Plant Physiol.
(1993) 103: 285-288
Rapid Communication
Activation of Cytosolic Pyruvate Kinase by
Polyethylene Glycol’
Florencio E. Podestá and William
Departments of Biology and Biochemistry, Queen’s University, Kingston, Ontario, Canada K7L 3 N 6
Homogeneous cytosolic pyruvate kinase from endosperm of
germinating castor oil (Ricinus communis 1. cv Hale) seeds was
potently activated by polyethylene glycol. The addition of 5%
(w/v) polyethylene glycol t o the pyruvate kinase reaction mixture
caused a 2.6-fold increase in maximal velocity and 12.5- and 2-
fold reductions in K,,, values for phosphoenolpyruvate and ADP,
respectively. Clycerol, ethylene glycol, and bovine serum albumin
also enhanced pyruvate kinase activity, albeit t o a lesser extent
than polyethylene glycol. The addition of 5% (w/v) polyethylene
glycol t o the elution buffer during high-performance gel filtration
chromatography of purified cytosolic pyruvate kinase helped to
stabilize the active heterotetrameric native structure of the enzyme.
A higher degree of inhibition by MgATP, but lower sensitivity t o
the inhibitors 3-phosphoglycerate and fructose-1,6-bisphosphate,
was also observed in the presence of 5% (w/v) polyethylene glycol.
It i s concluded that (a) plant cytosolic pyruvate kinase activity and
regulation, like that of other regulatory pyruvate kinases, i s modi-
fied by extreme dilution in the assay medium, probably as a result
of deaggregation of the native tetrameric enzyme, and (b) ATP i s
probably the major metabolic effector of germinating castor en-
dosperm cytosolic pyruvate kinase i n vivo.
PK (EC 2.7.1.40) is considered to be a key regulatory
enzyme of the glycolytic pathway. Purified PK, from 5-d-old
germinating COS has been shown to exist as a heterotetramer
having a native molecular mass of about 240 kD and con-
sisting of equal amounts of two types of related but non-
identical subunits of 57 and 56 kD (Plaxton, 1988, 1989).
Detailed kinetic studies with the homogeneous PK, from
germinating COS have not revealed any allosteric activators
for the enzyme but suggested a regulatory mechanism based
on ATP levels and pH-dependent alterations in the enzyme’s
response to various metabolite inhibitors (Podestá and Plax-
ton, 1991, 1992). Cytosolic enzymes such as PL, however,
are present in vivo at far higher concentrations than are used
during routine in vitro assays (Fulton, 1982). This difference
is believed to be particularly significant for enzymes consid-
ered to be important for metabolic regulation, because their
structure, and hence their kinetic properties, may be affected
by both homologous and heterologous protein-protein inter-
actions (Fulton, 1982; Medina et al., 1985; Manetas, 1990;
AragÓn and Sols, 1991; Salvucci, 1992; Timasheff, 1992).
This work was supported by the Natural Sciences and Engineer-
ing Research Council of Canada.
* Corresponding author; fax 1-613-545-6617.
285
C. Plaxton*
The interactions between enzymes that probably exist at the
high protein concentrations prevailing in vivo can be specif-
ically promoted in vitro by the addition of compatible solutes
(e.g. PEG) to the reaction mixture. The mechanism apparently
involves exclusion of the protein from the binary solvent,
thus increasing the local enzyme concentration and favoring
protein aggregation (Medina et al., 1985; AragÓn and Sols,
1991; Salvucci, 1992; Timasheff, 1992). The in vitro activity
of rat liver PK, for instance, as well as that of a variety of
other nonplant and plant regulatory enzymes, is enhanced
by the presence of PEG (Medina et al., 1985; Manetas, 1990;
AragÓn and Sols, 1991; Salvucci, 1992). In this report we
demonstrate that the kinetic and regulatory characteristics of
a plant PK, are significantly influenced by the addition of
PEG to the assay medium and argue that its effect is elicited
through the stabilization of the enzyme’s native heterotetra-
meric structure.
MATERIALS A N D METHODS
Chemicals and Plant Material
ADP, ATP, PEG (average M,SOOO), Mes, Bis-Tris-propane,
Fru-1,6-bisP, BSA, and rabbit muscle lactic dehydrogenase
were from Sigma Chemical Co. NADH and 3-P-glycerate
were from Boehringer-Mannheim. A11 other reagents were of
analytical quality purchased from BDH Chemicals. A11 solu-
tions were prepared with Milli-Q-processed water. Seeds of
the castor plant (Ricinus communis L. cv Hale) were obtained
from Bothwell Enterprises (Plainview, TX).
Enzyme Purification and Assay
PK, was purified to homogeneity from 5-d-old germinating
COS as described before (Plaxton, 1988), with the modifica-
tion introduced by Podestá and Plaxton (1991). Enzymic
activity was assayed at 3OoC in a reaction mixture of 1 mL
containing 25 m Bis-Tris-propane/Mes/HCl (pH 7.2), 20
m KCl, 10 mM free Mg2+(from MgC12), PEP, and ADP as
indicated in the table and figure legends, 2 mM DTT, 0.15
ITLM NADH, various amounts of PEG or other co-solutes, and
2 units of desalted lactic dehydrogenase. PEG was added
Abbreviations: COS, castor oil seed endosperm; I50, concentration
of inhibitor necessary to reduce enzymic activity by 50%; PEP,
phosphoenolpyruvate; PK, pyruvate kinase; PK,, cytosolic pyruvate
kinase.
286 Podestá and Plaxton
from a 50% (w/v) stock solution in distilled water. Fresh
stock solutions of PEP and ADP were prepared weekly, and
their respective concentrations were verified spectrophoto-
metrically as described by Podestá and Plaxton (1992). One
unit of PK, activity is defined as the amount of enzyme
resulting in the production of 1 pmol of pyruvate min-’.
Assays were initiated by the addition of 0.03 to 0.06 pg of
purified PK,. Activity in a11 assays was proportional to the
amount of enzyme added and remained linear with respect
to time.
Kinetic Studies
Substrate saturation data were fitted to the Hill equation
using a nonlinear least-squares regression computer kinetics
program kindly provided by Dr. Stephen Brooks (Brooks,
1992). Because our recent study (Podestá and Plaxton, 1992)
demonstrated that free (uncomplexed) PEP and ADP are the
catalytically effective forms of the substrates for the COS
PK,, the respective K, values for free PEP and ADP are
reported. Concentrations of total MgClz, PEP, ADP, and ATP
added to achieve the free Mg”, PEP, and ADP or MgATP
concentrations required were calculated as described previ-
ously (Podestá and Plaxton, 1992). Iso values were obtained
by fitting inhibition data to the Job equation using the afore-
mentioned computer kinetics program (Brooks; 1992).
Cel Filtration HPLC
Gel filtration HPLC was performed using an 8- X 150-mm
Shodex GW-804 column connected to a Waters 625 system,
equipped with a model 486 absorbance detector (set to 280
nm) and a U6K injector. Peak integration and molecular mass
estimates were performed using the Waters Baseline 810
Chromatography Workstation software (version 3.3). The
following standards were used in the calibration of the col-
umn: thyroglobulin (660 kD), apoferritin (440 kD), catalase
(232 kD), P-amylase (215 kD), alcohol dehydrogenase (150
kD), malic dehydrogenase (73 kD), BSA (66 kD), chymotryp-
sinogen (25 kD), and ribonuclease A (13.7 kD). The standard
buffer used for a11 runs contained 50 mivl Hepes/NaOH (pH
7.0) and 5 mM MgC12, to which the additions mentioned in
the text were made. Flow rate was 0.5 mL min-’, and the
sample volume was 20 pL.
RESULTS
Effect of Organic Co-Solutes on PK, Activity
As shown in Figure 1, COS PK, was activated by PEG,
glycerol, and ethylene glycol in a concentration-dependent
manner when assayed under optimal conditions. Maximal
activation occurred at 10 to 15% (v/v) for glycerol (1.4-fold),
10% (v/v) for ethylene glycol (1.7-fold), and 10% (w/v) for
PEG (2.6-fold), and higher concentrations reduced the extent
of activation (Fig. 1).If the PK, was assayed at a subsaturating
concentration of substrates (0.1 mM PEP, 5 p~ ADP), then
activation by 10% (v/v) glycerol, 10% (v/v) ethylene glycol,
and 10% (w/v) PEG was increased to 1.8-, 2.3-, and 2.9-fold,
respectively. BSA, tested at 1 and 5 mg mL-’ in the assay
medium, elicited a 1.4-fold activation of PK, under optimal
Plant Physiol. Vol. 103, 1993
- 300 I , , , ,
-
O
C
o
O PEG
0 Llhylclle glycol
V Glycerol
c
.-
.-
.-
>
t
assay conditions, whereas no effect was observed at subsa-
turating PEP and ADP.
Subsequent experiments were canied out with the best
activator, PEG. A PEG concentration of 5% (w/v) was rou-
tinely used to avoid practical problems arising from the high
viscosity of the reaction medium at higher PEG concentra-
tions. The presence of this polyalcohol substantially increased
the enzyme’s affinity for its substrates (Fig. 2, Table [). The
PEP and ADP saturation kinetics were not affected to the
same degree, however, because 5% (w/v) PEG caused the K,
values for free PEP and ADP to be reduced by aboui 12.5-
and 2-fold, respectively (Table I). A 2.6-fold increase in V,,,
was calculated by extrapolation to infinite concentrations of
free PEP and ADP (Table I). Hill coefficients were not signif-
icantly different from 1 in a11 cases. &,O values for severa1
metabolite inhibitors of COS PK, in the presence and absence
of 5% (w/v) PEG at pH 7.2 are listed in Table 11. Notably,
the Iso value for MgATP at pH 7.2 was lowered by approxi-
mately 2.2-fold by the addition of 5% (w/v) PEG. By contrast,
sensitivity of the enzyme to two other potentially important
inhibitory metabolites (3-P-glycerate and Fru-1,6-bisI’) (Po-
destá and Plaxton, 1991) was slightly reduced in the presence
of the aggregating reagent. The extent of MgATP inhibition
at pH 6.5 was not significantly altered by the inclusion of 5%
(w/v) PEG (Iso 2.2 mM in aqueous solution versus 2.1 mM in
5% [w/v] PEG).
Effect of PEG on PK, Quaternary Structure
PK, quatemary structure was investigated by gel filtration
HPLC (data not shown). Injection of 10 pg of homogeneous
COS PK,, followed by elution with the standard acpeous
buffer, resulted in about 73% of the protein eluting in an
inactive form at an estimated molecular mass of 56 kD
(attributable to the free subunits), 8% eluted as a 232-kD
heterotetramer, and 19% eluted as a dimer of 112 kD. If 5%
(w/v) PEG was added to the column equilibration/elution
buffer, the tetramer and dimer peaks increased to 14 and
22%, respectively, with 64% of the total protein still eluting
at 56 kD. Although the 232-kD peak was active, no PK
activity was ever recovered from the free subunits and only
traces could be observed in the 112-kD peak, probably re-
 Castor Oil Seed Cytosolic Pyruvate Kinase 287

'2'
.- 300 Table II. Effect of PEC addition on the inhibition of germinating
a,
+ castor oil seed PK, by severa1 metabolites
C? / 5 0 values were determined at pH 7.2 using approximate K,
.-t> Q
200 concentrations of free PEP (100 and 10 p~ in the absence and
O € presence of PEG, respectively).All data represent the means (+ SE)
.-O';. .- I of at least three independent determinations.
.-
O E ioc
b
I50

:l-
m o
E additions
3-
- 0
O 100 200 300 600
[free P E P ] ( y M )
Figure 2. Effect of PEC on the PEP saturation kinetics of germinating
castor oil seed PK,. Activity was measured with 1 m M total ADP at
pH 7.2 in the absence (O)or presence (O)of 5% (w/v) PEC.
flecting a tailing off in the elution of the active heterotetra-
meric form. Injection of a highly concentrated aliquot of COS
PK, (500 pg), followed by elution with the standard aqueous
buffer, resulted in all of the enzyme eluting as a 232-kD
heterotetramer. This molecular mass is very close to the value
of 240 kD that was previously determined for the homoge-
neous native COS PK, by Superose 6 fast protein liquid
chromatography in the presence of 25% (v/v) glycerol (Plax-
ton, 1988).
DlSCUSSlON
PK, is believed to play a key role in regulating the metabolic
switch from gluconeogenesis to glycolysis that occurs when
germinating COS undergoes anaerobiosis (Kobr and Beevers,
1971; Podestá and Plaxton, 1991, 1992). Current evidence
suggests that the anoxia-dependent increase in the in vivo
activity of COS PK, (Kobr and Beevers, 1971) arises from a
release of the enzyme from metabolite inhibition rather than
via the action of allosteric activators (Podestá and Plaxton,
1991). In this paper we demonstrate that the presence of
organic co-solutes, but in particular PEG, a compound often
used to promote protein-protein interactions in vitro (Medina
et al., 1985; Manetas, 1990; AragÓn and Sols, 1991; Salvucci,
1992; Timasheff, 1992), greatly influences the kinetic and
structural properties of purified COS PK,.
Most notably, the addition of 5% (w/v) PEG to the COS
PK, assay mixture caused enzyme activity at saturating and
subsaturating substrate concentrations to be greatly enhanced
(Fig. 2, Table I). The same trend was observed when glycerol
Table 1. Effect of PEC addition on substrate saturation kinetics of
germinating castor oil seed PK
All assays were conducted at pH 7.2; the fixed total and free co-
substrate concentrations were: PEP, 1.6 and 0.5 mM, respectively;
ADP, 1 m M and 35 p ~ respectively.
, All data represent the means
(+SE) of at least three independent determinations.
+5% (w/v)
Parameter No Additions
PEG
133.0 + 9.5 348.0 f 7.7
V,, (units mg-')
93.8 + 2.3
K, free PEP (WM) 7.5 f 0.7
K, free ADP (WM) 3.9 f 0.7 1.9 f 0.1
Metabolite No + 5% (W/V)
PEC
flM
MgATP 5.8 f 1.0 +
2.6 0.5
Fru-1,6-bisP +
3.0 0.5 4.2 f 0.5
3-P-glycerate 3.6+ 1.1 +
4.3 0.5
or ethylene glycol was added to the assay medium. BSA
effects were less pronounced, which could be due to a dele-
terious effect of heterologous protein-protein interactions.
Overall, the results indicate not only a modification of the
affinity of PK, for its substrates but also that there is more
active enzyme when PEG or other co-solutes are present.
This conclusion is substantiated by the fact that the presence
of 5% (w/v) PEG helps to stabilize the enzyme's active
heterotetrameric structure during gel filtration HPLC of di-
luted PK,.
Another significant effect of PEG addition was the marked
reduction in the PK,'s Zso for MgATP from about 5.8 to 2.6
m~ (measured at pH 7.2; Table 11). This implies that the PK,
activity of aerobic germinating COS should be almost com-
pletely suppressed in vivo, because a cytosolic ATP concen-
tration of 6 mM (declining to 2 mM under anoxia) can be
calculated (Kobr and Beevers, 1971; Podestá and Plaxton,
1991).
The above findings indicate that the kinetic and regulatory
properties of COS PK, are influenced by its own concentra-
tion. In fact, Plaxton (1988) reported that the activity of COS
PK, is highly labile when the purified enzyme is subjected to
dilution and attributed a heterotetrameric native structure to
the homogeneous enzyme. Our results also show that only
from the tetrameric form could PK activity be recovered after
gel filtration HPLC. These findings suggest that kinetic data
obtained in a dilute aqueous medium may not accurately
reflect the true catalytic properties of COS PK,. Indeed, the
cytosolic protein concentration in 5-d-germinated COS may
be as high as 100 mg mL-' and that of PK, between 0.055
and 0.1 mg mL-'.' By contrast, typical in vitro activity assays
contain COS PK, concentrations of between 30 and 60 ng
mL-'. This suggests that an increase in homologous protein-
protein interactions, probably promoting aggregation, is re-
sponsible for the PEG activation of COS PK,. By contrast,
Cytosolic protein and PK, concentrations in vivo were estimated
on the basis that (a) the concentration of soluble protein in 5-d-old
germinating COS is about 20 mg g-' fresh weight, (b) the PK, activity
of 5-d-germinating COS ranges from 1 to 2 units g-' fresh weight,
and (c) the maximal specific activity of COS PK, ranges from 200 to
350 units mg-' (Plaxton, 1988;Podesti and Plaxton, 1991; this work).
It was assumed that 50% of the total soluble protein is cytosolic and
that the cytosol represents 10% of the total cell volume.
288 Podestá and Plaxton
most other enzymes that have been examined i n the presence
of PEG, or by various in situ approaches, do not display a n y
significant alteration i n their kinetic behavior (Medina et al.,
1985; AragÓn a n d Sols, 1991). The exception seems to be t h e
case of certain regulatory enzymes such as mammalian phos-
phofructokinase a n d liver-type PK (Medina e t al., 1985;
AragÓn a n d Sols, 1991) a n d plant PEP carboxylase a n d
Rubisco activase (Manetas, 1990; Salvucci, 1992). This cor-
relation places COS PK, among the regulatory enzymes a n d
thus further emphasizes its preeminent role in the regulation
of plant glycolysis (Kobr a n d Beevers, 1971; Podestá a n d
Plaxton, 1991, 1992). Although this investigation provides a
means of examining plant PK, kinetic properties in an in vitro
environment that may be more closely related to the i n situ
situation, further studies are needed to elucidate whether
protein-protein interactions play a role i n the regulation of
this enzyme in vivo.
Received April 7, 1993; accepted May 26, 1993.
Copyright Clearance Center: 0032-0889/93/103/0285/04.
LITERATURE CITED
AragÓn JJ, Sols A (1991) Regulation of enzyme activity in the cell:
effect of enzyme concentration. FASEB J 5 2945-2950
Plant Physiol. Vol. 103, 1993
Brooks SPJ (1992) A simple computer program with statistical tests
for analysis of enzyme kinetics. Biotechniques 13: 906-91 1
Fulton AB (1982) How crowded is the cytoplasm? Cell 30 343-347
Kobr MJ, Beevers H (1971) Gluconeogenesis in the castor bean
endosperm. 1. Changes in glycolytic intermediates. Plant Physiol
47: 48-52
Manetas Y (1990) A re-examination of NaCl effects on phosphoen-
olpyruvate carboxylase at high (physiological)enzyme concentra-
tions. Physiol Plant 7 8 225-229
Medina R, AragÓn JJ, Sols A (1985) Effect of polyethylene glycol
on the kinetic behaviour of pyruvate kinase and other potentially
regulatory liver enzymes. FEBS Lett 180 77-80
Plaxton WC (1988) Purification of pyruvate kinase from germinating
castor bean endosperm. Plant Physiol86: 1064-1069
Plaxton WC (1989) Molecular and immunological characterization
of plastid and cytosolic pyruvate kinase isozymes from castor-oil-
plant endosperm and leaf. Eur J Biochem 181: 443-451
Podestá FE, Plaxton WC (1991) Kinetic and regulatory properties of
cytosolic pyruvate kinase from germinating castor beans. Biochem
J 279 495-501
Podestá FE, Plaxton WC (1992) Plant cytosolic pyruvate kinase: a
kinetic study. Biochim Biophys Acta 1160 213-220
Salvucci ME (1992) Subunit interactions of Rubisco activase: poly-
ethylene glycol promotes self-association, stimulates ATPase and
activation activities, and enhances interactions with Rubisco. Arch
Biochem Biophys 298: 688-696
Timasheff SN (1992) Solvents effects on protein stability. Curr Opin
Struc. Biol. 2: 35-39