2001;276:14767C72. with TSHR-adenovirus, but the pattern was more consistent with increased background noise than with spreading from primary epitopes to dominant secondary epitopes. In conclusion, these data suggest that factors other than particular TSHR T cell epitopes (such as adenovirus-induced expression of conformationally intact TSHR protein), contribute to the generation of thyroid stimulating antibodies with consequent hyperthyroidism in TSHR-adenovirus immunized mice. Keywords: T cell epitopes, interferon-, tumour necrosis factor-, linear antibody epitopes, thyrotropin-receptor, murine Graves disease INTRODUCTION Immunizing mice with DNA encoding the TSHR in a replication deficient adenovirus is an effective approach for inducing Graves disease in BALB/c mice [1C3]. Mice of this strain also develop hyperthyroidism after multiple injections of TSHR-expressing B cells [4] or dendritic cells [5], confirming that the BALB/c strain is genetically susceptible to experimentally induced Graves disease. In contrast, BALB/c mice vaccinated with TSHR-DNA in a plasmid do not develop hyperthyroidism [6C8]. In one study, TSHR antibodies developed in virtually all TSHR-DNA vaccinated BALB/c mice [6]. In contrast, we and others observed low or undetectable levels of antibodies generated by vaccination with TSHR-plasmid [1,7C9]. Overall therefore the lack of hyperthyroidism in this strain is likely to be related, at least in part, to low TSHR antibody levels induced by TSHR-DNA plasmid vaccination. Antibody production by B lymphocytes requires help from T lymphocytes in the form of cytokines. Previously, to understand why TSHR-DNA vaccination is relatively ineffective at inducing TSHR PDGFRA antibodies, we investigated the T cell response to TSHR antigen. Challenging spleen cells from animals injected with TSHR-DNA with TSHR antigen induced responses measured by cell proliferation (thymidine incorporation) and production of several T Estradiol dipropionate (17-Beta-Estradiol-3,17-Dipropionate) helper-1 (Th1)-cytokines, including interferon-gamma (IFN-) and, at lower levels, interleukin-2 (IL-2) and tumour necrosis factor (TNF-) [7,9,10]. No Th2 cytokines (such as IL-4 or IL-5) were detected. Because earlier studies had suggested that Th2 responses were characteristic features of Graves disease models (for example [11], we attempted to enhance antibody responses by immune deviation away from Th1. However, TSHR antibody levels were not enhanced in IFN- knock-out mice or by employing an IFN-?decoy together with TSHR-DNA [9]. Recently, we used a panel of synthetic overlapping peptides, Estradiol dipropionate (17-Beta-Estradiol-3,17-Dipropionate) corresponding to the amino acid sequence of the TSHR, to analyse T cell responses in TSHR-DNA immunized mice. Spleen cells from BALB/c and NOD.H-2h4 mice responded to different peptides, as would be expected from their different Major Histocompatibility Complex (MHC) genes [10]. However, an unexpected finding was the restricted number of peptides recognized by mice of either strain. In particular, two groups of nonoverlapping peptides were recognized by all BALB/c mice [10]. A cardinal feature of induced immunity is spreading of T cell epitopes as the immune response progresses. Such spreading can involve recognition of new epitopes on the same antigen or even intermolecular spreading to other antigens (for example [12C14]. In contrast to TSHR-DNA vaccinated mice, there are no data on TSHR T cell epitopes in the much more effective Graves disease model induced by TSHR-adenovirus immunization. In the present study we tested the hypothesis that, in the same strain of mice (BALB/c), TSHR-adenovirus immunization would induce a wider, or different, spectrum of TSHR T cell epitopes, hence accounting for the very high incidence of TSHR antibodies with functional activity leading to hyperthyroidism. METHODS AND MATERIALS Immunization of mice with TSHR-adenovirus Generation, amplification and purification of recombinant adenovirus Estradiol dipropionate (17-Beta-Estradiol-3,17-Dipropionate) expressing human TSHR (AdCMVTSHR) were performed as previously described [1]. Female BALB/c mice (6C7 weeks; Jackson Laboratories, Bar Harbor, Maine, USA) were injected intramuscularly once or three times at 3 week intervals with 50 l PBS containing 1011 particles of TSHR-adenovirus (Ad-TSHR) or control adenovirus expressing -galactosidase (Ad-Con). Mice were euthanized at the following times: 1 week after a single injection of adenovirus, or 8 weeks after the third injection; in addition, blood was drawn from the tail vein one week after the second injection. The mice immunized three times with TSHR adenovirus were previously characterized for TSHR antibodies and hyperthyroidism [3]. All animal studies were approved by the local Institutional Animal Care and Use Committee and were performed in accordance with the highest standards of care in a pathogen-free facility. ELISA for TSHR antibodies Sera were examined for recognition of TSHR-289, a variant of the receptor expressed in eukaryotic cells that corresponds approximately to the extracellular A-subunit recognized by human.