Abstract
Insulin-binding B cells are implicated in Type 1 Diabetes (T1D) pathology. Antigen-specific immunotherapy (ASIT) holds promise in T1D. However, ASIT-targeted suppression of insulin-binding B cells is hampered by insulin's hormonal activity and the resulting binding and endocytosis of insulin by insulin receptors (INSR). To evaluate ASIT strategies that target insulin-binding B cells in vivo, non-hormonally active insulin variants are needed. In this work, we aimed to improve upon prior non-hormonal insulin variants by making mutations to the insulin precursor, proinsulin, and including a c-terminal sortase (SrtA) tag (LPETGGHG) to enable facile site-selective bioconjugation to scaffolds or payloads. Of the insulin variants investigated that retained low-nM binding to the murine-derived insulin autoantibody mAb 125, proinsulin(F25D)-SrtA had the lowest INSR binding and activity and the greatest fibrillation resistance. Compared to desoctapeptide insulin, a previously proposed non-hormonal insulin variant, proinsulin(F25D)-SrtA demonstrated 50-fold lower INSR binding and 100-fold greater fibrillation lag time. However, insulin(F25D)-SrtA bound to the anti-insulin antibody 12M4 isolated from a presymptomatic T1D individual, whereas proinsulin(F25D)-SrtA and desoctapeptide insulin did not, highlighting the potential for anti-insulin B cells to develop in human T1D that would escape this ASIT moiety. The characteristics of proinsulin(F25D)-SrtA make it a well-suited non-hormonal insulin variant for insulin-binding B cell targeting and warrants additional study with other anti-insulin B cell specificities derived from T1D individuals.