Variable sensitivity to complement-dependent cytotoxicity in murine models of neuromyelitis optica.

Background: Studies of neuromyelitis optica (NMO), an autoimmune disease of the central nervous system (CNS), have demonstrated that autoantibodies against the water channel aquaporin-4 (AQP4) induce astrocyte damage through complement-dependent cytotoxicity (CDC). In developing experimental models of NMO using cells, tissues or animals from mice, co-administration of AQP4-IgG and normal human serum, which serves as the source of human complement (HC), is required. The sensitivity of mouse CNS cells to HC and CDC in these models is not known.

Methods: We used HC and recombinant monoclonal antibodies (rAbs) against AQP4 to investigate CDC on mouse neurons, astrocytes, differentiated oligodendrocytes (OLs), and oligodendrocyte progenitors (OPCs) in the context of purified monocultures, neuroglial mixed cultures, and organotypic cerebellar slices.

Results: We found that murine neurons, OLs, and OPCs were sensitive to HC in monocultures. In mixed murine neuroglial cultures, HC-mediated toxicity to neurons and OLs was reduced; however, astrocyte damage induced by an AQP-specific rAb #53 and HC increased neuronal and oligodendroglial loss. OPCs were resistant to HC toxicity in neuroglial mixed cultures. In mouse cerebellar slices, damage to neurons and OLs following rAb #53-mediated CDC was further reduced, but in contrast to neuroglial mixed cultures, astrocyte damage sensitized OPCs to complement damage. Finally, we established that some injury to neurons, OLs, and OPCs in cell and slice cultures resulted from the activation of HC by anti-tissue antibodies to mouse cells.

Conclusions: Murine neurons and oligodendroglia demonstrate variable sensitivity to activated complement based on their differentiation and culture conditions. In organotypic cultures, the protection of neurons, OLs, and OPCs against CDC is eliminated by targeted astrocyte destruction. The activation of human complement proteins on mouse CNS cells necessitates caution when interpreting the results of mouse experimental models of NMO using HC.