The cannabinoid receptor: biochemical, anatomical and behavioral characterization.

The actions of the active principle of marihuana, delta 9-tetrahydrocannabinol, are mimicked by synthetic cannabinoid agonists showing high potency and enantio-selectivity in behavioral assays. These drugs have been used to characterize cannabinoid receptor binding, biochemistry and pharmacology, leading to a better understanding of the effects of cannabinoids in the CNS of humans and experimental animals.

Conformational analysis of the prototype nonclassical cannabinoid CP-47,497, using 2D NMR and computer molecular modeling.

In an effort to determine the stereochemical requirements for pharmacological activity among the series of nonclassical cannabinoids synthesized at Pfizer, we have studied the conformational properties of the parent bicyclic analog CP-47,497. For this study, we have used a combination of solution NMR and theoretical computational approaches. The energetically favored conformation has the phenolic ring almost perpendicular to the cyclohexanol ring which exists in a chair conformation. The OH bond of the phenol is preferentially coplanar with the aromatic ring and points toward the C2 ring proton, while the dimethylheptyl side chain adopts a conformation almost perpendicular to the aromatic ring. The conformational features of this nonclassical cannabinoid analog closely resemble those of its classical counterparts. The only apparent difference is the small dihedral angle (psi 1 = 62 degrees) between the planes of the two rings of CP-47,497 compared to that of the tricyclic tetrahydro- or hexahydrocannabinol analogs (psi 1 = 137 degrees). However, CP-47,497 can be perfectly superimposed over the respective tricyclic analog by rotation around the Ph-cyclohexyl bond (C6-C7 bond) and assume a conformation which is energetically higher than the preferred one by 3.0 kcal/mol. It can be argued that such a conformation may be acquired by the nonclassical analog during its interaction with the active site.

A cannabinoid derived prototypical analgesic.

The synthesis and analgesic testing of 3-[4-(1,1-dimethylheptyl)-2-hydroxyphenyl]cyclohexanol (1) are described. Prior (SAR) studies led us to conclude that the pyran ring of 9-nor-9 beta-hydroxyhexahydrocannabinol (HHC) was not necessary for the expression of biological activity in this series of cannabinoids. Analysis of models and the use of molecular mechanics calculations suggested that a simpler compound, such as 1, would possess the biological activity of HHC. Compound 1 was prepared in nine steps from [3-(benzyloxy)phenyl]acetonitrile (2). Biological testing in five models of pain shows that compound 1 and morphine are equally potent as analgesics and demonstrates that the pyran ring of HHC is not necessary for biological activity. Further simplification of 1 was pursued by the synthesis of 4-[4-(1,1-dimethylheptyl)-2-hydroxyphenyl]-2-pentanol (17), but this derivative exhibits significantly reduced analgesic activity.

Selective and potent analgetics derived from cannabinoids.

Based on the hypothesis that analgetic activity is a dissociable feature of the cannabinoid molecule, we examined modifications of the side chain, the phenolic moiety, and, most significantly, structures that lack the benzopyran functionality present in THC and (--)-9-nor-9 beta-hydroxyhexahydrocannabinol (HHC). A new grouping, the 1-methyl-4-phenylbutyloxy C-3 side chain, elaborates a unique lipopholic region. Replacement of the phenol substituent produced several derivatives which retain analgetic activity in the codeine potency range. Introduction of a weakly basic nitrogen at C-5 and deletion of the axial methyl group in the B ring, two structural changes forbidden by traditional cannabinoid SAR, resulted in a unique family of benzoquinolines with potent analgetic activity. The prototype of this series, levonantradol, exhibits potent and stereospecific analgetic and antiemetic activity.

Enhancement of brain [3H]flunitrazepam binding and analgesic activity of synthetic cannabimimetics.

Novel, synthetic cannabimimetics and delta 9-tetrahydrocannabinol were found to enhance the binding of [3H]flunitrazepam to mouse brain in vivo. This property, suggestive of facilitation of binding to benzodiazepine receptors, is consistent with the potentiation of the anticonvulsant activity of diazepam against pentylenetetrazol by these compounds. The relative potencies of delta 9-tetrahydrocannabinol and the new cannabimimetics for enhancing [3H]flunitrazepam binding in vivo could also be correlated with their relative analgesic efficacies. Similar pharmacological stereospecificity was displayed for both binding enhancement and analgesic effects. The following order of decreasing potency was observed: N-methyllevonantradol and (-)-CP-55,244 greater than levonantradol, canbisol, CP-42,096 and (-)-CP-55,940 greater than 9-beta-normethyl-9-beta-hydroxyhexahydrocannabinol, nabilone and CP-47,497 greater than delta 9-tetrahydrocannabinol. Dextronantradol, (+)-CP-55,940 and (+)-CP-55,244 were considerably less active than the respective (-)-enantiomers; cannabidiol was inactive. Extensive investigation of structure versus activity led to N-methyllevonantradol and the 3-(2-hydroxyphenyl)cyclohexanols derivative, (-)-CP-55,244, which are approximately 1000-fold more potent than delta 9-tetrahydrocannabinol.

Behavioral, biochemical, and molecular modeling evaluations of cannabinoid analogs.

Numerous cannabinoids have been synthesized that are extremely potent in all of the behavioral assays conducted in our laboratory. An important feature in increasing potency has been the substitution of a dimethylheptyl (DMH) side chain for the pentyl side chain. Our previous studies have shown that (-)-11-OH-delta 8-THC-dimethylheptyl was 80-1150 times more potent than delta 9-THC. Stereospecificity was demonstrated by its (+)-enantiomer which was more than 1400-7500 times less potent. A related series of DMH cannabinoid analogs has recently been synthesized and preliminary evaluations reported here. (-)-11-OH-delta 9-THC-DMH was found to be equipotent with (-)-11-OH-delta 8-THC-DMH. The aldehyde (-)-11-oxo-delta 9-THC-DMH was 15-50 times more potent than delta 9-THC. Surprisingly, (-)-11-carboxy-delta 9-THC-DMH was also active, being slightly more potent than delta 9-THC. In the bicyclic cannabinoid series, the length and bulk of the side chain were found to be equally important. Aminoalkylindoles, which are structurally dissimilar from classical cannabinoids, have been found to exhibit a pharmacological profile similar to delta 9-THC. Though not extremely potent in vivo, they appear to represent an entirely new approach to studying the actions of the cannabinoids. The structural diversity and wide-ranging potencies of the analogs described herein provide the opportunity to develop a pharmacophore for the cannabinoids using molecular modeling techniques.

The conformational properties of the highly selective cannabinoid receptor ligand CP-55,940.

During a search for novel drugs possessing analgesic properties but devoid of the psychotropic effects of marijuana, a group of molecules designated as nonclassical cannabinoids was synthesized by Pfizer. Of these nonclassical cannabinoids CP-55,940 has received the most attention principally because it was used as the high affinity radioligand during the discovery and characterization of the G-protein-coupled cannabinoid receptor. In an effort to obtain information on the stereoelectronic requirements at the cannabinoid receptor active site, we have studied the conformational properties of CP-55,940 using a combination of solution NMR and computer modeling methods. Our data show that for the most energetically favored conformation, (i) the aromatic phenol ring is perpendicular to the cyclohexane ring, and the phenolic O-H bond is coplanar with the aromatic ring and points away from the cyclohexyl ring; ii) the dimethylheptyl chain adopts one of four preferred conformations in all of which the chain is almost perpendicular to the phenol ring; and iii) an intramolecular H-bond between the phenolic and hydroxypropyl groups allows all three hydroxyl groups of CP-55,940 to be oriented toward the upper face of the molecule. Such an orientation by the OH groups may be a characteristic requirement for cannabimimetic activity.

Cannabimimetic activity from CP-47,497, a derivative of 3-phenylcyclohexanol.

CP-47,497 (cis-3-[2-hydroxy-4(1,1-dimethylheptyl)phenyl]-cyclohexan-1-ol) is characterized as a cannabimimetic agent, with 3 to 28 times greater potency than delta 9-tetrahydrocannabinol delta 9-THC), based on the following. In common with delta 9-THC and other active structures closely related chemically to delta 9-THC, CP-47,497 exerts analgesic, motor depressant, anticonvulsant and hypothermic effects. It elicits vocalization in palpated rats and ataxia in dogs. In drug discrimination studies in rats, the stimulus properties of delta 9-THC (3.2 mg/kg i.p.) are generalized to CP-47,497, with an absolute threshold dose 3 to 14 times lower than the threshold dose of delta 9-THC itself, depending on route. Furthermore, rats are unable to discriminate between the stimulus properties of equated i.p. doses of delta 9-THC and CP-47,497 after prolonged training. Despite its potent behavioral effects, CP-47,497, like delta 9-THC, does not resemble standard antipsychotic, antidepressant, antianxiety or hypnotic drugs in simple drug interaction tests. Based on its pharmacology, CP-47,497 exemplifies a simplified structure capable of producing many effects in common with those of delta 9-THC and 9-normethyl-9 beta-OH-hexahydrocannabinol.

Pharmacological profile of a series of bicyclic cannabinoid analogs: classification as cannabimimetic agents.

Opening of the pyran ring of delta 9-tetrahydrocannabinol (THC) produces cannabidiol, a bicyclic cannabinoid devoid of many pharmacological properties produced by delta 8-THC or delta 9-THC. Interestingly, the bicyclic compound CP-47,497 (VI) has been described as producing many of the pharmacological effects produced by delta 9-THC, and another related bicyclic analog CP-55,940 (XIV) has been used to successfully define a cannabinoid binding site. A series of 16 bicyclic analogs of VI and XIV were evaluated and compared with the pharmacological profile of cannabidiol, delta 8-THC and delta 9-THC. The goals of the studies described herein were to determine whether these bicyclic analogs possess similar pharmacological properties of delta 9-THC, to compare pharmacological activity after s.c. and i.v. administration, and to evaluate the structure-activity relationship of this series of analogs for further insight into cannabinoid mechanism of action. Each analog was evaluated for its ability to produce hypoactivity, hypothermia, antinociception and catalepsy in mice. The ED50 values generated from these assays were averaged to provide an index of activity. The ED50 values for delta 9-THC varied from 1.0 to 1.5 mg/kg, giving an overall index of activity of 1.3. The index for delta 8-THC was 6.0, making this isomer 4-fold less potent. Although several bicyclic analogs (V, VI, VII, VIII, XI, XII, XIV and XVI) proved to be truly cannabimimetic, three (IV, IX and X) were sufficiently unique to be classified as noncannabimimetic. The index of activity of cannabimimetic bicyclic analogs varied from 0.2 to 2.2, although some minor differences between the bicyclics and delta 9-THC exist.(ABSTRACT TRUNCATED AT 250 WORDS)

Nonclassical cannabinoid analgetics inhibit adenylate cyclase: development of a cannabinoid receptor model.

Extensive structure-activity relationship studies have demonstrated that specific requirements within the cannabinoid structure are necessary to produce potent analgesia. A three-point association between the agonist and the receptor mediating analgesia consists of: 1) the C ring hydroxyl, 2) the phenolic A ring hydroxyl, and 3) the A ring alkyl hydrophobic side chain. Potent tricyclic and bicyclic structures were synthesized as "nonclassical" cannabinoid analgetics that conform to this agonist-receptor three-point interaction model. At the cellular level, centrally active cannabinoid drugs inhibit adenylate cyclase activity in a neuroblastoma cell line. The structure-activity relationship profile for inhibition of adenylate cyclase in vitro was consistent with this same three-point association of agonists with the receptor. A correlation exists between the potency of drugs to produce analgesia in vivo and to inhibit adenylate cyclase in vitro. Enantio- and stereoselectivity were exhibited by the nonclassical cannabinoid compounds for both the analgetic response and the ability to inhibit adenylate cyclase. The magnitude of the enantioselective response was equal for both the biochemical and physiological endpoints. Based on the parallels in structure-activity relationships and the enantioselective effects, it is postulated that the receptor that is associated with the regulation of adenylate cyclase in vitro may be the same receptor as that mediating analgesia in vivo. A conceptualization of the cannabinoid analgetic receptor is presented.

Characterization and localization of cannabinoid receptors in rat brain: a quantitative in vitro autoradiographic study.

A potent, synthetic cannabinoid was radiolabeled and used to characterize and precisely localize cannabinoid receptors in slide-mounted sections of rat brain and pituitary. Assay conditions for 3H-CP55,940 binding in Tris-HCl buffer with 5% BSA were optimized, association and dissociation rate constants determined, and the equilibrium dissociation constant (Kd) calculated (21 nM by liquid scintillation counting, 5.2 nM by quantitative autoradiography). The results of competition studies, using several synthetic cannabinoids, add to prior data showing enantioselectivity of binding and correlation of in vitro potencies with potencies in biological assays of cannabinoid actions. Inhibition of binding by guanine nucleotides was selective and profound: Nonhydrolyzable analogs of GTP and GDP inhibited binding by greater than 90%, and GMP and the nonhydrolyzable ATP analog showed no inhibition. Autoradiography showed great heterogeneity of binding in patterns of labeling that closely conform to cytoarchitectural and functional domains. Very dense 3H-CP55,940 binding is localized to the basal ganglia (lateral caudate-putamen, globus pallidus, entopeduncular nucleus, substantia nigra pars reticulata), cerebellar molecular layer, innermost layers of the olfactory bulb, and portions of the hippocampal formation (CA3 and dentate gyrus molecular layer). Moderately dense binding is found throughout the remaining forebrain. Sparse binding characterizes the brain stem and spinal cord. Densitometry confirmed the quantitative heterogeneity of cannabinoid receptors (10 nM 3H-CP55,940 binding ranged in density from 6.3 pmol/mg protein in the substantia nigra pars reticulata to 0.15 pmol/mg protein in the anterior lobe of the pituitary). The results suggest that the presently characterized cannabinoid receptor mediates physiological and behavioral effects of natural and synthetic cannabinoids, because it is strongly coupled to guanine nucleotide regulatory proteins and is discretely localized to cortical, basal ganglia, and cerebellar structures involved with cognition and movement.

Potent analgetics derived from 9-nor-9 beta-hydroxyhexahydrocannabinol.

Based on the report of morphine-like analgetic activity of 9-nor-9 beta-hydroxyhexahydrocannabinol (HHC), we undertook a study of structural modifications of the C-3 side chain of HHC to optimize the analgetic activity. We ultimately examined four distinct classes of side chains: (1) alkyl (la-lc), (2) arylalkyl (ld-lh), (3) alkoxy (li-lj) and (4) arylalkyloxy (lk-lo). Three of these derivatives (lb, lf, ll) possessed analgetic activity 10X morphine. These studies demonstrate that the C-3 side chain of HHC can be modified in a structure-dependent fashion to yield potent, nonopioid analgetics. In addition, the effect of the 1 methyl-4-phenylbutyloxy side chain is unique among the side chains examined.

Structure-activity relationships defining the ACD-tricyclic cannabinoids: cannabinoid receptor binding and analgesic activity.

Previous studies of the structure-activity relationships (SAR) for binding of a series of AC-bicyclic cannabinoid structures to the cannabinoid receptors in rat brain (believed to comprise the CB1 subtype) demonstrated the importance of the A-ring aryl C-3 side chain and phenolic hydroxyl substituents, and elucidated the importance of a C-ring hydroxyalkyl substituent [Melvin et al. Mol. Pharmacol. 44, 1008-1015 (1993)]. The present investigation examines the SAR surrounding this region (D-ring) of the molecule that is not present in the structure of delta(9)-THC and other classical cannabinoid compounds. Both rigid fused ring benzo and cyclohexyl derivatives (creating the D-ring) retained binding affinity for the cannabinoid receptor. Extension of ketone or hydroxyl substituents from the C2 position of the D-ring resulted in a 3-fold increase in binding affinity over the unsubstituted structure. However, the fused ring structure is not critical for the interaction with the receptor in as much as opening the ring did not decrease the potency. Extension of the D-ring C-2 alcohol by one carbon in length resulted in a pair of structures, for which the greatest affinity for the CB1 receptor occurred for the hydroxymethyl group in the axial conformation [(+/-)-CP-55,244]. Upon resolution, the latter provided a pair of enantiomers: (-)-CP-55,244 was approximately 3-fold more potent than the racemic mixtures, and (+)-CP-55,244 failed to bind to the CB1 receptor with an IC50 below 1 mM. Opening of the D-ring of these structures resulted in a loss of binding affinity. This study demonstrates that the potency could be optimized in (-)-CP-55,244 for both binding to the CB1 receptor and the biological activity of analgesia. In addition, the rigid positioning of the hydroxypropyl moiety of CP-55,940 enforced by the decalin ring structure of CP-55,244 increased the enantioselectivity by greater than 100-fold. These data define the critical stereochemistry for a region of the nonclassical ACD-tricyclic cannabinoid structure that contributes a potential hydrogen bonding component to the ligand-receptor interaction mechanism. Inasmuch as this region of the molecule is not present on classical ABC-tricyclic cannabinoid compounds, these studies elucidate a unique agonist recognition site on the CB1 receptor.

Disposition and metabolism of tenidap in the rat.

Tenidap is a new antirheumatic drug currently undergoing clinical evaluation. It inhibits production and activity of cytokines in vivo and causes significant reductions in plasma markers of disease activity in rheumatoid arthritis. After the oral administration of C-14 labeled tenidap, bile, urine and plasma were examined by HPLC and atmospheric pressure tandem mass spectrometry. Label is excreted primarily in bile/feces and the remainder in urine, with good recoveries. Numerous metabolites were identified and the structures of most were confirmed by comparison with authentic synthetic samples. Hydroxylation in several positions on both the oxindole and thienyl rings of tenidap represents the major routes of metabolism; most of these metabolites are subsequently conjugated. The glucuronide of 5'-hydroxytenidap, excreted primarily in bile, is the major metabolite, constituting about one third of the oral dose recovered. Other pathways include dihydroxylation and methoxylation on the thienyl ring. An unusual reduction of hydroxytenidap took place, resulting in the formation of a novel thiolactone analog. Anaerobic incubation with rat cecal contents generated the thiolactone metabolite, suggesting the involvement of gut microflora.

Pharmacology and stereoselectivity of structurally novel cannabinoids in mice.

The pharmacological effects of three stereoisomeric pairs of structurally novel cannabinoids were tested after i.v. administration in mice for depression of spontaneous activity and the production of hypothermia, antinociception and catalepsy. The (-)-enantiomers were as much as 770 times more potent than delta 9-6a,10a-trans-tetrahydrocannabinol and were 7 to 2000 times more potent than their respective (+)-enantiomers. The order of potency for cannabinoid-induced effects was spontaneous activity greater than antinociception greater than hypothermia greater than or equal to catalepsy. Levonantradol was active between 0.123 to 1.5 mg/kg, whereas dextronantradol, its (+)-enantiomer was inactive. (-)-CP 55,244 and (-)-CP55,940 analogs which lack the dihydropyran ring were 5 to 775 times more potent than delta 9-6a,10a-trans-tetrahydrocannabinol and 30 to 2000 times more potent than their respective (+)-enantiomers. Some separation of effects was demonstrated with (+)-CP 55,243 and (-)-CP 56,667 which were inactive in producing hypothermia and catalepsy but were active in the spontaneous activity and tail-flick procedures. The high degree of enantioselectivity and potency of these nonclassical cannabinoids are indicative of a highly specific mechanism of action such as a receptor.

Determination and characterization of a cannabinoid receptor in rat brain.

The determination and characterization of a cannabinoid receptor from brain are reported. A biologically active bicyclic cannabinoid analgetic CP-55,940 was tritium-labeled to high specific activity. Conditions for binding to rat brain P2 membranes and synaptosomes were established. The pH optimum was between 7 and 8, and specific binding could be eliminated by heating the membranes to 60 degrees. Binding to the P2 membranes was linear within the range of 10 to 50 micrograms of protein/ml. Specific binding (defined as total binding displaced by 1 microM delta 9-tetrahydrocannabinol (delta 9-THC) or 100 nM desacetyllevonantradol) was saturable. The Kd determined from Scatchard analysis was 133 pM, and the Bmax for rat cortical P2 membranes was 1.85 pmol/mg of protein. The Hill coefficient for [3H]CP-55,940 approximated 1, indicating that, under the conditions of assay, a single class of binding sites was determined that did not exhibit cooperativity. The binding was rapid (kon approximately 2.6 x 10(-4) pM-1 min-1) and reversible (Koff approximately 0.016 min-1) and (koff' greater than 0.06 min-1). The two Kd values estimated from the kinetic constants approximately 55 pM and exceeded 200 pM, respectively. The binding of the agonist ligand [3H]CP-55,940 was decreased by the nonhydrolyzable GTP analog guanylylimidodiphosphate. The guanine nucleotide induced a more rapid dissociation of the ligand from the binding site, consistent with an allosteric regulation of the putative receptor by a G protein. The binding was also sensitive to MgCl2 and CaCl2. Binding of [3H]CP-55,940 was displaced by cannabinoid drugs in the following order of potency: CP-55,940 greater than or equal to desacetyllevonantradol greater than 11-OH-delta 9-THC = delta 9-THC greater than cannabinol. Cannabidiol and cannabigerol displaced [3H]CP-55,940 by less than 50% at 1 microM concentrations. The (-)-isomer of CP-55,940 displaced with 50-fold greater potency than the (+)-isomer. This pharmacology is comparable to both the inhibition of adenylate cyclase in vitro and the analgetic activity of these compounds in vivo. The criteria for a high affinity, stereoselective, pharmacologically distinct cannabinoid receptor in brain tissue have been fulfilled.

Structure-activity relationships for cannabinoid receptor-binding and analgesic activity: studies of bicyclic cannabinoid analogs.

Cannabimimetic compounds, such as delta 9-tetrahydrocannabinol (delta 9-THC), evoke analgesia in addition to other behavioral responses in humans and animals. The cannabinoid receptor mediating this response has been characterized by its ability to bind the cannabinoid agonist [3H]CP-55,940 and to inhibit adenylyl cyclase via Gi. An investigation of structural requirements for antinociceptive activity of cannabinoid structures led to the development of a simple bicyclic cannabinoid agonist, CP-47,497, that possessed a spectrum of cannabinoid activities in animals that resembled that of delta 9-THC. The present investigation examines several series of CP-47,497 analogs for their binding affinity at the cannabinoid receptor and their ability to evoke analgesia in rodents. Analogs substituted at the C-3 alkyl side chain exhibited maximal affinity for the cannabinoid receptor with side chains of seven or eight carbons in length. Analgesic potency paralleled the receptor-binding affinity. The cyclohexyl ring was optimized as a six- or seven-membered ring structure for binding as well as analgesic activity. Cyclohexyl alkyl side chain extensions of up to four carbons in length had little influence on the affinity for the receptor or analgesic activity. Hydroxyalkyl side chains exhibited optimal binding affinity and antinociceptive activity at three or four carbon atoms in length; however, polar groups closer to the ring diminished binding to the receptor. The importance of the phenolic and cyclohexyl hydroxyl groups for binding affinity was demonstrated. In general, analgesic activity correlated well with the affinity of these analogs for the cannabinoid receptor. Exceptions could be explained by metabolic transformations likely to occur in vivo.

Cannabinoid structure-activity relationships: correlation of receptor binding and in vivo activities.

Although a receptor exists for cannabinoid drugs, it is uncertain which pharmacological actions this receptor mediates. This structure-activity relationship investigation was initiated to determine which effects might correspond to binding affinity for the cannabinoid receptor, as well as to explore the binding requirements of this site. The ability of nearly 60 cannabinoids to displace [3H]CP-55,940 [(-)-3-[2-hydroxy-4-(1,1-dimethylheptyl) phenyl]-4-[3-hydroxy propyl] cyclohexan-1-ol] was determined before establishing correlations between receptor affinity and in vivo pharmacological potency. Analysis of [3H]CP-55,940 binding indicated a Hill coefficient of 0.97, a Bmax of 499 pM (3.3 pmol/mg of protein) and an apparent Kd of 924 pM. Closer inspection indicated the binding assay exhibited "zone B" characteristics, and use of correction equations indicated a true Kd for CP-55,940 of 675 pM. The structure-activity relationship indicated the importance of side chain structure to high-affinity binding, with the most potent analogs (K1 < 10 nM) possessing either a dimethylheptyl side-chain, a similarly complex branched side chain or a halogen substituent at the 5' position. Comparative analysis of K1 values to in vivo potency in a mouse model indicated a high degree of correlation between parameters for the depression of spontaneous locomotor activity (r = 0.91) and for the production of antinociception (r = 0.90), hypothermia (r = 0.89) and catalepsy (r = 0.85). Similarly high correlations were demonstrated between binding affinity and in vivo potency in both the rat drug discrimination model (r = 0.81) and for psychotomimetic activity in humans (r = 0.88).(ABSTRACT TRUNCATED AT 250 WORDS)