Rho GTPase-activating protein 10 (ARHGAP10/GRAF2) is a novel autoantibody target in patients with autoimmune encephalitis

Journal of Neurology https://doi.org/10.1007/s00415-022-11178-9 ORIGINAL COMMUNICATION Rho GTPase‑activating protein 10 (ARHGAP10/GRAF2) is a novel autoantibody target in patients with autoimmune encephalitis Sven Jarius1 · Lars Komorowski2 · Jens U. Regula3,4 · Jürgen Haas1 · Stefanie Brakopp2 · Brigitte Wildemann1 Received: 22 February 2022 / Revised: 4 May 2022 / Accepted: 5 May 2022 © The Author(s) 2022 Abstract Background In 2010, we described a novel immunoglobulin G (IgG) autoantibody (termed anti-Ca after the index case) targeting Rho GTPase-activating protein 26 (ARHGAP26, also termed GTPase regulator associated with focal adhesion kinase [GRAF], or oligophrenin-like protein 1 [OPHN1L]) in autoimmune cerebellar ataxia (ACA). Later, ARHGAP26IgG/anti-Ca was reported in patients with limbic encephalitis/cognitive decline or peripheral neuropathy. In several of the reported cases, the syndrome was associated with cancer. ARHGAP10/GRAF2, which is expressed throughout the central nervous system, shares significant sequence homology with ARHGAP26/GRAF. Mutations in the ARHGAP10 gene have been linked to cognitive and psychiatric symptoms and schizophrenia. Objective To assess whether ARHGAP26-IgG/anti-Ca co-reacts with ARHGAP10. Methods Serological testing for ARHGAP10/GRAF2 autoantibodies by recombinant cell-based assays and isotype and IgG subclass analyses. Results 26/31 serum samples (84%) from 9/12 (75%) ARHGAP26-IgG/anti-Ca-positive patients and 4/6 ARHGAP26-IgG/ anti-Ca-positive CSF samples from four patients were positive also for ARHGAP10-IgG. ARHGAP10-IgG (termed antiCa2) remained detectable in the long-term (up to 109 months) and belonged mainly to the complement-activating IgG1 subclass. Median ARHGAP26-IgG/anti-Ca and median ARHGAP10-IgG/anti-Ca2 serum titres were 1:3200 and 1:1000, respectively, with extraordinarily high titres in some samples (ARHGAP26-IgG/anti-Ca: up to 1:1000,000; ARHGAP10IgG: up to 1:32,000). ARHGAP26/anti-Ca serum titres exceeded those of ARHGAP10-IgG in all samples but one. A subset of patients was positive also for ARHGAP10-IgM and ARHGAP10-IgA. CSF/serum ratios and antibody index calculation suggested intrathecal production of ARHGAP26-IgG/anti-Ca and anti-ARHGAP10. Of 101 control samples, 100 were completely negative for ARHGAP10-IgG; a single control sample bound weakly (1:10) to the ARHGAP10-transfected cells. Conclusions We demonstrate that a substantial proportion of patients with ARHGAP26-IgG/anti-Ca-positive autoimmune encephalitis co-react with ARHGAP10. Further studies on the clinical and diagnostic implications of ARHGAP10-IgG/ anti-Ca2 seropositivity in patients with autoimmune encephalitis are warranted. Keywords Cerebellar ataxia · Rho GTPase-activating protein 26 (ARHGAP26) · Rho GTPase-activating protein 10 (ARHGAP10) · Autoantibodies · Anti-Ca · Anti-Ca2 · Antibodies · Immunoglobulin G (IgG) · Autoimmune encephalitis · Limbic encephalitis · Polyneuropathy · Cognitive decline · Antigen · GTPase regulator associated with focal adhesion kinase (GRAF) · GRAF2 · Oligophrenin-like protein 1 (OPHN1L) · Medusa head ataxia Abbreviations ACA Autoimmune cerebellar ataxia AI Antibody index ARHGAP26 Rho GTPase-activating protein 26 CBA Cell-based assay CDT Clock-drawing test * Sven Jarius Extended author information available on the last page of the article CSF Cerebrospinal fluid FITC Fluorescein isothiocyanate GAD65 Glutamic acid decarboxylase, 65-kD isoform GRAF GTPase regulator associated with focal adhesion kinase HEK293 Human embryonic kidney 293 cells IgG/M/A Immunoglobulin G/M/A ITPR1 Inositol 1,4,5-trisphosphate receptor 1 IVIG Intravenous immunoglobulins 13 Vol.:(0123456789) Journal of Neurology IVMP Intravenous methylprednisolone MMSE Mini-mental status examination NMDAR N-methyl D-aspartate receptors OCB Oligoclonal bands OPHN1L Oligophrenin-like protein 1 PC Purkinje cells PLEX Plasma exchange TBA Tissue-based assay Introduction Background In 2010, we identified a novel high-titre serum reactivity (termed anti-Ca) targeting the Rho GTPase-activating protein 26 (ARHGAP26, also termed GTPase regulator associated with focal adhesion kinase [GRAF], or oligophreninlike protein 1 [OPHN1L]) in a patient with autoimmune cerebellar ataxia (ACA) [11]. The patient had presented with rapidly progressive ACA leading to marked cerebellar atrophy and severe disability. The antibodies were present at high titres, were produced intrathecally, and belonged to the IgG1 subclass, suggesting that ARHGAP26-IgG/antiCa may be not only of diagnostic but also of pathogenetic impact. ARHGAP26 is expressed predominantly in Purkinje cells (PC) in the cerebellum, but also by a subset of neurons in the hippocampus. Following up on our original report, we described five further ARHGAP26-IgG/anti-Ca-positive patients with ACA [2, 10, 28]. Two of these patients, as well as another patient without ACA, had signs of possible limbic encephalitis [13], indicating that the spectrum of neurological manifestations associated with ARHGAP26IgG/anti-Ca may be broader than originally thought. In three further cases, isolated cognitive decline was noted [1]. In addition, 17 further ARHGAP26/anti-Ca-positive patients were reported by the Mayo Clinic in 2020, 15 of whom had subacute progressive cerebellar ataxia and 2 peripheral neuropathy [15], and several as yet unpublished additional cases were identified in our laboratories over the subsequent years. Considering that a substantial number of these cases were associated with cancer (including squamous cell carcinoma of the lung, ovarian cancer, prostate cancer, gastric adenocarcinoma, B-cell lymphoma and thymoma), ARHGAP26IgG/anti-Ca-positive encephalitis is considered a facultative paraneoplastic neurological syndrome. Expression of ARHGAP26 at protein or RNA level has been shown for a multitude of solid tumours and cancer cell lines [24]. Several paralogues of ARHGAP26 have been reported. We were thus interested in whether ARHGAP26-IgG/antiCa may cross-react with other members of the ARHGAP family or related proteins. A database search revealed the 13 Rho GTPase 10 (ARHGAP10/GRAF2) gene as a particularly important paralogue of ARHGAP26/GRAF [26]. Accordingly, some regions within the proteins coded by these two genes show significant sequence homology [26]. Animal data suggest that ARHGAP10 is widely expressed throughout the brain, including the cerebellum [14, 24]. Therefore, we hypothesised that cross-reactivity of ARHGAP26-IgG/anti-Ca may render ARHGAP10 an additional immune target in patients with ACA and, possibly, other forms of autoimmune encephalitis associated with ARHGAP26-IgG/anti-Ca seropositivity. To explore this hypothesis, we developed a cell-based immunoassay employing recombinant ARHGAP10 as antigenic substrate and tested sera from both ARHGAP26-IgG/ anti-Ca-positive patients and controls for autoantibodies to ARHGAP10. We found that serum and CSF IgG from (...truncated)