The role of compartmentalized β‐AR/cAMP signaling in the regulation of lipolysis in white and brown adipocytes

Abstract

White and brown adipocytes are central mediators of lipid metabolism and thermogenesis, respectively. Their function is tightly regulated by all three β‐adrenergic receptor (β‐AR) subtypes which are coupled to the production of the second messenger 3′,5′‐cyclic adenosine monophosphate (cAMP). While known for decades in other cell types, compartmentation of adipocyte β‐AR/cAMP signaling by spatial organization of the pathway and by cAMP degrading phosphodiesterases (PDEs) as well as its role in the regulation of lipolysis is only beginning to emerge. Here, we provide a short overview of recent findings which shed light on compartmentalized signaling using live cell imaging of cAMP in adipocytes and discuss possible future directions of research which could open up new avenues for the treatment of metabolic disorders.

Adipocyte function is regulated by all three β‐adrenergic receptor subtypes via the second messenger cAMP. This type of signaling is highly confined in subcellular microdomains or signalosomes by the action of various phosphodiesterase (PDE) families. This review highlights recent insights into cAMP compartmentation in white and brown adipocytes gained by live cell imaging techniques and how it regulates lipolysis under healthy and diabetic conditions as well as during adipocyte maturation.

Abbreviations

AC

adenylyl cyclase

AKAPs

A‐kinase anchoring proteins

ATP

adenosine triphosphate

BAT

brown adipose tissue

cAMP

3′,5′‐cyclic adenosine monophosphate

cGMP

3′,5′‐cyclic guanosine monophosphate

EPAC

exchange protein directly activated by cAMP

FFA

free fatty acids

FRET

Förster resonance energy transfer

GPCRs

G protein‐coupled receptors

Gs

stimulatory heterotrimeric guanine nucleotide regulatory proteins

HSL

hormone sensitive lipase

ISO

isoproterenol

OPA

Optic atrophy 1

PDEs

phosphodiesterases

PKA

cAMP‐dependent protein kinase

T2D

type 2 diabetes

TG

triglycerides

UCP1

uncoupled protein 1

WAT

white adipose tissue.

β‐ARs

beta‐adrenergic receptors

Introduction

Dysregulated lipid metabolism leads to lipotoxicity and hyperlipidemia. Two states that are strongly associated with the metabolic diseases are obesity and type 2 diabetes (T2D). Furthermore, these alterations have been shown to further exacerbate not only these metabolic diseases but also the development of associated pathologies, such as cardiovascular disease [1, 2, 3]. While lipotoxicity and hyperlipidemia are characterized by an excess of fatty acids, in the form of stored triglycerides (TGs) within adipocytes and other organs and of an elevated level of free fatty acids (FFAs) in the plasma, the pathophysiology of these states extends beyond an excess supply or presence of fats. Indeed, our understanding of how lipid metabolism is regulated in health and dysregulated in disease is an exciting, evolving field of research that aims to provide integral insights in understanding and treating lipotoxicity and hyperlipidemia.

Adipocytes are subdivided into three types, white adipocytes which store energy as a single large lipid droplet, and brown and beige cells (brown‐like adipocytes in white adipose tissue) which exhibit multiple small lipid droplets and express genes for the thermogenic program. Lipolysis of stored TGs in adipocytes is stimulated via beta‐adrenergic receptor (β‐AR)/3′,5′‐cyclic adenosine monophosphate (cAMP) signaling and inhibited via cAMP hydrolyzing enzymes, phosphodiesterases (PDEs). PDEs are a superfamily of cyclic nucleotide hydrolyzing enzymes consisting of 11 families with PDE4, PDE7 and PDE8 capable of hydrolyzing exclusively cAMP, and PDE1, PDE2, PDE3, PDE10 and PDE11 known as dual‐substrate PDEs which hydrolyze both cAMP and 3′,5′‐cyclic guanosine monophosphate (cGMP) [4, 5, 6]. The ‘fight‐or‐flight’ signals catecholamines: adrenaline (through hormonal stress responses) and noradrenaline (through sympathetic nervous system) activate β‐ARs on multiple target cells. Physiologically, β‐ARs are canonical seven transmembrane G‐protein coupled receptors (GPCRs) coupled to intracellular heterotrimeric G‐proteins, controlling the functions of white, beige, and brown adipocytes. The three β‐ARs subtypes are expressed in adipocytes, β₁‐, β₂‐ and β₃‐AR [7, 8]. Upon extracellular stimulation, the β‐ARs couple predominantly to stimulatory G‐proteins (G_s), activating adenylyl cyclases (ACs) at the plasma membrane resulting in the production of cAMP from adenosine triphosphate (ATP). cAMP then goes on to activate effector molecules such as protein kinase A (PKA), exchange protein activated by cAMP (EPAC), cyclic‐nucleotide‐gated ion channels and Popeye domain containing proteins [9, 10], resulting in a broad range of downstream functional responses, such as regulating lipolysis, glucose uptake, thermogenesis, and adipocyte differentiation. While all three β‐ARs subtypes can be stimulated via the same hormones to produce cAMP, the downstream cAMP signaling and functional responses are reported to be different.

These divergent cAMP functions are thought to be made possible by the distinct compartmentation of cAMP within functionally relevant micro‐ or nanodomains, alternatively referred to as signalosomes. Indeed, in cardiac cells and neurons for example, rather than being freely diffusible, cAMP signaling has been shown to be compartmentalized at the subcellular level. This is possible due to the subcellular organization of cAMP degrading enzymes PDEs, effector proteins and A‐kinase anchoring proteins (AKAPs) (that tether PDEs and effector proteins [11, 12] at the subcellular level in microdomains/signalosomes) [13, 14, 15], thus allowing for the high spatial and temporal fidelity of β‐AR/cAMP signaling observed in the heart and neurons. While a similar β‐AR/cAMP compartmentation is hypothesized in adipocytes, we are only in the early stages of our understanding how this compartmentalized β‐AR/cAMP signaling occurs within adipocytes and how/whether this may be dysregulated in disease.

Role of cAMP signaling in white adipocytes

Considering the function of β‐AR/cAMP signaling specifically in white adipocytes, in the early 1970s, cAMP was reported to promote lipolysis of stored TGs [16]. β‐AR stimulated increases in cAMP lead to the activation of lipases, such as hormone sensitive lipase (HSL), that catalyzes the lipolysis of TGs to glycerol and FFAs, which can then be used by other tissues to produce ATP. Later on, it was shown that in the fed state, when this energy supply is no longer needed, insulin acts to decrease lipolysis via the activation of PDE3, which reduces cAMP levels and subsequent lipolytic activity [17]. This work solidified the essential action of cAMP as an on and off switch for lipolysis and the action of insulin and PDE3 to mediate this switch. In the early 1980s, an additional level of β‐AR regulation was uncovered in studies using healthy adult males infused either with the unselective β‐AR agonist isoproterenol (ISO) or relatively β₁‐, β₂‐, and β₃‐AR subtype‐specific agonists (dobutamine, terbutaline and CGP12177, respectively). This approach could identify varying increases in glycerol levels in the white adipose tissue depending on the β‐AR subtype stimulated, suggesting that the β‐AR subtypes differ in their effectiveness to stimulate lipolysis [18]. Subsequent work aimed at understanding β‐AR‐cAMP mediated regulation of lipolysis has been hindered by our ability to consolidate findings from subsequent studies that have been performed in many different murine and cell line models, which differ in their affinities for β‐AR agonists and antagonists, expression levels of β‐ARs and PDEs and indeed, even in the β‐AR subtypes that regulate lipolysis. In human white adipose tissue for example, it has been reported that the β₁‐AR subtype and to a lesser degree the β₂‐AR subtype are the main β‐AR subtypes responsible for stimulating lipolysis, with the β₃‐AR subtype reported to exhibit little to no lipolytic function in healthy adipose tissue [19]. Supporting this, the β₁‐AR and β₂‐ARs reported to exhibit a greater affinity to noradrenaline compared to the β₃‐AR [20]. In contrast, in murine models and in many cell lines, the β₃‐AR subtype is reported to be the predominant β‐AR subtype that stimulates lipolysis [21]. It has therefore become essential for us to carefully consider the model we chose and its appropriateness to address the research question in hand, in order to obtain results with high translational potential and prevent us from adding to the discrepancies between study findings/conclusions.

Evidence for the presence of cAMP microdomains/signalosomes in white adipocytes

As mentioned, the presence of β‐AR/cAMP microdomains in other cell types, such as in the heart have been extensively studied. In this past work, cutting edge live cell imaging research has been applied to cardiomyocytes allowing for the identification of distinct alterations in this β‐AR/cAMP microdomains to be involved in the underlying pathophysiology of traditional heart failure and obesity and T2D associated heart failure with preserved ejection fraction [22, 23, 24]. However, as cardiomyocytes and adipocytes are functionally and structurally distinct, these past findings relating to the presence, formation and function of cAMP micro‐ or nanodomains cannot be directly applied to adipocytes.

There have been some key insights supporting the presence of cAMP micro‐ or nanodomains in the adipocyte. An interesting study for example, investigating the interactions between insulin and β‐AR/cAMP signaling in adipocytes identified that disruption of cAMP‐dependent PKA scaffolding proteins that tether PKA in the cell has a similar effect on blunting β‐AR/cAMP signaling, as chronic insulin signaling [25]. The authors hypothesized that blunted β‐AR/cAMP signaling observed in chronic insulin signaling states is mediated, in part, by a loss of PKA scaffolding proteins within close proximity to β‐ARs. There is also evidence to support the subcellular distribution of PDEs within distinct cellular locations of the adipocyte, with PDE3B for example, found to localize within caveolae plasma membrane fractions of primary adipocytes, an important site for insulin signaling [26]. Indeed, in our work, we have identified for PDE3B to localize within plasma membrane microdomains of white adipocytes derived from human mesenchymal stem cells, and for this localization to be disrupted in insulin resistance states (in which PDE3B also lost its ability to blunt lipolysis) shifting from the plasma membrane to endoplasmic fractions [27]. Metformin treatment was shown to restore PDE3B localization to the plasma membrane fractions, which was accompanied by a restoration of insulin mediated inhibition of lipolysis and PDE3B activity.

We sought to further investigate the presence and function of micro‐ or nanodomains in the adipocyte with the application of advanced live cell imaging techniques in white adipocytes derived from human mesenchymal stem cells [28]. This live cell imaging technique is based on a highly sensitive Förster resonance energy transfer (FRET) biosensor which can report intracellular cAMP levels in real time [29, 30]. In this work, for the first time, we were able to not only identify the presence of β‐AR/cAMP microdomains in white adipocytes derived from human mesenchymal stem cells but visualize them and gain key insights into how alterations of these β‐AR microdomains lead to lipolytic dysregulation in insulin resistance [27].

Prior to the application of FRET in white adipocytes, previous work had been restricted to assessing cAMP levels and dynamics in whole white adipose tissue or cell lysates using classical biochemical methods such as enzyme‐linked immunosorbent assays and radioimmunoassays. While these studies provided key insights into the role of β‐AR/cAMP signaling in lipolysis, these methods lack the sensitivity and high spatial and temporal resolution needed to detect functionally relevant subtle changes in cAMP concentrations and to track these changes within single cells and within distinct subcellular compartments. Perhaps one of the most interesting findings from our work with FRET in white adipocytes, was that selective stimulation of each of the three β‐ARs induced comparable cytosolic cAMP amplitudes, however the lipolytic responses varied for which we identified was due to distinct PDE coupling and activities within each β‐AR microdomain [27].

This study showed that in healthy white adipocytes, basal lipolysis was mainly regulated by PDE3 and to a lesser extent by PDE4 [27]. ISO stimulated lipolysis was predominantly mediated by β₁‐AR signaling and controlled by PDE4. In contrast, β₂‐AR stimulated lipolysis was strongly suppressed by PDE3 and PDE4 activities, whereas PDE3 was almost exclusively responsible for the inhibition of β₃‐AR stimulated lipolysis (Fig. 1A). In insulin resistant adipocytes (induced via high glucose treatment), we identified an increased basal lipolysis and shift from catecholamine stimulated lipolysis via the β₁‐AR to the β₃‐AR, resulting in increased catecholamine induced cAMP amplitudes and associated lipolysis. With the application of selective PDE inhibitors, we identified this shift to β₃‐AR control to be mediated via a loss of basal and β₃‐AR coupled PDE3 activity (Fig. 1B). Interestingly, when we applied metformin treatment to the cells we were able to restore insulin sensitivity. This was accompanied by a restoration of PDE3 coupling which normalized both basal and β₃‐AR stimulated lipolysis. Therefore, this work has uncovered at least three distinct microdomains of cytosolic cAMP engaged by the β₁, β₂ and β₃‐ARs regulated by specific PDE families. The requirement to maintain PDE mediated regulation of cAMP within these microdomains was evident in insulin resistant adipocytes, where a loss of PDE3 coupled to the β₃‐AR/cAMP microdomain was sufficient to induce a β₃‐AR mediated lipolytic response that was not present in healthy control adipocytes. The therapeutic potential of targeting cAMP microdomains in insulin resistance was demonstrated by the ability for metformin treatment to restore β₃‐AR associated PDE3 activity and blunt lipolysis induced via β₃‐AR stimulation. Therefore, direct therapeutic targeting of local regulatory mechanisms could be beneficial for diseases with dysregulated lipolytic signaling [27].

Fig. 1.

Regulation of β‐AR/cAMP compartmentation in human white adipocytes by PDEs. (A) In a healthy human white adipocyte (WA) cell line derived from mesenchymal stem cells, lipolysis is predominantly mediated via β₁‐AR signaling, while β₂‐AR and β₃‐AR/cAMP responses are strongly compartmentalized by PDE4D and PDE3B, respectively. (B) In insulin‐resistant adipocytes, after high glucose treatment, levels of PDE3B are significantly reduced which allows a switch to β₃‐AR‐induced lipolysis due to loss of β₃‐AR‐PDE3B coupling. Signaling by other receptor isoforms can be also fine‐tuned by local reduction of PDE4 activity at β₂‐AR and global reduction of relatively low PDE2A activity. The described switch mechanism can be counteracted by metformin, an anti‐lipolytic drug, which restored the reduced PDE3B activity.

In addition to being an essential reservoir of energy, white adipocytes and white adipose tissue also possess essential endocrine functions, secreting numerous hormones, growth factors, enzymes, cytokines, complement factors, and matrix proteins. While this review is focused on lipolysis, it is important to note that these functions are not distinct but are interconnected. Therefore, the function of cAMP micro‐ or nanodomain formation and consequences for disruptions in these micro‐ or nanodomains in obesity and T2D may not be restricted to lipolysis. Indeed, for example, cAMP has been shown in 3T3‐L1 cells to stimulate exocytosis and adipokine release, that was observed to be EPAC and not PKA dependent [31]. Supporting the requirement for cAMP micro or nanodomains within the adipocyte to allowing targeted cAMP activity.

Role of cAMP signaling in brown adipocytes

Brown adipose tissue (BAT) is a highly vascularized organ containing brown fat cells highly enriched with mitochondria. The main function of BAT is to dissipate chemical energy in the form of heat (thermogenesis) [32]. Moreover, BAT secretes adipokines (‘BATokines’) [33] and exosomes [34, 35] which might influence appetite, metabolic crosstalk with other organs and regulate whole‐body metabolism. Thermogenesis is mediated by a mitochondrial thermogenic protein called uncoupling protein 1 (UCP1), which is activated by cold exposure, adrenergic stimulation and other pharmacological alterations. The existence of brown fat has been reported in adult humans [36, 37, 38, 39]. In obesity and T2D, BAT has been shown to alleviate metabolic complications such as dyslipidemia, impaired insulin secretion, and insulin resistance in numerous models. A recent study employing ¹⁸F‐fluorodeoxyglucose positron emission tomography‐computed tomography scans in large human cohorts for example, identified for the presence of BAT to independently correlate with lower odds of T2D, dyslipidemia and cardiovascular diseases. Beneficial effects of BAT were more pronounced in overweight or obese individuals, suggesting that BAT may be involved in mitigating clinically relevant complications associated with obesity and T2D [40, 41].

Considering evidence to target cAMP signaling in treating obesity/T2D, selective inhibition/deletion of PDEs shows promise as an effective strategy to prevent and treat obesity and associated metabolic disorders. Möllmann and colleagues administered roflumilast, a selective PDE4 inhibitor and FDA approved drug for the treatment of pulmonary obstructive disorders to mice fed with a high‐fat diet. They were able to observe a significant reduction in body weight gain, higher energy expenditure, improved glucose tolerance, reduced insulin resistance and reduced hepatic steatosis [42]. Authors explained that the beneficial effects of weight reduction are mediated mainly via actions on PKA activity in hepatocytes, increased cAMP response element‐binding protein phosphorylation and increased peroxisome proliferator‐activated receptor‐gamma coactivator‐1α expression [42]. In accordance with the important role of PDE4 in the regulation of cAMP‐dependent pathways, PDE4 inhibition is reported to be a promising new target for weight loss strategies given the anti‐inflammatory, antidiabetic and fat mass reducing effects of roflumilast in mice and humans [43, 44, 45, 46, 47].

These past studies showing favorable effects of PDE4 inhibition may be relevant to both white and brown adipocytes. Indeed, in our work, we have identified significant β₂‐AR induced lipolytic effects upon PDE4 inhibition in white adipocytes derived from human mesenchymal stem cells, providing key insights into the potential mechanisms of roflumilast induced weight loss in humans [27]. In mouse brown adipocytes, we have identified for PDE4 to be among the most highly expressed PDEs in brown preadipocytes [48]. In future work, it would be interesting to perform experiments to quantify the mass of BAT after cold exposure to activate BAT and to test the hypothesis that PDE4 inhibition has a positive consequence on brown preadipocyte proliferation to improve the thermogenic capacity. Indeed, more studies are needed in vivo and in human patient cohorts to clarify whether energy expenditure, BAT mass and activity are regulated by PDE4.

PDE3A is highly expressed in the platelets and in the cardiovascular tissues, while PDE3B is predominantly expressed in organs that regulate energy homeostasis including adipose tissues, liver, and pancreas. PDE3B is highly expressed in BAT in addition to WAT depots [48]. PDE3 accounts for around a half of total PDE activity in BAT homogenates obtained from young rats and it is mainly located at the plasma membrane fraction [49]. Cilostazol (PDE3 inhibitor) administration to obese mice has been shown to enhance the expression of thermogenic markers in BAT [50]. Another study reported that male and female SvJ129 PDE3B‐knockout mice, accumulate less weight under obesogenic conditions and display browning of gonadal WAT, higher adiponectin levels, enhanced mitochondrial biogenesis and thermogenesis [51]. Collectively, these data suggest a crucial role played by PDE3B for energy homeostasis. Further work is needed to explain this metabolic phenotype in the PDE3B‐knockout mice.

Other dual specificity‐ and cAMP‐PDE family members, PDE1 and PDE8A, have also been reported to be expressed in mouse BAT and cultured mature brown adipocytes; however, neither UCP1 expression nor the lipolysis rate was induced by inhibition of either of these PDEs [52]. However, glucose uptake in PDE8A knockout mice was enhanced by treatment with PDE3 and PDE4 inhibitors, suggesting that PDE1 and PDE8A do not localize within cAMP micro‐ or nanodomains that regulate lipolytic responses [52].

In summary, several PDEs have been reported to have a role for stimulating thermogenesis in murine and human brown adipocytes. Given the role of PDEs in the regulation of whole‐body energy homeostasis, fat metabolism and modulatory effects of inflammation in WAT and infiltration of WAT by macrophages, increasing our understanding of how PDEs act to regulate cAMP levels and activity may aid in our efforts to identify effective strategies to prevent and treat obesity and associated metabolic disorders.

Evidence for cAMP micro‐ or nanodomains/signalosomes in brown adipocytes

As mentioned above in the sections discussing evidence for β‐AR/cAMP micro‐nanodomains/signalosomes in white adipocytes, there is a significant body of evidence supporting the hypothesis of spatial compartmentation of GPCR/cAMP signaling in distinct cellular compartments, such as the plasma membrane lipid rafts, caveolins and clathrin‐coated pits [53, 54, 55]. With this compartmentation allowing for a high degree of organization to confer GPCR/cAMP signaling specificity, accuracy and speed.

Previous work in white adipocyte models may also provide some insights into the evidence for the presence of cAMP micro‐ or nanodomains/signalosomes in brown adipocytes. For example, the Taskens' group has reported that Optic atrophy 1 (OPA1), a protein located at the mitochondria, to exhibit an additional function as a dual‐specificity AKAP protein associated with lipid droplets in 3T3‐L1 cells [56]. They revealed that OPA1 is critical for orchestrating a supramolecular complex containing PKA type 1 and perilipin. Therefore, OPA1 is a key regulator of hormonal phosphorylation status of perilipin and lipolysis triggered by catecholamines. Since OPA1 is expressed in BAT in higher levels in comparison to WAT, it remains to be determined as to whether a similar organization exists in brown adipocytes.

In addition, Summers' group has reported that in mice, the β₃‐AR associates with a scaffolding protein caveolin‐1 in brown adipocytes to facilitate stimulatory Gs signaling rather than inhibitory Gi signaling, within lipid raft membrane compartments [57]. These conclusions were made based on disruption of membrane rafts with filipin III or knocking down of caveolin‐1 combined with mutagenesis studies on β₃‐AR and in situ proximity assays. On the other hand, with Caveolin‐1 reported to be critical for proper β₃‐AR signaling and stimulation of intracellular cAMP in brown adipocytes, ablation of caveolin‐1 does not influence endothelial growth factor, platelet‐derived growth factor and lysophosphatidic acid signaling transduction, further supporting the presence of distinct micro‐ or nanodomains in the brown adipocyte allowing for targeted functional effects of cAMP signaling [58].

Changes of cAMP microdomains/signalosomes during brown adipocyte maturation

The dynamics of β‐AR and PDE gene expression and activity during brown preadipocyte maturation into mature cells are likely to be a crucial factor for fine‐tuning cAMP microdomains. As in white adipocytes, species‐specific differences in the expression of β‐AR subtypes and their relative abundance and importance are reported [20]. In murine models, the β₁‐AR is reported to be the most abundant β‐AR subtype expressed in premature brown cells. After maturation, β₃‐AR is upregulated and becomes the most abundant subtype expressed in mature murine brown adipocytes [7, 48]. In human brown adipocytes, the main β‐AR subtype mediating catecholamine effects on BAT activation is still debated and an area of active research. This might be attributed to variable expression patterns of β‐ARs depending on the age, sex and the body mass index of the donor, as well as, the degree of exposure to environmental and physiological stimuli (exposure to cold stress or during activation of sympathetic nervous system). Additionally, human BAT is a heterogeneous organ consisting of different types/colors of adipocytes and samples are obtained from multiple anatomical locations [59, 60]. Although one study reported that β₁‐AR is the predominant subtype expressed in BAT from adult humans and it is the primary receptor subtype mediating the stimulation of UCP1 expression and lipolysis in human brown adipocytes [61], Cypess et al., suggested that β₃‐ARs are the main regulators of lipolysis and thermogenesis in human brown cells [62, 63]. In contrast, Blondin et al. [64] reported that β₂‐AR is the major regulator of thermogenesis in human BAT.

In addition to considering β‐AR subtype expression, PDEs are also differentially regulated during brown adipocyte maturation in murine cells. Based on real‐time FRET cAMP imaging [48], we have identified that PDE4 is the main PDE tightly shaping cAMP micro nano/domains produced by β₁‐, β₂‐ and β₃‐AR stimulation in murine brown preadipocytes. However, upon maturation, PDE3 (mostly the PDE3B but, interestingly, to some extent also PDE3A subfamily) is upregulated and becomes the most abundant PDE expressed in mature murine cells. FRET cAMP imaging experiments provided direct evidence that PDE3 and PDE4 contribute at comparable levels to regulate local cAMP pools generated by β₁ and β₂‐AR stimulation in murine brown cells, whereas PDE2 exerted only minor effects (Fig. 2). In contrast, cAMP pools initiated by stimulation of β₃‐AR in murine mature cells exhibited little control by PDE3 and PDE4. Consistent with these observations, the lipolysis rate of cells stimulated with β₃‐AR was not further enhanced after inhibition of PDE3 and PDE4 [48]. Furthermore, the catecholamine‐stimulated lipolysis is mediated primarily by PKA rather than EPAC1 [65]. These findings indicate that cAMP pools generated by selective stimulation of β₁ and β₂‐AR subtypes generated local cAMP pools that were functionally controlled by PDE3 and PDE4 and were segregated from cAMP pools generated by β₃‐ARs signaling. These results suggest that selective PDE coupling with specific β‐AR subtypes, rather than the expression level of PDEs, is responsible for driving functional outcomes [48]. Different PDE isoforms might be either targeted to specific locations via variable mechanisms, including association with AKAP scaffolding proteins or dynamically recruited subsequently to receptor stimulation by an agonist. In contrast to β₁‐ and β₂‐ARs, the C‐terminal tail of β₃‐ARs lacks the β‐arrestin binding motif and key phosphorylation sites by PKA and GPCR kinases making the receptor relatively resistant to classical short‐term desensitization [66, 67, 68]. This structural difference together with the observation that cAMP pools produced by β₃‐AR exhibit little regulation by the studied PDEs, may contribute to the robust lipolysis, thermogenesis, and fat reduction properties derived by β₃‐AR stimulation in rodents.

Fig. 2.

Model of regulation of β‐AR/cAMP compartmentation in murine brown adipocytes (BA) by PDEs. (A) In murine brown preadipocytes, stimulation of specific β‐ARs leads to local cAMP pools/compartmentalized cAMP signals rather than homogenous distribution throughout the cell. Cytosolic cAMP pools (violet) produced by stimulation of β‐ARs are predominantly controlled by PDE4. (B) In mature adipocytes, PDE3 is upregulated and acts together with PDE4 to regulate cAMP pools produced by β_1/2‐AR stimulation (violet). β₃‐AR is the predominant receptor subtype expressed in mature cells. cAMP signals produced by β₃‐AR (green) are poorly controlled by both PDE3 and PDE4 despite their high expression. As a consequence, spatially segregated cAMP signals activate PKA to facilitate lipolysis leading to the breakdown of triglycerides stored in the lipid droplets and the release of glycerol and free fatty acids.

Simultaneous inhibition of PDE3 and PDE4 has been shown to promote UCP1 expression and lipolysis in differentiated brown adipocytes, as well as glucose uptake in BAT at basal conditions in the absence of adrenergic stimulation [52]. This observation might explain the stimulatory effects of the non‐selective PDE inhibitor IBMX on induction of differentiation of brown adipocyte precursors.

Altogether, lipolytic events operate in a coordinated and specific fashion in brown adipocytes, which may be due β‐AR subtypes specific stimulation of cAMP, that is spatially confined into distinct micro‐ or nanodomains by facilitation of targeted localized PDE activities depending on the stage of adipocytes differentiation.

Remaining major unknowns and future perspectives

Since classical β‐adrenergic agonism is associated with cardiovascular side effects, alternative and novel/orphan GPCRs that are involved in the regulation of adipocytes and energy homeostasis might be promising targets for therapies directed at rebalancing energy expenditure. Examples for such alternatives are adenosine receptors and GPR3 which have recently been shown to facilitate lipolysis, activate browning of white adipocytes and induce energy expenditure in thermogenic fat cells in mice and humans [8, 20, 69, 70, 71, 72, 73]. More work needs to be done to uncover the potential cAMP compartmentation of these GPCRs and of the PDEs that may be functionally coupled to them, and to identify additional novel players for cAMP compartmentation in fat cells. cGMP is another second messenger that promotes the activation and recruitment of the brown adipocytes [74, 75, 76, 77] and it also regulates several PDEs. Therefore, it is appealing to study the cross‐talk and interplay between cAMP and cGMP signaling pathways [78, 79] and to reveal the major regulators of cyclic nucleotide compartmentation in adipocytes.

Almost completely unexplored remains the molecular basis of compartmentalized cAMP signaling in adipocytes, including the exact nature and components of macromolecular complexes or signalosomes regulating lipolysis by GPCRs. The identity of AKAPs, PDEs, kinases, phosphatases and other molecules associated with each β‐AR subtype will help to understand the reasons for various functional responses from different receptors and identify new potential therapeutic targets for metabolic diseases. Better understanding of these molecular components and of local, microdomain‐specific cAMP signaling and its alterations in metabolic disorders should allow tailoring more specific therapeutics which can improve cellular metabolism and ameliorate major pathologies such as lipotoxicity, hyperlipidemia diabetes, and cardiovascular disorders.

Conflict of interest

The authors declare no conflict of interest.

Author contributions

All authors developed the concept, wrote, and edited the manuscript.