Advances in computer-aided drug design with applications to glycopeptide antibiotics | Posted on:2010-03-24 | Degree:Ph.D | Type:Thesis | University:Yale University | Candidate:Leung, Siegfried Sik Fai | Full Text:PDF | GTID:2444390002981666 | Subject:Chemistry | Abstract/Summary: | PDF Full Text Request | This dissertation delineates the advancement in computer-aided drug design with applications to glycopeptide antibiotics. The glycopeptides of interest, vancomycin and teicoplanin, are antibiotics of last resort for treatments against multi-drug resistant infections. As an effort to combat the uprising epidemic of vancomycin-resistant infections, computational design of second generation derivatives of these two antibiotics were undertaken.The mechanisms of action of vancomycin and teicoplanin involve interrupting the cell wall biosynthesis of gram-positive bacteria by binding to peptidoglycan cell wall precursors at their D-Ala-D-Ala dipeptide termini. Common vancomycin-resistant bacterial strains, VanA and VanB, impair the activities of vancomycin-type antibiotics up to 1000 fold by reprogramming the terminal dipeptide to the depsipeptide, D-Ala-D -Lac, through an inducible mechanism. The focus of this dissertation is to seek modifications that allow each antibiotic to bind efficiently to both D-Ala-D-Ala and D-Ala- D-Lac sequences. Binding affinities for tripeptide mimics containing these sequences were computed as relative binding free energies using Monte Carlo/free energy perturbation (MC/FEP) methodology.For the study of vancomycin, the accuracy of the MC/FEP approach was first confirmed by reproducing the experimental trend in binding affinities of vancomycin aglycon with different tripeptides. The study of potential backbone modification between residues 4 and 5 indicates that the original peptide and the previously prepared amine linkages are the most promising. The findings of the subsequent investigation into side-chain modifications suggest that binding affinities of both sequences may be enhanced by hydrophobic substitutions at residue 7 that improve contacts with the ligands' D-Ala methyl side chains. Applying both backbone and side-chain modifications is also predicted to be benign.For teicoplanin, its aglycon form is shown to have similar binding modes for the dipeptide sequences as vancomycin aglycon based on their comparable predicted binding affinities for the tripeptides. Electronegative substitutions at the phenyl ring of residue 1 are predicted to be favorable for both sequences by strengthening the anion-pi interactions with the ligands' C-termini. Hydrophobic substitutions at residues I and 7 that enhance contacts with the ligands' D-Ala residues and displace surrounding water molecules are predicted to be constructive as well. | Keywords/Search Tags: | Antibiotics, Vancomycin, Binding affinities, Predicted | PDF Full Text Request | Related items |
| |
|