Variability within mammalian sorbitol dehydrogenases. The primary structure of the human liver enzyme

The primary structure of sorbitol dehydrogenase from human liver has been determined by peptide analysis in order to relate the variability of this enzyme to that of the others within the alcohol dehydrogenase family. The structure obtained reveals 355 residues with an acyl-blocked N-terminus and an unexpected microheterogeneity at position 237 (Gln/Leu). The residue identity between sheep and human liver sorbitol dehydrogenase is 89%. This variability is similar to that of class I alcohol dehydrogenases, but distinctly different from that of class III alcohol dehydrogenases, the structures of which are much more conserved. Consequently, class III alcohol dehydrogenase is thus far unique within this family of dehydrogenases, suggesting a particularly strict requirement for that structure. The variability within sorbitol dehydrogenase involves all segments of the molecule but is largely at surface positions and clusters in one such region, covering positions 214-240, corresponding to a segment of the coenzyme-binding domain. Ligands to the active-site zinc and most residues lining the coenzyme-binding and substrate-binding pockets are conserved. However, provided conformational models are reliable, a charge difference may affect the interactions at the inner part of the substrate pocket, another charge difference may affect the interdomain region, and a size difference the adenine pocket. The primary structure of human liver sorbitol dehydrogenase further shows that the absence of three of the four ligands to a second zinc atom present in alcohol dehydrogenases is a general property of sorbitol dehydrogenase.