effect (Poon et al., 2004b). Remarkably, the rapid activation of HO2 by protein
kinase C (PKC) phosphorylation parallels the availability of nNOS. Both are constitutive enzymes localized
in neurons, and nNOS is activated by calcium entry into cells binding to calmodulin. Similarly, PKC
phosphorylation of HO2 and the transient increase in intracellular bilirubin would provide a way for a
rapid response to calcium entry, this being a major activator of PKC. Recent evidence has demonstrated that
bilirubin and biliverdin possess strong antioxidant activities toward peroxyl radical, hydroxyl radical, and
hydrogen peroxide. Exposure of bilirubin and biliverdin to agents that release NO or nitroxyl resulted in a
concentrationand timedependent loss of bilirubin and biliverdin. Increasing concentrations of thiols
prevented bilirubin and biliverdin consumption by nitroxyl, indicating that bile pigments and thiol groups
can compete and/or synergize the cellular defence against NOrelated species. In view of the high inducibil-
ity of heme oxygenase1 by NOreleasing agents in different cell types, these findings highlight novel
antinitrosative characteristics of bilirubin and biliverdin, suggesting a potential function for bile pigments
against the damaging effects of uncontrolled NO production (Kaur et al., 2003).
Caloric restriction in mammals has been recognized as the best characterized and most reproducible
strategy for extending maximum life span, retarding physiological aging, and delaying the onset of age
related pathological situations. The overwhelming majority of studies using caloric restriction have used
shortlived rodent species, although current work using monkeys should reveal whether this paradigm is
also relevant for manipulating the rate of primate aging. The mechanisms by which restricted calorie intake
modifies the rate of aging and cellular pathology have been the subject of much controversy, although an
attenuation of accumulating oxidative damage appears to be a central feature (Hursting et al., 2003). A
major effect of calorierestricted feeding now appears to be on the rate of production or leak of free radicals
from mitochondrial sites, although the details of the adaptation and the signaling pathway that induce this
effect are currently unknown. General consensus, however, has been achieved that caloric restriction feeding
regimes reduce the rate of accrual of oxidative damage as measured by lipid peroxidation, nuclear and
mtDNA damage, and protein carbonyl formation. An analysis of published studies that used a degree of
food restriction in the range of 40%50% ad libitum intake revealed a significant positive correlation
between survival parameters, such as mean, maximum, and average survival time, and duration of caloric
restriction. The longer the animals are maintained on low calorie intake during the postweaning period of
the life span, the greater is the survival (Mattson et al., 2002). It is unclear whether caloric restriction
protects against random oxidative damage per se or is protective for those vulnerable proteins of key
pathways, such as those containing ironsulfur centers of the ETC or DNAbinding signaling proteins. This
is directly related to the question whether oxidative damage in genomic and mtDNA is primarily random as
a function of age or whether there is a specific pattern of distribution of ROS which may vary depending on
the tissue or the state of the cell cycle within any particular cell. It is generally accepted that agerelated
accrual of ROSinduced damage represents a balance between generation and defences, such as antioxidant
enzymes, repair systems, and turnover. It has been demonstrated that caloric restriction reduces cellular
injury and improves heat tolerance of old animals by lowering radical production and preserving cellular
ability to adapt to stress through antioxidant enzyme induction and translocation of these proteins to the
Aging and oxidative stress response in the CNS 6 123
nucleus (Calabrese et al., 2001a). It has been also demonstrated that mitochondria from calorierestricted
animals produce less ROS per nanomole of O2 during state 4 respiration, and recent work on ETC
complexes suggests a modification in the Km for complex III associated with a retention of highaffinity
binding sites for complex IV as a possible mechanism operating in reducing superoxide generation
(Mattson et al., 2003). It is conceivable that low calorieinduced changes in unsaturated fatty acid
composition of the mitochondrial membranes not only may protect against ROSinduced lipid peroxida-
tion but also may influence the binding properties of ETC proteins embedded in the membrane and the
related transport processes. However, several questions need to be addressed such as the signaling pathway
underlying the adaptive responses triggered by caloric restriction, or the effect of chronic caloric restriction
on either the bioenergetics of individual mitochondria or the mitochondrial number and turnover rate.
Highdensity oligonucleotide array studies have recently provided compelling evidence that aging results in
a differential gene expression pattern indicative of a marked stress response associated with lower expression
of metabolic and biosynthetic genes, and also, these alterations are either completely or partially prevented
by caloric restriction. In addition, the transcriptional patterns of calorierestricted animals suggest that
caloric restriction retards the aging process by causing a metabolic shift toward increased protein turnover
and decreased macromolecular damage (Martin et al., 2003; Strauss, 2003; Calabrese et al., 2004b).
The major CNS disorders associated with ROS/RNSmediated damage are listed in > Table 62.
124 6 Aging and oxidative stress response in the CNS
. Table 62
Neurodegenerative disorders associated with freeradical damage
1. Amyotrophic lateral sclerosis
2. Alzheimers disease
3. Parkinsons disease
4. Multiple sclerosis
5. Friedreichs ataxia
6. Downs syndrome
7. Huntingtons disease
8. Ischemia/reperfusion
neurons. A prediction of this model is that the antioxidant vitamin E should prevent or modulate these Ab
induced effects to neurons (Butterfield et al., 1997). Consistent with this model, this freeradical scavenger
was shown to block Abinitiated lipid peroxidation in cortical synaptosomes (Butterfield et al., 1999).
Further, protein oxidation induced by Ab in astrocyte cultures and assessed by increased protein carbonyl
content was abrogated by the more soluble form of vitamin E, trolox (Koppal et al., 1998). Vitamin E also
blocked Abinduced inhibition of transmembrane protein function, including ionmotive ATPases, glucose
and glutamate transporters, Gprotein coupled signal transduction, and the energyrelated enzyme creatine
kinase, and the methionine residue 35 of Ab142 and Ab140 was shown to be critical to the oxidative stress
properties of these peptides (Butterfield and Kanski, 2002). Human Ab142, expressed in vivo in transgenic
Caenorhabditis elegans nematodes, led to protein oxidation in the living animal, and methionine was
important in this process as well (Yatin et al., 1999).
A risk factor for AD is the presence of allele 4 of apolipoprotein E (apoE) (Roses, 1996). Synaptosomes
from apoEknockout mice, containing no gene for apoE, show increased susceptibility to oxidative stress
induced by Ab, (Lauderback et al., 2001), while synaptosomes from knockout mice containing human
apoE4 with no mouse background show significantly increased Abinduced oxidative stress compared with
synaptosomes from human apoE2 or apoE3 knockin mice (Lauderback et al., 2002). Thus, apoE may serve
an antioxidant function, but apoE4 may be less able than apoE2 or apoE3 to do so (Lauderback et al., 2002).
This notion was tested using 1monthold control and apoEdeficient mice. Both received dietary vitamin E
for 12 months. Vitamin Efed animals had better behavioral outcomes of spatial motor activity and
decreased levels of lipid peroxidation relative to apoEdeficient mice fed a normal diet (Veinbergs et al.,
2000). The sum of these studies suggests a decreased risk for and diminished oxidative stress in AD in
persons taking high dose dietary, or perhaps supplemental, vitamin E (and vitamin C to regenerate vitamin
E from the tocopherol radical).
Brains of AD patients undergo many changes, such as disruption of protein synthesis and degradation,
classically associated with the heatshock response, which is one form of stress response. Hsps are proteins
serving as molecular chaperones involved in the protection of cells from various forms of stress. Increasing
interest has focused on identifying dietary compounds that can inhibit, retard, or reverse the multistage
pathophysiological events underlying AD pathology. AD, in fact, involves a chronic inflammatory response
associated with both brain injury and Abassociated pathology. All of the above evidence suggests that
stimulation of various repair pathways by mild stress has significant effects on delaying the onset of various
ageassociated alterations in cells, tissues, and organisms. Spice and herbs contain phenolic substances with
potent antioxidative and chemopreventive properties, and it is generally assumed that the phenol moiety is
responsible for the antioxidant activity. In particular, curcumin, a powerful antioxidant derived from the curry
spice turmeric, has emerged as a strong inducer of the heatshock response. In light of this finding, curcumin
supplementation has been recently considered as an alternative, nutritional approach to reduce oxidative
damage and amyloid pathology associated with AD (Butterfield et al., 2002a, c; Calabrese et al., 2004b).
Conceivably, dietary supplementation with vitamin E or with polyphenolic agents, such as curcumin
and its derivatives, can forestall the development of AD, consistent with a major metabolic component to
this disorder. Nutritional biochemical research is providing optimism that this devastating brain disorder of
aging may be significantly delayed and/or modulated.
Decreased glutathione peroxidase and catalase activities associated to increased SOD activity leads to increased
levels of hydrogen peroxide (Dexter et al., 1991). This, in dopaminergic cells, is primarily produced by MAO
via deamination of dopamine and also nonenzymatically by autoxidation of dopamine. Hydrogen peroxide, by
reacting with reduced forms of transition metals, e.g., iron (II) or copper (I), gives rise to the powerful oxidant
hydroxyl radical, and oxidative damage to nigral membrane lipids, proteins, and DNA ensues. The role of iron
in brain oxidative injury has been extensively considered (Mattson et al., 2002). Dexter and coworkers (1991)
reported a 31%35% increase in the total iron content in parkinsonian SN compared with control tissue,
which was associated with decreased levels of the iron storage protein ferritin, contrasting with a signifi-
cant decrease of the levels of ironbinding protein in the CSF. A shift of iron II/iron III ratio in the
substantia nigra from almost 2:1 in the normal brain to 1:2 in the parkinsonian brain is also served (Li
and Dryhurst, 1997). Hence, a distinct possibility exists that excessive freeradical generation occurs in this
region, leading to the death of nigral neurons. In addition, substantia nigra is a dopaminerich brain area,
and catechols, including DOPA and dopamine, have been demonstrated to be cytotoxic in vitro, presum-
ably by formation of covalent bonds between their quinone forms and macromolecules of vital impor-
tance, primarily represented by thiol groups (Spencer et al., 2002). In fact, an intermediate in the
autoxidation of catechols to quinone is the free radical semiquinone. Both autoxidation steps generate
reduced forms of molecular oxygen such as superoxide anion and hydrogen peroxide which, in addition to
hydrogen peroxide produced by the MAOdependent catabolism of dopamine, contribute to maintain
considerable levels of the highly reactive hydroxyl radical, which on reacting with free thiol groups may
contribute to the decreased levels of GSH and a corresponding increase in GSSG found in the SN
(Calabrese et al., 2002b, 2004a). This is of special importance considering that nigral cells also contain
neuromelanin, a pigmented substance related to lipofuscin and derived from dopamine. Neuromelanin has
been demonstrated to have high affinity for iron III, and this ironmelanin interaction might have
pathogenetic implications. In fact, the synthesis of neuromelanins from dopamine is known to produce
more oxidative damage than the synthesis from other catecholamines (Spencer et al., 1994) and, in
addition, neuromelanins polymerize from pheomelanin in a process that requires cysteine for synthesis,
thus competing with gglutamyl cysteine synthetase which utilizes cysteine for GSH synthesis. Under these
circumstances the GSH system in the substantia nigra could result in a position of increased demand and
decreased synthetic capability, and hence contribute to the high vulnerability of this region to peroxidative
injury (Calabrese et al., 2001a, 2002b, 2004a). This is confirmed by the study of Perry and coworkers (1982),
which showed that GSH levels in the SN were significantly lower than in other brain regions. Moreover, a 40%
decrease in GSH in the SN of PD, associated with significant increase in oxidized glutathione, has been also
reported (Sian et al., 1994). Recently, it has been demonstrated in PD patients that the proportion of
dopaminergic neurons with immunoreactive NFkB in their nuclei was more than 70fold than that in control
subjects (Hunot et al., 1997). A possible relationship between the nuclear localization of NFkB in mesence-
phalic neurons of PD patients and oxidative stress in such neurons has been shown in vitro with primary
cultures of rat mesencephalon, where translocation of NFkB is preceded by a transient production of free
radicals during apoptosis induced by activation of the sphingomyelindependent signaling pathway with C2
ceramide (FranceLanord et al., 1997). Data suggest that this oxidantmediated apoptogenic transduction
pathway may play a role in the mechanism of neuronal death in PD (Schapira et al., 1990; Mc Naught et al.,
2001; Dawson and Dawson, 2002; Moore et al., 2003). Moreover, a potential role for excitotoxic processes in PD
has been strengthened by the observation that there appears to be a mitochondrially encoded defect in complex
I activity of the ETC (Schapira et al., 1990). An impairment of oxidative phosphorylation will enhance
vulnerability to excitotoxicity (Xin et al., 2000). Substantia nigra neurons possess NmethylDaspartate
receptors, and there are glutamatergic inputs into the substantia nigra from both the cerebral cortex and
the subthalamic nucleus. After activation of excitatory amino acid receptors, it has been suggested that there
is an influx of calcium followed by activation of nNOS, which can then lead to the generation of
peroxynitrite (Bechtold and Brown, 2003). Consistent with such a mechanism, studies of MPTP neurotox-
icity in both mice and primates have shown that inhibition of nNOS exerts neuroprotective effects, raising
the prospect that excitatory amino acid antagonists for nNOS inhibitors might be useful in the treatment of
PD (Dawson and Dawson, 1995, 2002).
128 6 Aging and oxidative stress response in the CNS
Recent clinical and animal studies also indicate that NO and ONOO play a central role in
the pathogenesis of MS (reviewed in Calabrese et al., 2001a). It has also been shown that in CSF and
plasma nitrite nitrate (stable end products of NO metabolism) levels are elevated in patients with MS
(Giovannoni, 1998).
ROS and RNS have a major role in the mediation of cell damage, and free sulfhydryl groups are vital in
cellular defence against endogenous or exogenous oxidants (Calabrese et al., 2000a, 2001b). The possible
links between MS and oxidant/antioxidant balance in cell perturbation may be suggested by several factors,
including increased incidence of MS in populations consuming high proportions of animal fat, (Calabrese
et al., 1994), increased malonaldehyde levels in blood, and decreased glutathione peroxidase activity in MS
erythrocytes, lymphocytes and granulocytes (Calabrese et al., 1998) and, in addition, an inappropriate
expression of Hsps on oligodendrocytes (Calabrese et al., 2002a). This last event could represent a possible
initiating factor at the level of MS lesions, capable of modulating the subsequent susceptibility or resistance
of cells to oxidative stress. Moreover, a decrease in sulfhydryl groups and increased amounts of lipid
peroxidation products have also been measured in the CSF and plasma of MS patients (Calabrese et al.,
2002a). Nitrosative stress in isolated astrocytes in vitro causes modifications in the endogenous thiol pool
associated with induction of Hsp32 or HO1, which is prevented by antioxidants, suggesting a biochemical
link between nitrosative stress, sulfhydryl function, and the heatshock pathway (Scapagnini et al., 2002d;
Calabrese et al., 2004b). In addition, this evidence suggests that redoxactive compounds such as glutathi-
one and the overall oxidant/antioxidant balance in the CNS are potentially of great importance in MS,
although as yet there has been few studies addressing the relationships between NO, ONOO, and
glutathione in MS. The chemical composition of human CSF is considered to reflect brain metabolism
(Thompson, 1988), and we have recently demonstrated in MS patients decreased levels of protein sulfhydryl
groups associated with an increase in RNS and peroxidative products (Calabrese et al., 2002a, 2003b). More
recently, we have provided experimental evidence that increased levels of RNS are present in the CSF of MS
patients, and this is associated with increased nitrosylation of sulfhydryl moieties. Our results are consistent
with evidence indicating increased protein nitrosylation in MS patients (Cross et al., 1997) and pose
intriguing implications regarding clinical manifestations in MS, which are potentially linked to a failure
of action potentials to propagate along damaged axons and involve inflammatory processes as primary
causative factors in addition to demyelination. In favor of this possibility is the evidence that NO donors are
capable of blocking conduction in rat demyelinated axons (Garthwaite et al., 2002). All this would suggest a
broader potential role for NO in the symptomatic manifestations of MS. Whether or not NO is central to
the pathogenesis of MS remains to be clarified, owing to its role of being a doubleedged sword..
Consistently, a recent study (Sellebjerg et al., 2002) has demonstrated an association between high CSF
levels of NO metabolites with severe disease activity in relapsing/remitting MS, and high concentrations of
NO metabolites were associated with more pronounced treatment responses after methylprednisolone
treatment. However, other studies have shown no significant correlation between NO metabolites and
disability score, disease progression index, MRI activity, and development of cortical atrophy on MRI.
(Yuceyar et al., 2001). We have also demonstrated in MS patients an increase in nitrosative stress, which was
associated with a significant decrease of both protein SH groups and GSH, with increased levels of GSSG
and nitrosothiols (Calabrese et al., 2002a). Interestingly, treatment of MS patients with acetylcarnitine
resulted in decreased CSF levels of NOreactive metabolites and protein nitration and in a significantly
higher content of both GSH and GSH/GSSG ratio. In addition, urinary nitrites, which were higher in MS
patients than in controls, decreased significantly after treatment with acetylcarnitine. Several studies have
shown the capability of carnitines to interfere with changes in oxidant/antioxidant balance and metabolism
induced by oxidants (Hagen et al., 1998a, b). Although, so far, the exact mechanisms of action of
acetylcarnitine are still unknown, current research points to its ability to enhance neuronal mitochondrial
bioenergetics (Calabrese and Rizza, 1999), which in turn may influence cellular oxidant/antioxidant balance
(Calabrese et al., 2002d). We have recently shown in astrocytes exposed to LPS and INFginduced
nitrosative stress that acetylcarnitine protects against cytokinemediated mitochondrial chain respiratory
complex impairment and the associated increase in protein and lipid peroxidation. The increase in
astroglial antioxidative potential observed after acetylcarnitine treatment involves a secondary line of
antioxidant defenses, represented by stressresponsive genes, such as HO1 and the mitochondrial Hsp60
130 6 Aging and oxidative stress response in the CNS
and SOD (Calabrese et al., 2005). Furthermore, as a brain energy enhancer, acetylcarnitine could improve
survival of damaged neurons (Scapagnini et al., 2002b), and metabolic studies conducted noninvasively in
humans with NMR indicate that acetylcarnitine helps the brain to maintain the constant supply of energy
needed for effective homeostasis.
MS is a progressive inflammatory neurodegenerative disease. However, despite increasing research
efforts and although several explanations have been proposed for destruction of myelin and oligodendro-
cytes in MS, there is still no proven mechanism of injury. The possibility of manipulating these complex
glial cell functions and controlling their pathologic interactions with immune cells probably will illuminate
how myelin damage can be contained and how the injured tissue can be repaired.
enzyme deficiencies and timedependent mitochondrial iron accumulation. In contrast, a model expressing
25%35% of wildtype frataxin levels, by virtue of a (GAA)230 expansion inserted in the first intron of the
mouse gene, has no obvious phenotype (Bradley et al., 2000).
Over the last 5 years, it has become clear that mitochondrial iron accumulation generates oxidative
stress and results in damage to critical biological molecules.
Studies using the budding yeast Saccharomyces cerevisiae have provided a further understanding of the
consequences of frataxin loss (Radisky et al., 1999). Deletion of the yeast frataxin homolog YFH1 results in a
tenfold increase in iron within the mitochondria along with increased ROS production (Lodi et al., 2002).
This leads to loss of mitochondrial function and the appearance of a petite phenotype in nearly all strains
that have been examined (Radisky et al., 1999). Bradley and colleagues (2000) demonstrated an impaired
oxidative phosphorylation system with severe and significant deficiencies of mitochondrial respiratory chain
complexes I and II/III and aconitase activity in cardiac muscle from patients with FA; mtDNA levels were
reduced in FA heart and skeletal muscle and increased iron deposition was present in FA heart, liver, and spleen
in a pattern consistent with a mitochondrial location. In addition, there is the appearance of nDNA damage
(Bradley et al., 2000). Moreover, aconitase deficiency is suggestive that oxidative stress may induce a self
amplifying cycle of oxidative damage associated with mitochondrial dysfunction, which may also contribute to
cellular toxicity. Iron deposition and enzyme deficiencies have been reported in postmortem heart and brain
tissues (Foury and Talibi, 2000) of FA patients. The role of oxidative damage in the pathogenesis of FA is also
supported by the finding that idebenone, an antioxidant similar to ubiquinone, can reduce myocardial
hypertrophy and also decrease markers of oxidative stress in FA patients (Lodi et al., 2002).
Upregulation of protein manganese superoxide dismutase (MnSOD) fails to occur in FA fibroblasts
exposed to iron. This finding, together with the absence of activation of the redoxsensitive factor NFkB,
suggests that NFkBindependent pathway which may not require freeradical signaling is responsible for
the reduced induction of MnSOD. This impairment could constitute both a novel defense mechanism
against ironmediated oxidative stress in cells with mitochondrial iron overload and, conversely, an
alternative source of free radicals that could contribute to the disease pathology. Iron chelator drugs and
antioxidant drugs have therefore been proposed for the treatment of FA. Drugs that reduce oxidative stress
have a limited effect on the progression of the disease pathology, probably because these cannot properly
remove iron accumulation. The potential role of iron chelator analogues (e.g., 2pyridylcarboxaldehyde
isonicotinoyl hydrazone (PCIH)) as agents to remove mitochondrial iron deposits have been recently under
investigation (Jauslin et al., 2003). These ligands have been specifically designed to enter and target
mitochondrial iron pools, which is a property lacking in desferrioxamine, the only chelator in widespread
clinical use. This latter drug may not have any beneficial effect in FA patients, probably because of its
hydrophilicity that prevents mitochondrial access. Indeed, standard chelation regimens will probably not
work in FA, as these patients do not exhibit gross iron loading. Considering that there is no effective
treatment for FA, it is essential that the therapeutic potential of iron chelators focusses on the mitochon-
drial iron pools as their primary target. Remarkably, in an in vitro model of regulated human frataxin
overexpression, it was shown that downregulation of the expression of mitogenactivated protein kinase
kinase 4 was associated with a decreased phosphorylation of cJun Nterminal kinase. In addition, to
understand whether this alteration might result in cell death, the caspase pathway was investigated in FA
cells, revealing in FA patients a significantly higher activation of caspase9 after serum withdrawal compared
with controls. These findings suggest the presence, in FA patient cells, of a hyperactive stresssignaling
pathway. The role of frataxin in FA pathogenesis could be explained, at least in part, by this hyperactivity.
Pilot studies have shown the potential effect of antioxidant therapy using idebenone or coenzyme Q10 with
vitamin E administration and provide a strong rationale for designing larger randomized clinical trials
(Lodi et al., 2003). There is now strong evidence to suggest that mitochondrially localized antioxidant
ameliorates cardiomyopathy in FA patients, as well other lipophilic antioxidants can protect FA cells from
cell death, indicating novel treatment strategies for FA and presumably for other neurodegenerative diseases
with mitochondrial impairment. Antioxidants targeted to mitochondria appear a promising approach to
effectively slow disease progression.
132 6 Aging and oxidative stress response in the CNS
5.8 Ischemia/Reperfusion
Ischemic brain damage is accompanied by an energy deficiency state and selective neuronal loss (Heales
et al., 1999). Under such conditions, there is an increase in the extracellular concentration of glutamate,
which may be neurotoxic due to activation of nNOS. Excess NO generation, causing impairment of energy
metabolism and other metabolic processes, may also downregulate glutamate (NMDA) receptors, thereby
minimizing the effect of glutamate. In addition, NO can cause vasodilation, and hence increase cerebral
blood flow to the infarcted area (Heales et al., 1999). These effects may provide an explanation for the
contradictory results that have been obtained when nonspecific NOS inhibitors have been evaluated in
various models of ischemia. Reperfusion, following ischemia, may exacerbate the generation of oxidizing
species, particularly superoxide. In a model of graded ischemia, loss of brain mitochondrial function, at the
levels of complexes I, II, IIIII, and ATP synthetase, has been reported (Powell and Jackson, 2003).
Reperfusion was associated with restoration of activity of these mitochondrial components, followed
after 2 h by a dramatic loss of complex IV activity (Brooks et al., 2002). The exact mechanism for this
loss of complex IV activity is not known, but could involve the oxidative and/or nitrosative stressmediated
reactions (Brooks et al., 2002). In fact, ischemia is accompanied by the formation of gliotic scar, principally
comprised of reactive astrocytes, which in large amount express iNOS (Heales et al., 1999). Thus, excessive
generation of glialderived ONOO may constitute an important contributing factor to the mitochondrial
damage associated with ischemia. Loss of brain ATP levels and mitochondrial complex IIIII and IV activity
has been demonstrated in a rat model of perinatal asphyxia (Bolanos et al., 1998); in addition, administra-
tion of an NOS inhibitor to the mothers prevented impairment of brain energy metabolism in the hypoxic
pups (Bolanos et al., 1998). Notably, ischemic preconditioning, which has been demonstrated to increase
Hsp expression, preserves brain mitochondrial functions during middle cerebral artery occlusion (Zhang
et al., 2003).
The first half of the twentieth century has seen a rapid increase in the life expectancy of individuals in
industrialized nations because of improved sanitation, public health, housing, nutrition, medical technol-
ogy, and pharmaceuticals. The second half of this century has been characterized by a growing concern with
the challenge produced by the increasing prevalence of old people in the society. Aging is a very common
feature in living organisms and can be described as the total effect of those intrinsic changes in an organism
that adversely affect its vitality and render it more susceptible to the many factors that can cause death.
Typically, mortality rate accelerates with time, but it is not clear whether this effect is the result of external
or internal causes of death. The full extent of aging in a population becomes apparent when most important
external hazards are removed, such as captive or laboratory conditions, and average longevity is usually
greatly extended (Calabrese et al., 2001a). Even if an organism is immortal it has nonzero probability of
dying because of extrinsic causes such as starvation, predation, and accidents. The probability of survival
decreases in the course of life and, since natural selection is effective only through the reproductive output
of individuals, the strength of natural selection decreases with age (Calabrese et al., 2001a).
The first genetic theories on the evolution of aging were proposed in 1957 by Medawar and Williams
almost simultaneously to the mechanistic theories of aging, such as the freeradical and the somatic
mutation theory, suggested by Harman (1956) and Szilard (1977), respectively. A synthesis of evolutionary
and mechanistic theories occurred in 1977 within the framework of the soma theory of aging postulated by
Kirkwood (1977). This theory provides a direct connection between evolutionary and physiological aspects
of aging, by recognizing the primary importance of the allocation of metabolic energy resources between
growth, somatic maintenance, and reproduction. It is suggested that longevity is determined through the
setting of longevity assurance mechanisms so as to provide an optimal compromise between investments in
somatic maintenance (including stress resistance) and in reproduction. As a corollary, increasing mainte-
nance promotes the survival and longevity of the organism only at the expense of significant metabolic
134 6 Aging and oxidative stress response in the CNS
investments that could otherwise be used to accelerate processes such as growth and reproduction. The
disposable soma theory of the evolution of aging also proposes that a high level of accuracy is maintained
in immortal germ line cells, or alternatively, defective germ cells, if any, are eliminated. The evolution of an
increase in longevity in mammals may be due to a concomitant reduction in the rates of growth and
reproduction, the socalled essential life and an increase in the accuracy of synthesis of macromolecules.
The theory can be tested by measuring accuracy in germ line and somatic cells and also by comparing
somatic cells from mammals with different longevities. Notably, the HO gene is evolutionarily different in
birds and mammals, with the biliverdin reductasebilirubin step present in the latter but absent in the
former.. Consistently, the organism sacrifices the potential for indefinite survival in favor of earlier and
more prolific fecundity. From an evolutionary perspective, aging is a nonadaptive phenomenon, since it
limits the reproductive potential of an individual. For this reason aging should be opposed by natural
selection, and hence the argument that it evolved to provide offspring with living space is now receiving
little credence. A clear prediction is that the actual mechanisms of senescence are stochastic, involving most
likely processes such as random accumulation of somatic mutations or oxidative damage to macromole-
cules. In the words of an anonymous poet, we are born as copies, but we die as originals.
It is becoming increasingly clear that genetic factors are prominently involved in aging, the major
lines of empirical evidence being: (1) the life span which in human populations shows significant
heritability; (2) different species have different intrinsic life spans due to genomic differences; (3) human
populations possess inherited progeroid disorders, such as Werners syndrome, a disease characterized by
premature agerelated disorders, including atherosclerosis, type II diabetes, osteoporosis, and cancers;
and (4) clear evidence of genetic effects on life span have been demonstrated in invertebrate model systems,
such as D. melanogaster and C. elegans. In this organism, five different genomic regions appear to be
associated with longevity, as assessed by quantitative genetic analysis (Rothschild and Jazwinski, 1988).
Also, in S. cerevisiae 13 longevity genes have been identified and cloned. Of these 13 genes, 11 have human
homologues (Rothschild and Jazwinski, 1988). At least, three categories of genes are predicted to affect
aging and longevity. They are: (1) genes that regulate levels of somatic maintenance and repair; (2)
pleiotropic genes, whose expression involves tradeoffs between earlylife fitness benefits and latelife fitness
disadvantages, which do not encompass somatic maintenance; and (3) lateacting deleterious mutations
that have escaped elimination as consequence of the decline in the force of natural selection at old ages
(Calabrese et al., 2001a). Efficient functioning of maintenance and repair process seems to be crucial for
both survival and physical quality of life. This complex network of the socalled longevity assurance
processes is composed of several genes, termed vitagenes (> Table 62). The homeodynamic property of
living systems is a function of such a vitagene network. Because aging is characterized by the failure of
homeodynamics, a decreased efficiency and accuracy of the vitagene network can influence gerontogenic
processes. It is not clear how various components of the vitagene network operate and influence each other
in a concordant or a discordant manner. Since aging is characterized by a progressive failure of maintenance
and repair, it is reasoned that genes involved in homeodynamic repair pathways, such as the HO1 or Hsp70
genes, are the most likely candidate vitagenes.
A promising approach for the identification of critical vitagenerelated processes is represented by the
hormesislike positive effect of stress, including regular muscle exercise (Butterfield et al., 2002a, b;
Calabrese et al., 2004b) caloric restriction, which can result in activation of the Hsp signal pathway and,
consequently, in stress tolerance. In particular, there is strong evidence that the HO/CO and biliverdin
bilirubin redox system might work critically as a therapeutic funnel in a number of physiopathological
situations where the sensing of redoxactive events is coupled to acquiring major resistance to the effects of
stressful and pathogenic conditions (> Figure 61). HO1 activity seems to be required for the action of
several other therapeutic molecules. In each case, the expression of HO1 or administration of one of its
metabolic products substitutes for the actions of the other protective molecule (Otterbein et al., 2003a).
Aging and oxidative stress response in the CNS 6 135
. Figure 61
Redox regulation of gene expression involving the vitagene system. Proposed role for the vitagene member
heatshock proteins (Hsps) in modulating cellular redox state and cell stress tolerance. Various proteotoxic (or
genotoxic) conditions cause depletion of free Hsps that lead to activation of stress kinase and proinflammatory
and apoptotic signaling pathways. Hsp70 prevents stressinduced apoptosis by interfering with the SAPK/JNK
signaling and by blocking caspase proteolytic cascade. Nitrosativedependent thiol depletion triggers HO1
induction, and increased HO1 activity is translated into augmented production of carbon monoxide (CO) and
the antioxidant bilirubin. These molecules may counteract increased NOS activity and NOmediated cytotoxici-
ty. In addition, HO1 may directly decrease NO synthase protein levels by degrading the cofactor heme.
Abbreviations: PLA2, phospholipase A2; IL6, interleukin6; AP1, activator protein1; SAPK, stressactivated
protein kinase; JNK, cjun Nterminal kinase; NFkB, nuclear factor kappaB; GSNO: Snitrosoglutathione; HO1,
heme oxygenase1
In many inflammatory situations, the ability of IL10 to suppress TNFa expression in macrophages
requires the presence of HO1 and the generation of CO; HO1 expression or CO administration has the
same effects as IL10 (Lee and Chau, 2002; Soares et al., 2004). In concert with this conceivable possibility,
the protective effect of IL10 in a lethal endotoxic shock mice model is strongly dependent on the expression
of HO1 and the generation of CO (Lee and Chau, 2002; Soares et al., 2004). Moreover, rapamycin appears
not to exert its antiproliferative effects on smooth muscle cells unless HO1 is present (Lee and Chau, 2002;
Akamatsu et al., 2004), and it has been proven that, in order for NO to protect mice livers from hepatitis
induced by TNFa and galactosamine, upregulation of HO1 seems to be essential (Otterbein et al., 2003b).
Also, alcohol has antiinflammatory effects as TNFa is suppressed and IL10 is increased (Otterbein et al.,
2003b; Yamashita et al., 2004). However, protection is lost when HO1 is blocked (Foresti et al., 1997). In
addition, the antiinflammatory effect of 15deoxyD12,14prostaglandin J2 has been shown to require the
activity of HO1 (Lee and Chau, 2002; Otterbein et al., 2003b). Notably, during heat shock, which leads to
upregulation of several Hsps endowed with cytoprotective actions, entire cytoprotection is lost if HO1 is
blocked with SnPPIX. Last, relevant to brain physiopathology, dietary and medicinal phytochemicals that
can inhibit, retard, or reverse the multistage pathogenic events associated with degenerative damage,
136 6 Aging and oxidative stress response in the CNS
particularly polyphenols such as curcumin, caffeic acid, and ferulic acid, all capable of exerting powerful
antiinflammatory actions, have been shown to function by upregulating HO1 (Scapagnini et al., 2002d,
2004; Poon et al., 2004a). The fact that in all these situations specific molecules or biological phenomena
appear to lose most, if not all, of their effects when HO1 is absent represents a compelling evidence that the
HO1 system may represent a final common mediator of many biological events associated to cell stress
response and, as such, working as a critical vitagene, which links redoxdependent pathways of stress
tolerance to a versatile biological program of cell life.
8 Conclusion
Modulation of endogenous cellular defense mechanisms via the stress response signaling represents an
innovative approach to therapeutic intervention in diseases causing tissue damage, such as neurodegenera-
tion (Calabrese et al., 2005; Sultana et al., 2005). Efficient functioning of maintenance and repair processes
seems to be crucial for both survival and physical quality of life. This is accomplished by a complex network
of the socalled longevity assurance processes, which are composed of several genes termed vitagenes.
Consistently, by maintaining or recovering the activity of vitagenes it can be possible to delay the aging
process and decrease the occurrence of agerelated diseases with resulting prolongation of a healthy life span.
As one of the most important neurodegenerative disorders, AD is a progressive disorder with cognitive and
memory decline, speech loss, personality changes, and synapse loss. With the increasingly aging population
of the United States, the number of AD patients is predicted to reach 14 million in the midtwentfirst
century in the absence of effective interventions (Butterfield et al., 2002a). This will pose an immense
economic and personal burden on the people of this country. Similar considerations apply worldwide, except
in subSaharan Africa where HIV infection rates seem to be leading to decreased incidence of AD (Butterfield
et al., 2002b). There is now strong evidence to suggest that factors such as oxidative stress and disturbed
protein metabolism and their interaction in a vicious cycle are central to AD pathogenesis. Brainaccessible
antioxidants, potentially, may provide the means of implementing this therapeutic strategy of delaying the
onset of AD, and more in general all degenerative diseases associated with oxidative stress. As one potentially
successful approach, potentiation of endogenous secondary antioxidants systems can be achieved by
interventions which target the HO1/CO and/or Hsp70 systems. In this chapter, the importance of the stress
response signaling and, in particular, the central role of HO1 together with the redoxdependent mechan-
isms involved in cytoprotection are outlined. The beneficial effects of HO1 induction result from heme
degradation and cytoprotective regulatory functions of biliverdin/bilirubin redox cycling. Thus, HO1 can
amplify intracellular cytoprotective mechanisms against a variety of insults. Consequently, induction of HO
1 by increasing CO and/or biliverdin availability can be of clinical relevance.
CO has been studied for >100 years and, until the last few years, has been touted as a molecule to avoid,
owing to its toxic effects exerted mostly on hemoglobin and cytochrome oxidase functions (Otterbein,
2002). However, these toxic effects are seen at concentrations of CO well above concentrations used
experimentally. Beneficial effects are obtained with relative low doses of CO (250 ppm for one to few
hours) in rodents (Otterbein et al., 2003b). Carboxyhemoglobin levels generated in such a model are not
too high from those of heavy smokers. If this beneficial effect is confirmed also in humans, limited exposure
of patients to CO might be considered as therapy for various syndromes, particularly to prevent restenosis
after angioplasty or treatment of an organ donor and/or the organ to suppress ischemiareperfusion injury
and to prolong allograft survival. Very importantly, HO1 and CO can suppress the development of
atherosclerotic lesions associated with chronic rejection of transplanted organs (Otterbein et al., 2003b;
Akamatsu et al., 2004). Interestingly, the recently discovered CORMS appears to afford similar protective
action, thereby providing an alternative therapeutic approach for those pathophysiological conditions
where CO administration may prove to be beneficial (Motterlini et al., 2002b, 2003). Furthermore,
administration of biliverdin or bilirubin after the first few weeks of life is proven not to have toxicity and
doses as much as 2.5 mg/dl used in experimental paradigms are only slightly above normal levels, yet
endowed with cytoprotective effects (Yamashita et al., 2004). Although clinical application of the HO
Aging and oxidative stress response in the CNS 6 137
system should be fully considered, a better understanding of how HO mediates its action will guide
therapeutic strategies to enhance or suppress HO effects. Remarkably, the recent envisioned role of
Hsp70 as a vehicle for intracytoplasmic and intranuclear delivery of fusion proteins or DNA to modulate
gene expression (Wheeler et al., 2003), along with the evidence that binding of HO protein to HO1 DNA
modifies HO expression via nonenzymatic signaling events (Weng et al., 2003) associated to CO and P38
dependent induction of Hsp70, opens intriguing perspectives, as it is possible to speculate that synergy
between these two systems might impact cell proliferation and apoptotic processes during oxidative stress,
hence contributing to programmed cell life or programmed cell death (> Figure 61), depending on the
relative extent of activation.
Presented here is strong evidence that a cross talk between stress response genes is critical for cell stress
tolerance, highlighting a compelling reason for a renewed effort to understand the central role of this most
extraordinary defense system in biology and medicine. All of the above evidence supports also the notion
that stimulation of various maintenance and repair pathways through exogenous intervention, such as mild
stress or nutritional compounds targeting the heatshock signal pathway, may have biological significance as
a novel approach to delay the onset of various ageassociated alterations in cells, tissues, and organisms
(Poon et al., 2004a, b). Hence, by maintaining or recovering the activity of vitagenes (Calabrese et al., 2001a,
2004b, 2005) it can be possible to delay the aging process and decrease the occurrence of agerelated diseases
with resulting prolongation of a healthy life span.
Acknowledgments
This work was supported in part by grants from the Wellcome Trust, MIURCluster Biomedicine, and FIRB
RBNE01ZK8F, and by grants from the National Institute of Health (D.A.B.). The authors acknowledge
helpful discussions with John Clark (Institute of Neurology, UCL, London, UK) and with Roberto
Motterlini (Northwick Park Institute for Medical Research, Harrow, UK); Enrico Rizzarelli and Eduardo
Puleo (Department of Chemistry, University of Catania).
References
Akamatsu Y, Haga M, Tyagi S, Yamashita K, GracaSouza AV, Babcock M, de Silva D, Oaks R, DavisKaplan S, Jiralerspong S,
et al. 2004. Heme oxygenase1derived carbon monoxide et al. 1997. Regulation of mitochondrial iron accumulation
protects hearts from transplant associated ischemia reper- by Yfh1p, a putative homolog of frataxin. Science 276:
fusion injury. FASEB J 18: 771-772. 1709-1712.
Alam J. 2002. Heme oxygenase1: Past, present, and future. Bagasra O, Michaels FH, Zheng YM, Bobroski LE, Spitsin SV,
Antioxid Redox Signal 4: 559-562. et al. 1995. Activation of the inducible form of nitric oxide
Alam J, Cook JL. 2003. Transcriptional regulation of the heme synthase in the brains of patients with multiple sclerosis.
oxygenase1 gene via the stress response element pathway. Proc Natl Acad Sci USA 92: 12041-12045.
Curr Pharm Des 9: 2499-2511. Balogun E, Hoque M, Gong P, Killeen E, Green CJ, et al. 2003.
Anisimov VN, Alimova IN, Baturin DA, Popovich IG, Curcumin activates the haem oxygenase1 gene via regula-
Zabezhinski MA, et al. 2003. Dosedependent effect of tion of Nrf2 and the antioxidantresponsive element. Bio-
melatonin on life span and spontaneous tumor incidence chem J 371: 887-895.
in female SHR mice. Exp Gerontol 38: 449-461. Beal MF. 1998. Excitotoxicity and nitric oxide in Parkinsons
Aquino DA, Capello E, Weisstein J, Sanders V, Lopez C, et al. disease pathogenesis. Ann Neurol 44: S110-S114.
1977. Multiple sclerosis: Altered expression of 70and 27kDa Beal MF. 2003. Mitochondria, oxidative damage, and inflam-
heat shock proteins in lesions and myelin. J Neuropathol Exp mation in Parkinsons disease. Ann NYAcad Sci 991: 120-131.
Neurol 56: 664-672. Beal MF, Ferrante RJ, Browne SE, Matthews RT, Kowall NW,
Aviv A, Levy D, Mangel M. 2003. Growth, telomere dynamics et al. 1997. Increased 3nitrotyrosine in both sporadic
and successful and unsuccessful human aging. Mech Age- and familial amyotrophic lateral sclerosis. Ann Neurol 42:
ing Dev 124: 829-837. 644-654.
138 6 Aging and oxidative stress response in the CNS
Bechtold DA, Brown IR. 2003. Induction of Hsp27 and Butterfield DA, Castegna A, Drake J, Scapagnini G, Calabrese
Hsp32 stress proteins and vimentin in glial cells of the rat V. 2002a. Vitamin E and neurodegenerative disorders asso-
hippocampus following hyperthermia. Neurochem Res 28: ciated with oxidative stress. Nutr Neurosci 5: 229-239.
1163-1173. Butterfield DA, Castagna A, Pocernich CB, Drake J, Scapag-
Beckman JS, Carlson M, Smith CD, Koppenol WH. 1993. nini G, et al. 2002b. Nutritional approaches to combat
ALS, SOD and peroxynitrite. Nature 364: 584-586. oxidative stress in Alzheimers disease. J Nutr Biochem 13:
Beckman KB, Ames BN. 1998. Mitochondrial aging: Open 444-461.
questions. Ann NY Acad Sci 854: 118-127. Butterfield DA, Lauderback CM 2002. Lipid peroxidation and
Beschorner R, Adjodah D, Schwab JM, Mittelbronn M, Pedal protein oxidation in Alzheimers disease brain: Potential
I, et al. 2000. Longterm expression of heme oxygenase1 causes and consequences involving amyloid beta-peptide-
(HO1, HSP32) following focal cerebral infarctions and associated free radical oxidative stress. Free Radic Biol Med
traumatic brain injury in humans. Acta Neuropathol 100: 32: 1050-1060.
377-384. Butterfield DA, Howard BJ, Yatin S, Allen KL, Carney JM.
Biesalski HK. 2002. Free radical theory of aging. Curr Opin 1997. Free radical oxidation of brain proteins in accel-
Clin Nutr Metab Care 5: 5-10. erated senescence and its modulation by Ntertbutyla
Bohr VA, Dianov GL. 1999. Oxidative DNA damage proces- phenylnitrone. Proc Natl Acad Sci USA 94: 674-678.
sing in nuclear and mitochondrial DNA. Biochimie 81: Butterfield DA, Koppal T, Subramaniam R, Yatin S. 1999.
155-160. Vitamin E as an antioxidant/free radical scavenger against
Bolanos JP, Almeida A, Medina JM. 1998. Nitric oxide med- amyloid bpeptideinduced oxidative stress in neocortical
iates brain mitochondrial damage during perinatal anoxia. synaptosomal membranes and hippocampal neurons in
Brain Res 787: 117-122. culture: Insights into Alzheimers disease. Rev Neurosci
Bradley JL, Blake JC, Chamberlain S, Thomas PK, Cooper JM, 10: 141-149.
et al. 2000. Clinical, biochemical and molecular genetic Calabrese V, Rizza V. 1999. Formation of propionate after
correlations in Friedreichs ataxia. Hum Mol Genet 9: shortterm ethanol treatment and its interaction with the
275-282. carnitine pool in rat. Alcohol 19: 169-176.
Bredt DS. 1999. Endogenous nitric oxide synthesis: Calabrese V, Bates TE, Giuffrida Stella AM. 2000a. NO
Biological functions and pathophysiology. Free Radic Res synthase and NOdependent signal pathways in brain
31: 577-596. aging and neurodegenerative disorders: The role of
Brooks KJ, Hargreaves I, Bhakoo K, Sellwood M, O Brien F, oxidant/antioxidant balance. Neurochem Res 65:
et al. 2002. Delayed hypothermia prevents decreases in 1315-1341.
Nacetylaspartate and reduced glutathione in the cerebral Calabrese V, Copani A, Testa D, Ravagna A, Spadaro F, et al.
cortex of the neonatal pig following transient hypoxia 2000b. Nitric oxide synthase induction in astroglial cell
ischaemia. Neurochem Res 27: 1599-1604. cultures: Effect on heat shock protein 70 synthesis and
Browne SE. 1997. Mitochondrial dysfunction and oxidative oxidant/antioxidant balance. J Neurosc Res 60: 613-622.
damage in Huntingtons disease. Mitochondria and Free Calabrese V, Testa D, Ravagna A, Bates TE, Giuffrida Stella
Radicals in Neurodegenerative Diseases. Flint Beal M, AM. 2000c. Hsp70 induction in the brain following ethanol
Howell N, BodisWollner I, editors. New York: WileyLiss. administration in the rat: Regulation by glutathione redox
Brwon RH. 1994. Clinical implications of basic research: A state. Biochem Biophys Res Commun 269: 397-400.
transgenicmouse model of amyotrophic lateral sclerosis. N Calabrese V, Bella R, Testa D, Spadaro F, Scorfani A, et al.
Engl J Med 331: 1091-1092. 1998. Increased cerebrospinal fluid and plasma levels of
Butterfield DA. 2002. Amyloid bpeptide142induced oxida- ultraweak chemiluminescence are associated with changes
tive stress and neurotoxicity: Implications for neurodegen- in the thiol pool and lipidsoluble fluorescence in multiple
eration in Alzheimers disease brain. A review. Free Radic sclerosis: The pathogenic role of oxidative stress. Drugs Exp
Res 36: 1307-1313. Clin Res 24: 125-131.
Butterfield DA. 2004. Proteomics: A new approach to investi- Calabrese V, Raffaele R, Casentino E, Rizza V. 1994. Changes
gate oxidative stress in Alzheimers disease brain. Brain Res in cerebrospinal fluid levels of malonaldehyde and gluta-
1000: 1-7. thione reductase activity in multiple sclerosis. J Clin Phar-
Butterfield DA, Kanski J. 2002. Methionine residue 35 is macol Res 4: 119-123.
critical for the oxidative stress and neurotoxic properties Calabrese V, Ravagna A, Colombrita C, Guagliano E, Scapag-
of Alzheimers amyloidpeptide 142. Review Peptides 23: nini G, et al. 2005. Acetylcarnitine induces heme oxygenase
1299-1309. in rat astrocytes and protects against oxidative stress:
Aging and oxidative stress response in the CNS 6 139
Involvement of the transcription factor Nrf2. J Neurosci aging and neurodegenerative disorders: The role of vita-
Res 79: 509-521. genes. In vivo 18: 23-45.
Calabrese V, Scapagnini G, Colombrita C, Ravagna A, Pennisi G, Calabrese V, Giuffrida Stella, AM, Butterfield DA, Scapag-
et al. 2003a. Redox regulation of heat shock protein nini G. 2004c. Redox regulation in neurodegeneration
expression in aging and neurodegenerative disorders asso- and longevity: Role of the heme oxygenase and Hsp70
ciated with oxidative stress: A nutritional approach. Amino systems in brain stress tolerance. Antiox Redox Signal 5:
acids 27: 15-23. 895-913.
Calabrese V, Scapagnini G, Ravagna A, Bella R, Butterfield Campuzano V, Contermini L, Molto MD, Pianese L, Cassee M,
DA, et al. 2003b. Disruption of thiol homeostasis and et al. 1996. Friedreichs ataxia: Autosomal recessive disease
nitrosative stress in the cerebrospinal fluid of patients caused by an intronic GAA triplet repeat expansion. Science
with active multiple sclerosis: Evidence for a protective 271: 1423-1427.
role of acetylcarnitine. Neurochem Res 28: 1321-1328. Carney JM, StarkeReed PE, Oliver CN, Landum RW, Cheng
Calabrese V, Butterfield DA, Giuffrida Stella AM. 2003c. MS, et al. 1991. Reversal of agerelated increase in brain
Nutritional antioxidants and the hem oxygenase pathway protein oxidation, decrease in enzyme activity, and loss in
of stress tolerance: Novel targets for neuroprotection in temporal and spatial memory by chronic administration of
Alzheimers disease. It J Biochem 52: 177-181. the spintrapping compound Ntertbutylaphenylnitrone.
Calabrese V, Scapagnini G, Giuffrida Stella AM, Bates TE, Proc Natl Acad Sci USA 88: 3633-3636.
Clark JB. 2001a. Mitochondrial involvement in brain func- Castegna A, Lauderback CM, MohmmadAbdul H,
tion and dysfunction: Relevance to aging, neurodegenera- Butterfield DA. 2004. Modulation of phospholipid asym-
tive disorders and longevity. Neurochem Res 26: 739-764. metry in synaptosomal membranes by the lipid peroxida-
Calabrese V, Scapagnini G, Catalano C, Bates TE, Dirotta F, tion products, 4hydroxynonenal and acrolein: Implications
et al. 2001b. Induction of heat shock protein synthesis in for Alzheimers disease. Brain Res 1004: 193-197.
human skin fibroblasts in response to oxidative stress: Catania MV, Giuffrida R, Seminara G, Barbagallo G, Aronica E,
Regulation by a natural antioxidant from rosemary extract. et al. 2003. Upregulation of neuronal nitric oxide synthase
Int J Tissue React 23: 121-128. in vitro stellate astrocytes and in vivo reactive astrocytes
Calabrese V, Scapagnini G, Ravagna A, Bella R, Foresti R, et al. after electrically induced status epilepticus. Neurochem Res
2002a. Nitric oxide synthase is present in the cerebrospinal 28: 607-615.
fluid of patients with active multiple sclerosis and is asso- Chakrabarty A, Emerson MR, Le Vine SM. 2003. Heme oxy-
ciated with increases in CSF protein nitrotyrosine, Snitro- genase1 in SJL mice with experimental allergic encephalo-
sothiols and with changes in glutathione levels. J Neurosci myelitis. Mult Scler 9: 372-381.
Res 70: 580-587. Chen K, Gunter K, Maines MD. 2000. Neurons overexpressing
Calabrese V, Scapagnini G, Ravagna A, Fardello RG, Giuffrida heme oxygenase1 resist oxidative stressmediated cell
Stella AM, et al. 2002b. Regional distribution of heme death. J Neurochem 75: 304-312.
oxygenase, HSP70, and glutathione in brain: Relevance Chou SM, Wang HS, Komai K. 1996. Colocalisation of NOS
for endogenous oxidant/antioxidant balance and stress and SOD1 in neurofilament accumulation within motor
tolerance. J Neurosci Res 68: 65-75. neurones of amyotrophic lateral sclerosisan immunohis-
Calabrese V, Scapagnini G, Ravagna A, Giuffrida Stella AM, tochemical study. J Chem Neuroanat 10: 249-258.
Butterfield DA. 2002c. Molecular chaperones and their Christen Y. 2000. Oxidative stress and Alzheimer disease. Am J
roles in neural cell differentiation. Dev Neurosci 24: 1-13. Clin Nutr 71: 621S-629S.
Calabrese V, Scapagnini G, Latteri S, Colombrita C, Ravagna A, Coceani F, Kelsey L, Seidlitz E, Marks GS, McLaughlin BE,
et al. 2002d. Longterm ethanol administration enhances et al. 1997. Carbon monoxide formation in the ductus
agedependent modulation of redox state in different brain arteriosus in the lamb: Implications for the regulation of
regions in the rat: Protection by acetyl carnitine. Int J muscle tone. Br J Pharmacol 120: 599-608.
Tissue React 24: 97-104. Collard JF, Cote F, Julien JP. 1995. Defective axonal transport
Calabrese V, Scapagnini G, Ravagna A, Colombrita C, in a transgenic mouse model of amyotrophic lateral sclero-
Spadaro F, et al. 2004a. Increased expression of heat shock sis. Nature 375: 61-64.
proteins in rat brain during aging: Relationship with mito- Colombrita C, Calabrese V, Giuffrida Stella AM, Mattei F,
chondrial function and glutathione redox state. Mech Age Alkon DL, et al. 2003. Regional rat brain distribution of
Dev 125: 325-335. heme oxygenase1 and manganese superoxide dismutase
Calabrese V, BoydKimball D, Scapagnini G, Butterfield DA. mRNA: Relevance of redox homeostasis in the aging pro-
2004b. Nitric oxide and cellular stress response in brain cesses. Exp Biol Med 228: 517-524.
140 6 Aging and oxidative stress response in the CNS
Corregidor C, De Pasamonte J. 1996. Cerebrospinal fluid Dringen R, Hirrlinger J. 2003. Glutathione pathways in the
nitrate levels in patients with Alzheimers disease. Acta brain. Biol Chem 384: 505-516.
Neurol Scand 94: 411-414. Durr A, Cossee M, Agid Y, Campuzano V, Mignard C, et al.
Cottrell DA, Turnbull DM. 2000. Mitochondria and ageing. 1996. Clinical and genetic abnormalities in patients with
Curr Opin Clin Nutr Metab Care 3: 473-478. Friedreichs ataxia. N Engl J Med 335: 1169-1175.
Cottrell DA, Blakely EL, Johnson MA, Ince PG, Borthwick El Khoury J, Hickman SE, Thomas CA, Loike JD, Silverstein SC.
GM, et al. 2001. Cytochrome c oxidase deficient cells accu- 1998. Microglia, scavenger receptors, and the pathogenesis
mulate in the hippocampus and choroid plexus with age. of Alzheimers disease. Neurobiol Aging 19: S81-S84.
Neurobiol Aging 22: 265-272. Ewing JF, Maines MD. 1992. In situ hybridization and immu-
Cross AH, Manning PT, Keeling RM, Schmidt RE, Misko TP. nohistochemical localization of heme oxygenase2 mRNA
1998. Peroxynitrite formation within the central nervous and protein in normal rat brain: Differential distribution of
system in active multiple sclerosis. J Neuroimmunol 88: isozyme 1 and 2. Mol Cell Neurosci 3: 559-570.
45-56. Ferri P, Cecchini T, Ciarloni S, Ambrogini P, Cuppini R, et al.
Cross AH, Manning PT, Stern MK, Misko TP. 1997. Evidence 2003. Vitamin E affects cell death in adult rat dentate gyrus.
for the production of peroxynitrite in inflammatory CNS J Neurocytol 32: 1155-1164.
demyelination. J Neuroimmunol 80: 121-130. Fink SL, Chang LK, Ho DY, Sapolsky RM. 1997. Defective
Dawson TM, Dawson VL. 2002. Neuroprotective and neuror- herpes simplex virus vectors expressing the rat brain stress
estorative strategies for Parkinsons disease. Nat Neurosci 5: inducible heat shock protein 72 protect cultured neurons
1058-1061. from severe heat shock. J Neurochem 68: 961-969.
Dawson VL, Dawson TM. 1995. Physiological and toxicologi- Floyd RA, Hensley K. 2002. Oxidative stress in brain aging.
cal actions of nitric oxide in the central nervous system. Implications for therapeutics of neurodegenerative dis-
Adv Pharmacol 34: 323-342. eases. Neurobiol Aging 23: 795-807.
Demple B, Harrison L. 1994. Repair of oxidative damage to Foresti R, Motterlini R. 1999. The heme oxygenase pathway
DNA: Enzymology and biology. Annu Rev Biochem 63: and its interaction with nitric oxide in the control of cellu-
915-948. lar homeostasis. Free Rad Res 31: 459-475.
Dexter DT, Carayon A, JavoyAgid F, Agid Y, Wells FR, et al. Foresti R, Clark JE, Green CJ, Motterlini R. 1997. Thiol
1991. Alterations in the levels of iron ferritin and other compounds interact with nitric oxide in regulating heme
trace metals in Parkinsons diseases affecting the basal ganglia. oxygenase1 induction in endothelial cells. Involvement of
Brain 114: 1953-1975. superoxide and peroxynitrite anions. J Biol Chem 272:
DinkovaKostova AT, Massiah MA, Bozak RE, Hicks RJ, Talalay 18411-18417.
P. 2001. Potency of Michael reaction acceptors as inducers Foresti R, Goatly H, Green CJ, Motterlini R. 2001. Role of
of enzymes that protect against carcinogenesis depends on heme oxygenase1 in hypoxiareoxygenation: Requirement
their reactivity with sulfhydryl groups. Proc Natl Acad Sci of substrate heme to promote cardioprotection. Am J
USA 98: 3404-3409. Physiol Heart Circ Physiol 281: H1976-H1984.
Dore S. 2002. Decreased activity of the antioxidant heme oxy- Forster MJ, Dubey A, Dawson KM, Stutts WA, Lal H, et al.
genase enzyme: Implications in ischemia and in Alzheimers 1996. Agerelated losses of cognitive function and motor
disease. Free Radic Biol Med 32: 1276-1282. skills in mice are associated with oxidative protein damage
Dore S, Goto S, Sampei K, Blackshaw S, Hester LD, et al. 2000. in the brain. Proc Natl Acad Sci USA 93: 4765-4769.
Heme oxygenase2 acts to prevent neuronal death in brain Foury F, Talibi D. 2000. Mitochondrial control of iron
cultures and following transient cerebral ischemia. Neuro- homeostasis. A genome wide analysis of gene expression
science 99: 587-592. in a yeast frataxin deficient mutant. J Biol Chem 276:
Drake J, Link CD, Butterfield DA. 2003a. Oxidative stress 7762-7768.
precedes fibrillar deposition of Alzheimers disease amyloid FranceLanord V, Brugg B, Michel PP, Agid Y, Ruberg M.
bpeptide142 in a transgenic Caenorhabditis elegans model. 1997. Mitochondrial free radical signal in ceramide
Neurobiol Aging 24: 415-420. dependent apoptosis: A putative mechanism for neuronal
Drake J, Sultana R, Aksenova M, Calabrese V, Butterfield DA. death in Parkinsons disease. J Neurochem 69: 1612-1621.
2003b. Elevation of mitochondrial glutathione by gglutamyl- Fries JF. 2002. Successful agingan emerging paradigm of
cysteine ethyl ester protects mitochondria against peroxyni- gerontology. Clin Geriatr Med 18: 371-382.
triteinduced oxidative stress. J Neurosci Res 74: 917-927. Friguet B. 2002. Protein repair and degradation during aging.
Drake J, Petroze R, Castegna A, Ding Q, Keller JN, et al. 2004. Scientific World Journal 2: 248-254.
4Hydroxynonenal oxidatively modifies histones: Implica- Fujita K, Yamaucji M, Shibayama M, Ando M, Honda M,
tions for Alzheimers disease. Neurosci Lett 356: 155-158. et al. 1996. Decreased cytochrome c oxidase activity but
Aging and oxidative stress response in the CNS 6 141
unchanged superoxide dismutase and glutathione peroxi- restores mitochondrial function and ambulatory activity.
dase activities in the spinal cords of patients with amyo- Proc Natl Acad Sci USA 95: 9562-9566.
trophic lateral sclerosis. J Neurosci Res 45: 276-281. Halliwell B. 1999. Oxygen and nitrogen are procarcinogens.
Ganguli M, Chandra V, Kamboh MI, Johnston JM, Dodge Damage to DNA by reactive oxygen, chlorine and nitrogen
HH, et al. 2000. Apolipoprotein E polymorphism and species: Measurement, mechanism and the effects of nutri-
Alzheimer disease: The IndoUS CrossNational Dementia tion. Mutat Res 443: 37-52.
Study. Arch Neurol 57: 824-830. Halliwell B. 2001. Role of free radicals in the neurodegenera-
Garthwaite G, Goodwin DA, Batchelor AM, Leeming K, tive diseases: Therapeutic implications for antioxidant
Garthwaite J. 2002. Nitric oxide toxicity in CNS white treatment. Drugs Aging 18: 685-716.
matter: An in vitro study using rat optic nerve. Neurosci- Halliwell B. 2002. Hypothesis: Proteasomal dysfunction: A
ence 109: 145-155. primary event in neurogeneration that leads to nitrative
Gaubatz JW, Tan BH. 1993. Agerelated studies on the remov- and oxidative stress and subsequent cell death. Ann NY
al of 7methylguanine from DNA of mouse kidney tissue Acad Sci 962: 182-194.
following NmethylNnitrosourea treatment. Mutat Res Harding AE. 1981. Friedreichs ataxia: A clinical and genetic
295: 81-91. study of 90 families with an analysis of early diagnostic
Gaubatz JW, Tan BH. 1994. Aging affects the levels of DNA criteria and intrafamilial clustering of clinical features.
damage in postmitotic cells. Ann NY Acad Sci 719: 97-107. Brain 104: 598-620.
Gegg ME, Beltran B, SalasPino S, Bolanos JP, Clark JB, et al. Harman D. 1981. The aging process. Proc Natl Acad Sci USA
2003. Differential effect of nitric oxide on glutathione me- 78: 7124-7128.
tabolism and mitochondrial function in astrocytes and Harman DA. 1956. A theory based on free radical and radia-
neurones: Implications for neuroprotection/neurodegen- tion chemistry. J Gerontol 11: 298-300.
eration? J Neurochem 86: 228-237. Harris N, Costa V, Mac Lean M, Mollapour M, Moradas
Genova ML, Pich MM, Biondi A, Pernacchia A, Falasca A, Ferreira P, et al. 2003. MnSOD overexpression extends
et al. 2003. Mitochondrial production of oxygen radical the yeast chronological (G0) life span but acts independent-
species and the role of coenzyme Q as an antioxidant. Exp ly of Sir2p histone deacetylase to shorten the replicative life
Biol Med 228: 506-513. span of dividing cells. Free Radic Biol Med 34: 1599-1606.
Genova ML, Pich MM, Pernacchia A, Bianchi C, Biondi A, Hausladen A, Privalle CT, Keng T, Deangelo J, Stamler JS.
et al. 2004. The mitochondrial production of reactive 1996. Nitrosative stress: Activation of the transcription
oxygen species in relation to aging and pathology. Ann NY factor OxyR. Cell 86: 719-729.
Acad Sci 1011: 86-100. Heales SJR, Bolanos JP, Stewart VC, Brookes PS, Land JM,
Giovannoni G. 1998. Cerebrospinal fluid and serum nitric et al. 1999. Nitric oxide, mitochondria and neurological
oxide metabolites in patients with multiple sclerosis. Mult disease. Biochim Biophys Acta 1410: 215-228.
Scler 4: 27-30. HillKapturczak N, Sikorski EM, Voakes C, Garcia J, Nick HS,
Glaum SR, Miller RJ. 1993. Zinc protoporphyrinIX blocks et al. 2003. An internal enhancer regulates heme and
the effects of metabotropic glutamate receptor activation cadmiummediated induction of human heme oxygenase1.
in the rat nucleus tractus solitarii. Mol Pharmacol 43: Am J Physiol Renal Physiol 285: F515-F523.
965-969. Hon T, Hach A, Lee HC, Cheng T, Zhang L. 2000. Functional
Gong P, Stewart D, Hu B, Vinson C, Alam J. 2002. Multiple analysis of heme regulatory elements of the transcriptional
basicleucine zipper proteins regulate induction of the activator Hap1. Biochem Biophys Res Commun 273: 584-591.
mouse heme oxygenase1 gene by arsenite. Arch Biochem Hu JR, Ferreira A, Van Eldik LJ. 1997. S100 b induces neuro-
Biophys 405: 265-274. nal cell death through nitric oxide release from astrocytes.
Graser T, Vedernikov YP, Li DS. 1990. Study on the mecha- J Neurochem 69: 2294-2301.
nism of carbon monoxide induced endotheliumindepen- Hunot S, Brugg B, Richard D, Michel PP, Muriel MP, et al.
dent relaxation in porcine coronary artery and vein. 1997. Nuclear translocation of NFkB is increased in dopa-
Biomed Biochim Acta 49: 293-296. minergic neurons of patients with Parkinsons disease. Proc
Grehan N. 1894. Le Gas du Sang. Paris. Natl Acad Sci USA 94: 7531-7536.
Hagen TM, Wehr CM, Ames BN. 1998a. Mitochondrial decay Hursting SD, Lavigne JA, Berrigan D, Perkins SN, Barrett JC.
in aging. Reversal through supplementation of acetyl 2003. Calorie restriction, aging, and cancer prevention:
Lcarnitine and Ntertbutylaphenylnitrone. Ann NY Mechanisms of action and applicability to humans. Annu
Acad Sci 854: 214-223. Rev Med 54: 131-152.
Hagen TM, Ingersoll RT, Wehr CM, Lykkesfeldt J, Vinarsky V, Hyun DH, Lee MH, Halliwell B, Jenner P. 2002. Proteasomal
et al. 1998b. AcetylLcarnitine fed to old rats partially dysfunction induced by 4hydroxy2,3transnonenal, an
142 6 Aging and oxidative stress response in the CNS
endproduct of lipid peroxidation: A mechanism contri- Koppal T, Subramaniam R, Drake J, Prasad RP, Dhillon H,
buting to neurodegeneration? J Neurochem 83: 360-370. et al. 1998. Vitamin E protects against Alzheimers amyloid
Hyun DH, Lee M, Halliwell B, Jenner P. 2003. Proteasomal peptide2535induced changes in neocortical synaptosomal
inhibition causes the formation of protein aggregates con- membrane lipid structure and composition. Brain Res 786:
taining a wide range of proteins, including nitrated pro- 270-273.
teins. J Neurochem 86: 363-373. Lauderback CM, Hackett JM, Keller JN, Varadarajan S,
Ignarro LJ. 2002. Nitric oxide as a unique signaling molecule Szweda L, et al. 2001. Vulnerability of synaptosomes from
in the vascular system: A historical overview. J Physiol apoE knockout mice to structural and oxidative modifica-
Pharmacol 53: 503-514. tions induced by Ab140: Implications for Alzheimers dis-
Izaki K, Kinouchim H, Watanabem K, Owada Y, Okubo A, ease. Biochemistry 40: 2548-2554.
et al. 2001. Induction of mitochondrial heat shock protein Lauderback CM, Kanski J, Hackett JM, Maeda N, Kindy MS,
60 and 10 mRNAs following transient focal cerebral ische- et al. 2002. Apolipoprotein E modulates Alzheimers Ab142
mia in the rat. Brain Res Mol Brain Res 88: 14-25. induced oxidative damage to synaptosomes in an allele
Jauslin ML, Meier T, Smith RA, Murphy MP. 2003. Mitochon- specific manner. Brain Research 924: 90-97.
driatargeted antioxidants protect Friedreich ataxia fibro- Lee TS, Chau LY. 2002. Heme oxygenase1 mediates the anti
blasts from endogenous oxidative stress more effectively inflammatory effect of interleukin10 in mice. Nat Med 8:
than untargeted antioxidants. FASEB J 17: 1972-1974. 240-248.
Jenkins B, Koroshetz W, Beal MF, Rosen B. 1993. Evidence for Li H, Dryhurst G. 1997. Irreversible inhibition of mitochon-
an energy metabolism defect in Huntingtons disease using drial complex I by 2aminoethyl3,4dyhydro5hydroxy2
localized proton spectroscopy. Neurology 43: 2689-2695. benzothiazine3carboxylic acid (DHBT): A putative nigral
Kalyuzhny AE. 2002. Simultaneous in situ detection of DNA endotoxin of relevance to Parkinsons disease. J Neurochem
fragmentation and RNA/DNA oxidative damage using 69: 1530-1541.
TUNEL assay and immunohistochemical labeling for Lim GP, Chu T, Yang F, Beech W, Frautschy SA, et al. 2001.
8hydroxy20 deoxyguanosine (8OHdG). Methods Mol The curry spice curcumin reduces oxidative damage and
Biol 203: 219-234. amyloid pathology in an Alzheimer transgenic mouse. J
Kanakiriya SK, Croatt AJ, Haggard JJ, Ingelfinger JR, Tang SS, Neurosci 21: 8370-8377.
et al. 2003. Heme: A novel inducer of MCP1 through Lindsey JW, Kerman RH, Wolinsky JS. 1997. T cellT cell
HOdependent and HOindependent mechanisms. Am J activation in multiple sclerosis. Mult Scler 3: 238-242.
Physiol Renal Physiol 284: F546-F554. Liu N, Wang X, McCoubrey WK, Maines MD. 2000. Devel-
Kaur H, Hughes MN, Green CJ, Naughton P, Foresti R, et al. opmentally regulated expression of two transcripts for
2003. Interaction of bilirubin and biliverdin with reactive heme oxygenase2 with a first exon unique to rat testis:
nitrogen species. FEBS Lett 543: 113-119. Control by corticosterone of the oxygenase protein expres-
Kazazian HH. 2000. Genetics. L1 retrotransposons shape the sion. Gene 241: 175-183.
mammalian genome. Science 289: 1152-1153. Liu Y, Zhu B, Luo L, Li P, Paty DW, et al. 2001. Heme
Kelly S, Zhang ZJ, Zhao H, Xu L, Giffard RG, et al. 2002. Gene oxygenase1 plays an important protective role in experi-
transfer of HSP72 protects cornu ammonis 1 region of the mental autoimmune encephalomyelitis. Neuroreport 12:
hippocampus neurons from global ischemia: Influence of 1841-1845.
Bcl2. Ann Neurol 52: 160-167. Lodi R, Rajagopalan B, Bradley JL, Taylor DJ, Crilley JG, et al.
Keyse SM, Tyrrell RM. 1989. Heme oxygenase is the major 2002. Mitochondrial dysfunction in Friedreichs ataxia:
32kDa stress protein induced in human skin fibroblasts by From pathogenesis to treatment perspectives. Free Radic
UVA radiation, hydrogen peroxide and sodium arsenite. Res 36: 461-466.
Proc Natl Acad Sci USA 86: 99-103. Lodi R, Rajagopalan B, Schapira AHV, Hart PE, Crilley JG,
Kimpara T, Takeda A, Yamaguchi T, Arai H, Okita N, et al. et al. 2003. Coenzyme Q10 and vitamin E treatment
2000. Increased bilirubins and their derivatives in cerebro- of patients with Friedreich ataxia. A 4 year clinical and
31
spinal fluid in Alzheimers disease. Neurobiol Aging 21: PMRS follow up study. International Society for Mag-
551-554. netic Resonance in Medicine. 11th Scientific Meeting and
Kirkwood TB. 1977. Evolution of ageing. Nature 270: 301-304. Exhibition. Toronto, Ontario, Canada. 1016 page 638.
Kish SJ, Bergeron C, Rajput A, Dozic S, Mastrogiacomo F, Maines MD. 1997. The heme oxygenase system: A regulator of
et al. 1992. Brain cytochrome oxidase in Alzheimers dis- second messenger gases. Annu Rev Pharmacol Toxicol 37:
ease. J Neurochem 59: 776-779. 517-554.
Knight JA. 2000. The biochemistry of aging. Adv Clin Chem Maines MD. 2000. The heme oxygenase system and its func-
35: 1-62. tions in the brain. Cell Mol Biol 46: 573-585.
Aging and oxidative stress response in the CNS 6 143
Mancuso C, Bonsignore A, Di Stasio E, Mordente A, Motter- Motterlini R, Foresti R, Bassi R, Calabrese V, Clark JE, et al.
lini R. 2003. Bilirubin and Snitrosothiols interaction: Evi- 2000a. Endothelial heme oxygenase1 induction by hyp-
dence for a possible role of bilirubin as a scavenger of nitric oxia. Modulation by inducible nitricoxide synthase and
oxide. Biochem Pharmacol 66: 2355-2366. Snitrosothiols. J Biol Chem 275: 13613-13620.
Mandavilli BS, Rao KS. 1996. Accumulation of DNA damage Motterlini R, Foresti R, Bassi R, Green CJ. 2000b. Curcumin,
in aging neurons occurs through a mechanism other than an antioxidant and antiinflammatory agent, induces heme
apoptosis. J Neurochem 67: 1559-1565. oxygenase1 and protects endothelial cells against oxidative
Markesbery WR. 1997. Oxidative stress hypothesis in Alzhei- stress. Free Radic Biol Med 28: 1303-1312.
mers disease. Free Radic Biol Med 23: 134-147. Motterlini R, Green CJ, Foresti R. 2002a. Regulation of heme
Martin D, Rojo AI, Salinas M, Diaz R, Gallardo G, et al. 2004. oxygenase1 by redox signals involving nitric oxide. Anti-
Regulation of heme oxygenase1 expression through the oxid Redox Signal 4: 615-624.
phosphatidylinositol3kinase/Akt pathway and the Nrf2 Motterlini R, Clark JE, Foresti R, Sarathchandra P, Mann BE,
transcription factor in response to the antioxidant phyto- et al. 2002b. Carbon monoxidereleasing molecules: Char-
chemical carnosol. J Biol Chem 279: 8919-8929. acterization of biochemical and vascular activities. Circ Res
Martin GM, La Marco K, Strauss E, Kelner KL. 2003. Research 92: 17-24.
on aging: The end of the beginning. Science 299: 1339-1341. Motterlini R, Mann BE, Johnson TR, Clark JE, Foresti R,
Mattson MP. 2003. Will caloric restriction and folate protect et al. 2003. Bioactivity and pharmacological actions of
against AD and PD? Neurology 60: 690-695. carbon monoxidereleasing molecules. Curr Pharm Des 9:
Mattson MP, Duan W, Guo Z. 2003. Meal size and fre- 2525-2539.
quency affect neuronal plasticity and vulnerability to dis- Narasimhan P, Swanson RA, Sagar SM, Sharp FR. 1996. As-
ease: Cellular and molecular mechanisms. J Neurochem 84: trocyte survival and HSP70 heat shock protein induction
417-431. following heat shock and acidosis. Glia 17: 147-159.
Mattson MP, Duan W, Chan SL, Cheng A, Haughey N, et al. Ojaimi J, Masters CL, McLean C, Opeskin K, McKelvie P, et al.
2002. Neuroprotective and neurorestorative signal trans- 1999. Irregular distribution of cytochrome c oxidase pro-
duction mechanisms in brain aging: Modification by genes, tein subunits in aging and Alzheimers disease. Ann Neurol
diet and behavior. Neurobiol Aging 23: 695-705. 46: 656-660.
Mautes AE, Bergeron M, Sharp FR, Panter SS, Weinzierl M, Okubo A, Kinouchi H, Owada Y, Kunizuka H, Itoh H, et al.
et al. 2000. Sustained induction of heme oxygenase1 in the 2000. Simultaneous induction of mitochondrial heat shock
traumatized spinal cord. Exp Neurol 166: 254-265. protein mRNAs in rat forebrain ischemia. Brain Res Mol
Mayer RJ. 2003. From neurodegeneration to neurohomeosta- Brain Res 84: 127-134.
sis: The role of ubiquitin. Drug News Perspect 16: 103-108. Otterbein LE. 2002. Carbon monoxide: Innovative antiin-
McLaughlin B, Hartnett KA, Erhardt JA, Legos JJ, White RF, flammatory properties of an ageold gas molecule. Antioxid
et al. 2003. Caspase 3 activation is essential for neuropro- Redox Signal 4: 309-319.
tection in preconditioning. Proc Natl Acad Sci USA 100: Otterbein LE, Soares MP, Yamashita K, Bach F. 2003a. Heme
715-720. oxygenase1: Unleashing the protective properties of heme.
Mc Naught KS, Olanow CW, Halliwell B, Isacson O, Jenner P. Trends Immunol 24: 449-466.
2001. Failure of the ubiquitinproteasome system in Par- Otterbein LE, Zuckerbraun BS, Haga M, Liu F, Song R, et al.
kinson disease. Nat Rev Neurosci 2: 589-594. 2003b. Carbon monoxide suppresses arteriosclerotic
Mecocci P, Beal MF, Cecchetti R, Polidori MC, Cherubini A, lesions associated with chronic graft rejection and with
et al. 1997. Mitochondrial membrane fluidity and oxidative balloon injury. Nat Med 9: 183-190.
damage to mitochondrial DNA in aged and AD human Papa S, Skulachev VP. 1997. Reactive oxygen species, mitochon-
brain. Mol Chem Neuropathol 31: 53-64. dria, apoptosis and aging. Mol Cell Biochem 174: 305-319.
Mecocci P, Polidori MC, Ingegni T, Cherubini A, Chionne F, Paradies G, Petrosillo G, Gadaleta MN, Ruggiero FM. 1999.
et al. 1998. Oxidative damage to DNA in lymphocytes from The effect of aging and acetylLcarnitine on the pyruvate
AD patients. Neurology 51: 1014-1017. transport and oxidation in rat heart mitochondria. FEBS
Moore DJ, Dawson VL, Dawson TM. 2003. Role for the Lett 454: 207-209.
ubiquitinproteasome system in Parkinsons disease and Paradies G, Petrosillo G, Pistolese M, Ruggiero FM. 2002.
other neurodegenerative brain amyloidoses. Neuromolecu- Reactive oxygen species affect mitochondrial electron
lar Med 4: 95-108. transport complex I activity through oxidative cardiolipin
Mosser DD, Caron AW, Bourget L, DenisLarose D. 1997. Role damage. Gene 286: 135-141.
of the human heat shock protein hsp70 in protection against Perluigi M, Poon FH, Hensley K, Pierce WM, Klein JB, et al.
stressinduced apoptosis. Mol Cell Biol 17: 5317-5327. 2005. Proteomic analysis of 4hydroxy2nonenalmodified
144 6 Aging and oxidative stress response in the CNS
proteins in G93ASOD1 transgenic miceA model of famil- Rosenberg PA, Li Y, Ali S, Altiok N, Back SA, et al. 1999.
ial amyotrophic lateral sclerosis. Free Radic Biol Med 38: Intracellular redox state determines whether nitric oxide is
960-968. toxic or protective to rat oligodendrocytes in culture. J
Perry G, Taddeo MA, Petersen RB, Castellani RJ, Harris PL, Neurochem 73: 476-484.
et al. 2003. Adventiouslybound redox active iron and cop- Roses AD. 1996. Apolipoprotein E alleles as risk factors in
per are at the center of oxidative damage in Alzheimer Alzheimers disease. Annu Rev Med 47: 387-400.
disease. Biometals 16: 77-81. Rothschild H, Jazwinski S. 1988. Human longevity determi-
Perry TL, Godin DV, Hansen S. 1982. Parkinsons disease: A nant genes. J La State Med Soc 150: 272-274.
disorder due to nigral glutathione deficiency? Neurosci Lett Sastre J, Pallardo FV, Vina J. 2003. The role of mitochondrial
33: 305-310. oxidative stress in aging. Free Radic Biol Med 35: 1-8.
Piantadosi CA, Zhang J, Levin ED, Folz RJ, Schmechel DE. Sayre LM, Perry G, Harris PL, Liu Y, Schubert KA, et al. 2000.
1997. Apoptosis and delayed neuronal damage after carbon In situ oxidative catalysis by neurofibrillary tangles and
monoxide poisoning in the rat. Exp Neurol 147: 103-114. senile plaques in Alzheimers disease: A central role for
Pocernich C, BoydKimball D, Poon F, Thongboonkerd V, bound transition metals. J Neurochem 74: 270-279.
Lynn BC, et al. 2005. Proteomic analysis of human astro- Scapagnini G, Butterfield DA, Colombrita C, Sultana R,
cytes expressing the HIV protein TAT. Brain Res Mol Brain Pascale A, et al. 2004. Ethyl ferulate, a lipophilic polyphe-
Res 133: 307-316. nol, induces HO1 and protects rat neurons against oxida-
Poon F, Calabrese V, Scapagnini G, Butterfield DA. 2004a. tive stress. Antioxid Redox Signal 6: 811-818.
Free radicals: Key to brain aging and heme oxygenase as a Scapagnini G, Giuffrida Stella AM, Abraham NG, Alkon D,
cellular response to oxidative stress. J Gerontol 59: 478-493. Calabrese V. 2002a. Differential expression of Heme oxyge-
Poon HF, Calabrese V, Scapagnini G, Butterfield DA. 2004b. nase1 in rat brain by endotoxin (LPS). Heme Oxygenase in
Free radicals and brain aging. Clin Geriatr Med 20: 329-359. Biology and Medicine. Abraham et al. editors. New York:
Poon FH, Frasier M, Shreve N, Calabrese V, Wolozin B, et al. Kluwer Academic/Plenum Publishers; pp. 121-134.
2005. Mitochondrial associated metabolic proteins are Scapagnini G, DAgata V, Calabrese V, Pascal A, Colombrita
selectively oxidized in A30P asynuclein transgenic mice C, et al. 2002b. Gene expression profiles of heme oxygenase
a model of familial Parkinsons disease. Neurobiol Dis 18: isoforms in the rat brain. Brain Res 954: 51-59.
492-498. Scapagnini G, Foresti R, Calabrese V, Giuffirida Stella AM,
PorteroOtin M, Bellmunt MJ, Ruiz MC, Barja G, Pamplona Green CJ, et al. 2002c. Caffeic acid phenethyl ester and
R. 2001. Correlation of fatty acid unsaturation of the major curcumin: A novel class of heme oxygenase1 inducers.
liver mitochondrial phospholipid classes in mammals to Mol Pharmacol 61: 554-561.
their maximum life span potential. Lipids 236: 491-498. Scapagnini G, Ravagna A, Bella R, Colombrita C, Pennisi G,
Powell CS, Jackson RM. 2003. Mitochondrial complex I, et al. 2002d. Longterm ethanol administration enhances
aconitase, and succinate dehydrogenase during hypoxia agedependent modulation of redox state in brain and
reoxygenation: Modulation of enzyme activities by MnSOD. peripheral organs of rat: Protection by acetyl carnitine.
Am J Physiol Lung Cell Mol Physiol 285: L189-L198. Int J Tissue React 24: 89-96.
Premkumar DR, Smith MA, Richey PL, Petersen RB, Schapira AH. 1998. Mitochondrial dysfunction in neurode-
Castellani R, et al. 1995. Induction of heme oxygenase1 generative disorders. Biochim Biophys Acta 1366: 225-233.
mRNA and protein in neocortex and cerebral vessels in Schapira AHV, Cooper JM, Dexter D. 1990. Mitochondrial
Alzheimers disease. J Neurochem 65: 1399-1402. complex I deficiency in Parkinson disease. J Neurochem 54:
Quiles JL, Martinez E, Ibanez S, Ochoa JJ, Martin Y, et al. 823-827.
2002. Ageingrelated tissuespecific alterations in mito- Schipper HM, Chertkow H, Mehindate K, Frankel D, Melmed
chondrial composition and function are modulated by die- C, et al. 2000. Evaluation of heme oxygenase1 as a systemic
tary fat type in the rat. J Bioenerg Biomembr 34: 517-524. biological marker of sporadic AD. Neurology 54: 1297-1304.
Radisky DC, Babcock MC, Kaplan J. 1999. The yeast frataxin Sellebjerg F, Giovannoni G, Hand A, Madsen HO, Jensen CV,
homologue mediates mitochondrial iron efflux. Evidence et al. 2002. Cerebrospinal fluid levels of nitric oxide meta-
for a mitochondrial iron cycle. J Biol Chem 274: 4497-4499. bolites predict response to methylprednisolone treatment
Raju VS, McCoubrey WK, Maines MD. 1997. Regulation of in multiple sclerosis and optic neuritis. J Neuroimmunol
heme oxygenase2 by glucocorticoids in neonatal rat brain: 125: 198-203.
Characterization of a functional glucocorticoid response Selman C, Gredilla R, Phaneuf S, Kendaiah S, Barja G, et al.
element. Biochim Biophys Acta 1351: 89-104. 2003. Shortterm caloric restriction and regulatory proteins
Risch N, Merikangas K. 1996. The future of genetic studies of of apoptosis in heart, skeletal muscle and kidney of Fischer
complex human diseases. Science 273: 1516-1517. 344 rats. Biogerontology 4: 141-147.
Aging and oxidative stress response in the CNS 6 145
Sergent O, Griffon B, Morel I, Chevanne M, Dubos MP, et al. Takeda A, Perry G, Abraham NG, Dwyer BE, Kutty RK, et al.
1977. Effect of nitric oxide on ironmediated oxidative 2000. Overexpression of heme oxygenase in neuronal
stress in primary rat hepatocyte culture. Hepatology 25: cells, the possible interaction with Tau. J Biol Chem 275:
122-127. 5395-5399.
Sharman EH, Bondy SC. 2001. Effects of age and dietary Tenhunen R, Marver HS, Schmid R. 1969. Microsomal heme
antioxidants on cerebral electron transport chain activity. oxygenase. Characterization of the enzyme. J Biol Chem
Neurobiol Aging 22: 629-634. 244: 6388-6394.
Sian J, Dexter DT, Lees AJ. 1994. Alterations in glutathione Thompson EJ. 1988. The CSF Proteins: A Biochemical
levels in Parkinsons disease and other neurodegenerative dis- Approach. Elsevier; London: pp. 16-24.
orders affecting the basal ganglia. Ann Neurol 36: 348-355. Toshniwal PK, Zarling EJ. 1992. Evidence for increased lipid
Sjorstrand T. 1949. Endogenous formation of carbon monoxide peroxidation in multiple sclerosis. Neurochem Res 17:
in man under normal and pathological conditions. Scand J 205-207.
Clin Lab Invest 1: 201-214. Tran EH, Hardin PH, Verge G, Owens T. 1997. Astrocytes and
Smith MA, Harris PRL, Sayre LM, Beckman JS, Perry G. 1997. microglia express inducible nitric oxide synthase in mice
Widespread peroxynitrite mediated damage in Alzheimers with experimental allergic encephalomyelitis. J Neuroim-
disease. J Neurosci 17: 2653-2657. munol 74: 121-129.
Soares MP, Seldon MP, Gregoire IP, Vassilevskaia T, Berberat Turcanu V, Dhouib M, Poindron P. 1998. Nitric oxide
PO, et al. 2004. Heme oxygenase1 modulates the expres- synthase inhibition by heme oxygenase decreases macro-
sion of adhesion molecules associated with endothelial cell phage nitric oxidedependent cytotoxicity: A negative feed-
activation. J Immunol 172: 3553-3563. back mechanism for the regulation of nitric oxide
Soti C, Csermely P. 2002. Chaperones and aging: Role in production. Res Immunol 149: 741-744.
neurodegeneration and in other civilizational diseases. Turner CP, Panter SS, Sharp FR. 1999. Antioxidants prevent
Neurochem Int 41: 383-389. focal rat brain injury as assessed by induction of heat shock
Spencer J, Jenner A, Aruoma O, Evans P, Jenner P, et al. 1994. proteins (Hsp70, HO1/Hsp32, Hsp47) following sub-
Intense oxidative DNA damage promoted by LDOPA and arachnoid injections of lysed blood. Brain Res Mol Brain
its metabolites. Implication for neurodegenerative diseases. Res 65: 87-102.
FEBS Lett 353: 246-250. Tyrrell R. 1999. Redox regulation and oxidant activation of
Spencer JP, Whiteman M, Jenner P, Halliwell B. 2002. heme oxygenase1. Free Radic Res 31: 335-340.
5Scysteinylconjugates of catecholamines induce cell Valentim LM, Rodnight R, Geyer AB, Horn AP, Tavares A,
damage, extensive DNA base modification and increase in et al. 2003. Changes in heat shock protein 27 phosphoryla-
caspase3 activity in neurons. J Neurochem 81: 122-129. tion and immunocontent in response to preconditioning to
Stadtman ER. 2001. Protein oxidation in aging and age oxygen and glucose deprivation in organotypic hippocam-
related diseases. Ann NY Acad Sci 928: 22-38. pal cultures. Neuroscience 118: 379-386.
Stamler JS, Hausladen A. 1998. Oxidative modifications in Veinbergs I, Mallory M, Sagara Y, Masliah E. 2000. Vitamin E
nitrosative stress. Cell 78: 931-936. supplementation prevents spatial learning deficits and den-
Stamler JS, Singel DI, Loscalzo J. 1992. Biochemistry of nitric dritic alterations in aged apolipoprotein Edeficient mice.
oxide and its redox activated forms. Science 258: 1898-1902. Eur J Neurosci 12: 4541-4546.
Stocker R, Yamamoto Y, McDonagh AF, Glazer AN, Ames BN. Verma A, Hirsch DJ, Glatt CE, Ronnett GV, Snyder SH. 1993.
1987. Bilirubin is an antioxidant of possible physiological Carbon monoxide: A putative neural messenger. Science
importance. Science 235: 1043-1046. 259: 381-384.
Strauss E. 2003. Longevity research. Single signal unites treat- Wang R, Wu L. 2003. Interaction of selective amino acid
ments that prolong life. Science 300: 881-883. residues of K(Ca) channels with carbon monoxide. Exp
Sultana R, Ravagna A, MohmmadAbdul H, Calabrese V, Biol Med 228: 474-480.
Butterfield DA. 2005. Ferulic acid ethyl ester protects neu- Weng YH, Yang G, Weiss S, Dennery PA. 2003. Interaction
rons against amyloid bpeptide142induced oxidative stress between heme oxygenase1 and 2 proteins. J Biol Chem
and neurotoxicity: Relationship to antioxidant activity. J 278: 50999-51005.
Neurochem 92: 749-758. Wheeler DS, Dunsmore KE, Wong HR. 2003. Intracellular
Sun J, Hoshino O, Takaku K, Nakajima O, Muto A, et al. 2002. delivery of HSP70 using HIV1 Tat protein transduction
Hemoprotein Bach1 regulates enhancer availability of domain. Biochem Biophys Res Commun 301: 54-59.
heme oxygenase1 gene. EMBO J 21: 5216-5224. Xin W, Chen XM, Li H, Dryhurst G. 2000. Oxidative meta-
Szilard L. 1977. On the nature of the aging process. Proc Natl bolites of 5Scysteinylnorepinephrine are irreversible inhi-
Acad Sci USA 45: 35-45. bitors of mitochondrial complex I and the aketoglutarate
146 6 Aging and oxidative stress response in the CNS
dehydrogenase and pyruvate dehydrogenase complexes: Yuceyar N, Taskiran D, Sagduyu A. 2001. Serum and cerebro-
Possible implications for neurodegenerative brain disor- spinal fluid nitrite and nitrate levels in relapsingremitting
ders. Chem Res Toxicol 13: 749-760. and secondary progressive multiple sclerosis patients. Clin
Yamashita K, McDaid J, Ollinger R, Tsui TY, Berberat PO, Neurol Neurosurg 103: 206-211.
et al. 2004. Biliverdin, a natural product of heme catabo- Zhang HX, Du GH, Zhang JT. 2003. Ischemic precondition-
lism, induces tolerance to cardiac allografts. FASEB J 18: ing preserves brain mitochondrial functions during the
765-767. middle cerebral artery occlusion in rat. Neurol Res 25:
Yatin SM, Kink CD, Butterfield DA. 1999. Invitro and invivo 471-476.
oxidative stress associated with Alzheimers amyloid bpep-
tide142. Neurobiol Aging 20: 325-330.
7 Parallels Between
Neurodevelopment and
Neurodegeneration: A Case Study
of Alzheimers Disease
X. Zhu . G. Casadesus . K. M. Webber . C. S. Atwood . R. L. Bowen . G. Perry . M. A. Smith
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
1.1 Review Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152
Abstract: An accumulating body of evidence shows significant similarities between cellular processes
involved in neurodegeneration and those involved in neurodevelopment. Most striking, in Alzheimers
disease, cell cycle re-entry appears to be a pathological signature of the disease that also occur in brain
development. Such inappropriate reactivation of a fetal program is likely to play a key role in both the
etiology and pathogenesis of disease.
List of Abbreviations: AD, Alzheimers disease; ApoE, apolipoprotein E; APP, amyloid- protein precur-
sor; A, amyloid-; Cdk, cyclin-dependent kinase; HPG axis, hypothalamic-pituitary-gonadal axis; HRT,
hormone replacement therapy; LH, luteinizing hormone; PS, presenilin; ROS, reactive oxygen species
1 Introduction
Interestingly, it appears that these ADassociated proteins also play central roles in early neural
development (Bothwell and Giniger, 2000). For example, the overexpression of human AbPP in
Drosophila leads to development of blisteredwing phenotype (Fossgreen et al., 1998). Mice genetically
deficient in AbPP, APLP1, or APLP2 are viable with only subtle neurological deficits, suggesting
redundancy in function among the AbPP family members (Heber et al., 2000). More recently, mutant
mice lacking all three AbPP genes displayed early lethality as well as high incidence of cortical dysplasia
Parallels between neurodevelopment and neurodegeneration: A case study of Alzheimers disease
7 149
suggesting a critical role in neural development (Koo, 2002). However, the actual function of AbPP
remains unresolved. Interestingly, in vitro studies suggest that AbPP may function as a cellcycle
regulator because AbPP and its proteolytic fragments (i.e., Ab peptide and AbPP) are mitogenic
(Schubert et al., 1989; Milward et al., 1992; Copani et al., 1999; Hoffmann et al., 2000; Schmitz
et al., 2002). Further, its adaptor proteins including Fe65 and AbPPBP1 have been shown to function
as regulators of the cell cycle. AbPPBP1 is a cellcycle protein that normally negatively regulates the
progression of cells into the S phase and positively regulates the progression into mitosis (Chen et al.,
2000), whereas Fe65 is a nuclear protein that negatively regulates G1 to Sphase progression by
inhibiting the key Sphase enzymes (Bruni et al., 2002). Therefore, it is likely that AbPP may act as
a cell surface receptor to relay cell cyclerelated signals.
2.1.2 Presenilins
The role of PS1 in neurodevelopment is more firmly established at least partially because of its involvement
in the gcleavage of the Notch receptor, which plays critical roles in determining cell fate during early
development (Selkoe and Kopan, 2003). It is shown that the PS1deficient mouse displays profound
developmental abnormalities, including failed somite formation and altered neurogenesis (Shen et al.,
1997; Wong et al., 1997). Additionally, the loop domain of PS1 serves as a scaffold for bcatenin degradation
(Willert and Nusse, 1998; Kang et al., 2002). It is known that bcatenin can migrate to the nucleus and
interact with transcription factors like TCF/LEF1 to induce the expression of genes such as Cyclin D1 and
c-myc, which are crucial in the G1/Sphase transition (Willert and Nusse, 1998). In fact, transgenic mice
that express a constitutively active bcatenin in neuroepithelial precursor cells develop enlarged brains
because a greater proportion of neural precursors reenter the cell cycle after mitosis (Chenn and Walsh,
2002). Therefore, PS1 may also be involved in cellcycle regulation. Indeed, the association of presenilins
with centrosome and centromeres suggests that they play a role in cell division and segregation of
chromosomes (Li et al., 1997). In addition, it has been shown that the overexpression of both PS1 and
PS2 proteins resulted in G1phase arrest of the cell cycle (Janicki and Monteiro, 1999; Janicki et al., 2000),
whereas PS1 deficiency results in accelerated entry into (Soriano et al., 2001) and significantly prolonged
length (Yuasa et al., 2002) of the S phase of the cell cycle. This evidence suggests that PS1 and PS2 play
important roles in cellcycle control.
2.1.3 Apolipoprotein E
Finally, the apolipoprotein E4 (ApoE4) alleles, risk factor of lateonset AD, which confer increased
susceptibility both to AD and prostate cancer, implicate an association between the ApoE4 allele and a
propensity toward developing a dysregulated cell cycle (Yuasa et al., 2002). In support of such a notion,
ApoE is itself associated with the amyloid deposits found in pituitary adenomas.
neurogenesis during development and neurodegeneration during AD. Indeed, this may well reflect the fact
that ADassociated proteins are involved in cellcycle regulation, which may be disrupted in the pathogen-
esis or progression of the disease.
AD in a topographical manner that overlaps with phosphotau. Moreover, free radicals produced during
oxidative stress are speculated to be pathologically important in AD and other neurodegenerative
diseases (Cross et al., 1987; Perry et al., 1998). Dysregulated cellcycle control may also lead to
increased cytoplasmic oxidative damage observed in susceptible neurons in AD, as it may adversely
affect mitochondria function. It is reported that mitochondrial DNA and protein are increased (Hirai
et al., 2001), which may be due to the same mitogenic signal that triggers cellcycle reentry and nuclear
DNA duplication. However, the incompetence of postmitotic neurons to complete the celldivision
cycle may result in the failure of organelle kinesis after mitochondrial DNA duplication that probably
leads to mitochondrial abnormalities. It is clear that dysregulated mitochondria are major sources of
ROS for surrounding cytoplasm.
climacteric years and is ineffective or even detrimental when administered after such time (Rapp et al.,
2003) or during the latent preclinical stage of AD, which usually occurs much later in life (Resnick and
Henderson, 2002).
Although falling levels of estrogen might explain the higher incidence of AD in females than in males, a
decline in sex steroids does not account for why males with Downs syndrome have a higher risk of
developing ADtype changes than females, as the levels of sex steroids in both males and females with
Downs syndrome are comparable to those found in the general population (Schupf et al., 1998). This
reversal in genderbased predisposition of AD combined with questions regarding a critical period of
HRTbased protection against dementia has prompted us to investigate the potential role of other
hypothalamicpituitarygonadal (HPG) axis hormones in the etiology of AD (Bowen et al., 2000, 2002;
Bowen 2001; Smith et al., 2003), which, along with sex steroids, regulate reproductive function, and also
show expression in many nonreproductive tissues including, most notably, the brain. In this regard,
previous reports demonstrate a twofold increase in the gonadotropin LH in AD patients when compared
to agematched control subjects (Bowen et al., 2000; Short et al., 2001). More important, the fact that the
highest density of LH receptors in the brain is found within the hippocampus (Lei et al., 1993; AlHader
et al., 1997a, b), the most vulnerable region in AD, and gonadotropins are known to cross the bloodbrain
barrier (Lukacs et al., 1995), we speculate that elevated LH levels may be a driving pathogenic force in
mitogenic dysfunction seen in AD (Bowen et al., 2002).
As LH is a powerful mitogen (Harris et al., 2002), and given the temporal and spatial overlap with mitotic
changes in AD (Bowen et al., 2002; unpublished observations), it is plausible that elevations in LH seen in
AD patients are responsible for the inappropriate cellcycle reentry in ADvulnerable neurons. In support of
such a notion, we found significant elevations of LH in ADvulnerable neuronal populations when
compared to agedmatched control cases (Bowen et al., 2002). More important, such increases appear to
be a very early change in disease history and parallel the ectopic expression of cellcycle and oxidative
markers that represent one of the initiating pathological changes preceding neuronal degeneration by
decades (Nunomura et al., 2001; Ogawa et al., 2003). Therefore, therapeutic strategies that are targeted
toward the release of LH may prove successful in treatment of AD. In this regard, it is likely that the
ineffectiveness of estrogen replacement therapy in the treatment of AD (Mulnard, 2000; Mulnard et al.,
2000) may be due to the incapacity of this hormone to restore the HPG axis to its premenopausal state.
Therefore, treatments such as leuprolide acetatea gonadotropinreleasing hormone agonist, which has
been shown to suppress LH to undetectable levels by downregulating pituitary gonadotropinreleasing
hormone receptorsfor mature patients (i.e., postmenopausal women) might be a more effective
method for AD. The results of phase III clinical trials of leuprolide acetate for the treatment of AD are
anticipated in late 2007.
5 Conclusions
As highlighted earlier, factors such as luteinizing hormone, which drive neuronal proliferation during
development, may also play a key role in neuronal degeneration during AD. What irony, that factors used
to build the brain are the same that deconstruct the brain after a lifetime of use. We hope that the
recognition of this fact will enable efforts to forestall deconstruction and preserve cognitive integrity to
the grave.
Parallels between neurodevelopment and neurodegeneration: A case study of Alzheimers disease
7 153
Acknowledgements
Work in the authors laboratories is supported by the National Institutes of Health and the Alzheimers
Association. CSA, GP, and MAS are consultants for and own equity in Voyager Pharmaceutical Corporation.
References
AlHader AA, Lei ZM, Rao CV. 1997a. Novel expression of Chenn A, Walsh CA. 2002. Regulation of cerebral cortical size
functional luteinizing hormone/chorionic gonadotropin by control of cell cycle exit in neural precursors. Science
receptors in cultured glial cells from neonatal rat brains. 297: 365-369.
Biol Reprod 56: 501-507. Connor B, Young D, Lawlor P, Gai W, Waldvogel H, et al.
AlHader AA, Lei ZM, Rao CV. 1997b. Neurons from fetal rat 1996. Trk receptor alterations in Alzheimers disease. Mol
brains contain functional luteinizing hormone/chorionic Brain Res 42: 1-17.
gonadotropin receptors. Biol Reprod 56: 1071-1076. Connor B, Beilharz EJ, Williams C, Synek B, Gluckman PD,
Allen SJ, MacGowan SH, Treanor JJ, Feeney R, Wilcock GK, et al. 1997. Insulinlike growth factorI (IGFI) immunore-
et al. 1991. Normal betaNGF content in Alzheimers disease activity in the Alzheimers disease temporal cortex and
cerebral cortex and hippocampus. Neurosci Lett 131: 135-139. hippocampus. Mol Brain Res 49: 283-290.
Birecree E, Whetsell WO Jr, Stoscheck C, King LE Jr, Nanney Copani A, Condorelli F, Caruso A, Vancheri C, Sala A, et al.
LB. 1988. Immunoreactive epidermal growth factor recep- 1999. Mitotic signaling by betaamyloid causes neuronal
tors in neuritic plaques from patients with Alzheimers death. FASEB J 13: 2225-2234.
disease. J Neuropathol Exp Neurol 47: 549-560. Coulson EJ, Paliga K, Beyreuther K, Masters CL. 2000. What
Bothwell M, Giniger E. 2000. Alzheimers disease: Neurode- the evolution of the amyloid protein precursor supergene
velopment converges with neurodegeneration. Cell 102: family tells us about its function. Neurochem Int 36:
271-273. 175-184.
Bowen RL. 2001. Sex hormones, amyloid protein, and Alzhei- Cross CE, Halliwell B, Borish ET, Pryor WA, Ames BN, et al.
mer disease. JAMA 286: 790-791. 1987. Oxygen radicals and human disease. Ann Intern Med
Bowen RL, Isley JP, Atkinson RL. 2000. An association of 107: 526-545.
elevated serum gonadotropin concentrations and Alzheimer Crutcher KA, Scott SA, Liang S, Everson WV, Weingartner J.
disease? J Neuroendocrinol 12: 351-354. 1993. Detection of NGFlike activity in human brain tissue:
Bowen RL, Smith MA, Harris PL, Kubat Z, Martins RN, et al. Increased levels in Alzheimers disease. J Neurosci 13: 2540-
2002. Elevated luteinizing hormone expression colocalizes 2550.
with neurons vulnerable to Alzheimers disease pathology. Fenton H, Finch PW, Rubin JS, Rosenberg JM, Taylor WG,
J Neurosci Res 70: 514-518. et al. 1998. Hepatocyte growth factor (HGF/SF) in Alzhei-
Breitner JC, Silverman JM, Mohs RC, Davis KL. 1988. Famil- mers disease. Brain Res 779: 262-270.
ial aggregation in Alzheimers disease: Comparison of risk Ferguson KL, Callaghan SM, OHare MJ, Park DS, Slack RS.
among relatives of earlyand lateonset cases, and among 2000. The RbCDK4/6 signaling pathway is critical in neu-
male and female relatives in successive generations. Neu- ral precursor cell cycle regulation. J Biol Chem 275: 33593-
rology 38: 207-212. 33600.
Bruni P, Minopoli G, Brancaccio T, Napolitano M, Faraonio Fossgreen A, Bruckner B, Czech C, Masters CL, Beyreuther K,
R, et al. 2002. Fe65, a ligand of the Alzheimers beta et al. 1998. Transgenic Drosophila expressing human amy-
amyloid precursor protein, blocks cell cycle progression loid precursor protein show gammasecretase activity and a
by downregulating thymidylate synthase expression. blisteredwing phenotype. Proc Natl Acad Sci USA 95:
J Biol Chem 277: 35481-35488. 13703-13708.
Busser J, Geldmacher DS, Herrup K. 1998. Ectopic cell cycle GomezPinilla F, Cummings BJ, Cotman CW. 1990. Induction
proteins predict the sites of neuronal cell death in Alzheimers of basic fibroblast growth factor in Alzheimers disease
disease brain. J Neurosci 18: 2801-2807. pathology. Neuroreport 1: 211-214.
Chen Y, McPhie DL, Hirschberg J, Neve RL. 2000. The amy- Hardy J. 1997. Amyloid, the presenilins and Alzheimers dis-
loid precursor proteinbinding protein APPBP1 drives the ease. Trends Neurosci 20: 154-159.
cell cycle through the SM checkpoint and causes apoptosis Harris D, Bonfil D, Chuderland D, Kraus S, Seger R, et al.
in neurons. J Biol Chem 275: 8929-8935. 2002. Activation of MAPK cascades by GnRH: ERK and
154
7 Parallels between neurodevelopment and neurodegeneration: A case study of Alzheimers disease
Jun Nterminal kinase are involved in basal and GnRH Lukacs H, Hiatt ES, Lei ZM, Rao CV. 1995. Peripheral and
stimulated activity of the glycoprotein hormone LHbeta intracerebroventricular administration of human chorionic
subunit promoter. Endocrinology 143: 1018-1025. gonadotropin alters several hippocampusassociated beha-
Heber S, Herms J, Gajic V, Hainfellner J, Aguzzi A, et al. 2000. viors in cycling female rats. Horm Behav 29: 42-58.
Mice with combined gene knockouts reveal essential and Manly JJ, Merchant CA, Jacobs DM, Small SA, Bell K, et al.
partially redundant functions of amyloid precursor protein 2000. Endogenous estrogen levels and Alzheimers dis-
family members. J Neurosci 20: 7951-7963. ease among postmenopausal women. Neurology 54:
Henderson VW, PaganiniHill A, Emanuel CK, Dunn ME, 833-837.
Buckwalter JG. 1994. Estrogen replacement therapy in Masliah E, Mallory M, Alford M, Deteresa R, Saitoh T. 1995.
older women. Comparisons between Alzheimers disease PDGF is associated with neuronal and glial alterations of
cases and nondemented control subjects. Arch Neurol 51: Alzheimers disease. Neurobiol Aging 16: 549-556.
896-900. McGonigal G, Thomas B, McQuade C, Starr JM, MacLennan
Hirai K, Aliev G, Nunomura A, Fujioka H, Russell RL, et al. WJ, et al. 1993. Epidemiology of Alzheimers presenile
2001. Mitochondrial abnormalities in Alzheimers disease. dementia in Scotland, 197488. BMJ 306: 680-683.
J Neurosci 21: 3017-3023. McShea A, Harris PL, Webster KR, Wahl AF, Smith MA. 1997.
Hoffmann J, Twiesselmann C, Kummer MP, Romagnoli P, Abnormal expression of the cell cycle regulators P16 and
Herzog V. 2000. A possible role for the Alzheimer amyloid CDK4 in Alzheimers disease. Am J Pathol 150: 1933-1939.
precursor protein in the regulation of epidermal basal cell Milward EA, Papadopoulos R, Fuller SJ, Moir RD, Small D,
proliferation. Eur J Cell Biol 79: 905-914. et al. 1992. The amyloid protein precursor of Alzheimers
Janicki SM, Monteiro MJ. 1999. Presenilin overexpression disease is a mediator of the effects of nerve growth factor on
arrests cells in the G1 phase of the cell cycle. Arrest poten- neurite outgrowth. Neuron 9: 129-137.
tiated by the Alzheimers disease PS2(N141I)mutant. Am J Mulnard RA. 2000. Estrogen as a treatment for Alzheimer
Pathol 155: 135-144. disease. JAMA 284: 307-308.
Janicki SM, Stabler SM, Monteiro MJ. 2000. Familial Alzhei- Mulnard RA, Cotman CW, Kawas C, van Dyck CH, Sano M,
mers disease presenilin1 mutants potentiate cell cycle et al. 2000. Estrogen replacement therapy for treatment of
arrest. Neurobiol Aging 21: 829-836. mild to moderate Alzheimer disease: A randomized con-
Jorm AF, Jolley D. 1998. The incidence of dementia: A meta trolled trial. Alzheimers disease cooperative study. JAMA
analysis. Neurology 51: 728-733. 283: 1007-1015.
Jorm AF, Korten AE, Henderson AS. 1987. The prevalence of Nagy Z, Esiri MM, Smith AD. 1997a. Expression of cell
dementia: A quantitative integration of the literature. Acta division markers in the hippocampus in Alzheimers dis-
Psychiatr Scand 76: 465-479. ease and other neurodegenerative conditions. Acta Neuro-
Kang DE, Soriano S, Xia X, Eberhart CG, De Strooper B, et al. pathol (Berl) 93: 294-300.
2002. Presenilin couples the paired phosphorylation of Nagy Z, Esiri MM, Cato AM, Smith AD. 1997b. Cell cycle
betacatenin independent of axin: Implications for beta markers in the hippocampus in Alzheimers disease. Acta
catenin activation in tumorigenesis. Cell 110: 751-762. Neuropathol (Berl) 94: 6-15.
Kawas C, Resnick S, Morrison A, Brookmeyer R, Corrada M, Nunomura A, Perry G, Aliev G, Hirai K, Takeda A, et al. 2001.
et al. 1997. A prospective study of estrogen replacement Oxidative damage is the earliest event in Alzheimer disease.
therapy and the risk of developing Alzheimers disease: J Neuropathol Exp Neurol 60: 759-767.
The Baltimore longitudinal study of aging. Neurology 48: Ogawa O, Lee HG, Zhu X, Raina A, Harris PL, et al. 2003.
1517-1521. Increased p27, an essential component of cell cycle control,
Koo EH. 2002. The betaamyloid precursor protein (APP) in Alzheimers disease. Aging Cell 2: 105-110.
and Alzheimers disease: Does the tail wag the dog? Traffic Perry G, Castellani RJ, Hirai K, Smith MA. 1998. Reactive
3: 763-770. oxygen species mediate cellular damage in Alzheimer dis-
Lei ZM, Rao CV, Kornyei JL, Licht P, Hiatt ES. 1993. Novel ease. J Alzheimers Dis 1: 45-55.
expression of human chorionic gonadotropin/luteinizing Raina AK, Zhu X, Rottkamp CA, Monteiro M, Takeda A, et al.
hormone receptor gene in brain. Endocrinology 132: 2000. Cyclin toward dementia: Cell cycle abnormalities
2262-2270. and abortive oncogenesis in Alzheimer disease. J Neurosci
Li J, Xu M, Zhou H, Ma J, Potter H. 1997. Alzheimer pre- Res 61: 128-133.
senilins in the nuclear membrane, interphase kinetochores, Rapp SR, Espeland MA, Shumaker SA, Henderson VW,
and centrosomes suggest a role in chromosome segregation. Brunner RL, et al. 2003. Effect of estrogen plus progestin
Cell 90: 917-927. on global cognitive function in postmenopausal women:
Parallels between neurodevelopment and neurodegeneration: A case study of Alzheimers disease
7 155
The Womens Health Initiative memory study: A rando- Tang MX, Jacobs D, Stern Y, Marder K, Schofield P, et al. 1996.
mized controlled trial. JAMA 289: 2663-2672. Effect of oestrogen during menopause on risk and age at
Resnick SM, Henderson VW. 2002. Hormone therapy and risk onset of Alzheimers disease. Lancet 348: 429-432.
of Alzheimer disease: A critical time. JAMA 288: 2170-2172. Tham A, Nordberg A, Grissom FE, CarlssonSkwirut C,
Rocca WA, Hofman A, Brayne C, Breteler MM, Clarke M, Viitanen M, et al. 1993. Insulinlike growth factors
et al. 1991. Frequency and distribution of Alzheimers disease and insulinlike growth factor binding proteins in cerebro-
in Europe: A collaborative study of 19801990 prevalence spinal fluid and serum of patients with dementia of the
findings. The EURODEMPrevalence Research Group. Alzheimer type. J Neural Transm Park Dis Dement Sect 5:
Ann Neurol 30: 381-390. 165-176.
Schmitz A, Tikkanen R, Kirfel G, Herzog V. 2002. The Vincent I, Jicha G, Rosado M, Dickson DW. 1997. Aberrant
biological role of the Alzheimer amyloid precursor protein expression of mitotic cdc2/cyclin B1 kinase in degenerating
in epithelial cells. Histochem Cell Biol 117: 171-180. neurons of Alzheimers disease brain. J Neurosci 17:
Schubert D, Cole G, Saitoh T, Oltersdorf T. 1989. Amyloid 3588-3598.
beta protein precursor is a mitogen. Biochem Biophys Res Vincent I, Zheng JH, Dickson DW, Kress Y, Davies P. 1998.
Commun 162: 83-88. Mitotic phosphoepitopes precede paired helical filaments
Schupf N, Kapell D, Nightingale B, Rodriguez A, Tycko B, in Alzheimers disease. Neurobiol Aging 19: 287-296.
et al. 1998. Earlier onset of Alzheimers disease in men with Willert K, Nusse R. 1998. Betacatenin: A key mediator of
Down syndrome. Neurology 50: 991-995. Wnt signaling. Curr Opin Genet Dev 8: 95-102.
Selkoe D, Kopan R. 2003. Notch and Presenilin: Regulated Wong PC, Zheng H, Chen H, Becher MW, Sirinathsinghji DJ,
intramembrane proteolysis links development and degen- et al. 1997. Presenilin 1 is required for Notch1 and
eration. Annu Rev Neurosci 26: 565-597. DII1 expression in the paraxial mesoderm. Nature 387:
Shen J, Bronson RT, Chen DF, Xia W, Selkoe DJ, et al. 1997. 288-292.
Skeletal and CNS defects in Presenilin1deficient mice. Yang Y, Geldmacher DS, Herrup K. 2001. DNA replication
Cell 89: 629-639. precedes neuronal cell death in Alzheimers disease. J Neu-
Short RA, Bowen RL, OBrien PC, GraffRadford NR. 2001. rosci 21: 2661-2668.
Elevated gonadotropin levels in patients with Alzheimer Yang Y, Mufson EJ, Herrup K. 2003. Neuronal cell death is
disease. Mayo Clin Proc 76: 906-909. preceded by cell cycle events at all stages of Alzheimers
Smith TW, Lippa CF. 1995. Ki67 immunoreactivity in disease. J Neurosci 23: 2557-2563.
Alzheimers disease and other neurodegenerative disorders. Yuasa S, Nakajima M, Aizawa H, Sahara N, Koizumi K, et al.
J Neuropathol Exp Neurol 54: 297-303. 2002. Impaired cell cycle control of neuronal precursor cells
Smith MA, Perry G, Atwood CS, Bowen RL. 2003. Estrogen in the neocortical primordium of presenilin1deficient
replacement and risk of Alzheimer disease. JAMA 289: 1100. mice. J Neurosci Res 70: 501-513.
Soriano S, Kang DE, Fu M, Pestell R, Chevallier N, et al. 2001. Zhu X, Rottkamp CA, Raina AK, Brewer GJ, Ghanbari HA,
Presenilin 1 negatively regulates betacatenin/T cell factor/ et al. 2000. Neuronal CDK7 in hippocampus is related
lymphoid enhancer factor1 signaling independently of to aging and Alzheimer disease. Neurobiol Aging 21: 807-
betaamyloid precursor protein and notch processing. 813.
J Cell Biol 152: 785-794. Zhu X, Raina AK, Perry G, Smith MA. 2004a. Alzheimers
Stopa EG, Gonzalez AM, Chorsky R, Corona RJ, Alvarez J, disease: The twohit hypothesis. Lancet Neurol 3: 219-226.
et al. 1990. Basic fibroblast growth factor in Alzheimers Zhu X, McShea A, Harris PLR, Raina AK, Castellani RJ, et al.
disease. Biochem Biophys Res Commun 171: 690-696. 2004b. Elevated expression of a regulator of the G2/M
Styren SD, Mufson EJ, Styren GC, Civin WH, Rogers J. 1990. phase of the cell cycle, neuronal CIP1 associated regulator
Epidermal growth factor receptor expression in demented of cyclin B, in Alzheimers disease. J Neurosci Res 75:
and aged human brain. Brain Res 512: 347-352. 698-703.
8 Differentiation and
DeDifferentiationNeuronal
CellCycle Regulation During
Development and AgeRelated
Neurodegenerative Disorders
T. Arendt
3.2.2 Oxidative Stress and the Autocrine Loop of SelfPerpetuating Mitogenic Activation . . . . . . . . . . 186
3.2.3 The Replay of Developmental Programs in Alzheimers Disease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
3.2.4 CellCycle Activation in Alzheimers Disease and Other Neurodegenerative Disorders . . . . . . . . . . . . 190
4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 159
Abstract: The eukaryotic cell cycle is an evolutionary highly conserved process that results in cell division
into two daughter cells. Nondividing cells rest outside the cell cycle. The cell cycle consists of an ordered set
of events that are tightly regulated by defined temporal and spatial expression, subcellular localization, and
degradation of cell cycle regulators. The transition through the cell cycle is controlled by stop and go
signals at defined checkpoints. These checkpoints allow alternate decisions between further progression
through the cell cycle and growth arrest, in order to provide potentials for DNA repair or induction of
apoptosis. Postmitotic neurons in the CNS are generated from neuroepithelial stem cells, which are
multipotent cells that differentiate into progenitor cells of neurons and glial cells. Differentiated neurons
are postmitotic cells that have permanently withdrawn from the cell cycle. Currently accumulating
evidence, however, shows that differentiated neurons express a number of cell cycle regulators which
suggest a role of these proteins beyond cell cycle regulation. Under conditions of neurodegeneration in
Alzheimers disease, neurons leave their differentiated stage and re-enter the cell cycle. This process which is
driven by an abnormal activation of mitogenic pathways eventually results in cell death. Alzheimers disease
should, thus, be added to the list of proliferative disorders such as cancer, cardiovascular disease, infections
and autoimmune diseases, where cell cycle regulators are recognized targets for treatment. Here we attempt
to give an overview on the current state of knowledge on the regulation of the cell cycle in neurons under
physiological and pathophysiological conditions.
List of Abbreviations: AD, Alzheimers disease; ATP, adenosine triphosphate; APC, anaphase-promoting
complex; APP, amyloid precorsor protein; Akt/PKB, phosphoinositide 3-kinase/protein kinase B; ApoE,
apoliporotein E; Abl, v-abl Abelson murine leukemia viral oncogene homolog 1 [Homo sapiens]; A,
amyloid-beta protein; ATM, ataxia telangiectasia mutated kinase; ATR, ATM-Rad-3 related kinase; Bad,
Bcl-2-associated death protein; Bak, Bcl-2 agonist killer 1; Bax, Bcl-2-interactive cell death susceptibility
regulator; Bcl-2, B-cell leukemia/lymphoma-2; Bcl-x, B-cell lymphoma 2-associated protein X; bFGF, basic
fibroblast growth factor; Bid, BH3 interacting domain death agonist; Bim, Bcl-2-interacting mediator of cell
death; BH, Bcl-2 homology; Bmf, Bcl-2 modifying factor; Bok, Bcl-2 related ovarian killer protein; CAK,
cdk-activating kinase; Cdc, cell-division-control; Cdk, Cyclin dependent kinase; C/EBP, CCAAT/enhancer-
binding protein transcription factor; Cip/Kip family, Cdk interacting protein/kinase inhibitory protein
(Cip/Kip) family; c-myc, avian virus inducing Myelocytomatosis, c = cellular; CNS, central nervous system;
DA, Dalton; Dab, Drosophila disabled homolog; DNA, deoxyribonucleic acid; DP5/Hrk, death protein5/
harakiri; ECM, extracellular matrix; E2F, transcription factor; EGF, epidermal growth factor; EGL, external
germinal layer; ELISA, enzyme linked immunosorbent assay; Elk-1, member of ETS oncogene family (ETS
erythroblast transformation specific domain); ERK, extracellular signal-regulated kinase; ERK1, Mitogen-
activated protein kinase p44 extracellular signal-regulated kinase; ERK2, Mitogen-activated protein kinase
p42 extracellular signal-regulated kinase; Ets-2, v-ets erythroblastosis virus E26 oncogene homolog
2 (avian); FAK, focal adhesion kinase; FKBP12, FK506-binding protein; FTD, Fronto-temporal dementia-
17 (FTD linked to 17q21); FAS, tumor necrosis factor ligand family, apoptose related molecule (CD95 or
APO-1), encodes region of fatty acid synthase activity; FAS; multifunctional protein; GAP-43, growth-
associated protein 43; GAPs, GTPase-activating proteins; G0, G1, G2, gap phase 0, 1, 2; GDP, guanine
diphosphate; GEFs, guanine nucleotide exchange factors; GSK3, glycogen synthase kinase 3-beta; GTP,
guanine triphosphate; HCMV, human cytomegalovirus; HGF/SF, hepatocyte growth factor/scatter factor;
HIV, human immunodeficiency virus; H3, histone 3; HP1, heterochromatin protein 1; HPV, human
papilloma virus; HSV, herpes simplex virus; IL, interleukin; IGF-1, IGF-2, insulin-like growth factor-1, -2;
INK4, inhibitors of cyclin-dependent kinase 4; JNK, c-Jun NH2-terminal kinase; Ki-67, cell-cycle related
nuclear protein; LEF/TCF, lymphoid enhancer factor/T-cell factor; LTP, long term potentiation; M, mitosis;
MAPK, mitogen activated protein kinase; MARCKS, myristoylated alanine-rich C-kinase substrate; MCI,
Mild Cognitive Impairment; Myt1, myelin transcription factor 1; MEK, Map kinase extracellular signal-
regulated kinase; mTOR/FRAP, mammalian target of rapamycin/FKBP12 rapamycin associated protein;
Mcl-1, myeloid cell leukemia sequence 1, antiapoptotic protein (Bcl-2 homologue); NCAM, neural cell
adhesion molecule; NMDA receptor, N-methyl-D-aspartate receptor; NR2A, subunit of NMDA receptor;
NO, nitric oxide; NOS, NO synthase; nNOS, Type I or neuronal isoform of NOS; iNOS, Type II or inducible
isoform of NOS; eNOS, Type III or endothelial isoform of NOS; NGF, nerve growth factor; NF-kB, nuclear
160
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
factor kappa-B, subunit 1; OA, okadaic acid; PCNA, proliferating cell nuclear antigen; p53, tumor protein
p53; PHFs, paired helical filaments; pRb, phospho-retinoblastoma protein; PI3K, phosphatidylinositol
3-kinase; PKC, protein kinase C; PP1, protein phosphatase 1; PHAS-1, Phosphorylated heat- and acid-
stable protein 1; PLA-2, phospholipase A2; p70S6k, p70 ribosomal protein S6 kinase; PAK, p21-activated
kinase; p75, p75 neurotrophin receptor; PC12, pheochromocytoma cell line; Puma, p53-upregulated
modulator of apoptosis; PMAIP1/Noxa, phorbol-12-myristate-13-acetate-induced protein1/BH3-containing
mitochondrial protein; PDGF, platelet derived growth factor; R, restriction point of the cell cycle; rhoA, ras-
related protein A; Ras, rat sarcoma virus; Raf, proto-oncogene serine/threonine-protein kinase; Rsk,
ribosomal protein S6 kinase; S, DNA synthesis phase of cell cycle; SAA, serum amyloid; SSeCKS, Src-
suppressed protein kinase C substrate with metastasis suppressor activity; Skp2, S-phase kinase-associated
protein 2 (p45); SUV39H1, Histone methyltransferase Suv39h1; SOS, son of sevenless protein; SMADs,
Mothers against decapentaplegic homolog (vertebrates) (derived from MAD [Mothers against dpp
CG12399-PA in Drosophila melanogaster] and SMA [in C. elegans]; STAT3, signal transducers and
activators of transcription; SOD1G37R, Superoxide dismutase 1 (transgenic mice overexpressing mutant);
Spike, small protein with inherent killing effect; Src, v-src sarcoma (Schmidt-Ruppin A-2) viral oncogene
homolog (avian) [Homo sapiens]; TGF, transforming growth factor; t-TGase, tissue-transglutaminase; TNF,
tumor necrosis factor; TUNEL, TdT-mediated dUTP-biotin nick end labeling (TdT = terminal deoxynu-
cleotidyl transferase); TrkA, tropomyosin receptor kinase A; Wnt, Wnt oncogene analog, wingless type
The cell cycle is an evolutionary, conserved, and ordered set of events resulting in the cell dividing into two
daughter cells. Nondividing cells rest outside the cell cycle in G0. They account for the major part of
nonproliferating cells. The cell cycle consists of two major consecutive events, basically characterized by
DNA replication and segregation of replicated chromosomes into two separate cells. Corresponding phases
of the cell cycle are the synthesis (S) phase in which DNA replication occurs and the mitosis (M) phase
in which the cell undergoes division into two daughter cells. G0 and S are separated by gap1 (G1) phase in
which cells increase in size and produce RNA and protein to get ready for replication, whereas S and M are
separated by gap2 (G2) phase in which cells continue to grow and produce new protein for subsequent cell
division (> Figure 8-1).
The transition from one phase to another and, thus, the progression of the cell cycle is tightly
regulated by defined temporal and spatial expression, subcellular localization, and degradation of cell
cycle regulators.
. Figure 8-1
The phases of the cell cycle. The site of action of regulatory CDK/cyclin complexes is indicated
Roberts, 1994; Hunter and Pines, 1994; King et al., 1994; Sherr and Roberts, 1995; Poon et al., 1995; Rivard
et al., 1996; Coats et al., 1996; Morgan, 1997; Pavletich, 1999).
The initial segment of the cell cycle, the first gap (G1) phase, is the site of integration of mitotic signals that result
from liganddependent activation of both growth factor receptors and integrins (Miyamoto et al., 1996; Assoian,
1997; Giancotti, 1997; Lin et al., 1997; Renshaw et al., 1997; Moro et al., 1998; Roovers et al., 1999; Assoian and
Schwartz, 2001; Danen and Yamada, 2001), which converge on the activation of G1cyclindependent kinases
(Cdks) Cdk4/6 and Cdk2 (Sherr 1993, 1994). The resultant Cdk activities determine whether mitogenic signals
are propagated downstream inducing phosphorylation of key substrates required for progression through G1 and
entry into Sphase. The major function of G1phase kinases is to partially phosphorylate the retinoblastoma
protein (pRb), thereby contributing to the inactivation of its repressive function. Consequently, the E2F
transcription factor can initiate transcription of cyclin E. Second, the cyclin D family sequesters the Cdk
inhibitor p27Kip1 from the previously inactive cyclinE/Cdk2 complex (> Figure 8-3).
. Table 8-1 162
Functional implications of cyclindependent kinases and associated cyclins
8
Cell cycle Cell differentiation Transcription Specific neuronal functions Cell death Other functions
CDK1 cyclinB prophase to metaphase amyloid
transition regulation of induced
topoisomerase 2 cytotoxicity
CDK2 cyclinA G1/S transition S phase regulation of Sp1 apoptosis in
cyclinE G2 phase centrosome mediated thymocytes,
duplication transcription mesangial cells
DNA damage
induced
apoptosis
CDK3 cyclinE G1/S transition
CDK4 cyclin D G1phase excitotoxin
induced neuronal
death
CDK5 p35, lens cell neurite outgrowth apoptosis Golgi membrane
p25 differentiation neurone migration traffic Insulin
P39, myogenesis induction of acetylcholine exocytosis by
p29 receptors dopamine and cells
glutamate signaling phototransduction
neurotransmitter release
CDK6 cyclin D G1phase neuronal cell
death
CDK7 cyclin H Cdk1, Cdk2 activation basal
transcription
CDK8 cyclin C regulation of basal
CDK7cyclin H transcription
CDK9 cyclinK MyoDmediated RNA signal
cyclinT myocyte transcription transduction
differentiation
CDK11 cyclin L RNA processing dopamine and glutamate apoptosis
or transcription signaling
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
. Figure 8-2
In proliferating cells, the temporal association of Cdks to cyclins is tightly regulated to control the progression
of the cell cycle (modified after Nguyen et al., 2002)
When quiescent cells enter the G1 phase, genes encoding Dtype cyclins (D1, D2, and D3) are induced
in response to mitogenic signals in conjunction with the appropriate external stimuli such as cell adhesion.
During progression through G1, the level of Dtype cyclins increases, and, in a mitogenregulated manner,
these proteins associate with, and activate their catalytic partners, Cdk4 or Cdk6 (Sherr, 1993; Matsushime
et al., 1994; Vidal and Koff, 2000). The major function of Dtype cyclins is to provide a link between the
mitogenic stimulation and the autonomously progressing cellcycle machinery. Unlike the other cyclins,
cyclin D is not expressed periodically, but is synthesized as long as mitogenic stimulation persists (Assoian and
Zhu, 1997) and, thus, is usually absent from the cell cycle that progresses independent of the presence of
mitogenic signals. Conversely, constitutive activation of cyclin D can contribute to oncogenic cell transforma-
tion. Cyclin D1 mRNA translation can be inhibited by rapamycin, a mTOR/FRAP (target of rapamycin/FkBP12
rapamycinassociatedprotein) inhibitor. In contact inhibited cells, cyclin D is sequestered to the cytoplasm by
binding to SSeCKS/gravin protein, a major PKC substrate with tumor suppressor activity (Lin et al., 2000).
The first wave of cyclin Ddependent kinase activity is followed in late G1 by an increase in cyclin E,
which associates with Cdk2 (Dulic et al., 1992; Koff et al., 1992). The Cdk2/cyclin E complex is not only
responsible for the G1/S transition but also regulates centrosome duplication. Unlike the Dtype cyclin
dependent kinases, assembly of cyclin E and Cdk2 into active kinase is not mitogendependent. Cdk2
reinforces Cdk4 to complete Rb phosphorylation and induces the degradation of p27Kip1. The cell cycle is
now irreversibly committed to enter the S phase. CyclinE/Cdk2 phosphorylates a variety of substrates,
thereby contributing to relieve the Rbmediated repression. Cyclin E targets components necessary for DNA
synthesis, contributing to the assembly of the DNA replication complexes. Cyclin E, like cyclin D, is targeted
for degradation by phosphorylation. The active cyclin ECdk2 complex itself phosphorylates cyclin E,
which is subsequently recognized by the ubiquitin ligase SCF (Skp2) and targeted for degradation.
Proteolysis of cyclin E marks the transition into the S phase (Won and Reed, 1996).
Assembly of cyclin A with Cdk2 is required during the S phase. The Cdk2/cyclin A complex phosphor-
ylates various substrates allowing DNA replication. At the S/G2 transition, cyclin A associates with Cdk1.
In late G2 phase, cyclin B appears, complexes with Cdk1, and triggers the G2/M transition, where cyclin
A is degraded (Peters, 1998). This resets the system and reestablishes the requirement for mitogenic cues to
induce Dtype cyclins for the next cycle.
Taken together, cyclins can influence when and where the Cdks are active. They are more than a
conformational on/off switch for Cdk activity and influence Cdk activity by mediating critical interactions
between cyclin/Cdk complexes and cellular proteins such as substrates, inhibitors, and activators.
164
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
. Figure 8-3
The cell division cycle and its major regulatory elements
by the Cdk7cyclin H complex, also called Cdkactivating kinase (CAK) (Lew and Kornbluth, 1996). These
phosphorylations induce conformational changes and further enhance binding of cyclins (Jeffrey et al.,
1995).
Activity of Cdks is negatively regulated by Cdkinhibitors, which bind directly to Cdks or to Cdk/cyclin
complexes (Biggs and Kraft, 1995; CordonCardo, 1995; Poon et al., 1995; Sherr and Roberts, 1995; Boice
and Fairman, 1996; Coats et al., 1996; Guan et al., 1996; Rivard et al., 1996). Based on structural features,
Cdk inhibitors are grouped into two families, the inhibitors of Cdk4 (INK4) family and the Cip/Kip family.
The INK4 family currently consists of four members, p16INK4a, p15INK4b, p18INK4c, and p19INK4d, which
commonly share the presence of ankyrin repeats (Serrano et al., 1993; Hannon and Beach, 1994; Guan et al.,
1994; Chan et al., 1995; Hirai et al., 1995). The INK4A locus encodes two independent but overlapping
genes, p16INK4a and p19ARF, the mouse homolog of p14ARF, which act in partly overlapping pathways
(Carnero et al., 2000). All these proteins can inhibit cyclin D associated kinases, Cdk4 and Cdk6. INK4
proteins compete with Dtype cyclins for binding to the Cdk subunit (McConnell et al., 1999; Parry et al.,
1999). The inhibition of cyclin binding is mediated by a conformational change in the kinase that prevents
association of the cyclins and distorts the ATP binding site (Russo et al., 1998). The INK4family of cyclin
dependent kinase inhibitors might be involved in the regulation of pathways that control cell growth and
proliferation as well as cell death. Deregulation of these Cdki proteins results in either uncontrolled
proliferation and neoplastic transformation or activation of apoptosis (Freeman et al., 1994; Kranenburg
et al., 1996; Liu et al., 1996a). The inhibitory action of the INK4 proteins is largely dependent on the
presence of pRb in the cell.
Three proteins currently comprise the following Cdk interacting protein/kinase inhibitory protein (Cip/
Kip) family: p21Cip1, p27Kip1, and p57Kip2 (Gu et al., 1993; Harper et al., 1993; Polyak et al., 1994a, b; Lee
et al., 1995a). These proteins share a homologous inhibitory domain, which is necessary for binding to
preformed complexes of cyclinCdks, thereby directly inhibiting the kinases. Compared to the INK4family,
they have a wider inhibiting specificity, affecting the activities of cyclin D, E, and Adependent kinases. In vivo
they preferentially act on Cdk2 complexes. Through binding to cyclin D/Cdk4 complexes, p27Kip1 and p21Cip1
can be sequestered without inhibiting the activity of the complex (Sherr and Roberts, 1995). Kip/Cip
proteins might even be able to promote activation of cyclin D/Cdk complexes by stabilizing the complexes
and directing their nuclear translocation (LaBaer et al., 1997; Cheng et al., 1999). When mitogens are
withdrawn and cyclin D is downregulated, this pool of Kip/Cip proteins is released, thereby inducing G1
phase arrest through inhibition of cyclin E/Cdk2. P21cip1 also associates with proliferating cell nuclear
antigen (PCNA), an essential binding partner of DNApolymerase delta, which might be an additional
mechanism through which p21Cip1 can inhibit DNA synthesis (Zhang et al., 1993; Luo et al., 1995).
In addition to binding to Cdks, p21Cip1 and p57Kip2 also bind to the PCNA (Gulbis et al., 1996). PCNA
is involved in DNA replication and binding of p21Cip1 or p57Kip2, prevents PCNAmediated DNA synthesis,
and arrests the cell cycle (Watanabe et al., 1998).
Cdkis are regulated by both external and internal signals. The expression of p21Cip1 is under the
transcriptional control of the p53 tumor suppressor gene (el Deiry et al., 1993). Expression and activation
of p15INK4b and p27Kip1 is increased by TGF, contributing to growth arrest (Reynisdottir et al., 1995).
Both p21Cip1 and p27Kip1 undergo phosphorylation, which determines their subcellular location. Recogni-
tion of the phosphorylated form of p27Kip1 by the ubiquitin ligase SCF (Skp2) targets p27Kip1 for
degradation.
Correct cellcycle progression is also regulated by the subcellular sorting of cellcycle regulating proteins.
Cyclin B contains a nuclear exclusion signal and is actively exported from the nucleus until the beginning of
166
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
the prophase. The Cdk inactivating kinases Wee1 and Myt1 are located in the nucleus and Golgi complex
and protect the cell from premature mitosis (Liu et al., 1997). Subcellular sorting of different proteins is also
regulated through sequestration by binding to the 1433 protein (Peng et al., 1997; Yang et al., 1999).
Many regulating proteins are subject to rapid degradation during the cell cycle. Proteins are irrevers-
ibly destroyed, which permits the orderly transition through the cycle, based on the completion of one
phase before entry into the next phase. The specific degradation of many proteins, including Cdks, cyclins,
and Cdk inhibitors, is achieved by targeted ubiquitination (Murray, 1995). Critical for the initiation of
mitosis and cellcycle exit is a specific E3type ubiquitin ligase, referred to as the anaphasepromoting
complex (APC) or the cyclosome (Peters et al., 1996). APC selectively targets proteins for ubiquitination
that contain a specific consensus sequence, which is found in cyclins and other mitotic proteins (Glotzer
et al., 1991).
Regulation of a Cdk that is not related to cellcycle progression, such as Cdk5, differs markedly from that of
cellcycle Cdks (Nguyen et al., 2002). The association of Cdk5 with one of its neuronspecific coactivators,
p35 or p39, is required for neurite outgrowth, axonal migration, cortical lamination, control of cell
adhesion, axonal transport, synaptic activity, neuronal adaptive changes, and motor functions (Dhavan
and Tsai, 2001). Inhibition of Cdk5 or expression of a dominant negative form of the kinase in cultured
neurons prevents neurite outgrowth, whereas overexpression of p35/Cdk5 induces the formation of longer
neurites (Xiong et al., 1997; Nikolic et al., 1998). P35null mice display a loss of stratified cellular
organization (Chae et al., 1997; Kwon and Tsai, 1998), and an even more severe disruption of embryonic
cortical layering is observed in mice lacking Cdk5 (Ohshima et al., 1996). The cortical phenotypes of Cdk5,
p35, or p39null mice resemble a severe developmental disorder in humans, referred to as type 1
lissencephaly (Reiner, 2000). P35/Cdk5 is also involved in associative learning (Fischer et al., 2002). The
role of Cdk5 in plastic neuronal changes involves control of presynaptic function by phosphorylation of
Munc18, amphysin, and synapsin I (Dhavan and Tsai, 2001). Cdk5 also regulates neurotransmission and
longterm potentiation by phosphorylating the NR2A subunit of NMDA receptors (Li et al., 2001). The
involvement of Cdk5 in neurodegeneration has been suggested by reports of Cdk5 expression during
apoptosis and by the presence of the kinase in neurofibrillary tangles in AD and frontotemporal dementia
17 (FTD) (Yamaguchi et al., 1996; Pei et al., 1998). An abnormal localization of Cdk5 is also found in canine
motor neuron disease, Parkinsons disease, and amyotrophic lateral sclerosis (Nakamura et al., 1997; Bajaj
et al., 1998; Green et al., 1998).
Rbrelated members of the pRb pocket protein gene family. They are also substrates of the G1phase
Cdks, share structural and biochemical properties, and, like pRb, bind and modulate the activity of E2F
transcription factors (Ewen, 1998; Ferreira et al., 1998; Mulligan and Jacks, 1998). P107 and p130 are
structurally more closely related to each other than to Rb, and are thought to function redundantly in G0
and G1 to repress E2Fresponsive genes that are distinct from those regulated by Rb.
The E2F transcription factor consists of a heterodimer between E2F and DP family proteins. Six
members of the E2F family (E2F1 to E2F6) and three DPs (1, 2, 3) have been identified so far (Nervins,
1998). E2F1, E2F2, and E2F3 are structurally similar and form complexes exclusively with the retinoblas-
toma protein, whereas E2F4 and E2F5 interact primarily with p107 and p130. E2F1, E2F2, and E2F3 but
not E2F4 and E2F5 contain cyclinbinding motifs, which promote stable binding of cyclin A/Cdk2. E2F6
lacks both pocket proteinbinding and cyclinbinding sequences and acts as a transcriptional repressor for
E2Fresponsive genes (Morkel et al., 1997).
Rbmediated repression occurs through at least three mechanisms (Brehm et al., 1998; MagnaghiJaulin
et al., 1998; Luo et al., 1998; Swanton, 2004). First, Rb induces modification of histones in the nucleosome
by recruiting the histone deacetylase to E2F sites. Histone deacetylation causes tighter association between
DNA and nucleosomes and restricts access of the transcriptional apparatus to DNA, thereby repressing
transcription. Second, Rb regulates nucleosome structure by recruiting ATPdependent remodeling com-
plexes to promoters. Third, Rb influences chromatine structure by recruiting the SUV39H1 methylase and
HP1 to the cyclin E promoter leading to histone H3 methylation and silencing of cyclin E.
Cdkdependent phosphorylation of Rb (pRb) disrupts its association with E2F family members, relieves
pRbmediated repression, and allows E2Fdependent transcription of several genes involved in cell cycle
and growth regulation. Promoters of many genes have been found to contain consensus E2F binding sites,
including enzymes of DNA metabolism, protooncogenes, and cellcycle regulatory proteins such as cyclins
E and A both of which are required to catalyze the G1/S transition (Sherr, 1994; DeGregori et al., 1995;
Schulze et al., 1995; Weinberg, 1995; Dyson, 1998; Nevins, 1998; Black et al., 1999; Cheng et al., 1999).
Hypophosphorylated pRb also binds to other proteins, such as members of the Abl family, which seem to be
closely implicated in adhesiondependent growth control. The E2Fdependent induction of cyclin E, which
in turn stimulates Rb phosphorylation through Cdk2, provides a feedback loop that contributes to the G1/S
transition. This event of pRb phosphorylation and release of E2F corresponds to progression through the
restriction point R (Pardee, 1974).
In addition to regulation of E2F activity through phosphorylation of pocket proteins, direct phosphor-
ylation of E2F also influences its activity. Both cyclin A/Cdk2 and cyclin A/Cdc2 can phosphorylate E2F1
on multiple sites (Mudryj et al., 1991; Kitagawa et al., 1995; Adams et al., 1996; Dynlacht et al., 1997), which
might contribute to its downregulation in late S/G2 phase. Besides its involvement in cell growth and cell
cycle regulation, E2F also plays a role in apoptosis. Similar to growth regulation, E2Fmediated apoptosis is
regulated by pocket proteins (Shan et al., 1996).
Other Cdk substrates include histone H1 (Bradbury et al., 1974), DNA polymerase alpha primase
(Voitenleitner et al., 1997), their own regulators Wee1 and Cdc25, and cytoskeletal proteins such as lamins,
microtubules, and vimentin, which are required for correct mitosis (Hoffmann et al., 1993; Blangy et al., 1995).
leads to genomic instability. These checkpoints allow to alternate decisions between further progression
through the cycle and growth arrest, in order to provide potentials for DNA repair or induction of
apoptosis. Apoptosis as an alternative to cellcycle progression is thus, aimed at preventing cell prolifera-
tion. In proliferating cells, this might be a protective mechanism against transmission of nonreparable
DNA. Depending on the chemical nature of the damage on DNA, different types of repair programs and
cellular responses are evoked (Laiho and Latonen, 2003). In highly differentiated cells, such as neurons, this
mechanism might be preserved to prevent inappropriate generation of new cells that cannot easily be
integrated into functional circuitry (Copani et al., 2001; Liu and Greene, 2001). Thus, an apoptotic
response might be inherent to an inappropriate activation of the cell cycle (Evan et al., 1995; Evan and
Littlewood, 1998).
DNA quality controls, i.e., damage checkpoints, are positioned before DNA replication, at the G1S
transition, and prior to mitosis, at the G2M transition. Further checkpoints are positioned during the S
and M phase. Cellcycle arrest induced by DNA damage at the G1S checkpoint is p53dependent. DNA
damage activates the product of the ataxiateleangectasia gene (ATM/ATR) and DNAdependent protein
kinase family, which leads to posttranslational stabilization of a normally labile p53 protein (Giaccia and
Kastan, 1998) (> Section 2.3.3).
Sphase checkpoints control initiation and elongation phase of DNA replication (Paulovich and Hartwell,
1995). Progression of the mammalian cell cycle requires the activity of Cdc25 family of phosphatases,
which dephosphorylate the inhibitory phosphorylations of Cdks and, thus, enable Cdk activities
required for S phase and mitosis. A G2 checkpoint prevents entry into mitosis by Cdk1 inhibition
through phosphorylation or sequestration outside the nucleus. The G2M checkpoint also involves p53,
which induces the Cdk inhibitor p21Cip1 as well as a member of the 1433 family, which sequesters cyclin
B outside the nucleus (Hermeking et al., 1997). In mitosis, the errorfree segregation of sister chromatides is
essential for division. In response to unreplicated or damaged DNA, the cell arrests at a DNA structure
checkpoint in mitosis. The spindle assembly checkpoint induces mitotic arrest in case of improper
alignments of chromosomes or defects in microtubule attachment (Amon, 1999). A further checkpoint,
dependent on both p53 and pRb, acts subsequent to mitotic errors to block proliferation of cells that have
entered G1 with tetraploid status (Borel et al., 2002).
Extracellular physiological signals affect the decision to transit the restriction point and this decision is
mediated through phosphorylation of pRb. This means, extracellular signals directly control phosphoryla-
tion of pRb. Mitogendependent induction of these mechanisms is often mediated through the RasMAP
kinase pathway (Filmus et al., 1994; Liu et al., 1995).
Small GTPases of the Ras family, which act as molecular switches in intracellular signaling (Bourne
et al., 1990), regulate cellular proliferation and mediate the mitogenic response of a variety of growth factors
(Mulcahy et al., 1985). In mammals, the Ras superfamily of GTPases contains more than 65 members.
According to their function and structure, this superfamily has been subdivided into six families repre-
sented by Ras, Rho, Rab, Ran, Rad, and Arf (Barbacid, 1987; Valencia, 1991). In mammals, three genes
encode the ras family: Hras, Kras, and Nras.
The characteristic of Ras function is its regulated transit between an inactive state, where it is bound to
guanine dinucleotides (GDPs), and an active state, bound to guanine trinucleotides (GTPs). The GDP/GTP
exchange is regulated by guanine nucleotide exchange factors (GEFs), which catalyze the GDP/GTP
exchange, and GTPaseactivating proteins (GAPs), which enhance the intrinsic capacity of Ras to hydrolyze
GTP into GDP, thereby returning Ras to the inactive state (Lowy and Willumsem, 1993; Schlessinger, 2000).
To be functional, Ras proteins must be recruited to the inner side of the plasma membrane. Upon ligand
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 169
binding, growth factor receptors dimerize, autophosphorylate tyrosine residues in their own cytoplasmic
domain, and recruit intracellular signaling molecules to the membrane. Among these recruited molecules
are GEFs for Ras protein, which promote binding to GTP (Schlessinger, 2000). Through conformational
changes that occur upon the exchange of GDP for GTP, the GTPbound form of Ras interacts with Raf,
phosphatidylinositol 3kinase (PI3K), and RalGdS, driving intracellular signaling cascades (> Figure 8-4).
Activated Raf stimulates MAPkinase extracellular signalregulated kinase (MEK) (Kyriakis et al., 1992),
which in turn activates mitogenactivated protein kinases p44 extracellular signalregulated kinase (ERK1)
and p42 (ERK2) (Gomez and Cohen, 1991). ERKs have a wide spectrum of substrates including cytoplas-
mic proteins such as SOS, MEK, Rsk, PHAS1, or PLA2, and nuclear proteins that are phosphorylated upon
ERK translocation to the nucleus, mainly transcription factors such as Elk1, Ets2, C/EBP, and SMADs
(Whitmarsh and Davis, 1996). The main substrates of PI3K are Akt/PKB and p70 ribosomal protein S6
kinase (p70S6K) (Downward, 1998). Activation of Akt generates an antiapoptotic signal and regulates
glycogen synthase kinase 3b (GSK3) among other substrates.
MAPkinase and PI3K pathways activate transcription factors to induce numerous genes, including
those of cyclin D1 (Hill and Treisman, 1999). The promoter of the cyclin D1 gene contains mitogen
responsive Ets and AP1 binding sites (Albanese et al., 1995), known to respond to Ras signaling (Galang
and Hauser, 1994). Ras signaling promotes nuclear location of cyclin D1 by negatively regulating GSK3,
which phosphorylates cyclin D1. This modification redistributes cyclin D1 to the cytoplasmic compartment
and labels cyclin D1 for degradation by the ubiquitinproteosome (Diehl et al., 1998). Ras also regulates the
cyclin D assembly. In quiescent cells, stimulated with growth factors, Ras signaling promotes transition
from G0 to G1 and from G1 to S. Proliferating cells make an Rasdependent commitment for completion of
the next cycle while they are in G2 phase of the preceding cycle (Hitomi and Stacey, 2001). Thus, Ras
activity during G2 phase induces cyclin D1 expression, which apparently persists independently of Ras until
the beginning of the next S phase (Sa et al., 2002).
Ras also controls cellcycle progression through control of the expression of other cyclins such as D3, A,
and E, and members of the E2F family of transcription factors (Fan and Bertino, 1997) as well as through
downregulation of p27Kip1 (Aktas et al., 1997).
Mutations conferring constitutive Ras activation are found in nearly 30% of human tumors (Bos,
1989). Oncogenic Ras promotes uncontrolled mitogenesis independently of the presence of growth factors
(Feramisco et al., 1984) but is unable to transform cells. Constitutively activated Ras, moreover, not only
regulates cellular proliferation, but also renders cells susceptible to apoptosis (Chen and Faller, 1995;
Gulbins et al., 1995; Mayo et al., 1997; Serrano et al., 1997; Downward, 1998; Joneson and BarSagi,
1999; Frame and Balmain, 2000; Liou et al., 2000; Khokhlatchev et al., 2002).
Depending on the activation of other signaling pathways, activation of Ras signaling pathways can also
arrest the cell cycle rather than activate proliferation (Kohl and Ruley, 1987; Ridley et al., 1988; Marshall,
1995; Serrano et al., 1997; Sewing et al., 1997; Woods et al., 1997). As a result, Ras suppresses oncogenic
transformation and induces a condition phenotypically identical to premature cellular senescence (Serrano
et al., 1997; Lin et al., 1998; Hahn et al., 1999). It selectively inhibits genes involved in mitosis, DNA
replication, segregation, and repair, and at the same time upregulates genes associated with cell death and
potentially involved in neurodegenerative disorders and tumorigenesis such as APP, serum amyloid (SAA),
and tissue transglutaminase (tTGase) (Chang et al., 2000). This process involves the upregulation of the
cyclindependentkinase inhibitors p15INK4b, p16INK4a, p21Waf1/Cip1, p19ARF, and p53 (Serrano et al., 1997;
Olson et al., 1998; Palmero et al., 1998; Ferbeyre et al., 2000; Fogal et al., 2000; Malumbres et al., 2000; Guo
et al., 2000a; Pearson et al., 2000; Zhong et al., 2000a, b; Gottifredi and Prives, 2001) and, thus, resembles
changes seen in early AD (Gartner et al., 1995, 1999; Arendt et al., 1996, 1998d, e). Activation of the p21Ras
cascade, thus, plays an essential role in oncogenesis and cellular senescence as well as AD (> Section 3.2.1).
Cellular effects of Ras activation are strictly cell type dependent and show pleiotropic features,
depending on the activity level of Ras signaling (Crespo and Leon, 2000). Whereas high levels of Raf
activity are able to induce cellular senescence, less intense Raf activation induces cyclin D1 expression and
drives cells into proliferation (Woods et al., 1997).
170
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
Activation of the cell cycle through growth factors also requires signals by the extracellular matrix. Entry
into G1 phase and further progression toward the S phase is a major downstream event of synergistic
signaling by mitogenic compounds and integrindependent cell adhesion. All of the important mitogenic
signaling cascades downstream of the Ras and Rho family small GTPases and the PI3kinasePKB/Akt
pathway are regulated by integrin mediated cell adhesion (Danen and Yamada, 2001) (> Figure 8-4). They
control critical molecular switches such as induction of cyclin D1 resulting in activation of Cdk4/6, the
suppression of p21Cip1 and p27Kip1 inducing Cdk2 activity (Fang et al., 1996; Schulze et al., 1996; Stromblad
et al., 1996; Zhu et al., 1996; Assoian, 1997; Resnitzky, 1997; Wu and Schonthal, 1997) and the subsequent
phosphorylation of Rb (Weinberg, 1995; Sherr and Roberts, 1999). Some of these matrixrelated effects
involve matrixdependent organization of the cytoskeleton. Adhesiondependent cellcycle progression
involves both integrin binding to the ECM and integrinmediated cytoskeletal organization (Hansen
et al., 1994; Assoian and Zhu, 1997; Chen et al., 1997). An organized cytoskeleton is, thus, a requirement
throughout the mitogendependent portion of the G1phase; and the adhesiondependent expression of
cyclin D and the phosphorylation of pRb are mediated by the cytoskeleton (Bohmer et al., 1996; Assoian
and Zhu, 1997). Cell adhesion to extracellular matrix (ECM) is mediated by binding to the cell surface
integrin receptors, which activate intracellular signaling cascades and mediate tensiondependent changes
in cell shape and cytoskeletal structure. Although growth control has focused on integrin and growth factor
signaling, the cell shape might play an equally critical role in cellcycle progression, which acts by
subjugating the molecular machinery that regulates the G1/S transition (Huang et al., 1998).
Adhesion to substratum has two separate effects on cells both of which could underlie the anchorage
requirement for cellcycle progression, i.e., the initial adhesion event and the subsequent clustering of
occupied integrins. Integrins might act as mechanoreceptors, transmitting mechanical information from
the extracellular matrix to the cytoskeleton (Wang et al., 1993; Hansen et al., 1994). Alternatively, an
organized cytoskeleton might be required to force integrins to remain clustered at focal contacts, and
integrin clustering seems to be a prerequisite for integrin signaling.
Signaling through Ras and integrin regulates cyclin D/Cdk4 and cylin E/Cdk2 at different levels. (1) Expression
of cyclin D1, often the rate/limiting step in the activation of cylin D/Cdk4/6, critically depends on synergistic
. Figure 8-4
Synopsis of current models of information flow (light gray) from cell surface events (dark gray) involved in
intercellular communication to gene expression. These pathways mediate structural plasticity and dynamic
regulation of intercellular connectivity as well as cellcycle activation and cellular proliferation (speckled). One
of the major downstream targets is cyclin D1, which is under positive control of bcatenin/LEF/TCF, RhoA, PAK,
JNK, and ERK1/2. RhoA also represses p21Cip1 and p27Kip1, inhibitors of Cdk2. Thus cyclin D/Cdk4/6 and cyclin E/
Cdk2 become active and phosphorylate E2Fpocket protein complexes (E2FpRb and E2Fp107), allowing for
E2Fdependent expression of cyclin A. This phosphorylation of pRb correlates with cellcycle progression out of
the growth factor dependent portion of the G1 phase. At multiple levels, these pathways are also potentially
involved in t phosphorylation (Cdk5, GSK3, JNK, ERK1/2) (hatched). (The individual regulatory relationships
depicted in this cartoon have been demonstrated in various published reports using different types of cells. These
signaling pathways, however, have neither been demonstrated in their entirety within a single experimental
system, nor is their action firmly established for neurons; for reference see: Assoian and Zhu, 1977; BenZeev, 1977;
Giancotti, 1997; Yamada and Geiger, 1997; Howe et al., 1998; Schlaepfer and Hunter, 1998; Schlaepfer et al., 1999;
Schoenwaelder and Burridge, 1999; Schwartz and Baron, 1999; Novak and Dedhar, 1999; Orford et al., 1999; Aplin
et al., 1999; Dedhar et al., 1999; Patapoutian and Reichert, 2000; Petit and Thiery, 2000; Dedhar, 2000; Welsh and
Assioan, 2000; Assoian and Schwartz, 2001; Baki et al., 2001) Modified after Arendt (2003)
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
. Figure 8-4 (continued)
8 171
172
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
signaling through the Ras/MAPK cascade, integrin, cadherin, and Wnt dependent signaling (Albanese et al.,
1995; Lavoie et al., 1996; Miyamoto et al., 1996; Sherr, 1996; Winston et al., 1996; Zhu et al., 1996; Aktas et al.,
1997; Kerkhoff and Rapp, 1997; Lin et al., 1997; Renshaw et al., 1997; Weber et al., 1997; Cheng et al., 1998;
Huang et al., 1998; Short et al., 1998; Aplin et al., 1999; Gille and Downward, 1999; Roovers et al., 1999; Roovers
and Assoian, 2000) (> Figure 8-4). (2) Assembly of newly synthesized cyclin D1 with Cdk4 similarly
depends on the Ras/MAPK cascade (Peeper et al., 1997; Cheng et al., 1998). (3) Turnover of Dtype cyclins,
moreover, is Rasdependent and mediated by the PI3kinaseAktpathway, which regulates the phosphory-
lation of cyclin D1 by GSK3b (Diehl et al., 1998). Inhibition of this pathway leads to cyclin D phosphory-
lation, enhancing its nuclear export and the ubiquitindependent proteasomal degradation. (4) Cell
adhesion is a prerequisite for efficiently downregulating the steadystate levels of p21Cip1 and p27Kip1
(Fang et al., 1996; Schulze et al., 1996; Zhu et al., 1996; Welsh and Assoian, 2000; Welsh et al., 2001).
P27Kip1 was first identified as a Cdk2inhibitory factor detected in contactinhibited cells (Koff et al., 1993;
Polyak et al., 1994a, b) and many factors, including cellcell contact, induce accumulation of p27Kip1
(Hengst and Reed, 1998). Under these conditions, p19INK4d and p27Kip1 cooperate to maintain differen-
tiated neurons in a quiescent state (Zindy et al., 1999). Adhesion lowers the levels of p21Cip1 and p27Kip1,
thus permitting activation of cyclin E/Cdk2 (Fang et al., 1996; Zhu et al., 1996) whereas high levels of
p27Kip1 persist in cells that were prevented from spreading by restriction of the size of the substratum,
despite normal MAPK activation (Huang et al., 1998).
Key intermediates of signaling of integrindependent cell adhesion upon cellcycle regulation are the
integrinlinked kinase (Radeva et al., 1997), which, through inhibition of GSK3 (Delcommenne et al.,
1998; Troussard et al., 1999), can also regulate both the expression and degradation of cyclin D1 (Diehl
et al., 1998) and Focal adhesion kinase (FAK), which controls cellcycle progression through G1 via JNK
(Oktay et al., 1999) and through RasERK1/2 (Chan et al., 1994; Cobb et al., 1994; Schaller et al., 1994;
Schlaepfer et al., 1994; Xing et al., 1994; Schlaepfer and Hunter, 1996; Zhao et al., 1998; Sastry et al., 1999;
Barberis et al., 2000; Assoian and Schwartz, 2001) (> Figure 8-4).
Downstream G1 events, including phosphorylation of pRb leading to release of E2Ftranscription
factors from their complex with Rb and cyclin A expression also require both soluble mitogens and cell
interaction with the extracellular matrix (Kang and Krauss, 1996; Bohmer et al., 1996; Koyama et al., 1996;
Schulze et al., 1996; Zhu et al., 1996; Assoian, 1997; Aplin et al., 1999).
. Figure 8-5
Proliferation of stem cells during early stages of neuronal tube development (symmetric) and during neuro-
genesis (asymmetric). Modified after Yoshikawa (2000)
occur as a result of erroneous DNA replication. In later stages, neuronal apoptosis might be due to limited
availability of trophic support or lack of synaptic input from target cells (Oppenheim, 1991). Neurons of
particular types arise from cells that divide at particular times. Different neuronal fates, thus, appear to be
determined at specific developmental stages or after a certain number of cell divisions. Some fates are
determined independently of active cycling, and the mature phenotype of a neuron is often determined even
postmitotically. Other fates, however, are determined when a cell is actively proliferating, and the cellcycle
status might have a critical influence over the determination process. The balance between proliferation and
differentiation in the nervous system is, thus, critical for histogenesis, and there might indeed exist
bidirectional mechanisms for coordinating the cell cycle with the process of neuronal determination and
differentiation (Edlund and Jessell, 1999; Ohnuma et al., 2001).
A variety of molecules such as Cdc2, Cdk1, Cdk5, Dtype cyclins, p27Kip1, p57Kip2, or Rb may have
complementary but separable roles in regulating the cell cycle and promoting neuronal determination/
174
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
differentiation (Hamel et al., 1992; Okano et al., 1993; Ohnuma et al., 1999; Dyer and Cepko, 2000; Levine
et al., 2000) (see also > Table 8-1). In addition, various determination/differentiation factors such as
TGFa, TGFb, EGF, or Notch might regulate cellcycle factors through transcriptional or posttranscrip-
tional mechanisms (Dorsky et al., 1997; Hanahan and Weinberg, 2000).
Apoptosis is a highly conserved mechanism by which eukaryotic cells actively commit suicide. Apoptosis in
the nervous system has been recognized as an essential process during development where it plays an
essential role in the control of the final number of neurons (Oppenheim, 1991). Apoptosis might also be
responsible for the loss of neurons during normal aging (Tagliatella et al., 1996; Kaufmann et al., 2001).
Apoptosis involves a series of stereotyped, morphologically welldefined phases. Its morphological
manifestations include nuclear and cytoplasmic condensation, intranucleosomal DNA cleavage, nuclear
membrane breakdown and plasma membrane blebbing, and the formation of apoptotic bodies. In
proliferating cells, apoptosis acts to prevent transmission of nonrepairable DNA modifications to the
progeny. Apoptosis, thus, prevents cell proliferation if it is inappropriate. This role is potentially maintained
in postmitotic cells such as differentiated neurons in which unscheduled proliferation would otherwise
generate neurons that cannot be easily integrated into functional circuitry. In proliferating cells, it is mostly
DNA damage that triggers cellcycle arrest, and if DNA damage is too extensive to be repaired, apoptosis.
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 175
During development, death is the default (Benn and Woolf, 2004). Immature neurons are compe-
tent to die (Deshmukh and Johnson, 1998) by an active apoptotic pathway unless it is shut off by sufficient
trophic support (Oppenheim, 1991). Thus, apoptosis during development results from the absence of
enough trophic factor support or can actively be triggered by signaling through the lowaffinity neuro-
trophin receptor p75NTR (Carter and Lewin, 1997; Miller and Kaplan, 2001) or the FAS receptor (Raoul
et al., 1999, 2000).
In mature neurons, two main pathways lead to apoptosis, an extrinsic or death receptorinitiated
pathway, and an intrinsic or mitochondrial pathway (Green, 1998; Strasser et al., 2000; Benn and Woolf,
2004). Both pathways converge at the mitochondrion on the release of apoptogenic proteins (Shimizu et al.,
2001; Rostovtseva et al., 2004; Sharpe et al., 2004). Extrinsic apoptosis is initiated by the activation of
plasma membrane death receptors of the tumor necrosis factor (TNF) receptor superfamily. The extrinsic
death receptor pathway either directly activates effector caspases through JNK or converges with the
intrinsic pathway at the mitochondrion. This intrinsic pathway results from mitochondrial damage that
leads to the release of cytochrome c from the mitochondria, which triggers the formation of the apoptotic
complex and the subsequent activation of effector caspases. Death pathways differ between neurons. The
specific deathsignaling pathway that leads to degeneration depends on the neuronal population that is
affected, as well on the nature, stage, cause, and extent of the deathinducing insult.
Apoptosis, proliferation, and differentiation are coupled through potentially multifunctional up-
stream signaling pathways. According to the cellular context, activation of identical pathways is translated
into such diverse cellular responses as growth arrest, cycling, cell death, or survival (Blagosklonny, 2003).
The cell cycle and apoptosis share both identical molecular regulators such as Rb, E2F, and p53 and cellular
phenomena such as detachment from the matrix, condensation of chromatin, disassembly of nuclear
lamina, and membrane blebbing (Ucker et al., 1991; Raff et al., 1993; Evan et al., 1995; Liu and Green,
2001). Deregulation of the cell cycle can either directly induce apoptosis or can increase the sensitivity to
apoptotic inducers (Evan et al., 1995; Evan and Littlewood, 1998). Successful proliferation, as an alternative
to apoptosis, requires active suppression of the apoptotic program.
The activation of various components of the cell cycle in postmitotic neurons might contribute to
apoptosis (King and Cidlowski, 1995). The induction of proteins that control the G1/S phase transition in
proliferating cells such as cyclin D1 occurs in dying neurons (Freeman et al., 1994; Kranenburg et al., 1996),
and G1 cyclindependent kinases can trigger neuronal death (Park et al., 1997a, b, 1998a). Similarly, proteins
that control the G2/M phase transition such as Cdc2 might be involved in apoptosis (Shi et al., 1994; Vincent
et al., 1997; Yu et al., 1998). Neuronal activity deprivation in cerebellar granule neurons can stimulate the
activity of Cdc2, which triggers apoptosis (Konishi et al., 2002). This process involves phosphorylation of the
BH3only protein Bad, thereby activating Badmediated apoptosis by inhibiting its sequestration through
1433 protein.
In neuronal apoptosis, induced in a variety of experimental in vitro paradigms, changes in Cdk
activities, levels of cyclins and Cdkis, and deregulation of Rb/E2F activity have been observed (Park et al.,
1997a, b, 1998a, b; Padmanabhan et al., 1999; OHare et al., 2000) (> Table 8-2). Induction of Cdks, for
example, occurs in vivo in mature adult neurons in mouse, rat, or rabbit during ischemia, excitotoxic cell
death, spinal cord injury, or various mutant mouse models associated with neuronal death such as mutant
SOD1G37R, weaver mouse, harlequin mouse, or staggerer mouse. It is also seen in vitro in neuroblastoma
cells or primary neuronal cultures deprived of trophic factor support or treated with DNA damaging agents
or amyloid (Giovanni et al., 1999, 2000; Park et al., 2000). The induction of Cdks in neurons is associated
with their dedifferentiation and triggers neuronal death that can be blocked by Cdk inhibitors or dominant
negative Cdks (Park et al., 2000; Liu and Greene, 2001).
NGF deprivation, for example, leads to increased Cdc2 activity and cyclin B expression in PC12 cells
(Gao and Zelenka, 1995), as well as elevated cyclin D1 levels in sympathetic neurons (Freeman et al., 1994)
whereas pharmacological inhibitors of Cdks, upregulation of Cdk inhibitors, or dominant negative Cdk4 or
Cdk6 promote survival of NGFdeprived sympathetic neurons (Park et al., 1997a, b, 1998a, b). Apoptosis,
induced by DNA damage, requires Cdk4 and Cdk6 activity and subsequent phosphorylation of pRb and
activation of the pRb/E2F/DP pathway (Park et al., 2000; Liu and Greene, 2001) and is suppressed by the G1/S
transition blockers deferoxamine and mimosine as well as by the Cdk inhibitors flavopiridol and olomoucine
176
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
. Table 8-2
Evidence of cellcycle activation under experimental conditions of in vivo and in vitro cell death
In Vivo Paradigms of Cell Death
Ischemia (transient focal or global ischemia in mouse, rat, or rabbit)
increased neuronal expression of cellcycle regulators/proliferation Guegan et al. (1997)
associated proteins Hayashi et al. (2000)
Katchanov et al. (2001)
Cdk4 Li et al. (1997a, b)
Cdk2 Osuga et al. (2000)
cyclin D1 Sakurai et al. (2000)
cyclin B1 Timsit et al. (1999)
cyclin G1 Tomasevic et al. (1998)
p21Cip1 van Lookeren Campagne
MPM2 and Gill (1998)
PCNA Gill (1998)
increased phosphorylation of Rb Wang et al. (2002)
neuroprotection through post/periischemic administration of Cdk inhibitors Wen et al. (2004)
Combined ischemia/hypoxia
increased neuronal expression of cellcycle regulators/proliferation Kuan et al. (2004)
associated proteins
Cdk2
Ki67
increased phosphorylation of Rb
DNAsynthesis (BrdU incorporation)
Kainateinduced seizures
increased neuronal expression of cellcycle regulators Ino and Chiba (2001)
Cdk4 Park et al. (2000)
cyclin D1 Timsit et al. (1999)
increased phosphorylation of Rb
Spinal cord injury (rat)
increased neuronal expression of cellcycle regulators/proliferation Di Giovanni et al. (2003)
associated proteins
Cdk4
cyclin D1
cyclin G
Gadd45a
cmyc
PCNA
increased phosphorylation Rb, E2F5
Amyotrophic lateral sclerosis caused by mutant SOD1G37R mice
increased neuronal expression of cellcycle regulators/proliferation Nguyen et al. (2003)
associated proteins
Cdk4
cyclin D1
increased phosphorylation of Rb
Weaver mouse (apoptosis in the external germinal layer (EGL)
of the cerebellum)
increased neuronal expression of cellcycle regulators/proliferation Migheli et al. (1999)
associated proteins
Cdk4
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 177
cclin D
cyclin A
PCNA
harlequin mouse mutation (progressive degeneration of terminally
differentiated cerebellar and retinal neurons)
increased neuronal expression of cellcycle regulators/proliferation Klein et al. (2002)
associated proteins
PCNA
Targetrelated neuronal death during development (apoptotic staggerer
granule cells)
increased neuronal expression of cellcycle regulators/proliferation Herrup and Busser (1995)
associated proteins
cyclin D
PCNA
(Park et al., 1997). Activation of the cyclin D/Cdk4/6RbE2F pathway might, thus, be a critical component of
the apoptotic mechanism induced by DNAdamage or trophic factor deprivation in postmitotic neurons (Liu
and Greene, 2001; Kruman et al., 2004).
2.3.1.1 Caspases Caspases are the executioners of the apoptotic program (Thornberry and Lazebnik,
1998). Caspases stands for cysteine aspartate proteases. They cause cell death by degrading critical structural
elements such as lamins and gelsolin, and by activating latent enzymes, such as DNAses, through proteolysis
of their inhibitors. Caspases are present in healthy cells as inactive precursors (zymogens) that can be
activated by caspasemediated cleavage. Proteolysis releases two fragments of about 10 kDa and 20 kDa that
are assembled into the active tetrameric enzyme (OConnor et al., 2000). There are 14 members in the
mammalian family of caspases, divided into three subfamilies on the basis of the peptidesequence
preferences of their substrates. Certain initiator caspases, such as caspase8 and caspase9 have low
enzymatic activity in their zymogen form and contain regulatory aminoterminal prodomains that can
bind to specific adapter proteins (Salvesen and Dixit, 1997; Thornberry and Lazebnik, 1998). Caspase8 is
associated with apoptosis involving death receptors, and loss of adhesion, whereas caspase9 is involved in
stressinduced apoptosis (Earnshaw et al., 1999). Stressed cells might receive conflicting signals for cell
division and growth arrest, which might trigger apoptosis (Arsura and Sonenshein, 1996; OConnor et al.,
2000). In response to apoptotic stimuli, adapter proteins bind these initiator caspases and form oligomeric
aggregates, which facilitate autocatalytic processing of the cymogens (Martin et al., 1998; Srinivasula et al.,
1998). Initiator caspases subsequently process effector caspases, creating a proteolytic cascade that
executes cell death. In addition, there are noncaspase proteases such as cathepsins, calpains, and
granzyme B, involved in apoptosis.
2.3.1.2 Anoikis The adhesionrelated form of apoptosis is called anoikis (Frisch and Francis, 1994). It is
triggered by inappropriate or inadequate contacts between the cell and the extracellular matrix
and essentially displays all the features of apoptosis such as membrane blebbing, nuclear fragmentation,
and DNA degradation.
Integrinmediated cell anchorage, furthermore, has a vital role in the control of apoptosis (Frisch and
Ruoslahti, 1997; Aplin et al., 1998). They regulate cell viability through their interaction with the extracel-
lular matrix and can convert cellmatrix interactions into intracellular signals. These mechanisms involve
FAK (Frisch et al., 1996b) and Rasdependent signaling (Chen and Guan, 1994; Datta et al., 1997; King et al.,
1997; Khwaja et al., 1997) as well as the Wnt1/catenin pathway (Frisch and Ruoslahti, 1997; Kuhl and
Wedlich, 1997). FAK when activated can suppress anoikis. Phosphatidylinositol 3kinase/Akt and mitogen
activated protein kinase may mediate the anoikissuppressing effects. Activation of TrkB, the major receptor
for brainderived neurotrophic factor, can also suppress anoikis through stimulation of PI3 kinase signaling
(Douma et al., 2004). Conversely, the stressactivated protein kinase/Jun aminoterminal kinase pathway
promotes anoikis (Zhan et al., 2004). Control of anoikis may vary substantially from one cell lineage to
another and Ras may fulfill both positive and negative regulatory functions depending on the cell type
(Meredith et al., 1993; Frisch and Francis, 1994; Hall et al., 1994; Polakowska et al., 1994; Ruoslahti and
Reed, 1994; Coucouvanis and Martin, 1995; Boudreau et al., 1995; Hungerford et al., 1996; Vachon et al.,
1996, Assoian, 1997; Frisch and Ruoslahti, 1997; Khwaja et al., 1997; Meredith and Schwartz, 1997; McGill
et al., 1997; Eckert et al., 2004).
Anoikis may also involve phosphorylation of FAK and p53dependent mechanisms (Zhang et al., 2004).
The proapoptotic protein Bim, a member of the Bcl2family, is a critical mediator of anoikis in epithelial
cells (Reginato et al., 2003). Detachmentinduced expression of Bim requires a lack of 1integrin engage-
ment, downregulation of EGFreceptor expression, and inhibition of ERK signaling (Reginato et al., 2003).
The Bcell leukemia/lymphoma2 (Bcl2) families of proteins are major gatekeepers of apoptosis (Cory
and Adams, 2002). They control cell survival through both caspasedependent and caspaseindependent
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 179
pathways and can act via regulation of permeabilization of the outer mitochondrial membrane (Cory and
Adams, 2002; Akhtar et al., 2004). The Bcl2 family consists of three main subgroups, which are categorized
according to their anti or proapoptotic function, and the presence or absence of the Bcl2 homology (BH)
domains (Korsmeyer, 1999).
Antiapoptotic members, such as Bcl2, BclxL, Bclw, and myeloid cell leukemia1 (Mcl1) contain four
Bcl2homology domains. They can also delay entry into the cell cycle (OReilly et al., 1996). The
proapoptotic members of this family fall into two categories: Baxlike proteins, such as Bcell lymphoma
2associated protein X (Bax), Bcl2 agonist killer 1 (Bak), Bcl2related ovarian killer protein (Bok), and BclxS
that contain multiple BH domains, or BH3only members, such as Bcl2associated death protein (Bad),
BH3 interacting domain death agonist (Bid), Bcl2interacting mediator of cell death (Bim), Bcl2 mod-
ifying factor (Bmf) death protein 5/harakiri (DP5/Hrk), Puma (p53upregulated modulator of apoptosis),
small protein with inherent killing effect (Spike), and phorbol12myristate13acetateinduced protein
(PMAIP1/Noxa) (Puthalakath and Strasser, 2002; Akhtar et al., 2004). Although dimerization among the pro
and antiapoptotic members of the Bcl2 family plays a major role in the regulation of apoptosis, the expression
levels of Bcl2 determine the cells fate. Bcl2 family members are involved in neuronal programmed cell death and
play a significant role in diseases of the nervous system such as hypoxicischemic cell death, seizureinduced
neurodegeneration, motor neuron disease, and neurotoxic and mechanical neuronal injury (for review see
Akhtar et al., 2004).
2.3.3 P53
Another major element in the regulation of apoptosis is p53. Its activation prevents cell proliferation by
inducing cellcycle arrest at the G1S checkpoint or apoptosis through activation of proapoptotic genes
such as bax.
Although the constitutive level of p53 is low, it can rapidly be induced through DNA damage (Levine,
1997), which stimulates the transcription of regulators such as p21Cip1, mdm2, and bax (Agarwal et al.,
1998). P21Cip1 inhibits Cdk activation and results in cellcycle arrest, thus, preventing replication of
damaged DNA (Ko and Prives, 1996). This function of p53 in shutting down the proliferation in response
to DNA damage reflects its role as guardian of the genome (Lane, 1992). Mdm2 provides a feedback loop
regulating degradation of p53 by ubiquitination (Oren, 1999). In severely damaged cells unable to repair
DNA, p53 induces bax, fas, and other genes involved in apoptotic signaling (Miyashita and Reed, 1995; Yin
et al., 1997; Gottlieb and Oren, 1998).
Nitric oxide (NO) is a signaling free radical, which displays antiadhesive properties (Radomski et al., 1990)
and can be considered as a cause for anoikis (Monteiro et al., 2004). It has the capacity to signal and trigger
proapoptotic events in a variety of cell types. NO can inhibit cell adhesion, interfere with the assembly of
focal adhesion complexes, and disrupt the cellextracellular matrix interactions. NO is enzymatically
derived from Larginine and O2. For the enzymes NO synthases (NOS), three isoforms can be distin-
guished, Type I or neuronal isoform (nNOS), Type II or inducible isoform (iNOS), and Type III or
endothelial isoform (eNOS). Lipopolysacharides, cytokines, and toxins can induce the expression of the
iNOS with the production of large amounts of NO whereas the neuronal and endothelial isoforms are
Ca2/calmodulindependent, constitutively expressed enzymes that produce lower amounts of NO
(Nathan, 1992).
NO might have pleiotropic effects on the induction of apoptosis and cell survival, depending on factors
such as levels of NO production, coactivation of other signaling pathways, and other cell typespecific
features. In a variety of cell types such as lymphocytes, hepatocytes, and endothelial cells, NO, at tissue
concentrations, can effectively block lipid peroxidation, and protect against oxidative damage mediated cell
death and apoptosis (Padmaja and Huie, 1993; Liu and Stamler, 1999). Among other mechanisms, these
180
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
effects involve the activation of p21Ras signaling through both MAPkinase and PI3 pathway (Lander et al.,
1995; Deora et al., 1998; McFall et al., 2001). As opposed to the protective role of NO when produced at low
fluxes, NO generated at high fluxes through iNOS is cytotoxic (Brune et al., 1999). NO has the capacity to
trigger proapoptotic events. It can inhibit cell adhesion, and it interferes with the assembly of focal adhesion
complexes, thereby disrupting cellextracellular matrix interactions leading to anoikis. NO can potentially
uncouple cooperative signaling by growth factors and matrix receptors, which ultimately leads to apoptosis
(Frisch and Screaton, 2001; Monteiro et al., 2004). This underlies the importance of coupling integrin and
growth factorsignaling pathways for regulating cellular survival.
2.3.5.1 CMyc CMyc is a nuclear phosphoprotein that acts as a transcription factor stimulating both cellcycle
progression and apoptosis (Penn et al., 1990; Facchini and Penn, 1998). Activity of cMyc is regulated through
phosphorylation and interaction with other cellular proteins such as Max (Lutterbach and Hann, 1994).
It can exert its effects on the cell cycle by controlling transcription of genes coding for Cdc25A, cyclin
D1, cyclin D2, cyclin E, cyclin A, Cdk1, Cdk2, Cdk4, and E2F (Born et al., 1994; Kim et al., 1994; Beier et al.,
2000). It can also suppress transcription of the genes for Cdk inhibitors p15INK4b, p21Cip1, and p27Kip1
(Grandori et al., 2000). As cMyc responds directly to mitogenic signals, it might have a critical
role especially in the transition from G0 to S phase (Spencer and Groudine, 1991). CMyc expression,
however, is maintained throughout the cell cycle and might have a role in G2 (Hann et al., 1985; Mateyak
et al., 1997).
CMyc also plays a key role in apoptosis, which might involve p53dependent and independent path-
ways. CMyc can transactivate the p53 gene promoter and increase the halflife of p53 (Hermeking and Eick,
1994). CMycinduced apoptosis might also involve activation of Fas receptor (Wang et al., 1998) and
inhibition of Bcl2 and Mcl1 (Reynolds et al., 1994). Phosphorylated cMyc can be detected immunocyto-
chemically in dystrophic neurites and neurons with neurofibrillary degeneration in AD, and in neurons and
glial cells with abnormal t deposition in Picks disease, progressive supranuclear palsy, and corticobasal
degeneration (Ferrer et al., 2001).
2.3.5.2 E2F A number of cellcycle molecules that have been associated with neuronal death converge on
regulation of the E2F family of transcription factors. E2F acts as a gene silencer in neurons and that repression is
required for neuronal survival (Liu and Greene, 2001). Induction of apoptosis induced by DNA damaging
agents or trophic factor withdrawal is characterized by depression of E2F responsive genes (Liu and Greene,
2001). Among those genes that are derepressed in the process of cell death are the transcription factors B and
Cmyb. E2F can also induce the expression of the proapoptotic factor Apaf1 (Moroni et al., 2001). E2F1
deficient mice are more resistant to focal ischemia (MacManus et al., 1999). PRb/E2F and p53
pathways may be directly linked in the cell cycle and apoptosis. Activated p53 induces p21Cip1,
thereby preventing Cdk activation and subsequent phosphorylation of pRb. In contrast, free E2F induces
p53 transactivation, connecting the pRb/E2F directly to p53dependent apoptosis (Hiebert et al., 1995).
2.3.5.3 STATs Many cytokine receptors transduce signals that promote cell division, block apoptosis, and
induce differentiation. These pathways involve distinct members of the JAK family of kinases and the STAT
family of transcriptional regulators (Ihle, 1996). Depending on the specific stimulus or cell type, STATs can
mediate either proapoptotic or antiapoptotic signals (Battle and Frank, 2002). Regulation of apoptotic
pathway by STATs is largely due to transcriptional activation of genes, which encode proteins that are
involved in cell death processes, such as BclxL, caspases, Fas, and TRAIL as well as those that regulate cell
cycle progression such as p21Waf1.
2.3.5.4 NFkB The transcription factor NFkB is activated by a broad range of signals including cytokines,
mitogens, free radicals, and other stress signals. NFkB upregulates several survival factors, and might also directly
promote antiapoptotic mechanisms, such as upregulation of Bcl2 family members (Baichwal and Baeuerle, 1997;
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 181
Hinz et al., 1999; Glasgow et al., 2001). In contrast, there is also evidence for apoptosispromoting activity of
NFkB, depending on the nature of the apoptosisinducing stimulus (Kaltschmidt et al., 2000).
Cellcycle regulators are recognized targets for the treatment of proliferative disorders such as cancer,
cardiovascular disease, infections, or autoimmune diseases (Brooks and La Thangue, 1999; Nabel, 2002;
Vermeulen et al., 2003a) (> Table 8-3). In cancer, cellcycle deregulation occurs through mutations affecting
proteins critically involved in different levels of the cell cycle such as Cdk, cyclins, Cdkactivating enzymes,
Cdk inhibitors, Cdk substrates, and checkpoint proteins (Sherr, 1996; McDonald and el Deiry, 2000).
. Table 8-3
Disorders with involvement of cellcycle regulators that are potential targets for treatment with cellcycle
inhibitors
Cancer
Cardiovascular disease
atherosclerosis
cardiac hypertrophy
Nervous system
Alzheimers disease
amyotrophic lateral sclerosis
stroke
Viral infections
human cytomegalovirus (HCMV)
human papillomavirus (HPV)
human immunodeficiency virus (HIV)
herpes simplex virus (HSV)
Fungal infections
Protozoan disease
malaria
leishmaniosis
trypanosomes
Psoriasis
Glomerulonephritis
Recent evidence indicates that the control of proliferation and differentiation, i.e., cellcycle control, is also
involved in degeneration of nondividing cells, such as mature neurons that might be of critical relevance to
the pathomechanism of AD and related neurodegenerative disorders (Arendt, 1993; Heintz, 1993; Nagy,
2000; Copani, 2001).
The control of cell division and cellular differentiation has repeatedly been linked in the past to
neurodegeneration in AD, and, based on this potential relationship, a variety of hypotheses have been
formulated such as the concept of hyperdifferentiation (Bouman, 1934), retrodifferentiation (Scheibel
and Scheibel, 1975), dysdifferentiation (Woolf and Butcher, 1990), or dedifferentiation (Arendt, 1993;
Heintz, 1993). About 10 years ago, it became increasingly clear that in AD, the mitogenic force is elevated
and neuronal mitogenic pathways are aberrantly activated in the early stages of the disease (Gartner et al.,
1995). From that time on, further insight into the pathomechanism progressed further downstream,
suggesting a potential involvement of cellcycle control as a result of abnormal mitogenic signaling (Arendt
et al., 1996; Vincent et al., 1996; Nagy et al., 1997).
182
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
Cell death in the nervous system, irrespective of whether caused by inflammation, ischemia, or primary
degeneration, selectively affects special types of neurons whereas the others are spared. In AD, this selective
neuronal vulnerability has systemic character. Neurodegeneration, initially very constantly affects circum-
scribed brain areas such as cholinergic basal forebrain projection neurons and the entorhinal cortex and
progresses from there throughout the brain, subsequently involving other areas in a defined sequence. The
reasons for this selective neuronal vulnerability are unknown, and several explanations have been given.
Paul Flechsig (1896) stressed the ontogenetic aspect of selective neuronal vulnerability when he advanced
the idea that variations in vulnerability of different groups of neurons are to be traced back in large part to
developmental conditions, a concept later defined by Cecile and Oscar Vogt (1951) as the principle of
pathoclisis. More recently, the phylogenetic dimension of selective neuronal vulnerability was recognized
(Ashford and Jarvik, 1985; Rapoport, 1989; DiPatre, 1991; Braak and Braak, 1996; Arendt et al., 1998;
Arendt, 2003). Brain areas and neuronal types highly vulnerable against neurofibrillary degeneration in AD
not only mature rather late during ontogenic development, but also have been acquired late during
phylogenetic development (or have completely been reorganized lately). Accordingly, these brain structures
typically subserve higher brain functions such as learning, memory, selfawareness, and others and as such
exhibit a particularly high degree of synaptic plasticity (> Figure 8-6).
During development, after proliferation, migration and differentiation are completed, neurons become
integrated into a neuronal network. This integration is based on intercellular communication, which is
largely regulated through external cues. Connectivity and attachment of a cell are mechanisms that, during
evolution from single cellular systems to multicellular systems, have been acquired to assess signals from the
environment and respond appropriately by proliferation, differentiation, and cell death. In a multicellular
organism, external cues are thus, used for morphoregulation, i.e., the assembly of individual cells into
highly ordered tissues. Terminally differentiated neurons that have permanently withdrawn from the cell
cycle, however, make use of these mechanisms primarily developed to control connectivity for their genuine
function, i.e., for information processing in a multicellular network. Thus, cellcell interactions in the
nervous system subserve a phylogenetically newly acquired purpose, the formation of a dynamic network.
Once the network has been set up, the cell cycle is shut down and information from neighboring cells is
translated into continuous changes of synaptic strength and morphology, a process referred to as synaptic
plasticity. This means that, at least partially, identical signaling mechanisms are used to control and regulate
such divergent effectors as cellcycle control and synaptic plasticity.
From the phylogenetic perspective, it might be a great achievement for a neuron to control its synaptic
plasticity; to do this on the potential expense of differentiation control, however, might put the neuron on a
permanent risk. Neurons acquired particularly late during evolution of the human brain (e.g., cortical
associative circuits), which subserve typical human higher cortical functions such as learning, memory,
reasoning, consciousness, selfawareness, etc., need to display a particularly high degree of synaptic
plasticity, which might explain the particularly high sensitivity toward the loss of differentiation control
and cell death (> Figure 8-6). What makes the human brain so special, a large association cortex
characterized by a high degree of plasticitya prerequisite for selfconsciousness, i.e., the development
of a belief of itselfat the same time creates a vulnerability unique to humans (> Figure 8-2). Identical
signaling pathways might be used in the first stage of synaptic change in both development and maturity,
and higher cognitive functions such as learning and memory might be based on adaptive modifications of
an ancient mechanism initially evolved to wire the brain. The activitydependent modification of synapses
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 183
. Figure 8-6
Hierarchy of cortical vulnerability to neurofibrillary degeneration in the cerebral cortex. a: Neuropathological
staging of AD according to Braak and Braak (1991). b: This progression of neurofibrillary degeneration throughout
different cortical areas in AD follows a certain sequence (stage I to VI) that represents systematic differences in the
vulnerability (hierarchy of vulnerability), which matches inversely the hierarchic pattern of complexity of cortical
information processing, structural plasticity, and phylogenetic and ontogenetic age. Those neuronal systems that
play a crucial role in higher brain functions and therefore become increasingly predominant as the evolutionary
process of encephalization progresses, such as the hippocampus and the neocortical association areas, retain a
high degree of structural plasticity throughout their life. These areas of the brain take the longest to mature
during childhood and adolescence. Exactly the same brain structures display the highest degree of vulnerability
during aging and in AD. Modified after, Arendt et al. (1998); Braak and Braak (1991, 1996)
is, thus, a powerful mechanism for shaping and modifying the response properties of neurons, but it is also
dangerous (Abbott and Nelson, 2000) as it still hides its beasty potency, the risk of phylogenetic regression
into cellcycle activation, which ultimately results in cell death (Arendt, 2003). Synaptic plasticity might
thus, not only be beneficial providing the basis for information processing and memory formation in the
184
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
brain, it might also be crucially involved in a variety of pathological conditions, including amnesia,
dementia, epilepsy, and psychosis (BenAri and Represa, 1990; McEachern and Shaw, 1996; Jeffery and
Reid, 1997).
Alterations of neuroplasticity are also a constant feature of AD (Ashford and Jarvik, 1985; Cotman and
Anderson, 1988; Butcher and Woolf, 1989; Phelps, 1990; Flood and Coleman, 1990; DiPatre, 1991; Geddes
and Cotman, 1991; Masliah et al., 1991c; Swaab, 1991; Walsh and Opello, 1992; Roberts et al., 1993; Neill,
1995; DeWitt and Silver, 1996, Mirmiran et al., 1996; Foster, 1999; Kondo et al., 1999; Mesulam, 1999; Teter
and Ashford, 2002).
Neurodegeneration in AD is associated with aberrant neuritic growth (Arendt et al., 1986). Abnormal
growth profiles preferentially affect neurons that are potentially vulnerable to neurodegeneration, such as
cholinergic basal forebrain neurons and cortical pyramidal cells (Arendt et al., 1995a, b, e) (> Figure 8-7).
. Figure 8-7
Aberrant morphological growth profiles in AD on neurons of the cholinergic basal nucleus of Meynert. Growth
profiles are localized on dendritic endings (d) or cover the soma (b, c). The appearance of these growth profiles
is associated with an upregulation of the NGF receptors TrkA (a) and p75 (b). (a: TrkA in situ hybridization;
b: p75 immunocytochemistry; c, d: Golgi impregnation). Scale bar: 30mm. Modified after Arendt and Bruckner
(1992); Arendt et al. (1995a)
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 185
Aberrant sprouts can be detected in early stages of the disease (Ihara, 1988) and precede the formation of
paired helical filaments or occur even without massive neuronal loss (Arendt and Bruckner, 1992b). They
might thus, represent a very early event of primary significance, inherent to the pathomechanism rather
than a secondary event triggered by ongoing degeneration.
As opposed to the continuous growth of dendritic elements during aging (Buell and Coleman, 1979),
growth in AD is aberrant with respect to its localization, morphological appearance, and composition of
cytoskeletal elements (Arendt et al., 1986, 1995a, b, d; McKee et al., 1989). Dystrophic neurites, mainly
dendritic but occasionally also axonal in origin, form a constant component of AD pathology. The original
identification of these neurites, in the early years of research on Alzheimers disease by Fischer (1907), has
been corroborated more recently by Golgi studies, ultrastructural evidence, and the identification of
elevated growthassociated proteins (for review see Arendt, 2003). Growth profiles on cholinergic basal
forebrain neurons, for example, are typically associated with receptors for NGF, indicating the potential
ability to respond to an increased trophic force (Arendt and Bruckner, 1992a, b) (> Figure 8-7).
The presence in the AD brain of growthassociated and growthpromoting proteins, as well as their receptors,
might be an indication of an increased mitogenic force particularly pronounced within the microenvironment
of plaques. The growing list of compounds contains GAP43, MARCKS, spectrin, heparansulfate, laminin,
NCAM, various cytokines, and neurotrophic factors such as NGF, bFGF, EGF, IL1, IL2, IL6, IGF1, IGF2,
PDGF, vascular endothelial growth factor, and HGF/SF (for review see Arendt, 2003).
3.2.1.1 Ras A number of these neurotrophic and potentially mitogenic compounds that are elevated early
in the course of the disease mediate their cellular effects through activation of receptor tyrosine kinases,
which recruit p21Ras, a GTPase that belongs to the Ras superfamily of smallGproteins and lead to the
sequential activation of Raf, MEK, and ERK (Seger and Krebs, 1995) (> Section 1.5.1). Ras proteins are
involved in the central coordination of a variety of functions including cytoskeletal organization, gene
expression, cellcycle progression, membrane trafficking, cell adhesion, migration and polarity, and
synaptic plasticity (Heumann et al., 2000; Arendt et al., 2004). During brain development, p21Ras
participates in the regulation of the G0/G1 transition of the cell cycle and might thus, be a critical
regulator for cellular proliferation and differentiation (Borasio et al., 1989). In the adult nervous system,
it plays a role in reactive dendritic proliferation and neosynaptogenesis (Phillips and Belardo, 1994; Holzer
et al., 2001a, b). Thus, Ras is a central regulator of synaptic plasticity in the adult brain (Arendt et al., 2004).
The RasERK/MAPK pathway is an evolutionary, conserved signaling mechanism that plays a fundamental
role in the regulation of cellular proliferation and differentiation and control of neuroplasticity, and is also
involved in modulating the expression and posttranslational processing of APP and t protein (Greenberg
et al., 1994; Mills et al., 1997; Sadot et al., 1998a), two proteins critically involved in AD pathomechanism.
In AD, p21Ras is highly expressed in vulnerable brain areas prior to its affection by neurofibrillary
degeneration (Gartner et al., 1999), which indicates an activation of the RasMAPK pathway in the early
stages of the disease (Arendt et al., 1995c) (> Figure 8-8).
3.2.1.2 MAPK The activation of MAPK that is associated by its nuclear translocation plays an essential role
in the expression of many immediate, early, and late response genes (> Section 1.5.1). MAPK activation and
nuclear translocation are required for longterm facilitation in Aplysia (Martin et al., 1997b) and LTP in
vertebrates (Impey et al., 1998). MAPK signaling might thus be a critical regulator for both shortterm
synaptic function and transcription of genes required for longterm plasticity. Ras/MAPK signaling
components are highly enriched in the adult CNS, and expression of many MAPK regulators is largely
186
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
restricted to the CNS, where they are highly abundant in association areas implicating a role in memory
consolidation and synaptic plasticity.
ERK1/2 and other members of the MAPK family (Drewes et al., 1992; Goedert et al., 1992) are able to
phosphorylate the microtubuleassociated protein t on threonine and serine residues found phosphory-
lated in PHFs. In AD, both MAPKK and MAPK are activated in the early disease stages (Arendt et al., 1995c)
(> Figure 8-8). This activation of the MAPK pathway can be modeled in transgenic mice expressing
permanently activated ras (> Figure 8-9).
Neuronal nitric oxide synthase (nNOS) was originally thought to be a constitutively expressed enzyme. It
becomes increasingly clear now, however, that its levels are dynamically regulated in response to neuronal
development, plasticity, and injury (Dawson et al., 1998). The transcriptional induction of nNOS, which is
controlled by neurotrophins and other growth factors, is in turn involved in regulating the expression of
immediate early genes in neurons, thereby controlling neuronal growth and differentiation (Peunova and
Enikolopov, 1993, 1995). NO may play a major role in nervous system morphogenesis and plasticity and
may be involved in activitydependent establishment of connections in both developing and regenerating
neurons (Dawson et al., 1998; Luth and Arendt, 1998a; Bogdan, 2001). Under developmental conditions,
NO may trigger growth arrest, a process that at least in certain cell types might involve inhibition of Cdk2,
a key regulator of the G1 and S phases of the cell cycle (see later). These antiproliferative effects of NO
involve the repression of cyclin A reexpression as well as an induction of the cyclindependent kinase
inhibitor p21Cip1 (Gansauge et al., 1998). The high degree of coexpression of nNOS with p16INK4a (Luth
et al., 2000) indicates that further regulators of the G1S transition might be involved in the NOinduced
cellcycle arrest or that additional mechanisms of proliferation and differentiationregulating mechanisms
are activated in parallel in the course of neurodegeneration in AD. Thus NO serves as an inducer of cell
cycle arrest, initiating the switch to cytostasis during differentiation (Peunova and Enikolopov, 1995;
Dawson et al., 1998), a process that can alternatively lead to apoptosis (Gansauge et al., 1998).
Although the molecular mechanism for the control of NO in proliferation, differentiation, cellular
survival, and death is not understood in detail, recent evidence indicates that activation of p21Ras is
critically involved in downstream signaling as a potential endogenous NOredoxsensitive effector molecule
(Lander et al., 1997b; Yun et al., 1998). Endogenous NO and intermediates generated through oxidative
stress can drive the Ras/MAPK cascade directly by direct activation of RasGTPase activity (Lander et al.,
1997; Bogdan 2001). NO might thus be a key mediator linking cellular activity to gene expression and long
lasting neuronal responses through the activation of p21Ras by redox sensitive modulation (Dawson et al.,
1998), a process with potential implications both in development and degeneration.
In AD, nNOS is aberrantly expressed in potentially vulnerable neurons of the isocortex and entorhinal
cortex. Since these neurons express nNOS before they are affected by neurofibrillary degeneration (Luth
and Arendt 1998b; Luth et al., 2000), transcriptional induction of nNOS might be an early event in the
process of neurodegeneration. As expression of nNOS in AD is highly colocalized with p21Ras (Luth et al.,
2000) (> Figure 8-10), an autocrine loop may exist within cells, whereby NO activates p21Ras, which in
turn leads to cellular activation and stimulation of NOS expression (Lander et al., 1997). In parallel,
. Figure 8-8
Activation of p21Rasdependent MAPkinase pathway in AD. Expression of p21Ras, bRaf, MAP kinase kinase
(MEK1), and the MAP kinases ERK1/2 quantified by ELISA and detected by immunohistochemistry in Brodmann
area 10. All four markers are elevated early in the course of the disease. P21Ras is highly enriched in plaques,
plaqueassociated astrocytes, as well as in tanglebearing neurons. BRaf and the associated protein 1433 are
also found in tangles. The subcellular translocation of MEK1 and ERK1/2 into neuronal nuclei indicate enzyme
activation. Modified after Arendt et al. (1995d); Arendt (2003); Gartner et al. (1995)
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 187
. Figure 8-9
Changes in the subcellular distribution of MAP kinase (ERK1/2) in AD, and in a transgenic mice model
expressing permanently activated p21Ras under control of the synapsin promoter (synRas mouse). Both in
AD and transgenic mice, MAPK is translocated from the cytoplasm (asterisk) to the nucleus (arrows) indicating
an activation of the enzyme. Scale bar: 10 mm. Modified after Arendt et al. (1995c); Holzer et al. (2001)
. Figure 8-10
Double immunofluorescence for nNOS/p21Ras and nNOS/p16INK4a. The aberrant expression of nNOS in pyra-
midal neurons is highly colocalized with p21Ras (top row) and p16INK4a (bottom row), respectively. Interneurons
expressing nNOS constitutively (arrow), on the contrary, neither express p21Ras nor p16INK4a, modified after
Luth et al. (2000)
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 189
downstream mechanisms such as the cell cycle might be activated as indicated by the coexpression of
cellcycle regulators (> Figure 8-10). The coexpression of NOS and p21Ras in neurons vulnerable to
neurofibrillary degeneration early in the course of AD might thus provide the basis for a feedback
mechanism that might exacerbate the progression of neurodegeneration in a selfpropagating manner
(> Figure 8-11). This selfperpetuation of a process likely to be associated with limited prospects of
physiological control and termination might be the critical switch converting two potentially neuropro-
tective mechanisms such as NO (Farinelli et al., 1996) and p21Ras (Heumann et al., 2000) dependent
signaling into a disease process, leading to slowly but continuously progressing neuronal death.
Several genes and proteins and their posttranslational modification associated with AD have recently been
found to play critical roles in neuronal development, particularly neuronal migration, and axon extension
(Arendt, 2000, 2001a, b; Bothwell and Giniger, 2000; Mehler and Gokhan, 2001). Neurodegeneration,
. Figure 8-11
Schematic illustration of the intracellular signaling events triggered by Morphodysregulation in AD that involve
an aberrant activation of p21Ras/MAPkinase signaling, a loss of differentiation control, the subsequent reentry
and partial completion of the cell cycle, and eventually result in cell death. Modified after Arendt et al. (1998),
J Neural Transm (Suppl 54), 147158; Arendt (2003)
190
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
furthermore, is associated with a complex structural reorganization, which involves a replay of develop-
mental programs. The aberrant neuritic growth in AD, as a likely indication of defect synapse turnover, is
accompanied by microtubular reorganization (McKee et al., 1989) associated with the reexpression of a
number of developmentally regulated proteins involved in morphoregulation indicating that molecules,
overexpressed in AD, might play a role in structural remodeling of the adult brain (Arendt, 2001b).
Based on this evidence, we have suggested previously that the process of continuous synaptic reorgani-
zation becomes defective in AD (Arendt, 1993). In this pathogenetic process, a subset of neurons retaining a
high degree of plasticity and that are presumably in a labile state of differentiation, are forced into a
condition of dedifferentiation that is characterized by an expression of developmental regulated genes,
posttranslational modifications, and an accumulation of gene products to an extent that goes beyond those
observed during regeneration. This replay of developmentally regulated mechanisms might be the end stage
of disturbed structural brain selforganization and a slowly progressing morphodysregulation. This
process of dedifferentiation involves molecular events that, in dividing cell populations, would lead to
cellular transformation and is thus, not compatible with the state of a neuron being irreversibly blocked
from reentry into the cell cycle. It might, therefore, lead to neuronal death. From this hypothesis, it can be
predicted that these molecular events that are involved in neoplastic transformation might also play a key
role in the pathomechanism of AD (Heintz, 1993). These mechanisms are notably a dysfunction of
mitogenic signal transduction and cellcycle control.
Proteins that normally function to control the cellcycle progression in actively dividing cells may play roles
in the death of terminally differentiated postmitotic neurons (Farinelli and Greene, 1996; Park et al., 1996).
Thus, a dysregulation of Cdks and their regulating partners, such as cyclins and Cdk inhibitors, indicating
an activation of the cell cycle in postmitotic neurons has been observed in AD (Smith et al., 1995; Arendt
et al., 1996, 1998b, c, 2000; McShea et al., 1997; Vincent et al., 1997; Nagy et al., 1997a, 1988; JordanSciutto
et al., 1999; Arendt, 2000; Dranovsky et al., 2001), cerebral ischemia, and during trophic factor deprivation
(Gill and Windebank, 1998; Tomasevic et al., 1998; van Lookeren Campagne and Gill, 1998; Timsit
et al., 1999; Osuga et al., 2000; Sakurai et al., 2000; Katchanov et al., 2001) (> Figures 8-12 and > 8-13,
> Table 8-4). Correspondingly, pharmacological inhibitors of cell cycle or ectopic expression of Cdkis can
protect neurons against death (Farinelli and Greene, 1996; Kranenburg et al., 1996; Park et al., 1996;
Giovanni et al., 1999; Padmanabhan et al., 1999; Osuga et al., 2000; Katchanov et al., 2001; Knockaert et al.,
2002) (> Figure 8-14).
3.2.4.1 CyclinDependent Kinases Several cyclindependent kinases critical for progression through the cell
cycle (Koh et al., 1995) such as Cdk1 (Cdc2), Cdk4, and Cdk5 are deregulated in AD (Ledesma et al., 1992;
Vincent et al., 1997; Busser et al., 1998; Nagy et al., 1998; Patrick et al., 1999; Arendt et al., 2000). The Cdk1
(Cdc2) kinase, moreover, is able to phosphorylate t protein at sites known to be phosphorylated in AD, and
thus potentially contributes to the generation of PHFt (Ledesma et al., 1992; Kobayashi et al., 1993). APP,
furthermore, is phosphorylated both in vitro and in intact cells by a Cdk1 (Cdc2)like kinase in a cellcycle
dependent manner, which is associated with altered production of potentially amyloidogenic fragments
containing the entire b/A4 domain (Suzuki et al., 1994). Ab in turn might potentially act as a proliferating
signal, driving cultured rat primary neurons into the cell cycle (Copani et al., 1999), although in vivo, this
has not been replicated (Gartner et al., 2003).
3.2.4.2 Cyclins The level of cyclin D1, a critical regulator of the transition from the G0 to G1 phase of the
cell cycle that acts through activation of Cdk4, is increased in neurons prone to neurodegeneration in AD
(Busser et al., 1998; Arendt et al., 2000). Cyclins other than D1 such as cyclins E and A involved in
regulation of G1/S transition as well as cyclin B regulating G2/M transition (Nagy et al., 1997a; Vincent
et al., 1997; Smith et al., 1999; Arendt et al., 2000) are also elevated.
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 191
. Figure 8-12
The cell division cycle and its major regulatory elements. Main switches for the activation and progression
of the cell cycle are elevated in neurons prone to neurofibrillary degeneration early in the course of AD (shown
in dark gray)
INK4a
3.2.4.3 Cdk Inhibitors p16 , a prominent representative of the INK4 family, involved in pathways for
control of cell growth and proliferation (Kamb et al., 1994), is also increased in AD as are other members of
the INK4 family of the cyclindependent kinase inhibitors interacting with Cdk4/6 such as p15INK4b,
p18INK4c, and p19INK4d. However, alterations of p21Cip1 and p27Kip1 are less constant (Arendt et al., 1996,
1998b, c).
192
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
. Figure 8-13
Major regulators of the activation and orderly progression through the cell cycle are expressed in vulnerable
neurons in AD (compare also > Figure 7-11). They are already highly abundant in pyramidal neurons prior to
PHF formation. In more advanced stages of the disease, they are associated with PHFs (Brodmann area 22).
Modified after Arendt et al. (2000), Neurobiol Aging 21, 783796
3.2.4.4 Retinoblastoma Protein In AD, increased immunoreactivity for hyperphosphorylated pRb and for
E2F has been described (JordanSciutto et al., 2002; Hoozemans et al., 2004). The upregulation of E2F1 as
it has been detected in pyramidal neurons of Downs patients exhibiting the neuropathological features of
AD at the same time as Ab deposition occurs might thus, be of potential relevance to the process of
neurodegeneration (Motonaga et al., 2001).
3.2.4.5 The Cell Cycle Integrates Intercellular and Intracellular Signaling and Links Development, Oncogenesis,
and Neurodegeneration The high capacity of structural neuronal plasticity in the adult brain might
predispose neurons to tangle formation in AD (Arendt et al., 1995a, 1998a; Arendt 2001a). This high
potential of neuroplasticity associated with the necessity of synaptic turnover and reorganization might
require properties inherent to both growth cones and synaptic connections (Pfenninger et al., 1991). Highly
plastic neurons might thus retain immature features and might not be fully differentiated or truly
postmitotic, i.e., arrested in G0, an assumption supported by recent findings on the expression of cyclin B
and E in hippocampal neurons of healthy elderly (Nagy et al., 1997a; Smith et al., 1999).
It is, therefore, suggested that the reexpression of developmentally regulated genes, the induction of
posttranslational modifications and the accumulation of gene products to an extent, which goes beyond
that observed during regeneration, and the aborted attempt of differentiated neurons to activate the cell
cycle, apparently is a critical event in the pathomechanism of AD (Arendt, 1993, 2001a; Arendt et al., 1998a,
1998c, 2000; Busser et al., 1998; McShea et al., 1997; Masliah et al., 1993c; Nagy et al., 1997a, b, 1998; Smith
et al., 1999; Smith and Lippa, 1995; Vincent et al., 1996, 1997, 1998). This condition is associated with a
re-expression of developmentally regulated genes, the induction of posttranslational modifications and
accumulation of gene products to an extent which goes beyond that observed during regeneration and
might, thus, be due to a loss of differentiation control that normally is involved in the regulation of
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 193
. Table 8-4
Evidence for cell-cycle activation in neurodegenerative disorders
Alzheimers disease
increased neuronal expression of cell-cycle regulators and proliferation Arendt et al. (1996, 1998d, e,
associated proteins 2000)
Cdk1 Blalock et al. (2004)
Cdk4 Busser et al. (1998)
Cdk7 Ding et al. (2000)
Dranovsky et al. (2001)
cyclin A Harris et al. (2000)
cyclin B Husseman et al. (2000)
cyclin C Illenberger et al. (1998)
cyclin D1 Jenkins et al. (1998)
cyclin E Jordan-Sciutto et al. (1999, 2002)
cyclin G1 Kondratick and Vandre (1996)
Ledesma et al. (1992)
Cdk5/p35, p25 Masliah et al. (1993)
McShea et al. (1997)
polo-like kinase Nagy et al. (1997a, b, 2002)
Ogawa et al. (2003)
CDC25A Pei et al. (2002)
CDC25B Pruess and Mandelkow (1998)
Ranganathan et al. (2001)
neuronal CIP-1-associated regulator of cyclin B (CARB) Smith and Lippa (1995)
Smith et al. (1999)
p15INK4b Stieler et al. (2001)
p16INK4a Suzuki et al. (1994)
p18INK4c Tsujioka et al. (1999)
p19INK4d Ueberham et al. (2003)
p21Cip1 Urcelay et al. (2001)
p27Kip1 Vincent et al. (1996, 1997, 1998,
2001)
p105 Yang et al. (2001)
Ki-67 Zhu et al. (2000, 2004)
PCNA
increased phosphorylation of Rb
mitotic phosphoepitopes are absent in immature neurons of the human brain
but reappear in potentially vulnerable neurons prior to neurofibrillary
degeneration
PHFs are associated with mitotic specific phosphoepitopes
regulation of the phosphorylation and metabolism of APP is cell-cycle
dependent
partial tetraploidy of neurons
abnormal mitotic regulation in peripheral cells such as lymphoblasts,
peripheral lymphocytes, or skin fibroblasts
Mild Cognitive Impairment (MCI)
increased neuronal expression of cell-cycle regulators/proliferation associated Yang et al. (2003)
proteins
cyclin D
cyclin B
PCNA
194
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
Vascular dementia
increased neuronal expression of cell-cycle regulators
cyclin B Smith et al. (1999)
Stroke
increased neuronal expression of cell-cycle regulators Love (2003)
Cdk4/
Cdk2
cyclin D1
Parkinsons disease
increased neuronal expression of cell-cycle regulators/proliferation associated Brion and Couck (1995)
proteins
Cdk5 Jordan-Sciutto et al. (2003)
ppRb Smith and Lippa (1995)
Ki-67
Lewy body disease
increased neuronal expression of cell-cycle regulators/proliferation associated Brion and Couck (1995)
proteins
Cdk5 Smith and Lippa (1995)
Ki-67
Picks disease
increased neuronal expression of cell-cycle regulators/proliferation associated Nagy et al. (1997-Acta 94)
proteins
Cdk1 Husseman et al. (2000)
cyclin B Nagy et al. (1997-Acta 93)
Ki-67 Smith and Lippa (1995)
Frontotemporal dementia with Parkinsonism (FTDP-17)
increased neuronal expression of cell-cycle regulators Husseman et al. (2000)
Cdk1
cyclin B
Downs syndrome
increased neuronal expression of cell-cycle regulators/proliferation associated Hermon et al. (2001)
proteins
Cdk1 Husseman et al. (2000)
Ki-67 Nagy et al. (1997a, b)
cyclin B Petersen et al. (2000)
higher incidence of leukemia in Downs patients Schneider and Epstein (1972)
higher incidence of colo-rectal cancer in siblings of Downs patients Smith and Lippa (1995)
altered mitotic index in fibroblasts from Downs patients Vincent et al. (1997)
maternal meiosis II errors in Downs patients are associated with presenilin-1
polymorphism
Temporal lobe epilepsy
increased neuronal expression of cell-cycle regulators Nagy and Esiri (1998)
cyclin B
Progressive supranuclear palsy
increased neuronal expression of cell-cycle regulators/proliferation associated Husseman et al. (2000)
proteins
Cdk1/cyclin B Borghi et al. (2002)
Cdk5 Smith and Lippa (1995)
Ki-67
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 195
Corticobasal degeneration
increased neuronal expression of cell-cycle regulators Husseman et al. (2000)
Cdk1
cyclin B
Huntingtons disease
increased neuronal expression of cell-cycle regulators Husseman et al. (2000)
Cdk1
cyclin B
Amyotrophic lateral sclerosis
increased neuronal expression of cell-cycle regulators Ranganathan and Bowser (2003)
cyclin D Husseman et al. (2000)
cyclin B1 Nakamura et al. (1997)
Cdk5 Bajaj et al. (1998)
phosphorylation of Rb Patzke and Tsai (2002)
Ranganathan et al. (2001)
NiemannPicks disease type C (NPC)
increased neuronal expression of cell-cycle regulators/proliferation associated Husseman et al. (2000)
proteins
Cdk1 Bu et al. (2002 a, b)
cyclin B
Cdk5/p25
presence of mitotic phosphoepitopes: TG-3, MPM-2, H-5
. Figure 8-14
Neuroprotection by ectopic expression of the cyclindependent kinase inhibitor p16INK4a. Microexplants
prepared from rat embryonic day 17 were challenged with 10nM OA (24h), a strong inducer of apoptosis.
Explants transfected with p16INK4a prior to OA treatment showed a much reduced rate of apoptosis (TUNEL,
dUTPrhodamine)
196
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
neuronal plasticity. A direct link between cellcycle reactivation and cell death is supported by observations
on the neuroprotective action of overexpression of the cyclindependent kinase inhibitor p16INK4a, which
locks neurons in a differentiated stage and prevents cellcycle reentry (Arendt, 2003).
Thus, it might be a labile fixation of plastic neurons in G0, which allows for ongoing morphoregu-
latory processes after development is completed. The delicate balance, however, between G0arrest and
G1entry might be prone to a variety of potential disturbances during the lifetime of an individual.
Morphodysregulation in AD, accompanied by aberrancies in intracellular mitogenic signaling, might
thus be a slowly progressing dysfunction, which eventually overrides this differentiation control and results
in dedifferentiation, a condition in conflict with the otherwise mature background of the nervous
system. Cell cycle and differentiation control might thus provide the link between structural brain self
organization and neurodegeneration (Arendt, 1993; Heintz, 1993) both of which in the human brain have
reached a phylogenetic level unique in nature.
4 Conclusions
Taken together, in multicellular organisms, the cell number is regulated spatially by extracellular signals
through cell interactions controlling proliferation and survival in local neighborhood. Instructions from
neighboring cells can induce cell proliferation, differentiation, or death. These stimuli include cellcell and
cellECM adhesion, growth factors, cytokines, neuropeptides, and mechanical factors. Signals from G
proteincoupled receptors, tyrosine kinase receptors, and integrins cooperate to integrate information from
multiple stimuli, which regulate cellcycle progression. To allow for a regulation of these processes, a tight
link is necessary between cell attachment mechanisms and control of proliferation and differentiation, i.e.,
the cellcycle machinery.
Contrary to many other cells that make up a multicellular organism, neurons use the molecular
circuitry developed to sense their relationship with other cells to reorganize their connectivity according
to the requirements for information processing within a cellular network. To form connections and to
reshape them continuously, cell division needs to be shut down. The molecular machinery behind it is now
used to regulate synaptic plasticity. This switch from proliferation control to plasticity control, which
occurs during the process of differentiation of a neuron, implies the permanent risk of erroneously
converting signals derived from plastic synaptic changes into positional cues that will activate the cell
cycle (Arendt, 2003). This replay of developmental mechanisms, i.e., phylogenetic and ontogenetic regres-
sion in the adult nervous system, contrary to development, ultimately results in cell death. Maintaining
neurons in a differentiated but still highly plastic phenotype will thus be the challenge to prevent
neurodegeneration.
References
Abbott LF, Nelson SB. 2000. Synaptic plasticity: Taming the D1 and the Cdk inhibitor p27KIP1. Mol Cell Biol 17:
beast. Nat Neurosci 3(Suppl): 1178-1183. 3850-3857.
Adams PD, Sellers WR, Sharma SK, Wu AD, Nalin CM, et al. Albanese C, Johnson J, Watanabe G, Eklund N, Vu D, et al.
1996. Identification of a cyclinCdk2 recognition motif 1995. Transforming p21Ras mutants and cEts2 activate
present in substrates and p21like cyclindependent kinase the cyclin D1 promoter through distinguishable regions. J
inhibitors. Mol Cell Biol 16: 6623-6633. Biol Chem 270: 23589-23597.
Agarwal ML, Taylor WR, Chernov MV, Chernova OB, Stark Amon A. 1999. The spindle checkpoint. Curr Opin Genet Dev
GR. 1998. The p53 network. J Biol Chem 273: 1-4. 9: 69-75.
Akhtar RS, Ness JM, Roth KA. 2004. Bcl2 family regulation of Aplin AE, Howe AK, Juliano RL. 1999. Cell adhesion mole-
neuronal development and neurodegeneration. Biochem cules, signal transduction and cell growth. Curr Opin Cell
Biophys Acta 1644(23): 189-203. Biol 11: 737-744.
Aktas H, Cai H, Cooper GM. 1997. Ras links growth factor Aplin AE, Howe A, Alahari SK, Juliano RL. 1998. Signal
signaling to the cell cycle machinery via regulation of cyclin transduction and signal modulation by cell adhesion
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 197
receptors: The role of integrins, cadherins, immunoglobulin Arendt Th, Bruckner MK, Gertz HJ, Marcova L. 1998a. Cor-
cell adhesion molecules, and selectins. Pharmacol Rev 50: tical distribution of neurofibrillary tangles in Alzheimers
197-263. disease matches the pattern of neurones that retain their
Arendt T. 2003. Synaptic plasticity and cell cycle activation in capacity of plastic remodelling in the adult brain. Neuro-
neurons are alternative effector pathways the Dr. Jekyll and science 83: 991-1002.
Mr. Hyde theory of Alzheimers disease or the yin and yang Arendt Th, Holzer M, Bruckner MK, Janke C, Gartner U.
of neuroplasticity. Progr Neurobiol 71: 83-248. 1998b. The use of okadaic acid in vivo and the induction
Arendt T, Bruckner MK, Gertz HJ, Marcova L. 1998. Cortical of molecular changes typical for Alzheimers disease. Neu-
distribution of neurofibrillary tangles in Alzheimers roscience 85: 1337-1340.
disease matches the pattern of neurones that retain their Arendt Th, Holzer M, Fruth R, Bruckner MK, Gartner U.
capacity of plastic remodelling in the adult brain. Neuro- 1995c. Paired helical filamentlike phosphorylation of tau,
science 83: 991-1002. deposition of b/A4amyloid and memory impairment in
Arendt T, Gartner U, Seeger G, Barmahenko G, Palm K, et al. rat induced by chronic inhibition of phosphatase 1 and 2A.
2004. Neuronal activation of Ras regulates synaptic con- Neuroscience 69: 691-698.
nectivity. Eur J Neurosci 19: 2953-2966. Arendt Th, Holzer M, Fruth R, Bruckner MK, Gartner U.
Arendt Th. 1993. Neuronal dedifferentiation and degenera- 1998c. Phosphorylation of tau, Abformation, and apopto-
tion in Alzheimers disease. Biol Chem HoppeSeyler 374: sis after in vivo inhibition of PP1 and PP2A. Neurobiol
911-912. Aging 19: 3-13.
Arendt Th. 2000. Alzheimers disease as a loss of differentiation Arendt Th, Holzer M, Gartner U, Bruckner MK. 1998e.
control in a subset of neurons that retain immature features Aberrancies in signal transduction and cell cycle related
in the adult brain. Neurobiol Aging 21: 783-796. events in Alzheimers disease. J Neural Transm 54(Suppl):
Arendt Th. 2001a. Disturbance of neuronal plasticity is a 147-158.
critical pathogenetic event in Alzheimers disease. Int J Arendt Th, Marcova L, Bigl V, Bruckner MK. 1995e. Dendritic
Dev Neurosci 19: 231-245. reorganization in the basal forebrain under degenerative
Arendt Th. 2001b. Alzheimers disease as a disorder of conditions and its defects in Alzheimers disease. I. Den-
mechanisms underlaying structural brain selforganization. dritic organisation of the normal human basal forebrain.
Commentary. Neuroscience 102: 723-765. J Comp Neurol 351: 169-188.
Arendt Th, Bruckner MK. 1992a. Is Alzheimers disease asso- Arendt Th, Rodel L, Gartner U, Holzer M. 1996. Expression of
ciated with reexpression of a developmental protein pat- the cyclindependent kinase inhibitor p16 in Alzheimers
tern? Neurochem Int 21(Suppl): B21. disease. Neuroreport 7: 3047-3049.
Arendt Th, Bruckner MK. 1992b. Perisomatic sprouts immu- Arendt Th, Holzer M, Grossmann A, Zedlick D, Bruckner MK.
noreactive for nerve growth factor receptor and neurofi- 1995d. Increased expression and subcellular transloca-
brillary degeneration affect different neuronal populations tion of the mitogen activated protein kinase kinase and
in the basal nucleus in patients with Alzheimers disease. mitogenactivated protein kinase in Alzheimers disease.
Neurosci Lett 148: 63-66. Neuroscience 68: 5-18.
Arendt Th, Holzer M, Gartner U. 1998d. Neuronal expression Arendt Th, Holzer M, Stobe A, Gartner U, Luth HJ, et al.
of cycline dependent kinase inhibitors of the INK4 family 2000. Activated mitogenic signalling induces a process of
in Alzheimers disease. J Neural Transm 105: 949-960. dedifferentiation in Alzheimers disease that eventually
Arendt Th, Zvegintseva HG, Leontovich TA. 1986. Dendritic results in cell death. Ann NY Acad Sci 920: 249-255.
changes in the basal nucleus of Meynert and in the diagonal Arsura M, Wu M, Sonenshein GE. 1996. TGF beta 1 inhibits
band nucleus in Alzheimers diseasea quantitative Golgi NFkappa B/Rel activity inducing apoptosis of B cells:
investigation. Neuroscience 19: 1265-1278. Transcriptional activation of I kappa B alpha. Immunity
Arendt Th, Bruckner MK, Bigl V, Marcova L. 1995a. Dendritic 5: 31-40.
reorganization in the basal forebrain under degenerative Ashford JW, Jarvik L. 1985. Alzheimers disease: Does neuron
conditions and its defects in Alzheimers disease. II. Ageing, plasticity predispose to axonal neurofibrillary degenera-
Korsakoff s disease, Parkinsons disease, and Alzheimers tion? N Engl J Med 313: 388-389.
disease. J Comp Neurol 351: 189-222. Assoian RK. 1997. Anchoragedependent cell cycle progres-
Arendt Th, Bruckner MK, Bigl V, Marcova L. 1995b. Dendritic sion. J Cell Biol 136: 1-4.
reorganization in the basal forebrain under degenerative Assoian RK, Schwartz MA. 2001. Coordinate signaling by
conditions and its defects in Alzheimers disease. III. The integrins and receptor tyrosine kinases in the regulation
basal forebrain compared to other subcortical areas. of G1 phase cellcycle progression. Curr Opin Genet Dev
J Comp Neurol 351: 223-246. 11: 48-53.
198
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
Assoian RK, Zhu X, 1997. Cell anchorage and the cytoskeleton ciation of human Eg5, a kinesinrelated motor essential for
as partners in growth factor dependent cell cycle progres- bipolar spindle formation in vivo. Cell 83: 1159-1169.
sion. Curr Opin Cell Biol 9: 93-98. Bogdan C. 2001. Nitric oxide and the regulation of gene
Baichwal VR, Baeurle PA, 1997. Activate NFkappa B or die? expression. Trends Cell Biol 11: 66-75.
Curr Biol 7: R94. Boice JA, Fairman R. 1996. Structural characterization of the
Bajaj NP, AlSarraj ST, Anderson V, Kibble M, Leigh N, et al. tumor suppressor p16, an ankyrinlike repeat protein. Pro-
1998. Cyclindependent kinase5 is associated with lipofus- tein Sci 5: 1776-1784.
cin in motor neurones in amyotrophic lateral sclerosis. Borasio GD, John J, Wittinghofer A, Barde YA, Sendtner M,
Neurosci Lett 245(1): 45-48. et al. 1989. Ras p21 protein promotes survival and fiber
Baki L, Marambaud P, Efthimiopoulos S, Georgakopoulos A, outgrowth of cultured embryonic neurons. Neuron 2:
Wen P, et al. 2001. Presenilin1 binds cytoplasmic epithelial 1087-1096.
cadherin, inhibits cadherin/p120 association, and regulates Borel F, Lohez O, Lacroix FB, Margolis RL. 2002. Multiple
stability and function of the cadherin/catenin adhesion centrosomes arise from tetraploidy checkpoint failure
complex. Proc Natl Acad Sci USA 98: 2381-2386. and mitotic centrosome clusters in p53 and RB pocket
Barbacid M. 1987. Ras genes. Ann Rev Biochem 56: 779-827. proteincompromised cells. Proc Natl Acad Sci USA 99:
Barberis L, Wary KK, Fiucci G, Liu F, Hirsch E, et al. 2000. 9819-9824.
Distinct roles of the adaptor protein Shc and focal adhesion Borghi R, Giliberto L, Assini A, Delacourte A, Perry G, et al.
kinase in integrin signaling to ERK. J Biol Chem 275: 2002. Increase of Cdk5 is related to neurofibrillary pathol-
36532-36540. ogy in progressive supranuclear palsy. Neurology 58:
Battle TE, Frank DA. 2002. The Role of STATs in Apoptosis. 589-592.
Curr Mol Med 2: 381-392. Born TL, Frost JA, Schonthal A, Prendergast GC, Feramisco
Beier R, Burgin A, Kiermaier A, Fero M, Karsunky H, et al. JR. 1994. cMyc cooperates with activated Ras to induce the
2000. Induction of cyclin ECdk2 kinase activity, E2F Cdc2 promoter. Mol Cell Biol 14: 5710-5718.
dependent transcription and cell growth by Myc are genet- Bos J. 1989. Ras oncogenes in human cancer: A review. Cancer
ically separable events. EMBO J 19: 5813-5823. Res 49: 4682-4689.
BenAri Y, Represa A. 1990. Brief seizure episodes induce Bothwell M, Giniger E. 2000. Alzheimers disease: Neuro-
longterm potentiation and mossy fibre sprouting in the development converges with neurodegeneration. Cell 102:
hippocampus. Trends Neurosci 13: 312-318. 271-273.
Benn SC, Woolf CJ. 2004. Adult neuron survival strategies Boudreau N, Sympson CJ, Werb Z, Bissell MJ. 1995. Suppres-
slamming on the brakes. Nat Rev Neurosci 5(9): 686-700. sion of ICE and apoptosis in mammary epithelial cells by
BenZeev A. 1997. Cytoskeletal and adhesion proteins as extracellular matrix. Science 267(5199): 891-893.
tumor suppressors. Curr Opin Cell Biol 9: 99-108. Bouman L. 1934. Senile plaques. Brain 57: 128-142.
Biggs JR, Kraft AS. 1995. Inhibitors of cyclindependent ki- Bourne HR, Sanders DA, McCormick F. 1990. The GTPase
nase and cancer. J Mol Med 73: 509-514. superfamily: A conserved switch for diverse cell functions.
Bito H, Deisseroth K, Tsien RW. 1996. CREB phosphorylation Nature 348: 125-132.
and dephosphorylation: A Ca(2) and stimulus duration Braak H, Braak E. 1996. Development of Alzheimerrelated
dependent switch for hippocampal gene expression. Cell neurofibrillary changes in the neocortex inversely recapitu-
87: 1203-1214. lates cortical myelogenesis. Acta Neuropathol 92: 197-201.
Black AR, AzizkhanClifford J. 1999. Regulation of E2F: A Bradbury EM, Inglis RJ, Matthews HR. 1974. Control of cell
family of transcription factors involved in proliferation division by very lysine rich histone (F1) phosphorylation.
control. Gene 237: 281-302. Nature 247: 257-261.
Blagosklonny MV. 2002. Are p27 and p21 cytoplasmic onco- Brehm A, Miska EA, McCance DJ, Reid JL, Bannister AJ, et al.
proteins? Cell Cycle 1(6): 391-393. 1998. Retinoblastoma protein recruits histone deacetylase
Blagosklonny MV. 2003. Apoptosis, proliferation, differen- to repress transcription. Nature 391: 597-601.
tiation: In search of the order. Semin. Cancer Biol 13(2): Brion JP, Couck AM, 1995. Cortical and brainstemtype Lewy
97-105. bodies are immunoreactive for the cyclindependent kinase
Blalock EM, Geddes JW, Chen KC, Porter NM, Markesber WR, 5. Am J Pathol 147(5): 1465-1476.
et al. 2004. Incipient Alzheimers disease: Microarray correla- Brooks G, La Thangue NB. 1999. The cell cycle and drug
tion analyses reveal major transcriptional and tumor suppres- discovery: The promise and the hope. DDT 4: 455-464.
sor responses. Pro Natl Acad Sci USA 101: 2173-2178. Brune B, von Knethen A, Sandau KB. 1999. Nitric oxide
Blangy A, Lane HA, DHerin P, Harper M, Kress M, et al. (NO): An effector of apoptosis. Cell Death Differ 6:
1995. Phosphorylation by p34Cdc2 regulates spindle asso- 969-975.
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 199
Bu B, Li J, Davies P, Vincent I. 2002a. Deregulation of Cdk5, regulated by mitogenactivated protein kinase kinase
hyperphosphorylation, and cytoskeleta pathology in the (MEK1). Proc Natl Acad Sci USA 95: 1091-1096.
NiemannPick type C murine model. J Neurosci 22: 6515- Cheng M, Olivier P, Diehl JA, Fero M, Roussel MF, et al. 1999.
6525. The p21(Cip1) and p27(Kip1) CDK inhibitors are essen-
Bu B, Klunemann H, Suzuki K, Li J, Bird T, et al. 2002b. tial activators of cyclin Ddependent kinases in murine
NiemannPick disease type C yields possible clue for why fibroblasts. EMBO J 18: 1571-1583.
cerebellar neurons do not form neurofibrillary tangles. Coats S, Flanagan WM, Nourse J, Roberts JM. 1996. Require-
Neurobiol Dis 11: 285-297. ment of p27Kip1 for restriction point control of the fibro-
Buell SJ, Coleman PD. 1979. Dendritic growth in the aged blast cell cycle. Science 272: 877-880.
human brain and failure of growth in senile dementia. Cobb BS, Schaller MD, Leu TH, Parsons JT. 1994. Stable
Science 206: 854-856. association of pp60src and pp59fyn with the focal adhe-
Bunz F. 1998. Requirement for p53 and p21 to sustain G2 sionassociated protein tyrosine kinase, pp125FAK. Mol
arrest after DNA damage. Science 282: 1497-1501. Cell Biol 14: 147-155.
Busser J, Geldmacher DS, Herrup K. 1998. Ectopic cell Copani A, Condorelli F, Caruso A, Vancheri C, Sala A, et al.
cycle proteins predict the sites of neuronal cell death in 1999. Mitotic signaling by betaamyloid causes neuronal
Alzheimers disease brain. J Neurosci 18: 2801-2807. death. FASEB J 13: 2225-2234.
Butcher LL, Woolf NJ. 1989. Neurotrophic agents may exac- Copani A, Uberti D, Sortino MA, Bruno V, Nicoletti F, et al.
erbate the pathologic cascade of Alzheimers disease. Neu- 2001. Activation of cell cycleassociated proteins in neuro-
robiol Aging 10: 557-570. nal death: A mandatory of dispensable path? Trends Neu-
Carnero A, Hudson JD, Price CM, Beach DH. 2000. rosci 24: 25-31.
p16INK4A and p19ARF act in overlapping pathways in Coqueret O. 2003. New roles for p21 and p27 cellcycle inhi-
cellular immortalization. Nat Cell Biol 2: 148-155. bitors: A function for each cell compartment? Trends Cell
Carter BD, Lewin GR. 1997. Neurotrophins live or let die: Biol 13: 65-70.
Does p75NTR decide? Neuron 18(2): 187-190. Coqueret O, Gascan H. 2000. Functional interaction of STAT3
Chae T, Kwon YT, Bronson R, Dikkes P, Li E, et al. 1997. Mice transcription factor with the cell cycle inhibitor p21WAF1/
lacking p35, a neuronal specific activator of Cdk5, display CIP1/SDI1. J Biol Chem 275: 18794-18800.
cortical lamination defects, seizures, and adult lethality. CordonCardo C. 1995. Mutations of cell cycle regulators.
Neuron 18(1): 29-42. Biological and clinical implications for human neoplasia.
Chan PY, Kanner SB, Whitney G, Aruffo A. 1994. A trans- Am J Pathol 147: 545-560.
membraneanchored chimeric focal adhesion kinase is Cory S, Adams JM. 2002. The Bcl2 family: Regulators of the
constitutively activated and phosphorylated at tyrosine cellular lifeordeath switch. Nat Rev Cancer 2: 647-656.
residues identical to pp125FAK. J Biol Chem 269: 20567- Cotman CW, Anderson KJ. 1988. Synaptic plasticity and
20574. functional stabilization in the hippocampal formation:
Chan FK, Zhang J, Cheng L, Shapiro DN, Winoto A. 1995. Possible role in Alzheimers disease. Adv Neurol 47:
Identification of human and mouse p19, a novel CDK4 and 313-335.
CDK6 inhibitor with homology to p16ink4. Mol Cell Biol Coucouvanis E, Martin GR. 1995. Signals for death and sur-
15: 2682-2688. vival: A twostep mechanism for cavitation in the vertebrate
Chang BD. 2000. Effects of p21Waf1/Cip1/Sdi1 on cellular embryo. Cell 83: 279-287.
gene expression: Implications for carcinogenesis, senes- Crespo P, Leon J. 2000. Ras proteins in the control of the
cence, and agerelated diseases. Proc Natl Acad Sci USA cell cycle and cell differentiation. Cell Mol Life Sci 57:
97: 4291-4296. 1613-1636.
Chen CY, Faller DV. 1995. Direction of p21 ras generated Danen EHJ, Yamada KM. 2001. Fibronectin, Inegrins, and
signals towards cell growth or apoptosis is determined by Growth Control. J Cell Physiol 189: 1-13.
protein kinase C and Bcl2. Oncogene 11: 1487-1498. Datta SR, Dudek H, Tao X, Masters S, Fu H, et al. 1997. Akt
Chen HC, Guan JL. 1994. Association of focal adhesion kinase phosphorylation of BAD couples survival signals to the
with its potential substrate phosphatidylinositol 3kinase. cellintrinsic death machinery. Cell 91: 231-241.
Proc Natl Acad Sci USA 91: 10148-10152. Dawson TM, Sasaki M, GonzalesZulueta M, Dawson VL.
Chen CS, Mrksich M, Huang S, Whitesides GM, Ingber DE. 1998. Regulation of neuronal nitric oxide synthase and
1997. Geometric control of cell life and death. Science 276: identification of novel nitric oxide signaling pathways.
1425-1428. Progr Brain Res 118: 3-11.
Cheng M, Sexl V, Sherr CJ, Roussel MF. 1998. Assembly Dedhar S. 2000. Cellsubstrate interactions and signaling
of cyclin Ddependent kinase and titration of p27Kip1 through ILK. Curr Opin Cell Biol 12: 250-256.
200
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
Dedhar S, Williams B, Hannigan G. 1999. Integrinlinked Dulic V, Lees E, Reed SI. 1992. Association of human cyclin E
kinase (ILK): A regulator of integrin and growthfactor with a periodic G1S phase protein kinase. Science 257:
signalling. Trends Cell Biol 9: 319-323. 1958-1961.
Delcommenne M, Tan C, Gray V, Rue L, Woodgett J, et al. Durfee T, Becherer K, Chen PL, Yeh SH, Yang Y, et al.
1998. Phosphoinositide3OH kinasedependent regulation 1993. The retinoblastoma protein associates with the
of glycogen synthase kinase 3 and protein kinase B/AKT by protein phosphatase type 1 catalytic subunit. Genes Dev
the integrinlinked kinase. Proc Natl Acad Sci USA 95: 7: 555-569.
11211-11216. Dyer MA, Cepko CL. 2000. p57(Kip2) regulates progenitor
Deora AA, Win T, van Haesebroeck B, Lander HM. 1998. A cell proliferation and amacrine interneuron development
redoxtriggered raseffector interaction. Recruitment of in the mouse retina. Development 127: 3593-605.
phosphatidylinositol 3kinase to Ras by redox stress. Dynlacht BD. 1997. Regulation of transcription by proteins
J Biol Chem 273: 29923-29928. that control the cell cycle. Nature 389: 149-152.
Deshmukh M, Johnson EM Jr. 1998. Evidence of a novel event Earnshaw WC, Martins LM, Kaufmann SH. 1999. Mammali-
during neuronal death: Development of competencetodie in an caspases: Structure, activation, substrates, and functions
response to cytoplasmic cytochrome c. Neuron 21: 695-705. during apoptosis. Ann Rev Biochem 68: 383-424.
De Witt DA, Silver J. 1996. Regenerative failure: A potential Eckert LB, Repasky GA, Ulku AS, McFall A, Zhou H, et al.
mechanism for neuritic dystrophy in Alzheimers disease. 2004. Involvement of Ras activation in human breast
Exp Neurol 142: 103-110. cancer cell signaling, invasion, and anoikis. Cancer Res 64:
Dhavan R, Tsai LH. 2001. A decade of CDK5. Nat Rev Mol 4585-4592.
Cell Biol 2: 749-759. Edlund T, Jessell TM. 1999. Progression from extrinsic to
Di Giovanni S, Knoblach SM, Brandoli C, Aden SA, Hoffman intrinsic signaling in cell fate specification: A view from
EP, et al. 2003. Gene profiling in spinal cord injury shows the nervous system. Cell 96: 211-224.
role of cell cycle in neuronal death. Ann Neurol 53 (4): ElDeiry WS, Tokino T, Velculescu VE, Levy DB, Parsons R,
454-468. et al. 1993. WAF1, a potential mediator of p53 tumor
Diehl JA, Cheng M, Roussel MF, Sherr CJ. 1998. Glycogen suppression. Cell 75: 817-825.
synthase kinase3beta regulates cyclin D1 proteolysis and Evan G, Littlewood T. 1998. A matter of life and cell death.
subcellular localization. Genes Dev 12: 3499-3511. Science 281: 1317-1322.
Ding XL, Husseman J, Tomashevski A, Nochlin D, Jin LW, Evan GI, Brown L, Whyte M, Harrington E. 1995. Apoptosis
et al. 2000. The cell cycle Cdc25A tyrosine phosphatase is and the cell cycle. Curr Opin Cell Biol 7: 825-834.
activated in degenerating postmitotic neurons in Alzhei- Evans T, Rosenthal ET, Youngblom J, Distel D, Hunt T. 1983.
mers disease. Am J Pathol 157: 1983-1990. Cyclin: A protein specified by maternal mRNA in sea ur-
Di Patre PL. 1991. Cytoskeletal alterations might account chin eggs that is. Cell 33: 389-396.
for the phylogenetic vulnerability of the human brain to Facchini LM, Penn LZ. 1998. The molecular role of Myc in
Alzheimers disease. Med Hypotheses 34: 165-170. growth and transformation: Recent discoveries lead to new
Dorsky RI, Chang WS, Rapaport DH, Harris WA. 1997 Regu- insights. FASEB J 12: 633-651.
lation of neuronal diversity in the Xenopus retina by Delta Fan J, Bertino JR. 1997. Kras modulates the cell cycle via both
signalling. Nature 385: 67-70. positive and negative regulatory pathways. Oncogene 14:
Douma S, van Laar T, Zevenhoven J, Meuwissen R, van 2595-2607.
Garderen E, et al. 2004. Suppression of anoikis and induc- Fang F, Orend G, Watanabe N, Hunter T, Ruoslahti E. 1996.
tion of metastasis by the neurotrophic receptor TrkB. Na- Dependence of cyclin ECdk2 kinase activity on cell an-
ture 430: 1034-1039. chorage. Science 271: 499-502.
Downward J. 1998. Lipidregulated kinases: Some common Farinelli SE, Greene LA. 1996. Cell cycle blockers mimosine,
themes at last. Science 279: 673-674. ciclopirox, and deferoxamine prevent the death of PC12
Dranovsky A, Vincent I, GregariI L, Schwarzman A, Colflesh cells and postmitotic sympathetic neurons after removal of
D, et al. 2001. Cdc2 phosphorylation of nucleolin demar- trophic support. J Neurosci 16: 1150-1162.
cates mitotic stages and Alzheimers disease pathology. Farinelli SE, Park DS, Green LA. 1996. Nitric oxide delays the
Neurobiol Aging 22: 517-528. death of trophic factordeprived PC12 cells and sympathet-
Drewes G, LichtenbergKraag B, Doring F, Mandelkow EM, ic neurons by a cGMPmediated mechanism. J Neurosci 16:
Biernat J, et al. 1992. Mitogen activated protein (MAP) 2325-2334.
kinase transforms tau protein into an Alzheimerlike state. Feramisco JR, Gross M, Kamata T, Rosenberg M, Sweet RW.
EMBO J 11: 2131-2138. 1984. Microinjection of the oncogene form of the human
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 201
Hras (T24) protein results in rapid proliferation of quies- Fujita S. 1960. Mitotic pattern and histogenesis of the central
cent cells. Cell 38: 109-117. nervous system. Nature 185: 702-703.
Ferbeyre G, de Stanchina E, Querido E, Baptiste N, Prives C, Galang CK, Der CJ, Hauser CA. 1994. Oncogenic Ras can
et al. 2000. PML is induced by oncogenic ras and promotes induce transcriptional activation through a variety of pro-
premature senescence. Genes Dev 14: 2015-2027. moter elements, including tandem cEts2 binding sites.
Ferreira R, MagnaghiJaulin L, Robin P, HarelBellan A, Oncogene 9: 2913-2921.
Trouche D. 1998. The three members of the pocket proteins Gansauge S, Nussler AK, Berger HG, Gansauge F. 1998. Nitric
family share the ability to repress E2F activity through oxideinduced apoptosis in human pancreatic carcinoma
recruitment of a histone deacetylase. Proc Natl Acad Sci cell lines is associated with a G1arrest and increase of the
USA 95: 10493-10498. cyclindependent kinase inhibitor p21WAF1/CIP1. Cell
Ferrer I, Blanco R, Carmona M, Puig B. 2001. Phos- Growth Differ 9(8): 611-617.
phorylated cMyc expression in Alzheimer disease, Picks Gao CY, Zelenka PS. 1995. Induction of cyclin B and H1
disease, progressive supranuclear palsy and corticobasal kinase activity in apoptotic PC12 cells. Exp Cell Res 219:
degeneration. Neuropathol Appl Neurobiol 27(5): 343-351. 612-618.
Filmus J, Robles AI, Shi W, Wong MJ, Colombo LL, et al. Gartner U, Holzer M, Th. Arendt1999. Elevated expression of
1994. Induction of cyclin D1 overexpression by activated p21Ras is an early event in Alzheimers disease and precedes
ras. Oncogene 9: 3627-3633. neurofibrillary degeneration. Neuroscience 91: 1-5.
Fischer O. 1907. Miliare Necrosen mit drusigen Wucherungen Gartner U, Holzer M, Heumann R, Arendt Th. 1995. Induc-
der Neurofibrillen, eine regelmassige Veranderung der tion of p21Ras in Alzheimer pathology. Neuroreport 6:
Hirnrinde bei seniler Demenz. Monatsschr Psychiatr 1441-1444.
Neurol 22: 361-372. Gartner U, Bruckner MK, Krug S, Schmetsdorf S, Staufenfeld
Fischer A, Sananbenesi F, Schrick C, Spiess J, Radulovic J. M, et al. 2003. Amyloid deposition in APP23 mice is asso-
2002. Cyclindependent kinase 5 is required for associative ciated with the expression of cyclins in astrocytes but not in
learning. J Neurosci 22(9): 3700-3707. neurons. Acta Neuropathol 6: 535-544.
Flechsig P. 1896. Gehirn und Seele. Druck von Alexander Geddes JW, Cotman CW. 1991. Plasticity in Alzheimers dis-
Edelmann, Leipzig. ease: Too much or not enough? Neurobiol Aging 12: 330-333.
Flood DG, Coleman PD. 1990. Hippocampal plasticity in Giaccia AJ, Kastan MB. 1998. The complexity of p53 modula-
normal aging and decreased plasticity in Alzheimers dis- tion: Emerging patterns from divergent signals. Genes Dev
ease. Prog Brain Res 83: 435-443. 12: 2973-2983.
Fogal V, Gostissa M, Sandy P, Zacchi P, Sternsdorf T, et al. Giancotti FG. 1997. Integrin signaling: Specificity and control
2000. Regulation of p53 activity in nuclear bodies by a of cell survival and cell cycle progression. Curr Opin Cell
specific PML isoform. EMBO J 19: 6185-6195. Biol 9: 691-700.
Foster TC. 1999. Involvement of hippocampal synaptic plasticity Gill JS, Winderbank AJ. 1998. Cisplatininduced apoptosis in
in agerelated memory decline. Brain Res Rev 30: 236-249. rat dorsal root ganglion neurons is associated with
Frame S, Balmain A. 2000. Integration of positive and nega- attempted entry into the cell cycle. J Clin Invest 101:
tive growth signals during ras pathway activation in vivo. 2842-2850.
Curr Opin Genet Dev 10: 106-113. Gille H, Downward J. 1999. Multiple ras effector pathways
Freeman RS, Estus S, Johnson EMJR. 1994. Analysis of cell contribute to G(1) cell cycle progression. J Biol Chem 274:
cyclerelated gene expression in postmitotic neurons: Selec- 22033-22040.
tive induction of cyclin D1 during programmed cell death. Giovanni A, WirtzBrugger F, Keramaris E, Slack R, Park DS.
Neuron 12: 343-355. 1999. Involvement of cell cycle elements, cyclindependent
Frisch SM, Francis H. 1994. Disruption of epithelial cell kinases, pRB, and EF2DP, in bamyloid induced neuronal
matrix interactions induces apoptosis. J Cell Biol 124: death. J Biol Chem 274: 19011-19016.
619-626. Giovanni A, Keramaris E, Morris EJ, Hou ST, OHare M, et al.
Frisch SM, Ruoslahti E. 1997. Integrins and anoikis. Curr 2000. E2F1 mediates death of Bamyloidtreated cortical
Opin Cell Biol 9: 701-706. neurons in a manner independent of p53 and dependent
Frisch SM, Screaton RA. 2001. Anoikis mechanisms. Curr on Bax and caspase 3. J Biol Chem 275: 11553-11560.
Opin Cell Biol 13: 555-562. Glasgow JN, Qiu J, Rassin D, Grafe M, Wood T, et al. 2001.
Frisch SM, Vuori K, Ruoslahti E, ChanHui PY. 1996b. Con- Transcriptional regulation of the BCLX gene by NF
trol of adhesiondependent cell survival by focal adhesion kappaB is an element of hypoxic responses in the rat
kinase. J Cell Biol 134: 793-799. brain. Neurochem Res 26(6): 647-659.
202
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
Glotzer M, Murray AW, Kirschner MW. 1991. Cyclin is de- Hamel PA, Gallie BL, Phillips RA. 1992. The retinoblastoma
graded by the ubiquitin pathway. Nature 349(6305): 132- protein and cell cycle regulation. Trends Genet 8: 180-185.
138. Hanahan D, Weinberg RA. 2000. The hallmarks of cancer. Cell
Goedert M, Cohen ES, Jakes R, Cohen P. 1992. p42 MAP 100: 57-70.
kinase phosphorylation sites in microtubuleassociated pro- Hann SR, Thompson CB, Eisenman RN. 1985. cmyc onco-
tein tau are dephosphorylated by protein phosphatase 2A1 gene protein synthesis is independent of the cell cycle in
implication for Alzheimers disease. FEBS Lett 312: 95-99. human and avian cells. Nature 314: 366-369.
Gomez N, Cohen P. 1991. Dissection of the protein kinase Hannon GJ, Beach D. 1994. p15INK4B is a potential effector
cascade by which nerve growth factor activates MAP of TGFbetainduced cell cycle arrest. Nature 371: 257-261.
kinases. Nature 353: 170-173. Hansen LK, Mooney DJ, Vacanti JP, Ingber DE. 1994. Integrin
Gottifredi V, Prives C. 2001. P53 and PML: New partners in binding and cell spreading on extracellular matrix act at
tumor suppression. Trends Cell Biol 11: 184-187. different points in the cell cycle to promote hepatocyte
Gottlieb TM, Oren M. 1998. p53 and apoptosis. Semin Cancer growth. Mol Biol Cell 5: 967-975.
Biol 8: 359-368. Harper JW, Adami GR, Wei N, Keyomarsi K, Elledge SJ. 1993.
Grandori C, Cowley SM, James LP, Eisenman RN. 2000. The The p21 Cdkinteracting protein Cip1 is a potent inhibitor
Myc/Max/Mad network and the transcriptional control of of G1 cyclindependent kinases. Cell 75: 805-816.
cell behavior. Annu Rev Cell Dev Biol 16: 653-699. Harris TE, Albrecht JH, Nakanishi M, Darlington GJ. 2001.
Green DR. 1998. Apoptotic pathways: The roads to ruin. Cell CCAAT/enhancerbinding proteinalpha cooperates with
94: 695-698. p21 to inhibit cyclindependent kinase2 activity and
Green SL, Vulliet PR, Pinter MJ, Cork LC. 1998. Alterations in induces growth arrest independent of DNA binding.
cyclindependent protein kinase 5 (CDK5) protein levels, J Biol Chem 276: 29200-29209.
activity and immunocytochemistry in canine motor neu- Harris PL, Zhu X, Pamies C, Rottkamp CA, Ghanbari HA,
ron disease. J Neuropathol Exp Neurol 57(11): 1070-1077. et al. 2000. Neuronal pololike kinase in Alzheimer dis-
Greenberg SM, Koo EH, Selkoe J, Qiu WQ, Kosik KS. 1994. ease indicates cell cycle changes. Neurobiol Aging 21:
Secreted beta amyloid precursor protein stimulates mito- 837-841.
gen activated protein kinase and enhances tau phosphory- Hayashi T, Sakai K, Sasaki C, Zhang WR, Abe K. 2000.
lation. Proc Natl Acad Sci USA 91: 7104-7108. Phosphorylation of retinoblastoma protein in rat brain
Gu Y, Turck CW, Morgan DO. 1993. Inhibition of CDK2 after transient middle cerebral artery occlusion. Neuro-
activity in vivo by an associated 20K regulatory subunit. pathol Appl Neurobiol 26: 390-397.
Nature 16: 7007-710. Heichman KA, Roberts JM. 1994. Rules to replicate by. Cell
Guan KL, Jenkins CW, Li Y, OKeefe CL, Noh S, et al. 1996. 79(4): 557-562.
Isolation and characterization of p19INK4d, a p16 Heintz N. 1993. Celldeath and the cellcyclea relationship
related inhibitor specific to CDK6 and CDK4. Mol Biol between transformation and neurodegeneration. Trends
Cell 7: 57-70. Biochem Sci 18: 157-159.
Guegan C, Levy V, David JP, AjchenbaumCymbalista F, Sola Hengst L, Reed SI. 1998. Inhibitors of the Cip/Kip family.
B. 1997. cJun and cyclin D1 proteins as mediators of Curr Top Microbiol Immunol 227: 25-41.
neuronal death after a focal ischaemic insult. Neuroreport Hermeking H, Eick D. 1994. Mediation of cMycinduced
8: 1003-1007. apoptosis by p53. Science 265: 2091-2093.
Gulbins E, Bissonnette R, Mahboubi A, Martin S, Nishioka W, Hermeking H, Lengauer C, Polyak K, He TC, Zhang L, et al.
et al. 1995. Immunity 2: 341-351. 1997. 1433 sigma is a p53regulated inhibitor of G2/M
Gulbis JM, Kelman Z, Hurwitz J, ODonnell M, Kuriyan J. progression. Mol Cell 1: 3.
1996. Structure of the Cterminal region of p21(WAF1/ Hermon C, Alberman E, Beral V, Swerdlow AJ. 2001.
CIP1) complexed with human PCNA. Cell 87(2): 297-306. Mortality and cancer incidence in persons with Downs
Guo A, Salomoni P, Luo J, Shih A, Zhong S, et al. 2000a. The syndrome, their parents and siblings. Ann Hum Genet 65:
function of PML in p53dependent apoptosis. Nat Cell Biol 167-176.
2: 730-736. Herrup K, Busser JC. 1995. The induction of multiple cell
Hahn WC, Counter CM, Lundberg AS, Beijersbergen RL, cycle events precedes targetrelated neuronal death. Devel-
Brooks MW, et al. 1999. Creation of human tumour cells opment 121: 2385-2395.
with defined genetic elements. Nature 400: 464-468. Heumann R, Goemans C, Bartsch D, Lingenhohl K, Waldme-
Hall PA, Coates PJ, Ansari B, Hopwood D. 1994. Regulation of ier PC, et al. 2000. Transgenic activation of ras in neurons
cell number in the mammalian gastrointestinal tract: The promotes hypertrophy and protects from lesioninduced
importance of apoptosis. J Cell Sci 107: 3569-3577. degeneration. J Cell Biol 151: 1537-1548.
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 203
Hiebert SW, Packham G, Strom DK, Haffner R, Oren M, et al. Illenberger S, ZhengFischhofer Q, Preuss U, Stamer K,
1995. E2F1:DP1 induces p53 and overrides survival fac- Baumann K, et al. 1998. The endogenous and cell cycle
tors to trigger apoptosis. Mol Cell Biol 15: 6864-6874. dependent phosphorylation of tau protein in living cells:
Hinz M, Krappmann D, Eichten A, Heder A, Sheidereit C, Implications for Alzheimers disease. Mol Biol Cell 9:
et al. 1999. NFkappaB function in growth control: Regula- 1495-1512.
tion of cyclin D1 expression and G0/G1toSphase transi- Impey S, Obrietan K, Wong ST, Poser S, Yano S, et al. 1998.
tion. Mol Cell Biol 19: 2690-2698. Cross talk between ERK and PKA is required for Ca2
Hirai H, Roussel MF, Kato JY, Ashmun RA, Sherr CJ. 1995. stimulation of CREBdependent transcription and ERK
Novel INK4 proteins, p19 and p18, are specific inhibitors of nuclear translocation. Neuron 21: 869-883.
the cyclin Ddependent kinases CDK4 and CDK6. Mol Cell Ino H, Chiba T. 2001. Cyclindependent kinase 4 and cyclin
Biol 15: 2672-2681. D1 are required for excitotoxininduced neuronal cell death
Hitomi M, Stacey DW. 2001. Rasdependent cell cycle com- in vivo. J Neuro Sci 21: 6086-6094.
mitment during G2 phase. FEBS Lett. 490: 123-131. Jeffery KJ, Reid IC. 1997. Modifiable neuronal connec-
Hoffmann I, Clarke PR, Marcote MJ, Karsenti E, Draetta G. tions: An overview for psychiatrists. Am J Psychiatr 154:
1993. Phosphorylation and activation of human Cdc25C 156-164.
by Cdc2cyclin B and its involvement in the selfamplifica- Jeffrey PD, Russo AA, Polyak K, Gibbs E, Hurwitz J, et al.
tion of MPF at mitosis. EMBO J 12: 53-63. 1995. Mechanism of CDK activation revealed by the struc-
Holzer M, Gartner U, Klinz FJ, Narz F, Heumann R, et al. ture of a cyclinACDK2 complex. Nature 376: 313-20.
2001a. Activation of mitogen activated protein kinase cas- Jenkins EC, Ye L, Gu H, Wisniewski HM. 1998. Mitotic index
cade (MAPK)cascade and phosphorylation of cytoskeletal and Alzheimers disease. Neuroreport 9: 3857-3861.
proteins after neuronspecific activation of p21Ras. I. Joneson T, BarSagi D. 1999. Suppression of Rasinduced
MAPKcascade. Neuroscience 105: 1031-1040. apoptosis by the Rac GTPase. Mol Cell Biol 19: 5892-5901.
Holzer M, Rodel L, Seeger G, Gartner U, Narz F, et al. JordanSciutto KL, Malaiyandi LM, Bowser R. 2002. Altered
2001b. Activation of mitogen activated protein kinase distribution of cell cycle transcriptional regulators
(MAPK)cascade and phosphorylation of cytoskeletal pro- during Alzheimer disease. J Neuropathol Exp Neurol 61:
teins after neuronspecific activation of p21Ras. II. Cytoskel- 358-367.
etal proteins and dendritic morphology. Neuroscience 105: JordanSciutto KL, Morgan K, Bowser R. 1999. Increased
1040-1054. cyclin G1 immunoreactivity during Alzheimers disease.
Hoozemanns J, Veerhuis R, Annemieke JM, Rozemuller AJ, J Alzheimers Dis 1: 409-417.
Arendt T, et al. 2004. Neuronal COX2 expression and JordanSciutto KL, Dorsey R, Chalovich EM, Hammond RR,
phosphorylation of pRb precede p38 MAPK activation Achim CL. 2003. Expression patterns of retinoblastoma
and neurofibrillary changes in AD temporal cortex. Neu- protein in Parkinson disease. J Neuropathol Exp Neurol
robiol Dis 15: 492-499. 62: 68-74.
Howe A, Aplin AE, Alahari SK, Juliano RL. 1998. Integrin Kaltschmidt B, Kaltschmidt C, Hofmann TG, Hehner SP,
signaling and cell growth control. Curr Opin Cell Biol 10: Droge W, et al. 2000. The pro or antiapoptotic function
220-231. of NFkappaB is determined by the nature of the apoptotic
Huang S, Chen CS, Ingber DE. 1998. Control of cyclin D1, stimulus. Eur J Biochem 267: 3828-3835.
p27(Kip1), and cell cycle progression in human capillary Kamb A, Gruis NA, WeaverFeldhaus J, Liu Q, Harshman K,
endothelial cells by cell shape and cytoskeletal tension. Mol et al. 1994. A cell cycle regulator potentially involved in
Biol Cell 9: 3179-3193. genesis of many tumor types. Science 264: 436-440.
Hungerford JE, Compton MT, Matter ML, Hoffstrom BG, Kang JS, Krauss RS. 1996. Ras induces anchorageindepen-
Otey CA. 1996. Inhibition of pp125FAK in cultured fibro- dent growth by subverting multiple adhesionregulated cell
blasts results in apoptosis. J Cell Biol 135: 1383-1390. cycle events. Mol Cell Biol 16: 3370-3380.
Hunter T, Pines J. 1994. Cyclins and cancer. II: Cyclin and Katachanov J, Harms C, Gertz K, Hauck L, Waeber C, et al.
CDK inhibitors come of age. Cell 79: 573-582. 2001. Mild cerebral ischemia induces loss of cyclin
Husseman JW, Nochlin D, Vincent I. 2000. Mitotic activation: dependent kinase inhibitors and activation of cell cycle
A convergent mechanism for a cohort of neurodegenerative machinery before delayed neuronal cell death. J Neurosci
diseases. Neurobiol Aging 21: 815-828. 21: 5045-5053.
Ihara Y. 1988. Massive somatodendritic sprouting of cortical Kaufmann JA, Bickford PC, Taglialatela G. 2001. Oxidative
neurons in Alzheimers disease. Brain Res 459: 138-144. stressdependent upregulation of Bcl2 expression in the
Ihle JN. 1996. STATs: Signal transducers and activators of central nervous system of aged Fisher344 rats. J Neuro-
transcription. Cell 84: 331-334. chem 76: 1099-1108.
204
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
Khokhlatchev A, Rabizadeh S, Xavier R, Nedwidek M, Chen T, Kohl NE, Ruley HE. 1987. Role of cmyc in the transformation
et al. 2002. Identification of a novel Rasregulated proa- of REF52 cells by viral and cellular oncogenes. Oncogene 2:
poptotic pathway. Curr Biol 12: 253-265. 41-48.
Khwaja A, RodriguezViciana P, Wennstrom S, Warne PH, Kondo M, Imahori Y, Mori S, Ueda Y, Fujii R, et al. 1999.
Downward J. 1997. Matrix adhesion and Ras transforma- Aberrant plasticity in Alzheimers disease. Neuroreport
tion both activate a phosphoinositide 3OH kinase and 10(7): 1481-1484.
protein kinase B/Akt cellular survival pathway. EMBO J Kondratick CM, Vandre DD. 1996. Alzheimers disease neu-
16: 2783-2793. rofibrillary tangles contain mitosisspecific phosphoepi-
Kim YH, Buchholz MA, Chrest FJ, Nordin AA. 1994. Up topes. J Neurochem 67: 2405-2416.
regulation of cmyc induces the gene expression of the Konishi Y, Lehtinen M, Donovan N, Bonni A. 2002. Cdc2
murine homologues of p34Cdc2 and cyclindependent ki- phosphorylation of BAD links the cell cycle to the cell death
nase2 in T lymphocytes. J Immunol 152: 4328-4335. machinery. Mol Cell 9: 1005-1016.
King KL, Cidlowski JA. 1995. Cell cycle and apoptosis: Korsmeyer SJ. 1999. BCL2 gene family and the regula-
Common pathways to life and death. J Cell Biochem 58: tion of programmed cell death. Cancer Res 59(7 Suppl):
175-180. 1693-1700.
King RW, Jackson PK, Kirschner MW. 1994. Mitosis in tran- Kovesdi I, Reichel R, Nevins JR. 1986. Identification of a
sition. Cell 79: 563-571. cellular transcription factor involved in E1A transactiva-
King WG, Mattaliano MD, Chan TO, Tsichlis PN, Brugge JS. tion. Cell 45: 219-228.
1997 Phosphatidylinositol 3kinase is required for integrin Kranenburg O, Eb van der AJ, Zantema A. 1996. Cyclin D1 is
stimulated AKT and Raf1/mitogenactivated protein ki- an essential mediator of apoptotic neuronal cell death.
nase pathway activation. Mol Cell Biol 17: 4406-4418. EMBO J 15: 46-54.
Kitagawa M, Higashi H, SuzukiTakahashi I, Segawa K, Hanks Kuan CY, Schloemer AJ, Lu A, Burns KA, Wenig WL, et al.
SK, et al. 1995. Phosphorylation of E2F1 by cyclin ACdk2. 2004. Hypoxiaischemia induces DNA synthesis without
Oncogene 10: 229-236. cell proliferation in dying neurons in adult rodent brain.
Kitaura H, Shinshi M, Uchikoshi Y, Ono T, IguchiAriga SM, J Neurosci 24(47): 10763-10772.
et al. 2000. Reciprocal regulation via proteinprotein Kuhl M, Wedlich D. 1997. Wnt signalling goes nuclear. Bioes-
interaction between cMyc and p21(cip1/waf1/sdi1) in says 19: 101-104.
DNA replication and transcription. J Biol Chem 275: Kwon YT, Tsai LH. 1998. A novel disruption of cortical devel-
10477-10483. opment in p35(/) mice distinct from reeler. J Comp
Klein JA, LongoGuess CM, Rossmann MP, Seburn KL, Neurol 395(4): 510-522.
Hurd RE, et al. 2002. The harlequin mouse mutation down- Kyriakis JM, App H, Zhang XF, Banerjee P, Brautigan DL,
regulates apoptosisinducing factor. Nature 419: 367-374. et al. 1992. Raf1 activates MAP kinasekinase. Nature 358:
Knockaert M, Grenngard P, Meijer L. 2002. Pharmacological 417-421.
inhibitors of cyclindependent kinases. Trends Pharmacol Laiho M, Latonen L. 2003. Cell cycle control, DNA damage
Sci 23: 417-418. checkpoints and cancer. Ann Med 35(6): 391-397.
Ko LJ, Prives C. 1996. p53: Puzzle and paradigm. Genes Dev Lander C, Kind P, Maleski M, Hockfield S. 1997a. A family of
10: 1054-1072. activitydependent neuronal cellsurface chondroitin sul-
Kobayashi S, Ishiguro K, Omori A, Takamatsu M, Arioka M, fate proteoglycans in cat visual cortex. J Neurosci 17:
et al. 1993. A Cdc2related kinase PSSALRE/Cdk5 is ho- 1928-1939.
mologous with the 30kDa subunit of tau protein kinase II, Lander HM, Ogiste JS, Pearce SF, Levi R, Novogrodsky A.
a prolinedirected protein kinase associated with microtu- 1995. Nitric oxidestimulated guanine nucleotide exchange
bule. FEBS Lett 335: 171-175. on p21Ras. J Biol Chem 3270: 7017-7020.
Koff A, Ohtsuki M, Polyak K, Roberts JM, Massague J. 1993. Lander HM, Hajjar DP, Hempstead BL, Mirza UA, Chait BT,
Negative regulation of G1 in mammalian cells: Inhibition of et al. 1997b. A molecular redox switch on p21 (ras). Struc-
cyclin Edependent kinase by TGFbeta. Science 260: 536-539. tural basis for the nitric oxidep21 (ras) interaction. J Biol
Koff A, Giordano A, Desai D, Yamashita K, Harper JW, et al. Chem 272: 4323-4326.
1992. Formation and activation of a cyclin ECdk2 complex Lane DP. 1992. Cancer. p53, guardian of the genome. Nature
during the G1 phase of the human cell cycle. Science 257: 358: 15-16.
1689-1694. Lavoie JN, LAllemain G, Brunet A, Muller R, Pouyssegur J.
Koh J, Enders GH, Dynlacht BD, Harlow E. 1995. Tumor 1996. Cyclin D1 expression is regulated positively by the
derived p16 alleles encoding proteins defective in cellcycle p42/p44MAPK and negatively by the p38/HOGMAPK
inhibition. Nature 375: 506-510. pathway. J Biol Chem 271: 20608-20616.
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 205
Ledesma MD, Correas I, Avila J, DiazNido J. 1992. Implica- Liu JJ, Chao JR, Jiang MC, Ng SY, Yen JJ, et al. 1996. Ras
tion of brain Cdc2 and MAP2 kinases in the phosphoryla- transformation results in an elevated level of cyclin D1 and
tion of tau protein in Alzheimers disease. FEBS Lett 308: acceleration of G1 progression in NIH 3T3 cells. Mol Cell
218-224. Biol 15: 3654-3663.
Lee MH, Reynisdottir I, Massague J. 1995a. Cloning of Love S. 2003. Neuronal expression of cell cyclerelated proteins
p57KIP2, a cyclindependent kinase inhibitor with unique after brain ischaemia in man. Neurosci Lett 353: 29-32.
domain structure and tissue distribution. Genes Dev 9: Lowy DR, Willumsen BM. 1993. Function and regulation of
639-649. ras. Ann Rev Biochem 62: 851-891.
Levine AJ. 1997. p53, the cellular gatekeeper for growth and Ludlow JW, Glendening CL, Livingston DM, De Carprio JA.
division. Cell 88: 323-331. 1993. Specific enzymatic dephosphorylation of the retino-
Levine EM, Close J, Fero M, Ostrovsky A, Reh TA. 2000. p27 blastoma protein. Mol Cell Biol 13: 367-372.
(Kip1) regulates cell cycle withdrawal of late multipotent Lundberg AS, Weinberg RA. 1998. Functional inactivation of
progenitor cells in the mammalian retina. Dev Biol 219: the retinoblastoma protein requires sequential modifica-
299-314. tion by at least two distinct cyclinCdk complexes. Mol
Lew DJ. 2000. Cellcycle checkpoints that ensure coordination Cell Biol 18: 753-761.
between nuclear and cytoplasmic events in Saccharomyces Luo Y, Hurwitz J, Massague J. 1995. Cellcycle inhibition by
cerevisiae. Curr Opin Gen Dev 10: 47-53. independent CDK and PCNA binding domains in
Lew DJ, Kornbluth S. 1996. Regulatory roles of cyclin depen- p21Cip1. Nature 375: 159-161.
dent kinase phosphorylation in cell cycle control. Curr Luo RX, Postigo AA, Dean DC. 1998. Rb interacts with his-
Opin Cell Biol 8: 795-804. tone deacetylase to repress transcription. Cell 92: 463-473.
Li BS, Sun MK, Zhang L, Takahashi S, Ma W, et al. 2001. Luth HJ, Arendt Th. 1998a. Oxyde Nitrique et maladie
Regulation of NMDA receptors by cyclindependent dAlzheimer. Alzheimer Actualites No. 132: 6-11.
kinase5. Proc Natl Acad Sci USA 98(22): 12742-12747. Luth HJ, Arendt Th. 1998b. Nitric oxide and Alzheimers
Lin X, Nelson P, Gelman IH. 2000. SSeCKS, a major protein disease. J Brain Res 39: 245-251.
kinase C substrate with tumor suppressor activity, regulates LuthHJ, Holzer M, Gertz HJ, Arendt Th. 2000. Aberrant
G(1)!S progression by controlling the expression and expression of nNOS in pyramidal neurons in Alzheimers
cellular compartmentalization of cyclin D. Mol Cell Biol disease is highly colocalized with p21Ras and p16INK4a.
19: 7259-7272. Brain Res 852: 45-55.
Lin TH, Chen Q, Howe A, Juliano RL. 1997. Cell anchorage Lutterbach B, Hann SR. 1994. Hierarchical phosphorylation
permits efficient signal transduction between ras and its at Nterminal transformationsensitive sites in cMyc pro-
downstream kinases. J Biol Chem 272: 8849-8852. tein is regulated by mitogens and in mitosis. Mol Cell Biol
Lin AW, Barradas M, Stone JC, van Aelst L, Serrano M, et al. 14: 5510-5522.
1998. Premature senescence involving p53 and p16 is acti- Macmanus JP, Koch CJ, Jian M, Walker T, Zurakowski B.
vated in response to constitutive MEK/MAPK mitogenic 1999. Decreased brain infarct following focal ischemia in
signaling. Genes Dev 12: 3008-3019. mice lacking the transcription factor E2F1. Neuroreport 10:
Liou JS, Chen CY, Chen JS, Faller DV. 2000. Oncogenic Ras 2711-2714.
mediates apoptosis in response to protein kinase C inhibi- MagnaghiJaulin L, Groisman R, Naguibneva I. 1998. Retino
tion through the generation of reactive oxygen species. J blastoma protein represes transcription by recruiting a
Biol Chem 275: 39001-390011. histone deacetylase. Nature 391: 601-605.
Liu DX, Greene LA. 2001. Regulation of neuronal survival and Malumbres M, Perez De Castro I, Hernandez MI, Jimenez M,
death by E2Fdependent gene repression and derepression. Corral T, et al. 2000. Cellular response to oncogenic ras
Neuron 32: 425-438. involves induction of the Cdk4 and Cdk6 inhibitor p15
Liu L, Stamler JS. 1999. NO: An inhibitor of cell death. Cell (INK4b). Mol Cell Biol 20: 2915-2925.
Death Differ 6: 937-942. Marshall CJ. 1995. Specificity of receptor tyrosine kinase
Liu F, Stanton JJ, Wu Z, PiwnicaWorms H. 1997. The human signaling: Transient versus sustained extracellular signal
Myt1 kinase preferentially phosphorylates Cdc2 on threo- regulated kinase activation. Cell 80: 179-185.
nine 14 and localizes to the endoplasmic reticulum and Martin DA, Sieggel RM, Zheng L, Lenardo MJ. 1998. Mem-
Golgi complex. Mol Cell Biol 17: 571-583. brane oligomerization and cleavage activates the caspase
Liu W, Bi XN, Tocco G, Baudry M, Schreiber SS. 1996a. 8 (FLICE/MACHalpha1) death signal. J Biol Chem 273:
Increased expression of cyclin D1 in the adult rat 4345-4349.
brain following kainic acid treatment. Neuroreport 7: Martin KC, Michael D, Rose JC, Barad M, Casadio A, et al.
2785-2789. 1997b. MAP kinase translocates into the nucleus of the
206
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
presynaptic cell and is required for longterm facilitation in Mehler MF, Gokhan S. 2001. Developmental mechanisms in
Aplysia. Neuron 18: 899-912. the pathogenesis of neurodegenerative diseases. Progr Neu-
Masliah E, Hansen L, Albright T, Mallory M, Terry RD. 1991b. robiol 63: 337-363.
Immunoelectron microscopic study of synaptic pathology in Meredith JE, Jr, Schwartz M. 1997. Integrins, adhesion and
Alzheimers disease. Acta Neuropathol (Berl) 81: 428-433. apoptosis. Trends Cell Biol 7: 146-150.
Masliah E, Hansen L, Mallory M, Albright T, Terry RD. 1991c. Meredith JE, Jr, FazeliB, SchwartzMA. 1993. The extracellular
Abnormal brain spectrin immunoreactivity in sprouting matrix as a cell survival factor. Mol Biol Cell 4: 953-961.
neurons in Alzheimer disease. Neurosci Lett 129: 1-5. Mesulam MM. 1999. Neuroplasticity failure in Alzheimers
Masliah E, Mallory M, Alford M, Hansen LA, Saitoh T. 1993. disease: Bridging the gap between plaques and tangles.
Immunoreactivity of the nuclear antigen p105 is associated Neuron 24: 521-529.
with plaques and tangles in Alzheimers disease. Lab Invest Migheli A, Piva R, Casolino S, Atzori C, Dlouhy SR, et al.
69: 562-569. 1999. A cell cycle alteration precedes apoptosis of granule
Masliah E, Fagan AM, Terry RD, DeTeresa R, Mallory M, et al. cell precursors in the weaver mouse cerebellum. Am J
1991a. Reactive synaptogenesis assessed by synaptophysin Pathol 155: 365-373.
immunoreactivity is associated with GAP43 in the dentate Mijashita T, Reed JC. 1995. Tumor suppressor p53 is a direct
gyrus of the adult rat. Exp Neurol 113: 131-142. transcriptional activator of the human bax gene. Cell 80:
Mateyak MK, Obaya AJ, Adachi S, Sedivy JM. 1997. Pheno- 293-299.
types of cMycdeficient rat fibroblasts isolated by targeted Miller FD, Kaplan DR. 2001. Neurotrophin signalling path-
homologous recombination. Cell Growth Differ 8: 1039- ways regulating neuronal apoptosis. Cell Mol Life Sci 58(8):
1048. 1045-1053.
Matsushime H, Quelle DE, Shurtleff SA, Shibuya M, Sherr CJ, Mills J, Charest DL, Lam F, Beyreuther K, Ida N, et al. 1997.
et al. 1994. Dtype cyclindependent kinase activity in Regulation of amyloid precursor protein catabolism
mammalian cells. Mol Cell Biol 14: 2066-2076. involves the mitogenactivated protein kinase signal trans-
Mayo MW, Wang CY, Cogswell PL, RogersGraham KS, duction pathway. J Neurosci 17: 9415-9422.
Lowe SW, et al. 1997. Requirement of NFkB activation to Mirmiran M, van Someren EJ, Swaab DF. 1996. Is brain
suppress p53independent apoptosis induced by oncogene plasticity preserved during aging and in Alzheimers dis-
Ras. Science 278: 1812-1815. ease? Behav Brain Res 78: 43-48.
McConnell BB, Gregory FJ, Stott FJ, Hara E, Peters G. 1999. Miyamoto S, Teramoto H, Gutkind JS, Yamada KM. 1996.
Induced expression of p16(INK4a) inhibits both CDK4 Integrins can collaborate with growth factors for phosphor-
and CDK2associated kinase activity by reassortment ylation of receptor tyrosine kinases and MAP kinase acti-
of cyclinCDKinhibitor complexes. Mol Cell Biol 19: vation: Roles of integrin aggregation and occupancy of
1981-1989. receptors. J Cell Biol 135: 1633-1642.
McDonald ER 3rd, ElDeiry WS. 2000. Cell cycle control as a Monteiro HP, Silva EF, Stern A. 2004. Nitric oxide: A potential
basis for cancer drug development. Int J Oncol 16: 871-886. inducer of adhesionrelated apoptosisanoikis. Nitric
McEachern JC, Shaw CA. 1996. An alternative to the LTP Oxide 10: 1-10.
orthodoxy: A plasticitypathology continuum model. Morgan D. 1997. Cyclindependent kinases: Engines, clo-
Brain Res Rev 22: 51-92. cks, and microprocessors. Annu Rev Cell Dev Biol 13:
McFall A, Ulka A, Lambert QT, Kusa A, RogersGraham K, 261-292.
et al. 2001. Oncogenic Ras blocks anoikis by activation of a Morkel M, Wenkel J, Bannister AJ, Kouzarides T, Hagemeier C.
novel effector pathway independent of phosphatidylinosi- 1997. An E2Flike repressor of transcription. Nature 390:
tol 3kinase. Mol Cell Biol 21: 5488-5499. 567-568.
McGill G, Shimamura A, Bates RC, Savage RE, Fisher DE. Moro L, Venturino M, Bozzo C, Silengo L, Altruda F, et al.
1997. Loss of matrix adhesion triggers rapid trans- 1998. Integrins induce activation of EGF receptor: Role in
formationselective apoptosis in fibroblasts. J Cell Biol MAP kinase induction and adhesiondependent cell sur-
138: 901-911. vival. EMBO J 17: 6622-6632.
McKee AC, Kowall NW, Kosik KS. 1989. Microtubular reor- Moroni MC, Hickman ES, Denchi EL, Caprara G, Colli E,
ganization and dendritic growth response in Alzheimers et al. 2001. Apaf1 is a transcriptional target for E2F and
disease. Ann Neurol 26: 652-659. p53. Nat Cell Biol 3: 552-558.
McShea A, Harris PL, Webster KR, Wahl AF, Smith MA. Motonaga K, Itoh M, Hirayama A, Hirano S, Becker LE, et al.
1997. Abnormal expression of the cell cycle regulators 2001. Upregulation of E2F1 in Downs syndrome brain
P16 and CDK4 in Alzheimers disease. Am J Pathol 150: exhibiting neuropathological features of Alzheimertype
1933-1939. dementia. Brain Res 905(12): 250-253.
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 207
Mudryj M, Devoto SH, Hiebert SW, Hunter T, Pines J, et al. Nguyen MD, Bourdreau M, Kriz J, CouillardDespres S,
1991. Cell cycle regulation of the E2F transcription factor Kaplan DR, et al. 2003. Cell cycle regulators in the neuronal
involves an interaction with cyclin A. Cell 65: 1243-1253. death pathway of amyotrophic lateral sclerosis caused by
Mulcahy LS, Smith MR, Stacey DW. 1985. Requirement for mutant superoxide dismutase 1. J Neurosci 23: 2131-2140.
ras protooncogene function during serumstimulated Nikolic M, Chou MM, Lu W, Mayer BJ, Tsai LH. 1998. The
growth of NIH 3T3 cells. Nature 313: 241-243. p35/Cdk5 kinase is a neuronspecific Rac effector that
Mulligan G, Jacks T. 1998. The retinoblastoma gene family: inhibits Pak1 activity. Nature 395(6698): 194-198.
Cousins with overlapping interests. Trends Genet 14: Novak A, Dedhar S. 1999. Signaling through betacatenin and
223-229. Lef/Tcf. Cell Mol Life Sci 56: 523-537.
Murray KD, Gall CM, Jones EG, Isackson PJ. 1994. Differen- OConner L, Huang CS, OReilly L, Strasser A. 2000. Apopto-
tial regulation of brainderived neurotrophic factor and sis and cell division. Cell Biol 12: 257-263.
type II calcium/calmodulindependent protein kinase mes- Ogawa O, Lee HG, Zhu X, Raina A, Harris PL, et al. 2003.
senger RNA expression in Alzheimers disease. Neurosci- Increased p27, an essential component of cell cycle control,
ence 60: 37-48. in Alzheimers disease. Aging Cell 2: 105-110.
Nabel EG. 2002. CDKs and CKIs: Molecular targets for tissue OHare MJ, Hou ST, Morris EJ, Cregan SP, Xu Q, et al. 2000.
remodelling. Nat Rev Drug Discov 1: 587-598. Induction and modulation of cerebellar granule neuron
Nagy Z. 2000. Cell cycle regulatory failure in neurones: Causes death by E2F1. J Biol Chem 275(33): 25358-25364.
and consequences. Neurobiol Aging 21: 761-769. Ohnuma S, Philpott A, Harris WA. 2001. Cell cycle and cell fate
Nagy Z, Esiri MM. 1998. Neuronal cyclin expression in the in the nervous system. Curr Opin Neurobiol 11(1): 66-73.
hippocampus in temporal lobe epilepsy. Exp Neurol 150: Ohnuma S, Philpott A, Wang K, Holt CE, Harris WA. 1999.
240-247. p27Xic1, a Cdk inhibitor, promotes the determination of
Nagy Z, Esiri MM, Smith AD. 1997b. Expression of glial cells in Xenopus retina. Cell 99: 499-510.
cell division markers in the hippocampus in Alzheimers Ohshima T, Ward JM, Huh CG, Longenecker G, Veeranna
disease and other neurodegenerative conditions. Acta Neu- et al. 1996. Targeted disruption of the cyclindependent
ropathol 93: 294-300. kinase 5 gene results in abnormal corticogenesis, neuronal
Nagy Z, Esiri MM, Smith AD. 1998. The cell division cycle pathology and perinatal death. Proc Natl Acad Sci USA
and the pathophysiology of Alzheimers disease. Neurosci- 93(20): 11173-11178.
ence 87: 731-739. Okano HJ, Pfaff DW, Gibbs RB. 1993. RB and Cdc2 expres-
Nagy Z, Combrinck M, Budge M, McShane R. 2002. Cell cycle sion in brain: Correlations with 3Hthymidine incorpora-
kinesis in lymphocytes in the diagnosis of Alzheimers tion and neurogenesis. J Neurosci 13: 2930-2938.
disease. Neurosci Lett 8184. Oktay M, Wary KK, Dans M, Birge RB, Giancotti FG. 1999.
Nagy Z, Esiri MM, Cato AM, Smith AD. 1997a. Cell cycle Integrinmediated activation of focal adhesion kinase is
markers in the hippocampus in Alzheimers disease. Acta required for signaling to Jun NH2terminal kinase and
Neuropathol 94: 6-15. progression through the G1 phase of the cell cycle. J Cell
Nakamura S, Kawamoto Y, Nakano S, Akiguchi I, Kimura J. Biol 145: 1461-1469.
1997. p35nck5a and cyclindependent kinase 5 colocalize in Olson MF, Paterson HF, Marshall CJ. 1998. Signals from Ras
Lewy bodies of brains with Parkinsons disease. Acta Neu- and Rho GTPases interact to regulate expression of p21
ropathol (Berl) 94(2): 153-157. Waf1/Cip1. Nature 394: 295-299.
Nakamura S, Kawamoto Y, Nakano S, Ikemoto A, Akiguchi I, Oppenheim RW. Cell death during development of the ner-
et al. 1997. Cyclindependent kinase 5 in Lewy bodylike vous system. Ann Rev Neurosci 14: (1991) 453501.
inclusions in anterior horn cells of a patient with sporadic OReilly LA, Huang DC, Strasser A. 1996. The cell death
amyotrophic lateral sclerosis. Neurology 48: 267-270. inhibitor Bcl2 and its homologues influence control of
Nathan C. 1992. Nitric oxide as a secretory product of mam- cell cycle entry. EMBO J 15: 6979-6990.
malian cells. FASEB J 3051: 3051-3064. Oren M. 1999. Regulation of the p53 tumor suppressor pro-
Neill D. 1995. Alzheimers disease: Maladaptive synaptoplas- tein. J Biol Chem 274: 36031-36034.
ticity hypothesis. Neurodegeneration 4: 217-232. Orford K, Orford CC, Byers SW. 1999. Exogenous expression
Nevins JR. 1998. Toward an understanding of the functional of betacatenin regulates contact inhibition, anchorage
complexity of the E2F and retinoblastoma families. Cell independent growth, anoikis, and radiationinduced cell
Growth Differ 9: 585-593. cycle arrest. J Cell Biol 146: 855-868.
Nguyen MD, Mushynski WE, Julien JP. 2002. Cycling at the Osuga H, Osuga S, Wang F, Fetni R, Hogan MJ, et al. 2000.
interface between neurodevelopment and neurodegenera- Cyclindependent kinases as a therapeutic target for stroke.
tion. Cell Death Differ 9(12): 1294-1306. Proc Natl Acad Sci USA 97: 10254-10259.
208
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
Padmaja S, Huie RE. 1993. The reaction of nitric oxide with Patzke H, Tsai LH. 2002. Cdk5 sinks into ALS. Trends Neu-
organic peroxyl radicals. Biochem Biophys Res Commun rosci. 25(1): 8-10.
195: 539-544. Paulovich AG, Hartwell LH. 1995. A checkpoint regulates the
Padmanabhan J, Park DS, Greene LA, Shelanski ML. 1999. rate of progression through S phase in S. cerevisiae in
Role of cell cycle regulatory proteins in cerebellar granule response to DNA damage. Cell. 82: 841-7.
neuron apoptosis. J Neurosci 19(20): 8747-8756. Pavletich NP. 1999. Mechanisms of cyclindependent kinase
Palmero I, Pantoja C, Serrano M. 1998. p19ARF links the regulation: Structures of Cdks, their cyclin activators, and
tumour suppressor p53 to Ras. Nature 395: 125-126. Cip and INK4 inhibitors. J Mol Biol 287: 821-828.
Pandey S, Wang E. 1995. Cells en route to apoptosis are Pearson M, Carbone R, Sebastiani C, Cioce M, Fagioli M,
characterized by the upregulation of cfos, cmyc, cjun, et al. 2000. PML regulates p53 acetylation and pre-
Cdc2, and RB phosphorylation, resembling events of early mature senescence induced by oncogenic Ras. Nature 406:
cellcycle traverse. J Cell Biochem 58: 135-150. 207-210.
Pardee AB. 1974. A restriction point for control of normal Peeper DS, Upton TM, Ladha MH, Neuman E, Zalvide J, et al.
animal cell proliferation. Proc Natl Acad Sci USA 71: 1286- 1997. Ras signalling linked to the cellcycle machinery by
1290. the retinoblastoma protein. Nature 386: 177-181.
Pardee AB. 1989. G1 events and regulation of cell prolifera- Pei JJ, Braak H, Gong CX, GrundkeIqbal I, Iqbal K, et al.
tion. Science 246: 603-608. 2002. Upregulation of cell division cycle (Cdc) 2 kinase in
Park DS, Farinelli SE, Greene LA. 1996. Inhibitors of cyclin neurons with early stage Alzheimers disease neurofibrillary
dependent kinases promote survival of postmitotic neu- degeneration. Acta Neuropathol 104: 369-376.
ronally differentiated PC12 cells and sympathetic neurons. Pei JJ, GrundkeIqbal I, Iqbal K, Bogdanovic N, Winblad B,
J Biol Chem 271: 8161-8169. et al. 1998. Accumulation of cyclindependent kinase 5
Park DS, Levine B, Ferrari G, Greene LA. 1997a. Cyclin de- (Cdk5) in neurons with early stages of Alzheimers disease
pendent kinase inhibitors and dominant negative cyclin neurofibrillary degeneration. Brain Res 797(2): 267-277.
dependent kinase 4 and 6 promote survival of NGF Peng CY, Graves PR, Thoma RS, Wu Z, Shaw AS, et al. 1997.
deprived sympathetic neurons. J Neurosci 17: 8975-8983. Mitotic and G2 checkpoint control: Regulation of 1433
Park DS, Morris EJ, Greene LA, Geller HM. 1997b. G1/S cell protein binding by phosphorylation of Cdc25C on serine
cycle blockers and inhibitors of cyclindependent kinases 216. Science 277: 1501-5.
suppress camptothecininduced neuronal apoptosis. J Neu- Penn LJ, Laufer EM, Land H. 1990. cMyc: Evidence
rosci 17: 1256-1270. for multiple regulatory functions. Semin Cancer Biol 1:
Park DS, Morris EJ, Bremner R, Keramaris E, Padmanabhan J, 69-80.
et al. 2000. Involvement of retinoblastoma family members Peters JM. 1998. SCF and APC: The Yin and Yang of cell cycle
and E2F/DP complexes in the death of neurons evoked by regulated proteolysis. Curr Opin Cell Biol 10: 759-768.
DNA damage. J Neurosci 20: 3104-3114. Peters JM, King RW, Hoog C, Kirschner MW. 1996. Identifi-
Park DS, Morris EJ, Padmanabhan J, Shelanski ML, Geller cation of BIME as a subunit of the anaphasepromoting
HM, et al. 1998a. Cyclindependent kinases participate in complex. Science 274(5290): 1199-1201.
death of neurons evoked by DNAdamaging agents. J Cell Petersen MB, Karadima G, Samaritaki M, Avramopoulos D,
Biol 143: 457-467. Vassilopoulos D, et al. 2000. Association between preseni-
Park DS, Morris EJ, Stefanis L, Troy CM, Shelanski ML, et al. lin1 polymorphism and maternal meiosis II errors in
1998b. Multiple pathways of neuronal death induced by Down syndrome. Am J Med Genet 93: 366-372.
DNAdamaging agents, NGF deprivation, and oxidative Petit V, Thiery JP. 2000. Focal adhesions: Structure and dy-
stress. J Neurosci 18: 830-840. namics. Biol Cell 92: 477-494.
Parry D, Mahony D, Wills K, Lees E. 1999. Cyclin DCDK Peunova N, Enikolopov G. 1993. Amplification of calcium
subunit arrangement is dependent on the availability of induced gene transcription by nitric oxide in neuronal cells.
competing INK4 and p21 class inhibitors. Mol Cell Biol Nature 364: 450-453.
19: 1775-1783. Peunova N, Enikolopov G. 1995. Nitric oxide triggers a switch
Patapoutian A, Reichardt LF. 2000. Roles of Wnt proteins in to growth arrest during differentiation of neuronal cells.
neural development and maintenance. Curr Opin Neuro- Nature 375: 68-73.
biol 10: 392-399. Pfenninger KH, de la Houssaye BA, Helmke SM, Quiroga S.
Patrick GN, Zukerberg L, Nikolic M, de la Monte S, Dikkes P, 1991. Growthregulated proteins and neuronal plasticity.
et al. 1999. Conversion of p35 to p25 deregulates Cdk5 A commentary. Mol Neurobiol 5: 143-151.
activity and promotes neurodegeneration. Nature 402: Phelps CH. 1990. Neural plasticity in aging and Alzheimers
615-622. disease: Some selected comments. Prog Brain Res 86: 3-9.
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 209
Phillips LL, Belardo ET. 1994. Increase of cfos and ras onco- Rapoport SI. 1989. Hypothesis: Alzheimers disease is a phy-
proteins in the denervated neuropil of the rat dentate gyrus. logenetic disease. Med Hypoth 29: 147-150.
Neuroscience 58: 503-514. Reginato MJ, Mills KR, Paulus JK, Lynch DK, Sgroi DC, et al.
Polakowska RR, Piacentini M, Bartlett R, Goldsmith LA, 2003. Integrins and EGFR coordinately regulate the pro
Haake AR. 1994. Apoptosis in human skin development: apoptotic protein Bim to prevent anoikis. Nat Cell Biol 5
Morphogenesis, periderm, and stem cells. Dev Dyn 199: (8): 733-740.
176-188. Reiner O. 2000. LIS1. Lets interact sometimes. . . (part 1).
Polyak K, Kato JY, Solomon MJ, Sherr CJ, Massague J, et al. Neuron 28(3): 633-636.
1994a. p27Kip1, a cyclinCdk inhibitor, links transforming Renshaw MW, Ren XD, Schwartz MA. 1997. Growth factor
growth factorbeta and contact inhibition to cell cycle ar- activation of MAP kinase requires cell adhesion. EMBO J
rest. Genes Dev 8: 9-22. 16: 5592-5599.
Polyak K, Lee MH, ErdjumentBromage H, Koff A, Roberts Resnitzky D. 1997. Ectopic expression of cyclin D1 but not
JM, et al. 1994b. Cloning of p27Kip1, a cyclindependent cyclin E induces anchorageindependent cell cycle progres-
kinase inhibitor and a potential mediator of extracellular sion. Mol Cell Biol 17: 5640-5647.
antimitogenic signals. Cell 78: 59-66. Reynisdottir I, Polyak K, Iavarone A, Massague J. 1995. Kip/
Poon RY, Toyoshima H, Hunter T. 1995. Redistribution of the Cip and Ink4 Cdk inhibitors cooperate to induce cell cycle
CDK inhibitor p27 between different cyclin.CDK com- arrest in response to TGFbeta. Genes Dev 9: 1831-45.
plexes in the mouse fibroblast cell cycle and in cells arrested Reynolds JE, Yang T, Qian L, Jenkinson JD, Zhou P, et al. 1994.
with lovastatin or ultraviolet irradiation. Mol Biol Cell 6: Mcl1, a member of the Bcl2 family, delays apoptosis
1197-1213. induced by cMyc overexpression in Chinese hamster
Preuss U, Mandelkow EM. 1998. Mitotic phosphorylation of ovary cells. Cancer Res 54: 6348-6352.
tau protein in neuronal cell lines resembles phosphoryla- Ridley AJ, Paterson HF, Noble M, Land, H. 1988. Ras
tion in Alzheimers disease. Eur J Cell Biol 76: 176-184. mediated cell cycle arrest is altered by nuclear oncogenes
Puthalakath H, Strasser A. 2002. Keeping killers on a tight to induce Schwann cell transformation. EMBO J 7: 1635-
leash: Transcriptional and posttranslational control of the 1645
proapoptotic activity of BH3only proteins. Cell Death Rivard N, LAllemain G, Bartek J, Pouyssegur J. 1996. Abro-
Differ 9(5): 505-512. gation of p27Kip1 by cDNA antisense suppresses quiescence
Radeva G, Petrocelli T, Behrend E, LeungHagesteijn C, Fil- (G0 state) in fibroblasts. J Biol Chem 271: 18337-18341.
mus J, et al. 1997. Overexpression of the integrinlinked Roberts GW, Nash M, Ince PG, Royston MC, Gentleman SM.
kinase promotes anchorageindependent cell cycle progres- 1993. On the origin of Alzheimers disease: A hypothesis.
sion. J Biol Chem 272(21): 13937-13944. Neuroreport 4: 7-9.
Radomski MW, Palmer RM, Moncada S. 1990. An Larginine/ Roovers K, Davey G, Zhu X, Bottazzi ME, Assoian RK. 1999.
nitric oxide pathway present in human platelets regulates Alpha5beta1 integrin controls cyclin D1 expression by sus-
aggregation. Proc Natl Acad Sci USA 87: 5193-5197. taining mitogenactivated protein kinase activity in growth
Raff MC, Barres BA, Burne JF, Coles HS, Ishizaki Y, et al. 1993. factortreated cells. Mol Biol Cell 10: 3197-3204.
Programmed cell death and the control of cell survival: Rostovtseva TK, Antonsson B, Suzuki M, Youle RJ,
Lessons from the nervous system. Science 262: 695-700. Colombini M, et al. 2004. Bid, but not Bax, regulates
Ranganathan S, Bowser R. 2003. Alterations in G(1) to S VDAC channels. J Biol Chem 279(14): 13575-13583.
phase cellcycle regulators during amyotrophic lateral scle- Ruoslahti E, Reed JC. 1994. Anchorage dependence, integrins,
rosis. Am J Pathol 162: 823-835. and apoptosis. Cell 77(4): 477-478.
Ranganathan S, Scudiere S, Bowser R. 2001. Hyperphosphor- Russo T, Faraonio R, Minopoli G, de Candia P, de Renzis S,
ylation of the retinoblastoma gene product and altered et al. 1998. Fe65 and the protein network centered around
subcellular distribution of E2F1 during Alzheimers dis- the cytosolic domain of the Alzheimers betaamyloid pre-
ease and amyotrophic lateral sclerosis. J Alzheimers Dis 3: cursor protein. FEBS Lett 434: 1-7.
377-385. Sa G, Hitomi M, Harwalkar J, Stacey AW, Chen G, Stacey DW
Raoul C, Henderson CE, Pettmann B. 1999. Programmed cell et al. 2002. Ras is active throughout the cell cycle, but is able to
death of embryonic motoneurons triggered through the Fas induce cyclin D1 only during G2 phase. Cell Cycle 1: 50-58.
death receptor. J Cell Biol 147(5): 1049-1062. Sadot E, Jaaro H, Seger R, Ginzburg I. 1998a. Rassignaling
Raoul C, Pettmann B, Henderson CE. 2000. Active killing pathways: Positive and negative regulation of tau expres-
of neurons during development and following stress: A sion in PC12 cells. J Neurochem 70: 428-431.
role for p75(NTR) and Fas? Curr Opin Neurobiol 10(1): Sakurai M, Hayashi T, Abe K, Itoyama Y, Tabayashi K, et al.
111-117. 2000. Cyclin D1 and Cdk4 protein induction in motor
210
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
neurons after transient spinal cord ischemia in rabbits. Serrano M, Hannon GJ, Beach D. 1993. A new regulatory
Stroke 31: 200-207. motif in cellcycle control causing specific inhibition of
Salvesen GS, Dixit VM. 1997. Caspases: Intracellular signaling cyclin D/CDK4. Nature 366: 704-707.
by proteolysis. Cell 91: 443-446. Serrano M, Lin AW, McCurrach ME, Beach D, Lowe SW.
Sastry SK, Lakonishok M, Wu S, Truong TQ, Huttenlocher A, 1997. Oncogenic ras provokes premature cell senescence
et al. 1999. Quantitative changes in integrin and focal associated with accumulation of p53 and p16INK4a. Cell
adhesion signaling regulate myoblast cell cycle withdrawal. 88: 593-602.
J Cell Biol 144: 1295-1309. Sewing A, Wiseman B, Lloyd AC, Land H. 1997. Highinten-
Sauer FC. 1935. Mitosis in the neural tube. J Comp Neurol 62: sity Raf signal causes cell cycle arrest mediated by p21 Cip
377-405. 1. Mol Cell Biol 17: 5588-5597.
Schaller MD, Hildebrand JD, Shannon JD, Fox JW, Vines RR, Shan B, Durfee T, Lee WH. 1996. Disruption of RB/E2F1
et al. 1994. Autophosphorylation of the focal adhesion interaction by single point mutations in E2F1 enhances
kinase, pp125FAK, directs SH2dependent binding of Sphase entry and apoptosis. Proc Natl Acad Sci USA 93:
pp60src. Mol Cell Biol 14: 1680-1688. 679-684.
Scheibel ME, Scheibel AB. 1975. Structural changes in the Sharpe JC, Arnoult D, Youle RJ. 2004. Control of mitochon-
aging brain. In: Aging Clinical, morphologic and neuro- drial permeability by Bcl2 family members. Biochem Bio-
chemical aspects in the aging central nervous system, Vol. 1. phys Acta 1644(23): 107-113.
Harold Brody, Harman D, Mark Ordy J, editors. New York: Sherr CJ. 1993. Mammalian G1 cyclins. Cell 73: 1059-1065.
Raven press; pp. 1137. Sherr CJ. 1994. G1 phase progression: Cycling on cue. Cell 79:
Schlaepfer DD, Hunter T. 1996. Evidence for in vivo phos- 551-555.
phorylation of the Grb2 SH2domain binding site on focal Sherr CJ. 1996. Cancer cell cycles. Science 274: 1672-1677.
adhesion kinase by Srcfamily proteintyrosine kinases. Mol Sherr CJ, Roberts JM. 1995. Inhibitors of mammalian G1
Cell Biol 16: 5623-5633. cyclindependent kinases. Genes Dev 9: 1149-1163.
Schlaepfer DD, Hunter T. 1998. Integrin signalling and tyro- Shi L, Nishioka WK, Thng J, Bradbury EM, Litchfield DW,
sine phosphorylation: Just the FAKs? Trends Cell Biol 8: et al. 1994. Premature p34Cdc2 activation required for
151-157. apoptosis. Science 263: 1143-1145.
Schlaepfer DD, Hauck CR, Sieg DJ. 1999. Signaling through Shimizu S, Matsuoka Y, Shinohara Y, Yoneda Y, Tsujimoto Y.
focal adhesion kinase. Prog Biophys Mol Biol 71: 435-478. 2001. Essential role of voltagedependent anion channel in
Schlaepfer DD, Hanks SK, Hunter T, Geer van der P. various forms of apoptosis in mammalian cells. J Cell Biol
1994. Integrinmediated signal transduction linked to Ras 152(2): 237-250.
pathway by GRB2 binding to focal adhesion kinase. Nature Smith TW, Lippa CF. 1995. Ki67 immunoreactivity in
372: 786-791. Alzheimers disease and other neurodegenerative disorders.
Schlessinger J. 2000. Cell signaling by receptor tyrosine J Neuropathol Exp Neurol 54: 297-303.
kinases. Cell 103: 211-225. Smith MZ, Nagy Z, Esiri MM. 1999. Cell cyclerelated protein
Schmetsdorf S, Gartner U, Arendt T. 2005. Expression of cell expression in vascular dementia and Alzheimers disease.
cyclerelated proteins in developing and adult mouse hip- Neurosci Lett 271: 45-48.
pocampus. Int J Dev Neurosci 23(1): 101-112. Spencer CA, Groudine M. 1991. Control of cmyc re-
Schneider EL, Epstein CJ. 1972. Replication rate and lifespan gulation in normal and neoplastic cells. Adv Cancer Res
of cultured fibroblasts in Downs syndrome. Proc Soc Exp 56: 1-48.
Biol Med 141: 1092-1094. Srinivasula SM, Ahmad M, FernandesAlinemri T, Alnemri
Schoenwaelder SM, Burridge K. 1999. Bidirectional signaling ES. 1998. Autoactivation of procaspase9 by Apaf1
between the cytoskeleton and integrins. Curr Opin Cell Biol mediated oligomerization. Mol Cell 1: 949-957.
11: 274-286. Stieler JT, Lederer C, Bruckner MK, Wolf H, Holzer M, Gertz
Schulze A, ThomeZerfass K, Berge`s J, Middendorp S, Jansen HJ, Arendt T. 2001. Impairment of mitogenic activation of
Durr P, et al. 1996. Anchoragedependent transcription of peripheral blood lymphocytes in Alzheimers disease. Neu-
the cyclin A gene. Mol Cell Biol 16(9): 4632-4638. roreport. 12(18): 39693972.
Schulze A, Zerfass K, Spitkovsky D, Middendorp S, Berge`s J, Strasser A, OConnor L, Dixit VM. 2000. Apoptosis signaling.
et al. 1995. Cell cycle regulation of the cyclin A gene pro- Annu Rev Biochem 69: 217-245.
moter is mediated by a variant E2F site. Proc Natl Acad Sci Strazza M, Luddi A, Brogi A, Carbone M, Riccio M, et al.
USA 92: 11264-11268. 2004. Activation of cell cycle regulatory proteins in the
Seger R, Krebs EG. 1995. The MAPK signaling cascade. FASEB apoptosis of terminally differentiated oligodendrocytes.
J 9: 726-735. Neurochem Res 29: 923-931.
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 211
Stroebe H, 1895. Die allgemeine Histologie der degenerativen Urcelay E, Ibarreta D, Parrilla R, Ayuso MS, MartinRequero
und regenerativen Prozesse im centralen und peripheren A. 2001. Enhanced proliferation of lymphoblasts from
Nervensystem nach den neuesten Forschungen. Central- patients with Alzheimer dementia associated with calmod-
blatt f Allg Pathologie pathol Anatomie 6: 849-960. ulindependent activation of the Na/H exchanger. Neu-
Stromblad S, Becker JC, Yebra M, Brooks PC, Cheresh DA. robiol Dis 8: 289-298.
1996. Suppression of p53 activity and p21WAF1/CIP1 ex- Vachon PH, Loechel F, Xu H, Wewer UM, Engvall E. 1996.
pression by vascular cell integrin alphaVbeta3 during an- Merosin and laminin in myogenesis; specific requirement
giogenesis. J Clin Invest 98: 426-433. for merosin in myotube stability and survival. J Cell Biol
Suzuki T, Oishi M, Marshak DR, Czernik AJ, Nairn AC, et al. 134: 1483-1497.
1994. Cell cycledependent regulation of the phosphoryla- Valencia A, Chardin P, Wittinghofer A, Sander C. 1991. The
tion and metabolism of the Alzheimer amyloid precursor ras protein family: Evolutionary tree and role of conserved
protein. EMBO J 13: 1114-1122. amino acids. Biochemistry 30: 4637-4648.
Swaab DF. 1991. Brain aging and Alzheimers disease, wear van Lookeren Campagne M, Gill R. 1998. Increased expres-
and tear versus use it or lose it. Neurobiol Aging 12: sion of cyclin G1 and p21WAF/CIP1 in neurons following
317-324. transient forebrain ischemia: Comparison with early DNA
Swanton C. 2004. Cellcycle targeted therapies. Lancet Oncol damage. J Neurosci Res 53: 279-296.
5(1): 27-36. Vermeulen K, Bernemann ZN, van Bockstaele DR.. 2003b.
Taglialetela G, Gegg M, PerezPolo JR, Williams LR, Rose GM. Cell cycle and apoptosis. Cell Prolif 36: 165-175.
1996. Evidence for DNA fragmentation in the CNS of aged Vermeulen K, van Bockstaele DR, Berneman ZN. 2003a. The
Fischer344 rats. Neuroreport 7: 977-980. cell cycle: A review of regulation, deregulation and thera-
Tannoch VJ, Hinds PW, Tsai LH. 2000. Cell cycle control. Adv peutic targets in cancer. Cell Prolif 36: 131-149.
Exp Med Biol 465: 127-140. Vidal A, Koff A. 2000. Cellcycle inhibitors: Three families
Taya Y. 1997. RB kinases and RBbinding proteins: New united by a common cause. Gene 247: 1-15.
points of view. Trends Biochem Sci 22: 14-17. Vincent I, Rosado M, Davies P. 1996. Mitotic mechanisms in
Teter B, Ashford W. 2002. Neuroplasticity in Alzheimers Alzheimers disease? J Cell Biol 132: 413-425.
disease. J Neurosci Res 70: 402-437. Vincent I, Jicha G, Rosado M, Dickson DW. 1997. Aberrant
Thornberry NA, Lazebnik Y. 1998. Caspases: Enemies within. expression of mitogenic Cdc2/cyclin B1 kinase in degener-
Science 281: 1312-1316. ating neurons of Alzheimers disease brain. J Neurosci 17:
Timsit S, Rivera S, Ouaghi P, Guischard F, Tremblay E, et al. 3588-3598.
1999. Increased cyclin D1 in vulnerable neurons in the Vincent I, Zheng JH, Dickson DW, Kress Y, Davies P. 1998.
hippocampus after ischaemia and epilepsy: A modulator Mitotic phosphoepitopes precede paired helical filaments
of in vivo programmed cell death? Eur J Neurosci 11: in Alzheimers disease. Neurobiol Aging 19: 287-296.
263-278. Vincent I, Bu B, Hudson K, Husseman J, Nochlin D, et al.
Tomasevic G, Kamme F, Wieloch T. 1998. Changes in 2001. Constitutive Cdc25B tyrosine phosphatase activity in
proliferating cell nuclear antigen, a protein involved in adult brain neurons with M phasetype alterations in
DNA repair, in vulnerable hippocampal neurons following Alzheimers disease. Neuroscience 105: 639-650.
global cerebral ischemia. Brain Res Mol Brain Res 60(2): Vogt C, Vogt O. 1951. Importance of neuroanatomy in the
168-176. field of neuropathology. Neurology 13: 205218.
Troussard AA, Tan C, Yoganathan TN, Dedhar S. 1999. Cell Voitenleitner C, Fanning E, Nasheuer HP. 1997. Phosphoryla-
extracellular matrix interactions stimulate the AP1 tran- tionof DNApolymerase alphaprimase by cyclinAdependent
scription factor in an integrinlinked kinase and glycogen kinases regulates initiation of DNA replication in vitro. Onco-
synthase kinase 3dependent manner. Mol Cell Biol 19: gene 14: 1611-1615.
7420-7427. Walsh TJ, Opello KD. 1992. Neuroplasticity, the aging brain,
Tsujioka Y, Takahashi M, Tsuboi Y, Yamamoto T, Yamada T. and Alzheimers disease. Neurotoxicology 13: 101-110.
1999. Localization and expression of Cdc2 and Cdk4 in Wang N, Butler JP, Ingber DE. 1993. Mechanotransduction
Alzheimer brain tissue. Dement Geriatr Cogn Disord across the cell surface and through the cytoskeleton. Sci-
10(3): 192-198. ence 260: 1124-1127.
Ucker DS. 1991. Death by suicide: One way to go in mamma- Wang R, Brunner T, Zhang L, Shi Y. 1998. Fungal metabolite
lian cellular development? New Biol 3: 103-109. FR901228 inhibits cMyc and Fas ligand expression. Onco-
Ueberham U, Hessel A, Arendt T. 2003. Cyclin C expression is gene 17: 1503-1508.
involved in the pathogenesis of Alzheimers disease. Neu- Wang F, Corbett D, Osuga H, Osuga S, Ikeda JE, et al. 2002.
robiol Aging 24: 427-435. Inhibition of cyclindependent kinases improves CA1
212
8 Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
neuronal survival and behavioral performance after kinase mediated by the Src SH2 domain. Mol Biol Cell 5:
global ischemia in the rat. J Cereb Blood Flow Metab 22: 413-421.
171-182. Xiong W, Pestell R, Rosner MR. 1997. Role of cyclins in
Watanabe H, Pan ZQ, SchreiberAgus N, De Pinho RA, neuronal differentiation of immortalized hippocampal
Hurwitz J, et al. 1998. Suppression of cell transformation cells. Mol Cell Biol 17(11): 6585-6597.
by the cyclindependent kinase inhibitor p57KIP2 requires Yamada KM, Geiger B. 1997. Molecular interactions in cell
binding to proliferating cell nuclear antigen. Proc Natl adhesion complexes. Curr Opin Cell Biol 9: 76-85.
Acad Sci USA 95(4): 1392-1397. Yamaguchi H, Ishiguro K, Uccida T, Takashima A, Lemere CA,
Weber JD, Raben DM, Phillips PJ, Baldassare JJ. 1997. Sus- et al. 1996. Preferential labeling of Alzheimer neurofibril-
tained activation of extracellularsignalregulated kinase 1 lary tangles with antisera for tau protein kinase (TPK)
(ERK1) is required for the continued expression of cyclin I/glycogen synthase kinase3 beta and cyclindependent
D1 in G1 phase. Biochem J 326: 61-68. kinase 5, a component of TPK II. Acta Neuropathol (Berl)
Weinberg RA. 1995. The retinoblastoma protein and cell cycle 92(3): 232-241.
control. Cell 81: 323-330. Yang Y, Geldmacher DS, Herrup K. 2001. DNA replication
Weintraub SJ, Prater CA, Dean DC. 1992. Retinoblastoma precedes neuronal cell death in Alzheimers disease. J Neu-
protein switches the E2F site from positive to negative rosci 21: 2661-2668.
element. Nature 358: 259-261. Yang Y, Mufson EJ, Herrup K. 2003. Neuronal cell death is
Weintraub SJ, Chow KNB, Luo RX, Zhang SH, He S, et al. preceded by cell cycle events at all stages of Alzheimers
1995. Mechanism of active transcriptional repression by the disease. J Neurosci 23: 2557-2563.
retinoblastoma protein. Nature 375: 812-815. Yang J, Winkler K, Yoshida M, Kornbluth S. 1999. Mainte-
Welsh CF, Assoian RK. 2000. A growing role for Rho family nance of G2 arrest in the Xenopus oocyte: A role for 1433
GTPases as intermediaries in growth factor and adhesion mediated inhibition of Cdc25 nuclear import. EMBO J 18:
dependent cell cycle progression. Biochim Biophys Acta 2174-2183.
1471: M21-29. Yin C, Knudson CM, Korsmeyer SJ, van Dyke T. 1997. Bax
Welsh CF, Roovers K, Villanueva J, Liu YQ, Schwartz MA, suppresses tumorigenesis and stimulates apoptosis in vivo.
et al. 2001. Timing of cyclin D1 expression within G1 phase Nature 385: 637-640.
is controlled by Rho. Nature Cell Biol 3: 950-957. Yoshikaqa K. 2000. Cell cycle regulators in neural stem cells
Wen Y, Yang S, Liu R, BrunZinkernagel AM, Koulen P, et al. and postmitotic neurons. Neurosci Res 37: 1-14.
2004. Transient cerebral ischemia induces aberrant neuro- Yu D, Jing T, Liu B, Yao J, Tan M, et al. 1998. Overexpression
nal cell cycle reentry and Alzheimers diseaselike tauopa- of ErbB2 blocks Taxolinduced apoptosis by upregulation
thy in female rats. J Biol Chem 279: 22684-22692. of p21Cip1, which inhibits p34Cdc2 kinase. Mol Cell 2:
Whitmarsh AJ, Davis RJ. 1996. Transcription factor AP1 581-591.
regulation by mitogenactivated protein kinase signal Yun HY, GonzalezZulueta M, Dawson VL, Dawson TM.
transduction pathways. J Mol Med 74: 589-607. 1998. Nitric oxide mediates NmethylDaspartate recep-
Won KA, Reed SI. 1996. Activation of cyclin E/CDK2 is torinduced activation of p21 ras. Proc Natl Acad Sci USA
coupled to sitespecific autophosphorylation and ubiqui- 95: 5773-5778.
tindependent degradation of cyclin E. EMBO J 15: Zhan M, Zhao H, Han ZC. 2004. Signalling mechanisms of
4182-4193. anoikis. Histol Histopathol 19: 973-983.
Woods D, Parry D, Cherwinski H, Bosch E, Lees E, et al. 1997. Zhang HS, Postigo AA, Dean DC. 1999. Active transcriptional
Rafinduced proliferation or cell cycle arrest is determined repression by the RbE2F complex mediates G1 arrest trig-
by the lvel of Raf activity with arrest mediated by p21Cip1. gered by p16INK4a, TGFbeta, and contact inhibition. Cell
Mol Cell Biol 17: 5598-5611. 97(1): 5361.
Woolf NJ, Butcher LL. 1990. Dysdifferentiation of structurally Zhang H, Xiong Y, Beach D. 1993. Proliferating cell nuclear
plastic neurons initiates the pathologic cascade of Alzhei- antigen and p21 are components of multiple cell cycle
mers disease: Toward a unifying hypothesis. Brain cholin- kinase complexes. Mol Biol Cell 4: 897-906.
ergic systems. Steriade M, Biesold D, editors. Oxford Zhang Y, Lu H, Dazin P, Kapila Y. 2004. Squamous cell
University Press; pp. 387-438. carcinoma cell aggregates escape suspensioninduced,
Wu RC, Schonthal AH. 1997. Activation of p53p21waf1 p53mediated anoikis: Fibronectin and integrin alphav me-
pathway in response to disruption of cellmatrix interac- diate survival signals through focal adhesion kinase. J Biol
tions. J Biol Chem 272(46): 29091-29098. Chem 279: 48342-48349.
Xing Z, Chen HC, Nowlen JK, Taylor SJ, Shalloway D, et al. Zhao JH, Reiske H, Guan JL. 1998. Regulation of the cell cycle
1994. Direct interaction of vSrc with the focal adhesion by focal adhesion kinase. J Cell Biol 143: 1997-2008.
Neuronal cellcycle regulation during development and agerelated neurodegenerative disorders
8 213
Zhong S, Salomoni P, Pandolfi PP. 2000b. The transcrip- Zhu X, McShea A, Harris PL, Raina AK, Castellani RJ, et al.
tional role of PML and the nuclear body. Nat Cell Biol 2: 2004. Elevated expression of a regulator of the G2/M phase
E85-90. of the cell cycle, neuronal CIP1associated regulator of
Zhong S, Muller S, Ronchetti S, Freemont PS, Dejean A, et al. cyclin B, in Alzheimers disease. J Neurosci Res 75: 698-703.
2000a. Role of SUMO1modified PML in nuclear body Zhu X, Rottkamp CA, Raina AK, Brewer GJ, Ghanbari HA,
formation. Blood 95: 2748-2752. et al. 2000. Neuronal CDK7 in hippocampus is related to
Zhu X, Ohtsubo M, Bohmer RM, Roberts JM, Assoian RK. aging and Alzheimer disease. Neurobiol Aging 21: 807-813.
1996. Adhesiondependent cell cycle progression linked to Zindy F, Cunningham JJ, Sherr CJ, Jogal S, Smeyne RJ, et al.
the expression of cyclin D1, activation of cyclin ECdk2, 1999. Postnatal neuronal proliferation in mice lacking
and phosphorylation of the retinoblastoma protein. J Cell Ink4d and Kip1 inhibitors of cyclindependent kinases.
Biol 133: 391-403. Proc Natl Acad Sci USA 96: 13462-13467.
9 mRNA Modulations in
Stress and Aging
E. Meshorer . H. Soreq
9 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
Abstract: Ample information suggests multileveled relationships between mRNA modulations and long
term stress responses, especially with regard to alternative splicing. The nature of these relationships is
predictably complex and multileveled. Here, we describe the contribution of both hormonal and stress
induced changes to alternative splicing, and address the molecular pathways that lead to the longlasting
stressassociated modulation in alternative splicing as well as the cellular changes that follow. We cover the
general areas of stress responses (both to psychological and chemical insults) and their longterm con-
sequences, the mechanism of splicing and alternative splicing, and the phenomenon of neuritic transloca-
tion of messenger RNA (mRNA). To present a more detailed example, we describe the alternative splicing
patterns of acetylcholinesterase (AChE) premRNA, which is neuronally expressed, stressresponsive, and
yields dendritically translocated products, and therefore served as an adequate case study for this purpose.
The ACHE gene and its variant AChE protein products are thus presented as both subjects to and putative
mediators of longterm consequences of stress and aging. We attempt to illustrate, citing from the current
literature, how these diverse areas are nonetheless intertwined.
. Figure 9-1
The HPA axis. The regions involved in the activation of the HPA axis are shown. The area enlarged is the
hypothalamic region. The hypothalamus is shaded gray in the inset. See text for details. PVN, paraventricular
nucleus; CRF, corticotrophinreleasing factor, ACTH, adrenocorticotropic hormone; PFC, prefrontal cortex; GC,
glucocorticoids
and vasopressin (AVP) from the cell bodies of the parvicellular hypophysiotropic neurons of the hypothalamic
paraventricular nucleus (PVN) (Sapolsky et al., 2000). The axons of these neurons terminate at the hypotha-
lamic median eminence, where CRF secreted from these neuronal processes diffuses into the hypophyseal
portal circulation, and through the binding of CRF to its specific receptor, CRFR1, it triggers the anterior
pituitary gland to release adrenocorticotropic hormone (ACTH) from specialized cells called corticotropes
into the general circulation (Sapolsky et al., 2000). ACTH then reaches the kidneys adrenal cortex, where it
binds specific cell surface receptors on the adrenal cells in the zona fasciculata, and through cAMPdependent
pathways stimulates the synthesis and secretion of glucocorticoids (GCs) (cortisol in humans and corticoste-
rone in rodents) and mineralocorticoids (aldosterone) into the general circulation (Sapolsky et al., 2000).
GCs, acting back on the hypothalamic PVN, activate a negative feedback loop, which inhibits CRF and hence
ACTH production, consequently decreasing the synthesis and secretion of GCs from the adrenal gland.
218
9 mRNA modulations in stress and aging
as IL1 and IL6, the same ones that are suppressed by GCs or conversely abundantly produced in response
to immunological stimuli (microbes, lipopolysaccharides), can activate the HPA axis and exert additional
effects on the central nervous system (Pruett, 2003).
(Sheline et al., 1996), schizophrenia (Bogerts et al., 1993), agingrelated syndromes (Convit et al., 1995),
Alzheimers disease (de Leon et al., 1993), and finally posttraumatic stress disorder (PTSD) (Bremner et al., 1995).
It is tempting to speculate that GCs and the processes induced under their excess accumulation are
responsible for the hippocampal atrophy associated with the syndromes listed here. Indeed, longlasting
. Figure 9-2
The hippocampus and prefrontal cortex (PFC). Two of the major stressresponding areas in the human brain are
illustrated. The hippocampus is shown as a dark crescent and the PFC is marked with a circle.
220
9 mRNA modulations in stress and aging
elevated glucocorticoid levels were shown to associate with diverse psychological syndromes such as
anxiety, depression, and related mood disorders (HoehnSaric et al., 1991; Yehuda et al., 1993; Calvo and
Volosin 2001; Kalinichev et al., 2002). However, considerable controversy still exists regarding the specific
contributions of the HPA axis to PTSD. Thus, PTSD patients may have normal to low levels of circulating
GCs, while presenting features reminiscent of those of normal individuals with excess GCs (reviewed in
(Yehuda, 2002). The contribution of the initial burst of GCs during trauma or stress leading to the PTSD
phenotype, hence, remained enigmatic.
An additional brain region central to the stress response is the prefrontal cortex (PFC, > Figure 9-2).
Similar to the hippocampus, the PFC is rich in GRs and participates in the feedback regulation loop of the
HPA (Lupien and Lepage, 2001). The PFC is known to serve cognitive and emotional functioning, both of
which are disrupted by stress (Moghaddam, 2002), and chronic stress was shown to induce impairments of
spatial working memory, a type of memory most associated with the PFC (Mizoguchi et al., 2000). Using
in vivo brain imaging methods, such as positron emission tomography (PET) or single photon emission
computed tomography (SPECT), the PFC was identified as a major area affected in a variety of depressive
and mental states (Ebert and Ebmeier, 1996). Indeed, PFC malfunctioning was shown in patients with
PTSD (Shin et al., 2001). PFC efferents and most of the afferents to the PFC, especially those that arrive
from the thalamus and the hippocampus, are glutamatergic. Glutamatemediated neurotransmission is
hence a major player in regulating PFC stress responses (Bagley and Moghaddam, 1997). Nevertheless, the
PFC is by no means restricted to glutamatergic neurotransmission. Levels of other neuromodulators were
found to be modified following stress in the PFC, attesting to the complex functionality of the PFC during
stress. These include serotonin (Kawahara et al., 1993), dopamine (Hamamura and Fibiger, 1993), mono-
amines (Flugge et al., 1997), excitatory amino acids (Moghaddam, 1993), norepinephrine (Nakane et al.,
1994), GABAA (Gruen et al., 1995), and acetylcholine (Mark et al., 1996). The latter poses an interesting
and important link between the glutamatergic and the cholinergic systems in the brain. Cholinergic
glutamatergic interactions have been associated with higher brain functions such as LTP, memory, and
behavior (Aigner, 1995), all of which are affected by stress.
2 PremRNA Splicing
In eukaryotic cells, especially in mammalian cells, the newly transcribed premRNA is subjected to a series
of processing events before leaving the nucleus, reaching its subcellular location, and translating into
protein. These include 50 capping, splicing, cleavage, and 30 polyadenylation (Rio, 1992). The precise
regulation of these events is critical for ensuring the accuracy and the diversity of gene expression.
Particularly interesting is the splicing machinery, which recognizes and removes introns to produce the
mature mRNA (Hastings and Krainer, 2001). In up to 60% of the genes, the splicing machinery further
generates different transcripts from the same transcription unit by selecting or excluding different exons,
a process known as alternative splicing. The different transcripts created must then be spatially and
temporally regulated. Alternative splicing is the key process responsible for the large proteomic diversity,
estimated to produce hundreds of thousands of proteins (Hodges et al., 2002), from merely 20,000 to
30,000 genes by alternative splicing of the majority of the premRNAs (Lander et al., 2001). This estimation
is based on the comparison of expressed sequence tags (ESTs) with genomic data. Therefore, the real
number is likely to be higher, as numerous lowabundance mRNAs are not taken into account, and because
ESTs are biased toward the mRNA 30 ends (Lander et al., 2001). One extreme example includes the
Drosophila Down syndrome cell adhesion molecule (Dscam), encoding an axon guidance receptor shown
to possess more than 38,000 alternatively spliced transcripts (Schmucker et al., 2000).
intermediate. In the second reaction, the two exons are ligated and the lariat is released for nuclear
degradation. In general, for an intron to be recognized, it must include, in addition to the conserved
branchpoint sequence, conserved 50 and 30 splice sites (usually GU and AG, respectively), and a polypyr-
imidine (Py) tract, located close to the 30 end of the intron. These elements recruit the required molecules
carrying out the splicing reaction.
The splicing machinery is mainly governed by small nuclear ribonucleoprotein particles (snRNPs),
which recognize and bind conserved elements within the premRNA. SnRNPs are RNAprotein complexes
made out of five small nuclear RNAs (snRNAs) (U1, U2, U4, U5, and U6) and associated proteins, which
form the core spliceosome (Hastings and Krainer, 2001). As shown in > Figure 9-3, the first steps of
. Figure 9-3
Spliceosome assembly. Early complex (E) involves binding of U1 snRNP to the 50 SS (GU) and the two subunits
(65 and 35) of U2AF to the Py tract and 30 SS (AG), respectively. Recognition of the branch point (A) by U2 snRNP
creates the A complex. Further binding of the U4/U6U5 trisnRNP breaches the association of U1, which is
released with U4, to form the B complex. The 50 splice site then juxtaposes near the branch point to form the C
complex, and the lariatshaped intron is removed
spliceosome assembly involve 50 splice site recognition by U1 and binding of U2 auxiliary factor (U2AF) to
the Py tract and 30 splice site. These early interactions form the E complex. Further base pairing of U2 with
the branchpoint forms the A complex. Subsequent steps include association of the U4/U6U5 trisnRNP
with the spliceosome. Before the first catalytic step, the U4U6 interaction is disrupted and an invariant
motif in U6 replaces U1 snRNA at the 50 splice site (forming the B complex). A specific base pairing between
U2 and U6 snRNAs then juxtaposes the branchpoint and the 50 splice site. U5 snRNP holds onto the 50 exon
intermediate, which has been cut free, and the spliceosome is reconfigured to bring the 30 splice site into the
catalytic core and align the exons for the second catalytic step (C complex).
The spliceosome involves, apart from the above, some additional 150 spliceosomeassociated proteins
(Zhou et al., 2002). NonsnRNPs relevant for splicing catalysis include heterogeneous nuclear ribonucleo-
proteins (hnRNPs), additional RNAbinding proteins, and most important, a family of SR proteins (Zahler
et al., 1992), which promote splicing activity (Zahler et al., 1993) and affect alternative splicing patterns
(Caceres et al., 1994).
2.2 SR Proteins
SR proteins are characterized by a modular structure with a conserved Cterminal arginine/serinerich
domain (RS domain) that interacts with components of the basic splicing machinery, and one or two
RNArecognition motifs (RRMs) (Tacke and Manley, 1999). The RRMs recognize loosely conserved RNA
222
9 mRNA modulations in stress and aging
sequences, identified as either exonic or intronic splicing enhancers (ESE/ISE) or suppressors (ESS/ISS)
(Cartegni et al., 2002). When SR proteins bind to their corresponding ESEs, they recruit the early splicing
component, i.e., U1 snRNP and U2AF to their correct sites (> Figure 9-4). In alternative splicing, ESEs are
. Figure 9-4
ESESR protein interactions. A premRNA containing both introns (white) and exons (dark) is shown. Exons
contain exonic splicing enhancers (ESE) recognized and bound by the RNA recognition motif (RRM) of SR
proteins (ovals). SR proteins recruit U1 snRNP to the 50 splice site and U2AF to the Py tract and the 30 splice site.
Subsequently, U2AF recruits U2 snRNP to the branchpoint (A). SR proteins can also bind intronic splicing
enhancers (ISE) and mediate proteinprotein interactions through their RS domain
usually found in the alternatively spliced exon, and function by recruiting splicing factors to a suboptimal 30
splice site, thus stimulating splicing of an upstream intron or inclusion of an alternative exon (Cartegni
et al., 2002). It was demonstrated, in vitro, that SR proteins alone can influence splice site selection and
hence regulate alternative splicing (Ge and Manley, 1990; Krainer et al., 1990). In vivo, it is likely that the
finetuned balance between SR proteins, hnRNPs, splice sites, and enhancer/silencer elements can be
modulated to achieve a change in the exon usage of premRNA (Hastings and Krainer, 2001). The levels
of SR proteins and hnRNPs vary among tissues (Hanamura et al., 1998) and can be further modulated by
releasing proteins from intranuclear storage compartments, such as the speckle domains at the nuclear
internal boundaries through protein phosphorylation (Mintz and Spector, 2000) and by tissuespecific
factors, especially in the nervous system (Dredge et al., 2001).
Using computational algorithms together with in vivo and in vitro validation, several functional ESEs
were identified (Schaal and Maniatis, 1999; Liu et al., 2000; Cartegni et al., 2002). These were found to be
68 bases long, purinerich sequences that are usually degenerate. In addition to serving in alternative
splicing, these elements may have crucial roles in identifying correct exons for constitutive splicing, as the
other relevant conserved sequences in exons (i.e., splice sites, Py tract, and branchpoint) are weak. If this is
indeed the case, it would explain the somewhat paradoxically high occurrence of ESEs in exons rather than
introns, and would support the model, which suggests that the recognition unit of the spliceosome is the
exon itself rather than the intron (Robberson et al., 1990).
splicing of HSP27 was impeded during heat shock (Bond, 1988), the splicing of the small HSP21 remained
unaffected but the length of its poly(A) tail, which usually confers longer halflife of mammalian mRNAs
(Soreq et al., 1974; Jacobson and Peltz, 1996), was increased (Osteryoung et al., 1993), demonstrating a
finely regulated mechanism and suggesting stressinduced stabilization of HSP21 mRNA. Later, two of the
HSPs participating in the reactivation of splicing following heat shock were identified as HSP70 and
HSP104 (Vogel et al., 1995).
Additional factors contributing to stressinduced splicing alterations include the family of hnRNPM
proteins and the splicingrelated protein 2H9, shown to participate in basal splicing processes as well as in
the early stages of stressinduced splicing arrest (Gattoni et al., 1996; Mahe et al., 1997). Other reports
showed disruption of the U4/U6U5 trisnRNPs complex in HeLa cells under heat stress (Bond, 1988;
Shukla et al., 1990), as well as inactivation of a splicing factor associated with U4/U6U5 (Utans et al., 1992).
Indeed, a heatshockinduced shift from speckled to diffusive distribution of snRNP antigens was reported
in the nucleus of mouse fibroblasts, whereas the splicing factor SC35 retained its normal distribution
(Spector et al., 1991). Remarkably, spatial association between splicing factors and transcriptionally active
genes in nuclear sites of transcription was shown to be intronindependent, as evident by the association of
both an intronrich, HSP90 alpha and the intronless HSP70 protein with nuclear speckles upon exposure
of cells to heat shock (Jolly et al., 1999). These findings indicate an intrinsic, finely tuned targeting capacity
for splicing proteins to potential splice sites during cellular stress.
brain revealed that some of them translocate from the nucleus to the cytosol after an ischemic event, which
could explain the observed changes in splice site selection (Daoud et al., 2002). Activating transcription
factor 3 (ATF3) was shown to possess a specific stressassociated splice variant, ATF3DeltaZip2, modulating
the activity of the normal ATF3 protein during stress responses (Hashimoto et al., 2002). Splice site
selection can hence serve both as a physiologic adaptation to a change in the external conditions and as a
contribution to pathophysiologic events.
Intriguingly, some of the evoked changes in alternative splicing produce transcripts that have antago-
nistic cellular actions. For example, the heatshock factor 4 transcript can generate either an activator or a
repressor of downstream heatshock genes by alternative splicing (Tanabe et al., 1999). Also, alternative
splicing of the apoptotic gene, bclx, substitutes a large protein product, BclxL, which inhibits cell death,
for a smaller one, BclxS, which antagonizes this property under certain conditions (Boise et al., 1993). The
gene encoding neural cell adhesion molecule (NCAM) was shown to undergo alternative splicing during
nitric oxide (NO)induced apoptosis and during neuronal differentiation, producing a smaller protein
product than the normal one, through the action of cJun/AP1 (Feng et al., 2002). This smaller species,
NCAM140, protects cells against NOinduced apoptosis, illustrating the significance of alternative splicing
regulation during stress responses. Most important, a direct link between splicing and apoptosis was
recently shown with the isolation of a protein complex termed the apoptosis and splicingassociated
protein (ASAP). ASAP was shown to disassemble during apoptosis, releasing the splicing activator RNPS1,
which could possibly participate in the splicing and/or alternative splicing of target genes that are activated
during apoptosis (Schwerk et al., 2003), strengthening the stress/apoptosis and splicing relationship.
An examination of the cases studied so far (e.g., GR, Slo, PS2, ATF3, HSF4, Bclx, NCAM, and AChE,
see later) supports the notion that alternative splicing serves as an adaptation strategy. Thus, alternative
splicing in these cases yields specific splice variants, which facilitate beneficial stress responses, at least in the
short range. For example, the alternative products of GR, Slo, ATF3, HSF4, Bclx, and NCAM actively
participate in adapting to the novel conditions, either by generating an antagonizer to the action of its
normally spliced counterpart (e.g., ATF3, HSF4, Bclx) or by generating a protective protein (e.g., GR,
NCAM). In the case of the potassium ion channel, Slo, a calciumcalmodulin dependent protein kinase
(CaMK) IV responsive RNA element (CaRRE) was identified. To demonstrate its function, this element was
experimentally transferred from the 30 splice site upstream of STREX, the alternatively spliced product of
Slo, to a heterologous gene. The otherwise constitutive acceptor sequence became sensitive to CaMK IV
(Xie and Black, 2001). This sets an intriguing example for the role of the premRNA sequence itself in
stress responses, which in this particular case, introduces longterm adaptive changes to the neuronal
electrophysiological properties.
between the alternative RNA targets of ASF/SF2 and SC35. Both proteins should hence be tested in the
poststress brain.
Identified more than 15 years ago (Davis et al., 1987), dendritic transport of mRNAs is known today to be a
feature of over 400 different transcripts, comprising about 5% of the total expressed neuronal mRNAs
(Eberwine et al., 2001). Dendritic targeting of mRNAs and local protein synthesis were both shown to be
associated with hippocampal synaptic plasticity (Kang and Schuman, 1996), as well as with memory storage
(Mayford et al., 1996), predicting involvement with aging and stress.
As expected from a process regulating synaptic function and plasticity, targeting of mRNAs to dendrites
was shown to be stimulusinduced (Tongiorgi et al., 1997; Steward et al., 1998). Yet more specifically,
potassium depolarization was demonstrated to induce calciumdependent transport of the mRNAs for
brain derived neurotrophic factor (BDNF) and the tropomyosinrelated kinase B (TrkB) BDNF receptor
into dendrites. This translocation did not require newly synthesized RNA, yet was followed by local protein
synthesis (Tongiorgi et al., 1997). In another study, the mRNA for the immediateearly gene activity
regulated cytoskeletonassociated protein (Arc) was shown to translocate selectively to activated dendritic
segments in the dentate gyrus. In this case as well, mRNA translocation was followed by new synthesis of the
Arc protein (Steward et al., 1998). Activitydependent mRNA targeting may thus provide a mechanism for
synaptic modulation, which requires local protein translation of specific transcripts. The recent generation
of mice lacking dendritic transport ability and local protein synthesis of calcium and calmodulin
dependent protein kinase II alpha (CaMKIIa) provided a direct proof that local translation is required
for LTP and memory consolidation (Miller et al., 2002).
The translocation of mRNAs into neuronal processes is generally considered to be associated with cis
acting elements, primarily within the 30 untranslated region (UTR) of the mRNA (Steward and Schuman,
2001). However, the length of the sequence required for dendritic targeting varies considerably. For
example, a 21 nucleotide long sequence in the 30 UTR of the myelin basic protein (MBP) mRNA was
shown to be sufficient for dendritic translocation (Ainger et al., 1997), whereas in another study,
640 nucleotides from the 30 UTR of the microtubuleassociated protein 2 (MAP2) were shown to be essential
for mRNA targeting (Blichenberg et al., 1999). For aCaMKII, 94 nucleotides from the 30 UTR were shown
to mediate its dendritic localization in cultured hippocampal neurons (Mori et al., 2000), whereas in a
different study, a 1,230 nucleotide region of the 30 UTR was shown to be responsible (Blichenberg et al.,
2001). These apparent discrepancies could be settled if the secondary structure, rather than primary
sequence, was responsible for targeting. Indeed, careful analyses demonstrated that dendritically localized
mRNAs do not share a common consensus sequence (Tiedge et al., 1999). It is therefore likely that
RNA folding into secondary structures might act as the driving mechanism (Ainger et al., 1997; Blichenberg
et al., 1999).
Alternative splicing of a growing number of gene products can be found in numerous types of cancerous
cells, although not abundant in the corresponding normal cells (HerreraGayol and Jothy, 1999; Caballero
et al., 2001; Perry et al., 2002), probably reflecting a general modification of the processing machinery
during tumorogenesis (Scorilas et al., 2001; Stoilov et al., 2002). In humans, SC35 expression is altered
under other diseases as well, e.g., HIV infection (Maldarelli et al., 1998) and also during pregnancy (Nie
et al., 2000). ASF/SF2 too is spatiotemporally regulated in the uterine myometrium during pregnancy
(Pollard et al., 2000), again, probably insinuating a more general phenomenon for the regulation of
splicingrelated genes during various stresses or under adaptation to altered conditions. A striking example
of a splicingrelated protein, which directly affects the severity of disease state, is illustrated by the putative
splicing factor SCNM1 (sodium channel modifier 1). A single point mutation changes a donor splice site in
226
9 mRNA modulations in stress and aging
the sodium channel gene Scn8a leading to chronic movement disorder. In one mouse strain, C57BL/6J, an
additional modifier nonsense mutation in SCNM1 exacerbates the Scn8a disorder into a lethal neurological
disease. The effect of the modifier mutation is a reduction of the abundance of correctly spliced sodium
channel transcripts below the threshold that is essential for survival (Buchner et al., 2003). This example
demonstrates how genetic variation in a splicing factor influences disease susceptibility.
A large number of mutations lead to aberrant splicing (Stoilov et al., 2002). These can directly disrupt
or create acceptor or donor splice sites, or alternatively, disrupt exonic or intronic splicing enhancers or
suppressors. Intriguingly, in a few cases, diseaseassociated alternative or aberrant splicing was reported, but
the culprit gene was found to be devoid of any splicingassociated mutations. These are particularly relevant
examples for this discussion, because a splicing regulatory protein is likely to be at fault, causing the
abnormal phenotype. We present several examples to illustrate each of these cases.
Familial hypercholesterolemia is an autosomal dominant genetic lipoprotein disorder caused by defects
within the lowdensity lipoprotein receptor gene. Some of its many mutations were shown to disrupt
acceptor splice sites: an A to G substitution in the penultimate nucleotide of intron 16 (Lombardi et al.,
1993) and a G to C transposition at the last nucleotide of intron 7 (Yu et al., 1999). In both cases, a cryptic
splice site was activated, leading to the formation of a mutated receptor protein. Another example of a splice
site mutation that leads directly to a disease phenotype is molybdenum cofactor deficiency (MoCoD), an
inherited autosomal recessive disease that leads to early childhood death. Two bicistronic genes, MOCS1
and MOCS2, are responsible for the generation of the molybdoenzymes, sulfite oxidase, xanthine dehy-
drogenase, and aldehydeoxidase. An MOCS1 splice site mutation leads to the deficiency of all these
molybdenum cofactorrelated enzymes (Reiss et al., 1999).
An increasing number of alterations, both in normal and in alternative splicing, are being linked to
defects in enhancer/silencer sequences (Philips and Cooper, 2000; Cartegni et al., 2002). A number of these
mutations are present in an exon, but do not change the protein sequence. Because they cause aberrant
splice site selection, they can cause a disease, although they do not modify mRNA translation. Examples of
these mutations include a T to C (L284L) mutation in tau exon 10 that disrupts an exonic splicing enhancer,
causing frontotemporal dementia with parkinsonism linked to chromosome 17 (FTDP17) (DSouza et al.,
1999), a C to G (R28R) mutation in porphobilinogen deaminase that results in exon 3 skipping, and that
causes one of the many forms of porphyria (Llewellyn et al., 1996) and an A to G mutation in pyruvate
dehydrogenase resulting in Leighs encephalomyelopathy (De Meirleir et al., 1994).
Mutations that alter the splicing process can occur outside of both splice sites and enhancer/silencer
elements. An example is the mutation described in a patient with ataxiatelangiectasia (ATM), which
represents a deletion in an intronsplicing processing element crucial for accurate intron removal (Pagani
et al., 2002). The deletion of four nucleotides in this element abolishes a binding site for U1 snRNP and
leads to activation of a cryptic exon, thus producing an abnormal mRNA transcript.
Changes in splicing factors can also be phenotypic. The appearance of splicingassociated diseases is
often correlated with impaired plasticity and longevity of the affected cells. For example, mutations in the
human homologs of the splicing factors PRP31 or PRPC8 may cause retinitis pigmentosa, a progressive loss
of rods and cones, which causes loss of over 90% of vision during childhood (McKie et al., 2001; Vithana
et al., 2001).
An example of a disease with unexplained aberrant splicing includes sporadic amyotrophic lateral
sclerosis (ALS), which was shown to be associated with a change in the splicing patterns of the mRNAs for
the glutamate transporter EAAT2 (Lin et al., 1998), as well as nitric oxide synthase (NOS) (Catania et al.,
2001). This indicates a basic defect in the premRNA processing machinery, which gives rise to altered
isoforms associated with the disease.
To a certain extent, premRNA processing thus appears to have the potential to adapt the information
stored in the genome to the physiologic requirements of circumstance, place, and time. The failure of such
adaptation is frequently traced to defects in processing. These defects can manifest themselves directly in a
disease or may remain silent until some internal or environmental stimulus (e.g., stress) or time (e.g., old
age) allows the mutation to become apparent. Furthermore, the sequential nature of premRNA processing
raises the interesting possibility that pleiotropic diseases with a variable phenotype may be caused by
specific compositions of transacting factors.
mRNA modulations in stress and aging
9 227
. Table 9-1
Agingassociated changes in alternative splicing
Gene Organism Tissue Affected process Reference
APP, APLPs rat Brain amyloid plaque (Sandbrink et al., 1997)
Dopamine D2 rat Neostriatal dopaminergic (Merchant et al., 1993)
receptor subregions neurotransmission
NCAM rat Heart neural adhesion (Andersson et al., 1993)
Insulin receptor rat Liver, muscle, heart insulin responses (Vidal et al., 1995)
MGF rat Skeletal muscle insulin responses (Owino et al., 2001)
COX1 rat Stomach gastric cytoprotection (Vogiagis et al., 2000)
NF1 Human Blood rRas signaling, (Wimmer et al., 2000)
neural architecture
emerges from these studies is of a general control switch that causes multiple aginginduced impairments
in alternative splicing. For example, such impairments could all reflect one or a few changes in regulatory
upstream factor(s). Support for this concept can be found in more recent microarray screens for changes in
gene expression during aging. Significant agerelated tissuespecific changes were found in the expression
levels of several RNA processing genes in aged, compared to young, gastrocnemius muscle, neocortex, and
cerebellum of C57BL/6 mice (Lee et al., 1999). Changes spanned SR proteins, hnRNPs, as well as 30 end
228
9 mRNA modulations in stress and aging
. Table 9-2
Agingassociated alterations in the expression of murine premRNA processing genes
Fold Affected premRNA Reference to
Gene changea Tissue processing step reported function
SF3A2 3.4 Cerebellum U2 snRNP binding (Ruby et al., 1993)
(SAP62,
PRP11)
U2AF65 3.2 Gastrocnemius muscle Splice site recognition, splicing (Wang et al., 1995)
Sox17 2.4 gastrocnemius muscle Transcription, Splicing (Ohe et al., 2002)
PRP22 (HRH1) 2.5 neocortex Spliceosome disassembly, (Company et al.,
mRNA release 1991)
PTB 2.3 cerebellum Alternative splicing (Wagner and
GarciaBlanco 2001)
PolyA RNA 2.1 gastrocnemius muscle 30 end processing, mRNA
export protein export
hnRNP H3 2.0 cerebellum Splicing, heatshock splicing (Mahe et al., 1997)
(hnRNP 2H9) arrest
CStF 0.7 neocortex Transcript cleavage
DDX18 0.7 neocortex DEADbox protein, specific
function unknown
PRP16 0.6 cerebellum Aberrant lariat discard (Burgess and
Guthrie 1993)
PABI 0.4 cerebellum mRNA stability
PRP22 0.4 cerebellum Spliceosome disassembly, (Company et al.,
mRNA release 1991)
a
From Lee et al. (1999)
processing factors (> Table 9-2). Because both SR proteins and hnRNPs influence splice site selection,
misdirected regulation of their expression levels can lead to alternative or aberrant splicing of many
downstream target sequences.
One of the first essential splicing factors to be identified was the 65 kD subunit of the U2 auxiliary factor
(U2AF65). U2AF65 binding to the conserved Py tract at the 30 end of introns is required for the subsequent
binding of U2 snRNA to the 50 splice site. Another subunit protein of this complex, U2AF35, binds the 30
splice site. U2AF65 mRNA levels were found to increase by over threefold in the gastrocnemius muscle of
aged mice (> Table 9-2). As U2AF65 contains both the RNA recognition motif and the RS domain of
SRproteins, it can regulate splicing in a concentrationdependent manner. Uncontrolled regulation of the
expression level of U2AF65 may therefore lead to aberrant regulation of splicing activity and alternative
splicing of numerous target genes (Wang et al., 1995) (> Figure 9-5).
The Py tractbinding protein (PTB, hnRNP I) also binds intronic Py tracts. PTB was shown to possess a
direct role in premRNA alternative splicing (Wagner and GarciaBlanco, 2001), and its expression levels
were reported to increase over twofold in the cerebellum of aged mice, which can potentially harm finely
regulated alternative splicing of many premRNAs (> Figure 9-5).
Other factors that may have a direct or indirect role in influencing splicing during aging are PRP22,
PRP16, SF3A2, DDX18, and hnRNP H3 (Table 1-2). The expression level of the splicing factor PRP22, for
example, is 2.5fold increased in the neocortex of aged C57BL/6 mice, whereas the same factor displays
2.7fold decrease in the cerebellum of these mice. Although the function of several of these factors is still
largely obscure, this expanded list points to a central involvement of splicing changes in the aging process
(> Figure 9-5).
Another checkpoint, which may affect agingrelated changes in premRNA splicing, lies upstream, at
the earlier phases controlling the initiation and efficacy of transcription. This involves the expression levels
mRNA modulations in stress and aging
9 229
. Figure 9-5
Misdirected regulation of premRNA processing during aging. Factors that were reported to change during
aging are depicted in the figure, along with the different steps of premRNA processing, in which they are
involved in normal and aged cells. a: The naked premRNA transcript. Exons are dark and introns are marked
white. The 50 (GU) and the 30 (AG) splice sites, the branchpoint (A), and the polypyrimidine tract (Py) are shown.
The 50 end of the premRNA is capped shortly after initiation of transcription, b: Initial snRNP complex
formation. U2AF65 and PTB were shown to be significantly overproduced in the gastrocnemius muscle and
cerebellum, respectively, of aged C57BL/6 mice, thus, potentially affecting splicing, c: Spliceosome assembly
and attraction of SR proteins. The splicingrelated factor/RNA helicase PRP16 was shown to be twofold
decreased in the cerebellum, d: Intron release and exon joining. Altered expression level of the different
factors shown earlier may lead to the alternatively spliced transcript shown, e: 30 end processing. The cleavage
and polyadenylation complex, which consists of CPSF, which recognizes and binds the poly(A) site (AAUAAA),
CStF, CFI, CFII, and PAP, which catalyze the first, slow, polyadenylation step are shown. PABII then binds the
short, newly formed, poly(A) tail and rapidly adds additional 200250 adenines. CStF and PAP were both about
1.5fold decreased in the neocortex of C57BL/6 aged mice, potentially harming the 30 end processing pathway
and the splicing pathway. Abbreviations: Py, polypyrimidine tract; snRNP, small nuclear ribonucleoprotein
particle; U2AF, U2 auxiliary factor; PTB, polypyrimidine tractbinding protein; SRproteins, serine, argininerich
proteins; CPSF, cleavage and polyadenylation specificity factor; CStF, cleavage stimulation factor; CF, cleavage
factor; PAP, poly(A) polymerase; PAB, poly(A)binding protein.
230
9 mRNA modulations in stress and aging
of a large number of transcription factors, also shown to be modified in aged tissues. As transcription
and splicing are tightly coupled processes, such changes are likely to affect the splicing machinery as well.
The transcription factor SOX17, for example, with a direct role in premRNA splicing (Ohe et al., 2002),
demonstrated 2.4fold overexpression in the aged gastrocnemius muscle. Altered expression of tran-
scriptionassociated genes during aging thus adds further support to the concept of premRNA splicing
as an agingaffected process. Indeed, agingrelated diseases often reflect abnormal upstream factors,
impaired regulation, or aberrant finetuning of mRNA processing in general and premRNA splicing in
particular.
Transcriptional and posttranscriptional poststress responses induce a cascade of changes in the levels and
properties of certain brain proteins that may play major roles in the subsequent events. This particularly
relates to proteins involved in relevant neurotransmission pathways. For example, acetylcholine (ACh)
levels were shown to be transiently elevated in the mammalian brain during stress responses (Imperato
et al., 1991). Similar increases in ACh were also reported following exposure to cholinesterase inhibitors
mRNA modulations in stress and aging
9 231
(Dazzi et al., 1995). Indeed, psychological stress elicits strikingly similar neuropsychological effects to those
observed after acute or chronic exposure to anticholinesterases (Rosenstock et al., 1991). In addition,
acute stress and pharmacological inhibition of acetylcholinesterase (AChE), the ACh hydrolyzing enzyme
(> Figure 9-6), suppress the production of the AChsynthesizing enzyme choline acetyltransferase (ChAT)
. Figure 9-6
Acetylcholine hydrolysis by AChE. AChE cleaves acetylcholine to acetate and choline at a rate of 300,000
molecules per minute, making it one of the most efficient enzymes in nature. Its active gorge includes a serine
residue responsible for the enzymes catalytic activity (Soreq and Seidman, 2001)
(Kaufer et al., 1998; Anguelova et al., 2000), while inducing rapid increases in AChE mRNA and protein
levels in response to forced swim stress in the mouse brain (Kaufer et al., 1998). This suggests a bimodal
mechanism for suppressing the cholinergic excitation, which follows stress insults.
. Figure 9-7
30 alternative splicing of AChE. The distal enhancer glucocorticoid response element (GRE), the human AChE
gene, and the three 30 splicing products (S, E, R) are shown. Linkage of exons 2, 3, and 4 is common to all
variants (a). The R transcript includes at its 30 terminus pseudointron 4 (bottom) and exon 5; The E transcript is
generated by splicing out intron 4, and option b yields the S transcript by connecting exon 4 to 6 (top).
232
9 mRNA modulations in stress and aging
isoforms, which differ at their Ctermini, dictating different cell adherence and noncatalytic properties for
each of the three. The Cterminus of human AChES consists of a 40 residue peptide, which includes a
cysteine 4 residues upstream from the Cterminus. This cysteine enables the formation of dimers and
tetramers, which are able to further bind a cholinesterasespecific collagen tail (ColQ) unique to AChE in
neuromuscular junctions (Massoulie, 2002) or a prolinerich membrane anchor (PRiMA) in the central
nervous system (Perrier et al., 2002). The former generates asymmetric AChE forms through interactions
between the ColQ tails of two or three different tetrameric units, and the two structural subunits anchor
AChES to synapses and neuromuscular junctions (Massoulie, 2002). AChEE forms dimers exclusively,
through a cysteine residue at its 43 amino acid Cterminus (position 8 of 43). This peptide is subsequently
cleaved after amino acid 14 (557 from the Nterminus) to enable linkage of the remaining Cterminus
through a glycosylphosphatidyl inositol (GPI) to the erythrocytic membrane. AChER has a shorter,
26 amino acids long, Cterminal peptide, which lacks a cysteine residue and hence remains monomeric
and soluble (Soreq and Seidman, 2001).
The core sequence of human and rodent AChE, common to all variants, consists of 543 amino acids
encoded by exons E2, E3, and E4. Exon 2 codes for the three amino acids responsible for the catalytic
activity of AChE inside the active gorge. This catalytic triad consists (in humans) of Ser203, Glu334, and
His447. Thus, AChES and AChER share the same capacities to hydrolyze acetylcholine (Soreq and
Seidman, 2001); however, the catalytic function of AChER in vivo should be extrasynaptic because it
does not attach to the synaptic membrane. This variant is hence more likely to tune the modulatory roles
of acetylcholine, presumed to reside outside synapses (Legay et al., 1995).
Several lines of evidence suggest that AChES and AChER have distinct roles in the normal brain,
as well as in the poststress processes (Seidman et al., 1995; Karpel et al., 1996; Shohami et al., 2000;
Sternfeld et al., 2000; Birikh et al., 2003; Brenner et al., 2003). Transgenic mice overexpressing one of the
corresponding human AChE isoforms (Beeri et al., 1995; Sternfeld et al., 1998) display distinct character-
istics. Mice overexpressing AChES show accelerated stressrelated pathology (Sternfeld et al., 2000)
including loss of dendritic arborizations and spines, leading to progressive deterioration (Beeri et al.,
1997), neuromotor malfunctions (Andres et al., 1997), hypersensitivity to anticholinesterases (Shapira
et al., 2000), and to closedhead injury (Shohami et al., 2000), and vulnerability to a switched dayandnight
cycle (Cohen et al., 2002). Intriguingly, transgenic mice expressing human AChES, which display changes
in the balance of their cholinergic neurotransmission (Erb et al., 2001) possess high levels of HSP70
(Sternfeld et al., 2000) associated with neuronal splicing abnormalities, reflected in high levels of their
murine AChER mRNA variant (Cohen et al., 2002). Further research will be required to explore the
possibility of an involvement of cholinergic signaling pathways in the maintenance of neuronal mRNA
splicing. AChERexpressing mice display normal neuromuscular function and their brains are relatively
protected from the stressassociated hallmarks of pathology, which predict agedependent neurodeteriora-
tion (Sternfeld et al., 2000). They further display intensified LTP, parallel to the stressinduced LTP
enhancement that associates with AChER mRNA accumulation in the hippocampus of nontransgenic
FVB/N mice (Nijholt et al., 2003). Although it is not yet clear whether this stressinduced alternative
splicing of AChE occurs as a cause or an effect, this phenotype may predict physiological relevance for the
exclusive overexpression of AChER following stress in the consolidation of fearful memories. Apart from
this apparent protective function, AChER may mediate at least some of the adverse cellular changes
associated with delayed stress responses. Intriguingly, AChER mRNA was found in apical dendrites of
neurons from all cortical layers following exposure to cholinesterase inhibitors (Kaufer et al., 1998),
suggesting possible local regulation and translation of AChE in the synapse (Steward and Schuman,
2001) also in noncholinergic neurons.
et al., 2000). In addition, repeated forced swim stress induced AChE mRNA overproduction in spermatozoa
during early spermatogenesis and elevated AChE protein levels at later stages of sperm formation (Mor
et al., 2001). In mice subjected to forced swim stress, serum AChE was subject to Cterminal proteolysis
whereas human CD34 hematopoietic progenitor cells presented cortisolinduced AChE mRNA synthesis
(Grisaru et al., 2001). Also in humans, increased levels of AChE were reported in the cerebrospinal fluid
(CSF) of hospitalized patients, but not in agematched controls. These increases were associated with blood
brain barrier (BBB) disruption, increased CSF albumin levels, and elevated serum cortisol values (Tomkins
et al., 2001). Interestingly, all the abovementioned instances of AChE overexpression involve elevations of a
single isoform of this enzyme, AChER, produced by 30 alternative splicing of the AChE premRNA.
The stressinduced changes in the alternative splicing of AChE premRNA and the overproduction of
AChER were shown to be beneficial under acute stress responses, when AChER tunes down the stress
induced hyperarousal (Kaufer and Soreq, 1999). However, prolonged stress responses likely present a
different case. Numerous physiological functions are adversely suppressed under prolonged stress
responses, with critical consequences for the affected organism sometimes. Likewise, transgenic animal
studies (Cohen et al., 2002; Birikh et al., 2003), animal disease models (Shohami et al., 2000; Brenner et al.,
2003), and human clinical trials with an antisense agent destroying AChER mRNA (Argov et al., 2005)
suggest that prolonged overproduction of AChER may take its toll.
The major product of AChEs 30 alternative splicing, AChES, was found to be abundantly expressed in
numerous tissues and developmental states. As the main variant responsible for terminating cholinergic
transmission, AChES forms tetramers through its unique Cterminus, allowing increased density of the
anchored enzyme in the synaptic cleft (25003000 moleculesmm2, (Anglister et al., 1994). This involves
interaction with structural proteins, such as ColQ (Krejci et al., 1997) or PRiMA (Perrier et al., 2002).
AChER expression, on the other hand, was found to increase during brain development, under poststress
situations (Soreq and Seidman, 2001; Massoulie, 2002), and in glial tumors (Perry et al., 2002). The
overexpression of AChER following stress could possibly be beneficial for shortterm responses, ridding
of excess ACh secreted now at higher rates. However, the longterm excess of this protein is likely to be
causally involved in the longterm hypersensitization associated with prolonged stress responses (Meshorer
et al., 2002). AChE is also modified in AD, shifting its protein product from tetramers to monomers (Talesa,
2001), and AD patients treated with anticholinesterases display increased levels of AChER in their CSF
(DarrehShori et al., 2004).
Studies in our laboratory demonstrated rapid, yet longlasting stressinduced shift from neuronal
AChES to AChER mRNA transcripts, which was associated with the accumulation of their protein
products in neurites, diverting the composition of neuritic AChE from the Cterminally adhered synaptic
membrane isoform to AChER, the variant equipped with a hydrophilic nonadhered Cterminus (Meshorer
et al., 2002). This shift was common to different neurons and stressors, suggesting that stimulusinduced
changes in alternative splicing and dendritic mRNA translocationboth considered major contributors to
neuronal plasticityare intimately interrelated for AChE. In vivo, such changes were associated with long
term cholinergic hypersensitivity in both nonstressed controls and transgenics overproducing human
AChES and murine AChER (Meshorer et al., 2002), indicating that they may contribute toward the
longknown neuronal hypersensitivity that follows stressful experiences (Antelman et al., 1980) for which
there was no mechanistic explanation.
Several steps along the pathway for gene expression may contribute toward the neuritic AChER
overexpression under stressful insults. First, elevated glucocorticoid levels activate transcription through
GRE interactions (Sapolsky et al., 2000). That corticosterone induces AChE overproduction in cultured
PC12 cells and primary cerebellar neurons suggests their participation in the initiation of neuronal AChER
mRNA overexpression.
Second, stressinduced alternative splicing [e.g., through Nova, Buckanovich et al. (1993)] might
regulate the production of AChER by binding to the Nova short consensus intronic motif, UCAUY (Jensen
234
9 mRNA modulations in stress and aging
et al., 2000), whereas the brainenriched polypyrimidine tractbinding protein (brPTB) can antagonize this
Novagenerated alternative splicing (Polydorides et al., 2000). AChE premRNA possesses both the UCAUY
motif and a possible polypyrimidine tract in conserved 30 intronic locations of both the human (positions
7,2427,246 and 7,2207,245 respectively in GeneBank accession number AF002993) (> Figure 9-8) and
mouse ACHE genes (positions 13,24213,246 and 13,21013,244, AF312033), suggesting their possible
relevance to this process.
. Figure 9-8
Novarelated sequences on AChE premRNA. Two Nova recognition sites (UCAUY) and a possible polypyrimi-
dine (Py) tract, which can potentially be recognized by the brainenriched polypyrimidine tractbinding protein
(brPTB), which antagonizes Novainduced alternative splicing, are shown. Note that among the variants of
AChE, these elements are absent in AChES mRNA
Third, stressinduced AChER mRNA accumulation is accompanied by neuritic translocation and most
likely, local AChER synthesis. Translocation into neuronal processes presumably depends on cisacting
elements, primarily within the 30 untranslated region (Wallace et al., 1998). Dendritic targeting sequences,
however, do not share common consensus motifs (Tiedge et al., 1999) and the 30 UTR of AChER mRNA
includes none of the known dendritic targeting sequences. Rather, RNA secondary structures were
proposed to direct targeting (Ainger et al., 1997; Blichenberg et al., 1999). The differential regulation of
AChER mRNA targeting into dendrites may thus be associated with its unique 30 sequence and secondary
structure. Alternatively, or additionally, AChER mRNA translocation may reflect the stressinduced
accumulation of many more nascent AChER mRNA transcripts. This is compatible with the intense
nuclear labeling of AChER mRNA, reflecting a higher premRNA:mRNA ratio for this transcript.
Fourth, neuritic stabilization of AChER mRNA may be involved. The normally short halflife charac-
teristic of AChER mRNA (Chan et al., 1998) is reminiscent of that of cytokines, where an AUUUA motif
or a longer Urich element in the 30 UTR is responsible for mRNA destabilization through the binding
of transacting proteins [reviewed by (Guhaniyogi and Brewer, 2001)]. A candidate Urich element
(UUUAUUUUUUUUUU) is present in the 30 UTR of mouse AChER mRNA (positions 13,02013,007
in GeneBank accession number AF312033) but not in AChES mRNA; however, its importance to the
stability of these transcripts remains to be tested.
Neuritic AChER mRNA is further resistant to antisense oligonucleotidecatalyzed degradation
(Meshorer et al., 2002). This further supports the notion of its sitedependent stabilization, and is probably
due to the low nuclease levels (especially RNase H, which targets DNAmRNA hybrids) in neurites. The
subcellular distribution of RNase H in neurons is not yet known, but tagged overexpressed chimeric RNase
H displayed in the protozoa Crithidia fasciculata a gradient pattern, from high levels in the nucleus to low
levels in the cytoplasm (Engel et al., 2001), which may extrapolate to low RNase H levels in neurites.
Cholinergicglutamatergic interactions were demonstrated in several brain regions (Giovannini et al.,
1994) to affect CNS functions such as LTP, memory, and behavior (Aigner, 1995), functions that show
hyperexcitation following stress. Electrophysiological recordings in the hippocampal CA1 region (Meshorer
et al., 2002) demonstrated weekslong hypersensitivity of prestressed mice to anticholinesterases, attribut-
ing a role for these changes in the cholinergic impact over poststress sensitization (Bell et al., 1994).
mRNA modulations in stress and aging
9 235
9 Conclusions
The realm of premRNA splicing modulations in aging, disease, and stress is now beginning to emerge.
Although accumulating evidence creates a clear association between splicing aberrations and disease states,
the biochemical, cellular, and molecular pathways are far from being understood. Harnessing the use of
highthroughput techniques, such as splicingspecific microarrays (Clark et al., 2002) and fiberoptic arrays
(Yeakley et al., 2002) over the coming years promises to bring about novel and exciting examples to add to
this growing list, and to shed new light on the intracellular pathways that generate alternatively spliced
products during cellular crisis. Studies aimed at the production of abnormal transcripts will pave the way
for novel RNAtargeted therapies that are already being tested in various disease models. These include
antisense oligonucleotides, ribozymes, RNAi, or spliceosomemediated RNA transsplicing (SMaRT)
(Stoilov et al., 2002). The use of such technologies will hopefully enable delicate manipulations of gene
expression in disease and aging, suppressing the alternatively spliced culprits, while maintaining accurate
levels of the normal proteins.
236
9 mRNA modulations in stress and aging
References
Aigner TG. 1995. Pharmacology of memory: Cholinergic Beato M, Truss M, Chavez S. 1996. Control of transcription by
glutamatergic interactions. Curr Opin Neurobiol 5: 155- steroid hormones. Ann NY Acad Sci 784: 93-123.
160. Beeri R, Andres C, Lev Lehman E, Timberg R, Huberman T,
Ainger K, Avossa D, Diana AS, Barry C, Barbarese E, et al. et al. 1995. Transgenic expression of human acetylcholines-
1997. Transport and localization elements in myelin basic terase induces progressive cognitive deterioration in mice.
protein mRNA. J Cell Biol 138: 1077-1087. Curr Biol 5: 1063-1071.
Allikmets R, Shroyer NF, Singh N, Seddon JM, Lewis RA, et al. Beeri R, Le Novere N, Mervis R, Huberman T, Grauer E, et al.
1997. Mutation of the Stargardt disease gene (ABCR) 1997. Enhanced hemicholinium binding and attenuated
in agerelated macular degeneration. Science 277: 1805- dendrite branching in cognitively impaired acetylcholines-
1807. terasetransgenic mice. J Neurochem 69: 2441-2451.
Altar CA. 1999. Neurotrophins and depression. Trends Phar- Bell IR, Schwartz GE, Amend D, Peterson JM, Stini WA. 1994.
macol Sci 20: 59-61. Sensitization to early life stress and response to chemical
Amir S, Brown ZW, Amit Z. 1980. The role of endorphins in odors in older adults. Biol Psychiatry 35: 857-863.
stress: Evidence and speculations. Neurosci Biobehav Rev 4: Birikh KR, Sklan EH, Shoham S, Soreq H. 2003. Interaction
77-86. of readthrough acetylcholinesterase with RACK1 and
Andersson AM, Olsen M, Zhernosekov D, Gaardsvoll H, PKCbeta II correlates with intensified fearinduced conflict
Krog L, et al. 1993. Agerelated changes in expression of behavior. Proc Natl Acad Sci USA 100: 283-288.
the neural cell adhesion molecule in skeletal muscle: Blichenberg A, Rehbein M, Muller R, Garner CC, Richter D,
A comparative study of newborn, adult and aged rats. et al. 2001. Identification of a cisacting dendritic targeting
Biochem J 290: 641-648. element in the mRNA encoding the alpha subunit of Ca2/
Andres C, Beeri R, Friedman A, LevLehman E, Henis S, calmodulindependent protein kinase II. Eur J Neurosci 13:
et al. 1997. Acetylcholinesterasetransgenic mice display 1881-8.
embryonic modulations in spinal cord choline acetyltrans- Blichenberg A, Schwanke B, Rehbein M, Garner CC,
ferase and neurexin Ibeta gene expression followed by late Richter D, et al. 1999. Identification of a cisacting
onset neuromotor deterioration. Proc Natl Acad Sci USA dendritic targeting element in MAP2 mRNAs. J Neurosci
94: 8173-8178. 19: 8818-8829.
Anglister L, Stiles JR, Salpeter MM. 1994. Acetylcholinester- Bogerts B, Lieberman JA, Ashtari M, Bilder RM, Degreef G,
ase density and turnover number at frog neuromuscular et al. 1993. Hippocampusamygdala volumes and psycho-
junctions, with modeling of their role in synaptic function. pathology in chronic schizophrenia. Biol Psychiatry 33:
Neuron 12: 783-794. 236-246.
Anguelova E, Boularand S, Nowicki JP, Benavides J, Smirnova Boise LH, GonzalezGarcia M, Postema CE, Ding L,
T. 2000. Upregulation of genes involved in cellular stress Lindsten T, et al. 1993. bclx, a bcl2related gene that
and apoptosis in a rat model of hippocampal degeneration. functions as a dominant regulator of apoptotic cell death.
J Neurosci Res 59: 209-217. Cell 74: 597-608.
Antelman SM, Eichler AJ, Black CA, Kocan D. 1980. Inter- Bond U. 1988. Heat shock but not other stress inducers leads
changeability of stress and amphetamine in sensitization. to the disruption of a subset of snRNPs and inhibition of
Science 207: 329-331. in vitro splicing in HeLa cells. EMBO J 7: 3509-3518.
Argov Z, McKee D, Agus S, Soreq H, BenYoseph O, et al. Botti SA, Felder CE, Sussman JL, Silman I. 1998. Electrotac-
2005. EN101: A novel antisense therapeutic strategy for tins: A class of adhesion proteins with conserved electro-
myasthenia gravis. Amer Assoc Neurol abstract. static and structural motifs. Protein Eng 11: 415-420.
Arimura A, Takaki A, Komaki G. 1994. Interactions between Bremner JD. 1999. Does stress damage the brain? Biol Psychi-
cytokines and the hypothalamicpituitaryadrenal axis dur- atry 45: 797-805.
ing stress. Ann NY Acad Sci 739: 270-281. Bremner JD, Randall P, Scott TM, Bronen RA, Seibyl JP, et al.
Bagley J, Moghaddam B. 1997. Temporal dynamics of gluta- 1995. MRIbased measurement of hippocampal volume in
mate efflux in the prefrontal cortex and in the hippocam- patients with combatrelated posttraumatic stress disorder.
pus following repeated stress: Effects of pretreatment with Am J Psychiatry 152: 973-981.
saline or diazepam. Neuroscience 77: 65-73. Brenner T, HamraAmitay Y, Evron T, Boneva N, Seidman S,
Ballinger S. 1990. Stress as a factor in lowered estrogen et al. 2003. The role of readthrough acetylcholinesterase
levels in the early postmenopause. Ann NY Acad Sci 592: in the pathophysiology of myasthenia gravis. FASEB J 17:
95-113. 214-222.
mRNA modulations in stress and aging
9 237
Brown GM, Seggie JA, Chambers JW, Ettigi PG. 1978. Psy- Cohen S, Tyrrell DA, Smith AP. 1991. Psychological stress
choendocrinology and growth hormone: A review. Psycho- and susceptibility to the common cold. N Engl J Med
neuroendocrinology 3: 131-153. 325: 606-612.
Buchner DA, Trudeau M, Meisler MH. 2003. SCNM1, a Company M, Arenas J, Abelson J. 1991. Requirement of the
putative RNA splicing factor that modifies disease severity RNA helicaselike protein PRP22 for release of messenger
in mice. Science 301: 967-969. RNA from spliceosomes. Nature 349: 487-493.
Buckanovich RJ, Posner JB, Darnell RB. 1993. Nova, the Convit A, Leon MJ, Tarshish C, Santi S, Kluger A, et al. 1995.
paraneoplastic Ri antigen, is homologous to an RNAbind- Hippocampal volume losses in minimally impaired elderly.
ing protein and is specifically expressed in the developing Lancet 345: 266.
motor system. Neuron 11: 657-672. Daoud R, Mies G, Smialowska A, Olah L, Hossmann KA, et al.
Buckingham JC. 2000. The role of glucocorticoids in stress. 2002. Ischemia induces a translocation of the splicing factor
Encyclopedia of stress. Fink, G, editor. San Diego: Acade- tra2beta 1 and changes alternative splicing patterns in the
mic; pp. 2: 261-269. brain. J Neurosci 22: 5889-5899.
Burgess SM, Guthrie C. 1993. A mechanism to enhance DarrehShori T, HellstromLindahl E, FloresFlores C, Guan
mRNA splicing fidelity: The RNAdependent ATPase ZZ, Soreq H, et al. 2004. Longlasting acetylcholinesterase
Prp16 governs usage of a discard pathway for aberrant splice variations in anticholinesterasetreated Alzheimers
lariat intermediates. Cell 73: 1377-1391. disease patients. J Neurochem 88: 1102-1113.
Caballero OL, de Souza SJ, Brentani RR. Simpson AJ. 2001. Davis L, Banker GA, Steward O. 1987. Selective dendritic
Alternative spliced transcripts as cancer markers. Dis Mar- transport of RNA in hippocampal neurons in culture.
kers 17: 67-75. Nature 330: 477-479.
Caceres JF, Stamm S, Helfman DM, Krainer AR. 1994. Dazzi L, Motzo C, Maira G, Sanna A, Serra M, et al. 1995.
Regulation of alternative splicing in vivo by overexpression Enhancement of acetylcholine release by flumazenil in the
of antagonistic splicing factors. Science 265: 1706-1709. hippocampus of rats chronically treated with diazepam but
Calvo N, Volosin M. 2001. Glucocorticoid and mineralocorti- not with imidazenil or abecarnil. Psychopharmacology
coid receptors are involved in the facilitation of anxietylike (Berl) 121: 180-185.
response induced by restraint. Neuroendocrinology 73: de Leon MJ, Golomb J, George AE, Convit A, Tarshish CY,
261-271. et al. 1993. The radiologic prediction of Alzheimer disease:
Cannon WB. 1939. The Wisdom of the body. New York: The atrophic hippocampal formation. AJNR Am J Neuror-
W.W. Norton. adiol 14: 897-906.
Cartegni L, Chew SL, Krainer AR. 2002. Listening to silence De Meirleir L, Lissens W, Benelli C, Ponsot G, Desguerre I,
and understanding nonsense: Exonic mutations that affect et al. 1994. Aberrant splicing of exon 6 in the pyruvate
splicing. Nat Rev Genet 3: 285-298. dehydrogenaseE1 alpha mRNA linked to a silent mutation
Catania MV, Aronica E, Yankaya B, Troost D. 2001. Increased in a large family with Leighs encephalomyelopathy. Pediatr
expression of neuronal nitric oxide synthase spliced var- Res 36: 707-712.
iants in reactive astrocytes of amyotrophic lateral sclerosis Dhabhar FS, McEwen BS. 1999. Enhancing versus suppressive
human spinal cord. J Neurosci 21: RCI48 effects of stress hormones on skin immune function. Proc
Chan RY, Adatia FA, Krupa AM, Jasmin BJ. 1998. Increased Natl Acad Sci USA 96: 1059-1064.
expression of acetylcholinesterase T and R transcripts dur- Diamond DM, Rose GM. 1994. Stress impairs LTP and
ing hematopoietic differentiation is accompanied by paral- hippocampaldependent memory. Ann NY Acad Sci 746:
lel elevations in the levels of their respective molecular 411-414.
forms. J Biol Chem 273: 9727-9733. Dredge BK, Polydorides AD, Darnell RB. 2001. The splice of
Chandler VL, Maler BA, Yamamoto KR. 1983. DNA sequen- life: Alternative splicing and neurological disease. Nat Rev
ces bound specifically by glucocorticoid receptor in vitro Neurosci 2: 43-50.
render a heterologous promoter hormone responsive DSouza I, Poorkaj P, Hong M, Nochlin D, Lee VM, et al.
in vivo. Cell 33: 489-499. 1999. Missense and silent tau gene mutations cause fron-
Clark TA, Sugnet CW, Ares M Jr. 2002. Genomewide analysis totemporal dementia with parkinsonismchromosome
of mRNA processing in yeast using splicingspecific micro- 17 type, by affecting multiple alternative RNA splicing
arrays. Science 296: 907-910. regulatory elements. Proc Natl Acad Sci USA 96:
Cohen O, Erb C, Ginzberg D, Pollak Y, Seidman S, et al. 2002. 5598-5603.
Neuronal overexpression of readthrough acetylcholines- Duman RS, Malberg J, Thome J. 1999. Neural plasticity to
terase is associated with antisensesuppressible behavioral stress and antidepressant treatment. Biol Psychiatry 46:
impairments. Mol Psychiatry 7: 874-885. 1181-1191.
238
9 mRNA modulations in stress and aging
Ebert D, Ebmeier KP. 1996. The role of the cingulate gyrus in Grisaru D, Deutsch V, Shapira M, Pick M, Sternfeld M, et al.
depression: From functional anatomy to neurochemistry. 2001. ARP, a peptide derived from the stressassociated
Biol Psychiatry 39: 1044-1050. acetylcholinesterase variant has hematopoietic growth pro-
Eberwine J, Miyashiro K, Kacharmina JE, Job C. 2001. Local moting activities. Mol Med 7: 93-105.
translation of classes of mRNAs that are targeted to neuro- Gruen RJ, Wenberg K, Elahi R, Friedhoff AJ. 1995. Altera-
nal dendrites. Proc Natl Acad Sci USA 98: 7080-7085. tions in GABAA receptor binding in the prefrontal cortex
Engel ML, Hines JC, Ray DS. 2001. The Crithidia fasciculata following exposure to chronic stress. Brain Res 684:
RNH1 gene encodes both nuclear and mitochondrial iso- 112-114.
forms of RNase H. Nucleic Acids Res 29: 725-731. Guhaniyogi J, Brewer G. 2001. Regulation of mRNA stability
Erb C, Troost J, Kopf S, Schmitt U, Loffelholz K, et al. 2001. in mammalian cells. Gene 265: 11-23.
Compensatory mechanisms enhance hippocampal acetyl- Habelhah H, Shah K, Huang L, OstareckLederer A,
choline release in transgenic mice expressing human ace- Burlingame AL, et al. 2001. ERK phosphorylation drives
tylcholinesterase. J Neurochem 77: 638-646. cytoplasmic accumulation of hnRNPK and inhibition of
Feng Z, Li L, Ng PY, Porter AG. 2002. Neuronal differentiation mRNA translation. Nat Cell Biol 3: 325-330.
and protection from nitric oxideinduced apoptosis require Hamamura T, Fibiger HC. 1993. Enhanced stressinduced
cJundependent expression of NCAM140. Mol Cell Biol dopamine release in the prefrontal cortex of amphet-
22: 5357-5366. aminesensitized rats. Eur J Pharmacol 237: 65-71.
Flugge G, Ahrens O, Fuchs E. 1997. Monoamine receptors in Hanamura A, Caceres JF, Mayeda A, Franza BR Jr, Krainer AR.
the prefrontal cortex of Tupaia belangeri during chronic 1998. Regulated tissuespecific expression of antagonistic
psychosocial stress. Cell Tissue Res 288: 1-10. premRNA splicing factors. RNA 4: 430-444.
Furey ML, Pietrini P, Haxby JV. 2000. Cholinergic enhance- Hashimoto Y, Zhang C, Kawauchi J, Imoto I, Adachi MT, et al.
ment and increased selectivity of perceptual processing 2002. An alternatively spliced isoform of transcriptional
during working memory. Science 290: 2315-2319. repressor ATF3 and its induction by stress stimuli. Nucleic
Gallego ME, Gattoni R, Stevenin J, Marie J, ExpertBezancon Acids Res 30: 2398-2406.
A. 1997. The SR splicing factors ASF/SF2 and SC35 Hastings ML, Krainer AR. 2001. PremRNA splicing in the
have antagonistic effects on intronic enhancerdependent new millennium. Curr Opin Cell Biol 13: 302-309.
splicing of the betatropomyosin alternative exon 6A. Heck S, Kullmann M, Gast A, Ponta H, Rahmsdorf HJ, et al.
EMBO J 16: 1772-1784. 1994. A distinct modulating domain in glucocorticoid
Gattoni R, Mahe D, Mahl P, Fischer N, Mattei MG, et al. 1996. receptor monomers in the repression of activity of the
The human hnRNPM proteins: Structure and relation transcription factor AP1. EMBO J 13: 4087-4095.
with early heat shockinduced splicing arrest and chromo- HerreraGayol A, Jothy S. 1999. Adhesion proteins in the
some mapping. Nucleic Acids Res 24: 2535-2542. biology of breast cancer: Contribution of CD44. Exp Mol
Ge H, Manley JL. 1990. A protein factor, ASF, controls cell Pathol 66: 149-156.
specific alternative splicing of SV40 early premRNA in Hodges PE, Carrico PM, Hogan JD, ONeill KE, Owen JJ, et al.
vitro. Cell 62: 25-34. 2002. Annotating the human proteome: The Human Pro-
Gershuny BS, Thayer JF. 1999. Relations among psychologi- teome Survey Database (HumanPSD) and an indepth
cal trauma, dissociative phenomena, and traumarelated target database for G proteincoupled receptors (GPCR
distress: A review and integration. Clin Psychol Rev 19: PD) from Incyte Genomics. Nucleic Acids Res 30: 137-141.
631-657. HoehnSaric R, McLeod DR, Lee YB, Zimmerli WD. 1991.
Giovannini MG, Camilli F, Mundula A, Pepeu G. 1994. Glu- Cortisol levels in generalized anxiety disorder. Psychiatry
tamatergic regulation of acetylcholine output in different Res 38: 313-315.
brain regions: A microdialysis study in the rat. Neurochem Hofmann TG, Schmitz LM. 2002. The promoter context
Int 25: 23-26. determines mutual repression or synergism between
Gould E, Tanapat P, McEwen BS, Flugge G, Fuchs E. 1998. NFkappaB and the glucocorticoid receptor. Biol Chem
Proliferation of granule cell precursors in the dentate gyrus 383: 1947-1951.
of adult monkeys is diminished by stress. Proc Natl Acad Imperato A, PuglisiAllegra S, Casolini P, Angelucci L. 1991.
Sci USA 95: 3168-3171. Changes in brain dopamine and acetylcholine release dur-
Grifman M, Galyam N, Seidman S, Soreq H. 1998. Functional ing and following stress are independent of the pituitary
redundancy of acetylcholinesterase and neuroligin in adrenocortical axis. Brain Res 538: 111-117.
mammalian neuritogenesis. Proc Natl Acad Sci USA 95: Ishihara T, Hong M, Zhang B, Nakagawa Y, Lee MK, et al.
13935-13940. 1999. Agedependent emergence and progression of a
mRNA modulations in stress and aging
9 239
tauopathy in transgenic mice overexpressing the shortest Krainer AR, Conway GC, Kozak D. 1990. The essential pre
human tau isoform. Neuron 24: 751-762. mRNA splicing factor SF2 influences 50 splice site selection
Jacobson A, Peltz SW. 1996. Interrelationships of the pathways by activating proximal sites. Cell 62: 35-42.
of mRNA decay and translation in eukaryotic cells. Annu Krejci E, Thomine S, Boschetti N, Legay C, Sketelj J, et al.
Rev Biochem 65: 693-739. 1997. The mammalian gene of acetylcholinesterase
Jensen KB, Dredge BK, Stefani G, Zhong R, Buckanovich RJ, associated collagen. J Biol Chem 272: 22840-22847.
et al. 2000. Nova1 regulates neuronspecific alternative Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, et al.
splicing and is essential for neuronal viability. Neuron 25: 2001. Initial sequencing and analysis of the human genome.
359-371. Nature 409: 860-921.
Jeong KH, Jacobson L, Widmaier EP, Majzoub JA. 1999. Lee CK, Klopp RG, Weindruch R, Prolla TA. 1999. Gene
Normal suppression of the reproductive axis following expression profile of aging and its retardation by caloric
stress in corticotropinreleasing hormonedeficient mice. restriction. Science 285: 1390-1393.
Endocrinology 140: 1702-1708. Lee VM, Goedert M, Trojanowski JQ. 2001. Neurodegenera-
Jolly C, Morimoto RI. 1999. Stress and the cell nucleus: tive tauopathies. Annu Rev Neurosci 24: 1121-1159.
Dynamics of gene expression and structural reorganization. Legay C, Huchet M, Massoulie J, Changeux JP. 1995. Devel-
Gene Expr 7: 261-270. opmental regulation of acetylcholinesterase transcripts in
Jolly C, Vourch C, RobertNicoud M, Morimoto RI. 1999. the mouse diaphragm: Alternative splicing and focaliza-
Intronindependent association of splicing factors with tion. Eur J Neurosci 7: 1803-1809.
active genes. J Cell Biol 145: 1133-1143. Licinio J, Wong ML. 1999. The role of inflammatory media-
Kalinichev M, Easterling KW, Plotsky PM, Holtzman SG. tors in the biology of major depression: Central nervous
2002. Longlasting changes in stressinduced corticosterone system cytokines modulate the biological substrate of de-
response and anxietylike behaviors as a consequence of pressive symptoms, regulate stressresponsive systems, and
neonatal maternal separation in LongEvans rats. Pharma- contribute to neurotoxicity and neuroprotection. Mol
col Biochem Behav 73: 131-140. Psychiatry 4: 317-327.
Kang H, Schuman EM. 1996. A requirement for local protein Lin CL, Bristol LA, Jin L, DykesHoberg M, Crawford T, et al.
synthesis in neurotrophininduced hippocampal synaptic 1998. Aberrant RNA processing in a neurodegenerative dis-
plasticity. Science 273: 1402-1406. ease: The cause for absent EAAT2, a glutamate transporter,
Karpel R, Sternfeld M, Ginzberg D, Guhl E, Graessmann A, in amyotrophic lateral sclerosis. Neuron 20: 589-602.
et al. 1996. Overexpression of alternative human acetylcho- Liu HX, Chew SL, Cartegni L, Zhang MQ, Krainer AR. 2000.
linesterase forms modulates process extensions in cultured Exonic splicing enhancer motif recognized by human SC35
glioma cells. J Neurochem 66: 114-123. under splicing conditions. Mol Cell Biol 20: 1063-1071.
Kaufer D, Friedman A, Seidman S, Soreq H. 1998. Acute stress Llewellyn DH, Scobie GA, Urquhart AJ, Whatley SD, Roberts
facilitates longlasting changes in cholinergic gene expres- AG, et al. 1996. Acute intermittent porphyria caused by
sion. Nature 393: 373-377. defective splicing of porphobilinogen deaminase RNA: A
Kaufer D, Soreq H. 1999. Tracking cholinergic pathways from synonymous codon mutation at 22 bp from the 50 splice
psychological and chemical stressors to variable neurode- site causes skipping of exon 3. J Med Genet 33: 437-438.
terioration paradigms. Curr Opin Neurol 12: 739-743. Lombardi P, Hoffer MJ, Top B, de Wit E, Gevers Leuven JA,
Kawahara H, Yoshida M, Yokoo H, Nishi M, Tanaka M. 1993. et al. 1993. An acceptor splice site mutation in intron 16 of
Psychological stress increases serotonin release in the rat the low density lipoprotein receptor gene leads to an elon-
amygdala and prefrontal cortex assessed by in vivo micro- gated, internalization defective receptor. Atherosclerosis
dialysis. Neurosci Lett 162: 81-84. 104: 117-128.
KeltikangasJarvinen L, Ravaja N, Raikkonen K, Hautanen A, Lupien SJ, Lepage M. 2001. Stress, memory, and the hippo-
Adlercreutz H. 1998. Relationships between the pituitary campus: Cant live with it, cant live without it. Behav Brain
adrenal hormones, insulin, and glucose in middleaged Res 127: 137-158.
men: Moderating influence of psychosocial stress. Metabo- Lupien SJ, de Leon M, de Santi S, Convit A, Tarshish C, et al.
lism 47: 1440-1449. 1998. Cortisol levels during human aging predict hip-
Kjaer A, Knigge U, Olsen L, Vilhardt H, Warberg J. 1991. pocampal atrophy and memory deficits. Nat Neurosci 1:
Mediation of the stressinduced prolactin release by hypo- 69-73.
thalamic histaminergic neurons and the possible involve- Magarinos AM, Verdugo JM, McEwen BS. 1997. Chronic
ment of vasopressin in this response. Endocrinology 128: stress alters synaptic terminal structure in hippocampus.
103-110. Proc Natl Acad Sci USA 94: 14002-14008.
240
9 mRNA modulations in stress and aging
Magarinos AM, McEwen BS, Flugge G, Fuchs E. 1996. Chron- Meshorer E, Soreq H. 2002. Splicing modulations in senes-
ic psychosocial stress causes apical dendritic atrophy of cence. Aging Cell 1: 10-16.
hippocampal CA3 pyramidal neurons in subordinate tree Meshorer E, Erb C, Gazit R, Pavlovsky L, Kaufer D, et al. 2002.
shrews. J Neurosci 16: 3534-3540. Alternative splicing and neuritic mRNA translocation under
Magiakou MA, Mastorakos G, Webster E, Chrousos GP. 1997. longterm neuronal hypersensitivity. Science 295: 508-512.
The hypothalamicpituitaryadrenal axis and the female Miller S, Yasuda M, Coats JK, Jones Y, Martone ME, et al.
reproductive system. Ann NY Acad Sci 816: 42-56. 2002. Disruption of dendritic translation of CaMKIIalpha
Mahe D, Mahl P, Gattoni R, Fischer N, Mattei MG, et al. 1997. impairs stabilization of synaptic plasticity and memory
Cloning of human 2H9 heterogeneous nuclear ribonucleo- consolidation. Neuron 36: 507-519.
proteins. Relation with splicing and early heat shock Mintz PJ, Spector DL. 2000. Compartmentalization of RNA
induced splicing arrest. J Biol Chem 272: 1827-1836. processing factors within nuclear speckles. J Struct Biol 129:
Maldarelli F, Xiang C, Chamoun G, Zeichner SL. 1998. The 241-251.
expression of the essential nuclear splicing factor SC35 is Mizoguchi K, Yuzurihara M, Ishige A, Sasaki H, Chui DH,
altered by human immunodeficiency virus infection. Virus et al. 2000. Chronic stress induces impairment of spatial
Res 53: 39-51. working memory because of prefrontal dopaminergic dys-
Mark GP, Rada PV, Shors TJ. 1996. Inescapable stress function. J Neurosci 20: 1568-1574.
enhances extracellular acetylcholine in the rat hippocam- Moghaddam B. 1993. Stress preferentially increases extra-
pus and prefrontal cortex but not the nucleus accumbens neuronal levels of excitatory amino acids in the prefrontal
or amygdala. Neuroscience 74: 767-774. cortex: Comparison to hippocampus and basal ganglia.
Massoulie J. 2002. The origin of the molecular diversity and J Neurochem 60: 1650-1657.
functional anchoring of cholinesterases. Neurosignals 11: Moghaddam B. 2002. Stress activation of glutamate neuro-
130-143. transmission in the prefrontal cortex: Implications for
Mayeda A, Krainer AR. 1992. Regulation of alternative pre dopamineassociated psychiatric disorders. Biol Psychiatry
mRNA splicing by hnRNP A1 and splicing factor SF2. Cell 51: 775-787.
68: 365-375. Mor I, Grisaru D, Titelbaum L, Evron T, Richler C, et al. 2001.
Mayford M, Bach ME, Huang YY, Wang L, Hawkins RD, et al. Modified testicular expression of stressassociated read-
1996. Control of memory formation through regulated through acetylcholinesterase predicts male infertility.
expression of a CaMKII transgene. Science 274: 1678-1683. FASEB J 15: 2039-2041.
McCormick JA, Lyons V, Jacobson MD, Noble J, Diorio J, Mori Y, Imaizumi K, Katayama T, Yoneda T, Tohyama M.
et al. 2000. 50 heterogeneity of glucocorticoid receptor mes- 2000. Two cisacting elements in the 30 untranslated region
senger RNA is tissue specific: Differential regulation of of alphaCaMKII regulate its dendritic targeting. Nat
variant transcripts by earlylife events. Mol Endocrinol 14: Neurosci 3: 1079-1084.
506-517. Nakane H, Shimizu N, Hori T. 1994. Stressinduced norepi-
McEwen BS. 1999a. Stress and hippocampal plasticity. Annu nephrine release in the rat prefrontal cortex measured by
Rev Neurosci 22: 105-122. microdialysis. Am J Physiol 267: R1559-1566.
McEwen BS. 1999b. Stress and the aging hippocampus. Front Nguyen T, Sudhof TC. 1997. Binding properties of neuroligin
Neuroendocrinol 20: 49-70. 1 and neurexin 1 beta reveal function as heterophilic cell
McEwen BS. 2002. Sex, stress and the hippocampus: Allosta- adhesion molecules. J Biol Chem 272(41): 26032-26039.
sis, allostatic load and the aging process. Neurobiol Aging Nie GY, Li Y, Batten L, Griffiths B, Wang J, et al. 2000. Uterine
23: 921-939. expression of alternatively spliced mRNAs of mouse splic-
McEwen BS, Weiss JM, Schwartz LS. 1968. Selective retention ing factor SC35 during early pregnancy. Mol Hum Reprod
of corticosterone by limbic structures in rat brain. Nature 6: 1131-1139.
220: 911-912. Nijholt I, Farchi N, Kye M, Sklan EH, Shoham S, et al. 2004.
McGaugh JL. 2000. Memorya century of consolidation. Stressinduced alternative splicing of acetylcholinesterase
Science 287: 248-251. results in enhanced fear memory and longterm potentia-
McKie AB, McHale JC, Keen TJ, Tarttelin EE, Goliath R, et al. tion. Mol Psychiatry 9: 174-183.
2001. Mutations in the premRNA splicing factor gene Ohe K, Lalli E, SassoneCorsi P. 2002. A direct role of SRY and
PRPC8 in autosomal dominant retinitis pigmentosa SOX proteins in premRNA splicing. Proc Natl Acad Sci
(RP13). Hum Mol Genet 10: 1555-1562. USA 99: 1146-1151.
Merchant KM, Dobie DJ, Dorsa DM. 1993. Differential loss Orsini C, Castellano C, Cabib S. 2001. Pharmacological evi-
of dopamine D2 receptor mRNA isoforms during aging in dence of muscariniccholinergic sensitization following
Fischer344 rats. Neurosci Lett 154: 163-167. chronic stress. Psychopharmacology (Berl) 155: 144-147.
mRNA modulations in stress and aging
9 241
Osteryoung KW, Sundberg H, Vierling E. 1993. Poly(A) tail structure and mutations in molybdenum cofactor deficien-
length of a heat shock protein RNA is increased by severe cy type B. Am J Hum Genet 64: 706-711.
heat stress, but intron splicing is unaffected. Mol Gen Genet Rio DC. 1992. RNA processing. Curr Opin Cell Biol 4:
239: 323-333. 444-452.
Owino V, Yang SY, Goldspink G. 2001. Agerelated loss of Robberson BL, Cote GJ, Berget SM. 1990. Exon definition
skeletal muscle function and the inability to express may facilitate splice site selection in RNAs with multiple
the autocrine form of insulinlike growth factor1 (MGF) exons. Mol Cell Biol 10: 84-94.
in response to mechanical overload. FEBS Lett 505: Robertson GL. 1976. The regulation of vasopressin function
259-263. in health and disease. Recent Prog Horm Res 33: 333-385.
Pagani F, Buratti E, Stuani C, Bendix R, Dork T, et al. 2002. A Rosenstock L, Keifer M, Daniell W, McConnell R,
new type of mutation causes a splicing defect in ATM. Nat Claypoole K. 1991. Chronic central nervous system effects
Genet 30: 426-429. of acute organophosphate pesticide intoxication. The Pes-
Partridge L, Gems D. 2002. Mechanisms of ageing: Public or ticide Health Effects Study Group. Lancet 338: 223-227.
private? Nat Rev Genet 3: 165-175. Ruby SW, Chang TH, Abelson J. 1993. Four yeast spliceoso-
Perrier AL, Massoulie J, Krejci E. 2002. PRiMA: The mem- mal proteins (PRP5, PRP9, PRP11, and PRP21) interact to
brane anchor of acetylcholinesterase in the brain. Neuron promote U2 snRNP binding to premRNA. Genes Dev 7:
33: 275-285. 1909-1925.
Perry C, Sklan EH, Birikh K, Shapira M, Trejo L, et al. 2002. Sandbrink R, Monning U, Masters CL, Beyreuther K. 1997.
Complex regulation of acetylcholinesterase gene expression Expression of the APP gene family in brain cells, brain
in human brain tumors. Oncogene 21: 8428-8441. development and aging. Gerontology 43: 119-131.
Persico AM, Schindler CW, OHara BF, Brannock MT, Uhl Sapolsky RM, Romero LM, Munck AU. 2000. How do gluco-
GR. 1993. Brain transcription factor expression: Effects of corticoids influence stress responses? Integrating permis-
acute and chronic amphetamine and injection stress. Brain sive, suppressive, stimulatory, and preparative actions.
Res Mol Brain Res 20: 91-100. Endocr Rev 21: 55-89.
Persico AM, Schindler CW, Zaczek R, Brannock MT, Uhl GR. Sato N, Hori O, Yamaguchi A, Lambert JC, ChartierHarlin
1995. Brain transcription factor gene expression, neuro- MC, et al. 1999. A novel presenilin2 splice variant in
transmitter levels, and novelty response behaviors: Altera- human Alzheimers disease brain tissue. J Neurochem 72:
tions during rat amphetamine withdrawal and following 2498-2505.
chronic injection stress. Synapse 19: 212-227. Schaal TD, Maniatis T. 1999. Selection and characterization
PetersenMahrt SK, Estmer C, Ohrmalm C, Matthews DA, of premRNA splicing enhancers: Identification of novel
Russell WC, et al. 1999. The splicing factorassociated SR proteinspecific enhancer sequences. Mol Cell Biol 19:
protein, p32, regulates RNA splicing by inhibiting ASF/ 1705-1719.
SF2 RNA binding and phosphorylation. EMBO J 18: Schmucker D, Clemens JC, Shu H, Worby CA, Xiao J, et al.
1014-1024. 2000. Drosophila Dscam is an axon guidance receptor
Philips AV, Cooper TA. 2000. RNA processing and human exhibiting extraordinary molecular diversity. Cell 101:
disease. Cell Mol Life Sci 57: 235-249. 671-684.
Pollard AJ, Sparey C, Robson SC, Krainer AR, EuropeFinner Schwerk C, Prasad J, Degenhardt K, ErdjumentBromage H,
GN. 2000. Spatiotemporal expression of the transacting White E, et al. 2003. ASAP, a novel protein complex
splicing factors SF2/ASF and heterogeneous ribonuclear involved in RNA processing and apoptosis. Mol Cell Biol
proteins A1/A1B in the myometrium of the pregnant 23: 2981-2990.
human uterus: A molecular mechanism for regulating re- Scorilas A, Kyriakopoulou L, Katsaros D, Diamandis EP. 2001.
gional protein isoform expression in vivo. J Clin Endocrinol Cloning of a gene (SRA1), encoding for a new member
Metab 85: 1928-1936. of the human Ser/Argrich family of premRNA splicing
Polydorides AD, Okano HJ, Yang YY, Stefani G, Darnell RB. factors: Overexpression in aggressive ovarian cancer. Br J
2000. A brainenriched polypyrimidine tractbinding pro- Cancer 85: 190-198.
tein antagonizes the ability of Nova to regulate neuron Seidman S, Sternfeld M, Ben AzizAloya R, Timberg R,
specific alternative splicing. Proc Natl Acad Sci USA 97: KauferNachum D, et al. 1995. Synaptic and epidermal
6350-6355. accumulations of human acetylcholinesterase are encoded
Pruett SB. 2003. Stress and the immune system. Pathophysi- by alternative 30 terminal exons. Mol Cell Biol 15:
ology 9: 133-153. 2993-3002.
Reiss J, Dorche C, Stallmeyer B, Mendel RR, Cohen N, et al. Selye H. 1936. A syndrome produced by diverse nocuous
1999. Human molybdopterin synthase gene: Genomic agent. Nature 138: 32
242
9 mRNA modulations in stress and aging
Selye H. 1978. The stress of life, McGrawHill Trade. New Stoilov P, Meshorer E, Gencheva M, Glick D, Soreq H, et al.
York, USA. 2002. Defects in premRNA processing as causes of and
Shapira M, TurKaspa I, Bosgraaf L, Livni N, Grant AD, et al. predisposition to diseases. DNA Cell Biol 21: 803-818.
2000. A transcriptionactivating polymorphism in the Sun Q, Mayeda A, Hampson RK, Krainer AR, Rottman FM.
ACHE promoter associated with acute sensitivity to anti 1993. General splicing factor SF2/ASF promotes alternative
acetylcholinesterases. Hum Mol Genet 9: 1273-1281. splicing by binding to an exonic splicing enhancer. Genes
Sheline YI, Wang PW, Gado MH, Csernansky JG, Vannier Dev 7: 2598-2608.
MW. 1996. Hippocampal atrophy in recurrent major de- Sureau A, Gattoni R, Dooghe Y, Stevenin J, Soret J. 2001. SC35
pression. Proc Natl Acad Sci USA 93: 3908-3913. autoregulates its expression by promoting splicing events
Shin LM, Whalen PJ, Pitman RK, Bush G, Macklin ML, et al. that destabilize its mRNAs. EMBO J 20: 1785-1796.
2001. An fMRI study of anterior cingulate function Sureau A, Soret J, Guyon C, Gaillard C, Dumon S, et al. 1997.
in posttraumatic stress disorder. Biol Psychiatry 50: Characterization of multiple alternative RNAs resulting
932-942. from antisense transcription of the PR264/SC35 splicing
Shohami E, Kaufer D, Chen Y, Seidman S, Cohen O, et al. factor gene. Nucleic Acids Res 25: 4513-4522.
2000. Antisense prevention of neuronal damages following Tacke R, Manley JL. 1999. Determinants of SR protein speci-
head injury in mice. J Mol Med 78: 228-236. ficity. Curr Opin Cell Biol 11: 358-362.
Shukla RR, Dominski Z, Zwierzynski T, Kole R. 1990. Inacti- Talesa VN. 2001. Acetylcholinesterase in Alzheimers disease.
vation of splicing factors in HeLa cells subjected to heat Mech Ageing Dev 122: 1961-1969.
shock. J Biol Chem 265: 20377-20383. Tanabe M, Sasai N, Nagata K, Liu XD, Liu PC, et al. 1999. The
Song JY, Ichtchenko K, Sudhof TC, Brose N. 1999. Neuroli- mammalian HSF4 gene generates both an activator and a
gin 1 is a postsynaptic celladhesion molecule of excitatory repressor of heat shock genes by alternative splicing. J Biol
synapses. Proc Natl Acad Sci USA 96: 1100-1105. Chem 274: 27845-27856.
Soreq H, Seidman S. 2001. Acetylcholinesterase new roles Tiedge H, Bloom FE, Richter D. 1999. RNA, whither goest
for an old actor. Nat Rev Neurosci 2: 294-302. thou? Science 283: 186-187.
Soreq H, Nudel U, Salomon R, Revel M, Littauer UZ. 1974. In Tomkins O, Kaufer D, Korn A, Shelef I, Golan H, et al. 2001.
vitro translation of polyadenylic acidfree rabbit globin Frequent bloodbrain barrier disruption in the human
messenger RNA. J Mol Biol 88: 233-245. cerebral cortex. Cell Mol Neurobiol 21: 675-691.
Spector DL, Fu XD, Maniatis T. 1991. Associations between Tongiorgi E, Righi M, Cattaneo A. 1997. Activitydependent
distinct premRNA splicing components and the cell dendritic targeting of BDNF and TrkB mRNAs in hippo-
nucleus. EMBO J 10: 3467-3481. campal neurons. J Neurosci 17: 9492-9505.
Starkman MN, Gebarski SS, Berent S, Schteingart DE. 1992. Tsigelny I, Shindyalov IN, Bourne PE, Sudhof TC, Taylor P.
Hippocampal formation volume, memory dysfunction, 2000. Common EFhand motifs in cholinesterases and
and cortisol levels in patients with Cushings syndrome. neuroligins suggest a role for Ca2 binding in cell surface
Biol Psychiatry 32: 756-765. associations. Protein Sci 9: 180-185.
Stefanski V. 2000. Social stress in laboratory rats: Hormonal Uno H, Tarara R, Else JG, Suleman MA, Sapolsky RM. 1989.
responses and immune cell distribution. Psychoneuroen- Hippocampal damage associated with prolonged and fatal
docrinology 25: 389-406. stress in primates. J Neurosci 9: 1705-1711.
Sternfeld M, Patrick JD, Soreq H. 1998. Position effect var- Utans U, Behrens SE, Luhrmann R, Kole R, Kramer A. 1992.
iegations and brainspecific silencing in transgenic mice A splicing factor that is inactivated during in vivo heat
overexpressing human acetylcholinesterase variants. shock is functionally equivalent to the [U4/U6. 5] triple
J Physiol (Paris) 92: 249-255. snRNPspecific proteins. Genes Dev 6: 631-641.
Sternfeld M, Shoham S, Klein O, FloresFlores C, Evron T, van der Houven van Oordt W, DiazMeco MT, Lozano J,
et al. 2000. Excess readthrough acetylcholinesterase atte- Krainer AR, Moscat J, et al. 2000. The MKK(3/6)p38
nuates but the synaptic variant intensifies neurodeter- signaling cascade alters the subcellular distribution of
ioration correlates. Proc Natl Acad Sci USA 97: 8647-8652. hnRNP A1 and modulates alternative splicing regulation.
Steward O, Schuman EM. 2001. Protein synthesis at synaptic J Cell Biol 149: 307-316.
sites on dendrites. Annu Rev Neurosci 24: 299-325. Vidal H, Auboeuf D, Beylot M, Riou JP. 1995. Regulation
Steward O, Wallace CS, Lyford GL, Worley PF. 1998. Synaptic of insulin receptor mRNA splicing in rat tissues. Effect of
activation causes the mRNA for the IEG Arc to localize fasting, aging, and diabetes. Diabetes 44: 1196-1201.
selectively near activated postsynaptic sites on dendrites. Vithana EN, AbuSafieh L, Allen MJ, Carey A, Papaioannou
Neuron 21: 741-751. M, et al. 2001. A human homolog of yeast premRNA
mRNA modulations in stress and aging
9 243
splicing gene, PRP31, underlies autosomal dominant reti- Xie J, Black DL. 2001. A CaMK IV responsive RNA element
nitis pigmentosa on chromosome 19q13.4 (RP11). Mol mediates depolarizationinduced alternative splicing of ion
Cell 8: 375-381. channels. Nature 410: 936-939.
Vogel JL, Parsell DA, Lindquist S. 1995. Heatshock proteins Xie J, McCobb DP. 1998. Control of alternative splicing
Hsp104 and Hsp70 reactivate mRNA splicing after heat of potassium channels by stress hormones. Science 280:
inactivation. Curr Biol 5: 306-317. 443-446.
Vogiagis D, Glare EM, Misajon A, Brown W, OBrien PE. Yeakley JM, Fan JB, Doucet D, Luo L, Wickham E, et al. 2002.
2000. Cyclooxygenase1 and an alternatively spliced mRNA Profiling alternative splicing on fiberoptic arrays. Nat Bio-
in the rat stomach: Effects of aging and ulcers. Am J Physiol technol 20: 353-358.
Gastrointest Liver Physiol 278: G820-827. Yehuda R. 2002. Current status of cortisol findings in
Wagner EJ, GarciaBlanco MA. 2001. Polypyrimidine tract posttraumatic stress disorder. Psychiatr Clin North Am
binding protein antagonizes exon definition. Mol Cell 25: 341-368, vii.
Biol 21: 3281-3288. Yehuda R, Boisoneau D, Mason JW, Giller EL. 1993. Gluco-
Wallace CS, Lyford GL, Worley PF, Steward O. 1998. Differen- corticoid receptor number and cortisol excretion in mood,
tial intracellular sorting of immediate early gene mRNAs anxiety, and psychotic disorders. Biol Psychiatry 34: 18-25.
depends on signals in the mRNA sequence. J Neurosci 18: Yost HJ, Lindquist S. 1986. RNA splicing is interrupted by
26-35. heat shock and is rescued by heat shock protein synthesis.
Wang HY, Xu X, Ding JH, Bermingham JR Jr, Fu XD. 2001. Cell 45: 185-193.
SC35 plays a role in T cell development and alternative Yu L, HeereRess E, Boucher B, Defesche JC, Kastelein J, et al.
splicing of CD45. Mol Cell 7: 331-342. 1999. Familial hypercholesterolemia. Acceptor splice site
Wang J, Xiao SH, Manley JL. 1998. Genetic analysis of the SR (G!C) mutation in intron 7 of the LDLR gene: Alternate
protein ASF/SF2: Interchangeability of RS domains and RNA editing causes exon 8 skipping or a premature stop
negative control of splicing. Genes Dev 12: 2222-2233. codon in exon 8. LDLR(Honduras1) [LDLR1061(1)
Wang Z, Hoffmann HM, Grabowski PJ. 1995. Intrinsic U2AF G!C]. Atherosclerosis 146: 125-131.
binding is modulated by exon enhancer signals in parallel Zahler AM, Lane WS, Stolk JA, Roth MB. 1992. SR proteins: A
with changes in splicing activity. RNA 1: 21-35. conserved family of premRNA splicing factors. Genes Dev
WegRemers S, Ponta H, Herrlich P, Konig H. 2001. Regula- 6: 837-847.
tion of alternative premRNA splicing by the ERK MAP Zahler AM, Neugebauer KM, Lane WS, Roth MB. 1993.
kinase pathway. EMBO J 20: 4194-4203. Distinct functions of SR proteins in alternative premRNA
Wimmer K, Eckart M, Rehder H, Fonatsch C. 2000. Illegiti- splicing. Science 260: 219-222.
mate splicing of the NF1 gene in healthy individuals Zhou Z, Licklider LJ, Gygi SP, Reed R. 2002. Comprehensive
mimics mutationinduced splicing alterations in NF1 proteomic analysis of the human spliceosome. Nature 419:
patients. Hum Genet 106: 311-313. 182-185.
10 Molecular Mechanisms of
Dendritic Spine Plasticity in
Development and Aging
M. R. Kreutz . I. Konig . M. Mikhaylova . C. Spilker . W. Zuschratter
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 246
5 Molecular Dynamics of the Plastic Spine or What are Spines Good for . . . . . . . . . . . . . . . . . . . . . . . . . 251
7 Structural Alterations of Spine Number and Formation in Brain Disease States . . . . . . . . . . . . . . . 254
Abstract: The postsynaptic density (PSD) of spinous excitatory synapses is characterized by an electron
dense filamentous meshwork of cytoskeletal proteins that serve three major functions: (1) the organization
of glutamate receptors, (2) the clustering of synaptic adhesion molecules, and (3) the coupling of synaptic
membrane proteins to intracellular signaling cascades. Consequently, in recent years a plethora of protein
protein interactions of PSD scaffolding proteins has been mapped and their possible function for the
formation and integrity of the synapse delineated. It is assumed that structural alterations in the PSD
precede morphological alterations of the spine. Detailed studies of such processes are of particular
significance because they will probably lead to a new conceptual framework in our appreciation of how
changes in the protein composition and proteinprotein interactions in the PSD can regulate spine
morphogenesis. This should provide novel answers to the longstanding questionwhat molecular
mechanisms underlie synaptic plasticity and in turn learning and memory processes?
List of Abbreviations: AKAP79, Akinase anchor protein 79; CaMKII, calcium/calmodulindependent
protein kinase II; CNS, central nervous system; GAP, GTPaseactivating protein; GEF, guanine nucleotide
exchange factors; LTD, longterm depression; LTP, longterm potentiation; NCAM, neural cell adhesion
molecule; PAK, p21activated kinase; PSD, postsynaptic density; SER, smooth endoplasmic reticulum;
SynCAM, synaptic cell adhesion molecule
1 Introduction
Synapses are sites of a specialized cellcell contact between neuronal cells and represent the major structure
involved in chemical neurotransmission in the nervous system. In higher mammals, the majority of brain
excitatory synapses using the neurotransmitter glutamate is found on the principal neuron of the cortex
and hippocampus, the pyramidal cell. Pyramidal cells are characterized by a complex dendritic cytoarch-
itecture harboring approximately 104105 synaptic contact sites with other neurons. It is estimated that only
1% of all synaptic contacts of cortical pyramids is concerned with the wiring to subcortical areas
(Creutzfeldt, 1995), implying that the predominant synapse of the mammalian telencephalon is concerned
with input from a closely related neuron in terms of cell lineage, morphology, and functional character-
istics. This fact is mainly emphasized because our knowledge about synaptic plasticity is to a large degree
based on this type of synaptic input.
Within the central nervous system (CNS), glutamatergic synapses are the major means by which excitatory
signals are conveyed from one neuron to another. As such, glutamatergic synapses are crucially important for
CNS function. Furthermore, it is widely believed that glutamatergic synapses are important loci for modifying
the functional properties of CNS networks, possibly providing the basis for phenomena collectively referred to
as learning and memory. Given their importance, it is not surprising that enormous efforts are being made to
understand the formation, structure, function, and regulation of glutamatergic synapses. To date, significant
progress has been made in our understanding of their ultrastructure, molecular composition, and physiological
properties, as well as the principles of how these synapses are initially assembled.
The term synaptic plasticity covers many different aspects of usedependent synaptic modifications.
Synaptic plasticity is therefore used in this chapter in a broader sense covering aspects of synaptic signal
transmission as well as structural alterations in the molecular makeup of the synapse related to synaptic
signaling events. The scope of this chapter is therefore (1) to introduce to current concepts on the molecular
mechanisms in the formation of spine synapses, (2) to survey current concepts on activitydependent
remodeling of the synaptic cytoskeleton, and (3) to summarize the present knowledge about changes in
spine synapses in brain disease states and during aging. Given that the topic is rather broad, extensive
reference to other recent reviews that cover certain aspects in more detail is given.
Spines are dendritic membrane protrusions that build synaptic contacts to the presynaptic membrane
compartment. The spine density of a typical pyramidal neuron varies between 4 and 10 spines/10 mm
dendritic length. As outlined above, gross calculations suggest that an average principal neuron harbors up
Molecular mechanisms of dendritic spine plasticity in development and aging 10 247
to 10,000 synaptic contacts. > Figure 10-1 depicts a reconstruction of such a pyramidal neuron with its
spine protrusions in two dimensions to give an impression of the morphological complexity that is imposed
by the existence of spinous synaptic contacts.
. Figure 10-1
A Lucifer yellowfilled neuron from the cortex of a gerbil (Meriones unguiculatus) shows numerous dendritic spines
along its dendric arborization. Note the variable shape and size of the dendritic spines in the inset. Imaging of
the neuron and dendrite was done by a confocal laser scanning microscope, with which approximately 80 focal
planes were taken from the specimen, and subsequently volume rendered by a 3D imaging program
Spines have different shapes and their threedimensional structure varies substantially. Mature spines
are typically 0.52 mm in length but can also reach up to 56 mm. Accordingly, the spine volume ranges from
less than 0.01 mm3 to 0.8 mm3. Large spines contain mitochondria and also smooth endoplasmic reticulum
(SER), an internal store of Ca2 (> Figure 10-2). Many attempts have been made in the past to assign spines
into discrete groups. At present, spine synapses are classified commonly into one of the following: filopodia
like protrusions, thin, stubby, mushroomshaped, or cupshaped spines (Hering and Sheng, 2001) (see
> Figure 10-3). This classification scheme, however, should be treated with some caution. Spines can
change their morphology rather rapidly (see also below) and in many cases it is unclear whether certain
spine types exist only as a developmental transition state. Moreover, not all spines can be clearly assigned to
one group, and it is also unclear whether this classification reflects differences in spine function. Accord-
ingly, recent measurements of spine dimension do support the existence of distinct spine categories
(Wallace and Bear, 2004). Despite these limitations, the obvious question that arises is how the different
shape and the molecular dynamics of dendritic spines are brought about.
An important role for the threedimensional organization of the spine synapse is played by an electron
dense structure beneath the postsynaptic membrane termed postsynaptic density (PSD) (> Figure 10-3).
The PSD of the spinous synapse consists of a specialized cytoskeleton, which is crucially involved in the
248 10 Molecular mechanisms of dendritic spine plasticity in development and aging
. Figure 10-2
Dendritic spine and shaft synapse from human hippocampus. Left: A dendritic spine (S) forms an asymmetric
contact with a presynaptic terminal (Pre), while another synaptic bouton contacts the dendritic shaft (D). Note,
both synaptic terminals are filled with numerous clear synaptic vesicles. In addition, the spine synapse contains
some mitochondria (Mi). Scale bar: 0.5 mm. Right: A dendritic shaft synapse, which contains a huge number of
clear synaptic vesicles and mitochondria (Mi), forms multiple contacts with a dendrite (D). Arrowheads indicate
postsynaptic contact zones. Scale bar: 0.5 mm
. Figure 10-3
Different morphologies of dendritic spines depending on type of neurons, developmental stage, and activity
status. From left to right: filopodium, thin, stubby, mushroomshaped, and cupshaped spines
organization of the neurotransmitter receptive apparatus and the adhesion of the postsynapse to presynap-
tic terminals (Sheng, 2001) (see > Figure 10-4). Furthermore, this structure contains specialized and
elaborate molecular scaffolds that link synaptic neurotransmission to various signaling cascades and the
cytoskeleton (McGee and Bredt, 2003; Kennedy et al., 2005). This is of particular importance since the spine
head is highly enriched in actin filaments (Matus et al., 1982), which mediate spine shape changes and
motility (Fischer et al., 1998). Fluoresancerecovery after photobleaching (FRAP) experiments of EGFP
actin in dendritic spines indicate that 85% of the actin in dendritic spines is dynamic, with an average cycle
Molecular mechanisms of dendritic spine plasticity in development and aging 10 249
. Figure 10-4
Molecular structure of the postsynaptic density (PSD) of excitatory synapses. Since excitatory synapses contain
thousands of proteins that form a complex, highly ordered functional meshwork, only a subset of molecules
and protein interactions are shown, focusing on the PSD. Among the key components of the PSD are on one
hand proteins of the MAGUK familylike SAP90/PSD95, Chapsyn110/PSD93, SAP97, and SAP102, which form
complexes with NMDA and AMPAtype glutamate receptors, kainate receptors, cell adhesion proteins like
neuroligin, or signaling proteins like AKAP79 (Akinase anchor protein 79) and kalirin7. AKAP79 is able to
target protein kinase A (not shown) to specific substrates in the PSD. Kalirin7 is a GDP/GTP exchange factor
(GEF) for the small GTPase Rac1, which belongs to the Rho subfamily of small GTPases. Additional signaling
proteins illustrated here are calcium/calmodulindependent protein kinase II (CaMKII), which can interact with
NMDA receptors, and caldendrin, which was shown to bind to Ltype calcium channels. On the other hand,
proteins of the ProSAP/Shank family are important scaffolding molecules that connect neurotransmitter
receptors in the postsynaptic membrane via direct or indirect protein interactions with the actin cytoskeleton
and are therefore believed to play a crucial role in the functional organization of the PSD during synaptogen-
esis, synaptic plasticity, and also in spine morphology. Among the proteins that interact directly with ProSAPs/
Shanks are, for example, Homer proteins, which provide a connection to metabotropic glutamate receptors
(mGluR1a) in the postsynaptic membrane and to inositol1,4,5trisphosphate receptors (InsP3R) in the mem-
brane of the SER. Cortactin and afodrin both connect ProSAPs/Shanks directly to the actin cytoskeleton. GKAP/
SAPAPs provide the link between MAGUK proteins and ProSAPs/Shanks. Recently, it was shown that ProSAPiP,
a novel ProSAPinteracting protein, binds to SPAR (spineassociated Rap GTPaseactivating protein). This
interaction provides an additional link for ProSAPs/Shanks to the actin cytoskeleton via SPAR and also to
MAGUKs, since SPAR was shown to be directly associated with PSD95 or PSD93. Additionally, it is assumed
that SPAR can bind to cellsurface proteins like ephrins or neuroligins
250 10 Molecular mechanisms of dendritic spine plasticity in development and aging
time of 44 s (Star et al. 2002). The actin cytoskeleton is linked to the PSD via a number of protein
interactions including the actinbinding proteins, cortactin and afodrin, and it is thought that this
interface will be responsible for the molecular dynamics of the spine (Oertner and Matus, 2005). At present
it is not yet entirely clear which regulatory events within the PSD control these events (but see also below).
Molecular and ultrastructural analysis of the PSD has lead to the identification of several hundred PSD
constituents (see Li et al., 2004) and to the notion that the PSD is a multilayered complex, composed of
membrane molecules (such as glutamate receptors), primary scaffolding molecules (such as SAP90/PSD95),
and higherorder scaffolding molecules (such as ProSAPs/Shanks) that bind subcomplexes and cytoskeletal
components into even more elaborate structures (> Figure 10-4). These scaffolding molecules harbor a
variety of domains for proteinprotein interactions, including ankyrin repeats, a GUK domain, a SH3
domain, a PDZ domain, a SAM domain, and a prolinerich region (> Figure 10-4). Accordingly, a large
number of proteinprotein interactions have been ascribed to these domains supporting the view that their
multidomain structure underlies their supposed function as scaffolders of the PSD (see > Figure 10-4 for
more details). Part of this function is to integrate signaling molecules like kinases or GTPases into the
complex network of the synaptic cytoskeleton. Such signaling components will eventually connect synaptic
activation to its downstream effectors and can also feedback to the structural organization of the PSD and
the morphology of the spine.
Much progress has been made in recent years in understanding the mechanisms that underlie synaptogen-
esis of excitatory brain synapses (for reviews see Garner et al., 2002; Ziv and Garner, 2004). With the
identification of the protein components of the synapse, their functional characterization with molecular
and genetic approaches, and with recent advances in optical imaging of living neurons it has become
increasingly clear that synapse assembly is a multistep process. In most cases, the development of dendritic
spines occurs concurrently with the growth of the presynaptic terminals, suggesting that cellcell interac-
tions and extrinsic cues likely induce the formation of dendritic spines. Thus, spines and filopodia actively
contact presynaptic afferents (Ziv and Smith, 1996) and can subsequently induce the formation of the
releasecompetent presynaptic membrane, but also the opposite, i.e., ingrowing axons initiating
the emergence of dendritic protrusions, which has been also described (Jontes et al., 2000). Moreover, it
was also suggested that dendritic filopodia participate in the formation of shaft synapses by recruitment of
the presynaptic contact to the dendrite (Sorra and Harris, 2000). Unexpectedly, it was shown that
synaptogenesis in brain as well as in culture can proceed in the complete absence of synaptic transmission
(Verhage et al., 2000; Varoqueaux et al., 2002). This implies that synaptic activity is not required for the
establishment of synaptic contacts. In any case, however, the question remains how the initial contact site
between the axon and the dendrite is specified.
The molecular mechanisms of synapse formation in the brain are not yet completely understood. For
the formation of the presynaptic membrane, synaptic precursor vesicles have been described in the axon
that are 80 nm in diameter and characterized by a dense core appearance. They contain the large presynaptic
multidomain proteins Piccolo and Bassoon, which participate in the formation of the cytoskeletal matrix at
the active zone of mature synapses (Ziv and Garner, 2004). Additionally, other core components of the
active zone are present on these transport vesicles indicating that larger parts of the presynapse will already
arrive as one complex at the initial site of synaptic contact. Interestingly, no corresponding precursor
transport vesicles have been found for the postsynapse (Bresler et al., 2004). Whether a particular molecule
determines the precise site on a dendrite where an axon will form a synapse remains also a largely
unresolved issue. An important role, however, seems to be played by the cell recognition molecules
neurexins and neuroligins (Dean et al., 2003). The most convincing evidence for an instructive role of
the postsynaptic compartment comes from the work of Scheiffele and coworkers (2000), who showed that
expression of the postsynaptic cell adhesion molecule neuroligin in nonneuronal cells is sufficient to trigger
the development of presynaptic structures in contacting axons. An induction of synapse formation after
Molecular mechanisms of dendritic spine plasticity in development and aging 10 251
overexpression of neuroligin by activation of its binding partner neurexin was also found in cultured
hippocampal neurons (Dean et al., 2003).
Many other cell recognition molecules accumulate at synapses during development, which indirectly
suggests that they also might be involved in synaptogenesis or synapse stabilization through their adhesive
properties. These molecules include Ncadherin, which is initially associated with all types of synapses in
cell culture but then becomes restricted to excitatory synapses (Benson and Tanaka, 1998), and members of
the immunoglobulin superfamily, such as synaptic cell adhesion molecule (SynCAM) (Biederer et al.,
2002), nectins (Mizoguchi et al., 2002), and neural cell adhesion molecule (NCAM) (Sytnyk et al., 2002).
Like neuroligin, overexpression of SynCAM in nonneuronal cells induces formation of synapses on the
transfected cells by axons of cocultured neurons (Biederer et al., 2002), whereas inhibition of nectinbased
adhesion results in a decrease of synapse size and a concomitant increase in synapse number (Mizoguchi
et al., 2002). Finally, the ephrinB and EphB tyrosine kinase receptor system has also been implicated in the
development of excitatory synapses through phosphorylation of the cellsurfaceexposed heparin sulphate
proteoglycan syndecan (Ethell et al., 2001) or interaction with the NMDA receptor (Dalva et al., 2000).
These findings suggest that EphB receptors participate not only during the initial formation of synapses but
also in their maturation.
5 Molecular Dynamics of the Plastic Spine or What are Spines Good for
Computer modeling as well as experimental evidence supports the idea that the dendritic spine serves as a
biochemical microcompartment. Ca2 transients can be restricted to single spines, thus isolating the effect
of activation of specific synapses (Yuste et al., 2000; Sabatini et al., 2002). Under physiological conditions,
Ca2 diffusion across the spine neck is negligible, and the spine head functions as a separate compartment
on long timescales, allowing localized Ca2 buildup during trains of synaptic stimuli (Sabatini et al., 2002).
Along these lines it has been proposed by many authors that morphology of synapses and dendritic spines
affects postsynaptic integration of signals, for example, allowing neurons to compensate for the attenuation
of signals from synapses distant from the soma (reviewed by Hausser, 2001). It was also reported that thick
spines (mushroom type) whose head is fully developed remain mostly stable (reviewed by Hering and
Sheng, 2001). Therefore, it was proposed that memory in the cerebral neuronal circuit is possibly stored in
the spine morphology (Oertner and Matus, 2005) and that the spine separation is a specialization for
regulating anatomical plasticity.
A stunning observation made with the invention of novel in vivo imaging techniques in the last decade
is, however, that the structure of dendritic spines is plastic, in the sense that spines can take various shapes
both during neuronal activity and at rest. Spine movements based on alterations in the cytoskeleton have
been postulated on theoretical grounds first by Crick (1982). This hypothesis was based on the idea that
spine structurestabilityfunction relationships are well suited for a molecular account of usedependent
modifications in the brain largely referred to as learning and memory. The first observation that spines are
very mobile structures that seem to constantly change their shape and size at least in culture was made by
the pioneering work of Andrew Matus and coworkers (Fischer et al., 1998, 2000). Furthermore, de novo
spine formation was shown to occur after induction of hippocampal longterm potentiation (LTP) (Engert
and Bonhoeffer, 1999; Toni et al., 1999; but see also Fiala et al., 2002a for contradictory results).
More recent work showed that synapse formation and elimination occur throughout life, although the
magnitude of such changes at distinct developmental stages seems to be quite different in vivo. Interestingly,
thin spines appear and disappear over a few days, while most thick spines persist for months (Grutzendler
et al., 2002; Holtmaat et al., 2005; Zuo et al., 2005). Utilizing transcranial twophoton microscopy, it was
shown that during postnatal week two to four spine retractions exceed additions, resulting in a net loss of
spines (Grutzendler et al., 2002). The fraction of persistent spines grows gradually during development and
into adulthood providing evidence that synaptic circuits continue to stabilize even in the adult brain, long
after the closure of known critical periods (Grutzendler et al., 2002; Holtmaat et al., 2005). Moreover,
filopodialike dendritic protrusions, extending and retracting over hours, are abundant in young animals
252 10 Molecular mechanisms of dendritic spine plasticity in development and aging
but virtually absent in the adults (Trachtenberg et al., 2002; Holtmaat et al., 2005; Zuo et al., 2005)
(> Figure 10-5). In summary, the published data suggest that spines, initially plastic during development,
become remarkably stable in the adult with the majority of them lasting throughout life, and providing a
potential structural basis for longterm information storage. This, however, does not seem to preclude
synapse formation in the adult cortex in the context of experiencedependent plasticity. Thus, experience
dependent plasticity of cortical receptive fields was reported to be accompanied by increased synapse
turnover (Trachtenberg et al., 2002).
. Figure 10-5
Time lapse 3D imaging of a GFPtransfected hippocampal neuron grown in vitro for 10 days. The 3D image
sequence shows ejecting and retracting filopodia as well as highly motile spines along a dendrite during an
observation period of approximately 2 h
Molecular mechanisms of dendritic spine plasticity in development and aging 10 253
To this end, it is largely unknown how synaptic signaling is transferred to the underlying alterations in the
spinous actin cytoskeleton and how altered spine morphology may feedback onto the efficacy of neuro-
transmission. It is assumed that structural alterations in the PSD precede the morphological alterations and
some important mechanisms have been investigated. To this end, it was shown that activitydependent
remodeling of the PSD is accompanied by altered protein turnover, which is remarkably rapid and robust
(Ehlers, 2003). In turn, this remodeling occurs with corresponding increases or decreases in ubiquitin
conjugation of synaptic proteins and requires proteasomemediated degradation (Colledge et al., 2003;
Ehlers, 2003). Functionally, these modifications alter synaptic signaling to major downstream effectors (Pak
et al., 2001, Pak and Sheng, 2003). In addition, a longlasting modification of synaptic strength requires
protein synthesis (Kang and Schuman, 1996; Steward and Worley, 2002), suggesting that newly synthesized
proteins contribute to the modification of synaptic efficacy. The detection of polyribosomes beneath
synapses at the base of dendritic spines (Steward and Levy, 1982) indicates that protein synthesis may
occur locally in dendrites under the control of the synapse itself. This concept implies that mRNAs are
present in dendrites, and that dendritically synthesized proteins can mediate activitydependent structural
and functional modifications of the postsynaptic apparatus. Therefore, active protein synthesis at or in
spines has been proposed to function as a synaptic tag for transcriptiondependent memory formation
(Frey and Morris, 1997). A variety of experimental evidence has underscored the involvement of local
protein synthesis in paradigms of synaptic plasticity. Ostroff and coworkers (2002) could show that
polyribosomes redistribute from dendritic shafts to synapses during LTP in rat hippocampal slices. The
importance of dendritic protein translation for synaptic plasticity has been demonstrated for calcium/
calmodulindependent protein kinase IIa (CaMKIIa) (Miller et al., 2002).
Another important mechanism of synaptic plasticity is through movement of neurotransmitter recep-
tors and regulatory proteins to and from the synapse (Malinow and Malenka, 2002; Kasai et al., 2003).
Several activitytriggered biochemical signaling mechanisms control these movements and their outcome
can therefore be quite diverse (Sheng and Kim, 2002; Kennedy et al., 2005). Activation of glutamate
receptors can cause an increase or decrease of spine size depending upon the activation scheme. These
differing responses may be due to activation of diverse downstream molecules. For example, spine collapse
requires calcineurin activation (Halpain et al., 1998), whereas an increase in size, which is likely due to an
increased recruitment of AMPA receptors to the postsynaptic membrane, requires CaMKII (Jourdain et al.,
2003; Matsuzaki et al., 2004). Interestingly, this pattern is reminiscent of longterm depression (LTD) and
LTP regulation and their morphological correlates in the regulation of spine size (Matsuzaki et al., 2004;
Zhou et al., 2004). LTD has been shown to be accompanied by shrinkage of spine size while induction of
NMDA receptordependent LTP had the opposite effect (Matsuzaki et al., 2004; Zhou et al., 2004). The
preferential sites of LTP induction are small spines, which express a small number of AMPA receptors. In
accord, it was hypothesized that they might correspond to the postsynaptic structures of socalled silent
synapses (Matsuzaki et al., 2004). During LTPinduction, NMDA receptor activation might also engage
molecules that directly affect actin redistribution. For instance, profilin, an actinbinding protein, moves to
the spine head following NMDA receptor activation and suppresses actin dynamics (Ackermann and
Matus, 2003), which leads to spine stabilization. Another actinbinding protein, cortactin, shows a similar
movement to the dendrite after NMDA activation (Hering and Sheng, 2003). A knockdown of cortactin,
however, leads to a decrease in spine density, whereas overexpression results in an elongation of spines
(Hering and Sheng, 2003). In addition to its binding site for actin, cortactin also binds to ProSAPs/Shanks,
thereby providing a physical link between the NMDA receptor and the actin cytoskeleton that could
potentially control spine morphology (Naisbitt et al., 1999). Accordingly, the protein levels of ProSAPs/
Shanks at the synapse control spine shape (Naisbitt et al., 1999).
Other components of the PSD, like SAP90/PSD95, also indirectly link glutamate receptors to the actin
cytoskeleton. This is brought about by its interaction with the small GTPase Ras, through the GTPase
activating protein (GAP) SynGAP (Chen et al., 1998), and Rap, through another GAP, SPAR (Pak et al.,
2001). Other small GTPases, specifically those of the Rho family, play a role in the regulation of spine
254 10 Molecular mechanisms of dendritic spine plasticity in development and aging
motility and morphogenesis (Tashiro et al., 2000). They are activated by guanine nucleotide exchange
factors (GEFs). For one of these GEFs, kalirin7, it was shown that its overexpression can increase spine
density (Penzes et al., 2001), whereas reduced kalirin7 levels are accompanied by shortened dendrites and
significantly lower numbers of dendritic spines (Ma et al., 2003). The putative downstream effectors of
small GTPases that control spine morphology include the p21activated kinase (PAK). Phosphorylated PAK
colocalizes with SAP90/PSD95 in dendritic spines, and transgenic mice expressing a dominant negative
form of PAK show a decreased spine density, morphological spine abnormalities, as well as enhanced LTP
and reduced LTD (Hayashi et al., 2004). Moreover, recent studies show that the interaction of kalirin7 with
Eph tyrosine kinase receptor has important consequences for spine morphology and incidence. These
effects are most likely mediated by the downstream activation of PAK and small GTPases. Transsynaptic
binding of the ligand ephrinB to the EphB receptor leads to a recruitment of kalirin7 to the activated
spine. Kalirin7 subsequently activates Rac1 via its GEF activity, which then results in enhanced PAK activity
and actin remodeling (Penzes et al., 2003), leading to an increased spine density and size. In summary, these
studies are especially important because they will probably lead to a new conceptual framework in our
understanding of how changes in the protein composition of the PSD can regulate spine morphogenesis.
Further work will provide novel answers for the old questionwhich molecular mechanisms underlie
synaptic plasticity and thereby indirectly learning and memory processes?
First accounts of altered spine morphology in brain diseases date back to the 1980s. Given the delicate balance
between the supply and breakdown of PSD molecules, it is likely that seemingly minor genetically or environ-
mentally related perturbations of postsynaptic maintenance mechanisms may eventually lead to synaptic
dysfunction, disintegration, and ultimately elimination, resulting in pathological synaptic loss and degenerative
conditions. Indeed, PSD and dendritic spine pathologies have been associated with a rapidly increasing number
of CNS diseases, ranging from neurodegenerative disease states to epilepsy and schizophrenia (> Figure 10-6).
For example, the loss of synaptic contacts is an early hallmark of both Alzheimer and Huntington disease
(Selkoe, 2002; Smith et al., 2005). The major hereditary mental retardation syndromes, fragile X and Down,
are accompanied by changes in spine morphology, in particular a decrease in mature dendritic spines with well
formed PSDs and an increase in filopodialike protrusions that lack PSDs (Kaufmann and Moser, 2000;
Irwin et al., 2000). The development of epileptic seizures is associated with rather drastic changes in the protein
composition of the PSD (Wyneken et al., 2001, 2003) and the loss of dendritic spines (Mizrahi et al., 2004),
whereas chronic administration of the psychostimulants cocaine and amphetamine produce increased
spine density in the nucleus accumbens (Robinson and Kolb, 1997, 1999). In schizophrenia, one of the
most prevalent psychiatric diseases, dysplastic events at the dendritic spine level not only are a common
observation but also are thought to be causally related to the onset of the disease (Costa et al., 2001;
Blanpied and Ehlers, 2004). More recently, downregulation of the expression of the synaptic scaffolding
protein SAP90/PSD95 has been causally linked to synaptic dysfunction in cocaineinduced drug addic-
tion (Yao et al., 2004). This list could be extended further, although a question arises whether such
alterations are really the cause or merely a consequence of a particular disease (Fiala et al., 2002b).
At present, our knowledge about alterations in dendritic spines of the aged brain is still sparse. Most reports
are concerned with the morphology and number of spines in the aged brain. But very little is known about
the underlying mechanisms of such alterations. Aging is often accompanied by cognitive impairments like
memory and spatiallearning deficits in senescent humans, nonhuman primates, and rodents. In contrast to
initial studies of neuronal densities, more recent studies based on stereological methods and counting of
Molecular mechanisms of dendritic spine plasticity in development and aging 10 255
. Figure 10-6
Spine pathologies. (b) A reduction of the density of dendritic spines as compared to (a) the spines occur in cases of
deafferentation, most mental retardation syndromes, severe alcohol abuse, epilepsy, Alzheimers disease, and
others. (c) An increase in dendritic spine density was shown for some types of deafferentation syndromes, the
fragile X syndrome, sudden infant death syndrome, and the abuse of psychostimulants. (d) Sensory deprivation
leads to a reduction of spine size. The same spine phenotype is described for Downs syndrome and schizophrenia.
Morphological changes that result in a (e) distortion of spine shape are also typical for most mental retardation
symptoms, epilepsy, severe alcohol abuse, poisoning with neurotoxicants, and spongiform encephalitis
total neuron number of brain material of several species have established that only little, if any, loss of
excitatory and inhibitory neurons occurs in different brain regions in normal aging and that this loss cannot
account for the observed agerelated functional deficits in multiple neurotransmitter systems. Studies have
demonstrated subregional specificities in sensitivity to the effects of advancing age, and it has become
increasingly clear that the mechanisms underlying agerelated changes are of a more subtle nature and
sometimes compensatory rather than detrimental to cognitive functioning (Morrison and Hof, 1997).
Disruption and agerelated structural alterations of the myelin sheaths of neocortical axons have been
described and correlated with age and degree of impairment in monkeys (Peters et al, 2000). In both
rodents and humans, changes have been reported in dendritic arbor, spines, and synapse morphology
(Morrison and Hof, 1997). Spine number and density seem to decrease with age (Duan et al., 2003), and
dendritic length seems to decline in cortical neurons (Jacobs et al., 1997). Rather than changes in spine
number, reduction in size of perforated postsynaptic densities, whose formation is postulated to be a
structural correlate of enhanced efficacy of synaptic transmission, has been reported for hippocampal
axospinous synapses (Nicholson et al., 2004). Studies in the monkey revealed decreased synaptic excitation
and increased synaptic inhibition of layer 2 and 3 pyramidal cells in the prefrontal cortex (Luebke et al.,
2004). Thus, alterations first appear on the level of synaptic connectivity and further manifest themselves on
the subcellular and molecular level.
BertoniFreddari and coworkers (2003, 2004, 2005) recently described ageinduced changes in cellular
energy metabolism. They observed an increase in the number of enlarged synaptic mitochondria with
generally decreased metabolic competence in cerebellar cortex of adult rats, as well as an agerelated decline
256 10 Molecular mechanisms of dendritic spine plasticity in development and aging
in metabolic competence of small and mediumsized synaptic mitochondria. Age also affects Ca2
homeostasis and changes in regulation of intracellular Ca2 concentrations underlie some agerelated
deficits in plasticity and cognition (reviewed in Foster, 1999; Foster and Kumar, 2002; Rosenzweig and
Barnes, 2003; Toescu and Verkhratsky, 2004). A number of mechanisms for handling Ca2 are thought to be
modified during aging, including intracellular buffering, extrusion, and influx of Ca2. Consequently,
hippocampal synaptic function is altered as seen by a shift in synaptic plasticity thresholds with LTP
requiring a higher stimulation threshold for induction and more induction sessions to saturate
LTP mechanisms in aged animals, and LTD showing a reduction in threshold, such that robust LTD is
observed for lower frequency stimulation patterns in aged compared with young animals. Thus, aged
animals acquire new information only slowly and with more repetition and forget rapidly. In this context, it
is hypothesized that a postsynaptic increase in Ca2 influx through Ltype VDCCs and reduced Ca2 influx
through NMDARs with age not only modifies the threshold for induction of Ca2dependent synaptic
plasticity in favor of LTD induction but also increases the duration and amplitude of afterhypolarization
following cell discharge activity and with that alters information processing. The effects seem to be
mediated by agerelated changes in the Ca2signaling pathways, including the activity of protein phos-
phatases and kinases involved in the expression of synaptic plasticity. The Ca2 and CaMdependent
phosphatase calcineurin and protein phosphatase 1, which is regulated by calcineurin, are enzymes critical
for LTD and according to Foster and coworkers (2001), activity of both increases with advanced age in an
Lchannel related manner. They showed that the phosphorylation state of calcineurin substrates (e.g., of
CREB) was reduced and that increased calcineurin activity correlated with memory function. This finding
is challenged by a more recent study by Agbas and coworkers (2005) who showed that calcineurin activity
decreases in the aging brain due to increased aggregation of the catalytic subunit of calcineurin caused by
oxidation.
Other components of the synaptic transmission machinery have also been analyzed for agerelated
changes. Whereas several groups examined the expression of membrane channel subtypes and receptor
subunits in cortical and hippocampal neurons of different species and documented regional and subtype/
subunit and sometimes species specific alterations (Tanaka and Ando, 2001; Hof et al., 2002; Bai et al., 2004;
Iwamoto et al., 2004), levels of presynaptic proteins like synaptotagmin1, synaptophysin, and syntaxin in
rat cerebral cortex were not significantly reduced with age (Iwamoto et al., 2004). Interestingly, in animal
models some of the described agerelated changes and effects on cognitive function can be reversed either
directly or in a complementary way by the administration of trophic factors that otherwise decline with age,
like insulinlike growth factor1 and estrogen (Sonntag et al., 2000; Adams and Morrison, 2003; Adams
et al., 2004; Shi et al., 2005).
Acknowledgments
The work of M.R.K. is supported by the BMBF, DAAD, DFG, the Land SachsenAnhalt, and the Schram
Foundation.
References
Ackermann M, Matus A. 2003. Activityinduced targeting of Agbas A, Zaidi A, Michaelis EK. 2005. Decreased activity and
profilin and stabilization of dendritic spine morphology. increased aggregation of brain calcineurin during aging.
Nat Neurosci 6: 1194-2000. Brain Res 1059: 59-71.
Adams MM, Morrison JH. 2003. Estrogen and the aging Bai L, Hof PR, Standaert DG, Xing Y, Nelson SE, et al. 2004.
hippocampal synapse. Cereb Cortex 13: 1271-1275. Changes in the expression of the NR2B subunit during
Adams MM, Fink SE, Janssen WGM, Shah RA, Morrison JH. aging in macaque monkeys. Neurobiol Aging 25: 201-208.
2004. Estrogen modulates synaptic NmethylDaspartate Benson DL, Tanaka H. 1998. Ncadherin redistribution dur-
receptor subunit distribution in the aged hippocampus. J ing synaptogenesis in hippocampal neurons. J Neurosci 18:
Comp Neurol 474: 419-426. 6892-6904.
Molecular mechanisms of dendritic spine plasticity in development and aging 10 257
BertoniFreddari C, Fattoretti P, Giorgetti B, Solazzi M, Engert F, Bonhoeffer T. 1999. Dendritic spine changes asso-
Balietti M, et al. 2004. Cytochrome oxidase activity in hip- ciated with hippocampal longterm synaptic plasticity.
pocampal synaptic mitochondria during aging: A quantitative Nature 399: 19-21.
cytochemical investigation. Ann N Y Acad Sci 1019: 33-36. Ethell IM, Irie F, Kalo MS, Couchman JR, Pasquale EB, et al.
BertoniFreddari C, Fattoretti P, Giorgetti B, Spazzafumo L, 2001. EphB/syndecan2 signaling in dendritic spine mor-
Solazzi M, et al. 2005. Agerelated decline in metabolic phogenesis. Neuron 31: 1001-1013.
competence of small and mediumsized synaptic mito- Fiala JC, Allwardt B, Harris KM. 2002a. Dendritic spines
chondria. Naturwissenschaften 92: 82-85. do not split during hippocampal LTP or maturation. Nat
BertoniFreddari C, Fattoretti P, Paoloni R, Caselli U, Neurosci 5: 297-298.
Giorgetti B, et al. 2003. Inverse correlation between mito- Fiala JC, Spacek J, Harris KM. 2002b. Dendritic spine pathol-
chondrial size and metabolic competence: A quantitative ogy: Cause or consequence of neurological disorders? Brain
cytochemical study of cytochrome oxidase activity. Natur- Res Rev 39: 29-54.
wissenschaften 90: 68-71. Fischer M, Kaech S, Knutti D, Matus A. 1998. Rapid actin
Biederer T, Sara Y, Mozhayeva M, Atasoy D, Liu X, et al. 2002. based plasticity in dendritic spines. Neuron 20: 847-854.
SynCAM, a synaptic adhesion molecule that drives synapse Fischer M, Kaech S, Wagner U, Brinkhaus H, Matus A. 2000.
assembly. Science 297: 1525-1531. Glutamate receptors regulate actinbased plasticity in den-
Blanpied TA, Ehlers MD. 2004. Microanatomy of dendritic dritic spines. Nat Neurosci 3: 887-894.
spines: Emerging principles of synaptic pathology in Foster TC. 1999. Involvement of hippocampal synaptic plas-
psychiatric and neurological disease. Biol Psychiatry 55: ticity in agerelated memory decline. Brain Res Brain Res
1121-1127. Rev 30: 236-249.
Bresler T, Shapira M, Boeckers T, Dresbach T, Futter M, et al. Foster TC, Kumar A. 2002. Calcium dysregulation in the aging
2004. Postsynaptic density assembly is fundamentally dif- brain. Neuroscientist 8: 297-301.
ferent from presynaptic active zone assembly. J Neurosci 24: Foster TC, Sharrow KM, Masse JR, Norris CM, Kumar A.
1507-1520. 2001. Calcineurin links Ca2 dysregulation with brain
Chen HJ, RojasSoto M, Oguni A, Kennedy MB. 1998. A aging, J Neurosci 21: 4066-4073.
synaptic RasGTPase activating protein (p135 SynGAP) Frey JU, Morris RG. 1997. Synaptic tagging and longterm
inhibited by CaM kinase II. Neuron 20: 895-904. potentiation. Nature 385: 533-536.
Colledge M, Snyder EM, Crozier RA, Soderling JA, Jin Y, et al. Garner CC, Zhai G, Gundelfinger ED, Ziv NE. 2002. Molecu-
2003. Ubiquitination regulates PSD95 degradation and lar mechanisms of CNS synaptogenesis. Trends Neurosci
AMPA receptor surface expression. Neuron 40: 595-607. 25: 243-250.
Costa E, Davis J, Grayson DR, Guidotti A, Pappas GD, et al. Grutzendler J, Kasthuri N, Gan WB. 2002. Longterm
2001. Dendritic spine hypoplasticity and downregulation of dendritic spine stability in the adult cortex. Nature 420:
reelin and GABAergic tone in schizophrenia vulnerability. 812-816.
Neurobiol Dis 8: 723-742. Halpain S, Hipolito A, Saffer L. 1998. Regulation of Factin
Creutzfeldt O. 1995. Cortex cerebri: Performance, structural stability in dendritic spines by glutamate receptors and
and functional organization of the cortex. Oxford: Oxford calcineurin. J Neurosci 18: 9835-9844.
University Press. Hausser M. 2001. Synaptic function: Dendritic democracy.
Crick F. 1982. Do dendritic spines twitch? Trends Neurosci 5: Curr Biol 11: 10-12.
44-46. Hayashi ML, Choi SY, Rao BS, Jung HY, Lee HK, et al. 2004.
Dalva MB, Takasu MA, Lin MZ, Shamah SM, Hu L, et al. Altered cortical synaptic morphology and impaired mem-
2000. EphB receptors interact with NMDA receptors and ory consolidation in forebrainspecific dominantnegative
regulate excitatory synapse formation. Cell 103: 945-956. PAK transgenic mice. Neuron 42: 773-787.
Dean C, Scholl FG, Choih J, De Maria S, Berger J, et al. 2003. Hering H, Sheng M. 2001. Dendritic spines: Structure,
Neurexin mediates the assembly of presynaptic terminals. dynamics and regulation. Nat Rev Neurosci 2: 880-888.
Nat Neurosci 6: 708-716. Hering H, Sheng M. 2003. Activitydependent redistribution
Duan H, Wearne SL, Rocher AB, Macedo A, Morrison JH, and essential role of cortactin in dendritic spine morpho-
et al. 2003. Agerelated dendritic and spine changes in genesis. J Neurosci 23: 11759-11769.
corticocortically projecting neurons in macaque monkeys. Hof PR, Duan H, Page TL, Einstein M, Wicinski B, et al. 2002.
Cereb Cortex 13: 950-961. Agerelated changes in GluR2 and NMDAR1 glutamate
Ehlers MD. 2003. Activity level controls postsynaptic compo- receptor subunit protein immunoreactivity in corticocorti-
sition and signaling via the ubiquitinproteasome system. cally projecting neurons in macaque and patas monkeys.
Nat Neurosci 6: 231-242. Brain Res 928: 175-186.
258 10 Molecular mechanisms of dendritic spine plasticity in development and aging
Holtmaat AJ, Trachtenberg JT, Wilbrecht L, Shepherd GM, Matus A, Ackermann M, Pehling G, Byers HR, Fujiwara K.
Zhang X, et al. 2005. Transient and persistent dendritic 1982. High actin concentrations in brain dendritic spines
spines in the neocortex in vivo. Neuron 45: 279-291. and postsynaptic densities. Proc Natl Acad Sci USA 79:
Irwin SA, Galvez R, Greenough WT. 2000. Dendritic spine 7590-7594.
structural anomalies in fragileX mental retardation syn- McGee AW, Bredt DS. 2003. Assembly and plasticity of the
drome. Cereb Cortex 10: 1038-1044. glutamatergic postsynaptic specialization. Curr Opin Neu-
Iwamoto M, Hagishita T, ShojiKasai Y, Ando S, Tanaka Y. robiol 13: 111-118.
2004. Agerelated changes in the levels of voltage Miller S, Yasuda M, Coats JK, Jones Y, Martone ME, et al.
dependent calcium channels and other synaptic proteins 2002. Disruption of dendritic translation of CaMKIIa
in rat brain cortices. Neurosci Lett 366: 277-281. impairs stabilization of synaptic plasticity and memory
Jacobs B, Driscoll L, Schall M. 1997. Lifespan dendritic and consolidation. Neuron 36: 507-519.
spine changes in areas 10 and 18 of human cortex: A Mizoguchi A, Nakanishi H, Kimura K, Matsubara K, Ozaki
quantitative Golgi study. J Comp Neurol 386: 661-680 Kuroda K, et al. 2002. Nectin: An adhesion mole-
Jontes JD, Buchanan J, Smith SJ. 2000. Growth cone and cule involved in formation of synapses. J Cell Biol 156:
dendrite dynamics in zebrafish embryos: Early events in 555-565.
synaptogenesis imaged in vivo. Nat Neurosci 3: 231-237. Mizrahi A, Crowley JC, Shtoyerman E, Katz LC. 2004. High
Jourdain P, Fukunaga K, Muller D. 2003. Calcium/calmodulin resolution in vivo imaging of hippocampal dendrites and
dependent protein kinase II contributes to activitydependent spines. J Neurosci 24: 3147-3151.
filopodia growth and spine formation. J Neurosci 23: 10645- Morrison JH, Hof PR. 1997. Life and death of neurons in the
10649. aging brain. Science 278: 412-419.
Kang H, Schuman EM. 1996. A requirement for local protein Naisbitt S, Kim E, Tu JC, Xiao B, Sala C, et al. 1999. Shank, a
synthesis in neurotrophininduced hippocampal synaptic novel family of postsynaptic density proteins that binds to
plasticity. Science 273: 1402-1406. the NMDA receptor/PSD95/GKAP complex and cortactin.
Kasai H, Matsuzaki M, Noguchi J, Yasumatsu N, Nakahara H. Neuron 23: 569-582.
2003. Structurestabilityfunction relationships of dendrit- Nicholson DA, Yoshida R, Berry RW, Gallagher M, Geinisman
ic spines. Trends Neurosci 28: 360-368. Y. 2004. Reduction in size of perforated postsynaptic
Kaufmann WE, Moser HW. 2000. Dendritic anomalies in densities in hippocampal axospinous synapses and age
disorders associated with mental retardation. Cereb Cortex related spatial learning impairments. J Neurosci 24:
10: 981-991. 7648-7653.
Kennedy MB, Beale HC, Carlisle HJ, Washburn LR. 2005. Oertner TG, Matus A. 2005. Calcium regulation of actin
Integration of biochemical signaling in spines. Nat Rev dynamics in dendritic spines. Cell Calcium 37: 477-482.
Neurosci 6: 423-434. Ostroff LE, Fiala JC, Allwardt B, Harris KM. 2002. Polyribo-
Li KW, Hornshaw MP, Van Der Schors RC, Watson R, Tate S, somes redistribute from dendritic shafts into spines with
et al. 2004. Proteomics analysis of rat brain postsynaptic enlarged synapses during LTP in developing rat hippocam-
density. Implications of the diverse protein functional pal slices. Neuron 35: 535-545.
groups for the integration of synaptic physiology. J Biol Pak DT, Sheng M. 2003. Targeted protein degradation and
Chem 279: 987-1002. synapse remodeling by an inducible protein kinase. Science
Luebke JI, Chang YM, Moore TL, Rosene DL. 2004. Normal 302: 1368-1373.
aging results in decreased synaptic excitation and increased Pak DT, Yang S, RudolphCorreia S, Kim E, Sheng M. 2001.
synaptic inhibition of layer 2/3 pyramidal cells in the mon- Regulation of dendritic spine morphology by SPAR, a PSD
key prefrontal cortex. Neuroscience 125: 277-288. 95associated RapGAP. Neuron 31: 289-303.
Ma XM, Huang J, Wang Y, Eipper BA, Mains RE. 2003. Penzes P, Beeser A, Chernoff J, Schiller MR, Eipper BA, et al.
Kalirin, a multifunctional Rho guanine nucleotide 2003. Rapid induction of dendritic spine morphogenesis by
exchange factor, is necessary for maintenance of hippocam- transsynaptic ephrinBEphB receptor activation of the
pal pyramidal neuron dendrites and dendritic spines. J RhoGEF kalirin. Neuron 37: 263-274.
Neurosci 23: 10593-10603. Penzes P, Johnson RC, Sattler R, Zhang X, Huganir RL, et al.
Malinow R, Malenka RC. 2002. AMPA receptor trafficking 2001. The neuronal RhoGEF Kalirin7 interacts with PDZ
and the control of synaptic plasticity. Annu Rev Neurosci domaincontaining proteins and regulates dendritic mor-
25: 103-126. phogenesis. Neuron 29: 229-242.
Matsuzaki M, Honkura N, EllisDavis GC, Kasai H. 2004. Peters A, Moss MB, Sethars A. 2000. Effects of aging on
Structural basis of longterm potentiation. Nature 429: myelinated nerve fiber in monkey primary visual cortex. J
761-766. Comp Neurol 419: 364-376.
Molecular mechanisms of dendritic spine plasticity in development and aging 10 259
Robinson TE, Kolb B. 1997. Persistent structural modifica- Tashiro A, Minden A, Yuste R. 2000. Regulation of dendritic
tions in nucleus accumbens and prefrontal cortex neurons spine morphology by the rho family of small GTPases:
produced by previous experience with amphetamine. Antagonistic roles of Rac and Rho. Cereb Cortex 10:
J Neurosci 17: 8491-8497. 927-938.
Robinson TE, Kolb B. 1999. Alterations in the morphology of Toescu EC, Verkhratsky A. 2004. Ca2 and mitochondria as
dendrites and dendritic spines in the nucleus accumbens substrates for deficits in synaptic plasticity in normal brain
and prefrontal cortex following repeated treatment with ageing. J Cell Mol Med 8: 181-190.
amphetamine or cocaine. Eur J Neurosci 11: 1598-1604. Toni N, Buchs PA, Nikonenko I, Bron CR, Muller D. 1999.
Rosenzweig ES, Barnes CA. 2003. Impact of aging on hippo- LTP promotes formation of multiple spine synapses be-
campal function: Plasticity, network dynamics, and cogni- tween a single axon terminal and a dendrite. Nature 402:
tion. Prog Neurobiol 69: 143-179. 421-425.
Sabatini BL, Oertner TG, Svoboda K. 2002. The life cycle of Trachtenberg JT, Chen BE, Knott GW, Feng G, Sanes JR, et al.
Ca2 ions in dendritic spines. Neuron 2002 33: 439-452. 2002. Longterm in vivo imaging of experiencedependent
Scheiffele P, Fan J, Choih J, Fetter R, Serafini T. 2000. Neuro- synaptic plasticity in adult cortex. Nature 420: 788-794.
ligin expressed in nonneuronal cells triggers presynaptic Varoqueaux F, Sigler A, Rhee JS, Brose N, Enk C, et al. 2002.
development in contacting axons. Cell 101: 657-669. Total arrest of spontaneous and evoked synaptic transmis-
Selkoe DJ. 2002. Alzheimers disease is a synaptic failure. sion but normal synaptogenesis in the absence of Munc13
Science 298: 789-791. mediated vesicle priming. Proc Natl Acad Sci USA 99:
Sheng M. 2001. Molecular organization of the postsynaptic 9037-9042.
specialization. Proc Natl Acad Sci USA 98: 7058-7061. Verhage M, Maia AS, Plomp JJ, Brussaard AB, Heeroma JH,
Sheng M, Kim MJ. 2002. Postsynaptic signaling and plasticity et al. 2000. Synaptic assembly of the brain in the absence of
mechanisms. Science 298: 776-780. neurotransmitter secretion. Science 287: 864-869.
Shi L, Linville MC, Tucker EW, Sonntag WE, BrunsoBechtold Wallace W, Bear MF. 2004. A morphological correlate of
JK. 2005. Differential effects of aging and insulinlike synaptic scaling in visual cortex. J Neurosci 24: b6928-
growth factor1 on synapses in CA1 of rat hippocampus. b6938.
Cereb Cortex 15: 571-577. Wyneken U, Marengo JJ, Villanueva S, Soto D, Sandoval R,
Smith R, Brundin P, Li JY. 2005. Synaptic dysfunction in et al. 2003. Epilepsyinduced changes in signalling systems
Huntingtons disease: A new perspective. Cell Mol Life Sci of human and rat postsynaptic densities. Epilepsia 44:
62: 1901-1912. 243-246.
Sonntag WE, Bennett SA, Khan AS, Thornton PL, Xu X, et al. Wyneken U, Smalla KH, Marengo JJ, Soto D, de la Cerda A,
2000. Age and insulinlike growth factor1 modulate N et al. 2001. Kainateinduced seizures alter protein compo-
methylDaspartate receptor subtype expression in rats. sition and NmethylDaspartate receptor function of
Brain Res Bull 51: 331-338. rat forebrain postsynaptic densities. Neuroscience 102:
Sorra KE, Harris KM. 2000. Overview on the structure, com- 65-74.
position, function, development, and plasticity of hippo- Yao WD, Gainetdinov RR, Arbuckle MI, Sotnikova TD, Cyr
campal dendritic spines. Hippocampus 10: 501-511. M, et al. 2004. Identification of PSD95 as a regulator of
Star EN, Kwiatkowski DJ, Murthy VN. 2002. Rapid turnover dopaminemediated synaptic and behavioral plasticity.
of actin in dendritic spines and its regulation by activity. Neuron 41: 625-638.
Nat Neurosci 5: 239-246. Yuste R, Majewska A, Holthoff K. 2000. From form to func-
Steward O, Levy WB. 1982. Preferential localization of poly- tion: Calcium compartmentalization in dendritic spines.
ribosomes under the base of dendritic spines in granule Nat Neurosci 3: 653-659.
cells of the dentate gyrus. J Neurosci 2: 284-291. Zhou Q, Homma KJ, Poo MM. 2004. Shrinkage of dendritic
Steward O, Worley P. 2002. Local synthesis of proteins at spines associated with longterm depression of hippocam-
synaptic sites on dendrites: Role in synaptic plasticity and pal synapses. Neuron 44: 749-757.
memory consolidation? Neurobiol Learn Mem 78: 508-527. Ziv NE, Garner CC. 2004. Cellular and molecular mechan-
Sytnyk V, Leshchynska I, Delling M, Dityateva G, Dityateva A, isms of presynaptic assembly. Nat Rev Neurosci 5: 385-399.
Schachner M. 2002. Neural cell adhesion molecule pro- Ziv NE, Smith SJ. 1996. Evidence for a role of dendritic
motes accumulation of TGN organelles at sites of neuron- filopodia in synaptogenesis and spine formation. Neuron
to-neuron contacts. J Cell Bio 159: 649-661. 17: 91-102.
Tanaka Y, Ando S. 2001. Agerelated changes in the subtypes Zuo Y, Lin A, Chang P, Gan WB. 2005. Development of long
of voltagedependent calcium channels in rat brain cortical term dendritic spine stability in diverse regions of cerebral
synapses. Neurosci Res 39: 213-220. cortex. Neuron 46: 181-189.
11 Alzheimers Disease
BACE Proteases
S. Roner . S. F. Lichtenthaler
Abstract: The aberrant proteolytical processing of the amyloid precursor protein (APP) gives rise to
bamyloid peptides, which accumulate as perivascular or parenchymal deposits in brains of Alzheimers
disease (AD) patients. The generation of bamyloid peptides is initiated by APP cleavage by an aspartyl
protease named betasite APPcleaving enzyme 1 (BACE1), followed by gsecretase cleavage. The discovery
of BACE1 some years ago and subsequent studies on the regulation of its expression and enzymatic activity
allowed for a better understanding of APP processing, bamyloid generation and the pathogenesis of AD in
general. Some of the unique features of BACE1, such as regulatory elements in the 50 region of BACE1
mRNA and factors that contribute to its cell typespecific activation helped to explain epidemiological
observations and opened new rational therapeutic opportunities. However, although BACE1 has been
shown to be the only protease in brain with significant bsecretase activity, there are a number of alternative
substrates in addition to APP, indicating that a careful estimation of possible side effects of BACE1 inhibi-
tors is required. Nevertheless, experimental data from transgenic mouse models of AD, BACE1 knockout
mice and pharmacological studies corroborate the potential usefulness of drugs that interfere with BACE1
expression and/or enzymatic activity for the treatment of AD.
List of Abbreviations: Ab, bamyloid peptide; AD, Alzheimers disease; ADAM, a disintegrin and metal-
loprotease; APLP, amyloid precursorlike protein; APP, amyloid precursor protein; BACE, betasite APP
cleaving enzyme; CCS, copper chaperone for superoxide dismutase1; ChAT, choline acetyltransferase; DS,
Downs syndrome; ER, endoplasmatic reticulum; FAD, familial forms of Alzheimers disease; GFAP, glial
fibrillary acidic protein; GGA, Golgilocalized gearcontaining ARF binding; HEK, human embryonic
kidney; MDCK, MadinDarby canine kidney; NSAIDs, nonsteroidal antiinflammatory drugs; PLSCR 1,
phospholipid scramblase 1; PPARg, peroxisome proliferatoractivated receptorg; PPRE, PPARg responsive
element; pro, prodomain; RTN3, reticulon 3; RTN4b, reticulon 4b; TGN, transGolgi network; SP,
signal peptide; TACE, tumor necrosis factor aconverting enzyme; TMD, transmembrane domain;
50 UTR, 50 untranslated region; uORF, upstream open reading frame
Alzheimers disease (AD) is the most prevalent neurodegenerative disease, affecting about 20 million people
worldwide [for an overview see Selkoe and Schenk (2003)]. It is characterized by progressive loss of
memory, declining cognitive function and, ultimately, leads to decreasing physical functions and death.
The neuropathological hallmarks of AD are the senile plaques and the neurofibrillary tangles, which are
protein aggregates deposited in the brain. The neurofibrillary tangles represent intraneuronal bundles of
paired helical filaments, which mainly consist of the microtubuleassociated protein tau in an abnormally
phosphorylated form [(for a review see Iqbal et al. (2005)]. The extracellular amyloid plaques mainly
consist of the 42 residue long amyloid bpeptide (Ab, bA4) (Glenner and Wong, 1984; Masters et al., 1985),
which is proteolytically derived from the much larger amyloid precursor protein (APP, bAPP) (Kang et al.,
1987). The generation and subsequent aggregation of Ab seem to be at the origin of the disease and are
believed to trigger a complex pathological cascade that causes neuronal dysfunction, the appearance of
the neurofibrillary tangles, inflammatory processes, neuronal loss, cognitive impairment, and finally the
onset of the disease. This pathological cascade, which is now widely accepted, is also referred to as the
amyloid hypothesis of AD [for overview see Hardy and Selkoe (2002), Citron (2004)].
Although drugs are currently available that may ameliorate latestage symptoms like the cognitive deficits
for a short time, no drugs are on the market that specifically target the cellular mechanisms of the disease,
namely the proteolytic generation of the Ab peptide from APP. APP is a type I membrane protein with unclear
biological function. Suggested functions for APP include a role in cell adhesion and copper homeostasis [for a
review see De Strooper and Annaert (2000)], cellular motility (Sabo et al., 2001) and being a cell surface
receptor (Cao and Sudhof, 2001) or a receptor for the motor protein kinesin (Kamal et al., 2001).
The proteolytic processing of APP involves protease activities that are referred to as a, b, and
gsecretase because their molecular identity was initially not known. The three proteases process APP in
two different pathways (> Figure 11-1a), one of which is called antiamyloidogenic because it prevents Ab
generation. The other one is termed amyloidogenic as it leads to the generation of the Ab peptide. Both
Alzheimers disease BACE proteases
11 263
. Figure 11-1
Proteolytic processing pathways of APP. (a) Amyloidogenic pathway: Cleavage of APP by the protease
bsecretase (BACE1) occurs at the Nterminus of the Abdomain and yields the secreted sAPPb and a Cterminal
fragment of APP of 99 amino acids (C99). C99 is further cleaved within its transmembrane domain by
gsecretase, leading to the secretion of the Ab peptide and the generation of the APP intracellular domain
(AICD). The consecutive cleavage of APP by b and gsecretase constitutes the amyloidogenic pathway as it
generates Ab. Antiamyloidogenic pathway: Cleavage of APP by asecretase or potentially BACE2 within the Ab
domain yields the neurotrophic and neuroprotective sAPPa as well as the Cterminal fragment C83. The
asecretase is a member of the ADAM family of metalloproteases. Similar to the processing of C99, C83 is
further processed by gsecretase within its transmembrane domain, leading to the generation of AICD and to
the secretion of the p3 peptide that is not deposited in amyloid plaques. (b) Amino acid sequence of APP within
and around the Ab peptide domain. Indicated are the major cleavage sites of the different protease activities
and the double mutation directly preceding the bsecretase cleavage site. This mutation is responsible for a
familial form of AD found in a Swedish family. Amino acids are shown in the oneletter code and numbered
according to the Ab sequence
pathways and their cleavage products (including Ab) are part of the normal proteolytic processing of APP
[for an overview see Selkoe and Schenk (2003)], which occurs in most cells. In the amyloidogenic pathway,
APP is first cleaved by the bsecretase at the Nterminus of the Ab domain (> Figure 11-1a). This cleavage
generates the soluble sAPPb and a Cterminal fragment, which undergoes a second cleavage by a protease
called gsecretase. The bsecretase was recently identified as the membranebound aspartyl protease BACE1
(also named memapsin2, Asp2). gSecretase also belongs to the aspartyl family of proteases but is an
unusual protease activity in that it cleaves APP within its hydrophobic transmembrane domain. gSecretase
is a heterotetrameric protein complex consisting of presenilin, nicastrin, PEN2, and APH1 [for a review
see Haass (2004)]. gSecretase cleavage occurs mainly after residue 40 and to a lower extent after residue
42 of the Ab sequence, thereby generating the 40 and 42 amino acid long Ab40 and Ab42 peptides (> Figure
11-1b). Due to its more hydrophobic nature and its higher tendency to aggregate, it is the Ab42 peptide but
not the Ab40 peptide, which is assumed to be the key player in the pathogenesis of AD. In fact, the amyloid
plaques mainly consist of Ab42. The physiological relevance of b and gsecretase in Ab generation is clearly
established. Mice deficient in the bsecretase or the gsecretase activity do not generate Ab, making both
proteases suitable drug targets for AD [reviewed in Haass (2004)].
264
11 Alzheimers disease BACE proteases
In contrast, the antiamyloidogenic pathway starts with APP cleavage by asecretase, which cuts
within the Ab domain (> Figure 11-1a, b). Thus, asecretase cleavage precludes Ab generation. Following
acleavage the Cterminal APP fragment undergoes intramembrane cleavage by the gsecretase complex.
This leads to the generation of the p3 peptide (Haass et al., 1993) (> Figure 11-1a), which is assumed to be
benign as it is not found in the AD amyloid plaques. aSecretase is a member of the ADAMfamily
of proteases (A disintegrin and metalloprotease) [for a review see Allinson et al. (2003)], and is either
ADAM10 (Lammich et al., 1999), ADAM17/tumor necrosis factor aconverting enzyme (TACE) (Buxbaum
et al., 1998) or even ADAM9 (Koike et al., 1999). At present, it is unclear, whether only one of them or all
three together constitute the physiologically relevant asecretase. An asecretaselike proteolytic cleavage at
residues 19 and 20 of Ab (> Figure 11-1b) is catalyzed by the protease BACE2 (Farzan et al., 2000; Yan et al.,
2001), which is a close homolog of the bsecretase BACE1. Because BACE2 cleaves within the Ab sequence,
it does not contribute to Ab formation, which is consistent with the finding, that a BACE1 knockout
alone in mice is sufficient to completely suppress Ab generation (Cai et al., 2001; Luo et al., 2001; Roberds
et al., 2001).
The essential role of the secretases in the generation or prevention of the Ab peptide is underlined by
genetic evidence. Mutations in the APP gene linked to familial forms of Alzheimers disease (FAD) are
found at or very close to the secretase cleavage sites and generally increase Ab generation. The Swedish
double mutation, which was found in a Swedish FAD family, is located directly Nterminally to the
bsecretase cleavage site (> Figure 11-1b) and allows a much more efficient bsecretase cleavage of APP
and turnover to Ab. Mutations close to the gsecretase site do not increase the total amount of Ab being
generated but increase the amount of the longer Ab42 species relative to the Ab40 peptide. Additionally,
FAD mutations were found close to the asecretase cleavage site, and appear to reduce asecretase cleavage,
resulting in an increased bsecretase cleavage of APP and thus in more Ab generation.
The following chapters highlight various aspects of BACE1 and BACE2 protease biology, starting with a
description of the discovery of both proteases.
Early on in the research of the APP secretases, cell biological and pharmacological studies revealed that
bsecretase is an enzyme different from a and gsecretase. It became clear that the bsecretase cleavage
requires an acidic environment and is localized both in endosomes and in the secretory pathway (Koo and
Squazzo, 1994; Haass et al., 1995). Moreover, it only cleaved membranebound APP, suggesting that
bsecretase itself is a membrane protein. Additionally, bsecretase activity was found in most tissues
analyzed and showed a particularly strong activity in brain. Moreover, it cleaved Swedish mutant APP
much more efficiently than wildtype APP (Citron et al., 1992, 1995). Over the years, different candidate
bsecretases were suggested, but none fulfilled all the criteria for bsecretase.
All three APP secretases were identified within a short time. After the identification of ADAM17 and
ADAM10 as asecretases (Buxbaum et al., 1998; Lammich et al., 1999), the presenilins were shown in early
1999 to be the catalytic subunit of gsecretase (Wolfe et al., 1999). At the end of the same year, bsecretase
was the last secretase to be identified. Vassar and colleagues were the first ones to report the identification of
the bsecretase enzyme, which they named bsite APP cleaving enzyme (BACE1, also called memapsin2 or
Asp2) (Vassar et al., 1999). Four subsequent reports by other groups identified the same enzyme (Hussain
et al., 1999; Sinha et al., 1999; Yan et al., 1999; Lin et al., 2000). The fact that all groups used different
experimental approaches to identify BACE1 made a strong case for BACE1 being bsecretase.
Vassar and colleagues used an expression cloning approach to identify bsecretase from a cDNA library
derived from human embryonic kidney 293 cells (HEK293) (Vassar et al., 1999). As a reporter cell line they
used HEK293 cells stably expressing Swedish mutant APP and transiently transfected them with pools of
cDNAs. Pools that increased Ab secretion into the conditioned medium were further subdivided by the sib
selection approach until they finally identified the individual cDNA that increased Ab generation. This
cDNA turned out to be BACE1.
Alzheimers disease BACE proteases
11 265
Another group used an affinitypurification strategy and identified bsecretase from human brain using
a peptidebased transitionstate analog of the APP sequence (Sinha et al., 1999). This compound inhibits
bsecretase activity and binds tightly to the protein. After several enrichment steps the enzyme was
Nterminally sequenced. The obtained amino acid sequence was then used to clone the cDNA of bsecretase.
Three additional groups hypothesized that bsecretase is an aspartyl protease and applied a genomics
approach searching EST databases to identify BACE1 (Hussain et al., 1999; Yan et al., 1999; Lin et al., 2000).
BACE1 is a type I membrane protein with 501 amino acids (> Figure 11-2). Its gene is encoded on
chromosome 11. The Nterminal signal peptide is followed by a prodomain, a catalytic domain comprising
the two catalytic aspartic acid residues [amino acids 9396 (DSGT) and amino acids 289292 (DTGS)]
characteristic for aspartyl proteases, a transmembrane domain, and a short cytoplasmic tail. Several
experiments show that BACE1 fulfills all criteria for being bsecretase (Hussain et al., 1999; Sinha et al.,
1999; Vassar et al., 1999; Yan et al., 1999; Lin et al., 2000). Mutation of the catalytic active site aspartic acid
. Figure 11-2
Domain structures of BACE1 and BACE2. N and Ctermini (N, C) of both proteins are indicated as well as the
transmembrane domain (TMD) and the cleavage sites of the signal peptide (SP) and the prodomain (pro). Also
shown are the two catalytic aspartic acid residues (D) and the Cterminal sequences of both proteins
residues abolishes its activity, providing evidence that BACE1 is indeed an aspartyl protease. Moreover,
BACE1 acts as a protease in vitro assays and cleaves APP at the correct peptide bond. The catalytic residues
are in the luminal domain of BACE1, and thus have the correct topology for cleaving within the luminal
domain of APP. Overexpression of BACE1 increases bsecretase cleavage of APP and Ab generation.
Conversely, BACE1 knockout mice do not generate Ab anymore (Cai et al., 2001; Luo et al., 2001; Roberds
et al., 2001). BACE1 is ubiquitously expressed, with higher expression levels observed in brain and
particularly in neurons, which is in agreement with the increased bsecretase activity found in this cell
type. Additionally, BACE1 has an acidic pHoptimum of 4.5, and localizes to the Golgi, TGN, and the
endosomes, where bsecretase activity is found. Most important, BACE1 cleaves the Swedish mutant APP
much more efficiently than wildtype APP (Lin et al., 2000; GruningerLeitch et al., 2002). Taken together,
BACE1 fulfills the criteria for being bsecretase.
Shortly after the discovery of BACE1, a homolog was identified by EST database searches and named
BACE2 (also referred to as memapsin1, Asp1, or DRAP) (Saunders et al., 1999; Yan et al., 1999; Acquati
et al., 2000; Lin et al., 2000; Solans et al., 2000). The BACE2 protein contains 518 amino acids, and its
sequence shares 46% identity and 62% similarity with the corresponding sequence of BACE1. Both
proteases are evolutionarily highly conserved. The sequence identity between the human and the mouse
protein is 95% for BACE1 and 89% for BACE2. BACE1 and BACE2 belong to the pepsin family of aspartyl
proteases, and are the only membranebound family members. BACE2 shares the same domain structure as
BACE1 (> Figure 11-2).
Initial important information on the regulation of BACE1/BACE2 expression and APP processing might be
obtained from the BACEs gene structure and alternative splicing events. Because BACE1 is the predominant
bsecretase in brain, and BACE2 is expressed peripherally under basal conditions, most studies performed
so far focused on BACE1 only.
266
11 Alzheimers disease BACE proteases
BACE1 mRNA encoding the fulllength, 501 aa BACE1 protein is expressed at high levels in brain and in
pancreas, but significant concentrations of enzymatically active BACE1 proteins are found exclusively
in brain (Hussain et al., 1999; Sinha et al., 1999). BACE1 mRNAs are synthesized as nine exons and eight
introns from a 30.6 kb region of the human chromosome 11q23.211q23.3 (Saunders et al., 1999) and the
presence of multiple mRNAs that terminate at different sites has been attributed to variable polyadenylation
(Sambamurti et al., 2004). Additionally, fulllength BACE1 mRNA may undergo alternative splicing to give
rise to three truncated BACE1 mRNAs with the deletion of 132, 75, and 207 nucleotides, respectively
(Tanahashi and Tabira, 2001; see also > Figure 11-3). Removal of 132 bp of exon 3 results in the expression
. Figure 11-3
Alternative splicing of BACE1 mRNA. Schematic representation of BACE1 coding exons and alternatively spliced
transcripts. Alternative splicing leads to the removal of 132 (44), 75 (25) or 207 (69) nucleotides (amino acids)
and results in the expression of BACE1457; BACE1476, and BACE1432 isoforms. In the transcript BACE1457
twothirds of exon 3 are spliced, resulting in the removal of two glycosylation sites (arrows) of the
corresponding protein. In the transcript BACE1476, 75 nucleotides of exon 4 are spliced and in the BACE1
432 mRNA both mentioned nucleotide sequences of exon 3 and exon 4 are removed. Asterisks, activesite
motifs; SP, signal peptide sequence; pro, prodomain; TMD, transmembrane domain
of a 457 aa BACE1 protein lacking two glycosylation sites between the two active site motifs (> Figure 11-3).
This BACE1457 mRNA was detected at high levels not only in pancreas (Bodendorf et al., 2001) but also at
much lower concentrations in brain (Tanahashi and Tabira, 2001). The resultant BACE1457 protein is
retained in a proenzymatic and endoglycosidase Hsensitive state in the endoplasmatic reticulum (ER;
Bodendorf et al., 2001; Tanahashi and Tabira, 2001). Similar observations were made for the BACE1476
isoform lacking 75 nucleotides of exon 4 (Tanahasi and Tabira, 2001). Because the BACE1501 isoform is
resistant to endoglycosidase H treatment (Capell et al., 2000; Huse et al., 2000), different posttranslational
modifications of the distinct BACE1 isoforms are discussed (see also > Section 5). Most importantly, the
deletion of 25 or more aa between the two activesite motifs of BACE1 impairs APP processing at the
bsecretase site. This might partly explain the discrepancy between high levels of BACE1 mRNA but low
BACE1 enzymatic activity with regard to APP cleavage in pancreas. However, despite the existence of
truncated BACE1 mRNAs in pancreas, the fulllength BACE1 isoform has been shown by RNAse protection
assays to be the major pancreatic BACE1 transcript (Ehehalt et al., 2002). On the other hand, the presence of
BACE1 isoforms with reduced or lacking bsecretase activity might be interpreted as a first hint for the
subsistence of alternative BACE1 substrates, which are processed by tissuespecific BACE1 isoforms in
distinct organs.
Alzheimers disease BACE proteases
11 267
What is the function of BACE2? Although it cleaves APP more efficiently within the bamyloid
sequence than at the bsecretase site (see also > Figure 11-1), it might play a role in specific forms of AD.
When comparing by database analysis the chromosomal localization of the BACE2 gene in different species
it came to our attention that BACE2 (but not BACE1) and APP map to the same chromosome in mice, rats,
and humans (> Figure 11-4). The localization of the APP gene on human chromosome 21 (Kang et al.,
. Figure 11-4
Chromosomal localization of BACE1, BACE2, and APP in mouse, rat, and human. BACE2, but not BACE1, and its
substrate APP map to the same chromosome in all three species
1987; Tanzi et al., 1987) and the finding that virtually all patients more than the age of 40 suffering Downs
syndrome (DS; trisomy 21) invariantly develop the clinical and pathological characteristics of AD, support
the hypothesis that increased APP expression plays a significant role in the progression of the disease [for
review see Kola and Hertzog (1997)]. In addition to APP, its secretase BACE2 is localized on human
chromosome 21 and maps to the Down critical region in 21q22.3 (Acquati et al., 2000; Solans et al., 2000)
and, therefore, the increased expression of both, substrate and its processing secretase might increase the
generation of Ab peptides and contribute to the pathogenesis of AD. Indeed, the increased secretion of
BACE2 from fibroblasts of patients suffering from DS has been reported (Barbiero et al., 2003). BACE2
overexpression was also observed in the DS fetal brains and in human neural embryonic DS stem cells in
which conditioned media BACE2 was secreted (Barbieri et al., 2003). Moreover, changes in the expression
of BACE2 were observed immunohistochemically in the frontal cortex of DS patients. The immunoreac-
tivity for BACE2 was particularly detected in neurofibrillary tanglebearing neurons from the elderly DS
brains with ADtype neuropathology, but was not detected in those of DS brains without ADtype
neuropathology or in those of control brains of any age (Motonaga et al., 2002). This suggests the possibility
that the elevated expression of BACE2 is involved in the ADtype neuropathology of DS. However, recent
studies using transfection and knockdown strategies revealed that overexpression of BACE2 reduces
amyloidogenic APP processing and increases the generation of nonamyloidogenic APP fragments whereas
elimination of BACE2 produced opposite effects (Sun et al., 2005, 2006a). It has to be established in
transgenic mice how BACE2 functions in brain in vivo. One way to address that question is to overexpress
BACE2 neuronally on the background of hAPPtg/BACE1/ mice and to quantify Ab generation and Ab
plaque formation. An alternative mechanism possibly contributing to AD pathology in DS patients was
revealed recently. DS brains displayed an higher ratio of mature to immature forms of BACE1 in the Golgi,
resulting in increased bsecretase activity (Sun et al., 2006b). This indicates that overproduction of Ab in DS
might be caused by abnormal BACE1 protein trafficking and maturation.
268
11 Alzheimers disease BACE proteases
The transcriptional regulation of BACE2 appears to be quite different from that of BACE1. The BACE2
gene is controlled by a TATAless promoter and although SP1 can regulate BACE1 and BACE2 genes there is
very little similarity between the two promoters as revealed by comparative sequence analysis and tran-
scription factor prediction (Sun et al., 2005). Moreover, these authors identified a neuronspecific repressive
element between 652 bp to 200 bp of the BACE2 promoter. Although it is not known yet under which
conditions the neuronal BACE2 repression can be overruled, such information might be important with
regard to diseaserelated APP processing (see earlier) and a potential upregulation of neuronal BACE2
expression as an AD treatment strategy.
Observations on altered BACE1 protein concentrations in brains of sporadic AD patients (see later) indicate
transcriptional and/or translational regulation of BACE1 expression in brain. The BACE1 promoter is
highly conserved between rat, mouse, and human (> Figure 11-5) and contains a number of putative
transcription factor binding sites (LangeDohna et al., 2003). This would suggest that regulatory mechan-
isms of BACE1 expression are common between these species and that mice and rats represent valuable
animal models to reveal such mechanisms and to test therapeutic strategies aimed at reducing BACE1
expression in AD. Interestingly, the BACE1 promoter contains several putative transcription factor binding
sites that are known to be involved in the expression of other ADrelated proteins. For example, NFkB sites
are present in the promoters of APP (Grilli et al., 1995), choline acetyltransferase (ChAT; ToliverKinsky
et al., 2000) and presenilins [for review see Glasgow and PerezPolo (2000)]. Likewise, SP1 sites have been
reported for the APP (Lahiri and Robakis, 1991; Pollwein, 1993; Lukiw et al., 1994; Hoffman and Chernak,
1995), ChAT (Hahn et al., 1992; Inoue et al., 1993), presenilin1 (Mitsuda et al., 1997), and presenilin
2 (Pennypacker et al., 1998) promoters. Based on results derived from luciferase reporter assays it was
suggested that an upstream NFkB site might act as a repressor of BACE1 transcription (LangeDohna et al.,
2003). A detailed analysis using BACE1 promoter constructs carrying mutations of the NFkB site revealed
a unique cell typespecific regulation of BACE1 promoter activity. In neuronlike cells, such as retinoic acid
differentiated SHSY5Y cells or nerve growth factordifferentiated PC12 cells, the NFkB site indeed acts
as repressor of the BACE1 promoter whereas in C6 glioma cells it acts as activator (Bourne et al., 2005). This
has important therapeutic implications and might explain in part the cell typespecific upregulation BACE1
in reactive astrocytes of APP transgenic mice with amyloid plaque formation (Roner et al., 2001),
in chronic lesion paradigms (HartlageRubsamen et al., 2003) and in brains of AD patients (Hartlage
Rubsamen et al., 2003; Leuba et al., 2005). These results are also consistent with observations demonstrating
that nonsteroidal antiinflammatory drugs (NSAIDs) decreased the BACE1 mRNA levels, protein expres-
sion, and also BACE1 enzymatic activity (Sastre et al., 2003). Pharmacological studies suggested that this
effect is mediated through activation of the peroxisome proliferatoractivated receptorg (PPARg; Sastre
et al., 2003). In line with that, the absence of PPARg potentiates BACE1 mRNA levels by increasing BACE1
promoter activity. Conversely, overexpression of PPARg as well as NSAIDs and pioglitazone, a PPARg
activator, reduced BACE1 promoter activity (Sastre et al., 2006). These effects appear to be mediated by a
PPARg responsive element (PPRE) within the BACE1 promoter. Mutagenesis of this PPRE element
abolished the binding of PPARg to the PPRE and increased BACE1 gene promoter activity (Sastre et al.,
2006). Furthermore, proinflammatory cytokines decreased the transcription of the PPARg gene, an effect
that was suppressed by NSAIDs. Interestingly, brain extracts from AD patients showed decreased PPARg
expression and reduced binding to PPRE in the BACE1 promoter. From these results, it was concluded that
the protective mechanism of NSAIDs in AD is exerted through activation of PPARg and decreased BACE1
gene transcription (Sastre et al., 2006). This conclusion is supported by data from a transgenic mouse
model of AD, in which treatment with NSAIDs or pioglitazone suppressed proinflammatory markers,
reduced BACE1 mRNA and protein levels and lowered Ab plaque load (Heneka et al., 2005). Transcrip-
tional regulation of BACE1 expression occurs also via a functional SP1 element 850 nucleotides upstream
of the transcription start site (Christensen et al., 2004). In HEK293T cells and in PC12 cells this SP1
site appears to act as an activator of BACE1 expression as shown by reduced BACE1 expression after
Alzheimers disease BACE proteases
11 269
. Figure 11-5
Rat, mouse, and human BACE1 promoter sequences. Comparison of the rat BACE1 promoter sequence with that
of its mouse and human homologs. There is 92% similarity between rat and mouse and 81% similarity between
rat and human within the region from bp 1 to 600. The similarity between rat and mouse by about 86%
extends between bp 600 and 1000, but the similarity between the rat and the human sequence is only 39%
in this fragment. At a greater distance than bp 1000, this similarity is reduced significantly (after LangeDohna
et al., 2003; modified)
270
11 Alzheimers disease BACE proteases
targeted disruption of the SP1 gene and by the potentiated BACE1 expression after overexpression of SP1
(Christensen et al., 2004).
In addition to a possible regulation of BACE1 expression by transcriptional mechanisms, there is ample
evidence for posttranscriptional modulation of BACE1 enzymatic activity. For example, BACE1 protein
levels are significantly upregulated up to 2.7fold in the brains of sporadic AD patients compared with
nonAD controls (Fukumoto et al., 2002; Holsinger et al., 2002; Yang et al., 2003), whereas mRNA levels were
unchanged (Yasojima et al., 2001; Holsinger et al., 2002; Preece et al., 2003). This indicates that posttran-
scriptional mechanisms, such as translational control or altered BACE1 protein degradation, regulate
BACE1 protein levels. In fact, tissue culture experiments have provided evidence for both mechanisms in
controlling BACE1 protein levels. An altered BACE1 degradation has been described by Kovacs and
coworkers (Puglielli et al., 2003). They reported that the lipid second messenger ceramide, which is elevated
in the brain of AD patients, increases the halflife of BACE1 and thereby increases Ab generation (Puglielli
et al., 2003). Translational control of BACE1 expression has been observed independently by different
groups (De Pietri Tonelli et al., 2004; Lammich et al., 2004; Rogers et al., 2004). The studies were based on
the observation that the 50 untranslated region (50 UTR) of the BACE1 mRNA contains unusual sequence
features that are often found in mRNAs showing tight translational control. In most vertebrate mRNAs, the
50 UTR is 10200 nucleotides long, is unstructured, not very GCrich and does not contain upstream open
reading frames (uORFs). In contrast, less than 10% of vertebrate genes, including many regulatory proteins
like protooncogenes, have 50 UTRs that are longer than 200 nucleotides and are often GCrich (7090%).
This indicates a high degree of secondary structure, which may impede the efficient scanning of the
ribosome (Kozak, 1987; Willis, 1999). Additionally, these long 50 UTRs may contain uORFs [for a review
see Clemens and Bommer (1999)] that can inhibit the translation of the main ORF. The 50 UTR of BACE1,
which is highly conserved in the human, rat, and murine sequence, belongs to the class of the long 50 UTRs
similar to many regulatory genes. It is 446 nucleotides long, has a GCcontent of 77%, and contains 3 uORFs
(> Figure 11-6). Importantly, all three different BACE1 mRNA species detected by Northern Blot analysis
. Figure 11-6
The mRNA domain structure of BACE1. The BACE1 mRNA consists of the 50 untranslated region (50 UTR),
followed by the coding region (open reading frame with the start codon AUG) and the 30 UTR. The 50 UTR has
the indicated high GCcontent and contains three upstream open reading frames (uORFs), represented as boxes
contain the 50 UTR (Lammich et al., 2004). The role of the 50 UTR in a possible translational control of
BACE1 mRNA was tested in different cell lines as well as in vitro translation experiments (De Pietri Tonelli
et al., 2004; Lammich et al., 2004; Rogers et al., 2004). The use of BACE1 constructs lacking part or all the
50 UTR revealed that the presence of the 50 UTR decreased BACE1 translation by up to 90%. Moreover, when
fused to luciferase, the 50 UTR of BACE1 strongly suppressed luciferase translation and activity, revealing
that the translation repressing activity must be encoded within the 50 UTR. The uORFs within the 50 UTR
were not the major reason for the inhibition of translation. Instead, the experiments suggest that due to its
high GCcontent the 50 UTR forms a constitutive translation barrier that prevents the ribosomes from
efficiently translating the BACE1 mRNA. Additionally, Zacchetti and colleagues showed that the 50 UTR
dependent translational repression may be alleviated in activated astrocytes (De Pietri Tonelli et al., 2004),
leading to an increased expression of BACE1 protein. In fact, activated astrocytes expressing elevated levels
of BACE1 are found around amyloid plaques in an animal model of AD and in the AD brain (Roner et al.,
Alzheimers disease BACE proteases
11 271
2001; HartlageRubsamen et al., 2003) (see also > Section 6). Thus, pathophysiological conditions leading
to the activation of astrocytes may increase BACE1 expression and thereby Ab generation and could
exacerbate AD pathogenesis.
A variety of cell biological studies have provided important insights into the posttranslational modifications
of BACE1 and the control of its intracellular transport. Like in the studies addressing transcriptional and
translational control of BACE1 expression (> Section 4), an elucidation of BACE1 modifications or
transport may allow to find new therapeutic approaches for AD. They may be used to lower BACE1 activity
or shift the APP cleavage away from the Ab generating bsecretase toward an increased asecretase cleavage
that prevents Ab generation.
within the ectodomain of BACE1 has been described between residues Leu228 and Ala299. This leads to two
BACE1 fragments, which remain bound to each other and which retain BACE1 activity. This cleavage was
mainly observed in liver, pancreas, and muscle, but not in brain (Huse et al., 2003).
Like APP, mature BACE1 cycles between the TGN, the plasma membrane, and the endocytic pathway,
particularly early and late endosomes (Huse et al., 2000; Walter et al., 2001). In fact, APP and BACE
colocalize at the plasma membrane and in endosomes (Kinoshita et al., 2003). BACE2 shows a similar
cellular distribution as BACE1, but a larger proportion of BACE2 is found at the plasma membrane
(Hussain et al., 2000; Yan et al., 2001). Close to its Cterminus BACE1 contains an evolutionarily conserved
dileucine amino acid motif (> Figure 11-2), which conforms to endocytic trafficking signals found in other
proteins, such as mannose6 phosphate receptor (Johnson and Kornfeld, 1992). Deletion of this motif alters
BACE1 trafficking, such that more BACE1 is localized at the cell surface and less is found in endosomes
(Huse et al., 2000).
Phosphorylation of BACE1 by casein kinase 1 at Ser498 seems to be another way of regulating BACE1
sorting in the cell. Replacing Ser498 by aspartic acid (Ser498Asp), which mimics phosphorylated BACE1 is
predominantly found in the TGN and in late endosomes, whereas an alanine at position 498 (Ser498Ala)
mimics nonphosphorylated BACE1 and is mainly localized in early endosomes (Walter et al., 2001;
Pastorino et al., 2002). Despite the differences in cellular sorting, both BACE1 mutants have similar effects
on APP processing when transfected into HEK293 cells, revealing that the phosphorylation of BACE1 does
not have a major effect on its catalytic activity (Pastorino et al., 2002).
In cultured cells, overexpressed APP and BACE1 have been detected partly within and partly outside
cholesterolrich membrane domains, socalled detergentresistant membrane microdomains or rafts. The
question whether BACE1 cleavage of APP occurs within or outside the cholesterolrich domains is of
particular interest, since cholesterol lowering drugs have been shown to reduce Ab generation. In contrast
to initial studies suggesting that APP cleavage by BACE1 occurs within the detergentresistant membranes
(Riddell et al., 2001; Tun et al., 2002; Cordy et al., 2003; Ehehalt et al., 2003), a recent report suggests that
this cleavage event takes place outside the detergentresistant membranes, at least under endogenous
expression levels (AbadRodriguez et al., 2004). Further studies are needed to fully understand the role of
cholesterol and cholesterolrich membrane domains in BACE1 activity and AD pathogenesis.
suggesting that the GGABACE1 interaction is required for the recycling of endocytosed BACE1 to the cell
surface (He et al., 2005).
Another binding partner of BACE1 that was identified in a yeast two hybrid screen is the brain specific
type II membrane protein BRI3, which colocalizes with BACE1 in HEK293 cells but also in brain (Wickham
et al., 2005). Moreover, BRI3 coimmunoprecipitates with endogenous BACE1 in HEK293 cells. The
maximal interaction between both proteins requires the cytoplasmic domain of BACE1 to be palmitoylated.
The functional consequences of this interaction and the function of BRI3 itself are as yet unknown.
Given the link of copper to the pathogenesis of AD [for an overview see Multhaup et al. (2002),
Cuajungco et al. (2005)], the identification of the copper chaperone for superoxide dismutase1 (CCS) is of
particular interest. CCS binds to the cytoplasmic tail of BACE in rat brain extracts, and both proteins are
cotransported through axons. Surprisingly, the cysteine residues in the cytoplasmic domain of BACE1 bind
a single Cu(I) atom (Angeletti et al., 2005). Whether the binding of copper or CCS to BACE1 has an effect
on APP processing remains to be studied.
Proteins that alter BACE1 activity by binding to BACE1 are reticulon 3 (RTN3) and its homolog RTN4b
(NoGo) and heparan sulfate. Heparan sulfate was found to be a natural extracellular ligand for BACE1
(Scholefield et al., 2003). This interaction inhibits BACE1 cleavage of APP but not the asecretase cleavage
of APP. The underlying molecular mechanism is not yet known. Heparan sulfate and BACE1 colocalize in
the Golgi and at the plasma membrane and coimmunoprecipitate at endogenous expression levels in HEK
293 cells.
RTN3 binds BACE1 in neurons, whereas RTN4b binds BACE1 in oligodendrocytes. The interaction of
BACE1 with RTN3 was not only found in cells overexpressing both proteins but also at endogenous expression
levels in brain. Most important, increasing the expression of RTN3 inhibited the secretion of Ab, whereas
lowering the expression of RTN3 by RNA interference increased the secretion of Ab, suggesting that reticulon
proteins are negative modulators of BACE1 by blocking access of BACE1 to APP (He et al., 2004).
Yet another binding partner of BACE1 is BACE1 itself (Schmechel et al., 2003; Westmeyer et al., 2004).
The homodimerization of BACE1 was described to occur already in the ER and to increase BACE1
proteolytic activity (Westmeyer et al., 2004). The dimerization takes place between the ectodomains and
requires BACE1 to be membraneanchored. Homodimerization of BACE1 was observed both in cultured
cells overexpressing BACE1 and in brain tissue.
Taken together as shown in > Table 111, several BACE1 binding proteins have been identified. Some of
them seem to influence BACE1 intracellular trafficking (PLSCR 1, GGAs), whereas other ones affect BACE1
. Table 11-1
Posttranslational modifications and binding partners of BACE1
Posttranslational modifications of BACE1 Binding partners of BACE1
Removal of signal peptide in the ER gsecretase (presenilin and nicastrin)
Nglycosylation at four sites Phospholipid scramblase 1 (PLSCR 1)
Removal of prodomain GGA (Golgilocalized gearcontaining ARF binding)
Three disulfide bridges BRI3
Ectodomain shedding Copper chaperone for superoxide dismutase 1 (CCS)
Palmitoylation Heparan sulfate
Phosphorylation Reticulon 3 and 4b (NoGo)
Sulfated at mature Nsugars BACE1 (homodimer formation)
Proteolytic cleavage within its ectodomain in liver,
pancreas and muscle
activity (heparan sulfate, RTNs, BACE1 homodimerization). Yet another group consists of proteins with
poorly understood function with respect to their binding to BACE1 (gsecretase subunits nicastrin and
presenilin, CCS, BRI3).
274
11 Alzheimers disease BACE proteases
6.2 Aging
During aging BACE1 mRNA (Bigl et al., 2000; Irizarry et al., 2001; Apelt et al., 2004) and protein levels
(Roner et al., 2001; Apelt et al., 2004; Fukumoto et al., 2004) are stable in brains of mice and humans.
BACE1 enzymatic activity, in contrast, is increased with aging in mouse, monkey, and human brain
(Fukumoto et al., 2004). In the light of aging being the most stringent risk factor for developing AD and
the concentrations of Ab peptides in brain strongly correlating with age, this is an important observation
and indirect evidence for the contribution of posttranslational mechanisms in the regulation of BACE1
activity and the pathogenesis of AD (see also > Section 5).
hand, it is well established that a quantitative shift APP processing toward the generation of Ab peptides
interferes with cholinergic neurotransmission. For example, solubilized Ab peptides strongly inhibit the
potassiumstimulated release of acetylcholine from hippocampal slices (Kar et al., 1996) and decreases
ChAT activity but not acetylcholinesterase activity in the cholinergic SN56 cell line (Pedersen et al., 1996).
Ab peptides also decrease the intracellular acetylcholine concentration (Hoshi et al., 1997) and impair
M1associated signaling (Kelly et al., 1996) in primary septal or cortical cultures. All these effects of Ab on
markers of cholinergic neurotransmission were observed at pM to nM concentrations and without obvious
evidence of neurotoxicity. The effects of Ab on reduction of pyruvate dehydroxygenase activity (Hoshi et al.,
1997), the key enzyme for the generation of acetylcholine used for neurotransmitter synthesis and citrate
cycle, might explain both, cholinergic hypoactivity and metabolic dysfunction of cholinergic neurons after
exposure to Ab at low concentrations.
. Figure 11-7
Astrocytic BACE1 expression in Tg2576 mice and AD. Confocal images of double immunofluorescence labeling
for BACE1 and GFAP in the parietal cortex of 17monthold Tg2576 mice and in Area 22 of the AD cortex. BACE1
immunoreactivity is shown by the green fluorescence; GFAPimmunoreactivity is encoded by red fluorescence.
In the overlay channel, colocalization of BACE1 and GFAP is indicated by yellow/orange color. Reactive astro-
cytes expressing BACE1 are indicated by arrowheads; the asterisks mark the position of Ab plaques (after
HartlageRubsamen et al., 2003; modified)
276
11 Alzheimers disease BACE proteases
LPS accelerates Ab deposition in APPV717F transgenic mice (Qiao et al., 2001). However, at the present
stage it is not clear to what extent reactive astrocytes contribute to the production of Ab peptides and to Ab
plaque formation. First, although activity measurements indicate that reactive astrocytes do express
enzymatically active BACE1 (Roner et al., 2005), it still has to be shown that the fulllength, nontruncated
BACE1 isoform is expressed. Second, to contribute to Ab plaque formation, the substrate APP has to be
available to astrocytic BACE1. It is known that primary astrocytes themselves do express APP and generate
significant amounts of Ab peptides (Gray and Patel, 1993; Amara et al., 1999; Beck et al., 2000; Blasko et al.,
2000; Docagne et al., 2004). Additionally, APP is also expressed by reactive astrocytes in experimental
models of chronic gliosis [see, e.g., Martins et al. (2001)] and this induced astrocytic APP expression results
in the increased generation of Ab peptides and BACE1 cleavagederived A4CT fragments (Bates et al., 2002;
Lesne et al., 2003). Alternatively, astrocytic BACE1 might be secreted and cleave membranous APP at the
neuronal surface. Such a shedding of BACE1 has been reported recently (Benjannet et al., 2001; Hussain
et al., 2003).
contributes to the rolling of leukocytes along the vessel wall before they transmigrate through the
endothelium. The type II membrane protein ST6Gal I is involved in B cell expansion and is required for
the correct glycosylation of CD22 ligands. VGSCs are an abundant type of ion channels that are responsible
for the initiation and propagation of action potentials in neurons. ST6Gal I, PSGL1, and VGSCb meet
three basic criteria for BACE1 substrates: (a) their cleavage occurs under conditions of endogenous BACE1
expression, (b) exogenous expression of BACE1 strongly increases their cleavage and, importantly, (c) their
cleavage is reduced or abolished in BACE1/ cells or animals (Lichtenthaler et al., 2003; Kitazume et al.,
2005; Wong et al., 2005), which establishes the physiological relevance of their cleavage. Interestingly, both
ST6Gal I and PSGL1 are cleaved Cterminally of a leucine residue, which is in good agreement with the in
vitro substrate specificity of BACE1 and with the fact that BACE1 cleavage of the Swedish mutant APP also
occurs after a leucine residue. Thus, additional substrates for BACE1 to be identified may also be cleaved
preferentially after a leucine residue. The exact cleavage sites of BACE1 in APLP1 and APLP2 remain to be
determined.
BACE1 is not only able to cleave type I membrane proteins (APP, PSGL1, APLP1, and APLP2) but also
a type II membrane protein (ST6Gal I), suggesting that, similar to the ADAM proteases, BACE1 may
contribute to the ectodomain shedding of a larger number of membrane proteins of type I or type II
membrane orientation. However, BACE1 is likely to have a more restricted set of substrate proteins, because
it is not involved in the shedding of two other ADAM protease substrates, the TNFa receptor 2 and the cell
adhesion protein Lselectin (Lichtenthaler et al., 2003). The proteolytic processing of APP and PSGL1
reveals additional insights into cleavage preferences of BACE1. APP and PSGL1 are both cleaved by BACE1,
but also by an ADAM metalloprotease. For both proteins most of the cleavage seems to occur by the ADAM
proteases, whereas BACE1 cleavage constitutes only a smaller fraction of the total cleavage. An additional
similarity is the localization of the BACE1 cleavage site in the membraneproximal domain. In contrast to
APP, it is the BACE1 cleavage site in PSGL1 that is located more closely to the membrane than the putative
ADAM protease cleavage site (Lichtenthaler et al., 2003). This is in opposite order compared with the
proteolytic processing of APP. However, this finding agrees well with the known properties of BACE1 and
ADAM proteases. The ADAM protease cleavage of APP occurs without a strict sequence specificity but at a
fixed distance from the membrane (Sisodia, 1992), whereas the BACE1 cleavage seems to be more sequence
specific but not strongly dependent on the distance of the cleavage site from the membrane (Citron et al., 1995).
Taken together, it is clear now, that BACE1 does not exclusively cleave APP, but has a wider substrate
spectrum. Much less is known about BACE2 substrates. It is not known whether the three BACE1 substrates
ST6Gal I, APLP1, and APLP2 can be cleaved by BACE2, but BACE2 clearly cleaves two other BACE1
substrates, namely APP and PSGL1. This raises the possibility, that BACE1 and BACE2 may have similar
or identical substrates. However, both proteases do not necessarily cleave their substrates at identical peptide
bonds. For example, BACE1 cleaves APP at the Nterminus of Ab, whereas BACE2 cleavage occurs within the
Ab domain and thus functions as an alternative asecretase, precluding Ab formation. However, BACE2
cleavage of APP may not reduce Ab levels in the brain, given the low expression level of BACE2 in this organ.
At present, the specific function of BACE1 in the cleavage of its substrates remains unclear, but different
possibilities are plausible. Given that several BACE1 substrates are also substrates for ADAM metallopro-
teases, it is possible, that both proteases cleave the substrates for different purposes. BACE1 seems to be
constitutively active, whereas ADAM proteases are tightly regulated in their activity by different cellular
signal transduction pathways (Allinson et al., 2003). Therefore, BACE1 cleavage, which partly occurs in
the secretory pathway, could provide a lowlevel constitutive cleavage and secretion of membrane
proteins. In contrast, under conditions where a high level of secretion is needed (e.g., of a growth factor
or cytokine) the activity of ADAM proteases might be upregulated. Another function of BACE1 could be
the initiation of substrate protein degradation. In this scenario, BACE1 would cleave off most of the
ectodomain of its substrate and leave the remaining membranebound stub for additional degradation by
the gsecretase omplex. This idea is in agreement with the finding that BACE1 cleavage occurs after APP
endocytosis in the endosomes and that the resulting secreted ectodomain of APP (sAPPb) does not have
any known biological functions. In contrast, sAPPa has neuroprotective and neurotrophic properties
(Furukawa et al., 1996; Meziane et al., 1998) and is generated by ADAM proteases at or close to the
plasma membrane.
278
11 Alzheimers disease BACE proteases
References
AbadRodriguez J, Ledesma MD, Craessaerts K, Perga S, copper chaperone for superoxide dismutase1 and binds
Medina M, et al. 2004. Neuronal membrane cholesterol copper. J Biol Chem 280: 17930-17937.
loss enhances amyloid peptide generation. J Cell Biol 167: Apelt J, Bigl M, Wunderlich P, Schliebs R. 2004. Agingrelated
953-960. increase in oxidative stress correlates with developmental
Acquati F, Accarino M, Nucci C, Fumagalli P, Jovine L, et al. pattern of betasecretase activity and betaamyloid plaque
2000. The gene encoding DRAP (BACE2), a glycosylated formation in transgenic Tg2576 mice with Alzheimerlike
transmembrane protein of the aspartic protease family, pathology. Int J Dev Neurosci 22: 475-84.
maps to the down critical region. FEBS Lett 468: 59-64. Barbiero L, Benussi L, Ghidoni R, Alberici A, Russo C, et al.
Allinson TM, Parkin ET, Turner AJ, Hooper NM. 2003. 2003. BACE2 is overexpressed in Downs syndrome. Exp
ADAMs family members as amyloid precursor protein Neurol 182: 335-345.
alphasecretases. J Neurosci Res 74: 342-352. Bates KA, Fonte J, Robertson TA, Martins RN, Harvey AR.
Amara FM, Junaid A, Clough RR, Liang B. 1999. TGF-beta(1), 2002. Chronic gliosis triggers Alzheimers diseaselike pro-
regulation of alzheimer amyloid precursor protein mRNA cessing of amyloid precursor protein. Neuroscience 113:
expression in a normal human astrocyte cell line: mRNA 785-796.
stabilization. Mol Brain Res 71: 42-49. Beck M, Bruckner MK, Holzer M, Kaap S, Pannicke T, et al.
Angeletti B, Waldron KJ, Freeman KB, Bawagan H, Hussain I, 2000. Guineapig primary cell cultures provide a model
et al. 2005. BACE1 cytoplasmic domain interacts with the to study expression and amyloidogenic processing of
Alzheimers disease BACE proteases
11 279
endogenous amyloid precursor protein. Neuroscience 95: Christensen MA, Zhou W, Qing H, Lehman A, Philipsen S,
243-254. et al. 2004. Transcriptional regulation of BACE1, the beta
Benjannet S, Elagoz A, Wickham L, Mamarbachi M, amyloid precursor protein betasecretase, by SP1. Mol Cell
Munzer JS, et al. 2001. Posttranslational processing of Biol 24: 865-874.
betasecretase (betaamyloidconverting enzyme) and its Citron M. 2004. betasecretase inhibition for the treatment of
ectodomain shedding. The pro and transmembrane/ Alzheimers diseasepromise and challenge. Trends Phar-
cytosolic domains affect its cellular activity and amyloid macol Sci 25: 92-97.
beta production. J Biol Chem 276: 10879-10887. Citron M, Teplow DB, Selkoe DJ. 1995. Generation of amyloid
Bennett BD, Denis P, Haniu M, Teplow DB, Kahn S, et al. beta protein from its precursor is sequence specific. Neuron
2000. A furinlike convertase mediates propeptide cleavage 14, 661-670.
of BACE, the Alzheimers betasecretase. J Biol Chem 275: Citron M, Oltersdorf T, Haass C, McConlogue L, Hung AY,
37712-37717. et al. 1992. Mutation of the bamyloid precursor protein in
Bigl M, Apelt J, Luschekina EA, LangeDohna C, Roner S, familial Alzheimers disease increases bprotein produc-
et al. 2000. Expression of betasecretase mRNA in transgen- tion. Nature 360: 672-674.
ic Tg2576 mouse brain with Alzheimer plaque pathology. Clemens MJ, Bommer UA. 1999. Translational control: The
Neurosci Lett 292: 107-110. cancer connection. Int. J Biochem Cell Biol 31: 1-23.
Blasko I, Veerhuis R, StampferKountchev M, SaurweinTeissl Cordy JM, Hussain I, Dingwall C, Hooper NM, Turner AJ.
M, Eikelenboom P, et al. 2000. Costimulatory effects of 2003. Exclusively targeting betasecretase to lipid rafts by
interferongamma and interleukin1beta or tumor necrosis GPIanchor addition upregulates betasite processing of
factor alpha on the synthesis of Abeta140 and Abeta142 the amyloid precursor protein. Proc Natl Acad Sci USA
by human astrocytes. Neurobiol Dis 7: 682-689. 100: 11735-11740.
Blobel CP. 2002. Functional and biochemical characterization Cuajungco MP, Frederickson CJ, Bush AI. 2005. Amyloidbeta
of ADAMs and their predicted role in protein ectodomain metal interaction and metal chelation. Subcell Biochem 38:
shedding. Inflamm Res 51: 83-84. 235-254.
Bodendorf U, Fischer F, Bodian D, Multhaup G, Paganetti P. De Pietri Tonelli D, Mihailovich M, Di Cesare A, Codazzi F,
2001. A splice variant of betasecretase deficient in the Grohovaz F, et al. 2004. Translational regulation of BACE1
amyloidogenic processing of the amyloid precursor pro- expression in neuronal and nonneuronal cells. Nucl Acids
tein. J Biol Chem 276: 12019-12023. Res 32: 1808-1817.
Bourne KZ, LangeDohna C, Roner S, PerezPolo JR. 2006. De Strooper B, Annaert W. 2000. Proteolytic processing and
Rat BACE1 promoter nuclear factorkB binding character- cell biological functions of the amyloid precursor protein.
ization. J Neurosci Res, in press. J Cell Sci 113: 1857-1870.
Buxbaum JD, Liu KN, Luo Y, Slack JL, Stocking KL, et al. De Strooper B, Craessaerts K, Dewachter I, Moechars D,
1998. Evidence that tumor necrosis factor alpha converting Greenberg B, et al. 1995. Basolateral secretion of amyloid
enzyme is involved in regulated alphasecretase cleavage of precursor protein in MadinDarby canine kidney cells is
the Alzheimer amyloid protein precursor. J Biol Chem 273: disturbed by alterations of intracellular pH and by intro-
27765-27767. ducing a mutation associated with familial Alzheimers
Cai H, Wang Y, McCarthy D, Wen H, Borchelt DR, et al. 2001. disease. J Biol Chem 270: 4058-4065.
BACE1 is the major betasecretase for generation of Abeta De Strooper B, Saftig P, Craessaerts K, Vanderstichele H,
peptides by neurons. Nat Neurosci 4: 233-234. Guhde G, et al. 1998. Deficiency of presenilin1 inhibits
Cao X, Sudhof TC. 2001. A transcriptively active complex of the normal cleavage of amyloid precursor protein. Nature
app with Fe65 and histone acetyltransferase Tip60. Science 391: 387-390.
293: 115-120. DiCarlo G, Wilcock D, Henderson D, Gordon M, Morgan D.
Capell A, Meyn L, Fluhrer R, Teplow DB, Walter J, et al. 2002. 2001. Intrahippocampal LPS injections reduce Abeta
Apical sorting of betasecretase limits amyloid betapeptide load in APP PS1 transgenic mice. Neurobiol Aging 22:
production. J Biol Chem 277: 5637-5643. 1007-1012.
Capell A, Steiner H, Willem M, Kaiser H, Meyer C, et al. 2000. Docagne F, Gabriel C, Lebeurrier N, Lesne S, Hommet Y, et al.
Maturation and propeptide cleavage of betasecretase. 2004. SP1 and Smad transcription factors cooperate to
J Biol Chem 275: 30849-30854. mediate TGFbetadependent activation of the Amyloid
Charlwood J, Dingwall C, Matico R, Hussain I, Johanson K, beta precursor protein gene transcription. Biochem J 383:
et al. 2001. Characterization of the glycosylation profiles of 393-399.
Alzheimers betasecretase protein Asp2 expressed in a Ehehalt R, Keller P, Haass C, Thiele C, Simons K. 2003.
variety of cell lines. J Biol Chem 276: 16739-16748. Amyloidogenic processing of the Alzheimer betaamyloid
280
11 Alzheimers disease BACE proteases
precursor protein depends on lipid rafts. J Cell Biol 160: Haass C, Hung AY, Schlossmacher MG, Oltersdorf T, Teplow
113-123. DB, et al. 1993. bAmyloid peptide and a 3kDa fragment
Ehehalt R, Michel B, De Pietri Tonelli D, Zacchetti D, Simons are derived by distinct cellular mechanisms. J Biol Chem
K, et al. 2002. Splice variants of the betasite APPcleaving 268: 3021-3024.
enzyme BACE1 in human brain and pancreas. Biochem Haass C, Lemere CA, Capell A, Citron M, Seubert P, et al.
Biophys Res Commun 293: 30-37. 1995. The Swedish mutation causes earlyonset Alzheimers
Farzan M, Schnitzler CE, Vasilieva N, Leung D, Choe H. 2000. disease by betasecretase cleavage within the secretory path-
BACE2, a betasecretase homolog, cleaves at the beta site way. Nat Med 1: 1291-1296.
and within the amyloidbeta region of the amyloidbeta Hahn M, Hahn SL, Stone DM, Joh TH. 1992. Cloning of the
precursor protein. Proc Natl Acad Sci USA 97: 9712-9717. rat gene encoding choline acetyltransferase, a cholinergic
Fukumoto H, Cheung BS, Hyman BT, Irizarry MC. neuronspecific marker. Proc Natl Acad Sci USA 89:
2002. Betasecretase protein and activity are increased 4387-4391.
in the neocortex in Alzheimer disease. Arch Neurol 59: Haniu M, Denis P, Young Y, Mendiaz EA, Fuller J, et al. 2000.
1381-1389. Characterization of Alzheimers betasecretase protein
Fukumoto H, Rosene DL, Moss MB, Raju S, Hyman BT, et al. BACE. A pepsin family member with unusual properties.
2004. Betasecretase activity increases with aging in human, J Biol Chem 275: 21099-21106.
monkey, and mouse brain. Am J Pathol 164: 719-725. Hardy J, Selkoe DJ. 2002. The amyloid hypothesis of Alzhei-
Furukawa K, Sopher BL, Rydel RE, Begley JG, Pham DG, et al. mers disease: Progress and problems on the road to thera-
1996. Increased activityregulating and neuroprotective peutics. Science 297: 353-356.
efficacy of alphasecretasederived secreted amyloid pre- Harrison SM, Harper AJ, Hawkins J, Duddy G, Grau E, et al.
cursor protein conferred by a Cterminal heparinbinding 2003. BACE1 (betasecretase) transgenic and knockout
domain. J Neurochem 67: 1882-1896. mice: Identification of neurochemical deficits and behav-
Gau JT, Steinhilb ML, Kao TC, DAmato CJ, Gaut JR, ioral changes. Mol Cell Neurosci 24: 646-655.
et al. 2002. Stable betasecretase activity and presynaptic HartlageRubsamen M, Zeitschel U, Apelt J, Gartner U,
cholinergic markers during progressive central nervous Franke H, et al. 2003. Astrocytic expression of the Alzhei-
system amyloidogenesis in Tg2576 mice. Am J Pathol 160: mers disease bsecretase (BACE) is stimulusdependent.
731-738. Glia 41: 169-179.
Glasgow J, PerezPolo JR. 2000. One path to cell death in the Hartmann D, de Strooper B, Serneels L, Craessaerts K,
nervous system. Neurochem Res 25: 1373-1383. Herreman A, et al. 2002. The disintegrin/metalloprotease
Glenner GG, Wong CW. 1984. Alzheimers disease: Initial ADAM10 is essential for Notch signalling but not for alpha
report of the purification and characterization of a novel secretase activity in fibroblasts. Hum Mol Genet 11:
cerebrovascular amyloid protein. Biochem Biophys Res 2615-2624.
Commun 120: 885-890. Hattori C, Asai M, Oma Y, Kino Y, Sasagawa N, et al. 2002.
Gray CW, Patel AJ. 1993. Regulation of betaamyloid precur- BACE1 interacts with nicastrin. Biochem Biophys Res
sor protein isoform mRNAs by transforming growth Commun 293: 1228-1232.
factorbeta 1 and interleukin1 beta in astrocytes. Mol He X, Chang WP, Koelsch G, Tang J. 2002. Memapsin 2 (beta
Brain Res 19: 251-256. secretase) cytosolic domain binds to the VHS domains of
Grilli M, Ribola M, Alberici A, Valerio A, Memo M, et al. GGA1 and GGA2: Implications on the endocytosis mecha-
1995. Identification and characterization of a kappa B/Rel nism of memapsin 2. FEBS Lett 524: 183-187.
binding site in the regulatory region of the amyloid precur- He X, Li F, Chang WP, Tang J. 2005. GGA proteins mediate
sor protein gene. J Biol Chem 270: 26774-26777. the recycling pathway of memapsin 2 (BACE). J Biol Chem
GruningerLeitch F, Schlatter D, Kung E, Nelbock P, Dobeli H. 280: 11696-11703.
2002. Substrate and inhibitor profile of BACE (beta He W, Lu Y, Qahwash I, Hu XY, Chang A, et al. 2004.
secretase) and comparison with other mammalian aspartic Reticulon family members modulate BACE1 activity and
proteases. J Biol Chem 277: 4687-4693. amyloidbeta peptide generation. Nat Med 10: 959-965.
Haass C. 2004. Take fiveBACE and the gammasecretase He X, Zhu G, Koelsch G, Rodgers KK, Zhang XC, et al. 2003.
quartet conduct Alzheimers amyloid betapeptide genera- Biochemical and structural characterization of the inter-
tion. EMBO J 23: 483-488. action of memapsin 2 (betasecretase) cytosolic domain
Haass C, Koo EH, Teplow DB, Selkoe DJ. 1994. Polarized with the VHS domain of GGA proteins. Biochemistry 42:
secretion of betaamyloid precursor protein and amyloid 12174-12180.
betapeptide in MDCK cells. Proc Natl Acad Sci USA 91: Hebert SS, Bourdages V, Godin C, Ferland M, Carreau M,
1564-1568. et al. 2003. Presenilin1 interacts directly with the betasite
Alzheimers disease BACE proteases
11 281
amyloid protein precursor cleaving enzyme (BACE1). Neu- Hussain I, Powell DJ, Howlett DR, Chapman GA, Gilmour L,
robiol Dis 13: 238-245. et al. 2000. ASP1 (BACE2) cleaves the amyloid precursor
Heneka MT, Sastre M, DumitrescuOzimek L, Hanke A, protein at the betasecretase site. Mol Cell Neurosci 16:
Dewachter I, et al. 2005. Acute treatment with the PPAR 609-619.
{gamma} agonist pioglitazone and ibuprofen reduces glial Inoue H, Baetge EE, Hersh LB. 1993. Enhancer containing
inflammation and A{beta}142 levels in APPV717I trans- unusual GC boxlike sequences on the human choline
genic mice. Brain 128: 1442-1453. acetyltransferase gene. Mol Brain Res 20: 299-304.
Hoffman PW, Chernak JM. 1995. DNA binding and regu- Iqbal K, Alonso Adel C, Chen S, Chohan MO, ElAkkad E,
latory effects of transcription factors SP1 and USF at the rat et al. 2005. Tau pathology in Alzheimer disease and other
amyloid precursor protein gene promoter. Nucl Acids Res tauopathies. Biochim Biophys Acta 1739: 198-210.
23: 2229-2235. Irizarry MC, Locascio JJ, Hyman BT. 2001. betasite APP
Holsinger RM, McLean CA, Beyreuther K, Masters CL, cleaving enzyme mRNA expression in APP transgenic
Evin G. 2002. Increased expression of the amyloid pre- mice: Anatomical overlap with transgene expression and
cursor betasecretase in Alzheimers disease. Ann Neurol static levels with aging. Am J Pathol 158: 173-177.
51: 783-786. Johnson KF, Kornfeld S. 1992. The cytoplasmic tail of the
Holsinger RM, McLean CA, Collins SJ, Masters CL, Evin G. mannose 6phosphate/insulinlike growth factorII recep-
2004. Increased betasecretase activity in cerebrospinal tor has two signals for lysosomal enzyme sorting in the
fluid of Alzheimers disease subjects. Ann Neurol 55: 898- Golgi. J Cell Biol 119: 249-257.
899. Kamal A, AlmenarQueralt A, LeBlanc JF, Roberts EA,
Hooper NM, Karran EH, Turner AJ. 1997. Membrane protein Goldstein LS. 2001. Kinesinmediated axonal transport of
secretases. Biochem J 321: 265-279. a membrane compartment containing betasecretase and
Hoshi M, Takashima A, Murayama M, Yasutake K, Yoshida N, presenilin1 requires APP. Nature 414: 643-648.
et al. 1997. Nontoxic amyloid b peptide (142) suppresses Kametaka S, Shibata M, Moroe K, Kanamori S, Ohsawa Y,
acetylcholine synthesispossible role in cholinergic et al. 2003. Identification of phospholipid scramblase 1
dysfunction in Alzheimers disease. J Biol Chem 272: as a novel interacting molecule with betasecretase (beta
2038-2041. site amyloid precursor protein (APP) cleaving enzyme
Hsiao K, Chapman P, Nilsen S, Eckman C, Harigaya Y, et al. (BACE)). J Biol Chem 278: 15239-15245.
1996. Correlative memory deficits, Abeta elevation, and Kang J, Lemaire HG, Unterbeck A, Salbaum JM, Masters CL,
amyloid plaques in transgenic mice. Science 274: 99-102. et al. 1987. The precursor of Alzheimers disease amyloid
Huse JT, Pijak DS, Leslie GJ, Lee VM, Doms RW. 2000. A4 protein resembles a cellsurface receptor. Nature 325:
Maturation and endosomal targeting of betasite amyloid 733-736.
precursor proteincleaving enzyme. The Alzheimers dis- Kar S, Seto D, Gaudreau P, Quirion R. 1996. bamyloidrelated
ease betasecretase. J Biol Chem 275: 33729-33737. peptides inhibit potassiumevoked acetylcholine release from
Huse JT, Byant D, Yang Y, Pijak DS, DSouza I, et al. 2003. rat hippocampal slices. J Neurosci 16: 1034-1040.
Endoproteolysis of betasecretase (betasite amyloid pre- Kelly JF, Furukawa K, Barger SW, Rengen MR, Mark RJ, et al.
cursor proteincleaving enzyme) within its catalytic do- 1996. Amyloid b peptide disrupts carbacholinduced mus-
main. A potential mechanism for regulation. J Biol Chem carinic cholinergic signal transduction in cortical neurons.
278: 17141-17149. Proc Natl Acad Sci USA 93: 6753-6758.
Huse JT, Liu K, Pijak DS, Carlin D, Lee VM, et al. 2002. Kinoshita A, Fukumoto H, Shah T, Whelan CM, Irizarry MC,
betasecretase processing in the transGolgi network prefer- et al. 2003. Demonstration by FRET of BACE interaction
entially generates truncated amyloid species that accumulate with the amyloid precursor protein at the cell surface and in
in Alzheimers disease brain. J Biol Chem 277: 16278-16284. early endosomes. J Cell Sci 116: 3339-3346.
Hussain I, Christie G, Schneider K, Moore S, Dingwall C. Kitazume S, Nakagawa K, Oka R, Tachida Y, Ogawa K, et al.
2001. Prodomain processing of Asp1 (BACE2) is autocata- 2005. In vivo cleavage of alpha2,6sialyltransferase by Alz-
lytic. J Biol Chem 276: 23322-23328. heimer betasecretase. J Biol Chem 280: 8589-8595.
Hussain I, Hawkins J, Shikotra A, Riddell DR, Faller A, et al. Kitazume S, Tachida Y, Oka R, Kotani N, Ogawa K, et al. 2003.
2003. Characterization of the ectodomain shedding of the Characterization of alpha 2,6sialyltransferase cleavage by
betasite amyloid precursor proteincleaving enzyme 1 Alzheimers betasecretase(BACE1). J Biol Chem 278:
(BACE1). J Biol Chem 278: 36264-36268. 14865-14871.
Hussain I, Powell D, Howlett DR, Tew DG, Meek TD, et al. Kitazume S, Tachida Y, Oka R, Shirotani K, Saido TC, et al.
1999. Identification of a novel aspartic protease (Asp 2) as 2001. Alzheimers betasecretase, betasite amyloid precur-
betasecretase. Mol Cell Neurosci 14: 419-427. sor proteincleaving enzyme, is responsible for cleavage
282
11 Alzheimers disease BACE proteases
secretion of a Golgiresident sialyltransferase. Proc Natl Lo AC, Haass C, Wagner SL, Teplow DB, Sisodia SS. 1994.
Acad Sci USA 98: 13554-13559. Metabolism of the Swedish amyloid precursor protein
Koike H, Tomioka S, Sorimachi H, Saido TC, Maruyama K, variant in MadinDarby canine kidney cells. J Biol Chem
et al. 1999. Membraneanchored metalloprotease MDC9 269: 30966-30973.
has an alphasecretase activity responsible for processing Lukiw WJ, Rogaev EI, Wong L, Vaula G, McLachlan DR, et al.
the amyloid precursor protein. Biochem J 343: 371-375. 1994. ProteinDNA interactions in the promoter region
Kola I, Hertzog PJ. 1997. Animal models in the study of the of the amyloid precursor protein (APP) gene in human
biological function of genes on human chromosome 21 and neocortex. Mol Brain Res 22: 121-131.
their role in the pathophysiology of Down syndrome. Luo Y, Bolon B, Damore MA, Fitzpatrick D, Liu H, et al. 2003.
Human Mol Genet 6: 1713-1727. BACE1 (betasecretase) knockout mice do not acquire
Koo EH, Squazzo SL. 1994. Evidence that production and compensatory gene expression changes or develop neural
release of amyloid betaprotein involves the endocytic path- lesions over time. Neurobiol Dis 14: 81-88.
way. J Biol Chem 269: 17386-17389. Luo Y, Bolon B, Kahn S, Bennett BD, BabuKhan S, et al. 2001.
Kozak M. 1987. An analysis of 50 noncoding sequences Mice deficient in BACE1, the Alzheimers betasecretase,
from 699 vertebrate messenger RNAs. Nucl Acids Res 15: have normal phenotype and abolished betaamyloid gener-
8125-8148. ation. Nat Neurosci 4: 231-232.
Lahiri DK, Robakis NK. 1991. The promoter activity of the Marcinkiewicz M, Seidah NG. 2000. Coordinated expression
gene encoding Alzheimer betaamyloid precursor protein of betaamyloid precursor protein and the putative beta
(APP) is regulated by two blocks of upstream sequences. secretase BACE and alphasecretase ADAM10 in mouse and
Mol Brain Res 9: 253-257. human brain. J Neurochem 75: 2133-2143.
Lammich S, Kojro E, Postina R, Gilbert S, Pfeiffer R, et al. Martins RN, Taddei K, Kendall C, Evin G, Bates KA, et al.
1999. Constitutive and regulated alphasecretase cleavage of 2001. Altered expression of apolipoprotein E, amyloid
Alzheimers amyloid precursor protein by a disintegrin precursor protein and presenilin1 is associated with
metalloprotease. Proc Natl Acad Sci USA 96: 3922-3927. chronic reactive gliosis in rat cortical tissue. Neuroscience
Lammich S, Schobel S, Zimmer AK, Lichtenthaler SF, Haass 106: 557-569.
C. 2004. Expression of the Alzheimer protease BACE1 is Masters CL, Simms G, Weinman NA, Multhaup G, McDonald
suppressed via its 50 untranslated region. EMBO Rep 5: BL, et al. 1985. Amyloid plaque core protein in Alzheimer
620-625. disease and Down syndrome. Proc Natl Acad Sci USA 82:
LangeDohna C, Zeitschel U, Gaunitz F, PerezPolo JR, Bigl V, 4245-4249.
et al. 2003. Cloning and expression of the rat BACE1 Meziane H, Dodart JC, Mathis C, Little S, Clemens J, et al.
promoter. J Neurosci Res 73: 73-80. 1998. Memoryenhancing effects of secreted forms of the
Lesne S, Docagne F, Gabriel C, Liot G, Lahiri DK, et al. 2003. betaamyloid precursor protein in normal and amnestic
Transforming growth factorbeta 1 potentiates amyloid mice. Proc Natl Acad Sci USA 95: 12683-12688.
beta generation in astrocytes and in transgenic mice. Mitsuda N, Roses AD, Vitek MP. 1997. Transcriptional regu-
J Biol Chem 278: 18408-18418. lation of the mouse presenilin1 gene. J Biol Chem 272:
Leuba G, Wernli G, Vernay A, Kraftsik R, Mohajeri MH, et al. 23489-23497.
2005. Neuronal and nonneuronal quantitative BACE Motonaga K, Itoh M, Becker LE, Goto Y, Takashima S. 2002.
immunocytochemical expression in the entorhinohippo- Elevated expression of betasite amyloid precursor protein
campal and frontal regions in Alzheimers disease. Dement cleaving enzyme 2 in brains of patients with Down syn-
Geriatr Cogn Disord 19: 171-183. drome. Neurosci Lett 326: 64-66.
Li Q, Sudhof TC. 2004. Cleavage of amyloidbeta precursor Multhaup G, Scheuermann S, Schlicksupp A, Simons A,
protein and amyloidbeta precursorlike protein by BACE1. Strauss M, et al. 2002. Possible mechanisms of APP
J Biol Chem 279: 10542-10550. mediated oxidative stress in Alzheimers disease. Free
Lichtenthaler SF, Dominguez DI, Westmeyer GG, Reiss K, Radic Biol Med 33: 45-51.
Haass C, et al. 2003. The cell adhesion protein Pselectin Ohno M, Sametsky EA, Younkin LH, Oakley H, Younkin SG,
glycoprotein ligand1 is a substrate for the aspartyl protease et al. 2004. BACE1 deficiency rescues memory deficits and
BACE1. J Biol Chem 278: 48713-48719. cholinergic dysfunction in a mouse model of Alzheimers
Lin X, Koelsch G, Wu S, Downs D, Dashti A, et al. 2000. disease. Neuron 41: 27-33.
Human aspartic protease memapsin 2 cleaves the beta Pastorino L, Ikin AF, Lamprianou S, Vacaresse N, Revelli JP,
secretase site of betaamyloid precursor protein. Proc Natl et al. 2004. BACE (betasecretase) modulates the processing
Acad Sci USA 97: 1456-1460. of APLP2 in vivo. Mol Cell Neurosci 25: 642-649.
Alzheimers disease BACE proteases
11 283
Pastorino L, Ikin AF, Nairn AC, Pursnani A, Buxbaum JD. APPbinding protein, regulate cell movement. J Cell Biol
2002. The carboxylterminus of BACE contains a sorting 153: 1403-1414.
signal that regulates BACE trafficking but not the forma- Sambamurti K, Kinsey R, Maloney B, Ge YW, Lahiri DK.
tion of total A(beta). Mol Cell Neurosci 19: 175-185. 2004. Gene structure and organization of the human beta
Pedersen WA, Kloczewiak MA, Blusztajn JK. 1996. Amyloid b secretase (BACE) promoter. FASEB J 18: 1034-1036.
protein reduces acetylcholine synthesis in a cell line derived Sastre M, Dewachter I, Landreth GE, Willson TM,
from cholinergic neurons of the basal forebrain. Proc Natl Klockgether T, et al. 2003. Nonsteroidal antiinflammatory
Acad Sci USA 93: 8068-8071. drugs and peroxisome proliferatoractivated receptor
Pennypacker KR, Fuldner R, Xu R, Hernandez H, Dawbarn D, gamma agonists modulate immunostimulated processing
et al. 1998. Cloning and characterization of the presenilin of amyloid precursor protein through regulation of beta
2 gene promoter. Mol Brain Res 56: 57-65. secretase. J Neurosci 23: 9796-9804.
Peschon JJ, Slack JL, Reddy P, Stocking KL, Sunnarborg S, Sastre M, Dewachter I, Roner S, Bogdanovic N, Rosen E,
et al. 1998. An essential role for ectodomain shedding in Borghgraef P, Evert B, DumitrescuOzimek L, Thal D,
mammalian development. Science 282: 1281-1284. Landreth G, Walter J, Klockgether T, van Leuven F, Heneka
Pollwein P. 1993. Overlapping binding sites of two different MT. 2006. NSAIDs repress BACE1 gene promoter activity
transcription factors in the promoter of the human gene for by activation of the peroxisome proliferatoractivated
the Alzheimer amyloid precursor protein. Biochem Bio- receptorg (PPARg). Proc Natl Acad Sci USA 103: 443448.
phys Res Commun 190: 637-647. Saunders AJ, Kim TW, Tanzi RE. 1999. BACE Maps to
Preece P, Virley DJ, Costandi M, Coombes R, Moss SJ, et al. Chromosome 11 and a BACE homolog, BACE2, reside in
2003. Betasecretase (BACE) and GSK3 mRNA levels in the obligate Down syndrome region of chromosome 21.
Alzheimers disease. Mol Brain Res 116: 155-158. Science 286: 1255a.
Puglielli L, Ellis BC, Saunders AJ, Kovacs DM. 2003. Ceramide Schmechel A, Zentgraf H, Scheuermann S, Fritz G, Pipkorn R,
stabilizes betasite amyloid precursor proteincleaving et al. 2003. Alzheimer betaamyloid homodimers facilitate
enzyme 1 and promotes amyloid betapeptide biogenesis. A beta fibrillization and the generation of conformational
J Biol Chem 278: 19777-19783. antibodies. J Biol Chem 278: 35317-35324.
Qiao X, Cummins DJ, Paul SM. 2001. Neuroinflammation Scholefield Z, Yates EA, Wayne G, Amour A, McDowell W,
induced acceleration of amyloid deposition in the et al. 2003. Heparan sulfate regulates amyloid precursor
APPV717F transgenic mouse. Eur J Neurosci 14: protein processing by BACE1, the Alzheimers beta
474-482. secretase. J Cell Biol 163: 97-107.
Riddell DR, Christie G, Hussain I, Dingwall C. 2001. Com- Selkoe D, Kopan R. 2003. Notch and presenilin: Regulated
partmentalization of betasecretase (Asp2) into low intramembrane proteolysis links development and degen-
buoyant density, noncaveolar lipid rafts. Curr Biol 11: eration. Annu Rev Neurosci 26: 565-597.
1288-1293. Selkoe DJ, Schenk D. 2003. Alzheimers disease: Molecular
Roberds SL, Anderson J, Basi G, Bienkowski MJ, Branstetter understanding predicts amyloidbased therapeutics. Annu
DG, et al. 2001. BACE knockout mice are healthy despite Rev Pharmacol Toxicol 43: 545-584.
lacking the primary betasecretase activity in brain: Impli- Shi XP, Chen E, Yin KC, Na S, Garsky VM, et al. 2001. The pro
cations for Alzheimers disease therapeutics. Hum Mol domain of betasecretase does not confer strict zymogen
Genet 10: 1317-1324. like properties but does assist proper folding of the protease
Rogers GW Jr, Edelman GM, Mauro VP. 2004. Differential domain. J Biol Chem 276: 10366-10373.
utilization of upstream AUGs in the betasecretase mRNA Shiba T, Kametaka S, Kawasaki M, Shibata M, Waguri S, et al.
suggests that a shunting mechanism regulates translation. 2004. Insights into the phosphoregulation of betasecretase
Proc Natl Acad Sci USA 101: 2794-2799. sorting signal by the VHS domain of GGA1. Traffic 5:
Roner S, Apelt J, Schliebs R, PerezPolo JR, Bigl V. 2001. 437-448.
Neuronal and glial betasecretase (BACE) protein expres- Sinha S, Anderson JP, Barbour R, Basi GS, Caccavello R, et al.
sion in transgenic Tg2576 mice with amyloid plaque 1999. Purification and cloning of amyloid precursor pro-
pathology. J Neurosci Res 64: 437-446. tein betasecretase from human brain. Nature 402: 537-540.
Roner S, Lange-Dohna C, Zeitschel U, Perez-Polo JR. 2005. Sisodia SS. 1992. Betaamyloid precursor protein cleavage by
Alzheimers disease -secretase BACE1 is not a neuron- a membranebound protease. Proc Natl Acad Sci USA 89:
specific enzyme. J Neurochem 92: 226-234. 6075-6079.
Sabo SL, Ikin AF, Buxbaum JD, Greengard P. 2001. The Solans A, Estivill X, de La Luna S. 2000. A new aspartyl
Alzheimer amyloid precursor protein (APP) and Fe65, an protease on 21q22.3, BACE2, is highly similar to Alzheimers
284
11 Alzheimers disease BACE proteases
amyloid precursor protein betasecretase. Cytogenet Cell von Arnim CA, Tangredi MM, Peltan ID, Lee BM, Irizarry
Genet 89: 177-184. MC, et al. 2004. Demonstration of BACE (betasecretase)
Sun A, Koelsch G, Tang J, Bing G. 2002. Localization of beta- phosphorylation and its interaction with GGA1 in cells by
secretase memapsin 2 in the brain of Alzheimers patients fluorescencelifetime imaging microscopy. J Cell Sci 117:
and normal aged controls. Exp Neurol 175: 10-22. 5437-5445.
Sun X, He G, Song W. 2006a. BACE2, as a novel APP theta- Walter J, Fluhrer R, Hartung B, Willem M, Kaether C, et al.
secretase, is not responsible for the pathogenesis of Alzhei- 2001. Phosphorylation regulates intracellular trafficking of
mers disease in Down syndrome. FASEB J 20: 1369-1376. betasecretase. J Biol Chem 276: 14634-14641.
Sun X, Tong Y, Qiung H, Chen CH, Song W. 2006b. Westmeyer GG, Willem M, Lichtenthaler SF, Lurman G,
Increased BACE1 maturation contributes to the pathogen- Multhaup G, et al. 2004. Dimerization of betasite beta
esis of Alzheimers disease in Down syndrome. FASEB J 20: amyloid precursor proteincleaving enzyme. J Biol Chem
1361-1368. 279: 53205-53212.
Sun X, Wang Y, Qing H, Christensen MA, Liu Y, et al. 2005. Wickham L, Benjannet S, Marcinkiewicz E, Chretien M,
Distinct transcriptional regulation and function of the Seidah NG. 2005. Betaamyloid protein converting enzyme 1
human BACE2 and BACE1 genes. FASEB J 19: 739-749. and brainspecific type II membrane protein BRI3: Binding
Tanahashi H, Tabira T. 2001. Three novel alternatively spliced partners processed by furin. J Neurochem 92: 93-102.
isoforms of the human betasite amyloid precursor protein Willis AE. 1999. Translational control of growth factor
cleaving enzyme (BACE) and their effect on amyloid beta and protooncogene expression. Int J Biochem Cell Biol
peptide production. Neurosci Lett 307: 9-12. 31: 73-86.
Tanzi RE, Gusella JF, Watkins PC, Bruns GA, St George Wolfe MS, Xia W, Ostaszewski BL, Diehl TS, Kimberly WT,
Hyslop P, et al. 1987. Amyloid bprotein gene: cDNA, et al. 1999. Two transmembrane aspartates in presenilin1
mRNA distribution, and genetic linkage near the Alzheimer required for presenilin endoproteolysis and gamma
locus. Science 235: 880-884. secretase activity. Nature 398: 513-517.
ToliverKinsky T, Wood T, PerezPolo JR. 2000. Nuclear factor Wong HK, Sakurai T, Oyama F, Kaneko K, Wada K, 2005.
kappaB/p49 is a negative regulatory factor in nerve growth Beta subunits of voltagegated sodium channels are novel
factorinduced choline acetyltransferase promoter activity substrates of BACE1 and gammasecretase. J Biol Chem
in PC12 cells. J Neurochem 75: 2241-2251. Apr 11.
Tun H, Marlow L, Pinnix I, Kinsey R, Sambamurti K. 2002. Yan R, Bienkowski MJ, Shuck ME, Miao H, Tory MC,
Lipid rafts play an important role in A beta biogenesis by et al. 1999. Membraneanchored aspartyl protease with
regulating the betasecretase pathway. J Mol Neurosci 19: Alzheimers disease betasecretase activity. Nature 402:
31-35. 533-537.
Tyler SJ, Dawbarn D, Wilcock GK, Allen SJ. 2002. Alpha and Yan R, Munzner JB, Shuck ME, Bienkowski MJ. 2001. BACE2
betasecretase: Profound changes in Alzheimers disease. functions as an alternative alphasecretase in cells. J Biol
Biochem Biophys Res Commun 299: 373-376. Chem 276: 34019-34027.
Vassar R, Bennett BD, BabuKhan S, Kahn S, Mendiaz EA, Yang LB, Lindholm K, Yan R, Citron M, Xia W, et al.
et al. 1999. Betasecretase cleavage of Alzheimers amyloid 2003. Elevated betasecretase expression and enzymatic
precursor protein by the transmembrane aspartic protease activity detected in sporadic Alzheimer disease. Nat Med
BACE. Science 286: 735-741. 9: 3-4.
von Arnim CA, Kinoshita A, Peltan ID, Tangredi MM, Herl L, Yasojima K, McGeer EG, McGeer PL. 2001. Relationship
Lee BM, Spoelgen R, Hshieh TT, Ranganathan S, Battey FD, between beta amyloid peptide generating molecules and
Liu CX, Bacskai BJ, Sever S, Irizarry MC, Strickland DK, neprilysin in Alzheimer disease and normal brain. Brain
Hyman BT. 2005. The low density lipoprotein receptor- Res 919: 115-121.
related protein (LRP) is a novel beta-secretase (BACE1)
substrate. J Biol Chem 280: 17777-17785.
12 Protein Misfolding, a Common
Mechanism in the Pathogenesis of
Neurodegenerative Diseases
L. Vergara . K. Abid . C. Soto
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
Abstract: Neurodegenerative diseases are some of the most debilitating disorders, affecting abstract
thinking, skilled movements, emotional feelings, cognition, and memory. Recent and compelling evi-
dence indicates that the misfolding, aggregation, and accumulation of proteins in the brain may be the
cause of these diseases. The aim of this chapter is to review the literature around the molecular mechanism
and role of misfolded protein aggregates in neurodegeneration and the possibilities for therapeutic
intervention.
List of Abbreviations: AD, Alzheimers disease; ALS, amyotrophic lateral sclerosis; APP, amyloid-
precursor protein; BACE, b-amyloid cleaving enzyme 1; BSE, bovine spongiform encephalopathy; CJD,
Creutzfeldt-Jakob disease; CWD, chronic wasting disease; fCJD, Familial CJD; GSS, Gerstmann-Straussler-
Scheinker; HD, Huntington disease; iCJD, Iatrogenic CJD; ND, neurodegenerative diseases; NFT,
Neurofibrillary tangles; PD, Parkinson disease; RNAi, RNA interference; SCA, spinocerebellar ataxia;
sCJD, Sporadic CJD; SOD1, superoxide dismutase 1; TSE, transmissible spongiform encephalopathies;
vCJD, Variant CJD
1 Introduction
As a consequence of the longer life expectancy brought by the advances in medical care, the last decade has
observed an increase in the incidence of a number of neurodegenerative diseases (ND) including Alzhei-
mers disease (AD), Parkinson disease (PD), Huntington disease (HD) (and related polyglutamine dis-
orders including several forms of spinocerebellar ataxia or SCA), transmissible spongiform
encephalopathies (TSEs, which include several human and animal diseases), and amyotrophic lateral
sclerosis (ALS). This diverse group of diseases can affect abstract thinking, skilled movements, emotional
feelings, cognition, memory, and other abilities (Martin, 1999). Despite their differences in clinical
symptoms and disease progression, these disorders do share some common features: most of them (except
HD and SCA) have both sporadic and inherited origins, all of them appear later in life (usually after the
fourth or fifth decade), and their pathology is characterized by neuronal loss and synaptic abnormalities
(Martin, 1999).
Recent, but compelling evidences suggest that protein misfolding and aggregation is the most likely
cause of ND (Soto, 2003). The physiology of cells, tissues, and organisms depends on the correct activity of
a network of thousands of proteins. The function of a protein depends on its three-dimensional structure,
which is determined by its amino acid sequence. In ND, some normal proteins become improperly folded
and begin a process of polymerization to form large protein aggregates, commonly referred to as amyloid-
like deposits (Soto, 2003). Amyloid was the name originally given to extracellular protein deposits found in
AD and systemic amyloid disorders, but it has been recently used to refer also to some of the intracellular
aggregates. Amyloid is a generic term that refers to aggregates organized in a cross-b structure, which
contain specific morphological and tinctorial characteristics, such as binding to congo red and thioflavin S,
higher resistance to proteolytic degradation, and a fibrillar appearance under electron microscopy (straight,
unbranched, 10-nm-wide fibrils) (> Figure 12-1) (Glenner, 1980; Sipe, 1992).
Misfolding and aggregation of different proteins is responsible for distinct diseases, and no
structural homology or similar amino acid sequences have been demonstrated among the proteins
implicated in different ND. Protein misfolding and aggregation not only occurs in the brain, and indeed
several systemic disorders have also this molecular mechanism of pathogenesis, including diabetes type II,
serpin deficiency disorders, hemolytic anemia, amyloid polyneuropathy, cystic fibrosis, dialysis-related
amyloidosis, secondary or reactive amyloidosis, and more than 15 other diseases (Carrell and Gooptu,
1998; Soto, 2001).
In this chapter we will review the characteristics and molecular mechanisms implicated in the
pathogenesis of ND. We will also outline some of the strategies under development for the treatment of
these yet incurable disorders.
Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases 12 287
. Figure 12-1
Morphological, tinctorial, and structural characteristic of amyloid. (a) Tissue staining with antibodies specific for
the protein forming the aggregates. (b) Electron microscopy shows typical straight, unbranched fibrils, which
are 510 nm in diameter and 100 nm in length. (c) Structure of the typical amyloid-binding dyes Congo red and
Thioflavine T. (d) Cong red staining visualized under polarized light shows green/yellow birefringency. X-ray
fiber diffraction studies show a characteristic pattern known as cross-b, represented by signals at 4.7A and 9.7A.
(e) A diagrammatic picture for the cross-b structure showing the periodicity observed in X-ray diffraction
studies. (f) Computer modeling of the putative tridimensional structure of a prototype amyloid fibril viewed
from the side and from the top
Neuropathologic and genetic studies, along with the development of transgenic animal models, have
provided strong evidence for the involvement of protein misfolding in ND. Nearly 100 years ago, the
German neuropathologist Alois Alzheimer described for the first time the typical aggregates observed in the
288 12 Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases
brain parenchyma of demented people (Alzheimer et al., 1995). These aggregates have now been
well characterized as formed by protein fibrils called amyloid plaques. The end point of protein misfolding
is usually aberrant protein aggregation and accumulation as amyloid-like deposits in the brain (Glenner,
1980; Sipe, 1992; Soto, 2003; Dimakopoulos and Dimakopoulos, 2005; Ross and Poirier, 2005). The
involvement of amyloid deposition in the pathogenesis of ND has been suggested by the colocalization
of protein aggregates with sites of tissue degeneration and its correlation with the symptoms of the
disease. Furthermore, protein deposits have become so frequent in affected individuals that their presence
is used for definitive diagnosis (Westermark, 1995; Gillmore et al., 1997). On the other hand, absolute
proof is still awaiting confirmation that aggregation itself is the culprit of the disease and not only an
inseparable epiphenomenon (Carrell and Gooptu, 1998; Tran and Miller, 1999; Goldberg and Lansbury,
2000; Caughey and Lansbury, 2003). Indeed, today the most accepted view is that the process of misfolding
and early stages of oligomerization, rather than the mature compacted aggregates deposited in the brain, are
the real culprits in neurodegeneration (Glabe and Kayed, 2006; Lansbury and Lashuel, 2006; Haass
and Selkoe, 2007).
Genetic studies have provided strong support for the role of protein misfolding in ND (Hardy and
Gwinn-Hardy, 1998; Selkoe et al., 2002). In addition to sporadic and infectious origin, ND can also be
caused by inherited defects. Epidemiological analysis has shown the presence of several loci in the human
genome responsible for diverse ND. Importantly, in most of the inherited cases, the mutated gene encodes
the protein that undergoes misfolding and aggregation. Thus it is believed that specific mutations in critical
amino acids may destabilize the normal protein folding, increasing its propensity to misfold and aggregate.
Mutations in the respective fibrillar proteins have been associated with familial forms of all ND, and also
with many other forms of systemic and brain disorders associated with protein misfolding and aggregation,
including amyloid polyneuropathy, cardiac amyloidosis, visceral amyloidosis, cerebral hemorrhage with
amyloidosis of the Dutch and Icelandic type, cerebral amyloidosis of the British and Danish type,
thromboembolic disease, angioedema, emphysema, sickle cell anemia, and diabetes type 2 (Jacobson and
Buxbaum, 1991; Kelly, 1996; Buxbaum and Tagoe, 2000). The familial forms usually have an earlier onset
and greater severity than sporadic cases.
As described later in this chapter, another strong evidence for the role of protein misfolding and
aggregation in ND has been the successful production of animal models by solely introducing the mutated
human gene encoding for the protein undergoing misfolding. In most of cases, this is enough to produce a
disease in the animals with similar pathological, clinical, and biochemical features as the human disease.
for APP and the presenilins (PS1 and PS2) (Selkoe, 2001, 2004b). The presence of the e4 allele of
apolipoprotein E (ApoE) has been shown to be a major predisposition factor for the late-onset form of
AD (Rebeck et al., 1993; Gearing et al., 1996; Lane and Farlow, 2005). APP is a transmembrane cell surface
glycoprotein that can be found in several forms arising both from alternative splicing and posttranslational
modifications (Walter et al., 1997; Selkoe, 2004b). The APP695 isoform is the most abundant in neurons;
other forms are expressed in other cell types of the brain, as well as in other tissues. Although the function
of APP is unclear, its soluble form has been described as an autocrine (Saitoh et al., 1989) and a
neuroprotective (Mattson et al., 1993) factor. The presence of an ectodomain containing a KPI (serine
protease inhibitor domain of Kunitz type) motif has raised the possibility that APP may function as
a protease inhibitor. Indeed, APP has been shown to inhibit the serine protease factor XIa in the clotting
cascade (Smith et al., 1990). Other suggested functions include participation in cellcell and cellsubstrate
adhesions (Schubert et al., 1989; Qiu et al., 1995), and more recent findings have shown that APP may be
involved in the brain regulation of neural progenitor cell proliferation (Caille et al., 2004; Conti and
Cattaneo, 2005) as well as in axonal outgrowth after injury in a fly model (Leyssen et al., 2005). No evidence
has been obtained to support the hypothesis that a loss of cellular function of APP may explain AD
pathogenesis. Indeed, knockout mice with the APP gene are viable and fertile; however, they exhibit a
decreased locomotor activity and forelimb grip strength, both events occurring later in adult life (Zheng
et al., 1995).
APP can be cleaved by three different complexes of proteases; a, b, and g secretases. The Ab fragment
comprises the first 1214 residues of the transmembrane domain plus 28 extracellular residues (Sisodia,
1992). Cleavage by a-secretase gives rise to a large soluble fragment (sAPPa) and a smaller peptide of 83 aa
(c83), which is retained in the membrane. The c83 peptide can be further cleaved by g-secretase to generate
a small peptide (p3) that encompasses two-thirds of Ab, but is not involved in amyloid formation (Chyung
et al., 2005; Agiostratidou et al., 2006). Alternative cleavage of APP is carried out by b-secretase, also called
b-amyloid cleaving enzyme 1 (BACE), an aspartyl protease. BACE cleavage gives rise to the sAPPb (which is
soluble and secreted from the cell membrane) and the 99 c-terminal amino acids retained in the membrane.
The c99 fragment is heterogeneously cleaved by g-secretase to produce Ab 40 and 42 peptides, ending at the
amino acid 40 and 42, respectively (Selkoe, 2004a). Ab42 is the most prevalent form found in amyloid
plaques. It is more hydrophobic than Ab40 and thus more prone to aggregation, despite the fact that it is
normally less abundant than its counterpart in healthy individuals (Selkoe, 1998, 2004a). Even though Ab is
embedded in the membrane at the time cleavage occurs, it is subsequently released to the extracellular space
were it is able to initiate the formation of fibrils resulting in the genesis of amyloid plaques. Mutations
observed in APP-, PS1-, and PS2-encoding genes result in an increased production of total Ab peptides,
selective increase of the Ab42 variant, or in Ab forms more prone to misfold (Soto, 1999; Selkoe, 2001).
Presenilin 1 missense mutations have been shown to be responsible of causing the earliest and most
aggressive forms of AD. The genes of PS1 and PS2 were shown to encode for the two presenilins that
form the catalytic core of the g-secretase complex (Wolfe et al., 1999). PS1 and PS2 are highly homologous
and functionally redundant (Wolfe et al., 1999; Haass and Steiner, 2002; Brunkan and Goate, 2005).
Mutations within the PS genes might modify the structure of the presenilin, affecting the precision of the
g-secretase activity and usually resulting in an increase in production of the Ab42 peptide (Selkoe, 2001).
Even though more than 100 mutations that cause AD have been found for the PS1 gene alone, genetically
based AD represents a minority compared with sporadic cases.
NFTs, the other specific characteristic lesion found in the brain of AD patients, are long bundles of
fibers that form large vesicles within the cytoplasm. These fibers, structurally organized in pairs of
approximately 10 nm in length, are composed of the microtubule-associated Tau proteins, whose main
function is microtubule stabilization, but it has also been reported to be involved in membrane interaction
or acting as an anchor for enzymes (Sontag et al., 1999). The majority of the Tau proteins found in NFTs
are highly insoluble and hyperphosphorylated (Grundke-Iqbal et al., 1986; Kosik et al., 1986). NFTs are
biochemically similar to amyloid plaques, although these two well-known lesions of AD have been observed
independent of each other in affected patients. Unlike amyloid plaques, NFT formation has not been found
to be because of known mutations within the genes associated with familial AD (Armstrong, 2006).
Inherited mutations in the protein Tau lead to fronto-temporal dementia with parkinsonism (Goedert
290 12 Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases
and Spillantini, 2000). In this disease, mutated Tau proteins accumulate in the cytoplasm as NFT without
extracellular amyloid plaques. Even though this disease is far less common than AD, it leads to similar
clinical manifestations (Hutton, 2001).
basal ganglia, HD especially targets neurons of the striatum, particularly those in the caudate nuclei and the
pallidum. In severe cases, the brain cortex is also largely affected. Nuclear protein aggregates composed of
the huntingtin protein are a typical feature of HD (DiFiglia et al., 1997).
The large majority of HD cases are inherited in an autosomal dominant way, due to a triplet
repeat expansion of CAG (coding for the amino acid glutamine) in exon 1 of the gene encoding for
the protein huntingtin (Wanker, 2000). The HD gene comprises a region coding for a stretch of poly-
glutamine residues close to the N terminus of the protein. In the case of the triplet expansion observed in
patients, the longer glutamine portion promotes protein aggregation in the brain (Wanker, 2000). These
deposits are known to trigger oxidative and endoplasmic stress along with proteosomal and mitochondrial
dysfunction (Meriin et al., 2002; Nishitoh et al., 2002). Individuals with 35 or fewer triplet repeats
do not develop HD, in contrast, those with more than 40 repeats always develop the disease. Several
biological activities have been attributed to huntingtin, including cytoskeletal anchoring, axonal transport,
endocytosis, and intracellular trafficking (Cattaneo et al., 2001; Smith et al., 2005). Evidence also points to a
role in brain development and neuroprotection. Ablation of huntingtin in transgenic mice results in
embryonic death, and conditional deletion results in neurodegeneration, suggesting that huntingtin has a
protective role (Reiner et al., 2003). Furthermore, huntingtin has been shown to function as a caspase
substrate that is actively cleaved during apoptosis in cultured cells (Cattaneo et al., 2001). In vitro
experiments have revealed the ability of huntingtin to protect CNS cells from a variety of apoptotic stimuli,
including serum withdrawal, death receptors, and pro-apoptotic proteins activation (Cattaneo et al., 2001;
Reiner et al., 2003).
The nature of the transmissible agent is perhaps the most extensively studied aspect of TSE research
(Soto and Castilla, 2004). Compelling evidence suggest that the infectious agent is composed mainly (or
exclusively) of a misfolded protein that can propagate in the absence of nucleic acid. The proteinaceous
nature of the agent led to its denomination as prion, and extensive research has enabled to identify that a
b-sheet-rich misfolded form of the prion protein (termed PrPSc) is the main component (Prusiner, 1998).
Infection with PrPSc produces the autocatalytic conversion of the normal form of the prion protein
(termed PrPC) into more of PrPSc in a process that results in the formation of a toxic structure responsible
for brain damage and disease. The fact that administration of highly purified PrPSc (>99%) is sufficient to
induce the disease in experimental animals provides a strong support not only for the prion hypothesis, but
also for the key role of protein misfolding in neurodegeneration.
Several animal models have provided tremendous amount of information in understanding the mechan-
isms underlying ND (Bilen and Bonini, 2005; Marsh and Thompson, 2006). Drosophila has been used for
modeling most syndromes, because the availability of powerful genetic tools allows testing many possible
genetic modifiers. In addition, the architecture of the fly nervous system presents several similarities with
the mammalian CNS, like compartmentalized brain functions such as olfaction, vision, or memory. Fly
models are engineered by cloning the genes responsible for the diseases observed in humans into transpo-
sable p-element vectors under the control of a yeast transcription system and injected into fly embryos
to obtain transgenic animals (Bilen and Bonini, 2005). Neuronal degeneration is typically measured
by observation of the structure of the photoreceptor neurons of the eyes and motor function. Like in
humans, transgenic drosophila models for human genes involved in ND develop progressive neuropathol-
ogy (Bilen and Bonini, 2005; Marsh and Thompson, 2006). For example, upon expressing polyQ-
tract extended protein or mutant alpha synuclein, tau, or Ab the number of intact photoreceptors
rapidly decreased after the day of eclosion (Jackson et al., 1998). Aggregates of misfolded proteins are
also found in transgenic flies, expressing the mutant human proteins, supporting the fact that fly is a
suitable model for neurodegeneration (Bilen and Bonini, 2005; Marsh and Thompson, 2006). However,
major limitations exist, the most significant ones including differences in morphology and functionality of
the brain, differences in metabolism, a lack of or reduced inflammatory response, or the absence of the
bloodbrain barrier.
Another animal model used to study ND is the worm Caenorhabditis elegans (Link, 2006; Brignull et al.,
2007). This organism has an invariant number of 959 cells, of which, remarkably around one-third (302)
are neuronal. Microinjection of mutated proteins results in several models of human disease (Kraemer
et al., 2003; Miyasaka et al., 2005; Wu et al., 2005). Transgenic C. elegans expressing Ab presents deposits of
the peptide in muscle cells with formation of vacuolation and motor impairment (Link, 1995). The
Huntingtin gene fused with expanded triplet repeats has been shown to aggregate in the cytoplasm of the
worm and to induce progressive cell death. Interestingly, the toxicity of polyQ-expanded protein was shown
to occur at a threshold of approximately 40 glutamine residues, similar to the human disease (Morley et al.,
2002). RNA interference (RNAi) technology has recently demonstrated that up to 186 genes might be
involved in the response to protein misfolding events in transgenic worms expressing yellow fluorescent
protein fused with an expanded polyQ tract (Nollen et al., 2004).
Considering the complexity of neurological diseases, it is no surprise that transgenic mice have been the
most popular model to study the contribution of protein misfolding in the pathogenesis of these diseases
(Price et al., 2000). Several pathological, biochemical, and clinical features of diverse ND have been
reproduced in transgenic mice models expressing mutant forms of the human genes encoding the
corresponding fibrillar protein (Tom Van Dooren et al., 5 A.D.; Araki et al., 1994; Janson et al., 1996; Price
et al., 1998; Weissmann et al., 1998; Emilien et al., 2000; Gurney, 2000; Price et al., 2000; Grieb, 2004).
Transgenic mice overexpressing the human APP containing diverse mutations responsible for AD progres-
sively develop many of the pathological hallmarks observed in humans, including cerebral amyloid
deposits, neuritic dystrophy, astrogliosis, and cognitive and behavioral alterations (Duff, 1998; Price
294 12 Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases
et al., 1998; Emilien et al., 2000). Similarly, transgenic mice overexpressing the human-mutated SOD1 gene
involved in ALS pathology have shown to develop signs corresponding to the disease found in ALS patients
(Price et al., 1997, 1998). Some of these mice developed motor neuron dysfunction and typical pathological
alterations, including the presence of hyaline inclusion bodies in degenerating axons, muscle atrophy,
astrocyte damage, and extensive loss of large myelinated axons of motor neuronal cells (Grieb, 2004).
Transgenic mice expressing the wild-type form of the human a-synuclein gene develop several of the
clinicopathological features of PD (Hashimoto et al., 2003). Among them are Lewy bodies accumulated in
neurons of the neocortex, hippocampus, and substantia nigra, loss of dopaminergic terminals in the basal
ganglia, and the associated motor impairments (Klement et al., 1998; Masliah et al., 2000). The transgenic
mice model of HD expressing exon 1 of the human huntingtin and carrying 115156 CAG repeat
expansions develop pronounced neuronal intranuclear inclusions (Davies et al., 1997). These typical
cerebral lesions were found to contain huntingtin and ubiquitin protein aggregates that appeared prior
to the development of the neurological phenotype and were strikingly similar to those observed in HD
patients. Furthermore, these mice develop a progressive neurological dysfunction with a movement
disorder and weight loss similar to HD (Sathasivam et al., 1999). One of the first animal models mimicking
a human misfolding protein disorder was obtained when the mutated gene PrnP, coding for human prion
protein (PrP), was overexpressed in a transgenic mice (Hsiao et al., 1990). Spontaneous neurologic disease
with spongiform degeneration was observed in these mice. Moreover, these abnormalities were found to be
transmissible to some transgenic mice by inoculation of the sick brain homogenate.
Thus, the critical role for protein misfolding and aggregation as the culprit of neurodegeneration is
strongly supported by transgenic animal models in which expression of the human proteins involved in
abnormal folding resulted in typical clinical and pathological features. However, time course studies of the
appearance of neurological signs of the disorder in some of the transgenic models have shown that visible
protein aggregates appear after clinical symptoms and significant tissue damage (Janson et al., 1996;
Moechars et al., 1999; Van Leuven, 2000). These observations suggest the presence of a misfolded soluble
intermediate that does not form large detectable aggregates and is not accumulated in the brain, which
could be the real culprit for the pathogenesis (Tran and Miller, 1999; Goldberg and Lansbury, 2000;
Caughey and Lansbury, 2003). This hypothesis implies that the formation of the characteristic aggregates
could be considered as a protective mechanism working by sequestering the possible toxic misfolded
proteins (Caughey and Lansbury, 2003).
No sequence or structural homology has been found among the native proteins implicated in ND. However,
there is accumulating evidence that the aggregates formed by the different misfolded proteins have a similar
molecular structure (Soto, 2003; Glabe and Kayed, 2006; Sawaya et al., 2007). The process of misfolding and
aggregation of most of the proteins involved in ND has been modeled in vitro. Large secondary structural
differences between the monomeric native protein and the aggregated material have been observed using
low-resolution structural studies (Makin and Serpell, 2005). The native protein is generally folded into a a-
helix structure while the misfolded protein is rich in b-sheet conformation. High-resolution studies have
been difficult due to the insoluble and noncrystalline nature of the aggregated proteins. However, recent
studies using X-ray fiber diffraction and solid-state nuclear magnetic resonance have confirmed the b-sheet
structure of the protein aggregates implicated in ND (Makin and Serpell, 2005; Nelson and Eisenberg, 2006;
Tycko, 2006).
Structural studies performed on misfolded proteins reveal that amyloid-like fibrils are composed of
several protofilaments that consist of hydrogen-bonded b-sheets with the b-strands running perpendicular
to the long fiber axis, a structure known as a cross-b conformation (Makin and Serpell, 2005; Makin et al.,
2005). These structural analyses demonstrate that a large conformational rearrangement of the polypeptide
chain occurs during misfolding and aggregation. However, it is presently unclear whether the misfolding
triggers protein aggregation or whether it is the protein oligomerization that induces the conformational
changes. This issue is of the highest interest, especially for the design of effective therapeutic strategies.
Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases 12 295
Although the detailed mechanism for the formation of fibrillar amyloid-like aggregates is not entirely
clear, the initiating event is protein misfolding, which results in the formation of aggregation-prone
structures that can eventually grow by an autocatalytic mechanism (Soto, 2003). Kinetic studies have
suggested that the critical event is the formation of protein oligomers that act as seeds to further propagate
protein misfolding (Soto et al., 2006). This is the basis for the currently accepted nucleation-dependent
polymerization model of amyloid formation (Gajdusek, 1994; Harper and Lansbury, 1997; Soto et al.,
2006). Diverse proteins have been shown to follow this crystallization-like process, including Ab,
huntingtin, and a-synuclein among others. According to this model, aggregation starts after the protein
concentration exceeds a point known as the critical concentration (Harper and Lansbury, 1997). Unfavor-
able interactions between monomers determine a slow phase (termed lag phase) in which oligomers are
formed, providing an ordered nucleus to catalyze further growth of the polymers. The addition of
preformed nuclei (seeds) serves as templates for the reaction and, as a result, the initial, slow phase of
primary nucleation is eliminated (Harper and Lansbury, 1997; Soto et al., 2006).
In vitro studies have been useful for understanding the structural requirements for protein misfolding
and the fragments of the protein mostly implicated in the conformational changes. Using mutated peptides
or shorter Ab fragments it has been shown that the internal hydrophobic region between amino acids 17
and 21 is essential in the early steps of Ab misfolding and aggregation (Soto, 1999). This finding
demonstrates that Ab assembly is partially driven by hydrophobic interactions (Hilbich et al., 1992; Soto
et al., 1995; Wood et al., 1995; Lazo et al., 2005), which is consistent with the higher ability of Ab peptides to
aggregate with two or three extra hydrophobic amino acids at the carboxy terminus (Jarrett et al., 1993).
Indeed, hydrophobic forces appear to be crucial in the aggregation process of most of the misfolded
proteins. In PrPc misfolding, the hydrophobic fragment 106126 of PrPc has been mapped to represent the
most relevant sequence for protein oligomerization (Tagliavini et al., 1993; Ziegler et al., 2006). A similar
observation has been reported concerning the fibrillogenesis of a-synuclein. Although less is known about
this process, evidence indicates that the amino-terminal fragment 187 might be crucial (Serpell et al., 1995,
2000). However, the aggregation process involved in huntingtin and other polyglutamine-containing
proteins seems to be different. In HD and other polyglutamine diseases, both disease and protein aggrega-
tion are associated with an inherited expansion of CAG (the codon for glutamine) repeats. In the case of
HD, aggregation of huntingtin in vitro depends on the length of the polyglutamine repeat (Scherzinger
et al., 1997; Burke et al., 2003), and the aggregation process is driven by the glutamine residue, which
contains an amide group that provides a polar side chain and the potential to form a hydrogen bond with
water in a model termed polar zipper (Perutz et al., 1994; Chan et al., 2005). In this model, b-sheets are
formed and stabilized by the collective strength of cooperative hydrogen bonding involving the amide
groups of the glutamine residues. Therefore, protein aggregation can arise from two different (and, in some
respects, opposite) driving forces: hydrophobic interactions and polar hydrogen bonding among side-chain
groups.
In addition to mature fibrils, several other structures have been described as part of the protein
misfolding and aggregation process, including soluble oligomers, pores, annular structures, spherical
micelles, and protofibrils (Caughey and Lansbury, 2003; Glabe and Kayed, 2006; Haass and Selkoe,
2007). Interestingly, these diverse structures have been identified in the amyloidogenesis process of various
disease-associated proteins, suggesting common misfolding pathways (Caughey and Lansbury, 2003; Glabe
and Kayed, 2006; Haass and Selkoe, 2007). However, currently, the biological relevance of these inter-
mediates is not clear, and it is even unknown whether or not they exist in the diseased brain. Furthermore,
although it is likely that these metastable species assemble in a stepwise process, the relative implication of
each of them is difficult to assess because they are too unstable to study (Glabe and Kayed, 2006; Teplow
et al., 2006). Recently, a set of antibodies that recognize specifically different type of aggregated species, such
as oligomers, annular assemblies, protofibrils, and fibrils, have been produced (Glabe and Kayed, 2006;
Kayed and Glabe, 2006). Strikingly, these antibodies are not sequence specific, but conformation specific
and thus they can recognize intermediate species formed by different proteins (Glabe and Kayed, 2006;
Kayed and Glabe, 2006). In summary, the biophysical studies of the intermediates in the amyloid formation
process indicate that diverse species with progressive degree of aggregation are present simultaneously and
in a dynamic equilibrium between each other (Caughey and Lansbury, 2003; Glabe and Kayed, 2006;
296 12 Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases
Teplow et al., 2006). This makes evaluation of the relative contribution of different protein structures to
neurodegeneration very difficult.
Besides of protein deposits, the typical pattern of brain damage in ND includes selective neuronal death,
synaptic alterations, and astrogliosis (Martin, 1999). The different clinical picture associated with distinct
ND is most likely determined by the diverse brain regions most affected in each case. Although it was widely
thought that neuronal apoptosis was the most important problem in neurodegeneration, recent evidence
from different diseases suggest that extensive neuronal death may not be the initial cause of the disease
(Haass and Selkoe, 2007). Indeed, clinical symptoms can be observed before significant neuronal loss and a
better temporal and topographic correlation is found with synaptic dysfunction (Haass and Selkoe, 2007).
Although protein misfolding and aggregation is undoubtedly associated with neurodegeneration and
disease, the mechanism by which this process leads to brain damage is unknown. The most widely accepted
theory of brain degeneration in ND proposes that misfolding and aggregation results in the acquisition of a
neurotoxic function by the misfolded protein (Soto, 2003). Although the possibility that neuronal damage
may be produced by the loss of activity of the protein upon the conformational changes has been extensively
studied, most of the evidence argues against this mechanism (Soto, 2003). The mechanism by which
misfolded aggregates produce synaptic dysfunction and neuronal death is unknown. It is also unknown
which of the different polymeric structures formed in the process of amyloidogenesis is the triggering factor
of brain damage (Lansbury and Lashuel, 2006; Haass and Selkoe, 2007) (> Figure 12-1). For many years it
was thought that big amyloid-like protein deposits were the species responsible for brain damage. However,
the hypothesis that large aggregates accumulated in the brain are toxic has been challenged by histopatho-
logical, biochemical, and cell biology studies (Lansbury and Lashuel, 2006; Haass and Selkoe, 2007). The
current view is that the process of misfolding and early stages of oligomerization, rather than the mature
compacted aggregates deposited in the brain, are the real culprits in neurodegeneration (Glabe and Kayed,
2006; Lansbury and Lashuel, 2006; Haass and Selkoe, 2007). This hypothesis is supported by results showing
that purified oligomeric species and protofibrils are toxic to cultured neurons, inhibit hippocampal long-
term potentiation, impair synaptic functions, and disrupt cognition and learned behavior in rats (for list of
references, see Glabe and Kayed (2006), Lansbury and Lashuel (2006), and Haass and Selkoe (2007)).
Several mechanisms have been proposed for the neurotoxic activity of misfolded oligomers, and it is
likely that different pathways operate, depending on whether the proteins accumulate intra- or extracellu-
larly (Soto, 2003). Extracellular oligomers might activate a signal transduction pathway leading to apoptosis
by interacting with specific cellular receptors. Intracellular polymers might damage cells by, for example,
recruiting factors essential for cell viability. Another well-supported mechanism by which oligomeric and
annular pore-like structures may damage cells is by membrane disruption and depolarization, resulting in
alterations of ion homeostasis and disregulation of cellular signal transduction (Glabe and Kayed, 2006).
Finally, misfolded proteins could induce cellular oxidative stress by producing free radical species, resulting
in protein and lipid oxidation, elevation of intracellular calcium, and mitochondrial disfunction (Behl et al.,
1994; Hsu et al., 2000).
In spite of the extensive knowledge collected in recent years toward understanding the pathogenesis of ND,
there are no effective therapies directed to prevent or reverse the disease progression. If protein misfolding
and aggregation play a central role in the pathogenesis of ND, then the strategy to follow should be the
prevention and/or the correction of the misfolding process. At least three different strategies have been
proposed to intervene against protein misfolding and aggregation (Soto, 2003): (1) stabilization of the
native protein conformation; (2) inhibition and reversion of protein conformational changes; and (3)
increase the clearance of the misfolded protein.
Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases 12 297
Stabilization of the native folding of the normal protein has been attempted in several protein
misfolding disorders. In the case of AD, nicotine and cotinine have been shown to inhibit Ab amyloid
formation by specifically binding to the peptide and stabilizing its alpha helix structure (Salomon et al.,
1996; Ono et al., 2002). In prion-infected neuroblastoma cells, treatment with chemical chaperones
(reagents known to stabilize native conformation of proteins) results in prevention of PrP misfolding
(Tatzelt et al., 1996; Bennion et al., 2004). Protein engineering has been proposed as an approach to create
sequence-modified proteins with higher stability, lower tendency to misfold, and ability to trans-suppress
the aggregation of wild-type proteins (Villegas et al., 2000). Although the idea is attractive, its therapeutic
application would require sophisticated gene therapy technologies.
Several small molecules, peptide fragments, specific antibodies, and proteins capable of interacting with
b-sheet aggregates have been shown to specifically interact and inhibit protein aggregation (Aguzzi et al.,
2001; LeVine, 2002; Sacchettini and Kelly, 2002; Bose et al., 2005; Soto and Estrada, 2005). Several small
chemical molecules have been reported to have this activity, including Congo red, 40 -iodo-40 -deoxydoxor-
ubicin, curcumin, rosmarinic acid, rifampicine, polyphenol, tetracycline, minocyclin, and ferulic acid
among others (Forloni et al., 2001, 2002; Cardoso et al., 2003; Ono et al., 2005; Familian et al., 2006).
The majority of the active compounds produce their effect by competitively blocking proteinprotein
interactions. These compounds can be classified in two classes depending on their ability to either interact
with the complexes in between monomers or to bind at the edge of b-sheet oligomers, preventing their
growth. One of the major problems with this approach is to determine the mechanism of action of
these competitive inhibitors. Depending on their activity, it is likely that some inhibitors may lead to
accumulation of toxic intermediates and thus may aggravate the disease. Another class of inhibitors
compounds act by destabilization of the pathological b-sheet conformation of misfolded proteins. One
example of this strategy are b-sheet breaker peptides that have been shown to inhibit and reverse protein
misfolding in vitro and in vivo (Soto and Estrada, 2005). b-Sheet breakers are short synthetic peptides that
can specifically interact with the protein fragment that is undergoing misfolding but has been designed to
be unable to get incorporated into the b-sheet structure. b-Sheet breaker peptides have been designed
to block the conformational changes and aggregation of both Ab and PrP in vitro and in various animal
models (Soto et al., 1998, 2000; Permanne et al., 2002).
Another strategy is to increase the clearance of misfolded and aggregated proteins. This alternative
represents an interesting approach based on findings showing that accumulation of protein aggregates
is dependent on a balance between deposition and clearance. In a conditional transgenic model of
HD, preventing the production of mutant fragments caused nuclear inclusions to disappear, indicating
that cells can metabolize the aggregated material (Yamamoto et al., 2000). Perhaps the most promising
strategy to increase the clearance of misfolded proteins is the immunization approach first described for
AD (Schenk et al., 1999). Aggregates of synthetic Ab protein were used as antigens to induce the immune
system to produce antibodies to clear them. Immunization can reduce amyloid load, cerebral damage,
and behavioral impairments in transgenic animal models of AD (Schenk et al., 1999; Janus et al., 2000;
Morgan et al., 2000; Boche et al., 2005; Gandy and Heppner, 2005; Goni and Sigurdsson, 2005; McGeer
et al., 2005; McGavern, 2006). A similar approach has been used for the treatment of TSE (Sigurdsson et al.,
2002; Cashman and Caughey, 2004; Boche et al., 2005; Goni and Sigurdsson, 2005; Magri et al., 2005).
However, a clinical trial to evaluate the efficacy of the immunization strategy in humans affected by AD was
stopped because of several cases of meningoencephalitis (Schenk, 2002). Little is known about the reason
for this side effect, but future research should bring both more knowledge about this problem and also
should result in new strategies to minimize brain inflammation after vaccination. Immunization with
fragments of Ab that preferentially stimulate the B cells (and not the T cells) or passive immunization
with antibodies against the misfolded proteins might be safer strategies (Schenk, 2002). Enhancement of
the clearance of amyloid-like deposits has been also attempted by removing some accessory constituents
found in plaques. The underlying idea is that other factors that tightly bind to the aggregates might
increase their insolubility and resistance to proteolytic degradation. Strategies for removing specifically the
amyloid-P component, proteoglycans, and metal ions have shown the best results in models of AD
and systemic amyloidosis (Kisilevsky et al., 1995; Cherny et al., 2001; Pepys et al., 2002; Lee et al., 2004;
Maynard et al., 2005).
298 12 Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases
Over the last 10 years considerable progress has been made in our understanding of the molecular
mechanisms responsible for ND. The observation that several syndromes are histologically characterized
by accumulation in the brain of normal proteins has led to the hypothesis that most ND are caused by a very
similar cascade of events. Misfolding of distinct proteins appears to be the initiating step that eventually
leads to neuronal dysfunction and clinical symptoms specific for each disease. Formation of amyloid
structures in the brain is the visible consequence of abnormalities in protein folding, but recent data
suggest that the plaques themselves may not be the triggering factor in neurodegeneration. The similarities
observed in the different ND provide hope for a common therapeutic strategy to treat these devastating
illnesses. The development of a therapy based on the misfolding concept may also enable a better
understanding of the mechanisms that eventually lead to amyloid formation.
The concept that protein misfolding leads to severe illness is relatively new; thus, one important
research for the future is to find out whether other diseases are associated with similar alterations. For
example, a report demonstrated that atherosclerosis might indeed be a consequence of ApoB misfolding
induced by high cholesterol level (Ursini et al., 2002). This raises the possibility that other common diseases
might directly or indirectly be a consequence of the misfolding of proteins. During many years it was
thought that only certain proteins could undergo misfolding and eventually form amyloid-like structures,
but this concept has been reconsidered since recent studies have shown that many, if not all, proteins can
form amyloid-like structures under appropriate conditions (Fandrich et al., 2001; Stefani and Dobson,
2003). Even though most of the data were obtained for proteins incubated in nonphysiological pH or salt
concentrations, it cannot be excluded that in certain cellular compartments, these conditions could be
reached transitorily, leading to a seeding event that would trigger the aggregation process. One important
open question is whether protein misfolding always leads to disease or rather a similar phenomenon may
play a normal biological function. The recent demonstration that several proteins can misfold and aggregate
into amyloid-like structures (Lindquist, 1997; Wickner et al., 1999; Iconomidou et al., 2000; Wosten and de
Vocht, 2000; Chapman et al., 2002; Kenney et al., 2002; Berson et al., 2003; Bieler et al., 2005), changing the
biological activity of the protein suggest that the acquisition of alternative protein folding is indeed not only
a disease-associated process. The possibility that many proteins may adopt multiple conformations to exert
different functions might revolutionize our understanding of biology.
Acknowledgments
This project was supported in part by NIH grant R01 AG028821 and an award from CART Foundation.
References
Agiostratidou G, Muros RM, Shioi J, Marambaud P, Robakis mutant transthyretin (Met 30) gene. Pathological and immu-
NK. 2006. The cytoplasmic sequence of E-cadherin pro- nohistochemical similarity to human familial amyloidotic
motes non-amyloidogenic degradation of A beta precur- polyneuropathy, type I. Mol Neurobiol 8: 15-23.
sors. J Neurochem 96: 1182-1188. Armstrong RA. 2006. Plaques and tangles and the pathogene-
Aguzzi A, Glatzel M, Montrasio F, Prinz M, Heppner FL. 2001. sis of Alzheimers disease. Folia Neuropathol 44: 1-11.
Interventional strategies against prion diseases. Nat Rev Balter M. 2001. Infectious diseases. Uncertainties plague pro-
Neurosci 2: 745-749. jections of vCJD toll. Science 294: 770-771.
Alzheimer A, Stelzmann RA, Schnitzlein HN, Murtagh FR. Behl C, Davis JB, Lesley R, Schubert D. 1994. Hydrogen
1995. An English translation of Alzheimers 1907 paper, peroxide mediates amyloid beta protein toxicity. Cell 77:
Uber eine eigenartige Erkankung der Hirnrinde. Clin 817-827.
Anat 8: 429-431. Bennion BJ, DeMarco ML, Daggett V. 2004. Preventing mis-
Araki S, Yi S, Murakami T, Watanabe S, Ikegawa S, et al. 1994. folding of the prion protein by trimethylamine N-oxide.
Systemic amyloidosis in transgenic mice carrying the human Biochemistry 43: 12955-12963.
Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases 12 299
Berson JF, Theos AC, Harper DC, Tenza D, Raposo G, et al. fibrils in vitro producing noncytotoxic species: Screening
2003. Proprotein convertase cleavage liberates a fibrillo- for TTR fibril disrupters. FASEB J 17: 803-809.
genic fragment of a resident glycoprotein to initiate mela- Carrell RW, Gooptu B. 1998. Conformational changes and
nosome biogenesis. J Cell Biol 161: 521-533. diseaseserpins, prions and Alzheimers. Curr Opin Struct
Bertoli-Avella AM, Oostra BA, Heutink P. 2004. Chasing Biol 8: 799-809.
genes in Alzheimers and Parkinsons disease. Hum Genet Cashman NR, Caughey B. 2004. Prion diseasesClose to
114: 413-438. effective therapy? Nat Rev Drug Discov 3: 874-884.
Bieler S, Estrada L, Lagos R, Baeza M, Castilla J, et al. 2005. Cattaneo E, Rigamonti D, Goffredo D, Zuccato C, Squitieri F,
Amyloid formation modulates the biological activity of a et al. 2001. Loss of normal huntingtin function: New devel-
bacterial protein. J Biol Chem 280: 26880-26885. opments in Huntingtons disease research. Trends Neurosci
Bilen J, Bonini NM. 2005. Drosophila as a model for human 24: 182-188.
neurodegenerative disease. Ann Rev Genet 39: 153-171. Caughey B, Lansbury PT. 2003. Protofibrils, pores, fibrils, and
Boche D, Nicoll JAR, Weller RO. 2005. Immunotherapy for neurodegeneration: Separating the responsible protein
Alzheimers disease and other dementias. Curr Opin aggregates from the innocent bystanders. Annu Rev Neu-
Neurol 18: 720-725. rosci 26: 267-298.
Bose M, Gestwicki JE, Devasthali V, Crabtree GR, Graef IA. Chan JCC, Oyler NA, Yau WM, Tycko R. 2005. Parallel beta-
2005. Nature-inspired drug-protein complexes as inhibi- sheets and polar zippers in amyloid fibrils formed by resi-
tors of A beta aggregation. Biochem Soc Trans 33: 543-547. dues 1039 of the yeast prion protein Ure2p. Biochemistry
Brignull HR, Morley JF, Morimoto RI. 2007. The stress of 44: 10669-10680.
misfolded proteins: C. elegans models for neurodegenera- Chapman MR, Robinson LS, Pinkner JS, Roth R, Heuser J,
tive disease and aging. Adv Exp Med Biol 594: 167-189. et al. 2002. Role of Escherichia coli curli operons in direct-
Brown P, Cathala F, Castaigne P, Gajdusek DC. 1986. ing amyloid fiber formation. Science 295: 851-855.
Creutzfeldt-Jakob disease: Clinical analysis of a consecutive Cherny RA, Atwood CS, Xilinas ME, Gray DN, Jones WD,
series of 230 neuropathologically verified cases. Ann Neurol et al. 2001. Treatment with a copper-zinc chelator markedly
20: 597-602. and rapidly inhibits beta-amyloid accumulation in Alzhei-
Brown P, Preece M, Brandel JP, Sato T, McShane L, et al. 2000. mers disease transgenic mice. Neuron 30: 665-676.
Iatrogenic Creutzfeldt-Jakob disease at the millennium. Chyung JH, Raper DM, Selkoe DJ. 2005. gamma-Secretase
Neurology 55: 1075-1081. exists on the plasma membrane as an intact complex that
Bruce ME, Will RG, Ironside JW, McConnell I, Drummond D, accepts substrates and effects intramembrane cleavage. J
et al. 1997. Transmissions to mice indicate that new vari- Biol Chem 280: 4383-4392.
ant CJD is caused by the BSE agent. Nature 389: 498-501. Clayton DF, George JM. 1998. The synucleins: A family of
Bruijn LI, Becher MW, Lee MK, Anderson KL, Jenkins NA, proteins involved in synaptic function, plasticity, neurode-
et al. 1997. ALS-linked SOD1 mutant G85R mediates dam- generation and disease. Trends Neurosci 21: 249-254.
age to astrocytes and promotes rapidly progressive disease Collinge J. 2001. Prion diseases of humans and animals: Their
with SOD1-containing inclusions. Neuron 18: 327-338. causes and molecular basis. Annu Rev Neurosci 24: 519-550.
Brunkan AL, Goate AM. 2005. Presenilin function and Conti L, Cattaneo E. 2005. Controlling neural stem cell divi-
gamma-secretase activity. J Neurochem 93: 769-792. sion within the adult subventricular zone: An APPealing
Budka H, Aguzzi A, Brown P, Brucher JM, Bugiani O, et al. job. Trends Neurosci 28: 57-59.
1995. Neuropathological diagnostic criteria for Creutzfeldt- Davies SW, Turmaine M, Cozens BA, DiFiglia M, Sharp AH,
Jakob disease (CJD) and other human spongiform ence- et al. 1997. Formation of neuronal intranuclear inclusions
phalopathies (prion diseases). Brain Pathol 5: 459-466. underlies the neurological dysfunction in mice transgenic
Burke MG, Woscholski R, Yaliraki SN. 2003. Differential for the HD mutation. Cell 90: 537-548.
hydrophobicity drives self-assembly in Huntingtons dis- Detwiler LA. 1992. Scrapie. Rev Sci Tech 11: 491-537.
ease. Proc Natl Acad Sci USA 100: 13928-13933. DiFiglia M, Sapp E, Chase KO, Davies SW, Bates GP, et al.
Buxbaum JN, Tagoe CE. 2000. The genetics of the amyloi- 1997. Aggregation of huntingtin in neuronal intranuclear
doses. Annu Rev Med 51: 543-569. inclusions and dystrophic neurites in brain. Science 277:
Caille I, Allinquant B, Dupont E, Bouillot C, Langer A, et al. 1990-1993.
2004. Soluble form of amyloid precursor protein regulates Dimakopoulos AC, Dimakopoulos AC. 2005. Protein aggre-
proliferation of progenitors in the adult subventricular gation in Alzheimers disease and other neoropathological
zone. Development 131: 2173-2181. disorders. Curr Alzheimer Res 2: 19-28.
Cardoso I, Merlini G, Saraiva MJ. 2003. 40 -Iodo-40 -deoxydox- Duff K. 1998. Recent work on Alzheimers disease transgenics.
orubicin and tetracyclines disrupt transthyretin amyloid Curr Opin Biotech 9: 561-564.
300 12 Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases
Emilien G, Beyreuther K, Masters CL, Maloteaux JM. 2000. 12 patients and review of the literature. J Child Neurol 21:
Prospects for pharmacological intervention in Alzheimer 223-229.
disease. Arch Neurol 57: 454-459. Grieb P. 2004. Transgenic models of amyotrophic lateral scle-
Familian A, Boshuizen RS, Eikelenboom P, Veerhuis R. rosis. Folia Neuropathol 42: 239-248.
2006. Inhibitory effect of minocycline on amyloid beta fibril Grundke-Iqbal I, Iqbal K, Tung YC, Quinlan M, Wisniewski
formation and human microglial activation. Glia 53: HM, et al. 1986. Abnormal phosphorylation of the
233-240. microtubule-associated protein tau (tau) in Alzheimer
Fandrich M, Fletcher MA, Dobson CM. 2001. Amyloid fibrils cytoskeletal pathology. Proc Natl Acad Sci USA 83:
from muscle myoglobin. Nature 410: 165-166. 4913-4917.
Forloni G, Colombo L, Girola L, Tagliavini F, Salmona M. Gurney ME. 2000. What transgenic mice tell us about neuro-
2001. Anti-amyloidogenic activity of tetracyclines: Studies degenerative disease. Bioessays 22: 297-304.
in vitro. FEBS Lett 487: 404-407. Gurney ME, Pu H, Chiu AY, Dal Canto MC, Polchow CY, et al.
Forloni G, Iussich S, Awan T, Colombo L, Angeretti N, et al. 1994. Motor neuron degeneration in mice that express a
2002. Tetracyclines affect prion infectivity. Proc Natl Acad human Cu, Zn superoxide dismutase mutation. Science
Sci USA 99: 10849-10854. 264: 1772-1775.
Forno LS. 1996. Neuropathology of Parkinson Disease. J Haass C, Selkoe DJ. 2007. Soluble protein oligomers in neu-
Neuropathol Exp Neurol 55: 259-272. rodegeneration: Lessons from the Alzheimers amyloid
Gajdusek DC. 1994. Nucleation of amyloidogenesis in infec- beta-peptide. Nat Rev Mol Cell Biol 8: 101-112.
tious and noninfectious amyloidoses of brain. Ann N Y Haass C, Steiner H. 2002. Alzheimer disease gamma-secretase:
Acad Sci 724: 173-190. A complex story of GxGD-type presenilin proteases. Trends
Gandhi S, Wood NW. 2005. Molecular pathogenesis of Cell Biol 12: 556-562.
Parkinsons disease. Hum Mol Genet 14: 2749-2755. Hardy J, Gwinn-Hardy K. 1998. Genetic classification of pri-
Gandy S, Heppner FL. 2005. Alzheimers amyloid immuno- mary neurodegenerative disease. Science 282: 1075-1079.
therapy: Quo vadis? Lancet Neurol 4: 452-453. Harper JD, Lansbury PT, Jr. 1997. Models of amyloid seeding
Gearing M, Mori H, Mirra SS. 1996. Abeta-peptide length and in Alzheimers disease and scrapie: Mechanistic truths and
apolipoprotein E genotype in Alzheimers disease. Ann physiological consequences of the time-dependent solubil-
Neurol 39: 395-399. ity of amyloid proteins. Annu Rev Biochem 66: 385-407.
Gillmore JD, Hawkins PN, Pepys MB. 1997. Amyloidosis: A Hashimoto M, Rockenstein E, Masliah E. 2003. Transgenic
review of recent diagnostic and therapeutic developments. models of alpha-synuclein pathology: Past, present, and
Br J Haematol 99: 245-256. future. Ann N Y Acad Sci 991: 171-188.
Glabe CG, Kayed R. 2006. Common structure and toxic Hilbich C, Kisters-Woike B, Reed J, Masters CL, Beyreuther K.
function of amyloid oligomers implies a common mecha- 1992. Substitutions of hydrophobic amino acids reduce the
nism of pathogenesis. Neurology 66: S74-S78. amyloidogenicity of Alzheimers disease beta A4 peptides.
Glenner GG. 1980. Amyloid deposits and amyloidosis. The J Mol Biol 228: 460-473.
beta-fibrilloses (first of two parts). N Engl J Med 302: Hilton DA. 2000. vCJDpredicting the future. Neuropathol
1283-1292. Appl Neurobiol 26: 405-407.
Goedert M, Jakes R, Crowther RA, Hasegawa M, Smith MJ, Hsiao KK, Scott M, Foster D, Groth DF, DeArmond SJ, et al.
et al. 1998. Intraneuronal filamentous tau protein and 1990. Spontaneous neurodegeneration in transgenic mice
alpha-synuclein deposits in neurodegenerative diseases. with mutant prion protein. Science 250: 1587-1590.
Biochem Soc Trans 26: 463-471. Hsu LJ, Sagara Y, Arroyo A, Rockenstein E, Sisk A, et al. 2000.
Goedert M, Spillantini MG. 2000. Tau mutations in fronto- alpha-Synuclein promotes mitochondrial deficit and oxi-
temporal dementia FTDP-17 and their relevance for Alz- dative stress. Am J Pathol 157: 401-410.
heimers disease. Biochim Biophys Acta 1502: 110-121. Huang XM, Chen P, Kaufer DI, Troster AI, Poole C. 2006.
Goldberg MS, Lansbury PT. 2000. Is there a cause-and-effect Apolipoprotein e and dementia in Parkinson diseaseA
relationship between alpha-synuclein fibrillization and meta-analysis. Arch Neurol 63: 189-193.
Parkinsons disease? Nat Cell Biol 2: E115-E119. Huang XM, Chen PC, Poole C. 2004. APOE-epsilon 2 allele
Goni F, Sigurdsson EM. 2005. New directions towards safer associated with higher prevalence of sporadic Parkinson
and effective vaccines for Alzheimers disease. Curr Opin disease. Neurology 62: 2198-2202.
Mol Ther 7: 17-23. Hutton M. 2001. Missense and splice site mutations in tau
Gonzalez-Alegre, P, Afifi AK. 2006. Clinical characteristics of associated with FTDP-17: Multiple pathogenic mechan-
childhood-onset (juvenile) Huntington disease: Report of isms. Neurology 56: S21-S25.
Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases 12 301
Iconomidou VA, Vriend G, Hamodrakas SJ. 2000. Amyloids Kraemer BC, Zhang B, Leverenz JB, Thomas JH, Trojanowski
protect the silkmoth oocyte and embryo. FEBS Lett 479: JQ, et al. 2003. Neurodegeneration and defective neurotrans-
141-145. mission in a Caenorhabditis elegans. model of tauopathy.
Jackson GR, Salecker I, Dong XZ, Yao X, Arnheim N, et al. Proc Natl Acad Sci USA 100: 9980-9985.
1998. Polyglutamine-expanded human huntingtin trans- Lambrechts D, Lafuste P, Carmeliet P, Conway EM. 2006.
genes induce degeneration of Drosophila photoreceptor Another angiogenic gene linked to amyotrophic lateral
neurons. Neuron 21: 633-642. sclerosis. Trends Mol Med 12: 345-347.
Jacobson DR, Buxbaum JN. 1991. Genetic aspects of amyloid- Lane RM, Farlow MR. 2005. Lipid homeostasis and apolipo-
osis. Adv Hum Genet 20: 69-11. protein E in the development and progression of Alzhei-
Janson J, Soeller WC, Roche PC, Nelson RT, Torchia AJ, et al. mers disease. J Lipid Res 46: 949-968.
1996. Spontaneous diabetes mellitus in transgenic mice Lansbury PT, Lashuel HA. 2006. A century-old debate on
expressing human islet amyloid polypeptide. Proc Natl protein aggregation and neurodegeneration enters the
Acad Sci USA 93: 7283-7288. clinic. Nature 443: 774-779.
Janus C, Pearson J, McLaurin J, Mathews PM, Jiang Y, et al. Lazo ND, Grant MA, Condron MC, Rigby AC, Teplow DB.
2000. A beta peptide immunization reduces behavioural 2005. On the nucleation of amyloid beta-protein monomer
impairment and plaques in a model of Alzheimers disease. folding. Protein Sci 14: 1581-1596.
Nature 408: 979-982. Lee JY, Friedman JE, Angel I, Kozak A, Koh JY. 2004. The
Jarrett JT, Berger EP, Lansbury PT, Jr. 1993. The C-terminus of lipophilic metal chelator DP-109 reduces amyloid patholo-
the beta protein is critical in amyloidogenesis. Ann N Y gy in brains of human beta-amyloid precursor protein
Acad Sci 695: 144-148. transgenic mice. Neurobiol Aging 25: 1315-1321.
Jeffrey M, Goodbrand IA, Goodsir CM. 1995. Patho- LeVine H. 2002. The challenge of inhibiting Abeta polymeri-
logy of the transmissible spongiform encephalopathies zation. Curr Med Chem 9: 1121-1133.
with pecial emphasis on ultrastructure. Micron 26: Leyssen M, Ayaz D, Hebert SS, Reeve S, De Strooper B,
277-298. et al. 2005. Amyloid precursor protein promotes post-
Johnson RT, Gibbs CJ, Jr. 1998. Creutzfeldt-Jakob disease and developmental neurite arborization in the Drosophila
related transmissible spongiform encephalopathies. N Engl brain. EMBO J 24: 2944-2955.
J Med 339: 1994-2004. Lindquist S. 1997. Mad cows meet psi-chotic yeast: The
Kayed R, Glabe CG. 2006. Conformation-dependent anti- expansion of the prion hypothesis. Cell 89: 495-498.
amyloid oligomer antibodies. Methods Enzymol 413: Link CD. 1995. Expression of human beta-amyloid peptide in
326-344. transgenic Caenorhabditis elegans. Proc Natl Acad Sci USA
Kelly JW. 1996. Alternative conformations of amyloidogenic 92: 9368-9372.
proteins govern their behavior. Curr Opin Struct Biol 6: Link CD. 2006. C. elegans models of age-associated neurode-
11-17. generative diseases: Lessons from transgenic worm models
Kenney JM, Knight D, Wise MJ, Vollrath F. 2002. Amyloi- of Alzheimers disease. Exp Gerontol 41: 1007-1013.
dogenic nature of spider silk. Eur J Biochem 269: MacDonald ST, Sutherland K, Ironside JW. 1996. Prion
4159-4163. protein genotype and pathological phenotype studies in
Kisilevsky R, Lemieux LJ, Fraser PE, Kong X, Hultin PG, et al. sporadic Creutzfeldt-Jakob disease. Neuropathol Appl
1995. Arresting amyloidosis in vivo using small-molecule Neurobiol 22: 285-292.
anionic sulphonates or sulphates: Implications for Alzhei- Magri G, Clerici M, DallAra P, Biasin M, Caramelli M, et al.
mers disease. Nat Med 1: 143-148. 2005. Decrease in pathology and progression of scrapie
Klement IA, Skinner PJ, Kaytor MD, Yi H, Hersch SM, et al. after immunisation with synthetic prion protein peptides
1998. Ataxin-1 nuclear localization and aggregation: Role in hamsters. Vaccine 23: 2862-2868.
in polyglutamine-induced disease in SCA1 transgenic mice. Makin OS, Atkins E, Sikorski P, Johansson J, Serpell LC. 2005.
Cell 95: 41-53. Molecular basis for amyloid fibril formation and stability.
Kosik KS, Joachim CL, Selkoe DJ. 1986. Microtubule- Proc Natl Acad Sci USA 102: 315-320.
associated protein tau (tau) is a major antigenic compo- Makin OS, Serpell LC. 2005. Structures for amyloid fibrils.
nent of paired helical filaments in Alzheimer disease. Proc FEBS J 272: 5950-5961.
Natl Acad Sci USA 83: 4044-4048. Marsh JL, Thompson LM. 2006. Drosophila in the study of
Kovacs GG, Trabattoni G, Hainfellner JA, Ironside JW, Knight neurodegenerative disease. Neuron 52: 169-178.
RS, et al. 2002. Mutations of the prion protein gene pheno- Martin JB. 1999. Molecular basis of the neurodegenerative
typic spectrum. J Neurol 249: 1567-1582. disorders. New Engl J Med 340: 1970-1980.
302 12 Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases
Masliah E, Rockenstein E, Veinbergs I, Mallory M, Hashimoto Ono K, Hirohata M, Yamada M. 2005. Ferulic acid destabi-
M, et al. 2000. Dopaminergic loss and inclusion body lizes preformed beta-amyloid fibrils in vitro. Biochem
formation in alpha-synuclein mice: Implications for neu- Biophys Res Commun 336: 444-449.
rodegenerative disorders. Science 287: 1265-1269. Pasinelli P, Brown RH. 2006. Molecular biology of amyo-
Mattson MP, Cheng B, Culwell AR, Esch FS, Lieberburg I, trophic lateral sclerosis: Insights from genetics. Nat Rev
et al. 1993. Evidence for excitoprotective and intraneuronal Neurosci 7: 710-723.
calcium-regulating roles for secreted forms of the beta- Pepys MB, Herbert J, Hutchinson WL, Tennent GA,
amyloid precursor protein. Neuron 10: 243-254. Lachmann HJ, et al. 2002. Targeted pharmacological
Maynard CJ, Bush AI, Masters CL, Cappai R, Li QX. 2005. depletion of serum amyloid P component for treatment
Metals and amyloid-beta in Alzheimers disease. Int J Exp of human amyloidosis. Nature 417: 254-259.
Pathol 86: 147-159. Permanne B, Adessi C, Saborio GP, Fraga S, Frossard MJ, et al.
McGavern DB. 2006. Immunotherapeutic relief from persis- 2002. Reduction of amyloid load and cerebral damage in a
tent infections and amyloid disorders. Neurology 66: transgenic mouse model of Alzheimers disease by treatment
S59-S64. with a beta-sheet breaker peptide. FASEB J 16: 860-863.
McGeer EG, McGeer PL, McGeer EG, McGeer PL. 2005. Perutz MF, Johnson T, Suzuki M, Finch JT. 1994. Glutamine
Abeta immunotherapy and other means to remove amy- repeats as polar zippers: Their possible role in inherited
loid. Curr Drug Targets CNS Neurol Disord 4: 569-573. neurodegenerative diseases. Proc Natl Acad Sci USA 91:
Meriin AB, Zhang X, He X, Newnam GP, Chernoff YO, et al. 5355-5358.
2002. Huntington toxicity in yeast model depends on poly- Price DL, Koliatsos VE, Wong PC, Pardo CA, Borchelt DR,
glutamine aggregation mediated by a prion-like protein et al. 1996. Motor neuron disease and model systems:
Rnq1. J Cell Biol 157: 997-1004. Aetiologies, mechanisms and therapies. Ciba Found Symp
Miyasaka T, Ding Z, Gengyo-Ando K, Oue M, Yamaguchi H, 196: 3-13.
et al. 2005. Progressive neurodegeneration in C. elegans Price DL, Sisodia SS, Borchelt DR. 1998. Genetic neurodegen-
model of tauopathy. Neurobiol Dis 20: 372-383. erative diseases: The human illness and transgenic models.
Moechars D, Dewachter I, Lorent K, Reverse D, Baekelandt V, Science 282: 1079-1083.
et al. 1999. Early phenotypic changes in transgenic mice Price DL, Wong PC, Borchelt DR, Pardo CA, Thinakaran G,
that overexpress different mutants of amyloid precursor et al. 1997. Amyotrophic lateral sclerosis and Alzheimer
protein in brain. J Biol Chem 274: 6483-6492. disease. Lessons from model systems. Rev Neurol (Paris)
Morgan D, Diamond DM, Gottschall PE, Ugen KE, Dickey C, 153: 484-495.
et al. 2000. A beta peptide vaccination prevents memory Price DL, Wong PC, Markowska AL, Lee MK, Thinakaren G,
loss in an animal model of Alzheimers disease. Nature 408: et al. 2000. The value of transgenic models for the study
982-985. of neurodegenerative diseases. Ann NY Acad Sci 920:
Morley JF, Brignull HR, Weyers JJ, Morimoto RI. 2002. The 179-191.
threshold for polyglutamine-expansion protein aggrega- Prusiner SB. 1998. Prions. Proc Natl Acad Sci USA 95:
tion and cellular toxicity is dynamic and influenced by 13363-13383.
aging in Caenorhabditis elegans. Proc Natl Acad Sci USA Prusiner SB, Scott MR. 1997. Genetics of prions. Annu Rev
99: 10417-10422. Genet 31: 139-175.
Nelson R, Eisenberg D. 2006. Recent atomic models Qiu WQ, Ferreira A, Miller C, Koo EH, Selkoe DJ. 1995. Cell-
of amyloid fibril structure. Curr Opin Struct Biol 16: surface beta-amyloid precursor protein stimulates neurite
260-265. outgrowth of hippocampal neurons in an isoform-depen-
Nishitoh H, Matsuzawa A, Tobiume K, Saegusa K, Takeda K, dent manner. J Neurosci 15: 2157-2167.
et al. 2002. ASK1 is essential for endoplasmic reticulum Rebeck GW, Reiter JS, Strickland DK, Hyman BT. 1993.
stress-induced neuronal cell death triggered by expanded Apolipoprotein E in sporadic Alzheimers disease:
polyglutamine repeats. Gen Dev 16: 1345-1355. Allelic variation and receptor interactions. Neuron 11:
Nollen EA, Garcia SM, van Haaften G, Kim S, Chavez A, et al. 575-580.
2004. Genome-wide RNA interference screen identifies pre- Reiner A, Dragatsis I, Zeitlin S, Goldowitz D. 2003. Wild-type
viously undescribed regulators of polyglutamine aggrega- huntingtin plays a role in brain development and neuronal
tion. Proc Natl Acad Sci USA 101: 6403-6408. survival. Mol Neurobiol 28: 259-275.
Ono K, Hasegawa K, Yamada M, Naiki H. 2002. Nicotine Ross CA, Poirier MA. 2005. What is the role of protein
breaks down preformed Alzheimers beta-amyloid fibrils aggregation in neurodegeneration? Nat Rev Mol Cell Biol
in vitro. Biol Psychol 52: 880-886. 6: 891-898.
Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases 12 303
Sacchettini JC, Kelly JW. 2002. Therapeutic strategies for amyloid fibril by image reconstruction from electron
human amyloid diseases. Nat Rev Drug Discov 1: 267-275. micrographs. J Mol Biol 254: 113-118.
Saitoh T, Sundsmo M, Roch JM, Kimura N, Cole G, et al. Sherman MY, Goldberg AL. 2001. Cellular defenses against
1989. Secreted form of amyloid beta protein precursor is unfolded proteins: A cell biologist thinks about neurode-
involved in the growth regulation of fibroblasts. Cell 58: generative diseases. Neuron 29: 15-32.
615-622. Sigurdsson EM, Brown DR, Daniels M, Kascsak RJ, Kascsak R,
Salomon AR, Marcinowski KJ, Friedland RP, Zagorski MG. et al. 2002. Immunization delays the onset of prion disease
1996. Nicotine inhibits amyloid formation by the beta- in mice. Am J Pathol 161: 13-17.
peptide. Biochemistry 35: 13568-13578. Sipe JD. 1992. Amyloidosis. Annu Rev Biochem 61: 947-975.
Sathasivam K, Hobbs C, Mangiarini L, Mahal A, Turmaine M, Sisodia SS. 1992. Beta-amyloid precursor protein cleavage by
et al. 1999. Transgenic models of Huntingtons disease. a membrane-bound protease. Proc Natl Acad Sci USA 89:
Philos Trans R Soc Lond B Biol Sci 354: 963-969. 6075-6079.
Savitt JM, Dawson VL, Dawson TM. 2006. Diagnosis and Small DH, Cappai R. 2006. Alois Alzheimer and Alzheimers
treatment of Parkinson disease: Molecules to medicine. disease: A centennial perspective. J Neurochem 99: 708-710.
J Clin Invest 116: 1744-1754. Smith PG, Cousens SN, Huillard Aignaux JN, Ward HJ, Will
Sawaya MR, Sambashivan S, Nelson R, Ivanova MI, Sievers RG. 2004. The epidemiology of variant Creutzfeldt-Jakob
SA, et al. 2007. Atomic structures of amyloid cross-beta disease. Curr Top Microbiol Immunol 284: 161-191.
spines reveal varied steric zippers. Nature (In press). Smith R, Brundin P, Li JY. 2005. Synaptic dysfunction in
Schenk D. 2002. Amyloid-beta immunotherapy for Alzhei- Huntingtons disease: A new perspective. Cell Mol Life Sci
mers disease: The end of the beginning. Nat Rev Neurosci 62: 1901-1912.
3: 824-828. Smith RP, Higuchi DA, Broze GJ, Jr. 1990. Platelet coagulation
Schenk D, Barbour R, Dunn W, Gordon G, Grajeda H, et al. factor XIa-inhibitor, a form of Alzheimer amyloid precur-
1999. Immunization with amyloid-beta attenuates Alzhei- sor protein. Science 248: 1126-1128.
mer disease-like pathology in the PDAPP mouse. Nature Sontag E, Nunbhakdi-Craig V, Lee G, Brandt R, Kamibayashi
400: 173-177. C, et al. 1999. Molecular interactions among protein phos-
Scherzinger E, Lurz R, Turmaine M, Mangiarini L, Hollen- phatase 2A, tau, and microtubules. Implications for the
bach B, et al. 1997. Huntingtin-encoded polyglutamine regulation of tau phosphorylation and the development
expansions form amyloid-like protein aggregates in vitro of tauopathies. J Biol Chem 274: 25490-25498.
and in vivo. Cell 90: 549-558. Soto C. 1999. Plaque busters: Strategies to inhibit amyloid
Schubert D, Jin LW, Saitoh T, Cole G. 1989. The regulation of formation in Alzheimers disease. Mol Med Today 5:
amyloid beta protein precursor secretion and its modula- 343-350.
tory role in cell adhesion. Neuron 3: 689-694. Soto C. 2001. Protein misfolding and disease; protein refold-
Selkoe DJ. 1998. The cell biology of beta-amyloid precursor ing and therapy. FEBS Lett 498: 204-207.
protein and presenilin in Alzheimers disease. Trends Cell Soto C. 2003. Unfolding the role of protein misfolding in
Biol 8: 447-453. neurodegenerative diseases. Nat Rev Neurosci 4: 49-60.
Selkoe DJ. 2001. Alzheimers disease: Genes, proteins, and Soto C, Castano EM, Frangione B, Inestrosa NC. 1995. The
therapy. Physiol Rev 81: 741-766. alpha-helical to beta-strand transition in the amino-termi-
Selkoe DJ. 2004a. Alzheimer disease: Mechanistic understand- nal fragment of the amyloid beta-peptide modulates amy-
ing predicts novel therapies. Ann Int Med 140: 627-638. loid formation. J Biol Chem 270: 3063-3067.
Selkoe DJ. 2004b. Cell biology of protein misfolding: The Soto C, Castilla J. 2004. The controversial protein-only hy-
examples of Alzheimers and Parkinsons diseases. Nat Cell pothesis of prion propagation. Nat Med 10: S63-S67.
Biol 6: 1054-1061. Soto C, Estrada L. 2005. Amyloid inhibitors and b-sheet break-
Selkoe DJ, Podlisny MB, Selkoe DJ, Podlisny MB. 2002. Deci- ers. Alzheimers Disease: Cellular and Molecular Aspects
phering the genetic basis of Alzheimers disease. Annu Rev of Amyloid Beta. Harris R, Fahrenholz F, editors. Springer.
Genomics Hum Genet 3: 67-99. Soto C, Estrada L, Castilla J. 2006. Amyloids, prions and the
Serpell LC, Berriman J, Jakes R, Goedert M, Crowther RA. inherent infectious nature of misfolded protein aggregates.
2000. Fiber diffraction of synthetic alpha-synuclein fila- Trends Biochem Sci 31: 150-155.
ments shows amyloid-like cross-beta conformation. Proc Soto C, Kascsak RJ, Saborio GP, Aucouturier P, Wisniewski T,
Natl Acad Sci USA 97: 4897-4902. et al. 2000. Reversion of prion protein conformational
Serpell LC, Sunde M, Fraser PE, Luther PK, Morris EP, et al. changes by synthetic beta-sheet breaker peptides. Lancet
1995. Examination of the structure of the transthyretin 355: 192-197.
304 12 Protein misfolding, a common mechanism in the pathogenesis of neurodegenerative diseases
Soto C, Sigurdsson EM, Morelli L, Kumar RA, Castano EM, Wanker EE. 2000. Protein aggregation and pathogenesis of
et al. 1998. beta-Sheet breaker peptides inhibit fibrillogen- Huntingtons disease: Mechanisms and correlations. Biol
esis in a rat brain model of amyloidosis: Implications for Chem 381: 937-942.
Alzheimers therapy. Nat Med 4: 822-826. Weber T. 2000. Clinical and laboratory diagnosis of Creutz-
Spillantini MG, Goedert M. 1998. Tau protein pathology feldt-Jakob disease. Clin Neuropathol 19: 249-250.
in neurodegenerative diseases. Trends Neurosci 21: 428-433. Weissmann C, Fischer M, Raeber A, Bueler H, Sailer A, et al.
Spillantini MG, Schmidt ML, Lee VM, Trojanowski JQ, Jakes 1998. The use of transgenic mice in the investigation of
R, et al. 1997. Alpha-synuclein in Lewy bodies. Nature 388: transmissible spongiform encephalopathies. Rev Sci Tech
839-840. 17: 278-290.
Stefani M, Dobson CM. 2003. Protein aggregation and Wells GA. 1993. Pathology of nonhuman spongiform ence-
aggregate toxicity: New insights into protein folding, phalopathies: Variations and their implications for patho-
misfolding diseases and biological evolution. J Mol Med genesis. Dev Biol Stand 80: 61-69.
81: 678-699. Westermark P. 1995. Diagnosing amyloidosis. Scand J Rheu-
Tagliavini F, Prelli F, Verga L, Giaccone G, Sarma R, et al. 1993. matol 24: 327-329.
Synthetic peptides homologous to prion protein residues Wickner RB, Edskes HK, Maddelein ML, Taylor KL,
106147 form amyloid-like fibrils in vitro. Proc Natl Acad Moriyama H. 1999. Prions of yeast and fungi. Proteins as
Sci USA 90: 9678-9682. genetic material. J Biol Chem 274: 555-558.
Tatzelt J, Prusiner SB, Welch WJ. 1996. Chemical chaperones Wilkinson KD. 2000. Ubiquitination and deubiquitination:
interfere with the formation of scrapie prion protein. Targeting of proteins for degradation by the proteasome.
EMBO J 15: 6363-6373. Semin Cell Dev Biol 11: 141-148.
Teplow DB, Lazo ND, Bitan G, Bernstein S, Wyttenbach T, Will RG. 2003. Acquired prion disease: Iatrogenic CJD, vari-
et al. 2006. Elucidating amyloid beta-protein folding and ant CJD, kuru. Br Med Bull 66: 255-265.
assembly: A multidisciplinary approach. Acc Chem Res 39: Will RG, Ironside JW, Zeidler M, Cousens SN, Estibeiro K,
635-645. et al. 1996. A new variant of Creutzfeldt-Jakob disease in
Tom Van Dooren ID, Borghgraef P, Leuven FV (5 A.D.). Trans- the UK. Lancet 347: 921-925.
genic mouse models for APP processing and Alzheimers Will RG, Zeidler M, Stewart GE, Macleod MA, Ironside JW,
disease: Early and late defects. Alzheimers Disease: Cellular et al. 2000. Diagnosis of new variant Creutzfeldt-Jakob
and Molecular Aspects of Amyloid beta. Harris R, Fahren- disease. Ann Neurol 47: 575-582.
holz F, editors. Springer. Wolfe MS, Xia W, Ostaszewski BL, Diehl TS, Kimberly WT,
Tran PB, Miller RJ. 1999. Aggregates in neurodegenerative et al. 1999. Two transmembrane aspartates in presenilin-1
disease: Crowds and power? Trends Neurosci 22: 194-197. required for presenilin endoproteolysis and gamma-
Trojanowski JQ, Lee VMY. 2003. Parkinsons disease and secretase activity. Nature 398: 513-517.
related alpha-synucleinopathies are brain amyloidoses. Wood SJ, Wetzel R, Martin JD, Hurle MR. 1995. Prolines and
Ann N Y Acad Sci 991: 107-110. amyloidogenicity in fragments of the Alzheimers peptide
Tycko R. 2006. Molecular structure of amyloid fibrils: Insights beta/A4. Biochemistry 34: 724-730.
from solid-state NMR. Q Rev Biophys 39: 1-55. Wosten HA, de Vocht ML. 2000. Hydrophobins, the fungal
Ursini F, Davies KJ, Maiorino M, Parasassi T, Sevanian A. coat unravelled. Biochim Biophys Acta 1469: 79-86.
2002. Atherosclerosis: Another protein misfolding disease? Wu Y, Luo Y, Wu Y, Luo Y. 2005. Transgenic C. elegans as a
Trends Mol Med 8: 370-374. model in Alzheimers research. Curr Alzheimer Res 2:
Van Leuven F. 2000. Single and multiple transgenic mice as 37-45.
models for Alzheimers disease. Prog Neurobiol 61: Yamamoto M, Gotz ME, Ozawa H, Luckhaus C, Saito T, et al.
305-312. 2000. Hippocampal level of neural specific adenylyl cyclase
Villegas V, Zurdo J, Filimonov VV, Aviles FX, Dobson CM, type I is decreased in Alzheimers disease. Biochim Biophys
et al. 2000. Protein engineering as a strategy to avoid Acta 1535: 60-68.
formation of amyloid fibrils. Protein Sci 9: 1700-1708. Zheng H, Jiang M, Trumbauer ME, Sirinathsinghji DJ,
Vonsattel JP, DiFiglia M. 1998. Huntington disease. J Neuro- Hopkins R, et al. 1995. beta-Amyloid precursor protein-
pathol Exp Neurol 57: 369-384. deficient mice show reactive gliosis and decreased locomo-
Walter J, Capell A, Hung AY, Langen H, Schnolzer M, et al. tor activity. Cell 81: 525-531.
1997. Ectodomain phosphorylation of beta-amyloid pre- Ziegler J, Viehrig C, Geimer S, Rosch P, Schwarzinger S. 2006.
cursor protein at two distinct cellular locations. J Biol Putative aggregation initiation sites in prion protein. FEBS
Chem 272: 1896-1903. Lett 580: 2033-2040.
13 AntiAging Strategies
J. A. Joseph . J. R. PerezPolo
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
Abstract: There are age-associated motor and cognitive deficits, even in the absence of neurodegenerative
disease, that result from alterations in the striatal dopamine or cholinergic systems, respectively. In both
instances, oxidative stress and inflammation are contributing factors to the observed behavioral impair-
ments. Muscarinic acetycholine receptors (mAChR) play important roles in neuronal and vascular func-
tion. There is a loss of sensitivity in classic receptor systems that include the dopaminergic and muscarinic
systems. It appears that a major factor that may be important in the loss of sensitivity in mAChR with
aging may be increased sensitivity to oxidative stress that is at least partially determined by changes in
membranes. Also possibly contributing to the loss of cholinergic integrity and ultimately behavioral
function in aging are perturbations in ChAT function. There are fewer cholinergic neurons in the aged
basal forebrain that express lower levels of cholinergic activity. For example, decreased ACh synthesis in
cholinergic neurons of the aged rat basal forebrain may result in part from altered binding of NF-kB to a
cognate DNA-binding repressor sites within the ChAT promoter. Given that NF-kB DNA-binding activity is
elevated in the aged rat basal forebrain and hippocampus, the result is age-associated decreases in ChAT
expression that have consequences on cognitive function in the aged. Given the role of oxidative stress in
aging, Anti-oxidant based fruit and vegetable supplementation could forestall the consequences of oxidative
stress on behavioral and neuronal deficits associated with aging.
List of Abbreviations: BC, black currant; BY, boysenberry; CB, cranberry; CHL, cholesterol; DA,
dopamine; DP, dried plums; ERK, extracellular signalregulated kinase; GR, grape; MPT, mitochondrial
pore transition; ORAC, oxygenradical absorbance capacity assay; PKC, protein kinase Ca; RIP, receptor
interacting protein; SAM, Sadenosylmethionine; SB, strawberry; TNFR, tumor necrosis factor receptor;
TRAF2, TNFRassociated factor 2; VaD, vascular disease
1 Introduction
There is a plethora of research indicating the occurrence of numerous behavioral deficits during aging, even
in the absence of neurodegenerative disease. These decrements can be expressed, ultimately, as alterations in
both motor (Joseph et al., 1983; Kluger et al., 1997) and cognitive behaviors (Bartus, 1990). The alterations
in motor function may include decreases in balance, muscle strength, and coordination (Joseph et al.,
1983), while memory deficits are seen on cognitive tasks that require the use of spatial learning and memory
(Ingram et al., 1994; ShukittHale et al., 1998). Indeed, these characterizations have been supported by a
great deal of research both in animals (Bartus, 1990; Ingram et al., 1994; ShukittHale et al., 1998) and
humans (West, 1996; Muir, 1997). Agerelated deficits in motor performance are thought to be the result of
alterations in the striatal dopamine (DA) system as the striatum shows marked neurodegenerative changes
with age (Joseph, 1992) or in the cerebellum which also shows agerelated alterations (Bickford et al., 1992;
Bickford, 1993). Memory alterations appear to occur primarily in secondary memory systems and are
reflected in the storage of newly acquired information (Bartus et al., 1982; Joseph, 1992). It is thought that
the hippocampus mediates allocentric spatial navigation (i.e., place learning), and that the prefrontal cortex
is critical for acquiring the rules that govern performance in particular tasks (i.e., procedural knowledge),
while the dorsomedial striatum mediates egocentric spatial orientation (i.e., response and cue learning)
(McDonald and White, 1994; Zyzak et al., 1995; Devan et al., 1996). It appears that oxidative stress (OS)
(ShukittHale, 1999) and inflammation (HaussWegrzyniak et al., 1999, 2000) are contributing factors to
the behavioral decrements seen in aging.
There are numerous factors involved in mediating these behavioral deficits in aging that include
decrements in calcium buffering (Landfield and Eldridge, 1994), as well as declines in the sensitivity of
several receptor systems including the (1) adrenergic (Gould and Bickford, 1997) and (2) dopaminergic
(Joseph et al., 1978; Levine and Cepeda, 1998) systems. However, one of the most notable changes are those
involving the muscarinic cholinergic system. There is a long association between cognitive behavior and
cholinergic functioning in the literature (Wortwein et al., 1998), and there are many reports of significant
declines in cholinergic functioning in aging. While we realize that there is an involvement of the nicotinic
Antiaging strategies 13 307
cholinergic system in learning and memory that mediates cognition, those changes that have been observed
in the muscarinic system in aging and the possible mechanisms involved in these changes have not been
well defined.
Research indicates that muscarinic acetylcholine receptors (mAChRs) are intimately involved in various
aspects of both neuronal (e.g., amyloid precursor protein (APP) processing) (Rossner et al., 1998) and
vascular functioning (Elhusseiny et al., 1999). The finding concerned with mAChR/vascular interaction is
important, since it also appears that vascular disease and Alzheimers disease (AD) can occur in concert in
aging and that vascular disease may exacerbate cognitive dysfunction in AD (Morris et al., 2000; Deschamps
et al., 2001). In fact, data suggest that there may be complex interactions between vascular dementia (VaD)
and AD (Morris, 2000; Deschamps et al., 2001).
There is a long history of research showing a loss of sensitivity in classic receptor systems including the
dopaminergic and muscarinic (see Joseph et al., 1992). Numerous studies have shown that the PIlinked
muscarinic receptors (mAChRs) show deficits in signal transduction, which contributes to their loss in
sensitivity (Joseph and Roth, 1991). Joseph et al. (1991) have shown that mAChR agonist enhancement of
IP3 formations is reduced in striatal tissue obtained from senescent Wistar rats. Additional studies have
suggested that the site of the deficit in agonist responsiveness in mAChR in aging may be at the mAChR/G
protein interface (Joseph et al., 1988), which may involve deficits in mAChR/Gprotein coupling/uncou-
pling (Yamagami et al., 1992). This is an important consideration since research also indicates that
decrements in mAChR/Gprotein coupling/uncoupling, as assessed via carbacholstimulated GTPase activ-
ity, may also be greater in (Cutler et al., 1994) than that seen in aging. Additionally, there also appear to be
reductions in anterior, cortical PI levels, as well as a 70% loss of IP3 binding in the temporal cortex and
hippocampus (Young et al., 1988) and a decrease in the number of mAChRs that are coupled to G proteins
in the highaffinity states (Flynn et al., 1991).
The mechanisms involved in these deficits have yet to be determined, but such factors as decreases
in membrane fluidity may also be involved. Findings showed that when striatal tissue was incubated with
Sadenosylmethionine (SAM), a potent membranefluidizing agent, it increased the sensitivity of the
mAChR to agonist stimulation in striatal slices, while exposure of the slices to cholesterol (CHL), which
increases membrane viscosity, decreased this sensitivity (Joseph et al., 1991). There are decreases in
membrane fluidity in aging that involves the translocation of CHL to the outer leaflet of the membrane
where it is more readily oxidized. It also appears that sphingomyelin metabolites such as sphingosine1
phosphate may increase sensitivity to OS in aging. Recent studies have also suggested an involvement of
lipid rafts with OS sensitivity (Shen et al., 2004). The findings of Shen and coworkers (2004) indicate that
receptorinteracting protein (RIP) and tumor necrosis factor receptor (TNFR)associated factor
2 (TRAF2), which are two primary molecules involved TNF signaling, may be necessary for ROSinduced
cell death of TNFR1. The colocalization of RIP with a membrane lipid raft marker (GM1) suggested that
these rafts might participate in the formation of a complex between RIP and TRAF2 during OS. The study
suggests that this complex may be part of a novel signaling pathway that mediates oxidantinduced cell
death. Thus, it appears that a major factor that there may be important in the loss of sensitivity in mAChR
with aging may be increased sensitivity to OS, which is at least partially determined by changes in
membranes.
Another component of the cholinergic signaling system that is affected in AD pathophysiology is
regulation of the deposition of bamyloid (Ab) peptides due to increased APP processing, resulting in
increased levels of Ab. For example, AChE inhibitor treatments increase the secretion of nonamyloidogenic
processing products and decreases Ab formation. (Mori et al., 1995; Pakaski et al., 2001). There is also
in vivo evidence that APP processing underlies cholinergic regulation in the rat brain (Rossner et al., 1998).
When M1MAChRs are selectively stimulated in tissue slices, the nonamyloidogenic processing of
APP increases, suggesting that activation of the M2MAChR subtype results in a suppression of the
308 13 Antiaging strategies
nonamyloidogenic processing pathway (Farber et al., 1995). Furthermore, BACE1 expression is regulated
in different ways through MAChR. Since increased M1MAChR increases BACE1 expression but increased
M2MAChR decreases BACE1 expression (Zuchner et al., 2004), cholinergic function is important for APP
processing and AD pathophysiology. This is also consistent with the observation that there is demonstrable
downregulation of M2MAChR in those brain regions of AD patients which contain senile plaques (Boulay
et al., 1996; Lai et al., 2001; Apelt et al., 2003).
Several experiments conducted over the last several years have indicated that there are considerable parallels
between the loss of sensitivity seen in aging in muscarinic receptors and those seen following heavyparticle
irradiation. Thus, studies have shown that, just as was seen in aging, striatal tissue obtained from young rats
exposed to various doses of 56Fe irradiation (0.11 Gy) showed reductions in oxotremorine enhancement of
Kevoked dopamine release (KEDAR) (Joseph et al., 1992), as well as decreases in carbacholstimulated
IP3 activity (Joseph et al., 1993a) and carbacholstimulated GTPase activity, an indicator of receptor/
Gprotein coupling/uncoupling. Further studies showed that the deficit appeared to be at the muscarinic
receptor/Gprotein interface (Joseph et al., 1993a). As early as 1991, Stark postulated that certain free
radicals were able to gain access into the interior of cellular membranes and interact with the lipid matrix or
with membranebound proteins, altering their integrity. Importantly, in this regard, subsequent research
indicated that as was seen previously in aging, increasing membrane fluidity via exposure of striatal tissue
obtained from the irradiated animals to SAM reversed the irradiatedinduced deficits in KERDA ( Joseph
et al., 1999b).
Even more important are recent findings that suggest that there may be differences among MAChR
subtypes in response to OS in aging. Clearly, if this is the case, it might provide some explanation for
findings such as those showing that there is differential aging among various brain regions. Areas such as
the hippocampus (Nyakas et al., 1997; Kaufmann et al., 2001), cerebellum (Hartmann et al., 1996;
Kaufmann et al., 2001), and striatum (Joseph et al., 1996; Kaasinen et al., 2000) show profound alterations
in aging in such factors as morphology, electrophysiology, and receptor sensitivity. Interestingly, studies
that have examined regional differences in mAChR subtypes have shown that the mAChRs that show
increased vulnerability to OS appear to be localized in areas that show profound alterations in aging; these
include the dentate, striatum, and forebrain (e.g., see Levey, 1996 for review). Thus, we assessed differential
vulnerability to OS among the various mAChR subtypes by transfecting one of five MAChR subtypes
(M1M5) in COS7 cells. The results showed that when the cells were exposed to DA (Joseph et al., 2002) or
Ab (Joseph and Fisher, 2003), differences in OS sensitivity, expressed as a function of Ca2 buffering as assessed
by examining the ability of the cell to extrude or sequester Ca2 following oxotremorineinduced depolari-
zation, were observed. The COS7 cells transfected with M1, M2, or M4AChRs showed greater sensitivity to
DA or Abinduced disruptions in calcium buffering than those transfected with M3 or M5MAChRs. An
additional study ( Joseph et al., 2004) has suggested that the locus of the differential sensitivity to OS among
the various receptor subtypes may involve the i3 loop, and studies are being carried out to determine
possible differences among the i3 loops of these mAChR subtypes that could contribute to these differences.
Importantly, the findings with respect to COS7 cells suggest that, as mentioned above, brain areas
showing increased effects of aging contain high concentrations of M1, M2, or M4 receptors (Flynn and
Mash, 1993; Levey, 1994; Hersch et al., 1999). Additionally, it should be noted that the loss in calcium
buffering could lead to further decrements in cell functioning and OS, since the intracellular compartmen-
talization of high amounts of calcium (Brini, 2003) and mitochondrial pore transition (MPT) (Green and
Kroemer, 2004) can mediate both cell necrosis and apoptosis, with MPT possibly releasing cytochrome
c. Thus, in the case of aging, OS appears to impinge on systems that are already compromised in their
ability to regulate Ca2 flux. Additionally, it also appears that the membrane alterations and decrements
in calcium buffering may reduce the effectiveness of endogenous antioxidants in the aged organism
(Joseph et al., 1998a, 2001).
Antiaging strategies 13 309
Also possibly contributing to the loss of cholinergic integrity and ultimately behavioral function in aging
are perturbations in choline acetyltransferase (ChAT) function. There are fewer cholinergic neurons in the
aged basal forebrain, and they express lower levels of ChAT and ACh uptake than younger rats. These
differences are consistent with ageassociated increases in proapoptotic markers and decreases in ChAT
activity (Drachman and Leavitt, 1974; Davies and Maloney, 1976; Drachman and Sahakian, 1980; Bartus
et al., 1982; Whitehouse et al., 1982; Davies, 1985; Fischer et al., 1987; Williams et al., 1993; De Lacalle et al.,
1996; Clemens et al., 1997). For example, cholinergic deafferentations carried out by transection of the
fimbria fornix or selective killing of basal forebrain cholinergic neurons by immunolesion fail to increase
neurotrophin protein levels in aged rat brains (Whitehouse et al., 1982; Fischer et al., 1987; Wortwein et al.,
1998). Taken together, it would appear that there is an agingassociated impairment of stress response
mechanisms. Determining the different contributions of cholinergic cell losses versus changes in the
cholinergic phenotype in the aged brain remains a disputed area.
For example, it could be that the activities of transcription factors that regulate gene expression
responsible for cholinergic function and apoptotic responses to OS and trauma are altered in the aged
brain (Tong et al., 2002). Although the activation of the transcription factor NFkB has been determined in
the aged rat and AD brains, the assessments have been limited to binding to oligonucleotide preparations
displaying the IgGkB promoter consensus sequence and have not differentiated among the specific
sequences present in individual promoters.
Choline acetyltransferase is rate limiting for acetylcholine synthesis (Nachmansohn and Machado,
1943). ChAT is expressed in basal forebrain cholinergic neurons, important in memory and cognition,
whose deterioration is associated with cognitive deficits. Decreases in ChAT and acetylcholine release are
pathological markers in AD and agingassociated dementias (Armstrong et al., 1983; Davies, 1985; Williams
et al., 1993). The cDNAs and regulatory regions for mouse, rat (Ishii et al., 1990), human (Toussaint et al.,
1992), and porcine (Berrard et al., 1987) ChATs have been cloned. They are highly homologous, and while
the 50 flanking sequences have greater interspecies variation, they share common regulatory elements, such
as a silencer element restricting expression to cholinergic neurons (Lonnerberg et al., 1996). ChAT
promoters also have a general enhancer and an NGFresponsive enhancer (Bejanin et al., 1992; Pu et al.,
1993). The mouse ChAT promoter 50 flanking sequence is alternatively spliced to yield seven different
mRNAs (R14, N12, and the most predominant M form) that differ in their 50 noncoding regions
(Misawa et al., 1992). Misawa and coworkers (1992) cloned a 4,014 base pair (bp) region of the promoter,
including the 1,470 nucleotides reported by Pu and coworkers (1993). Here, promoter positions are
numerically referenced as described by Misawa and coworkers (1993). The 50 flanking region lacks a
consensus TATA box, but contains TATAlike elements and multiple transcription initiation start sites.
There is a strong enhancer element that is not modulated by NGF (3021 to 2706) and an NGFinducible
enhancer (3416 to 3021). Region 3516 to 3148 confers celltype specificity, acting as a repressor
element in noncholinergic cells (Pu et al., 1993). Although there are AP1 (2934 to 2928 and 2858 to
2852), serum response element (3159 to 3145), cAMP response element (3083 to 3076), NGFIA
(3190 to 3182), and NFkB (3174 to 3165) consensus sites within the NGFresponsive element, it is not
known if the cognate transcription factors regulate ChAT. Transcription of Rtype mRNAs is controlled by
regulatory elements upstream of the R promoter, and transcription of N and Mtype mRNAs by elements
surrounding the N and M promoters. The most common form of ChAT mRNA, which is present in the
brain and relevant to this project, is the Mform. Misawa and coworkers (1992) cloned a 4014bp flanking
region, which includes the N and Mtype exons and the 1470 nucleotides reported by Pu and coworkers
(1993); numbering of mouse ChAT promoter nucleotide positions will be as in Misawa et al. (1992).
The transcription factor NFkB belongs to a family of homo and heterodimeric proteins that are
related by a conserved 300 residue NH2 terminal Rel/homology domain and includes p50, p65 (or RelA),
p52 (or p49), cRel, and RelB proteins (Bours et al., 1993; Nolan et al., 1993; Ruben et al., 1991; Schmid
et al., 1991; Bours et al., 1993). Heterodimerization of NFkB proteins produces species with different
intrinsic DNAbinding specificities (p52, p50) and transactivation properties (p65, cRel), an important
310 13 Antiaging strategies
functional distinction (Baeuerle, 1991; Siebenlist et al., 1994; Chen et al., 1998). The most commonly
described active subunit combination (Flohe et al., 1997) is the p50/p65 heterodimer, present in the
cytoplasm, bound to an inhibitory IkB subunit. There are several inhibitory IkB proteins (IkBa, IkBb,
IkBe, IkBg, and Bcl3). Upon stimulation, IkBa is phosphorylated, ubiquitinated, and degraded (Ghosh
and Baltimore, 1990; Liu et al., 1993), exposing nuclear localization signals on NFkB proteins, which allow
translocation to the nucleus for DNA binding. NFkB is stimulated by OS or receptor ligands via increased
IkBa degradation and NFkB nuclear translocation, but also via an independent pathway involving Bcl3
(Gozal et al., 1998; Zhang et al., 1998; Qiu et al., 2001). Bcl3, of the IkB family, functions in the nucleus by
drawing NFkB p50 or p52 homodimers away from NFkBbinding sites in promoters (Bours et al., 1993;
Franzoso et al., 1993; Fujita et al., 1993; Nolan et al., 1993; Bundy and McKeithan, 1997). Thus, Bcl3 can
inhibit p50 or p52 homodimer binding to promoter sites (Franzoso et al., 1993; Nolan et al., 1993;
Siebenlist et al., 1994; Zhang et al., 1998), allowing cRel/p50 or p65/p50 binding and activation (Heissmeyer
et al., 1999; Qiu et al., 2001).
We and others have shown that NFkB DNAbinding activity is elevated in the aged rat basal forebrain
(ToliverKinsky et al., 1997; 2000). Our hypothesis is that ageassociated alterations in these two regulatory
factors contribute to the ageassociated decreases in ChAT expression, which has consequences on cognitive
function in the aged. We have shown that a 20bp ChAT (NFkB) oligonucleotide probe containing the
ChAT (NFkB) site binds to NFkB p49, somewhat less to p65, but not at all to p50, cRel, or RelB. This
observation supports the idea that different consensus sequences are regulated by different subunits in a
genespecific and tissuespecific manner. When PC12 cells are transfected with two ChAT promoter
reporter constructs containing the wildtype ChAT promoter and a mutant construct where the NFkB
binding site is replaced with a Bgl2 restriction site, there was a threefold increase in promoter activity in
the cultures transfected with the mutant reporter plasmid as compared with those transfected with the
wildtype plasmid, consistent with the assignment of a repressor function to the (ChAT) NFkB site.
Furthermore, the NFkB proteins p49, p65, and p50 were overexpressed in NGFtreated PC12 cells, and
ChAT promoter and enzymatic activities were measured. The overexpression of p49 and of p49 and p65
significantly decreased total ChAT activity as compared with controls. The assignment of a repressor role to
NFkB on the ChAT promoter agrees with reports of increased NFkB binding and decreased ChAT in aged
and AD brains (Coyle et al., 1983; Vile et al., 1995; Helenius et al., 1996; Taglialatela et al., 1996; Terai et al.,
1996; Boissiere et al., 1997a, b; Kitamura et al., 1997). We measured NFkB activity by EMSA and Western
blots on nuclear fractions from brains of 3 and 30monthold male F344 BN hybrid rats. p49 appeared to
be higher in the basal forebrain nuclear extracts prepared from 30monthold rats.
Thus, NFkB DNAbinding activity increases in the aged basal forebrain where it may contribute to
decreased ChAT expression. Interestingly, this increase supports the hypothesis that changes in NFkB are
gradual and proportional to the aging process.
We have attempted to outline the possible role of OS in mediating decrements in cholinergic functioning
in aging. It is evident that methods to reduce stress may forestall or even reduce these deficits in the
aged organism. While there are numerous studies suggesting that various antioxidant supplements (see
Casadesus et al., 2002 for review) may be effective in this regard, our research suggests that the combina-
tions of antioxidant/antiinflammatory polyphenolics found in fruits and vegetables may show efficacy in
aging. All plants, including fruit or vegetable bearing plants, synthesize a vast array of chemical compounds
that are not necessarily involved in the plants metabolism. These secondary compounds instead serve a
variety of functions that serve to enhance the plants survivability. These compounds may be responsible for
the putative multitude of beneficial effects of fruits and vegetables on healthrelated issues, two of the most
important of which may be their antioxidant and antiinflammatory properties.
The anthocyanins are among the plant polyphenols that have potent antioxidant and antiinflammatory
activities. These are natural pigments responsible for the orange, red, and blue colors of fruits, flowers,
Antiaging strategies 13 311
vegetables, and other storage tissues in plants (Wang et al., 1999; Seeram et al., 2001a, b). Anthocyanins
have been reported to affect many of the parameters discussed above by inhibiting lipid peroxidation and
the activity of cycloxygenase1 and 2 (COX) enzymes (Seeram et al., 2001b, 2003).
The chemistry of the anthocyanins can be reduced to six major anthocyanidins: delphinidin, cyanidin,
pelargonidin, petunidin, peonidin, and malvidin. Among berry fruits, blueberries contain high levels of a
wide variety of anthocyanins including glycosides of four of the six major anthocyanidins: malvidin,
petunidin, peonidin, and cyanidin (Kalt et al., 1999). Therefore, blueberries may have very potent effects
on the fires of aging.
Anthocyanins are a subset of a larger class of polyphenols known as flavonoids. Over 4000 flavonoids
have been identified in plants. They are also abundant in seeds, fruits, and plantderived oils such as olive
oils, as well as tea and red wine. Thus, they are part of the human diet, and plants and spices containing
them have been used for many years in Eastern medicine. As might be expected from the above discussion
of the anthocyanins, flavonoids have been reported to inhibit lipid peroxidation in several biological
systems including mitochondria and microsomes (Bindoli et al., 1977; Cavallini et al., 1978) as well as
erythrocytes (Sorata et al., 1984; MaridonneauParini et al., 1986) and liver (Kimura et al., 1984). They
appear to be potent inhibitors of both NADPH and CCl4induced lipid peroxidation (Afanasev et al.,
1989). It appears that the ironchelating ability of the flavonoids may be very important in mediating their
potent inhibitory effects on 5LOX (Hoult et al., 1994), while CO inhibition appears to involve other
mechanisms.
The antioxidant effects of flavonoids may be derived in part from their ability to upregulate antioxidant
enzymes (e.g., glutathione) or enzymes related to glutathione synthesis. One mechanism that may be opera-
tional in these beneficial effects may be the direct enhancement of transcription factors that enhance antioxi-
dant enzymes or their signaling cascades. It is known, for example, that the enzymes for glutathione (reviewed
in Zipper and Mulcahy, 2000; Schroeter et al., 2002) or heme oxygenase (Chen and Maines, 2000) synthesis
exhibit extracellular signalregulated kinase 1/2 (ERK1/2) dependency in the regulation of their expression,
while Cu/ZnSOD is regulated by ELK1 (Chang et al., 1999), and MnSOD expression contains binding sites for
Sp1, AP1, and CREB (Das et al., 1995; Chang et al., 1999), which are ERK1/2 (Sgambato et al., 1998; Chang
and Karin, 2001) regulated. Finally, it also appears that flavonoids regulating ERK1/2 may influence iNOS
activity. Thus, there is a great deal of evidence to suggest that a possible link exists between the antioxidant
activity of flavonoids and their putative MAPkinases altering activity.
Since MAPKs are involved in numerous biological activities, the findings that flavonoids may influence
such signaling suggests that their potential benefits may involve properties other than those involving
antioxidant or antiinflammatory effects. As examples, delphinidin inhibits endothelial cell proliferation and
cellcycle progression by ERK1/2 activation (Martin et al., 2003), while grape seed proanthocyanidin can
reduce ischemia/reperfusioninduced activation of JNK1 and cJun and reduce cardiomyocyte apoptosis
(Sato et al., 2001). Additional research indicates that phytochemicals can regulate MAP kinase and other
signaling pathways at the level of transcription (Frigo et al., 2002).
These findings, coupled with a plethora of studies showing the involvement of ERK in diverse forms
of memory, such as contextual fear conditioning (English and Sweatt, 1996), longterm potentiation
(English and Sweatt, 1997), striataldependent learning and memory (Mazzucchelli and Brambilla, 2000),
hippocampaldependent spatial memory (Selcher et al., 1999), and inhibitory avoidance (Schafe et al.,
1999), suggest that interventions that influence MAPK signaling may have beneficial effects on cognition.
Given the findings reviewed above showing alterations in signaling as a function of age, the putative
signaling modifying properties of flavonoids may prove to be invaluable in altering the neuronal
and behavioral effects of aging. Thus, we believed that given multiple properties of fruits and vegetables,
they might show considerable efficacy in reducing the deleterious effects of aging on neuronal function
and behavior.
In the mAChRtransfected COS7 cells, research showed that M1MAChR transfected cells that had
been pretreated with various high antioxidants (boysenberry (BY); cranberry (CB); black currant (BC);
strawberry (SB); dried plums (DP); and grape (GR)) and exposed to Ab2535 or DA. Calcium buffering was
examined following oxotremorineinduced depolarization in M1MAChRtransfected COS7 cells, and on
cell viability following DA (4 h) exposure. The results suggested that, while there were differential levels of
312 13 Antiaging strategies
protection among the various fruit extracts, all of them decreased to varying degrees, the decrements in
calcium buffering induced by Ab or DA (Joseph et al., 2004).
Extending these findings, additional research showed that feeding high antioxidants (via the oxygen
radical absorbance capacity assay (ORAC)) (Cao et al., 1997; Wang et al., 1996; Prior et al., 1998) could both
prevent the onset of agerelated deficits in several indices (e.g., cognitive behavior, Morris water maze
performance) (Joseph et al., 1998) or reverse them (Joseph et al., 1999).
Interestingly, examinations of various brain regions from the supplemented animals showed only small
levels of antioxidant activity, which was insufficient to account for the observed significant beneficial effects
of blueberry (BB) supplementation on motor and cognitive function. Findings from this (Joseph et al.,
1999) and a subsequent study (Youdim et al., 2000) suggested that there are beneficial properties, in
addition to those involving antioxidants, of BBs on both motor and cognitive behavior; these may involve
alterations in neuronal signaling and communication.
This was observed recently in a study (Joseph et al., 2003) carried out in APP/PS1 transgenic mice,
which serve as a model for AD since these mutations promote the production of Ab, and subsequently
Alzheimerlike plaques in several brain regions, which are accompanied in middle age by cognitive
deficits. A group of these mice was given BB supplementation beginning at four months of age (as in
Joseph et al., 1999) and continued until they were 12 months of age, when their performance was tested in a
Ymaze. The results indicated that mice supplemented with BB exhibited Ymaze performance that was
similar to those seen in nontransgenic mice and significantly greater than that seen in the nonsupplemented
transgenic animals. Interestingly, there was a dichotomy between the plaque burden and behavior in the
BBsupplemented transgenic mice. No differences between the supplemented and nonsupplemented APP/
PS1 mice in the number of plaques were observed, even though behavioral declines were prevented in the
BBsupplemented animals.
One possible reason that the behavior did not reflect the morphology may be that there was enhanced
signaling present in the BBsupplemented transgenic mice, which acted to prevent or circumvent any
putative deleterious effects of the amyloid plaques on behavior. The evidence for this possibility is provided
by data showing that the BBsupplemented APP/PS1 mice exhibited greater levels of hippocampal ERK as
well as striatal and hippocampal protein kinase Ca (PKC) than that seen in the transgenic mice maintained
on the control diet. As pointed out above, ERK and PKC have been shown to be important in mediating
cognitive function, especially conversion of shortterm to longterm memory (Micheau and Riedel, 1999).
Enhancement was also seen in the BBsupplemented group in the sensitivity of muscarinic receptors
(i.e., increasing striatal, carbacholstimulated GTPase activity), which has been found to be associated
with learning and memory in numerous studies.
These findings, combined with additional preliminary research showing that BB supplementation in
addition to altering ERK activity may also increase hippocampal neurogenesis (Casadesus et al., 2004),
suggests that at least part of the efficacy of the BB supplementation may be to enhance neuronal function in
areas of the brain affected by aging or disease. This would allow more effective intra and interarea
communication and ultimately facilitate both cognitive and motor function.
References
Afanasev IB, Dorozhko AI, Brodskii AV, Kostyuk VA, Armstrong DA, Saper CB, Levey AI, Wainer BH, Terry RD.
Potapovitch AI. 1989. Chelating and free radical scav- 1983. Distribution of cholinergic neurons in rat brain: Dem-
enging mechanisms of inhibitory action of rutin and quer- onstration by immunocytochemical localization of choliner-
cetin in lipid peroxidation. Biochem Pharmacol 38(11): gic acetyltransferase. J Comp Neurol 216: 1184-1190.
1763-1769. Baeuerle PA. 1991. The inducible transcription activator
Apelt J, Kumar A, Schliebs R. 2003. Impairment of NFkB: Regulation by distinct protein subunits. Biochem-
cholinergic neurotransmission in adult and aged trans- ica et Biophysica Acta 1072: 63-80.
genic Tg2576 mouse brain expressing the Swedish muta- Bartus RT. 1990. Drugs to treat agerelated neurodegenera-
tion of human bamyloid precursor protein. Brain Res 953: tive problems. The final frontier of medical science? J Am
17-30. Geriatr Soc 38: 680-695.
Antiaging strategies 13 313
Bartus RT, Dean RL, Beer B, Lippa AS. 1982. The cholinergic and cognitive behavior by shortterm blueberry supple-
hypothesis of geriatric memory dysfunction. Science 271: mentation in aged rats. Nutr Neurosci 7: 309-316.
408-417. Cavallini L, Bindoli A, Siliprandi N. 1978. Comparative eval-
Bejanin S, Habert E, Berrard S, Edwards JB, Loeffler JP, et al. uation of antiperoxidative action of silymarin and other
1992. Promoter elements of the rat choline acetyltrans- flavonoids. Pharmacol Res Commun 10(2): 133-136.
ferase gene allowing nerve growth factor inducibility in Chang L, Karin M. 2001. Mammalian MAP kinase signalling
transfected primary cultured cells. J Neurochem 58(4): cascades. Nature 410(6824): 37-40.
1580-1583. Chang MS, Yoo HY, Rho HM. 1999. Positive and negative
Berrard S, Brice A, Lottspeich F, Braun A, Barde YA, et al. regulatory elements in the upstream region of the rat
1987. cDNA cloning and complete sequence of porcine Cu/Znsuperoxide dismutase gene. Biochem J 339 (Pt 2):
choline acetyltransferase: In vitro translation of the 335-341.
corresponding RNA yields an active protein. Proc Natl Chen FE, Huang DB, Chen YQ, Ghosh G. 1998. Crystal
Acad Sci USA 84(24): 9280-9284. structure of p50/p65 heterodimer of transcription factor
Bickford P. 1993. Motor learning deficits in aged rats are NFkB bound to DNA. Nature 391: 410-413.
correlated with loss of cerebellar noradrenergic function. Chen K, Maines MD. 2000. Nitric oxide induces heme oxyge-
Brain Res 620: 133-138. nase1 via mitogenactivated protein kinases ERK and p38.
Bickford P, Heron C, Young DA, Gerhardt GA, De La Garza R. Cell Mol Biol 46(3): 609-617.
1992. Impaired acquisition of novel locomotor tasks in Clemens JA, Stephenson DT, Smalstig EB, Dixon E, Little SP.
aged and norepinephrinedepleted F344 rats. Neurobiol 1997. Global ischemia activates nuclear factor kB in fore-
Aging 13: 475-481. brain neurons of rats. Stroke 28: 1073-1080.
Bindoli A, Cavallini L, Siliprandi N. 1977. Inhibitory action of Coyle JT, Price PH, De Long MR. 1983. Alzheimers disease: A
silymarin of lipid peroxide formation in rat liver mitochon- disorder of cortical cholinergic innervation. Science 219:
dria and microsomes. Biochem Pharmacol 26: 2405-2409. 1184-1189.
Boissiere F, Faucheux B, Agid Y, Hirsch EC. 1997a. Choline Cutler R, Joseph JA, Yamagami K, VillalobosMolina R, Roth
acetyltransferase mRNA expression in the striatal neurons of GS. 1994. Area specific alterations in muscarinic stimu-
patients with Alzheimers disease. Neurosci Lett 225: 169-172. lated low Km GTPase activity in aging and Alzheimers
Boissiere F, Hunot S, Faucheux B, Duyckaerts C, Hauw JJ, disease: Implications for altered signal transduction. Brain
et al. 1997b. Nuclear translocation of NFkB in cholinergic Res 664: 54-60.
neurons of patients with Alzheimersdisease. Neuroreport Das KC, LewisMolock Y, White CW. 1995. Activation of
8(13): 2849-2852. NFkB and elevation of MnSOD gene expression by thiol
Boulay SF, Sood VK, Rayeq MR, Cohen VI, Zeeberg BR, et al. reducing agents in lung adenocarcinoma (A549) cells. Am J
1996. Autoradiographic evidence that 3quinuclidinyl4 Physiol 269 (5 Pt 1): L588-L602.
fluorobenzilate (FQNB) displays in vivo selectivity for the Davies P. 1985. A critical review of the role of the cholinergic
M2 subtype. Neuroimage 3: 35-39. system in human memory and cognition. Ann N Y Acad Sci
Bours V, Franzoso G, Azarenko V, Park S, Kanno T, et al. 1993. 444: 212-217.
The oncoprotein Bcl3 directly transactivates through kB Davies P, Maloney AJF. 1976. Selective loss of central cholin-
motifs via association with DNAbinding p50B homodi- ergic neurons in Alzheimers disease. Lancet 2: 1403.
mers. Cell 72: 729-739. De Lacalle S, Cooper JD, Svendsen CN, Dunnett SB,
Brini M. 2003. Ca2 signalling in mitochondria: Mechanism Sofroniew MV. 1996. Reduced retrograde labeling with
and role in physiology and pathology. Cell Calcium fluorescent tracer accompanies neuronal atrophy of basal
34(45): 399-405. forebrain cholinergic neurons in aged rats. Neuroscience
Bundy DL, McKeithan TW. 1997. Diverse effects of BCL3 75(1): 19-27.
phosphorylation on its modulation of NFkB p52 homo- Deschamps V, BarbergerGateau P, Peuchant E, Orgogozo JM.
dimer binding to DNA. J Biol Chem 272: 33132-33139. 2001. Nutritional factors in cerebral aging and dementia:
Cao G, Sofic E, Prior RL. 1997. Antioxidant and prooxidant Epidemiological arguments for a role of oxidative stress.
behavior of flavonoids: Structureactivity relationships. Neuroepidemiology 20: 7-15.
Free Radic Biol Med 22(5): 749-760. Devan BD, Goad EH, Petri HL. 1996. Dissociation of
Casadesus G, ShukittHale B, Joseph JA. 2002. Qualitative hippocampal and striatal contributions to spatial navi-
versus quantitative caloric intake: Are they equivalent gation in the water maze. Neurobiol Learn Mem 66:
paths to successful aging. Neurobiol Aging 23: 747-769. 305-323.
Casadesus G, ShukittHale B, Stellwagen HM, Zhu X, Lee Drachman DA, Leavitt J. 1974. Human memory and the
HG, et al. 2004. Modulation of hippocampal plasticity cholinergic system. Arch Neurol 30: 113-121.
314 13 Antiaging strategies
Drachman DA, Sahakian BJ. 1980. Memory and cognitive Green DR, Kroemer G. 2004. The pathophysiology of mito-
function in the elderly. A preliminary trial of physostig- chondrial cell death. Science 305(5684): 626-629.
mine. Arch Neurol 37: 674-675. Hartmann H, Velbinger K, Eckert A, Muller WE. 1996.
Elhusseiny A, Cohen Z, Olivier A, Stanimirovic DB, Hamel E. Regionspecific downregulation of free intracellular calci-
1999. Functional acetylcholine muscarinic receptor sub- um in the aged rat brain. Neurobiol Aging 17: 557-563.
types in human brain microcirculation: Identification and HaussWegrzyniak B, Vannucchi MG, Wenk GL. 2000. Behav-
cellular localization. J Cereb Blood Flow Metab 19: 794-802. ioral and ultrastructural changes induced by chronic neu-
English JD, Sweatt JD. 1996. Activation of p42 mitogen roinflammation in young rats. Brain Res 859: 157-166.
activated protein kinase in hippocampal long term poten- HaussWegrzyniak B, Vraniak P, Wenk GL. 1999. The effects
tiation. J Biol Chem 271(40): 24329-24332. of a novel NSAID on chronic neuroinflammation are age
English JD, Sweatt JD. 1997. A requirement for the mitogen dependent. Neurobiol Aging 20: 305-313.
activated protein kinase cascade in hippocampal long term Heissmeyer V, Krappmann D, Wulczyn FG, Scheidereit C.
potentiation. J Biol Chem 272(31): 19103-19106. 1999. NFkB p105 is a target of IkB kinases and controls
Farber SA, Nitsch RM, Schulz JG, Wurtman RJ. 1995. Regu- signal induction of Bcl3p50 complexes. EMBO J 18: 4766-
lated secretion of bamyloid precursor protein in rat brain. 4778.
J Neurosci 15: 7442-7451. Helenius M, Hanninen M, Lehtinen SK, Salminen A. 1996.
Fischer W, Wictorin K, Bjorklund A, Williams LR, Varon S, Changes associated with aging and replicative senescence in
et al. 1987. Amelioration of cholinergic neuron atrophy and the regulation of transcription factor nuclear factorkB.
spatial memory impairment in aged rats by nerve growth Biochem J 318: 603-608.
factor. Nature 329: 65-68. Hersch SM, Gutekunst CA, Rees HD, Heilman CJ, Levey AI.
Flohe L, BrigeliusFlohe R, Saliou C, Traber MG, Packer L. 1994. Distribution of M1M4 muscarinic receptor proteins
1997. Redox regulation of NFkB activation. Free Radic in the rat striatum: Light and electron microscopic immu-
Biol Med 22(6): 1115-1126. nocytochemistry using subtypespecific antibodies. J Neu-
Flynn DD, Mash DC. 1993. Distinct kinetic binding proper- rosci 14: 3351-3363.
ties of N[3H]methylscopolamine afford differential label- Hoult JR, Moroney MA, Paya M. 1994. Actions of flavonoids
ing and localization of M1, M2, and M3 muscarinic receptor and coumarins on lipoxygenase and cyclooxygenase. Meth-
subtypes in primate brain. Synapse 14(4): 283-296. ods Enzymol 234: 443-454.
Flynn DD, Weinstein DA, Mash DC. 1991. Loss of high Ingram DK, Jucker M, Spangler E. 1994. Behavioral manifes-
affinity agonist binding to M1 muscarinic receptors in tations of aging. Pathobiology of the Aging Rat. Vol. 2.
Alzheimers disease: Implications for the failure of cholin- Mohr U, Cungworth DL, Capen CC, editors. Washington:
ergic replacement therapies. Ann Neurol 29(3): 256-262. ILSI Press; pp. 149-170.
Franzoso G, Bours V, Azarenko V, Park S, TomitaYamaguchi Ishii K, Oda Y, Ichikawa T, Deguchi T. 1990. Complementary
M, et al. 1993. The oncoprotein Bcl3 can facilitate NFkB DNAs for choline acetyltransferase from spinal cords of rat
mediated transactivation by removing inhibiting p50 and mouse: Nucleotide sequences, expression in mammali-
homodimers from select kB sites. EMBO J 12: 3893-3901. an cells, and in situ hybridization. Mol Brain Res 7: 151-159.
Frigo DE, Duong BN, Melnik LI, Schief LS, CollinsBurow BM, Joseph JA. 1992. The putative role of free radicals in the loss of
et al. 2002. Flavonoid phytochemicals regulate activator neuronal functioning in senescence. Integr Physiol Behav
protein1 signal transduction pathways in endometrial Sci 27: 216-227.
and kidney stable cell lines. J Nutr 132(7): 1848-1853. Joseph JA, Fisher DR. 2003. Muscarinic receptor subtype
Fujita T, Nolan GP, Liou HC, Scott ML, Baltimore D. 1993. determines vulnerability to amyloid b toxicity in trans-
The candidate protooncogene bcl3 encodes a transcrip- fected COS7 cells. J Alzheimers Dis 5: 197-208.
tional coactivator that activates through NFkB p50 homo- Joseph JA, Roth GS. 1992. Loss of muscarinic regulation of
dimers. Genes Dev 7: 1354-1363. striatal dopamine function in senescence. Neurochem Int
Ghosh S, Baltimore D. 1990. Activation in vitro of NFkB by 20 Suppl: 237S-240S.
phosphorylation of its inhibitor I kB. Nature 344: 678-682. Joseph JA, Roth GS. 1994. Oxidative stress and reduced
Gould TJ, Bickford P. 1997. Agerelated deficits in the cerebel- receptor responsiveness senescence. Trophic Regulation of
lar b adrenergic signal transduction cascade in Fischer 344 the Basal Ganglia: Focus on Dopamine Neurons. Fuxe K,
rats. J Pharmacol Exp Ther 281: 965-971. Agnati LF, Leon A, Ottoson D, editors. Pergamon Press
Gozal E, Simakajornboon N, Gozal D. 1998. NFkB induction Oxford.
during in vivo hypoxia in dorsocaudal brain stem of rat: Effect Joseph JA, Arendash G, Gordon M, Diamond D, ShukittHale
of MK801 and LNAME. J Appl Physiol 85(1): 372-376. B, et al. 2003. Blueberry supplementation enhances
Antiaging strategies 13 315
signaling and prevents behavioral deficits in an Alzheimer Joseph JA, ShukittHale B, McEwen J, Rabin B. 1999b. Mag-
disease model. Nutr Neurosci 6: 153-162. nesium activation of GTP hydrolysis or incubation in S
Joseph JA, Bartus RT, Clody D, Morgan D, Finch C, et al. adenosylLmethionine reverses iron56particleinduced
1983. Psychomotor performance in the senescent rodent: decrements in oxotremorine enhancement of Kevoked
Reduction of deficits via striatal dopamine receptor up striatal release of dopamine. Radiat Res 152: 637-641.
regulation. Neurobiol Aging 4: 313-319. Joseph JA, ShukittHale B, Denisova NA, Martin A, Perry G,
Joseph JA, Berger RE, Engel BT, Roth GS. 1978. Agerelated et al. 2001. Copernicus revisited: Amyloid b in Alzheimers
changes in the nigrostriatum: A behavioral and biochemi- disease. Neurobiol Aging 22: 131-146.
cal analysis. J Gerontol 33: 643-649. Joseph JA, VillalobosMolina R, Denisova N, Erat S, Cutler R,
Joseph JA, Cutler R, Roth GS. 1993a. Changes in G protein et al. 1996. Age differences in sensitivity to H2O2 or NO
mediated signal transduction in aging and Alzheimers induced reductions in Kevoked dopamine release from
disease. Ann N Y Acad Sci 24(695): 42-45. superfused striatal slices: Reversals by PBN or Trolox. Free
Joseph JA, Hunt WA, Rabin BM, Dalton TK, Harris AH. Radic Biol Med 20: 821-830.
1993b. Deficits in the sensitivity of striatal muscarinic Kaasinen V, Vilkman H, Hietala J, Nagren K, Helenius H, et al.
receptors induced by 56Fe heavyparticle irradiation: 2000. Agerelated dopamine D2/D3 receptor loss in extra-
Further ageradiation parallels. Radiat Res 135(2): striatal regions of the human brain. Neurobiol Aging 21(5):
257-261. 683-688.
Joseph JA, Dalton TK, Roth GS, Hunt WA. 1988. Alterations Kalt W, Forney CF, Martin A, Prior RL. 1999. Antioxidant
in muscarinic control of striatal dopamine autoreceptors in capacity, vitamin C, phenolics, and anthocyanins after
senescence: A deficit at the ligandmuscarinic receptor in- fresh storage of small fruits. J Agric Food Chem 47(11):
terface? Brain Res 454: 149-155. 4638-4644.
Joseph JA, Denisova N, Fisher D, ShukittHale B, Bickford P, Kaufmann JA, Bickford PC, Taglialatela G. 2001. Oxidative
et al. 1998a. Agerelated neurodegeneration and oxidative stressdependent upregulation of Bcl2 expression in the
stress: Putative nutritional intervention. Neurobiologic central nervous system of aged Fisher344 rats. J Neuro-
Clinics of North America: The Neurobiology of Aging 16: chem 76: 1099-1108.
747-755. Kimura M, Hatono S, Une M, Fukuoka C, Kuramoto T, et al.
Joseph JA, ShukittHale B, Denisova NA, Prior RL, Cao G, 1984. Synthesis, intestinal absorption and metabolism of
et al. 1998b. Longterm dietary strawberry, spinach or vita- sarcosine conjugated ursodeoxycholic acid. Steroids 43(6):
min E supplementation retards the onset of agerelated 677-687.
neuronal signaltransduction and cognitive behavioral def- Kitamura Y, Shimohama S, Ota T, Matsuoka Y, Nomura Y,
icits. J Neurosci 18: 8047-8055. et al. 1997. Alteration of transcription factors NFkB
Joseph JA, Fisher DR, Carey AN. 2004. Fruit extracts antago- and STAT1 in Alzheimers disease brains. Neurosci Lett
nize Ab or DAinduced deficits in Ca2 flux in M1trans- 237: 17-20.
fected COS7 cells. J Alzheimers Dis 6: 403-411; discussion Kluger A, Gianutsos JG, Golomb J, Ferris SH, George AE,
443449. et al. 1997. Patterns of motor impairment in normal aging,
Joseph JA, Fisher DR, Strain J. 2002. Muscarinic receptor mild cognitive decline, and early Alzheimers disease. J
subtype determines vulnerability to oxidative stress in Gerontol B Psychol Sci Soc Sci 52: P28-P39.
COS7 cells. Free Radic Biol Med 32: 153-161. Lai MK, Lai OF, Keene J, Esiri MM, Francis PT, et al.
Joseph JA, Gupta M, Han Z, Roth GS. 1991. The deleteri- 2001. Psychosis of Alzheimers disease is associated with
ous effects of aging and kainic acid may be selective for elevated muscarinic M2 binding in the cortex. Neurology
similar striatal neuronal populations. Aging (Milano) 3(4): 57: 805-811.
361-371. Landfield PW, Eldridge JC. 1994. The glucocorticoid hypoth-
Joseph JA, Hunt WA, Rabin BM, Dalton TK. 1992. Possible esis of agerelated hippocampal neurodegeneration: Role of
accelerated striatal aging induced by 56Fe heavyparticle dysregulated intraneuronal calcium. Ann N Y Acad Sci 746:
irradiation: Implications for manned space flights. Radiat 308-321; discussion 321326.
Res 130: 88-93. Levey AI. 1996. Muscarinic acetylcholine receptor expression
Joseph JA, ShukittHale B, Denisova NA, Bielinski D, in memory circuits: Implications for treatment of Alzhei-
Martin A, et al. 1999a. Reversals of agerelated declines mer disease. Proc Natl Acad Sci USA 93: 13541-13545.
in neuronal signal transduction, cognitive, and motor Levine MS, Cepeda C. 1998. Dopamine modulation of
behavioral deficits with blueberry, spinach, or strawberry responses mediated by excitatory amino acids in the neos-
dietary supplementation. J Neurosci 19: 8114-8121. triatum. Adv Pharmacol 42: 724-729.
316 13 Antiaging strategies
Liu J, Sen R, Rothstein TL. 1993. Abnormal kBbinding Pakaski M, Papp H, Rakonczay Z, Fakla I, Kasa P. 2001. Effects
protein in the cytoplasm of a plasmacytoma cell line of acetylcholinesterase inhibitors on the metabolism of amy-
that lacks nuclear expression of NFkB. Mol Immunol 30: loid precursor protein in vitro. Neurobiology (Bp). 9: 55-57.
479-489. Prior RL, Cao G, Martin A, Sofic E, McEwen J. 1998. Antioxi-
Lonnerberg P, Schoenherr CJ, Anderson DJ, Ibanez CF. 1996. dant capacity influenced by total phenolic and anthocyanin
Cell typespecific regulation of choline acetyltransferase content, maturity, and variety of Vaccinium species. J Agric
gene expression. J Biol Chem 52: 33358-33365. Food Chem 46: 2686-2693.
MaridonneauParini I, Braquet P, Garay RP. 1986. Heteroge- Pu H, Zhai P, Gurney M. 1993. Enhancer, silencer, and growth
neous effect of flavonoids on K loss and lipid peroxidation factor responsive regulatory sequences in the promoter for
induced by oxygenfree radicals in human red cells. Phar- the mouse choline acetyltransferase gene. Mol Cell Neu-
macol Res Commun 1: 61-72. rosci 4: 131-142.
Martin S, Favot L, Matz R, Lugnier C, Andriantsitohaina R. Qiu J, Grafe MR, Schmura SM, Glasgow JN, Kent TA, et al.
2003. Delphinidin inhibits endothelial cell proliferation 2001. Differential NFkB regulation of bclx gene expres-
and cell cycle progression through a transient activation sion in hippocampus and basal forebrain in response to
of ERK1/2. Biochem Pharmacol 65(4): 669-675. hypoxia. J Neurosci Res 64: 223-234.
Mazzucchelli C, Brambilla R. 2000. Rasrelated and MAPK Rossner S, Ueberham U, Schliebs R, PerezPolo JR, Bigl V.
signalling in neuronal plasticity and memory formation. 1998. The regulation of amyloid precursor protein metab-
Cell Mol Life Sci 57(4): 604-611. olism by cholinergic mechanisms and neurotrophin recep-
McDonald RJ, White NM. 1994. Parallel information proces- tor signaling. Prog Neurobiol 56: 541-569.
sing in the water maze: Evidence for independent memory Ruben SM, Dillon PJ, Schreck R, Henkel T, Chen CH, et al.
systems involving dorsal striatum and hippocampus. Neu- 1991. Isolation of a relrelated human cDNA that poten-
robiol Learn Mem 61: 260-270. tially encodes the 65kD subunit of NFkB. Science 251:
Micheau J, Riedel G. 1999. Protein kinases: Which one is the 1490-1493.
memory molecule? Cell Mol Life Sci 55(4): 534-548. Sato M, Bagchi D, Tosaki A, Das DK. 2001. Grape seed
Misawa H, Ishii K, Deguchi T. 1992. Gene expression of proanthocyanidin reduces cardiomyocyte apoptosis by
mouse choline acetyltransferase. J Biol Chem 267(28): inhibiting ischemia/reperfusioninduced activation of
20392-20399. JNK1 and CJUN. Free Radic Biol Med 31(6): 729-737.
Misawa H, Takahashi R, Deguchi T. 1993. Transcriptional Schafe GE, Nadel NV, Sullivan GM, Harris A, Le Doux JE.
regulation of choline acetyltransferase gene by cyclic AMP. 1999. Memory consolidation for contextual and auditory
J Neurochem 60: 1383-1387. fear conditioning is dependent on protein synthesis, PKA,
Mori F, Lai CC, Fusi F, Giacobini E. 1995. Cholinesterase and MAP kinase. Learn Mem 6(2): 97-110.
inhibitors increase secretion of APPs in rat brain cortex. Schmid RM, Perkins ND, Duckett CS, Andrews PC, Nable GJ.
Neuroreport 6: 633-636. 1991. Cloning of an NFkB subunit which stimulates HIV
Morris MC, Scherr PA, Hebert LE, Bennett DA, Wilson RS, transcription in synergy with p65. Nature 352: 733-736.
et al. 2000. The crosssectional association between blood Schroeter H, Boyd C, Spencer JP, Williams RJ, Cadenas E,
pressure and Alzheimers disease in a biracial community et al. 2002. MAPK signaling in neurodegeneration: Influ-
population of older persons. J Gerontol A Biol Sci Med Sci ences of flavonoids and of nitric oxide. Neurobiol Aging
55: M130-M136. 23(5): 861-880.
Muir JL. 1997. Acetylcholine, aging, and Alzheimers disease. Seeram NP, Bourquin LD, Nair MG. 2001a. Degradation
Pharmacol Biochem Behav 56: 687-696. products of cyanidin glycosides from tart cherries and
Nachmansohn D, Machado AL. 1943. The formation of ace- their bioactivities. J Agric Food Chem 49(10): 4924-4929.
tylcholine. A new enzyme: Choline acetylase. J Neuro- Seeram NP, Momin RA, Nair MG, Bourquin LD. 2001b.
physiol 6: 397-403. Cyclooxygenase inhibitory and antioxidant cyanidin glyco-
Nolan GP, Fujita T, Bhatia K, Huppi C, Liou HC, et al. 1993. sides in cherries and berries. Phytomedicine 8(5): 362-369.
The bcl3 protooncogene encodes a nuclear IkBlike mol- Seeram NP, Cichewicz RH, Chandra A, Nair MG. 2003.
ecule that preferentially interacts with NFkB p50 and p52 Cyclooxygenase inhibitory and antioxidant compounds
in a phosphorylationdependent manner. Mol Cell Biol 13: from crabapple fruits. J Agric Food Chem 51(7): 1948-1951.
3557-3566. Selcher JC, Atkins CM, Trzaskos JM, Paylor R, Sweatt JD.
Nyakas C, Oosterink BJ, Keijser J, Felszeghy K, de Jong GI, 1999. A necessity for MAP kinase activation in mammalian
et al. 1997. Selective decline of 5HT1A receptor binding spatial learning. Learn Mem 6(5): 478-490.
sites in rat cortex, hippocampus and cholinergic basal fore- Sgambato V, Pages C, Rogard M, Besson MJ, Caboche J.
brain nuclei during aging. J Chem Neuroanat 13: 53-61. 1998. Extracellular signalregulated kinase (ERK) controls
Antiaging strategies 13 317
immediate early gene induction on corticostriatal stimula- Vile GF, TanewIlitschew A, Tyrrell R. 1995. Activation of
tion. J Neurosci 18(21): 8814-8825. NFkB in human skin fibroblasts by the oxidative stress
Shen HM, Lin Y, Choksi S, Tran J, Jin T, et al. 2004. Essential generated by UVA radiation. Photochem Photobiol 62:
roles of receptorinteracting protein and TRAF2 in 463-468.
oxidative stressinduced cell death. Mol Cell Biol 24(13): Wang H, Cao G, Prior RL. 1996. Total antioxidant capacity of
5914-5922. fruits. J Agric Food Chem 44: 701-705.
ShukittHale B. 1999. The effects of aging and oxidative stress Wang H, Nair MG, Strasburg GM, Chang YC, Booren AM,
on psychomotor and cognitive behavior. Age 22: 9-17. et al. 1999. Antioxidant and antiinflammatory activities of
ShukittHale B, Mouzakis G, Joseph JA. 1998. Psychomotor anthocyanins and their aglycon, cyanidin, from tart
and spatial memory performance in aging male Fischer 344 cherries. J Nat Prod 62(5): 802.
rats. Exp Gerontol 33: 615-624. West RL. 1996. An application of prefrontal cortex function
Siebenlist U, Frazoso G, Brown K. 1994. Structure, regulation, theory to cognitive aging. Psychol Bull 120: 272-292.
and function of NFkB. Ann Rev Cell Biol 10: 405-455. Whitehouse PJ, Price DL, Struble RG, Clark AW, Coyle JT,
Sorata Y, Takahama U, Kimura M. 1984. Protective effect of et al. 1982. Alzheimers disease and senile dementia: Loss of
quercetin and rutin on photosensitized lysis of human neurons in basal forebrain. Science 215: 1237-1239.
erythrocytes in the presence of hematoporphyrin. Biochim Williams LR, Rylett RJ, Ingram DK, Joseph JA, Moises HC,
Biophys Acta 799(3): 313-317. et al. 1993. Nerve growth factor affects the cholinergic
Taglialatela G, Hibbert CJ, WerrbachPerez K, PerezPolo JR. neurochemistry and behavior of aged rats. Prog Brain Res
1996. Suppression of p140trkA does not abolish nerve 98: 251-256.
growth factor rescue of apoptotic PC12 cells and enables Wortwein G, Yu J, Toliver T, PerezPolo JR. 1998. Responses of
brainderived neurotrophic factor to promote cell survival. young and aged rat CNS to partial cholinergic immunole-
J Neurochem 66: 1826-1835. sions and NGF treatment. J Neurosci Res 52: 322-333.
Takeda A, Onodera H, Sugimoto A, Kogure K, Obinata M, Yamagami K, Joseph JA, Roth GS. 1992. Decrement in mus-
et al. 1993. Coordinated expression of messenger RNAs for carinic receptor stimulated low Km GTPase in striata and
nerve growth factor, brainderived neurotrophic factor and hippocampus from aged rat. Brain Res 57: 327-331.
neurotrophin3 in the rat hippocampus following transient Youdim KA, ShukittHale B, Martin A, Wang H, Denisova N,
forebrain ischemia. Neuroscience 55: 23-31. et al. 2000. Shortterm dietary supplementation of blueber-
Terai K, Matsuo A, McGeer PL. 1996. Enhancement of immu- ry polyphenolics: Beneficial effects on aging brain perfor-
noreactivity for NFkB in the hippocampal formation mance and peripheral tissue function. Nutr Neurosci 3:
and cerebral cortex of Alzheimers disease. Brain Res 735: 383-397.
159-168. Young LT, Kish SJ, Li PP, Warsh JJ. 1988. Decreased brain [3H]
ToliverKinsky T, Papaconstantinou J, PerezPolo JR. 1997. inositol1,4,5trisphosphate binding in Alzheimers disease.
Ageassociated alterations in hippocampal and basal fore- Neurosci Lett 94(12): 198-202.
brain nuclear factor kB activity. J Neurosci Res 48: 580-587. Zhang J, Patel JM, Block ER. 1998. Hypoxiaspecific upregu-
ToliverKinsky T, Wood T, PerezPolo JR. 2000. Nuclear factor lation of calpain activity and gene expression in pulmonary
kB/p49 is a negative regulatory factor in nerve growth artery endothelial cells. Am J Physiol 275: L461-L468.
factorinduced choline acetyltransferase promoter activity Zipper LM, Mulcahy RT. 2000. Inhibition of ERK and p38
in PC12 cells. J Neurochem 75: 2241-2251. MAP kinases inhibits binding of Nrf2 and induction
Tong L, ToliverKinsky T, Edwards M, Rassin DK, Werrbach of GCS genes. Biochem Biophys Res Commun 278(2):
Perez K, et al. 2002. Attenuated transcriptional responses to 484-492.
oxidative stress in the aged rat brain. J Neurosci Res 70(3): Zuchner T, PerezPolo JR, Schliebs R. 2004. bsecretase
318-326. BACE1 is differentially controlled through muscarinic ace-
Toussaint JL, Geoffrey V, Schmit M, Werner A, Garnier JM, tylcholine receptor signaling. J Neurosci Res 77: 250-257.
et al. 1992. Human choline acetyltransferase (CHAT): Par- Zyzak DR, Otto T, Eichenbaum H, Gallagher M. 1995.
tial gene sequence and potential control regions. Genomics Cognitive decline associated with normal aging in rats: A
12(2): 412-416. neuropsychological approach. Learn Mem 2: 1-16.
Index
Acetylcholine (Ach), 220, 230232 myelin breakdown in, 71, 73 hybridization, in situ, 274
AChE, 230235 neurodegeneration, 181, 182, 184, immunohistochemistry, 274
ACTH, 217 186, 190 interaction, protein-protein,
AD. See Alzheimers disease neuropathological and clinical 272, 273
Advanced glycosylation end products features of, 288290 isoforms, 266
(AGEs), 125 pathogenesis of, 125 localization, subcellular, 274, 276
Aging, 217, 219, 225, 227230 secretases, 264 mice
age-related increases in iron tau phosphorylation, 150, 151 knock-out, 265, 278
levels, 73 AMD, 230 transgenic, 274276
age-related loss of myelin, 71 Amyloid beta-peptide, 125, 288 modification, posttranslational
ferritin, 73 b-amyloid cleaving enzyme 1, 289 glycosylation, 271
iron content, 73 b-amyloid peptide phosphorylation, 272
reduction of fast conducting CNS amyloid precursor protein, 307 mRNA, alternative splicing,
axons, 71 free radicals, 308 265, 266
Aging brain and dendritic spine, Amyloid precursor protein, 288 promoter, 268, 269
254256 Amyotrophic lateral sclerosis, 226, regulation
Aging processes 286, 291 transcriptional, 268
causes of, 104 due to abnormal proteins, translational, 268
free-radical theory, 105 107, 108 structure, genomic, 265
genetic theories on evolution Anticholinesterases, 231, 233, 234 transport, intracellular,
of, 133 Antioxidant proteins 271, 272
mitochondrial theory of aging, cellular expression and activity of, BACE. See b-amyloid cleaving enzyme 1
108110 104, 110 BBB. See Blood brain barrier
ALS. See Amyotrophic lateral sclerosis HO, 113 bdnf gene, transgenic mouse, 12, 13
Alzheimer disease, 78, 82, 83, 85, 105, Antisense, 233235 BDNF. See Brain-derived neurotrophic
219, 223, 230, 286 Apolipoprotein E (ApoE), 289 factor
aging, 274 Apoptosis Betaine, 83
amyloid, 149 anoikis, 178 homocysteine
amyloid beta peptides (Ab) Bcl-2, 178180 methyltransferase, 80
deposit, 71 caspases, 178 Bilirubin
Ab toxicity in, 73 p53, 179 as antioxidant, 121, 122
amyloid precursor protein, 262 transcription factors, 161, end products of heme
cell 180, 181 degradation, 113
cycle, 181, 190196 APP. See Amyloid precursor protein nonconjugated,
death, 189, 190, 196 ASF/SF2, 224, 225 supraphysiological levels
damaging free radicals in AVP, 217 of, 121
pathogenesis of myelin, 73, 74 Ab. See Amyloid beta-peptide Biliverdin, antinitrosative
effects of developmental- BACE characteristics of, 122
dependent oligodendrocyte expression Biomarker, of DNA damage, 106
myelination in, 74 cell type-specific, 268, Blood brain barrier, 233
genetic factors, 148, 149 274, 275 Brain
Hsps proteins in, 121 mRNA, 266, 268, 270, 274 GSH system in, 120
mitotic catastrophe in, 150 protein, 268 heme oxygenase in, 117, 120
Brain aging cyclin dependent kinases, 150, CNS disorders. See Central nervous
DNA oxidative damage 160162, 175, 190 system disorders
genetic instability, 107 inhibitors of cyclin dependent CO. See Carbon monoxide
methylating agents, 106 kinases, 165, 174, 175, 180, 181, Cortical neuronal migration, Trk
production of ROS, 106 190, 191 signaling pathways in, 9
protein oxidative damage mitotic catastrophe, 150 Corticosterone, 217, 233
neurodegenerative diseases restriction point, 160, 167, 168 Cortisol, 217, 233
due to, 107, 108 retinoblastoma protein, 161, 166, COX enzyme, 95
proteolysis, 107 167, 192 COX-2. See Cyclooxygenase-2
Brain damage mechanism, in tau phosphorylation, 150, 151 Creutzfeldt-Jakob disease, 292
ND, 296 Cell differentiation CRF, 217
Brain-derived neurotrophic factor de-differentiation, 175, 190, 196 Cu/Zn-superoxide dismutase
(BDNF) development, 166, 172175 (Cu/Zn-SOD)
cerebellar granule neuron mechanisms, molecular, 167, 168, antioxidative genes,
migration, role in, 10 173, 186 overexpression of, 105, 106
motor neuron death by, 5 Cellular apoptosis, 90 cytosolic form, mutation in, 124
mutant mouse Central nervous system (CNS), 90 in DS patients, 132
cortical layer thickness, 8 development, TrkB and TrkC CWD. See Chronic wasting disease
motor neuron survival, 5 role in, 5 Cyclooxygenase-2, 93
neuronal migration in neocortex, neuronal survival, neurotrophins g-cystathionase, 7981
7, 8 role in, 3 Cystathionine, 7881, 84
Brain disease, spine number and neurons in, 11 Cystathionine b-synthase, 7981, 84
formation in, 254 NOS and its isoforms in, 111 Cystathioninuria, 81
Brain stress tolerance Central nervous system disorders Cysteic acid, 7880
heat-shock pathway of associated with ROS/RNS- Cysteine, 7981, 83, 84
Hsp27, Ubiquitin, and mediated damage, 123 Cysteine oxidase, 79, 80
Hsp47, 116 AD, 125126 Cysteinesulfinic acid, 7880
Hsp60 and Hsp32, 117 ALS, 124 Cysteinesulfinic acid decarboxylase,
Hsp70, stress proteins, DS, 132 79, 80
115, 116 FA, 130132 Cytochrome oxidase, mitochondrial,
in mammalian cells, 115 HD, 132 activity and aging process, 106
Branchpoint, 220222, 229 ischemia/reperfusion, 133 Cytokine-induced nitrosative stress, 116
Caenorhabditis elegans, 293 MS, 128130 Cytokine responses and cell survival,
Calcium/calmodulin-dependent PD, 126, 127 91, 92
protein kinase II-a, 253 Cerebellar granule neurons, migration Decoy treatments, 96
Caloric restriction of, 9, 10 Dendritic arborization, 11
rate of aging and cellular Cerebellar purkinje cells, dendritic Dendritic development, neurotrophin
pathology, 122 differentiation in, 12 role in
signaling pathway, 123 Cerebellum, layered structure in BDNF mouse strain, 12, 13
CaMKII-a. See Calcium/calmodulin- formation of, 9 neurotrophin receptors and
dependent protein kinase II-a Cerebral cortex, layered architecture downstream signaling molecules,
Cannon, W., 216 formation of 13, 14
Carbon monoxide during development and non-pyramidal interneuron
antiinflammatory properties, adulthood, 8 dendrites, 11, 12
113115 postmitotic neuron migration, 7 pyramidal neurons, 11
neurogasobiology of, 110, 111 Cerebrovascular disease, 78, 82 Dendritic spine
Cardiolipin, 109 Chaperone-buffered silent and aging brain, 254256
Cardiovascular disease, 78, 81 mutations, 105 synapses
Caspase-1, 91 Cholinergic system formation of, 250, 251
Caspr2, 7173 acetylcholine, 309 structure of, 246, 247
Cell cycle aging, 309, 310 Dendritic transport, 225
in Alzheimer disease, 149152 alzheimers disease, 309, 310 DFP, 232
check points, 159, 167, 168, choline acetyltransferase, DNA damage
179, 181 309, 310 enzmes to repair, 107
control elements, 149151 cognitive deficits, 309 due to oxidative stress, 106
cyclin(s), 150, 160163, 165167, Chronic wasting disease, 292 DNA polymerases, 107
169, 172175, 181, 190 CJD. See Creutzfeldt-Jakob disease Dopamine, 220, 227
Index 321
Downs syndrome, 132 GEFs. See Guanine nucleotide exchange cognitive function, 81, 82
Drosophila, in neurodegenerative factors dementias, 82
disease modeling, 293 Gene expression regulation and NGF, depression, 82
DS. See Downs syndrome 2233 early development, 80, 82
EAE. See Experimental allergic General adaptation syndrome, 216 fetal death, 82
encephalomyelitis Genome-based theory, 104 hip fractures, 82
Electron traffic jam upstream, 106 Gerstmann-Straussler-Scheinker, 292 ischemic stroke, 81, 82
Embryogenesis, TrkB expression GFP-transfected hippocampal neuron, neural tube defects, 78, 81, 82
during, 8 3D imaging of, 252 NMDA receptor, 83, 84
Endothelial NOS (eNOS), 95 Gliosis programmed cell death, 82, 83
Epinephrine, 217, 219 Glucocorticoids, 217, 218, 220, 233 protein bound, 84
ESEs, 222 Glutamate, 220, 226 thromboembolic episodes, 81
ESTs, 220 Glutathione, 80, 84 vascular dementia, 82
Experimental allergic cellular detoxification, 119 venous thrombosis, 82
encephalomyelitis, HO-1 expression reduced and oxidized, 120 Homocysteine methyltransferase,
in, 121 GR, 218, 223, 224 79, 80
FA. See Friedreichs ataxia GSH. See Glutathione Homocystinuria, 78, 81, 82, 84
Familial CJD (fCJD), 292 GSS. See Gerstmann-Straussler- HPA, 217220
Fight or flight, 216 Scheinker Hsp 27, 116
Fluoresance-recovery after GTPase activating protein, 253 HSP70, stress proteins family
photobleaching, 248 Guanine-nucleotide exchange factor(s), components of, 115
Folic acid, 8082, 85 13, 254 pathological condition
Forebrain development HD. See Huntington disease induction, 116
Emx1-positive Heme oxygenase (HO) pathway Huntington disease, 132, 286,
neuroepithelium, 67 defensive mechanism, 120 290, 291
fate mapping studies, 66 in antidegenerative mechanism, Hydrophobic interactions, in protein
genomic screened homeobox2 AD, 105 aggregation, 295
(Gsh2), 62, 67 HNE. See 4-hydroxy-trans-nonenal 4-hydroxy-trans-nonenal, 108
lateral ganglionic eminence hnRNPs, 221, 222, 227, 228 Hypersensitivity, 232235
(LGE), 62, 66, 67 HO system Hypotaurine, 7880
medial ganglionic eminence colocalization of, 118 Hypoxia, 82
(MGE), 62, 6668 endogenous, 121 Hypoxia ischemia (HI), 90
Nkx 2.1, 67, 68 expression, regulation of Iatrogenic CJD (iCJD), 292
Nkx2.1-Cre transgenic mouse HO-1, 119 IL-1 RI. See IL-1 type-I receptor
line, 67 HO-2, 118 IL-1 type-I receptor, 91
FRAP. See Fluoresance-recovery after intracellular modulators for, Imaging, 220
photobleaching 119, 120 Inducible nitric oxide synthase (NOS),
Free-radical hypothesis of aging induction of, 118 93, 95
causative factors, 105 inhibition of, 121 Inflammatory gene expression, 95, 96
oxidative damage, DNA and isoforms of, 117 Inflammatory responses, 91
proteins, 106, 107 NADPH cytochrome c P450 Inherited genetic defects, 78
oxidized dysfunctional proteins, reductase activity, degradation of iNOS. See Inducible nitric oxide
accumulation of, 108 heme, 117 synthase; Inducible NOS
proteasomal dysfunction, 107 upregulation in nigral Intron, 220223, 226, 229231
Friedreichs ataxia, 130, 131 neurons, 121 Ischemic brain damage, 133
Fruits and vegetables Homeostasis, 90, 216 Juxtaparanode junction, cytoskeletal
anthocyanins, 310, 311 Homocysteic acid, 78, 83 proteins
flavonoids, 311 Homocysteine connexin 29, 71
polyphenols, 310, 311 age, 82 Kv1.1, 71, 73
FTDP-17, 226, 230 aging, 82 Kv1.2, 71
GABAergic interneurons alzheimers disease, 78, 82, 83, 85 Kv1.6, 71
in medial ganglionic eminence amyloid beta, 83 Kv2, 71
(MGE), 7 amyotrophic lateral sclerosis, 83 tranasient axonal glycoprotein-1
tangential migration of, NT-4 and apoptosis, 82, 83 (Tag1), 71
BDNF role in, 9 atherosclerosis, 81, 85 Lipid peroxidation, in
GAP. See GTPase activating protein cardiovascular disease, 81 neurodegenerative disorders, 108
GCs. See Glucocorticoids cerebrovascular disease, 78, 82 Lipopolysaccharides (LPS), 94
322 Index