Structure and dynamics of the drug-bound bacterial transporter EmrE in lipid bilayers.

The dimeric transporter, EmrE, effluxes polyaromatic cationic drugs in a proton-coupled manner to confer multidrug resistance in bacteria. Although the protein is known to adopt an antiparallel asymmetric topology, its high-resolution drug-bound structure is so far unknown, limiting our understanding of the molecular basis of promiscuous transport. Here we report an experimental structure of drug-bound EmrE in phospholipid bilayers, determined using 19F and 1H solid-state NMR and a fluorinated substrate, tetra(4-fluorophenyl) phosphonium (F4-TPP+). The drug-binding site, constrained by 214 protein-substrate distances, is dominated by aromatic residues such as W63 and Y60, but is sufficiently spacious for the tetrahedral drug to reorient at physiological temperature. F4-TPP+ lies closer to the proton-binding residue E14 in subunit A than in subunit B, explaining the asymmetric protonation of the protein. The structure gives insight into the molecular mechanism of multidrug recognition by EmrE and establishes the basis for future design of substrate inhibitors to combat antibiotic resistance.

Modeling the antiferromagnetic MnIIMnII system within the protein phosphatase-5 catalytic site.

Protein phosphatase-5 (PP5), a novel target for inhibition in a search for new antitumor drugs, contains a homobimetallic Mn(II)Mn(II) system in its catalytic site. The ground electronic state is an antiferromagnetically-coupled singlet. We report optimizations of a known inhibitor within a 42-residue model of the PP5 catalytic site under several two-level hybrid ONIOM computational models. Using the high-resolution crystal structure of a PP5/inhibitor complex as reference, we compare geometric parameters as the qualities of the "high-level" and "low-level" wavefunctions are successively improved by using the correct antiferromagnetic (AF) singlet state. We find that the UB3LYP AF wavefunction for the high-level region is necessary for experimental fidelity. A closed-shell semi-empirical method (RPM6) can be used for the low-quality part of the hybrid scheme to afford geometries which are qualitatively on par with that obtained using the more time-consuming open-shell UB3LYP AF wavefunction. As the AF state can be elusive for such a large system, the ferromagnetic (F) state can also be used in the low-quality calculations without impacting the geometry.