Structural studies of MHC class 1 complexes : implications for NK- and T-cell recognition

The cell-cell recognition events and the functional characteristics of the cell surface receptors that govern the responses of NK- and T-cells are central to molecular immunology. This thesis addresses the structural basis of such molecular recognition, more specifically the interaction between peptides and MHC class I molecules, as well as the interaction between such complexes and the inhibitory Ly49 receptors of NK cells or triggering antigen specific receptors of T-cells (TCR). A better understanding of the details of these receptor-ligand interactions may help us to develop tools to activate or inhibit specific NK- and T-cell subsets. The structure of MHC class I H-2Dd in complex with the HIV-derived peptide PI 8-110 was determined by X-ray crystallography at 2.4Å resolution following cloning, expression and refolding of bacterially expressed, non-glycosylated soluble MHC-peptide complexes. The cleft architecture displayed different structural features from that of other mouse MHC class I molecules, explaining the unusual peptide-binding motif. Comparative analysis of the H-2Dd structure and the previously published H-2Db structure was used to predict recognition motifs for the Ly49A receptor of NK cells. The structure was also used to interpret earlier studies of H-2Dd restricted cytotoxic T-cells. Specificity of the interaction of Ly49 NK cell receptors and their ligands was investigated using fluorescent tetrameric recombinant MHC-peptide complexes. These studies indicated that MHC class I associated glycan was not required for Ly49A-H-2Dd interaction, and that H-2Kb tetramer binding to Ly49C receptors was strongly influenced by the peptide presented by the class I molecule. Additionally, tetramer binding allowed visualization of receptor interactions previously undetected in functional studies, such as the binding of H-2Db to Ly49A and Ly49C, indicating the usefulness of this technology in dissecting molecular recognition events. The MHC class I tetramer technology was further used to investigate structural consequences of exchange of the mouse [beta]2-Microglobulin ([beta]2m) subunit with a human version. The nature of the [beta]2m subunit was demonstrated to profoundly affect the ability of MHC class I molecules to bind NK cell receptors. This effect Of [beta]2m may be indirect and occur through conformational changes in the MHC class I molecule. It may also be interpreted in favor of a direct interaction between Ly49 receptors and [beta]2m. Practically, as [beta]2m can be passively exchanged from serum in cellular in vitro culture systems, these data indicate that autologous serum should always be employed in functional assays. Finally, the mechanism of viral escape was dissected at a structural level, by solving the structures of two MHC class I alleles complexed with the same immunodominant peptide from lymphocytic choriomeningitis virus (LCMV), known to be selectively modified during the course of viral infection. The peptide bound the two MHC molecules in diametrically opposed conformations, where one side chain could serve as a downpointing anchor in one case and as a solvent exposed TCR residue in the other case. Through a single modification of an amino acid the virus is able to affect, in different but critical ways, the peptide binding or recognition of both MHC complexes which would otherwise be used to mount a protective cytolytic immune response. These data thus demonstrate that the functional consequences of subtle structural alterations may be enormous in a pathogen-host interplay, and that structural studies provide conclusive evidence to support other forms of investigative analysis