[Recent progress in multiple sclerosis research: astrocytopathy in demyelinating diseases].

Multiple sclerosis (MS) is a demyelinating disease of the central nervous system (CNS) while neuromyelitis optica (NMO) is an inflammatory disease of the CNS that selectively affects the optic nerves and spinal cord. Recently, a specific IgG against NMO, designated NMO-IgG, was discovered, and the relevant antigen was found to be aquaporin 4 (AQP4), one of the major water channel proteins in the CNS. The sensitivity of NMO-IgG/anti-AQP4 antibodies for NMO varies from 30% to 80%, while specificity is 90-100%. Pathological studies on NMO patients have revealed perivascular immune complex (IgM, IgG and C9neo) deposition and extensive loss of AQP4 in active lesions, while myelin basic protein (MBP) staining was relatively preserved. IgG from NMO-IgG-seropositive NMO patients induces astrocyte death in culture in the presence of complement, and reproduces astrocyte loss in vivo when MBP-specific T cells are co-transferred to cause experimental autoimmune encephalomyelitis. Therefore, it is postulated that the complement-activating anti-AQP4 antibodies have a pivotal role in the development of NMO lesions through astrocyte necrosis, and that demyelination is a secondary event. Baló's disease is characterized by alternating rings of demyelination and preserved myelin. As additional MS-like lesions often coexist in Baló's cases, Baló's disease is regarded as a variant of MS. However, Baló's concentric rings are also observed in NMO cases and in Asian opticospinal MS patients in the cerebral white matter, spinal cord and optic chiasm. In demyelinated areas, many hypertrophic astrocytes are present, in close contact with oligodendrocytes that often show apoptotic features. In the outermost layer of preserved myelin, stress proteins involved in tissue preconditioning are abundant in oligodendrocytes. The peri-plaque white matter is thus assumed resistant to subsequent attack, thereby leaving a layer of preserved myelin. In some patients, Baló's concentric rings develop systematically in a centrifugal direction, while other patients show simultaneous enhancement of multiple rings. Therefore, tissue preconditioning and successive ring formation does not fully explain the mechanism of the disease. We recently reported that AQP4 was extensively lost in glial fibrillary acidic protein-positive hypertrophic astrocytes, both in demyelinated and myelinated layers of all actively demyelinating lesions in four Filipino Baló's patients. None of six other patients with magnetic resonance imaging-confirmed Baló's disease was seropositive for anti-AQP4 antibodies. I therefore propose that AQP4 astrocytopathy, in the absence of anti-AQP4 antibodies, is characteristic of Baló's disease. Since a similar loss of AQP4 without perivascular deposition of immunoglobulin and complement is also observed in autopsied CNS tissues from NMO and MS cases, I consider that autoantibody-independent astrocytopathy may widely occur in human CNS demyelinating diseases, including Baló's disease, MS and NMO.