(-)-Trans-Δ9-tetrahydrocannabinol (Δ9-THC) is the main compound responsible for the intoxicant activity of Cannabis sativa L. The length of the side alkyl chain influences the biological activity of this cannabinoid. In particular, synthetic analogues of Δ9-THC with a longer side chain have shown cannabimimetic properties far higher than Δ9-THC itself. In the attempt to define the phytocannabinoids profile that characterizes a medicinal cannabis variety, a new phytocannabinoid with the same structure of Δ9-THC but with a seven-term alkyl side chain was identified. The natural compound was isolated and fully characterized and its stereochemical configuration was assigned by match with the same compound obtained by a stereoselective synthesis. This new phytocannabinoid has been called (-)-trans-Δ9-tetrahydrocannabiphorol (Δ9-THCP). Along with Δ9-THCP, the corresponding cannabidiol (CBD) homolog with seven-term side alkyl chain (CBDP) was also isolated and unambiguously identified by match with its synthetic counterpart. The binding activity of Δ9-THCP against human CB1 receptor in vitro (Ki = 1.2 nM) resulted similar to that of CP55940 (Ki = 0.9 nM), a potent full CB1 agonist. In the cannabinoid tetrad pharmacological test, Δ9-THCP induced hypomotility, analgesia, catalepsy and decreased rectal temperature indicating a THC-like cannabimimetic activity. The presence of this new phytocannabinoid could account for the pharmacological properties of some cannabis varieties difficult to explain by the presence of the sole Δ9-THC.
Cannabis sativa has always been a controversial plant as it can be considered as a lifesaver for several pathologies including glaucoma1 and epilepsy2, an invaluable source of nutrients3, an environmentally friendly raw material for manufacturing4 and textiles5, but it is also the most widely spread illicit drug in the world, especially among young adults6.
Its peculiarity is its ability to produce a class of organic molecules called phytocannabinoids, which derive from an enzymatic reaction between a resorcinol and an isoprenoid group. The modularity of these two parts is the key for the extreme variability of the resulting product that has led to almost 150 different known phytocannabinoids7. The precursors for the most commonly naturally occurring phytocannabinoids are olivetolic acid and geranyl pyrophosphate, which take part to a condensation reaction leading to the formation of cannabigerolic acid (CBGA). CBGA can be then converted into either tetrahydrocannabinolic acid (THCA) or cannabidiolic acid (CBDA) or cannabichromenic acid (CBCA) by the action of a specific cyclase enzyme7. All phytocannabinoids are biosynthesized in the carboxylated form, which can be converted into the corresponding decarboxylated (or neutral) form by heat8. The best known neutral cannabinoids are undoubtedly Δ9-tetrahydrocannabinol (Δ9-THC) and cannabidiol (CBD), the former being responsible for the intoxicant properties of the cannabis plant, and the latter being active as antioxidant, anti-inflammatory, anti-convulsant, but also as antagonist of THC negative effects9.
All these cannabinoids are characterized by the presence of an alkyl side chain on the resorcinyl moiety made of five carbon atoms. However, other phytocannabinoids with a different number of carbon atoms on the side chain are known and they have been called varinoids (with three carbon atoms), such as cannabidivarin (CBDV) and Δ9-tetrahydrocannabivarin (Δ9-THCV), and orcinoids (with one carbon atom), such as cannabidiorcol (CBD-C1) and tetrahydrocannabiorcol (THC-C1)7. Both series are biosynthesized in the plant as the specific ketide synthases have been identified10.
Our research group has recently reported the presence of a butyl phytocannabinoid series with a four-term alkyl chain, in particular cannabidibutol (CBDB) and Δ9-tetrahydrocannabutol (Δ9-THCB), in CBD samples derived from hemp and in a medicinal cannabis variety11,12. Since no evidence has been provided for the presence of plant enzymes responsible for the biosynthesis of these butyl phytocannabinoids, it has been suggested that they might derive from microbial ω-oxidation and decarboxylation of their corresponding five-term homologs13.
The length of the alkyl side chain has indeed proved to be the key parameter, the pharmacophore, for the biological activity exerted by Δ9-THC on the human cannabinoid receptor CB1 as evidenced by structure-activity relationship (SAR) studies collected by Bow and Rimondi14. In particular, a minimum of three carbons is necessary to bind the receptor, then the highest activity has been registered with an eight-carbon side chain to finally decrease with a higher number of carbon atoms14. Δ8-THC homologs with more than five carbon atoms on the side chain have been synthetically produced and tested in order to have molecules several times more potent than Δ9-THC15,16.
To the best of our knowledge, a phytocannabinoid with a linear alkyl side chain containing more than five carbon atoms has never been reported as naturally occurring. However, our research group disclosed for the first time the presence of seven-term homologs of CBD and Δ9-THC in a medicinal cannabis variety, the Italian FM2, provided by the Military Chemical Pharmaceutical Institute in Florence. The two new phytocannabinoids were isolated and fully characterized and their absolute configuration was confirmed by a stereoselective synthesis. According to the International Non-proprietary Name (INN), we suggested for these CBD and THC analogues the name “cannabidiphorol” (CBDP) and “tetrahydrocannabiphorol” (THCP), respectively. The suffix “-phorol” comes from “sphaerophorol”, common name for 5-heptyl-benzen-1,3-diol, which constitutes the resorcinyl moiety of these two new phytocannabinoids.
A number of clinical trials17,18,19 and a growing body of literature provide real evidence of the pharmacological potential of cannabis and cannabinoids on a wide range of disorders from sleep to anxiety, multiple sclerosis, autism and neuropathic pain20,21,22,23. In particular, being the most potent psychotropic cannabinoid, Δ9-THC is the main focus of such studies. In light of the above and of the results of the SAR studies14,15,16, we expected that THCP is endowed of an even higher binding affinity for CB1 receptor and a greater cannabimimetic activity than THC itself. In order to investigate these pharmacological aspects of THCP, its binding affinity for CB1 receptor was tested by a radioligand in vitro assay and its cannabimimetic activity was assessed by the tetrad behavioral tests in mice.
Identification of cannabidiphorol (CBDP) and Δ9-tetrahydrocannabiphorol (Δ9-THCP) by liquid chromatography coupled to high-resolution mass spectrometry (LC-HRMS)
The FM2 ethanolic extract was analyzed by an analytical method recently developed for the cannabinoid profiling of this medicinal cannabis variety12,24. As the native extract contains mainly the carboxylated forms of phytocannabinoids as a consequence of a cold extraction25, part of the plant material was heated to achieve decarboxylation where the predominant forms are neutral phytocannabinoids. The advanced analytical platform of ultra-high performance liquid chromatography coupled to high resolution Orbitrap mass spectrometry was employed to analyze the FM2 extracts and study the fragmentation spectra of the analytes under investigation. The precursor ions of the neutral derivatives cannabidiphorol (CBDP) and Δ9-tetrahydrocannabiphorol (Δ9-THCP), 341.2486 for the [M-H]− and 343.2632 for the [M + H]+, showed an elution time of 19.4 min for CBDP and 21.3 min for Δ9-THCP (Fig. 1a). Their identification was confirmed by the injection of a mixture (5 ng/mL) of the two chemically synthesized CBDP and Δ9-THCP (Fig. 1b) as it will be described later. As for their carboxylated counterpart, the precursor ions of the neutral forms CBDP and Δ9-THCP break in the same way in ESI+ mode, but they show a different fragmentation pattern in ESI− mode. Whilst Δ9-THCP shows only the precursor ion [M-H]− (Fig. 1d), CBDP molecule generates the fragments at m/z 273.1858 corresponding to a retro Diels-Alder reaction, and 207.1381 corresponding to the resorcinyl moiety after the break of the bond with the terpenoid group (Fig. 1c). It is noteworthy that for both molecules, CBDP and Δ9-THCP, each fragment in both ionization modes differ exactly by an ethylene unit (CH2)2 from the corresponding five-termed homologs CBD and THC. Moreover, the longer elution time corroborates the hypothesis of the seven-termed phytocannabinoids considering the higher lipophilicity of the latter.
Isolation and characterization of natural CBDP and Δ9-THCP
In order to selectively obtain a cannabinoid-rich fraction of FM2, n-hexane was used to extract the raw material instead of ethanol, which carries other contaminants such as flavonoids and chlorophylls along with cannabinoids26. An additional dewaxing step at −20 °C for 48 h and removal of the precipitated wax was necessary to obtain a pure cannabinoids extract. Semi-preparative liquid chromatography with a C18 stationary phase allowed for the separation of 80 fractions, which were analyzed by LC-HRMS with the previously described method. In this way, the fractions containing predominantly cannabidiphorolic acid (CBDPA) and tetrahydrocannabipgorolic acid (THCPA) were separately subject to heating at 120 °C for 2 h in order to obtain their corresponding neutral counterparts CBDP and Δ9-THCP as clear oils with a >95% purity. The material obtained was sufficient for a full characterization by 1H and 13C NMR, circular dichroism (CD) and UV absorption.
Stereoselective synthesis of CBDP and Δ9-THCP
(-)-trans-Cannabidiphorol ((-)-trans-CBDP) and (-)-trans-Δ9-tetrahydrocannabiphorol ((-)-trans-Δ9-THCP) were stereoselectively synthesized as previously reported for the synthesis of (-)-trans-CBDB and (-)-trans-Δ9-THCB homologs11,12,24. Accordingly, (-)-trans-CBDP was prepared by condensation of 5-heptylbenzene-1,3-diol with (1 S,4 R)-1-methyl-4-(prop-1-en-2-yl)cycloex-2-enol, using pTSA as catalyst, for 90 min. Longer reaction time did not improve the yield of (-)-trans-CBDP because cyclization of (-)-trans-CBDP to (-)-trans-Δ9-THCP and then to (-)-trans-Δ8-THCP occurred. 5-heptylbenzene-1,3-diol was synthesized first as reported in the Supporting Information (Supplementary Fig. SI-1). The conversion of (-)-trans-CBDP to (-)-trans-Δ9-THCP using diverse Lewis’ acids, as already reported in the literature for the synthesis of the homolog Δ9-THC27,28,29, led to a complex mixture of isomers which resulted in an arduous and low-yield isolation of (-)-trans-Δ9-THCP by standard chromatographic techniques. Therefore, for the synthesis of (-)-trans-Δ9-THCP, its regioisomer (-)-trans-Δ8-THCP was synthesized first by condensation of 5-heptylbenzene-1,3-diol with (1 S,4 R)-1-methyl-4-(prop-1-en-2-yl)cycloex-2-enol, as described above, but the reaction was left stirring for 48 hours. Alternatively, (-)-trans-CBDP could be also quantitatively converted to (-)-trans-Δ8-THCP in the same conditions. Hydrochlorination of the Δ8 double bond of (-)-trans-Δ8-THCP, using ZnCl2 as catalyst, allowed to obtain (-)-trans-HCl-THCP, which was successively converted to (-)-trans-Δ9-THCP in 87% yield by selective elimination on position 2 of the terpene moiety using potassium t-amylate as base (Fig. 2a).
The chemical identification of synthetic (-)-trans-CBDP and (-)-trans-Δ9-THCP, and their unambiguous 1H and 13C assignments were achieved by NMR spectroscopy (Supplementary Table SI-1,2 and Supplementary Fig. SI-2,3). Since (-)-trans-CBDP and (-)-trans-Δ9-THCP differ from the respective homologs (CBD, CBDB, CBDV, Δ9-THC, Δ9-THCB and Δ9-THCV) solely for the length of the alkyl chain on the resorcinyl moiety, no significant differences in the proton chemical shifts of the terpene and aromatic moieties were observed for CBD and Δ9-THC homologs. The perfect match in the chemical shift of the terpene and aromatic moieties between the synthesized (-)-trans-CBDP and (-)-trans-Δ9-THCP and the respective homologues11,24,30, combined with the mass spectra and fragmentation pattern, allowed us to unambiguously confirm the chemical structures of the two new synthetic cannabinoids. The trans (1R,6R) configuration at the terpene moiety was confirmed by optical rotatory power. The new cannabinoids (-)-trans-CBDP and (-)-trans-Δ9-THCP showed an [α]D20 of −145° and −166°, respectively, in chloroform. The [α]D20 values were in line with those of the homologs11,31, suggesting a (1R,6R) configuration for both CBDP and Δ9-THCP. A perfect superimposition between the 1H (Fig. 2b,e) and 13C NMR spectra (Fig. 2c,f) and the circular dichroism absorption (Fig. 2d,g) of both synthetic and extracted (-)-trans-CBDP and (-)-trans-Δ9-THCP was observed, confirming the identity of the two new cannabinoids identified in the FM2 cannabis variety.
Binding affinity at human CB1 and CB2 receptors
The binding affinity of (-)-trans-Δ9-THCP against purified human CB1 and CB2 receptors was determined in a radioligand binding assay, using [3H]CP55940 or [3H]WIN 55212-2 as reference compounds, and dose-response curves were determined (Fig. 3a,b). (-)-trans-Δ9-THCP binds with high affinity to both human CB1 and CB2 receptors with a Ki of 1.2 and 6.2 nM, respectively. (-)-trans-Δ9-THCP resulted 33-times more active than (-)-trans-Δ9-THC (Ki = 40 nM), 63-times more active than (-)-trans-Δ9-THCV (Ki = 75.4 nM) and 13-times more active than the newly discovered (-)-trans-Δ9-THCB (Ki = 15 nM) against CB1 receptor12,14. Moreover, the new identified (-)-trans-Δ9-THCP resulted about 5- to 10-times more active against CB2 receptor (Ki = 6.2 nM), in contrast with (-)-trans-Δ9-THC, (-)-trans-Δ9-THCB and (-)-trans-Δ9-THCV, which instead showed a comparable binding affinity with a Ki ranging from 36 to 63 nM (Fig. 3a)12,14.
The highest activity of (-)-trans-Δ9-THCP, compared to the shorter homologues, was investigated by docking calculation. The X-ray structure of the active conformation of hCB1 receptor in complex with the agonist AM11542 (PDB ID: 5XRA) was used as reference for docking since marked structural changes in the orthosteric ligand-binding site are observed in comparison with the conformation of the receptor bound to an antagonist32,33. AM11542 is a synthetic Δ8 cannabinoid with high affinity against hCB1 receptor (Ki = 0.11 nM) possessing a 7′-bromo-1′,1′-dimethyl-heptyl aliphatic chain at C3 of the resorcinyl moiety. As expected, due to the close chemical similarity, the predicted binding mode of (-)-trans-Δ9-THCP (Fig. 3c) reflected that of AM11542 in the CB1 crystal structure (Fig. SI-6a,b)18. (-)-trans-Δ9-THCP bound in the active conformation of CB1 in an L-shaped pose. The tetrahydro-6H-benzo[c]chromene ring system is located within the main hydrophobic pocket delimited by Phe174, Phe177, Phe189, Lys193, Pro269, Phe170, and Phe268. In particular, the aromatic ring of the resorcinyl moiety is involved in two edge-to-face π-π interactions with Phe170 and Phe268, whereas the phenolic hydroxyl group at C1 is engaged in a H-bond with Ser383 (Fig. 3c). Interestingly, the heptyl chain at C3 extended into a long hydrophobic tunnel formed by Leu193, Val196, Tyr275, Iso271, Leu276, Trp279, Leu359, Phe379, and Met363 (Fig. 3c,d). Because the predicted pose of the tricyclic tetrahydrocannabinol ring system is conserved among the four THC homologs (Supplementary Fig. SI-7a–c), the length of the alkyl chain at C3 of the resorcinyl moiety could account for the different binding affinity observed among the four cannabinoids. (-)-trans-Δ9-THCP (Fig. 3c) and (-)-trans-Δ9-THC (Supplementary Fig. SI-7a) share the same positioning of the alkyl ‘tail’ within the hydrophobic channel12,33,34. However, the long heptyl chain of Δ9-THCP is able to extend into the tunnel along its entire length, maximizing the hydrophobic interactions with the residues of the side channel. In contrast, the tunnel is only partially occupied by the shorter pentyl chain of (-)-trans-Δ9-THC, accounting for the higher affinity of Δ9-THCP (Ki = 1.2 nM) compared to Δ9-THC (Ki = 40 nM). A different positioning of the ‘tail’ was instead predicted for the shorter alkyl chain homologues, Δ9-THCV and Δ9-THCB. The propyl and butyl chain of Δ9-THCV and Δ9-THCB, respectively, are too short to effectively extend within the hydrophobic channel. As stated in our previous work12, these shorter chains accommodate within a small hydrophobic pocket delimitated by Phe200, Leu359, and Met363 (Supplementary Fig. SI–7b,c). This side pocket is located at the insertion between the main hydrophobic pocket and the long channel (Fig. 3d) and seems to accommodate small hydrophobic substituents (i.e. gem-dimethyl or cycloalkyl) introduced at C1′ position of the side chain of several synthetic cannabinoids, rationalizing the notable enhancement in potency and affinity for these derivatives35,36,37,38,39.
In vivo determination of the cannabinoid profile of Δ9-THCP
The cannabinoid activity of Δ9-THCP was evaluated by the tetrad of behavioural tests on mice. The tetrad includes the assessment of spontaneous activity, immobility index (catalepsy), analgesia and changes in rectal temperature. Decrease of locomotor activity, catalepsy, analgesia and hypothermia are well-known signs of physiological manifestations of cannabinoid activity40. After intraperitoneal (i.p.) administration, Δ9-THCP at 2.5 mg/kg markedly reduced the spontaneous activity of mice in the open field, while at 5 and 10 mg/kg it induced catalepsy on the ring with the immobility as compared to the vehicle treated mice (Fig. 4b,c) (0 mg/kg: 6888 cm ± 474.8, 10 mg/kg: 166.8 cm ± 20.50, 5 mg/kg: 127.5 cm ± 31.32, 2.5 mg/kg: 4072 cm ± 350.8, p = 0.0009). Moreover, Δ9-THCP administration induced a significant increase, at 10 and 5 mg/kg, in the latency for moving from the catalepsy bar (Fig. 4e) (0 mg/kg: 15.20 sec ± 4.33, 10 mg/kg: 484.5 sec ± 51.58, 5 mg/kg: 493.4 sec ± 35.68, 2.5 mg/kg: 346.1 sec ± 35.24, p = 0.0051). In the hot plate test (Fig. 4f), Δ9-THCP (10 and 5 mg/kg) induced antinociceptive effect, whereas at 2.5 mg/kg there was a trend in the induction of antinociception, which resulted not statistically significant as compared to the vehicle treated mice (0 mg/kg: 19.20 sec ± 2.65, 10 mg/kg: 57.0 sec ± 2.0, 5 mg/kg: 54.38 sec ± 2.86, 2.5 mg/kg: 40.22 sec ± 5.8, p = 0.0044). Δ9-THCP administration induced a dose dependent significant decrease, only at 10 mg/kg, in body temperature as compared to vehicle (0 mg/kg: 0.40 °C ± 0.25, 10 mg/kg: −7.10 °C ± 0.43, 5 mg/kg: −5.28 °C ± 0.36, 2.5 mg/kg: −4,12 °C ± 0.38, p = 0.0009) (Fig. 4d).
Subscribe to the newsletter news
We hate SPAM and promise to keep your email address safe