Direct hydrogen bonds between sidechains are obviously important to structural stability in proteins. From time to time I see cases of water-mediated bonds in which a single water molecule seems to play an important role (sometimes taking the place of a missing ligand atom in an apo structure, for example). But what about larger chains and networks of water? Assuming a structure is high enough in resolution and well-ordered enough to observe such things, has anyone systematically studied the structural importance of multiple water interactions (I do know of a paper by Faerman and Karplus back in 94, but perhaps there is more recent work).
Actin gives an example of such. It has two large "domains" with a metal-nucleotide binding cleft between them. Without the metal-nucleotide the protein is unstable and slowly and irreversibly denatures.
When a water network coming out from the Me-nucleotide is analysed (I did it manually), the reason becomes very clear - secondary and tertiary hydration shells provide a web of interactions that ultimately "join" the two large domains, thus, presumably, providing the stability.
Dima
