CrossTalk proposal: Weighing the evidence for Class A GPCR dimers, the evidence favours dimers

Evidence has accumulated for functionally relevant dimers of class A G protein-coupled receptors (GPCRs). Much work remains to irrefutably demonstrate their functional importance in vivo, but overall available data strongly support an important role for GPCR dimers. Binding experiments provided initial evidence that GPCRs were multimeric, with allostery explaining cooperativity measured between different equivalents of ligand (De Lean et al. 1980). However, this was overshadowed by the discovery that G proteins allosterically modulated receptors, focusing attention on receptor–G protein interactions. Ligand binding studies explicitly suggest that oligomeric receptors explain such cooperativity (Chidiac et al. 1997; Ma et al. 2007). Conceivably, a ‘shared’ G protein between two GPCR monomers might be the agent through which such cooperativity is manifested (Chabre et al. 2009). However, allostery between ligand binding sites in receptor homodimers can be measured even when G proteins have been removed from receptor preparations (Wreggett & Wells, 1995). In a recent study, cooperativity in ligand binding in M2 muscarinic receptor dimers, reconstituted with G protein, was lost when receptor monomers were studied (Redka et al. 2013). This argues that allostery is manifested between receptors through direct physical contact. Structural studies suggest that GPCRs are likely to be arranged asymmetrically with respect to a shared heterotrimeric G protein (Mesnier & Banères, 2004; Jastrzebska et al. 2013). Studies using binding-impaired mutant D2 dopamine and 5-HT4 receptors have also demonstrated asymmetries between GPCR dimers and G protein coupling (Han et al. 2009; Pellissier et al. 2011). G proteins or effectors also have allosteric impacts on ligand binding, as parts of larger complexes, with unique patterns of allostery depending on which proteins are in contact with each other during signal transduction. Historically, biochemical evidence provided initial insight into how arrangments of receptors and their signalling partners might be organized in cells. Approaches such as radiation inactivation, co-immunoprecipitation, cross-linking, gel filtration and reconstitution of chimeric receptors all suggested GPCRs can exist as dimers (Milligan & Bouvier, 2005). Criticism raised regarding these approaches suggested that cell disruption and protein solubilization resulted in artefactual aggregations despite rigorous controls. Purified receptors reconstituted into proteoliposomes that can accommodate only monomers yield functional receptors that activate G proteins (Whorton et al. 2007, 2008). Although these results clearly show that class A GPCRs such as the β2-adrenergic receptor can function as monomers in vitro, they do not argue either for or against the existence of GPCR dimers in vivo. Interestingly, reconstitution of purified receptors under conditions leading to dimers, resulted in complexes stabilized by inverse agonists, suggesting that dimers are relevant to receptor activation state (Fung et al. 2009). For rhodopsin, reconstituted oligomers resulted in accelerated kinetics of transducin activation compared to monomers, suggesting again that the oligomeric state may control state transitions during receptor activation (Whorton et al. 2008). Regulation of the active state was also shown in cell-based assays assessing dopamine D2 receptor dimerization (Han et al. 2009). Here, a minimal signalling unit composed of two receptors and a single G protein was maximally activated by agonist binding to a single protomer, whereas agonist binding to the second protomer blunted signalling. Diverse functional consequences of co-expressing and/or stimulating two different receptors suggests multiple functionally relevant hetero-dimers (see Milligan & Bouvier, 2005). The proposed functional consequences of hetero-dimerization include altered ligand binding and signalling selectivity and heterodimer-selective receptor trafficking. A possible role for molecular crosstalk between receptors, rather than true physical interactions, cannot be excluded. Mutating the putative dimerization interfaces can lead to intracellular retention of receptor (Kobayashi et al. 2009) suggesting that dimerization may play a role in quality control and export of GPCRs. In response to criticisms raised regarding biochemical approaches used to study GPCR dimerization, several biophysical methods were developed to examine GPCR signalling in living cells. These include resonance energy transfer approaches such as FRET or BRET and protein complementation assays that suggested that many GPCRs form homo- and heterodimers (see Milligan & Bouvier, 2005). Although these techniques are powerful tools to monitor protein complexes in living cells, these assays detect proximity between proteins and not direct protein–protein interactions. Multiple approaches with rigorous controls must be performed to support the existence of GPCR dimers. Another complication in using proximity-based assays to study GPCR dimerization is that, with few exceptions, dimerization is not influenced by ligand binding, rendering use of such pharmacological tools uninformative. This has led to the suggestion that GPCR dimerization is constitutive as observed for other multimeric membrane proteins. More recently, studies using single particle tracking and image analysis brought further support to the existence of class A GPCR dimers in cellulo (Fonseca & Lambert, 2009; Hern et al. 2010; Calebiro et al. 2013). In addition to allowing visualization of dimers in real time, these studies revealed that dimerization may be a dynamic process at the plasma membrane with dimers in an equilibrium with monomers. Varying levels of dimer stability were observed for different receptors suggesting that this proportion varies among receptors (Calebiro et al. 2013). Structural evidence supporting dimerization of class A GPCRs was provided by atomic force microscopy of rhodopsin (Fotiadis et al. 2003) that revealed arrays of dimers. Similar images were obtained from native samples (Liang et al. 2003). More recently, transmission electron microscopy confirmed rhodopsin as a dimer that interacts with a single G protein (transducin) (Jastrzebska et al. 2013). No doubt such organization has important mechanistic implications for receptor homo- and heterodimers as suggested by studies on cooperativity discussed above. Dimers were detected in some early rhodopsin crystals (Whitten et al. 2000) but were largely attributed to crystallization artifacts. The first β2-adrenergic receptor structure was also solved as a dimer (Rosenbaum et al. 2007) but interactions involved limited protein–protein contacts and occurred mainly through a cholesterol molecule present at the interface, again raising questions about the dimer's biological relevance. A number of GPCR crystal stuctures have been solved showing dimeric organization, μ- and κ-opioid receptors (Manglik et al. 2012; Wu et al. 2012) and CXCR4 chemokine receptors (Wu et al. 2010). Because crystallization involves purifying receptors, stripping away partners and native lipid environments and letting them pack without chaperones shaping their assembly, it is unclear if absence or presence of dimers in crystals represents native quaternary structure. Additional studies will define this organization in situ. Ultimately, proof that class A GPCR dimers are physiologically relevant must come from in vivo studies. Recent work brings substantial support to the functional importance of dimerization. A study using knockout mice showed that MT1 and MT2 melatonin receptors function as heterodimers in rod photoreceptors (Baba et al. 2013). Loss of either receptor or over-expression of non-functional MT2 mutants abolished melatonin signalling. Heterodimeric 5-HT2A/mGluR2 receptors have been co-immunoprecipitated from mouse brain and are important for integrating effects of antipsychotics (Gonzalez-Maeso et al. 2008; Fribourg et al. 2011). Finally, fluorescent ligands/labelled antibodies in time-resolved fluorescence experiments showed GPCR dimers in native tissues (Albizu et al. 2010; Cottet et al. 2013). Taken together, support outbalances skepticism regarding the function and relevance of GPCR dimers. Dimers offer the simplest explanation for the ligand-binding cooperativity observed in native systems. Understanding the full scope of GPCR dimerization will remain a very active area of research. Readers are invited to give their views on this and the accompanying CrossTalk articles in this issue by submitting a brief comment. Comments may be posted up to 6 weeks after publication of the article, at which point the discussion will close and authors will be invited to submit a ‘final word’. To submit a comment, go to http://jp.physoc.org/letters/submit/jphysiol;592/12/2439 Michel Bouvier is currently a senior scientist at the Institute for Research in Immunology and Cancer (IRIC) where he holds the Canada Research Chair in Signal Transduction and Molecular Pharmacology. His conceptual and experimental contributions to the field of G protein-coupled receptors (GPCRs) have led to major paradigm shifts (e.g. discovery of inverse agonism, receptor oligomers and pharmacological chaperones) with significant impact on both drug discovery and fundamental notions about pharmacology as a discipline. Terry Hébert is a Professor in the Department of Pharmacology and Therapeutics at McGill University. He has held a Chercheur National Award from the Fonds de Recherche en Santé du Québec (FRSQ) and was a McDonald Scholar of the Heart and Stroke Foundation of Canada. His focus is on the ontogeny, formation and trafficking of GPCR-based signalling complexes with a view toward understanding the architecture, wiring and integration of signalling pathways. He has developed new methods for in cellulo assays designed to facilitate drug discovery. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article. None declared.