Skip to main content

CNS autoimmunity mediated by pathogenic T cell and antibody responses: immunogenetics, pathogenesis and therapy of MOG-induced EAE

Deliverables

Multiple sclerosis (MS) is a common inflammatory/demyelinating disease of the central nervous system (CNS) in Europe and current treatments are only modestly beneficial. Myelin oligodendrocyte glycoprotein (MOG) is potentially relevant as target auto-antigen in MS. It can recruit both an encephalitogenic T cell response and a demyelinating B cell response. It is therefore of interest to explore the immunopathogenesis of MOG induced experimental autoimmune encephalomyelitis (EAE), its potential therapy and the role of MOG directed autoimmunity in MS.To reveal mechanisms of importance for MS, EAE models should mimic human disease as close as possible. The project therefore aimed at establishing such models with the use of MOG as inducing auto-antigen in rats. We established MOG induced EAE in different rat strains, being more MS like than any previously described rodent model. These rats displayed neuropathological lesions mimicking those found in MS and a chronic relapsing disease course. We have documented that the immunopathogenesis depend on both a pathogenic T and B cell response against critical epitopes of MOG, in turn controlled by the MHC haplotype. Therapeutic strategies need to take care of both arms of the autoimmunity. Thus, tolerogenic strategies dampening a pathogenic T cell response against MOG, under certain circumstances enhanced the B cell response, with aggravated disease outcome as result. Molecular mimicry with a milk protein, butyrophilin, suggested potential effects of milk products on MOG autoimmunity. Experiments in Fc receptor gene deleted mice demonstrated a crucial role of this receptor in autoimmune neuro-inflammation. Tolerogenic strategies to limit disease progression in MOG induced EAE: The results obtained during the three year period strengthen our initial hypothesis that the development of an MS-like pathology in MOG-induced EAE is dependent on the induction of both MOG-specific T and B cell responses. Functionally, the pathogenic T cell response initiates inflammation in the CNS, recruiting immune effector cells and disrupting the blood brain barrier, allowing antibody and other serum proteins to enter the CNS compartment. MOG-specific antibodies then initiate demyelination and further amplify the local inflammatory response by the production of complement derived pro-inflammatory factors. As a consequence of this antibody/complement driven inflammatory cascade, severe clinical disease and extensive demyelination develop in MOG-EAE even if the MOG-specific T cell response is minimal and by itself unable to induce disease (partner 2 and 3, Stefferl et al,1999). Analysis of the fine specificity of the demyelinating MOG-specific antibody response demonstrated that it recognises a restricted repertoire of conformation dependent, non-linear epitopes that are conserved between species (partners 1, 2 and 3, Brehm et al 1999, Lindert et al,1999). In contrast, the identity of the epitopes recognized by the encephalitogenic Th T cell response is defined by the MHC class II haplotype and is therefore species/strain specific (partners 1, 2 and 3, Weissert et al 1998, Stefferl et al 1999, Abdul-Majid et al,2000). The duality of the pathogenic autoimmune response to MOG imply that successful tolerogenic and/or therapeutic strategies must efficiently control both T cell and antibody dependent effector mechanisms (Partners 1, 2 and 3). Ongoing experiments are investigating the tolerogenic potential of treatment with soluble full length MOG and MOG-derived peptides in both prophylactic and therapeutic settings. The data obtained so far indicate that regardless of the route of application (intravenous, intra peritoneal, nasal or oral) exposure to the intact antigen can down modulate (but not abolish) the encephalitogenic T cell response, but paradoxically stimulate a pathogenic antibody response. Cessation of treatment is then followed by an increase in disease activity that is often fulminant and lethal. However the use of truncated recombinant MOG proteins or synthetic peptides avoids the induction/enhancement of pathogenic antibody responses and can completely inhibit disease induction by specifically targeting the encephalitogenic T cell response (Borquin et al2000; Lobell et al. 2000). The importance of avoiding the induction/enhancement of pathogenic anti-MOG antibody responses is also seen in studies analysing the therapeutic potential of DNA vaccination in MOG-EAE. In SJL/J mice DNA vaccination with constructs encoding the full length protein resulted in the induction of a high titer pathogenic antibody response that exacerbated EAE induced by either MOG or irrelevant myelin antigens such as PLP (Borquin et al,2000). In contrast, constructs encoding the immuno-dominant T cell epitopes of MOG for Lew.1AV1 and Lew.1N rats specifically targeted the effector T cell response in these rat strains and protected them against MOG peptide induced disease (Partners 1, 2 and 3, Lobell et al,2000). These results imply that universal strategies to re-establish self-tolerance to MOG and suppress disease activity in MOG-EAE will be successful if they efficiently suppress the effector T cell response while avoiding the induction/enhancement of a demyelinating antibody responses. In view of the importance of the antibody response in potentiating disease activity we are currently using site directed mutagenesis to identify those amino acid residues responsible for the maintaining the structure of the conformational epitopes targeted by demyelinating antibodies. We anticipate that this will lead to the development of a generation of modified MOG derivatives that can be used to block the effector T cell response across a broad range of MHC class II haplotypes (Partners 1, 2 and 3). Modulation of MOG specific autoimmune responses by environmental factors: The etiology of multiple sclerosis (MS) is believed to involve environmental factors that disrupt immunological self-tolerance to CNS myelin in genetically susceptible individuals, but the identity and mode of action of the factors involved remain unknown. Intriguingly, sequence homologies involving the extracellular immunoglobulin-like domain of MOG (MOGIgd) identified MOG as a member of an extended family of "B7-like"proteins. Of particular interest was the observation that the highest level of sequence identity was with a homologous extracellular Ig domain of butyrophilin (BTN) (Stefferl et al,2000), a major protein of the milk fat globule membrane (MFGM). This observation led us to speculate that immunological cross-reactivity or "molecular mimicry" with BTN may influence the function of the MOG-reactive autoimmune repertoire. We demonstrated that this was indeed the case, identifying for the first time a dietary antigen that can have a significant impact on the functional activity of the autoimmune response to an immunodominant myelin antigen (Stefferl et al., 2000). In the Dark Agouti (DA) rat active immunization with native BTN triggers an inflammatory response in the CNS characterized by the formation of scattered meningeal and perivascular infiltrates of T cells and macrophages. This pathology was mediated by a MHC class II restricted T cell response that recognizes a homologous peptide sequence in MOG and BTN. Conversely, molecular mimicry with BTN can also be exploited to suppress disease activity in MOG-induced EAE. MOG-induced EAE can be suppressed in DA rats by both i.v. and trans-mucosal treatment with BTN (Stefferl et al 2000). This observation suggests that dietary exposure to BTN may modulate the MOG-specific autoimmune response in man. We therefore initiated an investigation of molecular mimicry between MOG and BTN in normal healthy controls and patients with MS. Western blotting and ELISA studies identified anti-BTN IgG antibodies in approximately 30% of all donors irrespective of their disease status. Isolating MOGIgd-specific antibodies from seropositive donors and testing their reactivity with a panel of BTN peptides by ELISA determined whether or not this response cross-reacted with MOG. We found that MOG-specific antibodies isolated from approximately 60% of the MS patients cross-reacted with one or more BTN peptides. In contrast, anti-MOG antibodies isolated from the pooled sera of approximately 3000 healthy donors failed to bind to any BTN peptides, although they recognized a variety of MOG peptide epitopes (Guggenmos et al, 2000). This result identifies for the first time an enhanced level of immunological cross-reactivity between a candidate auto-antigen and a common environmental factor in multiple sclerosis. The functional relevance of this enhanced cross-reactivity is now being investigated at the level of the T cell response to BTN. Key regulatory immunological mechanisms as examined in gene deleted mice Immunisation of DBA/1 mice with MOG1-125 /CFA induces EAE after 9 days. Pathology reveals extensive demyelination with infiltration of macrophages and T cells and a MOG-specific antibody response. We assessed the in vivo consequences of deficiency of receptors for immunoglobulin (FcgRs) which are important for antibody-mediated immunity in this mouse model of multiple sclerosis. were immunised id with MOG/CFA. Clinical EAE scores were monitored in female DBA/1 mice lacking FcgRI/ FcgRIII or FcgRII, serum levels of anti-MOG antibodies were determined and MOG-specific T cell function addressed in vitro by cellular recall assays and in vivo by adoptive transfer. Degree of CNS pathology was determined histologically. All mice developed comparable T and B cell specific immune responses following MOG immunisation, as assessed by anti-MOG ELISA, cell proliferation and the ability of T cells from knockout mice to transfer EAE to naive wild-type recipients. Mice lacking FcgRI / FcgRIII did not develop EAEas assessed clinically and histologically. Mice lacking FcgRII had a trend to a worsened disease. We conclude that distinct FcgRs differentially regulate MOG-EAE susceptibility in DBA/1 mice (Abdul-Majid et al 2000). In the same DBA/1 mouse model, we had access to mice deleted of their CD 4 and CD8 genes. This is of interest since the vast majority of studies in MS and EAE has focused on pathogenic CD4 + cells. Recently there are genetic evidence suggesting that the MHC class I genotype in humans may modulate the risk/course of MS. Similarly, there are scattered reports in the literature on MHC class I regulation of EAE. Interestingly, we could document a regulation of MOG EAE by CD 8+ cells, since CD4 -/- mice as well as CD8-/- mice developed an ameliorated form of EAE as compared to their littermate wild type controls. Antibody depletions confirmed this conclusion (Wefer et al. 2000).
Myelin oligodendrocyte glycoprotein (MOG) is potentially relevant as target auto-antigen in MS. It can recruit both an encephalitogenic T cell response and a demyelinating B cell response. It is therefore of interest to explore the genetic regulation of MOG induced experimental autoimmune encephalomyelitis (EAE). We established MOG induced EAE in different rat strains, being more MS like than any previously described rodent model. These rats displayed neuropathological lesions mimicking those found in MS and a chronic relapsing disease course. There was a striking genetic regulation by genes within and outside the MHC as in human MS. Eight non-MHC genome regions have been defined. These are now subject for positional cloning attempts and synteny studies in human MS materials. Definition of susceptibility genes: The importance of MOG-induced EAE as an animal model of MS was once again underlined by the induction of the full pathogenic spectrum of MS in the experimental setting (P1, 2 and 3, Storch et al,1998) This highlights the suitability of our model system for all following investigations. 1. Mapping of non-MHC genes (P1, 2, 3): To examine whether the genome regions identified for MOG-EAE also regulate other models of autoimmune disease, a whole genome scan was performed on the DAxACI F2 cross using peripheral nerve myelin (PNM) induced experimental autoimmune neuritis (EAN) (P1, Dahlman et al 2000). Interestingly, susceptibility to this disease segregated in the F2 cross with high levels of anti PNM IgG2b and IgG2c isotype levels, supporting the hypothesis that susceptibility genes for organ specific inflammatory disease influence auto-antigen specific T helper 1 and T helper 2 differentiation (P1). Importantly, we identified genome regions that regulate both EAN and EAE on chromosome 4,10, 12, and 18, as well as one EAN selective genome region on chromosome 17. To identify additional genome regions regulating MOG EAE we performed a large F2 cross using MHC-matched Lew.1AV1x PVG.1AV1 rat strains. The Lew rat is the most commonly used EAE permissive rat strain and we identified five further QTL’s on chromosome 1, 7, 13, 17 and X predisposing for MOG-EAE. Functional analysis of these QTL's revealed an influence on chronicity of disease in one region (Becanovic 2000). Apart from representing the starting point for the identification of more disease regulating genes, the data are of interest also in another sense. Both the Lewis and the DA rat strains are EAE permissive, but due to different genom regions. This clearly demonstrates a genetic heterogeneity in the regulation of an MS like disease. Thus completely different genes can dispose for the same final inflammatory neurological disease. 2. Experimental strategies to define MHC genome region involvement (P1, 2, 3): We used rat strains with a constant EAE permissive background genome and differing MHC haplotypes selected to provide maximum MHC class II gene diversity. The studies were performed cooperatively (P1 and P2, animal experiments, MHC genetics, immunological analysis, P3 detailed neuropathology), and led to new perspectives on MHC influences on a chronic MS like disease (Weissert et al,1998). In addition, we have expanded our animal model to studies in the mouse to make advanced genetic tools available for future research (P1 and P3, Abdul-Majid, 2000). We first reported a hierarchical and allele-specific influences of the MHC haplotype on disease susceptibility, recruitment of MOG specific immunocompetent cells, clinical course and CNS pathology. Major haplotype specific effects mapped to the class II gene region. However, these effects were modified by genes both in the class I and III regions (Stefferl et al,1999). These results formed the basis for further studies on genetic and mechanistic MHC regulation of MOG EAE.
Multiple sclerosis (MS) is a common inflammatory/demyelinating disease of the central nervous system (CNS) in Europe and current treatments are only modestly beneficial. Myelin oligodendrocyte glycoprotein (MOG) is potentially relevant as target autoantigen in MS. It is therefore of interest to explore the immunopathogenesis of MOG induced experimental autoimmune encephalomyelitis (EAE), and the role of MOG directed autoimmunity in MS. In addition, axonal damage is important for permanent neurological deficits in MS and therefore subject of study in this project. Pathogenic mechanisms of anti MOG B and T cell responses in genetically defined congenic rat strains were sought, and we studied the potential role of such responses in human MS: We established MOG induced EAE in different rat strains, being more MS like than any previously described rodent model. These rats displayed neuropathological lesions mimicking those found in MS and a chronic relapsing disease course. Experimentally, we have documented that the immunopathogenesis depend on both a pathogenic T and B cell response against critical epitopes of MOG, in turn controlled by MHC and non-MHC genes. Therapeutic strategies need to take care of both arms of the autoimmunity. In human MS, there are increased numbers of MOG autoimmune T cells and increased levels of antibodies directed against MOG of potential relevance for disease. MOG EAE resulted in axonal damage mimicking that seen in MS. A profound expression of an alfa 1B subunit was demonstrated in dystrophic axons, suggesting an important role in axonal degeneration, and a possible target for therapy. MOG-specific immune responses in cerebrospinal fluid (CSF) and blood of humans: The MOG specific T cell repertoire have been analysed by using overlapping peptides of human MOG. Peptide specific responses were quantified with short term in vitro stimulation and measurement of single cells producing interferon gamma in response to the peptides (Partners 1 and 2). This assay enables measurement of low frequencies of autoreactive T cells and is 20-50 fold more sensitive than traditional limiting dilution techniques. It furthermore avoids any long-term artefacts that may appear in T cell cloning procedures. HLA-DR2 increases the risk for MS. The responses were measured in HLA DR 2 + MS patients and in HLA DR 2 + healthy controls. MS patients displayed higher frequencies of MOG reactive T cells than healthy controls, consistent with a role of MOG autoimmunity also in human disease. One immunodominant MOG epitopes, MOG 63-87, could be identified in DR2 + patients. This finding is of importance for the design of immunoselective therapy in the future (Wallstrfm et al. 1998). We are currently fine mapping this region using 12 amino acid long overlapping peptides. Preliminary results from 20 patients analysed so far revealed a number of important T cell epitopes within the MOG 63-87 region. In addition, we are analysing another important HLA haplotype, namely DR3. Preliminary results reveal an increased frequency of MOG reactive T cells in DR3+ patients as compared to HLA matched controls and suggest broad MOG peptide recognition. These studies were paralleled by an analysis of the MOG-specific T cell repertoire using T cell lines selected from MS patients and healthy controls (Partner 2,Lindert et al,1999). The T cell lines were analysed with respect to epitope specificity, MHC restriction and T cell receptor gene usage. These studies revealed that MOG reactive T cells are a normal component of the healthy immune repertoire. These T cell lines were found to be heterogeneous with respect to all parameters investigated and no differences were observed between cell lines derived from control and MS donors. Intriguingly however these MOG-specific T cell lines differed from a panel of MBP-specific T cell lines in that the majority exhibited a Th0/Th2 cytokine phenotype (low IFN_, high levels of IL-4, IL-10), rather than a Th1-like phenotype (high IFNg, low levels of IL-4 and IL-10). This observation suggests that these antigen specific responses differ with respect to their function in vivo. MOG-specific autoantibodies in multiple sclerosis: Our previous studies demonstrated that MOG-specific autoantibodies mediate demyelination in both rodent and primate models of multiple sclerosis. Analysis of sera obtained from MS patients and appropriate controls revealed that although the frequency of seropositive donors was highest within the MS cohort (60-80%), significant MOG-specific antibodies titres were also found in patients with non-demyelinating CNS diseases (30-40%) as well as in 5-18% of normal healthy controls (Reindl et al,1999, Lindert et al,1999). These observations reveal that CNS injury can result in the loss of B cell self tolerance to MOG and that this is most pronounced in patients with chronic inflammatory CNS disease. It was therefore imperative to develop a method to differentiate pathogenic (demyelinating) from non-pathogenic antibody responses to MOG in patients with MS. An approach that may provide a useful diagnostic, or para-clinical marker in patients with MS. Our experiments in experimental rodents revealed that antibody mediated demyelination in MOG-induced experimental autoimmune encephalomyelitis (EAE) was associated with a conformation dependent antibody response. This antibody response could be identified by FACS by incubating human MOG transfected target cells with dilutions of the sera under investigation (Brehm et al,1999). In experimental animals this pathogenic antibody response could only be induced by sensitisation with the intact extracellular domain of MOG, or alternatively vaccination with MOG encoding DNA constructs (Bourquin et al., 2000; Brehm et al., 2000). Strikingly, immunisation with synthetic MOG peptides or truncated MOG fusion proteins was able to induce an inflammatory T cell mediated response in the CNS, but failed to induce a pathogenic demyelinating antibody response (Brehm et al., 2000). Analysis of MOG-specific antibodies isolated from MS patients and normal, but seropositive, healthy controls revealed that both groups recognise synthetic MOG peptides by ELISA, only one of the seventeen MS samples contained antibodies that recognised the native protein in vitro (Haase et al,2000). This surprising observation has two important implications. First, potentially pathogenic MOG-specific autoantibodies are only present in a small proportion of MS patients and secondly the under-representation of this conformation-dependent response suggests that the former response is under far more stringent regulatory control. The identification of additional targets for autoantibody mediated demyelination and the regulatory mechanisms that control these responses are a major priority for future research.

Searching for OpenAIRE data...