Structural biology techniques and applications High resolution electron microscopy/tomography (EM/ET) has tremendous potential to obtain molecular structures of complex macromolecular assemblies that function as dynamic cellular machines. While maps derived from ordered two-dimensional protein crystals approach atomic resolution, the maps derived in single particle microscopy are most often at much lower resolution, in the range of 10 to 30 . A variety of computational docking algorithms have been developed to perform fitting into low-resolution maps, and have been reviewed by Wriggers and Chacon. We have developed a core-weighting approach to fit atomic structures into low-resolution EM maps of biomolecular assembly with multiple components. This core-weighted fitting method significantly improves the sensitivity to distinguish the correct fit. Combination of EM/ET and the new computational method provides us opportunities to study biomolecular machineries. Structural characteristics of chemotaxis receptor assemblies Bacterial chemoreceptors respond to changes in concentration of extracellular ligands by undergoing conformational changes that initiate a series of signaling events, leading ultimately to regulation of flagellar motor rotation. We used cryo-electron tomography combined with the 3D averaging method developed from this lab to determine the in situ 3D structure of receptor assemblies. We demonstrate that in wild-type Caulobacter crescentus cells chemoreceptors are organized as trimers of receptor dimers, forming signaling complexes in the cytoplasmic membrane that are packed in disordered hexagonal arrays at the cell pole. Experimental structure determination of human leukocyte antigen (HLA) DRA, DRB3*0101 MHC receptors recognize, bind and present samples of pathogens to T cells to initiate the immune response. The MHC region is by far the most gene dense and most polymorphic in the whole human genome. Over several hundred class II MHC molecules are known. However, only a small handful of structures are available. We determined the structure of HLA-DRA, DRB3*0101, also called HLA-DR52a. Using biophysical calculations and fine mapping of the electrostatics onto the binding groove, we showed that class II MHC alleles that predispose to autoimmune diabetes have strong electrostatic potential in their binding pocket, specifically pocket 9 (P9). Using sequence analysis and bioinformatics approaches, we show that two polar residues in the pocket are responsible for the electrostatic potential. We propose a new paradigm for autoimmune diabetes susceptibility: class II HLA-DQ and DR alleles that possess polar residues at β9 and β37 exert an electrostatic grip on the peptides with polar, particularly acidic residues that characterize diabetogenic epitopes. This results in tight peptide binding affinity and prolonged peptide kinetic off rates and altered cytokine expression that favor inflammation. We have extended our findings to the paralogous protein family HLA-DP. Molecular modeling with Map Objects Traditional molecular modeling is performed at atomic resolution, which relies on X-ray and NMR experiments to provide structural information. When dealing with biomolecular assemblies of millions of atoms, atomic description of molecular objects becomes inefficient. We developed a method that uses map objects as a molecular modeling tool to efficiently derive structural information from experimental maps. In addition, this technique allows convenient manipulation of map objects and performs conformational searching directly using these objects. This method was applied in a structural study of the perioxiredoxin complex. Structural biology techniques and applications To utilize the structural information of the low resolution images, more effective and sensitive computational methods need to be developed. So far we have developed the core-weighted fitting method to derive atomic structures of macromolecular assemblies from single-particle EM images and have successfully applied this method in the E2 icosahedral core of pyruvate dehydrogenase and molecular architecture and mechanism of an icosahedral pyruvate dehydrogenase complex. The rapid development of EM technique and application provide tremendous potential and need for new computational methods. The following are several directions we shall pursue in the coming years. Ongoing activities include: 1. Identification of multiple conformational maps from electron tomography. 2. Molecular modeling and simulation of chemotaxis assembly. High resolution 3. Multiscale simulation with EMAP. 4. Signal transaction in chemotaxis assembly. 5. Quantum mechanical calculations of modeling of excited state structures and solvatochromism in fluorescent probes for biological imaging 6. Map based simulation
Showing the most recent 10 out of 21 publications
