EDITOR’S NOTE: In this analysis, Vincenzo Di Marzo discusses recent developments in understanding endocannabinoid actions and metabolism, and efforts to target these lipid mediators as a novel approach to analgesia. Read his review, and the related PRF news stories (FLAT-tening Pain, COX-2 Inhibitors Give Endocannabinoids a Lift), and then share your own thoughts by leaving a comment.
That Cannabis preparations such as marijuana can alleviate pain in humans has been known for millennia, whereas the realization that plant cannabinoids, and Δ9-tetraydrocannabinol (THC) in particular, strongly reduce nociception in animal models of acute, visceral, inflammatory, and chronic pain is relatively more recent. The cloning in the 1990s of G protein-coupled receptors for THC, the cannabinoid type-1 and -2 receptors (CB1 and CB2), strongly expressed in neurons and immune cells, respectively, provided a molecular mechanism for these therapeutically promising effects and raised expectations that they could be soon translated into new analgesic and anti-inflammatory medicines. Yet, the possibility arose that direct activation of CB1 and CB2 with either natural or synthetic agonists could also produce unwanted effects in terms of psychotropic and immunosuppressive actions, respectively. Therefore, the discovery of endocannabinoids, i.e., endogenous agonists of cannabinoid receptors, and of the molecular mechanisms and enzymes controlling their biosynthesis and inactivation, appeared to provide new solutions to this dilemma. In fact, it soon became clear that, unlike neurotransmitters, hormones, and other mediators, and similar to other derivatives of arachidonic acid (eicosanoids) often involved in pain and inflammation, the two most studied endocannabinoids, N-arachidonoyl-ethanolamine (anandamide) and 2-arachidonoyl-glycerol (2-AG), are not biosynthesized and stored in secretory vesicles to be released following cell stimulation, but are instead produced “on demand,” i.e., they are biosynthesized and then immediately released to activate CB1 and CB2 receptors only locally (Di Marzo, 2008). Furthermore, it emerged that endocannabinoid levels might become increased or reduced during chronic or inflammatory pain only in those tissues directly affected by these pathological conditions. Thus, devising selective inhibitors of endocannabinoid inactivation would represent a selective and possibly safer way to activate cannabinoid receptors only when and where needed to counteract pain and inflammation. Several efforts have been made by medicinal chemists and pharmacologists in this direction over the last decade, culminating with the first Phase I and II trials on some of such compounds, and, alas, with the realization that even such a strategy may be problematic (Alvarez-Jaimes and Palmer, 2011).
In fact, endocannabinoids are lipids, and as such, they are “redundant” in all aspects of their biology and pharmacology. We now know that redundant are the biochemical routes and enzymes through which endocannabinoids are produced and inactivated, and the same applies to their molecular targets. Anandamide, for example, is produced together with other N-acylethanolamines, which can be hydrolyzed by the same enzyme as the endocannabinoid, i.e., fatty acid amide hydrolase (FAAH), and interact with non-cannabinoid receptors, including the peroxisome proliferator-activated receptor α(PPARα) and the transient receptor potential vanilloid type-1 (TRPV1) channel, both of which are involved in pain and inflammation. Anandamide itself can efficaciously activate and desensitize the TRPV1 channel (Di Marzo, 2008). On the other hand, 2-AG was recently found to activate GABAA receptors via interaction with their δ-subunits (Sigel et al., 2011). This endocannabinoid, by many regarded as the true CB1 agonist in the CNS, is inactivated by monoacylglycerol lipase (MAGL), but also by other hydrolases, such as ABHD6 and ABHD12, as well as FAAH (Di Marzo, 2008). Furthermore, 2-AG has been known for decades to act also as an alternative precursor for arachidonic acid release (and eicosanoid) production in sensory neurons (Gammon et al., 1989), and recent studies have confirmed that, among several 2-AG hydrolytic enzymes, MAGL might be the one responsible for this function (Nomura et al., 2011). Finally, both anandamide and 2-AG are very good substrates for COX-2, through which they serve as precursors for novel lipid mediators known as prostaglandin-ethanolamides (prostamides) and prostaglandin-glycerol esters (PG-GEs). These compounds have been proposed to act via new receptors and to exert a stimulatory function on inflammation (Woodward et al., 2008). In view of these emerging data, one cannot help wondering what really happens in a cell when inhibiting MAGL or FAAH. Does FAAH inhibition result in the “indirect” and site- and time-selective activation not only of CB1 and CB2, but also of PPARαand TRPV1, as well as in the elevation of prostamide tissue levels, and does all this interfere with the beneficial effects of cannabinoid receptor activation? Does MAGL inhibition lead to reduction of eicosanoid production and strengthening of GABAA activity as extra “bonuses” reinforcing the inhibition of pain and inflammation exerted by indirect CB1 and CB2 activation, or does it cause elevation of PG-GE levels as a “collateral event” counteracting the action of cannabinoid receptors? If redundancy in endocannabinoid catabolic pathways and molecular targets hinders the development of new analgesics from FAAH and MAGL inhibitors, is there a way to circumvent this problem?
Some studies that recently appeared in the literature, together with published and unpublished collaborative studies from my lab, seem to provide the first ways out of this conundrum. For example, we showed that a synthetic compound, arachidonoylserotonin, that inhibits at the same time as FAAH and TRPV1, can be more efficacious at counteracting inflammatory and chronic pain than more potent and selective FAAH inhibitors or TRPV1 antagonists (Maione et al., 2007; Costa et al., 2010). Indeed, widely used analgesics and anti-pyretic drugs, such as paracetamol and dipyrone, have recently emerged as “multi-target” drugs, as they can be transformed in vivo into metabolites that can still inhibit COX but also activate and desensitize TRPV1 or the transient receptor potential ankyrin-type-1 (TRPA1) channels, as well as inhibit FAAH and mildly activate CB1 receptors (Hogestatt et al., 2005; Mallet et al., 2008; Mallet et al., 2010; Andersson et al., 2011; Rogosch et al., 2011). The non-psychotropic cannabinoids, cannabichromene and cannabidiol, when injected into the periaqueductal grey, can reinforce the descending pathway of anti-nociception in rats, probably by virtue of their stimulatory activity at TRPA1 and inhibitory activity on endocannabinoid and adenosine metabolism (Maione et al., 2011). Thus, both synthetic and natural products may be able to provide us with new strategies to cope with endocannabinoid redundancy and try to exploit it pharmacologically with new or old “dirty” drugs.
A typical example of this came from a recent study by Duggan and colleagues (Duggan et al., 2011), recently the focus of an interesting article on this website, COX-2 Inhibitors Give Endocannabinoids a Lift. The authors, by exploiting a multidisciplinary approach including data obtained with X-ray crystallography, computational modeling, pharmacology, and biochemistry and analytical chemistry techniques, provided a possible explanation as to why a particular class of NSAIDs, the (R)-profens, are efficacious anti-inflammatory agents, even though they are only weak inhibitors of COX-catalyzed formation of prostanoids. In fact, it turns out, from the very elegant experiments carried out by Duggan and coworkers, that some (R)-profens, i.e., the R enantiomers of ibuprofen, naproxen, and flurbiprofen, are potent inhibitors of the metabolism of endocannabinoids, rather than arachidonic acid, by COX-2, and, therefore, they do not only prolong the lifespan of anandamide and 2-AG, but also prevent the formation of prostamides and PG-GEs in vitro. The selectivity of (R)-profens is based on the cooperativity recently described to occur between the two subunits of COX-2 and potentially allows administration of doses of these compounds that will inhibit the inactivation of endocannabinoids without impairing the production of other COX metabolites, thereby avoiding the side effects typical of COX inhibitors (Duggan et al., 2011). Indeed, the potential benefit of the use of these compounds emerges more convincingly if one considers as-yet unpublished results from my group showing that, in a mouse model of inflammatory pain of the knee, prostamide F2α, a compound that we found to exhibit potent pro-algesic actions, can indeed be produced in the spinal cord (Piscitelli et al., 2011). It is tempting to speculate that the failure of a selective FAAH inhibitor at reducing pain in human osteoarthritis, described at the 2010 IASP meeting in Montreal, might have been due also to the concomitant elevation of the levels of pro-inflammatory prostamides, thus perhaps suggesting that co-administration of FAAH and (R)-profens might have instead done the trick. Also, the emerging concept that MAGL inhibitors, especially if used at submaximal doses, might replace COX-2 inhibitors, since they appear to inhibit 2-AG/MAGL-driven prostaglandin production in the CNS but not in the GI tract (Nomura et al., 2011), might be revisited if other preliminary data, suggesting instead that MAGL inhibition, at least in the brain, may open the way to the formation of pro-inflammatory PG-EGs, are confirmed (Sagredo et al., 2011). In view of these complications, the development and use of drug combinations, and, even better, of molecules with more targets, such as, for example, dual COX-FAAH or COX-MAGL inhibitors, are to be fostered. It’s interesting to note that inhibition of endocannabinoid inactivation produces gastro-protection against NSAID-induced gastric hemorrhages in animal models (Kinsey et al., 2011), thus strengthening the hypothesis that dual inhibitors of COX and endocannabinoid hydrolytic enzymes could be not only more efficacious, but also safe.
Another strategy that has been so far explored only in preclinical studies is the one based on the inhibition of the putative “endocannabinoid membrane transporter” (EMT), the existence of which has been dwelled upon for over 15 years. In fact, hydrolysis or oxidation of endocannabinoids is effected at the intracellular level, and although these mediators are lipophilic and can cross the plasma membrane by simple passive diffusion, there is indirect evidence for the existence of a membrane transporter facilitating this process according to the gradient of concentrations and in a manner subject to regulation by other mediators (Di Marzo, 2008). However, the fact that no specific protein has been identified so far to facilitate endocannabinoid transport across the membrane has cast doubts over the existence of the EMT and hindered the development of selective as well as potent inhibitors of this process. Such compounds would have an advantage over MAGL and FAAH inhibitors, that is, the capability to interfere more selectively with anandamide and 2-AG inactivation (with respect to, e.g., other N-acylethanolamines) and to block, at the same time, the interaction of these compounds with intracellular binding sites (as in the case of TRPV1) or oxidizing enzymes, such as COX-2, whilst leading to a more selective prolongation of their activity at their most important extracellular targets, i.e., CB1 and CB2 receptors.
A recent study by Fu et al. (2011) sheds light on the possibility of developing new drugs in this direction. In fact, the authors identified an intracellular protein that specifically facilitates anandamide cellular reuptake in neurons and astrocytes, and named it “FAAH-like anandamide transporter” (FLAT). FLAT is a splicing variant of FAAH that is catalytically inactive, i.e., unable to catalyze the hydrolysis of anandamide, 2-AG, or other N-acylethanolamines. Nevertheless, this protein binds anandamide, but not 2-AG or other N-acylethanolamines, with high affinity, and acts as an intracellular sink for anandamide transport across the membrane. FLAT is certainly not the long-sought EMT, primarily because of its cytosolic location and of it not being able to recognize 2-AG (Fu et al., 2011). Furthermore, other intracellular proteins had been previously suggested to participate in intracellular anandamide trafficking, although these proteins are clearly not specific for endocannabinoids (Maccarrone et al, 2010).
In fact, the discovery of FLAT is potentially very important from the therapeutic point of view, since selectively inhibiting the intracellular trafficking and, as a consequence, the cellular uptake of anandamide, but not 2-AG, might offer some advantages over the inhibition of the inactivation of both endocannabinoids, as exemplified by data obtained with a dual FAAH/MAGL inhibitor (Long et al., 2009). Whilst blocking the putative EMT might prolong both anandamide and 2-AG actions at cannabinoid receptors, selectively inhibiting FLAT would only retard anandamide reuptake and delivery to FAAH, thus producing similar therapeutic effects as FAAH inhibitors, but without affecting the levels of other N-acylethanolamines or overexposing intracellular targets or enzymes to undegraded anandamide. Indeed, Fu and colleagues did screen a large library of synthetic compounds for their activity to bind to FLAT and inhibit anandamide cellular uptake, and came up with a novel compound, ARN272, which competitively inhibited FLAT, but not FAAH, and exerted potent anti-hyperalgesic actions in models of acute and inflammatory pain in a way attenuated by a CB1 receptor antagonist (Fu et al., 2011).
In conclusion, redundancy in endocannabinoid inactivation, of which the capability of anandamide to be oxidized by COX-2 or the existence of alternative splicing variants of FAAH are just recent examples, may complicate the development of safe and efficacious analgesics from inhibitors of this process, but also offers exciting new challenges for the design of multi-target therapies and of inhibitors of anandamide intracellular trafficking. Thus, the last word on new painkillers developed by targeting endocannabinoid inactivation is far from being set.
Alvarez-Jaimes LJ, Palmer JA. The Role of Endocannabinoids in Pain Modulation and the Therapeutic Potential of Inhibiting their Enzymatic Degradation. (2011) Curr Pharm Biotechnol. Apr 5. [Epub ahead of print]
Andersson DA, Gentry C, Alenmyr L, Killander D, Lewis SE, Andersson A, Bucher B, Galzi JL, Sterner O, Bevan S, Högestätt ED, Zygmunt PM. TRPA1 mediates spinal antinociception induced by acetaminophen and the cannabinoid Δ(9)-tetrahydrocannabiorcol. (2011) Nat Commun. 2011 Nov 22;2:551.
Costa B, Bettoni I, Petrosino S, Comelli F, Giagnoni G, et al. The dual fatty acid amide hydrolase/TRPV1 blocker, N-arachidonoyl-serotonin, relieves carrageenan-induced inflammation and hyperalgesia in mice. (2010) Pharmacol Res 61: 537-546.
Di Marzo V. Targeting the endocannabinoid system: to enhance or reduce? (2008) Nat Rev Drug Discov 7: 438-455.
Duggan KC, Hermanson DJ, Musee J, Prusakiewicz JJ, Scheib JL, et al. (R)-Profens are substrate-selective inhibitors of endocannabinoid oxygenation by COX-2. (2011) Nature Chemical Biology 7: 803–809.
Fu J, Bottegoni G, Sasso O, Bertorelli R, Rocchia W, et al. A catalytically silent FAAH-1 variant drives anandamide transport in neurons. (2011) Nat Neurosci. [Epub ahead of print]
Gammon CM, Allen AC, Morell P. Bradykinin stimulates phosphoinositide hydrolysis and mobilization of arachidonic acid in dorsal root ganglion neurons. (1989) J Neurochem 53: 95-101.
Hogestatt ED, Jonsson BA, Ermund A, Andersson DA, Bjork H, et al. Conversion of acetaminophen to the bioactive N-acylphenolamine AM404 via fatty acid amide hydrolase-dependent arachidonic acid conjugation in the nervous system. (2005) J Biol Chem 280: 31405-31412.
Kinsey SG, Nomura DK, O'Neal ST, Long JZ, Mahadevan A, et al. Inhibition of monoacylglycerol lipase attenuates nonsteroidal anti-inflammatory drug-induced gastric hemorrhages in mice. (2011) J Pharmacol Exp Ther 338: 795-802.
Long JZ, Nomura DK, Vann RE, Walentiny DM, Booker L, et al. Dual blockade of FAAH and MAGL identifies behavioral processes regulated by endocannabinoid crosstalk in vivo. (2009) Proc Natl Acad Sci U S A. 106:20270-5.
Maccarrone M, Dainese E, Oddi S. Intracellular trafficking of anandamide: new concepts for signaling. (2010) Trends Biochem Sci. 35(11):601-8.
Maione S, De Petrocellis L, de Novellis V, Moriello AS, Petrosino S, et al. Analgesic actions of N-arachidonoyl-serotonin, a fatty acid amide hydrolase inhibitor with antagonistic activity at vanilloid TRPV1 receptors. (2007) Br J Pharmacol 150: 766-781.
Maione S, Piscitelli F, Gatta L, Vita D, De Petrocellis L, et al Non-psychoactive cannabinoids modulate the descending pathway of antinociception in anaesthetized rats through several mechanisms of action. (2011) Br J Pharmacol 162: 584-596.
Mallet C, Barriere DA, Ermund A, Jonsson BA, Eschalier A, et al. TRPV1 in brain is involved in acetaminophen-induced antinociception. (2010) PLoS One 5.
Mallet C, Daulhac L, Bonnefont J, Ledent C, Etienne M, et al. Endocannabinoid and serotonergic systems are needed for acetaminophen-induced analgesia. (2008) Pain 139: 190-200.
Nomura DK, Lombardi DP, Chang JW, Niessen S, Ward AM, et al. Monoacylglycerol lipase exerts dual control over endocannabinoid and fatty acid pathways to support prostate cancer. (2011) Chem Biol 18: 846-856.
Piscitelli F, Bisogno,T., Di Marzo, V. Endocannabinoidomics: targeted lipidomics methods for the study of endocannabinoid-related molecules in the CNS. (2011) "I Cannabinoidi: dalla Biologia alla Clinica",Cagliari, Italy.
Rogosch T, Sinning C, Podlewski A, Watzer B, Schlosburg J, et al. Novel bioactive metabolites of dipyrone (metamizol). (2011) Bioorganic & Medicinal Chemistry, In Press, Available online 25 November 2011
Sagredo O, Pazos, R., Bisogno, T., Piscitelli, F., Petrosino, S., Valdeolivas, S., Hernández, M., Di Marzo, V., Fernández-Ruiz, J. POTENTIAL OXIDATION OF 2-AG BY COX-2 ENHANCES MALONATE TOXICITY IN THE STRIATUM: RELEVANCE FOR CANNABINOID TREATMENTS IN HUNTINGTON’S DISEASE. (2011) IACM 6th Conference on Cannabinoids in Medicine and 5th European Workshop on Cannabinoid Research.
Sigel E, Baur R, Racz I, Marazzi J, Smart TG, et al. The major central endocannabinoid directly acts at GABAA receptors. (2011) Proc Natl Acad Sci U S A 108: 18150-18155.
Woodward DF, Carling RW, Cornell CL, Fliri HG, Martos JL, et al. The pharmacology and therapeutic relevance of endocannabinoid derived cyclo-oxygenase (COX)-2 products. (2008) Pharmacol Ther 120: 71-80.