Identification of a 31,500 molecular weight islet cell protease as cathepsin B.

A method for the preparation of a radioisotopically labeled active-site directed reagent for proteases (125I-Tyr-Ala-Lys-ArgCH2Cl) is described, and an example of its use as a sensitive method for identifying trypsin-like proteases is provided. This high specific activity reagent was then used in an attempt to identify proteases in rat islets of Langerhans involved in the conversion of proinsulin to insulin. Previous studies have indicated that the endoprotease involved in proinsulin conversion is a cysteine proteinase and that 125I-Tyr-Ala-Lys-ArgCH2Cl affinity labels an islet crude granule fraction protein having a molecular weight of 31,500. Here we demonstrate, using a probe of higher specific activity, that the major affinity-labeled proteins of the islet crude granule fraction, when displayed by sodium dodecyl sulfate gel electrophoresis, have molecular weights of approximately 39,000 (5%), 31,500 (53%), and 5,000-6,000 (37%), with several other minor proteins (less than 5%) also labeled. The two predominant labeled proteins were mainly soluble rather than membrane bound, and they exhibited patterns of competition with various inhibitors that were similar to the pattern shown by the conversion of proinsulin to insulin in vitro. A rabbit antibody to rat liver cathepsin B immunoprecipitated both affinity-labeled 31,500 and 5,000-6,000 molecular weight proteins, and on the basis of this and structural considerations the 31,500 molecular weight cysteine protease is identified as cathepsin B. The 5,000-6,000 molecular weight peptide is an NH2-terminal, active site cysteine-containing, proteolytic fragment of the 31,500 molecular weight protein. Because cathepsin B is not per se a candidate for the proinsulin convertase because of its excessively broad substrate specificity, these studies suggest that a similar enzyme or a modified form of this enzyme is active within the secretory progranules, whereas the more typical cathepsin B may be largely confined to lysosomal contaminants in our granule preparations.