In non-insulin-dependent diabetes, circulating insulin-related immunoreactivity (IRI) is often composed of a higher fraction of the incompletely converted forms proinsulin and des-31,32 proinsulin. The present study describes an immunoadsorption method for measuring the proportions of proinsulin, its two split products, and insulin in human pancreatic tissue and for determining their rates of formation in human isolated islets. The method uses two junction-specific monoclonal proinsulin antibodies in a protein G fractionation; it is validated by > or = 90% specificity and recovery. The peptide contents measured in tissue extracts were comparable to those determined in a previously developed immunoradiometric assay. In the nine tissue extracts from nondiabetic donor organs, 97% of IRI corresponded to insulin, 1% to proinsulin, 2% to the des-31,32 proinsulin conversion product, and 0.1% to des-64,65 proinsulin. Two samples from non-insulin-dependent diabetics under sulfonylurea treatment contained a fourfold lower content of IRI but the peptide distribution was comparable except for a low percentage (0.3) of proinsulin in one case. In pulse-chase experiments on three-preparations of human islets isolated from nondiabetic donors, proinsulin represented the major (> 90%) IRI that was synthesized at the end of the 30-min pulse; a subsequent 90-min chase at either 2.5 or 10 mM glucose resulted in conversion of 75% of proinsulin to des-31,32 (20%) and des-64,65 (2%) intermediates and to insulin (50%); after a 180-min chase, 88% of proinsulin was converted to insulin, but 10% remained present as proinsulin. In a pulse-chase experiment on islets isolated from tissue with a high proportion of des-31,32 intermediate (5% instead of 2%), the conversion process was slower (45% after 90 min and 70% after 180 min) and resulted in a higher fraction of des-31,32 intermediate, suggesting that the elevated tissue content in this intermediate is caused by a reduced PC2 converting activity. These data confirm that des-31,32 proinsulin represents the major conversion intermediate in normal human islets and indicate the existence of slow converters, possibly as a result of decreased enzymatic processing of the prohormone's AC junction.