Cannabinoid Receptor Ligands

Professor R. G. Pertwee
Department of Biomedical Sciences, Institute of Medical Sciences, University of Aberdeen, Foresterhill, Aberdeen AB25 2ZD, Scotland

Roger Pertwee is Professor of Neuropharmacology at the University of Aberdeen. His research interests include the pharmacology of cannabinoid receptors, the physiological and pathophysiological roles of endogenous cannabinoid receptor ligands, and the therapeutic potential of cannabinoids.

CB₁ mRNA has been detected both in the central nervous system and in certain peripheral tissues including pituitary gland, immune cells, reproductive tissues, gastrointestinal tissues, superior cervical ganglion, heart, lung, urinary bladder and adrenal gland.¹² Some CB₁ receptors are located at central and peripheral nerve terminals^12,13,14 and, when activated, these receptors seem to suppress the neuronal release of one or other of a range of excitatory and inhibitory transmitters that include acetylcholine, noradrenaline, dopamine, 5­hydroxy-tryptamine, g-aminobutyric acid, glutamate and aspartate.¹⁴ CB₂ mRNA is present mainly in immune cells with particularly high levels in B-cells and natural killer cells.¹⁵ Little is yet known in any detail about the physiological or pathophysiological roles of CB₂ receptors. Most likely, these include immunomodulation which may depend at least in part on CB₂ receptor-mediated suppression of proinflammatory cytokine release and enhancement of antiinflammatory cytokine release from immune cells.^16,17 Thus one major role that CB₁ and CB₂ receptors may have in common is modulation of ongoing release of chemical messengers.

The central distribution pattern of CB₁ receptors is heterogeneous and accounts for several prominent pharmacological properties of CB₁ receptor agonists, for example their ability to impair cognition and memory and to alter the control of motor function. Thus the cerebral cortex, hippocampus, lateral caudate-putamen, substantia nigra pars reticulata, globus pallidus, entopeduncular nucleus and the molecular layer of the cerebellum are all populated with particularly high concentrations of CB₁ receptors.^12,18 In line with the analgesic properties of cannabinoid receptor agonists, CB₁ receptors are also found on pain pathways in brain and spinal cord and at the peripheral terminals of primary sensory neurones.¹⁴ Although the concentration of CB₁ receptors is considerably less in peripheral tissues than in the central nervous system, this does not imply that peripheral CB₁ receptors are unimportant. This is because some peripheral tissues may contain high concentrations of CB₁ receptors, localized in discrete regions such as nerve terminals that form only a small part of the total tissue mass. Peripheral tissues in which CB₁ receptors are expressed on neurones include the heart, vas deferens, urinary bladder and small intestine.¹²

As detailed elsewhere,^12,19 both CB₁ and CB₂ receptors are coupled through G_i/o proteins, negatively to adenylate cyclase and positively to mitogen-activated protein kinase. In addition, CB₁ receptors are coupled to ion channels through G_i/o proteins, positively to A-type and inwardly rectifying potassium channels and negatively to N-type and P/Q-type calcium channels and to D-type potassium channels.^12,19,20 The coupling to A-type and D-type potassium channels is thought to be through adenylate cyclase.^20,21 Inwardly rectifying potassium channels can also serve as a signalling mechanism for the CB₂ receptor, at least in Xenopus oocytes that have been transfected with such channels together with this receptor type.^22,23 In addition, there is evidence from experiments with rat hippocampal CA1 pyramidal neurones that CB₁ receptors are negatively coupled to M-type potassium channels.²⁴ CB₁ receptors may also mobilize arachidonic acid and close 5-HT₃ receptor ion channels¹² and, under certain conditions, couple to G_s proteins to activate adenylate cyclase^25,26 and/or to reduce outward potassium K current, possibly through arachidonic acid-mediated stimulation of protein kinase C.²⁷ The questions of whether CB₁ receptor coupling to G_s proteins has physiological importance and of whether such coupling increases after G_i/o protein sequestration by co-localized non-cannabinoid G_i/o protein-coupled receptors have yet to be resolved. CB₁ receptors have also been reported to be positively coupled to phospholipase C through G proteins in COS7 cells co-transfected with CB₁ receptors and G_a subunits²² and negatively coupled to voltage-gated L-type calcium channels in cat cerebral arterial smooth muscle cells.²⁸ One other recent finding is that CB₁ receptors on cultured cerebellar granule neurones can operate through a phospholipase C-sensitive mechanism to enhance NMDA-elicited calcium release from inositol 1,4,5-triphosphate-gated intracellular stores.²⁹

The discovery of cannabinoid receptors was followed in 1992 by the demonstration that arachidonoyl ethanolamide (anandamide) is an endogenous ligand for these receptors.³⁰ Other endogenous cannabinoids (endocannabinoids) have since been identified, of which the most important is 2-arachidonoyl glycerol.^31-33 Both anandamide and 2-arachidonoyl glycerol undergo depolarization-induced synthesis/release from neurones and are removed from the extracellular space by a carrier-mediated, saturable uptake process that is present in the membranes of neurones and astrocytes.^31,34-37 There is also evidence that anandamide release in rat dorsal striatum can be triggered by the activation of dopamine D₂, D₃ and/or D₄ receptors.³⁸ Once within the cell, anandamide is hydrolysed to arachidonic acid and ethanolamine by fatty acid amide hydrolase (FAAH).^31,36,39 FAAH can also catalyse the hydrolysis of 2-arachidonoyl glycerol,^31,40 an indication that it has esterase as well as amidase activity. The distribution in rat brain of FAAH immunoreactivity is heterogeneous, as is the brain distribution of FAAH mRNA and FAAH enzymatic activity. In line with the putative roles of anandamide and 2-arachidonoyl glycerol as endogenous cannabinoids, these measures of FAAH brain distribution exhibit considerable though not complete overlap with the brain distribution of CB₁ receptors.^34,39,41-43 Unlike the endocannabinoid membrane transporter, which remains to be fully characterized, FAAH has been cloned⁴⁴ and FAAH knockout mice developed (personal communication from Dr Benjamin Cravatt). At least one other anandamide-hydrolysing enzyme also exists, probably in lysozomes.⁴³ Cannabinoid receptors and their endogenous ligands together constitute what is now usually referred to as the ‘endocannabinoid system’. These receptors, together with FAAH and the endocannabinoid membrane transporter, constitute important new molecular targets and the remainder of this review focuses on ligands that have been developed to interact selectively with these targets.

CB₁ and CB₂ receptor agonists

Cannabinoid receptor agonists fall essentially into four chemical groups: classical, nonclassical, aminoalkylindole and eicosanoid,⁴⁵ (Table 1). Classical cannabinoids consist of dibenzopyran derivatives and are either plant-derived cannabinoids or their synthetic analogues. The most investigated of these include the psychotropic plant cannabinoids, D⁹-tetrahydrocannabinol (D⁹-THC) and D⁸-THC, and the synthetic cannabinoid, 11-hydroxy-D⁸-THC-dimethylheptyl (HU-210). The nonclassical cannabinoids were developed by a Pfizer research team.⁴⁸ They are quite similar in structure to classical cannabinoids, consisting as they do of bicyclic and tricyclic analogues of D⁹-THC that lack a pyran ring. Important examples are CP 55940, CP 55244, CP 50556 (L-nantradol) and desacetyl-L-nantradol. The aminoalkylindole group was developed by a Sterling Winthrop research team,^49,50 the prototype of this group being WIN 55,212-2 (R-(+)-WIN55212). This group contains compounds that are structurally quite different from classical or nonclassical cannabinoids and, indeed, there is evidence that WIN 55,212-2 binds differently to the CB₁ receptor than classical and nonclassical cannabinoids, albeit it in a manner that still permits mutual displacement between WIN 55,212-2 and non-aminoalkylindole cannabinoids at CB₁ and CB₂ binding sites.⁴⁵ The prototypic member of the eicosanoid group of cannabinoid receptor agonists is the endocannabinoid, anandamide (see above). Cannabinoid receptor agonists often contain chiral centres and these generally confer marked stereoselectivity in pharmacological assays. WIN 55,212-2 is more active than WIN 55,212-3 (S-(-)-WIN55212) and classical and nonclassical cannabinoids with the same absolute stereochemistry as (-)-D⁹-THC at 6a and 10a (6aR, 10aR ) have the greater activity. Anandamide itself does not contain any chiral centres. However, some of its synthetic analogues do, one example being methanandamide, the R-(+)-isomer of which has nine times greater affinity for CB₁ receptors than the S-(-)-isomer.⁵¹

Table 1. Pharmacological properties of some important cannabinoid receptor agonists

2-AG, 2-arachidonoyl glycerol. See text for full names of other agonists. *PMSF present. ** Inhibitors of the enzymic hydrolysis of 2-AG present.⁴⁶ Binding assays were performed with [³H]CP 55940 or [³H]HU-243 (2-AG) using rat brain^† and rat or mouse spleen^†† membranes or membranes from cultured cells transfected with CB₁ or CB₂ receptors. For further information see text and references 45 and 47.

The ability of cannabinoid receptor agonists that are commonly used in the laboratory to interact with cannabinoid receptors has been reviewed elsewhere.^45,52,53 Essentially, (-)-D⁹-THC binds equally well to CB₁ and CB₂ receptors and is a partial agonist at both these receptor types. It has even less efficacy at CB₂ than at CB₁ receptors and, indeed, has been reported in one CB₂ bioassay system to behave as an antagonist.⁵⁴ (-)-D⁸-THC resembles (-)-D⁹-THC both in its affinities for CB₁ and CB₂ receptors and in its CB₁ receptor efficacy. CP 55940 and WIN 55,212-2 have CB₁ and CB₂ affinities in the low nanomolar range and exhibit relatively high efficacy at both these receptor types. Results from binding experiments with CB₁- and CB₂-transfected cells indicate that CP 55940 has essentially the same affinity for CB₁ and CB₂ receptors. WIN 55,212-2 has slightly greater affinity for CB₂ than for CB₁ receptors whilst anandamide binds marginally more readily to CB₁ than to CB₂ receptors and, when protected from enzymic hydrolysis, exhibits a CB₁ affinity similar to that of (-)-D⁹-THC. Anandamide also resembles (-)-D⁹-THC in behaving as a partial agonist at CB₁ and CB₂ receptors and in exhibiting lower CB₂ than CB₁ efficacy. One commercially available classical cannabinoid, HU-210, has efficacies at CB₁ and CB₂ receptors that match those of CP 55940 and WIN 55,212-2 and affinities for CB₁ and CB₂ receptors that exceed those of these other cannabinoids. As a result it is a particularly potent cannabinoid receptor agonist. Its pharmacological effects in vivo are also exceptionally long-lasting. The enhanced affinity and efficacy shown by HU-210 at cannabinoid receptors can be largely attributed to the replacement of the pentyl side chain of D⁸-THC with a dimethylheptyl group.

One other classical cannabinoid receptor agonist that merits special mention is 3-(5´-cyano-1´,1´-dimethyl-pentyl)-1-(4-N-morpholinobutyryloxy)-D⁸-THC hydrochloride (O-1057).⁵⁵ This stands out from established cannabinoid receptor agonists in being readily soluble in water. The potency of O-1057 relative to that of CP 55940 is just 2.9 times less at CB₁ receptors and 6.4 times less at CB₂ receptors. The availability of a water-soluble cannabinoid will facilitate cannabinoid delivery not only in the laboratory but also in the clinic, particularly where administration to patients is to be by injection or aerosol inhalation. Another important advance has been the development of cannabinoid receptor agonists that, unlike the agents already mentioned, show marked differences in their abilities to interact with CB₁ and CB₂ receptors.^45,47 For the development of CB₁-selective agonists, the starting point has been the anandamide molecule, the marginal CB₁ selectivity of which can be significantly enhanced by inserting a fluorine atom on the terminal 2´ carbon to form O-585 and/or by replacing a hydrogen atom on the 1´ or 2 carbon with a methyl group to form R-(+)-methanandamide or O-689.⁴⁵ Another important consequence of inserting a methyl group on the 1´ or 2 carbon is greater resistance to the hydrolytic action of FAAH and, indeed, R-(+)-methanandamide was first synthesized in order to meet the need for a metabolically more stable anandamide analogue. The most CB₁-selective agonists so far developed are arachidonyl-2´-chloroethylamide (ACEA) and arachidonylcyclo-propylamide (ACPA), both of which behave as potent CB₁ receptor agonists with reasonably high efficacy.⁵⁶ However, unlike methanandamide and O-689,⁴⁵ neither ACEA nor ACPA show any sign of reduced susceptibility to enzymic hydrolysis. This is presumably because they lack a methyl substituent on the 1´ or 2 carbon and, indeed, it was recently shown that the addition of a methyl group to the 1´ carbon of ACEA does markedly decrease the susceptibility of this molecule to FAAH-mediated hydrolysis.⁵⁷ This structural change also reduces the affinity of ACEA for CB₁ receptors by about 14-fold. As to CB₂-selective agonists, the best to have been developed so far are all classical cannabinoids: L-759633, L-759656, JWH-133 and HU-308. Each of these agents not only binds more readily to CB₂ than to CB₁ receptors but also behaves as a potent CB₂-selective agonist in functional assays.^47,58,59 Conformational requirements for the interaction of endocannabinoids not only with cannabinoid receptors but also with FAAH and with the anandamide membrane transporter (see below) have been reviewed recently by Reggio and Traore.⁵²

Selective CB₁ receptor antagonists/inverse agonists

The first of these, the diarylpyrazole, SR141716A, was developed by Sanofi.^45,60 This is a highly potent and selective CB₁ receptor ligand (Table 2) that readily prevents or reverses CB₁-mediated effects both in vitro and in vivo. There is convincing evidence that SR141716A is not a “silent” antagonist. Thus, as well as attenuating effects of CB₁ receptor agonists, SR141716A can by itself also elicit responses in some CB₁ receptor-containing tissues that are opposite in direction from those elicited by CB₁ receptor agonists. Whilst such “inverse cannabimimetic effects” may in some instances be attributable to a direct antagonism of responses evoked at CB₁ receptors by released endocannabinoids, there is evidence that this is not the only possible mechanism and that SR141716A is in fact an inverse agonist.^63-65 Thus SR141716A may produce inverse cannabimimetic effects in at least some tissues by somehow reducing the constitutive activity of CB₁ receptors (the coupling of CB₁ receptors to their effector mechanisms that it is thought can occur in the absence of exogenously added or endogenously produced CB₁ agonists). Another notable CB₁-selective antagonist that exhibits inverse CB₁ receptor properties in some assay systems is LY320135.⁴⁵ This agent, which was developed by Eli Lilly, shares the ability of SR141716A to bind much more readily to CB₁ than CB₂ receptors (Table 2). However it has less affinity for CB₁ receptors than SR141716A and, at concentrations in the low micromolar range, also binds to muscarinic and 5HT₂ receptors. Although neither SR141716A nor LY320135 are commercially available, it is possible to purchase structural analogues of SR141716A. These are AM 251 and AM 281 which have been found to be respectively three and eight times less potent than SR141716A in displacing [³H]SR141716A from binding sites on mouse cerebellar membranes.⁶⁶ AM 281 has also been reported to bind more readily to CB₁ than CB₂ receptors (Table 2) and to attenuate the ability of WIN 55,212-2 or CP 55940 to decrease rat locomotor activity, to inhibit single population spikes and evoked acetylcholine release in rat hippocampal slices, to inhibit electrically-evoked contractions of the myenteric plexus-longitudinal muscle preparation of guinea-pig ileum and to suppress guinea-pig intestinal peristalsis.^62,67-70 Like SR141716A, AM 281 can behave as an inverse agonist when administered alone.^68-70

Table 2. Pharmacological properties of some important receptor partial agonists/antagonists/inverse agonists

A, surmountable antagonist; iA, ireversible antagonist; I, inverse agonist; P, partial agonist. Main pharmacological classification for each ligand is shown in bold. See text for full names of ligands. *Less inverse agonist efficacy at CB₂ receptors than SR144528. MAFP is also a potent irreversible inhibitor of FAAH (see text). Binding assays were performed with [³H]CP 55940 using rat forebrain^† and mouse spleen^†† membranes or membranes from cultured cells transfected with CB₁ or CB₂ receptors. For further information see text and references 45, 47, 61 and 62.

In some in vitro bioassay systems, for example rat and guinea-pig isolated arteries and mouse isolated vas deferens, SR141716A has been reported to be less potent against anandamide than against non-eicosanoid CB₁ receptor agonists.^71,72 There are also reports that in certain in vivo bioassay systems, SR141716A fails to antagonize effects of anandamide altogether when it is given at doses that markedly antagonize the effects of other CB₁ receptor agonists in the same bioassays.^73-75 For the isolated tissue preparations at least, this discrepancy often seems to stem from the ability of anandamide to activate vanilloid as well as cannabinoid receptors.^71,72,76,77 Analogues of anandamide that share this ability to bind to both cannabinoid and vanilloid receptors include two anandamide membrane transport inhibitors, AM 404 and arvanil.^71,78-80 Of these, arvanil is a hybrid of anandamide and the established vanilloid receptor agonist, capsaicin. As is to be expected, the structural requirements for activation of cannabinoid and vanilloid receptors are not the same.^71,72,76,77 Whether the ability of anandamide to activate vanilloid receptors has physiological as well as pharmacological significance remains to be established.

CB₂ receptor antagonists/inverse agonists

The most notable of these is the Sanofi compound, SR144528, a diarylpyrazole that binds with markedly higher affinity to CB₂ than CB₁ receptors (Table 2).⁸¹ There is good evidence that, like SR141716A, SR144528 is not a “silent” antagonist but rather an inverse agonist that can, by itself, produce inverse cannabimimetic effects at CB₂ receptors.^59,81 Another CB₂-selective antagonist/inverse agonist is 6-iodopravadoline (AM 630; Table 2) which potently reverses CP 55940-induced inhibition of forskolin-stimulated cyclic AMP production by human CB₂-transfected CHO cell preparations (EC₅₀ = 129 nM) and enhances forskolin-stimulated cyclic AMP production by the same cell line when administered by itself (EC₅₀ = 230 nM).⁵⁹ The inverse efficacy of AM 630 at CB₂ receptors appears to be less than that of SR144528.⁸² As to the ability of AM 630 to interact with CB₁ receptors, results from several investigations when taken together suggest that this ligand has mixed agonist-antagonist properties and that it is a low-affinity partial CB₁ agonist.^45,59,83-85 There is also one report that it can behave as a low-potency inverse agonist at CB₁ receptors.⁸⁶

One cannabinoid receptor ligand that is close to being a silent cannabinoid receptor antagonist, albeit at both CB₁ and CB₂ receptors, is the classical cannabinoid, 6´-azidohex-2´-yne-D⁸-THC (O-1184; Table 2). In addition to a terminal N₃ group, the alkyl side chain of this ligand contains a carbon-carbon triple bond, a structural modification that reduces CB₁ and CB₂ efficacy but not CB₁ or CB₂ affinity.⁸² O-1184 behaves as a high-affinity low-efficacy agonist at CB₁ receptors and as a high-affinity low-efficacy inverse agonist at CB₂ receptors.^82,87 O-1238, in which the carbon-carbon triple bond of O-1184 is replaced by a carbon-carbon double bond, has higher efficacy than O-1184 at CB₁ receptors and behaves as a high-affinity low-efficacy partial agonist at CB₂ receptors (Table 2).⁸²

Radiolabelled cannabinoid receptor ligands

Tritiated cannabinoid receptor ligands that have been most widely used in binding assays or for autoradiography are the CB₁-selective [³H]SR141716A (CB₁ K_d = 0.19 to 1.24 nM) and [³H]CP 55940,
[³H]WIN 55,212-2 and [³H]HU-243, all three of which bind more or less equally well to CB₁ and CB₂ receptors. Typical K_d values for [³H]CP 55940, [³H]WIN 55,212-2 and [³H]HU-243 are 0.07 to 4 nM, 1.9 to 16.2 nM and 0.045 nM respectively at CB₁ receptors and 0.2 to 7.4 nM, 2.1 to 3.8 nM and
0.061 nM respectively at CB₂ receptors.⁴⁵ Thus [³H]HU-243 is particularly potent. Three radiolabelled ligands have also been developed as potential probes for human single photon emission computed tomography (SPECT) or positron emission tomography (PET) experiments. These are
¹²³I-labelled analogues of AM 251 (CB₁ K_d = 0.23 to 0.62 nM) and AM 281^66,88,89 and an ¹⁸F-labelled analogue of SR141716A (SR144385).⁹⁰ Particularly promising results have been obtained from animal experiments with [¹²³I]AM 281.⁶⁶

Inhibitors of the enzymic hydrolysis of endocannabinoids

The presence of FAAH in many cannabinoid bioassay systems has created the need for FAAH inhibitors that can be used to protect endocannabinoids from enzymic hydrolysis.¹² The need for FAAH inhibitors also stems from the possibility that FAAH inhibitors might have therapeutic potential as indirect cannabinoid receptor agonists, the expectation being that drugs that activate the endocannabinoid system by increasing the concentration of endocannabinoids at cannabinoid receptors would be more selective than direct agonists. This is because they are unlikely to affect all parts of the endocannabinoid system at one time, producing instead effects only at sites where on-going production of endocannabinoids is taking place. The drug that has been most widely used to inhibit the enzymic hydrolysis of anandamide is the non-specific serine protease inhibitor, phenylmethyl-sulphonyl fluoride.¹²

FAAH inhibitors with much greater potency are now available, two notable examples being palmitylsulphonyl fluoride (AM 374) and stearylsulphonyl fluoride (AM 381). Both ligands inhibit FAAH irreversibly and show good separation between potency for FAAH inhibition and ability to bind to CB₁ receptors. Thus the EC₅₀ value of AM 374 is 7 nM for inhibition of FAAH and 520 nM for displacement of [³H]CP 55940 from specific binding sites on rat forebrain membranes whilst the corresponding EC₅₀ values of AM 381 are 4 nM and 18.5 mM respectively.⁹¹ AM 374 potentiates both anandamide-induced inhibition of evoked [³H]acetylcholine release in rat hippocampal slices⁹² and anandamide-induced suppression of rat operant lever pressing and open field locomotor activity.⁹³ When administered by themselves, both AM 374 and the anandamide membrane transport inhibitor AM 404 (see below) share the ability of cannabinoid receptor agonists to ameliorate spasticity in mice with chronic relapsing experimental allergic encephalomyelitis (CREAE), an induced syndrome that serves as an animal model of multiple sclerosis.^94,95 This they most likely do by augmenting levels of endogenously released endocannabinoid(s) as brain and spinal cord concentrations of both anandamide and 2-AG are greater in spastic CREAE mice than in control animals.⁹⁵ The ability of AM 381 to potentiate anandamide remains to be reported.

Methyl arachidonyl fluorophosphonate (MAFP) is another potent irreversible inhibitor of FAAH (EC₅₀ = 2.5 nM).⁶¹ However, it also potently displaces [³H]CP 55940 from specific binding sites on rat brain membranes in an irreversible manner (EC₅₀ = 20 nM)⁶¹ and produces insurmountable antagonism at CB₁ receptors (Table 2).⁹⁶ One recently developed analogue of MAFP, O-1887, shows much greater separation of FAAH inhibitory potency (EC₅₀ = 15 nM) from potency for binding to CB₁ receptors (CB₁ K_i > 10 mM).⁹⁷ Interestingly, this compound shares the ability of PMSF to elicit cannabimimetic responses when administered to mice by itself.^97,98 Whether this is through inhibition of the enzymic hydrolysis of endogenously released cannabinoids has yet to be investigated. Other reasonably potent FAAH inhibitors are diazomethyl arachidonyl ketone (EC₅₀ = 520 nM)⁹⁹ and (E)-6-(bromomethylene) tetrahydro-3-(1-naphthalenyl)-2H-pyran-2-one (EC₅₀ = 800 nM),¹⁰⁰ which act irreversibly, and arachidonyl trifluoromethyl ketone (EC₅₀ = 900 nM), which is a tight-binding but reversible FAAH inhibitor that also inhibits cytosolic phospholipase A₂.^91,101 Arachidonyl trifluoromethyl ketone and diazomethyl arachidonyl ketone have been reported to bind to CB₁ receptors with K_i values of 0.65 and 1.3 mM respectively^99,102 and there is also a report that the EC₅₀ of arachidonyl trifluoromethyl ketone for displacement of [³H]CP 55940 from CB₁ binding sites is 2.5 mM.⁹¹ Significantly more potent as FAAH inhibitors are a series of a-keto bicyclic heterocycles with alkyl or phenylalkyl side chains that have recently been developed by Boger et al.¹⁰³ These inhibit the enzyme competitively, some with K_i values in the picomolar or low nanomolar range. The structure of the most potent of this new generation of FAAH inhibitors is shown in Figure 6 (Compound 59; K_i = 140 pM). Other pharmacological properties of these inhibitors, for example their ability to interact with cannabinoid or vanilloid receptors or to potentiate endocannabinoids, have yet to be reported.

It has been proposed that since endogenous substrates for FAAH such as linoleoylethanolamide, oleoylethanolamide and oleamide serve as reversible inhibitors of this enzyme,^36,104-106 anandamide may be protected from enzymic hydrolysis by endogenously produced FAAH inhibitors.³² This putative “entourage” effect may well extend to 2-arachidonoyl glycerol which can be protected from the esterase action of FAAH (and so potentiated) by the endogenous fatty acid derivatives, 2-linoleyl glycerol and 2-palmitoyl glycerol.^32,46,107 Other inhibitors of FAAH have also been identified and details of these appear elsewhere.^{18,39,43,57,97,101,102,108-110}

Inhibitors of endocannabinoid membrane transport

One notable membrane transport inhibitor to have been developed is N-(4-hydroxyphenyl) arachidonylamide (AM 404). This has been reported to inhibit anandamide uptake by rat cultured cortical neurones (EC₅₀ = 1 mM) and astrocytes (EC₅₀ = 5 mM) and to potentiate anandamide both in vitro and in vivo.^35,111 When administered to rats by itself, AM 404 increases plasma levels of anandamide and shares the ability of this endocannabinoid to decrease locomotor activity, depress plasma levels of prolactin and alter tyrosine hydroxylase activity in the hypothalamus (increase) and substantia nigra (decrease).^112-114 The inhibitory effect of AM 404 on locomotor activity has been found to be susceptible to antagonism by SR141716A.^113,114 AM 404 does not, however, elicit two other typical responses to CB₁ receptor agonists in rats: catalepsy and signs of analgesia in the hot plate test.¹¹³

Structure-activity experiments with AM 404 analogues have revealed major differences between the structural requirements of the transporter for ligand recognition and those for ligand translocation.³⁵ At concentrations in the low micromolar range, it displaces [³H]CP 55940 from specific binding sites on rat forebrain membranes (K_i = 1.76 mM).¹¹⁵ Even so, there are no reports that AM 404 behaves as a cannabinoid receptor agonist or antagonist. There is also one report that AM 404 inhibits FAAH with an EC₅₀ of 0.5 mM.⁵⁷ However, this conflicts with other reports that the EC₅₀ of AM 404 for inhibition of FAAH is 22 mM¹¹⁶ or exceeds 30 mM.¹¹¹ Submicromolar concentrations of AM 404 can activate vanilloid receptors,^78,80 prompting the need for a more selective inhibitor of endocannabinoid membrane transport. This need was recently addressed by the development of the AM 404 analogue, VDM11, which exhibits the same potency as AM 404 as an endocannabinoid membrane transport inhibitor in C6 rat glioma cells (EC₅₀ = 10 mM) but markedly less efficacy than AM 404 at vanilloid receptors.¹¹⁶ Although membrane transport-inhibiting concentrations of VDM11 undergo some degree of binding to CB₁ receptors, these concentrations (5-10 mM) displace [³H]SR141716A from CB₁ binding sites by less than 50%.¹¹⁶ VDM11 also exhibits relatively low potency as an inhibitor of FAAH (EC₅₀ > 50 mM). Another ligand that inhibits the endocannabinoid membrane transporter (EC₅₀ = 3.6 mM) more readily than it inhibits FAAH (EC₅₀ = 32 mM), at least in RBL-2H3 cells, is arvanil.⁷⁹ However this agent not only has reasonably high affinity for CB₁ receptors (K_i = 0.5 mM) but is also a potent vanilloid receptor agonist (human and rat VR1 EC₅₀ = 0.5 nM).^71,79,116 Details of other membrane transport inhibitors are to be found elsewhere.^{35,37,57,79,110,117}

One other important recent advance has been the development of a fluorescent substrate for anandamide uptake (SKM 4-45-1) that should serve as a useful experimental tool for transporter studies.¹¹⁷ This agent is more potent as an inhibitor of endocannabinoid membrane transport (EC₅₀ = 7.8 mM) than as an FAAH inhibitor (EC₅₀ > 10 mM). Whilst the ability of SKM 4-45-1 to interact with vanilloid receptors remains to be investigated, it is known that, at 3 mM, it does not displace [³H]CP 55940 from CB₁ binding sites on rat brain membranes.¹¹⁷

Other cannabinoid receptors?

Calignano et al¹¹⁸ have postulated the existence of an SR144528-sensitive non-CB₂ cannabinoid receptor (‘CB₂-like’ receptor). This hypothesis is based on evidence that even though palmitoylethanolamide lacks significant affinity for CB₁ or CB₂ receptors,^30,119-121 its ability to produce signs of anti-hyperalgesia in the mouse formalin paw test is readily attenuated by the CB₂-selective antagonist/inverse agonist, SR144528.¹¹⁸ The existence of CB₂-like receptors in the mouse vas deferens has also been proposed.¹²² More recently, evidence was obtained for the presence in vascular endothelium of an SR141716A-sensitive non-CB₁, non-CB₂, non-vanilloid receptor that is unresponsive to established non-eicosanoid CB₁/CB₂ receptor agonists but can be activated both by the eicosanoid cannabinoids, anandamide and methanandamide, and by certain classical cannabinoids that do not act through CB₁ or vanilloid receptors (“abnormal cannabidiol” and its more potent analogue, (O-1602).^123,124 Interestingly, two effects of abnormal cannabidiol, hypotension and mesenteric vasodilation, were found to be antagonized by the non-psychotropic classical cannabinoid, cannabidiol. So too was anandamide-induced mesenteric vasodilation. Finally, the existence in the brain of non-CB₁, non-CB₂, SR141716A-insensitive G protein-coupled receptors for anandamide has recently been proposed to explain results obtained from experiments with CB₁ knockout mice.¹²⁵

Conclusions

The field of cannabinoid research is now at a particularly exciting stage. Thus, whilst a number of major advances have recently been made, many important questions still remain unanswered or incompletely addressed so prompting the need for more research at both non-clinical and clinical levels. Of particular importance at the pharmacological level is the need:

• to understand the modes of action of anandamide and 2-arachidonoyl glycerol more fully and to characterize their effects and the processes that terminate these effects more completely,
• to develop silent cannabinoid receptor antagonists possibly through further elucidation of the structural features that attenuate agonist and inverse agonist efficacy at cannabinoid receptors without reducing cannabinoid receptor affinity,
• to seek out any additional endocannabinoids or novel types of cannabinoid receptor,
• to achieve a fuller pharmacological characterization of some existing inhibitors of endocannabinoid membrane transport and enzymic hydrolysis,
• to develop inhibitors of endocannabinoid membrane transport and enzymic hydrolysis that show greater separation between their sought-after inhibitory actions and other pharmacological actions, particularly the ability to activate or block cannabinoid, vanilloid or other receptors or to inhibit other enzymes or transporters,
• to obtain more detailed information about endocannabinoid release in both health and disease and hence gain a better appreciation of the therapeutic potential of drugs that inhibit endocannabinoid membrane transport and/or endocannabinoid enzymic hydrolysis,
• to search for beneficial and unwanted consequences of selective inhibition of both endocannabinoid membrane transport and the enzymic hydrolysis of endocannabinoids.

Note added in proof: the presence in the brain of a novel CB₁-selective endocannabinoid, 2-arachidonoyl glyceryl ether (noladin ether), has just been announced.¹²⁶

Published May 2001