Abstract
Recent work from our laboratory, using different models of experimental neuromyelitis optica spectrum disorder (NMOSD), has led to a number of observations that might be highly relevant for NMOSD patients. For example: (i) in the presence of neuromyelitis optica immunoglobulin G, astrocyte-destructive lesions can be initiated by CD4+ T cells when these cells recognize aquaporin 4 (AQP4), but also when they recognize other antigens of the central nervous system. The only important prerequisite is that the T cells have to be activated within the central nervous system by "their" specific antigen. Recently activated CD4+ T cells with yet unknown antigen specificity are also found in human NMOSD lesions. (ii) The normal immune repertoire might contain AQP4-specific T cells, which are highly encephalitogenic on activation. (iii) The retina might be a primary target of AQP4-specific T cells and neuromyelitis optica immunoglobulin G: AQP4-specific T cells alone are sufficient to cause retinitis with low-grade axonal pathology in the retinal nerve fiber/ganglionic cell layer. A thinning of these layers is also observed in NMOSD patients, where it is thought to be a consequence of optic neuritis. Neuromyelitis optica immunoglobulin G might target cellular processes of Müller cells and cause their loss of AQP4 reactivity, when AQP4-specific T cells open the blood-retina barrier in the outer plexiform layer. Patchy loss of AQP4 reactivity on Müller cells of NMOSD patients has been recently described. Cumulatively, our findings in experimental NMOSD suggest that both CD4+ T cell and antibody responses directed against AQP4 might play an important role in the pathogenesis of tissue destruction seen in NMOSD.