Mechanistic Studies of the Inhibition of Insulin Fibril Formation by Rosmarinic Acid

Abstract

The self-assembly of insulin to form amyloid fibrils has been widely studied because it is a significant problem in the medical management of diabetes and is an important model system for the investigation of amyloid formation and its inhibition. A few inhibitors of insulin fibrillation have been identified with potencies that could be higher. Knowledge of how these work at the molecular level is not known but important for the development of more potent inhibitors. Here we show that rosmarinic acid completely inhibits amyloid formation by dimeric insulin at pH 2 and 60 °C. In contrast to other polyphenols, rosmarinic acid is soluble in water and does not degrade at elevated temperatures, and thus we were able to decipher the mechanism of inhibition by a combination of solution-state H NMR spectroscopy and molecular docking. On the basis ofH chemical shift perturbations, intermolecular nuclear Overhauser effect enhancements between rosmarinic acid and specific residues of insulin, and slowed dynamics of rosmarinic acid in the presence of insulin, we show that rosmarinic acid binds to a pocket found on the surface of each insulin monomer. This results in the formation of a mixed tetramolecular aromatic network on the surface of insulin dimer, resulting in increased resistance of the amyloidogenic protein to thermal unfolding. This finding opens new avenues for the design of potent inhibitors of amyloid formation and provides strong experimental evidence for the role of surface aromatic clusters in increasing the thermal stability of proteins.

Mechanistic Studies of the Stabilization of Insulin Helical Structure by Coomassie Brilliant Blue

“…However, in the absence of zinc ion at physiological pH (~7.0) it is primarily present as a dimer and aggregates very easily at elevated temperatures. At pH ~ 2.0 and room temperature, HI forms dimmers (due to charge repulsion between Zn 2+ and protonated His B10 ) 51 , as confirmed by solution-state NMR 43,47,52 and X-ray crystallography. 9 Dimeric structure of human insulin in sodium phosphate buffer (pH 2.0) was indicated by a series of intermolecular NOEs between the two monomers of the dimer (Supporting Figure S6).…”

“…These NOEs corroborated well with the previous study. 47,52 Apart from these, no other intermolecular NOEs between HI and CBBG were observed.…”

“…56 Previous studies have also indicated that the amyloid fibril formation for globular proteins, such as HI, occurs via partially unfolded intermediates that gradually associate to form the oligomers, eventually into well-ordered mature fibrils. 9,14,23,27,33,47,57,58 A partial unfolding of the monomeric globular protein produces a molten globule like structure that is comparable to the amyloidogenic protein intermediates, acting as the nucleating form. 29,50,55,[59][60][61] In the absence of zinc at physiological pH, HI is primarily present as a conformationally stable dimer;…”

“…46 The recorded spectrum was similar to previously published NOESY data recorded in the similar experimental conditions. 47 Surprisingly, after 24 h of incubation of HI sample at 60 o C, all the NOE cross-peaks were completely broadened ( Figure 6C), suggesting the formation of a larger assembly, which were beyond the detection in NMR time scale due to slower tumbling and rapid T 2 relaxation. 21,[47][48][49][50] HI sample incubated with equimolar concentrations of CBBG at room temperature also showed NOESY spectra with numerous NOE peaks ( Figure 6B).…”

“…47 Surprisingly, after 24 h of incubation of HI sample at 60 o C, all the NOE cross-peaks were completely broadened ( Figure 6C), suggesting the formation of a larger assembly, which were beyond the detection in NMR time scale due to slower tumbling and rapid T 2 relaxation. 21,[47][48][49][50] HI sample incubated with equimolar concentrations of CBBG at room temperature also showed NOESY spectra with numerous NOE peaks ( Figure 6B). However, incubating the solution for 24 h at high temperature (60 o C) the NOESY spectrum ( Figure 6D) was very similar to the spectra ( Figure 6B) obtained before heating the solution.…”