Immune mechanisms in autoimmune thyroiditis

Searching for pathogenic epitopes in thyroglobulin: parameters and caveats

Last year marked the 40th anniversary of the discovery that thyroglobulin (Tg) is a major autoantigen in autoimmune thyroiditis. The Tg molecule presents unique challenges for epitope mapping owing to its large size and extensive iodination. Consequently, pathogenic determinants have only recently been identified. Here, George Carayanniotis and Varada Rao summarize the approaches used to determine pathogenic Tg T-cell epitopes and discuss caveats in this unusual quest.

Contrasting immunopathogenic properties of highly homologous peptides from rat and human thyroglobulin

The current lack of amino acid sequence data for mouse thyroglobulin (Tg) necessitates mapping of pathogenic T-cell epitopes on heterologous Tg in mouse experimental autoimmune thyroiditis (EAT). A prevailing assumption has been that epitopes sharing a high degree of amino acid homology among heterologous Tg are likely to exhibit the same immunopathogenic properties in the same host. In this report, we have examined this concept while working with the 18-mer rat(r)Tg(2695-13) peptide that was previously shown to elicit 'A'-restricted T cells and EAT in SJL mice. A major immunopathogenic T-cell epitope was localized within the 12-mer rTg(2695-06). It was found that the human 12-mer homologue that carries two Ser substitutions at Glu2703 and Thr2704 exhibited contrasting properties: it failed to activate Th1 cells in lymphokine and proliferation assays; it did not cross-react with rTg(2695-06) at the T-cell level; and it induced only focal thyroiditis following adoptive transfer of specific lymph node cells. These data highlight the caveat involved in extrapolating results of pathogenic T-cell epitope mapping across heterologous Tgs, even when such epitopes share a high degree of amino acid homology.

IFN-gamma has a protective role against thyroid-specific autoantibody production in severe combined immunodeficient (SCID) mice xenografted with Graves' thyroid tissue

We studied the effects of exogenous human IFN-gamma or neutralizing monoclonal antibody (mAb) to IFN-gamma on xenografted human Graves' thyroid tissue in severe combined immunodeficient (SCID) mice to investigate a possible role of IFN-gamma in the pathogenesis of human Graves' disease. Human thyroid tissues from four patients with Graves' disease were xenografted into SCID mice. Two weeks after xenografting, mice were divided into three groups with human IgG levels similar to each other. Mice in the first group were treated with human IFN-gamma daily for 6 weeks; mice in the second (similar) group were treated with an mAb to IFN-gamma; mice in the third group were given mouse IgG only (control group). Blood samples were taken every 2 weeks for human IgG and thyroid-specific autoantibodies (Tg-Ab, TPO-Ab, and thyroid-stimulating antibody). After 6 weeks' treatment, mice were killed, and the thyroid xenograft was examined for thyrocyte HLA-DR expression. Human IgGs were produced equally in all three groups; mice treated with IFN-gamma showed significantly lower amounts of thyroid autoantibodies than those in the control group. Thyrocyte HLA-DR expression was markedly increased in xenografts from mice with IFN-gamma administration. On the other hand, anti-IFN-gamma mAb injection caused only slight suppression of HLA-DR expression on xenografted thyroid cells. In conclusion, IFN-gamma may down-regulate the production of thyroid-specific autoantibodies but not human IgG, at least under these circumstances; there thus may be specific inhibitory effects of IFN-gamma against thyroid-specific autoantibody production of intrathyroidal plasma cells, and this animal model may help to elucidate the possible role of cytokines in the pathogenesis of Graves' disease.

Interferon-gamma is required during the initiation of an organ-specific autoimmune disease

Autoimmune gastritis induced by neonatal thymectomy of mice is a CD4+ T cell-mediated organ-specific autoimmune disease. The characteristic features of autoimmune gastritis, which include a mononuclear infiltrate within the gastric mucosa, loss of parietal and chief cells and circulating autoantibodies to the gastric H+/K+ ATPase, appear 6-10 weeks after thymectomy. Here we have assessed the role of interferon-gamma (IFN-gamma) in the pathogenesis of the gastric lesion. Splenic T cells derived from mice with gastritis produced three- to tenfold more IFN-gamma than T cells from normal animals after stimulation with anti-CD3 antibodies. Treatment of neonatally thymectomized mice at weekly intervals for 6 or 12 weeks with a neutralizing rat monoclonal antibody to mouse IFN-gamma abolished the production of anti-gastric autoantibodies and decreased the incidence of gastric mononuclear infiltrates from the 69% observed in normal rat immunoglobulin (Ig)-injected mice to 16%. Further, in mice treated with only a single dose of anti-IFN-gamma immediately after thymectomy at 3 days after birth, the incidence of autoimmune gastritis was 1/19 compared to 8/19 in normal rat Ig-injected mice. Prevention of autoimmunity by neutralization of IFN-gamma several weeks prior to the detection of a pathological lesion strongly suggests that IFN-gamma plays an essential role in the initiation of the gastric autoimmune response.

The immunology of thyroid cancer

This article reviews the immunologic aspects of thyroid cancer, including thyroid-specific effector elements on thyroid cancer, and discusses the potential antigens, unique molecular markers, and transcription factors that could serve as targets for immunotherapy. The potential role of this type of treatment for thyroid cancer is examined also.

Non-tolerant B cells cause autoimmunity in anti-CD8 IgG2a-transgenic mice

Using a pair of gamma 2a/chi immunoglobulin genes, transgenic mice were generated to study tolerance induction in B cells that express IgG2a autoantibodies. The transgenic IgG2a specifically binds CD8 alpha chains of the CD8.2 allotype expressed on the surface of CD8+ T cells, but not CD8 molecules expressed by the CD8.1 allele. Thus, IgG2a transgenic mice expressing the CD8.1 allele were used as controls to monitor B cell development and mice expressing CD8.2 were used to study B cell tolerance. Both types of mice showed transgenic gamma 2a expression on the surface of B cells. Expression of endogenous heavy chain alleles was strongly inhibited in immature B cell subsets, whereas mature B cells co-expressed transgenic gamma 2a and endogenous IgM/D. The transgenic chi chain expression leads only to partial allelic exclusion of endogenous light chains. B cells that express high levels of transgenic CD8.2-specific IgG2a were identified using soluble CD8-Ig. In CD8.1+ and in CD8.2+ mice, we found no differences in expression and maturation of transgenic anti-CD8.2 IgG2a+ B cells. High levels of serum anti-CD8.2 IgG2a antibodies led to the elimination of CD8+ T cells, causing a severe defect in cytotoxic immune responses. These results show that tolerance induction is incomplete in the CD8.2+ mice, either because IgG2a+ B cells are resistant to censoring mechanisms or because the secreted CD8-specific IgG2a antibodies render the CD8 autoantigen inaccessible to the B cells. This contrasts strongly with the efficient induction of B cell tolerance in mice expressing anti-CD8.2 IgM autoantibodies.

Molecular heterogeneity of antigen- or idiotype-induced anti-thyroglobulin monoclonal autoantibodies

To define the molecular basis of the cognitive interaction in experimental autoimmune thyroiditis (EAT), we sequenced the variable regions of monoclonal autoantibodies to thyroglobulin (Tg), specific or not for the F40D peptide, a Tg peptide capable of inducing EAT in CBA/J mice. Three MoAbs were obtained by immunization with syngeneic Tg of CBA/J (3B8G9, 2F6F2) or C57Bl/6 (4D11F4) mice. 3B8G9 was specific for F40D peptide, whereas 2F6F2 and 4D11F4 were not. Two others were raised in CBA/J mice by manipulation of idiotypic pathways: B12 resulted from the immunization with one Ab2 beta, bearing the internal image of one F40D epitope, and TA2 from the immunization with F40D-specific cytotoxic HTC2 T cells. B12 and TA2 were both specific for F40D. All hybridomas expressed different members of the J558 VH family, except 3B8G9 which expressed a Q52 VH gene segment. These data led us to hypothesize that regulatory anti-id autoantibodies used members of one VH family located in the 5'-end of the VH locus, whereas EAT-associated autoantibodies used a member of one of the most D-proximal VH family. As expected, no homologies were found when anti-F40D monoclonal autoantibodies were compared with two other monoclonal autoantibodies displaying a different epitopic specificity. Among the anti-F40D monoclonal autoantibodies, one histidine residue located in position 35 of the CDR1 region was constantly found. Moreover, TA2 and B12 exhibited two common amino acids in their CDR3 regions, one glycine and one tyrosine, in positions 98 and 99, respectively. Striking homologies were found between TA2 and one anti-polyGAT MoAb, and between 3B8G9 and some anti-phenyloxazolone (phOx) monoclonal autoantibodies. Lastly, the VK sequence from 4D11F4 was identical at the amino acid level to the VK sequence from another monoclonal autoantibody, 81B1, which was previously raised towards syngeneic Tg in CBA/J mice. Our data imply that anti-idiotypic regulatory circuits in EAT might be generated by a heterogeneous population of B cells rather than obtained by a single dominant B cell population.

Biphasic effect of interferon-gamma in murine collagen-induced arthritis

Interferon-gamma (IFN-gamma) exerts both enhancing and suppressing influences on collagen-induced arthritis (CIA), depending on the route and protocol of administration. To study the role of IFN-gamma on the autoimmune process of CIA, we treated DBA/1 mice with two different rat monoclonal antibodies (mAb) to murine IFN-gamma. Treatments, given twice weekly for 4 weeks, consisted of intraperitoneal injections of either mAb. In early treatments, starting from the day of immunization with type II collagen (CII), the severity of arthritis was reduced in both groups of anti-IFN-gamma-treated mice compared with control groups. Moreover, anti-CII antibody levels decreased in the sera of these mice. CIA was also down-regulated in mice treated from days 14 or 28 post immunization. In contrast, late treatments with anti-IFN-gamma mAb either induced aggravating effects, or did not affect the course of the disease. On the other hand, administration of high doses (8 x 10(4) U three times/week) of rat recombinant IFN-gamma exerted a transient increase of CIA severity. These findings suggest that IFN-gamma may play a critical role during both the induction and the course of CIA, first enhancing the immune response, and then regulating the arthritis process.

Proliferation and autoantibody production by mouse thyroglobulin (MTg)-specific B cells activated in vitro by MTg and MTg-specific T cells

Immunization of genetically susceptible strains of mice with mouse thyroglobulin (MTg) and adjuvant results in sensitization of MTg-specific CD4+ T cells that can induce experimental autoimmune thyroiditis (EAT) after in vitro activation with MTg. These MTg specific CD4+ T cells also provide help to MTg-specific B cells which produce anti-MTg autoantibodies in vivo. This study was undertaken to develop a system with which to measure MTg-specific helper function in vitro. MTg-activated T cells were shown to provide help to MTg-primed B cells in vitro as assessed by B-cell proliferation and anti-MTg autoantibody production. These responses were MTg-specific and required MTg priming of both the T- and B-cell populations. An MTg-specific CD4+ T-cell line also induced MTg-specific B-cell proliferation and was capable of inducing IgG1, IgG2a and IgG2b synthesis in MTg- primed B cells.

A conserved epitope on H+,K(+)-adenosine triphosphatase of parietal cells discerned by a murine gastritogenic T-cell clone

BACKGROUND/AIMS

H+,K(+)-adenosine triphosphatase (H+,K(+)-ATPase) of parietal cells is an organ-specific enzyme recognized by autoantibodies found in human and murine autoimmune gastritis (AIG). Murine AIG can be induced in BALB/c mice by thymectomy 3 days after birth and is a T cell-mediated disease. This study examined the specificity of T cells that cause AIG and the role of H+,K(+)-ATPase in this disease.

METHODS

From an AIG mouse, a gastritogenic T-cell clone (II-6) was established, and its reactivity to synthetic peptides of H+,K(+)-ATPase was tested.

RESULTS

II-6 cells are CD4+, V beta 14+, and interferon gamma producers. Adoptive transfer of II-6 cells to syngeneic nude mice resulted in AIG without the production of autoantibodies to parietal cells. The II-6 cells were responsive not only to murine but also to human and porcine parietal cells. Their proliferation was also induced by amino acids 891-905 (alpha 891) and 892-906 (alpha 892) of the alpha subunit of porcine and human H+,K(+)-ATPase, respectively.

CONCLUSIONS

The T-cell response to a single epitope of H+,K(+)-ATPase, the amino acid sequence of which is conserved among at least three mammals tested, is sufficient to cause AIG. Autoantibodies to parietal cells are not detected in these AIG mice.

Assessment of collagen type II induced arthritis in mice by whole blood chemiluminescence

Lucigenin-enhanced chemiluminescence can be measured in 100 microliters samples of whole, unseparated mouse blood. A procedure for doing so is here described in detail, using a standard clinical luminometer. The assay measures the TPA-induced oxidative burst from granulocytes and macrophages, which is believed to depend on the overall level of inflammation in the body. It is here applied to mice suffering from type II collagen-induced arthritis, and its relation to overt disease symptoms (the arthritis score) is characterised during the course of the disease. A correlation between the assay and the arthritis score is found at the height of the disease (r = 0.42, p = .039), but not at early or very late time points, although there is a strong hint that the results of an early assay may predict the subsequent disease course. The assay provides a rapid, convenient, quantitative and economical method of assessing disease activity, which can be carried out repeatedly on the same individual. It should be applicable in other mouse models of chronic inflammatory disease. It may find application for rapid screening of novel anti-rheumatic drugs and treatments.

Distinct genetic pattern of mouse susceptibility to thyroiditis induced by a novel thyroglobulin peptide

Experimental autoimmune thyroiditis (EAT), induced by thyroglobulin (Tg) and adjuvant, is major histocompatibility complex-controlled and dependent on Tg-reactive T cells, but the immunopathogenic T-cell epitopes on Tg remain mostly undefined. We report here the thyroiditogenicity of a novel rat Tg peptide (TgP2; corresponding to human Tg amino acids 2695-2713), identified by algorithms as a site of putative T-cell epitope(s). TgP2 causes EAT in SJL (H-2s) but not in C3H or B10.BR (H-2k), BALB/c (H-2d), and B10 (H-2b) mice. This reveals a new genetic pattern of EAT susceptibility, since H-2k mice are known to be high responders (susceptible) after Tg challenge. Following in vivo priming with TgP2, T cells from only SJL mice proliferated significantly and consistently to TgP2 in vitro, whereas TgP2-specific IgG was observed in all strains tested. Adoptive transfer of TgP2-primed SJL lymph node cells to naive syngeneic recipients induced a pronounced mononuclear infiltration of the thyroid, which was more extensive than that observed after direct peptide challenge. TgP2 is non-immunodominant, since priming of SJL mice with rTg did not consistently elicit T-cell responses to TgP2 in vitro and a TgP2-specific T-cell hybridoma did not respond to antigen presenting cells pulsed with rTg. The data support the notion that Tg epitopes need not be either iodinated or immunodominant in order to cause severe thyroiditis and that the genetic pattern of the disease they induce can be distinct from that of Tg-mediated EAT.

A common T-cell epitope between human thyroglobulin and human thyroid peroxidase is related to murine experimental autoimmune thyroiditis

We have investigated functional common T-cell epitopes between human thyroglobulin (hTg) and human thyroid peroxidase (hTPO) in mice. Four hTg peptides, Tg-P1, Tg-P2, Tg-P3 and Tg-P4, in which 5 amino acid residues are identical to those of hTPO, and 1 hTPO peptide, TPO-P4 relevant to Tg-P4, were prepared. Among these peptides, only Tg-P4 (residues 2730-2743) and TPO-P4 (residues 118-131) were highly antigenic and both peptides shared the common T-cell epitope. In addition, when the spleen cells from mice immunized with mouse Tg (mTg) were restimulated in vitro by Tg-P4 or TPO-P4 as well as by mTg, these cells transferred thyroiditis to naive recipient mice. These findings indicate that this common T-cell epitope between hTg and hTPO is immunogenic and related to the development of murine experimental autoimmune thyroiditis.

Cellular basis of T-cell autoreactivity in autoimmune diseases

There is no doubt that T cells play a key role in the pathogenesis of autoimmune diseases (AD) both as effector and regulatory cells. Despite spectacular progress in the understanding of natural tolerance to self, owing particularly to transgenic technology, important questions remain open regarding the pathogenesis of AD, the conditions favoring the transition from benign or 'physiological' autoimmunity to deleterious autoimmunity, and the precise effector mechanisms. This review on the cellular basis of T-cell-mediated AD begins with an enumeration of the main arguments in favor of direct T-cell involvement, special emphasis being given to two animal models which have been most extensively investigated: experimental allergic encephalomyelitis, and the nonobese diabetic mouse. The question as to whether pathogenic T cells use a restricted repertoire of V beta genes is examined in the context of these two models. From here we proceed to an evaluation of the mechanisms of onset of AD, discussing both extrinsic and intrinsic factors responsible for the breakdown in T-cell tolerance and reviewing the arguments in favor of suppressor T cells being actively involved in the prevention of autoimmunity. The last two sections are devoted to the effector mechanisms responsible for tissue injury in organ-specific AD and to T-cell-directed therapeutic interventions, respectively. We discuss the two main pathogenic hypotheses based on direct intervention of cytotoxic T cells or indirect involvement of inflammatory cytokines and macrophages, and evaluate the importance of ecotaxis in leading autoreactive T cells to the site of injury. We conclude on a brief and nonexhaustive list of strategies aimed at selectively neutralizing potentially harmful T cells.

Traumatic root resorption in dentine-immunized mice

Traumatic root resorption in a mouse model has been shown to coincide with a decline in naturally occurring serum antibody levels to dentin. It has been proposed that root resorption may be dentin antibody mediated. The purpose of this study was to examine the traumatic root resorption response in mice after hyperimmunization with a crude tooth extract (dentin). The hypothesis of this study was that elevated dentin antibody titers would positively correlate with root resorption. Mice were immunized with mouse dentin and controls were sham immunized. All mice were boosted 4 weeks later with or without mouse dentin as appropriate. All mice were then boosted two more times at weekly intervals with mouse dentin and then twice at weekly intervals with rat dentin. The change to rat dentin was made to increase mouse serum antibody titers to dentin. Serum samples were obtained before the initial immunization and weekly after each boost and were examined for antibody-to-dentin antigen by the enzyme-linked immune sorbent assay (ELISA). One week after the second boost with rat dentin, all animals were exposed to the cryoprobe procedure. Mice were killed 10 days later, and serum tested for antibody to dentin antigen. The incisors were examined by scanning electron microscopy, and root resorption quantified. Root resorption was observed on the incisors in the sham-immunized mice but not in the dentin-immunized mice. A trend toward increased serum antibody titers to dentin in immunized mice was observed over time.(ABSTRACT TRUNCATED AT 250 WORDS)

Ig VH gene family usage in spleen cells of CBA/J mice immunized with experimental autoimmune thyroiditis (EAT) inducer antigens

Experimental autoimmune thyroiditis (EAT) is an autoimmune disorder of the thyroid gland induced in susceptible strains of mice by thyroglobulin (Tg). We recently showed that low Mr (< 10 kDa) Tg tryptic fragments and a 40 amino-acid peptide (F40D) from Tg could induce EAT as well as native Tg. Because it has been reported that autoantibodies (A-Abs) express VH families preferentially located in the D-proximal VH gene segment, we investigated whether A-Abs specific for one pathogenic peptide from Tg were also skewed towards D proximal VH gene segment. In that respect, we immunized CBA/J mice with EAT inducer antigens of decreasing sizes: Tg (660 M(r)), < 10 kDa Tg trypic fragments or F40D peptide (4.9 kDa M(r)) from Tg. The VH gene segments utilized by immune spleen cells were determined by hybridization to total spleen cell RNA previously deposited onto nylon membranes and densitometric scans. This study was conducted on days 7 and 9 after determination of the maximum amounts of mRNA coding for immunoglobulins and on day 28 when A-Ab levels are the highest. Results were compared to VH gene segment expression both in normal and adjuvant-injected mice. We found that immunization of CBA/J mice with EAT inducer antigens stimulate B cells the restriction of which, in terms of VH family usage, depends on the size of the immunizing antigen: the larger the antigen, the higher the numbers of VH families used. Moreover, we found that B cell stimulation consecutive to immunization with the peptidic antigen inducing EAT occurs in VH Q52 family, a VH encoded by D-proximal gene segment.

Antibody production in sharks and humans: a role for natural antibodies

Although gene segments specifying Igs of all vertebrates show clear homology, their arrangements differ markedly, thereby suggesting that the mechanisms for the generation of diversity and for the regulation of gene expression may be quite distinct. In the sandbar shark, light chain gene segments are distributed as apparently independent clusters consisting of V, J, and C elements that require rearrangement for expression. The usual distance between V and C in the clusters is 3 kb but larger clusters occur. The V, J, and C elements are clearly homologous to those of human lambda chains. Shark Igs resemble mammalian IgM in structure and gene similarity. IgM may comprise as much as 50% of serum proteins in the shark. By contrast, IgM in humans comprises less than 5%. Human autoantibodies usually are IgM. These show little dependence on thymic function for expression and tend to increase with age. We have carried out a study of the capacity of Igs of unimmunized sharks and people (normals and patients suffering from autoimmune diseases) to react against a panel of antigens, including those usually considered autoantibodies, such as thyroglobulin and single-stranded DNA. Sharks and humans possess IgM antibodies that react with thyroglobulin and ssDNA. Affinity-purified natural shark antibodies to thyroglobulin or ssDNA constitute small fractions of total IgM. They illustrate extensive cross-reactivity comparable to that shown by polyspecific IgM autoantibodies produced by human B cells (CD5+) that appear early in ontogeny.

The effects of a monoclonal antibody to interferon-gamma on experimental autoimmune thyroiditis (EAT): prevention of disease and decrease of EAT-specific T cells

CBA/J mice immunized with thyroglobulin (Tg) develop an experimental autoimmune thyroiditis (EAT) with lymphocytic infiltration of the thyroid glands, autoantibodies to Tg and occurrence of EAT-specific T cells. When these mice were treated for 4 weeks after immunization with 1 mg/week of a monoclonal antibody (mAb) that neutralizes the activity of interferon-gamma (IFN) a beneficial effect on the onset of EAT was observed. Characteristic features of EAT were significantly reduced, including the lymphocytic infiltrations of the thyroid glands and the serum levels of autoantibodies to Tg. Moreover, in lymphoid organs, mAb to IFN-gamma significantly reduced the percentages of Tg-specific CD8+ cells, labeled by the anti-clonotypic mAb AG7. These Tg-specific T cells seem responsible for thyroid damages and disease development, since EAT was simultaneously abrogated. These results show that IFN-gamma plays an essential role in the pathophysiology of EAT and suggest the possibility to treat autoimmune thyroid diseases with mAb to IFN-gamma or drugs able to antagonize the production and/or the action of this cytokine.

New antigenic clusters on human thyroglobulin defined by an expanded panel of monoclonal antibodies

Twenty-seven hybridomas secreting monoclonal antibodies (mAb) directed against new antigenic clusters on human thyroglobulin (hTg) were obtained by fusion of the mouse myeloma P3-X63-Ag8 653 with spleen cells from BALB/c mice immunized with a mixture of hTg and six anti-hTg mAb with the aim of masking the corresponding antigenic clusters previously reported. Fourteen mAb were selected, produced in ascitic fluid and characterized. All these mAb were of the IgG1 subclass. Five new antigenic clusters on the hTg molecule were defined by the 14 mAb, extending the initial antigenic map of hTg to 11 clusters. These mAb were used in an attempt to probe the interaction between hTg and the autoantibodies from patients with Hashimoto's thyroiditis who do not recognize antigenic cluster II, a cluster whose recognition by anti-hTg autoantibodies is significantly associated with thyroid disorders.

Decreases in alpha beta T cell receptor negative T cells and CD8 cells, and an increase in CD4+ CD8+ cells in active Hashimoto's disease and subacute thyroiditis

We examined peripheral lymphocyte subsets in patients with autoimmune thyroid disease, or subacute thyroiditis, in the active stage when possible. During destructive thyrotoxicosis arising from alpha beta T cell receptor (TCR) negative T (WT31-CD3+) cells and CD8 (CD4-CD8+) cells decreased and those of CD4+CD8+ cells increased slightly, resulting in proportional increases in CD4 (CD4+CD8-) cells, non-T, non-B (CD5-CD19-) cells, and the CD4/CD8 cell ratio. Changes were similar in active subacute thyroiditis. During stimulative thyrotoxicosis in active Graves' disease, the numbers of such T lymphocyte subsets were not changed, but only the number of CD5+ B (CD5+CD19+) cells increased markedly, resulting in proportional decreases in total T (CD3+) cells, alpha beta+ TCR T (WT31+CD3+) cells, CD8 cells, and non-T, non-B cells. A serial study of some of the patients showed opposite changes in alpha beta TCR- T cells, the CD4/CD8 cell ratio, and CD5+ B cells between the active stages of Graves' and Hashimoto's diseases. alpha beta TCR- T cells were mostly gamma delta TCR+ T (IIF2+ CD3+) cells in these patients. These data suggest that alpha beta TCR-T (gamma delta TCR+ T), CD8, and CD4+ CD8+ cells are important in thyroid destruction in Hashimoto's disease and subacute thyroiditis, and that CD5+ B cells are important in thyroid stimulation in Graves' disease.

T cells and human autoimmune thyroid disease: emerging data show lack of need to invoke suppressor T cell problems

Human T cells recognize self and foreign antigens when such antigens are processed into small peptides and bound to molecules coded for by genes of the HLA region on chromosome 6. The part of the T-cell surface which is responsible for such recognition is a set of molecules coded for by a variety of genes and known as the T-cell-receptor complex. In animal models, T cells are able to transfer autoimmune thyroiditis and T cells have, therefore, long been implicated in the etiology of human autoimmune thyroid disease (AITD). Information gained from the study of intrathyroidal T cells and thyroid antigen-specific T-cell clones has shown that in patients with Graves' disease, mainly helper T-cell clones have been obtained, whereas in autoimmune (Hashimoto's) thyroiditis cytolytic T-cell clones may be predominant. Such thyroid antigen-specific T cells have now been shown to recognize one or other of the three major thyroid-specific antigens; thyroglobulin, thyroid peroxidase, or the TSH receptor and efforts are currently in progress to characterize the T-cell epitopes of these major thyroid autoantigens. Recent findings of restricted T-cell receptor V gene use amongst intrathyroidal T cells confirm the primary role of T cells in human thyroid autoimmune processes leading to AITD. However, the mechanisms whereby such autoreactive T cells escape deletion and anergy, and how they become activated, remain uncertain. There is compelling evidence that the thyroid cell itself, by expressing HLA molecules, and presenting antigen directly to the T cells, may initiate disease, perhaps after an external insult.

Induction of severe granulomatous experimental autoimmune thyroiditis in mice by effector cells activated in the presence of anti-interleukin 2 receptor antibody

Spleen cells from CBA/J mice immunized with mouse thyroglobulin (MTg) and the adjuvant lipopolysaccharide induce experimental autoimmune thyroiditis (EAT) after transfer to recipient mice if they are first activated in vitro with MTg. EAT induced by cells cultured with MTg is generally moderate in severity and is characterized by a thyroid infiltration consisting primarily of mononuclear cells. Addition of the anti-interleukin 2 receptor (IL-2R) monoclonal antibodies (mAbs) M7/20, 3C7, or 7D4 to spleen cell cultures with MTg resulted in a cell population capable of inducing a more severe type of EAT characterized by extensive follicular destruction, granuloma formation, and the presence of multinucleated giant cells. Recipients of cells cultured with MTg and anti-IL-2R mAb also had higher anti-MTg autoantibody responses than recipients of cells cultured with MTg alone. Activation of cells capable of transferring severe granulomatous EAT and increased anti-MTg autoantibody responses required both MTg and M7/20 in culture and required addition of M7/20 within the first 8 h of the 72-h culture period. CD4+ T cells were required for the expression of both the severe granulomatous EAT lesions and the mononuclear cell infiltrates typically observed in murine EAT. The increased anti-MTg autoantibody responses in recipients of cells cultured with MTg and anti-IL-2R mAbs were not restricted to a particular immunoglobulin G (IgG) subclass and included antibody of the IgG1, IgG2A, and IgG2B subclasses. These results suggest that a subset of CD4+ T cells capable of inducing severe granulomatous EAT and increased anti-MTg autoantibody responses is preferentially activated when cells are cultured in the presence of anti-IL-2R mAb. Anti-IL-2R mAb may either prevent activation of cells that induce classical lymphocytic EAT or prevent activation of cells that normally function to downregulate EAT effector T cell activity.

Suppression in experimental autoimmune thyroiditis: the role of unique and shared determinants on mouse thyroglobulin in self-tolerance

Previous studies have shown that T cells from mice genetically susceptible to experimental autoimmune thyroiditis (EAT) recognize determinants shared between mouse thyroglobulin (Tg) and heterologous Tgs. Some shared determinants are thyroiditogenic; lymphocytes from mice immunized with mouse Tg (MTg) or human Tg (HTg) and reciprocally restimulated in vitro with either Tg can transfer EAT. Studies on the mechanisms of self-tolerance have shown that pretreatment with soluble MTg suppresses in vitro proliferation to MTg and EAT induction with MTg. To determine the role of share epitopes in maintaining tolerance, mice were pretreated with soluble HTg and immunized with HTg or MTg and adjuvant. Cells from HTg-pretreated. HTg-immunized mice showed suppressed in vitro proliferative response to HTg. Following MTg immunization, the cells showed suppressed in vitro response to MTg. However, in contrast to MTg pretreatment, the subsequent development of EAT in vivo was unaltered in severity following HTg pretreatment. Thus, determinants shared between HTg and MTg can induce suppression of in vitro responses to HTg and MTg, but not inhibit the onset of thyroiditis, suggesting that T cells recognizing MTg-unique epitopes expanded to mediate thyroiditis. We conclude that recognition of both unique epitopes expanded to mediate thyroiditis. We conclude that recognition of both unique and shared epitopes on MTg are essential for the overall maintenance of self-tolerance.

Autoantibody specific for a thyroglobulin epitope inducing experimental autoimmune thyroiditis or its anti-idiotype correlates with the disease

In order to elucidate the role of anti-thyroglobulin (Tg) autoantibodies (A-Ab) in experimental autoimmune thyroiditis (EAT), a kinetic study was conducted in EAT-susceptible (CBA/J) and non-susceptible (C57BL/6) strains of mice. From day 0 to 70 post-Tg immunization, titers of A-Ab to Tg and to the linear 5-10-kDa Tg tryptic fragment inducing EAT as well as anti-idiotypic A-Ab representing the internal image of the thyroidogeneic antigen were measured. EAT onset, development and recovery correlate with the presence of two subsets of A-Ab only in susceptible strains of mice. First, with the presence of anti-Tg A-Ab to one determinant borne by the linear 5-10-kDa Tg tryptic fragment, and second with the presence of anti-idiotypic A-Ab specific for the monoclonal anti-Tg A-Ab (3B8G9) paratope which binds to Tg determinant inducing EAT.

Expression of intercellular adhesion molecule-1 (ICAM-1) on human thyroid cells lines correlated with their binding of lymphoblasts

We have investigated the expression of intercellular adhesion molecule-1 (ICAM-1) by novel functional human thyroid cell lines (designated SGHTL). ICAM-1 is constitutively expressed and it is rapidly upregulated in response to each of the recombinant cytokines: gamma-interferon, interleukin-1 and tumour necrosis factor. This contrasts with the more slowly increased expression of major histocompatibility complex (MHC) class II antigens in response to gamma-interferon alone. We have also demonstrated binding of activated lymphocytes to SGHTL cells: this interaction is increased following treatment with these cytokines and is inhibited by monoclonal antibodies directed against ICAM-1 or lymphocyte function-associated antigen-1 (LFA-1) but not by antibodies against CD2 or MHC class II antigens. Hence, we conclude that the binding of lymphoblasts to human thyroid cells involves an LFA-1- and ICAM-1-dependent pathway as well as other basal and cytokine-inducible pathway(s). These do not appear to involve MHC class II antigens, CD2 or an LFA-1 ligand other than ICAM-1.

Primary hypothyroidism associated with interleukin-2 and interferon alpha-2 therapy of melanoma and renal carcinoma

Four patients out of twenty with renal cancer and melanoma undergoing cancer immunotherapy with interleukin 2 (IL-2) and interferon alpha-2 (IFN-alpha 2) had laboratory evidence of hypothyroidism starting at cycle three to six, with a decline in serum thyroxine below normal and, in three cases, a rise in serum thyrotropin and thyroglobulin. One hypothyroid patient had elevated serum antimicrosomal antibody titres before the start of treatment and two others responded similarly during therapy. Three of the sixteen euthyroid patients also developed elevated titres of this antibody. Partial or complete remission was observed in seven of the patients--three of the four with hypothyroidism showed tumour regression. Thus IL-2 and IFN-alpha 2 can cause hypothyroidism, presumably via induction or exacerbation of autoimmune thyroid reactions. The occurrence of hypothyroidism may be mediated by high-dose IL-2 (rather than by LAK cell therapy as previously suggested) and potentiated by IFN-alpha 2.