In silico identification and analysis of the binding site for aminocoumarin type inhibitors in the C-terminal domain of Hsp90.

Hsp90 contains two Nucleotide Binding Sites (NBS): one each in its N-terminal domain (NTD) and C-terminal domain (CTD), respectively. Previously we used computational techniques to locate a nucleotide-binding site in the CTD. Nucleotide binding at this site stabilized the structurally labile region within this domain, thus providing a rationale for increased resistance to thermal denaturation and proteolysis. A scan for ligand-binding sites in CTD revealed four potential sites with the requisite volume to accommodate aminocoumarins and -derived inhibitors. Only one of these reproducibly formed docked complexes with inhibitors and showed excellent interactions with residues lining the site. Fortuitously, it was identical to the aforementioned nucleotide-binding site thus providing an explanation for the reported direct competition between inhibitors and nucleotides. Further studies with carefully chosen inhibitors and some inactive analogues provided an explanation for the known Structure-Activity Relationships (SAR) of aminocoumarin and -derived inhibitors. We also performed similar studies of the NTD to discern the reason(s) for its inability to bind aminocoumarins, given the family resemblance to prokaryotic Top-IV and Gyr-B. Our studies permitted the identification of the putative inhibitor binding site in the CTD, an explanation for increased resistance to thermal denaturation and proteolysis upon inhibitor binding as well as direct competition with ATP.