GPCR Desensitization: Acute and Prolonged Phases

Abstract

G protein-coupled receptors (GPCRs) transduce a wide array of extracellular signals and regulate virtually every aspect of physiology. While GPCR signaling is essential, overstimulation can be deleterious, resulting in cellular toxicity or uncontrolled cellular growth. Accordingly, nature has developed a number of mechanisms for limiting GPCR signaling, which are broadly referred to as desensitization, and refer to a decrease in response to repeated or continuous stimulation. Short-term desensitization occurs over minutes, and is primarily associated with β-arrestins preventing G protein interaction with a GPCR. Longer-term desensitization, referred to as downregulation, occurs over hours to days, and involves receptor internalization into vesicles, degradation in lysosomes and decreased receptor mRNA levels through unclear mechanisms. Phosphorylation of the receptor by GPCR kinases (GRKs) and the recruitment of β-arrestins is critical to both these short- and long-term desensitization mechanisms. In addition to phosphorylation, both the GPCR and β-arrestins are modified post-translationally in several ways, including by ubiquitination. For many GPCRs, receptor ubiquitination promotes degradation of agonist-activated receptors in the lysosomes. Other proteins also play important roles in desensitization, including phosphodiesterases, RGS family proteins and A-kinase-anchoring proteins. Together, this intricate network of kinases, ubiquitin ligases, and adaptor proteins orchestrate the acute and prolonged desensitization of GPCRs.

Desensitization: a protective measure to block overstimulation

G protein-coupled receptors (GPCRs) transduce signals originating from the extracellular space to the biochemical and genetic machinery inside a cell and lead to responses including vision, olfaction, taste, vascular tone, muscle contraction, pain and behavior. While GPCR signaling is essential for normal functioning of cells, continued signaling or overstimulation can be deleterious to the survival of cells or may lead to uncontrolled growth of cells as in cancer. Accordingly, healthy living cells have tightly regulated mechanisms that blunt or desensitize GPCR signaling either transiently or over a period of time to sustain normal physiology.

Within seconds of an agonist binding to its GPCR, effector pathways downstream of the receptor are activated that include the classical second messenger cascades such as cAMP, calcium and diacylglycerol [1]. This response peaks within a few milliseconds to a few minutes depending upon the GPCR, then plateaus and rapidly declines even though the agonist-receptor complex persists [1]. Moreover, repeated stimulation of a GPCR with its agonist over minutes results in a response that is decreased compared to the initial response. These effects that limit repeated GPCR activation are referred to as desensitization (Figure 1). When desensitized, the GPCR either persists at the cell surface but becomes refractory to repeated stimulus or, in some instances, the GPCR is degraded and is restored later by protein synthesis and protein processing. When a GPCR is exposed to its agonist over a period of many minutes to hours or days, the response is significantly reduced and is associated with decreased receptor expression at the plasma membrane, through a related process referred to as downregulation [2–6]. Downregulation is a longer term process (over tens of minutes to hours) that is associated with: (1) receptor internalization into vesicles and its trafficking and destruction in the lysosomes; and (2) decreased mRNA levels effected by yet undefined mechanisms. The term tachyphylaxis refers to the decreased responsiveness of a cell to repeated stimulation with drug, and can be due to both processes of GPCR-transducer uncoupling (acute or short-term desensitization) and GPCR downregulation (long-term desensitization) (Figure 1). Signal initiation, propagation and termination processes are linked by a common cellular machinery involving GPCRs, the heterotrimeric G proteins, G protein-coupled receptor kinases (GRKs), β-arrestins, and a host of proteins that are recruited from the endocytic and ubiquitination pathways.

FIGURE 1. Acute and prolonged phases of desensitization.

Repeated stimulation with agonist results in a decreased response over a time frame of minutes (transducer uncoupling or acute desensitization). Over hours to days, this may result in decrease receptor expression at the plasma membrane (downregulation or prolonged desensitization). Together, these processes contribute to the phenomenon of tachyphylaxis.

GPCR phosphorylation: an early step in desensitization

Much of our understanding of GPCR desensitization has come from the groundbreaking experiments from the Lefkowitz group and others that have elucidated the mechanisms of desensitization at the β₂ adrenergic receptor (β₂AR). An initial important finding was that catecholamine-induced desensitization of adenylyl cyclase was associated with phosphorylation of the β₂AR [7]. Over longer periods of time, prolonged incubation of the receptors with isoproterenol was associated with a 50–60% decrease in the number of β₂AR binding sites in the plasma membrane and a pool of sequestered receptors in vesicles that were uncoupled from adenylyl cyclase [8]. Short-term desensitization was noted to occur through two parallel mechanisms. It was discovered that the receptor could be phosphorylated in response to incubation with cAMP or cAMP analogs that activated PKA in the absence of agonist [9], a process termed heterologous desensitization. However, this process only partially mimicked the effect of agonists in promoting receptor phosphorylation, suggesting another mechanism that regulated receptor phosphorylation. This other process, homologous desensitization, was agonist-specific [10] and was dependent on β₂AR phosphorylation that occurred in the absence of activation of adenylyl cyclase and generation of cAMP [11, 12]. While heterologous desensitization of the β₂AR was mediated primarily by PKA, a novel catecholamine-activated cAMP-independent kinase, termed βAR kinase (βARK, later named GRK2), was responsible for the process of homologous desensitization [11, 12].

Phosphorylation of the receptor was critical to both homologous and heterologous desensitization, as removal of distinct phosphorylation sites prevented or delayed each of these desensitization mechanisms [13] [14]. Phosphorylation of these sites was later shown to occur through the distinct activities of βARK, PKA and PKC [15]. For example, a mutated form of human β₂AR lacking a consensus PKA phosphorylation site [13] localized to one specific residue [16] had no effect on the ability of the receptor to mediate agonist-stimulation of adenylyl cyclase or undergo rapid internalization, but had an impairment in receptor phosphorylation and desensitization. Based on an analogy with rhodopsin, in which phosphorylation of the receptor by rhodopsin kinase allows the binding of visual arrestin that desensitized the receptor, the Lefkowitz group then demonstrated that retinal arrestin was capable partially restoring desensitization to phosphorylated β₂AR [17]. This suggested that analogous proteins (later discovered as β-arrestins 1 and 2) could desensitize the β₂AR physiologically. With the advent of cloning, a number of discoveries were made, including of the discovery of the GPCR superfamily [18], the cloning of the G protein receptor kinases (GRKs) [19, 20] and of β-arrestin 1 and 2 [21, 22]. The aforementioned seminal studies have largely led to our current understanding of GPCR desensitization and downregulation but other important mechanisms for desensitization have also been identified.

Proteins that affect GPCR desensitization

As noted above, there are two major protein classes, which regulate desensitization at the level of the GPCR: kinases that phosphorylate the receptor, e.g., GRKs for homologous desensitization and other kinases such as PKA and PKC for heterologous desensitization; and the β-arrestins, which recognize the phosphorylated receptor and are responsible for short-term and long-term desensitization. GRKs are members of the AGC kinase family [23] but are multidomain and multifunctional proteins that perform more roles than simply receptor phosphorylation [24]. There are seven GRK subtypes that are classified in three subfamilies based on sequence and functional similarity [25, 26]: GRK1/7, GRK2/3 and GRK 4/5/6. GRK1/7 are primarily visual GRKs and GRK4 is expressed primarily in the testes. GRKs 2/3 and GRK5/6 are thought to be ubiquitously expressed; however, the expression levels of GRKs are altered in diseased cells (for example GRK2 expression is elevated in failing cardiac myocytes [27]) and in specific tissues (for example, some dopaminergic cortical neurons have high GRK2 expression [28]).

All GRKs share an RGS-like domain, a central protein kinase domain and a variable carboxyl terminal domain (CTD), which functions in their juxtamembrane localization [29, 30]. In the GRK1 subfamily, the C-terminus can be modified, e.g, GRK1 is farnesylated while GRK7 is geranylgeranylated. GRK2/3 contains a PH domain that binds PIP₂ and G_βγ subunits, both of which cooperate to translocate the enzymes to the plasma membrane and activate the kinase [29]. The PH domain is critical for recruitment of GRK2/3 to the membrane and for phosphorylation of GPCRs [31]. The isolated PH domain of GRK2, named βARK CT based on its historical name, inhibits GRK2-dependent receptor phosphorylation by competing with GRK2 for G_βγ [31–33]. GRK4/5/6 either have polybasic domains that bind PIP₂ or have palmitoylated cysteine residues, as in GRK4 and the GRK6a splice variant [34]. Loss of GRK functions is associated with significant alterations in GPCR signaling. One of the best examples is GRK2, where loss of function, either by tissue specific knockout or by overexpression of a competitive peptide, results in unopposed adrenergic signaling [35, 36].

Since their initial discovery, a number of other roles outside of classic desensitization have been identified for GRKs. For example, the N-terminal RH domain of GRK2/3 binds G_q proteins preventing them from activating their effectors [37]. Kinase-dead GRK2 suppresses signaling of mGluR1/5 in cultured cells via the RH domain [38]. GRKs also have roles outside of regulating receptor signaling, such as a role for GRK2 in regulating the levels of reactive oxygen species in mitochondria [39] and GRK5 mediating pathologic cardiac hypertrophy through regulation of HDAC5 and interaction with NFAT [40]. Thus, while playing a critical role in regulating receptor desensitization, these kinases play a wide range of cellular roles.

Once a receptor is phosphorylated by a GRK, arrestins bind to the receptor with high affinity. Arrestin-1 and 4 are expressed in the eye, while arrestin-2 and 3 (β-arrestin 1 and 2) are expressed ubiquitously [41]. Both β-arrestins share 78% sequence homology [22] and are highly conserved across species, with ~50% sequence homology between vertebrates and invertebrates [42]. Similar to arrestin’s function in the visual system, β-arrestins were first identified for their capacity to desensitize β₂AR G protein signaling following agonist stimulation [21]. Through a number of investigations, it became apparent that the two β-arrestin isoforms shared the capability to interact with activated GPCRs, but that they differed in terms of their expression patterns, their specificity for different GPCRs, and their functional effects [43, 44]. We now appreciate that the β-arrestins regulate a diverse array of cellular processes through their scaffolding of different proteins to the receptor. These processes include MAPK (Mitogen-Activated Protein Kinase) signaling [45], receptor tyrosine kinase transactivation [46, 47], receptor trafficking [48, 49], and transcriptional regulation [50, 51] in addition to the canonical roles of GPCR desensitization and internalization [52]. The arrestins also promote internalization of receptors and recycling of receptors back to the cell surface [53]. The arrestins interact with the clathrin adaptor protein AP2 and with clathrin directly [54, 55]. β-arrestins interact with a GPCR through two interfaces, a “core” interaction at the base of the TM stack and the phosphorylated C-terminal tail [56]. While classically it was thought that β-arrestins would sterically prevent G proteins from interacting with the receptor, more recently it has been noted that β-arrestin and G proteins can bind to the receptor simultaneously, allowing the activation of G protein-mediated signaling [57]. This suggests that these proteins are capable of even broader functions than previously thought.

In addition to GRKs and β-arrestins, other proteins play important roles in desensitization. Arrestins can scaffold phosphodiesterases that degrade cAMP to localize near the agonist-activated β₂AR and turn off cAMP signals effectively [58, 59]. The RGS family of proteins, which consists of over 25 members, act as GTPase-activating proteins for heterotrimeric G proteins and inactivate G protein signaling [60, 61]. A-kinase anchoring proteins (AKAPs) act as scaffolds for localizing protein kinases and phosphatases to specific sub-cellular domains connecting them with their substrates, which include GPCRs, ion channels and neurotransmitters [62–64]. All of these proteins play important roles in regulating and limiting the cellular response to receptor stimulation.

Mammalian cells express proteins that have structural homology with the arrestins [65]. Some of these arrestin-domain containing proteins (ARRDCs) are integrated into GPCR trafficking pathways and their exact roles are being uncovered [66–68]. ARRDC2, 3 and 4 associate with internalized β₂ARs on endosomes, but are not required for β₂AR ubiquitination, internalization or lysosomal trafficking [68, 69]. However, they connect internalized β₂AR with early endosomes that contain ESCRT-0 proteins, namely HRS1 and STAM-1 [68, 69]. In addition, ARRDC3 attenuates β₂AR mobility into SNX27-containing recycling endosomal protrusions, and further promotes endosomal cAMP signaling although the exact mechanism is unknown [69].

Ubiquitination of β-arrestins: pleiotropic role in signaling and desensitization

In addition to phosphorylation, both the GPCR its adaptor β-arrestin are modified post-translationally in several ways, which regulates their stability and activity [53, 70]. One such well-studied modification is by the covalent attachment of ubiquitin(s) to the epsilon amino side-chains of internal lysines known as ubiquitination [71]. Tagging proteins by ubiquitin leads to a wide range of functional consequences such as protein trafficking, protein activation and protein degradation, although ubiquitination was originally identified as a degradation signal [72]. Both GPCR and β-arrestin become ubiquitinated upon agonist stimulation, which affects their association and intracellular trafficking in a distinct manner (Figure 2) [70, 73].

Figure 2. Roles of ubiquitination/deubiquitination in the life cycle of agonist-stimulated β₂AR.

(1) Within seconds of agonist exposure, β₂ARs stimulate G_s, and adenylyl cyclase, increasing cellular cAMP. (2) Agonist-occupied receptors are phosphorylated by GRKs on cytoplasmic domain seryl and/or threonyl residues, within seconds to minutes of agonist exposure. (3) Cytosolic β-arrestin2 (βarr2) translocates to phosphorylated receptors within 1–5 min after agonist treatment. Agonist-dependent β-arrestin ubiquitination (“U”) occurs immediately upon β-arrestin recruitment and is mediated by Mdm2 that is bound to β-arrestin. β-arrestin recruitment prevents further G protein coupling and β-arrestin ubiquitination allows it to form signaling and endocytic complexes, facilitating both receptor endocytosis and MAPK signaling. (4) β-arrestin conformational changes that occur upon receptor binding allow its interaction with Nedd4, which displaces Mdm2 from β-arrestin (5– 15 minutes after agonist treatment). (5) By interacting simultaneously with β₂AR, clathrin and AP-2, β-arrestin2 facilitates β₂AR endocytosis. (6) β-arrestin2 is deubiquitinated by USP33 starting at step 4. Nedd4 mediates ubiquitination of the internalizing β₂AR (10-15 minutes after agonist treatment). (7) Ubiquitinated β₂ARs move on into early endosomes (at > 15 minutes after activation). (8) β₂AR ubiquitination persists until about 6 h after agonist stimulation, when β₂ARs move into late endosomal/lysosomal compartments. (9) The level of ubiquitinated β₂ARs decreases, as ubiquitinated receptors are degraded in lysosomes (6 – 24 h or more after agonist stimulation). (10–12) From the early and/or late endosomes, receptors may take up an alternate path and enter recycling endosomes (<15–30 minutes or 6h after activation), in which β₂ARs become dephosphorylated by a phosphatase and deubiquitinated by USP20 (or USP33), and return to the plasma membrane as “naïve receptors”. (This figure has been modified from an originally published panel in the Journal of Biological Chemistry by Shenoy S. K. et al. 2008; 283(32):22166–22176.)

Distinct site-specific ubiquitination is induced in β-arrestin by different GPCRs upon agonist-stimulation. Mostly, this affects the ability of β-arrestin to associate with the GPCR, as well as with other partner proteins involved in endocytosis and signaling [73–75]. Although the influence of β-arrestin ubiquitination in classical G protein coupling remains unknown, stable ubiquitination of β-arrestin can suppress the G protein-dependent nuclear signaling, as shown for the Angiotensin II type1a receptor [74]. Ubiquitination of β-arrestin by the E3 ligase Mdm2 is critical for activating β-arrestin-dependent ERK downstream of the β₂AR, as well as for facilitating rapid endocytosis of the β₂AR [70, 76]. Furthermore, a β-arrestin-ubiquitin chimera enhances internalization and degradation of activated β₂AR [73]. Thus β-arrestin ubiquitination may serve as a mechanism to uncouple G proteins (acute desensitization) and to degrade internalized GPCRs (long-term desensitization), while yet promoting specific β-arrestin-dependent signaling.

Ubiquitin-dependent lysosomal degradation of GPCRs and long-term desensitization

GPCR ubiquitination, as shown initially for the β₂AR and the chemokine receptor CXCR4, is a critical tag that directs internalized receptors to lysosomes, although initial agonist-induced endocytosis of these receptors does not require receptor ubiquitination [70, 77]. Currently, 40 GPCRs have been characterized for a role of ubiquitination and the functional significance ranges from the regulation of lysosomal trafficking, recycling, and signaling, to those receptors in which ubiquitination appears to have no demonstrable function [78]. For many GPCRs, including the β₂AR, ubiquitination serves to promote degradation of agonist-activated receptors in lysosomes (Figure 2) [78]. Thus, while phosphorylation serves to uncouple the β₂AR from second messenger signaling, engendering short-term desensitization, receptor ubiquitination is required for long-term desensitization via protein degradation of agonist-activated β₂ARs.

For the β₂AR, ubiquitination on lysines in both the carboxyl tail and the third intracellular loop is critical for lysosomal targeting [79]. Both G proteins and β-arrestins have been predicted to interact with these two domains; therefore, the dynamics of ubiquitination might play a role in differential coupling or activation of these transducers in addition to ubiquitin-tagging of activated receptors for lysosomal degradation. Ubiquitination of agonist-activated β₂AR also requires both receptor phosphorylation and β-arrestin2 binding [70]. A mutant β₂AR in which all the phosphorylation sites are altered, such that there is no agonist-induced phosphorylation, shows impaired ubiquitination as well as significantly reduced β-arrestin interaction [70, 80]. In cells where the β-arrestin2 gene is deleted or its expression downregulated by RNAi, β₂AR ubiquitination is not induced by agonist-stimulation and the agonist-activated receptors are not targeted to lysosomes [68, 70]. β-arrestin2 thus functions as a critical adaptor for recruiting the E3 ubiquitin ligase NEDD4 that ubiquitinates agonist-activated β₂ARs [68, 81]. These findings have added to the existing paradigms of regulation of GPCR intracellular trafficking and the signaling roles of β-arrestin.

Recent studies have shown that under conditions of physiological stress, or during unfavorable growth conditions such as nutrient starvation of cells, agonist-activated β₂ARs are trafficked through autophagosomes in a ubiquitin-dependent manner [82]. Autophagy is a cellular degradation process in which cytoplasmic components are sequestered into newly formed membrane compartments (autophagosomes) and delivered into lysosomes for degradation [83, 84]. When cellular autophagy is triggered, the β₂AR endocytic pathway is rerouted to traffic internalized receptors through autophagy vesicles. While the exact significance of the autophagic trafficking of the β₂AR is unknown, it is intriguing that autophagy in cardiomyocytes is considered as a protective mechanism during hemodynamic stress and may influence the trafficking and signaling of cardiac βARs under stressed conditions [85].

Role of De-ubiquitinases

Ubiquitination is a reversible modification and specific enzymes called deubiquitinases trim or remove ubiquitin tags from substrate proteins [86]. Deubiquitination of β-arrestin2 by the ubiquitin specific protease 33 (USP33) diminishes β-arrestin’s capacity to associate with activated GPCRs, and with endocytic and signaling proteins [76]. β-arrestins that are impaired in ubiquitination are poorly recruited to activated GPCRs at the plasma membrane and are predicted to be weak in interdicting G protein coupling and promoting desensitization. On the other hand, deubiquitinated β-arrestins form unstable scaffolds for MAP kinases and promote little or no β-arrestin-dependent signaling. In this context, deubiquitinases such as USP33 can be regarded as ‘desensitizers’ of β-arrestin-dependent signaling.

Deubiquitination of GPCRs can regulate their lysosomal sorting, recycling and plasma membrane expression and may play an important role in receptor recycling and resensitization (Figure 2) [87]. USP33 and its homolog USP20 deubiquitinate agonist-activated β₂ARs, block β₂AR lysosomal degradation and facilitate detection of β₂ARs at the plasma membrane (due to enhanced recycling) even after prolonged agonist-stimulation [87]. USPs 20 and 33, but not their inactive forms, increase the cAMP signals obtained upon repeated stimulation of cells after an initial agonist treatment, suggesting that USPs 20 and 33 also promote efficient resensitization of β₂ARs [87]. Recent studies have shown that β₂AR activation triggers site-specific phosphorylation of USP20 by the cAMP-activated protein kinase A (PKA) at serine 333 of USP20 [82]. Only dephosphorylated USP20 shows robust activity for deubiquitinating the β₂AR. Additionally, overexpression of USP20 wild type or USP20 S333A (phosphorylation-impaired mutant) blocks the localization of internalized β₂AR in autophagic vesicles, whereas a USP20 S333D (phosphorylation-mimetic mutant) has no effect. Reversible phosphorylation of USP20 at S333 could be a critical factor that determines the rate of down-regulation of β₂ARs via autophagosomes during catecholamine stress [82].

GPCR bias and its effect on desensitization

Over the past decade, the concept of “biased agonism”, the ligand-dependent ability of a receptor to selectively signal through different effector pathways, has been an area of active study because of its potential of developing drugs with unique therapeutic profiles [88]. A number of drugs have been described that act as G protein-biased, i.e., selectively activate G proteins and not arrestins, or β-arrestin-biased ligands, i.e., selectively activate arrestins and not G proteins. Moreover, there is significant granularity within bias, as a G protein-biased ligand may selectively activate some G proteins and not others, and a β-arrestin-biased ligand may selectively activate some arrestin functions, such as internalization or signaling, and not others, such as desensitization [89–91]. Therefore, one would expect that biased agonists could have significantly different patterns of receptor desensitization than “balanced” agonists. Indeed, such effects have been shown. Chemokine receptors can frequently display bias between their multiple endogenous ligands [92]. For example, the two endogenous ligands of CCR7, CCL19 and CCL21, have been shown to be biased, differentially activating G proteins and β-arrestins [93]. This behavior is encoded by the selective engagement of GRKs, where CCL19 is phosphorylated by both GRK3 and GRK6 while CCL21 is phosphorylated by GRK6 alone [94]. Only CCL19 led to the redistribution of β-arrestin2 into endocytic vesicles and classical receptor desensitization, demonstrating that a biased response encoded in differential GRK recruitment (GRK3 in this case) can alter receptor desensitization. At the β₂AR and the CXCR4, detailed mapping of intracellular phosphorylation sites by mass spectrometry defined distinct sites for phosphorylation by individual GRKs, which were associated with distinct functions of β-arrestins [95, 96].

These findings and others [96, 97] have led to the hypothesis of a “receptor barcode” [95, 97, 98]. In this model, different kinases, and perhaps E3 ubiquitin ligases, modify the receptor C-terminus in a ligand-dependent manner. These post-translational modifications are “read” by β-arrestins and other effector proteins, resulting in distinct patterns of signaling and receptor trafficking. This model integrates our current understanding of GPCR signaling and provides a framework for understanding the allosteric regulation of β-arrestins and other GPCR effectors and relates them to biased signaling. Other examples of biased agonists with impaired desensitization include salmeterol at the β₂AR [99] and a selective D3 dopamine receptor agonist that did not induce desensitization and preferentially signaled through G proteins [100].

An example of a change in receptor signaling induced by ubiquitination is by the biased agonist “beta-blocker” carvedilol at the β₂AR, which stimulates β-arrestin2-dependent signaling and β₂AR internalization [101]. Carvedilol also induces ubiquitination of the β₂AR; however, this does not involve β-arrestin2 or NEDD4 (which are critical for agonist-induced ubiquitination), but requires a transmembrane RING-finger E3 ligase, MARCH2 (Membrane-associated RING-CH2) [102]. Interestingly, while agonist activation targets lysyl residues in the β₂AR for ubiquitination, carvedilol triggers MARCH2-mediated ubiquitination of non-lysine residues in the β₂AR. While agonist-induced ubiquitination is dispensable for β₂AR internalization, carvedilol-induced ubiquitination is required for receptor endocytosis [102]. Collectively, carvedilol while blocking G protein signaling, promotes β₂AR phosphorylation by GRK6 [95], β-arrestin binding and β-arrestin-dependent signaling [101]; these activities of carvedilol-bound β₂AR are further counteracted by MARCH2 binding, ubiquitination and β₂AR degradation ensuring long-term desensitization [102].

Additional mechanisms in desensitization

Over the past few years, endosomal signaling by GPCRs has been an active area of research [103, 104]. Classically, receptor internalization into endosomes has been associated with receptor downregulation and a loss of signaling. With the discovery of arrestin-mediated signaling, it was appreciated that receptors could signal from endosomes via β-arrestins [105]. This signaling, even when mediated through the same pathways, can have different spatial and temporal localization and is associated with distinct consequences [106]. More recently, the phenomenon of G protein-mediated endosomal signaling has been appreciated, with studies demonstrating such signaling by the PTH [107], V2 vasopressin [57, 108] and β₂ARs [109]. Similar to arrestin-mediated endosomal signaling, G protein-mediated endosomal signaling has distinct effects from G protein signaling at the plasma membrane, with β₂AR G protein endosomal signaling resulting in changes in transcriptional regulation [106]. Notably, a recent study employing cryo-electron microscopy and bioluminescence resonance energy transfer has demonstrated that both G proteins and arrestins can interact with a GPCR simultaneously in a “megaplex” conformation that is capable of activating G protein signaling [57]. This suggests that β-arrestin-mediated GPCR internalization that was thought to represent a path to receptor downregulation in the lysosomes may represent altered G protein signaling. The desensitization mechanisms that counteract such endosomal signaling are not completely understood, although in the case of the vasopressin V2 receptor, the endosomal cAMP signaling is terminated by SNX27 and the associated retromer complex [107, 108].

Concluding Remarks

Decades of research on GPCRs by the Lefkowitz lab and other research labs have enlightened us at a molecular level about how cells perceive extracellular stimuli and generate a tightly regulated biochemical response(s). This research by hundreds of scientists has uncovered a complex network of proteins that regulate GPCR signaling to ensure homeostasis. The strength, duration and continuity of GPCR signaling results from a delicate balance between the processes that turn OFF signaling (desensitization) and turn ON signaling (resensitization). GPCR resensitization is facilitated by mechanisms that reverse the events in desensitization (dephosphorylation, deubiquitination, recycling and re-synthesis); however, the mechanistic basis of resensitization is only partially understood [87, 110–113] and future work in this area should expand our understanding.

Agonist-induced GRK phosphorylation of GPCRs followed by β-arrestin binding ensure rapid uncoupling of G proteins and acute desensitization of second messenger signaling. On the other hand, additional mechanisms that include ubiquitin-dependent regulation of GPCRs and β-arrestins enable the prolonged phase of desensitization. While, historically, desensitization has referred to blockade of G protein coupling and second messenger responses, it is now clear that GPCRs can also signal via β-arrestins independent of G proteins and even signal via β-arrestin- and G protein-dependent mechanisms that are not fully understood. Changes in conformational dynamics and reversible ubiquitination of β-arrestins are potential mechanisms that can provoke or desensitize β-arrestin-dependent signaling. Understanding the mechanisms that underlie short- and long-term desensitization could provide novel insights that will aid in the development of novel drugs with increased efficacy and unique therapeutic profiles.