G Protein-Coupled Receptor Heterocomplexes in Neuropsychiatric Disorders

Abstract

G protein-coupled receptors (or GPCRs) represent the largest family of membrane proteins in the human genome and are the target of approximately half of all therapeutic drugs. GPCRs contain a conserved structure of seven transmembrane domains. Their amino terminus is located extracellularly, whereas the carboxy terminus extends into the cytoplasm. Accumulating evidence suggests that GPCRs exist and function as monomeric entities. Nevertheless, more recent findings indicate that GPCRs can also form dimers or even higher order oligomers. The differential pharmacological and signaling properties of GPCR heteromeric complexes hint that their physiological effects may be different as compared to those obtained in tissue cultures that express a particular GPCR. In this chapter, we review current data on the role of GPCR heteromerization in receptor signaling, as well as its potential implication in neuropsychiatric disorders such as schizophrenia, depression, and Parkinson’s disease.

1. INTRODUCTION

Among the different proteins found in cell membranes responsible for transmitting information from the extracellular environment into the cytoplasm, there are those ones belonging to the superfamily ofGprotein-coupled receptors (GPCRs).^1–4 It has been described in the human genome about 800 members of GPCRs,^5,6 and they are the target of nearly half of the drugs currently in use for the treatment of different human diseases.^3,7 GPCRs are single polypeptides embedded in the membrane with the amino terminus remaining in the extracellular side and the carboxy terminus in the cytoplasm.^8–11 Until recently, GPCRs have been assumed to function as monomers that, upon agonist binding, initiate the activation of heterotrimeric G proteins and consequently modulate the function of intracellular signaling cascades downstream. This is supported by findings that demonstrate functional G protein coupling of a single purified family A GPCR reconstituted into a phospholipid bilayer.^12,13 During the past two decades, however, there has been much evidence to suggest that GPCRs exist as dimers and higher order oligomers.^14–19 Of particular interest is the demonstration of heteromerization between two GPCRs belonging to different families. These protein complexes have been shown to affect diverse aspects of receptor function.^20–24 This chapter reviews some of the latest advances in our understanding of structure and function of GPCR heteromers, and their role in neuropsychiatric disorders such as schizophrenia, depression, and Parkinson’s disease.

2. STRUCTURE OF GPCR HETEROMERS

The three major families of GPCRs include family A (rhodopsin-like), family B (glucagon-related receptors), and family C (metabotropic glutamate (mGlu)-related receptors).^1,5,6 Much evidence indicates that family C GPCRs, including GABA_B^25–31 and mGlu receptors^30,32–36, exist and function as constitutive dimers. The GABA_B receptor has been shown to be a heterodimer of GABA_B-R1 and GABA_B-R2, each of which is unable to form a functional GABA_B receptor. The domains responsible for GABA_B heterodimeric receptor formation are located in the C-termini of both GABA_B-R1 and GABA_B-R2 receptors. Recent findings also suggest that mGlu receptors are expressed at the plasma membrane as strict dimers, and not higher order oligomers, that are covalently linked through a disulfide bond. In marked contrast, however, family A GPCRs are assembled as homomers and heteromers through structural mechanisms that involve noncovalent interactions between amino acid residues located in transmem-brane (TM) regions. As an example, it has been suggested using biochemical and biophysical approaches such as coimmunoprecipitation of epitope-tagged receptors and fluorescence resonance energy transfer (FRET) that α_1b-adrenergic receptors are assembled into oligomeric structures with symmetric contact points involving residues located in TM domains 1 and 4^14,37–39 (see also Refs. 40 and 41 for reviews about FRET and other techniques based on energy transfer). Expression of α_1b-adrenergic receptors as higher order oligomers was suggested based on the use of three-color FRET (3-FRET). Similar findings were obtained with the dopamine D2 receptor. Thus, the combination of bioluminescence/fluorescence complementation and energy transfer provided evidence that dopamine D2 receptors are expressed as higher order oligomers in the plasma membrane.^15,38,39 The use of cysteine cross-linking experiments also pointed toward the location of TM1 and TM4 at the symmetrical interfaces of the dopamine D2 receptor oligomers (i.e., TM1–TM1 interface and TM4–TM4 inter-face).^15,38,39 The critical role of TM4 in the formation of GPCR complexes is further supported by recent findings with the serotonin 5-HT_2A and the metabotropic glutamate 2 (mGlu2) receptors.^20,23,24 It was demonstrated that 5-HT_2A and mGlu2, but not mGlu3, form a GPCR heteromer in tissue culture. The differences between mGlu2 and mGlu3 receptors to be assembled as a GPCR heteromer with the 5-HT_2A receptor provided the rational for the design of mGlu2/mGlu3 chimeras that showed differences in their capacity to be expressed in close proximity with coexpressed 5-HT_2A receptors. The authors suggested that three residues located at the intracellular end of TM4 of mGlu2 are necessary to form a GPCR heteromer with the 5-HT_2A in HEK293 cells.²⁴ Together, these findings suggest that TM4 and TM1 contribute to the formation of GPCR homo- and heterocomplexes.

Another important question related to structure and function of GPCRs concerns the stability of dimeric/oligomeric states.⁴² The use of fluorescence recovery after photobleaching (FRAP) has demonstrated transient interactions on a timescale of seconds between the components of the β1-adrenergic receptor homomer. On the contrary, the β2-adrenergic receptor has been shown to be expressed as a stable oligomer in HEK293 cells.⁴³ A dynamic equilibrium between monomers and homodimers/olig-omers has also been recently suggested using FRAP in HEK293 cells expressing dopamine D2 receptors.⁴⁴ Similarly, the use of total internal reflection fluorescence microscopy in living cells point toward a dynamic nature of muscarinic M1 receptor dimer formation. Thus, muscarinic M1 receptors are expressed as monomers or dimers, but not oligomers. It was also shown that muscarinic M1 receptors do not form constitutive dimers, with ∼30% of the receptor molecules in close proximity at any given time.¹⁷

3. ROLE OF GPCR HETEROCOMPLEXES IN NEUROPSYCHIATRIC DISORDERS

Although is has been clearly established in reconstituted systems that single GPCR molecules are able to couple with and activate heterotrimeric G proteins, recent evidence also demonstrates close molecular proximity between different GPCR molecules that affects functional outcomes such as ligand binding profiles, patterns of G protein coupling, and subcellular location (Fig. 8.1). This has opened in recent years a new field of research that adds complexity to the mechanisms underlying the physiological and behavioral responses induced by signaling pathways downstream GPCRs (Table 8.1). As certain receptor subtypes, such as dopamine, serotonin, glutamate, and adenosine receptors, have been shown to form GPCR heteromers in vitro in tissue culture and in vivo in animal models, a better understanding of their structure, neuroanatomical location, and physiological function may represent a new target for the design of drugs for the treatment of neuropsychiatric disorders such as schizophrenia, depression, suicidal behavior, and drug abuse.^45–48 Here, we will review recent findings related to GPCR heteromers and their role in central nervous system (CNS) function.

Figure 8.1.

Schematic model of functional responses induced by GPCR homomers as compared to GPCR heteromers. Combinations of drugs modulate different patterns of signaling responses, and a cross talk between the components of the GPCR heteromeric complex leads to unique signaling properties.

Table 8.1.

Examples of GPCR Heterocomplexes with potential roles in neuropsychiatric disorders

GPCR heterocomplex	Experimental approaches	Neuropsychiatric relevance
A1–A2A	Cell lines, striatum (IP, BRET, FRET)	Drug tolerance
A1–D1	Cell lines, striatum (IP, BRET, BP)	Involved in PD and cocaine addiction
A1–mGlu1	Cell lines (IP)	Involved in schizophrenia
A1–5HT2A	Cortex (BP)	Involved in schizophrenia
A2A–D2	Cell lines, striatum (IP, BRET, FRET, Phar, BP)	Reversal of PD deficits, putative antipsychotic, and anticocaine properties
A2A–mGlu5	Cell lines, striatum (IP)	Possible PD, schizophrenia, and cocaine treatment
A2A–D2– mGlu5	Cell lines, striatum (IP, SRET)	Possible PD, schizophrenia, and cocaine treatment
A2A–CB1	Cell lines, striatum (IP, BRET, BP)
A2A–CB1–D2	Cell lines (BP, SRET)	Involved in schizophrenia
α2A–MOR	Cell lines, striatum (IP, BRET, FRET)	Involved in addiction
CB1–D2	Cell lines, striatum (IP, FRET, Phar, BP)	Possible PD, addiction, and schizophrenia treatment
CRH1–V1b	Cell lines (IP, BRET, FRET)	Involved in mood disorder, depression
D1–D2	Cell lines, striatum (IP, FRET)	Possible PD, schizophrenia, and treatment
D1–D3	Cell lines (BRET, FRET, Phar, BP)	Cocaine addiction
D2–D3	Cell lines, striatum (IP, BP)	Involved in PD and schizophrenia
D2–H3	Cell lines, striatum (BRET, Phar, BP)	Involved in PD
D4–MOR	Striatum (Phar)	Acquisition during drug addiction and risks for drug use initiation
mGlu2–5HT2A	Cell lines, cortex (IP, BRET, FRET, Phar, BP)	Possible schizophrenia treatment

IP, coimmunoprecipitation; BRET, bioluminescence resonance energy transfer; FRET, fluorescence resonance energy transfer; SRET, sequential resonance energy transfer; Phar, pharmacological allosteric effects in radioligand binding assays; PD, Parkinson’s disease; BP, behavioral paradigms.

4. ADENOSINE AND DOPAMINE RECEPTORS

4.1. Adenosine A1 and dopamine D1 receptors

Expression of adenosine A1 receptor and dopamine D1 receptor as a GPCR heteromer has been demonstrated using approaches such as coimmunoprecipitation, bioluminescence resonance energy transfer (BRET), and FRET.^49–53 It has also been shown that adenosine A1 receptor and dopamine D1 receptor are colocalized in soma and dendritic regions of cortical neurons. Selective adenosine A1 receptor antagonists reverse the hyperlocomotor behavioral effects induced by dopamine D1 receptor agonists, and opposite effects were shown after administration of adenosine A1 receptor agonists. These findings suggest adenosine A1 receptor as a potential target to modulate dopamine D1 receptor-dependent signaling. Importantly, suboptimal doses of adenosine A1 receptor antagonists potentiate the therapeutic-like effects of dopamine D1 antagonists in rodent models of Parkinson’s disease.

4.2. Adenosine A2A and dopamine D2 receptors

Kjell Fuxe at the Karolinska Institute and collaborators provided the first evidence of a ligand binding interaction between adenosine A2A and dopamine D2 receptors in rat striatum,^54,55 and they suggested intramembrane receptor/receptor interaction as a potential mechanism involved in this pharmacological cross talk. Based on these findings, they demonstrated the existence of adenosine A2A–dopamine D2 receptor heteromers using a variety of approaches such as coimmunoprecipitation, BRET, and FRET in vitro and in tissue culture.^56–60 An electrostatic interaction between the intracellular loop 3 of dopamine D2 receptor and the C-terminal tail of adenosine A2A receptor has been proposed as necessary for this protein–protein interaction.⁶⁰ Behavioral and microdialysis assays in whole animal models suggest the functional role of the adenosine A2A–dopamine D2 receptor heteromer in whole animal models through a mechanism that involves their coexpression in striatopallidal GABAergic neurons and nucleus accumbens.^61–72 Thus, adenosine A2A receptor antagonists reduce dopamine D2 receptor-dependent signaling and pharmacological properties,^54,73–75 and enhance therapeutic effects in rodent models of Parkinson’s disease.^65,76–81 The adenosine A2A receptor antagonist istradefylline (KW-6002) has been used in animal models of Parkinson’s disease.⁷⁷ Importantly, clinical studies with istradefylline showed a symptomatic improvement in Parkinson’s disease patients.^82,83 Bivalent ligands that bind both dopamine D2 and adenosine A2A receptors have also been proposed as a new approach to treat Parkinson’s disease.⁸⁴

The potential role of the adenosine A2–dopamine D2 receptor heteromer as a target of antipsychotic drugs has been suggested using the hyperlocomotor activity induced by amphetamine and the noncompetitive NMDA receptor antagonist phencyclidine as rodent model of schizophrenia and psychosis.^85,86 The adenosine A2A receptor agonist CGS21680 reduces the hyperlocomotor activity induced by amphetamine or phencyclidine.⁸⁷ The authors provide evidence suggesting that this antipsychotic-like behavioral effect may be mediated by inhibition of dopamine D2 receptor-dependent signaling through the adenosine A2A–dopamine D2 receptor heteromer in striatopallidal GABAergic neurons.⁸⁸

4.3. Adenosine A2A, dopamine D2, and mGlu5 receptors

In addition to adenosine and dopamine transmission, glutamate transmission plays an important role in the function of striatal GABAergic efferent neurons originating in the nucleus accumbens. In 1984, l-glutamate was found to reduce the affinity of dopamine D2 receptor agonist binding sites in rat striatal membrane preparations, providing the first evidence of the existence of glutamate receptor–dopamine D2 receptor heteromeric complexes.⁸⁹ It was also shown that adenosine A2A receptor agonists and group I mGlu receptor agonists synergistically reduce the affinity of dopamine D2 receptor agonists binding in striatal membrane preparations.⁹⁰ In line with these results, adenosine A2A receptor and mGlu5 receptor colocalize in primary cultures of rat striatal neurons⁹¹ and in striatal glutamate nerve terminals.^92,93 Similarly, adenosine A2A receptors and dopamine D2 receptors colocalize with mGlu5 receptors in rat striatum.⁹⁰ Additional data showed that adenosine A2A receptor and mGlu5 receptor agonists act synergistically to increase extracellular levels of GABA in the nucleus accumbens, which also potentiate the inhibitory effects of the dopamine D2 receptor.⁹⁴ In Parkinson’s disease, glutamate transmission is overactive mostly due to the reduced inhibitory effect of the dopamine D2 receptor,⁹⁵ and anti-parkinsonian drugs regulate functional responses that require adenosine A2A and mGlu5 receptors.^96,97

Moreover, recent data suggest that adenosine A2A antagonists and mGlu5 antagonists induce antiparkinsonian-like effects in animal models acting through the adenosine A2A–dopamine D2–mGlu5 receptor oligomer.⁹⁸ This is further supported by findings showing that the antiparkinsonian-like effects of mGlu5 receptor antagonists need coexpression of adenosine A2A and dopamine D2 receptors.⁹⁶ Since the intramembrane adenosine A2A–dopamine D2 receptor functional interaction is positively regulated by mGlu5 receptor agonists,^90,93,99,100 together, these results suggest that adenosine A2A receptor antagonists and mGlu5 receptor antagonists induce antiparkinsonian-like effects through a mechanism that requires expression of adenosine A2A, mGlu5, and dopamine D2 as a GPCR oligomeric functional unit.⁹⁸ Based on these findings, the authors proposed the combination of adenosine A2A and mGlu5 receptor antagonists to enhance their antiparkinsonian-like effects in rodent models of motor deficits, and suggested that this nondopaminergic therapy may avoid the adverse effects of dopaminergic drugs such as dyskinesia and cognitive dysfunction.^97,101,102

4.4. Adenosine A2A, dopamine D2, and cannabinoid CB1 receptors

The cannabinoid system represents an important inhibitory neuroregulator acting in the CNS.^103–107 Cannabinoid CB1 receptors are coexpressed with the dopamine D2 receptor ventral striatopallidal GABAergic neurons, and with the adenosine A2A receptor in corticostriatal glutamatergic terminals.^108–114 Activation of cannabinoid CB1 receptors by WIN55,212-2 leads to motor suppression in rodent models.¹⁰⁸ Importantly, coimmunoprecipitation and BRET experiments showed that CB1 and adenosine A2A receptors interact together in close molecular proximity in living cells and in rat striatum.¹¹⁵ It was also suggested that this GPCR heteromer is functional since some of the CB1 receptor-dependent locomotor effects were affected in adenosine A2A receptor knockout mice.¹¹⁶ Biochemical and cellular signaling assays in SH-SY5Y neuroglioblastoma cells together with behavioral tests in mice indicated that some of the effects induced by activation of the cannabinoid CB1 receptor in striatum depend upon adenosine A2A receptor-dependent signaling.^108,115 Several studies have reported antagonistic interactions between cannabinoid CB1 and dopamine D2 receptors.^117–119 Together with coimmunoprecipitation and FRET experiments,⁴⁶ the reduction of the affinity of dopamine D2 receptor agonists in the presence of cannabinoid CB1 receptor agonists point toward expression of these two receptors as a GPCR heteromer.⁴⁵ It has also been suggested that the antagonistic interactions between cannabinoid CB1 and dopamine D2 receptors as a GPCR heteromer may also involve the adenosine A2A receptor.^120–122 This is supported by the demonstration that adenosine A2A receptors directly interact with both dopamine D2⁵⁶ and cannabinoid CB1¹¹⁵ receptors. The importance of this functional cross talk has been further suggested by recent findings using a method that combined bio-molecular fluorescence complementation (BiFC) and BRET techniques and showed expression of dopamine D2, cannabinoid CB1, and adenosine A2A receptors as a higher order GPCR oligomeric complex in living cells.^123–125

In postmortem human brain of schizophrenic subjects, radioligand binding assays showed alterations in expression of adenosine A2A,¹²⁶ dopamine D2,¹²⁷ and cannabinoid CB1¹²⁸ receptors. In rodent models, chronic anti-psychotic treatment with the atypical antipsychotic drug clozapine induces downregulation of cannabinoid CB1 receptor expression in nucleus accumbens,^129,130 which has been suggested to be a consequence of a compensatory mechanism that reduces the endocannabinoid-mediated suppression of GABA release.¹³⁰ Together, these findings suggest that the adenosine A2A–cannabinoid CB1–dopamine D2 receptor heteromer may represent a potential target for new antipsychotic compounds.

4.5. Serotonin and glutamate receptors

Serotonin (5-HT) is one of the most ancient signaling molecules in evolution, and is involved in biological processes such as learning and memory, mood, food intake, sleep, reproduction, circadian rhythm, thermoregulation, pain perception, and social behavior.^131,132 Serotonin receptors also play an important role in cardiovascular, gastrointestinal, and endocrine function.¹³² Glutamate is the primary excitatory neurotransmitter in the CNS.¹³³ Pyramidal neurons represent approximately 80% of the neurons of the cortex, and glutamate serves as the neurotransmitter of the cortical pyramidal cells.^134,135 Glutamate receptors are classified either as ion channel receptors (ionotropic) or metabotropic GPCRs.^136–138 mGlu receptors have an important function in synaptic modulation throughout the CNS.¹³⁵

Recent findings demonstrate that 5-HT_2A and mGlu2 receptors colocalize at a subcellular level in mouse cortical pyramidal neurons.^20,23,24 It has also been shown that these two receptors form a GPCR heteromer in HEK293 cells with consequences on pharmacology, signaling, and behavioral effects of drugs that bind to either 5-HT_2A or mGlu2 receptors.^20,24 Radioligand binding assays showed that drugs that activate the mGlu2 receptor increase the affinity of hallucinogenic drugs for the 5-HT_2A receptor, and that drugs that activate the 5-HT_2A receptor decrease the affinity of agonists for the mGlu2 receptor.²⁰ Using ion channels in Xenopus oocytes as markers of G_q/11-dependent and G_i/o-dependent G protein signaling, recent findings demonstrate that both serotonergic and glutamatergic ligands balance the pattern of G protein signaling downstream of the 5-HT_2A–mGlu2 receptor heteromer in a way that predicts their anti-psychotic or propsychotic potential, which may provide the basis for the rationale design of new antipsychotic drugs that affect the function of this serotonin–glutamate receptor heteromer.²³

The behavioral effects of 5-HT_2A and mGlu2 as a GPCR heteromer have been recently demonstrated using hallucinogenic 5-HT_2A receptor agonists as a rodent model of psychosis. Hallucinogenic 5-HT_2A agonists, such as lysergic acid diethylamide (LSD) and mescaline, induce head-twitch behavior in mice, and this behavior is absent in 5-HT_2A receptor knockout mice.^139,140 It has been shown that the head-twitch behavioral response is significantly reduced in mGlu2 knockout mice.¹⁴¹ More importantly, viral-mediated overexpression of wild-type mGlu2 receptor in frontal cortex rescues the effects of LSD-like drugs in mGlu2 knockout mice,^24,141 and this does not occur by overexpression of mGlu2DTM4N, a construct that is unable to form a GPCR heteromer with the 5-HT_2A receptor.²⁴ These observations suggest that it is 5-HT_2A–mGlu2 receptor complex, and not the 5-HT_2A alone, which is the molecular target responsible for psychoactive-like behavioral effects of LSD-like hallucinogenic drugs in mouse models of pychosis.^23,24,141

Expression of the components of the 5-HT_2A–mGlu2 receptor complex has been shown to be dysregulated (i.e., increased 5-HT_2A receptor and decreased mGlu2 receptor) in postmortem human brain of antipsychotic-free schizophrenic subjects.²⁰ More recent findings provide evidence that the allosteric binding cross talk between 5-HT_2A receptor and mGlu2 receptor as a GPCR heterocomplex is upregulated in postmortem schizophrenia brain.^24,142 It is then possible that 5-HT_2A–mGlu2 receptor complex-dependent signaling effects may integrate serotonin and glutamate signaling, and therefore contribute to the abnormalities of thought and behavior in schizophrenia patients.

5. CONCLUSION

Understanding the structure and function of GPCR heteromers is essential for the discovery of mechanisms that define the signaling cross talk between the components of these protein complexes. Given that a large number of GPCRs are expressed in individual cells at any given time, further work is needed to better unravel the molecular rules that govern GPCR heteromeric formation, as well as the stability and lifetime of such structural units. Here, we have focused our attention on the potential role of GPCR heteromers in the treatment of neuropsychiatric disorders such as schizophrenia and Parkinson’s disease. Despite these advances, further research efforts will focus on a better understanding of the structure, pharmacology, and behavioral function of GPCR heteromers in vitro and in animal models, with the final goal of developing new therapeutic drugs that specifically affect their function.