The stability of salt bridges at high temperatures: implications for hyperthermophilic proteins.

Salt bridges have been proposed to play a crucial role in promoting hyperthermostability in proteins, yet they appear to make little contribution to protein stability at room temperature. The latter point has been rationalized previously on the basis that the association of two charged molecules to form a salt bridge incurs a substantial desolvation penalty, which is seldom completely compensated by favourable interactions within the salt bridge and with the rest of the protein. Here a continuum solvation model is used to investigate how this same argument applies at temperatures more appropriate to hyperthermophiles. The solvation model employed was previously parameterised to reproduce the hydration free energies of neutral and charged amino acid side-chains in the temperature range from 5-100 degreesC. A key result of the previous work was that the hydration free energies of charged side-chains are more adversely affected by increasing temperature than are the hydration free energies of hydrophobic side-chains of identical size and shape (isosteres). As is shown here, a direct consequence of the temperature dependence of the hydration free energies is that at high temperatures the desolvation penalty for formation of a salt bridge is markedly reduced in magnitude. As a result, the argument that relative to hydrophobic isosteres, salt bridges destabilise proteins, may no longer be true at high temperatures. We demonstrate this point first in the setting of a small model system, but then also show that the same argument is likely to carry over to real proteins. We compare three hyperthermophilic proteins with their mesophilic homologues and find that hydration effects preferentially stabilise the hyperthermophiles at high temperatures. When the hydration effects are incorporated into a model for the free energy of folding of the proteins, it is found that in each case, the hyperthermophile is predicted to remain stable to a temperature 20-25 deg.C higher than the corresponding mesophile. Higher thermal stability for the hyperthermophile is obtained even if the mesophile is more stable at room temperature. The results obtained therefore suggest one possible way in which the apparently destabilising effects of salt bridges at room temperature can be reconciled with their increased abundance in hyperthermophilic proteins. Copyright 1998 Academic Press.