The focus of this work has been to understand the molecular details that control the interaction of MHC molecules with T cell receptors, of MHC molecules with coreceptor molecules, and of MHC molecules with NK receptors. These studies are dependent upon both functional and biophysical analysis of the interaction of the molecules in question, and attempts are made to correlate binding properties with function and structure. Our studies of MHC/TCR interaction have been focused on several systems, the TCR of the 2C cytolytic T cell which kills H-2Ld- bearing cells and has been shown to recognize a self peptide LSPFPFDL; and the TCR from an H-2Dd-restricted HIV envelope glycoprotein restricted T cell hybridoma. In the both systems we have made a comprehensive study of the interaction of the T cell with MHC molecules complexed with a variety of variant peptides. In particular, we have been trying to understand the relationship of early activation events to the strength of the binding as measured by surface plasmon resonance. The general trend has clearly been for the affinity of the TCR for the MHC/peptide complex to correlate with the degree of activation of the T cell. However, a single variant peptide, LSPLPFDL, is a perfectly good agonist when bound to H-2Ld, but it fails to make a complex that binds the 2C TCR in our in vitro binding assay. Dr. M. Jelonek has characterized this peptide and its higher affinity variant QLSPLPFDL extensively in terms of the ability to activate the T cell as assessed by a number of assays including Ca++ mobilization and z chain phosphorylation. The peptide is perfectly good at the functional level, but fails to make a complex for which we can detect binding. It is possible that the level of discrimination for the binding assay is at a critical threshold. Further studies using other binding assays, including surface staining with multivalent MHC peptide complexes, should resolve this issue. In the course of evaluating the interaction of the MHC-I molecule H-2Ld with the T cell coreceptor CD8 in direct binding assays, we observed that H-2Ld, and not H-2Dd or H-2Kb bound the CD8 preparation only when the H-2Ld had been freed of bound peptides, and was blocked by H-2Ld-binding peptides. Further investigation confirmed the existence of an H-2Ld-binding motif at the amino terminus of CD8, and established that the binding interaction was through the H-2Ld peptide-binding cleft. A major effort of this project in the past 6 months has been the analysis of the interaction of the inhibitory natural killer cell receptor, Ly-49A, with the MHC-I molecule, H-2Dd. Using a bacterial expression system, we have expressed the extracellular domain of Ly-49A, and several truncation versions of this, purified these by biochemical techniques, established their serological fidelity by binding to monoclonal antibodies, and analyzed the binding to H-2Dd. We have established that the kinetics of binding of H-2Dd to the NK receptor are characterized by a Kd of about 7 to 9 x 10-6 M, and kinetic dissociation rate constant, kc of ~0.02 sec-1. Interestingly, the binding of H-2Dd to Ly-49A is dependent upon bound peptide, but in contrast to the interaction with T cell receptor is not peptide specific. This result confirms in a pure binding assay, the findings that we had obtained in collaboration with the laboratory of Dr. W. Yokoyama, which indicated that functional recognition of Ly-49A was also not peptide specific. The in vitro binding assays have also allowed us to evaluate the contributions of carbohydrate and Ca++ to the binding of Ly-49A to H-2Dd. We have extended our ability to engineer Ly-49A to several other members of the Ly-49 family, in particular, Ly-49C and Ly-49G2. We expect to resolve questions of the differential specificity of these important inhibitory receptors that regulate NK cell recognition.
Showing the most recent 10 out of 31 publications
