There is much interest in therapeutic manipulation of cytokine responses in autoimmunity, yet studies in mouse models have sometimes produced conflicting findings as to the role of particular mediators in disease. Examples include the contradictory findings regarding susceptibility to experimental allergic encephalomyelitis (EAE) or diabetes in knockout mice for various individual Th1 or Th2 cytokines or their receptors. An alternative approach to the analysis of Th1 and Th2 mechanisms in these diseases is to investigate strains carrying a null mutation for molecules involved in cytokine receptor signal transduction, signal transducer and activator of transcription (Stat4) and Stat6. Stat4 is pivotal in Th1 polarization, being activated when IL-12 binds the IL-12R and leading to the production of IFNgamma. We here report disease susceptibility in non-obese diabetic mice carrying a Stat4-null mutation. Knockout mice were almost completely protected from diabetes, only rarely showing pancreatic peri-islet infiltrates. Furthermore, there was near complete protection from the induction of EAE by either of the two encephalitogenic myelin epitopes. Despite this protection, Stat4-null mice showed clear epitope spread compared with controls during myelin oligodendrocyte glycoprotein-induced EAE as judged by T cell proliferation, although this was not associated with a strong Th1 response to the initial or spread epitope and, furthermore, there was no evidence of a switch to Th2 cytokines.

Although the primary cause of multiple sclerosis (MS) is unclear, evidence supports a role for autoimmune attack of myelin by T lymphocytes. However, it has been difficult to relate patterns of autoimmunity to pathogenesis. In mouse models, the case has been made for relapsing and remitting disease driven by epitope spread: an initial lesion leads to presentation of central nervous system antigens, in turn triggering the next wave of autoimmune T cells of different specificity, the response thus broadening. Few studies have been done to determine whether these events could be important over the longer time scale of human disease. We compared T cell responses with a panel of myelin epitopes in clinically isolated syndrome patients with a first attack, patients with MS with a mean disease duration of 0.95 years, and patients with MS having a mean disease duration of 15.9 years. T cells from patients with long-term disease recognize more myelin epitopes than patients with recent-onset disease. The epitope myelin basic protein 131-149, in particular, was more commonly recognized by patients with long-term disease. The data support the notion that the T cell response in MS broadens with time and is thus implicated in the ongoing pathogenic process. However, there was no clear correlation between disease severity and number of epitopes recognized. This may argue against a simple causal role of epitope spread in driving progression, as has been suggested in experimental allergic encephalomyelitis.