Group A streptococcal cell surface M proteins elicit highly protective, serotype-specific opsonic antibodies and many serotypes also elicit host cross-reactive antibodies, which may contribute to the pathogenesis of poststreptococcal autoimmune disease. To date, studies aimed at designing safe (non-host-cross-reactive, defined-epitope) M vaccines have focused almost exclusively on antibody epitopes. Here we identify T-cell epitopes recognized by T cells from BALB/c, C57BL/6, and CBA/Ca mice immunized with purified, recombinant serotype 5 M protein (rM5). The responses of rM5-specific, major histocompatibility complex class Il-restricted, T-cell clones to synthetic peptides representing most of the M5 sequence identified at least 13 distinct T-cell recognition sites, including sites recognized by more than one major histocompatibility complex haplotype of mice. Although none of these sites appeared to be strongly immunodominant, an N-terminal peptide, sM5[1-35], was recognized by lymph node T cells of rM5-immunized mice and by a larger proportion of rM5-specific T-cell clones than any other individual peptide. The fine specificity of these clones was mapped with subpeptides to a single site at or overlapping the sequence ELENHDL at residues 21 to 27, which is in close proximity to previously mapped protective antibody epitopes. Other T-cell recognition sites are distributed throughout the M protein and include several in the highly conserved C-terminal region of the molecule. One of these C-terminal sites, located within residues 300 to 319, was recognized by a significant proportion of T-cell clones from two strains of mice. Helper T-cell epitopes located in the C-terminal region of M5 are likely to be widely conserved between different M serotypes and could be particularly useful in designing multivalent, defined-epitope M vaccines.

Several indirect observations have indicated that lymphocyte in the thyroid may be an important site of TSH receptor antibody synthesis in Graves' disease and we now describe an investigation of this possibility using improved lymphocyte isolation and TSH receptor antibody assay procedures. Our studies demonstrate that thyroid lymphocytes spontaneously produce TSH receptor antibody in culture. Furthermore, experiments with mitogen tend to suggest that these cells, in contrast to lymphocytes from lymph nodes draining the thyroid, are part of an active immune response to the TSH receptor.

Thyroid autoantibody synthesis was investigated in cultures of lymphocytes isolated from several sources, including thyroid and lymph nodes from patients with hyperthyroid Graves' disease treated preoperatively with carbimazole or propranolol. The ability of thyroid lymphocytes to secrete immunoglobulins, including thyroid microsomal or thyroglobulin autoantibodies, was markedly reduced in lymphocyte suspensions obtained from patients treated with carbimazole compared with suspensions from patients treated with propranolol. This effect (which was greater in individuals treated with carbimazole for longer periods) was attributable to a significant reduction in the number of viable lymphocytes present after the 14-day culture interval. In contrast, the type of preoperative therapy had little effect on cultures of lymphocytes obtained from lymph nodes draining the thyroid. Although it is not yet clear whether carbimazole exerts its effects in vivo by direct immunosuppression or indirectly by altering the thyroid microenvironment, our observations indicate that the fall in serum levels of thyroid autoantibodies that occurs during carbimazole therapy is related to an effect of the drug on lymphocytes within the thyroid.