We recently reported that normal pancreatic islets or beta cell lines contain a subset of proinsulin molecules that have not achieved the three intramolecular disulfide bonds required for native insulin structure. Indeed, we found that non-native proinsulin monomers can form intermolecular disulfide-linked complexes, which are detectable by immunoblotting after nonreducing SDS-PAGE. However, we have recognized difficulties in the quantitative interpretation of these analyses, as we frequently identified conditions in which little or no proinsulin monomer was detected, and non-native disulfide-linked complexes amounted to greater cumulative proinsulin band intensity than the entire population of proinsulin molecules detected by immunoblotting upon reducing SDS-PAGE. Here we have developed several critical modifications of our methodology to rectify these discrepancies to undertake a more accurate determination of the pool of misfolded proinsulin molecules in pancreatic beta cells under conditions of health and disease. Using these modifications, we find that the size of the proinsulin monomer pool has previously been underestimated and the fraction of disulfide-linked complexes overestimated — and this can be readily explained by proinsulin antibody affinity in the immunoblotting protocol. Additionally by immunoblotting, we now resolve native and non-native proinsulin monomers, highlighting the inefficiency of non-native proinsulin transport through the secretory pathway.
A.Arunagiri: None. M.Alam: None. L.Haataja: None. P.J.Samy: None. S.Soleimanpour: None. L.S.Satin: None. P.Itkin-ansari: None. M.Liu: None. P.Arvan: n/a.
National Institutes of Health (R01-DK48280) , National Institutes of Health (U01-DK127747) , and National Institutes of Health (R24 DK110973)