The Proton Transfer Reactions of Muscle Pyruvate Kinase*

SUMMARY Rabbit muscle pyruvate kinase detritiates phosphoenolpyruvate 3-tritium under conditions of the net forward reaction prior to the release of pyruvate. This exchange requires that release of pyruvate is not rapid with respect to reversal of the proton transfer steps involved in the catalysis of its enolization. This conclusion is consistent with the low tritium isotope effect that is observed for the hydrogen exchange on pyruvate that is activated by ATP. An intrinsic isotope effect for the proton abstraction step as high as 26 can be expected since such high values were observed in the enzymatic enolization of pyruvate activated by other phos-phate compounds. Considerable exchange of tritium from tritiated water into remaining phosphoenolpyruvate during the course of the forward pyruvate kinase reaction is observed, indicating that the enzyme-bound pyruvate-ATP complex that is generated in the forward reaction can return to substrate form at a significant rate relative to product release. A kinetic analysis indicates that the release of phosphoenolpyruvate and ADP may be a rate-determining factor in this exchange rate. Thus,

This exchange requires that release of pyruvate is not rapid with respect to reversal of the proton transfer steps involved in the catalysis of its enolization.
This conclusion is consistent with the low tritium isotope effect that is observed for the hydrogen exchange on pyruvate that is activated by ATP.
An intrinsic isotope effect for the proton abstraction step as high as 26 can be expected since such high values were observed in the enzymatic enolization of pyruvate activated by other phosphate compounds.
Considerable exchange of tritium from tritiated water into remaining phosphoenolpyruvate during the course of the forward pyruvate kinase reaction is observed, indicating that the enzyme-bound pyruvate-ATP complex that is generated in the forward reaction can return to substrate form at a significant rate relative to product release.
A kinetic analysis indicates that the release of phosphoenolpyruvate and ADP may be a rate-determining factor in this exchange rate. Thus, the conclusions of earlier workers that muscle pyruvate kmase follows rapid equilibrium kinetics require re-examination.
The effects of varying pH, mono-and divalent cations, temperature, and alternative activators on the rates of proton exchange are reported.
The over-all reaction catalyzed by pyruvate kinase consists in the transfer of the phosphoryl group of phosphoenolpyruvate to Al)l' and the transfer of a proton from water to form the methyl group of pyruvate.
This dual functionality is illustrated by the following proposed reaction sequence in which enolpyruvate is considered to be an enzyme-bound intermediate. Previous work has established that these two functions can be made independent of each other. Pyruvate kinase transfers the phosphoryl from ATI' to fluoride (1) or hydroxylamine (a), * This cvork was supported by the Research Board of t,he University of Illinois, by Grants CA-07818, CA-06927, and RR-05539 from the United States Public Health Service, and by an appropriation from the Commonwealth of Pennsylvania. A preliminary report of this work was presented at the American Chemical Society National Meeting, September 14, 1970, Chicago, 111. $ Recipient of a United States Public Health Service Postdoctoral Fellowship. and pyruvate kinase enoliees pyruvate in the presence of compounds other than ATP, such as Pi, without phosphoryl transfer (3).
The present paper concerns the path that hydrogen follows in going from the medium to the -CH3 of pyruvate.
It has been shown that this hydrogen is introduced stereospecifically from the 2-s; face of C-3 (446), but the fact that this proton becomes equivalent to the other two hydrogens of the methyl group b:, rotation at the CL-C& bond, Sequence 2, makes pos&le the monitoring of paths and rates of proton transfer bv determining (2) the subsequent loss of hydrogen derived from PEP' to the medium.
In considering the paths of hydrogen in t,he over-all reaction, one can visualize the possibilities descrihcd in Scheme 3. The Rengenfs-Pyruvate kinase (rabbit muscle) and lactate dehydrogcnase (rabbit muscle) were purchased from Boehringer Mannheim Corp. or Calbiochem.
A reaction mixture contained the following in a volume of 6 ml: 300 pmoles of triethanolamine, pH 7.6; 120 pmoles of MgCl%; 60 pmoles of sodium [3-Tlpyruvate 3 x lo6 cpm per pmole; 90 pmoles of ATP; and 5 units of PEP synthase (7). After 1 hour at 26", the react~ion misture was diluted to 120 ml, placed on a Dowex 1 (Cl-) column (20 X 1.7 cm), which was then washed with water and 0.02 N HCl. The [3-T]PEP was eluted with 0.04 N HCl (7), pooled, neutralized with 2 N NaOlI, and stored frozen. Our best preparations of [3-T]l'EP contained less than 1 76 pyruvate and less than 0.2% volatile counts; greater than 98 % of the counts could be conrertcd by pyruvate kinase to pyruvate and water. Treatment of the [N-T]PEP with acid-washed charcoal removed ADP which eluted from the column with the PEP. To desalt the PEP, it was placed on a G-10 column (2 x 31 cm) equilibrated with 5 mM Tris (pH 7.5) and the PEP Iv-as eluted with the same buffer. Z-Phosphoenol-aketobutyrate (5) was the generous gift of Dr. D. B. Sprinson. All other chemicals were obtained commercially and used without further purification, except for TOH and DzO, which are redistilled before use. Values for pD are those of the measured "pH" + 0.4 (10).
Primary Hydrogen Isotope Effect in Pyruvate Enolization-A comparison was made of the rates of exchange of proton and tritium from [3-Tlpyrurate with deuterium in DtO in a single reaction mixture.
This approach is free of complications from secondary isotope effects on the rate of bond cleavage since tritium is present only in tracer quant.ity and DzO is a common medium.
Proton exchange was followed on the Varian HA-100-15 nuclear magnct,ic resonance spectrometer, as the decrease in relat'ive areas of the pyruvate methyl group signal (2.4 ppm) t'o the methylene peak (3.6 ppm) of Tris. Tritium eschnnge was followed as TOH production by transferring 0.05ml aliquots at various times to 0.95 ml of 0.1 in HCl, removing 0.1 ml of this mixture to a distilling tube containing 0.4 ml of 0.025 M Nas-citrate, and upon distillation under reduced pressure, counting the entire distillate.
Both measurements were made at a minimum of five time points with each reaction mixture and both gave linear first order plots.
Calculation of exchange rates from the three equivalent positions of the pyruvate methyl group was by the equation: ~1, = -%(pyruvate)(time 'rng enzymc)-' ln( 1 -fraction exchanged). Parfition of Tritium from [S-TJPEP between Pyruvafe and Water--In these experiments, [3-T]PEP and an excess of ADP (or other phosphoryl acceptor) were reacted with pyruvate kinase and a lactate dehydrogenase trap for the pyruvate.
The reaction was followed to completion spectrophotometrically by the change in A340 due to DPNH oxidation.
All experiments were conducted at 26" unless otherwise stated.
Tritium released into the solvent was determined by counting the distillate from an O.l-ml aliquot added to 0.4 ml of 0.1 II sodium citrate, pH 6. The citrate was added to minimize the volatility of lactic acid. No back reaction of free pyruvate, which would release tritium into the solvent, would be observed under these conditions because of the lactate dehydrogenase trap and the high K, for pyruvate in the back reaction (3). No effect on the partition was observed upon increasing the concentration of lactate dehydrogenase or upon simultaneous removal of ATP by reaction with glucose and hexokinase.
Partition of Tritium from Water between PEP or Pyruvate -In these experiments, PEP and ADP (at half the PEP concentration) were reacted in TOH with pyruvate kinase and a lactate dehydrogenase trap. When the reaction was complete with respect to ADP, the reaction mixtures were frozen, the TOH removed by sublimation, the residue dissolved in 50 ml of Hz0 and placed on a Dowex 1 (Cl-) column (5 X 1.7 cm) and, after a water wash, the [3-Tllactate was eluted with 0.005 N HCl and the [3-T]PEP with 0.5 M NH&l, according to the method of Bartlett (11). Concentrations of lactate and PEP in appropriate fractions were determined on aliquots by enzymatic procedures (12) and identical aliquots were counted for radioactivity to get specific activities.
Similar column chromatography and analysis of lactate derived from [3-T]PEP as a function of percentage reaction was used to assess any secondary tritium isotope effect in the pyruvate kinase reaction.

RESULTS
In a previous study, enolization of pyruvate was found to require the presence of all the components of the net reaction, K+, Mgt', and ATP, with the notable exception that ATP could be replaced by a number of simple dianions such as Pi and monoester phosphates (3). It 7vas considered of interest to ask whether [3-T]PEP was detritiated by the enzyme under conditions that were either incomplete for carrying out the net reaction or, if complete, the formed pyruvate was trapped with lactate dehydrogenase to prevent its further reaction. This question is of interest for several reasons, The possibility that protonation might be able to precede phosphoryl transfer (Reaction 4) is suggested in analogy t'o the mechanism for Hg'+catalyzed hydrolysis of PEP suggested by Benkovic and Schray (13), in which a proton replaces Hgzf. Previous experiments by Harrison et al. (14), in which the possiblility of a phosphoryl-enzyme intermediate was eliminated using the test of pyruvate to PEP exchange is more demanding of the mechanism than the tritium exchange test since the latter does not require that free pyruvate be formed from PEP in the absence of XDP.
Conditions necessary for the detritiation of [3-T]PEP are shown in Table I. B trapping system for pyruvate (lactate dehydrogenase and DPNH) was included to prevent net reversal of pyruvate formed in t'he complete system. It is observed that, under specific conditions of pyruvate formation a significant portion of the tritium of the substrate was exchanged for protons of the medium.
With an amount of enzyme capable of completing the net reaction in only one-tenth the incubation time used, lit,tle or no detritiation occurred if any one of the components required for net reaction was omitted.
In addition, replacement of K+ by tetramethylammonium, and replacement of ADP by AJ#lP, AMP + Pi, or AMP + NO3 did not lead to hydrogen exchange.
Since the formation of pyruvate from PEP is known to be stereospecific (4), any route for exchange requires protonation, randomization of hydrogen by rotation around the G-G bond and proton abstraction.
Hence, one can conclude that one or more of these steps depends on phosphoryl transfer to ADP.
As evidence that loss of tritium from PEP occurs only during the course of the net reaction, Fig. 1 shows a correlation of detritiation with the formation of pyruvate. Reaction Scheme 3 provides an explanation for the detritiation of PEP during the course of the net reaction.
The exchange requires only that Step 4 is not rapid compared with one or more of the paths connecting Ea with medium protons.
The loss of trit'ium from [3-T]PEP may greatly underestimate the amount of proton exchange that is occurring if Step 3 of Scheme 3 is rate-determining for the exchange process. To evaluate the contribution that an isotope effect would make, the kinetic isotope effect for hydrogen exchange from pyruvate was measured.
In order to do this, a mixture of [3JH]pyruvate and [S-Tlpyruvate (trace) was subjected to pyruvate kinase under exchange conditions in a medium of D20. The rates of loss of the proton signal of pyruvate and of the appearance of tritium in water were compared with either ATP or an ATP analog to activate the exchange.
Three analogs were chosen for their isoelectronic character, but wide range of basicity.
Thus, the pK, values of Pi are 2.1, 6.7, and 11.8 (15), those of methylphosphonate, 2.4 and 7.7 (16), and those of fluorophosphate, 0.5 and 4.8 (17). Table II documents the isotope effects observed in the enolization and the rates of tritium exchange.
The isotopic effect seen with ATP as the activator is quite small, vZR:2rZT = 2.5 to 4, compared with 15 to 25 observed with the analogs. Apparently the C-H bond-breaking step determines the rate of hydrogen exchange with the analog, but may not do so with ATP.
That the same step might not be rate determining with ATP and the analogs was suggested by earlier data (3) in which the pH dependence of the enolization process with Pi and the other analogs followed the pK, of the activator, whereas with 9TP the rate increased sharply in the region above pH 7.5 where ATP s fully ionized and product dissociation (k.J. The rate constant for proton exchange is made up of rate constants for all the exchange paths. This equation and the full expression for kZH can readily be derived usiug the approach of Yagil and Hoberman (18). In the corresponding expression for vZT, k-aT would replace k--3= in the derivation of L,,.
In Scheme 3, Step 3 is itself complex in that it includes rotation at G-C& of bound pyruvate as well as proton abstraction, which may or may not involve a basic group on the enzyme. Likewise, hi+! may be a complex constant in which stepwise dissociations of ,1TP and pyruvate are included. These extensions alter neither the form of Equation 1, nor the central importance of Eseq iu the pnititioning between proton exchange and product dissociation.
A small isotope effect in pyruvate enolization may result from either a slow product dissociation relative to proton exchange (i.e. Ich < IcZH) or from a rate-limiting dissociation of protou from the base on the enzyme (19). Attempting to find conditions where, with iZTP, the C-H bond is broken in the relatively slow step, the isotope effect was determined with either Co2+ or Mn2+, replacing Mg2f. Isotope effects (&X:&T) of 1.8 and 2.2 were observed for CoZf and >I& at pD 7.7, respectively, where rates of tritium exchange were 17 and 90.3 pmoles per miu per mg of protein.
The C--H bond-breaking step does not nppcar rate limiting under these condit'ions.
The alternative route for generating the ternary products complex (&) from tritiated PEP allows one to measure directly the ratio kZT:kd from a determinat,ion of the ratio, called Rr, of tritium in the water to that in the pgruvate in the over-all for- ward directions, as in Table I. Determinations of RT \vere made at several pH values and Tvith Mgzf, MnZS-, or Co2+, as the required divalent cation.
The forward velocities ( ITfor) and the extrapolated maximum rate of tritium exchange from pyruvate with ATP as activator mere determined under comparable conditions.
The necessary monovalent cation was K", present at its optimal concentration. Table III lists values for RT, Vror, and VzT, interpolated to a common PI-1 of 8. The interpolation was necessary because each of t'hese values was found to be sensitive to pII and the variations were dissimilar for each divalent ion (see Figs. 2 and 4).
It will be noted that RT, the partition of tritium of PEP between water and pyruvate, varies considerably with divalent metal ion. At this pH (and at all others examined, see Fig. 3): RT with Co2+ is greater than with iVIn2f, which is greater than with Mg2f.
High values of proton exchange prior to product release in the forward reaction should lead to a low isotope effect in ATP-dependent enolization of pyrurate according to Equation 6. Accordingly, the high RT values observed with Cozf and I%?+ are consistent with the small isotope effects observed in the corresponding enolization reactions.
The low & value with &Ig2' is coupled, however, with an isotope effect in the ATP-dependent enolization that is small relative to the tritium isotope effect of 20 to 25 in the enolization activated by compounds other than ATP.
These observations suggest that another step, such as proton release from t'he enzyme, may be rate limiting in the ATP-dependent enolization and may be responsible for the low RT.
A further deduction can be made from the data in Table III A further test that product dissociation is rate determining, not only in pyruvate enolization (one of the explanations for the low isotope effect with ATP), but also in the net forward reaction (as indicated by the similarity in Ezsr and &,J, consists in measuring the total reversal of Esss to free PEP.
This measurement of the appearance of tritium from water into free PEP during the forward reaction was made in the presence of lactate dehydrogenase and DPNH to make free pyruvate unavailable for reversal.
The forward reaction was followed in the presence of tritiated water until 5Oyh of the PEP was consumed; when the reaction was terminated, PEP and lactate were isolated, and their specific activities determined.
The partition of tritium from water between PEP and pyruvate is designated PT and is a relative measure of Esss reversing to free PEP during the for-  In parallel incubations, RT, the partition ratio of tritium from PEP between water and pyruvate, was measured and is a relative measure of Ezss reversing to EQss and whatever other steps are involved in tritium dissociation from the enzyme. The specific activities of PEP and lactate, the derived value of PT, and the independently measured value of RT under several conditions are included in Table IV.
It is evident that these values are stongly influenced by pH and divalent cation.
The discrimination against tritium incorporation into pyruvate has previously been studied at pH 7 by Simon et al. (20) with a discrimination factor of about 6, as found here also at pH 7.3 with Mg2f.
However, this is clearly dependent on the choice of conditions since rate limitation in product release is clearly shown with Coz+ at pII 8.8 by both the high value of RT and the specific activity of the lactate which exceeds that of the water.
It is also clear that tritium introduced in the ketonization step and retained in the enolization has found its way back to free PEP in rather significant amounts. Thus, with Mgz+ at pH 8.8 the reisolated PEP has 26% the specific activity of the lactate.
If complete isotopic equilibration between PEP and water had occurred prior to release of products, this value could have been 66.7c/0 barring an enrichment in tritium.
The question of which steps might be rate-limiting for the return of intermediates to PEP will be considered under "Discussion. The isotope effect measured from the discrimination against tritium from the medium in the formation of product is about 6.5 in this experiment, comparing t,he specific activity of the lactate and water hydrogen.
A sensitive test of the second possibility was made possible by the recent reports of Bondinell and Sprinson (5) and Stubbe and Kenyon (6) showing that Z-phosphoenol a-ketobutyrate is an alternate substrate for muscle pyruvate kinase. The second authors (6) reported that phosphoenolbutyrate had about the same K, as PEP, 2.5 X 10m6 M, but reacted at only about 0.1% the maximum rate. The following experiment was set up to test for the occurrence of intermolecular tritium transfer.
These were volatile at alkaline pH and hence represent the exchange with water.
Elution with 2 m&r HCl brought off 7.2 x lo6 cpm which would represent the acetate and propionate, and elution with 40 InM HCl brought off 255,000 cpm in the region where PEP was expected.
Since a comparable sample of the incubation mixture taken at zero time failed to show radioactivity in the PEP region, this latter value represents the tritiated PEP formed in exchange for phosphoenolbutyrate. This is a minimal value since, in spite of the low concentration of formed PEP relative to phosphoenolbutyrate, the PEP will tend to be reutilized preferentially in view of its lOOO-fold higher V,,, . A portion of the acetate-propionate mixture was resolved by partition chromatography on silicic acid (21). The acetate had a specific activity of 9.1 x IO5 cpm per pmole. The propionate, corresponding to 2.4 pmoles for the whole incubation, had a specific activity of 212 cpm per pmole. These counts exceed, by far, the number that could be introduced by way of the tritiated water which would have an average specific activity of I..6 cpm per patom of hydrogen.
The zero time sample was treated in the same way with 10 pmoles of propionate added as carrier. Approximately 550 cpm were recovered in the propionic acid peak of the silicic acid column.
This corresponds closely with the experimental result and indicates that no intermolecular tritium transfer occurred. Although Step 4 may be rate-determining in the forward reaction, the detritiation of PEP is far from complete under most conditions.
From the fact that, only a small tritium isotope effect can be observed in the hTP-dependent enolization of pyruvate, this suggests that the tritium exchange cannot simply occur by Step 3' since this would necessarily impose an isotope effect, but may in fact be limited by dissociation of the pyruvate-derived proton from the enzyme, Steps hl or hz. In an attempt to understand the factors that control proton release from the enzyme, a number of studies were undertaken of the sensitivity of RT to incubation condit,ions, and these will be described next.
Using increasing concentrations of urea, no increase was seen in RT (Table V). Cottam et al. (22) have previously shown that 2 M urea dissociates pyruvate kinase from tetramers to dimers, with 68% the activity and that the monomers formed in 4 M urea are devoid of activity.
The loosening of structure accompanying urea denaturation does not increase proton exchange preferentially.
On the contrary, the ratio, RT, decreases.
Furthermore, glycerol, which might favor the interaction of polar groups on the enzyme, raised the relative extent of enolization of bound pyruvate substantially.
These two results suggest that it is release of products rather than of proton that is primarily affected by urea and glycerol.
If product dissociation is facilitated in urea and attenuated in glycerol, the values of R, would show the observed changes. It follows from this that if a major pathway of hydrogen exchange involves Step hz, then the divalent cation must play a role susequent to the phosphoryl transler step, where these ions have been previously reported to act (25, 26). On the other hand, if t)he met)al ion has its effect at the phosphoryl transfer step only, then phosphoryl transfer from XTP to enolpyruvate, Step -2, must be a nrcessarg preliminary to the exchange; that is, Step hl and not hZ is on the major route for the exchange.
This order corresponds neither to the dissociation rate of the ATI'-met,al complex (28), nor to the affinity of the metal-enzyrnc: comples (25), nor to the metal octahedral radiu< (29) Williams series (30). Rather, as shown in Fig. 3, RT is a linear function of the electronegativity of the metal ion, as measured by the pK, of water in its coordination sphere (31). RT is found lo equal 1O-o.24 pK,.
Such an effect could result from the metal affecting the e1ectronegntivit.y of the base which, on the enzyme, is involved in the proton transfer step. The dissociated form of the base is found to facilitate proton exchange, and t,hc divalent cations appear to affect t,he concentration of 1.1~ dissociated form.
'l'he divalent metal and pH effects on the ATP-activated enolization of pyruvate by pyruvate kinase are illustrated in P'ig. 4. With YIgG, the enolization rate illcreases with increase in $1, previously shown by Rose (3), and similarly to the effect of pa on i&L The Co2f-and h4n2f-dependent rates eshibit pll optima at 7.3 and 8.1, respectively, in contrast 1.0 the obselvations with R1, (Fig. 2). These differences emphasize the fact that &, being a ratio ol rate constants, is independent of factors which determine the athsolute rate of &T. The tlifferential effect of the divalent cations on the pI-I profile of v,~ ma? result from the effects on the "affinities" for ,4TP and pyruvate or from formation of inhibitory compleses.
While the Ii, for AY'l' at pH 7.6 has been found to be 1.33, 0.38, and 8 IIIM for ;\rg2~f-, hln2f, and Co*, respectively, the effect of pH on this I\',, at~tl on other kinetic constants awaits further study. l'yr\lvate kinase requires a monovalent cation for activity. 'I'abl~ VI 1 shows the effect of various monovalent cations on the partition ratio, RT. It is of interest, that those cations that :V-tris(hydroxymethyljmcthy1 aminoethane sulfonate (6.6 to X.2). and sodium 12:.tris(hydroxymethyl)methyl aminopropane sulfonate (8.2 to 9). The reaction is susceptible to buffer effects with significant variations observed even with these buffers outside of t,he pH ranges indicated. lleaction time was 30 min and distill:~t~lo collnts in an aliquot were determined as described lrnder "l,&pc~~ime~ltal Procedure" and were corrected for nonellzymic erlolix:ltion at each point. (32) also give the highest amount of labilization of l,ritium of [3-T]PEP during the course of the net reaction. If, as was suggested aboT-e, the dissociation of product is rate limiting for the net reaction, it ma\-follow that the same property of the monovalent cation that illcreases the rate of this step must increase the hydrogen c~change process to an even greater extent,. Since the value of RT is dependent on the ratio of rate constants, it would not bc expected that RT would vary with the concentration of a coml)ollcnt that is required for the reaction. However, as seen in Table VIII, the concentrat.ion of K'. in the range well above its act ivatiou constant for the reaction (11 rnhl) (32) had a marked effect t,hat could not be attributed to changes in ionic strength.
These results show that the K+ at high concentration is involved in steps subsequent to the phosphoryl transl'er step and that /i-mzT ant1 lid are differentially affected.
This might be explained by a low affinity binding of Kf to ,some kind of effector site; however, an alternative explanation will be considered under "Discussion." The effect of various ~~hosphoryl acceptors on RT is shown in Table IX A major finding of the present study is that release of product is rate determining in the forward reaction path. This C:OIIC~Usion is derived from two sources. First, the concentration of the ternary products complex in the forward steady state at, substrate saturation, Eass, agrees well with the concelrtration of Eaey established from the side of pyruvate plus ATP at sat.ur:ttion, Table III. Similar experiments, not reported, hxvc I)cerr performed at pH 7.3 and 8.6 with similar results. The second evidence is more restricted and comes from the extent of dctritiation of [3-T]PEP at pH 8.8 with Co2+, Table IV. Hrrc, even ignoring a discrimination against tritium iu the reversal of Step 3, a factor of l&fold or more, the reversal of that step IS mrasured by RT is very rapid. Since the complete system is required for the detritiation of [3-Tlpyruvate (3), the irreversible dissociation of either product of the forward reaction would terminate the release into water of tritium derived from [3-T]I'EP.
It is possible that the ratedetermining step is isomcrization of the ternary produc+ WIYIplex, &, rather than the immediate release of products. effect of high levels of monovalent cation on the forward reaction rate. Further studies will be required to determine whether this explanation pertains to both phenomena. Likewise such an investigation will be required to determine to what extent the large differences in the extent of detritiation of PEP seen in Table VII with different monovalent cations is to be attributed to such a mechanism.
Although the liberation of products is rate-limiting for the maximum forward velocity, as shown in Table IV, this does not bring with it extensive isotopic equilibration of the substrates with the complex of enzyme and bound product's. Thus the back labeling of PEP from tritiated water, PT, is no greater than the net forward rate, and appreciably less under certain conditions. This fact can be brought into harmony with rate limitation at Step 4 in the forward reaction only if it is assumed that the release of PEP, or of ADP if that is required to precede it, is rate-limiting for the back labeling of PEP. This assumption is not inconsistent with a slow Step 4 since Step 1 is not involved in the kinetic expressions for the maximum forward velocity.
Initial rate studies by Reynard et al. (33) have been interpreted in support of rapid and random interactions of PEP and ,4DP with pyruvate kinase. This is based on the observation of competitive inhibition by ATP with either substrate and the occurrence of linear double reciprocal plots for both substrates wit,h I<, values of each found to be independent of t'he concentration of the invariant subst.rate. The argument for random interaction is supported by the observation of direct binding studies with both PEP and ADP by ultracentrifugation studies (31) and by kinetic protection and proton relaxation rate studies (26). Furthermore, in the latter work, good correlations between the calculated binary dissociation constants and the kinet,ic K, values were observed with Mn2+ as the activator.
Although these observations can be considered definitive in support of a random order of substrate interaction, the rest'riction of "equilibrium kinetics" may not be considered to be established. Thus, nonlinear double reciprocal plots were observed by Melchior (34) when (MgADP) was the varied substrate. In addition, the occurrence of linear replots for a random bisubstrate reaction is now considered to be of limited value in deciding whether a quasiequilibrium state pertains (34, 35). The conclusion drawn concerning the rate-limiting role of product release for the maximal forward velocity is unrelated to the particular sequence of phosphoryl and proton transfer that is chosen for the mechanism.
The arguments in support of the stepwise mechanism, Reaction 1 are based on the independent occurrence of the two processes and may not be universally accepted.
However, for the purposes of placing the data in a context for discussion, this mechanism, expanded in Scheme 3 is assunled to be correct.
Considering the steps between 1 and 4, it is observed from Tables IV to IX that the degree to which E3 undergoes enolization prior to dissociation of products, as indicated by the value of I&, varies greatly with the conditions of the incubation.
It is evident that exchange of the ---CH3 proton with medium occurs prior to release of PEP or pyrm-ate since an I<, of 2.24 (Table IV) is much greater than the observed return of PEP, PT = 0.149, Jvould allolr-and also much greater t,han could be explained if the methyl tritium became equilibrated with a protonic base, such as lysine, and in which exchange with t,he medium followed dissociation of products. Under these particular conditions, pH 8.8 1vit.h Co2+, the steps that are involved in enolization and exchange with the medium clearly do not contribute to rate determination of the forward velocity. On the ot'her hand, there must be a small contribution at pH 7.3 with Mg*f since, as noted in Table IV, the value of RT is quite small and a significant, although not maximal, discrimination against the utilization of tritium from the medium is evident in t,he formation of pyruvnte.
It is only indirectly established, from the experiments reported, that the enzyme functions as a base in the enolization process, although this is to be expected.
The failure to observe intermolecular tritium transfer from [3-Tjpyruvate to phosphoenolbutyrate tends to rule out the presence of EH+ per se in the ho step, Scheme 3. In this experiment the phosphoenolbutyrate was present> at 300-fold its K, for the formation of a-ketobutyrate and ,4TP and therefore should have provided an excellent trap for any tritium derived from pyruvate and retained by the enzyme. On the other hand, the very low ralue of the isotope effect of ATP-dependent enolization of pyruvate, together with the low value of R, under certain conditions, strongly suggests that the enolization cannot only involute direct loss of hydrogen to the medium, Step 3', since this \Tould always necessarily show a full isotope effect under such conditions. Among the other steps involved in the exchange loss of tritium from [3-T]PEP, t'he rotation around C&-C, of pyrurate is probably much faster than competing steps since such unhindered rotations are known to have rate constants of the order of 1Ol2 se0 (36), and it was determined that the RT was the same for both E-and Z-(3-T]-

PEP.
Concerning the nature of a base catalyst on the enzyme, it is not possible to be definitive at this time.
Hydroxide in the coordination sphere of the divalent metal could be acting as a base. This would provide a ready explanation for the effect of electronegatirity of metal ion on the partition ratio, Rr (Fig. 3).
Alternatively, an amino acid residue coordinated to the metal could serve as a base and show a response to the electronegativity of the metal.
The low slope of the linear free energy relationship of Fig. 3 suggests an indirect effect.
Of possible importance for future studies is that the enolization of pyruvate that is activated by analogs of ATP such as Pi occurs with a high isotope effect, indicating that the C-H bond cleavage is rate-determining.
Thus kinetic studies of the exchange may be expected to reflect the character of the basic group as well as other factors involved in this step. In this respect it is of interest that the pH dependence of enolization of pyruvate with various dianionic activators corresponded to the pK, of the activator (3).
As was noted in Fig. 1, the loss of tritium from PEP to the medium closely followed the course of the over-all reaction.
In a parallel study to this one, also at pH 7.3, the lactates formed early and at completion of the reaction were compared with respect to their specific activities and found to be within 3?4, of each other. Thus no secondary isotope effect could be demonstrated.
Rather large inverse isotope effects would be expected if cit'hcr Step 3 or 4 were rate determining (compare with studies of enolase in which were obtained values of kT:& = 1.26 as a kinetic effect and kT:lcH = 1.4 as an equilibrium effect (37)). However, to show a discrimination between tritiated and normal species of PEP in the course of many cycles of reaction it is required that intermediates prior to the rate-determining step be able to mix with the free PEP pool in order t,o restore the specific activity of the intermediates to normal. Likewise, discrimination due to a preliminary equilibrium depends on