Identification of autoantigens in multiple sclerosis

Abstract

Multiple sclerosis (MS) is an autoimmune disorder of the central nervous system in which cells from primarily the adaptive immune system infiltrate the brain and spinal cord, leading to inflammation and demyelination. Debuting primarily between 20-40 years of age and with a prevalence in Sweden of ~0.2%, it is one of the leading causes of disability in working-age adults. While the cause is still unknown, the risk of developing MS is influenced by an interplay of both genetic and environmental risk factors. Genetic and immunological data point towards CD4+ T cells being a primary driver of the disease. While some de facto targets, i.e., autoantigens, have been identified, the known autoantigen repertoire still contains considerable gaps. This remains a critical problem for developing autoantigen-targeted diagnostic tools and autoantigen-specific treatment strategies. This thesis aimed to identify novel autoantigens in MS.

Paper I addressed a common problem when studying autoantigen-specific T-cell responses: autoreactive T cells are rare, and antigens can be either weak stimulators or contain contaminants that are challenging to remove, resulting in high assay noise and low sensitivity. By covalently coupling recombinant protein antigens to 1 μm paramagnetic polystyrene beads, we show that contaminants can be removed while the ability to stimulate T-cell responses remains. This resulted in a sensitive assay with high signal-to-noise ratios, with a threshold for detection at 1 in 18 000 cells.

In Paper II, we used this novel method to examine T cell responses to myelin oligodendrocyte glycoprotein (MOG), an autoantigen for which previous results have conflicted. By examining peripheral blood mononuclear cells (PBMCs) from a cohort of persons with MS (pwMS) and matched healthy controls (HC), MOG-specific CD4+ T cells were detected in approximately half of all pwMS. Additionally, MOG-epitopes were presented by monocytes and restricted to HLA-DR. Lastly, using three different antibody-assays, we could not detect any significant portion of MOG-specific autoantibodies despite the presence of MOG-specific T cells.

Paper III addressed the main aim of this thesis, i.e., identifying novel autoantigens. This study combined the antigen-bead method with the Human Protein Atlas recombinant protein epitope signature tag library to screen for T-cell reactivity against a panel of 63 central nervous system-expressed proteins. In a smaller screening cohort, there were increased proinflammatory responses against four novel autoantigens targets: fatty acid binding protein 7 (FABP7), prokineticin-2 (PROK2), reticulon-3 (RTN3), and synaptosome associated protein 91 (SNAP91), as well as the previously described autoantigen MOG in pwMS. The screening results were validated using full-length versions of the targets in two larger cohorts, including pharmacologically untreated pwMS. The autoreactive profiles of individuals were heterogenous, but a panel of several autoantigens could distinguish between MS and non-MS with high accuracy. Immunophenotyping revealed MS-specific autoreactive cells to be mainly HLA-DR-restricted CD4+ T cells and responded with interferon-gamma and granulocyte-macrophage colony-stimulating factor production upon stimulation. The presence of autoantibodies was examined in a large cohort of patients and controls. Still, it was not increased in MS. Immunization of mice with the novel autoantigens induced T cell responses, leading to CNS-leukocyte migration and crossing of the blood-brain barrier, demonstrating encephalitogenic potential.

Paper IV explored a possible immunological link between Epstein-Barr virus infection and MS. We examined serological responses to alpha-crystallin B (CRYAB) and Epstein-Barr virus nuclear antigen 1 (EBNA1) in a cohort of 713 pwMS and 722 HC. Anti-CRYAB-antibodies were associated with MS with an odds ratio (OR) of 2.0, which had a synergistic effect with high EBNA1 responses (OR of 9.0). By depleting plasma of anti-EBNA1 antibodies, CRYAB responses were similarly removed, demonstrating cross-reactivity between the two antigens due to an amino acid sequence homology (RRPFF, CRYAB aa11-15 and EBNA1 aa402-406 respectively). In a mouse model, EBNA1-primed T cells were also CRYAB-reactive, and EBNA1 and CRYAB-responsive T cells were highly correlated and increased in natalizumab-treated pwMS, pointing towards a similar cross-reactivity in the T-cell compartment as well.

In conclusion, this thesis presents methods for sensitively assessing autoreactive T-cell responses, reexamining and confirming MOG and CRYAB as targets. It considerably expands the knowledge regarding the targets of the autoimmune attack in MS by adding four novel autoantigens to the known repertoire. Further, it demonstrates an underlying heterogeneity of the immunological landscape of MS and provides a mechanistic link between Epstein-Barr virus and MS. It demonstrates a first step in the development of autoantigen-specific methods for diagnostics and introduces novel targets for potentially effective antigen-specific immunotherapy.