Book of Abstracts: Albany 2007
June 19-23 2007
Protein Thermostability Probed by Statistical Potentials: Salt Bridges Versus Hydrophobic Interactions
The temperature dependence of the interactions that stabilize protein structures is a longstanding issue, whose elucidation would make possible to predict and rationally modify the thermostability of a target protein. This problem is tackled here by deriving distance-dependent amino acid pair potentials from datasets of monomeric proteins whose melting temperatures Tm are known. Four datasets are defined, characterized by different Tm ranges. Note that we use here an unusual definition of thermophilicity, based on the melting temperature of the proteins rather than on the living temperature of the host organisms. The temperature dependence of pair interactions was inferred from the differences in shape of the potentials derived from the four datasets. We focused on hydrophobic and salt bridge interactions, represented by contacts between I, L or V amino acids, and between E or D and K or R, respectively. In analyzing the R-involving salt bridges, two different geometries had to be distinguished, the fork-fork geometry where the side chains point towards each other (right of Fig.a) and the fork-stick geometry involving the Nε side chain atom of R (left of Fig.a). Our results show that the stabilizing weight of hydrophobic interactions remains constant as the temperature increases, compared to the other interactions (Fig.a-b). In contrast, salt bridges appear relatively more stabilizing at high temperature, with the possible exception of fork-fork geometries. A preference for more compact salt bridges is moreover noticeable in heat-resistant proteins, whereas the hydrophobic packing remains constant. The Tm -dependent potentials developed should be valuable for predicting thermal stability changes upon mutation.
Distance mean force potentials (ΔWds) derived from a dataset of low Tm (blue line) or high Tm (red dotted line) proteins, as a function of the spatial distance (r) between geometric centers of the residue side chains. (a) Glu-Arg salt bridge and (b) Ile-Leu hydrophobic interaction.
Unite de Bioinformatique