The most well-characterized of the deorphanized GPCRs, GPR55 (Ryberg et al., 2007), is expressed most highly in brain, adrenal gland, and digestive tract (although at reportedly low levels relative to that of CB1). When expressed in HEK293 cells, specific binding of the synthetic CB1 agonist CP55940 is observed, but not of the CB1/CB2 synthetic agonist WIN55212-2. Interestingly, a very small amount of specific GPR55 binding was observed using 50 nM of the CB1-selective antagonist/inverse agonist SR141716A (SR).

Using the same HEK293 cell line for GTPγS functional assays, the endogenous agonist 2-arachidonylglyerol (2-AG) potently activates GPR55 (EC50 = 3 nM), although virodhamine (EC50 = 12 nM) is a more complete agonist with about 50% greater efficacy. Notably, in a separate assay of intracellular calcium release, 2-AG and virodhamine showed no GPR55 agonism, suggesting this receptor is subject to agonist-dependent functional selectivity (Lauckner et al., 2008). Other potent agonists discovered in this initial screen of GPR55 efficacy include; palmitoylethanolamide (PEA), CP55940, Δ⁹-tetrahydrocannabinol (THC), noladin ether, and anandamide with EC50s of 4, 5, 8, 10, and 18 nM respectively, and efficacies similar to that of 2-AG. Interestingly, the non-psychoactive phytocannabinoid cannabidiol (CBD) was found to antagonize GPR55 activity at physiologically relevant concentrations.

GPR55 primarily couples through the relatively obscure Gα13. This interesting G-protein activates a cascade involving RhoA that, among other effects, ultimately alters actin polymerization and stability, implicating GPR55 signaling in processes related to neuronal morphology (Worzfeld et al., 2008). This potential role is further supported by effects to promote neurite retraction (Obara et al., 2011). More recently, evidence in transfected HEK293 cells indicates that signaling through G_q also occurs (Lauckner et al., 2008).

In ERK activation assays with GPR55-expressing HEK293 cells, lysophosphatidylinositol (LPI) is implicated as an important endogenous agonist (Oka et al., 2007). Notably, 2-AG and virodhamine were not effective in stimulating ERK in transfected HEK293 cells, possibly indicating ligand-dependent functional selectivity for different signal transduction pathways. LPI was also found to dose-dependently increase intracellular calcium in these cells with an EC50 ~ 1 μM, an effect possibly related to GPR55 activation of G_q. System-dependent efficacy of GPR55 signaling may also involve interaction with other receptors including CB1. GPR55 and CB1 have been shown to heterodimerize within rat and macaque striatum, possibly related to a role in motor behaviors. When co-expressed in HEK293 cells, CB1 and GPR55 antagonize respective agonist effects to promote ERK1/2 activation (Martínez-Pinilla et al., 2014).

The physiological significance of GPR55 expression in brain remains an open question; although, some evidence indicates a role in controlling ingestive behaviors (reviewed by Liu et al., 2015). A role in promoting appetite makes sense as it has been reported that GPR55 regulates peripheral metabolism and energy homoeostasis (reviewed by Simcocks et al., 2014). Deletion of GPR55 in mice has subtle effects on motor coordination, but it has shown no significant effect on several learning and memory tests, in contrast to the role of CB1 signaling, and despite dense GPR55 expression in hippocampus, striatum, and cortex (Wu et al., 2013). Spinal cord expression of GPR55 is upregulated in a chronic constriction injury model of neuropathic pain; this suggests that it may mediate some of the analgesic effects of N-arachidonoyl-serotonin (AA-5-HT, Malek et al., 2016). GPR55 activation by some agonists increases calcium release from intraneuronal stores and inhibits M-type potassium current, both tending to promote neuronal activity (Lauckner et al., 2008). These effects may be important in the context of CBD's ability to antagonize the receptor, as this non-psychoactive phytocannabinoid has recently been shown effective for the treatment of Dravet syndrome, a formerly intractable form of childhood epilepsy (Devinsky et al., 2017). Whether the efficacy of CBD in this syndrome involves GPR55 antagonism, or perhaps, involves some combination of the many other established targets of this drug (see Table 2) remains to be resolved. Other evidence for potential roles for GPR55 signaling includes: distinct expression in rod cells of primate retina suggesting a role in low-light vision (Bouskila et al., 2013) and; expression in microglia that, following LPI activation, acts to protect hippocampal neurons from excitotoxicity (Kallendrusch et al., 2013).

Interestingly, microglia migration is stimulated by activation of a second formerly orphan cannabinoid-related receptor, GPR18 (McHugh et al., 2010). This G_i/o-coupled receptor was originally reported to be highly-expressed in testes and spleen (Gantz et al., 1997); yet, it clearly is also functional within brainstem (Penumarti and Abdel-Rahman, 2014b). GPR18 is activated by N-arachidonoyl glycine (NAGly, Kohno et al., 2006) that is a product of anandamide metabolism (Aneetha et al., 2009; Bradshaw et al., 2009). It is also activated by abnormal cannabidiol (abn-CBD) that, when infused into regions of the brainstem, causes reduction in blood pressure in a manner that involves nitric oxide synthase and adiponectin signaling (Penumarti and Abdel-Rahman, 2014a). As with GPR55 discussed above (section GPR55 in CNS), GPR18 receptor expression is upregulated following spinal cord injury; this suggests that it may mediate some of the analgesic effects of N-arachidonoyl-serotonin (AA-5-HT, Malek et al., 2016). Notably, the principal active constituent of Cannabis, THC, is a potent agonist of GPR18 when expressed in HEK293 cells (McHugh et al., 2012). The lesser Cannabis constituent, CBD, along with synthetic CB agonists CP55940 and WIN55212-2, have no activity at physiologically-relevant concentrations. Potent GPR18 activation by THC and NAGly has been suggested to underlie efficacy to resolve inflammatory pain (Burstein et al., 2011; McHugh et al., 2014; Crowe et al., 2015). Some controversy about GPR18 signal transduction was raised when NAGly failed to inhibit Ca⁺⁺ currents following heterologous expression and agonist stimulation in rat superior cervical ganglion neurons (Lu et al., 2013). Whether this is a special case of system-dependent functional selectivity, or an indication that NAGly modulates neuronal activity via other receptor systems, remains an open question.

A third relevant G-protein-coupled receptor, GPR119, has been deorphanized. In humans, GPR119 was initially reported to be expressed only outside of CNS, most notably in pancreas, suggesting a peripheral metabolic function (Fredriksson et al., 2003). A possible CNS expression difference between human and rodent species is suggested in a patent application that indicates GPR119 is also expressed at high density in rat and mouse brain (Bonini et al., 2002). Mouse CNS expression is confirmed by high-affinity GPR119 binding and potent efficacy to reduce seizure-related hippocampal activity (Scott et al., 2014). At micromolar concentrations, the endogenous peroxisome proliferator-activated receptor alpha (PPARα) agonist oleoylethanolamide (OEA) activates GPR119, while anandamide appears to have partial efficacy in yeast cells expressing the receptor (Overton et al., 2006). Notably, 2-AG, CP55940, WIN55212-2, methanandamide, and JWH-133 did not activate GPR119. Supporting CNS activity, and a role in controlling ingestive behaviors, OEA and a synthetic GPR119 agonist with similar efficacy, but four-fold greater potency, PSN632408, were found to reduce food consumption and body weight in Sprague-Dawley rats (Overton et al., 2006). More recent evidence suggests that PSN632408 may lack selectivity for GPR119 in peripheral tissues (Ning et al., 2008). Despite evidence of GPR119 CNS activity, this receptor is being most closely studied in as a target for treating diabetes and other metabolic disorders (Ansarullah et al., 2013; Cornall et al., 2013, 2015).

A final G_i/o-coupled deorphanized receptor, GPR35, has been reported to transduce effects of 2-arachidonyl lysophosphatidic acid (2A-LPA) and kynurenic acid, both of which are present at relevant concentrations within neuronal tissues. 2A-LPA is a product of metabolism of the principal brain endocannabinoid, 2-arachidonyl glycerol (2-AG, Oka et al., 2010). Kynurenic acid is the product of tryptophan metabolism, and is most well-characterized as an endogenous antagonist of receptors for the principal excitatory neurotransmitter, NMDA (Stone et al., 2013). A very recent structure-activity study has led to identification of several potent GPR35 agonists and three structures critical for efficacy (Abdalhameed et al., 2017). GPR35 is expressed most predominantly in the gastrointestinal tract and in leukocytes (Wang et al., 2006). Low-level expression of GPR35 in brain reduces the likelihood that it significantly modulates neuronal activity within this region of the CNS. Within spinal cord, on the other hand, convincing evidence demonstrates functionally-relevant expression of GPR35, with distinct enrichment within murine dorsal horn ganglia (Cosi et al., 2011). In an acetic acid-induced pain assay, both kynurenic acid, and the phosphodiesterase inhibitor, zaprinast, were found to produce significant analgesia with EC50s of about 100 and 1 mg/kg, respectively. Note zaprinast is an effective agonist of GPR35 with a particularly high potency at rat vs. human forms of the receptor (Taniguchi et al., 2006). Homology between GPR35, and other LPA receptors, may suggest that 2A-LPA is more relevant than kynurenic acid as an endogenous modulator (Zhao and Abood, 2013). Clearly, more needs to be learned and understood about GPR35 signaling.