► GPR55 is a putative cannabinoid receptor. ► GPR55 shares little homology (10–15%) with cannabinoid-1 (CB₁) and cannabinoid-2 (CB₂) receptors. ► GPR55 expression is found in the gastrointestinal (GI) tract. ► GPR55 mRNA expression is increased in intestinal inflammation. ► GPR55 agonists reduce intestinal inflammation.

Despite sharing little homology (10–15%) with cannabinoid-1 (CB₁) and cannabinoid-2 (CB₂) receptors, the G protein-coupled receptor 55 (GPR55) was initially thought to be a new member of the cannabinoid receptor family. Apart from being activated by various exogenous cannabinoids, GPR55 is also activated by endocannabinoids like anandamide, which is found in organs with high GPR55 expression such as the brain and the gastrointestinal (GI) tract. The phylogenetic distance to the classical CB receptors and its pharmacological responsiveness to certain cannabinoids suggests that GPR55 may constitute a novel class of cannabinoid receptors. GPR55 influences mechanisms in the nervous system, vasculature, kidney and bone. Recent research revealed that GPR55 is also involved in cancer development and inflammatory pain. Because of its presence in the GI tract, several studies have started to focus on the involvement of GPR55 in the physiology and pathophysiology of the gut. The following article intends to discuss the potential role of GPR55 in GI functions.

Structure of GPR55

Several years ago, a GeneBank search for transmembrane regions characteristic of the GPCR family (High Throughput Genome; HTG) revealed several new 7TM/GPCRs, among them the GPR55 gene [2]. GPR55 was subsequently cloned and soon localized by Northern blot analysis in human brain and liver as well as in rat intestine and spleen [2]. The ability of GPR55 to respond to cannabinoids was later documented by two independent patents [3] indicating that GPR55 may represent a new cannabinoid receptor. Although GPR55 is activated by certain cannabinoids, the low sequence homology with CB receptors (10–15%) argues against a clear affiliation to the historically established cannabinoid receptor family, despite higher sequence homologies within the conserved regions [3]. Amino acid residues at the binding and activity sites of human CB₁ and CB₂ align poorly with human GPR55 implicating that GPR55 may not share the same binding pockets with the classical CB receptors [4]. However, the fact that many cannabinoid compounds are able to activate GPR55 may suggest that it belongs to a novel group of cannabinoid receptors that, like CB₁ and CB₂, could play a potential role in the physiology and pathophysiology of the GI tract [5]. Phylogenetically, GPR55 is part of the purine receptor cluster of the δ group of rhodopsin receptors [6] together with GPR18, another cannabinoid-responsive GPCR [7•,8]. From other GPCRs closely related to GPR55 (GPR35, GPR92 and GPR23), no reports have been published yet as to whether they are activated or inhibited by cannabinoids [1••]. Recent data indicate that lysophosphatidic acid (LPA) may act as their natural ligand [9–11]. Rodent GPR55 shows up to 78% homology with the human GPR55 [12•].

Pharmacology and physiology of GPR55

The pharmacology of GPR55 has not yet been resolved and data are too contradictory as to whether GPR55 can be unambiguously coined as the ‘third cannabinoid receptor’ [1••]. In GTPγS binding assays of overexpressed cell systems (mainly HEK293 cells), GPR55 activation was observed following application of many natural and synthetic cannabinoids as well as following application of endocannabinoids like anandamide but these results were not confirmed in other assays of receptor activation (e.g. β-arrestin activation [13]). Concerning the endogenous ligand(s) for GPR55, the only consistent data converge on a lysophospholipid, namely lysophosphatidylinositol (LPI). Especially, LPI carrying an arachidonic acid moiety is supposed to have the greatest impact on GPR55 activation [14]. Our knowledge on the physiological role of GPR55 and whether it could be part of the endocannabinoid system is in its infancy. Because of the lack of studies with specific antagonists, not much is known yet about the role of endogenous GPR55 to get a clear picture of the receptor’s functions. Some data have been reported from cell types endogenously expressing GPR55 such as from human neutrophils [15••], DRG neurons [16••], PC12 cells [17•], endothelial cells [18••] and also from cells genetically lacking GPR55 expression [19,20]. In addition, GPR55 expression has been found in certain cancer cells [21••,22] and in the endocrine pancreas, where it has been suggested to play a contributory role in insulin secretion [19]. In human neutrophils, GPR55 augments migratory responses of CB₂ to 2-arachidonoyl glycerol (2-AG) at the level of small GTPases, such as Rac2 and Cdc42 [15••] whereas in endothelial cells, activation of GPR55 causes Ca⁺⁺ mobilization, which is, depending on the status of integrin clustering, inhibited by CB₁activity [18••]. Both findings direct to the intriguing possibility that an interaction may exist between GPR55 and CB receptors in physiological and pathophysiological mechanisms triggered by endocannabinoids. In this context, it is noteworthy that the CB₁ antagonists SR141716A and AM251 also act as GPR55 agonists [23•]. The main differences between GPR55 and CB receptors not only lie in their low homology but also in the activation of different downstream G proteins and downstream signals. In neurons, for instance, CB₁ signals through G_i/o proteins, inhibits voltage gated Ca⁺⁺channels and activates A-type and inwardly rectifying K⁺ currents [24]. In contrast to CB receptors, GPR55 signals through G_α12 and G_q proteins, activates downstream small G proteins like RhoA and increases Ca⁺⁺ release and K⁺ type M-currents [16••]. In osteoclasts, LPI was also shown to stimulate Rho, an effect that was absent in GPR55^−/− mice [20]. It seems that GPR55 initiates mostly excitatory and not inhibitory effects, which is in contrast to the role of the classical CB receptors, which mostly initiate inhibitory effects. In line with this concept, GPR55 has been implicated in the development of neuropathic and inflammatory pain [25•]. The reader is referred to several recently published excellent reviews on GPR55 pharmacology and physiology [26••,27••,28••,29•,30].

Motility and inflammation

Similar to CB₁ and CB₂ (reviewed in [36]), expression of GPR55 can be found throughout the GI tract, although a detailed description of tissues and cells expressing GPR55 is still lacking. Real time and semiquantitative PCR show that GPR55 is present in duodenum, jejunum, ileum and the colon of rodents [12•,37,38•]. In particular, GPR55 expression is found in mucosal scrapings and in longitudinal-myenteric plexus preparations of the colon indicating that the receptor is most likely present in both, gastrointestinal epithelial cells and in gastrointestinal enteric neurons [38•]. Additionally, Lin et al. were able to detect GPR55 in enteric neurons of the rat ileum by immunohistochemistry [37]. The study showed that the expression of GPR55 was higher in rats treated with lipopolysaccharide (LPS) than in control animals, indicating that GPR55 may be involved in the response of the gut to intestinal inflammation [37]. The authors also suggested that immunohistochemical expression of GPR55 was increased in enteric neurons following LPS treatment, though they did not provide any quantitative data to corroborate this finding. Nevertheless, the detection of GPR55 in enteric neurons raises the possibility that GPR55 has a role in gastrointestinal functions, such as motility and secretion. In support of this hypothesis, the atypical cannabinoid O-1602 (a GPR55 agonist that has shown GTPγS activation in membranes of human recombinant GPR55-expressing cells with an EC₅₀ of 1.4 [39•] and 13 nm [12•]), reduced spontaneous contraction in the rat ileum at 0.1 μm while in the colon, only contractions that were induced by LPS were decreased in a dose dependent manner [37]. In contrast, O-1602 had no effect on the membrane potential in the jejunum of untreated mice [37]. This indicates that, depending on the part of the GI tract, GPR55 is probably involved in pathophysiological as well as physiological motor functions. Whether relaxation of gut segments by O-1602 is an entirely GPR55 dependent process is uncertain because of its off-target effects. Thus, O-1602 has been shown to retain vasodilatory [39•], orexigenic [40] and anti-inflammatory activities [38•] in GPR55^−/−mice. However, a very recent study demonstrated that O-1602 concentration-dependently reduced electrical field-induced contractions in the colon strips from wild-type and CB₁^−/−/CB₂^−/− knockout mice, an effect that was significantly inhibited in GPR55^−/− knockout mice [41]. Interestingly, under physiological conditions, the inverse CB₁ agonist AM251, identified as a GPR55 agonist [23•], increased upper GI transit and whole gut transit in mice, but had no effect on colonic expulsion [42].

On the basis of the assumption that cannabidiol (CBD) acts as a GPR55 antagonist (CBD was shown to antagonize GPR55 activity to the CB receptor agonist CP55940 with an IC₅₀ of 445 nm [12•]) the effect of CBD on gut motility was investigated after LPS treatment [37]. LPS increased GPR55 mRNA expression in rat duodenum and ileum and inhibited intestinal motility in mice while CBD counteracted the slowed GI motility, as measured by the upper GI transit time. CBD also normalized the LPS-induced inhibition of isolated rat smooth muscle strips. However, CBD had no effect on GI transit in normal mice or on the resting membrane potential of jejunum smooth muscle suggesting that an involvement of GPR55 in GI motility is more prevalent during inflammatory than healthy conditions. Upregulation of GPR55 during inflammation may demask the antagonistic effect of CBD on GPR55, though this remains speculative. In this respect, the behavior of GPR55 resembles that of the CB₂ receptor which is also increased in inflammatory conditions, such as IBD [32] and which modulates motor functions only in an inflammatory state [33]. The exact mechanism of GPR55 in regulating GI motility remains elusive until a specific GPR55 antagonist is available. In the meantime, experiments in GPR55^−/− mice may be an option to address these questions.

The potential role of GPR55 in intestinal inflammation has been also addressed in experimental colitis [38•]. There, the effects of O-1602 were investigated in CB₁^−/−/CB₂^−/− and GPR55^−/− mice that had been submitted to dextran sulfate sodium (DSS) which induces inflammation of the colon. O-1602 was still effective in preventing colitis in these knockout mice suggesting that the compound had caused an improvement via unidentified targets (Figure 1). The authors also noted that the severity of the colitis was slightly but significantly lower in the GPR55^−/− mice than in the respective wild types. The finding favors the hypothesis that GPR55 may play a pro-inflammatory role in GI inflammation.

Figure 1

Hematoxylin/eosin staining of colon wall sections from GPR55 knockout mice exposed to dextran sulfate sodium (DSS) and treated with 5 mg/kg O-1602 (or vehicle) twice daily. The putative GPR55 agonist O-1602 is still effective in decreasing the …

Secretion and mucosal homeostasis

A possible involvement of GPR55 in intestinal secretion remains speculative as this has not been addressed yet. GPR55 is found at sites of the GI tract known to be involved in secretory processes, that is, in the mucosa (mucosal scrapings) of the colon [38•] and in the enteric nervous system [37] which is known to control secretion via intrinsic neurons. A role for GPR55 in secretory processes and mucosal homeostasis is therefore conceivable.

Central regulation of gut motility, food intake and emesis

Since GPR55 is expressed in the brain [2] it could participate in central mechanisms that influence certain functions of the gut. Thus, GPR55 transcripts were demonstrated in the brainstem, which houses the dorsal motor nucleus of the vagal nerve, as well as in the hypothalamus where centers related to energy household are located [2,12•]. This suggests that GPR55 may play a role in food intake. Anandamide and other acylethanolamides such as palmitoylethanolamide (PEA) and oleoylethanolamide (OEA) activate CBs and GPR55 and induce or inhibit food intake [43,44]. OEA is widely produced in the intestine and functions as a potent inhibitor of food intake [45], however, unlike PEA, it is a weak GPR55 agonist [12•]. As with CB₁, GPR55 expression could be effected by different levels of anandamide. In the brain, for instance, levels of anandamide were shown to be regulated by fasting [46] and diet composition [47]. The same applies for the gut where anandamide was found to be increased in the small intestine, but not in the stomach, 24 hours after fasting [48]. A recent article describes that a missense polymorphism in the GPR55 gene was associated with Anorexia nervosa[49] supporting the idea that GPR55 may play a role in the regulation of food intake.

Emetic reflexes are controlled by CB receptors in the brainstem [50,51]. Vagal afferents express CB₁ receptors which are subject to plastic changes depending on the state of fasting [52]. It would be interesting to determine whether GPR55 is expressed in the nodose ganglion and whether it colocalizes or interacts with CB₁ receptors thereby influencing emesis. Against the idea that GPR55 plays a role in eating disorders it might be argued that GPR55^−/− mice do not differ from wild types in their body weights and the amount of their food intake [25•,40]. Likewise, the GPR55 agonist O-1602 induced food intake in mice when given icv [40] and this effect persisted in GPR55^−/− mice questioning a major role of GPR55 in food intake.

Is GPR55 involved in GI mechanoreception?

GPR55 was detected in large diameter dorsal root ganglia cells known to represent mechanoreceptors [16••]. This implicates that extrinsic sensory nerve fibers expressing GPR55 very likely innervate the GI tract. GPR55 may be therefore involved in mechanosensation of the GI tract and eventually in the control of GI pain. Splanchnic mechanoreceptors arising from dorsal root ganglia innervate the GI tract at a high percentage [53,54] and have polymodal functions, that is, they can also act as nociceptors [54] or obtain this ability under brief inflammation [55]. A role for GPR55 in mechanoreception and pain is supported by a study of Staton et al. who revealed a pro-inflammatory role of GPR55 in mechanical hyperalgesia by demonstrating that GPR55^−/− mice failed to develop mechanical hyperalgesia for 2 weeks after intraplantar administration of Freund’s complete adjuvants or for 28 days following partial ligation of the sciatic nerve [25•]. On the basis of these observations, GPR55 could play a pro-nociceptive role in the GI tract and contribute to hyperalgesic mechanisms in visceral neurons that underlie the development of irritable bowel syndrome [56]. However, O-1602 did not alter responses to noxious colorectal distension in rats [57] which somewhat argues against an involvement of GPR55 in visceral nociception, but as already mentioned above, the in vivo pharmacology of O-1602 may not depend on GPR55. Functional studies addressing the involvement of GPR55 in intestinal mechanoreception have not yet been performed as specific GPR55 antagonists are not yet available. The recent identification of a GPR55 agonist binding site is a promising move to be able to investigate specific GPR55 effects in the near future [58].

• • of special interest
• •• of outstanding interest

RS is supported by grants from the Austrian Science Fund (FWF P 22771), Austrian National Bank (OeNB 14429) and the Franz Lanyar Foundation (351). MS is supported by the Deutsche Forschungsgemeinschaft (DFG).