Adipocyte G Protein–Coupled Receptors as Potential Targets for Novel Antidiabetic Drugs Free

The increasing prevalence of obesity has been the major driver of the worldwide T2D epidemic (13,14). Obesity is characterized by the hypertrophy of adipocytes and adipose tissue inflammation, resulting in the release of free fatty acids (FFAs) and several proinflammatory cytokines into the bloodstream (14), supporting the concept that the metabolic deficits caused by obesity are due, at least in part, to chronic inflammatory processes in multiple metabolic tissues (15). This phenomenon is also referred to as “metaflammation,” which involves the cross talk between immune and metabolic pathways (15). At the same time, the secretion of anti-inflammatory adipokines, such as adiponectin, is decreased in the obese state (14). Together, these and other factors eventually cause peripheral insulin resistance, a key feature of T2D (14,16).

Based on function and appearance, adipocytes are classified into several subtypes (17). White adipocytes are essential for the storage of excess calories as triglycerides and the release of FFAs as a fuel source. Brown adipocytes are located in specific depots and express high levels of UCP1 and other thermogenic genes (18). These adipocytes function to convert chemical energy stored in substrates into heat, primarily to maintain body temperature (18). Beige (or brite) adipocytes can emerge within white adipose tissue (WAT) upon cold exposure or other conditions that involve the activation of the sympathetic nervous system (17–19). Functionally, beige adipocytes share many similarities with brown adipocytes (18,19). In humans, brown adipose tissue (BAT) was initially thought to be of physiological relevance only in newborns and infants (20). However, accumulating evidence indicates that BAT also regulates key metabolic functions in adults (20).

Given the central role of adipocytes in the pathophysiology of T2D (14), the lack of GPCR-based antidiabetic drugs that primarily target adipocytes is somewhat surprising. It is likely that such agents could prove clinically highly efficacious in the treatment of T2D and related metabolic disorders. It should be noted in this context that the therapeutic use of peroxisome proliferator–activated receptor-γ agonists of the thiazolidinedione family, a class of antidiabetic drugs that act by sensitizing adipocytes to the actions of insulin as well as by other mechanisms, underscores the capacity of adipocytes to reverse insulin resistance (21).

In this article, we discuss a series of recent studies suggesting that specific GPCR/G protein signaling cascades in white and thermogenic (brown and beige) adipocytes will emerge as useful targets for novel antidiabetic or antiobesity drugs. However, as discussed by Kusminski et al. (22), other, non-GPCR–based strategies are also under development for the therapy of obesity-associated T2D.

Previous studies have shown that adipocytes express dozens of GPCRs that are coupled to different functional classes of G proteins (3,23–26). This applies to both white adipocytes as well as adipocytes endowed with thermogenic activity (24) (Table 1). As a general rule, these receptors are expressed not only by adipocytes but also by many other cell types and tissues (3). Moreover, most GPCRs do not couple exclusively to a single functional class of heterotrimeric G proteins but show a certain degree of coupling promiscuity (27,28). The spectrum of heterotrimeric G proteins activated by a particular GPCR depends on multiple factors, including receptor density, cell type, and the differential expression of various GPCR-associated proteins (29). For this reason, it has been difficult to explore the in vivo consequences of activating a specific adipocyte G protein signaling pathway via classical pharmacological approaches.

The development of designer GPCRs known as DREADDs (designer receptors exclusively activated by designer drugs) has made it possible to turn on distinct G protein signaling cascades in a cell type–specific fashion in vivo (30,31). During the past two decades, DREADDs endowed with high selectivity for each of the four major functional classes of heterotrimeric G proteins were developed (27,30,31). The most commonly used DREADDs represent mutant muscarinic acetylcholine receptors that no longer respond to physiological concentrations of acetylcholine or other endogenous ligands (Fig. 1) (32). However, these designer GPCRs can be activated by small synthetic ligands, such as clozapine N-oxide (CNO) or deschloroclozapine (DCZ), which are otherwise pharmacologically inert, at least when used in the proper dose or concentration range (Fig. 1) (30,33). Interestingly, a recent study revealed the structural details underlying the ability of CNO and DCZ to selectively activate this new class of designer receptors (34). In the following sections, we will briefly discuss how the use of DREADD technology has advanced our knowledge about the in vivo roles of distinct GPCR/G protein pathways that are operative in adipocytes and how this novel information can be exploited for the development of novel therapeutic agents.

To explore how activation of G_s signaling in adipocytes affects whole-body glucose and energy homeostasis, two groups independently generated a mouse strain that selectively expressed a G_s-coupled DREADD (GsD) in mouse adipocytes (strain name adipo-GsD) (Figs. 1 and 2). Acute activation of adipocyte G_s signaling following treatment of adipo-GsD mice with CNO resulted in significantly reduced blood glucose levels, most likely due to elevated plasma FFA levels caused by G_s-mediated lipolysis (35,36). It is well-known that acutely elevated plasma FFA levels trigger a robust stimulation of insulin secretion, in part via activation of the FFA1 receptor (also called GPR40), a G_q-coupled receptor expressed at relatively high levels by pancreatic β-cells (37). In agreement with these findings, acute activation of adipocyte G_s signaling also resulted in a significant improvement in glucose tolerance, independent of the diet that the mice consumed (regular chow or a high-fat diet [HFD]) (35,36) (Table 2).

Chronic CNO treatment of HFD adipo-GsD mice (CNO was administered intraperitoneally [i.p.] for 4 weeks via single daily injections) resulted in reduced body weight and adiposity and improved glucose tolerance (36). These phenotypes were associated with decreased levels of blood/plasma glucose, insulin, FFAs, and leptin. In contrast, plasma adiponectin levels were increased following chronic activation of adipocyte G_s signaling (36). It is likely that the improvement in glucose homeostasis displayed by CNO-treated HFD adipo-GsD mice is a consequence of reduced adiposity, resulting in improved insulin sensitivity. Chronic CNO treatment of HFD adipo-GsD mice also promoted the beiging of WAT, increased total energy expenditure (TEE), and reduced food intake (36), suggesting that chronic activation of adipocyte G_s signaling reduces body weight and adiposity by enhancing energy expenditure and suppressing appetite. Wang et al. (36) demonstrated that the observed increase in TEE is most likely due to enhanced thermogenic activity of BAT and inguinal WAT (iWAT) caused by chronic G_s activation. In contrast, the mechanism through which adipocyte G_s signaling can inhibit food intake remains unclear at present.

In the mouse, the metabolic effects caused by selective activation of adipocyte G_s signaling can be mimicked by administration of β₃-adrenergic receptor (AR)–selective agonists such as CL316,243, consistent with the known coupling preference of the β₃-AR and other β-AR subtypes for G_s (for recent reviews, see references 38 and 39). In agreement with this observation, the ability of CL316,243 to stimulate TEE and lipolysis was greatly reduced in mice that selectively lacked the α-subunit of G_s in adipocytes (40). Interestingly, certain polymorphic variants of the human β₃-AR gene (ADRB3) have been linked to human obesity and diabetes (39). Moreover, studies with human primary adipocytes and immortalized brown/beige adipocytes have shown that the β₃-AR gene also plays a key role in regulating lipolysis and thermogenesis in human adipocytes (41). In line with these observations, the metabolic effects of mirabegron, a relatively selective β₃-AR agonist, have been explored in several clinical studies (for a review, see Chen et al. [42]; for information on mirabegron clinical trials, see https://clinicaltrials.gov/ct2/results?cond=&term=mirabegron+and+metabolic&cntry=&state=&city=&dist=&Search=Search). For example, O’Mara et al. (43) showed that chronic treatment with mirabegron resulted in an increase in BAT metabolic activity, suggesting that mirabegron will prove useful for improving impaired energy homeostasis in humans.

While β₃-AR is the predominant β-AR subtype expressed by mouse adipocytes, human adipocytes primarily express β₁- and β₂-ARs and only relatively low levels of β₃-ARs (38). Moreover, β₃-AR agonists target several other organs or tissues, including urinary bladder, heart, and brain (44). Thus, carefully designed dose-response studies are needed to explore whether β₃-AR agonists, including mirabegron, will prove clinically useful in the improvement of glucose and energy homeostasis in humans.

Besides β-ARs, adipocytes express many other GPCRs that preferentially couple to G_s, including, for example, the glucagon, melanoncortin-2, and A_2A adenosine receptors (36). Since these receptors are also present in many other tissues and cell types (3), targeting these receptors by receptor subtype–selective agonists is likely to result in considerable off-target effects. Adipocytes also express several orphan GPCRs at relatively high levels (23,45). Hopefully, future studies will clarify whether ligand-induced activation of one or more of these receptors can mimic, at least partially, the beneficial metabolic effects observed after chemogenetic stimulation of adipocyte G_s coupling.

To examine the in vivo metabolic consequences resulting from the activation of G_i/o-type G proteins expressed by adipocytes, Wang et al. (46) generated a mouse line that selectively expressed a G_i/o-coupled DREADD (GiD) in mouse adipocytes (strain name adipo-GiD) (Figs. 1 and 2). Acute, CNO-dependent activation of adipocyte G_i signaling in adipo-GiD mice resulted in marked reductions in plasma FFA, glycerol, and triglyceride levels (46), indicating that stimulation of adipocyte G_i signaling inhibits lipolysis, consistent with published data (23). At the same time, acute activation of this signaling pathway also led to significant improvements in glucose tolerance and insulin sensitivity (46), most likely resulting from reduced plasma FFA levels (47). These phenotypes were observed independently of the diet that the adipo-GiD mice consumed (regular chow or HFD) (46).

Chronic CNO treatment of HFD adipo-GiD mice (CNO was administered i.p. for 12 days via single daily injections) caused significantly reduced plasma FFA levels throughout the entire CNO administration period (46). This effect was accompanied by significant improvements in insulin sensitivity and glucose tolerance, indicating that stimulation of adipocyte G_i signaling can greatly ameliorate the metabolic impairments resulting from diet-induced obesity.

To further explore the physiological roles of adipocyte G_i signaling, Wang et al. (46) generated an additional mouse line that expressed the catalytic subunit of pertussis toxin (S1-PTX) selectively in mouse adipocytes (adipo-Gi-knockout [adipo-Gi-KO] mice). PTX functionally inactivates all α-subunits of the G_i/o family (except for Gα_z) by covalent modification of a C-terminal cysteine residue (4). Adipo-Gi-KO mice showed a series of striking metabolic phenotypes, particularly when maintained on an HFD (46). Despite reduced adiposity, HFD adipo-Gi-KO mice displayed severe impairments in glucose homeostasis and insulin sensitivity. Additional studies showed that these metabolic deficits resulted from impaired insulin action on multiple peripheral tissues or organs, including liver, skeletal muscle, and adipose tissue, such as BAT (46). Adipose tissues (iWAT, epidydimal WAT [eWAT], and BAT) from HFD adipo-Gi-KO mice showed enhanced expression of several proinflammatory cytokines and extensive infiltration by macrophages (iWAT and eWAT), indicating that adipocyte G_i/o deficiency triggers adipose tissue inflammation (46). In agreement with this observation, HFD adipo-Gi-KO mice showed significantly increased plasma levels of resistin and several proinflammatory cytokines known to contribute to peripheral insulin resistance (46). Plasma FFA levels were also significantly elevated in HFD adipo-Gi-KO mice, most likely due to the lack of G_i-mediated suppression of triglyceride breakdown (46). Since high plasma FFA levels are closely linked to insulin resistance (47), it is highly likely that elevated FFA levels contribute to the decrease in insulin sensitivity displayed by adipo-Gi-KO mice. This concept is supported by the observation that lowering of plasma FFA levels by treatment of HFD adipo-Gi-KO mice with an inhibitor of hormone-sensitive lipase, BAY 59-9435, resulted in significant improvements in insulin sensitivity and glucose homeostasis (46). These observations indicate that the metabolic phenotypes that result from disruption of adipocyte G_i signaling are generally the opposite of those found with CNO-treated adipo-GiD mice (Table 2).

The data discussed in the previous paragraphs suggest that drugs able to stimulate adipocyte G_i signaling will prove useful for improving lipid and glucose homeostasis for therapeutic purposes. Mouse adipocytes express many G_i-linked GPCRs, including, for example, the HCA₁ and HCA₂ receptors (also called GPR81 and GPR109A, respectively), the succinate receptor 1, the apelin and CB₁ cannabinoid receptor subtypes, many chemokine receptors, and several orphan receptors (46). Interestingly, the expression levels of several of these receptors are upregulated in adipocytes of obese mice, including, for example, the P2Y₁₄ and FFAR2 (GPR43) receptor subtypes and the GPR84 and GPR183 orphan receptors (46), raising the possibility that agents able to selectively stimulate these receptors will prove clinically useful at doses that do not have major effects on other tissues or cell types. Importantly, many of the G_i-coupled receptors expressed by mouse adipocytes are also expressed in human adipose tissues (23). Given the beneficial metabolic roles of activating adipocyte G_i signaling, these receptor subtypes represent potential targets for the treatment of T2D and related metabolic disorders.

To explore the metabolic outcomes of activating G proteins of the G_q/11 family in adipocytes, Kimura et al. (48) recently generated and analyzed mice that selectively expressed the Gq DREADD (GqD) in adipocytes (strain name adipo-GqD) (Figs. 1 and 2). After an acute i.p. injection of CNO, adipo-GqD mice showed marked reductions in blood glucose and plasma FFA levels in both lean and obese mice. In addition, acute CNO treatment of HFD adipo-GqD mice resulted in significant improvements in glucose tolerance and insulin sensitivity. Chronic CNO treatment (single daily injections for 11 days) of adipo-GqD mice also resulted in significant metabolic improvements in glucose homeostasis, most likely resulting from reduced plasma FFA levels (48).

Mechanistic in vivo and in vitro studies showed that stimulation of adipocyte G_q/11 signaling results in the Ca²⁺/CaM kinase kinase 2–mediated activation of AMP-activated protein kinase (AMPK), which in turn inhibits lipolysis via an inhibitory phosphorylation of hormone-sensitive lipase at S565 (48). This signaling pathway most likely is responsible for the decrease in plasma FFA levels following the stimulation of adipocyte G_q/11 proteins. Importantly, a similar pathway is also operative in human white adipocytes (48).

Additional studies demonstrated that G_q/11-mediated activation of AMPK also plays an important role in mediating the hypoglycemic effects caused by stimulation of adipocyte G_q/11 signaling (48). Studies with cultured mouse adipocytes showed that chemogenetic activation of G_q/11 leads to the phosphorylation of the Rab GTPase-activating protein AS160 (also called Tbc1d4), an event that facilitates the translocation of GLUT4 to the plasma membrane (48). It should be noted in this context that AS160 also plays a key role in mediating insulin-stimulated glucose uptake by GLUT4 in adipocytes and other insulin-sensitive cell types (49).

These observations highlight an interesting fact: while activation of adipocyte G_s or G_q/11 signaling causes reduced blood glucose levels, the two pathways have opposing effects on lipolysis. While stimulation of G_s signaling promotes lipolysis, activation of the G_q/11 pathway exerts antilipolytic activity in adipocytes (described above).

Since activation of adipocyte G_q/11 signaling mimics the hypoglycemic and antilipolytic effects of insulin, Kimura et al. (48) speculated that stimulation of this pathway can rescue the deficits in glucose homeostasis caused by impaired insulin receptor signaling in adipocytes. In agreement with this hypothesis, chemogenetic activation of adipocyte G_q/11 signaling restored normal glucose homeostasis in HFD mutant mice that harbored only one functional copy of the insulin receptor gene in adipocytes (48). These findings strongly suggest that strategies aimed at enhancing G_q/11 signaling in adipocytes may become therapeutically useful to ameliorate insulin resistance in T2D.

To explore the physiological relevance of G_q/11 endogenously expressed by adipocytes, Kimura et al. (48) also analyzed mice that lacked Gα_q selectively in adipocytes and Gα₁₁ in all body cells (adipo-G_q/11 KO mice). Independent of the diet (regular chow or HFD) that the adipo-G_q/11 KO mice consumed, these mutant mice displayed significant deficits in glucose homeostasis associated with elevated fasting plasma FFA levels (48). This observation indicates that G_q/11 proteins endogenously expressed by adipocytes contribute to the maintenance of glucose and lipid homeostasis (Table 2).

Like other cell types, adipocytes express many G_q/11-coupled receptors (3,23)_. Several of these receptors, including the cysteinyl leukotriene receptor 2 (CysLT₂ receptor), are expressed at considerably higher levels in adipocytes prepared from obese mice than from lean control littermates (48). Interestingly, CysLT₂ receptor mRNA levels determined in human subcutaneous fat also correlate well with the degree of adiposity (48). A proof-of-concept study demonstrated that i.p. injection of HFD wild-type mice with a selective CysLT₂ receptor agonist (NMLTC₄) resulted in significantly improved glucose homeostasis (48). This effect was absent in HFD adipo-G_q/11 KO mice, indicative of the involvement of G_q/11-coupled CysLT₂ receptors endogenously expressed by adipocytes. Additional work showed that adipocyte CysLT₂ receptor–G_q/11 signaling plays a similar functional role in both mouse and human white adipocytes (48). These new findings suggest that agents capable of selectively activating G_q/11 signaling in adipocytes will prove useful as novel antidiabetic drugs.

Brown and beige adipocytes (thermogenic adipocytes) play a key role in dissipating chemical energy stored in substrates (fatty acids, glucose, etc.) into heat, thus keeping body temperature in a normal range (50). Importantly, enhanced BAT activity has been linked to improved metabolic health and reduced body weight gain (24,51,52). BAT function is under the regulatory control of various GPCRs that are coupled to different functional classes of heterotrimeric G proteins (for a comprehensive review, see Sveidahl Johansen et al. [24] and references therein). For this reason, strategies aimed at modulating the activity of specific GPCR signaling pathways in thermogenic adipocytes are predicted to have translational potential for the treatment of T2D, obesity, and related metabolic disorders.

Physiologically, activation of the sympathetic nervous system, for example during cold exposure, stimulates BAT activity primarily via activation of the β-AR (β₃-AR in the mouse)/G_s cascade, resulting in a significant increase in TEE (53) (Fig. 3). However, during the past few years, additional G_s-coupled receptors have been identified that also contribute to BAT-dependent stimulation of oxygen consumption (thermogenesis). These receptors include, for example, receptors for adenosine (A_2A and A_2B receptors) (54,55), GIP (56), secretin (57,58), and glucagon (59).

Interestingly, a recent study described a novel mechanism through which the activity of thermogenic adipocytes can be enhanced via G_s signaling in a cell-autonomous fashion. Sveidahl Johansen et al. (60) reported that the expression of GPR3 (mouse gene name Gpr3), a receptor that activates G_s in a ligand-independent fashion, was greatly increased in mouse brown and beige adipocytes following cold exposure, resulting in a robust increase in energy expenditure. Additional studies showed that this phenomenon was independent of the presence of β-ARs but required a distinct lipolytic signal that is not generated by the classical β-AR–cAMP signaling cascade (60) (Fig. 3). Transcriptional induction of Gpr3 expression stimulated energy consumption in thermogenic adipocytes both in vitro and in vivo and did not require the activity of the sympathetic nervous system (Fig. 3). Gpr3 overexpression in murine brown and beige adipocytes resulted in multiple beneficial metabolic effects, including improved glucose tolerance and markedly reduced adiposity, when mice were maintained on an HFD. In vitro studies indicated that GPR3 is also functional in human BAT (60). These observations suggest that agents able to promote the expression of GPR3 in thermogenic adipocytes have translational potential for the therapy of disorders of energy and glucose homeostasis.

Since G_s-mediated increases in intracellular cAMP levels play a central role in enhancing the activity of thermogenic adipocytes, activation of pathways that counteract cAMP generation (e.g., G_i-mediated inhibition of adenylyl cyclase) are predicted to impair the function of brown and beige adipocytes. In agreement with this concept, Copperi et al. (61) recently identified a G_i-coupled receptor, GPR183 (also called EBI2), that acts as a negative regulator of brown adipocyte activity, including energy expenditure (Fig. 3). In vitro studies showed that treatment of murine or human brown adipocytes with a GPR183 agonist (7α,25-dihydroxycholesterol [7α,25-OHC]) inhibited β-AR–mediated activation of these cells. On the other hand, treatment with a GPR183 antagonist (NIBR189) or genetic inactivation of GPR183 significantly enhanced the ability of norepinephrine to stimulate cAMP production, lipolysis, and O₂ consumption in murine brown adipocytes, suggesting that 7α,25-OHC impairs the function of brown adipocytes in an autocrine fashion (61). In vivo studies with GPR183-deficient mice and the GPR183 antagonist (NIBR189) indicated that disruption of the 7α,25-OHC/GPR183 signaling cascade increases whole-body energy expenditure in response to acute cold exposure, most probably via enhancing BAT activation. These novel findings suggest that selective GPR183 antagonists will prove useful for stimulating energy expenditure and promoting weight loss for therapeutic purposes.

A recent study (62) reported that α_1A-AR, a G_q/11-coupled receptor, is expressed at relatively high levels in mouse brown adipocytes and human BAT. Interestingly, activation of α₁-AR signaling in thermogenic adipocytes could promote futile creatine cycling in thermogenic adipocytes stimulated by the β-AR-G_s–cAMP signaling cascade (62). Biophysical and pharmacological studies strongly suggested that the thermogenic effects of α_1A-AR signaling were mediated via activation of G_q/11 proteins (Fig. 3). In vivo studies with adipo-GqD mice indicated that stimulation of G_q/11 signaling in mature adipocytes caused a sustained increase in energy expenditure during simultaneous G_s activation (62). Mechanistic data indicated that this effect requires, at least partially, the activity of creatine kinase B, an enzyme that is essential for thermogenesis that results from the activation of the futile creatine cycle (63) (Fig. 3). However, studies with adipo-GqD mice and isolated brown adipocytes showed that stimulation of adipocyte G_q/11 or α_1A-AR signaling alone was unable to promote energy expenditure. These data suggest that drugs or drug combinations that stimulate both G_s and G_q/11 signaling in thermogenic adipocytes will prove beneficial to enhance energy expenditure for the treatment of various metabolic disorders.

Recent studies have led to important new insights into how the metabolic functions of white, brown, and beige adipocytes are regulated by distinct GPCR/G protein signaling pathways. The individual signaling molecules that participate in these pathways have been studied in great detail in the mouse and other animal models. Since activation of G_s signaling in both WAT and BAT leads to beneficial metabolic effects, GPCR agonists that stimulate G_s signaling in both fat depots are of particular therapeutic interest. The same is true for GPCR agonists that can activate G_q/11 signaling in both WAT and BAT. The situation is different as far as G_i signaling in the two major fat depots is concerned. In this case, antagonists that block GPCR-mediated inhibition of G_i signaling in BAT and GPCR agonists that promote G_i signaling in WAT are predicted to be of potential therapeutic usefulness.

It remains to be established in most cases whether these signaling cascades are also operative in human adipose tissues under physiological and pathophysiological conditions. Since GPCRs represent excellent drug targets, the recent findings discussed in this article are likely to stimulate the development of novel classes of GPCR-based, clinically useful drugs capable of modulating adipocyte function to improve glucose and energy homeostasis in various metabolic disorders.

Acknowledgments. The authors thank all present and past members of the Wess laboratory who contributed to part of the work discussed in this review.

Funding. The authors’ own research reviewed in this article was supported by the Intramural Research Program of the National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda, MD.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.