Ester Fraga

Unusually diverse T cell response to a repeating tripeptide epitope

The immune system utilizes a diverse T cell repertoire for the recognition of foreign antigens in the context of self MHC gene products. We have examined the potential diversity of the T cell response directed to a immunodominant repeating tripeptide epitope (EYA)5. This peptide represents one of the two T cell epitopes on the synthetic alpha-helical polypeptide antigen Poly 18, Poly EYK(EYA)5 in H-2d mice and does not require antigen processing prior to presentation to Poly 18-specific T cell hybridomas. The T cell response directed to the repeating tripeptide epitope (EYA)5 is extremely heterogenous even though the epitope has a relatively simple amino acid sequence. We have analyzed the fine specificity of 21 randomly chosen Poly 18-reactive, (EYA)5-specific and H-2d-restricted T cell hybridomas derived from H-2d, H-2bxd, and H-2b----H-2bxd Poly 18-responding mice to determine the number of unique antigen reactivity patterns represented by this T cell population. We used alanine- and/or lysine-substituted (EYA)5 peptides and a panel of haplotype-varied splenocytes and observed a great deal of microheterogeneity in response. We find that 13 of the 21 hybridomas have a distinct fine antigen specificity and T cell receptors. The binding of (EYA)5 to the antigen-binding groove of I-Ad appears to generate a highly diversified T cell response. Therefore, (EYA)5-I-Ad complex allows the activation of unrelated T cell clonotypes with the same overall antigen specificity and MHC restriction, but with distinct microheterogeneity in response and receptor usage.

Evidence for immunodominance between closely related epitopes in the selection of T cell repertoire: Hierarchy of T cell epitopes in a repeating sequence

In a protein antigen the number of epitopes that are presented by the MHC molecules to the T cells is generally limited. This phenomenon of immunodominance determines the T cell response to a given antigen. To understand the molecular basis of epitope selection we have analysed the hierarchy of T cell epitopes in a repeating synthetic polypeptide antigen Poly 18, Poly EYK(EYA)5 in the mice of H-2d haplotype. Because of its repeating nature, all the potential epitopes of Poly EYK(EYA)5 generated as a result of antigen processing will have extensive sequence overlap, therefore, providing an excellent system to investigate the molecular basis of T cell peptide epitope selection in vivo. We have synthesized a series of 12 and 15 amino acid peptides to mimic these epitopes. In H-2d mice Poly EYK(EYA)5 elicits T cell clones that optimally react with 15 amino acid peptides EYK(EYA)4 and/or (EYA)5. Similar results were obtained when related 12 amino acid peptides EYK(EYA)3 and/or (EYA)4 are used. (EYA)5 reactive T cell clones appear to be very heterogenous and much larger in number than EYK(EYA)4 reactive clones. (EYA)5 reactive clones could be elicited by at least three short Poly-18 derived epitopes (EYA)4, EYK(EYA)3 and (EYA)3EYK while EYK(EYA)4 reactive clones elicited only by the EYK(EYA)3 epitope. However, we observed the dominance of (EYA)5 reactive clones even when EYK(EYA)3 was used as an immunogen and this could be related to the degeneracy of their antigen specificity. Our earlier antigen competition studies suggest that (EYA)5 does not compete with EYK(EYA)4 epitope in binding to I-Ad. Therefore, there is no intramolecular competition between these epitopes to activate T cells. The epitope (EYA)3EYK appears to be subdominant since it can elicit Poly EYK(EYA)5 specific T cells upon immunization but does not appear to be part of Poly EYK(EYA)5 repertoire. Peptides such as (EYA)2EYKEYA or EYAEYK(EYA)2 with lysine substitution in the middle of the sequence were non immunogeneic. Similar results were obtained when the larger 15 amino acid peptides were used as antigen. Another level of epitope immunodominance is seen when substituted peptides of the two immunodominant epitopes are used. Some of these epitopes have potential to be part of the Poly 18 repertoire but they are greatly under represented when intact Poly 18 is used as antigen. The unusual hierarchy observed for immunodominance in these overlapping epitopes of EYK(EYA)5 sequence suggest a bias in the selection of T cell repertoire based upon the crossreactivity between potential epitopes generated as a result of antigen processing.

Role of the first external domain of I-Aβ chain in immune responses and diabetes in non-obese diabetic (NOD) mice

Diabetes in the non-obese diabetic (NOD) mouse is a multigenic autoimmune disease and is possibly controlled by three recessive loci, including one that is linked to the major histocompatibility complex (MHC). The first external domain of the Class II MHC I-A beta chain in these mice is unique and has been suggested as being responsible for autoimmunity. The I-A alpha chain in these mice is I-A alpha d, and they lack the expression of I-E molecules. We have investigated immune responses to various Ir gene control antigens in NOD mice to determine the influence of the NOD Ia and particularly the I-A beta chain. We find that sheep insulin is highly immunogenic while other insulins are weakly immunogenic in these mice. Hen egg lysozyme, pigeon cytochrome C and the synthetic polypeptide Poly 18, Poly EYK(EYA)5 antigen produce good antibody responses. Apart from H-2d, NOD are the only mice where Poly 18 antigen is immunogenic. In these mice Poly 18 induced good T-cell proliferative response, which was inhibited by anti-Ia antibody, and the mice were able to respond to tyrosine-containing polypeptide Poly EYA but not to the phenylalanine-containing antigen Poly EFA. We also found that synthetic peptide 48-60 of the NOD I-A beta chain is highly immunogenic in syngeneic NOD mice both for T cells and B cells. Using an I-A beta chain-specific monoclonal antibody, we are able to prevent induction of diabetes when the antibody was administrated in prediabetic, young mice. Our results suggest that the immune response to various antigens and autoimmune diabetes in NOD mice is directly influenced by the I-A beta chain.

Characterization of agretopes and epitopes involved in the presentation of beef insulin to T cells

Beef insulin-specific I-Ad-restricted T cell hybridomas were derived from the fusion of antigen-primed (BALB/c X B6)F1 T cells with BW5147 thymoma. Specificity analysis revealed that the A-chain loop region is involved in antigen recognition. Hybridoma A20.2.15 is specific for beef insulin and cross-reacted with sheep insulin, but not with pork insulin. Using synthetic peptides we showed that the A-chain loop containing peptide A1-A14 jointed to the B7-B15 peptide by a disulfide bond can activate this hybridoma. Fragments generated by enzyme digest further suggest that the peptide recognized on beef insulin appears to involve A-chain loop residues A5-A12 and B-chain residues B7-B13 that are linked by the A7-B7 disulfide bridge. We found that beef insulin needs to be processed prior to T cell activation. Glutaraldehyde fixation and chloroquine treatment of presenting cells abolished their capacity to present insulin. Beef insulin denatured by pH changes cannot activate, thus suggesting that simple denaturation is not sufficient for presentation by antigen presenting cells. Finally, the agretope on beef insulin is comprised of two functional regions B7-B13 on the B chain and the A-chain loop in the A-chain, while residues A8 and A10 are probably involved in interaction with the T cell receptor.

Fine-specificity analysis of antibodies directed to the C-terminal peptides of cytochrome c recognized by T-lymphocytes.

Recognition of cytochrome c by T-lymphocytes seems to involve the amino acid residues in the C-terminal region of the molecule. Lys-99 has particularly been identified as one of the critical residues in the recognition process. We have now raised antibodies against the C-terminal region of the cytochrome c molecule to map the residues that may be recognized by B-lymphocytes. These antibodies were generated in high-responder B10.A mice against either the 81-104 CNBr fragment of pigeon cytochrome c or against the synthetic spliced fragment (86-90)-(94-103) of the tobacco hornworm moth cytochrome c. A good antibody response was obtained for both fragments as measured by solid-phase radioimmunoassay. A series of peptides related to these fragments were synthesized for competitive inhibition to assess the antigenic sites on these molecules. In spite of substantial homology between the moth (86-90)-(94-103) and pigeon (81-104) fragments, the antibody populations raised against each fragment differed in their recognition patterns. Residues 99 (Lys), 103 (Ala) and 104 (Lys) were found to be crucial for binding of the anti-pigeon antibody to the pigeon 81-104 fragment. The fine specificity mapping of the antigenic sites on the moth (86-90)-(94-103) fragment indicated that along with some of the residues in the N-terminus (86-90), residue 99 (Lys) was involved in recognition of the moth (86-90)-(94-103) fragment by its antibody. This residue (Lys-99) also acts as a T-cell receptor contact site for both pigeon and moth cytochrome c. We therefore conclude that common patterns of recognition must exist between T and B-cells that recognize the C-terminal region of cytochrome c. Since Lys-99 is also present in mouse cytochrome c, the antigenic site must involve both self and non-self residues.