Modeling the interface between islet amyloid polypeptide and insulin-based aggregation inhibitors: correlation to aggregation kinetics and membrane damage.

Human islet amyloid polypeptide (hIAPP) forms cytotoxic fibrils in type-2 diabetes and insulin is known to inhibit formation of these aggregates. In this study, a series of insulin-based inhibitors were synthesized and assessed for their ability to slow aggregation and impact hIAPP-induced membrane damage. Computational studies were employed to examine the underlying mechanism of inhibition. Overall, all compounds were able to slow aggregation at sufficiently high concentrations (10× molar excess); however, only two peptides showed any inhibitory capability at the 1:1 molar ratio (EALYLV and VEALYLV). The results of density functional calculations suggest this is due to the strength of a salt bridge formed with the Arg11 side chain of hIAPP and the inhibitors' ability to span from the Arg11 to past the Phe15 residue of hIAPP, blocking one of the principal amyloidogenic regions of the molecule. Unexpectedly, slowing fibrillogenesis actually increased damage to lipid membranes, suggesting that the aggregation process itself, rather than the fibrilized peptide, may be the cause of cytotoxicity in vivo.

Correlation of LUMO localization with the alpha-amylase inhibition constant in a Tendamistat-based series of linear and cyclic peptides.

The glycosidase alpha-amylase is responsible for the hydrolysis of alpha(1-->4) glycosidic linkages found in dietary starch as one means for controlling blood sugar level. The effect of alpha-amylase is detrimental, however, in the disease state diabetes mellitus, where blood glucose levels are elevated due to a biochemical defect. Inhibition of the enzyme's activity would reduce glucose absorption by the small intestine. Our objective was to develop small peptides based on essential binding elements of the natural protein inhibitor, Tendamistat. These smaller analogs may be better studied structurally and conformationally to help us understand molecular-level interactions. In addition, we have been able to correlate the activity of our compounds with the lowest unoccupied molecular orbital (LUMO) localization in energy-minimized conformations. The positive charge/LUMO of most active inhibitors is localized on the central Arg residue of the required triplet. This provides a predictive model for the design of active molecules.