Stereochemistry of pyruvate kinase, pyruvate carboxylase, and malate enzyme reactions.

Abstract Using isotopically asymmetric phosphoenolpyruvate-3-d,t it was established that pyruvate kinase causes the addition of a proton to C-3 from the face designated by a counterclockwise sequence of phosphate, carboxyl, and vinyl groups, the si face. Having shown that the pyruvate carboxylase reaction has a normal primary isotope effect, it was possible to determine its stereochemistry with enantiomorphic pyruvate prepared in the pyruvate kinase reaction. Carboxylation occurs with retention of configuration at carbon-3 of pyruvate. With this information, it was possible to analyze the enantiomorphic pyruvates formed by reaction of (3S)- or (3R)l-malate-3-d,t with malate enzyme. The decarboxylation occurred with retention.

preparation of isotopically enantiomeric forms of pyruvate. With these forms, and applying the observation that the pyruvate carboxylase reaction discriminates against the heavier hydrogen isotopes, the stereochemistry of this reaction was directly determined by examination with fumarase of the malate formed in trapping the oxalacetate produced. Finally, with specific L-malate-d, t, the stereochemistry of the malate dehydrogenasedecarboxylating (herg referred to as malate enzyme) reaction in Hz0 was determined by analysis of the pyruvate with the pyruvate carboxylase-fumarase sequence.
The results of these studies are considered in terms of structural and mechanistic aspects of the three reactions. MATERIALS AND METHODS Preparation of CompoundsA mixture of 2 and 3 n-phosphoglycerates (PGA), containing both deuterium and tritium stereospecifically located at C-3, were prepared with purified glycolytic enzymes as previously described (3) in such a way as to produce (3X)PGA-3d,Z by starting with glucose-l-t and using DtO in the phosphoglucose isomerase step, and to produce (3R)PGA by starting with glucose-l-d and using 3HOH in the phosphoglucose isomerase step. The glucose-l-d was synthesized from D-ghcone-&la&one by Na-Hg reduction (7) in DzO. Both compounds should contain greater than 95% deuterium in their specific positions and tracer amounts of tritium mixed with predominant hydrogen in the position geminal with the deuterium.
The specific activities of tritium were: 10' cpm per pmole for the (3X)PGA and 5.7 x lo4 cpm per pmole for the (3R)PGA. L-Malate, containing both deuterium and a tracer of tritium, both stereospecifically located at C-3, were prepared as follows: (3R)L-malate-2d-3d, t was prepared by reaction in tritiated water of fumarase with fumarate-2,3-dz (89 atom Q, which was kindly provided by Dr. Sasha Englard. The malate was recovered from a Dowex I-formate column with 1 N formic acid. (3X)L-Malate-2t-3d, t was prepared by reaction in DsO of fumarase with fumarate-2,3-t*. The latter compound was prepared from D , L-malate-2-t (New England Nuclear) by extracting 2 The R/S system of Cahn, Ingold, and Prelog is used with the priority rule that higher mass number precedes lower (6). Since tritium is present only in tracer concentration, the chirality symbol R or S is understood to refer to that component of the mixture of isotopic species that contains all three hydrogen isotopes in the case when a methyl group is named and to the specified two hydrogen isotopes when a methylene group is named. from yeast was prepared from yeast by the method of Dixon and Kornberg (9). The other enzymes were obtained commercially from Boehringer Mannheim.
Care was taken to limit the amount of these enzymes as much as possible to limit the introduction of significant fumarase activity. The reaction mixture was dried from dilute acid to remove unreacted acetic acid and tritiated water.
Malate, 0.5 to 0.8 pmole, was isolated from a Dowex I-Cl-column by elution with 5 mn/r HCl.
To determine the tritium distribution in the two C-3 positions of malate, part of the malate was reacted with fumarase (3.5 units) in 40 mu KPO( buffer, pH 7.0, for 30 min at 25". The counts that are recovered in the distillate of such a reaction mixture are attributed to the C-3 pro-R (10) position of malate (11,12). A second sample of malate was converted to oxalacetate within 10 min at pH 9.5 with malate dehydrogenase (70 units) and 3-acetylpyridine-DPN and hence served as the assay for malate (13). As predicted from the data of Banks (14), both methylene hydrogens of oxalacetate fully exchange with the alkaline medium within 60 min and could be determined by distillable radioactivity.
The difference in counts between the two determinations is attributed to the C-3 pro-S position of malate.
Pyruvate formation was monitored on small samples with lactate dehydrogenase.
When the reaction was complete in about 10 min, HzOz (100 pmoles) was added to convert the pyruvate to acetate, which was eluted from Dowex-l-Cl-with 5 mM HCl. The neutralized acetate was concentrated in a vacuum and converted to malate by the acetate kinase-transacetylase-malate synthase system. The isolated malate was treated to determine the distribution of tritium, with the results shown in Table I.
The malate is observed to have 80% of the tritium-specific activity of the starting PGA.
The decrease in specific activity may be attributed to several sources: in the pyruvate kinase step at pH 6.5, about 3 to 5% of the counts of PEP-3-t are diverted into water during the conversion to pyruvate-3-t. 3 In the conversion of acetate-3-t to malate, about 15% of the counts are diverted into water as a result of the malate synthase step and any fumarase present.
Finally, any contaminating acetate that is picked up from the reagents of the incubation or the ion exchange step will contribute to the decreased specific activity of the malates.
The results of this study, taken together with the known stereochemical course of enolase and malate synthase, lead to the conclusion that the proton adds to the si face4 of C-3 of PEP in the pyruvate kinase reaction (see Scheme 1). Thus, the anti-elimination of water from (3S)PGA-3-d, t in the enolase reaction leads to the indicated form of PEP (3), and in the malate synthase reaction the proton of acetyl-CoA is displaced with inversion (1,2). To the extent that proton is displaced in this reaction in preference to deuterium, tritium will be enriched in the C-3 pro-R position of malate, as indicated by the transfer of radioactivity to water in the fumarase reaction. The degree of enrichments seen are comparable with those previously reported (15), and the more extreme results with (3S)phosphoglycerate suggests an isotope effect of kH/k, = 5.
It may be argued that the face of enolpyruvate, to which the proton is added in the pyruvate kinase reaction, is turned toward the medium, whereas the opposite face is bound to the surface of the enzyme.
It has been demonstrated in this laboratory3 that under certain conditions PEP-3-t can undergo loss of as much as 70% of its tritium to the medium during the net reaction in which pyruvate is trapped by rapid reduction with lactate dehydrogenase. Since all elements of the reaction, i.e. ATP, enolpyruvate, K+, and a divalent cation, particularly Cozf, are present in the complex at the time of proton release, the extensive reorganization of the complex that would be required if proton release occurred from the direction of the binding face of the enolpyru- This indicates a tions, in which phosphoryl transfer is to HzO, Pi, or GDP, CO2 discrimination factor of 4.2-fold in favor of protium relative to addition is invariably to the si face (5) and that this is also the tritium, and hence about 2.7-fold (17) relative to deuterium. case in the allylic replacement by erythrose-4-P to produce From this it follows that carboxylation of pyruvate of known 2-keto-3-deoxy-7-P heptulonic acid (4). These correlations chirality, prepared as in the pyruvate kinase study, would yield suggest that a unique arrangement of amino acids responsible malate-3-t that would be stereoselectively tritiated in a manner yoi "Biotin 0,C co/ D The pyruvates used in these experiments were prepared by reaction of phosphoglycerate mutase, enolase, and pyruvate kinase with the two forms of PGA-3-d, t in Hz0 as above. They were isolated on Dowex l-Cl-columns by elution with 10 mrvr HCl. Their stereochemical assignments are derived from the conclusion that pyruvate kinase adds a proton to the si face of PEP. The incubations contained, in 1 ml: Tris-Cl (0.2 M, pH 7.8); ATP (0.8 mM) ; MgCln (5 mM) ; KHCO, (20 mrvr) ; K&04 (2 mM>; acetyl-CoA (60 PM) ; DPNH (0.2 mM) ; malate dehydrogenase (7 units) ; pyruvate carboxylase (0.4 unit) ; and about 0.1 pmole of (3X)-or (3R)pyruvate. Within about 10 min, the absorbance decrease at 340 rnp was complete and the malates were recovered from Dowex l-Cl-columns by elution with 5 mu HCl. The malate was examined for tritium distribution as before.
The results of such an experiment using enantiomorphic forms of pyruvate are given in Table II. They indicate that the reaction occurs with retention of configuration. The degrees of tritium asymmetry correspond to intramolecular isotope effects, k&o of 3.5 and 1.8 for the two substrates. The disagreement is in the same direction as found in the previous experiment (Table I) for which the same sources of pyruvate were used. One might reasonably attribute this discrepancy to a lack of isotopic homogeneity in the glucose-l-d in the preparation of (3R)PGA-3-d, t and hence (3R)pyruvate. The stereochemistry of only one other biotin-containing carboxylase has been determined, namely propionyl-Coil carboxylase, which was shown to proceed by retention in the formation of (X)methylmalonyl-CoA (18,19). Additional examples should be analyzed before any suggestion of a stereochemical rule can be made. It is noteworthy, however, that enzymes of this class undergo proton activation of the substrate only when the biotinprotein is in the carboxylated state (19,20). This result alone might be interpreted in terms of a concerted mechanism, such as suggested by Mildvan and Scrutton (21) : 04$...,, This mechanism restricts the stereochemistry of proton replacement to that of retention, and hence the present results are consistent with this mechanism. Malate Enzyme-To study the stereochemistry of the decarboxylation of malate by this enzyme, the pyruvate formed from specifically labeled L-malate-3-d) t in Hz0 was converted back to malate with the pyruvate carboxylase-malate dehydrogenase sequence, and the malate was analyzed for its C-3 tritium distribution as before (Scheme 3).
In the conversion of malate to pyruvate, care was taken to limit the effect of a contaminating fumarase activity that was present in the enzyme from E. coli by carrying out the reaction to only about 28% conversion in a 0.3-ml incubation containing Tris-Cl (50 mu, pH 7.4) ; EDTA (1 II~M) ; dithiothreitol (1 mM); KC1 (0.1 M) ; TPN (15 111~) ; MnCls (5 n-& ; E. co&malate enzyme (0.03 unit) ; and 2 pmoles of (3R)-or (3S)malate-3-d, t. The reaction was monitored at 340 rnh and pyruvate was recovered on Dowex-l-Cl-by elution with 10 mu HCl. In the case of the (3R)malate, about 24% of the radioactivity applied to the column was present as water in the column breakthrough, presumably the result of action by the fumarase contaminant in , . . . ) 63,800 1 ii / ii / fi exchanging the C-3 pro-R tritium for a proton from the medium.
In (3S)malate, the exchange is between protons; hence the tritium in the pro-X position remains fixed. The pyruvate was converted to malate as before, and the distribution of radioactivity in C-3 of the synthesized malate was determined with fumarase.
The results shown in Table III indicate clearly that the decarboxylation occurs with retention of configuration.
The isotope discrimination seen with (3S)malate agrees with the larger effect observed in the pyruvate carboxylase study, indicating k&kD s 3.3.
A parallel study with the pigeon liver malate enzyme is reported in Table IV. In this experiment, the malates (1 mu) were incubated with triethanolamine-Cl (50 mM, pH 7.5), MgClz (4 mu), dithiothreitol (1 KIM), TPN (2 mM), and 0.1 unit of malate enzyme per ml. The reaction was completed within 30 min. Only about 2.5% of the radioactivity was labilized in either incubation at this stage. The further steps were like those in the previous experiment.
The results indicate that the pigeon liver malate enzyme reaction proceeds with retention as in the case of the E. coli enzyme.
The stereochemistries of several other enzymes that carry out pyridine nucleotide-dependent oxidative decarboxylations of P-hydroxyacids have been reported. Roth the TPN-specific (22,23) and DPN-specific (24) isocitrate dehydrogenases also show complete retention in the decarboxylation step. However, 6-P-gluconate hydrogenase (25) brings about inversion in the replacement of carboxyl, and a related reaction catalyzed by UDP-glucuronate carboxylase leads to inversion (26). The