Neuropathological and Neuroimmunological Studies of the Central Nervous System of Mice With Axial-Rotatory Experimental Autoimmune Encephalomyelitis

Experimental autoimmune encephalomyelitis (EAE) is an autoimmune inflammatory and demyelinating disease which is commonly used as a model for multiple sclerosis (MS). The “classical” clinical phenotype of EAE is one in which the animals develop ascending paralysis, commencing in the tail and progressing rostrally to the hind limbs and sometimes the front limbs. In this type of EAE, the inflammatory lesions occur predominantly in the spinal cord and the infiltrate is comprised primarily of mononuclear cells. However, as MS is a complex disease with multiple clinical symptoms and signs, it would be beneficial to develop models that more accurately represent the full spectrum of clinical manifestions of disease. This thesis describes in detail a novel form of EAE, called “axial rotatory EAE” (AR-EAE). AR-EAE provides a useful model for studying cerebellar and vestibular types of ataxia, which are a common feature of MS. It was found that AR-EAE could be actively induced by injecting myelin proteolipid protein (PLP) peptide PLP190–209 in complete Freund’s adjuvant into C3H/HeJ mice, or passively induced by transfer of PLP190–209-specific T cells. Clinically, disease progressed very rapidly and was characterised by a distinctive phenotype where the mice roll in an axial rotatory motion. This was shown to be due to the formation of inflammatory and demyelinating lesions predominantly in the brainstem (following both active and passive induction) and cerebellum (following active induction only), with only a small number of lesions in the spinal cord. The atypical localisation of the lesions in AR-EAE underlying the clinical phenotype led us to ask the question of why these lesions occur where they do in this model. Previous reports have suggested that the blood-brain barrier (BBB) shows a constitutive increase in permeability in certain areas of the CNS, which can direct inflammatory lesions to form in the adjacent brain parenchyma. Using HRP extravasation experiments we found that there was no constitutive increase in permeability in the BBB in C3H/HeJ mice that correlated with the lesion distribution in AR-EAE. BBB disruption occurred only at sites of inflammatory cell infiltration in the cerebellum and brainstem. These studies and others from the literature indicate that the pattern of lesion distribution can differ, depending on the myelin protein used to induce EAE and the genetic background of the animal. We hypothesized that a similar correlation might occur in MS and therefore we investigated the correlation between the specificity of the autoimmune responses by immune cells in the blood and cerebrospinal fluid (CSF), the HLA molecules carried, and the clinical features of MS patients. We found that MS patients with extensive involvement of the brainstem and cerebellum carried HLA-DR4, -DR7 or DR-13 molecules and show strongly increased CD4+ T-cell and antibody reactivity to PLP184–209. Furthermore, levels of PLP190–209-specific Abs in the blood of MS patients also correlated with the presence of cerebellar lesions. Because of these findings, and the findings from earlier in the thesis that, in the AR-EAE model, lesions only occur in the cerebellum in actively induced AR-EAE, and not following passive induction, we investigated the possible role of PLP-specific antibody in AR-EAE. One of the main differences between active and passive induction is that active induction involves the production of antibodies, whereas passive induction does not. We demonstrated that when AR-EAE was induced passively with PLP190–209-specific T cells, lesions only occurred in the brainstem, however, when PLP190–209-specific T cells were injected simultaneously with antibody specific for this peptide, lesions were directed towards the brainstem and cerebellum. The other remarkable feature of AR-EAE is that the inflammatory cells that populate these lesions are largely polymorphonuclear, particularly on day 1 of disease. Using RNA protection assays we have shown that the chemokines, CXCL1 and CXCL2, potent chemoattractants for polymorphonuclear cells, are upregulated in the CNS at the onset of clinical signs. Furthermore, using immunohistochemistry, we have shown that the receptor for these ligands, CXCR2, is upregulated on infiltrating neutrophils and activated microglia in the brain parenchyma affected by inflammation in AR-EAE. Blocking this receptor with a small molecule ligand was found to change the EAE phenotype from an axial rotatory to a typical ascending paralysis type, and significantly decreased neutrophil influx into, and inflammation in, the brainstem. Some studies have suggested that the development of a neutrophil-rich CNS inflammatory infiltrate in EAE is due to a downregulation or absence of IFN-. We have shown using RT-PCR, ELISA and Elispot assays that the absence of IFN- is not a requirement for the development of a neutrophil-rich CNS inflammatory infiltrate and axial rotation in the C3H/HeJ model: the encephalitogenic T cells in C3H/HeJ mice produce a mixture of T helper (Th)1 and Th17 cytokines at both the mRNA and protein level, including substantial amounts of both IFN- and IL-17A. Notably, however, the T cells in the CNS inflammatory infiltrate also produced relatively large amounts of Th2 cytokines. We suggest that Th2 cytokines and the chemokines that they induce are a requirement for the development of this form of EAE, both through their induction of a neutrophil-rich infiltrate, and possibly through the role of Th2 cytokines in promoting antibody production. Thus, the C3H/HeJ AR-EAE model may be useful for studying the immunopathology of acute MS predominantly affecting the brainstem and cerebellum. It may also be a useful model for the development of therapy for Marburg’s disease, particularly for testing the efficacy of inhibitors of CXC chemokines or their receptors on the development of acute CNS inflammation and demyelination