Encephalitis with Autoantibodies against the Glutamate Kainate Receptors GluK2.

OBJECTIVE: The objective of this study was to report the identification of antibodies against the glutamate kainate receptor subunit 2 (GluK2-abs) in patients with autoimmune encephalitis, and describe the clinical-immunological features and antibody effects. METHODS: Two sera from 8 patients with similar rat brain immunostaining were used to precipitate the antigen from neuronal cultures. A cell-based assay (CBA) with GluK2-expressing HEK293 cells was used to assess 596 patients with different neurological disorders, and 23 healthy controls. GluK2-ab effects were determined by confocal microscopy in cultured neurons and electrophysiology in GluK2-expressing HEK293 cells. RESULTS: Patients' antibodies precipitated GluK2. GluK2 antibody-specificity was confirmed by CBA, immunoprecipitation, GluK2-immunoabsorption, and GluK2 knockout brain immunohistochemistry. In 2 of 8 samples, antibodies reacted with additional GluK2 epitopes present in GluK1 or GluK3; in both, the reactivity was abrogated after GluK2 immuno-absorption. Six of 8 patients developed acute encephalitis and clinical or magnetic resonance imaging (MRI) features of predominant cerebellar involvement (4 presenting as cerebellitis, which in 2 patients caused obstructive hydrocephalus), and 2 patients had other syndromes (1 with cerebellar symptoms). One of the samples showed mild reactivity with non-kainate receptors (alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors [AMPAR] and N-methyl-D-aspartate receptors [NMDAR]) leading to identify 6 additional cases with GluK2-abs among patients with anti-AMPAR (5/71) or anti-NMDAR encephalitis (1/73). GluK2-abs internalized GluK2 in HEK293 cells and neurons; these antibody-effects were reversible in neurons. A significant reduction of GluK2-mediated currents was observed in cells treated with patients' GluK2 serum following the time frame of antibody-mediated GluK2 internalization. INTERPRETATION: GluK2-abs associate with an encephalitis with prominent clinicoradiological cerebellar involvement. The antibody effects are predominantly mediated by internalization of GluK2. ANN NEUROL 2021;90:107-123.

AMPA/kainate glutamate receptors contribute to inflammation, degeneration and pain related behaviour in inflammatory stages of arthritis.

OBJECTIVES: Synovial fluid glutamate concentrations increase in arthritis. Activation of kainate (KA) and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) glutamate receptors (GluRs) increase interleukin-6 (IL-6) release and cause arthritic pain, respectively. We hypothesised that AMPA and KA GluRs are expressed in human arthritis, and that intra-articular NBQX (AMPA/KA GluR antagonist) prevents pain and pathology in antigen-induced arthritis (AIA). METHODS: GluR immunohistochemistry was related to synovial inflammation and degradation in osteoarthritis (OA) and rheumatoid arthritis (RA). A single intra-articular NBQX injection was given at induction, and knee swelling and gait of AIA and AIA+NBQX rats compared over 21 days, before imaging, RT-qPCR, histology and immunohistochemistry of joints. Effects of NBQX on human primary osteoblast (HOB) activity were determined. RESULTS: AMPAR2 and KA1 immunolocalised to remodelling bone, cartilage and synovial cells in human OA and RA, and rat AIA. All arthritic tissues showed degradation and synovial inflammation. NBQX reduced GluR abundance, knee swelling (p<0.001, days 1-21), gait abnormalities (days 1-2), end-stage joint destruction (p<0.001), synovial inflammation (p<0.001), and messenger RNA expression of meniscal IL-6 (p<0.05) and whole joint cathepsin K (p<0.01). X-ray and MRI revealed fewer cartilage and bone erosions, and less inflammation after NBQX treatment. NBQX reduced HOB number and prevented mineralisation. CONCLUSIONS: AMPA/KA GluRs are expressed in human OA and RA, and in AIA, where a single intra-articular injection of NBQX reduced swelling by 33%, and inflammation and degeneration scores by 34% and 27%, respectively, exceeding the efficacy of approved drugs in the same model. AMPA/KA GluR antagonists represent a potential treatment for arthritis.

Developmental distribution of the glutamate receptor subunits KA2, GluR6/7, and delta 1/2 in the rat medial nucleus of the trapezoid body. A quantitative image analysis.

The medial nucleus of the trapezoid body (MNTB) acts as a relay nucleus in the transmission of auditory information from the cochlear nucleus (CN) to the lateral superior olive. Glutamate receptors mediate the excitatory synaptic transmission in the CN-MNTB projection. Here, we used immunohistochemistry to investigate the expression pattern of the kainate receptor subunits KA2 and GluR6/7 and the orphan glutamate receptor subunits delta 1/2 in principal neurons of the rat MNTB during early postnatal development (P2-59). To objectively quantify the intensity of immunoreactivity, images were scanned with a CCD camera and used for gray-value measurements. At all ages analyzed, each of the three antisera produced immunoreactivity in the somata of MNTB principal cells and in the neuropil. KA2 immunoreactivity of somata and neuropil remained nearly constant between P2 and 23. In contrast, the intensity of GluR6/7 immunoreactivity of somata and neuropil increased between P2 and 6, followed by a decrease until P10. Between P10 and 23, GluR6/7 immunoreactivity of neuropil remained nearly constant, whereas it increased in the somata. In both somata and neuropil, the intensity of delta 1/2 immunoreactivity decreased between P2 and 10, reaching a constant, low level by P10. Our results demonstrate the continuous presence of the glutamate receptor subunits KA2, GluR6/7 and delta 1/2 in the developing MNTB, yet quantitative changes occur which may be associated with functional differences.

GluR5,6,7 subunit immunoreactivity on apical pyramidal cell dendrites in hippocampus of schizophrenics and manic depressives.

Recent postmortem studies have suggested that changes in the regulation of kainate-sensitive glutamate receptors (kainate receptors) in the hippocampus may play a role in schizophrenia. To explore this possibility further, the distribution of immunoreactivity (IR) for the GluR5,6,7 subunits of the KR was assessed in a cohort consisting of 15 normal controls, 15 schizophrenics, and 9 manic depressives matched for age and postmortem interval (PMI). Cross sections of hippocampus showed abundant GluR5,6,7-IR on apical dendrites of pyramidal neurons in the stratum radiatum and stratum moleculare. In normal controls, both the numerical and length density of IR dendrites were much higher in sector CA2 than in sectors CA3 or CA1. When data for the individual groups were separately examined, the schizophrenics showed a 30-35% reduction in the density of GluR5,6,7-IR dendrites found in both stratum radiatum and stratum moleculare of sectors CA3 and CA2, as well as proximal and middle portions of CA1. In CA2, the magnitude of this decrease in schizophrenia was 2.5 times larger than that seen in any of the other sectors. For the manic depressive group, no significant differences were observed in any sectors or laminae examined. The potential confounding effects of either age, PMI, or neuroleptic exposure do not explain the reduced density of IR dendrites detected in the schizophrenic group. Taken together, the preferential reduction of GluR5,6,7-IR observed on apical dendrites of pyramidal neurons is consistent with a functional downregulation of the kainate receptor in the hippocampus of schizophrenic brain.

Ionotropic glutamate receptors during the development of the chick retina.

Glutamate is the main neurotransmitter of photoreceptors, bipolar cells, and ganglion cells of the vertebrate retina. Three main classes of ionotropic glutamate receptors comprising different subunits can be distinguished: AMPA (alpha-amino-3-hydroxy-5-methyl-4-isoxasolepropionate), KA (kainate), and NMDA (N-methyl-D-aspartate). This study was undertaken to characterize the AMPA (GluR1, GluR2/3, and GluR4), KA (GluR5/6/7), and NMDA (NR1) ionotropic glutamate receptor subunits and to determine their distribution during the development of the chick retina by Western blotting and immunohistochemistry. Western blotting analysis at 1 day after hatching indicated that the antibodies against GluR1, 2/3, 4, and 5/6/7 and NR1 recognized specifically a single band of 100-110 kDa. In turn, immunohistochemistry at P1 showed that all subunits were expressed in cells of the inner nuclear and ganglion cell layers of the chick retina, mostly amacrine and ganglion cells, and their processes in the inner plexiform layer. In addition, stained processes in the outer plexiform layer were observed with the antibodies against GluR2/3, GluR4, and GluR5/6/7. Although all subunits appeared around E5-E6 in the prospective ganglion cell layer, and later in the prospective inner nuclear layer, the distribution of cells containing these glutamate receptor subunits revealed distinct ontogenetic patterns. This multiplicity of glutamate receptors may contribute to different processes that occur in the chick retina during development.

Identification of the amino acids on a neuronal glutamate receptor recognized by an autoantibody from a patient with paraneoplastic syndrome.

Autoantibodies from a patient with paraneoplastic disease were identified previously to bind to the glutamate receptor (GluR) subunit GluR5 and to function as potential allosteric modulators of receptor activity (Gahring et al. [1995] Mol Med 1:245-253). In the present study we have used deletion mapping and mutagenesis to define the residues in GluR5 bound by this autoreactivity. The autoantibody contact residues include residues K497, N508, K510, E512, and to a lesser extent Q507. Residues 507-512 confer autoantibody specificity of the autoreactivity to GluR5. These residues have been shown in crystallographic studies (Armstrong et al. [1998] Nature 395:913-917) to participate in a loop structure, whereas residue K497 is located on a beta-strand. Notably, this binding spans tyrosine 504, a residue important in forming the agonist-binding site. We propose that autoantibody binding of essential residues in this GluR5 autoantigenic region defines a subunit-specific allosteric regulatory site on neuronal glutamate receptors and suggests how receptor dysfunction and region-specific neuronal death in the brain can progress in certain autoimmune neurological diseases.

Dendritic localization of Ca(2+)-permeable AMPA/kainate channels in hippocampal pyramidal neurons.

Although it is well established that cortical and hippocampal gamma-aminobutyric acid (GABA)-ergic neurons generally have large numbers of Ca(2+)-permeable alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA)/kainate channels (Ca-A/K channels), their presence on pyramidal neurons is controversial. Ca2+ permeability of AMPA channels is regulated by expression of a particular glutamate receptor subunit (GluR2), which confers Ca2+ impermeability to heteromeric channels. Most electrophysiology studies, as well as in situ hybridization and immunolabeling studies demonstrating expression of GluR2 mRNA or peptide in pyramidal neurons, have provided evidence against the presence of Ca-A/K channels on pyramidal neurons. However, observations that pyramidal neurons often appear to be labeled by kainate-stimulated Co2+ influx (Co2+(+) cells), a histochemical stain that identifies cells possessing Ca-A/K channels, suggests that they may have these channels. The present study futher examines cellular and subcellular distribution of Ca-A/K channels on hippocampal pyramidal neurons in slice as well as in culture. To this end, techniques of kainate-stimulated Co2+ influx labeling, supplemented by AMPA receptor subunit immunocytochemistry and fluorescent imaging of kainate-stimulated intracellular Ca2+ ([Ca2+]i) rises are employed. Co2+ labeling is often seen in pyramidal neuronal dendrites in both slice and in culture. In addition, although GluR1 and 4 staining in these neurons is often seen in the soma and dendrites, GluR2 label, when evident, is generally more restricted to the soma. Finally, measurement of kainate-stimulated [Ca2+]i rises in cultured neurons, assessed by using low affinity Ca2+ indicators in the presence of N-methyl-D-aspartate (NMDA) receptor and voltage-sensitive Ca2+ channel blockade, often shows dendritic rises to precede those in the somata. Thus, these data support the hypothesis that Ca-A/K channels are present in dendritic domains of many pyramidal neurons, and may help to provide resolution of the apparently conflicting data regarding their distribution.

Excitotoxicity in the enteric nervous system.

Glutamate, the major excitatory neurotransmitter in the CNS, is also an excitatory neurotransmitter in the enteric nervous system (ENS). We tested the hypothesis that excessive exposure to glutamate, or related agonists, produces neurotoxicity in enteric neurons. Prolonged stimulation of enteric ganglia by glutamate caused necrosis and apoptosis in enteric neurons. Acute and delayed cell deaths were observed. Glutamate neurotoxicity was mimicked by NMDA and blocked by the NMDA antagonist D-2-amino-5-phosphonopentanoate. Excitotoxicity was more pronounced in cultured enteric ganglia than in intact preparations of bowel, presumably because of a reduction in glutamate uptake. Glutamate-immunoreactive neurons were found in cultured myenteric ganglia, and a subset of enteric neurons expressed NMDA (NR1, NR2A/B), AMPA (GluR1, GluR2/3), and kainate (GluR5/6/7) receptor subunits. Glutamate receptors were clustered on enteric neurites. Stimulation of cultured enteric neurons by kainic acid led to the swelling of somas and the growth of varicosities ("blebs") on neurites. Blebs formed close to neurite intersections and were enriched in mitochondria, as revealed by rhodamine 123 staining. Kainic acid also produced a loss of mitochondrial membrane potential in cultured enteric neurons at sites where blebs tended to form. These observations demonstrate, for the first time, excitotoxicity in the ENS and suggest that overactivation of enteric glutamate receptors may contribute to the intestinal damage produced by anoxia, ischemia, and excitotoxins present in food.

Drugs of abuse and stress increase the expression of GluR1 and NMDAR1 glutamate receptor subunits in the rat ventral tegmental area: common adaptations among cross-sensitizing agents.

Behavioral and electrophysiological evidence suggests that glutamatergic neurotransmission plays an important role in some of the long-term effects of cocaine and other drugs of abuse on brain function. We therefore examined the effect of repeated cocaine treatment on glutamate receptor subunit expression in central dopamine (DA) pathways implicated in many of cocaine's behavioral actions. By immunoblotting procedures using subunit-specific antibodies, we found that repeated, but not acute, cocaine treatment increased the levels of immunoreactivity of GluR1 (an AMPA receptor subunit) and NMDAR1 (an NMDA receptor subunit) in the ventral tegmental area (VTA), a nucleus containing mesolimbic DA neurons. In contrast, chronic cocaine treatment did not alter levels of GluR2 (an AMPA receptor subunit), NMDA2A/B (NMDA receptor subunits), or GluR6/7 (kainate receptor subunits) in this brain region. Moreover, GluR1 and NMDAR1 levels were not regulated in other regions of the mesolimbic or nigrostriatal DA pathways, including the substantia nigra. Because several drugs of abuse and stress can elicit common and cross-sensitizing effects on mesolimbic DA function, we next examined whether repeated morphine and stress treatments would regulate these proteins similarly in the VTA. Although morphine delivered by subcutaneous pellet implantation had no significant effect on subunit levels, morphine delivered intermittently by subcutaneous injections of escalating doses elevated GluR1 levels in the VTA. Repeated restraint stress also paradigm (2 stressors/d under variable conditions) increased both GluR1 and NMDAR1 levels in this brain region. Unlike cocaine, morphine, and stress, repeated treatment with other psychotropic drugs (haloperidol, raclopride, sertraline, and desipramine) that lack reinforcing or sensitizing properties did not regulate GluR1 or NMDAR1 subunit levels in the VTA. Increased glutamate receptor subunit expression in the VTA may represent an important molecular mechanism by which drugs of abuse and stress exert common, long-term effects on mesolimbic DA function.

Biochemical and assembly properties of GluR6 and KA2, two members of the kainate receptor family, determined with subunit-specific antibodies.

To examine subunit assembly and biochemical properties of two members of the kainate family of glutamate receptors (GluR), antibodies were made to synthetic peptides corresponding to the carboxyl termini of GluR6 and KA2. Immunoblot analysis of membranes from human embryonic kidney cells transfected with glutamate receptor cDNAs showed that these antibodies are selective for their respective receptor subunit except that the antibody to GluR6 also recognizes GluR7, which is expected due to the sequence homology between the two subunits at the carboxyl terminus. In transfected cell membranes, immunoblot analysis with the antibody to GluR6 showed a major immunoreactive band at 118 kDa and minor bands at 103 and 28 kDa. The 103-kDa band appears to be a deglycosylated form of GluR6 since deglycosylation eliminates staining at 118 kDa and increases staining of the 103-kDa band. Immunoblot analysis of KA2 transfected cell membranes shows a major band at 123 kDa and minor bands at 109 and 37 kDa. Deglycosylation converts the 123-kDa band into a 109-kDa band. Analysis of brain tissues shows that both antibodies label single major bands which migrate at the same molecular masses as those from transfected cell membranes, 118 and 123 kDa for GluR6 and KA2, respectively. Immunoprecipitation studies showed that antibodies to GluR6 and KA2 selectively immunoprecipitated [3H]kainate binding activity, but not 3H-labeled alpha-amino-3-hydroxy-5-methyl-4-isoxazole proprionic acid (AMPA) binding activity, from Triton X-100-solubilized rat brain membranes. Furthermore, each antibody coimmunoprecipitated GluR6 and KA2 from cells co-transfected with GluR6 and KA2 cDNAs and from detergent-solubilized rat brain membranes, indicating that these two subunits can coassemble into a molecular complex. Interestingly, GluR1 and GluR2, subunits of the AMPA receptor, also co-immunoprecipitated with GluR6 in cells co-transfected with GluR6 and GluR1 or GluR2 cD-NAs. Such complexes appear to be present to a limited extent in the brain.