CB 2 Cannabinoid Receptors as a Therapeutic Target – What does the Future Hold ?
Amey Dhopeshwarkar and Ken Mackie
Molecular Pharmacology, 2014, 86, 430–437
The past decades have seen an exponential rise in our understanding of the endocannabinoid system, comprising CB1 and CB2 cannabinoid receptors, endogenous cannabinoids (endocannabinoids), and the enzymes that synthesize and degrade endocannabinoids. The primary focus of this review is the CB2 receptor. CB2 receptors have been the subject of considerable attention, primarily due to their promising therapeutic potential for treating various pathologies while avoiding the adverse psychotropic effects that can accompany CB1 receptor–based therapies. With the appreciation that CB2-selective ligands show-marked functional selectivity, there is a renewed opportunity to explore this promising area of research from both a mechanistic as well as a therapeutic perspective. In this review, we summarize our present knowledge of CB2 receptor signaling, localization, and regulation. We discuss the availability of genetic tools (and their limitations) to study CB2 receptors and also provide an update on preclinical data on CB2 agonists in pain models. Finally, we suggest possible reasons for the failure of CB2 ligands in clinical pain trials and offer possible ways to move the field forward in a way that can help reconcile the inconsistencies between preclinical and clinical data.
The endocannabinoid system consists of endogenous cannabinoids (endocannabinoids), cannabinoid receptors (primarily CB1 and CB2), and the enzymes that synthesize and degrade endocannabinoids. A complete [receptor(s), enzymes, and endocannabinoids] endocannabinoid system appears to be present in all vertebrates (Elphick and Egertová, 2005). D9-Tetrahydrocannabinol (D9-THC), the primary psychoactive component of cannabis, produces many of its psychoactive effects by engaging CB1 cannabinoid receptors. In addition to its psychoactivity, cannabis has been shown (or suggested) to be efficacious for multiple therapeutic indications and maladies (Grotenhermen and Müller-Vahl, 2012; Borgelt et al., 2013). Most of these appear to be mediated by D9-THC’s activation of either CB1 or CB2 receptors, though cannabidiol can be an important factor in the therapeutic efficacy of Cannabisbased medicines (Campos et al., 2012). These potential therapeutic effects of D9-THC have motivated a great deal of drug development over the past 40 years. Most of these efforts have taken the form of targeted manipulation of endocannabinoid engagement with cannabinoid receptors or inhibition of the enzymes that degrade endocannabinoids.
A major limitation for the therapeutic development of compounds that directly activate CB1 receptors is unwanted psychotropic effects (Volkow et al., 2014). These CB1-mediated psychotropic actions produce both practical and administrative hurdles that have severely curtailed the development of directacting
CB1 agonists. In contrast, activation of CB2 receptors does not appear to produce these psychotropic effects (Deng et al., 2014). Thus, the observation that CB2 receptor activation produces desirable actions in a range of preclinical models (Leleu-Chavain et al., 2012; Han et al., 2013) attracted considerable interest and generated much activity in both academic and commercial laboratories. For example, agonists targeting CB2 receptors have been proposed as therapies for the treatment or management of a range of painful conditions, including acute pain, chronic inflammatory pain, and neuropathic pain (Ehrhart et al., 2005). They may also be helpful in treating diseases that have a neuro-inflammatory or neurodegenerative component, such as multiple sclerosis (Pertwee, 2007; Dittel, 2008), amyotrophic lateral sclerosis (Kimet al., 2006; Shoemaker et al., 2007), Huntington’s disease (Sagredo et al., 2012), and stroke (Zhang et al., 2007; Pacher and Hasko, 2008). CB2 agonists have also been proposed as therapeutics in peripheral disorders that involve inflammation, including atherosclerosis (Mach et al., 2008), inflammatory bowel diseases (Izzo and Camilleri, 2008; Wright et al., 2008), ischemia / reperfusion injury (Bátkai et al., 2007), renal fibrosis (Barutta et al., 2011), and liver cirrhosis (Mallat et al., 2007; Izzo and Camilleri, 2008; Lotersztajn et al., 2008). Both epidemiologic and preclinical data suggest that activation of CB2 receptors may be protective in osteoporosis (Ofek et al., 2006). Finally, CB2 agonists have shown efficacy in preclinical cancer models (Guzman, 2003; Izzo and Camilleri, 2008;Wright et al., 2008). However, despite very favorable efficacy in a range of preclinical models, CB2 agonists have fared poorly in the clinic. In this review, we summarize our current state of knowledge of CB2 receptor signaling, review preclinical and clinical studies using CB2 agonists, discuss the mismatch between preclinical and clinical results, and suggest possible ways forward. As mentioned above, CB2 agonists may be beneficial for a variety of ailments. However, thisminireview focuses primarily on CB2 agonists for treating chronic pain. Nonetheless, many of the concepts discussed apply to the use of CB2 agonists for other therapeutic indications.
Like CB1 receptors, CB2 receptors are class A serpentine receptors that couple primarily to Gi/o proteins to modulate an array of signaling pathways: adenylyl cyclase,mitogen-activated protein kinase [MAPK (p44/42 and p38)], c-Jun N-terminal kinase,Akt kinase/protein kinase B, phosphoinositide 3-kinase / Akt nuclear factor k-light-chain-enhancer of activated B cells, nuclear factor of activated T cells, cAMP response element– binding protein/activating transcription factor, Janus kinase/ signal transducer and activator of transcription, sphingomyelinase, and caspase, as well as some potassium and calcium ion channels (Bouaboula et al., 1996; Pertwee, 1997; McAllister et al., 1999; Sugiura et al., 2000; Howlett et al., 2002; Molina- Holgado et al., 2002; Ehrhart et al., 2005; Herrera et al., 2005, 2006; Pertwee et al., 2010; Atwood et al., 2012a) (Fig. 1). Despite activation of a wide range of signaling pathways by CB2 receptors, characterization of CB2 receptor ligands has primarily focused on modulation of adenylyl cyclase and extracellular signal–regulated kinases 1/2 (ERK1/2), while other pathways, such as those involving arrestin, Akt, ceramide, and ion channel modulation, and the physiologic processes they mediate, are much less well studied.
CB2 receptor–mediated pertussis toxin–sensitive Gi/o protein stimulation leads to inhibition of adenylyl cyclase and decreased cAMP levels (Felder et al., 1995; Slipetz et al., 1995; Mukherjee et al., 2004). However, expression levels and the environment of expression influences strongly influence the coupling of CB2 to adenylyl cyclase inhibition. For example, stimulation of CB2 receptors on human lymphocytes that endogenously express CB2 receptors poorly inhibits forskolin-stimulated adenylyl cyclase compared with CB2-transfected human embryonic kidney or Chinese hamster ovary (CHO) cells (Bouaboula et al., 1996; Pertwee, 1997, 1999; Schatz et al., 1997; Herring et al., 1998; Gardner et al., 2002; Massi et al., 2003). Similarly, activation of CB2 receptors in mouse spleen cells that endogenously express CB2 receptors does not inhibit forskolin-stimulated adenylyl cyclase at physiologically relevant agonist concentrations (Kaminski, 1993; Kaminski et al., 1994). However, activation of CB2 receptors in the BV2microglial cell line inhibits adenylyl cyclase (Franklin et al., 2003). A particularly interesting example is the natural product 49-O-methylhonokiol, which shows inverse agonism for cAMP production and agonism for release of intracellular calcium (Schuehly et al., 2011).
MAPKs are enzymes involved in a wide variety of vital signaling cascades in many cellular responses, including cell proliferation, migration, transformation, and cell death. Bouaboula et al. (1996) were the first to report the time- and dosedependent activation of ERK1/2 by CB2 agonists in CHO cells transfected with CB2 receptors. They found this activation to be pertussis toxin–sensitive, indicating involvement of Gi/o protein, but adenylyl cyclase–independent. They further showed that activation of this signaling cascade results in phosphorylation of transcription factor Krox-24, thus indicating potential control of gene transcription by CB2 receptors. Unlike inhibition of adenylyl cyclase, ERK1/2 activation is routinely observed in both recombinant as well as nonrecombinant cells/systems (Beltramo, 2009). For example, robust ERK1/2 activation was also reported in immune cells as well as microglia and macrophages, thus confirming the likely physiologic relevance of this pathway (Beltramo, 2009; Merighi et al., 2012). p38 MAPK activation by a nonselective CB2 receptor agonist (D9-THC) was found to have a proapoptotic effect in the Jurkat human leukemia cell line (Herrera et al., 2005) and cytotoxicity in J774-1 macrophages (Yamaori et al., 2013). This effect was exclusively mediated by CB2 receptors (Herrera et al., 2005; Yamaori et al., 2013; Kauppinen et al., 2014). Interestingly, Yamaori et al. (2013), in the same cells, also found a c-Jun N-terminal kinase–mediated cytoprotective effect mediated by D9-THC activation of CB2 receptors. Thus, the same CB2 receptor ligand can activate different MAPKs with varied responses and outcomes (Lopez-Ilasaca, 1998).
Although initial experiments failed to detect functional coupling of CB2 receptors to G protein–gated inwardly rectifying potassium channels and calcium channels (Felder et al., 1995; Pertwee, 1997), other reports suggest that CB2 receptors can modulate the activity of these channels (Ho et al., 1999; Mc Allister et al., 1999; Atwood et al., 2012b). Atwood et al. (2012b) showed CB2 receptor-mediated inhibition of voltagegated calcium channels in AtT20 cells. CP55940 [(2)-cis-3-[2- hydroxy-4-(1,1-dimethylheptyl)phenyl]-trans-4-(3-hydroxypropyl) cyclohexanol effectively inhibited voltage-gated calcium channels, while WIN55212-2 [(R)-(1)-[2,3-dihydro-5-methyl-3-(4- morpholinylmethyl) pyrrolo [1,2,3-de]-1,4-benzoxin-6-yl]-1- naphthalenylmethanone mesylate] was inactive on its own and
antagonized CP55940 inhibition. Thus, the reason some of the earlier studies failed to find ion channel modulation can likely be attributed to the functional selectivity of the ligands used in the earliest studies (see below) (Atwood et al., 2012b).
Most G protein–coupled receptors (GPCRs) undergo some degree of internalization following agonist binding. Internalization CB2 as a Therapeutic Target 431 Downloaded from molpharm.aspetjournals.org at ASPET Journals on August 17, 2019 can play a role in downregulation of the GPCR’s ability to signal at the membrane (Ferguson, 2001). Additionally, internalized GPCRs can engage novel signaling pathways inaccessible to GPCRs residing on the surface membrane (Miller and Lefkowitz, 2001). Thus, internalization of a GPCR in response to a ligand can be considered a form of signaling. CB2 receptors exhibit variable internalization in response to an agonist, with some agonists promoting marked internalization and others being inactive (Grimsey et al., 2011; Atwood et al., 2012b; Petrov et al., 2013).