Biophysical characterization of protein folding and misfolding.

The HPr proteins were characterized as folding by a two-state folding mechanism. Here, we present a comparison of the equilibrium and kinetic folding for the HPr protein from Bacillus subtilis, E coli and a key variant from these proteins. For the wild-type protein we find that GHX is greater than GUDC, suggesting that the HPr does not fold by a simple two-state mechanism. This discrepancy is revealed by testing the two-state nature of the folding reaction of HPr with mutation. We show that removing a single charge side chain (Asp 69) converts the HPr protein back to a simple two-state mechanism. Ribonuclease Sa and two charge-reversal variants can be converted into amyloidin vitro by the addition of 2,2,2-triflouroethanol (TFE). We report here amyloid fibril formation for these proteins as a function of pH. The pH at maximal fibril formation correlates with the pH dependence of protein solubility, but not with stability, for these variants. Additionally, we show that the pH at maximal fibril formation for a number of ivwell-characterized proteins is near the pI, where the protein is expected to be the least soluble. This suggests that protein solubility is an important determinant of fibril formation.