Endosomal proteolysis of insulin by an acidic thiol metalloprotease unrelated to insulin degrading enzyme.

Although insulin is degraded as a consequence of receptor-mediated endocytosis, the location and nature of the responsible proteinase(s) remain controversial. Insulin degrading enzyme (IDE; EC 3.4.22.11), a mainly cytosolic neutral thiol metalloendopeptidase of 110 kDa, has been proposed to be the main cellular clearance mechanism. However, endosomes concentrate and degrade internalized insulin demonstrating that IDE is unlikely to be the relevant enzyme for endosomal proteolysis of internalized insulin in liver parenchyma. In purified endosomal fractions insulin was actively degraded at acid pH and IDE was undetectable as evaluated by immunoblotting, immunoprecipitation, or chemical cross-linking procedures. Affinity purified endosomal acidic insulinase displayed a pH optimum of 4-5.5, a lack of inhibition by EDTA and N-ethylmaleimide, and a partial metal-ion requirement (for Mn2+) all of which distinguished it from IDE. A small but detectable presence of IDE in particulate nuclear (N) and large granule (ML) fractions was observed by differential centrifugation. By analytical centrifugation, IDE cosedimented with the organelle containing the peroxisomal marker proteins catalase and thiolase (median density, 1.21 g.cm-3). By preparative centrifugation, highly purified peroxisomes were observed to be enriched in IDE. Since all cloned cDNAs of IDE (human, rat, and Drosophila) reveal a deduced classical peroxisomal targeting sequence A/SKL at their carboxyl termini this may account for the peroxisomal location of IDE. Taken together, our studies identify an insulin-degrading enzyme in endosomes which is distinct from IDE. The latter's presence in peroxisomes suggests that its physiological substrate(s) in vivo are polypeptides other than insulin.