Evidence showing that the two-chain form of vitronectin is produced in the liver by a selective furin cleavage.

The adhesive protein vitronectin (75 kDa) occurs in human blood fluid in a one-chain (Vn75) or a two-chain form (Vn65-10), and is produced by a specific cleavage (at Arg379-Ala380), by a proteinase not identified hitherto. These two forms were shown to be functionally different and therefore, this cleavage may have a regulatory significance in vivo. Here, we report the use of a tailored one-chain recombinant Vn, a specific protein kinase A phosphorylation at Ser378, and sequence analysis to show: (1) that none of the proteinases originating from blood, previously thought to be the endogenous proteinase (plasmin, thrombin, tPA, and uPA), is indeed the in vivo convertase; and (2) that furin, a serine endoproteinase residing in the secretory pathway of hepatocytes, where Vn is synthesized, specifically cleaves Vn at the endogenous cleavage site. Consequently, we propose that the Vn75 to Vn65-10 conversion takes place in the liver (not in blood) and is carried out by furin.