Identification of a Novel Serotonin/Glutamate Receptor Complex Implicated in Psychosis

Abstract

The psychosis associated with schizophrenia is characterized by alterations in sensory processing and perception^1,2. Some antipsychotic drugs were identified by their high affinity for serotonin 5-HT_2A receptors (2AR)^3,4. Drugs that interact with metabotropic glutamate receptors (mGluR) also show potential for the treatment of schizophrenia^5-7. The effects of hallucinogenic drugs, such as psilocybin and lysergic acid diethylamide (LSD), require the 2AR^8-10 and resemble some of the core symptoms of schizophrenia^10-12. Here we show that the mGluR2 interacts via specific transmembrane helix domains with the 2AR, a member of an unrelated G protein-coupled receptor (GPCR) family, to form functional complexes in brain cortex. The 2AR/mGluR2 complex triggers unique cellular responses when targeted by hallucinogenic drugs, and activation of mGluR2 abolishes hallucinogen specific signalling and behavioural responses. In postmortem human brain from untreated schizophrenic subjects, the 2AR is up-regulated and the mGluR2 is down-regulated, a pattern that could predispose to psychosis. These regulatory changes suggest that the 2AR/mGluR2 complex may be involved in the altered cortical processes of schizophrenia, and represents a promising new target for the treatment of psychosis.

The 2AR and mGluR2/3 show an overlapping distribution in brain cortex in autoradiography studies¹³. The mGluR2 and mGluR3 are not distinguished by autoradiographic ligands. We used fluorescent in situ hybridization (FISH) to determine whether either of these receptor subtypes are co-expressed by the same neurons. In layer V mouse somatosensory cortex (SCx), 2AR mRNA positive cells were mostly mGluR2 mRNA positive. The level of expression in SCx was much lower for mGluR3 mRNA, which rarely co-localized with 2AR mRNA (Fig. 1a). Control studies validated assay sensitivity and specificity, and similar 2AR/mGluR2 mRNA co-localization was found in cortical primary cultures (Figs. 1a,b,c, and Supplementary Fig. S1). Translation of 2AR protein in cortical pyramidal neurons was found to be necessary for normal mGluR2 expression. Mice with globally disrupted 2AR expression (htr2A−/− mice) showed reduced cortical mGluR2 binding and expression, while mice in which 2AR expression was selectively restored in cortical pyramidal neurons^8,14 showed control expression levels (Supplementary Table S1, and Supplementary Fig. S2). The effects of mGluR2/3 activation on 2AR responses have been generally attributed to synaptic mechanisms^5,6,13,15. However, the co-localization of 2AR and mGluR2 and the reduction of mGluR2 expression levels in htr2A−/− mice motivated us to examine whether a direct mechanism contributed to cortical crosstalk between these two receptor systems.

Figure 1. 2AR and mGluR2 co-localize and interact.

a, 2AR and mGluR2, but not mGluR3, co-express in neurons. Scale, top=50 μm, bottom=10 μm. Nuclei are blue. Inset: co-expressing neuron. b, FISH for mGluR3 in thalamus. Scale, top=25μm, bottom=10μm. c, mRNA levels by real-time PCR (n=6 per group). d, Specific co-immunoprecipitation of 2AR and mGluR2 in duplicate human cortex samples (arrows). e. BRET² shows specific 2AR and mGluR2 interaction in HEK293 cells. Data are mean±s.e.m. (n=3). The mGluR2/2AR curve is preferably fitted by a saturation curve, F test (p<0.001). The other co-transfection datasets show linear correlations. f, [³H]Ketanserin displacement curves in mouse SCx membranes (top panels). 2AR agonist affinities were higher in the presence of mGluR2/3 agonist 10 μM LY379. [³H]LY341495 displacement curves (bottom panels). mGluR2/3 agonist affinities were lower in the presence of 2AR agonist 10 μM DOI.

Recent studies have demonstrated that some GPCRs belonging to the same sequence classes can form dimers¹⁶ or, potentially, higher-order oligomers¹⁷. Although the 2AR and mGluR2 belong to different GPCR classes, we established the existence of 2AR/mGluR2 heterocomplexes by several methods: co-immunoprecipitation of human brain cortex samples (Fig. 1d) and of HEK293 cells transfected with epitope-tagged receptors (Fig. 2b), bioluminescence resonance energy transfer (BRET) (Fig. 1e, and Supplementary Fig. S3), and fluorescence resonance energy transfer (FRET) (Fig. 2d) studies in transfected cells.

Figure 2. mGluR2 transmembrane domains 4/5 mediate association with 2AR.

a, mGluR2/mGluR3 chimeras studied. b, c-myc-2AR and HA-mGluR2/mGluR3 chimera co-immunoprecipitations. Cells separately expressing each construct were also mixed. c, 2AR competition binding in cells stably expressing 2AR and transfected with mGluR2/mGluR3 chimeras. d, FRET in cells expressing 2AR-eCFP and either mGluR2, mGluR3 or mGluR3ΔTM4,5 chimera, all tagged with eYFP. Pseudo-colour images represent normalized values (FRET^N). eCFP + eYFP (n=19); 2AR-eCFP + mGluR2-eYFP (n=43); 2AR-eCFP + mGluR3-eYFP (n=31); 2AR-eCFP + mGluR3ΔTM4,5-eYFP, (n=27). **p < 0.01; ANOVA with Dunnett's post hoc test. e, DOI-stimulated [³⁵S]GTPγS binding in membranes followed by immunoprecipitation with anti-Gα_q/11 (top panels), or anti-Gα_i1,2,3 (bottom panels). Cells stably expressing 2AR were transfected with mGluR2, mGluR3 or mGluR3ΔTM4,5. The potency of DOI activating Gα_i1,2,3 was significantly increased when the 2AR was co-expressed with either mGluR2 or mGluR3ΔTM4,5, an effect abolished by 10 μM LY379 (p<0.001 by F test). Data are mean±s.e.m. of three experiments performed in triplicate. f, Ribbon backbone representation of the transmembrane helices of the 2AR/mGluR2 heteromer model from the intracellular face.

To determine whether the formation of the 2AR/mGluR2 complex has functional consequences, we first examined the effects in mouse SCx membranes of an mGluR2/3 agonist on the competition binding of several hallucinogenic 2AR agonists (Fig. 1f, top) and of a 2AR agonist on the competition binding of several mGluR2/3 agonists (Fig. 1f, bottom). The agonist affinities for the 2AR and mGluR2/3 were decreased when receptor/G protein complexes were uncoupled by GTPγS (Supplementary Fig. S4, and Supplementary Tables S2 and S3). Notably, the glutamate agonist LY379268 (LY379) increased the affinity of all three hallucinogens studied for the 2AR binding site. Furthermore, the 2AR agonist DOI decreased the affinity of the three mGluR2/3 agonists for the glutamate receptor binding site. The allosteric interactions observed were eliminated by antagonist for each modulator (see Supplementary Tables S2 and S3 and Supplementary Fig. S4 for additional concentrations of DOI and LY379, and elimination of the allosteric effects by antagonists). Although the glutamate agonists studied do not distinguish between the mGluR2 and mGluR3 subtypes¹⁸, the rarity of mGluR3 and 2AR mRNA co-expression in cortex, the absence of evidence for 2AR/mGluR3 complex formation by co-immunoprecipitation, BRET and FRET, and the detection of 2AR/mGluR2 complexes by these same assays, suggest that the crosstalk identified results from 2AR/mGluR2 complexes.

The differences in the capacity of the mGluR2 and mGluR3 to interact with the 2AR and their close sequence similarity provided the basis to identify the specific mGluR2 domains responsible for heterocomplex formation. Study of a series of molecular chimeras of the mGluR2 and mGluR3 (see Fig. 2a) demonstrated that the segment containing transmembrane (TM) helices 4 and 5 of the mGluR2 receptor was both necessary and sufficient for complex formation with the 2AR. The mGluR3 receptor chimera containing only this segment from the mGluR2 (mGluR3ΔTM4,5) was capable of co-immunoprecipitating with the 2AR (Fig 2b), mediating allosteric crosstalk (Fig. 2c) and maintaining close proximity with the 2AR as indicated by FRET (Fig. 2d). In contrast, mGluR2ΔTM4,5 did not show evidence of complex formation with the 2AR (Fig. 2, Supplementary Figs. S5, S6, and Supplementary Tables S4 and S5 for complete curves, analysis and evidence of membrane expression of all chimeras). The absolute and relative levels of expression of heterologous constructs were comparable to the physiological levels found in mouse SCx, and in cortical primary cultures (Supplementary Fig. S5 and Supplementary Table S4). Our data do not exclude the possibility that the predicted 2AR/mGluR2 heterodimer, a model of which is shown in Fig. 2f, assembles into tetramers or larger receptor oligomers^19,20.

The changes in high affinity binding caused by 2AR/mGluR2 crosstalk suggested that this complex may serve to integrate serotonin and glutamate signalling and modulate G protein coupling^21,22. This hypothesis was tested by measuring 2AR regulation of Gα_q/11 and Gα_i proteins. High-affinity activation of Gα_q/11 by the 2AR was reduced by co-expression of mGluR2 (Fig. 2e, and Supplementary Table S6). Interestingly, the activation of Gα_i by the 2AR was markedly enhanced by mGluR2 co-expression (Fig. 2e, and Supplementary Table S7). The mGluR2-dependent effects on both Gα_q/11 and Gα_i regulation by the 2AR were reversed in the presence of mGluR2 agonist (Fig. 2e, and Supplementary Tables S6 and S7). Consonant with the co-immunoprecipitation, allosteric modulation and FRET results, the functional assays of G protein activity also show that the TM4−5 segment of the mGluR2, when substituted into the mGluR3, was sufficient for signalling crosstalk to occur (Fig. 2e). These data support the presence of functional and physiological 2AR/mGluR2 complexes that integrate serotonin and glutamate neurotransmission to specify the pattern of G protein regulation.

Similar evidence for specification of G protein subtype regulation was also observed by the endogenous brain 2AR/mGluR2 complex with membranes from cortical primary cultures (Fig 3a). The pattern of G protein regulation in cortical pyramidal neurons has been shown to predict specific behavioural responses to 2AR agonists. Hallucinogenic drugs and non-hallucinogenic drugs activate the same population of 2ARs in cortical pyramidal neurons, but differ in the 2AR-dependent pattern of G protein regulation and gene induction they elicit^8,9. In brain cortical neurons, the signalling elicited by hallucinogenic and non-hallucinogenic 2AR agonists causes induction of c-fos and requires G_q/11-dependent phospholipase C activation. However, the signalling of hallucinogens such as DOI and LSD acting at the 2AR also induces egr-2, which is G_i/o-dependent. Thus c-fos expression results from any 2AR-signalling, and egr-2 induction is a specific marker for hallucinogen signalling via the 2AR^8,9. The finding that mGluR2 modulates the G_i protein coupling of the 2AR (Fig. 3a, and Supplementary Tables S6 and S7) suggested that this complex might be important for hallucinogen signalling. The induction of c-fos by hallucinogenic 2AR agonists or by structurally similar non-hallucinogenic 2AR agonists in vivo in mouse SCx and in cortical primary cultures (Fig. 3b, and Supplementary Figs. S8, S9 and S10) was not affected by the mGluR2/3 agonist LY379. In contrast, the hallucinogen-specific induction of egr-2 was selectively blocked by LY379 in both mouse cortex in vivo and in primary cortical cultures (Fig. 3b, and Supplementary Figs. S8, S9 and S10 for FISH results with LSD treatment, and real-time PCR gene assay results with DOI, DOM, DOB, LSD, lisuride and ergotamine). We also studied the effects of LY379 on the head-twitch response (HTR) behavior, which is hallucinogen-specific^8,9. Similar to its effects on G protein activation and gene induction, the glutamate agonist LY379 suppressed the induction of the HTR by either DOI or LSD (Supplementary Fig. S11). These results suggest that LY379 acts at the 2AR/mGluR2 complex to reduce the hallucinogen-specific G_i/o protein signalling and behaviour. To further establish the functional relevance of 2AR/mGluR2 crosstalk, we compared the responses to the mGluR2/3 antagonist LY341494 in htr2A+/+ and htr2A−/− mice. The locomotor and vertical activities elicited by LY341495 were significantly attenuated in the htr2A−/− mice (Fig. 3c), supporting the functional relevance of the 2AR/mGluR2 complex in vivo and suggesting that it also influences the endogenous response to glutamate.

Figure 3. 2AR/mGluR2 complex-dependent modulation of cellular and behavioural responses.

a , DOI-stimulated [³⁵S]GTPγS binding in primary culture membranes followed by immunoprecipitation with anti-Gα_q/11 or anti-Gα_i1,2,3 antibodies. DOI Gα_i1,2,3 activation potency was significantly decreased by 10 μM LY379. Data are mean±s.e.m. of three experiments performed in triplicate. b, FISH in mice injected with vehicle or 2 mg/kg DOI 15 min after being pre-injected with vehicle or 15 mg/kg LY379 (left panels), and in primary cultures treated with 10 μM DOI 15 min after being pre-treated with vehicle or 10 μM LY379 (bottom panels). Nuclei are blue. Scale, left=50 μm, right=10μm. c, Distance and vertical activity induced in htr2A+/+ and htr2A−/− mice by mGluR2/3 antagonist 6 mg/kg LY341495. In htr2A−/− mice, LY341495 effect on distance was reduced (p<0.05, Bonferroni's post hoc test of two-factor ANOVA), and on vertical activity was absent (n=30−32).

The findings that G_i/o protein regulation, which is necessary for the effects of hallucinogens⁸, is enhanced by the formation of the 2AR/mGluR2 complex and that activation of the mGluR2 component suppresses hallucinogen-specific signalling implicate this complex in the effects of hallucinogens. The neuropsychological effects of hallucinogenic drugs present commonalities with the psychosis of schizophrenia, and both conditions are accompanied by disruptions of cortical sensory processing^10,11,23-27. We investigated whether the components of the 2AR/mGluR2 signalling complex are dysregulated in brain cortex of subjects with schizophrenia. We determined the density of 2AR and mGluR2/3 binding sites in cortex from schizophrenic subjects and controls who were matched by gender, age, and postmortem delay (Supplementary Tables S8 and S9). The receptor densities in cortical membranes from untreated schizophrenic subjects were significantly altered, showing increased 2AR and reduced mGluR2/3 receptor levels (Fig. 4a, b). mRNA assays showed that expression of mGluR2 but not mGluR3 was reduced in schizophrenia cortex (Fig. 4e). The studies in mouse show that activation of the mGluR2 component of the 2AR/mGluR2 complex eliminates the hallucinogen-specific component of the signalling responses to LSD-like drugs. Thus the increased 2AR and decreased mGluR2 found in the brain in schizophrenia may predispose to a hallucinogenic pattern of signalling.

Figure 4. 2AR is increased and mGluR2 is decreased in schizophrenia.

a, b, Frontal cortex membrane receptor binding assays from untreated schizophrenic (n = 13) and matched control subjects (n = 13). In schizophrenia, [³H]ketanserin binding was higher and [³H]LY341495 binding was lower (p< 0.05; Student's t-test). c, d, Receptor binding in antipsychotic-treated schizophrenic (n = 12) and matched control subjects (n = 12). In treated schizophrenia, [³H]ketanserin binding was unaffected and [³H]LY341495 binding was lower (p < 0.05). e, mGluR2 mRNA expression is reduced in untreated schizophrenic subjects (n = 7) compared to matched control subjects (n = 7, p < 0.05, mean±s.e.m).

Many laboratories have attempted to determine the density of 2AR in postmortem brain from subjects with schizophrenia, and some studies have reported decreased or unchanged 2AR densities²⁸. To try to understand the basis for these discrepancies from our results, we first studied the effects of chronic antipsychotic treatment on the 2AR and mGluR2 in mouse. The chronic atypical antipsychotic clozapine specifically down-regulated the level of expression of 2AR and of mGluR2 in mouse SCx (Supplementary Fig. S12). The down-regulation of mGluR2 by clozapine required expression of the 2AR, as it did not occur in htr2A−/− mice (Supplementary Fig. S12), and was not induced by the chronic typical antipsychotic haloperidol (Supplementary Fig. S13). In concordance with the effects of clozapine in murine models, the density of 2AR was reduced to control levels in postmortem human brain cortex of schizophrenics treated with atypical antipsychotic drugs (Fig. 4c), and the mGluR2/3 binding sites were also down-regulated (Fig. 4d). The onset of psychosis in schizophrenia usually occurs in later adolescence or early adulthood¹. We studied the relationship of receptor densities with aging and both [³H]ketanserin and [³H]LY341495 binding displayed a highly significant negative correlation with age (Supplementary Fig. S14). Hallucinations and delusions typically attenuate with aging²⁹, which correlates with the lower density of the components of the 2AR/mGluR2 complex that we observed in older subjects. Consequently, the marked dysregulation of both 2AR and mGluR2 expression in schizophrenia would be unlikely to be observed in samples from heterogeneous groups including treated patients²⁸ or in studies including older patients^28,30.

These studies identify the 2AR/mGluR2 complex as a possible site of action of hallucinogenic drugs. The glutamate and serotonin systems have both been implicated in psychotic disorders, and the components of this complex are found to be differentially regulated in cortex from individuals with schizophrenia. The results are consistent with the hypothesis that the 2AR/mGluR2 complex integrates serotonin and glutamate signalling to regulate the sensory gating functions of the cortex, a process that is disrupted in psychosis.

METHODS

A detailed Methods section is available in Supplementary Information. Briefly, all reagents were purchased from commercial vendors except for LY379268 (Eli Lilly and Company). Mouse lines, treatment protocols, behavioural studies, dissections, and primary neuronal cultures, approved by Institutional Use and Care Committees, have been previously described^8,9. Protocols used for FISH⁸, binding assays⁸, real-time PCR⁸, FRET¹⁷ and co-immunoprecipitation¹⁷ were performed as previously described or with minor modifications. Epitope tagged, BRET², FRET and chimera receptor constructs were generated using standard cloning techniques and were confirmed by sequencing. BRET² using Renilla luciferase and Green Fluorescent Protein (GFP²) was performed in HEK293 cells. Matched schizophrenia and control human brains were obtained from autopsies performed in the Basque Institute of Legal Medicine, Bilbao, Spain in compliance with policies of research and ethical review boards for postmortem brain studies.