Human urinary kallikrein Complete amino acid sequence and sites of glycosylation

Abstract

Human glandular kallikrein was purified from urine and subjected to detailed structural characterization. The protein was carboxymethylated with iodoacetic acid and digested with TPCK-trypsin, Staphylococcal aureus V-8 protease and endo LysC peptidase. The resulting peptide fragments were separated by reverse-phase HPLC using C-4 columns and acetonitrile-trifluoroacetic acid gradient elution. The complete amino acid sequence of the carboxymethylated derivative was elucidated by sequence analysis and alignment of peptides derived from different proteolytic cleavages. A procedure using in situ CNBr cleavage of a large endo LysC peptidase-derived peptide followed by direct sequencing was carried out to provide overlap for two glycosylation sites at residues 78 and 84. Three Asn-linked glycosylation sites were confirmed by the direct sequence analysis of the isolated glycopeptides. However, the third glycosylation at Asn-144 occurs only in 60% of kallikrein molecules. Reverse-phase HPLC effectively separates two species of HUK which correspond to molecules glycosylated and non-glycosylated at Asn-144, respectively. The human urinary kallikrein contains 238 amino acid residues with Ile and Ser as N- and C-terminal amino acids, respectively. The primary structure is completely identical to that deduced from a human genomic DNA sequence (F.K. Lin et al., manuscript in preparation) and is different in one amino acid (Lys-162 vs. Glu-162) from that deduced from pancreatic or kidney cDNA sequence.