The overall focus of the Project, Molecular Interactions of Lymphoid Cell Receptors, is a broad one, based on the crucial importance of the recognition of MHC molecules by many receptors of the innate and adaptive immune systems. Our longterm interest and expertise in the structure and function of MHC-I and MHC-II molecules, and a battery of molecular, cellular, and genetic reagents that are available for these studies allow us to address questions of the fundamental mechanisms that govern the molecular function, the cellular reactions, and the genetics and evolution of these molecules. In addition, because of our experience in the analysis of molecular interactions, we have been able to explore other molecules that function in innate recognition using various biochemical techniques. Our overall approach is to exploit cloned genes to express recombinant forms of the molecules of interest, to analyze their function in in vitro cellular assays, to analyze their biophysical properties using state-of-the-art methodologies such as surface plasmon resonance and analytical ultracentrifugation, and to understand the three-dimensional structural basis of their function with high-resolution X-ray crystallographic structural models. Contemporary molecular modeling strategies, taking advantage of computer clusters at the NIH have permitted us to extend the findings of our structural studies of some murine molecules to homologous molecules of the human, and allow a rational interpretation of the structural basis for many mutants reported in the literature. ? ? In this project we review functional studies of the interaction of the MHC-I alpha3 domain with the CD8 coreceptor, that indicate the disparity in its function with low as compared with high affinity T cells, and our recent determination of the first high resolution structure of the complex of the murine CD8alpha/beta heterodimer with a classical MHC-I molecule and beta2-microglobulin will be discussed.? ? We have examined the function of the CD8 coreceptor interaction with MHC-I by studying quantitatively the functional contribution of MHC-I when used in an in vitro stimulation assay of CD8-dependent CTL clones that are restricted to H-2Dd and show specificity for the HIV IIIB peptide, P18-I10 (RGPGRAFVTI). The CTL were well-stimulated by MHC/peptide complexes consisting of H-2Dd and P18-I10, but were poorly stimulated by mutant H-2Dd bearing an E227K mutant of the binding site for CD8. However, addition of wild-type H-2Kb protein, that failed to bind the antigenic peptide, and failed to stimulate the CTL on its own, was synergistic, indicating that the interaction of CD8 with a non-cognate MHC-I on the target or APC surface still contributes in shifting the sensitivity of the resulting T cell activation. Our explanation for this observation is that the interaction of CD8 with non-cognate MHC on the APC can still promote the incorporation of CD8 into lipid rafts at the T cell/APC interface.? ? These experiments prompted us to reexamine the structural basis of CD8 interactions with MHC-I molecules. Though the structure of the first MHC-I molecule was determined in 1987, and the structure of the human CD8 alpha/alpha homodimer was determined in 1992, and structures of CD8 alpha/alpha in complex with MHC-I and of mouse CD8 alha/beta unliganded have been reported, there has been no successful determination of the structure of the CD8 alpha/beta heterodimer in complex with MHC-I. This is of considerable importance because CD8 alpha/beta, not CD8 alpha/alpha, is the major functional molecule expressed on mature CD8+ cells. Rational strategies to develop highly efficient CTL for killing tumors would depend on knowledge of the MHC/beta2-m/CD8alpha/beta complex. Binding and quantitative functional studies from several laboratories indicate that CD8 achieves its coreceptor function by improving the functional readout of T cell activation, but not by increasing avidity. That is, the measured affinities of CD8 alpha/alpha and CD8 alpha/beta for MHC-I/peptide complexes have been measured to be about the same. Extensive binding and functional analyses of a battery of CD8 alpha/beta mutants in several laboratories have led to the rather unsatisfying conclusion that CD8 alpha/beta may not bind to MHC-I in a single conformation, but rather may assume one of several different distinct conformations, accounting for several ambiguous results derived from the mutagenesis studies. To address the issue of the nature of the conformation of CD8 alpha/beta when bound to an MHC-I molecule, we explored several different approaches to engineering CD8 alpha/beta for structural studies, and finally succeeded with mouse CD8 alpha and CD8 beta extracellular domains engineered for expression as inclusion bodies in E. coli. These were then solubilized and refolded with minor variations of our standard methods, and stable, disulfide-linked CD8 alpha/beta heterodimers were produced, purified free from contaminating homodimers, and used in co-crystallization trials with bacterially expressed, refolded, highly purified H-2Dd/mouse beta2m/P18-I10 complexes. Crystals were obtained, the best of which diffracted to 2.6 Angstroms. The results of this structure indicate that the CD8beta chain is in the membrane distal position, and reveal numerous details of the interaction of the CD8 heterodimer with the MHC-I molecule. ? ? To expand our understanding of the human MHC-I/beta-2m/CD8alpha/beta interaction we have undertaken a molecular dynamics model of the human MHC-I/beta-2m/CD8alpha/beta complex based on the mouse structure. This will provide a structural basis for evaluating a host of published mutants.
Showing the most recent 10 out of 31 publications
