SummaryThe development of autoimmune disease is accompanied by the acquired recognition of new self-determinants, a process commonly referred to as determinant spreading. In this study, we addressed the question of whether determinant spreading is pathogenic for progression of chronic-relapsing experimental autoimmune encephalomyelitis (EAE), a disease with many similarities to multiple sclerosis (MS). Our approach involved a systematic epitope mapping of responses to myelin proteolipid protein (PLP) as well as assaying responses to known encephalitogenic determinants of myelin basic protein (MBP 87-99) and myelin oligodendrocyte g|yco-protein (MOG 92-106) at various times after induction of EAE in (SWRXSJL)F1 mice immunized with PLP 139-151. We found that the order in which new determinants are recognized during the course of disease follows a predictable sequential pattem. At monthly intervals after immunization with p139-151, responses to PLP 249-273, MBP 87-99, and PLP 173-198 were sequentially accumulated in all mice examined. Three lines of evidence showed that determinant spreading is pathogenic for disease progression: (a) spreading determinants mediate passive transfer of acute EAE in naive (S'W~• 1 recipients; (b) an invariant relationship exists between the development of relapse/progression and the spreading of recognition to new immunodominant encephalitogenic determinants; and (c) after EAE onset, the induction ofpeptidespecific tolerance to spreading but not to nonspreading encephalitogenic determinants prevents subsequent progression of EAE. Thus, the predictability of acquired self-determinant recognition provides a basis for sequential determinant-specific therapeutic intervention after onset of the autoimmune disease process.

Multiple sclerosis (MS) is an inflammatory demyelinating disease of the central nervous system (CNS). Most patients undergo an initial relapsing-remitting (RR-MS) course that transforms into a relentless neurodegenerative disorder, termed secondary progressive (SP)-MS. Reversible inflammation and demyelination account readily for the pattern of RR-MS but provide an unsatisfactory explanation for irrevocable decline in SP-MS. Axon loss is thought to be responsible for progressive, non-remitting neurological disability during SP-MS. There is considerable potential for neuroprotective therapies in MS, but their application awaits animal models in which axonal loss correlates with permanent neurological disability. In this report, we describe quantitative immunohistochemical methods that correlate inflammation and axonal loss with neurological disability in chronic-relapsing experimental autoimmune encephalomyelitis (EAE). At first attack, CNS inflammation, but not axon loss, correlated with the degree of neurological disability. In contrast, fixed neurological impairment in chronic EAE correlated with axon loss that, in turn, correlated with the number of symptomatic attacks. As proposed for MS, these observations imply a causal relationship between inflammation, axon loss, and irreversible neurological disability. This chronic-relapsing EAE model provides an excellent platform for 2 critical objectives: investigating mechanisms of axon loss and evaluating efficacy of neuroprotective therapies.

We have made the following observations regarding self-recognition during the development and progression of murine experimental autoimmune encephalomyelitis (EAE) and human multiple sclerosis (MS): 1) chronic progression of EAE is accompanied by a sequential, predictable cascade of neo-autoreactivity, commonly referred to as epitope spreading, presumably caused by endogenous self-priming during autoimmune-mediated tissue damage; 2) there is an invariant relationship between the progression of EAE and the emergence of epitope spreading; 3) progression of EAE can be inhibited by the induction of antigen-specific tolerance to spreading determinants after onset of initial neurologic symptoms; 4) CD4+ Th 1 cells responding to spreading determinants are autonomously encephalitogenic; 5) epitope spreading occurs during the development of MS and in some cases involves HLA-DP class II-restricted self-recognition; and 6) progression of both EAE and MS is accompanied by the decline of primary T-cell autoreactivity associated with disease onset and by the concurrent emergence of the epitope spreading cascade. Our studies directly challenge the traditional view that EAE and MS are initiated and maintained by autoreactivity directed against a single predominant myelin protein or determinant. Our results indicate that progression of EAE and MS involves a shifting of T-cell autoreactivity from primary initiating self-determinants to defined cascades of secondary determinants that sustain the inflammatory self-recognition process during disease progression.

Recent studies using murine animal model systems indicate that clinical progression of autoimmune disease may be due to the sequential accumulation of neoautoreactivity characterized by extensive plasticity of self recognition. In the present study, we addressed the question of whether a similar paradigm of self recognition is implicated in the development of multiple sclerosis (MS), a demyelinating disease with a presumed autoimmune etiology. Our approach was to determine serial changes over a 12-18-mo period in response to an epitope-mapping series of 265 12-mer peptides of myelin proteolipid protein (PLP) by patients with isolated monosymptomatic demyelinating syndromes (IMDS), a group of distinct clinical disorders with variable rates of progression to MS. Our data showed that an extensive array of proteolipid protein peptides could elicit autoreactivity. Moreover, differential autoreactive patterns were evident within IMDS patient subpopulations. Monocentric monophasic IMDS patients with no evidence of prior subclinical disease typically showed fully sustained autoreactivity characterized by extensive plasticity, epitope focusing, shifting, and spreading of responses to new self determinants. In contrast, multicentric monophasic IMDS patients with putative evidence of prior asymptomatic lesion formation typically showed partially sustained autoreactivity characterized by abrupt abrogation of responses to an extensive array of self determinants. No sustained autoreactivity was observed in normal control subjects or in patients with other neurologic diseases. Our results indicate that self recognition associated with the development of MS is a developmental process characterized by autoreactive diversity, plasticity, and instability.

Recombinant interferon beta (IFNbeta) benefits patients with relapsing remitting multiple sclerosis (MS), but the mechanisms of action are unknown. We studied in vivo immunologic effects of IFNbeta treatment and their relationship to clinical efficacy. Cytokines were measured in blood and CSF from MS patients participating in a placebo-controlled phase III clinical trial and an open-label phase IV [corrected] tolerability study of IFNbeta-1a. Additionally, immunologic studies were conducted in animals with proteolipid protein (PLP)-induced chronic relapsing experimental autoimmune encephalomyelitis. Single intramuscular (IM) injections of IFNbeta-1a (6 MIU, 30 microg) were associated with significant in vivo upregulation of interleukin-10 (IL-10) and IL-4 but not IFNgamma mRNA in peripheral blood mononuclear cells. Forty-eight hours after each IFNbeta-1a injection, serum IL-10 levels increased and remained elevated for 1 week. IFNbeta-1a recipients in the placebo-controlled phase III clinical trial showed significantly increased concentrations of CSF IL-10 after 2 years of treatment. This response correlated with a favorable therapeutic response. Exposure of PLP-reactive murine T-cell lines to IFNbeta resulted in increased antigen-driven expression of IL-4 and IL-10 and reduced encephalitogenicity. IFNbeta-1a injections induce systemic and intrathecal immunosuppressive cytokines. Myelin-specific T cells treated with IFNbeta-1a demonstrate increased immunosuppressive cytokine expression and reduced encephalitogenicity. The relationship between increased CSF IL-10 and response to therapy suggests that induction of IL-10 is a mechanism underlying IFNbeta-1a effects in MS patients.