Glutathione reductase from human erythrocytes: amino-acid sequence of the structurally known FAD-binding domain

Glutathione reductase (Mr 2 x 52 500), a flavoenzyme of known three-dimensional structure, catalyses the reduction of glutathione disulfide by NADPH. This paper describes the primary structure of the FAD-binding domain which ranges from AcAla-1 to Gly-157. The three CNBr-produced fragments (69, 10 and 80 residues) of the domain were fractionated further by enzymatic and chemical methods; isolated peptides were sequenced mainly by automatic solid-phase Edman degradation. The tryptic peptides were overlapped by chymotryptic peptides. A fragment which results from cleavage at the acid-labile bond between Asp-135 and Pro-136 supplied peptides for overlapping the CNBr-produced fragments. In addition, many peptides were ordered and overlapped by computerized comparison with a complete sequence guessed from the electron density map. With one exception the computer method and the chemical alignment gave the same results. The sequence data are discussed in the light of the secondary and tertiary structure (Schulz et al. (1978) Nature (Lond.) 273, 120--124]. The 17 N-terminal residues are not visible in the electron density map. Consequently our numbering scheme differs from that of Schulz et al. by approximately 20 residues. Acetylation of the N terminus and an unusual composition of the following residues may serve to protect the loose N-terminal section of the protein against proteolysis in situ. The four cysteinyl residues of the FAD domain are of special interest. Cys-2 at the tip of the N-terminal extension is likely to be involved in the aggregation behaviour of glutathione reductase. Cys-58 and Cys-63 (formerly Cys-41 and Cys-46) represent the enzyme's redox-active dithiol. Cys-90 with its location at the twofold axis forms a disulfide bridge with Cys-90 of the other peptide chain of the enzyme. This might be related to the fact that both peptide chains contribute to each of the two active centers. In view of the interchain disulfide bridge glutathione reductase should be regarded as a monomeric protein. The sequence of the FAD-binding domain was compared with the sequence of the NADPH-binding domain of glutathione reductase using a computer program. As discussed, the scarcity of sequence similarities does not argue against the assumption that the two nucleotide-binding domains of glutathione reductase originated by gene duplication. The pyrophosphate moiety of FAD binds to a part of the polypeptide chain which in geometric structure, in topology and in sequence resembles the phosphate loops of other nucleotide-binding proteins and of flavodoxin. Using the phosphate loop as a reference, the N-terminal sequence of five flavoproteins can be aligned. The results of Williams et al. on the sequence of lipoamide dehydrogenase (EC 1.6.4.3) and our data on glutathione reductase (EC 1.6.4.2) show clearly that these two mechanistically similar enzymes possess homologous structures.