TRANSHYDROGENASE REGULATION 11
Colowick, S. P., Kaplan, N. O., Neufeld, E. F., and Ciotti, M.
M. (1952), J . Biol. Chem. 195, 95-105.
Dieter, H., Koberstein, R., and Sund, H. (1974), FEBS Lett.
47, 90-93.
Ginsburg, A., and Mehler, A. H. (1966), Fed. Proc., Fed. Am.
SOC.Exp. Biol. 25, 407.
Goldbeter, A. (1974), J . Mol. Biol. 90, 185-190.
Hojeberg, B., and Rydstrom, J. (1976), Znt. Congr. Biochem.,
IOth, 07-4-117.
Kaplan, N. 0. (1957), Methods Enzymol. 3, 873-876.
Kaplan, N. 0. (1972), Haroey Lect. 66, 105-133.
Kaplan, N. O., Colowick, S. P., and Neufeld, E. F. (1952), J .
Biol. Chem. 195, 107-1 19.
Kaplan, N. O., Colowick, S. P., Neufeld, E. F., and Ciotti, M.
M. (1953), J . Biol. Chem. 205, 17-29.
Kirtley, M. E., and Koshland, D. E., Jr. (1967), J. Biol. Chem.
242, 4192-4205.
Koshland, D. E., Jr., Nemethy, G., and Filmer, D. (1966),
Regulatory Properties of the Pyridine Nucleotide Transhydrogenase
from Pseudomonas aeruginosa. Active Enzyme Ultracentrifugation
Studies?
FranGois Widmert.8 and Nathan 0. Kaplan*
ABSTRACT: Active enzyme ultracentrifugation studies of the
pyridine nucleotide transhydrogenase from Pseudomonas
aeruginosa (EC 1.6.1.1.) show that the enzymatic reaction is
catalyzed by a molecular species characterized by an ~ 2 0
value of about 34 S, whatever the reduced substrate may be
(tri- or diphosphopyridine nucleotide). The filamentous ag-
gregated form of the enzyme ( ~ 2 0 = , ~ 121 S and higher),
identified by previous investigations (Cohen, P. T., and Kaplan,
w e have reported, in the preceding paper of this issue
(Widmer and Kaplan, 1976), that the MWC allosteric model
(Monod et al., 1965) might be used as a framework to explore
the regulatory characteristics of the enzyme transhydrogenase
from Pseudomonas aeruginosa. The R state is favored by
TPNH,’ 2’-AMP, and several other 2’-phosphate nucleotides,
whereas TPN+ and inorganic phosphate show more affinity
for the T state. On the other hand, the structure of PATH, as
elucidated by ultracentrifugation (Cohen, 1967; Cohen and
Kaplan, 1970a) and electronmicroscopy (Louie et a]., 1972),
might prompt one to assume that association-dissociation
+ From the Department of Chemistry, University of California, San
Diego, La Jolla, California 92093. Receiued April 6, 1976. This work was
supported by grants from the American Cancer Society (BC-60-P) and
from the National Institutes of Health (USPHS) (CA 11683-0s).
* Postdoctoral fellow of the Swiss National Science Foundation.
5 Present address: NestlC Products Technical Assistance Co. Ltd.,
Research Department, Biochemistry Section, P.O. Box 88, CH-I 8 14 La
Tour-de-Peilz, Switzerland.
I Abbreviation: PATH, Pseudomonas aeruginosa pyridine nucleotide
transhydrogenase; for other abbreviations, see footnote 1 of the preceding
paper in this issue (Widmer and Kaplan, 1976).
Biochemistry 5, 365-38 5 .
Louie, D. D., and Kaplan, N. 0. (1970), J . Biol. Chem. 245,
5691 -5698.
Louie, D. D., Kaplan, N . O., and McLean, J. D. (1972), J .
Mol. Biol. 70, 65 1-664.
Monod, J., Wyman, J., and Changeux, J . P. (1965), J . Mol.
Biol. 12, 88-118.
P-L Biochemicals, Inc., Circular OR- 18 (1 965), Ultraviolet
Absorption Spectra of Pyridine Nucleotide Coenzymes and
Coenzyme Analogs.
Rydstrom, J., Hoek, J. B., and Hojeberg, B. (1973), Biochem.
Biophys. Res. Commun. 52, 421-429.
Walter, P., and Rubin, B. (1966), Biochem. Prep. 10, 166-
170.
Webb, J. L. (1963), in Enzymes and Metabolic Inhibitors, Vol.
2, New York, N.Y., Academic Press, 439, 446.
Widmer, F., and Kaplan, N. 0. (1976), Biochemistry 15,
(following paper in this issue).
N. 0. (1970), J . Biol. Chem. 245, 2825-2836; Louie, D. D.,
Kaplan, N. O., and Mc Lean, J. D. (1972), J . Mol. Biol. 70,
65 1 -664), appears, therefore, to be an inactive species. The
, ~ physiological implicationsof the enzyme are discussed. Several
lines of evidence lead to the conclusion that the transhydro-
genase might act as an essential link between carbohydrate
catabolism and the respiratory chain.
reactions could characterize the catalytic mechanism of the
enzyme. Such a regulatory mechanism would not be in
agreement with the fairly general rule that allosteric enzymes
have a fixed number of protomers, which is independent of any
allosteric transitions which might occur.
The main component of the sedimentation pattern of PATH
, ~ of 121 S,
in its native form is characterized by an ~ 2 0value
whereas aggregated material, not actually in solution, sedi-
ments with an even higher speed (Cohen and Kaplan, 1970a).
These two components should correspond to the rodlike shaped
polydisperse structure seen on electronmicrographs of the
native enzyme (Louie et al., 1972). On the other hand, Cohen
and Kaplan (1970a) found that in the presence of 1 mM 2’-
AMP or 1 mM TPN+ the sedimentation pattern is homoge-
neous, and characterized by a single component with an ~ 2 0 , ~
of 33.8 S. This component has been assumed by Louie et al.
( 1972) to correspond to the uniform population of cylindrical
particles (900 000 dalton) seen on electronmicrographs after
addition of 1 mM 2’-AMP. A very small amount of such units
are already seen on electronmicrographs of the native struc-
ture, and should correspond to the component with an ~ 2 0of, ~
BIOCHEMISTRY, VOL. 1 5 , NO. 21, 1976 4699

. . . . .
.,: . . . . . . . .. .. . .
.. .. ..
about 34 S, which can be observed in a low concentration on
the heterogeneous sedimentation pattern of native PATH
(Cohen and Kaplan, 197ba). Three different components have
thus been identified on this pattern; they have been called
components I, 11, and III (I: ~ 2 0 =. 33.8
~ S; 11: s~o,,, = 121 S;
111: aggregated material).
The active enzyme ultracentrifugation technique first de-
veloped by R. Cohen (1963) and modified by Kemper and
Everse ( I 973) allows for the determination of the value
of an enzyme in its catalytically active form; this determination
is done under kinetic assay conditions. To date, the following
reactions of PATH have been studied with this technique:
2'-AMP-activated DPNH-(TN)DPN+ reaction, 2'-AMP-
activated TPNH-(TN)DPN+ reaction, and TPNH-(TN)-
DPN+ reaction (Louie et al., 1972). The last named reaction
was characterized by an 320.wvalue of 1IO S, the other ones by
an s20.wvalue between 28 S and 30 S. It was therefore possible
to correlate the activation by 2'-AMP with the disaggregation
process. As for the third reaction, the ~ 2 0 of. ~1 IO S was is-
sumed to indicate that the rodlike structure is the active species
for the given reaction (Louie et al., 1972). This assumption,
we now believe, was not correct, mainly because no T P N H
regenerating system was used. Consequently, the concentra-
tions of the allosteric ligands T P N H and TPN+ were not kept
constant. For such a situation, it is likely that the ultracen-
trifugation pattern was affected by the continuous shift of the
allosteric equilibrium throughout the experiment. Further-
more, the ~ 2 0 value
. ~ yielded by an active enzyme ultracen-
trifugation can be an apparent value only (as for usual ultra-
centrifugation techniques) if the enzyme is characterized by
an association-dissociation equilibrium. All the available ev-
idence suggests that PATH is such a system.
The main goal of the present report was to ascertain whether
the filamentous aggregated form of PATH might really rep-
resent an active species. It appears from our results that dis-
aggregation into particles of 900 000 molecular weight is a
prerequisite to enzymatic activity.
Materials and Methods
Chemicals and Enzymes. DL-Sodium isocitrate was pur-
chased from Calbiochem. All other reagents, as well as PATH,
were obtained as indicaied in the preceding paper of this issue
(Widmer and Kaplan, 1976). TPN+-specific isocitrate dehy-
drogenase from pig heart was purified by Mr. F. E. Stolzen-
bach according to the procedure of Ochoa (1955).
4700 BIOCHEMISTRY, VOL. 1 5 , NO. 21, 1976
WIDMER A N D KAPLAN
Actiue Enzyme Ultracentrifugation. The experiments were
carried out according to the method of Kemper and Everse
(1973). using a Spinco Model E analytical centrifuge equipped
with a photoelectric scanning system. All the reagents for the
assays were prepared in 0.1 M Tris buffer (pH 7.5) containing
IO mM 2-mercaptoethanol. The PATH dilutions were pre-
pared in 0.01 M Tris buffer (pH 7.5). also containing IO m M
2-mercaptcethanol. Ten microliters of these dilutions were used
for each experiment a t the concentrations indicated in the
text.
Results
DPNH-(TN)DPN+Reaction. This reaction proceeds a t
a very low rate in the absence of 2'-AMP, but is not negligible,
particularly when the solutions are buffered with Tris. The
active enzyme ultracentrifugation of this reaction was carried
out with five different PATH concentrations, between 0.02 and
0.0035 enzyme units in the 10 pl initiating the reaction, and
with substrate concentrations of 0.1 mM.
A heterogeneous pattern was always obtained and on each
scan three different reg.ons could be identified, as exemplified
by Figure 1.
Part I: a well defined boundary is moving down with an ~ ~ 0 , ~
valueof 33.7S.ThesemilogplotoflogXvs. time(Xbeingthe
distance between the center of rotation and the boundary) is
straight for about 40 min, then slightly bends downward. This
s 2 0 . value
~ of 33.7 S fits quite well with the ~ 2 0value
. ~ of 33.8
S characterizing the TPN+ or 2'-AMP-induced structure of
PATH (Cohen and Kaplan, 1970a).
Part 11: the constant absorbance increase in this part should
be due to a uniformly distributed active enzyme species.
Part 111: a significant activity can be seen a t the bottom of
the cell as soon as the run starts, and the absorbance increase
appears as if due to a diffusion process.
When the concentration of enzyme is raised, the charac-
teristics of parts I1 and III become more conspicuous but, in
this case, all three parts are somewhat overlapping (compare
scans A and B un Figure 1). For a given enzyme concentration,
the absorbance values seen in parts I1 and 111 are increasing
with time.
The heterogeneous pattern just described can be reconciled
with the very first ultracentrifuge runs of native PATH by
Cohen and Kaplan (1970a). because parts I, 11, and III seen
on our scans can be explained by components I, 11, and 111
identified by these authors under usual ultracentrifugation
conditions (high enzyme concentration, absence of any sub-
strate or effector). In both cases, the heterogeneity of the ul-
tracentrifuge patterns can be interpreted as the indication of
a very slow interaction (if at all) among different enzyme
species. Therefore, the similarity between the native enzyme
pattern and the active enzyme pattern for the DPNH-
(TN)DPN+ reaction suggests that, when PATH is layered as
a thin film on the substrate solutions in the ultracentrifuge cell,
the nonuniform population remains characterized by the three
components described for the native enzyme by Cohen and
Kaplan (l970a). This may be related to the fact that the sub-
strates involved in the reaction are unable to alter the allosteric
equilibrium (Widmer and Kaplan, 1976).
A consistent explanation of our results is possible if one as-
sumes that only component I is significantly active for the
DPNH-(TN)DPN+ reaction. The straight log Xvs. time plot
characterizing the boundary of part I, which contains most of
the enzymatic xtivity, is indeed explained by the sedimenta-
tion of a homogeneous active enzyme species with an S Z O . of ~
33.7 S. The slight loss of activity detected after 40 min, and
TRANSHYDROGENASE REGULATION I1
elicited by the downward bending of the semilog plot, might
be accounted for by a slow reassociation p r o x s (in compliance
with the interconversion equilibrium), with a resulting faster
sedimentation of the created aggregates. The assumption of
a slow interconversion rate can also explain the features of parts
11 and Ill. The diffusion-like absorbance increase at the bottom
of the cell (part H I ) , which can be seen even in the very early
stages of the runs, definitely recalls to mind the very fast sed-
imentation of component 111 in the ultracentrifugation of na-
tive PATH (Cohen and Kaplan, 1970a). If the different species
present in the heterogeneous native population (not affected
b) DPNH or ( T U ) D k + )arechxdc&ed by interconver-
$ion equilibria. component 111 must disaggregate to some ex-
tent, in compliance with the equilibrium constants. Therefore.
it is our view that the enzymatic activity seen in part I l l is due
to 3 4 s particles created by the pariial disaggregation of
component 111. A similar line of argument leads to assume that
the enzymatic activity seen in part I1 might be due to 3 4 s
particles resulting from the partial disaggregation of an enzyme
species probably corresponding to the 121s component.
However, it is not yet fully explained why this activity seems
to be uniformly distributed in part I1 (this distribution could
be due to a pression and/or concentration effect).
The assumption that the 3 4 s particles are the only signifi-
cantly active species is substantiated by the fact that most of
the activity is found in part I, where this species is found as a
homogeneous population, while making up only a minute
percentage of the total protein in the ultracentrifuge cell
(Cohen and Kaplan, 1970a: Louie et al., 1972).
TPNH-( TN)DPN+ Reaction with Regenerating System
(Saturating Amount of T P N H ) . The substrate T P N H is as-
sumed to shift the R-T allosteric equilibrium towards the R
state, whereas the product TPN+ favors the T state (Widmer
and Kaplan, 1976). In such a situation, it is obvious that any
active enzyme ultracentrifugation pattern will be complicated
by the continuously varying TPNH and TPN+ concentrations
if no T P N H regenerating system is used. It is therefore ad-
visable to include such a system in the contents of the ultra-
centrifuge cells. In so doing, the concentrations of allosteric
ligands will be kept constant. The regenerating system we used
for our experiments was the TPN+-specific isocitrate dehy-
drogenase system, as described by Cohen and Kaplan (1970b).
For PATH, the experimental conditions consisted of 0.3 mM
TPNH, 0.1 mM (TN)DPN+, and 0.001 enzyme unit.
A homogeneous pattern was obtained (Figure 2A) with a
single symmetrical boundary Characterized by an ~ 2 0 of. ~35.0
S. No loss of activity was encountered during the runs and no
activity whatsoever could be detected in the rest of the cell. The
presence of 0.3 mM T P N H can, therefore, be assumed to in-
duce a homogeneous enzyme population; the ~ 2 0value . ~ is in
agreement with that of the 900 000 molecular-weight species.
The homogeneous pattern indicates that the structural alter-
ation induced by T P N H is very rapid, compared to the slow
interconversion rates assumed in the absence of allosteric li-
gands (see preceding paragraph).
I t has been reporled that 2':AMPactivation can no longer
. ,
be seen for theTPNH-(TNIDPN' reaction when theTPKH
concentration reaches 0.3 mM (Cohen and Kaplan, 1970a).
This fact can be reconciled with our present finding that at this
concentration the allosteric ligand T P N H induces the for-
mation of a homogeneous population similar to the 2'-AMP-
induced population.
TPNH-( TN)DPN+ Reaction without Regenerating Sys-
tem (Saturating Amount of TPNH). When the regenerating
system is omitted, the pattern is heterogeneous (Figure 2B)
BIOCHEMISTRY. VOL
... .. ..._ . ...
. ....
. ... ........-
.. . . :!
FIGUKt. 2 Scans of Ihc T P N H - ( T N ) D P N + racliun. about 40 nun after
the start of the runs (24 630 rpm: 0.001 P A T H units). ( A ) with T P N H
regenerating system: (B) without TPNH regenerating system. Monitoring:
as far Figure I
and strongly resembles the heterogeneous patterns of the
DPNH-(TN)DPN+ reaction (Figure I). However, there are
two major differences. First the log X vs. time plot for the
boundary corresponding to part I bends downward from the
outset (the initial slope of this plot yields an ~ 2 0of. about
~ 35
S). Secondly, the increase with time of the absorbance seen in
part I I is not uniform throughout the ultracentrifuge run. A
fast moving "shoulder" is elicited towards the end of the run.
which would inevitably mask the 35 S boundary (part I),
should PATH be used a t a higher concentration. This com-
plicated ultracentrifugation pattern in the absence of regen-
erating system is certainly due to the fact that the concentra-
tions of TPNH and TPN+ are not kept constant throughout
the experiment.
TPNH-( TN)DPN+ Reaction with Regenerating System
(Subsaturating Amount of TPNH). As a complement to the
two preceding studies, the TPNH-(TN)DPN+ reaction with
TPNH regenerating system was also studied at a relatively low
TPNH concentration ( I O ,LMTPNH; (TN)DPN+ and PATH
were maintained at their previous concentrations).
The significant decrease of T P N H concentration has the
expected consequence that a homogeneous enzyme population
is no longer induced, since the given concentration is far from
being saturating. The scans correspond somewhat to the pat-
tern illustrated in Figure 2B. However, there is a conspicuous
difference: the main component, which shows an ~ 2 0of. 33.3 ~
S , is characterized by a constant activity for at least 40 min,
i.e., the log X vs. time plot is a straight line for this period of
time (the subsequent downward bending is hardly seen). It is
therefore possible to conclude that. in this case, the active
species also corresponds to the low ~ 2 0 species.
. ~ The occur-
rence of heterogeneous material in the rest of the cell is un-
avoidable, since the concentration of the allosteric ligand
T P N H is well under the saturating level.
The following statements summarize the results of the UI-
lracentrilugation \tudies: (I)The native form of PATH is a
- ..
heteroeeneous ~onulation.The interconversion rdteb areslow.
compared to those of the centrifugal separation of the com-
ponents. (2) With the DPNH-(TN)DPN+ reaction, PATH
retains its native form ( I 2 1 s component and aggregated ma-
terial). The elicited transhydrogenase activity is accounted for
by the 3 4 s particles already present in the nonuniform native
population and by similar particles appearing during centrif-
ugal separation throughout the cell by the slow breaking down
of the large aggregates. (3) With the TPNH-(TN)DPN+
15, NO 21. 1976 4701

TPN’
D P ~ P Glyox late N
FADH;! DPN’
ADP -
FIGURE 3: Metabolic significance of PATH (WDH: Warburg-Dick-
ens-Horecker: ED: Entner-Doudoroff).
reaction, a rapid structural alteration of PATH occurs as soon
as the enzyme is layered on the substrates solutions. With a
constant saturating concentration of TPNH, the entire enzyme
population corresponds to the 34s species. If the concentration
of T P N H is not saturating, the 3 4 s induction is not complete
and some larger entities are observed. (4) The dilution to assay
conditions does not promote in itself the disaggregation of the
rodlike polydisperse structure characterizing the native isolated
form of PATH.
Discussion
Cohen and Kaplan (1970a), as well as Louie et al. (1972),
have shown that the polydisperse structure of PATH is a
characteristic of the native isolated form of the enzyme. It
appears from the present ultracentrifugation results that most,
if not all, of the transhydrogenase activity is associated with
the low-molecular form of the enzyme (34s species). This was
found to be the case with whatever reduced substrate was used
(DPNH or TPNH), which means that the filamentous ag-
gregates have little or no enzymatic activity. The present work
also shows that the association-dissociation equilibrium
characterizing the native PATH is not significantly affected
by the presence of DPNH and (TN)DPN+, whereas T P N H
is able to induce a homogeneous 3 4 s population. The dis-
aggregation of the inactive filamentous aggregates is therefore
related to the presence of a ligand favoring the hypothetical
R form of the enzyme, since TPNH-unlike DPNH and
(TN)DPN+-has been recognized as an allosteric ligand
(Widmer and Kaplan, 1976). In terms of the MWC model, the
inactive aggregates would, therefore, somehow correspond to
the nonfunctional T form, which might be prone to spontane-
ous aggregation in the absence of any allosteric ligands. This
does not correspond to any basic assumption or prediction of
the MWC model, which we have used as a framework to ex-
plain the catalytic properties of the transhydrogenase (Widmer
and Kaplan, 1976). One has to keep in mind, however, that the
model offers only an oversimplified first approximation of real
systems, and that the “state” of an allosteric protein may not
in fact be exactly the same whether it is actually bound or
unbound to as stabilizing ligand (Monod et al., 1965). It has
been shown that the low-molecular-weight form of PATH (34s
species) possesses 20-24 FAD’s, presumably corresponding
to the same number of protomers (Louie et al., 1972). Such a
large size is not a priori incompatible with the symmetry re-
quirement of the MWC model, since it has been shown that
this requirement can be fulfilled for oligomers containing as
many as 24 or 60 protomers (Hanson, 1966; Haschemeyer and
de Harven, 1974).
For PATH, association-dissociation processes would have
4702 BIOCHEMISTRY, VOL. 15, NO. 21, 1976
WIDMER A N D KAPLAN
little importance in vivo, since the enzyme is always in the
presence of T P N H and/or T P N + , which stabilize its 34s
forms. It is not possible to rule out that the association-disso-
ciation processes might possibly be the result rather than the
H ~
prime causes of the allosteric mechanisms of PATH. In this
connection, it is worth quoting that experiments with immo-
bilized beef liver glutamate dehydrogenase (EC 1.4.1.3) have
recently shown that allosteric modulation of the given enzyme
is independent of its association-dissociation proclivities
(Horton et al., 1974).
The Pseudomonaceae are aerobic organisms characterized
by a very active glucose oxidation (Doelle, 1969). For the
species Pseudomonas aeruginosa, the Embden-Meyerhoff
pathway is not functional, since the production of pyruvate
takes place by using the pentose pathway (Warburg-Dick-
ens-Horecker pathway) and the Entner-Doudoroff pathway
(Stern et al., 1960). The oxidation of pyruvate could occur by
means of the citric acid cycle, but it seems more likely that the
glyoxylate cycle is preferred. DPNH is used in a respiratory
chain which has 0’ as a terminal electron acceptor (see Figure
3) or N03- in the case of a low oxygen level (Fewson and
Nicholas, 1961).
It is our view that activation of PATH by 2’-AMP and re-
lated nucleotides is an in vitro property, and hence has no
physiological importance (Widmer and Kaplan, 1976).
Therefore, PATH can be considered as a true unidirectional
catalyst. In such a case, its physiological role is to promote
TPNH oxidation with a concomitant generation of DPNH.
This irreversible catalysis certainly corresponds to a specific
metabolic necessity. The pentose phosphate pathway and the
Entner-Doudoroff pathway make use of the enzymes glu-
cose-6-phosphate dehydrogenase (EC 1.1.1.49) and 6-phos-
phogluconate dehydrogenase (EC 1.1.1.44), which are TPNf
specific. In order for their reactions to proceed at an appre-
ciable rate, TPNH has to be by some means reoxidized. That
can undoubtedly occur through biosynthetic activity, but the
presence of path definitely provides a more direct way of TPN+
evolution. An equally important advantage is that hydrogen
can be readily transferred from T P N H to DPN+; DPNH can
then enter the respiratory chain and promote ATP synthesis.
Therefore, PATH might be endowed with a central role in the
metabolism of the microorganism (see Figure 3). ( 1 ) It pro-
vides the oxidation power needed for glucose catabolism. (2)
The direct link to the respiratory chain allows this catabolism
to be a source of ATP.’ An in vivo activation by 2’-AMP and
related nucleotides would suppress the unidirectional character
of PATH. This would be unfavorable to the microorganism,
because such an event would oppose TPN+ evolution and
disturb the link between carbohydrate oxidation and respira-
tory chain.
The regulatory enzymes, which have provided the basis for
working out of the MWC model, have two common charac-
teristics: they operate in metabolic pathways immediately after
a branching point, and for each of them the specific inhibitor
(allosteric effector) is the terminal metabolite of the corre-
sponding pathway (Monod et al., 1963). This observation ex-
plains the limitative point of view according to which allosteric
effectors do not have any direct chemical or metabolic relation
with the possible stubstrates, coenzymes or products of the
When the enzyme was discovered, the Pseudomonas strain was grown
on citrate as the sole source of carbon. I n such a case, the metabolic role
of the transhydrogenase might be quite similar, because the isocitrate
dehydrogenase present in the studied microorganism is a TPN+-specific
enzyme (Colowick et al.. 1952).
INTERACTION OF COENZYMES WITH ISOCITRATE DEHYDROGENASE
enzyme they act upon. For PATH, the situation is exactly the
opposite (the inhibitor TPN+ is a potential substrate) and can
be explained by the particular metabolic role of the enzyme
(see Figure 3). Unlike the above-mentioned regulatory en-
zymes, PATH is not the first catalyst of a multistage metabolic
pathway, but the essential link between two pathways. Since
the enzyme has to comply with the physiological necessity of
unidirectional catalysis, the direct control by the product
TPN+ appears to be the best way of regulation (if not the only
possible one). The fact that TPN+ behaves as negative effector
explains the irreversibility of TPN+ evolution, but is also re-
sponsible for the TPN+ inhibition of the TPNH-DPN+ re-
action. The latter effect could correspond to a kind of “meta-
bolic buffer effect” averting too large a consumption of TPNH,
which is also needed for biosynthetic purposes.
References
Cohen, P. T. (1967), Ph.D. Thesis, Brandeis University,
Waltham, Mass., Ann Arbor, Mich., University Microfilms
Inc., No. 67- 16542.
Cohen, P. T., and Kaplan, N. 0. (1970a), J. Biol. Chem. 245,
2825-2836.
Cohen, P. T., and Kaplan, N. 0. (1 970b), J. Biol. Chem. 245,
4666-4672.
Cohen, R . (1963), C. R . Hebd. Seances Acad. Sci., Ser. C 256,
Coenzyme Binding by Triphosphopyridine Nucleotide Dependent
Isocitrate Dehydrogenase from Beef Liver. Equilibrium and Kinetics
Studies?
Marie France Carlier* and Dominique Pantaloni
ABSTRACT: The binding of reduced nicotinamide adenine
dinucleotide phosphate (NADPH) to nicotinamide adenine
dinucleotide phosphate (NADP) dependent isocitrate dehy-
drogenase from beef liver cytoplasm was studied by several
equilibrium techniques (ultracentrifugation, molecular sieving,
ultrafiltration, fluorescence). Two binding sites (per dimeric
enzyme molecule) were found with slightly different disso-
ciation constants (0.5 and 0.12 pM) and fluorescence yields
(7.7 and 6.3). A ternary complex was formed between enzyme,
isocitrate, and NADPH, in which NADPH dissociation con-
stant was 5 pM, On the contrary, no binding of NADPH to the
enzyme took place in the presence of magnesium isocitrate.
Dialysis experiments showed the existence of 1 NADP binding
site/dimer, with a dissociation constant of 26 pM. When
NADPH was present with the enzyme in the proportion of 1
molecule/dimer, the dissociation constant of NADP was de-
Previous studies (Carlier and Pantaloni, 1973) have shown
that isocitrate dehydrogenase (threo-D,-isocitrate:NADP+
t From the Laboratoire d’Enzymologie du C.N.R.S., 91 190 Gif-sur-
Yvette, France. Receiued May 17, 1976.
Abbreviations used are: NAD, nicotinamide adenine dinucleotide;
NADP, N A D phosphate; NADPH, reduced NADP; EDTA, (ethyl-
enedinitri1o)tetraacetic acid.
BIOCHEMISTRY,
3513-3515.
Colowick, S. P., Kaplan, N. O., Neufeld, E. F., and Ciotti, M.
M. (1952), J. Biol. Chem. 195. 95-105.
Doelle, H. W. (1969), In Bacterial Metabolism, New York,
N.Y., Academic Press, 352-401.
Fewson, A., and Nicholas, D. J. D. (1961), Biochim. Biophys.
Acta 49, 335-349.
Hanson, K. R . (1966), J . Mol. Biol. 22, 405-409.
Haschemeyer, R. H., and de Harven, E. (1974), Annu. Rev.
Biochem. 43, 279-301.
Horton, H. R., Swaisgood, H . E., and Mosbach, K . (1974),
Biochem. Biophys. Res. Commun. 61, 1 1 18- 1 124.
Kemper, D., and Everse, J. (1973), Methods Enzymol. 27,
67-82.
Louie, D. D., Kaplan, N. O., and Mc Lean, J. D. (1972), J.
Mol. Biol. 70, 65 1-664.
M o d , J., Changeux, J. P., and Jacob, F. (1 963), J . Mol. Biol.
6 , 306-329.
Monod, J . , Wyman, J., and Changeux, J. P. (1965), J. Mol.
Biol. 12, 88-1 18.
Ochoa, S. (1955), Methods Enzymol. 1 , 699-704.
Stern, I. J., Wang, C. H., and Gilmour, C. M. (1960), J.
Bacteriol. 79, 601 -6 1 1.
Widmer, F., and Kaplan, N. 0. (1976), Biochemistry 15,
(preceding paper in this issue).
creased fourfold, reaching a value quantitatively comparable
to the Michaelis constant. The kinetics of coenzyme binding
was followed using the stopped-flow technique with fluores-
cence detection. NADPH binding to the enzyme occurred
through one fast reaction (kl = 20 pM-I s-l). Dissociation
of NADPH took place upon NADP binding; however, equi-
librium as well as kinetic data were incompatible with a simple
competition scheme. Dissociation of NADPH from the enzyme
upon magnesium isocitrate binding was preceded by the for-
mation of a transitory ternary complex in which the fluores-
cence of NADPH was only about 30% of that in the enzyme-
NADPH complex. The interaction between the coenzymes and
the involvement of ternary complexes in the catalytic rnecha-
nism are discussed in relation with what is known about the
regulatory role of the coenzyme (Carlier, M. F., and Pantaloni,
D. (1976), Biochemistry 15, 1761-1766).
oxidoreductase (decarboxylation) EC 1.I . 1.42) purified from
beef liver cytoplasm is a dimeric enzyme of molecular weight
48 000 X 2. In the absence of divalent metal cations, steady-
state kinetics exhibit catalytic activation by NADPH, the re-
action product (Carlier and Pantaloni, 1976a). It has been
demonstrated that NADPH did not play a redox role in this
activation and was probably involved in the second step (de-
VOL. 15, NO. 21, 1 9 7 6 4703