Novel peptide antigens provide insights into the recognition process by T cells

The studies described here involve molecular aspects of T cell recognition and the reactivity of T cell receptors with novel peptide antigens. The first study (Chapter 2) described a collection of novel peptide antigens for a single T cell receptor and their diverse requirements for the co-receptor CD8. One of the peptides showed no activity despite binding the TCR, and examination of the crystal structure of the complex suggested that an unusual TCR docking geometry could influence the T cell signaling complex. The second study (Chapter 3) described the peptide-binding effects of two polymorphic residues (Trp and Arg) at position 97 in mouse and human MHC proteins. These polymorphisms in the human gene HLA-B have recently been associated with HIV control (tryptophan), or progression (arginine) to AIDS. My findings showed that the same polymorphism in position 97 in the mouse protein Ld resulted in a very significant affect on the repertoire of self-peptides that bound to the alleles. Binding of fewer peptides (Trp97) would result in broader diversity of the T cell repertoire, which would include T cells that enable immunity to HIV. Conversely, binding a broader array of self-peptides (Arg97) would result in thymic deletion of a larger number of T cells, including those that would have been capable of controlling an HIV infection. Structures of various peptide-MHC complexes allowed me to interpret the possible basis of the peptide-binding results. In the third study (Chapter 4), novel self-peptide antigens for a single, alloreactive T cell, were isolated in an attempt to examine the peptide antigen(s) that selected the mouse clone called 2C. My results showed that contrary to earlier studies, there were dozens of naturally expressed and Ld-presented peptides that could mediate reactivity with 2C T cells. However, the self-peptide called p2Ca (LSPFPFDL) that was identified over 20 years ago as a putative ligand, was not one of them. The first project, described in Chapter 2, was performed in collaboration with Drs. Jarrett Adams and K. Christopher Garcia (Stanford University). Yeast display libraries that expressed diverse nonamer peptides, covalently linked to Ld, were used to identify novel peptide antigens that bound to the TCR called 42F3. To screen these peptides for biological activity, I generated 42F3 T cell hybridoma cell lines, with or without the CD8αβ co-receptor. Results of T cell activation assays revealed that the novel antigenic peptides were either CD8-dependent or CD8-independent, but this dependency was not completely associated with the affinity of the pep-Ld complex for the TCR, as had been expected. Furthermore, one of the peptides was completely inactive, even though it bound well as a complex to the TCR. This result, in the context of the crystal structure of the complex, suggested that the geometry of TCR docking may influence the association of the TCR with signaling components and thus the clustering of TCR-peptide/MHC complexes in the immune synapse. The third chapter in this thesis details a project examining the contributions of position 97 of the murine MHC allele Ld in affecting peptide binding. A 2010 study published by the International HIV Controllers Study group described a relationship between HIV disease control or progression and HLA-B alleles, and focused on specific amino acids at key positions of HLA-B, including position 97. The study found that HLA-B alleles that contained a tryptophan at position 97 were associated more with HIV control, whereas alleles containing an arginine at position 97 were associated more with HIV progression. I focused on the Trp97 and Arg97 alleles because our lab, and others, had found independently in the mouse Ld system that an arginine at position 97 affected the stability of the Ld molecule. Structurally, the Ld allele shares homology with HLA-B alleles; position 97 in Ld and HLA-B lies near the center of the peptide binding groove, with the side chain of the position 97 residue directed out into the peptide-binding groove, a prime location for affecting peptide binding. For the analysis detailed in Chapter 3, I generated immortalized cell lines expressing full length LdW97 and LdR97 and used competitive peptide binding assays together with a soluble in vitro reconstitution system to determine the relative stability and binding of a panel of peptides to these two variants. The results showed that LdR97 bound a broader repertoire of self-peptides than LdW97. An examination of available peptide/Ld structures suggested that this was not due to additional flexibility of the arginine, as previously suggested in the HLA-B system. Rather, proximal residues at positions 114, 116, and 156 formed different interactions with either W97 or R97, and these proximal residues showed greater flexibility in LdR97 structures compared to LdW97 structures. This suggests that the enhanced flexibility of these residues in LdR97 provides the freedom to accommodate a broader collection of self peptides. Comparisons of Ld and HLA-B structures containing Trp97 and Arg97, indicate that the results will translate to the HLA-B system. Thus, the result is consistent with a greater proportion of T cells being clonally deleted during negative selection against self-peptide/MHC-R97 complexes. This would restrict the diversity of the T cell repertoire in the periphery and reduce the availability of T cells that could be effective against HIV (thereby leading to a “progressor” phenotype with the Arg97 alleles). The fourth chapter in this thesis re-examines the naturally presented alloantigen that selected the mouse T cell clone called 2C. Thirty years ago, T cell clone 2C was raised against the Ld-expressing P815 mastocytoma cell line. About twenty years ago, a peptide called p2Ca was identified by acid extraction from BALB/c thymus tissue as a peptide that mediated reactivity between the 2C T cell clone and Ld-expressing target cells. While p2Ca has been assumed to be the natural ligand for 2C, our lab had strong evidence to suggest that peptide p2Ca was not endogenously processed and presented on MHC Ld on the surface of P815 cells. Accordingly, peptide p2Ca could not have been the alloantigen against which the 2C T cell clone was raised. To search for the bona fide peptide(s), I obtained naturally derived peptides from BALB/c (Ld-positive) liver tissue, using either an acid extraction method or an affinity purification method (with an anti-Ld-monoclonal antibody). The resulting peptides were fractionated based on hydrophobicity using reverse-phase HPLC, and HPLC fractions were sampled in cytotoxicity assays using 2C T cells and the T2-Ld target cell line. Surprisingly, there appeared to be many more active fractions containing peptides than described previously, with a minimum estimate of at least 30 different active peptides. Thus, the alloreactive T cell clone 2C has an even greater repertoire of ligands than thought