The complete amino acid sequence of human serum transferrin

The complete amino acid sequence of human serum transferrin has been determined by aligning the structures of the 10 CNBr fragments. The order of these fragments in the polypeptide chain is deduced from the structures of peptides overlapping methionine residues and other evidence. Human transferrin contains 678 amino acid residues and--including the two asparagine-linked glycans--has an overall molecular weight of 79,550. The polypeptide chain contains two homologous domains consisting of residues 1-336 and 337-678, in which 40% of the residues are identical when aligned by inserting gaps at appropriate positions. Disulfide bond arrangements indicate that there are seven residues between the last half-cystine in the first domain and the first half-cystine in the second domain and therefore, a maximum of seven residues in the region of polypeptide between the two domains. Transferrin--which contains two Fe-binding sites--has clearly evolved by the contiguous duplication of the structural gene for an ancestral protein that had a single Fe-binding site and contained approximately 340 amino acid residues. The two domains show some interesting differences including the presence of both N-linked glycan moieties in the COOH-terminal domain at positions 413 and 610 and the presence of more disulfide bonds in the COOH-terminal domain (11 compared to 8). The locations of residues that may function in Fe-binding are discussed.