Antibody-Mediated Encephalitis Mechanisms

Encephalitis associated with N-methyl-D-aspartate receptor antibodies can have a devastating impact on patients' cognitive and psychiatric function. Researchers at the University of Pennsylvania, Philadelphia, obtained cerebrospinal fluid (CSF) and sera at random from a series of 320 patients with clinical manifestations of the disease to explore the cellular mechanisms by which antibodies cause the loss of NMDA receptors.

Anti-NMDA receptor antibodies trigger symptoms of memory loss, deficits in learning and cognition, and psychosis in patients with anti-NMDA receptor encephalitis by crosslinking and internalizing synaptic NMDA receptor clusters, according to findings reported by Ethan G. Hughes, Ph.D., and his colleagues. (Dr. Hughes is now a postdoctoral fellow at Johns Hopkins University, Baltimore.)

Among patients with clinical manifestations of the disease, they found that purified anti-NR1 antibodies reduced the surface and total cluster density of NMDA receptors on cultured hippocampal neurons in a titer-dependent fashion. (The NR1 epitope is one of two protein subunits that form the NMDA receptor.) The researchers cultured the neurons with CSF samples that had been obtained at two different time points of the disease in two patients. In one of the patients, they found that CSF taken from the patient at symptom presentation had a higher titer of anti-NR1 antibodies than was present during symptom improvement; whereas in the second patient, they found that the CSF obtained during symptom worsening had a higher antibody titer than at symptom presentation. Periods in which anti-NR1 antibody titers were higher also had greater declines in NMDA receptor surface and total cluster density (J. Neurosci. 2010;30:5866–75).

Further experiments with hippocampal neurons cultured with CSF or purified anti-NR1 antibodies demonstrated that the patients' antibodies “dramatically reduced” the synaptic localization of NMDA receptor clusters in a titer-dependent fashion without affecting the overall structural integrity of excitatory neurons, synapses, or other synaptic proteins or receptors. The density of the NMDA receptor clusters returned to baseline levels 4 days after the antibodies were removed, demonstrating the reversibility of the antibodies' effects. Whole-cell patch recordings found that the antibodies decreased only NMDA receptor–mediated currents and not those mediated by other receptors. The patients' antibodies promoted the internalization of the receptors by binding to, capping, and cross-linking them.

In addition to the study of patients with the disease, the researchers directly infused human anti-NR1 antibodies into the hippocampus of adult rats for 2 weeks. They found that the antibodies had bound to the rat hippocampus in a predictable pattern that was dependent on NMDA receptor density and similar to the pattern seen in patients with anti-NMDA receptor encephalitis. The density of NMDA receptor clusters had declined significantly in regions where the antibodies had been deposited, similar to what was observed in sections of the hippocampus in two patients who died from anti-NMDA receptor encephalitis.

The fact that many patients who develop anti-NMDA receptor encephalitis are first admitted to psychiatric institutions with schizophrenia-like symptoms suggests that NMDA receptor hypofunction underlies the symptoms of anti-NMDA receptor encephalitis and many manifestations of schizophrenia, Dr. Hughes and his colleagues wrote.

The research was funded by grants from the National Institutes of Health.

Dr. Carter's comment: Anti-NMDA receptor encephalitis is the first of several “autoimmune synaptic encephalopathies” that have been described within the last 5 years. Other similar disorders with different receptor targets include anti-AMPA receptor encephalitis and anti-GABA receptor encephalitis. These disorders share certain clinical features such as female predominance, frequent association with neoplasm, and presentation with psychosis, behavioral changes, or intractable seizures. Importantly, these disorders are treatable and reversible if recognized early and treated aggressively with immunosuppressive therapies and removal of the neoplasm if present.

The experiments of Dr. Hughes and his associates elegantly clarify the cellular and synaptic mechanisms of anti-NMDA receptor encephalitis using both in vitro and in vivo models. They demonstrate that antibodies from patient sera or CSF reversibly reduce synaptic NMDA receptor clusters by binding, crosslinking, and internalizing NMDA receptors without damaging synaptic connections or the number or structural integrity of these synapses. Furthermore, this reduction in synaptic NMDA receptors shows a linear correlation with the titer of anti-NMDA receptor antibodies. Patient antibodies also reduce synaptic NMDA receptor currents, and reduce NMDA receptor cluster density in rodent and human hippocampus slices in vivo.

This report solidly establishes the role of anti-NMDA receptor antibodies in the pathogenesis of this disorder, creating a new paradigm of central nervous system synaptic autoimmunity. The reversibility of this process is also encouraging for clinicians who may see and treat patients with these disorders, which are likely highly underrecognized at present. The challenge for the future will be to translate these basic science advances into effective treatments for these disorders, and to recognize new forms and presentations of this widening family of autoimmune synaptic encephalopathies.