Antibodies to neuronal surface proteins in Tourette Syndrome: Lack of evidence in a European paediatric cohort.

Abstract

In Tourette Syndrome (TS) a role for autoantibodies directed against neuronal proteins has long been suspected, but so far results are still inconsistent. The aim of this study was to look for antibodies to specific or undefined neuronal proteins that could be involved in the aetiology of the disease. Sera from children with Tourette Syndrome or another chronic tic disorder (TS/TD), collected as part of the longitudinal European Multicenter Tics in Children Study, were investigated. Participants included 30 siblings of patients with TS/TD prior to developing tics (preclinical stage) and the same children after the first tic onset (onset), and 158 patients in the chronic phase undergoing an acute relapse (exacerbation). Presence of antibodies binding to rodent brain tissue was assessed by immunohistology on rat brain sections and by immunofluorescent staining of live hippocampal neurons. Live cell-based assays were used to screen for antibodies to NMDAR, CASPR2, LGI1, AMPAR and GABAAR. Immunohistology indicated evidence of antibodies reactive with brain tissue, binding mainly to the hippocampus, the basal ganglia or the cerebellum in 26/218 (12%), with 8% of the preclinical or onset sera binding to the dentate gyrus/CA3 region or cerebellum. Only two individuals (one pre-clinical, one chronic) had antibodies binding the NMDAR and the binding was only weakly positive. No other specific antibodies were detected. Despite some immunoreactivity towards neuronal antigens on brain tissue, this was not mirrored by antibodies binding to live neurons, suggesting the presence of non-specific antibodies or those that bind non-pathogenic intracellular epitopes. NMDAR or the other neuronal surface antibodies tested were very infrequent in these patients. The evidence for pathogenic antibodies that could be causative of TS is weak.