A model of arthritis was established by the injection of type II collagen into mice. Only mice bearing the H-2q haplotype were susceptible to the disease. Susceptibility was further mapped by the use of recombinant strains on the Iq locus. Type II collagen arthritis was observed in the (resistant X susceptible) F1 cross. Mice strains were designated high, intermediate, or low responders with respect to the anti-type II antibody levels measured by radioimmunoassay. Arthritis-susceptible strains were all classified as high antibody responders. The clinical and histological appearance of type II collagen arthritis in the mouse indicates that it may be a good animal model for the investigation of various immunogenetic traits in rheumatoid arthritis.

Immunization of DBA/1 mice with native chick type II collagen resulted in development of polyarthritis 4-5 wk later. Sera of these mice contained high levels of anticollagen antibodies, and immunoglobulin concentrates of their sera transferred arthritis to unimmunized recipients. Histopathologically, this passively transferred arthritis resembled the early disease of immunized donors. Immunofluorescence studies demonstrated the deposition of IgG and C3 on the articular surface but not in synovial tissue of arthritic joints. Transferred, isotopically labeled anticollagen antibodies rapidly localized to the limbs and to other cartilage-containing tissues. When transfer concentrate was administered to arthritis-resistant strains, they also developed arthritis. Indeed, immunoglobulin concentrates from rats with collagen-induced arthritis transferred arthritis to naive mice. The amount of concentrate required for transfer to B10.D2 resistant mice was reduced by immunizing them with collagen 4 wk before transfer. Although susceptibility to arthritis from immunization is H-2 linked, these studies clearly demonstrate that passive transfer of arthritis depends upon injection of specific antibody and not on other host factors.

One hypothesis that couples infection with autoimmune disease is molecular mimicry. Molecular mimicry is characterized by an immune response to an environmental agent that cross-reacts with a host antigen, resulting in disease 1,2 . This hypothesis has been implicated in the pathogenesis of diabetes, lupus and multiple sclerosis (MS) 1-4 . There is limited direct evidence linking causative agents with pathogenic immune reactions in these diseases. Our study establishes a clear link between viral infection, autoimmunity and neurological disease in humans. As a model for molecular mimicry, we studied patients with human T-lymphotropic virus type 1 (HTLV-1)-associated myelopathy/ tropical spastic paraparesis (HAM/TSP), a disease that can be indistinguishable from MS (refs. 5 -7 ). HAM/TSP patients develop antibodies to neurons 8 . We hypothesized these antibodies would identify a central nervous system (CNS) autoantigen. Immunoglobulin G isolated from HAM/TSP patients identified heterogeneous nuclear ribonuclear protein-A1 (hnRNP-A1) as the autoantigen. Antibodies to hnRNP-A1 cross-reacted with HTLV-1-tax, the immune response to which is associated with HAM/TSP (refs. 5 ,9 ). Immunoglobulin G specifically stained human Betz cells, whose axons are preferentially damaged 7 . Infusion of autoantibodies in brain sections inhibited neuronal firing, indicative of their pathogenic nature. These data demonstrate the importance of molecular mimicry between an infecting agent and hnRNP-A1 in autoimmune disease of the CNS. To test for molecular mimicry between an environmental agent and the central nervous system (CNS), we isolated immunoglobulin G (IgG) from the serum of patients with human Tlymphotropic virus type 1 (HTLV-1)-associated myelopathy/tropical spastic paraparesis (HAM/TSP) and tested it for reactivity with human tissues (Fig. 1a). There was intense staining of neurons in brain and no staining of glia, dorsal root ganglion (peripheral nervous system) or systemic organs. A monoclonal antibody to HTLV-1-tax (tax mAb) mimicked IgG staining of neurons. To identify the protein, cortical neurons were isolated, proteins extracted and subjected to SDS-PAGE and western-blot analysis. The IgG recognized a band of approximately 33 kD (Fig. 1b), whereas IgG isolated from controls did not. Importantly, the tax mAb that stained CNS neurons reacted with the antigen. All patients with HAM/TSP (13/13) developed antibodies recognizing the neuronal antigen 8 . Nine of ten HTLV-1-seropositive patients without neurological symptoms and 12 HTLV-1-seronegative controls showed no reactivity (P < 0.0001 versus HAM/TSP) 8 . Clinically, the HAM/TSP patients presented with progressive neurological disease in which corticospinal tract damage (weakness, spasticity and pathological reflexes) predominated 6 . Many of our patients were originally diagnosed with MS. In fact, one of our patients was diagnosed with MS for 20 years before HTLV-1 testing 10 .To establish a direct link between the immune response to HTLV-1 and the neurological...

Rheumatoid arthritis (RA) is an autoimmune disease that is strongly associated with the expression of several HLA-DR haplotypes, including DR1 (DRB1*0101). Although the antigen that initiates RA remains elusive, it has been shown that many patients have autoimmunity directed to type II collagen (CII). To test the hypothesis that HLA-DR1 is capable of mediating an immune response to CII, we have generated transgenic mice expressing chimeric (human/ mouse) HLA-DR1. When the DR1 transgenic mice were immunized with human CII (hCII), they developed a severe autoimmune arthritis, evidenced by severe swelling and erythema of the limbs and marked inflammation and erosion of articular joints. The development of the autoimmune arthritis was accompanied by strong DR1-restricted T and B cell responses to hCII. The T cell response was focused on a dominant determinant contained within CII(259–273) from which an eight amino acid core was defined. The B cell response was characterized by high titers of antibody specific for hCII, and a high degree of cross-reactivity with murine type II collagen. These data demonstrate that HLA-DR1 is capable of presenting peptides derived from hCII, and suggest that this DR1 transgenic model will be useful in the development of DR1-specific therapies for RA.