Atomistic biosimulations

点击翻译按钮获取中文摘要

Certain natural and synthetic small cationic peptides have the ability to kill bacteria. Some of these peptides do not also attack mammalian cells, and are therefore potential replacements for traditional antibiotics, which is very desirable because of the development of antibiotic resistance of bacteria to traditional antibiotics. A long term goal is to understand at the molecular level, using atomistic computer simulation methods, the mechanisms of how these antimicrobial peptides kill bacteria, often through interaction with the cytoplasmic membrane, and how some of them can discriminate between bacterial and mammalian cell membranes. Three projects in this area are proposed. The first will test test the hypothesis of Prof. Tom Gill (Dalhousie) that the natural peptide protamine crosses the E.coli outer membrane via the membrane protein OmpF. We will simulate the local free energy and local diffusion coefficient for protamine inside the pore of OmpF to determine if protamine can pass through. A second project is to discover the molecular basis for the extraordinary lethality of two new very small synthetic peptides designed by Prof. Bob Hancock (UBC). We are simulating the binding free energies of the peptides to model bacterial-like membranes, and also to model mammalian-like membranes, to determine the selectivity properties of the peptides. The third project is to study by simulation the aggregation properties of the new Hancock peptides to test the hypothesis that these peptides first aggregate in groups of about six and then form barrel-like open pores in the bacterial membrane, which leads to bacterial cell death. Another project involves the simulation of a putative model of the transport protein ProP, proposed by the group of Prof. Janet Wood (Guelph), embedded in a model E.coli inner bacterial membrane. ProP functions by sensing excess osmotic pressure outside the bacterial cell, compared to inside, and then transmitting molecules into the cell, to restore osmotic balance. Comparison of simulated properties from the model with actual properties will validate the model or lead to a second generation improved one.