REVIEW Antibodies to the N-Methyl-D-Aspartate Receptor and Other Synaptic Proteins in Psychosis Julia Deakin, Belinda R. Lennox, and Michael S. Zandi This review concentrates on the evidence for autoantibodies to cell surface synaptic proteins in psychosis and schizophrenia. We and others have recently found antibodies to the N-methyl-D-aspartate receptor in first-episode psychosis. We describe the evidence for pathogenicity and disease-relevance of these antibodies, which builds on the novel field in neuroimmunology of cell surface antibody– associated central nervous system disorders. Relevant autoantibodies in psychosis and schizophrenia are likely to be those directed to cell surface proteins, in which the likelihood of pathogenicity is greater. We discuss the evidence for this from the field of paraneoplastic neurologic syndromes and the discovery of novel cell surface antigen central nervous system autoimmune syndromes. Key Words: Autoantibody, encephalitis, encephalopathy, N-methyl-D-aspartate receptor (NMDAR), leucine-rich glioma- inactivated 1 (Lgi1), paraneoplastic, psychosis, schizophrenia, voltage-gated potassium channel complex (VGKC) W e review the evidence for disease-relevant autoantibodies to cell surface synaptic proteins in psychosis. We and others have recently found antibodies to the N-methyl- D-aspartate receptor (NMDAR) in first-episode psychosis in patients who were later diagnosed with schizophrenia (1,2). We describe the clinical relevance of these antibodies, building on the novel field in neuroimmunology of cell surface antibody–associated central nervous system disorders (3,4). Brain autoantibodies, often region-specific (e.g., to hippocampus or cerebellum) have been studied in schizophrenia since the 1930s, but no specific and relevant antibody has been found thus far (5–7). Many of such previously studied autoantibodies have turned out not to be clinically relevant or prevalent in high numbers of healthy control subjects, or the original work was not replicated. These limitations may be due in part to the method of antibody measurement in most cases, using nonspecific small peptide enzyme-linked immu- nosorbent assays (ELISA), or to the intracellular location of most antigens associated with antibodies. Intracellular antigens are more likely to be irrelevant and epiphenomenal markers of a destructive predominantly cellular process or nonspecific markers in serum of no relevance. Relevant autoantibodies in psychosis and schizo- phrenia are more likely to be directed to cell surface synaptic proteins in which the likelihood of pathogenicity is greater. The evidence for this has emerged from the field of paraneoplastic neurological syndromes and the discovery of novel cell-surface antigen central nervous system autoimmune syndromes. Cell Surface Antibodies in Paraneoplastic and Nonparaneoplastic Encephalitis: Discovery and Mechanisms An autoimmune paraneoplastic noninfectious encephalitis associated with cancer was first described in the 1960s (8,9). Paraneoplastic syndromes include cerebellar degeneration, opso- clonus myoclonus, and neuropathy/ganglionopathy. The anti- bodies associated with these disorders are often directed to intracellular antigens (with exceptions such as the voltage-gated calcium channel antibodies in the Lambert-Eaton syndrome), and therefore the antibodies are probably epiphenomena, raised to the debris of brain cellular destruction. Patients with intracellular antibodies responded poorly to tumor removal and immunother- apy, including plasmapheresis and corticosteroids (10). In 2001, Buckley, Vincent, and colleagues described autoimmune ence- phalitis in two patients with serum antibodies to voltage-gated potassium channel complexes (VGKC), measured by iodinated- dendrotoxin radioimmunoprecipitation assay (3,11), which was a major landmark in the field. Patients with these antibodies responded well to treatment and often had no evidence of an underlying tumour, even after prolonged follow-up. Later work by Vincent and colleagues, replicated by Dalmau and colleagues, demonstrated that most VGKC antibodies are directed toward cell surface proteins complexed to Kv potassium channel subunits, predominantly leucine-rich glioma inactivated-1 (Lgi1) and con- tactin associated protein-2 (Caspr2), complexed with potassium channel subunits (most detected by the radioimmunoprecipita- tion assay) (12,13). In 2007, Dalmau and colleagues found antibodies to the NMDAR in a group of women with ovarian teratoma and an encephalitis notable for its preceding psychotic phase, often initially mistaken for a psychiatric condition (14). The patients usually developed seizures, a rhythmic movement disorder, and prolonged coma but responded well to tumor removal and immunotherapy (15,16). The discovery of the antibody has led to reports of more than 600 patients, including men and children, in the 6 years since that first publication (4,16–19). The antibodies in most cases are directed to the amino terminal domain of the NR1 subunit (20). Tumors were found in 20% of the European series (16) and 40% of the U.S. series (19), and those who had tumors removed had a better functional outcome than those without tumors (23). Aggressive immunotherapy, with steroids, plasmapheresis, or intravenous immunoglobulin, and second-line B cell depletion therapies such as rituximab in refractory cases are associated with better outcomes in both retrospective series, but no clinical trials have taken place. Irani et al. describe a two-stage illness with “cortical” symptoms, including psychosis, dominating the initial phase of the disease (16). Approximately two thirds of patients, in both series, presented with psychiatric symptoms first, often to psychiatrists. Having stated the distinction of antibodies into those directed to intracellular and cell surface antigens, not all cell surface antibodies may be clinically relevant, and their interpretation From the Department of Psychiatry (JD), University of Cambridge, Cam- bridge; Department of Psychiatry (BRL), University of Oxford, Oxford; and Department of Clinical Neurosciences (MSZ), University of Cam- bridge, Cambridge, United Kingdom. Address correspondence to Belinda Lennox, D.M., M.R.C.Psych., Depart- ment of Psychiatry, Warneford Hospital, Oxford OX3 7JX, United Kingdom; E-mail: Belinda.Lennox@psych.ox.ac.uk. Received Apr 1, 2013; revised Jul 1, 2013; accepted Jul 17, 2013. 0006-3223/$36.00 BIOL PSYCHIATRY 2013;]:]]]–]]] http://dx.doi.org/10.1016/j.biopsych.2013.07.018 & 2013 Society of Biological Psychiatry