Behaviour and neuropathology in mice injected with human contactin-associated protein 2 antibodies
Abstract Serum antibodies that bind to the surface of neurons or glia are associated with a wide range of rare but treatable CNS diseases. In many, if not most instances, the serum levels are higher than CSF levels yet most of the reported attempts to reproduce the human disease in mice have used infusion of antibodies into the mouse cerebral ventricle(s) or intrathecal space. We used the intraperitoneal route and injected purified plasma IgG from either a CASPR2-antibody-positive patient (n = 10 mice) or healthy individual (n = 9 mice) daily for 8 days. Lipopolysaccharide was injected intraperitoneally on Day 3 to cause a temporary breach in the blood brain barrier. A wide range of baseline behaviours, including tests of locomotion, coordination, memory, anxiety and social interactions, were established before the injections and tested from Day 5 until Day 11. At termination, brain tissue was analysed for human IgG, CASPR2 and c-fos expression, lymphocyte infiltration, and neuronal, astrocytic and microglial markers. Mice exposed to CASPR2-IgG, compared with control-IgG injected mice, displayed reduced working memory during the continuous spontaneous alternation test with trends towards reduced short-term and long-term memories. In the open field tests, activities were not different from controls, but in the reciprocal social interaction test, CASPR2-IgG injected mice showed longer latency to start interacting, associated with more freezing behaviour and reduced non-social activities of rearing and grooming. At termination, neuropathology showed more IgG deposited in the brains of CASPR2-IgG injected mice, but a trend towards increased CASPR2 expression; these results were mirrored in short-term in vitro experiments where CASPR2-IgG binding to hippocampal neurons and to CASPR2-transfected HEK cells led to some internalization of the IgG, but with a trend towards higher surface CASPR2 expression. Despite these limited results, in the CASPR2-IgG injected mouse brains there was increased c-fos expression in the piriform-entorhinal cortex and hypothalamus, and a modest loss of Purkinje cells. There was also increased microglia density, morphological changes in both microglia and astrocytes and raised complement C3 expression on astrocytes, all consistent with glial activation. Patients with CASPR2 antibodies can present with a range of clinical features reflecting central, autonomic and peripheral dysfunction. Although the behavioural changes in mice were limited to social interactions and mild working-memory defects, the neuropathological features indicate potentially widespread effects of the antibodies on different brain regions.
