Fine-tuning of GPCR activity by receptor-interacting proteins

Preface

G protein-coupled receptors (GPCRs) mediate physiological responses to a large variety of different hormones, neurotransmitters, sensory stimuli and other ligands. The signaling and trafficking properties of GPCRs are often highly malleable depending on cellular context. Such fine-tuning of GPCR function can be attributed in many cases to receptor-interacting proteins that are differentially expressed in distinct cell types. In some cases, these GPCR-interacting partners directly mediate receptor signaling, whereas in other instances the interacting partners serve mainly as scaffolds to modulate G protein-mediated signaling. Furthermore, GPCR-interacting proteins can also exert profound regulation over receptor trafficking, localization and/or pharmacological properties.

Introduction

G protein-coupled receptors (GPCRs) comprise the largest family of transmembrane proteins in vertebrates and are the molecular targets for nearly half of the therapeutic drugs that are prescribed worldwide¹. The approximately 1000 members of this family can be divided into three main subfamilies (termed A, B, and C) based on sequence similarity, with all members exhibiting a conserved seven-transmembrane domain topology. The canonical view of how GPCRs regulate cellular physiology is that the binding of ligands (such as hormones, neurotransmitters or sensory stimuli) induces conformational changes in the receptors’ transmembrane and intracellular domains, thereby allowing receptor interactions with heterotrimeric G proteins. Activated receptors act as guanine nucleotide exchange factors (GEFs) for the Gα subunits, catalyzing the release of GDP and the binding of GTP to activate the G proteins. The activated G protein subunits (Gα and Gβγ) can then associate with downstream effectors to modulate various aspects of cellular physiology.

In addition to interacting with G proteins, agonist-bound GPCRs also associate with G protein-coupled receptor kinases (GRKs), leading to receptor phosphorylation. GRKs are a seven-member family of related kinases (GRK1-7) that have differential patterns of distribution across bodily tissues and distinct preferences for certain receptors^2-4. However, a common property of GPCR phosphorylation by GRKs is decreased receptor interactions with G proteins and enhanced receptor interactions with arrestins, which are members of a family of four closely-related scaffold proteins. GPCR interactions with arrestins serve to further turn off receptor signaling through G proteins while simultaneously turning on certain other signaling pathways that are initiated via arrestin-mediated recruitment of signaling proteins to activated receptors². Furthermore, arrestins can directly link active receptors to clathrin-coated pits to facilitate receptor endocytosis, which is an important process controlling the desensitization and resensitization of GPCR activity⁵.

GPCR interactions with G proteins, GRKs and arrestins have been intensively studied for a large number of receptors and exhaustively reviewed elsewhere^2-5. For this reason, these broadly-important interactions, which now represent a canonical model of GPCR regulation (Box 1), will not be further reviewed here. Similarly, the importance of homomeric and heteromeric interactions between GPCRs has also been thoroughly reviewed elsewhere^6-8 and will not be addressed here. The focus of this review will be on recent advances in the characterization of receptor-selective GPCR associations with various proteins outside of the four previously-mentioned families of general GPCR-interacting proteins (G proteins, GRKs, arrestins and other receptors). These receptor-selective partners can mediate GPCR signaling, organize receptor signaling through G proteins, direct GPCR trafficking, anchor GPCRs in particular subcellular areas and/or influence GPCR pharmacology. Since many of these receptor-selective partners exhibit limited patterns of tissue expression, these interactions can help to explain many examples of cell-specific fine-tuning of GPCR functional activity.

Main Text

Mediation of GPCR Signaling

For a GPCR-interacting protein to be considered as a mediator of GPCR signaling, it would seem a requirement that the protein’s interaction with the receptor should be regulated by agonist stimulation, since agonist-induced changes are the essence of receptor-initiated signaling. For example, GPCR interactions with G proteins, GRKs and arrestins are all strongly enhanced by agonist stimulation^1-5. Beyond G proteins, GRKs and arrestins, certain other GPCR-interacting proteins have been shown to interact with specific receptors in an agonist-promoted fashion to mediate particular aspects of receptor signaling, and these examples will be considered here in this section. In contrast, a number of other GPCR-interacting proteins associate with receptors in an agonist-independent manner. Such proteins can potentially modulate G protein-mediated signaling, as described in the next section, but they should not themselves be considered as mediators of receptor signaling if their interactions with receptors are not influenced by agonist stimulation. If a GPCR can be considered as analogous in some ways to a gun, ligand-dependent interactors that mediate signaling are analogous to bullets, whereas ligand-independent interactors that modulate signaling are analogous to silencers and scopes, which influence gun function but do not directly mediate the effects of guns on targets.

Agonist-promoted interactors that mediate signaling

Several different GPCRs have been shown to initiate cellular signaling through agonist-promoted interactions with members of the Janus (Jak) family of non-receptor protein tyrosine kinases. For example, the agonist-dependent activation of the AT1 angiotensin receptor can recruit a complex of Jak2 and the tyrosine phosphatase SHP-2 to associate with the receptor’s C-terminus, facilitating Jak2 phosphorylation and activation^{9, 10}. Activated Jak2 can then recruit and phosphorylate members of the STAT family of transcription factors, leading to STAT dissociation from Jak2 and translocation to the nucleus (Figure 1A). The agonist-promoted interaction between Jak2 and AT1 demonstrates an additional signaling avenue for AT1, beyond the receptor’s well-established coupling to Gαq, and can help to explain certain effects of AT1 stimulation on cellular physiology that are not explained by G protein-mediated signaling¹¹. Interestingly, AT1 receptor coupling to Gαq can induce rises in intracellular calcium that further potentiate Jak/STAT signaling by the receptor^{12, 13}, providing an example of how G protein-dependent and G protein-independent signaling mechanisms can in some cases work together synergistically. Another GPCR that can interact with Jak2 is the platelet activating factor (PAF) receptor, which associates with a Tyk2/Jak2 complex in an agonist-regulated manner^{14, 15}. A mutant version of the PAF receptor that does not couple to G proteins but still interacts with the Tyk2/Jak2 complex is fully capable of activating downstream STATs, demonstrating the physiological importance of the receptor’s recruitment of Jak2¹⁵.

Figure 1.

GPCR signaling can be mediated by receptor-interacting proteins. A) Certain GPCR-interacting proteins can act as mediators of agonist-induced receptor signaling, independent of G protein- or arrestin-mediated signaling pathways. An example of this phenomenon is the interaction of the non-receptor tyrosine kinase Jak2 with the angiotensin AT1 receptor. Association of Jak2 with the phosphatase SHP-2 and stimulation of AT1 with angiotensin II together promote Jak2 association with the AT1 receptor and initiation of Jak2-dependent signaling. Activated Jak2 can phosphorylate members of the STAT family of transcription factors, leading to STAT dimerization, translocation into the nucleus, and regulation of genes controlling cell growth. B) In other cases, the agonist-dependent dissociation of an interacting protein from a GPCR can alter the activity of an intracellular signaling pathway. For example, the interaction between the p85 regulatory subunit of phosphatidylinositol 3-kinase (PI3K) and the somatostatin type II receptor (SSTR2) is disrupted by agonist stimulation, leading to reduced PI3K-mediated Akt signaling and suppression of cell survival pathways.

GPCR interactions with proteins that possess PDZ domains can also in some cases mediate agonist-promoted GPCR signaling. PDZ domains, named for the first three proteins in which they were discovered (PSD-95, Discs-large and ZO-1), can mediate high-affinity interactions with specific motifs at the distal C-termini of target proteins¹⁶. For example, the PDZ protein Na⁺/H⁺ exchange regulatory factor 1 (NHERF-1; also known as EBP50) has been shown to associate in an agonist-promoted fashion with the C-terminus of the β₂-adrenergic receptor (β₂-AR)^{17, 18}. The recruitment of NHERF-1 to the β₂-AR disrupts the ability of NHERF-1 to inhibit the Na⁺/H⁺ exchanger type 3 (NHE3), providing a G protein-independent mechanism by which the β₂-AR can activate Na⁺/H⁺ exchange in kidney cells¹⁷. The κ opioid receptor (κOPR) is another GPCR that can regulate Na⁺/H⁺ exchange via agonist-induced interactions with NHERF-1^{19, 20}. Interestingly, the studies on κOPR provide an example of how interaction with a protein such as NHERF-1 can confer cell-specific signaling to a given GPCR, as κOPR stimulation robustly activates NHE3 activity in cell lines expressing high levels of NHERF-1 but not in other cell lines that lack significant NHERF-1 expression²⁰.

Agonist-disrupted interactors that mediate signaling

In addition to the aforementioned examples of agonist-promoted associations with receptor-interacting proteins mediating aspects of GPCR signaling, agonist-disrupted interactions between a GPCR and a cytoplasmic binding partner can also initiate cellular signaling. For example, it has been shown that agonist stimulation of the somatostatin receptor type 2 (SSTR2) disrupts its constitutive association with the phosphatidylinositol 3-kinase (PI3K) subunit p85 to negatively regulate PI3K signaling²¹ (Figure 1B). In the absence of agonist, association of p85 with the first intracellular loop of SSTR2 constitutively enhances PI3K activity to promote cell survival through the Akt pathway. However, following agonist stimulation of SSTR2, association of the receptor with p85 is disrupted, leading to decreased PI3K activity and sensitization of cells to stimuli that induce apoptosis²¹. Thus, even though the classical examples of proteins that mediate GPCR signaling (G proteins and arrestins) exhibit enhanced associations with receptors following agonist stimulation, it is evident that as long as a GPCR-interacting partner exhibits some type of change in its location and/or activity in response to agonist stimulation, this can be sufficient to lead to the initiation of cellular signaling.

Modulation of GPCR Signaling

Some GPCR-interacting partners increase the speed and efficiency of GPCR signaling by acting as scaffolds to tether downstream effectors in close proximity to the receptor. On the other hand, certain GPCR-interacting partners can decrease the intensity and/or time course of GPCR signaling by disrupting association with G proteins or, in some cases, recruiting negative regulators of GPCR signaling. By finely tuning the spatial and temporal resolution of signaling, certain GPCR interactors can dramatically affect the ability of GPCRs to transduce extracellular stimuli into changes in cellular physiology.

Interactors that enhance G protein-mediated signaling

The visual system of the fruit fly Drosophila melanogaster has long served as an important model system for studying fast and efficient G protein-mediated signaling. Visual signaling in Drosophila is mediated by light stimulation of rhodopsin, which couples to Gαq to activate phospholipase Cβ (PLCβ), leading to the generation of the second messengers inositol triphosphate (IP₃) and diacylglycerol and the opening of transient receptor potential (TRP) calcium channels. The influx of calcium and production of diacylglycerol lead to activation of protein kinase C (PKC), which plays a key role in the termination of visual signaling. Interestingly, almost all of the components of the Drosophila visual signaling pathway are known to be tethered together by a large PDZ-domain containing scaffold protein known as InaD, which associates with rhodopsin, TRP channels, PLCβ and PKC^22-28. By tethering these downstream effectors in close proximity to rhodopsin, InaD creates an efficient signaling complex that dramatically increases the speed and amplitude of physiological responses to light stimulation²⁹.

Analogous to Drosophila InaD, some mammalian PDZ scaffolds have been found to interact with GPCRs to enhance the efficiency of receptor-stimulated G protein signaling. For example, associations of NHERF-1 and/or the closely-related protein NHERF-2 with several different GPCRs, including the PTH1 parathyroid hormone receptor^30-34, LPA₂ lysophosphatidic acid receptor³⁵, P2Y1 purinergic receptor³⁶ and mGluR5 metabotropic glutamate receptor³⁷ can enhance PLC-mediated signaling by these receptors. Unlike the aforementioned interactions of NHERF-1 with β₂AR and κOPR, which are regulated by agonist, the associations of the NHERF proteins with PTH1-R, LPA₂-R, P2Y1-R and mGluR5 are not altered by agonist stimulation and mainly seem to facilitate enhanced G protein-mediated signaling by the receptors. Interestingly, many known NHERF-binding partners (beyond GPCRs and NHE3) are downstream components of Gαq/PLC signaling pathways, including Gαq³⁸, several TRP channels^39,40, various isoforms of PLCβ^41,36, PKC⁴² and protein kinase D (PKD)⁴³. The interaction of NHERF-2 with PTH1-R provides a particularly compelling example as to how receptor-interacting scaffolds can help to explain cell-specific fine-tuning of receptor signaling. PTH1-R signals mainly through adenylyl cyclase (AC) regulation in ROS 17/2.8 cells, which do not express detectable levels of NHERF proteins, but mainly through PLC regulation in ECV304 cells, which contain high levels of both NHERF-1 and NHERF-2³⁰. PTH1-R has long been enigmatic for seeming to signal primarily through AC regulation in some cell types but PLC regulation in other cells, and it now seems that such cell-specific differences in the receptor’s signaling activity can be accounted for in many cases by differential cellular expression of the NHERF proteins⁴⁴ (Fig. 2A).

Figure 2.

GPCR-interacting proteins can modulate G protein-mediated signaling. In the example shown in panel A, interaction between the NHERF scaffold proteins and the parathyroid hormone receptor (PTH1R) leads to a preferential enhancement of the downstream Gαq-mediated signaling by PTH1R. Through its various protein-protein interaction domains, NHERF not only binds to PTH1R, but also tethers multiple downstream signaling effectors in close proximity to the PTH1R, thereby creating an efficient complex for Gαq-mediated signaling. However, when PTH1R is in a different cell type or separate cellular compartment in which NHERF proteins are absent, PTH1R instead preferentially signals through Gαs to activate adenylyl cyclase. A contrasting example is shown in panel B. Periplakin can associate with the μ opioid receptor (μOPR) to attenuate G protein-mediated signaling. However, in cell types or cellular compartments where periplakin is not found, μOPR ligands can more robustly activate the receptor to stimulate signaling by G proteins.

In addition to the examples of the NHERF proteins and InaD, a variety of other PDZ scaffolds have been shown to associate with specific GPCRs to enhance certain signaling pathways. For example, associations of the multi-PDZ protein Mupp1 with GABA_B receptors⁴⁵ and MT₁ melatonin receptors⁴⁶ result in markedly enhanced Gαi-mediated signaling following receptor stimulation. Similarly, interactions of the PDZ scaffold MAGI-3 with Frizzled-4⁴⁷ and the LPA₂ receptor⁴⁸ enhance receptor-mediated activation of ERK/mitogen-activated protein (MAP) kinase pathways. Conversely, associations of the LPA₂ receptor with two related PDZ scaffolds, PDZ-RhoGEF and LARG, do not result in enhanced downstream MAP kinase activation but rather potentiate LPA₂-induced stimulation of Rho signaling to modify cytoskeleton dynamics⁴⁹. Thus, the LPA₂ receptor can preferentially couple to downstream PLC activation³⁵, MAP kinase activation⁴⁸ or Rho activation⁴⁹ in a cell-specific manner depending upon which LPA₂-interacting PDZ scaffold is expressed in a particular cell type.

Other prominent examples of GPCR-associated proteins that enhance the efficiency of G protein-mediated signaling are the Homer proteins, which associate with mGluR1 and mGluR5^50-53, and members of the A-kinase anchoring protein (AKAP) family, which interact with β-adrenergic receptors^54-58. In addition to associating with mGluRs, the Homer proteins also interact with intracellular IP₃ receptors, thereby serving to link the IP₃ receptors, mGluRs and other components together to greatly increase the efficiency of mGluR-stimulated calcium signaling^{52, 59, 60}. The AKAPs were originally named because of their associations with protein kinase A (PKA), and indeed the interactions of AKAP79⁵⁵ and AKAP250^{54, 56, 57} with β₂AR serve to tether PKA in the vicinity of the receptor and increase the efficiency of PKA-mediated phosphorylation of various substrates that are downstream of receptor activation, including the receptor itself. The consequences of the increased functional relationship between PKA and β₂AR include enhanced efficiency of receptor resensitization^{54, 57} and more robust β₂AR-mediated ERK signaling⁵⁵.

Interactors that reduce G protein-mediated signaling

In contrast to the examples described above of GPCR-interacting proteins that increase the efficiency of certain GPCR-stimulated signaling pathways, some proteins that associate with GPCRs can decrease the efficiency of G protein-mediated signaling. The arrestins are perhaps the most general example of this phenomenon. A more receptor-specific example is spinophilin, which has been shown to interact with the third intracellular loops of a handful of different GPCRs, including members of the dopamine, adrenergic and muscarinic acetylcholine receptor families^61-65. Spinophilin also binds to several members of the regulators of G-protein signaling (RGS) family of proteins, which tightly regulate the intensity and time course of GPCR signaling by accelerating the inherent GTPase activity of activated Gα subunits. Thus, spinophilin can tether RGS proteins in close proximity to receptors to attenuate receptor-stimulated G protein signaling^{63, 64}. Interestingly, RGS proteins can also sometimes associate directly with the intracellular regions of certain GPCRs to inhibit receptor signaling and exert cell-specific regulation of receptor activity^66-71. RGS proteins have been reviewed in detail elsewhere^{72, 73}, and will therefore not be discussed further here.

Calmodulin is another protein that can interact with a variety of different GPCRs to modulate receptor functional properties. Calmodulin is a widely-expressed calcium-binding protein that can associate in a calcium-sensitive manner with metabotropic glutamate^74-77, serotonin^78-80, dopamine^{81, 82} and other^83-85 receptors. The functional effects of calmodulin interaction vary depending on the receptor, but perhaps the most commonly-reported effect of calmodulin interaction with GPCRs is an attenuation of G protein coupling^{79, 81, 83, 85}. Since stimulation of many GPCRs can result in downstream increases in cellular calcium levels, the calcium-dependent interactions of calmodulin with GPCRs can in some cases represent a form of feedback inhibition that restrains receptor-initiated G protein signaling. However, G protein-independent signaling pathways may actually be potentiated by receptor interactions with calmodulin, as for example calmodulin association with the serotonin 5-HT2C receptor strongly promotes arrestin-mediated signaling (but not G protein-mediated signaling) by the receptor in both transfected cells and cultured neurons⁸⁰.

Two other proteins that can associate with a select number of GPCRs to tone down G protein-mediated signaling are periplakin and neurochondrin. Periplakin was first reported to associate with the C-terminus of the μ opioid receptor (μOPR)⁸⁶ and the melanin-concentrating hormone receptor 1 (MCH₁R)⁸⁷, while neurochondrin was found to interact with the same region of the MCH₁R C-terminus as periplakin⁸⁸. Both periplakin and neurochondrin have recently been shown to interact with a handful of other GPCRs in addition to μOPR and MCH1R⁸⁹. For all of the receptors that associate with periplakin and neurochondrin, the primary functional consequence is an attenuation of G protein-mediated signaling^86-89 (Fig. 2B). Since periplakin and neurochondrin exhibit discrete patterns of distribution in the brain and other tissues⁸⁹, it seems likely that they contribute to the cell-context-dependent sculpting of receptor signaling for the various receptors with which they interact. Considered together with the other GPCR partners discussed in this section, the emerging theme is that receptor-interacting proteins can exert bidirectional effects on the efficiency of G protein-mediated signaling to impart cell-specific fine-tuning of GPCR activity.

Regulation of GPCR Trafficking

GPCRs are typically trafficked to the plasma membrane to achieve functional activity. Following agonist stimulation, most receptors internalize into endosomes and may then either be targeted for lysosomal degradation or recycled back to the plasma membrane. A number of GPCR-interacting proteins have been shown to exert dramatic effects on both the biosynthetic trafficking and post-endocytic sorting of particular receptors.

Interactors that regulate biosynthetic trafficking

GPCRs must be properly folded after translation and in most cases transported to the plasma membrane to achieve functional activity. A number of GPCR-interacting proteins can regulate the folding, biosynthetic trafficking and surface expression of receptors in a cell-specific and receptor-specific manner. A classic example is the Drosophila melanogaster protein Nina A, which associates with rhodopsin to enhance the receptor’s folding and forward trafficking^{90, 91}. Similarly, Ran-binding protein 2 is a vertebrate homolog of Nina A that associates with vertebrate opsins to enhance their biosynthetic trafficking⁹². Another protein that regulates rhodopsin trafficking is the dynein light chain component Tctex-1, which directly associates with vertebrate rhodopsin to promote the receptor’s surface trafficking^{93, 94}. Other vertebrate proteins that have been found to act as chaperones to enhance GPCR surface expression in a receptor-specific manner include GEC1^{95, 96}, RACK1⁹⁷, DRIP78^{98, 99}, ATBP50¹⁰⁰, Usp4¹⁰¹, which are listed in Table 1. For each of these GPCR-interacting proteins, their expression level in a given cell type can strongly influence the level of functional receptor expression for the particular GPCRs with which they interact.

Interactor*	Associated GPCR	Location	Significance of Interaction	Ref. #
AKAP79	β2-AR	CT	Tethers PKA near receptor	55, 58
AKAP250	β2-AR	CT	Tethers PKA near receptor	54, 56, 57
ATBP50	AT2R	CT	Enhances receptor surface expression	100
Calmodulin	5-HT1A	i3L	Competes with PKC for receptor phosphorylation	78
Calmodulin	5-HT2A	i2L, CT	Impairs G protein-coupling	79
Calmodulin	5-HT2C	CT	Promotes arrestin-dependent ERK activation	80
Calmodulin	D2R	i3L	Modulates G protein signaling	81, 82
Calmodulin	mGluR5	CT	Promotes receptor surface expression	74, 75
Calmodulin	mGluR7	CT	Regulates receptor phosphorylation	76, 77
Calmodulin	PTH1R	CT	Inhibits receptor activity	85
Calmodulin	V2R	CT	Enhances receptor-induced calcium signaling	84
Calmodulin	μOPR	i3L	Inhibits G protein-coupling	83
DRIP78	AT1R	CT	Promotes receptor surface expression	99
DRIP78	D1R	CT	Promotes receptor surface expression	98
GASP1	CB1R	CT	Targets receptor to lysosomes for degradation	116, 117
GASP1	D2R	CT	Targets receptor to lysosomes for degradation	115
GASP1	δOPR	CT	Targets receptor to lysosomes for degradation	114
GEC1	EP3R	CT	Enhances receptor surface expression	96
GEC1	κOPR	CT	Enhances receptor surface expression	95, 96
Homer	mGluR1	CT	Regulates receptor signaling and localization	51, 52, 127, 153
Homer	mGluR5	CT	Regulates receptor signaling and localization	51, 52, 126, 128, 153, 154
InaD	Rhodopsin	CT	Enhances speed and efficiency of receptor signaling	22-29
Jak2	AT1	CT	Promotes Jak/STAT signaling	9, 10, 12, 13
Jak2	PAF	CT	Promotes Jak/STAT signaling	14, 15
LARG	LPA2	CT	Facilitates receptor-mediated activation of Rho	49
M10 MHC	V2R		Enhances receptor surface expression	108
MAGI-3	Frizzled-4	CT	Potentiates receptor-mediated ERK/MAPK activation	47
MAGI-3	LPA2	CT	Potentiates receptor-mediated ERK/MAPK activation	48
MPP3	5-HT2C	CT	Inhibits agonist-induced 5-HT2C internalization	139
MRAP	MC2R	NT, TM	Promotes forward trafficking of receptor	109-113
MRAPII	MC2R	NT, TM	Promotes forward trafficking of receptor	110, 113
Mupp1	5-HT2C	CT	Induces receptor clustering	155
Mupp1	GABAB	CT	Increases receptor stability	45
Mupp1	hSSTR3	CT	Targets receptor to epithelial tight junctions	144
Mupp1	MT1	CT	Enhances receptor-mediated Gαi signaling	46
Neurochondrin	MCH1R	CT	Disrupts G protein-mediated signaling	88
NHERF-1	PTH1R	CT	Enhances Gαq-mediated receptor signaling	31-34
NHERF-1	β2-AR	CT	Mediates activation of Na+/H+ exchange, promotes receptor recycling	17, 18, 122
NHERF-1	κOPR	CT	Mediates activation of Na+/H+ exchange, promotes receptor recycling	19, 20
NHERF-2	LPA2	CT	Enhances Gαq-mediated receptor signaling	35
NHERF-2	mGluR5	CT	Prolongs receptor-mediated calcium signaling	37
NHERF-2	P2Y1	CT	Prolongs receptor-mediated calcium signaling	36
NHERF-2	PTH1R	CT	Enhances Gαq-mediated receptor signaling	30
Nina A	Rhodopsin		Promotes receptor biogenesis and trafficking	90, 91
ODR-4	ODR-10		Facilitates receptor folding and surface expression	103
p85	SSTR2	i1L, CT	Mediates survival signaling by receptor	21
PDZ-RhoGEF	LPA2	CT	Facilitates receptor-mediated activation of Rho	49
Periplakin	MCH1R	CT	Impairs G protein-mediated signaling	87
Periplakin	μOPR	CT	Impairs G protein-mediated signaling	86
PICK1	mGluR7a	CT	Facilitates receptor clustering in presynaptic active zones	140, 141, 156
PSD-95	5HT2A	CT	Impairs receptor internalization and directs localization	134, 135
PSD-95	β1-AR	CT	Attenuates agonist-promoted receptor internalization	136, 138
RACK1	TXA2R	i1L, CT	Promotes forward trafficking of receptor	97
RAMP1	CRLR	NT, TM	Forms functional CGRP receptors	146
RAMP1	CT	NT, TM	Forms functional amylin receptors	149, 150
RAMP2	CRLR	NT, TM	Forms functional adrenomedullin receptors	146
RAMP3	CRLR	NT, TM	Forms functional adrenomedullin receptors	147, 148
RAMP3	CT	NT, TM	Forms functional amylin receptors	149, 150
RnBP2	Opsin		Promotes forward trafficking of receptor	92
REEP	OR		Promotes forward trafficking of receptors	104
RTP	OR		Promotes forward trafficking of receptors	104, 105
RTP/REEP	T2R		Promotes surface expression of receptors	106
RTP4	μ-δ OPR	CT	Promotes heterodimer surface expression	107
Shank	CL1	CT	Promotes receptor clustering	132
Shank	mGluR1	CT	Anchors receptor in mature dendritic spines	129
Shank	mGluR5	CT	Anchors receptor in mature dendritic spines	129
SNX1	PAR1	CT	Facilitates agonist-promoted receptor degradation	119, 120
Spinophilin	D2R	i3L	Reduces G protein and arrestin-mediated signaling	61
Spinophilin	mAChR	i3L	Reduces receptor-mediated calcium signaling	64
Spinophilin	α2-AR	i3L	Reduces receptor-mediated calcium signaling	62-65
Syntrophin	α1D-AR	CT	Enhances receptor stability	142, 143
Tctex-1	Rhodopsin	CT	Promotes apical delivery of receptor in polarized cells	93, 94
Usp4	A2AR	CT	Enhances receptor surface expression	101

^*A comprehensive list of interactors and their associated GPCRs. Interactors are organized alphabetically, followed by associated GPCR, location of interactor association on the receptor, functional significance of interaction, and references. Blanks in the table are because location of interaction is unknown. Not included on this list are GPCR interactions with G proteins, GRKs, arrestins, RGS proteins, or other receptors. Abbreviations used in the location column: CT, carboxyl-terminus; i3L, third intracellular loop; i2L, second intracellular loop; NT, amino-terminus; TM, transmembrane; i1L, first intracellular loop.

Olfactory receptors (ORs), the largest sub-family of GPCRs, have proven notoriously difficult to study in heterologous cells (e.g., cells other than olfactory sensory neurons) due to poor trafficking to the plasma membrane¹⁰². The severe trafficking deficits of ORs in heterologous cells suggest a key role for chaperone proteins in the cell-specific control of anterograde OR trafficking in olfactory sensory neurons. Indeed, the Caenorhabditis elegans protein ODR-4 is expressed exclusively in chemosensory neurons, where it regulates the forward trafficking of chemosensory receptors such as ODR-10¹⁰³. In vertebrates, two unrelated families of transmembrane proteins have been shown to associate with vertebrate ORs to enhance receptor trafficking to the plasma membrane: the receptor transporting protein (RTPs) and receptor expression enhancing protein (REEPs)^{104, 105}. Certain RTP and REEP isoforms are expressed exclusively in the olfactory epithelium, where they exert cell-specific enhancement of OR trafficking¹⁰⁴. However, other RTP and REEP isoforms are more widely expressed and can interact with other GPCRs, including T2R bitter taste receptors¹⁰⁶ and opioid receptors¹⁰⁷, to enhance receptor trafficking to the plasma membrane.

Analogous to the RTPs and REEPs, two other types of single-transmembrane proteins have been shown to control the biosynthetic trafficking of particular GPCRs. V2R vomeronasal receptors had proven very difficult to study in heterologous cells until the observation that these receptors interact with M10 MHC molecules, which associate with β2-microglobulins to strongly promote the receptors’ surface expression in heterologous cells¹⁰⁸. In a similar vein, the melanocortin 2 receptor (MC2R)-associated proteins, MRAP and MRAP2, can directly interact with MC2R and dramatically enhance the receptor’s surface expression^109-113. The physiological importance of MC2R-MRAP interactions in vivo has been well-established by studies revealing that naturally-occurring MRAP mutations cause defects in MC2R trafficking and functionality, resulting in an inherited disease known as familial glucocorticoid deficiency type 2 ¹⁰⁹.

Interactors that influence post-endocytic trafficking

Most GPCRs undergo significant endocytosis from the plasma membrane in response to agonist stimulation. In some cases, the receptors are recycled back to the plasma membrane, but in other cases the receptors are targeted to lysosomes for degradation⁵. GPCR internalization is heavily influenced by two of the canonical families of GPCR-interacting proteins, the GRKs and arrestins². However, certain other GPCR-interacting proteins can also regulate endocytic trafficking of GPCRs in a more receptor-selective fashion. For example, the GPCR-associated sorting proteins (GASPs) comprise a family of 10 proteins, with the founding member GASP1 originally identified in a yeast two-hybrid screen for proteins interacting with the C-terminus of the δ opioid receptor (δOPR)¹¹⁴. Association of δOPR with GASP1 strongly promotes receptor trafficking to lysosomes following agonist-stimulated endocytosis¹¹⁴ (Figure 3A), and has similar effects on a handful of other GPCRs, including D2 dopamine¹¹⁵ and CB1 cannabinoid receptors^{116, 117}. Most of the GASP family members are preferentially expressed in the central nervous system (CNS) ¹¹⁸, suggesting that they may act in a cell-specific manner to control the post-endocytic fate (recycling versus degradation) of certain CNS-enriched GPCRs.

Figure 3.

GPCR-interacting proteins can regulate the post-endocytic trafficking of GPCRs. Following agonist-induced receptor endocytosis, some GPCRs are targeted for proteolytic and/or lysosomal degradation, while other GPCRs rapidly recycle back to the plasma membrane. As shown in panel A, the interaction between the GPCR-interacting protein GASP1 and the δ-opioid receptor (δOPR) promotes the endocytic targeting of agonist-internalized δ-opioid receptors to lysosomes, where the receptors are degraded. In contrast, as shown in panel B, the interaction between the GPCR-interacting protein NHERF-1 and the β2-adrenergic receptor (β2-AR) promotes the rapid recycling of receptors following agonist-promoted internalization.

Beyond the members of the GASP family, a handful of other GPCR-interacting proteins have also been shown to influence the post-endocytic sorting of particular receptors. For example, sorting nexin 1 (SNX1) directly associates with the protease-activated receptor 1 (PAR1) to promote PAR1 post-endocytic trafficking to lysosomes^{119, 120}. SNX1 can also bind to the C-termini of several other GPCRs¹²¹, but how this influences the trafficking of those receptors is still unknown. In contrast to the effects of GASPs and SNX1, which associate with GPCRs to decrease receptor recycling to the plasma membrane, GPCR interactions with NHERF-1 have been found to promote receptor recycling following endocytosis^{19, 122-124}. As mentioned earlier, β₂AR and κOPR interactions with NHERF-1 mediate certain aspects of receptor signaling, while NHERF interactions with other GPCRs can modulate signaling by G proteins. Interestingly, mutant versions of β₂AR that cannot associate with NHERF-1 are targeted much more robustly than wild-type receptors to lysosomes after agonist stimulation¹²², revealing that their interaction with NHERF-1 promotes receptor recycling (Figure 3B). Association of κOPR with NHERF-1 also promotes receptor recycling over targeting to lysosomes¹⁹, and interaction of PTH1-R with NHERF-1 increases the plasma membrane retention of the receptor¹²⁵. Moreover, transplantation of the NHERF-1 binding motif onto the C-termini of receptors that do not normally interact with NHERF-1, such as δOPR, dramatically enhances the efficiency of receptor recycling back to the plasma membrane following endocytosis^{123, 124}. These findings provide an example of how certain GPCR-interacting proteins, such as NHERF-1, can act as signaling intermediates, regulators of G protein signaling, and regulators of receptor trafficking, analogous to the multiple roles played by arrestins in the signaling and trafficking of many GPCRs².

Control of GPCR Targeting

In addition to the effects of many of the aforementioned GPCR-interacting proteins on receptor signaling and trafficking, some of these interactions also control receptor anchoring to discrete regions of the plasma membrane. For example, Homer proteins not only enhance the efficiency of mGluR1/5-mediated calcium signaling (as described above), they also facilitate the clustering and anchoring of mGluR1/5 in post-synaptic dendritic spines^126-128. Consequently, mGluR1/5 receptors can be selectively localized near the post-synaptic density, positioning them to respond to glutamate that is released into the synaptic cleft. Members of the Shank family of PDZ scaffolds interact with both Homer proteins and mGluRs to further strengthen the anchoring of mGluRs to post-synaptic regions^129-131. The Shank family of scaffolding proteins also induce clustering of the latrotoxin-binding GPCR latrophilin-1 (also known as CL1 and CIRL-1) in heterologous cells and co-localize with latrophilin-1 at synapses in native brain tissue^{132, 133}.

A variety of PDZ scaffolds beyond the Shank proteins have also been shown to regulate GPCR clustering and anchoring in receptor-specific ways. For example, interactions of the post-synaptic density protein PSD-95 with the serotonin 5-HT2A receptor induce clustering of 5-HT2A in heterologous cells¹³⁴ and facilitate targeting of 5-HT2A to post-synaptic dendritic compartments in cultured cortical pyramidal neurons¹³⁵. PSD-95 also associates with the C-terminus of the β₁-AR to facilitate receptor clustering with other components of the post-synaptic density such as NMDA receptors¹³⁶. For both 5-HT2A and β₁AR, the anchoring to the plasma membrane induced by association with PSD-95, which is known to be palmitoylated and therefore tightly membrane-associated¹³⁷, greatly reduces agonist-promoted internalization of the receptors^{134, 136, 138}. In contrast, PSD-95 association with the serotonin 5-HT2C receptor actually facilitates agonist-dependent internalization, while another PDZ scaffold protein, MPP3, associates with 5-HT2C to prevent agonist-dependent internalization and stabilize the receptor at the cell surface in primary cortical neurons¹³⁹. As for other examples, interaction of the mGluR7a C-terminus with the PDZ scaffold PICK1 results in specific clustering of the receptor at pre-synaptic sites^{140, 141}, the α_1D-adrenergic receptor interacts with syntrophins to achieve linkage to the dystrophin-utrophin complex in smooth muscle cells^{142, 143}, and the SSTR3 somatostatin receptor interacts with the PDZ protein Mupp1 to be targeted to tight junctions in epithelial cells¹⁴⁴. These examples illustrate how certain GPCR-interacting partners can selectively target GPCRs to specialized cellular compartments to promote the receptors’ physiological activity.

Regulation of Ligand Binding

The vast majority of GPCR-interacting proteins described above associate with intracellular regions of the receptors and therefore have no direct effects on the pharmacological properties of the receptors, which are typically determined by the receptors’ extracellular and/or transmembrane domains. However, a handful of GPCR-interacting proteins have been shown to exert striking effects on the agonist selectivity of the receptors with which they interact. The most intensively-studied examples of GPCR-interacting partners that regulate receptor pharmacology are the receptor activity modifying proteins, RAMP1, −2 and −3 ¹⁴⁵. The RAMP proteins were initially identified in experiments designed to search for the receptor activated by calcitonin gene-related peptide (CGRP). Surprisingly, it was found that expression of a functional CGRP receptor required co-expression of an orphan GPCR, known as the calcitonin-receptor-like receptor (CRLR, also known as CL), with a receptor-interacting partner, the single-transmembrane protein RAMP1¹⁴⁶. When CRLR was co-expressed with RAMP2, this was shown to result in the formation of receptors activated not by CGRP but rather by a related peptide known as adrenomedullin¹⁴⁶. Subsequent work has revealed that CRLR can also interact with RAMP3 to form a distinct subtype of adrenomedullin receptor^{147, 148}. Moreover, expression of the calcitonin (CT) receptor with any of the RAMPs results in the formation of receptors with unique pharmacological properties, including preferential activation of some RAMP/CT receptor combinations by a distinct peptide known as amylin^{149, 150}. It is clear from work in this area that the pharmacological properties of the CT receptor and CRLR are heavily regulated in a cell-specific fashion depending on which of the RAMP proteins is expressed in a given cell type, and furthermore the RAMP proteins can also dramatically affect the surface expression levels of the CT receptor and CRLR¹⁴⁵. Thus, a comprehensive understanding of RAMP/GPCR interactions is essential for developing therapeutics that might target the various RAMP-interacting receptors¹⁴⁵.

Conclusions and Perspectives

The signaling and trafficking of most GPCRs involves receptor interactions with G proteins, GRKs, arrestins and other receptors. In addition to these widespread canonical GPCR associations, there are also many other types of GPCR-interacting proteins that can interact with particular receptors to fine-tune receptor activity. GPCRs are important drug targets, and since it is often desirable to achieve cell-specificity of drug action in order to minimize side effects, it can be clinically useful to consider the ways in which GPCRs can be differentially modulated by therapeutics depending on cellular context. Along these lines, there has been increasing interest over the past few years in studies on “biased agonists” that can preferentially activate G protein-mediated versus arrestin-mediated GPCR signaling pathways^{151, 152}. It is worth keeping in mind that agonists are likely to be differentially-biased depending on the cellular context, which is determined largely by the set of GPCR-interacting proteins that regulate receptor signaling, trafficking and/or localization in a particular cell type. Thus, uncovering the full set of GPCR-interacting proteins for receptors of clinical interest can provide novel therapeutic insights by shedding light on the fundamental mechanisms controlling the fine-tuning of GPCR activity.

Box 1. Canonical mechanisms of GPCR signaling.

A) In the classical view of GPCR signaling, agonist binds to extracellular and/or transmembrane regions of the receptor, leading to interactions with heterotrimeric G proteins. The receptor acts as a guanine nucleotide exchange factor, catalyzing the exchange of GDP for GTP on the Gα subunit and inducing dissociation of the Gα and Gβγ subunits. Activated GTP-Gα subunits, of which there are multiple sub-types including Gαs, Gαi, Gα12/13, and Gαq, subsequently bind to and regulate the activity of effectors such as adenylyl cyclase, RhoGEF, or phospholipase C, thereby generating second messengers that modulate further downstream effectors. The Gβγ subunits can also bind to and regulate certain downstream effectors, such as ion channels and phospholipase C, following their liberation from the heterotrimeric complex.

B) G protein-mediated signaling by agonist-activated receptors can be terminated via receptor phosphorylation by G protein-coupled receptor kinases (GRKs) and concomitant receptor association with arrestins, which interact with clathrin (black shapes) and the clathrin adaptor AP2 (green circles) to drive receptor internalization into endosomes². This physiological process of receptor internalization regulates the functional process of receptor desensitization. However, in addition to attenuating G protein-mediated signaling and facilitating receptor internalization, recruitment of arrestins to activated GPCRs can also lead in many cases to the initiation of distinct arrestin-mediated signaling pathways. Following association with arrestins, GPCRs may then be trafficked to lysosomes, where they are ultimately degraded, or to recycling endosomes for recycling back to the cell surface in the functional process of resensitization. Interestingly, “biased” agonists have been recently characterized that specifically activate G protein-mediated signaling pathways over arrestin-mediated receptor signaling pathways or vice versa^{151, 152}. This new concept illustrates the importance of characterizing all GPCR downstream signaling pathways in order to more fully exploit the therapeutic potential of clinically-important receptors.