Epithelial V-Like Antigen Mediates Efficacy of Anti-Alpha(4) Integrin Treatment in a Mouse Model of Multiple Sclerosis

Natalizumab inhibits the transmigration of activated T lymphocytes into the brain and is highly efficacious in multiple sclerosis (MS). However, from a pharmacogenomic perspective, its efficacy and safety in specific patients remain unclear. Here our goal was to analyze the effects of epithelial V-like antigen (EVA) on anti-alpha(4) integrin (VLA4) efficacy in a mouse model of MS, experimental autoimmune encephalomyelitis (EAE). EVA has been previously characterized in human CD4 T lymphocytes, mouse thymic development, and choroid plexus epithelial cells. Further analysis here demonstrated expression in B lymphocytes and an increase in EVA(+) lymphocytes following immunization. Following active induction of EAE using the MOG(35–55) active immunization model, EVA deficient mice developed more severe EAE and white matter tissue injury as compared to wild type controls. This severe EAE phenotype did not respond to anti-VLA4 treatment. In both the control antibody and anti-VLA4 conditions, these mice demonstrated persistent CNS invasion of mature B lymphocyte (CD19(+), CD21(+), sIgG(+)), increased serum autoantibody levels, and extensive complement and IgG deposition within lesions containing CD5(+)IgG(+) cells. Wild type mice treated with control antibody also demonstrated the presence of CD19(+), CD21(+), sIgG(+) cells within the CNS during peak EAE disease severity and detectable serum autoantibody. In contrast, wild type mice treated with anti-VLA4 demonstrated reduced serum autoantibody levels as compared to wild type controls and EVA-knockout mice. As expected, anti-VLA4 treatment in wild type mice reduced the total numbers of all CNS mononuclear cells and markedly decreased CD4 T lymphocyte invasion. Treatment also reduced the frequency of CD19(+), CD21(+), sIgG(+) cells in the CNS. These results suggest that anti-VLA4 treatment may reduce B lymphocyte associated autoimmunity in some individuals and that EVA expression is necessary for an optimal therapeutic response. We postulate that these findings could optimize the selection of treatment responders.