Beta-cell Granule Protein Profile by Split Reporter Ass*

DESCRIPTION (provided by applicant): Beta cell replacement holds great promise as a potential therapeutic approach in the treatment of type 1 diabetes. Unfortunately, the supply of human organs is limited and the implantation of engineered B-cells may have to serve as an alternative to islets. Recent approaches focusing on the differentiation of cells from pancreatic ducts or protocols inducing transdifferentiation of liver progenitor cells have demonstrated that these stem/progenitor cells, under suitable conditions, make insulin. The evidence that the activation of insulin expression has also resulted in the formation of fully functional secretory granules, is less clear. This is not surprising, the traditional assumption that granule components are unique to granules has been largely replaced by observations that some granule components are indeed unique, whereas others are not, and only their relative concentrations may be unique. Thus, it becomes quite difficult to determine when the insulin storage in cells undergoing differentiation simulates the storage of actual b cells. In addition, much of the characterization of granules done in cell lines or rodent islets may not faithfully reflect the composition of human islets. Recently, Ozawa and colleagues developed a split reporter assay for the Rapid Identification of Novel Genes (RING) to characterize cDNAs for organelle-localized proteins. The RING strategy, is based on the idea that interactions of a 'split reporter tagged' bait protein specifically localized to an organelle and a complementary 'split reporter tagged' protein targeted to the same compartment will reconstitute a fully functioning reporter and thus identify the localization of the test protein. This approach has been used very successfully in screening eDNA libraries for proteins with mitochondrial targeting information. Whereas Ozawa and colleagues, used a construct which consisted of a C-terminal fragment of EGFP and DnaE fused in tandem to the precursor of subunit VIII of cytochrome c oxidase as split reporter-'bait', this application proposes to characterize the composition of secretory granules of human b cells by test reporter constructs with known localization to secretory granules. For the immediate future, we will draw from the abundant literature of protein/protein interactions in secretory granules and preliminary data from our lab and test the cDNAs of well characterized granule proteins for use as reporter molecules. Next, we will reconstitute the split reporters by cotransducing b cells with 'bait' and the complementary split reporter tagged test-cDNAs and examine localization to secretory granules by immunofluorescence. The proposed work will provides a means to describe the composition of human secretory granules, and in the long term, facilitate comparison of secretory granule composition in healthy and diseased states, partially differentiated cells induced to make insulin and fully differentiated human beta cells.