The ostensible paradox of multidrug recognition.

The ability of multidrug-efflux transporters to recognize scores of dissimilar organic compounds has always been considered paradoxical because of its apparent contradiction to some of the basic dogmas of biochemistry. In order to understand, at least in principle, how a protein can recognize multiple compounds, we analysed the transcriptional regulator of the Bacillus subtilis multidrug transporter Bmr. This regulator, BmrR, binds multiple dissimilar hydrophobic cations and, by activating expression of the Bmr transporter, causes their expulsion from the cell. Crystallographic analysis of the complexes of the inducer-binding domain of BmrR with some of its inducers revealed that ligands penetrate the hydrophobic core of the protein, where they form multiple van der Waals and stacking interactions with hydrophobic amino acids and an electrostatic bond with the buried glutamate. Mutational analysis of the binding site suggests that each ligand forms a unique set of atomic contacts with the protein: each tested mutation exerted disparate effects on the binding of different ligands. The example of BmrR demonstrates that a protein can bind multiple hydrophobic compounds with micromolar affinities by using only electrostatic and hydrophobic interactions. Its ligand specificity can be further broadened by the flexibility of the binding site. It appears, therefore, that the commonly expressed fascination with the relaxed substrate specificity of multidrug transporters is misdirected and originates from an almost exclusive familiarity with the more sophisticated processes of specific molecular recognition that predominate among proteins analyzed to date.