Cell, Vol.
100, 377–386, February 4, 2000, Copyright 2000 by Cell Press
The Complete Sequence of a Heterochromatic
Island from a Higher Eukaryote
The Cold Spring Harbor Laboratory,
Washington University Genome Sequencing Center,
and PE Biosystems Arabidopsis
Sequencing Consortium*
Summary
Heterochromatin, constitutively condensed chromo-
somal material, is widespread among eukaryotes but
incompletely characterized at the nucleotide level. We
have sequenced and analyzed 2.1 megabases (Mb)
of Arabidopsis thaliana chromosome 4 that includes
0.5–0.7 Mb of isolated heterochromatin that resembles
the chromosomal knobs described by Barbara McClin-
tock in maize. This isolated region has a low density
of expressed genes, low levels of recombination and
a low incidence of genetrap insertion. Satellite repeats
were absent, but tandem arrays of long repeats and
many transposons were found. Methylation of these
sequences was dependent on chromatin remodeling.
Clustered repeats were associated with condensed
chromosomal domains elsewhere. The complete se-
quence of a heterochromatic island provides an op-
portunity to study sequence determinants of chromo-
some condensation.
Introduction
The chromosomes of higher eukaryotes are character-
ized by the presence of condensed regions of deeply
* The Cold Spring Harbor Laboratory, Washington
University Genome Sequencing
Center, and PE Biosystems (CSHL/WUGSC/
PEB†) Arabidopsis Sequencing Consortium
is comprised of the following researchers:
W. Richard McCombie,1 Melissa de la Bastide,1
Kristina Habermann,1 Laurence Parnell,1
Neilay Dedhia,1 Lidia Gnoj,1 Kristin Schutz,1
Emily Huang,1 Lori Spiegel,1 Cristy Yordan,2
Mundeep Sehkon,3 Jennifer Murray,3 Paul Sheet,3
Matt Cordes,3 Jane Threideh,3 Tamberlyn Stoneking,3
Joelle Kalicki,3 Tina Graves,3 Gwen Harmon,3
Jennifer Edwards,3 Phil Latreille,3 Laura Courtney,3
James Cloud,3 Amanda Abbott,3 Kelsi Scott,3
Doug Johnson,3 Pat Minx,3 Dan Bentley,3
Bob Fulton,3 Nancy Miller,3 Tracie Greco,3
Kim Kemp,3 Jason Kramer,3 Lucinda Fulton,3
Elaine Mardis,3 Mike Dante,3 Kym Pepin,3
LaDeana Hillier,3 Joanne Nelson,3 John Spieth,3
Joe Simorowski,2 Bruce May,2 Peter Ma,4
Ray Preston,1 Daniel Vil,1 Lei Hoon See,1
Monica Shekher,1 Anthony Matero,1 Ravi Shah,1
I’Kyori Swaby,1 Andrew O’Shaughnessy,1
Milka Rodriguez,1 Jane Hoffman,1 Sally Till,1
staining material, heterochromatin, that can be differen-
tiated cytologically from the surrounding lightly staining
chromosomal regions (Cavalli and Paro, 1998; Holm-
quist et al., 1998). In Drosophila and maize, heterochro-
matic regions have been shown to suppress recombina-
tion, to deter transposon insertion, and to influence the
expression of genes found nearby (Cook and Karpen,
1994; Henikoff, 1998; Dimitri and Junakovic, 1999). In
Drosophila, the composition of heterochromatin has
been extensively investigated in minichromosomes de-
rived from chromosome 4. Sample sequencing and FISH
(fluorescent in situ hybridization) have revealed that
much of the minichromosome comprised arrays of short
satellite repeats, but pulsed-field gel analysis revealed
islands of low complexity DNA (Le et al., 1995). Some
of these islands contained intact retrotransposable ele-
ments, but none were found to be unique to centromeric
regions (Sun et al., 1997). In maize, heterochromatin
has been extensively investigated since chromosome
staining procedures were sensitive enough to detect
them (McClintock, 1929, 1930; McClintock et al., 1981).
Heterochromatin is found in large pericentromeric do-
mains up to 100 megabases (Mb) in size as well as at
nucleolar organizers. Smaller islands of heterochroma-
tin, known as knobs and chromomeres, are distributed
along the chromosome arms, though they are often telo-
meric. Knobs occur at different positions in different
strains of maize and are stably polymorphic, allowing
their extensive use as cytogenetic landmarks (Richards
and Dawe, 1998). Classically, the correlation between
cytological and genetic crossover was established in
maize using these landmarks (Creighton and McClin-
tock, 1931). Both interstitial and terminal knobs can vary
Susan Granat,1 Nadim Shohdy,1 Amy Hasegawa,1
Aliyah Hameed,1 Mohammad Lodhi,1,5
Arthur Johnson,1,6 Ellson Chen,4 Marco Marra,3
‡
Richard K. Wilson,3 and Robert Martienssen2,
1
Lita Annenberg Hazen Genome Center
2
Cold Spring Harbor Plant Biology Group
Cold Spring Harbor Laboratory
Cold Spring Harbor, New York 11724
3
Genome Sequencing Center
Washington University
School of Medicine
St. Louis, Missouri 63108
4
Applied Biosystems
Foster City, California 94494
5
Axys
La Jolla, California 92037
6
Department of Forestry
North Carolina State University
Raleigh, North Carolina 27695
† Future citations of this work should use following form: (CSHL/
WUGSC/PEB Arabidopsis Sequencing Consortium, 2000).
‡ To whom correspondence should be addressed (e-mail: martiens@
cshl.org).
Cell
378

in size, suggesting that some property of the chromo-

somal location itself, such as the DNA sequence, might
be required for chromosome condensation and might
vary from one allele to another.
Sample sequencing and end labeling has revealed
that heterochromatic knobs in maize are composed of
180 bp and 350 bp tandem repeat arrays that comprise
70% of the sequence, with retrotransposons making up
the remaining 30% (Ananiev et al., 1998a, 1998b). It has
therefore been proposed that either the repeats or the
retrotransposons might be responsible for the hetero-
chromatic properties of these regions. Supernumerary
maize B chromosomes have also been extensively ana-
lyzed for their heterochromatic content (Alfenito and
Birchler, 1993). B chromosome–specific high copy re-
peats were isolated by sample cloning and shown to be
related to knob sequences. By taking advantage of their
unusual disjunction at meiosis, it was possible to associ-
ate these sequences with euchromatic and heterochro-
matic regions of the minichromosome (Kaszas and
Birchler, 1996). However, no more than a small percent-
age of the contiguous sequence of a heterochromatic
region has yet been determined in any higher eukaryote.
Furthermore, the complete sequence of these regions
is unlikely to be determined in the near future. In part,
this is because long regions of satellite repeats are re-
fractory to complete sequence analysis due to the diffi-
culty in assembling overlapping reads (Rubin, 1998; Be- Figure 1. Physical Map of the Heterochromatic Knob on Arabi-
van et al., 1999). Because these repeats are coextensive dopsis Chromosome 4
with the condensed chromatin, the sequence elements Maps of the 2.1 Mb region from the assembly of YACs (blue) and
responsible for the formation of heterochromatin cannot BACs (green) are aligned with the genetic map derived from recom-
binant inbreds (left). The heterochromatic knob (right) is shown sche-
be unambiguously assigned.
matically. It lies within the YAC CIC8B1 (Fransz et al., 2000).
In the model plant Arabidopsis thaliana, heterochro-
matin is largely confined to the pericentromeric regions
of each of the five chromosomes as well as to the nucleo-
lar organizers at the ends of chromosomes 2 and 4. Results
These regions have been shown to include hundreds of
kilobases of tandemly repeated sequences (Copenhaver Mapping and Sequencing
and Pikaard, 1996; Round et al., 1997). Sample sequenc- The ten chromosome arms of the Arabidopsis genome
ing and FISH have shown that pericentromeric repeats are almost completely represented by contiguous bac-
include 180 bp satellite repeats, as well as retrotranspo- terial artificial chromosome (BAC) clones assembled in
sons, most notably from the family known as Athila (Pel- order by restriction enzyme fingerprinting and hybridiza-
issier et al., 1996; Brandes et al., 1997; Heslop-Harrison tion (Marra et al., 1999; Mozo et al., 1999). Fingerprinting
et al., 1999). These sequences have also proved impos- allowed us to select an optimal set of these clones for
sible to assemble into a single contiguous stretch (Lin sequencing the short arm of chromosome 4 (Figure 1),
et al., 1999; Mayer et al., 1999). However, the short arm even though this chromosomal region is highly repeti-
of chromosome 4 has an isolated region of heterochro- tive, which prohibited assembly by hybridization. Each
matin resembling a maize chromomere or knob (Fransz of these BAC clones was then sequenced to a very
et al., 2000), and we report here its complete sequence high level of accuracy, and the resulting sequence was
as well as the 2.1 Mb region surrounding it. The knob analyzed for gene and repeat content. The sequenced
has a very low gene density, and the genes are inter- region was placed on the cytogenetic map using five
spersed with numerous classes of repetitive sequence. sequenced genetic markers and a physical map con-
These include retrotransposons and DNA transposons structed from yeast artificial chromosomes (YAC) (Schmidt
as well as a central domain of tandemly repeated 1950 et al., 1995). In an accompanying paper, Fransz and
bp elements. The repeats and transposons are heavily coworkers describe the use of these YACs as probes
methylated, and their methylation is dependent on chro- for FISH, allowing genetic and cytological regions to be
matin remodeling. We have used genetrap mutagenesis aligned directly (Fransz et al., 2000 [this issue of Cell]).
and recombinant inbred lines to show that the entire A diagram of the short arm, showing the clones selected
region is refractory to meiotic recombination as well as for sequencing and salient markers is shown at http://
to detectable genetrap insertions. Unlike heterochro- www.cshl.org/protarab/chrom4map.htm.
matic regions in other organisms, this region in Arabi- The region described here is between BACs F2N1 and
dopsis is relatively small and complex, allowing the com- T1J1 with the knob lying in the center of YAC CIC8B1
plete sequence to be determined. (Figure 1).
Arabidopsis Heterochromatin Sequence Analysis
379
Figure 2. Distribution of Genes, Transposons, and Repeats
The sequence of the 2080 kb region shown in Figure 1 was divided
into 250 kb intervals, and the total number of genes and repeats was
calculated for each interval. (A) Transcribed (green) and hypothetical
(yellow) genes; (B) dispersed (pink) and tandem (purple) repeats; (C)
DNA transposons (light blue), Athila (green), and other retrotranspo-
sons (dark blue). (D) The heterochromatic knob (green) is shown
along with the locations of genetrap and enhancer trap elements
(triangles).
Gene Distribution
Three hundred forty-six genes, seven tRNA genes, 88
intact and defective transposons, and 150 other repeats
were distributed along the length of the 2.1 Mb region.
One hundred eighty-five of the genes could be catego-
rized in functional classes (data not shown) and included
some genes that were previously thought unique to ani-
mals, such as an expressed homolog of the NMDA ligand-
gated ion-channel receptor, found in neurons (N. Kaplan
et al., unpublished data, GenBank entry AC002330; Lam
et al., 1998), and a homolog of the zinc finger apoptosis
inhibitor protein IAP-5. However, these classes were
not significantly different from those observed in the
chromosome as a whole (Mayer et al., 1999). Although
functional classes did not differ significantly between
euchromatic and heterochromatic regions (data not
shown), considerable differences in gene density were
observed (Figure 2). In the euchromatic regions, trans-
posons and other repeats comprised less than 10% of
the genes (Figures 2A and 2C). In the heterochromatic
region, repeats outnumbered genes by ten to one. In
the most condensed portion of the chromosome, there
is a stretch of 183,155 bp (from T5L23.27 to T24M8.10)
that has no predicted genes and only four hypothetical
genes (Figure 3). Unlike predicted genes, hypothetical
genes do not match any known gene or transcribed
sequence and were only found by gene modeling (see
Experimental Procedures). Eight DNA transposons and
21 retrotransposons were found in this stretch, along
with a large number of other repeats (Figure 3).
The frequency of expressed genes, as estimated by
matches to EST (expressed sequence tag) and cDNA
clones was also significantly biased (Figure 2A). In the
distal portion of the arm, the number of genes that
matched ESTs (37%) was similar to that reported for
the whole chromosome (Mayer et al., 1999). In the proxi-
mal half of the arm, only 13% of the genes matched
known transcribed sequences, and none of the 24 genes
in the central 450 kb of the knob corresponded to ESTs
(Figure 3). This bias was not evident in the distribution
of hypothetical genes (Figure 2A).
Genetrap Integration and Expression
We examined the number of gene traps and enhancer
traps that had integrated in this region from a collection
of approximately 2000 random integrations. These engi-
neered genetraps were based on the Ac/Ds transposons
from maize that had been introduced artificially into Ara-
bidopsis (Springer et al., 1995; Sundaresan et al., 1995).
Random integrations were mapped by amplifying and
sequencing DNA flanking the insertion site (see Experi-
mental Procedures). BLAST analysis indicated that 33
of the 2000 genetraps had integrated within the 2.1 Mb
region, consistent with expectations for a 130 Mb ge-
nome. However, none had integrated in heterochromatic
sequences, and all but one had integrated in the distal
euchromatin (Figure 2D). Seventeen of these 33 euchro-
matic genetraps landed within genes, and ten resulted
in expression of the reporter gene in 7-day-old seed-
lings. Two of the 33 insertions were within retrotranspo-
sons. A similar distribution was found in the genome as
a whole for the other 2000 genetraps examined (R. M.
et al., unpublished data).
Analysis of Repeats
Repeats and endogenous transposons were detected
by both sequence comparison and compositional analy-
sis (see Experimental Procedures). Sequence compari-
son detected reading frames with homology to plant and
animal transposase. This analysis revealed a number of
transposons not previously described in Arabidopsis,
including those of the Robertson’s Mutator (Bennetzen,
1996) and Mariner (Plasterk et al., 1999) classes (see
GenBank annotations). Inverted repeats were frequently
found at the ends of these elements, and transposons
were classified as defective if they had suffered re-
arrangements in the coding region (data not shown). All
15 copies of the Athila retrotransposon were found in
the proximal portion of the chromosome arm (Figure 2C)
along with other classes of repeat sequence (such as
Cell
380
mi167) previously thought to be restricted to centro-
meric regions (Pelissier et al., 1996; Thompson et al.,
1996). None of these elements were found in the distal
euchromatic portion. Other retrotransposons as well as
DNA transposons were clustered in the heterochromatin
(Figure 3) so that it more closely resembled the maize
genome than it did Arabidopsis (San Miguel et al., 1996).
However, these transposons were also found in the dis-
tal portion, although they were relatively rare. For exam-
ple, out of ten Mu transposons located on this chromo-
some arm, only one lay in the euchromatic region, and it
was inserted in the most condensed portion (see below).
Retrotransposons were also slightly enriched in this dis-
tal region (Figure 2C).
Compositional analysis of repeats was performed us-
ing PrintRepeats and DOTPLOT (Figure 4). Dispersed
repeats were distributed throughout the region and in-
cluded putative transposons as well as gene families.
Similar repeats found close to each other may reflect
the propensity of transposons to integrate close to the
site from which they came as well as local gene duplica-
tion. Several clusters of related genes were observed,
including one that contained five calcium-dependent
protein kinases and another with three NAM (no apical
meristem) transcription factors. These clusters were of-
ten interrupted by transposons or dispersed repeats,
which may have been involved in the duplication events.
Two clusters of genes encoding CHP-rich zinc finger
proteins were found in the distal region (Fig 4B). Thirty-
three genes in the Arabidopsis sequence database en-
code similar proteins, some of which are highly con-
served at the nucleotide as well as the amino acid level
and lack introns, suggesting recent duplication or trans-
position. Interestingly, this region of the chromosome
(corresponding to CIC9F6 in Figure 1) is twice as con-
densed as the surrounding regions during meiosis (re-
gion IIb in Fransz et al., 2000).
Tandem repeats were nonrandomly distributed rela-
tive to dispersed repeats (Figure 2B). In particular, 22.5
tandemly arranged copies of a 1950 bp repeat were
located in the heterochromatic knob (Figure 3). Each
Figure 3. Detailed Analysis of the Knob
Region
The diagram shows a 0.52 Mb region within
the interval that includes the heterochromatic
knob (T5L23 to T27D20, Figure 1). Hypotheti-
cal (black) and predicted (red) genes are
shown as thin arrows, tandem repeats as
block arrows, and transposons as triangles.
The only two genes corresponding to ESTs in
this region are highlighted as thick red arrows
(T5L23.27 and T27D20.10). The 183,000 bp
region that corresponds to the heterochro-
matic region and has no expressed or pre-
dicted genes is bracketed.
1950 bp element included a terminal 31 bp direct repeat
and varied by 2%–8% in nucleotide sequence from other
members of the array (Figure 4C). Gene modeling pro-
grams did not predict any exon structure in the knob
repeat, but BLASTN searches identified one sequence
on chromosome 5 and two sequences on chromosome
2 with significant homology (BLASTN score ⬎500, bases
576 to 1400). The two related sequences on chromo-
some 2 were annotated as genes, located approximately
400 kb apart (T19K21.3 and F17L24.8). BLAST searches
revealed that T19K21.3 appeared to be integrated within
a transposase gene from the Spm family (data not
shown; Lin et al., 1999). Taken together, we conclude
that the knob repeat may be related to a defective trans-
posable element. We next performed searches using
only the nucleotide sequence from the most conserved
portion of the repeat (nucleotides 1747–1904), which
shared 40% identity at the amino acid level with the other
three genes (Figure 4C). This 160 bp region matched six
additional BAC clones from the pericentromeric regions
of chromosomes 2 and 4. Of special note, 14 copies of
the sequence were included in a tandem array of 1 kb
degenerate repeats on T6L9, a BAC from the pericen-
tromeric heterochromatic region of the long arm of chro-
mosome 4 (Figure 4C). A dotplot analysis of the two
BACs is shown in Figure 4D.
DNA Methylation
We tested the methylation status of the knob-specific
repeats in cv. Columbia and cv. Landsberg by Southern
analysis (Figure 5). We found that the array of 1950 bp
knob repeats was present in both strains and was heav-
ily methylated at SalI and PvuII sites, which cut once in
each repeat (Figure 4C). Levels of methylation at SalI
sites in the repeat array were equivalent in Columbia
and Landsberg, but PvuII was found to digest Columbia
but not Landsberg repeat arrays (Figure 5A). Athila ele-
ments from the knob and throughout the genome were
also methylated at EcoRII, SalI, and HpaII sites. Repeat
methylation was examined in ddm1 (decrease in DNA
Arabidopsis Heterochromatin Sequence Analysis
381
methylation) mutants, which have unmethylated centro-
meric satellite and nucleolar organizer DNA (Vongs et
al., 1993). Single copy DNA retains its methylation status
in these mutants, at least for the first few generations
(Kakutani et al., 1996, 1999). Both Athila retrotranspo-
sons and the 1950 bp repeats were found to be exten-
sively, if not fully, demethylated in F2 ddm1 mutants
(Figures 5A and 5C). Interestingly, enzymes that cut once
within the 1950 bp repeat gave a ladder of monomers,
dimers, and trimers consistent with the repeats being
found in tandem arrays in both Landsberg and Co-
lumbia.
Recombination
Recombination across the region could be measured
using sequence-tagged sites from the recombinant in-
bred map (Schmidt et al., 1995). Figure 1 shows that
the level of recombination observed across the region
varied dramatically. In the euchromatic portion, each
cM spanned 70 to 150 kb, compared to 950 kb in the
Figure 4. Analysis of Repeated Sequences
(A) PrintRepeats output for the entire 2.1 Mb
region. Repeated sequences are joined by
arches. The knob region and the distal con-
densed region are shown schematically.
(B) PrintRepeats output for the condensed dis-
tal euchromatic region around BAC T15B16
(Figure 1). Arches join individual members of
gene families described in the text.
(C) Structure of the 1950 bp knob repeat,
aligned with the 1 kb degenerate repeat found
on the long arm BAC, T6L9.
(D) DOTPLOT comparison of the BACs T5H22
(short arm) and T6L6 (long arm) showing the
tandem arrays found on each BAC as well as
a comparison between them.
heterochromatic portion. Interestingly, the highest lev-
els of recombination were in the region of the chromo-
some identified by Fransz et al. (2000) as being the least
condensed during meiosis (YAC CIC4A7, Figure 1). This
relationship was examined further by plotting the cumu-
lative number of genes and transposons as a function
of genetic distance along the chromosome arm (Figure
6). This analysis revealed that physical distance, the
total number of genes, and especially the number of
transposons were unrelated to genetic distance. Only
the cumulative number of expressed genes varied lin-
early with genetic distance, indicating a correlation be-
tween gene expression and recombination.
Discussion
Heterochromatin Structure
In both Drosophila and mammalian cells, two classes
of heterochromatin, ␣ and ␤, have been defined by the
extent of condensation and other cytological criteria
Figure 5. Repeated Sequences in the Het-
erochromatic Knob Are Methylated
Southern hybridization analysis of Arabi-
dopsis DNA digested with various enzymes
and probed with the 1950 bp tandem repeat
(A) and the Athila retrotransposon (B and C).
DNA samples were as follows: (A) lane 1, Co-
lumbia digested with PvuII; lane 2, Landsberg
digested with PvuII; lane 3, Columbia di-
gested with SalI; lane 4, Landsberg digested
with SalI; lane 5, Columbia ddm1 digested
with SalI; lane 6, Landsberg ddm1 di-
gested with SalI. (B) Columbia DNA digested
with EcoRII (lane 1), BstNI (lane 2), HpaII (lane
3), and MspI (lane 4). (C) DNA samples di-
gested with EcoRII from Columbia (lane 1),
ddm1 homozygous plants (lane 2), ddm1/⫹
heterozygous plants (lane 3), and ddm1 ho-
mozygous plants after two further genera-
tions of inbreeding (lanes 5 and 6). Arrows
indicate fragments corresponding to fully di-
gested repeats.
Cell
382
Figure 6. Recombination Varies Linearly with Gene Expression
The cumulative number of genes, expressed genes, transposons,
and physical distance in 10 s of kb is plotted against recombination
distance throughout the region.
(reviewed by Wallrath, 1998). At the sequence level,
␣-heterochromatin has extensive arrays of tandem re-
peats occasionally interrupted with transposons. In
contrast, ␤-heterochromatin has scrambled arrays of
transposons with little evidence of satellite arrays (Holm-
quist et al., 1998). In this respect, the knob on chromo-
some 4 resembles ␤-heterochromatin in that there are
no tandem arrays of short repeats. In maize, sample
sequencing has revealed that heterochromatic knobs
are composed of 180 bp and 350 bp tandem repeat
arrays that comprise 70% of the sequence, with retro-
transposons making up the remaining 30% (Ananiev et
al., 1998b). This is a lower density of retrotransposons
than is found in the maize genome as a whole. In the
heterochromatic knob on chromosome 4 of Arabidopsis,
retrotransposons also make up 30% of the region, but
less than 5% of the genome as a whole. One class of
retroelements (the 11 kb Athila elements) are found only
in heterochromatic regions (Pelissier et al., 1996) and
are highly represented in the Arabidopsis knob (Figure
3), suggesting they are specifically targeted to hetero-
chromatin and may contribute to its structure. Other
classes of retrotransposons are found throughout the
region between the centromere and the knob, sug-
gesting they do not preferentially target heterochroma-
tin, but rather may be retained in regions with sup-
pressed recombination (see below).
Unlike in maize, neither the Arabidopsis 180 bp centro-
meric satellite repeat nor any other known satellite was
found in the heterochromatic knob. Instead, tandem re-
peats of a 1950 bp sequence were observed. A related
tandem repeat was found in the pericentromeric hetero-
chromatin on the other side of the centromere. This sug-
gests that tandem repeats, rather than satellite arrays,
might be responsible for local condensation via a mech-
anism similar to that reported in Drosophila and S.
pombe (Dorer and Henikoff, 1994; Grewal and Klar,
1997). In these studies, the ability of tandem repeats to
induce the formation of heterochromatin was demon-
strated by gene silencing and nuclease sensitivity as
well as association with the chromocentre, which is an
important property of functional heterochromatin in Dro-
sophila (Csink and Henikoff, 1996).
Heterochromatin Dynamics
Tandem repeats can form by unequal exchange (Hennig,
1999). In the case of Arabidopsis rDNA, such exchange
can be detected by examining repeat structure in recom-
binant inbred populations (Copenhaver and Pikaard,
1996). However, in the centromere (Round et al., 1997),
as well as in the knob region (see below), meiotic recom-
bination is substantially eliminated, suggesting that mei-
otic crossover can not be responsible for the amplifica-
tion of repeated arrays. The presence of transposons
in Drosophila heterochromatin has led to the suggestion
that satellite repeats might be amplified in part by un-
equal exchange following transposon excision (Thomp-
son-Stewart et al., 1994). Alternatively, the integration
of transposons multiple times at the same site, or within
one another, could also result in tandem or more com-
plex arrays (Figure 3). The knob repeat on chromosome 4
resembles a defective transposable element, consistent
with this idea. Repeat arrays could acquire condensa-
tion if the repeats carry chromatin binding sites that
cooperate to induce heterochromatinization (Fanti et al.,
1998). Such chromatin-binding sites may help to target
further transposon integrations (Pirrotta and Rastelli,
1994). In this way, knobs could form very quickly in
plant genomes during evolutionary bursts of transposon
activity. According to this view, heterochromatin forma-
tion is simply a function of transposon density that re-
sults in novel genetic properties such as silencing and
reduced levels of recombination. An alternative view,
that knobs are themselves megatransposons (Ananiev
et al., 1998a), cannot be excluded but is less likely in
this case given the complex composition of the knob and
the paucity of other knobs in the Arabidopsis genome.
Fransz et al. (2000) report that Arabisopsis thaliana
cv. Landsberg plants do not have a heterochromatic
knob at this position on chromosome 4, indicating that
knobs are polymorphic in Arabidopsis, as they are in
maize (McClintock et al., 1981). However, Southern hy-
bridization indicates that the knob repeat array is pres-
ent in the Landsberg genome (Figure 5) as are sequences
between the knob and the centromere (Copenhaver et
al., 1999; Fransz et al., 2000). One possibility is that the
knob has been fused to the centromere in Landsberg,
a possibility suggested by inversion of at least part of
the intervening region observed by FISH (Fransz et al.,
2000). Under conditions of “genome stress,” knobs have
a tendency to spontaneously fuse with telomeres and
centromeres as well as other knobs (McClintock, 1951).
DNA transposons in heterochromatic regions, such as
the elements reported here, could account for these
fusions if they were activated under similar conditions
(McClintock, 1978). This is because abortive transposi-
tions can result in fusion of donor and target sites and
inversion of the intervening sequences (Hudson et al.,
1990; English et al., 1993; Weil and Wessler, 1993).
The Genetic Properties of Heterochromatin
The heterochromatic knob on Arabidopsis chromosome
4 is composed largely of transposons and retrotranspo-
sons, many of which are also found near centromeres
(Thompson et al., 1996). Further, the knob maps to an
interval thought to contain the centromere on chromo-
some 4 (Copenhaver et al., 1998). However, it has been
Arabidopsis Heterochromatin Sequence Analysis
383
shown recently that the knob and the centromere are
distinct (Copenhaver et al., 1999). While the function
of heterochromatin remains a mystery, we have taken
advantage of the complete sequence to characterize
the genetic properties of the region.
Many fewer genes matched ESTs in the heterochro-
matic region than in the euchromatic region (Figure 2A).
Even more dramatically, the number of ESTs that
matched heterochromatic genes (four in 645,000 bp)
was 40-fold less than in the distal euchromatin (260
in 1,102,089 bp) and 20-fold less than in the proximal
euchromatin (35 in 305,447 bp). This quantitative mea-
sure strongly suggests that gene expression is severely
reduced in the heterochromatic knob (Adams et al.,
1991; McCombie et al., 1992a). Furthermore, genetraps
were not found in heterochromatic sequences. This was
not due to a propensity to avoid retrotransposons, as
integrations in dispersed retrotransposons elsewhere in
the genome were at normal levels. One possibility is
that the selectable marker gene carried by the genetrap
was silenced when integrated into heterochromatin so
that these integrations could not be detected (Sundare-
san et al., 1995). Another possibility is that genetrap
transposons preferentially integrate within expressed
genes that are rare or absent in heterochromatin. A simi-
lar relationship has been observed in Drosophila (Berg
and Spradling, 1991). It may be possible to target gene-
traps to integrate within heterochromatin by including
heterochromatic sequences in the transposon construct
(Kassis et al., 1992; Fauvarque and Dura, 1993; Taille-
burg et al., 1999), and we are currently exploring this
possibility.
The frequency of recombination varied greatly with
physical distance (Figure 1), but it varied almost linearly
with the density of expressed genes: EST matches oc-
curred eight times per cM on average throughout the
region (Figure 6). This observation is consistent with the
notion that recombination might be limited to euchro-
matic portions of the genome that encode expressed
genes. Such a relationship has been predicted on the
basis of genome size and genetic distance (Thuriaux,
1977; Bennett, 1998) and has received some support
from studies of intragenic recombination in maize (Ci-
vardi et al., 1994; Xu et al., 1995; Timmermans et al.,
1996; Schnable et al., 1998).
The paucity of expressed genes was also correlated
with the highly repeated and methylated nature of the
heterochromatic region. DNA methylation was depen-
dent on the DDM1 gene, which is required for methyla-
tion of heterochromatic regions in Arabidopsis (Vongs
et al., 1993). DDM1 is also known to influence gene
expression, and in each case the affected genes are
repeated as well as methylated (Jeddeloh et al., 1998;
Paszkowski and Mittelsten Scheid, 1998). DDM1 en-
codes a SNF2 chromatin remodeling enzyme (Jeddeloh
et al., 1999) and might influence gene expression by
altering the interrelationship of repeated sequences,
allowing access to methyltransferase and potentially to
transcriptional repressors, such as methyl-binding pro-
teins (Wolffe, 1998; Martienssen and Henikoff, 1999;
Wolffe and Hayes, 1999). The influence of ddm1 on
methylation of transposons might also lead to changes
in gene expression. This is because integration of trans-
posons in the vicinity of genes can bring them under
their control (McClintock, 1965; Martienssen, 1998b). In
at least one case, the genes affected by DDM1 undergo
a form of paramutation (Jeddeloh et al., 1998), an epige-
netic change in gene expression linked to transposons
as well as repeats (Martienssen, 1996). It is possible
that paramutation, and related imprinting mechanisms
(Martienssen, 1998a), might preferentially affect re-
peated regions of the genome that are relatively con-
densed and whose methylation depends on chromatin
remodeling.
Chromatin Condensation and Repeats
In the accompanying paper, Fransz et al. (2000) describe
the changes in chromosomal condensation that accom-
pany meiosis in the short arm of chromosome 4. While
condensation is most severe in the conspicuous knob,
it is also observed in a distal portion of the arm that
can therefore be considered “cryptic” heterochromatin
(Figure 4B). This region encompasses three families of
clustered genes interspersed with transposons and ret-
rotransposons and is twice as condensed as the sur-
rounding euchromatin (Fransz et al., 2000). Interestingly,
the clustered genes are not found in EST databases,
raising the possibility that they are downregulated in the
same way as the genes found in the knob. McClintock
(1951) predicted that the study of heterochromatin
would lead to the understanding of euchromatic regula-
tion of gene expression. Given the role that transposons
(McClintock’s “controlling elements”) and repeats play
in heterochromatin and in epigenetic gene regulation,
this prediction may yet be borne out in such regions.
Unlike in maize, or any other higher eukaryote, the
complete sequence of the Arabidopsis knob, including
the boundaries of the heterochromatic region, has been
determined. The complete and highly accurate se-
quence of this region of the genome should greatly aid
efforts to determine the mechanism behind chromatin
condensation, recombination suppression, and gene si-
lencing responsible for this ancient property of plant
genomes. Although there are many sequence-based
strategies for the analysis of complex genomes (Adams
et al., 1991; McCombie et al., 1992a; Waterston et al.,
1992), only complete sequencing linked to genetic and
cytological maps can provide an understanding of
higher order chromosome organization and function
(McCombie et al., 1992a; Sulston et al., 1992; Wilson et
al., 1994; Bevan et al., 1998; Ashburner et al., 1999).
Analysis of the first complete genome of an animal, C.
elegans (The C. elegans Sequencing Consortium, 1998),
has already led to new insights into genome structure
and evolution. As Arabidopsis genome sequencing
nears completion, we can expect a dramatic increase
in our understanding of higher order chromosome struc-
ture and function in flowering plants.
Experimental Procedures
Sequencing Methods
BAC clones F2N1, F3D13, F11O4, T15B16, T7B11, T10M13, T2H3,
T14P8, T10P11, T5J8, T4I9, F4C21, F9H3, T5L23, T5H22, T7M24,
T25H8, T24M8, T24H24, T27D20, T19B17, T26N6, F4H6, T19J18,
T4B21, and T1J1 were individually cultured for large-scale prepara-
tion, and the DNA was isolated via alkaline lysis, phenol/chloroform
extraction, and isopropanol precipitation. The purified DNA was
Cell
384
then sheared by nebulization with 10 psi of nitrogen gas. 1.5–3 kb
fragments were size selected on agarose gels, repaired with mung
bean nuclease, and ligated into M13. Cells were transformed by
electroporation and plaques picked from a lawn of JM101. Single-
stranded M13 phage DNA was isolated from overnight cultures in
96-well format via the Thermomax procedure (Mardis, 1994). The
purified DNA was then cycle sequenced using a combination of dye
primer and dye terminator chemistries and subsequently separated
and detected by electrophoresis through acrylamide gels run on
ABI 377 sequencers (McCombie at al., 1992b). Data were tracked
and processed and raw data quality monitored using a web-based
lab management data system, Kaleidaseq (Dedhia and McCombie,
1998). BAC clones were sequenced to approximately 8-fold cover-
age. The random fragments were basecalled using PHRED (Phil
Green) and assembled into contiguous pieces using the assembly
algorithm PHRAP (Phil Green). Assemblies were converted (using
software developed by Gabor Marth at the WUGSC) into the editing
programs Gap4 (Roger Staden) or CONSED (David Gordon) for man-
ual inspection and finalization. Directed reads were chosen to close
sequence gaps and achieve contiguity as well as resolve ambiguities
within the sequence. Vector junctions were also delineated. The
sequence was finished to an accuracy of less than one error per
100 kb by ensuring that each base was covered by at least two
independent subclones representing both strands of DNA or se-
quenced with both dye primer and dye terminator chemistries on
multiple clones representing the same strand. The final assemblies
were confirmed by comparison of the assemblies’ predicted restric-
tion fragments to actual restriction digests of the BAC clone, includ-
ing the fingerprints from the mapping group.
Gene Modeling
Gene modeling was accomplished through an integration of gene
prediction algorithms and similarity searching. An extensive analysis
of the performance of Arabidopsis exon and gene prediction algo-
rithms (L. P. et al., unpublished data) indicated that GenScan, MZEF,
and NetPlantGene perform best at delineating the exon-intron struc-
ture of a gene (Hebsgaard et al., 1996; Burge and Karlin, 1998;
Zhang, 1998). GRAIL is best at identifying terminal exons. A consen-
sus gene model is then built, guided by the results of similarity
searching. BLASTN identifies EST matches and BLASTX marks se-
quence with the potential to encode a protein similar to entries in
a protein database. (EST matches above 96% identity were scored
in which most mismatches were unreadable bases.) Intergenic seg-
ments and large (typically greater than 800 bp) introns are routinely
searched for the occurrence of repeats. DNA repeats (LINES, MITES,
solo LTRs, etc.) with the exception of simple sequence repeats
(SSRs) are ⬎90% identical and over 100 bp in length; SSRs must
extend for ⬎25 bp. Repeats with the potential to encode transposon
proteins are identified by similarity to transposase, reverse tran-
scriptases, and polyproteins. PrintRepeats analysis (Parsons, 1995)
was performed at a cutoff of 150. Dotplots were performed on line
at the Virtual Genome Centre at the University of Minnesota (http://
alces.med.umn.edu/rawdot.html) or by using the Wisconsin GCG
group sequence analysis package.
DNA Methylation Analysis
Arabidopsis genomic DNA was digested with restriction enzymes
and subjected to DNA gel blot analysis as described previously
(Springer et al., 1995). DNA from ddm1 mutants was from the F2
generation. Landsberg ddm1 heterozygotes were genotyped by
progeny test, and were only self-pollinated after six generations of
backcrossing (L. Das and R. M., unpublished data) so that demethyl-
ation was a result of loss of ddm1 function rather than inheritance
of unmethylated DNA (Vongs et al., 1993; Kakutani et al., 1999). DNA
gel blots were hybridized and washed at high stringency with probes
corresponding to bases 570 to 1340 of the 1950 bp repeat and to
bases 1123 to 1847 of the adjacent defective Athila element T5H22.4.
Probes were amplified from genomic DNA using the primer pairs
5⬘-ACGACGAGAGGCTCCATCTA-3⬘/5⬘-CTGTTCCTCGCAAGTTC
CTC-3⬘ and 5⬘-CCACACAGGTGGATCAACAG-3⬘/ 5⬘-TGATTGAT
GAGGACATCGGA-3⬘ respectively. Probes were gel purified and se-
quenced to verify their origin.
Gene Trap and Enhancer Trap Transposon Mutagenesis
Two thousand independent Arabidopsis plants (cv. Landsberg) were
isolated so that each contained one or occasionally more than one
insertion of the DsG gene trap or DsE enhancer trap transposable
elements described by Sundaresan et al. (1995) and Springer et al.
(1995). Genomic DNA was prepared from each plant and the inser-
tion sites were amplified using TAIL PCR and sequenced essentially
as described by Tsukegi et al. (1996). BLASTN searches were per-
formed against the Arabidopsis genome sequence in GenBank using
an automated annotation routine (B. M. and R. M., unpublished
data). Unique matches of greater than 96% identity among readable
bases were taken to indicate insertion of the element in the genomic
region indicated by the GenBank entry. Enhancer (ET) and gene (GT)
traps found in the region were: ET1967, GT5365, GT6223, GT4985,
ET587, GT5961, GT6437, ET5292, GT6437, GT5927, GT5912,
GT6230, ET1949, GT5222, GT331, ET120, GT4985, ET446, GT148,
ET5413, GT150, GT6149, ET37, ET5862, ET5867, ET5864, ET134,
GT4985, ET5506, GT6702, GT5962, ET238, ET5490, and GT2722.
Acknowledgments
We would like to thank Eric Richards for comments on the manu-
script and for ddm1 seed. We would like to thank Bruce May, John
Healy, and Andy Reiner for help with the genetrap sequence analy-
sis, and Juana-Marie Arroyo and Rulan Shen for sharing unpublished
genetrap data. We acknowledge the support of the United States
Department of Agriculture NRI (95-37300-1578), the National Sci-
ence Foundation (DBI-9813578), the Robertson Research Fund, West-
vaco Corporation, and the generous support of David L. Luke III.
Received October 26, 1999; revised January 10, 2000.
References
Adams, M.D., Kelley, J.M., Gocayne, J.D., Dubnick, M., Polymero-
poulos, M.H., Xiao, H., Merril, C.R., Wu, A., Olde, B., Moreno, R.F.,
et al. (1991). Complementary DNA sequencing: expressed sequence
tags and the human genome project. Science 252, 1651–1656.
Alfenito, M.R., and Birchler, J.A. (1993). Molecular characterization
of a maize B chromosome centric sequence. Genetics 135, 589–597.
Ananiev, E.V., Phillips, R.L., and Rines, H.W. (1998a). Complex struc-
ture of knob DNA on maize chromosome 9. Retrotransposon inva-
sion into heterochromatin. Genetics 149, 2025–2037.
Ananiev, E.V., Phillips, R.L., and Rines, H.W. (1998b). A knob-associ-
ated tandem repeat in maize capable of forming fold-back DNA
segments: are chromosome knobs megatransposons? Proc. Natl.
Acad. Sci. USA 95, 10785–10790.
Ashburner, M., Misra, S., Roote, J., Lewis, S.E., Blazej, R., Davis,
T., Doyle, C., Galle, R., George, R., Harris, N., et al. (1999). An explora-
tion of the sequence of a 2.9 Mb region of the genome of Drosophila
melanogaster. The Adh region. Genetics 153, 179–219.
Bennett, M.D. (1998). Plant genome values: how much do we know?
Proc. Natl. Acad. Sci. USA 95, 2011–2016.
Bennetzen, J.L. (1996). The mutator transposable element system
of maize. Curr. Top. Microbiol. Immunol. 204, 195–229.
Berg, C.A., and Spradling, A.C. (1991). Studies on the rate and site-
specificity of P element transposition. Genetics 127, 515–524.
Bevan, M., Bancroft, I., Bent, E., Love, K., Goodman, H., Dean, C.,
Bergkamp, R., Dirkse, W., Van Staveren, M., Stiekema, W., et al.
(1998). Analysis of 1.9 Mb of contiguous sequence from chromo-
some 4 of Arabidopsis thaliana. Nature 391, 485–488.
Bevan, M., Bancroft, I., Mewes, H.W., Martienssen, R., and McCom-
bie, R. (1999). Clearing a path through the jungle: progress in Arabi-
dopsis genomics. Bioessays 21, 110–120.
Brandes, A., Thompson, H., Dean, C., and Heslop-Harrison, J.S.
(1997). Multiple repetitive DNA sequences in the paracentromeric
regions of Arabidopsis thaliana L. Chromosome Res. 5, 238–246.
Burge, C.B., and Karlin, S. (1998). Finding the genes in genomic
DNA. Curr. Opin. Struct. Biol. 8, 346–354.
C. elegans Sequencing Consortium (1998). Genome sequence of
Arabidopsis Heterochromatin Sequence Analysis
385
the nematode C. elegans: a platform for investigating biology. Sci-
ence 282, 2012–2018.
Cavalli, G., and Paro, R. (1998). Chromo-domain proteins: linking
chromatin structure to epigenetic regulation. Curr. Opin. Cell. Biol.
10, 354–360.
Civardi, L., Xia, Y., Edwards, K.J., Schnable, P.S., and Nikolau, B.J.
(1994). The relationship between genetic and physical distances in
the cloned a1-sh2 interval of the Zea mays L. genome. Proc. Natl.
Acad. Sci. USA 91, 8268–8272.
Cook, K.R., and Karpen, G.H. (1994). A rosy future for heterochroma-
tin. Proc. Natl. Acad. Sci. USA 91, 5219–5221.
Copenhaver, G.P., and Pikaard, C.S. (1996). Two-dimensional RFLP
analyses reveal megabase-sized clusters of rRNA gene variants in
Arabidopsis thaliana, suggesting local spreading of variants as the
mode for gene homogenization during concerted evolution. Plant
J. 9, 273–282.
Copenhaver, G.P., Browne, W.E., and Preuss, D. (1998). Assaying
genome-wide recombination and centromere functions with Arabi-
dopsis tetrads. Proc. Natl. Acad. Sci. USA 95, 247–252.
Copenhaver, G.P., Nickel, K., Kuromori, T., Benito, M., Kaul, S., Lin,
X., Bevan, M., George Murphy, G., Harris, B., McCombie, W.R., et
al. (1999). Genetic definition and sequence analysis of Arabidopsis
centromeres. Science 286, 2468–2474.
Creighton, H., and McClintock, B. (1931). A correlation of cytological
and genetical crossing over in Zea mays. Proc. Natl. Acad. Sci. USA
17, 492–497.
Csink, A.K., and Henikoff, S. (1996). Genetic modification of hetero-
chromatic association and nuclear organization in Drosophila. Na-
ture 381, 529–531.
Dedhia, N.N., and McCombie, W.R. (1998). Kaleidaseq: a web-based
tool to monitor data flow in a high throughput sequencing facility.
Genome Res. 8, 313–318.
Dimitri, P., and Junakovic, N. (1999). Revising the selfish DNA hy-
pothesis: new evidence on accumulation of transposable elements
in heterochromatin. Trends Genet. 15, 123–124.
Dorer, D.R., and Henikoff, S. (1994). Expansions of transgene re-
peats cause heterochromatin formation and gene silencing in Dro-
sophila. Cell 77, 993–1002.
English, J., Harrison, K., and Jones, J.D. (1993). A genetic analysis
of DNA sequence requirements for Dissociation state I activity in
tobacco. Plant Cell 5, 501–514.
Fanti, L., Dorer, D.R., Berloco, M., Henikoff, S., and Pimpinelli, S.
(1998). Heterochromatin protein 1 binds transgene arrays. Chro-
mosoma 107, 286–292.
Fauvarque, M.O., and Dura, J.M. (1993). Polyhomeotic regulatory
sequences induce developmental regulator-dependent variegation
and targeted P-element insertions in Drosophila. Genes Dev. 7,
1508–1520.
Fransz, P., Armstrong, S., de Jong, J.H., Parnell, L.D., Van Drunen,
C., Dean, C., Zabel, P., Bisseling, T., and Jones, G.H. (2000). Inte-
grated cytogenetic map of chromosome arm 4S of A. thaliana: struc-
tural organization of heterochromatic knob and centromeric hetero-
chromatin. Cell 100, this issue, 367–376.
Grewal, S.I., and Klar, A.J. (1997). A recombinationally repressed
region between mat2 and mat3 loci shares homology to centromeric
repeats and regulates directionality of mating-type switching in fis-
sion yeast. Genetics 146, 1221–1238.
Hebsgaard, S.M., Korning, P.G., Tolstrup, N., Engelbrecht, J., Rouze,
P., and Brunak, S. (1996). Splice site prediction in Arabidopsis thali-
ana pre-mRNA by combining local and global sequence information.
Nucleic Acids Res. 24, 3439–3452.
Henikoff, S. (1998). Conspiracy of silence among repeated
transgenes. Bioessays 20, 532–535.
Hennig, W. (1999). Heterochromatin. Chromosoma 108, 1–9.
Heslop-Harrison, J.S., Murata, M., Ogura, Y., Schwarzacher, T., and
Motoyoshi, F. (1999). Polymorphisms and genomic organization of
repetitive DNA from centromeric regions of Arabidopsis chromo-
somes. Plant Cell 11, 31–42.
Holmquist, G.P., Kapitonov, V.V., and Jurka, J. (1998). Mobile genetic
elements, chiasmata, and the unique organization of beta-hetero-
chromatin. Cytogenet. Cell Genet. 80, 113–116.
Hudson, A.D., Carpenter, R., and Coen, E.S. (1990). Phenotypic ef-
fects of short-range and aberrant transposition in Antirrhinum ma-
jus. Plant Mol. Biol. 14, 835–844.
Jeddeloh, J.A., Bender, J., and Richards, E.J. (1998). The DNA meth-
ylation locus DDM1 is required for maintenance of gene silencing
in Arabidopsis. Genes Dev. 12 1714–1725.
Jeddeloh, J.A., Stokes, T.L., and Richards, E.J. (1999). Maintenance
of genomic methylation requires a SWI2/SNF2-like protein. Nat.
Genet. 22, 94–97.
Kakutani, T., Jeddeloh, J.A., Flowers, S.K., Munakata, K., and Rich-
ards, E.J. (1996). Developmental abnormalities and epimutations
associated with DNA hypomethylation mutations. Proc. Natl. Acad.
Sci. USA 93, 12406–12411.
Kakutani, T., Munakata, K., Richards, E.J., and Hirochika, H. (1999).
Meiotically and mitotically stable inheritance of DNA hypomethyla-
tion induced by ddm1 mutation of Arabidopsis thaliana. Genetics
151, 831–838.
Kassis, J.A., Noll, E., VanSickle, E.P., Odenwald, W.F., and Perrimon,
N. (1992). Altering the insertional specificity of a Drosophila trans-
posable element. Proc. Natl. Acad. Sci. USA 89, 1919–1923.
Kaszas, E., and Birchler, J.A. (1996). Misdivision analysis of centro-
mere structure in maize. EMBO J. 15, 5246–5255.
Lam, H.M., Chiu, J., Hsieh, M.H., Meisel, L., Oliveira, I.C., Shin, M.,
and Coruzzi, G. (1998). Glutamate-receptor genes in plants. Nature
396, 125–126.
Le, M.H., Duricka, D., and Karpen, G.H. (1995). Islands of complex
DNA are widespread in Drosophila centric heterochromatin. Genet-
ics 141, 283–303.
Lin, X., Kaul, S., Rounsley, S., Shea, T.P., Benito, M.I., Town, C.D.,
Fujii, C.Y., Mason, T., Bowman, C.L., Barnstead, M., et al. (1999).
Sequence and analysis of chromosome 2 of the plant Arabidopsis
thaliana. Nature, 402, 761–768.
Mayer, K., et al. The European Union Arabidopsis Genome Sequenc-
ing Consortium and the Cold Spring Harbor, Washington University
in St. Louis and PE Biosystems Arabidopsis Sequencing Consortium
(1999). Sequence and analysis of chromosome 4 of the plant Arabi-
dopsis thaliana. Nature 402, 769–777.
Mardis, E. (1994). High-throughput detergent extraction of M13 sub-
clones for fluorescent DNA sequencing. Nucleic Acids Res. 22,
2173–2175.
Marra, M., Kucaba, T., Sekhon, M., Hillier, L., Martienssen, R., Chin-
walla, A., Crockett, J., Fedele, J., Grover, H., Gund, C., et al. (1999).
A map for sequence analysis of the Arabidopsis thaliana genome.
Nat. Genet. 22, 265–270.
Martienssen, R. (1996). Epigenetic phenomena: paramutation and
gene silencing in plants. Curr. Biol. 6, 810–813.
Martienssen, R. (1998a). Chromosomal imprinting in plants. Curr.
Opin. Genet. Dev. 8, 240–244.
Martienssen, R. (1998b). Transposons, DNA methylation and gene
control. Trends Genet. 14, 263–264.
Martienssen, R., and Henikoff, S. (1999). The House and Garden
guide to chromatin remodelling. Nat. Genet. 22, 6–7.
McClintock, B. (1929). Chromosome morphology in Zea mays. Sci-
ence 69, 629.
McClintock, B. (1930). A cytological demonstration of the location
of an interchange between two non-homologous chromosomes of
Zea mays. Proc. Natl. Acad. Sci. USA 16, 791–796.
McClintock, B. (1951). Chromosome organization and genic expres-
sion. Cold Spring Harb. Symp. Quant. Biol. 16, 13–47.
McClintock, B. (1965). The control of gene action in maize. Brookha-
ven Symp. Biol. 18, 162–184.
McClintock, B. (1978). Mechanisms that rapidly reorganize the ge-
nome. Satadler Genet. Symp. 10, 25–48.
McClintock, B., Angel, T., Kato, Y., and Bluemenshein, A. (1981).
Chromosomal Constitution of Races of Maize (Chapingo, Mexico:
Colegio de Postgraduados).
Cell
386

McCombie, W.R., Adams, M.D., Kelley, J.M., FitzGerald, M.G., Ut- Tsugeki, R., Kochieva, E.Z., and Fedoroff, N.V. (1996). A transposon
terback, T.R., Khan, M., Dubnick, M., Kerlavage, A.R., Venter, J.C., insertion in the Arabidopsis SSR16 gene causes an embryo-defec-
and Fields, C. (1992a). Caenorhabditis elegans expressed sequence tive lethal mutation. Plant J. 10, 479–489.
tags identify gene families and potential disease gene homologues. Vongs, A., Kakutani, T., Martienssen, R.A., and Richards, E.J. (1993).
Nat. Genet. 1, 124–131. Arabidopsis thaliana DNA methylation mutants. Science 260, 1926–
McCombie, W.R., Heiner, C., Kelley, J.M., Fitzgerald, M.G., and Go- 1928.
cayne, J.D. (1992b). Rapid and reliable fluorescent cycle sequencing Wallrath, L.L. (1998). Unfolding the mysteries of heterochromatin.
of double-stranded templates. DNA Seq. 2, 289–296. Curr. Opin. Genet. Dev. 8, 147–153.
Mozo, T., Dewar, K., Dunn, P., Ecker, J.R., Fischer, S., Kloska, S.,
Waterston, R., Martin, C., Craxton, M., Huynh, C., Coulson, A., Hillier,
Lehrach, H., Marra, M., Martienssen, R., Meier-Ewert, S., and Alt-
L., Durbin, R., Green, P., Shownkeen, R., Halloran, N., et al. (1992).
mann, T. (1999). A complete BAC-based physical map of the Arabi-
A survey of expressed genes in Caenorhabditis elegans. Nat. Genet.
dopsis thaliana genome. Nat. Genet. 22, 271–275.
1, 114–123.
Parsons, J.D. (1995). Miropeats: graphical DNA sequence compari-
Weil, C.F., and Wessler, S.R. (1993). Molecular evidence that chro-
sons. Comput. Appl. Biosci. 11, 615–619.
mosome breakage by Ds elements is caused by aberrant transposi-
Paszkowski J., and Mittelsten Scheid, O. (1998). Plant genes: the tion. Plant Cell 5, 515–522.
genetics of epigenetics. Curr. Biol. 8, R206–R208.
Wilson, R., Ainscough, R., Anderson, K., Baynes, C., Berks, M.,
Pelissier, T., Tutois, S., Tourmente, S., Deragon, J.M., and Picard,
Bonfield, J., Burton, J., Connell, M., Copsey, T., et al. (1994). 2.2
G. (1996). DNA regions flanking the major Arabidopsis thaliana satel-
Mb of contiguous nucleotide sequence from chromosome III of C.
lite are principally enriched in Athila retroelement sequences. Genet-
elegans. Nature 368, 32–38.
ica 97, 141–151.
Wolffe, A.P. (1998). When more is less. Nat. Genet. 18, 5–6.
Pirrotta, V., and Rastelli, L. (1994). White gene expression, repressive
chromatin domains and homeotic gene regulation in Drosophila. Wolffe, A.P., and Hayes, J.J. (1999). Chromatin disruption and modi-
Bioessays 16, 549–556. fication. Nuc. Acids Res. 27, 711–720.
Plasterk, R.H., Izsvak, Z., and Ivics, Z. (1999). Resident aliens: the Xu, X., Hsia, A.P., Zhang, L., Nikolau, B.J., and Schnable, P.S. (1995).
Tc1/mariner superfamily of transposable elements. Trends Genet. Meiotic recombination break points resolve at high rates at the 5⬘
15, 326–332. end of a maize coding sequence. Plant Cell 7, 2151–2161.
Richards, E.J., and Dawe, R.K. (1998). Plant centromeres: structure Zhang, M. (1998). Identification of protein-coding regions in Arabi-
and control. Curr. Opin. Plant Biol. 1, 130–135. dopsis thaliana genome based on quadratic discriminant analysis.
Round, E.K., Flowers, S.K., and Richards, E.J. (1997). Arabidopsis Plant Mol. Biol. 37, 803–806.
thaliana centromere regions: genetic map positions and repetitive
DNA structure. Genome Res. 7 1045–1053. GenBank Accession Numbers
Rubin, G.M. (1998). The Drosophila genome project: a progress
report. Trends Genet. 14, 340–343. The GenBank accession numbers for the sequences reported in this
paper are as follows: AF007269 (F2N1), AF069300 (F3D13),
San Miguel, P., Tikhonov, A., Jin, Y.K., Motchoulskaia, N., Zakharov,
AF096370 (F11O4), AF104919 (T15B16), AC007138 (T7B11),
D., Melake-Berhan, A., Springer, P.S., Edwards, K.J., Lee, M., et al.
(1996). Nested retrotransposons in the intergenic regions of the AF001308 (T10M13), AF075597 (T2H3), AF069298 (T14P8),
maize genome. Science 274, 765–768. AC002330 (T10P11), AC004044 (T5J8), AF069442 (T4I9), AC005275
(F4C21), AF071527 (F9H3), AC005142 (T5L23), AF096372 (T5H22),
Schmidt, R., West, J., Love, K., Lenehan, Z., Lister, C., Thompson,
AF077408 (T7M24), AF128394 (T25H8), AF077409 (T24M8),
H., Bouchez, D., and Dean, C. (1995). Physical map and organization
AF075598 (T24H24), AF076274 (T27D20), AF069441 (T19B17),
of Arabidopsis thaliana chromosome 4. Science 270, 480–483.
AF076243 (T26N6), AF074021 (F4H6), AF149414 (T19J18), AF118223
Schnable, P.S., Hsia, A.P., and Nikolau, B.J. (1998). Genetic recom- (T4B21), and AF128393 (T1J1).
bination in plants. Curr. Opin. Plant Biol. 1, 123–129.
Springer, P.S., McCombie, W.R., Sundaresan, V., and Martienssen,
R.A. (1995). Gene trap tagging of PROLIFERA, an essential MCM2-
3-5-like gene in Arabidopsis. Science 268, 877–880.
Sulston J., Du, Z., Thomas, K., Wilson, R., Hillier, L., Staden, R.,
Halloran, N., Green, P., Thierry-Mieg, J., Qiu, L., et al. (1992). The
C. elegans genome sequencing project: a beginning. Nature 356,
37–41.
Sun, X., Wahlstrom, J., and Karpen, G. (1997). Molecular structure
of a functional Drosophila centromere. Cell 91, 1007–1019.
Sundaresan, V., Springer, P., Volpe, T., Haward, S., Jones, J.D.,
Dean, C., Ma, H., and Martienssen, R. (1995). Patterns of gene action
in plant development revealed by enhancer trap and gene trap trans-
posable elements. Genes Dev. 9, 1797–1810.
Taillebourg, E., and Dura, J.-M. (1999). A novel mechanism for P
element homing in Drosophila. Proc. Acad. Natl. Sci. USA 96, 6956–
6961.
Thompson, H., Schmidt, R., Brandes, A., Heslop-Harrison, J.S., and
Dean, C. (1996). A novel repetitive sequence associated with the
centromeric regions of Arabidopsis thaliana chromosomes. Mol.
Gen. Genet. 253, 247–252.
Thompson-Stewart, D., Karpen, G.H., and Spradling, A.C. (1994). A
transposable element can drive the concerted evolution of tandemly
repetitious DNA. Proc. Natl. Acad. Sci. USA 91, 9042–9046.
Thuriaux, P. (1977). Is recombination confined to structural genes
on the eukaryotic genome? Nature 268, 460–462.
Timmermans, M.C., Das, O.P., and Messing, J. (1996). Characteriza-
tion of a meiotic crossover in maize identified by a restriction frag-
ment length polymorphism-based method. Genetics 143, 1771–
1783.