Passive transfer of human aquaporin-4 IgG autoantibodies causes complement-independent spinal cord immunopathologies in mice

Abstract

Objective: To study the pathological effects of AQP4-IgG on glial cells and neurons in the absence of complement activation. Background: Neuromyelitis optica spectrum disorders (NMOSD) are autoimmune central nervous system (CNS) inflammatory demyelinating disorders causing blindness, paralysis and mortality. Autoantibodies against aquaporin-4 (AQP4-IgG) are specific and pathogenic in NMOSD. Neuroinflammation is initiated upon binding of AQP4-IgG to astrocytic AQP4. Previous experimental studies mostly focused on the later complement-mediated inflammation and pathologies upon autoantibody-AQP4 interaction. Little is known about the role of early complement-independent events on acute attack development. Design/Methods: Mice were used because human IgG does not activate mice complements. Animals were pretreated with complete Freund’s adjuvant (CFA) and pertussis toxin (PTx) to disrupt blood-brain barrier (BBB), then received daily intraperitoneal injections of IgG (4 mg) purified from AQP4-IgG-seropositive NMOSD patients (IgG-AQP4+) or healthy subjects (IgG-healthy) as control for 8 days. Cervical spinal cords were collected for immunofluorescent analysis. Results: Breakdown of BBB by CFA and PTx induced human IgG infiltration into spinal cord parenchyma. Deposition of complement membrane attack complex (C5b9) was absent after human IgG transfer. Mice received IgG-AQP4+ showed astrocytic injury evidenced by loss of AQP4 and glial fibrillary acidic protein (GFAP) immunoreactivities. These mice also displayed patchy decrease in excitatory amino acid transporter 2 (EAAT2, glutamate transporter) immunostaining, and extensive microglial activation and macrophage activation/infiltration evidenced by ionized calcium-binding adapter molecule 1 (Iba1) and CD68 staining, respectively. In parallel, mice received IgG-AQP4+ had spinal cord lesions with patchy demyelination and axonal loss on myelin basic protein and neurofilament staining. These pathologies were not observed in mice received IgG-healthy. Conclusions: Our findings suggest that passive transfer of AQP4-IgG led to complement-independent pathologies including AQP4, EAAT2 and astrocyte loss, microglia activation, macrophage activation/infiltration, demyelination and axonal loss that may be involved in early stage of NMOSD pathophysiologies.