Granulocyte-Macrophage Colony-Stimulating Factor: Linking the Adaptive and Innate Immune Systems in Autoimmune Demyelinating Disease

Abstract

Multiple sclerosis (MS) is a debilitating disorder of the central nervous system (CNS) characterized by motor, sensory, and visual deficits. Published literature supports the contention that MS is an autoimmune disease mediated by auto-reactive CD4+ T-helper (Th) cells which infiltrate the CNS from circulation. Infiltrating Th cells are reactivated within the CNS in an antigen-specific manner, driving recruitment, differentiation, and activation of circulating myeloid cells, and ultimately resulting in demyelination and axonopathy. Much work has been devoted to elucidating the roles of each of the known Th cell subsets, and the inflammatory mediators they produce, in CNS autoimmune disease. One Th cell mediator, granulocyte-macrophage colony-stimulating factor (GM-CSF), has been proposed to be a critical inflammatory cytokine that connects CNS-infiltrating Th cells with the pathogenic programming of tissue-invading myeloid cells. The aim of my dissertation project was to determine the mechanism of action of GM-CSF in the initiation, progression, and maintenance of experimental autoimmune encephalomyelitis (EAE), widely used as an animal model of MS. The Segal lab and others have previously shown that C57BL/6 mice deficient in GM-CSF are resistant to EAE. We found that lymph node cells from immunized GM-CSF deficient mice mount an impaired MOG35-55-specific proliferative and cytokine response. Insufficient Th priming could explain, in part, the resistance of those mice to EAE. To study the role of GM-CSF during the effector phase, we transferred encephalitogenic T cells from MOG35-55-primed wild-type (WT) mice into naïve GM-CSF receptor-deficient mice (Csf2r-/-). Although Csf2r-/- recipients developed EAE with similar incidence and initial disease trajectory as their WT counterparts, they underwent clinical remission. The total number of cells infiltrating the CNS at peak EAE were comparable between groups, but neutrophils, myeloid-derived dendritic cells (mDCs), and MOG35-55-specific T cells were reduced in Csf2r-/- recipients. This suggested that either a paucity of infiltrating neutrophils, mDCs, and/ or encephalitogenic T cells could be responsible for the remitting phenotype exhibited by Csf2r-/- recipients. To investigate the basis of the differences in the CNS infiltration, we compared chemokine expression in the spinal cords of WT and Csf2r-/- recipients during EAE. The myeloid cell chemoattractants CXCL1, CXCL2, and CCL2 were consistently comparable between groups, and CCL6 was comparable at disease onset but diverged thereafter. CCL6 levels progressively rose in the CNS of WT mice but dropped dramatically in Csf2r-/- mice by peak disease. CCR1, the sole receptor for CCL6, is expressed by subsets of leukocytes and has been detected on inflammatory cells in MS lesions. Its role in EAE remains to be elucidated. CCR1 blockade beginning at the time of T cell transfer reduced CNS-infiltration by monocytes and classical DCs (cDCs) and prevented the development of EAE. Conversely, CCR1 blockade following EAE onset triggered clinical remission associated with a reduction in CNS-infiltrating neutrophils, closely resembling the phenotype of Csf2r-/- recipients. Together, these studies provide insight into the pleiotropic, and extensive, roles GM-CSF plays in the development and maintenance of CNS autoimmunity. We propose the following model: encephalitogenic T cells are reactivated in the CNS by cDCs and secrete GM-CSF. GM-CSF promotes the production of CCL6 within the CNS and expression of CCR1 by peripheral myeloid cells. CCL6:CCR1 interactions promote the migration of monocytes and neutrophils across the blood-brain-barrier and into the CNS parenchyma where they directly damage the myelin sheath and axons.