Inositol dehydrogenase (IDH, EC 1.1.1.18) from <i>Bacillus subtilis</i> catalyzes the reversible NAD⁺-dependent oxidation of the axial hydroxyl group of <i>myo</i>-inositol to form 2-keto-<i>myo</i>-inositol, NADH and H⁺. IDH is the first enzyme in catabolism of myo-inositol, and <i>Bacillus subtilis</i> is able to grow on <i>myo</i>-inositol as the sole carbon source. Our laboratory has previously shown that this enzyme has an unusual active site that can accommodate large hydrophobic substituents at 1L-4-position of <i>myo</i>-inositol.<p> In this dissertation, the further characterization of this IDH is described, with focus on the mechanism, inhibition, kinetics, substrate binding, and alteration of substrate specificity. A kinetic isotope effect study revealed that the chemical step of the reaction was not rate-limiting. In order to probe the inositol-binding site, five inositol analogues were synthesized and evaluated as competitive inhibitors. Recently the crystal structures of the <i>apo</i>-IDH, <i>holo</i>-IDH and ternary complex have been solved. Using structural information, as well as modeling and sequence alignment approaches, we predicted the active site structure of the enzyme. On the basis of these predictions, coenzyme specificity was converted from entirely NAD⁺-dependent to 6-fold preference for NADP⁺ over NAD⁺ by site-directed mutagenesis. The critical residues for coenzyme recognition were therefore identified. Besides coenzyme specificity alteration, eleven amino acid residues in and around the proposed <i>myo</i>-inositol active site were also modified to test their roles in order to improve our understanding of substrate binding and activation.

Inositol dehydrogenase (IDH, EC 1.1.1.18) from <i>Bacillus subtilis</i> catalyzes the reversible NAD⁺-dependent oxidation of the axial hydroxyl group of <i>myo</i>-inositol to form 2-keto-<i>myo</i>-inositol, NADH and H⁺. IDH is the first enzyme in catabolism of myo-inositol, and <i>Bacillus subtilis</i> is able to grow on <i>myo</i>-inositol as the sole carbon source. Our laboratory has previously shown that this enzyme has an unusual active site that can accommodate large hydrophobic substituents at 1L-4-position of <i>myo</i>-inositol.<p> In this dissertation, the further characterization of this IDH is described, with focus on the mechanism, inhibition, kinetics, substrate binding, and alteration of substrate specificity. A kinetic isotope effect study revealed that the chemical step of the reaction was not rate-limiting. In order to probe the inositol-binding site, five inositol analogues were synthesized and evaluated as competitive inhibitors. Recently the crystal structures of the <i>apo</i>-IDH, <i>holo</i>-IDH and ternary complex have been solved. Using structural information, as well as modeling and sequence alignment approaches, we predicted the active site structure of the enzyme. On the basis of these predictions, coenzyme specificity was converted from entirely NAD⁺-dependent to 6-fold preference for NADP⁺ over NAD⁺ by site-directed mutagenesis. The critical residues for coenzyme recognition were therefore identified. Besides coenzyme specificity alteration, eleven amino acid residues in and around the proposed <i>myo</i>-inositol active site were also modified to test their roles in order to improve our understanding of substrate binding and activation.

Lactobacillus casei BL23 expresses two enzymes encoded by the genes iolG1 and iolG2. They have been putatively assigned as myo-inositol dehydrogenases by sequence comparison. The enzyme catalyzes the reversible conversion of myo-inositol to scyllo-inosose and the concurrent reduction of NAD+ to NADH. iolG1 was subsequently determined to be a myo-inositol dehydrogenase but iolG2 was determined to be a scyllo-inositol dehydrogenase. Sequence analysis and kinetics by themselves did not provide insight as to why the enzymes are functionally different. This manuscript provides a structural rationalization for the differences in stereoisomer selectivity by X- ray crystal structure analysis and comparison. High resolution apo, binary, and ternary crystal structures for iolG1 and iolG2 wild type enzymes were determined. For iolG1 the ternary structures were determined for myo-inositol and d-chiro-inositol and for iolG2 the scyllo-inositol bound structure was determined. The high resolution structure information revealed the composition of their respective active sites and showed that subtle differences in critical amino acids for each enzyme define the orientation of the inositol stereoisomer for inline transfer of a hydride to NAD+. Mutagenesis studies of a closely related myo-inositol dehydrogenase from Bacillus subtilis were carried out. The wild type structure for BsIDH had already been determined and characterized. A portion of the results in this manuscript briefly explore structures of dehydrogenase mutants which validate the structural role of residues involved in cofactor selectivity

Inositol dehydrogenase (IDH) catalyzes the oxidation of myo-inositol to scyllo-inosose using NAD+ as the coenzyme. IDH-related genes (Lp_iolG1 to Lp_iolG4) from Lactobacillus plantarum WCSF1 and (Lc_iolG1 and Lc_iolG2) from Lactobacillus casei BL23 were cloned into the vector pQE-80L, expressed in E. coli host cells and the proteins were purified to homogeneity. IDH activity of the purified enzymes was explored with myo-inositol and other structurally related compounds. It was found that IDH-related enzymes from L. plantarum WCSF1 did not exhibit any activity with tested substrates but, LcIDH1 and LcIDH2 from L. casei BL23 showed activity with myo-inositol and other related compounds. pH-rate profile studies have demonstrated the optimum pH for the reactions catalyzed by the active enzymes. Steady-state kinetics of the active enzymes was performed as with IDH from Bacillus subtilis (BsIDH), revealing that LcIDH1 is a myo-inositol dehydrogenase and LcIDH2 is a scyllo-inositol dehydrogenase. Both LcIDH1 and LcIDH2 are observed to be NAD+-dependent. Kinetic isotopic effect experiments for LcIDH1 have demonstrated that the chemical step in the reaction is partly rate-limiting. Substrate spectrum of LcIDH1 and LcIDH2 was explored and compared to BsIDH. Finally, a multiple sequence alignment of IDH-related enzymes was performed and the proposed consensus sequence motifs were considered to understand the activity differences between these enzymes.