The first primary structure for a sorbitol dehydrogenase has been determined by analysis of the tetrameric enzyme from sheep liver. The ['4C]carboxymethylated protein was cleaved with CNBr and proteolytic enzymes. Peptides were purified by several methods, often utilizing exclusion chromatography for pre-fractionation and reverse-phase high-performance liquid chromatography for final purification. Different methods of sequence analysis complemented each other, mainly the manual dimethylaminoazobenzene isothiocyanate method and the use of liquidphase sequencer degradations.All eight major CNBr fragments were purified and form the basis of the work. Three minor CNBr fragments derived from an acid cleavage and from a partly resistant Met-Thr bond were also obtained, as well as evidence for a contaminating homologous polypeptide. Most of the tryptic peptides were purified, including all with methionine residues, thus overlapping the CNBr fragments. Combined, all data permit the deduction of a 354-residue amino acid sequence for the polypeptide chain of sorbitol dehydrogenase. The N terminus is acyl-blocked, the C terminus is formed by a proline residue, tryptophan is the least common residue (two, at positions 50 and 301) and there are 10 cysteine residues, including the residue previously shown to be especially reactive (at postion 43). Similarities to 'long' alcohol dehydrogenases have functional implications.Liver sorbitol dehydrogenase was recently found to be similar to zinc-containing alcohol dehydrogenases [I], constituting an apparent structural intermediate [2] between the distantly related alcohol dehydrogenases from yeast and horse liver [3,4]. The results suggested wide and parallel evolutionary relationships between groups of dehydrogenases [5] and emphasized relationships in a metabolic pathway from glucose [6]. The complete primary structure of a sorbitol dehydrogenase is necessary for further correlations.In the present work, the amino acid sequence of the entire protein chain of sheep liver sorbitol dehydrogenase was determined. The work was centered on purification and analysis of all CNBr fragments, which were overlapped by identification of the methionine-containing tryptic peptides. Throughout, [14C]carboxymethylation of cysteine residues was used to obtain suitable markers in several peptides of all digests. A 354-residue primary structure for the protein chain was deduced. Evidence was obtained for the occurrence of a minor contaminating component which appeared distantly homologous to sorbitol dehydrogenase in one region. Otherwise the preparations are homogeneous, suggesting all subunits in the tetramer to be identical. Results to support the amino acid sequence are given in this study, together with aspects of methodological interest. Further correlations in a protein family of zinc-containing alcohol/polyol dehydrogenases are given in an accompanying paper [7].
