Induced reoxidation and reactivation of a reduced uridine diphosphate galactose-4-epimerase complex.

Abstract An inactive reduced UDP-galactose-4-epimerase complex, which is prepared by reduction with NaBH4 in the presence of substrates and is known to contain DPNH and UDP-hexoses in tightly bound form, is largely reoxidized and partially reactivated in the presence of 0.02 to 0.4 m cyclohexanone or cyclohexanol. Upon placing the epimerase-[4-β-3H]DPNH · UDP-hexose complex in the presence of 0.4 m cyclohexanone about 70% of the DPNH is reoxidized and about 40% of the catalytic activity is restored within 90 min at 27° and pH 8.5, while the major radioactive products are 3H2O and free tritiated DPN+, as well as enzyme-bound tritiated DPN+, but not tritiated cyclohexanol. There is no evidence that cyclohexanone is reduced, and cyclohexanol at similar concentrations causes a similar reactivation. It is probable that these compounds act by causing the tightly bound UDP-hexose molecules to be released from the reduced complex. The resultant epimerase · DPNH complex then undergoes spontaneous autoxidation and reactivation.


Induced Reoxidation and Reactivation of a Reduced Uridine Diphosphate
Galactose-P-epimerase Complex* (Received for publication, February 5, 1973, and in revised form, September 17, 1973) TIMPLE G. WEE AND  about 70% of the DPNH is reoxidized and about 40% of the catalytic activity is restored within 90 min at 27" and pH 8.5, while the major radioactive products are 3H20 and free tritiated DPN+, as well as enzyme-bound tritiated DPN+, but not tritiated cyclohexanol.
There is no evidence that cyclohexanone is reduced, and cyclohexanol at similar concentrations causes a similar reactivation. It is probable that these compounds act by causing the tightly bound UDP-hexose molecules to be released from the reduced complex.
The resultant epimerase .DPNH complex then undergoes spontaneous autoxidation and reactivation.
The Escherichia coli and Xaccharomyces jragilis enzymes contain 1 mole of tightly bound DPN+ per mole of enzyme when purified to near homogeneity, and this appears to constitute a full complement of pyridine nucleotide (l-3). The partially purified mammalian enzymes require added DPN+ to activate them, but they also appear to bind the coenzyme fairly tightly as indicated by very small K, values, in the lop7 M range, for added DPN+ (4).
Recent experiments, in particular, clearly establish that under certain conditions UDP-4-ketosugars are free intermediates (12). Reduced inactive epimerase.DPNH complexes are of interest because of their potential relationship to the DPNH con-* This work was supported by Grant AM 13502 from the National lnstitute of Arthritis, Metabolic and Digestive Diseases.
# Author to whom correspondence should be addressed.
taining complexes which occur as intermediates during the catalytic process. Such complexes can be prepared by treating the yeast or bacterial epimerases with NaBH4 in the presence of a UDP-sugar or with one of several aldohexoses or aldopentoses in the presence of UMP (10, 12, 13). These complexes contain tightly bound UDP-sugar or UMP and they can be reoxidized and reactivated by certain ketones, dTDP-or UDP-4-keto-6deoxyglucose and by 2-deoxyglucose or myo-inosose-2 (10, 14,15). The reactivation by nucleoside diphosphate-4-keto-6deoxyglucoses proceeds by direct transfer of the 4-P-hydrogen of enzyme-bound DPNH to produce the corresponding nucleoside diphosphate-6-deoxyglucose and nucleoside diphosphate-6-deoxygalactose.
The reactions of the reduced S. jragilis enzyme with 2-ketoglucose or myo-inosose-2 have not been shown to proceed by direct hydrogen transfer or to involve actual reduction of the ketones.
We present here the results of some studies on the cyclohexanone-induced reoxidation and partial reactivation of an E. coli epimerase.DPNH.UDP-hexose complex, which is found to proceed without direct hydrogen transfer and probably without reduction of the ketone.

EXPERIMENTAL PROCEDURE
Enzymes-UDP-galactose-4-epimerase was purified from E. coli cells as described earlier (12) following the procedure of Wilson and Hogness (2). Purified acetoacetate decarboxylase from Clostridium acetobutylicum was a generous gift from Professor F. H. Westheimer, Department of Chemistry, Harvard University.
The reduced inactive epimerase . DPNH 'UDP-hexose complex was prepared essentially as described (12) by reducing UDP-galactose-4-epimerase with 0.005 M NaBH4 in the presence of 0.002 M UDP-glucose and then reisolating the protein by gel filtration.
The activity loss upon reduction was typically 98 to 99%, but 5 to 10% of the activity was restored upon gel filtration.

AND DISCUSSION
Cyclohexanone-induced Reoxidation and Partial Reactivation of Reduced Complex-An inactive reduced form of E. coli UDPgalactose-4-epimerase prepared by NaBH4 reduction in the presence of substrates is described in earlier publications (10, 12). It appears to be similar to an abortive complex also described by Nelsestuen and Kirkwood (10) and by us (12), at least insofar as both contain tightly bound UDP-hexose and DPNH and both can be fully reactivated by reaction with TDP-or UDP-4keto-6-deoxyglucose.
In preliminary experiments we attempted to reactivate this comples with simple carbonyl compounds including pyruvate, acetone, acetaldehyde, and cyclohexanone at concentrations in the 0.01 to 0.02 M range at pH 8.5. Of these only cyclohexanone at 0.02 M caused the activity to increase gradually (16). At 0.4 M cyclohexanone we repeatedly found that 40 to 50% of the original activity was restored within 90 min. No activity increase could be detected in the absence of cyclohexanone or in the presence of any of the other ketones.
As shown in Fig. 1 the cyclohexanone-induced reactivation of this complex is accompanied by reoxidation of the enzymebound DPNH, and both follow essentially the same time course. The extent of DPNH reoxidation in this experiment is 73% of the calculated amount of DPNH present initially, whereas the extent of reactivation at its maximum is about 40% of that expected for complete reactivation of this enzyme preparation. The discrepancy results from the fact that cyclohexanone under these conditions causes the activity of the native enzyme to decrease gradually, which can be seen in the data beyond 90 min in Fig. 1.
Fate of T&urn in Reoxidation of Epimerase. [Q-/-3H]DPNH. UDP-hexose Complex-The data in Fig. 1 suggested that cyclohexanone may oxidize the reduced complex, however, we have found no evidence that cyclohexanol is produced.
When the epimerase . [4-/3-aH]DPNH .UDP-hexose complex was subjected to the conditions of Fig. 1 and then to  found in the small molecular weight fraction. Tritiated cyclohexanol could not be detected in this fraction; specifically the radioactivity could not be extracted into cyclohexanol. Moreover, when carrier cyclohexanol was added and reisolated as its crystalline 3,5-dinitrobenzoate, the derivative contained less than 2% of the radioactivity (16). The small molecular weight radioactivity was identified as tritiated DPNf and 31Tz0 which were present in variable but comparable amounts (16). The aH20 was identified by its volatility and by the fact t'hat the tritium was fully exchangeable with the protons of acetone in the presence of acetoacetate decarboxylase which catalyzes this reaction (17).

Autoxidation
Induced by Cyclohexanone and Cyciohexanol-The absence of tritiated cyclohexanol together with the finding that tritiated DPN+ is released from the complex suggest that the action of cyclohexanone may not be based upon its reactivity as a hydride acceptor.
DPNf does not dissociate from this enzyme except under denaturing conditions or in the presence of p-hydroxymercuribeneoate; we believe that the partial release of DPN+ is probably caused by some alteration in the structure of the protein induced by cyclohexanone under the reaction conditions.
If so, it is very likely that substrate molecules would also dissociate under these conditions, probably on a shorter time scale than the dissociation of DPN+ which is almost certainly bound more tightly.
The resulting epimerase. DPNH complex can be expected to undergo spontaneous autoxidation to active epimerase.DPN+ complex (10, 18). This interpretation is supported by Fig. 2, which shows that 0.076 M and 0.15 M cyclohexanol cause a reactivation similar to that induced by cyclohexanone.
Inasmuch as cyclohexanol is very unlikely to be reduced by DPNH and hydrogen transfer to cyclohexanone cannot be detected, yet both molecules promote reactivation, it appears that they act to induce the dissociation of UDP-hexose molecules. The epimerase.DPNH complex then undergoes autoxidation and reactivation. That the en-by guest on March 24, 2020 http://www.jbc.org/ Downloaded from zyme is never fully reactivated is explained by the fact that DPN+ is also released from the protein, probably very slowly and subsequent to UDP-hexose dissociation and autoxidation. This interpretation is further supported by the fact that both cyclohexanone and cyclohexanol at 0.3 M cause a very gradual decrease in the activity of the native enzyme. 1 It has recently been proposed (19) that autoxidation of the catalytic intermediate, the epimerase .DPNH .UDP-4-ketoglucose complex, to epimerase.DPN+ concomitant with the dissociation of UDP-4-ketoglucose can account for our earlier observation that the appearance of UDP-4-ketoglucose, in a form reducible by NaBH4, is a slow process by comparison with the catalytic time scale (11). This interpretation does not account for the fact that the activity of the enzyme is greatly decreased in this process (II), nor is it consistent with several other facts more recently reported and accounted for on the basis that free UDP-4-ketoglucose appears slowly as a result of the slow reversible appearance of an abortive complex (12). In addition, as shown here and elsewhere (10) the epimerase. DPNH .UDP-hexose complexes are not subject to autoxidation at a significant rate.