Purification and Properties of Bacillus subtilis Inositol Dehydrogenase*

Inositol 2-dehydrogenase (EC 1.1.1.18) activity appears during growth of Bacillus subtilis (strain 60015) in nutrient sporulation medium. Its synthesis is induced by myo-inositol and repressed by D-glucose. The enzyme has an apparent molecular weight of 155,000 to 160,000 as determined by sucrose density gradient centrifugation, and it is comprised of four subunits, each having a molecular weight of 39,000 as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The isoelectric point of the enzyme is 4.4 as determined by column isoelectric focusing. The enzyme shows the highest Vmax and lowest Km with myo-inositol as substrate but does not react with scyllo-inositol; it also reacts with the alpha anomer (but not the beta anomer) of D-glucose and with D-xylose. Apparently, the enzyme can remove only the single equatorial hydrogen of the cyclitol or pyranose ring. In contrast to the glucose dehydrogenase of spores, which reacts with D-glucose or 2-deoxy-D-glucose and with NAD or NADP, inositol dehydrogenase requires NAD and does not react with 2-deoxy-D-glucose.

it also reacts with the a anomer (but not the /3 anomer) of D-ghCOSe and with D-xylose. Apparently, the enzyme can remove only the single equatorial hydrogen of the cyclitol or pyranose ring. In contrast to the glucose dehydrogenase of spores, which reacts with D-glucose or 2-deoxy-D-glucose and with NAD or NADP, inositol dehydrogenase requires NAD and does not react with 2-deoxy-D-glucose.
Spore-forming bacteria, such as Bacillus subtilis, synthesize new proteins during the sporulation development such as glucose dehydrogenase (EC 1.1.1.47). Since this enzyme is synthesized only during sporulation, its appearance has been used as one of the markers characterizing the sequential biochemical processes during bacterial sporulation (l-lo). The properties of glucose dehydrogenase, isolated from spores, have been examined for Bacillus cereus (ll-13), Bacillus megaterium (14, 15), and B. subtilis (10). Recently, the synthesis of a glucose dehydrogenase-type enzyme was observed in vegetative cells of B. subtilis when they were grown in nutrient sporulation medium (NSMP). The authors demonstrated that the synthesis of this enzyme was repressed by D-&COSe and other readily metabolizable carbohydrates, and they localized the potential repressor on the metabolic chart (16). Electrophoresis of cell-free extracts indicated that the enzyme differs from spore glucose dehydrogenase ( 10).
The present paper reports the purification of this enzyme, shows that it reacts preferentially with myo-inositol, hence calls it an inositol 2-dehydrogenase (EC 1.1.1.18), and then presents some of its physical and catalytic properties. EXPERIMENTAL PROCEDURES' * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 USC. Section 1734 solely to indicate this fact.
' Portions of this paper (including the "Experimental Procedures," Figs. Sl to S16, Tables SI to SIII, and the References) are presented RESULTS Synthesis of Inositol Dehydrogenase in B. subtilis-During growth of B. subtilis (strain 60015) in NSMP2 medium, an NAD-dependent inositol dehydrogenase activity appeared at the time at which glycerol was depleted (Fig. 1). The increase coincided with that of a glucose dehydrogenase activity confirming earlier results (16). Later, the specific activity decreased while protein synthesis continued. In polyacrylamide gel electrophoresis of vegetative cell extracts, the activities observed with myo-inositol and D-ghCOSe co-migrated; this electrical mobility differed from that of glucose dehydrogenase isolated from spores (10). Both vegetative activities (for myoinositol and D-glucose) were induced (in the absence of Dglucose) by myo-inositol and maintained a constant ratio (Table SI), and they were both repressed during growth in the presence of D-ghCOSe.
In NSMP containing 10 mM myoinositol, the inositol dehydrogenase activity started to increase at the same Asoo value as in a culture grown without added myo-inositol, but it reached a higher specific activity and did not show the later decrease. Apparently, NSMP contained a small amount of myo-inositol which lasted only a limited time, owing to myo-inositol metabolism.
The specific activity of inositol dehydrogenase remained low during growth in a medium containing vitamin-free casamino acids, increased greatly after addition of inositol, and the increase was prevented by the presence of D-ghCOSe (Fig. Sl). A B. subtilis mutant (strain 61663), isolated for its inability to grow on myo-inositol as sole carbon source, produced both the inositol dehydrogenase and the vegetative glucose dehydrogenase activities only slowly; the specific activities eventually reached 10% of the level observed in the parent strain. From all these findings, we conclude that the vegetative glucose dehydrogenase activity is produced by the same protein as the inositol dehydrogenase activity. Physical and Catalytic Properties of Inositol Dehydrogenase-The purification procedure of inositol dehydrogenase isolated from B. subtilis is shown in the Supplement (Table  SII). The NAD-dependent dehydrogenase activities for inositol and glucose maintained a constant ratio during purifkation. The purified inositol dehydrogenase was homogeneous by chromatography on Ultrogel AcA-34 and further chromatography on o-aminohexyl-Sepharose, and it produced a single protein band which co-migrated with the enzymatic activity in disc polyacrylamide gel electrophoresis under various conin miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. ditions (see Supplement). The enzyme has a molecular weight of 155,000 to 160,000 (by sucrose density gradient centrifugation), a Stokes radius of 5.1 x lo-' cm (Sephadex G-200 gel filtration) and an axial ratio of 1.25. It appears to contain four equal subunits of M, = 39,000 and has an isoelectric point of pH 4.4. The optimum pH for enzyme stability is 6.5 (see Supplement).
The pH optimum of inositol dehydrogenation activity (with NAD), in 0.1 M Tris/acetate or Tris/Cl buffer was pH 9.5 (Fig.  2). In potassium/Hepes buffer, the pH curve was slightly shifted to the acid side giving a pH optimum of 9.2. The enzyme reacted (at 100 mM carbohydrate concentration) most with either enzyme using n-mannose, n-sorbitol, n-mannitol, GDP-D-mannose, ADP-n-glucose, D-mannose-&phosphate, CDP-Dglucose, n-glucose-6-phosphate, and 2-deoxy-n-glucose-&phosphate. No detectable activity was observed with inositol dehydrogenase using n-erythrose, n-glucosamine, n-sorbitol, N-acetyl-n-glucosamine, and n-erythritol.   strongly with myo-inositol (loo%), appreciably with D-glucose (25%) and n-xylose (14%), and it showed a trace of activity with D-ribose and D-f?tiCtOSe (Table I). The C-2 isomer of nyo-inositol (scyllo-inositol) was not a substrate (at 7 mM) nor did the compound (at 7 mM) inhibit the reaction with myo-inositol (10 mM). In contrast to inositol dehydrogenase, purified spore glucose dehydrogenase showed no activity with myo-inositol, but it reacted with 2-deoxy-n-glucose, and it could use both NAD and NADP as substrate; inositol dehydrogenase did not reduce NADP in the presence of any substrate. The apparent K, values of inositol dehydrogenase were 0.23 mM for NAD, 18 mM for myo-inositol, 167 mM for Dglucose, 190 mM for D-xylOSe, and 56 mM for a-D-glucose. The apparent K, for NAD did not appreciably depend on the carbohydrate substrate (Table SIII). To study whether the a or the p anomer of D-glucose reacts with inositol dehydrogenase, the reaction rates were determined with freshly dissolved (Y-or ,&D-glucose at 25°C in 0.1 M Tris/acetate, pH 7.0 (to reduce the rate of nonenzymatic mutarotation) (40). The rate of nonenzymatic mutarotation of the substrate was determined in a Cary recording spectropolarimeter. Within the error of measurements, the ,8 anomer of D-glucose did not react with inositol dehydrogenase (less than 5% of the (Y anomer) (Table II), in contrast to the spore glucose dehydrogenase which is apparently specific for P-D-glucose (Footnote 3 and Ref. 41).
To determine with which H group on inositol and glucose the purified inositol dehydrogenase reacts, the transfer of tritium from compounds tritiated at different positions to NAD, producing NAD"H, was measured. Near to completion of the reaction, the diluted reaction mixture was chromatographed on DEAE-cellulose which separated NADH from NAD and other reaction components.
The specific activity (cpm/pmol) of the NADH fraction was constant across the elution profile and the tube containing the maximal amount of NADH had about the same specific activity as that of the input ["Hz]inositol or n-["H,]glucose; n-["Hg]glucose showed no tritium transfer to give NAD"H ( glucose dehydrogenase also showed tritium transfer only with n-["H,]glucose. These findings indicated that the product of the reaction of inositol dehydrogenase with myo-inositol is 2inosose and that with n-glucose is D-glUCOnOlaCtOne (which is nonenzymatically converted to n-gluconate).
The reverse reaction of 2-inosose + NADH += myo-inositol + NAD is demonstrated in Fig. 2 which shows the pH dependence of the initial rates of both the forward and the reverse reactions. The product of the reaction with (unlabeled or uniformly 14Clabeled) o-glucose was identified as gluconic acid by paper chromatography.

DISCUSSION
B. subtilis can grow on myo-inositol as sole carbon source; the biochemical pathway of myo-inositol catabolism in B. subtilis is not known but may be similar to that in Aerobacter aerogenes (21) (now reclassified as Klebsiella aerogenes). Inositol dehydrogenase, the first enzyme in the catabolic pathway, is induced by myo-inositol and repressed (even in the presence of inositol) by D-ghCOSe. The molecular weight of the B. subtilis enzyme is 150,000 to 160,000, which is about twice the molecular weight (74,000) of the NAD-specific mammalian (brain) inositol dehydrogenase (22). The molecular weight for the K. aerogenes enzyme is similar to that of the B. subtilis enzyme as we have observed by Ultrogel AcA-34 chromatography of a partially purified enzyme preparation (obtained from Sigma Chemical Co.). The catalytic properties of the highly purified B. subtilis enzyme are similar to those of the inositol dehydrogenase of K. aerogenes (23-26) and Cryptococcus melibiosum (27). For example, the pH optima of both the B. subtilis (Fig. 1) and the K. aerogenes enzymes are pH 9.0 to 9.5 for the reaction from myo-inositol to 2-inosose and pH 7 for the reverse reaction. Larner et al. (24) have proposed that the high pH optimum of the forward reaction may be due to the liberation of H' as a product, causing the rate in this direction to be reduced by high hydrogen ion concentrations.
The apparent K, values for inositol and NAD (forward direction) at pH 9 are about lo-fold higher than the corresponding K,,, values for 2-inosose and NADH (reverse direc-tion) at pH 7 (or pH 9). Product inhibition studies of the B. subtilis enzyme show (Supplement) that NAD' and NADH mutually compete with each other, whereas 2-inosose is a noncompetitive inhibitor of both NAD' and inositol at low concentrations of inositol. At saturating concentrations of inositol (e.g. 200 IIIM), 2-inosose is an uncompetitive inhibitor of NAD. These data are consistent with an ordered Bi Bi reaction pathway (28), i.e. the enzyme reacts first with NAD and then with inositol yielding, first, 2-inosose and then NADH. These results are similar to those reported by Vidal-Leiria and Uden (27) for the inositol dehydrogenase of C. melibiosum (yeast). Fig. 3 displays the structures of the three substrates of inositol dehydrogenase, myo-inositol, a-D-ghCOSe, and a-Dxylose, in such a way that their configurational similarities can be immediately seen. It is apparent that position 1 of a-Dglucose (and a-n-xylose) corresponds to position 2 of myoinositol. It was known (21-23), and has been confirmed by us, that myo-inositol is oxidized by inositol dehydrogenase at position 2 to form 2-inosose. By measurements of the transfer of tritium from glucose-1-"H to NAD we have also shown that the enzyme oxidizes position 1 of  and produces gluconate (probably via gluconolactone).
The enzyme apparently reacts only with an equatorial hydrogen associated with an axial OH group as is demonstrated by the facts that myoinositol and a-D-&COW are oxidized, whereas scyllo-inositol and ,&n-glucose, which have only axial hydrogen groups, are not oxidized by the enzyme. The oxidation of D-ghCOSe by inositol dehydrogenase now explains the appearance of a glucose dehydrogenase-type activity observed during exponential growth of B. subtilis in nutrient broth (16). When this medium also contains D-glucose or glycerol, the enzyme appeared only after these carbohydrates had been used up. Without myo-inositol addition, the enzyme may be produced only as long as the small supply of myo-inositol in nutrient broth lasts.
Inositol dehydrogenase differs in many ways from the spore glucose dehydrogenase which can also catalyze the oxidation of D-glucose in the presence of NAD (10). Major differences are: in contrast to inositol dehydrogenase, the spore glucose dehydrogenase does not react with inositol but it reacts with 2-deoxy-n-glucose and with either NAD or NADP, and it rapidly loses its activity at pH 8 or higher; it has a lower molecular weight (115,000) and a slightly more basic isoelectric point (pH 4.9) than inositol dehydrogenase (pH 4.4) when they are electrophoresed together. Whereas inositol dehydrogenase can be produced by vegetative cells, the spore glucose dehydrogenase is normally produced only during sporulation and then apparently exclusively within forespores (10).