SummaryThe emergence of resistance to vancomycin and related glycopeptide antibiotics is spurring efforts to develop new antimicrobial therapeutics. High resolution structural information about antibioticligand recognition should prove valuable in the rational design of improved drugs. We have determined the X-ray crystal structure of the complex of vancomycin with N-acetyl-D-Ala-D-Ala, a mimic of the natural muramyl peptide target, and refined this structure at a resolution of 1.3 Å to R and R free values of 0.172 and 0.195, respectively. The crystal asymmetric unit contains three backback vancomycin dimers; two of these dimers participate in ligand-mediated face-face interactions that produce an infinite chain of molecules running throughout the crystal. The third dimer packs against the side of a face-face interface in a tight "side-side" interaction that involves both polar contacts and burial of hydrophobic surface. The trimer of dimers found in the asymmetric unit is essentially identical to complexes seen in three other crystal structures of glycopeptide antibiotics complexed with peptide ligands. These four structures are derived from crystals belonging to different space groups, suggesting that the trimer of dimers may not be simply a crystal packing artifact, and prompting us to ask if ligand-mediated oligomerization could be observed in solution. Using size exclusion chromatography, dynamic light scattering, and small-angle X-ray scattering, we demonstrate that vancomycin forms discrete supramolecular complexes in the presence of tripeptide ligands. Size estimates for these complexes are consistent with assemblies containing 4-6 vancomycin monomers.

[formula: see text] Complexes formed between fluorobenzene and N-methylformamide or benzene have been used as models of the interaction of fluoroaromatic drugs with carbonic anhydrase II. These structures have been investigated via ab initio and density functional methods, including HF, B3LYP, and MP2 procedures. The results of the calculations are consistent with the hypothesis, suggested originally by experimental X-ray crystal structures of the drug-receptor complexes, that favorable fluorine-hydrogen interactions affect binding affinity.

Ab initio (HF) and density functional theory (DFT) calculations of 19 F NMR chemical shifts were performed for models of fluoroaromatic inhibitors of carbonic anhydrase II (CA). DFT gave slightly better agreement with the experimentally measured chemical shifts of the actual inhibitors, suggesting that intramolecular dispersion does contribute significantly to the chemical shifts in these molecules. HF and DFT calculations for the stacked complex of hexafluorobenzene with benzene gave excellent agreement with experimental 19 F chemical shifts in this system. The fact that both approaches to this calculation were successful suggests that intermolecular dispersion is not an important contributor to 19 F chemical shifts in this system. Electron transfer and electrostatics must, therefore, be responsible for the changes in the 19 F NMR spectra observed on complexation. Finally, an unsuccessful attempt was made to apply HF and DFT methods to the calculation of the 19 F chemical shift of a pentafluorobenzyl-derived CA inhibitor bound to the protein in close proximity to a phenylalanine residue. A model of the inhibitor's aromatic ring interacting with the protein's aromatic residue gave a calculated chemical shift change that was much greater than that observed experimentally. Effects on the chemical shift from the field due to atoms omitted from the calculation, as well as from extensive rovibrational freedom, cannot easily be addressed in calculations of these large systems and are the likely reasons for the failure of these calculations.