Pertwee RG. The Therapeutic Potential of Drugs That Target Cannabinoid Receptors or Modulate the Tissue Levels or Actions of Endocannabinoids. AAPS Journal. 2005; 7(3): E625-E654. DOI:  10.1208/aapsj070364

Correspondence to:
Roger G. Pertwee
Tel: + 44-1224-555740
Fax: + 44-1224-555844
Email: rgp@abdn.ac.uk

Received: May 31, 2005;  Accepted: July 21, 2005;  Published: October 24, 2005

Abstract

There are at least 2 types of cannabinoid receptor, CB₁ and CB₂, both G protein coupled. CB₁ receptors are expressed predominantly at nerve terminals and mediate inhibition of transmitter release, whereas CB₂ receptors are found mainly on immune cells, their roles including the modulation of cytokine release and of immune cell migration. Endogenous agonists for cannabinoid receptors also exist. These “endocannabinoids” are synthesized on demand and removed from their sites of action by cellular uptake and intracellular enzymic hydrolysis. Endocannabinoids and their receptors together constitute the endocannabinoid system. This review summarizes evidence that there are certain central and peripheral disorders in which increases take place in the release of endocannabinoids onto their receptors and/or in the density or coupling efficiency of these receptors and that this upregulation is protective in some disorders but can have undesirable consequences in others. It also considers therapeutic strategies by which this upregulation might be modulated to clinical advantage. These strategies include the administration of (1) a CB₁ and/or CB₂ receptor agonist or antagonist that does or does not readily cross the blood brain barrier; (2) a CB₁ and/or CB₂ receptor agonist intrathecally or directly to some other site outside the brain; (3) a partial CB₁ and/or CB₂ receptor agonist rather than a full agonist; (4) a CB₁ and/or CB₂ receptor agonist together with a noncannabinoid, for example, morphine or codeine; (5) an inhibitor or activator of endocannabinoid biosynthesis, cellular uptake, or metabolism; (6) an allosteric modulator of the CB₁ receptor; and (7) a CB₂ receptor inverse agonist.

Keywords: cannabinoid receptors, endocannabinoids, fatty acid amide hydrolase inhibitors, autoprotection, therapeutic strategies

Introduction

Mammalian tissues are now known to express at least 2 types of cannabinoid receptor, both of which are G-protein coupled.^1,2 These are CB₁ receptors, cloned in Tom Bonner’s laboratory in 1990 (Matsuda et al³), and CB₂ receptors, cloned by Sean Munro in 1993.⁴ Although CB₁ receptors are expressed by certain nonneuronal cells and tissues, for example, the pituitary gland, immune cells, and reproductive tissues, they are found predominantly at central and peripheral nerve terminals where they mediate inhibition of transmitter release. CB₂ receptors occur mainly on immune cells, their functions including the modulation, both within and outside the central nervous system, of cytokine release and immune cell migration. Thus, one common role of CB₁ and CB₂ receptors appears to be the regulation of ongoing release of chemical messengers, CB₁ receptors mainly from neurones and CB₂ receptors from immune cells. The finding that mammalian tissues express cannabinoid receptors was followed by the discovery of endogenous ligands for these receptors (Endocannabinoids section). These endogenous cannabinoids or endocannabinoids are all eicosanoids, 2 notable examples being N-arachidonoylethanolamine (anandamide) and 2-arachidonoyl glycerol. Endocannabinoids together with cannabinoid CB₁ and CB₂ receptors constitute the endocannabinoid system.

The discovery of the endocannabinoid system has prompted research directed at establishing its physiological and pathophysiological roles. This research has provided evidence first, that there are certain disorders in which endocannabinoid levels, cannabinoid receptor density, and/or cannabinoid receptor coupling efficiency increase in particular tissues, and second, that this upregulation of the endocannabinoid system often leads to a suppression of unwanted signs and symptoms, and so is autoprotective. The main objectives of this review are to summarize the evidence that the endocannabinoid system can be protective, and then to go on to consider possible therapeutic strategies by which such protection might best be exploited in the clinic. These are strategies in which endocannabinoid-induced protection is mimicked with directly acting CB₁ and/or CB₂ receptor agonists or in which it is augmented with drugs expected to delay the disappearance of endocannabinoids following their endogenous release or to induce an allosteric enhancement of cannabinoid receptor activation by released endocannabinoids. Evidence that the endocannabinoid system may sometimes be responsible for the production of undesirable effects is also briefly discussed. The review begins with short overviews of first, the pharmacological actions of important cannabinoid receptor ligands; second, the processes by which endocannabinoids are produced and then removed from their sites of action; and third, the progress that has been made to date in the development of drugs that can selectively inhibit one or other of these removal processes.

Cannabinoid Receptor Ligands

Several ligands that bind selectively to cannabinoid receptors are now available.^1,2 These fall essentially into 2 categories: the exogenous cannabinoids that are found in cannabis or have been designed and synthesized by chemists and the endogenous cannabinoids that occur naturally in mammalian tissues.

Exogenous Ligands

As detailed elsewhere,^1,2 a large number of CB₁ and CB₂ receptor agonists and antagonists have now been developed. Among the agonists are several compounds that bind more or less equally well to CB₁ and CB₂ receptors, examples being the “classical” cannabinoids, Δ⁹-THC and (–)-11-hydroxy-Δ⁸-THC-dimethylheptyl (HU-210), the “nonclassical” cannabinoid, CP55940, and the aminoalkylindole cannabinoid, R-(+)-WIN55212, which has marginally greater CB₂ than CB₁ affinity. Of these nonselective agonists, HU-210, CP55940, and R-(+)-WIN55212 have the highest CB₁ and CB₂ relative intrinsic activities, and HU-210 has the highest affinity for both CB₁ and CB₂ receptors. Δ⁹-THC, which is the main psychotropic constituent of cannabis, has lower CB₁ and CB₂ affinities and relative intrinsic activities than these other cannabinoids. Indeed, at both receptor types, Δ⁹-THC exhibits the mixed agonist-antagonist properties that are typical of a partial agonist. Another cannabis constituent, the classical cannabinoid cannabinol, also behaves as a partial agonist, at least at CB₁ receptors.⁵ Agonists with significant selectivity for CB₁ or CB₂ receptors have also been developed. Notable CB₁-selective agonists include the eicosanoids R-(+)-methanandamide, arachidonyl-2’-chloroethylamide (ACEA), arachidonylcyclopropylamide (ACPA), and O-1812, which are all analogs of the endocannabinoid, anandamide (Endocannabinoids section). CB₂-selective agonists include L-759633, L-759656, and JWH-133, all structural analogs of Δ⁹-THC; other notable examples are the nonclassical cannabinoid, HU-308, and the aminoalkylindoles, JWH-015 and AM1241.

Turning now to cannabinoid receptor antagonists, the compounds most commonly used in research are the CB₁-selective SR141716A, AM251, and AM281 and the CB₂-selective SR144528 and AM630. These are usually all classified as inverse agonists since they can, in at least some cannabinoid receptor-containing systems, produce effects by themselves that are opposite in direction from those produced by agonists for these receptors.⁶ It is thought that they may produce these inverse cannabimimetic effects by shifting cannabinoid receptors from a constitutively active “on” state to one or more constitutively inactive “off” states. This putative mechanism relies on the assumption that cannabinoid receptors can exist in a constitutively active state in which they undergo some degree of spontaneous coupling to their effector mechanisms even in the absence of an endogenously released or exogenously added agonist. “Neutral” antagonists have also been developed.⁶ These share the ability of the inverse agonists to block responses to cannabinoid receptor agonists but lack the apparent ability of inverse agonists to produce inverse cannabimimetic effects in cannabinoid receptor-containing systems in the absence of any endogenously released or exogenously added cannabinoid receptor agonist. Neutral antagonists are, however, expected to share the ability of CB₁ and CB₂ receptor inverse agonists to produce inverse cannabimimetic effects in tonically active biological systems when this tonic activity arises from ongoing endocannabinoid release onto cannabinoid receptors.

There is evidence that some established exogenous cannabinoid receptor agonists and antagonists have non-CB₁, non-CB₂ pharmacological targets.^2,6,7 These include TRPV1 (vanilloid VR1) receptors for some eicosanoid agonists (see also Endocannabinoids section), adenosine A₁ receptors for SR141716A and AM251, and possibly also the following:

• central putative TRPV1-like receptors for CP55940, R-(+)-WIN55212 and SR141716A but not AM251;
• central putative non-CB₁, non-CB₂, non-TRPV1 G protein-coupled receptors for R-(+)-WIN55212 but not Δ⁹-THC, HU-210, or CP55940;
• putative non-CB₁, non-CB₂, non-TRPV1 receptors on perivascular sensory neurons for Δ⁹-THC and cannabinol but not HU-210 or CP55940;
• putative non-I₁, non-I₂ imidazoline receptors for CP55940, R-(+)-WIN55212, and SR141716A;
• putative abnormal-cannabidiol receptors on mesenteric arteries for R-(+)-methanandamide and SR141716A but not for Δ⁹-THC or R-(+)-WIN55212;
• gap junctions and L-type Ca²⁺ channels for SR141716A;
• Ca²⁺-activated (BK_Ca) and ATP-sensitive potassium channels for SR141716A; and
• putative allosteric sites for certain exogenous cannabinoid receptor ligands on muscarinic M₁ and M₄, 5-HT₂ and 5-HT₃ receptors.

Evidence has also recently emerged for the presence of an allosteric site on the CB₁ receptor. Thus, we have found a series of novel compounds to behave as allosteric CB₁ receptor modulators. These compounds do not displace [³H]CP55940 from CB₁ binding sites but do modulate the rate at which [³H]CP55940 dissociates from these sites.⁸ Whether there are also allosteric sites on CB₂ receptors remains to be established.

Endocannabinoids

The most investigated of the endocannabinoids have been the polyunsaturated fatty acid amide, N-arachidonoylethanolamine (anandamide), and the polyunsaturated monoacylglycerol, 2-arachidonoyl glycerol.^1,2,9 Of these, anandamide has marginally greater CB₁ than CB₂ affinity and, like the exogenous cannabinoid Δ⁹-THC, exhibits relatively low efficacy at CB₁ and CB₂ receptors, behaving as a partial agonist at both receptor types. As to 2-arachidonoyl glycerol, this has been found in several investigations to have affinities for CB₁ and CB₂ receptors similar to those of anandamide but to exhibit higher CB₁ and CB₂ efficacy than anandamide. There has, however, been one recent investigation with human CB₁ receptor-containing tissue in which 2-arachidonoyl glycerol was found both to exhibit lower CB₁ receptor affinity than anandamide and to lack detectable CB₁ receptor efficacy at concentrations of up to 10 µM.¹⁰

It is likely that anandamide and/or 2-arachidonoyl glycerol have neuromodulatory roles (eg, by serving as retrograde synaptic messengers) and immunomodulatory roles (eg, through the regulation of cytokine release and immune cell migration).^1,2,11-13 It is also likely that both these endocannabinoids play a part in certain pathological processes both within and without the central nervous system.

Cannabinoid CB₁ and CB₂ receptors are not the only pharmacological targets for anandamide.^2,7,14-16 Thus, it is generally accepted that the TRPV1 receptor is activated by anandamide and by synthetic analogs of this endocannabinoid such as methanandamide, arachidonyl-2’-chloroethylamide, and N-(4-hydroxybenzyl)arachidonoylamine (AM404), although not by 2-arachidonoyl glycerol or by noneicosanoid cannabinoid receptor agonists such as R-(+)-WIN55212, CP55940, or HU-210. As detailed elsewhere,^2,7 additional targets that have been proposed for anandamide include

• central putative non-CB₁, non-CB₂, non-TRPV1 G protein-coupled receptors that are also activated by R-(+)-WIN55212;
• putative non-CB₁, non-TRPV1 receptors on neurons of the small intestine;
• putative non-I₁, non-I₂ imidazoline receptors that are also activated by CP55940 and R-(+)-WIN55212;
• putative abnormal-cannabidiol receptors on mesenteric arteries that are also activated by R-(+)-methanandamide but not by 2-arachidonoyl glycerol;
• putative non-CB₁, non-CB₂, non-TRPV1 receptors on coronary arteries;
• allosteric sites on muscarinic M₁ and M₄ receptors and on 5-HT₃ receptors that are also activated by certain exogenous cannabinoids; and
• allosteric sites on glutamate GLU_A1 and GLU_A3 receptors.

Other endocannabinoids are the anandamide analogs, dihomo-γ-linolenoylethanolamide, docosatetraenoylethanolamide, and possibly the CB₁-selective agonist 2-arachidonylglyceryl ether (noladin ether).⁹N-arachidonoyl dopamine, N-oleoyl dopamine, and oleamide may also be endocannabinoids as there are reports that these endogenous compounds can bind to CB₁ receptors with K_i values that are in the midnanomolar (N-arachidonoyl dopamine) or low micromolar range.^9,17,18N-arachidonoyl dopamine can induce both antinociception in the mouse hot-plate test and thermal hyperalgesia, is distributed differently from anandamide in the brain, and shares the ability of anandamide to activate TRPV1 receptors.^9,17N-oleoyl dopamine is also a TRPV1 receptor agonist. Indeed, it activates these receptors with particularly high potency and has also been found to induce both thermal hyperalgesia and allodynia.⁹ The mono-unsaturated fatty acid amide, oleamide, has been shown to activate CB₁ receptors at concentrations above 10 or 100 nM, to block gap junction-mediated cell-cell communication, and to increase food intake, and has been postulated to be involved in sleep regulation.^9,18 One other endogenous fatty acid amide that interacts with cannabinoid receptors at concentrations in the low micromolar range is O-arachidonoylethanolamine (virodhamine).⁹ In one investigation this ligand was found to activate CB₂ receptors but to exhibit either partial agonist or antagonist activity at CB₁ receptors,¹⁹ whereas in another it behaved as a CB₁ receptor antagonist/inverse agonist.¹⁰

The Fate of Endocannabinoids and Its Modulation by Drugs

The Biosynthesis and Fate of Endocannabinoids

The biosynthesis of anandamide takes place on demand in response to elevations of intracellular calcium. The immediate precursor to anandamide is N-arachidonoyl phosphatidylethanolamine, which is formed from phosphatidylcholine and phosphatidylethanolamine and is converted to anandamide by the action of N-arachidonoyl phosphatidylethanolamine-phospholipase D.^16,20 Anandamide is removed from its sites of action by cellular uptake processes that may involve a transmembrane carrier protein, membrane-associated binding proteins, and/or simple diffusion.²¹ It is then metabolized intracellularly. In most tissues, this metabolism is catalyzed by fatty acid amide hydrolase (FAAH),^16,22,23 an enzyme that in central neurons is located mainly on the cytosolic surfaces of smooth endoplasmic reticulum cisternae and mitochondria.²⁴ Anandamide is also sometimes metabolized by another intracellular enzyme, palmitoylethanolamide-preferring acid amidase (PAA),^22,25,26 and indeed, there is evidence that it is PAA rather than FAAH that is primarily responsible for deactivating anandamide in the mouse duodenum.²⁷ Other enzymes that can metabolize anandamide are cyclooxygenase-2, lipoxygenases, and cytochrome P450.²⁸

Like anandamide, 2-arachidonoyl glycerol is not stored but rather synthesized on demand in a manner that can be triggered by elevations of intracellular calcium.^16,20 Its synthesis depends on the conversion of 2-arachidonate-containing phosphoinositides to diacylglycerols (DAGs), which are then converted to 2-arachidonoyl glycerol by the action of DAG lipase. Also, like anandamide, 2-arachidonoyl glycerol is thought to be removed from its sites of action by cellular uptake and then to be metabolized intracellularly.^16,20,28 Although 2-arachidonoyl glycerol can readily be metabolized by FAAH as well as by cyclooxygenase-2 and lipoxygenases, there is evidence that the enzyme mainly responsible for its metabolism in vivo may be monoacyl glycerol (MAG) lipase.^20,22,29,30 FAAH, MAG lipase, and cannabinoid CB₁ receptors have broadly similar distribution patterns, at least within some brain areas.²⁴ It is noteworthy, however, that while MAG lipase and cannabinoid CB₁ receptors are located mainly presynaptically in these brain areas, FAAH is found mainly postsynaptically in somata and dendrites of principal neurons. Unlike polyunsaturated fatty acid amides, mono-unsaturated and saturated fatty acid amides are thought to be poor substrates for the putative anandamide transporter.²⁷ In contrast, cellular uptake may be the main means by which noladin ether is removed from its site of action as this putative endocannabinoid is considered not to be a likely substrate for enzymic hydrolysis.³¹N-oleoyl dopamine is also thought to be recognized by the putative anandamide transporter and to be only a poor FAAH substrate.⁹ However, 3 other putative endocannabinoids, oleamide, N-arachidonoyl dopamine, and virodhamine, all appear to be reasonable FAAH substrates.²³

Several endogenous fatty acid amides, in addition to anandamide, are metabolized by FAAH.^22,23,26,27 These include not only polyunsaturated fatty acid amides such as N-arachidonoyl dopamine but also a range of mono-unsaturated and saturated compounds. One endogenous saturated fatty acid amide that serves as a substrate for FAAH, and also for PAA, is palmitoylethanolamide.^22,25 This has antiinflammatory and antinociceptive properties, can potentiate anandamide-induced antinociception and TRPV1 receptor activation, lacks significant affinity for CB₁ or CB₂ receptors, and is thought to be an agonist for the PPAR-α receptor and possibly also for a putative “CB₂-like” receptor.^7,26,32 A second pharmacologically active endogenous substrate of FAAH and PAA is the mono-unsaturated fatty acid amide, oleoylethanolamide, which can inhibit both anandamide cellular uptake and metabolism. Like palmitoylethanolamide, oleoylethanolamide can activate PPAR-α receptors and has negligible affinity for CB₁ or CB₂ receptors.⁹ There is evidence that oleoylethanolamide serves as a satiety hormone, that it is released in the upper small intestine in response to food intake, and that it induces hypophagia by acting through intestinal PPAR-α receptors to activate vagal sensory afferent neurons that innervate higher brain structures involved in the control of energy balance.^26,27 Oleoylethanolamide is also a TRPV1 receptor agonist.⁹

Other pharmacologically active endogenous fatty acid amides include linoleamide, which is a FAAH substrate, may be involved in sleep regulation, and does not interact appreciably with cannabinoid receptors; N-arachidonoyl serine, which activates the putative abnormal-cannabidiol receptor; and N-palmitoyl dopamine and N-stearoyl dopamine, which potentiate anandamide at TRPV1 receptors but do not inhibit its cellular uptake or its metabolism by FAAH.⁹ Two other endogenous ligands of interest are N-arachidonoyl glycine and the unsaturated fatty acid amide, linoleoylethanolamide, as both have been reported to inhibit FAAH. It has also been found that N-arachidonoyl glycine has antinociceptive and anti-edema activity, that it lacks significant affinity for cannabinoid CB₁ receptors, and that it is a FAAH substrate.^9,23,33

Finally, it is noteworthy that some effects of endogenously released anandamide and 2-arachidonoyl glycerol may be enhanced through an “entourage effect” that relies on the corelease of other endogenous fatty acid derivatives. These derivatives include palmitoylethanolamide and oleamide, which can potentiate anandamide, and 2-linoleyl glycerol and 2-palmitoyl glycerol, which can potentiate 2-arachidonoyl glycerol.²

Inhibitors of Endocannabinoid Cellular Uptake and Intracellular Metabolism

The finding that the actions of anandamide and 2-arachidonoyl glycerol are terminated by cellular uptake and intracellular enzymic hydrolysis has been followed by the discovery of several drugs that will inhibit one or other of these processes.^16,23 Several of these inhibitors have been used as pharmacological tools in animal experiments directed at elucidating the physiological and pathological roles of anandamide or 2-arachidonoyl glycerol when these are released endogenously. These compounds are

• palmitylsulphonyl fluoride (AM374), an irreversible FAAH inhibitor (IC₅₀ = 13 or 50 nM) that inhibits this enzyme at concentrations below those at which it binds to CB₁ receptors (K_i = 520 nM)²³;
• methyl arachidonoyl fluorophosphonate (MAFP), an irreversible FAAH inhibitor (IC₅₀ = 1 to 3 nM) that has been reported also to inhibit MAG lipase (IC₅₀ = 2 to 800 nM) in brain tissue, to bind irreversibly to CB₁ receptors at concentrations in the low nanomolar range,²³ and to behave as an irreversible CB₁ receptor antagonist in one investigation³⁴ but not in another³⁵;
• diazomethylarachidonoylketone (DAK), an irreversible FAAH inhibitor (IC₅₀ = 500 nM), that has been reported also to inhibit MAG lipase in vitro at 1 µM and to bind to CB₁ receptors (IC₅₀ = 1.3 µM)^23,36;
• the alkylcarbamic acid aryl esters, URB532 and URB597, that are irreversible FAAH inhibitors (IC₅₀ = 63 and 4.6 nM, respectively) that do not inhibit MAG lipase, acetylcholinesterase, butyrylcholinesterase, or the putative anandamide membrane transporter at concentrations of up to 30 or 300 µM and lack significant affinity both for CB₁ and CB₂ receptors and for several other established receptors³⁷;
• the α-ketoheterocycle, 1-oxo-1[5-(2-pyridyl)-2-yl]-7-phenylheptane (OL-135), that is a potent reversible FAAH inhibitor (IC₅₀ = 2.1 nM), has less potency as an inhibitor of triacylglycerol hydrolase (IC₅₀ = 620 nM), and does not inhibit MAG lipase or a selection of other serine hydrolases at concentrations of up to 10 µM or exhibit significant affinity for CB₁ or CB₂ receptors (K_i > 10 µM)³⁸;
• N-arachidonoyl glycine, a FAAH substrate that is found endogenously, inhibits anandamide metabolism (IC₅₀ = 4.1 or 7 µM), and lacks significant affinity for CB₁ receptors (K_i > 10 µM)^33,39;
• N-arachidonoyl serotonin that inhibits anandamide metabolism (IC₅₀ = 5.6, 9, or 12 µM), does not significantly inhibit anandamide cellular uptake at 25 µM, and lacks significant affinity for CB₁ receptors (K_i > 50 µM)^40,41;
• palmitoylisopropylamide that inhibits FAAH (IC₅₀ = 12.9 µM) and probably also anandamide cellular uptake and does not readily displace [³H]CP55940 or [³H]R-(+)-WIN55212 from CB₁ or CB₂ receptors (IC₅₀ > 100 µM)⁴²;
• N-(4-hydroxybenzyl)arachidonoylamine (AM404), an inhibitor of anandamide cellular uptake (IC₅₀ = 1 to 11 µM) that, however, also inhibits FAAH (IC₅₀ = 0.5-5.9 µM or 22 µM or >30 µM), binds to CB₁ receptors (K_i = 1.76 µM), and potently activates TRPV1 receptors (EC₅₀ = 26 to 50 nM)^43-48;
• Two structural analogs of AM404, VDM11 and VDM13, that inhibit anandamide cellular uptake (IC₅₀ = 6.1 to 11 µM, and 12 µM, respectively), have less potency as TRPV1 receptor agonists (EC₅₀ >> 10 µM) and also as FAAH inhibitors in some experiments (IC₅₀ > 50 µM and = 27 µM, respectively) but not others (VDM11 IC₅₀ = 1.2 to 3.7 µM, and bind to CB₁ and CB₂ receptors with K_i values that exceed 5 or 10 µM^47,48;
• the (S)- and (R)-1’-(4-hydroxybenzyl) derivatives of N-oleoylethanolamine (OMDM-1 and OMDM-2 respectively) that inhibit anandamide cellular uptake (K_i = 2.4 and 3 µM, respectively; IC₅₀ = 2.6 or >20 µM, and 3.2 or 17 µM, respectively), have less potency as FAAH inhibitors (K_i or IC₅₀ >50 or >100 µM) or as TRPV1 receptor agonists (EC₅₀ ≥ 10 µM) and bind to CB₁ receptors (K_i = 12 µM and 5 µM, respectively)^48,49;
• N-(3-furylmethyl)arachidonoylamine (UCM707) that inhibits anandamide cellular uptake (K_i = 0.8, 25, 41, or >100 µM), is a FAAH substrate and inhibitor (IC₅₀ = 30 µM or >100 µM), binds to CB₁ (K_i = 4.7 µM) and CB₂ receptors (K_i = 67 nM), and lacks appreciable affinity for TRPV1 receptors (K_i > 5 µM).^48,50,51

AM374 is an analog of the irreversible serine protease inhibitor, phenylmethylsulphonyl fluoride (PMSF), which itself inhibits both FAAH (IC₅₀ = 290 nM to 15 µM) and MAG lipase (IC₅₀ = 155 µM or >500 µM).²³ MAG lipase inhibitors with sufficient potency and selectivity for use as pharmacological tools have yet to be developed.²³

Results obtained with the FAAH and cellular uptake inhibitors listed above should be interpreted with particular caution as the pharmacological characterization of these compounds is far from complete. An added complication for some of these inhibitors is that they can activate TRPV1 receptors, an action that is known to stimulate anandamide biosynthesis, presumably by increasing the intracellular concentration of calcium,¹⁶ and that will most likely also stimulate the biosynthesis of other endocannabinoids including 2-arachidonoyl glycerol. Based on the available data, compounds that currently show greatest promise as pharmacological tools are AM374, URB532, URB597, and OL-135 for FAAH inhibition, and OMDM-1 for inhibition of the cellular uptake of anandamide.

Evidence for an Upregulation of the Endocannabinoid System in Some Disorders

Since the discovery of anandamide in 1992, there have been several reports that tissue concentrations of this endogenous fatty acid amide increase in some human disorders and in animal models of certain diseases or disorders. More specifically, such increases have been detected in

• brain, blood, or circulating monocytes of patients with disorders that include schizophrenia, stroke, endotoxic (septic) shock, embryo implantation failure, and cancer (Table 1);
• blood in bone cement implantation syndrome (Table 1);
• brain and spinal cord in a mouse model of multiple sclerosis (Table 2);
• lumbar spinal cord in a mouse model of amyotrophic lateral sclerosis (Table 2);
• basal ganglia in a rat model of Parkinson’s disease (Table 2);
• ventral mesencephalon in a rat model of Huntington’s disease (Table 2);
• periaqueductal gray in a rat model of inflammatory pain (Table 2);
• urinary bladder in a rat model of painful hemorrhagic cystitis (Table 2);
• hypothalamus and uterus in mouse models of obesity (Table 2);
• cardiovascular tissue in experimental models of septic shock, cardiogenic shock, and biliary cirrhosis (Table 3);
• small intestine in rodent models of paralytic ileus, ileitis, and secretory diarrhea (Table 3); and
• lesioned brain areas in rat models of cerebral ischemia and excitotoxicity (Table 4).

In some of these investigations, tissue levels of 2-arachidonoyl glycerol were also monitored. These were found to increase in human colorectal and pituitary adenomas or carcinomas (Table 1); in bone cement implantation syndrome (Table 1); in mouse models of multiple sclerosis, amyotrophic lateral sclerosis, aversive memory extinction, and obesity (Tables 2 and 3); in rat models of hepatic ischemia-reperfusion injury, myocardial infarction, and ileitis (Table 4); and in a model of cardiogenic shock (Table 3). Usually these increases were detected in the same tissues as those in which increases in anandamide levels occurred. However, there have been at least 2 instances in which increases in anandamide and 2-arachidonoyl glycerol were found to have taken place in different cell-types: in a mouse model of septic shock¹¹³ and in experiments with human cancer cells⁷³ (Tables 1 and 3). There have also been a few instances in which increases in anandamide levels were found not to have been paralleled by any detectable increases in 2-arachidonoyl glycerol (Tables 1, 2, 3, and 4) or in which an increase in 2-arachidonoyl glycerol was not accompanied by any increase in anandamide (Tables 2, 3, and 4). Increased levels of 2-arachidonoyl glycerol in brain tissue have sometimes been observed in experiments in which anandamide levels were not also monitored, for example, in experiments with a mouse model of closed head injury and with a rat model of excitotoxicity (Table 4).

Increases in tissue levels of anandamide and 2-arachidonoyl glycerol are most likely usually triggered by increases in intracellular calcium. However, there is evidence that these increases are sometimes further modulated by changes in the ability of tissues to synthesize these compounds and/or to dispose of them by cellular uptake or subsequent enzymic degradation. Thus, for example, in a mouse model of obesity in which increases in uterine levels of both anandamide and 2-arachidonoyl glycerol were detected, these increases were found to be accompanied by a decrease in uterine tissue both of FAAH and MAG lipase activity and of anandamide transport across membranes¹⁰⁴ (Table 2). An increase in uterine DAG lipase activity was noted as well, suggesting that the increased uterine levels of 2-arachidonoyl glycerol observed in these experiments were partly attributable to an increase in the capacity of the uterus to synthesize this endocannabinoid. A decrease in FAAH activity has also been observed to accompany increases in anandamide levels in the peripheral lymphocytes of women with implantation failure (Table 1) and in rat striatal tissue in a model of Parkinson’s disease (Table 2). Also observed in these rat experiments was a decrease in anandamide cellular uptake. However, in experiments using a lipopolysaccharide (LPS) model of endotoxic shock, in one study LPS-induced increases in anandamide levels were reported to be paralleled by a decrease in the gene expression and activity of FAAH (in human peripheral lymphocytes),¹¹⁵ while in another study, in which a much lower concentration of LPS was used, increases in anandamide levels were found to be paralleled by an increase in the expression and activity of this enzyme (in mouse macrophages)¹¹⁴ (Table 3). In the second of these investigations, increases were also detected in the activities of 2 enzymes that can catalyze anandamide biosynthesis (N-acyltranferase and N-arachidonoyl phosphatidylethanolamine-phospholipase D). An increase in FAAH activity has also been detected in the small intestine of mice in which intestinal inflammation and increased motility had been induced by orally administered croton oil¹¹⁷ (Table 3). In these experiments, high levels of anandamide (and 2-arachidonoyl glycerol) were detected in the small intestine of both croton oil-treated and control mice. Some increases in anandamide tissue levels take place in the absence of changes in FAAH expression or activity. Thus, for example, in experiments with human colorectal cancer biopsies (Table 1) and with mouse models of diarrhea and paralytic ileus (Table 3), elevations in anandamide levels observed in the cancerous tissue and in mouse small intestine were not associated with any detectable changes in FAAH expression or activity.

In some disorders or animal models in which endocannabinoid tissue concentrations have been reported to be high, elevated expression levels and/or densities of cannabinoid CB₁ receptors have also been detected. Such upregulation of the CB₁ receptor has been observed in the following:

• postmortem cerebral cortical tissue of suicide victims (Table 1),
• tissue from humans with advanced liver cirrhosis (Table 1),
• brain tissue in a rat model of traumatic head injury (Table 4), and
• small intestine in mouse models of intestinal inflammation, secretory diarrhea, and paralytic ileus (Table 3).

Increases in CB₁ receptor density have also sometimes been detected in experiments in which no increases in endocannabinoid levels were observed or no measurements of endocannabinoid tissue concentrations made. These increases in receptor density were detected in the following:

• brain tissue taken postmortem from patients with schizophrenia or Parkinson’s disease (Table 1),
• the striatum in marmoset and rat models of Parkinson’s disease (Table 2),
• the spinal cord in a rat model of neuropathic pain (Table 2),
• cortical neurons in a rat model of focal cerebral ischemia (Table 4),
• the hippocampus in a rat febrile seizure model (Table 4),
• the colon in a mouse model of colitis (Table 3),
• mesenteric arteries in a rat model of cirrhosis (Table 3), and
• human prostate cancer cells (Table 1).

It is noteworthy that spontaneously hypertensive rats have been reported to exhibit increased CB₁ receptor density in myocardium and aorta but a decrease in myocardial anandamide levels¹¹¹ (Table 3). Conversely, excitotoxicity induced by N-methyl-D-aspartate (NMDA) in rats has been found to provoke not only a rise in anandamide concentrations but a fall in CB₁ receptor density and expression level in some brain areas (Table 4). A fall in CB₁ receptor density has also been detected in cerebral cortex and caudate putamen in a rat model of multiple sclerosis (Table 2). However, in this investigation, the reduced population of cannabinoid receptors was found to signal with increased efficiency. Increases in the efficiency of CB₁ receptor signaling have also been observed in postmortem brain tissue of patients who had suffered from Parkinson’s disease or were suicide victims (Table 1) and in the striatum in a marmoset model of Parkinson’s disease (Table 2). Additional examples of human disorders or of animal models of disorders in which signs of downregulation rather than upregulation of the endocannabinoid system have been observed are to be found in Tables 1 to 4.

It is likely that expression levels of CB₂ receptors also change in certain disorders. Indeed, Zhang et al¹⁰⁰ have found that when signs of neuropathic pain were produced in rats by sciatic nerve ligation, nonneuronal CB₂-expressing cells became detectable in the lumbar spinal cord. These cells were probably activated microglia that had migrated into this region of the cord. CB₂ expression was detectable only in certain regions of the dorsal and ventral horns, the precise distribution pattern of these receptors within the spinal cord varying with the model of experimental neuropathic pain used. No change in CB₂ expression was observed in the spinal cords of rats with inflamed paws. There have also been reports that

• in Alzheimer’s disease brain tissue, CB₂ receptors are selectively overexpressed in neuritic plaque-associated microglia and are also expressed in tangle-like neurons and dystrophic neurites but not in normal brain tissue^52,53 (see also Table 1),
• CB₂-expressing immune cells are present in brain tissue samples from macaques with simian immunodeficiency virus-induced encephalitis but not in samples from control animals⁸¹ (see also Table 2),
• CB₂ receptors are expressed in atherosclerotic plaques of human and mouse diseased arteries but not in nondiseased arteries (Table 1),
• CB₂ receptors are highly upregulated in human cirrhotic liver (Table 1),
• CB₂ (and CB₁) receptor expression is higher in human prostate cancer cells than in normal cells⁷⁵ (see also Table 1), and
• CB₂ but not CB₁ receptor expression is higher in biopsies from human astrocytomas exhibiting high-grade malignancy than in biopsy tissue exhibiting lower grade malignancy.¹³¹

Finally, it is important to bear in mind that changes in endocannabinoid tissue levels or in the expression level or density of cannabinoid receptors seems to take place in response not only to certain pathological insults but also to some nonpathological processes or stimuli such as fasting, feeding, stress, aging, and exercise (Tables 2 and 3).

Evidence That the Endocannabinoid System Is Autoprotective

There is evidence not only that tissue concentrations of endocannabinoids, cannabinoid receptor density and/or cannabinoid receptor coupling efficiency increase in a range of different disorders but also that these increases serve to reduce the severity of signs and symptoms of some of these disorders or even to oppose disease progression. Support for the hypothesis that the endocannabinoid system has such an “autoprotective” role has so far come mainly from experiments concerned with pain, multiple sclerosis, cancer, intestinal, mental and cardiovascular disorders, excitotoxicity, traumatic head injury, and Parkinson’s disease.

Pain

Reports are emerging that FAAH inhibitors such as OL-135, URB532, URB597, and N-arachidonoylglycine are antinociceptive in various rodent models of pain. These models are the formalin paw test of persistent inflammatory pain^38,39 in which it is already known that central nervous system (CNS) levels of anandamide rise above control values in response to the formalin (Table 2), and the tail-immersion and hot-plate tests of acute thermal pain.^37,38,132 There are reports too that the endocannabinoid cellular uptake inhibitors, OMDM-2 and VDM11, produce antinociception in the rat hot-plate test.¹³³

That the antinociception induced by inhibitors of FAAH or endocannabinoid cellular uptake is mediated by endogenous fatty acid ethanolamides is supported by several findings. These include the observations that OL-135, URB597, and N-arachidonoylglycine elevate endogenous levels of these amides in rodents^37,38,134 and that exogenously administered anandamide can be potentiated by OL-135 in the mouse tail-immersion test³⁸ and by AM404 in the mouse hot-plate test.⁴⁴ Also consistent with this hypothesis are the findings first, that mice from which FAAH has been genetically deleted (FAAH^−/− mice) exhibit reduced sensitivity to noxious stimuli when these are applied in the tail-immersion, hot-plate, or formalin paw test or in the carrageenan model of inflammatory pain^135,136; and second, that FAAH^−/− mice have elevated tissue levels of anandamide and other fatty acid ethanolamides and exhibit greater sensitivity than wild-type mice to the antinociceptive effect of exogenously administered anandamide.¹³⁵

Some experiments have been performed with transgenic (FAAH-NS) mice that express FAAH only within the nervous system. These differ from FAAH^−/− mice in not exhibiting enhanced levels of anandamide and other fatty acid ethanolamides in the CNS or reduced nociception in the tail-immersion or hot-plate plate test.¹³⁷ Consequently, it is likely that signs of hypoalgesia in these tests that are observed in FAAH^−/− mice or are induced by FAAH or endocannabinoid cellular uptake inhibitors in wild-type mice depend on the activation of targets located in the nervous system. It is also likely that these targets are mainly or wholly cannabinoid CB₁ receptors. Thus, Kathuria et al³⁷ found that the antinociceptive effects of URB532 and URB597 in the mouse hot-plate test were attenuated by SR141716A at a dose that by itself had no effect on nociception. Similarly, it was found by Lichtman et al³⁸ that SR141716A attenuated OL-135 induced-antinociception in mice in the tail-immersion, hot-plate, and formalin paw tests. None of the FAAH inhibitors used in these 2 investigations exhibited significant affinity for CB₁ (or CB₂) receptors, suggesting that these inhibitors did not produce antinociception by activating cannabinoid receptors directly. It has also been found that after they have been injected with SR141716A, FAAH^−/− mice no longer exhibit hypoalgesic responses to noxious stimuli in the tail-immersion test, the hot-plate test, or the formalin paw test (both phases).^135,136 There is good evidence that CB₂ receptor activation can induce antinociception in the second phase of the formalin paw test (see below). Even so, it is unlikely that the signs of hypoalgesia that are exhibited by FAAH^−/− mice in this pain model are mediated by CB₂ receptors as the ability of SR141716A to eliminate these signs in such animals is not shared by SR144528.¹³⁶

The concept that cannabinoid CB₁ receptors can mediate antinociception produced by inhibitors of FAAH or endocannabinoid cellular uptake or by genetic deletion of FAAH is consistent with several findings that have been extensively reviewed elsewhere.^20,138-140 Briefly, these findings are first, that CB₁ receptors are located on pain pathways in the brain and spinal cord and on the central and peripheral terminals of primary afferent neurons that mediate both neuropathic and nonneuropathic pain, and second, that cannabinoid receptor agonists can induce signs of antinociception when injected either systemically or directly onto a CB₁ receptor-expressing region of a pain pathway. CB₁ receptor agonists exhibit antinociceptive activity in a wide range of experimental pain models and this antinociception has often been found to be antagonized by SR141716A, albeit sometimes in models in which this antagonist produces signs of hyperalgesia when administered by itself. This hyperalgesia could reflect either antagonism of endocannabinoids released onto CB₁ receptors or an ability of SR141716A to reduce the extent to which CB₁ receptors couple spontaneously to their effector mechanisms (Exogenous Ligands section) or both of these actions. Since SR141716A-induced inverse agonism is expected to occur particularly in tissues in which there is a relatively high expression of CB₁ receptors, it is noteworthy that CB₁ receptor expression levels in the spinal cord increase in a rat model of neuropathic pain in which SR141716A has been found to be hyperalgesic^99,141 (see also Table 2).

That CB₁ receptors have a role in pain perception is also supported by the results from experiments in which antisense methods have been used to achieve a “knockdown” of CB₁ receptors in the brain or spinal cord.^142-144 Additional support comes from findings that mice from which the CB₁ receptor has been genetically deleted (CB₁^−/− mice) exhibit reduced antinociception following a forced swim in water at 34°C¹⁴⁵ and increased tactile sensitivity.¹⁴⁶ Unexpectedly though, there are also reports that compared with wild-type animals, CB₁^−/− mice exhibit either no differences in nociceptive behavior in hot-plate, tail-immersion, tail-flick, tail-pressure, or abdominal-stretch tests^146-148 or even signs of hypoalgesia in the hot-plate test and the first phase of the formalin paw test.¹⁴⁸

There is evidence that CB₂ receptors can also mediate analgesia. Thus, for example, when administered systemically to rats or mice the CB₂-selective agonists HU308, GW405833, and AM1241 have all been found to exhibit antinociceptive activity in various pain models in a manner that is opposed by a CB₂-selective antagonist. Such apparent CB₂-mediated antinociception has been observed for HU308 in the second phase of the mouse formalin paw test,¹⁴⁹ for GW405833 in rat carrageenan and rat and mouse Freund’s complete adjuvant models of inflammatory pain and in rat models of neuropathic and incisional pain,^150,151 and for AM1241 in rat models of allodynia and of thermal and inflammatory pain and in mouse and rat models of neuropathic pain.^146,152-155 It is unlikely that these CB₂-selective agonists were also activating CB₁ receptors, as AM1241 was no less effective in reducing signs of neuropathic pain in CB₁^−/− mice than in wild-type animals¹⁴⁶ and as HU-308 and AM1241 were not antagonized by CB₁-selective antagonists. It is possible that CB₂ receptors that mediate antinociception are located on nonneuronal cells in the skin and that, when activated, these receptors modulate the endogenous release of molecules that target peripheral nociceptors.¹⁵² Indeed, evidence has recently emerged that suggests that one mechanism by which CB₂ selective agonists such as AM1241 produce antinociception, at least in a rat model of thermal pain, may be by stimulating the release of β-endorphin from CB₂-expressing cells in the skin such as keratinocytes onto μ-opioid receptors located on the terminals of primary afferent neurons.¹⁵⁶ However, it would be premature to exclude other possibilities, for example, that CB₂ receptors can also be expressed by sensory neurons¹⁵⁷ and that such CB₂ receptors, or indeed microglial CB₂ receptors that appear in the spinal cord after nerve injury¹⁰⁰ (see also Table 2), can mediate antinociceptive effects. It also remains possible that CB₂-selective agonists can produce antinociception in at least some pain models by activating “CB₂-like” receptors, putative receptors at which the endogenous fatty acid amide palmitoylethanolamide may induce hypoalgesia (reviewed elsewhere^7,139).

In some pain models, signs of hypoalgesia that seem to be produced by endogenously released endocannabinoids appear to be entirely CB₁ receptor mediated. However, there is also some evidence that antinociception may sometimes be induced instead by endogenous activation of CB₂ or (CB₂-like) receptors. Thus, in FAAH^−/− mice, there is an elevation in the tissue levels of palmitoylethanolamide,¹³⁵ which may produce antinociception by acting on a CB₂-like receptor, and in the carrageenan model of inflammatory pain, although not in the formalin paw test (see preceding paragraph), these knockout mice exhibit signs of hypoalgesia that are not reduced by SR141716A but are lessened, albeit only partially, by SR144528.¹³⁶ It has also been reported that URB597 decreases carrageenan-induced mouse paw edema in an SR144528-sensitive manner.¹⁵⁸ Moreover, when administered by itself, this CB₂-selective antagonist has been found to produce signs of hyperalgesia in rats both in the carrageenan model¹⁵⁰ and in the first phase of the formalin paw test.¹⁵⁹ However, there are also several reports that CB₂ receptor antagonists are not hyperalgesic in these rat models of inflammatory pain or in models of allodynia or thermal pain.^152-155,160 A CB₁- or CB₂ receptor antagonist may of course sometimes fail to enhance signs of pain in an experimental model because the pain score is already maximal. Conversely, it remains possible that in experiments in which SR144528 was hyperalgesic, it was acting as an inverse agonist rather than as an antagonist of endogenously released endocannabinoids.

There is evidence that cannabinoid receptor agonists can also relieve pain in humans, for example, in patients with neuropathic pain¹⁶¹ and in some patients with cancer, multiple sclerosis, spinal cord injury, blepharospasm, brachial plexus damage, or pain from limb amputation.^161-169 Whether inhibitors of endocannabinoid metabolism or cellular uptake will also be able to induce analgesia in the clinic has still to be established.

Finally, there is evidence that CB₂ receptor inverse agonists have therapeutic potential as antiinflammatory agents. Thus, although there is little doubt that CB₂ receptor agonists can produce antinociception in experimental models of various kinds of pain that include inflammatory pain, it has also been found that the CB₂ receptor inverse agonists, SR144528 and JTE-907, can inhibit carrageenan-induced mouse paw edema,¹⁷⁰ indicating them to be antiinflammatory agents, and that a third CB₂ receptor inverse agonist, Schering-Plough’s compound 4j, inhibits immune cell migration both in vitro and in vivo.^171,172 There is also some evidence that although FAAH inhibitors may be effective against thermal and inflammatory pain, they lack potential for the management of neuropathic pain. Thus, in experiments with sciatic nerve ligation models, it has been observed that FAAH^−/− mice do not show signs of hypoalgesia¹³⁶ and that rats exhibit an antinociceptive response to URB597 only when this is administered at a dose well above its threshold dose for FAAH inhibition.¹⁷³

Multiple Sclerosis

There are 2 reports that in a mouse model of multiple sclerosis, spasticity can be ameliorated both by the FAAH inhibitor, AM374, and by the inhibitors of endocannabinoid cellular uptake, AM404, VDM11, OMDM-1, and OMDM-2.^77,133 The model used, the chronic relapsing experimental allergic encephalomyelitis (CREAE) model, is one in which levels of anandamide and 2-arachidonoyl glycerol in both brain and spinal cord are higher than in unlesioned animals (Table 2). In another model of multiple sclerosis in which mice are inoculated intracerebrally with Theiler’s murine encephalomyelitis virus (TMEV), OMDM-1 has been found to oppose the impaired rotarod performance and reductions in spontaneous motor activity exhibited by the lesioned animals and to enhance levels of anandamide although not 2-arachidonoyl glycerol in the spinal cords of these animals.¹⁷⁴

There are several reasons for believing that the amelioration of spasticity induced in CREAE mice by inhibitors of FAAH or endocannabinoid cellular uptake is mediated at least in part by CB₁ and possibly also by CB₂ receptors. First, the antispastic effect of the FAAH inhibitor, AM374, has been found to be blocked by SR141716A and SR144528.⁷⁷ As AM374 is not itself expected to bind to cannabinoid receptors at the dose used, this finding suggests that it produced its inhibitory effect on spasticity indirectly by enhancing endocannabinoid concentrations at these receptors. Second, Pryce et al¹⁷⁵ have found that compared with wild-type CREAE mice, CB₁^−/− CREAE mice exhibit an earlier onset of spasticity, more immobility, residual paresis, spinal cord axonal loss and spinal neurodegeneration, and greater mortality. Third, there have been several reports that the exogenous administration to lesioned rodents of cannabinoid receptor agonists, including anandamide and 2-arachidonoyl glycerol, R-(+)-WIN55212 and Δ⁹-THC, the CB₁-selective agonists, R-(+)-methanandamide and ACEA, and the CB₂-selective agonists, JWH-133 and JWH-015, can reduce spasticity or other signs of neurological damage such as tremor and spasm^77,176-178 or ameliorate atonia, ataxia, gait abnormalities, paralysis, moribundity, and mortality^179-181 or improve rotarod performance.¹⁸² These experiments were performed with rodent models of multiple sclerosis in which demyelination was induced by inoculation either with TMEV^181,182 or with mixtures containing CNS tissue or myelin basic protein (CREAE/EAE models). Finally, the CB₁/CB₂ receptor agonist R-(+)-WIN55212 has been found to ameliorate clinical signs of demyelination in mice in a manner that is both stereoselective (CREAE and TMEV mice) and susceptible to antagonism by SR141716A and SR144528 (CREAE mice).^176,177,181

Evidence that cannabinoids can reduce the spasms, spasticity, or tremor of multiple sclerosis has also been obtained in clinical trials with multiple sclerosis patients^{165,166,183,184} (also reviewed elsewhere¹⁶⁴). The degree of spasm or spasticity was either scored by the investigators using an objective measure or assessed subjectively by the patients themselves. The negative results sometimes obtained in such experiments when spasticity has been scored objectively^166,184,185 may well be a reflection of the low sensitivity of the available methods.

Further support for the hypothesis that CREAE mice release endocannabinoids onto cannabinoid receptors and that the resultant activation of these receptors reduces spasticity comes from the finding that CREAE mice with mild spasticity became significantly more spastic when injected with SR141716A alone or in combination with SR144528.¹⁷⁶ However, since these agents are both inverse agonists, it is also possible that they acted in an endocannabinoid-independent manner by reducing the extent to which cannabinoid receptors were coupling spontaneously to their effector mechanisms. Because SR141716A and SR144528 increase spasticity in CREAE mice, the finding that these cannabinoid receptor antagonists oppose R-(+)-WIN55212-induced reductions in the spasticity of CREAE mice should be interpreted with caution.

Results obtained with CREAE/EAE or TMEV models of multiple sclerosis also suggest that cannabinoid CB₁ or CB₂ receptor activation by exogenously administered or endogenously released agonists may oppose the progression of multiple sclerosis by slowing the neurodegenerative process,¹⁷⁵ reducing inflammation,^{174,179,181,182} and promoting remyelination.¹⁸²

Cancer

Evidence is emerging that certain types of cancer cells overproduce anandamide and/or 2-arachidonoyl glycerol (Table 1), that these endocannabinoids act through cannabinoid receptors to inhibit cancer cell proliferation or invasion, and that this inhibitory effect can be enhanced by inhibitors of endocannabinoid metabolism or cellular uptake.

Ligresti et al⁴¹ have detected both these endocannabinoids in biopsy specimens from human CB₁-, CB₂-, and FAAH-expressing colorectal tissue and found that the levels of anandamide and 2-arachidonoyl glycerol were highest in precancerous adenomatous polyps, lower in carcinomas, and lowest in normal tissue. Experiments were also performed with cultured colorectal cancer cells (CaCo-2 cells) that express CB₁ receptors and FAAH but not CB₂ receptors and contain anandamide and 2-arachidonoyl glycerol. The proliferation of these cells was inhibited by anandamide, by 2-arachidonoyl glycerol, by certain synthetic CB₁ receptor agonists, by the FAAH inhibitor N-arachidonoyl serotonin, and by the endocannabinoid cellular uptake inhibitors, VDM11 and VDM13, in a manner that was sensitive to antagonism by SR141716A. Of interest, anandamide did not inhibit the proliferation of CaCo-2 cells that had differentiated into enterocytes. These differentiated cells exhibit much lower malignancy and invasiveness, have a lower endocannabinoid content, and express more FAAH than undifferentiated CaCo-2 cells.

More recently, Bifulco et al¹⁸⁶ showed that in vivo growth of rat thyroid tumors in athymic mice could be inhibited by intratumoral injections of VDM11 or N-arachidonoyl serotonin. The tumor content of 2-arachidonoyl glycerol was increased by both these drugs, whereas the tumor content of anandamide (and palmitoylethanolamide) was increased only by N-arachidonoyl serotonin. It was also found that in vitro proliferation of these tumor cells was inhibited both by VDM11 and by the cannabinoid receptor agonist, 2-methyl-arachidonyl-2’-fluoro-ethylamide, and that this inhibition was blocked by SR141716A but not by SR144528. 2-Arachidonoyl glycerol and N-arachidonoyl serotonin, which also inhibited the in vitro proliferation of these cells, were less susceptible to antagonism by SR141716A. These findings are in line with evidence from previous investigations that both cancer cell proliferation in vitro and angiogenesis, tumor growth, and metastatic spreading of cancer cells in vivo can be decreased by CB₁ receptor activation.^187-191 Evidence that CB₂ receptors mediate antitumor effects also exists.^188,190,191 Intriguingly, it is likely that as well as acting through CB₁ receptors to inhibit the growth of tumors in vivo and cancer cell proliferation in vitro, the anandamide analog, 2-methyl-arachidonyl-2’-fluoro-ethylamide, increases expression of CB₁ receptors in these tumors and cancer cells but decreases the levels of these receptors in healthy, noncancerous cells.^187,189 Such an action would presumably be shared by other cannabinoid receptor agonists when these are exogenously administered or endogenously released.

In another recent investigation, Nithipatikom et al³⁶ found first, that certain CB₁ and CB₂-expressing human cultured androgen-independent prostate cancer cells produce 2-arachidonoyl glycerol at high concentrations, and second, that in vitro invasion of these cells could be inhibited by 2-arachidonoyl glycerol, noladin ether, R-(+)-WIN55212 and R-(+)-methanandamide, and also by 2 inhibitors of 2-arachidonoyl glycerol metabolism, MAFP and DAK. SR141716A but not SR144528 increased invasion of the prostate cancer cells and reversed the inhibitory effect of MAFP. It was also found that MAFP-induced inhibition of invasion of these cells could be enhanced by 2-arachidonoyl glycerol. Evidence has also recently been obtained from experiments with human cultured androgen-sensitive prostate cancer cells for a CB₁ and CB₂ receptor-mediated induction of apoptosis and inhibition of cell growth by R-(+)-WIN55212 and for the reduction by this cannabinoid receptor agonist of angiogenesis.⁷⁵

Intestinal Disorders

There is evidence first, that certain disorders characterized by inflammation of the gastrointestinal tract or by diarrhea may be associated with an increase in intestinal endocannabinoid levels and/or in the expression of CB₁ receptors by myenteric neurons (Table 3), second, that the resultant hyperactivity of the endocannabinoid system ameliorates at least some of the symptoms of these diseases and, third, that this amelioration can be mimicked by CB₁ receptor agonists or enhanced by inhibitors of endocannabinoid metabolism.¹⁹²

Izzo et al¹¹⁹ have obtained evidence that the endocannabinoid system acts through overexpressed CB₁ receptors to oppose cholera toxin-induced accumulation of intestinal fluid in mice. Thus, levels of anandamide, although not of 2-arachidonoyl glycerol or palmitoylethanolamide, increased in the small intestine in response to cholera toxin as did the expression of CB₁ receptors. The intestinal fluid accumulation induced by cholera toxin was increased by SR141716A but not by SR144528 and was reduced by CP55940 and ACEA pretreatment in a manner that was sensitive to antagonism by SR141716A but not SR144528. Intestinal fluid accumulation was also prevented by VDM11 in an SR141716A-sensitive manner but was unaffected by the CB₂-selective agonist JWH-015 or by the TRPV1 antagonist capsazepine.

Results obtained by Massa et al¹¹⁸ from experiments with CB₁^−/−, FAAH^−/−, and wild-type mice suggest that the CB₁ component of the endogenous cannabinoid system decreases colonic inflammation induced by intrarectal administration of 2,4-dinitrobenzene sulphonic acid (DNBS). DNBS-induced colitis was more marked in CB₁^−/− mice and less marked in FAAH^−/− mice than in wild-type mice. Moreover, in wild-type mice, DNBS-induced colitis was reduced by HU-210 and enhanced by SR141716A and provoked an increase in the number of CB₁-expressing cells in mouse colonic myenteric plexus. There is also a report that intestinal inflammation and hypermotility induced by oral croton oil is associated with an increase in CB₁ expression level in mouse jejunum.¹¹⁷ The sensitivity of croton oil-treated mice to CP55940 and cannabinol-induced inhibition of gastrointestinal transit of an orally administered charcoal suspension increased as well. High levels of anandamide and 2-arachidonoyl glycerol were detected in the small intestines of both croton oil-treated and control mice, even though FAAH activity was 2-fold higher in inflamed small intestine than in control tissue.

Mental Disorders

Giuffrida et al⁵⁴ have found that levels of anandamide but not of palmitoylethanolamide or oleoylethanolamide are markedly higher in the cerebrospinal fluid of antipsychotic-naïve first-episode paranoid schizophrenics and of schizophrenics taking “atypical” antipsychotics than in the cerebrospinal fluid of healthy controls. In contrast, anandamide levels were not elevated in schizophrenics taking “typical” antipsychotics or in patients with dementia or affective disorders. In the antipsychotic-naïve schizophrenics there was a negative correlation between anandamide levels and severity of symptoms, a finding that is at least consistent with the hypothesis that anandamide has a protective role in schizophrenia. These findings raise the possibility that in schizophrenia, exaggerated dopamine release onto postsynaptic D₂-like receptors triggers release of anandamide, which then acts as a retrograde messenger to induce a CB₁ receptor-mediated attenuation of dopamine release. This chain of events would most likely be prevented by typical antipsychotics as these block D₂-like receptors but be unaffected by atypical antipsychotics as these act mainly on 5-HT_2A receptors. Giuffrida et al⁵⁴ have also suggested that heavy cannabis use may promote psychotic episodes in vulnerable individuals by desensitizing CB₁ receptors such that the ability of endogenously released anandamide to ameliorate signs and symptoms of schizophrenia is compromised.

Marsicano et al¹⁰⁸ performed experiments with an aversive memory model in which mice are trained to associate a tone with a foot-shock such that once they are conditioned, they freeze when re-exposed just to the tone. Their main findings were that this tone provoked an increase in anandamide and 2-arachidonoyl glycerol levels in the basolateral amygdala complex but not the medial prefrontal cortex of conditioned mice, that extinction of tone-induced freezing behavior was more pronounced in wild-type than CB₁^−/− mice, and that extinction was also impaired by SR141716A in wild-type mice. These results prompted the suggestion that the endocannabinoid system could represent a therapeutic target for the treatment of diseases associated with inappropriate retention of aversive memories or with inadequate responses to aversive situations such as post-traumatic stress disorders or phobias.

Excitotoxicity and Traumatic Brain Injury

There is some evidence from experiments with mice that anandamide is released onto CB₁ receptors during excitotoxicity (Table 4) and that this release has a protective role.¹²⁸ More specifically, it has been found that kainic acid elevates anandamide but not 2-arachidonoyl glycerol or palmitoylethanolamide in the hippocampus, that this excitotoxin induces more severe seizures when the CB₁ receptor is genetically deleted or blocked with SR141716A, and that CB₁ receptor-expressing mice can be protected from kainic acid-induced seizures by the endocannabinoid cellular uptake inhibitor, UCM707. In addition, the data obtained suggest that the CB₁ receptors mediating this protective effect are located on principal forebrain neurons rather than on GABAergic interneurons, the likely end result of anandamide-induced activation of these receptors being inhibition of glutamate release.

There is also evidence that the endocannabinoid system may protect against the consequences of traumatic brain injury, although in this case through the release of 2-arachidonoyl glycerol. Thus, results obtained using a mouse model of closed head injury suggest that brain levels of 2-arachidonoyl glycerol increase in response to traumatic brain injury (Table 4) and that when administered exogenously, this endocannabinoid can act through CB₁ receptors to reduce brain edema and improve neurobehavioral function and clinical recovery.^126,193

Parkinson’s Disease

It has been postulated that it may prove possible to alleviate symptoms of Parkinson’s disease by using inhibitors of the metabolism or cellular uptake of anandamide to raise endogenous levels of this fatty acid amide so that its inhibitory effect on glutamate release from corticostriatal neurons is enhanced.^84,85 This hypothesis was prompted by results obtained in experiments using a rat model of Parkinson’s disease in which nigrostriatal dopaminergic neurons are destroyed unilaterally by injecting 6-hydroxydopamine close to the substantia nigra on one side of the brain. These experiments showed that the frequency of glutamatergic spontaneous excitatory postsynaptic potentials (sEPSPs) was greater in corticostriatal slices of lesioned than of control (sham-operated) animals and that this increased rate of firing could be reduced by the fatty acid amide hydrolase inhibitors, PMSF and MAFP, and by the anandamide cellular uptake inhibitors, AM404 and VDM11. The direct CB₁ receptor agonist, HU-210, was found to mimic this effect of these inhibitors, while the CB₁-selective antagonist, SR141716A prevented the effects of AM404 and PMSF. It was also found that 6-hydroxydopamine elevates levels of anandamide in the striatum but not in the cerebellum. Striatal and cerebellar levels of 2-arachidonoyl glycerol were unaffected. That inhibitors of anandamide metabolism or cellular uptake may alleviate symptoms of Parkinson’s disease is also supported by a report that intraperitoneally administered AM404 ameliorates akinesia and sensorimotor orientation and reduces amphetamine-induced turning behavior in 6-hydroxydopamine-lesioned rats.⁸⁶ The effect of AM404 on amphetamine-induced turning behavior was abolished by the CB₁-selective antagonist, AM251. There is also evidence from experiments with 6-hydroxydopamine-lesioned rats that drugs that enhance the activity of the endocannabinoid system may have the capacity to suppress or prevent unwanted dyskinesias that are often induced in parkinsonian patients by the therapeutic agent, L-dihydroxyphenylalanine (L-DOPA).⁸⁷

Results from other investigations have raised the possibility that symptoms of Parkinson’s disease could be alleviated by CB₁ receptor antagonists. Thus there are 2 reports from the same laboratory that the akinesia, sensorimotor orientation, and asymmetric motor behavior exhibited by rats with 6-hydroxydopamine-induced unilateral nigral lesions can be significantly attenuated by SR141716A or AM251 when these are given either systemically or unilaterally into the striatum or globus pallidus on the lesioned side of the brain.^194,195 An additional finding, that inhibitory effects of systemic SR141716A or AM251 on amphetamine-induced asymmetric motor behavior could be prevented by AM404, was taken as evidence for the presence of a CB₁ receptor-mediated modulation of nigrostriatal dopaminergic tone in the lesioned animals and also suggests that these antagonists were acting to reduce this tone. SR141716A and AM251 were most effective when given to rats exhibiting particularly severe behavioral signs of nigral degeneration, suggesting that cannabinoid CB₁ receptor antagonists might be useful for treating advanced stages of Parkinson’s disease in humans.¹⁹⁵ That CB₁ receptor antagonists could be used for the management of Parkinson’s disease has also been proposed by Lastres-Becker et al.⁶¹ This suggestion was prompted by results they obtained in experiments with tissue both from parkinsonian patients and from marmosets lesioned with 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP). More specifically, they found that CB₁ receptors in parkinsonian and MPTP-lesioned basal ganglia bind more [³H]CP55940 and exhibit greater coupling efficiency than control tissue and that these changes were less marked in the basal ganglia of MPTP-treated animals that had been exposed to L-DOPA. It is noteworthy, however, that Meschler et al¹⁹⁶ have reported that SR141716A does not alleviate motor deficits induced by MPTP in cynomolgus monkeys.

Clearly no firm conclusions can yet be drawn about the role of the endocannabinoid system in Parkinson’s disease or about the extent to which modulation of this system with drugs could alleviate undesirable symptoms associated either with the disease itself or with the established anti-parkinsonian drug, L-DOPA.

Cardiovascular Disorders

There is evidence that anandamide and 2-arachidonoyl glycerol are major contributors to hypotension associated with hemorrhagic, septic, or cardiogenic shock. Thus first, it has been found that hypotension induced in anaesthetized rats by bleeding, by acute myocardial infarction, or by intravenous administration of bacterial lipopolysaccharide (LPS), a major pathogenic factor in septic shock, can be opposed by SR141716A or AM281, whereas exogenous 2-arachidonoyl glycerol administration induces hypotension in anaesthetized normotensive animals.^{113,116,197,198} Second, there are reports that macrophages obtained from the blood of rats that are normotensive or in hemorrhagic shock can synthesize anandamide¹⁹⁷ and that in vitro administration of LPS elevates levels of 2-arachidonoyl glycerol in rat platelets and of anandamide in rat and mouse macrophages and human lymphocytes^113-115 (Table 3). There is also a report that anandamide and 2-arachidonoyl glycerol are detectable in monocytes and platelets isolated from rats after acute myocardial infarction but not in cells obtained from control animals¹¹⁶ (Table 3). Third, blood from hemorrhaged rats and cells isolated from rat blood after acute myocardial infarction or onset of LPS-induced hypotension can cause prolonged hypotension in anaesthetized normotensive rats.^113,116,197 This hypotension was found to be considerably attenuated or completely abolished by SR141716A, pointing to an involvement of CB₁ receptors. Moreover the ability of LPS to increase anandamide levels in mouse macrophages in vitro can be enhanced by the nonselective FAAH inhibitor, PMSF, and a much greater decrease in blood pressure can be elicited by LPS-treated peritoneal macrophages when these cells are obtained from FAAH^−/− rather than from FAAH^+/+ mice.¹¹⁴

It is possible that CB₁-mediated hypotension may aid survival in hemorrhagic and cardiogenic shock. Thus, as well as attenuating hypotension induced in urethane-anaesthetized rats by bleeding or acute myocardial infarction, SR141716A has been found to reduce the survival time of these animals.^116,197 Conversely, survival of hemorrhaged rats is significantly increased by the cannabinoid receptor agonists, Δ⁹-THC and HU-210.¹⁹⁷ In LPS-treated rats, however, survival can be improved both by Δ⁹-THC and by SR141716A or AM281.^113,198

There is also evidence that CB₁ receptors may mediate a protective hypotensive effect in spontaneously hypertensive rats.¹¹¹ These are animals in which it has been found that blood pressure is elevated by SR141716A and reduced by the FAAH inhibitor, URB597, and by 2 inhibitors of endocannabinoid cellular uptake, AM404 and OMDM-2. The blood pressure of normotensive rats was not affected by any of these drugs. Compared with normotensive rats, spontaneously hypertensive rats also express more CB₁ receptors in the myocardium and aorta (Table 3) and show greater sensitivity to anandamide- and HU-210-induced hypotension.¹¹¹ However, FAAH expression is higher and myocardial levels of anandamide (but not of 2-arachidonoyl glycerol) are lower in spontaneously hypertensive rats than in normotensive animals (Table 3), raising the possibility that the endocannabinoid system may protect against hypertension in this animal model by increasing “target organ sensitivity” rather than by increasing endocannabinoid levels.¹¹¹

Finally, it is possible that the endocannabinoid system may protect against atherosclerosis as first, CB₂ receptors are expressed by macrophages and T-lymphocytes within atherosclerotic plaques of human and mouse diseased arteries (Table 1), and second, there is evidence that when administered orally at 1 mg/kg per day, Δ⁹-THC can act through CB₂ receptors to produce a significant inhibition of atherosclerosis progression in mice.⁷¹

Therapeutic Strategies

It is clear from the preceding 2 sections of this review that the endocannabinoid system is thought to upregulate in some disorders and that one consequence of this upregulation may be the alleviation of symptoms or even the removal of the underlying causes of some of these symptoms. This section begins by considering how direct agonists might best be used in the clinic to mimic protective effects of the endocannabinoid system. It then goes on to discuss some alternative potential strategies that would augment these protective effects indirectly by potentiating endogenously released endocannabinoids through effects on their metabolism, cellular uptake, or receptors.

Direct Cannabinoid Receptor Agonists

Two CB₁/CB₂ cannabinoid receptor agonists have been licensed for clinical use in some countries for several years. These are dronabinol (Marinol; Δ⁹-THC) and the synthetic Δ⁹-THC analog, nabilone, which can be prescribed as antiemetics (both drugs) and to stimulate appetite (dronabinol).¹⁹⁹ Both these drugs are administered orally. In addition, Sativex, a medicine that contains Δ⁹-THC and cannabidiol, was recently licensed for the management of neuropathic pain associated with multiple sclerosis. ^199,200 A significant number of patients also claim to obtain relief from symptoms of multiple sclerosis and, indeed, from various kinds of chronic pain, from arthritis and/or from neuropathy by self-medicating with cannabis.^201,202 As to novel synthetic CB₁/CB₂ or CB₂ cannabinoid receptor agonists that exhibit antinociceptive activity in animal models of pain, there are as yet no clinical data, although several of these have been synthesized and investigated preclinically by drug companies (eg, by GlaxoSmithKlein^150,151 and by Bayer²⁰³).

A strong case can be made for using cannabinoid receptor direct agonists in the clinic now, particularly to relieve neuropathic pain and to ameliorate the spasms and spasticity of multiple sclerosis. However, a strong argument can also be made for conducting research directed at establishing how in the longer term the benefit-to-risk ratio of drugs that rely on the activation of cannabinoid receptors for their sought-after clinical effects can be improved. Some potential strategies for meeting this objective are listed below.

• Exploit the ability of cannabinoid CB₁ receptors expressed outside the central nervous system to mediate sought-after effects such as pain relief (reviewed elsewhere^139,140), the inhibition of cancer cell proliferation and spread (Cancer section) and the amelioration of certain intestinal and cardiovascular disorders (Intestinal Disorders and Cardiovascular Disorders sections). One possibility would be to administer CB₁ receptor agonists transdermally and, indeed, there is already evidence that this strategy is effective in human subjects for reducing experimental pain.^204,205 A second possibility would be to develop CB₁ receptor agonists that do not readily cross the blood-brain barrier.
• Exploit the ability of cannabinoid CB₁ receptors expressed in the spinal cord to mediate pain relief by administering CB₁ receptor agonists intrathecally (reviewed elsewhere^139,140). This is already an accepted procedure that is, for example, used for the self-administration of baclofen by some multiple sclerosis patients.
• Exploit the ability of cannabinoid CB₂ receptors to mediate pain relief by administering a CB₂-selective agonist (Pain section). This would of course avoid all unwanted consequences of CB₁ receptor activation, provided the dose of agonist used was not excessive.
• Exploit the ability of a low dose of a cannabinoid receptor agonist such as Δ⁹-THC both to interact synergistically with a low dose of an opioid such as morphine or codeine for the production of analgesia and to oppose onset of opioid tolerance (reviewed elsewhere^139,206). The success of this approach will depend on the extent to which these drugs interact synergistically in the clinic to produce undesirable effects. It is worth noting that there is already a report that orally-administered Δ⁹-THC had a “morphine-sparing effect” for a patient with severe abdominal pain.²⁰⁷
• Exploit the apparent ability of drugs such as the nonpsychotropic cannabinoid, cannabidiol, to reduce some of the undesirable effects of CB₁ receptor activation and perhaps also provide additional beneficial effects.²⁰⁸ Indeed, the licensed medicine, Sativex, contains both cannabidiol and Δ⁹-THC.^199,200

For some clinical applications it may be advantageous to combine 2 or more of these potential strategies. For example, neuropathic pain might sometimes best be relieved by targeting both CB₁ and CB₂ receptors, perhaps by administering a CB₂ receptor agonist together with a CB₁ agonist that does not readily cross the blood-brain barrier.

Any agonist that possesses CB₁ or CB₂ selectivity is expected to target CB₁ or CB₂ receptors selectively only within a finite dose range and to activate both receptor types when administered at a dose that lies above this range. Such a finding has, for example, been made with the CB₂-selective agonist GW405833.¹⁵¹ For any particular agonist, the width of its CB₁ or CB₂ selectivity window will of course be affected by the CB₁ to CB₂ receptor ratios of the tissues in which sought-after (or unwanted) effects are induced by CB₁ or CB₂ receptor activation. It is noteworthy, therefore, that not all tissues express CB₁ and CB₂ receptors in equal numbers in health and that there is also evidence that disparate changes in CB₁ and CB₂ expression levels may be induced in some cells or tissues either pathologically or pharmacologically. Consequently, when an agonist that binds more readily either to CB₁ or to CB₂ receptors is administered to patients, it is likely that the CB₁ or CB₂ selectivity of this ligand will not be the same in all tissues that express both receptor types and also that any selectivity will be lost when a certain dose level is exceeded. It is also important to bear in mind that there is evidence that some cannabinoid receptor agonists can activate non-CB₁, non-CB₂ pharmacological targets. Since cannabinoid receptor agonists differ in the extent to which they appear to interact with each of these proposed additional targets, it follows that some ligands with apparently similar abilities to activate CB₁ and/or CB₂ receptors are likely to possess different pharmacological profiles.

Indirect Cannabinoid Receptor Agonists

Existing indirect cannabinoid receptor agonists are all drugs that delay the removal of endocannabinoid molecules from their sites of action. They can be subdivided into FAAH inhibitors that inhibit the metabolism of anandamide and certain other fatty acid amides, MAG-lipase inhibitors that inhibit the metabolism of 2-arachidonoyl glycerol and related compounds, and drugs that inhibit endocannabinoid cellular uptake (Inhibitors of Endocannabinoid Cellular Uptake and Intracellular Metabolism section). Following the discovery of allosteric sites on CB₁ receptors, it may also prove possible to develop a second category of indirect cannabinoid receptor agonists that act by allosterically enhancing the ability of endogenously released (and exogenously administered) ligands to activate cannabinoid CB₁ receptors. The argument for using indirect agonists as therapeutic agents is that there are disorders in which endocannabinoids appear to be released selectively onto just certain populations of cannabinoid receptors, some or all of which mediate symptom relief. Consequently, drugs that potentiate endocannabinoids are expected to augment the relief of symptoms produced by such disorders without eliciting as many unwanted cannabinoid receptor-mediated responses as direct agonists that of course will activate all accessible CB₁ and/or CB₂ receptors.

Currently, much attention is being directed at the possibility of developing FAAH inhibitors as medicines. As a result FAAH inhibitors with some degree of selectivity have been developed, and the ability of these compounds to ameliorate symptoms in animal models of pain, multiple sclerosis, excitotoxicity, Parkinson’s disease, and hypertension and to reduce cancer cell proliferation has been demonstrated. There is also evidence from animal experiments that FAAH inhibitors are more selective than direct CB₁ receptor agonists, at least after single administration. Thus, for example, in contrast to direct CB₁ agonists, FAAH inhibitors can produce antinociception in mice at doses that do not also induce hypomotility, hypothermia, or catalepsy.^20,37,38 It is noteworthy, however, that there are as yet no published clinical data for this group of compounds. Also, although FAAH inhibitors can induce signs of analgesia in animal models of acute and inflammatory pain, there is some doubt as to whether they would be effective against neuropathic pain. Some important pharmacological considerations that relate to FAAH inhibitors are listed below.

• Inhibition of FAAH is likely to augment anandamide levels not only at cannabinoid receptors but also at its other targets, for example the vanilloid TRPV1 receptor and the putative abnormal cannabidiol receptor, if this endocannabinoid is being released onto these other targets.
• Inhibition of FAAH is also likely to augment levels of certain other endogenous fatty acid amides when these are undergoing release. These amides include 2 cannabinoid receptor ligands, oleamide which might well increase drowsiness if its levels in the brain are elevated,²⁰⁹ and N-arachidonoyl dopamine. They also include ligands that do not readily bind to cannabinoid receptors such as palmitoylethanolamide, which produces antinociception possibly by acting on peripheral “CB₂-like” receptors (Pain section), oleoylethanolamide, which is thought to stimulate fat utilization through activation of PPAR-α receptors, and N-arachidonoyl glycine, which is itself a FAAH inhibitor that can induce antinociception in animals when administered exogenously.
• A FAAH inhibitor is not expected to enhance the ability all endocannabinoids to induce symptom relief. Thus, some putative endocannabinoids are metabolically stable (eg, noladin ether) or are metabolized mainly by enzymes other than FAAH (eg, 2-arachidonoyl glycerol).
• Inhibition of FAAH may increase the extent to which the fatty acid amide substrates of this enzyme are degraded by other enzymes. For example, anandamide is known also to be metabolized by PAA, by cyclo-oxygenase-2, by lipoxygenases, and by cytochrome P450. Moreover, as there is evidence that the lipoxygenase metabolites of anandamide are more potent than their parent compound as TRPV1 agonists,²¹⁰ it is possible that some patients receiving a FAAH inhibitor would experience either hyperalgesia due to TRPV1 activation or analgesia resulting from TRPV1 desensitization.

Turning now to inhibitors of endocannabinoid cellular uptake, there is evidence that these too have therapeutic potential. This comes from experiments in which such inhibitors have been found to decrease cancer cell proliferation and tumor growth and to ameliorate symptoms in animal models of pain, multiple sclerosis, excitotoxicity, Parkinson’s disease, cholera toxin-induced diarrhea, and hypertension. As to MAG lipase inhibitors, any assessment of their therapeutic potential must await the development of suitable compounds, there being a need for compounds that inhibit MAG lipase at doses that do not also inhibit FAAH or interact directly with cannabinoid receptors or with other pharmacological targets that might compromise their selectivity. There is also an urgent need for an allosteric enhancer of the CB₁ receptor that exhibits appropriate efficacy and selectivity. Such an enhancer will most likely differ from a FAAH inhibitor in the following ways:

• by possessing the ability to produce greater symptom alleviation in at least some disorders through the augmentation of CB₁ receptor-mediated responses to all endocannabinoids no matter how their actions are terminated and
• by lacking the ability to augment the activation by anandamide or other fatty acid amides of targets other than CB₁ receptors.

Cannabinoid Receptor Expression Levels and Signaling

There is evidence that upregulation of the endocannabinoid system leading to an alleviation of symptoms can take the form not only of enhanced endocannabinoid release but also of an increase in cannabinoid receptor density or coupling efficiency in tissues in which these receptors mediate symptom relief or inhibition of disease progression when activated by endogenously released endocannabinoids or by exogenously administered cannabinoid receptor agonists. Indeed, results from rat experiments suggest that the endocannabinoid system may protect against hypertension through an increase in CB₁ receptor density/sensitivity rather than through an increase in endocannabinoid release (Cardiovascular Disorders section). Other seemingly protective increases in CB₁ and/or CB₂ receptor density or coupling efficiency have been detected in human and mouse atherosclerotic plaques, in human prostate cancer cells, and in animal models of Parkinson’s disease, neuropathic pain, diarrhea, intestinal inflammation, and colitis.

It has been reported that cannabinoid receptor expression is higher in some cancer cells than in normal cells^75,187,189 and also that it sometimes correlates with the degree of cancer cell malignancy.¹³¹ It will be important to identify other disorders in which cannabinoid receptor upregulation is confined entirely or mainly to tissues in which activation of cannabinoid receptors is expected to alleviate symptoms or to inhibit disease progression. This is because such a pattern of receptor upregulation may well improve the selectivity of cannabinoid receptor agonists by enhancing their potency for the production of sought-after effects more than their potency for the production of unwanted effects. Selectivity improvements resulting from an uneven cannabinoid receptor upregulation of this kind are expected to be most evident for partial agonists such as Δ⁹-THC and cannabinol. Thus although an increase in receptor density will augment the potencies of both full and partial agonists, it will often also increase the size of the maximal response to a partial agonist without affecting the maximal response to a full agonist. This difference between the pharmacology of full and partial agonists can be seen in results from experiments with cannabinol and CP55940 in which an increase in the intestinal expression of CB₁ receptors (and intestinal inflammation) had been induced in mice by oral croton oil and in which the measured response was cannabinoid-induced CB₁ receptor-mediated inhibition of upper gastrointestinal transit of a charcoal suspension.¹¹⁷ Thus, it was found that this increase in CB₁ expression level was accompanied not only by a leftward shift in the log dose-response curve of cannabinol but also by an increase in the size of the maximal effect of this partial agonist. In contrast, the higher efficacy agonist, CP55940, exhibited an increase in its potency but no change in its maximal effect.

The upregulation of a subpopulation of cannabinoid receptors that alleviates symptoms or inhibits disease progression when activated should of course also augment protective effects of endogenously released endocannabinoids and hence of FAAH inhibitors and other types of indirect agonist. For anandamide, cannabinoid receptor upregulation is likely to produce an increase in maximal effect as well as potency as this endocannabinoid is a CB₁ and CB₂ receptor partial agonist. On the other hand, 2-arachidonoyl glycerol would be expected to exhibit only an increase in potency as it has been found in several investigations to exhibit much higher CB₁ and CB₂ receptor efficacy than anandamide (Endocannabinoids section).

Tolerance can develop to CB₁ receptor agonists and there is evidence that this is often caused by a reduction in CB₁ receptor density or signaling.²¹¹ Consequently, it is tempting to speculate that if a particular disease causes an increase in cannabinoid receptor density or signaling this receptor upregulation may oppose the ability of endogenously released or exogenously administered cannabinoid receptor agonists to induce tolerance. If such receptor upregulation is restricted largely to receptors that mediate an alleviation of symptoms or an inhibition of disease progression, then patients would be protected from the development of tolerance to sought-after but not to unwanted effects of a cannabinoid receptor agonist in a manner that would lead to a broadening of the agonist’s therapeutic window in response to its repeated administration. There is already good evidence that tolerance develops less readily to some effects of cannabinoid receptor agonists than to others (reviewed elsewhere²¹²) and, indeed, that some sought-after therapeutic effects of a CB₁ receptor agonist may be more resistant to tolerance development than some of its unwanted effects.²¹³ CB₁ receptor agonists may also sometimes widen their own therapeutic windows by inducing a selective increase in the density of CB₁ receptors that are mediating a sought-after effect such as inhibition of the growth and proliferation of cancer cells.

Damaging Upregulation of the Endocannabinoid System?

There are some disorders in which the endocannabinoid system appears to upregulate to induce undesirable symptoms, an indication that this system may sometimes malfunction. For example, it has been postulated that upregulation of the endocannabinoid system can

• impair fertility in some women as there are reports (1) that FAAH activity and expression fall in the lymphocytes of some women who miscarry or who fail to become pregnant after in vitro fertilization (IVF) and embryo transfer, (2) that this fall is accompanied in the IVF-embryo transfer group of women by a corresponding rise in blood anandamide^65,66 (see also Table 1), and (3) that anandamide can impair embryo implantation and development in mice²¹⁴;
• contribute toward obesity in some individuals as, for example, it has been found that hypothalamic levels of anandamide and/or 2-arachidonoyl glycerol are raised in obese db/db and ob/ob mice and fa/fa rats (Table 2), that CB₁ receptor expression levels are elevated in the adipose tissue of obese fa/fa rats (Table 2), and that SR141716A reduces the food intake of db/db and ob/ob mice, reduces the body weight of db/db mice, and decreases the food intake, body weight, and waist circumference of obese humans^102,215;
• contribute toward osteoporosis and other diseases in which there is bone loss, as bone mineral density is greater in CB₁^−/− mice than in wild type animals, as ovariectomy does not induce bone loss in CB₁^−/− mice, as AM251 and SR144528 protect against ovariectomy-induced bone resorption, as CB₁ and CB₂ receptor antagonists/inverse agonists significantly inhibit evoked osteoclast formation in CB₁ and CB₂ receptor-expressing mouse bone marrow cultures, and as anandamide and CP55940 can stimulate osteoclast formation in these cultures²¹⁶;
• contribute toward the production of cerebral injury in stroke as unilateral ischemia/reperfusion injury induced in rats by transient middle cerebral artery occlusion has been reported to be associated with an elevation of whole brain anandamide and as infarct volume and neurological impairment in lesioned rats were both found to be less after pretreatment with selective CB₁ receptor antagonists¹²³ (see also Table 4);
• contribute toward the life-threatening excessive hypotension of endotoxaemic shock triggered by advanced liver cirrhosis^68,112 (Table 3) or by the cemented hip arthroplasty procedure⁷⁶ (Table 1) (but may aid survival in hemorrhagic, septic, and cardiogenic shock);
• contribute toward the development of hyperreflexia and hyperalgesia during cystitis as it has been found that urinary bladder levels of anandamide increase in a rat model of this disorder (Table 2) and that reflex activity increased both when anandamide was applied to healthy bladders and when the FAAH/anandamide cellular uptake inhibitor, palmitoylisopropylamide, was applied to healthy or inflamed bladders¹⁰¹;
• contribute toward intestinal inflammation in ileitis as there is a report that ileal levels of anandamide (and 2-arachidonoyl glycerol) increase in a rat model of this disorder (Table 3), as inflammation and fluid accumulation in the rat ileum increase in response to intraluminal injection of either of these endocannabinoids, and as intraluminal injections of the FAAH/MAG lipase inhibitors, PMSF or MAFP, augment intra-ileal fluid accumulation induced by Clostridium difficile toxin A¹²⁰;
• contribute toward intestinal hypomotility in paralytic ileus as reductions in mouse intestinal motility induced by peritoneal irritation with acetic acid has been shown to be counteracted by SR141716A (although not by SR144528), to be exacerbated by the anandamide cellular uptake inhibitor, VDM11, in an SR141716A-sensitive manner, and to be accompanied in the small intestine by increases in anandamide levels and in the neural density of CB₁ receptor immunoreactivity¹²¹ (see also Table 3).

Endocannabinoids appear to produce many of their unwanted effects by interacting with cannabinoid CB₁ receptors. Thus, results from animal experiments suggest that endocannabinoids act on these receptors to induce (1) impairment of embryo implantation and development (reviewed elsewhere²¹⁴), (2) increases in food intake and body weight associated with obesity,¹⁰² (3) cerebral injury in stroke,¹²³ (4) hypotension in liver cirrhosis,⁶⁸ and (5) intestinal hypomotility in paralytic ileus.¹²¹ As to the CB₂ component of the endocannabinoid system, evidence is emerging that this too may sometimes induce unwanted effects by promoting inflammation in a manner that can be counteracted by CB₂ receptor inverse agonism. There are also unwanted effects of the endocannabinoids that appear to be mediated by TRPV1 rather than CB₁ or CB₂ receptors. These are urinary bladder hyperreflexia in cystitis, and intestinal inflammation and fluid accumulation in ileitis.^101,120 In these instances the undesirable effects can be attenuated by TRPV1 receptor antagonism but not CB₁ receptor antagonism which, indeed, was found to exacerbate anandamide-induced bladder hyperreflexia. Clearly then, there may be a place in the clinic for CB₁ and TRPV1 receptor antagonists and possibly also for CB₂ receptor inverse antagonists. Indeed, the CB₁ receptor antagonist, SR14171A (rimonabant; Acomplia), is already in the final stages of development as an anti-obesity agent.²¹⁵ Selective inhibitors of endocannabinoid biosynthesis, activators of endocannabinoid-metabolizing enzymes, and allosteric antagonists of the CB₁ receptor also have therapeutic potential for the management of some disorders with symptoms that are induced by endogenously released cannabinoids. However, such drugs have yet to be developed.

Future Directions

There is no doubt that the endocannabinoid system upregulates both in response to certain pathological changes or to stress and during normal physiological events such as aging, feeding, fasting, and exercise. There is also good evidence that this upregulation can take the form of an increase both in the release of endocannabinoid molecules onto their receptors and in the density or coupling efficiency of some of these receptors. Often, this upregulation appears to protect the organism from unwanted symptoms or even to slow the progression of a disease. However, there is also evidence that upregulation of the endocannabinoid system can sometimes trigger unwanted symptoms, an indication that this system has its own pathology and possibly also that it is sometimes influenced detrimentally by pathological events taking place in some other system from which it receives input.

As to the future, it will be important to obtain a more complete list of pathological processes that are modulated by the endocannabinoid system and to determine for each of these processes whether this modulation has desirable or undesirable consequences. It will also be important to establish both the nature of this modulation and the extent to which it is confined to tissues in which cannabinoid receptor activation affects symptoms or disease progression. This in turn will help identify the best pharmacological strategy for the management of any particular disorder in which the endocannabinoid system has upregulated, be this to administer a CB₁ and/or CB₂ receptor agonist or antagonist that does or does not readily cross the blood-brain barrier, a CB₁ and/or CB₂ receptor agonist intrathecally or directly to some other site outside the brain, a partial CB₁ and/or CB₂ receptor agonist rather than a full agonist, a CB₁ and/or CB₂ receptor agonist together with a noncannabinoid, an inhibitor or activator of endocannabinoid biosynthesis, cellular uptake or metabolism, an allosteric modulator of the CB₁ receptor, or a CB₂ receptor inverse agonist. Other important objectives for future research are

• to obtain a more complete understanding of the processes that determine the biosynthesis, release, and fate of endocannabinoids,
• to identify those drugs that are best at modulating these processes selectively or at producing selective allosteric enhancement or antagonism of cannabinoid receptors,
• to characterize the pharmacology of these drugs and to test them in the clinic,
• to investigate the extent to which tolerance develops over time to protective effects of endocannabinoids when these are released by themselves or in the presence of drugs that augment the tissue levels of released endocannabinoids,
• to establish in greater detail the part played by non-CB₁, non-CB₂ targets in both the protective and the undesirable consequences of endogenous cannabinoid release, and
• to gain a better understanding of the part played by spinal or peripheral CB₁ and CB₂ receptors in modulating the symptoms and progression of particular diseases.

Finally, since the completion of this review, evidence has emerged that the orphan G protein-coupled receptor, GPR55, is a cannabinoid receptor,^217,218 prompting a need for research that will identify the physiological and/or pathological roles of this receptor and characterize its pharmacology. The discovery of a selective MAG lipase inhibitor, URB602, has also just been announced.²¹⁹

Acknowledgments

The writing of this review was supported by grants from NIDA (DA09789) and from GW Pharmaceuticals (Salisbury, Wiltshire, UK).

References

1. Howlett AC, Barth F, Bonner TI, et al.  International Union of Pharmacology. XXVII. Classification of cannabinoid receptors. Pharmacol Rev. 2002;54:161-202.
PubMed  DOI: 10.1124/pr.54.2.161

2. Pertwee RG. Pharmacological actions of cannabinoids. In: Pertwee RG,  ed. Cannabinoids, Handbook of Experimental Pharmacolology. Heidelberg, Germany: Springer-Verlag; 2005;168:1-51. 

3. Matsuda LA, Lolait SJ, Brownstein MJ, Young AC, Bonner TI.  Structure of a cannabinoid receptor and functional expression of the cloned cDNA. Nature. 1990;346:561-564.
PubMed  DOI: 10.1038/346561a0

4. Munro S, Thomas KL, Abu-Shaar M.  Molecular characterization of a peripheral receptor for cannabinoids. Nature. 1993;365:61-65.
PubMed  DOI: 10.1038/365061a0

5. Pertwee RG.  Pharmacology of cannabinoid receptor ligands. Curr Med Chem. 1999;6:635-664.
PubMed 

6. Pertwee RG.  Inverse agonism and neutral antagonism at cannabinoid CB₁ receptors. Life Sci. 2005;76:1307-1324.
PubMed  DOI: 10.1016/j.lfs.2004.10.025

7. Pertwee RG.  Novel pharmacological targets for cannabinoids. Curr Neuropharmacol. 2004;2:9-29.
PubMed  DOI: 10.2174/1570159043476927

8. Price MR, Baillie G, Thomas A, et al. Allosteric modulation of the cannabinoid CB₁ receptor. Mol Pharmacol.  In press.. 

9. Bradshaw HB, Walker JM.  The expanding field of cannabimimetic and related lipid mediators. Br J Pharmacol. 2005;144:459-465.
PubMed  DOI: 10.1038/sj.bjp.0706093

10. Steffens M, Zentner J, Honegger J, Feuerstein TJ.  Binding affinity and agonist activity of putative endogenous cannabinoids at the human neocortical CB₁ receptor. Biochem Pharmacol. 2005;69:169-178.
PubMed  DOI: 10.1016/j.bcp.2004.08.033

11. Cabral GA, Staab A. Effects on the immune system. In: Pertwee RG,  ed. Cannabinoids, Handbook of Experimental Pharmacolology. Heidelberg, Germany: Springer-Verlag; 2005;168:385-423. 

12. Szabo B, Schlicker E. Effects of cannabinoids on neurotransmission. In: Pertwee RG,  ed. Cannabinoids, Handbook of Experimental Pharmacolology. Heidelberg, Germany: Springer-Verlag; 2005;168:327-365. 

13. Vaughan CW, Christie MJ. Retrograde signalling by endocannabinoids. In: Pertwee RG,  ed. Cannabinoids, Handbook of Experimental Pharmacolology. Heidelberg, Germany: Springer-Verlag; 2005;168:367-383. 

14. Price TJ, Patwardhan A, Akopian AN, Hargreaves KM, Flores CM.  Modulation of trigeminal sensory neuron activity by the dual cannabinoid-vanilloid agonists anandamide, N-arachidonoyl-dopamine and arachidonyl-2-chloroethylamide. Br J Pharmacol. 2004;141:1118-1130.
PubMed  DOI: 10.1038/sj.bjp.0705711

15. Baker CL, McDougall JJ.  The cannabinomimetic arachidonyl-2-chloroethylamide (ACEA) acts on capsaicin-sensitive TRPV1 receptors but not cannabinoid receptors in rat joints. Br J Pharmacol. 2004;142:1361-1367.
PubMed  DOI: 10.1038/sj.bjp.0705902

16. Di Marzo V, De Petrocellis L, Bisogno T. The biosynthesis, fate and pharmacological properties of endocannabinoids. In: Pertwee RG,  ed. Cannabinoids, Handbook of Experimental Pharmacolology. Heidelberg, Germany: Springer-Verlag; 2005;168:147-185. 

17. Bisogno T, Melck D, Bobrov MY, et al.  N-acyl-dopamines: novel synthetic CB₁ cannabinoid-receptor ligands and inhibitors of anandamide inactivation with cannabimimetic activity in vitro and in vivo. Biochem J. 2000;351:817-824.
PubMed  DOI: 10.1042/0264-6021:3510817

18. Leggett JD, Aspley S, Beckett SRG, D’Antona AM, Kendall DA, Kendall DA.  Oleamide is a selective endogenous agonist of rat and human CB₁ cannabinoid receptors. Br J Pharmacol. 2004;141:253-262.
PubMed  DOI: 10.1038/sj.bjp.0705607

19. Porter AC, Sauer J-M, Knierman MD, et al.  Characterization of a novel endocannabinoid, virodhamine, with antagonist activity at the CB₁ receptor. J Pharmacol Exp Ther. 2002;301:1020-1024.
PubMed  DOI: 10.1124/jpet.301.3.1020

20. Cravatt BF, Lichtman AH.  The endogenous cannabinoid system and its role in nociceptive behavior. J Neurobiol. 2004;61:149-160.
PubMed  DOI: 10.1002/neu.20080

21. Hillard CJ, Jarrahian A.  Cellular accumulation of anandamide: consensus and controversy. Br J Pharmacol. 2003;140:802-808.
PubMed  DOI: 10.1038/sj.bjp.0705468

22. Ueda N.  Endocannabinoid hydrolases. Prostaglandins Other Lipid Mediat. 2002;68-9:521-534.
DOI: 10.1016/S0090-6980(02)00053-9

23. Ho W-SV, Hillard CJ. Modulators of endocannabinoid enzymic hydrolysis and membrane transport. In: Pertwee RG,  ed. Cannabinoids, Handbook of Experimental Pharmacolology. Heidelberg, Germany: Springer-Verlag; 2005;168:187-207. 

24. Gulyas AI, Cravatt BF, Bracey MH, et al.  Segregation of 2 endocannabinoid-hydrolyzing enzymes into pre- and postsynaptic compartments in the rat hippocampus, cerebellum and amygdala. Eur J Neurosci. 2004;20:441-458.
PubMed  DOI: 10.1111/j.1460-9568.2004.03428.x

25. Ueda N, Yamanaka K, Yamamoto S.  Purification and characterization of an acid amidase selective for N-palmitoylethanolamine, a putative endogenous anti-inflammatory substance. J Biol Chem. 2001;276:35552-35557.
PubMed  DOI: 10.1074/jbc.M106261200

26. Lo Verme J, Gaetani S, Fu J, Oveisi F, Burton K, Piomelli D.  Regulation of food intake by oleoylethanolamide. Cell Mol Life Sci. 2005;62:708-716.
PubMed  DOI: 10.1007/s00018-004-4494-0

27. Fegley D, Gaetani S, Duranti A, et al.  Characterization of the fatty acid amide hydrolase inhibitor cyclohexyl carbamic acid 3′-carbamoyl-biphenyl-3-yl ester (URB597): effects on anandamide and oleoylethanolamide deactivation. J Pharmacol Exp Ther. 2005;313:352-358.
PubMed  DOI: 10.1124/jpet.104.078980

28. Kozak KR, Marnett LJ.  Oxidative metabolism of endocannabinoids. Prostaglandins Leukot Essent Fatty Acids. 2002;66:211-220.
PubMed  DOI: 10.1054/plef.2001.0359

29. Saario SM, Savinainen JR, Laitinen JT, Järvinen T, Niemi R.  Monoglyceride lipase-like enzymatic activity is responsible for hydrolysis of 2-arachidonoylglycerol in rat cerebellar membranes. Biochem Pharmacol. 2004;67:1381-1387.
PubMed  DOI: 10.1016/j.bcp.2003.12.003

30. Dinh TP, Kathuria S, Piomelli D.  RNA interference suggests a primary role for monoacylglycerol lipase in the degradation of the endocannabinoid 2-arachidonoylglycerol. Mol Pharmacol. 2004;66:1260-1264.
PubMed  DOI: 10.1124/mol.104.002071

31. Fezza F, Bisogno T, Minassi A, Appendino G, Mechoulam R, Di Marzo V.  Noladin ether, a putative novel endocannabinoid: inactivation mechanisms and a sensitive method for its quantification in rat tissues. FEBS Lett. 2002;513:294-298.
PubMed  DOI: 10.1016/S0014-5793(02)02341-4

32. De Petrocellis L, Davis JB, Di Marzo V.  Palmitoylethanolamide enhances anandamide stimulation of human vanilloid VR1 receptors. FEBS Lett. 2001;506:253-256.
PubMed  DOI: 10.1016/S0014-5793(01)02934-9

33. Sheskin T, Hanus L, Slager J, Vogel Z, Mechoulam R.  Structural requirements for binding of anandamide-type compounds to the brain cannabinoid receptor. J Med Chem. 1997;40:659-667.
PubMed  DOI: 10.1021/jm960752x

34. Fernando SR, Pertwee RG.  Evidence that methyl arachidonyl fluorophosphonate is an irreversible cannabinoid receptor antagonist. Br J Pharmacol. 1997;121:1716-1720.
PubMed 

35. Savinainen JR, Saario SM, Niemi R, Järvinen T, Laitinen JT.  An optimized approach to study endocannabinoid signaling: evidence against constitutive activity of rat brain adenosine A₁ and cannabinoid CB₁ receptors. Br J Pharmacol. 2003;140:1451-1459.
PubMed  DOI: 10.1038/sj.bjp.0705577

36. Nithipatikom K, Endsley MP, Isbell MA, et al.  2-Arachidonoylglycerol: a novel inhibitor of androgen-independent prostate cancer cell invasion. Cancer Res. 2004;64:8826-8830.
PubMed 

37. Kathuria S, Gaetani S, Fegley D, et al.  Modulation of anxiety through blockade of anandamide hydrolysis. Nat Med. 2003;9:76-81.
PubMed  DOI: 10.1038/nm803

38. Lichtman AH, Leung D, Shelton CC, et al.  Reversible inhibitors of fatty acid amide hydrolase that promote analgesia: evidence for an unprecedented combination of potency and selectivity. J Pharmacol Exp Ther. 2004;311:441-448.
PubMed  DOI: 10.1124/jpet.104.069401

39. Huang SM, Bisogno T, Petros TJ, et al.  Identification of a new class of molecules, the arachidonyl amino acids, and characterization of one member that inhibits pain. J Biol Chem. 2001;276:42639-42644.
PubMed  DOI: 10.1074/jbc.M107351200

40. Bisogno T, Melck D, De Petrocellis L, et al.  Arachidonoylserotonin and other novel inhibitors of fatty acid amide hydrolase. Biochem Biophys Res Commun. 1998;248:515-522.
PubMed  DOI: 10.1006/bbrc.1998.8874

41. Ligresti A, Bisogno T, Matias I, et al.  Possible endocannabinoid control of colorectal cancer growth. Gastroenterology. 2003;125:677-687.
PubMed  DOI: 10.1016/S0016-5085(03)00881-3

42. Jonsson K-O, Vandevoorde S, Lambert DM, Tiger G, Fowler CJ.  Effects of homologues and analogues of palmitoylethanolamide upon the inactivation of the endocannabinoid anandamide. Br J Pharmacol. 2001;133:1263-1275.
PubMed  DOI: 10.1038/sj.bjp.0704199

43. Khanolkar AD, Abadji V, Lin S, et al.  Head group analogs of arachidonylethanolamide, the endogenous cannabinoid ligand. J Med Chem. 1996;39:4515-4519.
PubMed  DOI: 10.1021/jm960152y

44. Beltramo M, Stella N, Calignano A, Lin SY, Makriyannis A, Piomelli D.  Functional role of high-affinity anandamide transport, as revealed by selective inhibition. Science. 1997;277:1094-1097.
PubMed  DOI: 10.1126/science.277.5329.1094

45. Jarrahian A, Manna S, Edgemond WS, Campbell WB, Hillard CJ.  Structure-activity relationships among N-arachidonylethanolamine (anandamide) head group analogues for the anandamide transporter. J Neurochem. 2000;74:2597-2606.
PubMed  DOI: 10.1046/j.1471-4159.2000.0742597.x

46. Zygmunt PM, Chuang H, Movahed P, Julius D, Högestätt ED.  The anandamide transport inhibitor AM404 activates vanilloid receptors. Eur J Pharmacol. 2000;396:39-42.
PubMed  DOI: 10.1016/S0014-2999(00)00207-7

47. De Petrocellis L, Bisogno T, Davis JB, Pertwee RG, Di Marzo V.  Overlap between the ligand recognition properties of the anandamide transporter and the VR1 vanilloid receptor: inhibitors of anandamide uptake with negligible capsaicin-like activity. FEBS Lett. 2000;483:52-56.
PubMed  DOI: 10.1016/S0014-5793(00)02082-2

48. Fowler CJ, Tiger G, Ligresti A, López-Rodríguez ML, Di Marzo V.  Selective inhibition of anandamide cellular uptake versus enzymatic hydrolysis—a difficult issue to handle. Eur J Pharmacol. 2004;492:1-11.
PubMed  DOI: 10.1016/j.ejphar.2004.03.048

49. Ortar G, Ligresti A, De Petrocellis L, Morera E, Di Marzo V.  Novel selective and metabolically stable inhibitors of anandamide cellular uptake. Biochem Pharmacol. 2003;65:1473-1481.
PubMed  DOI: 10.1016/S0006-2952(03)00109-6

50. López-Rodríguez ML, Viso A, Ortega-Gutiérrez S, et al.  Design, synthesis and biological evaluation of new endocannabinoid transporter inhibitors. Eur J Med Chem. 2003;38:403-412.
PubMed  DOI: 10.1016/S0223-5234(03)00045-X

51. Ruiz-Llorente L, Ortega-Gutiérrez S, Viso A, et al.  Characterization of an anandamide degradation system in prostate epithelial PC-3 cells: synthesis of new transporter inhibitors as tools for this study. Br J Pharmacol. 2004;141:457-467.
PubMed  DOI: 10.1038/sj.bjp.0705628

52. Benito C, Núñez E, Tolón RM, et al.  Cannabinoid CB₂ receptors and fatty acid amide hydrolase are selectively overexpressed in neuritic plaque-associated glia in Alzheimer’s disease brains. J Neurosci. 2003;23:11136-11141.
PubMed 

53. Ramírez BG, Blázquez C, Gómez del Pulgar T, Guzmán N, de Ceballos ML.  Prevention of Alzheimer’s disease pathology by cannabinoids: neuroprotection mediated by blockade of microglial activation. J Neurosci. 2005;25:1904-1913.
PubMed  DOI: 10.1523/JNEUROSCI.4540-04.2005

54. Giuffrida A, Leweke FM, Gerth CW, et al.  Cerebrospinal anandamide levels are elevated in acute schizophrenia and are inversely correlated with psychotic symptoms. Neuropsychopharmacology. 2004;29:2108-2114.
PubMed  DOI: 10.1038/sj.npp.1300558

55. Leweke FM, Giuffrida A, Wurster U, Emrich HM, Piomelli D.  Elevated endogenous cannabinoids in schizophrenia. Neuroreport. 1999;10:1665-1669.
PubMed 

56. Dean B, Sundram S, Bradbury R, Scarr E, Copolov D.  Studies on [³H]CP-55940 binding in the human central nervous system: regional specific changes in density of cannabinoid-1 receptors associated with schizophrenia and cannabis use. Neuroscience. 2001;103:9-15.
PubMed  DOI: 10.1016/S0306-4522(00)00552-2

57. Zavitsanou K, Garrick T, Huang XF.  Selective antagonist [³H]SR141716A binding to cannabinoid CB₁ receptors is increased in the anterior cingulate cortex in schizophrenia. Prog Neuropsychopharmacol Biol Psychiatry. 2004;28:355-360.
PubMed  DOI: 10.1016/j.pnpbp.2003.11.005

58. De Marchi N, De Petrocellis L, Orlando P, Daniele F, Fezza F, Di Marzo V.  Endocannabinoid signalling in the blood of patients with schizophrenia. Lipids Health Dis. 2003;2:1-9.
PubMed  DOI: 10.1186/1476-511X-2-5

59. Hungund BL, Vinod KY, Kassir SA, et al.  Upregulation of CB₁ receptors and agonist-stimulated [³⁵S]GTPγS binding in the prefrontal cortex of depressed suicide victims. Mol Psychiatry. 2004;9:184-190.
PubMed  DOI: 10.1038/sj.mp.4001376

60. Vinod KY, Arango V, Xie S, et al.  Elevated levels of endocannabinoids and CB₁ receptor-mediated G-protein signaling in the prefrontal cortex of alcoholic suicide victims. Biol Psychiatry. 2005;57:480-486.
PubMed  DOI: 10.1016/j.biopsych.2004.11.033

61. Lastres-Becker I, Cebeira M, de Ceballos ML, et al.  Increased cannabinoid CB₁ receptor binding and activation of GTP-binding proteins in the basal ganglia of patients with Parkinson’s syndrome and of MPTP-treated marmosets. Eur J Neurosci. 2001;14:1827-1832.
PubMed  DOI: 10.1046/j.0953-816x.2001.01812.x

62. Glass M, Faull R, Dragunow M.  Loss of cannabinoid receptors in the substantia-nigra in Huntington’s disease. Neurosci. 1993;56:523-527.
PubMed  DOI: 10.1016/0306-4522(93)90352-G

63. Richfield EK, Herkenham M.  Selective vulnerability in Huntington’s disease: preferential loss of cannabinoid receptors in lateral globus pallidus. Ann Neurol. 1994;36:577-584.
PubMed  DOI: 10.1002/ana.410360406

64. Schäbitz W-R, Giuffrida A, Berger C, et al.  Release of fatty acid amides in a patient with hemispheric stroke: a microdialysis study. Stroke. 2002;33:2112-2114.
PubMed  DOI: 10.1161/01.STR.0000023491.63693.18

65. Maccarrone M, Valensise H, Bari M, Lazzarin N, Romanini C, Finazzi-Agricò A.  Relation between decreased anandamide hydrolase concentrations in human lymphocytes and miscarriage. Lancet. 2000;355:1326-1329.
PubMed  DOI: 10.1016/S0140-6736(00)02115-2

66. Maccarrone M, Bisogno T, Valensise H, et al.  Low fatty acid amide hydrolase and high anandamide levels are associated with failure to achieve an ongoing pregnancy after IVF and embryo transfer. Mol Hum Reprod. 2002;8:188-195.
PubMed  DOI: 10.1093/molehr/8.2.188

67. Wang Y, Liu Y, Ito Y, et al.  Simultaneous measurement of anandamide and 2- arachidonoylglycerol by polymyxin B-selective adsorption and subsequent high-performance liquid chromatography analysis: increase in endogenous cannabinoids in the sera of patients with endotoxic shock. Anal Biochem. 2001;294:73-82.
PubMed  DOI: 10.1006/abio.2001.5015

68. Bátkai S, Járai Z, Wagner JA, et al.  Endocannabinoids acting at vascular CB₁ receptors mediate the vasodilated state in advanced liver cirrhosis. Nat Med. 2001;7:827-832.
PubMed  DOI: 10.1038/89953

69. Fernández-Rodriguez CM, Romero J, Petros TJ, et al.  Circulating endogenous cannabinoid anandamide and portal, systemic and renal hemodynamics in cirrhosis. Liver Int. 2004;24:477-483.
PubMed  DOI: 10.1111/j.1478-3231.2004.0945.x

70. Julien B, Grenard P, Teixeira-Clerc F, et al.  Antifibrogenic role of the cannabinoid receptor CB₂ in the liver. Gastroenterology. 2005;128:742-755.
PubMed  DOI: 10.1053/j.gastro.2004.12.050

71. Steffens S, Veillard NR, Arnaud C, et al.  Low dose oral cannabinoid therapy reduces progression of atherosclerosis in mice. Nature. 2005;434:782-786.
PubMed  DOI: 10.1038/nature03389

72. Maccarrone M, Attinà M, Cartoni A, Bari M, Finazzi-Agricò A.  Gas chromatography-mass spectrometry analysis of endogenous cannabinoids in healthy and tumoral human brain and human cells in culture. J Neurochem. 2001;76:594-601.
PubMed  DOI: 10.1046/j.1471-4159.2001.00092.x

73. Petersen G, Moesgaard B, Schmid PC, et al.  Endocannabinoid metabolism in human glioblastomas and meningiomas compared to human non-tumour brain tissue. J Neurochem. 2005;93:299-309.
PubMed  DOI: 10.1111/j.1471-4159.2005.03013.x

74. Pagotto U, Marsicano G, Fezza F, et al.  Normal human pituitary gland and pituitary adenomas express cannabinoid receptor type 1 and synthesize endogenous cannabinoids: first evidence for a direct role of cannabinoids on hormone modulation at the human pituitary level. J Clin Endocrinol Metab. 2001;86:2687-2696.
PubMed  DOI: 10.1210/jc.86.6.2687

75. Sarfaraz S, Afaq F, Adhami VM, Mukhtar H.  Cannabinoid receptor as a novel target for the treatment of prostate cancer. Cancer Res. 2005;65:1635-1641.
PubMed 

76. Motobe T, Hashiguchi T, Uchimura T, et al.  Endogenous cannabinoids are candidates for lipid mediators of bone cement implantation syndrome. Shock. 2004;21:8-12.
PubMed  DOI: 10.1097/01.SHK.0000094766.36694.49

77. Baker D, Pryce G, Croxford JL, et al.  Endocannabinoids control spasticity in a multiple sclerosis model. FASEB J. 2001;15:300-302.
PubMed 

78. Berrendero F, Sánchez A, Cabranes A, et al.  Changes in cannabinoid CB₁ receptors in striatal and cortical regions of rats with experimental allergic encephalomyelitis, an animal model of multiple sclerosis. Synapse. 2001;41:195-202.
PubMed  DOI: 10.1002/syn.1075

79. Cheng BC, Xu H, Calbay L, Pertwee RG, Coutts A, Forrester JV.  Specific CB₂ receptor agonist suppresses the murine model of experimental autoimmune uveitis. Invest Ophthalmol Vis Sci. 2004;45:552.
PubMed  DOI: 10.1167/iovs.03-0749

80. Witting A, Weydt P, Hong S, Kliot M, Möller T, Stella N.  Endocannabinoids accumulate in spinal cord of SOD1^G93A transgenic mice. J Neurochem. 2004;89:1555-1557.
PubMed  DOI: 10.1111/j.1471-4159.2004.02544.x

81. Benito C, Kim W-K, Chavarria I, et al.  A glial endogenous cannabinoid system is upregulated in the brains of macaques with simian immunodeficiency virus-induced encephalitis. J Neurosci. 2005;25:2530-2536.
PubMed  DOI: 10.1523/JNEUROSCI.3923-04.2005

82. Adriani W, Caprioli A, Granstrem O, Carli M, Laviola G.  The spontaneously hypertensive-rat as an animal model of ADHD: evidence for impulsive and non-impulsive subpopulations. Neurosci Biobehav Rev. 2003;27:639-651.
PubMed  DOI: 10.1016/j.neubiorev.2003.08.007

83. Di Marzo V, Hill MP, Bisogno T, Crossman AR, Brotchie JM.  Enhanced levels of endogenous cannabinoids in the globus pallidus are associated with a reduction in movement in an animal model of Parkinson’s disease. FASEB J. 2000;14:1432-1438.
PubMed  DOI: 10.1096/fj.14.10.1432

84. Gubellini P, Picconi B, Bari M, et al.  Experimental parkinsonism alters endocannabinoid degradation: implications for striatal glutamatergic transmission. J Neurosci. 2002;22:6900-6907.
PubMed 

85. Maccarrone M, Gubellini P, Bari M, et al.  Levodopa treatment reverses endocannabinoid system abnormalities in experimental parkinsonism. J Neurochem. 2003;85:1018-1025.
PubMed 

86. Fernandez-Espejo E, Caraballo I, Rodriguez de Fonseca F, et al.  Experimental parkinsonism alters anandamide precursor synthesis, and functional deficits are improved by AM404: a modulator of endocannabinoid function. Neuropsychopharmacology. 2004;29:1134-1142.
PubMed  DOI: 10.1038/sj.npp.1300407

87. Ferrer B, Asbrock N, Kathuria S, Piomelli D, Giuffrida A.  Effects of levodopa on endocannabinoid levels in rat basal ganglia: implications for the treatment of levodopa-induced dyskinesias. Eur J Neurosci. 2003;18:1607-1614.
PubMed  DOI: 10.1046/j.1460-9568.2003.02896.x

88. Herkenham M, Lynn AB, de Costa BR, Richfield EK.  Neuronal localization of cannabinoid receptors in the basal ganglia of the rat. Brain Res. 1991;547:267-274.
PubMed  DOI: 10.1016/0006-8993(91)90970-7

89. Romero J, Berrendero F, Pérez-Rosado A, et al.  Unilateral 6-hydroxydopamine lesions of nigrostriatal dopaminergic neurons increased CB₁ receptor mRNA levels in the caudate-putamen. Life Sci. 2000;66:485-494.
PubMed  DOI: 10.1016/S0024-3205(99)00618-9

90. Mailleux P, Vanderhaeghen J-J.  Dopaminergic regulation of cannabinoid receptor mRNA levels in the rat caudate-putamen: an in situ hybridization study. J Neurochem. 1993;61:1705-1712.
PubMed 

91. Zeng B-Y, Dass B, Owen A, et al.  Chronic L-DOPA treatment increases striatal cannabinoid CB₁ receptor mRNA expression in 6-hydroxydopamine-lesioned rats. Neurosci Lett. 1999;276:71-74.
PubMed  DOI: 10.1016/S0304-3940(99)00762-4

92. McCaw EA, Hu H, Gomez GT, Hebb ALO, Kelly MEM, Denovan-Wright EM.  Structure, expression and regulation of the cannabinoid receptor gene (CB1) in Huntington’s disease transgenic mice. Eur J Biochem. 2004;271:4909-4920.
PubMed  DOI: 10.1111/j.1432-1033.2004.04460.x

93. Page KJ, Besret L, Jain M, Monaghan EM, Dunnett SB, Everitt BJ.  Effects of systemic 3-nitropropionic acid-induced lesions of the dorsal striatum on cannabinoid and μ-opioid receptor binding in the basal ganglia. Exp Brain Res. 2000;130:142-150.
PubMed  DOI: 10.1007/s002210050016

94. Lastres-Becker I, Fezza F, Cebeira M, et al.  Changes in endocannabinoid transmission in the basal ganglia in a rat model of Huntington’s disease. Neuroreport. 2001;12:2125-2129.
PubMed  DOI: 10.1097/00001756-200107200-00017

95. Lastres-Becker I, Bizat N, Boyer F, Hantraye P, Fernández-Ruiz J, Brouillet E.  Potential involvement of cannabinoid receptors in 3-nitropropionic acid toxicity in vivo. Neuroreport. 2004;15:2375-2379.
PubMed  DOI: 10.1097/00001756-200410250-00015

96. Lastres-Becker I, Hansen HH, Berrendero F, et al.  Alleviation of motor hyperactivity and neurochemical deficits by endocannabinoid uptake inhibition in a rat model of Huntington’s disease. Synapse. 2002;44:23-35.
PubMed  DOI: 10.1002/syn.10054

97. Lastres-Becker I, Berrendero F, Lucas JJ, et al.  Loss of mRNA levels, binding and activation of GTP-binding proteins for cannabinoid CB₁ receptors in the basal ganglia of a transgenic model of Huntington’s disease. Brain Res. 2002;929:236-242.
PubMed  DOI: 10.1016/S0006-8993(01)03403-5

98. Walker JM, Huang SM, Strangman NM, Tsou K, Sañudo-Peña MC.  Pain modulation by release of the endogenous cannabinoid anandamide. Proc Natl Acad Sci USA. 1999;96:12198-12203.
PubMed  DOI: 10.1073/pnas.96.21.12198

99. Lim G, Sung B, Ji R-R, Mao J.  Upregulation of spinal cannabinoid-1-receptors following nerve injury enhances the effects of Win 55,212-2 on neuropathic pain behaviors in rats. Pain. 2003;105:275-283.
PubMed  DOI: 10.1016/S0304-3959(03)00242-2

100. Zhang J, Hoffert C, Vu HK, Groblewski T, Ahmad S, O’Donnell D.  Induction of CB₂ receptor expression in the rat spinal cord of neuropathic but not inflammatory chronic pain models. Eur J Neurosci. 2003;17:2750-2754.
PubMed  DOI: 10.1046/j.1460-9568.2003.02704.x

101. Dinis P, Charrua A, Avelino A, et al.  Anandamide-evoked activation of vanilloid receptor 1 contributes to the development of bladder hyperreflexia and nociceptive transmission to spinal dorsal horn neurons in cystitis. J Neurosci. 2004;24:11253-11263.
PubMed  DOI: 10.1523/JNEUROSCI.2657-04.2004

102. Di Marzo V, Goparaju SK, Wang L, et al.  Leptin-regulated endocannabinoids are involved in maintaining food intake. Nature. 2001;410:822-825.
PubMed  DOI: 10.1038/35071088

103. Bensaid M, Gary-Bobo M, Esclangon A, et al.  The cannabinoid CB₁ receptor antagonist SR141716 increases Acrp30 mRNA expression in adipose tissue of obese fa/fa rats and in cultured adipocyte cells. Mol Pharmacol. 2003;63:908-914.
PubMed  DOI: 10.1124/mol.63.4.908

104. Maccarrone M, Fride E, Bisogno T, et al.  Up-regulation of the endocannabinoid system in the uterus of leptin knockout (ob/ob) mice and implications for fertility. Mol Hum Reprod. 2005;11:21-28.
PubMed  DOI: 10.1093/molehr/gah130

105. Kirkham TC, Williams CM, Fezza F, Di Marzo V.  Endocannabinoid levels in rat limbic forebrain and hypothalamus in relation to fasting, feeding and satiation: stimulation of eating by 2-arachidonoyl glycerol. Br J Pharmacol. 2002;136:550-557.
PubMed  DOI: 10.1038/sj.bjp.0704767

106. Hill MN, Patel S, Carrier EJ, et al.  Downregulation of endocannabinoid signaling in the hippocampus following chronic unpredictable stress. Neuropsychopharmacology. 2005;30:508-515.
PubMed  DOI: 10.1038/sj.npp.1300601

107. Patel S, Roelke CT, Rademacher DJ, Hillard CJ.  Inhibition of restraint stress-induced neural and behavioural activation by endogenous cannabinoid signalling. Eur J Neurosci. 2005;21:1057-1069.
PubMed  DOI: 10.1111/j.1460-9568.2005.03916.x

108. Marsicano G, Wotjak CT, Azad SC, et al.  The endogenous cannabinoid system controls extinction of aversive memories. Nature. 2002;418:530-534.
PubMed  DOI: 10.1038/nature00839

109. Liu P, Bilkey DK, Darlington CL, Smith PF.  Cannabinoid CB₁ receptor protein expression in the rat hippocampus and entorhinal, perirhinal, postrhinal and temporal cortices: regional variations and age-related changes. Brain Res. 2003;979:235-239.
PubMed  DOI: 10.1016/S0006-8993(03)02872-5

110. Sparling PB, Giuffrida A, Piomelli D, Rosskopf L, Dietrich A.  Exercise activates the endocannabinoid system. Neuroreport. 2003;14:2209-2211.
PubMed  DOI: 10.1097/00001756-200312020-00015

111. Bátkai S, Pacher P, Osei-Hyiaman D, et al.  Endocannabinoids acting at cannabinoid-1 receptors regulate cardiovascular function in hypertension. Circulation. 2004;110:1996-2002.
PubMed  DOI: 10.1161/01.CIR.0000143230.23252.D2

112. Domenicali M, Ros J, Fernández-Varo G, et al.  Increased anandamide induced relaxation in mesenteric arteries of cirrhotic rats: role of cannabinoid and vanilloid receptors. Gut. 2005;54:522-527.
PubMed  DOI: 10.1136/gut.2004.051599

113. Varga K, Wagner JA, Bridgen DT, Kunos G.  Platelet- and macrophage-derived endogenous cannabinoids are involved in endotoxin-induced hypotension. FASEB J. 1998;12:1035-1044.
PubMed 

114. Liu J, Bátkai S, Pacher P, et al.  Lipopolysaccharide induces anandamide synthesis in macrophages via CD14/MAPK/phosphoinositide 3-kinase/NF-κB independently of platelet-activating factor. J Biol Chem. 2003;278:45034-45039.
PubMed  DOI: 10.1074/jbc.M306062200

115. Maccarrone M, De Petrocellis L, Bari M, et al.  Lipopolysaccharide downregulates fatty acid amide hydrolase expression and increases anandamide levels in human peripheral lymphocytes. Arch Biochem Biophys. 2001;393:321-328.
PubMed  DOI: 10.1006/abbi.2001.2500

116. Wagner JA, Hu K, Bauersachs J, et al.  Endogenous cannabinoids mediate hypotension after experimental myocardial infarction. J Am Coll Cardiol. 2001;38:2048-2054.
PubMed  DOI: 10.1016/S0735-1097(01)01671-0

117. Izzo AA, Fezza F, Capasso R, et al.  Cannabinoid CB₁-receptor mediated regulation of gastrointestinal motility in mice in a model of intestinal inflammation. Br J Pharmacol. 2001;134:563-570.
PubMed  DOI: 10.1038/sj.bjp.0704293

118. Massa F, Marsicano G, Hermann H, et al.  The endogenous cannabinoid system protects against colonic inflammation. J Clin Invest. 2004;113:1202-1209.
PubMed  DOI: 10.1172/JCI200419465

119. Izzo AA, Capasso F, Costagliola A, et al.  An endogenous cannabinoid tone attenuates cholera toxin-induced fluid accumulation in mice. Gastroenterology. 2003;125:765-774.
PubMed  DOI: 10.1016/S0016-5085(03)00892-8

120. McVey DC, Schmid PC, Schmid HHO, Vigna SR.  Endocannabinoids induce ileitis in rats via the capsaicin receptor (VR1). J Pharmacol Exp Ther. 2003;304:713-722.
PubMed  DOI: 10.1124/jpet.102.043893

121. Mascolo N, Izzo AA, Ligresti A, et al.  The endocannabinoid system and the molecular basis of paralytic ileus in mice. FASEB J. 2002;16:U131-U51.

122. Jin KL, Mao XO, Goldsmith PC, Greenberg DA.  CB₁ cannabinoid receptor induction in experimental stroke. Ann Neurol. 2000;48:257-261.
PubMed  DOI: 10.1002/1531-8249(200008)48:2<257::AID-ANA18>3.0.CO_2-P

123. Muthian S, Rademacher DJ, Roelke CT, Gross GJ, Hillard CJ.  Anandamide content is increased and CB₁ cannabinoid receptor blockade is protective during transient, focal cerebral ischemia. Neuroscience. 2004;129:743-750.
PubMed  DOI: 10.1016/j.neuroscience.2004.08.044

124. Kurabayashi M, Takeyoshi I, Yoshinari D, Matsumoto K, Maruyama I, Morishita Y.  2-Arachidonoylglycerol increases in ischemia-reperfusion injury of the rat liver. J Invest Surg. 2005;18:25-31.
PubMed 

125. Hansen HH, Schmid PC, Bittigau P, et al.  Anandamide, but not 2-arachidonoylglycerol, accumulates during in vivo neurodegeneration. J Neurochem. 2001;78:1415-1427.
PubMed  DOI: 10.1046/j.1471-4159.2001.00542.x

126. Panikashvili D, Simeonidou C, Ben-Shabat S, et al.  An endogenous cannabinoid (2-AG) is neuroprotective after brain injury. Nature. 2001;413:527-531.
PubMed  DOI: 10.1038/35097089

127. Sugiura T, Yoshinaga N, Kondo S, Waku K, Ishima Y.  Generation of 2-arachidonoylglycerol, an endogenous cannabinoid receptor ligand, in picrotoxinin-administered rat brain. Biochem Biophys Res Commun. 2000;271:654-658.
PubMed  DOI: 10.1006/bbrc.2000.2686

128. Marsicano G, Goodenough S, Monory K, et al.  CB₁ cannabinoid receptors and on-demand defense against excitotoxicity. Science. 2003;302:84-88.
PubMed  DOI: 10.1126/science.1088208

129. Chen K, Ratzliff A, Hilgenberg L, et al.  Long-term plasticity of endocannabinoid signaling induced by developmental febrile seizures. Neuron. 2003;39:599-611.
PubMed  DOI: 10.1016/S0896-6273(03)00499-9

130. Maccarrone M, Piccirilli S, Battista N, et al.  Enhanced anandamide degradation is associated with neuronal apoptosis induced by the HIV-1 coat glycoprotein gp120 in the rat neocortex. J Neurochem. 2004;89:1293-1300.
PubMed  DOI: 10.1111/j.1471-4159.2004.02430.x

131. Sánchez C, de Ceballos ML, Gómez del Pulgar T, et al.  Inhibition of glioma growth in vivo by selective activation of the CB₂ cannabinoid receptor. Cancer Res. 2001;61:5784-5789.
PubMed 

132. Burstein SH, Rossetti RG, Yagen B, Zurier RB.  Oxidative metabolism of anandamide. Prostaglandins Other Lipid Mediat. 2000;61:29-41.
PubMed  DOI: 10.1016/S0090-6980(00)00053-8

133. de Lago E, Ligresti A, Ortar G, et al.  In vivo pharmacological actions of 2 novel inhibitors of anandamide cellular uptake. Eur J Pharmacol. 2004;484:249-257.
PubMed  DOI: 10.1016/j.ejphar.2003.11.027

134. Burstein SH, Huang SM, Petros TJ, Rossetti RG, Walker JM, Zurier RB.  Regulation of anandamide tissue levels by N-arachidonylglycine. Biochem Pharmacol. 2002;64:1147-1150.
PubMed  DOI: 10.1016/S0006-2952(02)01301-1

135. Cravatt BF, Demarest K, Patricelli MP, et al.  Supersensitivity to anandamide and enhanced endogenous cannabinoid signaling in mice lacking fatty acid amide hydrolase. Proc Natl Acad Sci USA. 2001;98:9371-9376.
PubMed  DOI: 10.1073/pnas.161191698

136. Lichtman AH, Shelton CC, Advani T, Cravatt BF.  Mice lacking fatty acid amide hydrolase exhibit a cannabinoid receptor-mediated phenotypic hypoalgesia. Pain. 2004;109:319-327.
PubMed  DOI: 10.1016/j.pain.2004.01.022

137. Cravatt BF, Saghatelian A, Hawkins EG, Clement AB, Bracey MH, Lichtman AH.  Functional disassociation of the central and peripheral fatty acid amide signaling systems. Proc Natl Acad Sci USA. 2004;101:10821-10826.
PubMed  DOI: 10.1073/pnas.0401292101

138. Martin BR, Lichtman AH.  Cannabinoid transmission and pain perception. Neurobiol Dis. 1998;5:447-461.
PubMed  DOI: 10.1006/nbdi.1998.0218

139. Pertwee RG.  Cannabinoid receptors and pain. Prog Neurobiol. 2001;63:569-611.
PubMed  DOI: 10.1016/S0301-0082(00)00031-9

140. Walker JM, Hohman AG. Cannabinoid mechanisms of pain suppression. In: Pertwee RG,  ed. Cannabinoids, Handbook of Experimental Pharmacolology. Heidelberg, Germany: Springer-Verlag; 2005;168:509-54. 

141. Herzberg U, Eliav E, Bennett GJ, Kopin IJ.  The analgesic effects of R(+)-WIN 55,212-2 mesylate, a high affinity cannabinoid agonist, in a rat model of neuropathic pain. Neurosci Lett. 1997;221:157-160.
PubMed  DOI: 10.1016/S0304-3940(96)13308-5

142. Edsall SA, Knapp RJ, Vanderah TW, Roeske WR, Consroe P, Yamamura HI.  Antisense oligodeoxynucleotide treatment to the brain cannabinoid receptor inhibits antinociception. Neuroreport. 1996;7:593-596.
PubMed 

143. Richardson JD, Aanonsen L, Hargreaves KM.  Hypoactivity of the spinal cannabinoid system results in NMDA-dependent hyperalgesia. J Neurosci. 1998;18:451-457.
PubMed 

144. Dogrul A, Gardell LR, Ma S, Ossipov MH, Porreca F, Lai J.  ‘Knock-down’ of spinal CB₁ receptors produces abnormal pain and elevates spinal dynorphin content in mice. Pain. 2002;100:203-209.
PubMed  DOI: 10.1016/S0304-3959(02)00302-0

145. Valverde O, Ledent C, Beslot F, Parmentier M, Roques BP.  Reduction of stress-induced analgesia but not of exogenous opioid effects in mice lacking CB₁ receptors. Eur J Neurosci. 2000;12:533-539.
PubMed  DOI: 10.1046/j.1460-9568.2000.00929.x

146. Ibrahim MM, Deng H, Zvonok A, et al.  Activation of CB₂ cannabinoid receptors by AM1241 inhibits experimental neuropathic pain: pain inhibition by receptors not present in the CNS. Proc Natl Acad Sci USA. 2003;100:10529-10533.
PubMed  DOI: 10.1073/pnas.1834309100

147. Ledent C, Valverde O, Cossu G, et al.  Unresponsiveness to cannabinoids and reduced addictive effects of opiates in CB₁ receptor knockout mice. Science. 1999;283:401-404.
PubMed  DOI: 10.1126/science.283.5400.401

148. Zimmer A, Zimmer AM, Hohmann AG, Herkenham M, Bonner TI.  Increased mortality, hypoactivity, and hypoalgesia in cannabinoid CB₁ receptor knockout mice. Proc Natl Acad Sci USA. 1999;96:5780-5785.
PubMed  DOI: 10.1073/pnas.96.10.5780

149. Hanus L, Breuer A, Tchilibon S, et al.  HU-308: a specific agonist for CB₂, a peripheral cannabinoid receptor. Proc Natl Acad Sci USA. 1999;96:14228-14233.
PubMed  DOI: 10.1073/pnas.96.25.14228

150. Clayton N, Marshall FH, Bountra C, O’Shaughnessy CT.  CB₁ and CB₂ cannabinoid receptors are implicated in inflammatory pain. Pain. 2002;96:253-260.
PubMed  DOI: 10.1016/S0304-3959(01)00454-7

151. Valenzano KJ, Tafesse L, Lee G, et al.  Pharmacological and pharmacokinetic characterization of the cannabinoid receptor 2 agonist, GW405833, utilizing rodent models of acute and chronic pain, anxiety, ataxia and catalepsy. Neuropharmacology. 2005;48:658-672.
PubMed  DOI: 10.1016/j.neuropharm.2004.12.008

152. Malan TP, Ibrahim MM, Deng H, et al.  CB₂ cannabinoid receptor-mediated peripheral antinociception. Pain. 2001;93:239-245.
PubMed  DOI: 10.1016/S0304-3959(01)00321-9

153. Nackley AG, Makriyannis A, Hohmann AG.  Selective activation of cannabinoid CB₂ receptors suppresses spinal Fos protein expression and pain behavior in a rat model of inflammation. Neuroscience. 2003;119:747-757.
PubMed  DOI: 10.1016/S0306-4522(03)00126-X

154. Quartilho A, Mata HP, Ibrahim MM, et al.  Inhibition of inflammatory hyperalgesia by activation of peripheral CB₂ cannabinoid receptors. Anesthesiology. 2003;99:955-960.
PubMed  DOI: 10.1097/00000542-200310000-00031

155. Hohmann AG, Farthing JN, Zvonok AM, Makriyannis A.  Selective activation of cannabinoid CB₂ receptors suppresses hyperalgesia evoked by intradermal capsaicin. J Pharmacol Exp Ther. 2004;308:446-453.
PubMed  DOI: 10.1124/jpet.103.060079

156. Ibrahim MM, Porreca F, Lai J, et al.  CB₂ cannabinoid receptor activation produces antinociception by stimulating peripheral release of endogenous opioids. Proc Natl Acad Sci USA. 2005;102:3093-3098.
PubMed  DOI: 10.1073/pnas.0409888102

157. Ross RA, Coutts AA, McFarlane SM, et al.  Actions of cannabinoid receptor ligands on rat cultured sensory neurones: implications for antinociception. Neuropharmacology. 2001;40:221-232.
PubMed  DOI: 10.1016/S0028-3908(00)00135-0

158. Holt S, Comelli F, Costa B, Fowler CJ.  Inhibitors of fatty acid amide hydrolase reduce carrageenan induced hind paw inflammation in pentobarbital-treated mice: comparison with indomethacin and possible involvement of cannabinoid receptors. Br J Pharmacol. 2005;146:467-476.
PubMed  DOI: 10.1038/sj.bjp.0706348

159. Calignano A, La Rana G, Giuffrida A, Piomelli D.  Control of pain initiation by endogenous cannabinoids. Nature. 1998;394:277-281.
PubMed  DOI: 10.1038/28393

160. Beaulieu P, Bisogno T, Punwar S, et al.  Role of the endogenous cannabinoid system in the formalin test of persistent pain in the rat. Eur J Pharmacol. 2000;396:85-92.
PubMed  DOI: 10.1016/S0014-2999(00)00226-0

161. Notcutt W, Price M, Miller R, et al.  Initial experiences with medicinal extracts of cannabis for chronic pain: results from 34 ‘N of 1’ studies. Anaesthesia. 2004;59:440-452.
PubMed  DOI: 10.1111/j.1365-2044.2004.03674.x

162. Noyes R, Brunk SF, Baram DA, Canter A.  Analgesic effect of delta-9-tetrahydrocannabinol. J Clin Pharmacol. 1975;15:139-143.
PubMed 

163. Noyes R, Brunk SF, Avery DH, Canter A.  Analgesic properties of delta-9-tetrahydrocannabinol and codeine. Clin Pharmacol Ther. 1975;18:84-89.
PubMed 

164. Pertwee RG.  Cannabinoids and multiple sclerosis. Pharmacol Ther. 2002;95:165-174.
PubMed  DOI: 10.1016/S0163-7258(02)00255-3

165. Wade DT, Robson P, House H, Makela P, Aram J.  A preliminary controlled study to determine whether whole-plant cannabis extracts can improve intractable neurogenic symptoms. Clin Rehabil. 2003;17:21-29.
PubMed  DOI: 10.1191/0269215503cr581oa

166. Zajicek J, Fox P, Sanders H, et al.  Cannabinoids for treatment of spasticity and other symptoms related to multiple sclerosis (CAMS study): multicentre randomised placebo-controlled trial. Lancet. 2003;362:1517-1526.
PubMed  DOI: 10.1016/S0140-6736(03)14738-1

167. Gauter B, Rukwied R, Konrad C.  Cannabinoid agonists in the treatment of blepharospasm - a case report study. Neuroendocrinol Lett. 2004;25:45-48.
PubMed 

168. Brady CM, DasGupta R, Dalton C, Wiseman OJ, Berkley KJ, Fowler CJ.  An open-label pilot study of cannabis-based extracts for bladder dysfunction in advanced multiple sclerosis. Mult Scler. 2004;10:425-433.
PubMed  DOI: 10.1191/1352458504ms1063oa

169. Svendsen KB, Jensen TS, Bach FW.  Does the cannabinoid dronabinol reduce central pain in multiple sclerosis? Randomized double blind placebo controlled crossover trial. BMJ. 2004;329:253-257.
PubMed  DOI: 10.1136/bmj.38149.566979.AE

170. Iwamura H, Suzuki H, Ueda Y, Kaya T, Inaba T.  In vitro and in vivo pharmacological characterization of JTE-907, a novel selective ligand for cannabinoid CB₂ receptor. J Pharmacol Exp Ther. 2001;296:420-425.
PubMed 

171. Lavey BJ, Kozlowski JA, Hipkin RW, et al.  Triaryl bis-sulfones as a new class of cannabinoid CB₂ receptor inhibitors: identification of a lead and initial SAR studies. Bioorg Med Chem Lett. 2005;15:783-786.
PubMed  DOI: 10.1016/j.bmcl.2004.11.007

172. Lunn CA. Immune modulation by cannabinoids: targetting the cannabinoid CB₂ receptor. Symposium on cannabinoid receptors: pharmacological opportunities for obesity and inflammation: New York Academy of Sciences,  2004; Available at: http://www.nyas.org/ebriefreps/main.asp?intSubsectionID=1406. 

173. Costa B, Colleoni M, Trovato AE, Comelli F, Franke C, Giognoni G. Anandamide transport and hydrolysis as targets for the treatment of neuropathic pain: Effects of AM404 and URB597 in rats with chronic constriction injury of the sciatic nerve. 2005 Symposium on the Cannabinoids; June 24-27; Clearwater Beach, FL. Burlington, VT: International Cannabinoid Research Society.; 2005: 33. 

174. Mestre L, Correa F, Arévalo-Martin A, et al.  Pharmacological modulation of the endocannabinoid system in a viral model of multiple sclerosis. J Neurochem. 2005;92:1327-1339.
PubMed  DOI: 10.1111/j.1471-4159.2004.02979.x

175. Pryce G, Ahmed Z, Hankey DJR, et al.  Cannabinoids inhibit neurodegeneration in models of multiple sclerosis. Brain. 2003;126:2191-2202.
PubMed  DOI: 10.1093/brain/awg224

176. Baker D, Pryce G, Croxford JL, et al.  Cannabinoids control spasticity and tremor in a multiple sclerosis model. Nature. 2000;404:84-87.
PubMed  DOI: 10.1038/35003583

177. Brooks JW, Pryce G, Bisogno T, et al.  Arvanil-induced inhibition of spasticity and persistent pain: evidence for therapeutic sites of action different from the vanilloid VR1 receptor and cannabinoid CB₁/CB₂ receptors. Eur J Pharmacol. 2002;439:83-92.
PubMed  DOI: 10.1016/S0014-2999(02)01369-9

178. Wilkinson JD, Whalley BJ, Baker D, et al.  Medicinal cannabis: is Δ⁹-tetrahydrocannabinol necessary for all its effects? J Pharm Pharmacol. 2003;55:1687-1694.
PubMed  DOI: 10.1211/0022357022304

179. Lyman WD, Sonett JR, Brosnan CF, Elkin R, Bornstein MB.  Δ⁹-tetrahydrocannabinol: a novel treatment for experimental autoimmune encephalomyelitis. J Neuroimmunol. 1989;23:73-81.
PubMed  DOI: 10.1016/0165-5728(89)90075-1

180. Wirguin I, Mechoulam R, Breuer A, Schezen E, Weidenfeld J, Brenner T.  Suppression of experimental autoimmune encephalomyelitis by cannabinoids. Immunopharmacology. 1994;28:209-214.
PubMed  DOI: 10.1016/0162-3109(94)90056-6

181. Croxford JL, Miller SD.  Immunoregulation of a viral model of multiple sclerosis using the synthetic cannabinoid R(+)WIN55,212. J Clin Invest. 2003;111:1231-1240.
PubMed  DOI: 10.1172/JCI200317652

182. Arévalo-Martin N, Vela JM, Molina-Holgado E, Borrell J, Guaza C.  Therapeutic action of cannabinoids in a murine model of multiple sclerosis. J Neurosci. 2003;23:2511-2516.
PubMed 

183. Vaney C, Heinzel-Gutenbrunner M, Jobin P, et al.  Efficacy, safety and tolerability of an orally administered cannabis extract in the treatment of spasticity in patients with multiple sclerosis: a randomized, double-blind, placebo-controlled, crossover study. Mult Scler. 2004;10:417-424.
PubMed  DOI: 10.1191/1352458504ms1048oa

184. Wade DT, Makela P, Robson P, House H, Bateman C.  Do cannabis-based medicinal extracts have general or specific effects on symptoms in multiple sclerosis? A double-blind, randomized, placebo-controlled study on 160 patients. Mult Scler. 2004;10:434-441.
PubMed  DOI: 10.1191/1352458504ms1082oa

185. Killestein J, Hoogervorst ELJ, Reif M, et al.  Safety, tolerability, and efficacy of orally administered cannabinoids in MS. Neurology. 2002;58:1404-1407.
PubMed 

186. Bifulco M, Laezza C, Valenti M, Ligresti A, Portella G, Di Marzo V.  A new strategy to block tumor growth by inhibiting endocannabinoid inactivation. FASEB J. 2004;18:U320-U33.

187. Bifulco M, Laezza C, Portella G, et al.  Control by the endogenous cannabinoid system of ras oncogene-dependent tumor growth. FASEB J. 2001;15:U100-U16.

188. Bifulco M, Di Marzo V.  Targeting the endocannabinoid system in cancer therapy: a call for further research. Nat Med. 2002;8:547-550.
PubMed  DOI: 10.1038/nm0602-547

189. Portella G, Laezza C, Laccetti P, De Petrocellis L, Di Marzo V, Bifulco M.  Inhibitory effects of cannabinoid CB₁ receptor stimulation on tumor growth and metastatic spreading: actions on signals involved in angiogenesis and metastasis. FASEB J. 2003;17:U458-U74.

190. Guzmán M.  Cannabinoids: potential anticancer agents. Nat Rev Cancer. 2003;3:745-755.
PubMed  DOI: 10.1038/nrc1188

191. Guzmán M. Effects on cell viability In: Pertwee RG,  ed. Cannabinoids, Handbook of Experimental Pharmacolology. Heidelberg, Germany: Springer-Verlag; 2005;168:627-642. 

192. Izzo AA, Coutts AA. Cannabinoids and the digestive tract. In: Pertwee RG,  ed. Cannabinoids, Handbook of Experimental Pharmacolology. Heidelberg, Germany: Springer-Verlag; 2005;168:573-598. 

193. Panikashvili D, Mechoulam R, Beni SM, Alexandrovich A, Shohami E.  CB₁ cannabinoid receptors are involved in neuroprotection via NF-κB inhibition. J Cereb Blood Flow Metab. 2005;25:477-484.
PubMed  DOI: 10.1038/sj.jcbfm.9600047

194. El Banoua F, Caraballo I, Flores JA, Galan-Rodriguez B, Fernandez-Espejo E.  Effects on turning of microinjections into basal ganglia of D₁ and D₂ dopamine receptors agonists and the cannabinoid CB₁ antagonist SR141716A in a rat Parkinson’s model. Neurobiol Dis. 2004;16:377-385.
PubMed  DOI: 10.1016/j.nbd.2004.03.002

195. Fernandez-Espejo E, Caraballo I, Rodriguez de Fonseca F, et al.  Cannabinoid CB₁ antagonists possess antiparkinsonian efficacy only in rats with very severe nigral lesion in experimental parkinsonism. Neurobiol Dis. 2005;18:591-601.
PubMed  DOI: 10.1016/j.nbd.2004.10.015

196. Meschler JP, Howlett AC, Madras BK.  Cannabinoid receptor agonist and antagonist effects on motor function in normal and 1-methyl-4-phenyl-1,2,5,6- tetrahydropyridine (MPTP)-treated non-human primates. Psychopharmacology (Berl). 2001;156:79-85.
PubMed  DOI: 10.1007/s002130100728

197. Wagner JA, Varga K, Ellis EF, Rzigalinski BA, Martin BR, Kunos G.  Activation of peripheral CB₁ cannabinoid receptors in haemorrhagic shock. Nature. 1997;390:518-521.
PubMed  DOI: 10.1038/37371

198. Kadoi Y, Hinohara H, Kunimoto F, Kuwano H, Saito S, Goto F.  Effects of AM281, a cannabinoid antagonist, on systemic haemodynamics, internal carotid artery blood flow and mortality in septic shock in rats. Br J Anaesth. 2005;94:563-568.
PubMed  DOI: 10.1093/bja/aei106

199. Robson PJ. Human studies of cannabinoids and medicinal cannabis. In: Pertwee RG,  ed. Cannabinoids, Handbook Exp Pharmacol. Heidelberg, Germany: Springer-Verlag; 2005:719-756. 

200. Robson PJ, Guy GW. Clinical studies of cannabis-based medicines. In: Guy GW, Whittle BA, Robson PJ,  eds. The Medicinal Uses of Cannabis and Cannabinoids. London, UK: Pharmaceutical Press; 2005:229-269. 

201. Consroe P, Musty R, Rein J, Tillery W, Pertwee R.  The perceived effects of smoked cannabis on patients with multiple sclerosis. Eur Neurol. 1997;38:44-48.
PubMed 

202. Ware MA, Adams H, Guy GW.  The medicinal use of cannabis in the UK: results of a nationwide survey. Int J Clin Pract. 2005;59:291-295.
PubMed  DOI: 10.1111/j.1742-1241.2004.00271.x

203. De Vry J, Denzer D, Reissmueller E, et al.  3-[2-cyano-3-(trifluoromethyl)phenoxy]phenyl-4,4,4-trifluoro-1-butanesulfonate (BAY 59-3074): a novel cannabinoid CB₁ /CB₂ receptor partial agonist with antihyperalgesic and antiallodynic effects. J Pharmacol Exp Ther. 2004;310:620-632.
PubMed  DOI: 10.1124/jpet.103.062836

204. Rukwied R, Watkinson A, McGlone F, Dvorak M.  Cannabinoid agonists attenuate capsaicin-induced responses in human skin. Pain. 2003;102:283-288.
PubMed  DOI: 10.1016/S0304-3959(02)00401-3

205. Dvorak M, Watkinson A, McGlone F, Rukwied R.  Histamine-induced responses are attenuated by a cannabinoid receptor agonist in human skin. Inflamm Res. 2003;52:238-245.
PubMed 

206. Cichewicz DL.  Synergistic interactions between cannabinoid and opioid analgesics. Life Sci. 2004;74:1317-1324.
PubMed  DOI: 10.1016/j.lfs.2003.09.038

207. Holdcroft A, Smith M, Jacklin A, et al.  Pain relief with oral cannabinoids in familial Mediterranean fever. Anaesthesia. 1997;52:483-488.
PubMed  DOI: 10.1111/j.1365-2044.1997.139-az0132.x

208. Pertwee RG. The pharmacology and therapeutic potential of cannabidiol. In: Di Marzo V,  ed. Cannabinoids. New York, NY: Kluwer Academic/Plenum Publishers; 2004:32-83. 

209. Mendelson WB, Basile AS.  The hypnotic actions of the fatty acid amide, oleamide. Neuropsychopharmacology. 2001;25:S36-S39.
PubMed  DOI: 10.1016/S0893-133X(01)00341-4

210. Pertwee RG, Ross RA.  Cannabinoid receptors and their ligands. Prostaglandins Leukot Essent Fatty Acids. 2002;66:101-121.
PubMed  DOI: 10.1054/plef.2001.0341

211. Sim-Selley LJ.  Regulation of cannabinoid CB₁ receptors in the central nervous system by chronic cannabinoids. Crit Rev Neurobiol. 2003;15:91-119.
PubMed  DOI: 10.1615/CritRevNeurobiol.v15.i2.10

212. Lichtman AH, Martin BR. Cannabinoid tolerance and dependence. In: Pertwee RG,  ed. Cannabinoids, Handbook of Experimental Pharmacolology. Heidelberg, Germany: Springer-Verlag; 2005;168:691-717. 

213. De Vry J, Jentzsch KR, Kuhl E, Eckel G.  Behavioral effects of cannabinoids show differential sensitivity to cannabinoid receptor blockade and tolerance development. Behav Pharmacol. 2004;15:1-12.
PubMed  DOI: 10.1097/00008877-200402000-00001

214. Paria BC, Dey SK.  Ligand-receptor signaling with endocannabinoids in preimplantation embryo development and implantation. Chem Phys Lipids. 2000;108:211-220.
PubMed  DOI: 10.1016/S0009-3084(00)00197-3

215. Van Gaal LF, Rissanen AM, Scheen AJ, Ziegler O, Rossner S.  Effects of the cannabinoid-1 receptor blocker rimonabant on weight reduction and cardiovascular risk factors in overweight patients: 1-year experience from the RIO-Europe study. Lancet. 2005;365:1389-1397.
PubMed  DOI: 10.1016/S0140-6736(05)66374-X

216. Idris AI.  van’t Hof RJ, Greig IR, Ridge SA, Baker D, Ross RA, Ralston SH. Regulation of bone mass, bone loss and osteoclast activity by cannabinoid receptors. Nat Med. 2005;11:774-779.
PubMed  DOI: 10.1038/nm1255

217. Brown AJ, Ueno S, Suen K, Dowell SJ, Wise A. Molecular identification of GPR55 as a third G protein-coupled receptor responsive to cannabinoid ligands. Symposium on the Cannabinoids; June 24-27; Clearwater Beach, FL. Burlington, VT: International Cannabinoid Research Society; 2005: 16. 

218. Sjögren S, Ryberg E, Lindblom A, et al. A new receptor for cannabinoid ligands Symposium on the Cannabinoids; June 24-27; Clearwater Beach, FL. Burlington, VT: International Cannabinoid Research Society; 2005: 106. 

219. Hohmann AG, Suplita RL, Bolton NM, et al.  An endocannabinoid mechanism for stress-induced analgesia. Nature. 2005;435:1108-1112.
PubMed  DOI: 10.1038/nature03658