A Multiple Role for the Coenzyme in the Mechanism of Action of 6-Phosphogluconate Dehydrogenase THE OXIDATIVE DECARBOXYLATION OF 2-DEOXI--6l’HOSPHOGLUCONATE*

6-Phosphogluconate dehydrogenase from catalyzes the oxidative decarboxylation of 2-deoxy-6-phos-phogluconate. The 3-keto-2-deoxy-6-phosphogluconate, an intermediate of the reaction, is reduced to 2-deoxy-6-phos-phogluconate and decarboxylated to I-deoxyribulose 5-phos-phate when incubated with the enzyme and TPNH. The decarboxylation process does not occur in the absence of the reduced coenzyme, which does not have, in this step, an oxidation-reduction role. decarboxylation

The 3-keto-2-deoxy-6-phosphogluconate, an intermediate of the reaction, is reduced to 2-deoxy-6-phosphogluconate and decarboxylated to I-deoxyribulose 5-phosphate when incubated with the enzyme and TPNH. The decarboxylation process does not occur in the absence of the reduced coenzyme, which does not have, in this step, an oxidation-reduction role. Since TPNH also has a non-redox role in a tritium exchange reaction catalyzed by the enzyme, it appears that the coenzyme has a multiple role in the mechanism of action of 6-phosphogluconate dehydrogenase: a redox role in the dehydrogenation and another (or others) role(s) in the decarboxylation and tritium exchange reactions. The hydroxyl group present at carbon 2 of 6-phosphogluconate seems to have a dual role in the mechanism of action of the enzyme: one in the binding of the substrate to the enzyme, another in enhancing the decarboxylation of the dehydrogenation product. These findings are discussed with relations to the mechanism of action of isocitrate dehydrogenase and of the malic enzyme.
The enzymatic oxidative decarboxylation of 2-deoxy-6phosphogluconate is a new step for the metabolism of the metabolic inhibitor 2-deoxyglucose. boxylation of, respectively, oxalosuccinate (I, 2) and oxaloacctate (3-7); these two P-k&o acids have so far escaped detection as free intermediates of the over-all oxidative decarboxylntion of isocitrate and malate.
The failure to trap thcsc three P-keto acids has been justified with the hypothesis that they are enzyme-bound and are never released in the medium (8-10).
But thcrc is another explanation based on the rates of the partial reactions.
In the case of at least isocitrate dehydrogenase (9-12) and 6~phosphogluconate dehydrogenase (13), the dehydrogenation is the rate-limiting step, thus the decarbosylation is faster than the oxidation; in the backward reaction the reduction is faster than the carboxylation. For t,hese reasons there is no possibility to accumulate the intermediate. Furthermore, it is not possible to obtain the carboxylation without reduction, running the reaction backward in the absence of TI'NH, since the reduced coenzyme is essential (14, 15) in a step which very likely precedes the carboxylation, i.e. the tritium exchange reaction bctwcen the products of the enzymatic oxidative dccarboxylation and water. In order to accumulate a ,&k&o acid intermediate, one could try to decrease the rate of dccarbosylation, either by n chemical modification of the enzyme, or by using a substrate analogue which has a lower tendency to decarboxylatc.
Choosing this second approach, we have used as substrate of the B~phosphogluconate dehydrogenase from Cundid~ ufilis the 2-deoxy&phosphogluconate, which, lacking the electron withdrawing effect of the hydroxyl at carbon 2, should be more resistant to decarboxylntion than 6-phosphogluconate. The rcsuits of this approach arc reported in the present paper. EXPERIMENTAL PROC!I!X1URlZ 6-Phosphogluconate dehydrogenasc, type I (16), crystalline, was prcparrd from C. u!ilis as previously described (17). The enzyrne used in this work had n specific activity of 34 i.u. Prior to the experiments the crystals wcrc collected by centrifugation and dissolved in quartz-distilled mater, and the resulting protein solution was desalted through a Sephadex G-50 column equilibrated with 50 m&f Tris-HCI buffer, pH 7.5, containing 1 111~ EDT?!.
6-Pllosplloglllcorrnte dehydrogenase, type II from C. dilis, was prepared as previously reported (16). B-Phosphogluconate dehydrogenase was purified also from bakers' yeast and hog kidney. After barium precipitation of ATP and ADP, the radioactive 2-deoxyglucose 6-phosphate was purified by column chromatography on Dowes I-Cl. The purified compound was thrn oxidized with bromine (18) to 2-deoxy-6-phosphogluconate which was purified by a second Dowex l-Cl column chromatography.
The final product had a specific radioactivity of 81,600 cpm per pmole.

Chemical and Enzymatic
Determinations--2.Deoxyglucose B-phosphate was assayed enzymatically at pH 8.0 using TPN and glucose 6-phosphate dehydrogenasc from C. utilis. The enzymatic activity of 6-phosphogluconate dehydrogenase on 2-deosy6-phosphogluconate was assayed at ~1-1 8.0 in 50 m&r Tris-HCl buffer, in the presence of 1.8 mM TPN, taking readings at 340 nm in an ACTA III recording spectrophotometer at 5-s intervals.
The concentration of 2-deoxy&phosphogluconate was determined as above except that 2 mM MgCla was added to the reaction mixture to accelerate the dccarboxylation of the 3-keto acid intermediate.
Radioactivity measurements were carried out in a Packard Tri-Carb liquid scintillation counter, using Bray's solution (20). Determination by Radioactivity dleasurements of W-Deoxy-6phosphogluconate and S-Keto-%deoxy-6-phosphogluconate-These determinations were carried out following the observation that the keto compound was dccarboxylated, and thus loosed radioactivity, by boiling or by treatment with 4-aminoantipyrine, while 2.deoxy%phosphogluconate was resistant to these treatmcnts. Thus, in order to establish the concentrations of these two radioactive compounds in a solution containing both of them, the following procedure was used. One sample was diluted with acetate buffer (final concentration, 1 InM, pII 3.8), treated for 5 min at room temperature with 4-aliiiiioaiitip~rilie (final concentration, 20 rnnl) and then kept under vacuum for 5 min. The residual radioactivity was attributed to 2-dcoxy-6-phosphogluconate.
Another sample was treated as above, except that the treatment with aminoantipyrine was omitted; the residual radioactivity was attributed to the sum of 2-deoxy-fi-phosphogluconate and 3-kcto-2-deoxy-6-phosphogluconate; then by difference between these two values it was possible to determine the concentration of the radioactive 3-keto-2-deoxy-B-phosphogluconatc.
The vacuum treatment was required to eliminate the radioactive CO2 (volatile radioactivity) which was formed by the decarbosylation of t,he keto acid. C. utdis, in the presence of TPN, catalyzes the oxidative decarboxylation of 2-deoxy-6.phosphogluconatc. At p1-I 8.0 and in the presence of 1.8 mM TPN, the Ii, of the enzyme for 2-deoxy-6phosphogluconate was 0.55 mu as compared with 0.054 mM for 6-phosphogluconate (17). The Ii, for TPN was 0.020 rnM using either 2-deoxy-6-phosphogluconate or 6-phosphogluconate. When the dehydrogenase activity was tested in presence of a concentration of substrate corresponding to the R, value, 1 mg of enzyme catalyzed the oxidation of 0.23 and 17.2 pmoles per min of 2-deoxy-6-phosphogluconate and 6-phosphogluconate, respectively.

RESULW
Thus, in these experimental conditions, the dehydrogenation of 2-deoxy-6-phosphogluconate was 75fold slower than that of 6-phosphogluconate (Table I, Experiments 1 and 2). The ratio between the dehydrogenase activity on these two substrates was constant in all steps of the 340.fold purification procedure (17) of the enzyme.
The same ratio has been found also using the enzyme type II from C. utilis and the enzyme purified from bakers' yeast and hog kidney.
TPN was the specific coenzyme for this activity, DPN was inactive.
Evidence for an Intermediate in Oxidative Decarboxylation of d-Deoxy-6-phosphogluconate-When 2-deosy-6-phosphogluconatc was used as substrate, the reduction of TPN in the first minutes of the reaction was paralleled by a formation of a compound which gave a positive reaction with diazotised p-nitroaniline, and thus was a P-keto acid (19). This positive reaction disappeared if the reaction was allowed to proceed for longer time, indicating that the intermediate was consumed (Fig. 1). The observed decrease in the rate of TPNH formation was due to the inhibitory effect of TPXH (21). The disappearance of the intermediate is to be attributed to the increase of concentration of TPNH, which enhances (see below) the decarbosylation of the fl-keto acid.
Preparation and Isolation oJ' l&ennediate-The preparation of the intermediate of the oxidatiye clecarboxylation of 2-deoxy-6phosphogluconate was accomplished running the reaction in the presence of pyruvate and lactate dehydrogcnase, in order to recycle the TPNH formed during the oxidation of the 2-deoxy-6phosphogluconate.
In order to follow the reaction, two samples of 50 ~1 each were withdrawn at each time interval and one sample was boiled for 5 min. The two samples were then acidified and kept 5 min under vacuum to eliminate the radioactive CO* formed by the decarboxylation of the intermediate. The radioactivity found in the boiled sample was attributed to unreacted 2-deoxy-6-phosphogluconate, the radioactivity of the unboiled sample was attributed to the sum of unreacted 2-deoxy-S-phosphogluconate and of the P-keto acid; this was possible since ,Bketo acids are decarboxylated on boiling. As shown in Fig. 2 The fractions were analyzed for radioactivity and formazan formation.
At the end of the experiments, i.e. when no further increase of the concentration of the intermediate was observed, the incubation mixture was treated with 600 mg of charcoal to absorb the enzymes and the coenzyme and centrifuged.
The supernatant was then subjected to a column chromatography.
Only two radioactive peaks were obtained (Fig.  3). The material contained in the first radioactive peak was identified as unreacted 2-deoxy+phosphogluconate, since there was no loss of radioactivity upon boiling and the reaction with diazotized p-nitroaniline was negative.
The material of the second radioactive peak was identified as 3-keto-2-deoxy-6-phosphogluconate.
IdentiJication of Infermediate as S-Keto-2-deoxy-6-phosphogluconote--The material present in the second radioactive peak obtained after column chromatography was identified as 3-keto-2deoxy-6-phosphogluconate by the following criteria: (a) it gave a formazan derivative upon coupling with diazotized p-nitroaniline; (b) it was radioactive and thus contained carbon 1 of 2-deoxy-6phosphogluconate; (c) upon boiling there was complete loss of radioactivity and no more formazan formation; (d) by its chromatographic behavior, it appeared to be a phosphorylated compound containing a carboxyl group.
Additional evidence of a P-keto acid structure was furnished by the observation that the compound lost radioactivity (Fig. 4)  6-phosphogluconate and in part decarboxylated to l-deoxyribulose 5-phosphate.
The reduction to 2-deoxy+phosphogluconate was shown by the parallel oxidation of TPNH (followed spectrophotometrically) and the increase of radioactivity stable to heat and 4-aminoantipyrine treatments; the decarboxylation was shown by the decrease of nonvolatile radioactivity (Fig. 5). At the end of the reaction the lack of a positive reaction with diazotized p-nit,roaniline confirmed that all of the 3-keto-2-deoxy-6phosphogluconate had been consumed. At this point the addition of TPN (to shift the equilibrium existing between the 2.deoxy-6.phosphogluconate and 3-keto-2-deoxy-phosphogluconnte toward this last compound) and MgClz (to accelerate the decarboxylation of the fi-keto acid) caused a complete loss of radioactivity, indicating t,hat the radioactive compound produced by the enzymatic reduction of the 3-keto-2-deoxy+phosphogluconate was oxidized and decarboxylated and thus was the 2-deoxy-6-phosphogluconate.
In our experimental conditions, the rates of decarboxylation and reduction were of the same order of magnitude (Table I, Experiments 3 and 4). When either TPNH or enzyme were omitted in the incubation mixture, no decarboxylation occurred. There was also no decarboxylation if TPNH was substituted with TPN, DPN, or DPNH.
These results indicated that the reduced coenzyme was essential to the decarboxylation reaction. The fact that TPNH causes, in the presence of the enzyme, the decarboxylation of 3-keto-2-deoxy-6-phosphogluconate, explains why in the presence of pyruvate and lactate dehydrogenase, i.e. in conditions where there is no accumulation of TPNH, it is possible to accumulate the 3-keto-2-deoxy+phosphogluconate.
In previous papers we have reported some basic properties of 6-phosphogluconate dehydrogenase from C. utilis (for review see Reference 25). In order to obtain more information on the mechanism of action of the enzyme, on the nature of the intermediate product of the oxidative decarboxylation of 6-phosphogluconate, and to clarify why the intermediate products of the oxidative decarboxylations of isocitrate and ma.late were never trapped, m-e have used as substrate of the 6-phosphogluconate dehydrogenase the Z-deoxy-6-phosphogluconate and observed that this substrate analogue is first dehydrogenated to 3-keto-2deoxy-6-phosphogluconate and then decarboxylated to l-deoxyribulose 5-phosphate.
The dehydrogenase activity of the enzyme on 2-deoxy-6-phosphogluconate is only 1.5y, of that on 6-phosphogluconate. Hence the possibility that this low activity could be due to an enzymatic contaminant in an otherwise physicochemically homogeneous preparation cannot be overlooked. The evidences for a single enzyme catalyzing the oxidation of both 6-phosphogluconate and 2-deoxy-R-phosphogluconate are the following: (a) both activities require TPN, and the enzyme has the same affinity for t,he coenzyme using both substrates; (b) the two substrates differ only in one substituent at carbon 2; (c) the two substrates undergo same transformations; (d) the ratio between the dehydrogenase activity on the two substrates is constant during all steps of the 340-fold purification procedure of the enzyme; (e) the same ratio is obtained using the enzyme prepared from three different sources; (f) to our knowledge no other enzyme has been reported to catalyze the oxidative decarboxylation of 2deoxy-6-phosphogluconate.
Finally, the lower activity on 2deoxy-6-phosphogluconate can be partially due to the fact that, in this case, the decarboxylation is the rate-limiting step, whereas in the case of 6-phosphogluconate the rate-limiting step is the dehydrogenation (13). Following the enzymatic oxidation of 2-deoxy+phosphogluconate, the transient formation of an intermediate having the characteristics of a P-keto acid has been detected; this compound accumulates in the first minutes of the reaction and then disappears being decarboxylated.
If the reaction is instead carried out in the presence of pyruvate and lactate dehydrogenase, to efficiently recycle the TPNH, the substrate is oxidized to a pketo acid which is not, in the absence of TPNH, decarboxplated, and can hence be accumulat,ed, purified, and identified as 3-keto-2-deoxy-6-phosphogluconate.
The 3-keto-2-deoxy-B-phosphogluconate prepared in this way is a true intermediate of the oxida-tive decarboxylation of 2-deoxy-6-phosphogluconate; indeed, in t,he presence of TPNH and enzyme, it is reduced to 2-deoxy-6phosphogluconate and decarboxylated to Ldeoxyribulose 5phosphate.
Since the enzyme, in the absence of TI'NH, is unable to catalyze the decarboxylation of the 3-keto-2-deoxy+phosphogluconate, it appears that the presence of the reduced coenzyme is essential to the decarboxylation step. TPNH is essential also in the tritium exchange reaction between the ribulose 5-phosphatc and water, catalyzed by the 6-phosphogluconate dehydrogenase (IL?), and it has been proved that in this reaction the reduced coenzyme does not have a redox role (25). A redox role for the reduced coenzyme in the decarboxylation step would implicate the reduction of 3-lieto-2-deoxy-6-phosphogluconate to 2-deoxy-6-phosphogluconate which should in turn be decarboxylated bypassing the 3.keto-2.deoxy-6-phosphogluconate; this process is to be excluded, since it has been observed that the 2deoxy-6-phosphoglutonate, obtained by reduction of the P-keto compound, is, in our experimental conditions, not further transformed.
From our results it appears that, in the reaction catalyzed by the 6-phosphogluconate dehydrogenase, the coenzyme has at least a dual role: (a) a redox role in the first step (dehydrogenation) ; and (b) a nonredox role in the decarboxylation and in the tritium exchange reactions.
If the enzymatic decarbosylation of 3-keto-2-deoxy+phosphogluconate includes t,wo steps, (a) the loss of CO2 to form an enolate, and (b) the enzymatic ketonizat,ion to form the product which is liberated, it could be possible that the reduced coenzyme has a role only in the ketonization process and not in the decarboxylation.
Studies are in active progress in this laboratory to establish the mechanism(s), not connected with an osidoreduction, by which the reduced coenzyme stimulates both the decarboxylation and the tritium exchange reactions.
It is to be recalled that TPNH is essential for the tritium exchange reaction catalyzed by isocitrate dehydrogenase (26) and that TPN stimulates the decarboxylation of oxalacctate catalyzed by the malic enzyme (3-7).
Our results indicate also that, in the presence of enzyme and TPNH, the chemical equilibria existing between the 3-ket,o-2deoxy-6-phosphogluconate and its reduction and decarboxylation products are unfavorable to the accumulation of the intermediate. Also in the cases of 2-deoxygluconate dehydrogenase (23) and /%hydroxy acid dehydrogenase (24), the chemical equilibrium is shifted toward the reduced substrate.
Coupled with the considerations on the rates of the partial reactions (exposed in the introduction of this paper), this finding could explain why the oxalosuccinate, the oxalacetate, and the 3-keto-6-phosphogluconat'e were never trapped as intermediates of the oxidative decarboxylation of isocitrate, malatc, and 6-phosphogluconate. The 3-keto-6-phosphogluconate, very likely formed by the enzymatic oxidative decarboxylation of the 6-phosphogluconate, has an hydroxyl group at carbon 2; the electron-attracting effect of this group is expected to develop at carbon 2 of B-phosphogluconate, an electron deficiency which results in an easy release of t,he CO,; in these conditions the dehydrogenation could be the rate-limiting step, as experimentally found (13). Using instead as substrate the 2-deoxy&phosphogluconate, the decarbosylntion is slower and is the rate-limiting step, in fact, the formation of the ,L-keto acid has been observed.
It appears therefore that the hydroxyl group present at carbon 2 of 6-phosphogluconate has a dual role in the mechanism of action of the enzyme : an orientation role for the substrate in the active site of the enzyme (in fact the 6-phosphogluconate has a Km for the enzyme IO-fold lower than 2-deoxy-6-phosphogluconate) and a chemical role favoring the decarboxylation of the intermediate. We were unable so far to obtain a direct chemical identification of the product of the enzymatic osidative decarboxylation of 2-deoxy-6-phosphoglueonate.
This product should be the ldeosyribulose 5-phosphate on the basis of the analogy existing between the reaction we have studied and the oxidative decarboxylations of A-phosphogluconate (27), 2-deoxygluconate (23), and gulonate (24) which yield ribulose 5-phosphate, l-deosyribulosc, and xylulose. respectively.
To the best of our knowledge the I-deosyribulose 5-phosphate was never reported in the literature.
When the oxidative decarborylation of 2-deoxy-6.plrosphogluconate was carried out in tritiated water, the pento,se phosphate obtained contained 1 atom of tritium per residue of phosphate. Incubating this tritiated I-deoxyribulose 5-phosphate with TPNH and enzyme in the conditions described (15) for the tritium exchange between ribulose 5-phosphate and water, no appreciable exchange of tritium was obtained.
This result was not unexpected; comparing t,he possibilities of tritium exchange of ribulose 5-phosphate and I -deoxyribulose 5-phosphate, it appears that for the second compound the rate of tritium exchange should be one-third of that of the first compound, since all three hydrogens are equivalent.
Furthermore the hydroxyl group present at carbon 1 of ribuloae S-phosphate has both an orientation effect for the substrate in the active site of the enzyme and a labilization effect for the hydrogen exchangeable with the medium; in I-deoxyr,ibulosc 5-phosphate, which lacks this hydroxyl, both effects are absent and thus the rate of tritium exchange should be much lower.
Finally a last point is of interest from the metabolic point of view. 2.Deoxyglucose is a well known metabolic inhibitor (28). It can be phosphor,ylated to 2-deoxyglucose B-phosphate, oxidized to 2-deoxy-6-phosphogluconate and then metabolized in different ways (28). From our experimental results it appears that 2-deoxy-6-phosphogluconate can be also oxidatively decarbosylated to l-deosyr,ibulose 5-phosphate, following another step of the oxidative pathway of the metabolism of glucose. Research is in progress to elucidate the further metabolism of l-deoxyribulose 5-phosphate and the possible action of this compound as metabolic inhibitor.