We have purified a novel human serine proteinase, designated as prostasin, from seminal fluid (Yu et al., 1994). In the present study, we have cloned and characterized the full-length cDNA encoding prostasin and identified its tissue-specific expression and cellular localization. A cDNA fragment was obtained by polymerase chain reaction using degenerate oligonucleotide primers derived from the NH2-terminal and internal amino acid sequences. A full-length cDNA sequence encoding prostasin was obtained by amplification of the 5'- and 3'-ends of the cDNA. It contains a 1,032-base coding region, a 572-base 3'-noncoding region and a 138-base 5'-noncoding sequence. Prostasin cDNA encodes a protein of 343 amino acids, which consists of a 32-amino acid signal peptide and a 311-amino acid proprostasin. Proprostasin is then cleaved between Arg12 and Ile13 to generate a 12-amino acid light chain and a 299-amino acid heavy chain, which are associated through a disulfide bond. The deduced amino acid sequence of the heavy chain has 34-42% identity to human acrosin, plasma kallikrein, and hepsin. A potential N-glycosylation site at Asn127 and the catalytic triad of His53, Asp102, and Ser206 have been identified. The deduced prostasin has a unique 19-amino acid hydrophobic portion at the COOH terminus, which makes it suitable to anchor in the cell membrane. Carboxyl-terminal sequencing of purified prostasin indicates that the hydrophobic portion is removed and that there is a cleavage between Arg290 and Pro291 during secretion. Southern blot analysis, following a reverse transcription polymerase chain reaction, indicates that prostasin mRNA is expressed in prostate, liver, salivary gland, kidney, lung, pancreas, colon, bronchus, renal proximal tubular cells, and prostate carcinoma LNCaP cells. Cellular localization of prostasin mRNA was identified within epithelial cells of the human prostate gland by in situ hybridization histochemistry.