In recent work, we have also begun to use X-ray scattering experiments and molecular dynamics calculations to obtain new information about the role of hydration in protein folding. Our goal is to determine whether "long range" hydration forces exert a significant effect during protein folding, and whether experimental estimates of these forces are able to improve computer simulations of protein folding.

We are currently involved in the collection and analysis of electron diffraction and X-ray diffraction data which are used to compare the structure of bacteriorhodopsin (bR) in its "light-adapted" resting state to that of different intermediates in the bR photocycle. This protein is thought to pump protons across the cell membrane, using energy that is absorbed by its retinal chromophore to establish an electrochemical potential across the cell membrane.

Additional work is in progress with various other membrane proteins, in order to form crystals that are suitable for electron diffraction or X-ray diffraction.

A well known but nevertheless striking feature of protein folding is the fact that folding occurs much too rapidly to be accounted for as a random search through all allowed conformations of the peptide chain. In addition, it is significant to note that energy minimization calculations, using standard potential-energy functions, are plagued by the problem of getting stuck in false minima, i.e. structures different from the native fold. Our understanding of folding is evidently still quite incorrect; folding in the real-life random search does not get stuck in false (i.e., local) minima, in the way that computer searches do. One possible explanation is that "soft" but long-range interactions between the hydration shells of different amino acid sidechains bias the range of local peptide conformations for a given amino acid sequence, so that the number of accessible conformations is drastically reduced, and the peptide chain is steered away from false minima. Long-range (10A - 15A), hydration-mediated interactions between "macroscopic"hydrophobic surfaces and between "macroscopic" hydrophilic surfaces are, in fact, experimentally well characterized. Our X-ray scattering experiments are designed to characterize similar effects at the microscopic level of individual sidechain-to-sidechain interactions. The long-term goal is to then incorporate empirical estimates of "hydration-based forces" in molecular dynamics simulations of the early stages of protein folding.