Molecular analysis of V gene sequences encoding cytotoxic anti-streptococcal/anti-myosin monoclonal antibody 36.2.2 that recognizes the heart cell surface protein laminin

Anti-streptococcal/anti-myosin mAb 36.2.2 is unique among the cross-reactive anti-streptococcal mAbs due to its cytotoxicity for rat heart cells and its ability to strongly label the surface of heart cells in indirect immunofluorescence assays. In this study, cytotoxic mAb 36.2.2 was found to react strongly with the extracellular matrix protein laminin in immunoblots and inhibition assays, while 11 other cross-reactive anti-streptococcal mAbs did not react with laminin and were not cytotoxic. Cytotoxicity appeared to correlate with the presence of laminin on the surface of cells. Heavy and light chain variable region genes encoding mAb 36.2.2 were highly homologous to other V genes encoding anti-carbohydrate and/or autoantibodies. VH, JH, and Jkappa segments of mAb 36.2.2 may be encoded by germline gene segments. The VH segment may be identical with an as yet unidentified VH7183 family germline sequence, and the 36.2.2 Vkappa region gene is encoded by a Vkappa8 family member.     

IL-12 is involved in the induction of experimental autoimmune myasthenia gravis, an antibody-mediated disease

IL-12 has been shown to be involved in the pathogenesis of Th1-mediated autoimmune diseases, but its role in antibody-mediated autoimmune pathologies is still unclear. We investigated the effects of exogenous and endogenous IL-12 in experimental autoimmune myasthenia gravis (EAMG). EAMG is an animal model for myasthenia gravis, a T cell-dependent, autoantibody-mediated disorder of neuromuscular transmission caused by antibodies to the muscle nicotinic acetylcholine receptor (AChR). Administration of IL-12 with Torpedo AChR (ToAChR) to C57BL/6 (B6) mice resulted in increased ToAChR-specific IFN-gamma production and increased anti-ToAChR IgG2a serum antibodies compared with B6 mice primed with ToAChR alone. These changes were associated with earlier and greater neurophysiological evidence of EAMG in the IL-12-treated mice, and reduced numbers of AChR. By contrast, when IL-12-deficient mice were immunized with ToAChR, ToAChR-specific Th1 cells and anti-ToAChR IgG2a serum antibodies were reduced compared to ToAChR-primed normal B6 mice, and the IL-12-deficient mice showed almost no neurophysiological evidence of EAMG and less reduction in AChR. These results indicate an important role of IL-12 in the induction of an antibody-mediated autoimmune disease, suggest that Th1-dependent complement-fixing IgG2a anti-AChR antibodies are involved in the pathogenesis of EAMG, and help to account for the lack of correlation between anti-AChR levels and clinical disease seen in many earlier studies.     

Experimental autoimmune myasthenia gravis induction in B cell-deficient mice

Experimental autoimmune myasthenia gravis (EAMG) is an animal model for human myasthenia gravis (MG). Autoantibody-induced functional loss of nicotinic acetylcholine receptor (AChR) at the postsynaptic membrane is an important pathogenic feature of both MG and EAMG. To evaluate the extent at which the humoral immune response against AChR operates in the pathogenesis of EAMG, we immunized B cell knockout (muMT) and wild-type C57BL/6 mice with AChR and complete Freund's adjuvant. The ability of AChR-primed lymph node cells to proliferate and secrete IFN-gamma in response to AChR and its dominant peptide alpha146-162 were intact in muMT mice as in wild-type mice. Similar amounts of mRNA for IFN-gamma, IL-4 and IL-10 in AChR-reactive lymph node cells were detected in muMT and wild-type mice. However, muMT mice had no detectable anti-AChR antibodies and remained completely free from clinical EAMG. We conclude that B cells are critically required for the genesis of clinical EAMG, but not for AChR-specific T cell priming.     

The Th2 cytokine IL-4 is not required for the progression of antibody-dependent autoimmune myasthenia gravis

Experimental autoimmune myasthenia gravis (EAMG), a disorder of the neuromuscular junction, is mediated by autoantibodies against muscle nicotinic acetylcholine receptor (AChR). The roles of IFN-gamma (Th1) and IL-4 (Th2) cytokines in the initiation and progression of this disease are not fully understood. Recently, we have demonstrated that IFN-gamma is necessary for the initiation of tAChR-induced EAMG in mice. However, the role of IL-4 in the progression of clinical EAMG remained undetermined. In this study we have addressed the contribution of IL-4 in the disease progression in IL-4(-/-) C57BL/6j mice whose IL-4 gene has been disrupted. Following immunization with Torpedo (t) AChR, the IL-4(-/-) mice readily developed signs of muscle weakness and succumbed to clinical EAMG with kinetics similar to the susceptibility of IL-4(+/+) mice. The tAChR-primed lymph node cells from IL-4(-/-) mice vigorously proliferated to tAChR and to its dominant alpha146-162 sequence associated with disease pathogenesis. However, these T cells secreted higher levels of IFN-gamma and IL-2, suggesting the development of a Th1 default pathway in these mice. Nevertheless, the IL-4 mutation had no effect on the recruitment of CD4+ Vbeta6+ T cells specific to the dominant tAChR alpha146-162 sequence in vivo. Immune sera from IL-4(-/-) mice showed a dramatic increase in mouse AChR-specific IgG2a levels followed by a concomitant decrease in IgG1 levels, but these mice did not exhibit an accelerated disease. In conclusion, we have demonstrated for the first time that IL-4 is not required either for the generation of a pathogenic anti-AChR humoral immune response or for progression of clinical EAMG in mice.     

Prevention and treatment of experimental autoimmune myasthenia gravis with anti-rabbit thymocyte serum and gammaglobulin in rabbits

Twenty-two rabbits received Torpedo californica acetylcholine receptor (AChR) subcutaneously. Six rabbits were treated with 0.5 ml/kg/day of anti-rabbit thymocyte serum (ARTS), and four were treated with 5.0 mg/kg/day of anti-rabbit thymocyte gammaglobulin (ARTG) subcutaneously beginning concomitantly 1, 2 and 3 weeks after the initial AChR immunization. All 12 control animals died of experimental autoimmune myasthenia gravis (EAMG) within 52 days. None of the 10 treated rabbits developed clinical EAMG. ARTS- and ARTG-treated animals had significantly lower anti-AChR antibody titers than control animals at 3 weeks (pre-AChR booster, p < 0.01). At 6 weeks (post-AChR booster), only ARTS-treated animals had significantly lower titers than controls (p < 0.01). ARTS-treated animals developed sterile abscesses at injection sites, which were minimal in the ARTG-treated group. ARTS and ARTG prevent EAMG.     

The humoral response to xenografts is controlled by a restricted repertoire of immunoglobulin VH genes

BACKGROUND: The early phases of the host immune response to xenografts are dominated by anti-donor antibodies. The immunological pathways responsible for mediating the host humoral responses to xenografts are largely unknown, and this report addresses the nature of the immunoglobulin genes controlling the host antibody response to xenografts. METHODS: cDNA libraries established from rat anti-hamster monoclonal antibodies and splenic lymphocytes from LEW rats rejecting hamster heart xenografts were used to clone, sequence, and identify the immunoglobulin genes responsible for encoding rat xenoantibodies to hamster heart grafts. Libraries for germline variable region heavy chain (VH) genes encoding the anti-hamster xenograft antibodies were established by genomic DNA cloning and analyzed by nucleotide sequencing. The frequency of Ig VH gene usage for controlling the antibody responses to hamster xenografts was examined by colony-filter dot hybridization. The nucleic acid structure of these genes was then compared to their genomic progenitors to identify the number and structural diversity expressed by the Ig VH genes used to mediate the response. RESULTS: Rat monoclonal antibodies selected for their ability to precipitate the rejection of hamster xenografts exclusively use a closely related group of VH genes. The VH genes used by these antibodies are restricted to a single family of germline genes (VHHAR) for which 15 family members have been identified. The frequency of VHHAR gene usage in splenic IgM-producing B cells from LEW rats rapidly expands from 0.8% in naive animals to 13% in recipients 4 days after xenotransplantation. cDNA libraries expressing VHHAR genes were established from splenic lymphocytes derived from naive or xenograft recipients at 4 and 21 days after transplantation. Examination of 20 cDNA clones revealed that the majority (75%) of these clones express VHHAR genes displaying limited somatic mutation. CONCLUSIONS: The use of a closely related group of Ig VH genes in a germline configuration to control the early humoral response to xenografts suggests that this response may represent the utilization of a primitive, T cell-independent pathway of antibody production by the graft recipients.     

Biased VH family usage in primary gastric B cell lymphoma of mucosa-associated lymphoid tissue-type and the selected V beta expression of T cells infiltrating into the lymphoma cells

Seven cases of primary gastric low-grade B cell lymphoma of mucosa-associated lymphoid tissue (MALT) type, two cases of high-grade B cell lymphoma with a low-grade component and three cases of pure high-grade lymphoma were selected for the current study. The Ig VH gene use of lymphoma cells and the V beta repertoires of infiltrating T cells were investigated. The VH gene analysis showed multiple VH family usage in 12 cases, but the MALT-type lymphoma cell usage was found to be biased for the families that have a low number of VH genes (VHIV and V). Another analysis of lymphoma-infiltrating T cells showed restricted expressions of the V beta repertoire in all seven low-grade cases and three high-grade cases. In those 10 cases, a considerable number of CD4-positive T cells infiltrated into lymphoma cells and RAG-1 was also prominently expressed. Based on these findings, it was thus assumed that the normal counterpart of gastric B cell lymphoma of MALT type is different from the conventional B cell lymphoma, and the restricted expression of V beta repertoires is therefore considered to be a characteristic finding in low-grade B cell lymphomas of MALT type as well as in a proportion of high-grade lymphomas (the so called 'high-grade lymphoma of MALT type').     

The platelet-stimulating effect of adrenaline through alpha 2-adrenergic receptors requires simultaneous activation by a true stimulatory platelet agonist. Evidence that adrenaline per se does not induce human platelet activation in vitro

In the murine model for EAMG we investigated the relation between disease susceptibility and fine specificity of anti-AChR antibodies obtained from high susceptible C57Bl/6 and low susceptible BALB/c mice after immunization with Torpedo acetylcholine receptor (tAChR). Anti-AChR MoAbs with fine specificity for the main immunogenic region (MIR), the alpha-bungarotoxin (alpha-BT)/acetylcholine binding sites and other extra- and intracellular epitopes were isolated from both mouse strains. In total, nine out of 38 MoAbs obtained from C57Bl/6 mice were directed against extracellular epitopes on mouse AChR in contrast to only one out of 27 MoAbs from BALB/c mice. A difference in antibody repertoire may underlie the difference in pathogenic response observed between these mouse strains. These results indicate that strain-specific differences in disease susceptibility in murine EAMG may be related to differences in the available repertoire of potential pathogenic antibodies.     

Two high-affinity monoclonal IgG2a antibodies with differing thermodynamic stability demonstrate distinct antigen-induced changes in protein A-binding affinity

Injection of anti-AChR antibodies in passive transfer experimental autoimmune myasthenia gravis (EAMG) results in increased degradation of acetylcholine receptor (AChR) and increased synthesis of AChR alpha-subunit mRNA. Passive transfer of anti-Main Immunogenic Region (MIR) mAb 35 in aged rats does not induce clinical signs of disease nor AChR loss. The expression of the AChR subunit genes was analyzed in susceptible and resistant rats. In aged EAMG resistant rats, no increase in the amount of AChR alpha-subunit mRNA was measured. In vivo AChR degradation experiments did not show any increase in AChR degradation rates in aged resistant rats, in contrast to young susceptible rats. Taken together, these data demonstrate that resistance of the AChR protein to antibody-mediated degradation is the primary mechanism that accounts for the resistance to passive transfer EAMG in aged rats.     

Germline structure and differential utilization of Igha and Ighb VH10 genes

Ab heavy chains encoded by mouse VH10 genes have been of particular interest due to their frequent association with DNA binding. We reported previously that VH10 sequences are over-represented in the preimmune repertoire considering the apparent number of germline-encoded VH10 gene segments. In this report, we show that the VH10 family consists of three and two germline genes in the Igha and Ighb haplotypes, respectively. The complete nucleotide sequences of these five genes, including promoters and recombination signal sequences, were determined and allow unambiguous assignment of allelic relationships. The usage of individual VH10 genes varied significantly and ranged from 0.2% to an extraordinary 7.2% of the VH genes expressed by splenic B cells. Since the promoter and recombination signal sequence elements of all five VH10 genes are identical, we suggest that the few amino acid differences encoded by these five germline VH10 genes determine their representation in the preimmune repertoire. Rearrangements of the most frequently used VH10 gene have an apparent bias for histidine at position 95 of complementarity-determining region-3 (CDR3). These CDR3s are also biased for asparagine, an amino acid associated with the CDRs of DNA binding Abs. Together, these results suggest that high VH10 gene use is the result of B cell receptor-mediated selection and may involve DNA and/or ligands that share antigenic features with DNA.     

Preferential use of the VH4 Ig gene family by diffuse large-cell lymphoma

Ig heavy chain variable region (VH) genes expressed by human diffuse large-cell lymphoma (DLC) and follicular lymphoma (FL) were identified and analyzed with respect to germline gene families. In 67 cases of FL, VH region genes were expressed in a pattern similar to that of normal B cells, with a predominance of the large VH3 gene family being used. In contrast, of the 17 cases of DLC, there was an extremely biased use of VH genes. Of these DLC tumors, 88% expressed genes from the small VH4 gene family; and even among these tumors, there was a limited use of genes, with 11 cases producing Igs derived from the VH4.21 germline gene. Although most of the VH genes expressed by DLC tumor cells contained mutations with respect to their germline counterparts, almost all of these mutations occurred before the clonal expansion of the tumor. This contrasts with our previous findings of ongoing mutations in FL and represents a fundamental difference between these two malignancies. This preferential gene use implies an important role for the VH4 gene family, and specifically for VH4.21, in the genesis of DLC.     

Pathogenic autoimmunity to affinity-purified mouse acetylcholine receptor induced without adjuvant in BALB/c mice

Myasthenia gravis (MG) and experimental autoimmune myasthenia gravis (EAMG) are antibody-mediated disorders in which anti-acetylcholine receptor (anti-AChR) antibodies cause loss of muscle AChR and subsequent weakness. Many species are susceptible to induction of EAMG with purified xenogeneic AChR in adjuvant, but injection of Torpedo AChR without adjuvants can also induce evidence of EAMG. To see whether pathogenic autoimmunity could be induced in mice by isolated mouse AChR we injected BALB/c mice with several doses (1 pmole; about 0.1 microgram) of affinity-purified AChR (from the BC3H1 cell line but thought to be identical with denervated mouse muscle) intraperitoneally, without adjuvant, over a period of 10-22 weeks. Some of the mice became ill and died. High levels of serum anti-mouse AChR, directed mainly towards the main immunogenic region, were found and, in the survivors, correlated with loss of muscle AChR. Thus BALB/c mice can mount an autoimmune response to minute amounts of mouse AChR, without the use of adjuvants, and this response is very similar to that found in MG. This novel finding has implications regarding the etiology of the human disease.     

A single predominantly expressed polymorphic immunoglobulin VH gene family, related to mammalian group, I, clan, II, is identified in cattle

In order to understand the generation of antibody diversity in cattle, seven cDNAs, from heterohybridomas secreting bovine IgM and IgG1 antibodies, were cloned and structurally analyzed for rearranged bovine VDJ genes. All of the seven bovine VH genes, together with four available bovine VH gene sequences, shared a high nucleotide sequence homology (84.2-93.5%). Based upon the criteria of nucleic acid homology > or =80%, all of the bovine VH gene sequences isolated from the expressed antibody repertoire constitute a single VH gene family, which we have designated as bovine VH1 (Bov VH1). An analysis of 44 bovine IgM-secreting mouse x cattle heterohybridomas, originating from polyclonally-activated PBLs from bovine leukemia virus-infected cattle, revealed that all of these expressed Bov VH1 (100%) based upon DNA sequencing and Northern dot blot. The bovine VH genes showed highest DNA sequence similarity, ranging between 81.5 and 87.6%, with a single sheep VH gene family (related to human VH4) and are, thus, closest to the VH genes from another ruminant species. The Bov VH1 gene family is most homologous to the murine VH Q-52 (71.8-78%) and human VH4 (67.4-69.8%) gene families, which belong to mammalian group, I, clan, II, VH genes. The CDR3 length of rearranged bovine VDJ genes is characteristically long (15-23 amino acids). The bovine JH gene segments were most homologous to human JH4 (82.1-87.2%) and JH5 (84.6-89.7%) genes, suggesting the existence of at least two JH gene segments. An analysis of CDRs provides evidence that somatic hypermutations contribute significantly to the generation of antibody diversity in cattle. Southern blot analysis of BamH I, EcoR I and Hind III digested genomic DNA from four cattle breeds (Holstein, Jersey, Hereford and Charolais) revealed three RFLP patterns; the genomic complexity of Bov VH1 ranged between 13 and 15 genes. These observations provide evidence for polymorphism at the bovine Ig-VH locus, similar to that seen in mice and humans.     

TCR-Vbeta usage in the thymus and blood of myasthenia gravis patients

In myasthenia gravis (MG) the muscle acetylcholine receptor (AChR) is the target of an autoimmune response. The anti-AChR response may originate in the thymus, which is abnormal in most MG patients and contains anti-AChR T and B cells. Microbial superantigens (sAg) may trigger autoimmune responses and in this study we sought clues as to whether sAg play a role in the pathogenesis of MG. We investigated the frequency of use of the different TCR Vbeta families by the thymus and blood T cells in MG patients and in control subjects, using a multi-primer PCR assay. Identical TCR-Vbeta usage was found in the thymi of MG patients and controls, except Vbeta2, which showed a small increase in MG patients' thymi. Blood T cells of MG patients used Vbeta4, Vbeta6, Vbeta15, Vbeta16 and Vbeta24 significantly more than those of the controls. Vbeta4 and Vbeta6 are the gene families most frequently used by anti-AChR CD4(+) cells in MG patients. Blood T cells from MG patients used Vbeta12, Vbeta14, Vbeta17 and Vbeta18 significantly less than controls. MG patients used Vbeta4 and Vbeta6 significantly more in the blood than in the thymus, while the opposite occurred for Vbeta7, Vbeta12 and Vbeta14. Controls used Vbeta17 more and Vbeta24 less in the blood than in the thymus. The preferential expansion of Vbeta4 and Vbeta6 in MG patients might reflect the immunodominance of certain AChR epitopes, or the action of a sAg outside the thymus. The minimal differences in the TCR-Vbeta usage in the blood and thymus of control subjects might be due to expansion of T cell clones specific for common antigens. Identical Vbeta usage in the thymi of MG patients and controls does not support an important role of the thymus as the location of anti-AChR sensitization when MG is clinically evident. The differences observed in the Vbeta usage in blood and thymi of MG patients are likely to be due to preferential Vbeta usage by the anti-AChR T cells in the blood.     

Molecular basis for the interaction between human IgM and staphylococcal protein A

To examine the relationship between VH gene usage and reactivity of immunoglobulins, we cloned B cells from peripheral blood from adults and from human neonatal cord blood by EBV transformation. Nearly one-third of the B cell clones from both sources produced IgM reactive with staphylococcal protein A (SPA). None of such IgM reacted with other antigens, except for the crude extract of Staphylococcus aureus. All of 22 B cell clones producing IgM reactive with SPA expressed VH3 genes, while none of the control 15 clones secreting IgM nonreactive with SPA expressed VH3. The IgM proteins reactive with SPA could be clearly divided into two subjects based on the differential binding avidity to solid-phase SPA. Both kappa and lambda light chains were used in each subset of SPA-reactive IgM. Sequence analysis of the PCR products from seven VH3-IgM clones revealed that the VH3 genes were used in nearly germline configuration. The D and J gene usage was diverse. Comparison of amino acid sequences between antibodies with high and low avidity to SPA suggests that the differential avidity is related to amino acid sequence differences in the complementarity determining region 2 and framework region 3. The high frequencies of B cells committed to the production of SPA-reactive IgM in normal blood and the restricted use of VH3 heavy chain genes in nearly germline configuration in these cells support the notion that SPA behaves like a superantigen toward human B cells.     

Deletional mapping of fifteen mouse VH gene families reveals a common organization for three Igh haplotypes

In addition to the content of germ-line variable gene segments, the organization of V genes has been implicated in the development of the Ab repertoire. We have searched the expressed VH genes of BALB/c mice for additional VH gene families and utilized deletion mapping to explore the extent of VH gene family interspersion. We have identified and characterized one new VH gene family (VH15) and extended our previous studies of the Igha and Ighb haplotypes to include a third haplotype (Ighj) using a newly developed panel of pre-B cell lines (CXCB cell lines). We conclude that the Igha, Ighb, and Ighj haplotypes have a similar Igh-V locus structure. A refined deletional map for 15 VH gene families and an individual member of the VHSM7 family (H10) has been constructed based on the deletion profiles of 72 rearranged heavy chain loci. These results demonstrate previously unrecognized examples of interspersion among members of the VHS107, VH10, and VHSM7 families.     

Pharmacokinetic mechanisms for obtaining high renal coelimination of phencyclidine and a monoclonal antiphencyclidine antigen-binding fragment of immunoglobulin G in the rat

Myasthenia gravis (MG) is a neuromuscular disorder of man caused by a humoral response to the acetylcholine receptor (AChR). Most of the antibodies in MG and in experimental autoimmune myasthenia gravis (EAMG) are directed to the extracellular portion of the AChR alpha subunit, and within it, primarily to the main immunogenic region (MIR). We have cloned and expressed recombinant fragments, corresponding to the entire extracellular domain of the AChR alpha subunit (H alpha1-210), and to portions of it that encompass either the MIR (H alpha1-121) or the ligand binding site of AChR (H alpha122-210), and studied their ability to interfere with the immunopathological anti-AChR response in vitro and in vivo. All fragments were expressed as fusion proteins with glutathione S-transferase. Fragments H alpha1-121 and H alpha1-210 protected AChR in TE671 cells against accelerated degradation induced by the anti-MIR monoclonal antibody (mAb)198 in a dose-dependent manner. Moreover, these fragments had a similar effect on the antigenic modulation of AChR by other anti-MIR mAb and by polyclonal rat anti-AChR antibodies. Fragments H alpha1-121 and H alpha1-210 were also able to modulate in vivo muscle AChR loss and development of clinical symptoms of EAMG, passively transferred to rats by mAb 198. Fragment H alpha122-210 did not have such a protective activity. Our results suggest that the appropriate recombinant fragments of the human AChR may be employed in the future for antigen-specific therapy of myasthenia.     

Three-point correlation functions in uniformly and randomly driven diffusive systems

We have previously shown that human antibody (Ab) directed against the capsular polysaccharide of the important bacterial pathogen, Haemophilus influenzae type b (Hib) is encoded by a small group of VH3 gene family members. The majority of anti-Hib PS Ab use members of the smaller VH3b subfamily. To examine directly the available human VH3 repertoire, we have used PCR to amplify and clone candidate germ-line VH3b H chain V region genes from two unrelated subjects from whom anti-Hib polysaccharide mAb had been previously obtained. A single functional VH3b germ-line gene was obtained from one subject. This gene is identical throughout the coding region to the previously identified gene 9.1. Twelve distinct VH3b germ-line sequences, 87.6-99.8% homologous to one another, were obtained from the second subject. One of these genes, LSG1.1, is also identical to the 9.1 germ-line gene, and a second, LSG6.1, is identical to a previously reported cDNA, M85. These germ-line VH3b genes are 82.7-94.1% homologous to rearranged anti-Hib PS VH3b segments obtained from these subjects. Our findings further demonstrate that considerable polymorphism of VH segments exists in the human population. Despite the presence of very highly homologous VH elements in the germ line, particular genes are highly conserved within the outbred human population.     

Polymorphism at the HLA-DQ locus determines susceptibility to experimental autoimmune myasthenia gravis

Studies in myasthenia gravis (MG) patients demonstrate that polymorphism at the HLA-DQ locus influences the development of MG. Several studies using the mouse models also demonstrate the influence of class II molecules, especially the H2-A, which is the mouse homologue of HLA-DQ, in experimental autoimmune myasthenia gravis (EAMG). We used transgenic mice expressing two different DQ molecules, DQ8 (DQA1*0301/B1*0302) and DQ6 (DQA1*0103/B1*0601), to evaluate the role of HLA-DQ genes in MG. These mice do not express endogenous mouse class II molecules since they contain the mutant H2-A beta0 gene. The mice were immunized with Torpedo acetylcholine receptor, and EAMG was assessed by clinical evaluation and was confirmed by electrophysiology. Clinical scores for EAMG were highest in HLA-DQ8 transgenic mice, whereas the scores of HLA-DQ6 mice rarely exceeded grade 1. There was no incidence of EAMG in class II-deficient (H2-A beta0) mice. These results demonstrate that polymorphism at the HLA-DQ locus affects the incidence and the severity of EAMG. The manifestation of susceptibility to EAMG in the context of human class II molecules underscores the important roles of these molecules in the initiation and perpetuation of EAMG.