Differential binding of antibodies in PANDAS patients to cholinergic interneurons in the striatum

Abstract

Pediatric Autoimmune Neuropsychiatric Disorder Associated with Streptococcus, or PANDAS, is a syndrome of acute childhood onset of obsessive-compulsive disorder and other neuropsychiatric symptoms in the aftermath of an infection with Group A beta-hemolytic Streptococcus (GABHS). Its pathophysiology remains unclear. PANDAS has been proposed to result from cross-reactivity of antibodies raised against GABHS with brain antigens, but the targets of these antibodies are unclear and may be heterogeneous. We developed an in vivo assay in mice to characterize the cellular targets of antibodies in serum from individuals with PANDAS. We focus on striatal interneurons, which have been implicated in the pathogenesis of tic disorders. Sera from children with well-characterized PANDAS (n = 5) from a previously described clinical trial (NCT01281969), and matched controls, were infused into the striatum of mice; antibody binding to interneurons was characterized using immunofluorescence and confocal microscopy. Antibodies from children with PANDAS bound to ∼80% of cholinergic interneurons, significantly higher than the <50% binding seen with matched healthy controls. There was no elevated binding to two different populations of GABAergic interneurons (PV and nNOS-positive), confirming the specificity of this phenomenon. Elevated binding to cholinergic interneurons resolved in parallel with symptom improvement after treatment with intravenous immunoglobulin. Antibody-mediated dysregulation of striatal cholinergic interneurons may be a locus of pathology in PANDAS. Future clarification of the functional consequences of this specific binding may identify new opportunities for intervention in children with this condition.

Keywords: Animal model; Antibody; Autoimmunity; Interneurons; Obsessive-compulsive disorder; PANDAS; Tics.

Copyright © 2017 Elsevier Inc. All rights reserved.

Figures

Figure 1. PANDAS autoantibodies differentially bind to ChAT interneurons in the basal ganglia

Deposition of antibodies from control and PANDAS sera (collected at baseline from highly symptomatic patients) onto ChAT-positive interneurons (CINs) in vivo was evaluated by double immunofluorescent staining and confocal microscopy, using anti-human IgG (red) and anti-ChAT antibodies (green). Antibodies from both control and PANDAS serum were deposited onto CINs, but a significantly larger number of ChAT+IgG+ cells (arrowheads), as a fraction of total ChAT+ cells, was observed in PANDAS samples, compared to healthy controls (N=5 samples per group, 2 mice per sample; two-tailed t-test: t[8] = 12.90; p < 0.0001). Insets illustrate binding to a single ChAT+ cell at higher magnification. Scale bar = 100 μm.

Figure 2. PANDAS autoantibodies do not show elevated binding to parvalbumin-positive GABAergic interneurons

No differences between PANDAS and control serum in antibody deposition onto PV-positive interneurons were observed (two-tailed t-test: t[8] = 1.7, p = 0.13). Scale bar = 100 μm.

Figure 3. PANDAS autoantibodies do not show elevated binding to nNOS-positive GABAergic interneurons

No differences between PANDAS and control serum in antibody deposition onto nNOS interneurons were observed (two-tailed t-test: t[8] = 1.6, p = 0.15). Scale bar = 100 μm.

Figure 4. Resolution of elevated ChAT binding in post-IVIG serum

A. These patients all showed a significant clinical response to IVIG treatment. Individual patients’ changes in the Children’s Yale-Brown Obsessive Compulsive Scale (CY-BOCS) score across four evaluation sessions are shown (see also Table 1). Antibody binding to interneurons was assayed at baseline (Visit 1; Figures 1–3) and 12 weeks later (Visit 3). B. IgG binding to ChAT+ interneurons was markedly reduced in serum collected from all five PANDAS patients after IVIG treatment (evaluation 3), relative to baseline (paired t-test: t[4] = 4.4; p = 0.012). C. Change in IgG binding to CINs correlated significantly with improvement in CY-BOCS after IVIG treatment (r2 = 0.86, p = 0.023). Scale bar = 100 μm.