Furin is a mammalian propeptide-processing endoprotease in nonendocrine cells and has been demonstrated to be present in virtually all nonendocrine cells, including fibroblasts, epithelial cells, and hepatocytes. Furin cleaves the concensus processing site -Arg-4-X-3-Lys/Arg-2-Arg-1 decreases X+1-. Some subunit-containing precursor proteins, including an insulin receptor precursor, possess an additional basic residue at position -3, thus forming a tetrabasic processing site. This implies that a tetrabasic processing site must be easily cleavable in nonendocrine cells. We created a mutant proinsulin DNA with a peptide structure comprised of B- and A-chains linked to the C-peptide by a pair of tetrabasic residues, in the following order: B-chain-Arg-Arg-Lys-Arg-C peptide-Arg-Arg-Lys-Arg-A-chain. The native proinsulin structure was B-chain-Arg-Arg-C-peptide-Lys-Arg-A-chain. Both the native and mutant proinsulins were expressed in the following four cell lines: a monkey kidney-derived cell line (COS-7), a Chinese hamster ovary-derived cell line (CHO), a human liver cancer-derived cell line (HepG2), and a mouse fibroblast-like cell line (NIH3T3). We used these cell lines because they contain different quantities of furin mRNA, ranking as follows: NIH3T3 > HepG2 > COS > CHO. When mutant insulin was expressed in these cells, the conversion of proinsulin to mature insulin was approximately 85% in NIH3T3, 70% in HepG2, 60% in COS, and 50% in CHO. The conversion correlated well with the furin expression in each cell line as measured by the density of its Northern blot band. Moreover, in CHO, the cell line with the lowest furin expression, coexpression of mutant proinsulin with furin resulted in complete conversion of proinsulin to mature insulin.