ATP-dependent phosphorylation of α-substituted carboxylic acids catalyzed by pyruvate kinase

ARCHIVES OF BIOCHEMISTRY Vol. 228, No. 1, January,

AND BIOPHYSICS pp. 31-40, 1984

ATP-Dependent Phosphorylation of a-Substituted Acids Catalyzed by Pyruvate Kinase’ DAVID

E. ASH,2 PAULA

J. GOODHART,

AND

Carboxylic

GEORGE

H. REED3

Department of Biochemistry and Biophysics, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104 Received

April

12, 1983, and in revised

form

August

22, 1983

Pyruvate kinase from rabbit muscle catalyzes an ATP-dependent phosphorylation of glycolate to yield 2-phosphoglycolate (F. J. Kayne (1974) B&hem. Biophgs. Res. Cmmun. 59, 8-13). An investigation of anologous reactions with other a-substituted carboxylic acids reveals several new substrates for such a phosphorylation reaction. Thus the cy-hydroxy carboxylic acids L-lactate, D-lactate, DL-a-hydroxybutyrate, DL-CYhydroxyvalerate, L-glycerate, D-glycerate, DL-nitrolactate, and DL-8-chlorolactate are phosphorylated on the cu-hydroxy group to give the corresponding phosphoesters. Thioglycolate is also a slow substrate for phosphorylation of the thiol group to give the phosphothioglycolate, and DL-thiolactate is phosphorylated in a very slow reaction to give phosphothiolactate. P-Hydroxypyruvate is a substrate; but, unlike the reaction with pyruvate, with ,&hydroxypyruvate the equilibrium for the reaction lies in favor of ADP and the phosphorylated product which appears from 31P NMR data to be tartronate-semialdehyde-2-phosphate. 31PNMR spectroscopy has been used to verify the identity of the products for all of the reactions. Steady-state kinetic constants have been obtained for some of the more rapid reactions. The reactions with glycolate, Lglycerate, and P-hydroxypyruvate have k eatvalues that are close to that for phosphorylation of pyruvate in the reverse of the physiological reaction.

Pyruvate kinase (EC 2.7.1.40) has a relatively high specificity for P-enolpyruvate* when the reaction is viewed in the normal

physiological direction. Although several phosphorylated a-keto acids show substrate activities, the maximal velocities observed with these analogues are less than 1% of that for P-enolpyruvate (l-5). In the other direction, with ATP as a phosphoryl donor, the enzyme catalyzes a bicarbonatedependent phosphorylation of fluoride ion (6), and of hydroxylamine (7-9) as well as phosphorylation of glycolate (10). Moreover, the velocities of these ATP-dependent phosphorylations are comparable (within a factor of four or five) to that reported for phosphorylation of pyruvate in the reverse of the normal reaction (11). Other side reactions or partial reactions catalyzed by the enzyme include the enolization of pyruvate (12, 13), the decarboxylation of oxalacetate (14-16), and a bicarbonate-dependent ATPase (6).

1 This work was supported by a grant from the National Institutes of Health, AM1751’7. A preliminary account of this work has been presented (D. E. Ash and G. H. Reed (1981) Abstracts, Division of Biological Chemistry Meeting, American Chemical Society, Minneapolis, Minnesota). ’ Present address: Department of Biochemistry, Temple University School of Medicine, Philadelphia, Pennsylvania 19140. ’ To whom correspondence should be addressed. ’ Abbreviations used: P-enolpyruvate, phosphoenolpyruvate; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; nitrolactate, 3-nitro-2-hydroxypropionate; ATP-/S, adenosine-5’-(3-thiotriphosphate); DL-a-hydroxyvalerate, in,-2-hydroxypentanoate; DL-a-hydroxycaproate, DL-2-hydroxyhexanoate; thioglycolate, 2-thioacetate; DL-thioiactate, DL-2-thiopropionate. 31

0003-9861/84 Copyright All rights

$3.00

0 1984 by Academic Press. Inc. of reproduction in any form reserved.

32

ASH,

GOODHART.

Although the specificities of all of these reactions for inorganic cations as activators vary, all require divalent cations; and with the possible exception of the oxalacetate decarboxylase reaction, most require monovalent cations as well (17). Thus, if bicarbonate is viewed as a surrogate for the carboxylate group of the natural substrates (9, 18), cofactor requirements for all of the side activities parallel those of the normal reaction. The substrate activity of the two-carbon, cr-hydroxy carboxylic acid glycolate (lo), and the potent inhibition of the enzyme by D- and L-phospholactate (19,20), raised the possibility that the lactates, other a-hydroxy carboxylic acids, and perhaps other compounds with a nucleophilic substituent (Yto a carboxylate group might also serve as substrates for the reverse reaction of pyruvate kinase. The present paper reports the results of activity studies with a number of such compounds in an ATP-dependent phosphorylation reaction that is catalyzed by pyruvate kinase. The pattern of activities found with this class of compounds provides new insight into the steric latitude that is available at the active site of the enzyme and permits some preliminary structure-activity correlations. MATERIALS

AND

METHODS

Pyruvate kinase was isolated from rabbit skeletal muscle by the method of Tietz and Ochoa (6) with the substitution of an ammonium sulfate fractionation at 55% saturation for the final crystallization procedure. Stock solutions of the enzyme were stored at 4°C in 20 mM imidazole/HCl, pH 7.0, 1 mM EDTA. The enzyme preparations had specific activities of between 200 and 250 IU when measured at 22’C in the coupled assay with lactate dehydrogenase (21). At the high concentrations of enzyme that are required for some of the slower reactions, trace contaminating activity from adenylate kinase became apparent upon prolonged incubation of reaction mixtures. This contaminating activity was minimized by an additional chromatography of the enzyme on a column (2.5 X 90 cm) of Sephacryl S-200 (Pharmacia Fine Chemicals) with an equilibration and running buffer of 0.5 M Hepes/tetramethylammonium hydroxide, 0.1 M tetramethylammonium chloride, pH 7.5 (Buffer A). Nitrolactate was provided by Dr. D. J. T. Porter, University of Pennsylvania. All other reagents were obtained from Sigma Chemical Company. The purities

AND

REED

of thioglycolic acid and of DL-thiolactic acid were assayed by proton NMR spectroscopy and comparison with published spectra (Aldrich Library of NMR Spectra). Solutions of the calcium salts of D- and Lglyceric acid were passed through a column of Chelex100 (Bio-Rad Laboratories) in the sodium form to remove inhibitory calcium ion, and the absence of Ca(I1) was verified by atomic absorption spectroscopy. Chromatographic assays were used for preliminary screening of potential substrates. Reaction mixtures contained 2 mM ATP, 5 mM MgClz or MnCls in 0.075 M KCl, 0.05 M Hepes/KOH, pH 7.5 to 9.0 (Buffer B). The concentration of potential substrate was 25 mM, and the reactions were initiated by the addition of 20 to 100 pg of pyruvate kinase to 100 ~1 of reaction mixture. At various time points, aliquots of the mixtures and of control samples that lacked one essential component were spotted onto polyethyleneimine thinlayer chromatography plates (Brinkman Inc.). The plates were developed in 1.2 M LiCl or in 0.75 M potassium phosphate, pH 3.5, and the nucleotides spots were visualized under ultraviolet light. Some of the reactions were assayed for nucleotides by high-performance liquid chromatography with a Whatman Partisil-10 SAX anion exchange column eluted with 0.5 M ammonium phosphate at pH 4.3. The products of the reactions that showed formation of ADP above background levels were further characterized by *‘P NMR. For the NMR experiments the concentrations of the components were increased to the following levels: 20 mM ATP, 30 mM MgCls, 1 to 30 mg/ml enzyme, and 50 mM substrate. Prior to acquisition of spectra, the reactions were terminated by the addition of EDTA to a final concentration of 0.1 M. Kinetic constants for some of the more rapid phosphorylation reactions were determined using a continuous pH stat assay (10). Solutions for the pH stat assays contained 0.1 M KCI, 2 mM ATP, 5 mM MnClz or MgClz at pH 8.0. The concentration of pyruvate kinase varied from 0.02 to 0.5 mg/ml, and the reactions were initiated by addition of substrate. ‘H NMR (60 MHz) and *lP NMR (24.3 MHz) spectra were obtained with a Varian NV-14 spectrometer modified for Fourier transform spectroscopy. Samples were contained in 8-mm-o.d. tubes. A coaxial insert containing DzO was used for the heteronuclear lock. Some *iP NMR spectra were obtained at 145.8 MHz with a Bruker WH 360 spectrometer. ‘iP chemical shifts are reported relative to external 35% phosphoric acid. Positive shifts correspond to signals that are at lower fields than the reference signal. RESULTS

AND

DISCUSSION

Preliminary Screening of Potential Substrates Several a-substituted carboxylic acids were tested for substrate activity with py-

NEW

SUBSTRATES

FOR

ruvate kinase using the chromatographic assays for nucleotides. The results of these experiments are summarized in Table I. Product identification by 31P NMR (see Table I) verifies that, in each instance where activity is indicated, production of ADP is coupled directly to the phosphorylation of substrate and does not result from an ATPase activity. Several qualitative trends in reactivity were apparent from the chromatographic assays (data not shown). With the excep-

TABLE SUBSTRATE

ACTIVITY

PYRUVATE

tion of the activity present with dihydroxyfumarate, the reactions are preferentially activated by Mn(I1) versus Mg(I1). A preference for activation by Mn(I1) over Mg(I1) was also observed for the glycolate reaction (10). The velocities of the reactions with D-la&& and L-lactate increase over the pH range from 7.5 to 8.2, and the L isomers of lactate and of glycerate are phosphorylated at higher apparent rates than the corresponding D isomers. The reactions with the lactates are stimulated by

I

OF ~-SUBSTITUTED CARBOX~LJC ACIDS WITH PYRUVATE 31P NMR PARAMETERS OF PRODUCTS ‘iP NMR

Compound L-Lactate D-h&ate L-Glycerate D-Glycerate DL-&Chlorolactate P-Hydroxypyruvate DL-a-Hydroxybutyrate a-Hydroxyisobutyrate Thioglycolate DL-Thiolactate DL-a-Hydroxyvalerate DL-cu-Hydroxycaproate Tartronate D-Malate L-Malate DL-Nitrolactated Acetopyruvate 6-Fluoropyruvatee Phenylpyruvate Dihydroxyfumaratef

Activitya yes yes yes yes yes yes 3-s no yes yes yes no II0 no no yes no no no yes

33

KINASE

Chemical (wm)* 3.1 3.1 3.6 3.6 3.7 4.2 3.1 -

parameters species

KINASE

AND

of product

shift JH-P

U-W”

8.9 8.9 8.8 8.8 9.4 8.6 8.8

doublet doublet doublet doublet doublet doublet doublet -

16.4 15.8 3.6 -

6.7 triplet 10 doublet 8.6 doublet -

-

-

3.0 -

9

doublet -

“The sensitivity of the assays was sufficient to detect 50 PM ADP after 2 h incubation with 1 mg/ml pyruvate kinase. Background levels of ADP in solutions lacking divalent metal ion or substrate were less than 10 PM. *Chemical shifts are from external 85% HePOd. Precision in chemical shift measurements is approximately 0.1 ppm. ’ Precision in the coupling constants is approximately 0.5 Hz. The multiplet structure of the signal due to phosphorus-proton coupling is indicated. d Values taken from Ref. (23). e This compound is a substrate for pyruvate kinase-catalyzed enolization (22), and z-phosphoenol-3&oropyruvate is a substrate for the forward reaction (2, 3). ‘The actual substrate may be fl-hydroxypyruvate-the product of decarboxylation of dihydroxyfumarate.

34

ASH,

GOODHART,

addition of KC1 to solutions of the Li+ salts in Buffer A. Compounds with a carboxylate group at the /3 or y position (tartronate and malate) are inactive. Of the cY-keto acids that were tested only P-hydroxypyruvate gave measurable activity in the assay. The unfavorable equilibrium in the reverse reaction for the other cY-keto acids no doubt results in insufficient production of ADP for detection above the background. ATP \

Dihydroxyfumarate stimulates a slow conversion of ATP to ADP in a reaction that is preferentially activated by Mg(I1). It is unclear at this time whether or not the true substrate in this reaction is dihydroxyfumarate or P-hydroxypyruvate that would be produced in the decarboxylation of dihydroxyfumarate. Preliminary experiments indicate that the conversion of dihydroxyfumarate to P-hydroxypyruvate is stimulated by the presence of divalent cations and pyruvate kinase. Thus, pyruvate kinase may catalyze the decarboxylation of dihydroxyfumarate in a reaction analogous to the decarboxylation of oxalacetate (14, 15). Product Identification bg “P NMR In order to establish that the substrate activity attributed to the compounds in Table I was not the result of an induced ATPase activity, the products of each of the positive reactions were characterized by 31P NMR. Representative NMR spectra for several of the reaction mixtures are shown in Figs. l-3, and NMR data for the products are summarized in Table I. The spectra in Fig. 1 show that the reaction with L-lactate completely exhausts the ATP whereas under similar conditions there is considerable ATP left in the reaction mixture with D-lactate. However, there is a preferential inhibition of the en-

AND

REED

The reaction with nitrolactate has been reported elsewhere (23). The activities with &hydroxypyruvate and with dihydroxyfumarate require additional discussion. In contrast to the unfavorable equilibrium for the phosphorylation of pyruvate, /3-hydroxypyruvate is quantitatively converted to a phosphorylated product. On the basis of evidence provided by 31P NMR (see below) it appears that the reaction is driven by isomerizations as shown below: AOP f

zyme by D-phospholactate (20), and product inhibition would be more pronounced in the reaction with D-lactate. Proton-decoupled 31P NMR spectra for reaction mixtures with all of the a-hydroxy carboxylates that showed activity exhibited a single non-nucleotide product signal at +3.1 to +4.0 ppm at pH 8. Addition of potassium phosphate to a final concentration of 20 mM produces a separate signal for the internal Pi at +2.6 ppm, and this result confirms the absence of an ATPase activity. Furthermore, in proton-coupled spectra the product resonances are multiplets consistent with the patterns expected for proton phosphorus coupling in the product phosphoesters (see Table I). 31P NMR spectra for the reaction mixture containing @-hydroxypyruvate are shown in Fig. 2. The product resonance is a doublet with Jn-r of approximately 8.6 Hz. The magnitude of Jn-r is close to those for phosphorylated cY-hydroxy carboxylic acids, and the doublet splitting indicates that there is a single strongly coupled proton in the product. Thus, the ‘lP NMR data for the product of the reactions with /Shydroxypyruvate are consistent with the structure of the product shown in Reaction [l]. The proposed isomerizations would account for the quantitative conversion of /3hydroxypyruvate to tartronate-semialde-

NEW

SUBSTRATES

FOR

PYRUVATE

35

KINASE

L-lactate

ADP P-lactate

i

0

-5 CHEMICAL

I

+5

0

-IO

-I5

SHIFT

(pm)

, ADPB

-5 CHEMICAL

-20

4

-20

-27

%

ATP

-IO SHIFT

-I5 (ppm)

FIG. 1. *‘P NMR spectra for reaction mixtures with L-lactate and p-lactate. Solutions contained 150 mM Hepes/KOH, 0.1 M KCl, pH 8.2, 20 mM ATP, 30 mM MgC&, 1.8 mg/ml pyruvate kinase, and 50 mM L-lactate or D-lactate. Reactions were terminated after 29 h at room temperature by addition of EDTA to a final concentration of 0.1 M. Spectra were recorded at 24.3 MHz and represent 2000 transients taken with a 1.6-s acquisition time, 2-s pulse delay, and a flip angle of 45”. Resonances due to the phosphate groups of the nucleotides are ADP j3P and ATP yP, -6 ppm; ADP CUP and ATP CUP, -10 ppm, ATP BP, -21 ppm.

hyde-Bphosphate. The stereochemistry of the product is not yet known. 31P NMR spectra for the reaction mixtures with phydroxypyruvate undergo further changes5 sa1P NMR spectra for reaction mixtures of @-hydroxypyruvate recorded after the sample had been at room temperature for 24 h exhibited three new signals. In addition to the signal for the initial product, in proton-decoupled spectra there are signals at 4.4 and 4.0 ppm, each with intensity approximately half that of the 4.2-ppm product signal. There is also a signal at 3.3 ppm with an intensity about 15% that of the 4.2-ppm signal. Addition of phosphate to the sample produces a separate signal at 2.6 ppm.

upon incubation of the sample at room temperature for 24 hrs. The product shown in Reaction [l] must decompose or rearrange further during this time. It is likely that tartronate-semialdehyde-2-phosphate undergoes decarboxylation, but the decomposition products were not investigated further. 31P NMR spectra for reaction mixtures containing glycolate and thioglycolate are compared in Fig. 3. The proton-decoupled spectrum for the reaction mixture with glycolate exhibits a single resonance at +3.6 ppm (see also (24)). In proton-coupled

36

ASH,

GOODHART,

AND

REED

B-hydroxypyruvate

3

o

+5

-5

CHEMICAL

-10

-15

SHIFT

-20-25 ( ppm )

FIG. 2. ‘lP NMR spectra for a reaction mixture with @-hydroxypyruvate. The components of the solution were the same as given in Fig. 1, except 50 mM @hydroxypyruvate substituted for lactate. The spectrum was recorded after incubation of the mixture for 3 h at room temperature.

spectra the expected triplet pattern is observed. The 31P NMR spectrum for the reaction mixture with thioglycolate shows three groups of resonances in addition to those due to ATP and ADP. The resonances at +2.9 ppm and +4.1 ppm are assigned to Pi and AMP, respectively, on the basis of their chemical shifts and splitting patterns in proton-coupled spectra and by addition of Pi and AMP to the samples. The protondecoupled spectrum exhibits a single line at +16.4 ppm that becomes a poorly resolved triplet in the proton-coupled spectrum. This low field resonance is assigned to phosphothioglycolate in which the sulfur atom occupies the bridging position. The downfield shift of this resonance relative to that of phosphoglycolate is consistent with the known effects of sulfur substitution on the 31P chemical shifts of phosphorothioates. Shifts due to sulfur substitution in these compounds appear to correlate with the bond order of the P-S linkage, and the largest shifts are observed when the sulfur occupies a nonbridging position. For example, the 31P NMR resonance for the phosphorothioate analog of

AMP with a nonbridging sulfur atom has a chemical shift of +39.4 ppm relative to AMP (25). However, when the sulfur is in the bridging position, the chemical shift of the 31P signal is at +12.4 ppm relative to AMP (26). 31P NMR spectra for reaction mixtures with glycolate and ATPyS reveal a triplet signal for thiophosphoglycolate (nonbridging sulfur) at +39.6 ppm from the signal for phosphoglycolate. The NMR data confirm the identity of the product of the reaction with thioglycolate as phosphothioglycolate. Production of inorganic phosphate in the reaction mixture could be the result of a slow substrate dependent ATPase or of a slow decomposition of phosphothioglycolate. The appearance of AMP in this sample is probably due to the slight contamination of the enzyme with adenylate kinase-the activity of which becomes apparent because of the prolonged incubation of the sample with high concentrations of protein. 31P NMR spectra for reaction mixtures with DL-thiolactate that had been incubated with pyruvate kinase (30 mg/ml) and ATP for up to 24 h reveal a signal at +15.8

NEW

SUBSTRATES

FOR

PYRUVATE

37

KINASE

glycolofe

‘H

decoupled

1

-5

0

CHEMICAL

-IO

-I5

SHIFT

(km)

-2’5

-20

thioglycolote

product

20

I5

IO

5

0

CHEWCAL

FIG. 3. ‘lP NMR spectra the same as given in Fig. The glycolate mixture was incubated with 30 mg/ml

-5

SHIFT

-10

45

-20

-25

(pp,,,)

for reaction mixtures with glycolate and thioglycolate. Conditions are 1, except glycolate at 50 mrd or thioglycolate at 50 mM replaced lactate. incubated for 29 h at room temperature. The thioglycolate sample was pyruvate kinase at pH 9.0 for 27 h.

ppm that is a doublet in the proton-coupled spectra. There are also signals for Pi, ATP, and ADP. The presence of the low field doublet signal in the spectra confirms the formation of the expected product, phosphothiolactate. However, the reaction with thiolactate is much slower than the reaction with thioglycolate. Kinetic Meaxurmenti Release of a proton upon phosphorylation of glycolate and the other cy-hydroxy

carboxylic acids permits a continuous assay of the reactions with a pH stat (10). The expected potent product inhibition required that low extents of reaction be used in initial velocity measurements. The kinetic constants were obtained from linear regression analysis of double-reciprocal forms of the data (27). The kinetic constants given in Table II confirm the general trends in reactivity that were noted from the results of the chromatographic assays. The L-isomers of

38

ASH, GOODHART, TABLE STEADY-STATE Substrate Glycolate Glycolate” L-Lactate D-Lactate L-Glycerate D-Giycerate DL-@-Chlorolactate @Hydroxypyruvated Pyruvate’ L-Lactatef L-Lactate0 L-Lactate”

KINETIC

AND REED II

CONSTANTS FORPHOSPHORYLATIONOF SUBSTRATES~

Apparent Lt (0 1.79 f 3 0.047 + 0.009 f 0.23 f 0.079 k 0.06 f 1.24 + 2 0.122 + 0.020 k 0.016 +

0.02

Apparent Kn mJ)

Number of points

rb

10

0.97

0.002 0.002 0.02 0.011 0.01 0.14

3.5 f 0.4 2.3 3.8 + 0.3 0.38 f 0.14 0.84 f 0.16 2.06 + 0.41 2.4 + 0.6 4.31 + 0.74

13 13 7 7 9 7

0.98 0.79 0.95 0.96 0.99 0.97

0.008 0.003 0.004

16 + 0.7 8.6 k 2.1 11 + 4.6

11 7 15

0.99 0.89 0.82

‘Except where noted, the assays were carried out with MnCla at pH 8 and 21°C. Other conditions are given under Materials and Methods. The error limits are standard deviations from the regression analysis. * r is the correlation coefficient for the least-squares fit to the data. ‘Values in this line are from Ref. (10) and the data were taken at 30°C. d MgCls was used in assays with this substrate. ‘Value from Ref. (11). ‘Values for Mg(I1) activation at pH 8 and 37’C. 0 Values for Mg(I1) activation at pH 8 and 21“C. h Values for Mg(I1) activation at pH 7.2 and 37°C.

lactate and glycerate give higher k,, values than the respective D-iSOmW3 although Dlactate has a lower apparent K, than Llactate. This same trend in affinity is also observed in the inhibition constants for the phosphorylated forms of the lactates (20). For compounds of the same configuration and chain length, the glycerates give higher k,,{s than the lactates. Dougherty (28) has pointed out that the pH-rate profile for the glycolate reaction is consistent with the idea that the alkoxide is the reactive form of the substrate. The higher kcat)s for the glycerates (relative to the lactates) might be rationalized by an inductive effect of the fi-hydroxy group on the pK, of the (Yhydroxy group. However, the apparent k,,J Km for the racemic mixture of P-chlorolactate is lower than that for L-glycerate, so factors other than inductive effects must also be involved. Among the homologous series of straight-chain cr-hydroxy carboxylates, glycolate has the highest kcat; but com-

pounds up to at least five carbons in length (a-hydroxyvalerate) undergo reaction. (YHydroxycaproate did not give a measureable reaction although it does appear to inhibit other activities. The a-hydroxy dicarboxylates tartronate, D-malate, and L-malate do not undergo reaction whereas the structurally related keto-acid, oxalacetate, undergoes decarboxylation at a site that at least overlaps the site for the kinase activity (15). Muirhead et al. (29) have presented evidence from X-ray diffraction data (for the enzyme from cat muscle) that a glutamate residue is positioned such that its carboxyl group could participate in the protonationdeprotonation of the p carbon of the physiological substrates. EPR data for Mn(I1) complexes with the enzyme show that tartronate and D-malate do not bind at the active site. However, L-malate does bind, and produces a spectral change that is similar to that given by glycolate and the other active compounds (D. T. Lodato and

NEW

SUBSTRATES

FOR

G. H. Reed, unpublished observations). Thus, the failure of tartronate and D-malate to bind is compatible with the suggestion that there is a negatively charged side chain (e.g., glutamate) near the p carbon of the substrate. On the other hand, L-malate must be bound such that the repulsive charge-charge interaction between carboxyl groups is lessened. One might then expect L-malate and oxalacetate to be bound in a similar conformation at the active site. The reason for the inactivity of L-malate is not yet clear although it is possible that the y carboxyl group may interfere with the productive binding of ATP. Phosphorylation of L-lactate may have some physiological ramifications. Tissue levels of L-lactate rise above the apparent Km for the lactate phosphorylation reaction during anaerobic glycolysis (30), and the levels of ATP and of pyruvate kinase in the tissues that produce lactate are sufficient to promote significant product formation. Although the drop in cellular pH that coincides with lactate production would slow the progress of the phosphorylation, some phospholactate must be formed. It is likely that a phosphatase akin to phosphoglycolate phosphatase (31, 32) is active under these circumstances in preventing accumulation of phospholactate. In addition to interest in the substrate activity shown by these compounds, the chemical and stereochemical diversity of the active compounds may be exploited in other types of investigations. Unlike fluoride and hydroxylamine, which have high apparent Km’s for pyruvate kinase, most of the compounds in Table I bind with affinities that are conducive to spectroscopic investigation. For example, the characteristic EPR spectra for the Mn(II)-enzyme complexes with these compounds can provide insight into the structure of the ternary complexes and into the mode of their interaction with one of the essential divalent cations at the active site of the enzyme (33). Finally, phosphorylation catalyzed by pyruvate kinase provides a convenient method for synthesis of the phosphoesters of the parent a-substituted carboxylic acids. Some of these phosphoes-

PYRUVATE

KINASE

39

ter derivatives may be useful as inhibitors of other glycolytic enzymes (34). ACKNOWLEDGMENTS The authors thank Dr. David Porter for many valuable suggestions, and Dr. W. W. Cleland and Dr. T. Dougherty for communicating results prior to publication. REFERENCES 1. BODINELL, W., AND SPRINSON, D. (1970) B&hem. Biophys. Res. Commun. 40,1464-1467. 2. STLJBBE, J., AND KENYON, G. L. (1971) Biochemistry 10, 2669-2677. 3. STUBBE, J., AND KENYON, G. L. (1972) Biochemistry 11,338-345. 4. SOLING, H. D., WALTER, U., SAUER, H., AND KLEINKE, J. (1971) FEBS Lett. 19, 139-143. 5. WOODS, A. E., CHATMAN, V. B., AND CLARK, R. A. (1972) B&hem Biophys. Res. Commun 46, l-4. 6. TIETZ, A., AND OCHOA, S. (1958) Arch B&hem. Biophys. 78, 477-493. 7. KUPIECKI, F. P., AND COON, M. J. (1959) J. Biol. Chem. 234,2428-2432. 8. KUPIECKI, F. P., AND COON, M. J. (1960) J. Biol Chem 235,1944-1947. 9. COTTAM, G. L., KUPIECKI, F. P., AND COON, M. J. (1968) J. Biol Chem 243, 1630-1637. 10. KAYNE, F. J. (1974) B&hem. Biophys. Res. Commm 59, 8-13. 11. MCQUATE, J. T., AND UTTER, M. F. (1959) J. Biol. Chem, 234, 2151-2157. 12. ROSE, I. A. (1960) J. Biol. Chem. 235.1170-1177. 13. ROBINSON, J. L., AND ROSE, I. A. (1972) J. BioL Chem. 247,1096-1105. 14. CREIGHTON, D. J., AND ROSE, I. A. (1976) J. Biol. Chem 251, 61-68. 15. JURSINIC, S. B., AND ROBINSON, J. L. (1978) Biochim Biophys. Acta 523,358-367. 16. BIRUS, M., AND LEUSSING, D. L. (1982) Inorg. Chem 21, 374-380. 17. NOWAK, T., AND SUELTER, C. H. (1981) Mol. Cell. Biochem. 35,65-75. 18. BOYER, P. D. (1962) The Enzymes, 2nd ed., Vol. 6, pp. 95-113, Academic Press, New York. 19. NOWAK, T., AND MILDVAN, A. S. (1970) J. Biol Chem. 245, 6057-6064. 20. NOWAK, T., AND MILDVAN, A. S. (1972) Bioch,emi&?-g 11, 2819-2828. 21. BUCHER, T., AND PFLEIDERER, G. (1955) in Methods in Enzymology (Colowick, S. P., and Kaplan, N. O., eds.), Vol. 1, pp. 435-440, Academic Press, New York. 22. GOLDSTEIN, J. A., CHEUNG, Y. F., MARLETTA,

40

ASH, GOODHART, M. A., AND WALSH,

C. (1978) Biochemistry

17,

55674575. 23. PORTER, D. J. T., ASH, D. E., AND BRIGHT, H. J. (1983) Arch Biochem Biuphya 222,200-206. 24. NAGESWARA RAO, B. D., KAYNE, F. J., AND COHN, M. (1979) J. Biol. Chem 254, 2669-2696. 25. JAFFE, E. K., AND COHN, M. (1978) Biodwmistly

17,652-657. 26. ROSSOMANDO,

E. F., CORDIS,

G. A., AND MARKHAM,

G. D. (1983) Arch. B&hem Biophgs, 220, 7178. 27. ROBERTS, D. V. (1977) in Enzyme Kinetics, pp. 285-299, Cambridge University Press, London. 28. DOUGHERTY, T. (1982) Ph.D. Thesis, University of Wisconsin.

AND REED 29. MUIRHEAD, H., GRANT, J. P., LAWTON, M. A., MIDWINTER, C. A., NORTON, J. C., AND STUART, D. I. (1981) Biochem Sot Tram. 9.212-213. 30. BYLUND-FELLENIUS, A., WALKER, P. M.,ELANDER, A., HOLM, S., HOLM, J., AND SCHERSTEN, T. (1981) Biochem, J. 200,247-255.

31. ROSE, Z. B. (1981) Arch Biochem Biophys. 208, 602409. 32. BEUTLER, E., AND WEST, C. (1980) Anal. Biochem 106, 163-168. 33. ASH, D. E. (1982) Ph.D. Thesis, University of Pennsylvania. 34. WEISS, P. M., AND CLELAND, W. W. (1983) Fed Proc. 42,192l.