This project focuses on the study of membranes and carbohydrates by molecular dynamics computer simulation. Modeling studies, primarily on proteins and in collaboration with experimental groups, are also carried out.? ? Highlights of the year were simulations of lipid bilayers and micelles containing the fusion peptide from Influenza Hemagglutinin (reference 3, below), and simulations of bilayers and monolayers with the disacharide trehalose (4). ? ? In the former (3), simulations showed how the lipids rearrange themselves around the fusion peptide in a manner very distinct from lipids in a pure bilayer. Is is proposed that this deformation of lipids induces curvature in the bilayer, which is the essential first step of membrane fusion.? ? The simulations of trehalose (4) were aimed at answering two questions: (1) whether its effects are similar in monolayers (which are easy to study experimentally) and bilayers (which are more difficult to study); (2) the molecular basis for the cell preservation. Results of simulations indicated that the perturbations of trehalose on monolayers and bilayers are quantitiavely similar. This implies that conclusions based on monolayer studies can be reasonably applied to bilayers. The simulations also showed that trehalose displaces waters imbedded in the headgroup region of monolayers and bilayers while leaving the total number of hydrogen bonds approximately constant and other properties largely unchanged. Because the freezing of water in the cell membrane is a cause of cell rupture, the """"""""structure-preserving"""""""" replacement of water by trehalose is at the core of cell preservation. A second study (5) of trehalose/bilayer systems showed that trehalose lowers the area compressibility; i.e., it makes the membrane surface more flexible. Studies with other glucose, raffinose and other olgiosaccharides are ongoing. ? ? Another membrane study (2) entailed the development of a simulation-based structural model for the intrepratation of x-ray diffraction in lipid bilayers. Ths model and related techniques developed in this paper greatly facilate the estimation of lipid surface area from diffraction studies.? ? The group is also actively engaged in testing and development of force fields for the program CHARMM (Chemistry at HARvard Macomolecular Mechanics). This year's contribution involved proteins(1). Using lysozyme as a test cast, it was demonstrated that the Lipari-Szabo NMR order parameters and the root mean squared fluctuations agree well with experiment when the CMAP correction is included in simulations. Agree or simulation and experiment is qualitaively worse in the loop regions when CMAP is not included because the amplitudes of torsional fluctuations are too large.?
Showing the most recent 10 out of 13 publications
