INTRODUCTION
Cannabinoids: Discovery and Pharmacology
Cannabis sativa L is one of the earliest plants utilized by humans for fiber, food, and medicinal purposes, as well as in social and religious rituals Evidence of its medicinal use dates back to China around 2600 BC, where it was recommended for various ailments, including malaria and rheumatic pains In Assyria, circa 800 BC, cannabis was recognized as a significant drug, referred to as "the drug that takes away the mind." Over time, its use expanded to India, the Middle East, South Africa, and South America, primarily for pain relief and neurological disorders By the 19th century, cannabis gained popularity as a mainstream medicine in Western Europe, particularly in England, while in France, it was more commonly associated with recreational use.
Research into the active components of Cannabis sativa began in the late 19th century, but progress was hindered by the complexity of its compounds and the limitations of early separation techniques In 1899, Wood and colleagues isolated the first component, cannabinol (C21H26O2), but its structure was only partially determined by Cahn in 1932, and it was not identified as the primary active ingredient In the early 1940s, both the Todd group in the UK and the Adams group in the USA synthesized various cannabinol isomers based on Cahn's findings, focusing on isolating active natural constituents They correctly elucidated the structure of cannabinol, isolated cannabidiol—another inactive component—and unexpectedly discovered that racemic Δ6a,10a-tetrahydrocannabinol (synhexyl) exhibited activity in animal tests.
Figure 1 Cannabis Sativa L., Marinol ® , some natural cannabinoids and the synthetic cannabinoid nabilone
In the early 1960s, advancements in chromatography and NMR spectroscopy led to the establishment of the correct structure and stereochemistry of cannabidiol The isolation and elucidation of the major psychoactive component of Cannabis sativa, Δ9-tetrahydrocannabinol (Δ9-THC), were achieved by Gaoni and Mechoulam in 1964, followed by the synthesis of the natural (–)-Δ9-THC enantiomer in 1967 Since then, research into cannabinoids' chemistry, pharmacology, metabolism, and clinical implications has significantly expanded, resulting in the identification of over 480 natural components in the cannabis plant, including 70 classified as cannabinoids The term "cannabinoid," originally defined as C21 compounds found in Cannabis sativa, has now been broadened to include synthetic analogs and endogenous cannabinoids, reflecting the extensive growth of cannabinoid research with more than 10,000 publications on the subject.
The clinical use of cannabinoids has evolved since the 1980s, beginning with 9 -THC (Dronabinol) and nabilone (Cesamet) for nausea and vomiting relief in chemotherapy patients In 1992, Marinol was introduced to stimulate appetite in AIDS patients, followed by the 2005 approval of cannabidiol in a 1:1 mixture with 9 -THC (Nabiximols) for neuropathic pain relief in multiple sclerosis and cancer patients Despite their therapeutic benefits, the medical use of cannabinoids is limited by significant adverse effects on the respiratory, digestive, urinary, and central nervous systems, as well as risks of addiction, aggression, anxiety, sedation, depression, and suicide.
Raphael Mechoulam, a pioneer in cannabinoid research, supports the idea of marijuana as a medical option, stating, "My answer is 'yes'." However, he emphasizes the importance of regulating its use, similar to other potent drugs.
Cannabinoid Receptors
Cannabinoids were initially thought to interact non-specifically with membrane lipids due to their lipophilic nature, supported by evidence linking their ability to alter artificial membranes to their psychoactive effects In the mid-1980s, Howlett and colleagues demonstrated that psychotropic cannabinoids inhibit adenylate cyclase via pertussis toxin-sensitive Gi/o proteins, aligning with the established mechanism of G-protein-coupled receptors In 1988, Devane identified high-affinity binding sites for the radiolabeled cannabinoid [3H]-CP-55,940 in rat brain membranes, providing strong evidence that cannabinoids interact with a G-protein-coupled receptor, as their ability to displace [3H]-CP-55,940 correlates with their analgesic effects in vivo.
The discovery of cannabinoid receptors began with the cloning of the rat CB1 receptor by Matsuda in 1990, followed by the human CB1 receptor cloned by Gerard in 1991, and the CB2 receptor by Munro in 1993 Currently, two main types of cannabinoid receptors, CB1 and CB2, have been identified and characterized based on their tissue distribution, amino acid sequences, signaling mechanisms, and structural differences.
Ligands require specific criteria for activation, and research indicates the potential existence of additional non-CB1/CB2 receptors like GPR18, GPR55, and GPR119, as demonstrated in experiments involving CB1 and CB2 knockout mice However, to date, there have been no reports confirming the cloning of these receptors.
CB1 and CB2 receptors are essential membrane proteins belonging to the class-A (rhodopsin-like) G-protein coupled receptors (GPCRs) They consist of seven transmembrane α-helices, linked by alternating intracellular and extracellular loops, along with an extracellular N-terminus and an intracellular C-terminus.
Figure 3 illustrates helical-net representations of the human CB1 and CB2 sequences, with the initial amino acids from the N termini and the final amino acid from the C termini omitted for clarity (Onaivi et al., 2006).
The human CB1 and CB2 receptors consist of 472 and 360 amino acids, respectively, sharing 44% overall amino acid sequence homology and 68% homology within their transmembrane domains Research using autoradiography and positron emission tomography has shown that CB1 receptors are primarily located in the brain, with the highest density found in the hippocampus, cerebellum, and striatum, which aligns with the effects of cannabinoids on cognitive and motor functions Additionally, CB1 receptors are present in various peripheral tissues outside the central nervous system, including the gastrointestinal tract, pancreas, liver, kidneys, prostate, testis, uterus, eyes, lungs, and adipose tissue.
The relationship between tissue and heart health involves key factors such as energy balance, metabolism, nociception, and cardiovascular well-being While CB2 cannabinoid receptors are primarily found in the peripheral immune system, they can also be present in the central nervous system, specifically in perivascular microglial cells, as well as in peripheral tissues during inflammatory responses.
Activation of cannabinoid receptors inhibits adenylate cyclase, leading to decreased levels of cyclic adenosine monophosphate (cAMP) from adenosine triphosphate (ATP) and the activation of mitogen-activated protein (MAP) kinases, which are crucial for regulating cell growth, division, differentiation, and apoptosis The modulation of cAMP levels influences the phosphorylation of key enzymes and proteins, potentially linking to the physiological and psychological effects of cannabinoids, although the exact mechanisms remain unclear While CB2 receptor activation does not affect ion channel function, CB1 receptors influence several ion channels by stimulating potassium channels and inhibiting N- and P/Q-type calcium channels, which are vital for neurotransmission Additionally, cannabinoid receptors can interact with various other receptor systems, including opioid, vanilloid TRPV1, serotonin (5-HT)3, N-methyl-D-aspartate (NMDA), and nicotinic acetylcholine receptors.
Figure 4 Signalling-transduction of cannabinoid receptor, Rukwied et al 2005 41b
Understanding the interactions between ligands and target proteins is crucial for drug discovery and design Structure-based drug design typically relies on experimental structural data obtained from techniques like NMR or X-ray crystallography However, analyzing GPCRs such as CB1 and CB2 is challenging due to their conformational heterogeneity and rapid denaturation outside of membrane environments Consequently, while 1H NMR spectra of the CB1 receptor have been reported, crystal structures for both CB1 and CB2 remain elusive Therefore, developing specially designed molecules to investigate structure-activity relationships (SAR) is essential for uncovering the structural requirements for ligand-receptor interactions.
Bioassay Techniques
Current quantitative behavioral assays for assessing the modes of action of CNS drugs, particularly cannabinoids, include successful methods such as 'static ataxia' in dogs, 'overt behavior' in monkeys, and the recent 'mouse tetrad.' Additionally, in vitro assays have been established for cannabinoids, such as radioligand binding assays, cAMP production inhibition, [35S]GTPγS binding assays, and inhibition of electrically evoked contractions in isolated smooth muscle This section of the dissertation focuses on the radioligand binding assay used for receptor-ligand binding affinity measurements and introduces the novel ligand-assisted protein structure (LAPS) technique Some compounds prepared in our studies were aimed at supporting LAPS investigations of CB1 and CB2 receptors.
Competitive binding assays are widely used to assess the binding affinity between a ligand and its receptor binding sites
Figure 5 Tritium labelled non-classical cannabinoids used in binding assays
The novel ligand or inhibitor is evaluated for its competitive binding ability against radiolabeled ligands, including [3H]-CP-55,940, [3H]-WIN-55212-2, and [3H]-SR-141716A This assessment is conducted across various concentrations in a buffered solution that incorporates artificial membranes or tissues known to express either CB1 or CB2 receptors.
10 of the inhibitor at which 50% of the substrate is displaced is determined to be the IC50 value for the inhibitor
Figure 6 Example of competitive radioligand binding assay: displacement of [ 3 H]-SR-141716A binding to CB1 receptor in mouse brain homogenates by AM-2233 and WIN-55212-2, Deng, H et al 2005 49
To accurately describe the binding affinity of a ligand to its receptors, independent of the radioligand concentration used in assays, the absolute inhibition constant (Ki) is calculated using the Cheng-Prusoff equation: Ki = IC50 / (1 + [L]/KD) In this equation, [L] represents the fixed concentration of radioligand, while KD denotes the dissociation constant, indicating the concentration of radioligand that achieves half-maximal receptor activation It is important to note that this equation is applicable only when the inhibitor and substrate bind to the receptor competitively.
Ligand-assisted protein structure (LAPS) is an innovative experimental technique developed by Dr Makriyannis's group to identify key amino acid residues involved in ligand-receptor interactions This method is particularly beneficial for studying membrane-bound receptors such as CB1 and CB2, which cannot be crystallized with their ligands The LAPS experiment necessitates high-affinity ligands that are chemically modified to covalently bond with specific amino acids in the receptor's binding domain, facilitating the analysis of these crucial interactions.
The receptor-ligand covalent complex is analyzed and sequenced through mass spectrometry to identify interaction sites between the ligand and specific amino acids By utilizing the known primary amino acid sequence, the receptor pocket's location can be inferred Site-directed protein mutations provide further data to confirm the binding site The insights gained from these experiments enable the construction of computer models of the ligand-protein complex, laying the groundwork for rational drug design.
Tricylic Cannabinoids and Structure Activity Relationships
Two different numbering systems, dibenzopyran and monoterpenoid, are generally used for cannabinoids 53 These are shown in Figure 7
Figure 7 Numbering systems in ()- 9 -THC
In this dissertation, the dibenzopyran ring nomenclature will be used for tricylic cannabinoids
The structural classification of cannabinoids includes: (i) classical cannabinoids; (ii) nonclassical cannabinoids; (iii) hybrid cannabinoids; (iv) arachidonic acid based endocannabinoids and their analogs; (v) diarylpyrazole compounds, aminoalkylindole compounds, and other cannabinoids 54
Figure 8 Pharmacophores in ()- 9 -THC and ()-CP-55,940 (Note: h = human, r = rat)
Classical cannabinoids are natural or synthetic cannabinoids structurally related to ()-
9-THC is a class of compounds known as ABC-tricyclic terpenoids, characterized by the presence of a benzopyran moiety The pharmacological activity of cannabinoids is primarily attributed to three key pharmacophores: the northern aliphatic group (NAG), the phenolic hydroxyl (PH), and the lipophilic side chain (SC).
Nonclassical cannabinoids, developed by Pfizer, aim to simplify the structure of classical cannabinoids while enhancing their biological activity These unique ligands are identified by their AC bicyclic and ACD tricyclic structures, exemplified by compounds such as (−)-CP-55,940 and (−)-CP-55,244.
10) structures lacking the pyran B-ring of classical cannabinoids, but containing an additional pharmacophore: the southern aliphatic hydroxyl group (SAH), which is trans to the aromatic ring ()-CP-55,940 is a cannabinoid agonist that is considerably more potent than ()- 9 -THC in both behavioral tests and receptor binding assays 55
Figure 9 Some cannabinoids that illustrate their structural classification (Note: m = mouse)
Hybrid cannabinoids, first introduced by Makriyannis and Tius in 1994, blend the properties of classical and nonclassical cannabinoids, such as AM-4030 A notable feature of these hybrids is that the C-6 equatorial β-hydroxyethyl analog exhibits a higher affinity for the CB1 receptor compared to its α-axial epimer.
In 1992, Mechoulam first identified endogenous cannabinoid ligands, notably N-arachidoylethanolamine (anandamide), derived from porcine brain These endogenous ligands and their synthetic analogs consist of two main components: a polar head group, such as the ethanolamido found in anandamide, and a hydrophobic arachidonyl chain characterized by four non-conjugated cis double bonds.
Recent research has expanded the classification of cannabinoids to include structurally distinct compounds such as diarylpyrazole, aminoalkyl indole, diarylsulfonyl ester, and diarylmethyleneazetidine compounds Notably, a novel cannabinergic chemotype has been developed by replacing the C-ring in the classical THC structure with a hydrolyzable seven-membered lactone (AM-4809) This compound can be enzymatically hydrolyzed by plasma esterases, resulting in an inactive acid metabolite at CB receptors The design of AM-4809 aims to create a short-acting cannabinoid that becomes inactive after fulfilling its biological function, thereby minimizing the undesirable side effects associated with long-acting cannabinoids.
1.4.3 Structure Activity Relationships of Tricyclic Cannabinoids
In 1971, Mechoulam discovered that hexahydrocannabinol, characterized by a fully saturated C-ring and an equatorial methyl group at C9, exhibits similar psychoactivity to tetrahydrocannabinol (THC) in rhesus monkeys, while its axial epimer is significantly less active Subsequent studies by Wilson and May in 1975 and 1976 revealed that adding a hydroxyl group at the northern aliphatic position enhances potency in analgesia tests in mice, with certain compounds showing nearly five times the activity of both (–)-Δ8-THC and (–)-Δ9-THC Notably, a compound with an equatorial hydroxyl at C9 demonstrated strong analgesic properties in rodents, whereas its axial counterpart was inactive.
Figure 10 Cannabinoids illustrating the NAG
The equatorial northern aliphatic hydroxyl group is a key structural characteristic found in high-potency cannabinoids, encompassing both classical examples like AM-906 and non-classical variants such as CP-55,940 and levonantradol.
Introducing a carbonyl group at the C9 position, such as in nabilone, enhances cannabinoid activity Additionally, the (6aR, 10aR) stereochemistry is favored for optimal cannabinoid effects For example, HU-210 is recognized as one of the most potent classical cannabinoids, whereas its inactive enantiomer, HU-211, with (6aS, 10aS) stereochemistry, shows no activity with CB1 and CB2 receptors.
The phenolic hydroxyl group at C1 is crucial for maintaining CB1 affinity, as its substitution with a methoxy, hydrogen, or fluorine atom significantly reduces this affinity, while having a lesser impact on CB2 This understanding is fundamental for the development of CB2 selective cannabinoids.
Figure 11 Cannabinoids illustrating the PH
The C3 aliphatic side chain is a key pharmacophore in tricyclic cannabinoids, with significant research dating back to 1947-1948 by Adam and colleagues They found that cannabinoids featuring 1',1'-dimethylheptyl, 1'-methyloctyl, and 1',2'-dimethylheptyl side chains exhibit optimal affinity, notably 3-(1',1'-dimethylheptyl)-Δ6a,10a-tetrahydrocannabinol being 20 times more potent than its n-pentyl counterpart, while a mixture of isomeric 3-(1',2'-dimethylheptyl)-Δ6a,10a-tetrahydrocannabinols is 512 times more effective Among the various isomers, the (1'S,2'R) configuration stands out.
(37) and (1'R,2'S) are considerably more potent than the other isomers 74 Although the 3-(1',2'-
The 3-(1',1'-dimethylheptyl) analogs of cannabinoids have been extensively studied due to the availability of their precursor, 1,3-dimethoxy-5-(1,1-dimethylheptyl)benzene, which simplifies synthesis without the need for stereochemical control Among these, 3-(1',1'-dimethylheptyl)-Δ8-THC has demonstrated a high in vivo affinity, comparable to the most potent isomers The incorporation of the dimethylheptyl side chain has led to the development of highly potent cannabinoids like HU-210, JWH-051, CP-55940, and nabilone, all of which show nanomolar affinities for cannabinoid receptors Additionally, the length of the side chain influences CB1/CB2 selectivity, with shorter chains (ethyl, propyl, butyl) favoring CB2 selectivity, while longer chains (seven or eight carbons) tend to prefer CB1 For instance, various Δ8-THC analogs exhibit increasing CB1 affinity as the side chain lengthens, with 1',1'-dimethylheptyl-Δ8-THC showing a Ki value of 0.77 nM at CB1 Furthermore, JWH-133 exhibits high affinity and improved CB2 selectivity compared to its C1-deoxy analogs with longer side chains.
Figure 12 Cannabinoids illustrating the SC
A study investigating the stereochemical requirements of cannabinoid side chains revealed that introducing a double or triple bond to the C1'-C2' bond significantly enhances affinity for the CB1 receptor Cannabinoids featuring these unsaturated bonds, such as compounds 25, 40, and 41, demonstrate a stronger binding affinity compared to those with a saturated C1'-C2' bond, particularly the cis-alk-1-ene AM-906 (25), which exhibits superior affinity.
CB1/CB2 selectivity than other analogs 66
Figure 13 Cannabinoids illustrating the SC
Modifying the position of the double bond by adding a 1'-methylene group to the heptyl side chain maintains high affinity but limits selectivity for CB1 receptors Conversely, introducing a ketone or hydroxy group at C1' significantly reduces affinities for both CB1 and CB2 receptors These findings suggest that a hydrophobic group at the benzylic position enhances affinities for cannabinoid receptors Supporting this trend, the 1',1'-dithiolane analog (AMG-3) shows high affinities for both CB1 and CB2 receptors, with Ki values in the subnanomolar range.
Figure 14 Cannabinoids illustrating the SC
Research on various ring sizes of the C1'-cycloalkyl side chain revealed that C1'-cyclopropyl and C1'-cyclopentyl structures are optimal for binding to both CB1 and CB2 receptors The C1'-cyclobutyl variant demonstrated strong CB1 affinity and superior CB1/CB2 selectivity compared to smaller rings, while the C1'-cyclohexyl showed diminished affinity for both receptors This structural innovation has been utilized by the Makriyannis group in the development of AM-2389, a highly effective CB1 agonist known for its prolonged action in both in vitro and in vivo studies.
Earlier Synthesis Approaches Towards Tricyclic Cannabinoids
The early 1940s marked the beginning of cannabinoid synthesis, highlighted by Rodger Adams' laboratories where cannabinol and its isomers were first synthesized.
In 1967, Raphael Mechoulam reported the first stereospecific synthesis of cannabinoids, notably (–)-Δ9-THC, the primary psychoactive component of Cannabis sativa, along with its isomer (–)-Δ8-THC The structures of these tricyclic cannabinoids, including Δ9-THC and Δ8-THC, consist of an aromatic section and an alicyclic section, which were initially synthesized by condensing olivetol with a monoterpene, such as verbenol.
Figure 17 General structure of a classical tetrahydrocannabinoid, Razdan et al 1981 93
The Mechoulam synthesis is notable for utilizing the bulky dimethylmethylene bridge of verbenol, which ensures stereochemical control and results in exclusively trans products This method benefits from the availability of optically pure α-pinene in both (+) and (−) forms, allowing for the production of both natural (−) and unnatural (+) series In his synthesis, (−)-cis and (−)-trans-verbenols were condensed with olivetol in the presence of BF3·Et2O, yielding (−)-Δ8-THC with a 44% yield, though purification proved challenging Alternatively, using p-TsOH led to the formation of byproducts alongside the major product, but ultimately, this method was more efficient, producing (−)-Δ8-THC in an 80% yield after treatment with BF3·Et2O.
The reactions of the same product suggest a mechanism involving the same allylic cation In this method, (−)-Δ9-THC (1) is synthesized through hydrochlorination followed by dehydrochlorination of (−)-Δ8-THC (20).
Reagents and conditions: (a) BF3Et2O, CH2Cl2, rt, 44%; (b) p-TsOH, CH2Cl2, 60 (60%), 61
(15%), 62 (11%) ; (c) BF3Et2O, CH2Cl2, rt, 80%; (d) HCl, ZnCl2 cat, toluene, -15 o C; (e) NaH, THF, reflux
Scheme 1 Synthesis of (–)- 9 -THC and ()- 8 -THC by Mechoulam, Mechoulam et al 1967 16b
In 1967, Petrzilka achieved the total synthesis of the natural (−) series from (+)-cis/trans-p-mentha-2,8-dien-1-ol, establishing a widely adopted and scalable method for producing tricyclic cannabinoids Notably, the use of strong protic acids during the condensation and cyclization reactions leads to the formation of the thermodynamically stable (−)-Δ8-THC, necessitating further steps to obtain the desired (−)-cannabinoid.
Scheme 2 Synthesis of (–)- 9 -THC and ()- 8 -THC by Petrzilka, Petrzilka et al 1967 94
Based on the same principle, various optically active monoterpenes (Figure 18) including (+)- cis-chrysanthenol, 95a (+)-trans-2-carene epoxide, 95b and (+)-(1R, 4R)-p-menth-2-ene-1,8-diol 96 have been used to synthesize optically active natural THC's
Figure 18 Some optically active monoterpenes used to prepare THC
In 1977, Archer and coworkers developed the total synthesis of optically active 9- ketocannabinoid nabilone 5 by condensation of resorcinol with the mixture of diacetates 67a and
In Scheme 3, the formation of an undesired abnormal cannabinoid byproduct was notably absent, likely due to the steric hindrance caused by the bulky 1',1'-dimethylheptyl group Although the undesired cis-ketone 70 was produced from treatment 69 using p-TsOH·H2O, it could be isomerized to trans-5 through treatment with anhydrous AlCl3 in CH2Cl2.
Reagents and conditions: (a) isopropenyl acetate, p-TsOHH2O, reflux; (b) Pb(OAc)4, benzene, reflux, 39% over 2 steps; (c) p-TsOHH2O, CHCl3, rt, 70%; (d) SnCl4, CHCl3, rt, 84%; (e) p-
TsOHH2O, CHCl3, reflux, 61% (and 31% of compound 5); (f) AlCl3, CH2Cl2, rt, 70%
Scheme 3 Total synthesis of nabilone, Archer et al 1977 97
The Tius group has successfully utilized a specific approach for constructing the tricyclic ring system in hexahydrocannabinoid synthesis Various methods have been investigated, including the Michael addition of the resorcinol fragment to apoverbenone and the intramolecular hetero-Diels–Alder cycloaddition Our recent research has concentrated on modifying cannabinoid ligand pharmacophores, such as probing the NAG with halogens and C9-methyl carboxylate esters, as well as exploring substituents at the C6 position in hybrid cannabinoids Notably, the SC has been examined for unsaturation and functionalization with groups like NCS, N3, and halogens, as well as substitutions with bulky groups, including 1',1'-dimethylalkyl and various adamantyl derivatives.
THE TOTAL SYNTHESIS OF C1'-AZACYCLOALKYL 9-HYDROXY
Introduction
The aliphatic side chain of tricyclic cannabinoids is crucial for their affinity to cannabinoid receptors and overall pharmacological potency Cannabinoids featuring a C1'-tert-alkyl side chain exhibit strong binding to both CB1 and CB2 receptors Recent research by the Makriyannis group indicates that smaller ring sizes (three to five members) in the C1'-cycloalkyl side chain enhance receptor binding affinity, with cannabinoids containing a C1'-cyclobutyl side chain demonstrating notable selectivity for CB1 and CB2 receptors.
To enhance the water solubility of tricyclic cannabinoids and investigate receptor interactions, heteroatoms that facilitate hydrogen bonding are introduced into the hydrophobic side chain This section of the dissertation focuses on developing a series of C1'-azacycloalkyl cannabinoids, incorporating both four- and five-membered rings, along with their ammonium salts These tricyclic cannabinoids are characterized by an equatorial 9-hydroxy group, (6aR, 10aR) absolute stereochemistry, a phenolic hydroxy group, and a 1',1'-disubstituted n-heptyl side chain, all essential for achieving high affinity for CB1 and CB2 receptors.
Figure 20 General structure of C1'-azacycloalkyl cannabinoids and intermediate triflate 78
This article discusses the synthesis of hexahydrocannabinoids from a common precursor, triflate 78, derived from (−)-β-pinene using a method developed by Dr Darryl Dixon It details the non-diastereoselective synthesis of 2,2-disubstituted pyrrolidine, the synthesis of 3,3-disubstituted azetidine, and the diastereoselective synthesis of 2,2-disubstituted azetidine cannabinoids, along with their evaluation in receptor binding assays.
Synthesis of Advanced Intermediate Triflate
Aryl triflate 78 was obtained from commercially available ()- -pinene 71 in acceptable yields in 10 steps, following the procedure that has been developed in our group (Scheme 4)
Reagents and conditions: (a) i O3, CH3OH/CH2Cl2, -78 o C, ii Me2S, -78 o C to rt, 89%; (b) isopropenyl acetate, p-TsOHH2O, reflux, 6 h; (c) Pb(OAc)4, benzene, reflux, 2.5 h, 71% from
65; (d) p-TsOHH2O, CHCl3/(CH3)2CO, 0 o C to rt, 2 h, rt, 4 h; (e) Ac2O, Pyr, DMAP, CH2Cl2, rt,
12 h, 67% from 67; (f) KOH, CH3OH, 0 o C, 2 h, 96%; (g) TMSOTf, CH3NO2, 0 o C, 2.5 h, 83%; (h) PhNTf2, Et3N, CH2Cl2, rt, 14 h, 84%; (i) NaBH4, CH3OH, -5 o C, 1 h, 90%; (j) CH3OCH2Cl, i-
Pr2NEt, CH2Cl2, 0 o C, 45 min, rt, 2 h, 89%
Scheme 4 Synthesis of intermediate triflate 78, Dixon et al 2010 85
The total synthesis of the tricyclic cannabinoid nucleus begins with the condensation of phloroglucinol derivative 72 and a mixture of optically active terpene-derived diacetates 67 This process is followed by rearrangement-cyclization to achieve the desired structure To ensure the (6aR, 10aR) absolute configuration, the chiral diacetates 67 were synthesized from (–)-β-pinene through ozonolysis to produce nopinone 65, followed by enol acetylation to yield 66, and subsequent oxidation with lead (IV) acetate to form diacetates 67a and 67b.
The condensation step in the synthesis of nabilone, as described by Archer and colleagues, effectively utilizes chloroform as the solvent However, due to the limited solubility of phloroglucinol in chloroform, alternative solvents yielded poor results To address this, previous team members introduced a modification using acetone/chloroform and masking the phenolic hydroxyl groups with trimethylsilyl ethers This approach involved generating persilylated phloroglucinol in situ and condensing it with diacetates in the modified solvent, resulting in the formation of bicyclic compound 73 Although separating 73 from unreacted phloroglucinol proved challenging, peracetylation of the crude mixture yielded pure triacetate 74 with approximately 70% yield Hydrolysis of 74 with KOH in methanol successfully regenerated pure 73, and subsequent rearrangement-cyclization of 73 using TMSOTf in nitromethane produced tricyclic ketocannabinoid 75 in about 80% yield Finally, treatment of 75 with N-phenyltriflimide led to the regioselective formation of monotriflate 76.
9β-hydroxyl cannabinoids exhibit greater affinities for CB1 and CB2 receptors compared to their 9α-hydroxyl diastereomers To synthesize the equatorial alcohol 77, ketone 76 was reduced using sodium borohydride This was followed by the methoxymethylation of both the phenolic and aliphatic hydroxyl groups, resulting in the production of intermediate triflate 78 in high yields For a comprehensive understanding of the total synthesis of intermediate triflate 78, refer to Dixon et al., 2010.
2.3 Non-diastereoselective Synthesis of 2,2-Disubstituted Pyrrolidine Cannabinoids
The synthesis of 2,2-disubstituted pyrrolidine cannabinoids was initiated from the common intermediate triflate 78 The introduction of the pyrrolidine ring was achieved through desulfitative carbon–carbon cross-coupling of pyrrolidine-2-thione with arylboronic acid 80, which was derived from triflate 78 using Miyaura borylation followed by the hydrolysis of arylboronic acid pinacol ester Additionally, the n-hexyl group of compound 83 was incorporated via nucleophilic addition to a cyclic imine.
81 The details of these reactions will be discussed in what follows
Scheme 5 Retrosynthesis of 2,2-disubstituted pyrrolidine HHC
The details of the synthesis of 2,2-disubstituted pyrrolidine cannabinoids are summarized in Scheme 6
The synthesis process involves several key steps and conditions: (a) AcOK and PdCl2(dppf) in DMF at 90°C for 3.5 hours yields an 84% product; (b) NaIO4 and NH4OAc in a mixture of acetone and water at room temperature for 20 hours; (c) CuTC with Pd2(dba)3·CHCl3 and PPh3 in THF under microwave irradiation at 100°C for 2 hours results in a 62% yield from 79; (d) a two-step reaction involving LiCl in THF at room temperature for 30 minutes followed by n-C6H13Li from -10°C to room temperature for 2 hours achieves a 92% yield; (e) Sc(OTf)3 with 1,3-propanediol in acetonitrile under reflux for 48 hours gives a 91% yield; (f) finally, a reaction with (HCHO)n and Ti(OiPr)4 in diglyme at 60°C for 30 minutes followed by room temperature.
Scheme 6 Synthesis of 2,2-disubstituted pyrrolidine cannabinoids
The palladium-catalyzed coupling of aryl triflate 78 with bis(pinacolato)diboron yielded aryl boronic ester 79 in 84% efficiency, utilizing 4 mol % of PdCl2(dppf) as a catalyst, known for its effectiveness in borylation reactions involving aryl halogenates and triflates Pd(PPh3)4 was excluded due to previous findings indicating the formation of an undesired byproduct when coupled with phenyl groups Notably, the undesired biaryl byproduct was not observed in the Suzuki cross-coupling of aryl boronic ester 79 and triflate 78 The choice of base is crucial for the Miyaura borylation reaction's success, with KOAc identified as the optimal base for achieving high yields, as it enhances the transmetalation step through the high reactivity of the Pd-O bond.
The acetoxopalladium (II) intermediate features a bond characterized by the interaction of a soft acid and a hard base, driven by the boron atom's high affinity for acetate This interaction also helps prevent the formation of undesirable biaryl byproducts Utilizing stronger bases like K3PO4 enhances this process.
K2CO3 has been shown to facilitate undesired coupling reactions, leading to biaryl byproducts by activating the boron atom in aryl boronic ester products As a result, K2CO3 is not utilized for this reaction.
Model studies on the cross-coupling reactions involving phenylboronic acid, pinacol phenylboronate, and potassium trifluoro(phenyl)borate with pyrrolidine-2-thione indicate that only phenylboronic acid yields the desired product, implying that the hydrolysis of boronic ester 79 to boronic acid 80 is necessary However, the hydrolysis of boronic ester 79 proved to be unexpectedly difficult, as traditional methods for hydrolyzing pinacol arylboronates under strong acidic conditions, such as HCl and BBr3, were not effective.
Transesterification with diethanolamine and acidic hydrolysis were avoided due to the sensitivity of methoxymethyl ether groups and the benzopyran ring in compounds 79 and 80 to acidic conditions A different approach successfully converted boronic ester 79 to potassium trifluoroborate 85 using potassium bifluoride; however, the purification of 85 was unsatisfactory, preventing confirmation of its structural assignment.
In situ hydrolysis of compound 85 using bases like K2CO3 and LiOH, or fluorophiles such as TMSCl and water-silica gel, failed to yield a pure sample of boronic acid 80.
Reagents and conditions: (a) KHF2, CH3OH/H2O, rt, 30 min; (b) TMSCl, H2O, CH3CN, rt, 1 h; (c) K2CO3, CH3CN/H2O, rt, 20 h; (d) LiOH, CH3CN/H2O, rt, 20 h; (e) watersilica gel, rt, 24 h
Scheme 7 Hydrolysis of boronic ester 79
Boronic acid 80 exhibited instability and decomposed during hydrolysis and purification processes Kuivila and colleagues noted that boronic acids can undergo protodeboronation under basic conditions, suggesting that the equilibrium-generated boronate experiences direct protonolysis instead of cleavage via a Wheland intermediate, as indicated by the Hammett rho value analysis Additionally, Lennox and Lloyd-Jones found that in anhydrous environments, boronic acids tend to form stable trimeric anhydrides known as boroxines, a process that is entropically favored and releases 3.0 equivalents of water.
Scheme 8 Base catalysed protodeboronation of arylboronic acid (left) Dehydration of boronic acid to form aromatic boroxines (right), Lennox and Lloyd-Jones 2014 113
The hydrolysis of pinacol boronic ester is challenging due to the instability of boronic acid and the tendency of the released diol to reform the ester To shift the equilibrium forward, a strategy was implemented to remove the pinacol Sodium metaperiodate was used for the oxidative cleavage of the pinacol boronic ester in an aqueous ammonium acetate and acetone solution, resulting in a clean formation of boronic acid The oxidation of pinacol to acetone serves as the driving force for the hydrolysis toward boronic acid The boronic acid obtained from this oxidative cleavage was utilized in the subsequent step without the need for further purification after work-up and extraction.
Palladium(0)-catalyzed, copper(I)-mediated desulfitative carbon-carbon cross-coupling of crude arylboronic acid with pyrrolidine-2-thione in THF was successfully conducted in a microwave reactor at 100°C for 2 hours, yielding cyclic imine in 62% from boronic ester.
The Liebeskind-Srogl coupling reaction of boronic acid 80 under conventional reflux conditions in THF or dioxane was slow, taking approximately three days with a high catalyst loading of 10 mol% Pd2(dba)3, resulting in only 35-46% yield of cyclic imine 81 from precursor 79 In contrast, microwave-assisted coupling of 80 was completed in a shorter time, yielding higher product amounts with just 4 mol% Pd2(dba)3 This efficiency is attributed to microwave irradiation, which provides effective internal heating by directly interacting with molecules, thereby minimizing the formation of alternative products that could arise from gradual heating The proposed mechanism for the desulfitative coupling reaction involves the exchange of pyrrolidine-2-thione with Cu(I)-thiophene-2-carboxylate (CuTC), leading to intermediate 86, which can undergo Pd insertion to form 87, followed by complexation with CuTC and reaction with boronic acid.
80 or undergoes a reversed process via 88, to give the key intermediate 89
Scheme 9 Mechanism for the desulfitative thioamideboronic acid cross-coupling, Prokopcova et al 2007 116
Synthesis of 3,3-Disubstituted Azetidine Cannabinoids
The synthesis of 3,3-disubstituted azetidine cannabinoids begins with the common intermediate triflate 78 The azetidine ring is formed through the reductive cyclization of α-tosyloxymethyl nitrile 98 An aliphatic side chain is added to triflate 78 using palladium-catalyzed decarboxylative coupling, resulting in nitrile 96, which is then subjected to aldol condensation with paraformaldehyde.
Scheme 11 Retrosynthesis of 3,3-disubstituted azetidine HHC
The details of the synthesis of 3,3-disubstituted azetidine cannabinoids are summarized in Scheme 12
The synthesis process involves several key reagents and conditions: (a) employing [PdCl(C3H5)]2 with Xantphos in xylene at 130°C for 16 hours yielded an 82% result; (b) utilizing (HCHO)n with 40% Triton B/CH3OH in toluene at 60°C for 26 hours achieved a 95% yield; (c) the reaction of p-TsCl with Et3N and DMAP in CH2Cl2 at room temperature for 4 hours resulted in a 94% yield; (d) reduction with LiAlH4 in THF at room temperature for 3 hours; (e) a two-step process involving 37% HCHO(aq) with CH3OH at room temperature for 2 hours followed by NaBH4 at room temperature for an additional 2 hours provided a 62% yield from 98; (f) the use of Sc(OTf)3 in C2H5OH and CH3CN under reflux for 12 hours resulted in an 89% yield; and (g) finally, succinic acid with CH3OH at room temperature for 12 hours was also performed.
Scheme 12 Synthesis of 3,3-disubstituted azetidine cannabinoids
The initial exploration of incorporating an aliphatic side chain into the tricyclic cannabinoid nucleus utilized the Buchwald–Hartwig palladium-catalyzed α-arylation method Despite successful conditions demonstrated in a model study with 4-bromoanisole, the coupling of triflate 78 with ethyl cyanoacetates did not yield the expected products.
13) Most examples of the BuchwaldHartwig palladium-catalyzed -arylation have been reported with aryl bromides or aryl chlorides, but not with aryl triflates 129
Conditions and reagents: (a) NaHMDS/THF, Pd(OAc)2, rac-BINAP, toluene, 100 o C, 20 h; (b)
Na3PO4, Pd2Cl2(C3H5)2, (t-Bu)3PHBF4, toluene, 70 o C, 5-20 h
Scheme 13 Palladium-catalyzed -arylation of 4-bromoanisole and of intermediate triflate 78
Consequently, an alternative coupling reaction developed by Liu and co-workers was employed Palladium-catalyzed decarboxylative coupling of potassium 2-cyanooctanoate with aryl triflate
78 provided a diastereomeric mixture of -aryl nitriles 96 in 82% yield (Scheme 12) 130
Potassium 2-cyanooctanoate was prepared in two steps: monoalkylation of ethyl 2-cyanoacetate with 1.0 equivalent of 1-bromohexane in the presence of K2CO3 in DMF at 85 o C gave ethyl 2- cyanooctanoate in 85% yield, and hydrolysis of the ester with 1.0 equivalent of potassium tert- butoxide in a combination with 1.0 equivalent of water in EtOH at 60 o C gave anhydrous potassium carboxylate in nearly quantitative yield It is noteworthy that no extra base was required for the palladium-catalyzed decarboxylative cross coupling reaction The reaction worked well in the temperature range of 120130 o C, but a slightly higher reaction temperature
The decomposition of triflate 78 at 145 °C complicated the separation of compound 96 from 78 using conventional silica gel column chromatography, as they exhibited similar mobility on the column Fortunately, the reaction reached completion, which simplified the purification process of the product The mechanism for the palladium-catalyzed decarboxylative coupling reaction of 78, initially proposed by Jiang in the Liu group, is depicted in Scheme 14.
According to Jiang, the catalyst [PdCl(C3H5)]2 is reduced in situ and then combines with
Xantphos to generate Pd(0) complex Oxidative addition of Pd(0) to aryl triflate 78, followed by metathesis with potassium 2-cyanooctanoate led to palladium carboxylate 103
Scheme 14 Mechanism for the Pd-catalyzed decarboxylative -arylation, Jiang et al 2012 131
The decarboxylation step in the catalytic cycle is particularly noteworthy, as it has been shown by Tsuji and Tunge that palladium plays a direct role in this process, resulting in the formation of a metalated nitrile.
Scheme 15 Palladium-catalyzed decarboxylation, Tunge et al 2009 133
Anions of nitriles can coordinate with metals via the α-carbon or cyano nitrogen, and they can also bridge two metals in a 2-C,N fashion Consequently, it is assumed that intermediate 103 undergoes decarboxylation.
The palladium catalyst plays a crucial role in the reductive elimination process to yield the desired product An alternative mechanism, not previously discussed in literature, involves the rearrangement of the ketenimine form of 103 from Pd-O to Pd-C bonding, followed by reductive elimination before decarboxylation However, this pathway is less favorable as it fails to highlight the significance of the palladium catalyst in the decarboxylation step.
Scheme 16 Alternative mechanism for decarboxylative coupling reaction
The synthesis of 2,2-disubstituted azetidine cannabinoids progresses with the introduction of a hydroxymethyl group at the α-position of nitrile 96 This is achieved by condensing α-aryl nitrile 96 with excess paraformaldehyde in the presence of Triton B, a benzyltrimethylammonium hydroxide solution, in toluene at 60°C within a sealed tube, resulting in diastereomeric alcohols 97 with excellent yield.
The treatment of alcohol 97 with p-TsCl and excess Et3N, along with a stoichiometric amount of DMAP, successfully yielded tosylate 98 in 94% yield The subsequent reduction of cyanotosylate 98 using excess LiAlH4 in THF at room temperature, followed by spontaneous substitution and cyclization of the intermediate amino derivative, resulted in the formation of 3,3-disubstituted azetidine 99 Further, reductive methylation of the crude secondary amine 99 with 37% aqueous formaldehyde in methanol, followed by sodium borohydride reduction of the iminium ion, produced tertiary N-methylamine 100 with a yield of 62% from 98 Finally, the cleavage of methoxymethyl protecting groups from amine 100 was achieved using a stoichiometric amount of scandium triflate in the presence of excess ethanol, yielding amine 101 in 89%.
Ethanol effectively replaced 5.0 equivalents of 1,3-propanediol in this reaction, addressing the challenge of separating the desired amino alcohol product from 1,3-propanediol The advantage of using ethanol lies in its ease of removal through evaporation, streamlining the purification process.
101 with 1.0 equivalent of succinic acid gave ammonium hemisuccinate 102
Diastereoselective Synthesis of 2,2-Disubstituted Azetidine Cannabinoids
The synthesis of 2,2-disubstituted azetidine cannabinoids was designed starting from the common intermediate triflate 78
The initial proposal aimed for a nondiastereoselective synthesis, intending to introduce an aliphatic side chain through nucleophilic addition to cyclic imine 112, similar to the preparation of 2,2-disubstituted pyrrolidine 82 The plan involved synthesizing cyclic imine 116 from aryl nitrile 113 via the Kulinkovich reaction to yield cyclopropylamine 114, followed by diazo transfer and thermal rearrangement A model study using benzonitrile successfully produced 2-butyl-2-phenylazetidine with satisfactory yield However, the actual Kulinkovich reaction of nitrile 113 resulted in a complex mixture, likely due to the instability of the benzopyran ring and the methoxymethyl groups in the presence of BF3·Et2O.
Scheme 17 Nondiastereoselective synthesis of 2,2-disubstituted azetidine via Kulinkovich reaction and nucleophilic addition to cyclic imine
In another nondiastereoselective synthesis, the synthesis of the azetidine ring was planned to take place by SN2 cyclization of carbamate 120 or the corresponding primary amine
The Curtius rearrangement of carboxylic acid 119, intended to be synthesized from nitrile 96 through alkylation and hydrolysis, results in carbamate 120 However, attempts to hydrolyze the sterically hindered nitrile 117 to form either carboxylic acid 119 or its amide derivative were unsuccessful under various conditions Additionally, reducing the nitrile to aldehyde using DIBAL, followed by oxidation to the carboxylic acid, produced cyclic imine 122 It is likely that chloride displacement occurred with the intermediate aluminum salt Alternatively, alkylation of the dianion from carboxylic acid 118 yielded five-membered ring lactone 123, while alkylation with protected 2-bromoethanol successfully generated the desired product.
Scheme 18 Nondiastereoselective synthesis of 2,2-disubstituted azetidine via Curtius rearrangement and SN2 cyclization
Since the nondiastereoselective synthesis of the 2,2-disubstituted azetidines had failed, we considered that we might have better success with an alternative, diastereoselective strategy
The azetidine ring can be synthesized from β-amino ester 107 through a condensation reaction to form azetidinone 108, followed by a reduction process The highly diastereoselective preparation of β-amino ester 107 from ketone 104 is rooted in the sophisticated methodologies established in previous research.
Ellman utilized the N-tert-butanesulfinyl group as a chiral auxiliary for the enolate addition to the derived imine Ketone 104 was synthesized from the common intermediate triflate 78 through a palladium-catalyzed decarboxylative coupling, followed by oxidative decyanation The retrosynthetic analysis based on these principles is depicted in Scheme 19.
Scheme 19 Diastereoselective retrosynthesis of 2,2-disubstituted azetidine cannabinoids
The details of the synthesis of 2,2-disubstituted azetidine cannabinoids are summarized in Scheme 20
Reagents and conditions: (a) i NaHMDS, THF, rt, 30 min, ii O2 (gas), -78 o C, 30 min, iii
Na2SO3 (aq), 0 o C, 30 min, 74%; (b) Ti(OEt)4, THF, reflux, 19 h, 92%; (c) CH3COOC2H5, LDA, TiCl(Oi-Pr)3, THF, -78 o C, 1 h, 81%; (d) HCl/1,4-dioxane, CH3OH, 10 o C, 2 h, 85%; (e)
CH3MgBr, Et2O, rt, 1.5 h, 78%; (f) LiBF4, CH3CN, H2O, 72 o C, 18 h, 87%; (g) LiAlH4, THF, 65 oC, 24 h; (h) i 37% HCHO (aq), CH3OH, rt, 3 h, ii NaBH4, rt, 2 h, 70% from 108; (i) Dowex 50W-X8, CH3OH, rt, 30 h, 77%
Scheme 20 Diastereoselective synthesis of 2,2-disubstituted azetidine cannabinoids
Oxidation of -aryl nitrile 96 with molecular oxygen (gas tank) in the presence of sodium bis(trimethylsilyl)amide in THF at -78 o C, followed by reduction of intermediate sodium
-cyanohydroperoxide with aqueous sodium sulfite and spontaneous decyanation provided
50 ketone 104 in 74% yield 141 The mechanism for the oxidative decyanation of -aryl nitrile 96 , which was originally proposed by Watt et al., 142 is illustrated in Scheme 21
Sheme 21 Mechanism for the oxidative decyanation of -arylnitrile
The mechanism by which molecular oxygen oxidizes nitrile carbanion 125 remains unclear and has not been thoroughly explored in previous studies It is likely that nitrile carbanion 125 donates an electron to molecular oxygen, resulting in the formation of nitrile radical 126, which is subsequently oxidized to produce the nitrile peroxide anion.
The synthesis of ketone 104 can be efficiently achieved through the reduction of α-hydroperoxynitrile with aqueous sodium sulfite, followed by spontaneous decyanation, offering a streamlined two-step process from intermediate triflate 78 This method, which involves palladium-catalyzed decarboxylative coupling and subsequent oxidative decyanation, contrasts with earlier, more complex approaches developed by Dr Darryl Dixon and Dr Naoyuki Shimada Additionally, the palladium-catalyzed cyanation of triflate 78 using zinc cyanide and PMHS leads to nitrile 130, which is then reduced by DIBAL to yield aldehyde 131 Finally, the nucleophilic addition of n-hexylmagnesium bromide to aldehyde 131, followed by oxidation with manganese dioxide, results in high yields of ketone 104.
Reagents and conditions: (a) Zn(CN)2, Pd(PPh3)4, PMHS, DMF, 60 o C, 8 h; 96% (Dr Dixon's result); (b) DIBAL, CH2Cl2, toluene, -78 o C, 1 h, 89% (Dr Shimada's result); (c) n-C6H13MgBr,
Et2O, 0 o C, 1 h, 96%; (d) MnO2, CH2Cl2, rt, 2 d, 92%
Scheme 22 Original approach for the synthesis of ketone 104
The diastereoselective synthesis of 2,2-disubstituted azetidine cannabinoids was initiated through a well-structured approach by Ellman This process involved the titanium (IV)-mediated condensation of the chiral auxiliary (R)-(+)-tert-butanesulfinamide with ketone 104 in THF at reflux, resulting in the formation of N-sulfinyl imine.
In a yield of 92%, compound 105 was synthesized without the undesired deprotection of methoxymethyl groups or aldol condensation of the ketone, even in the presence of Ti(OEt)4 Additionally, ketimine 105 demonstrated resistance to hydrolysis during aqueous work-up and conventional purification processes The reaction utilized approximately 2.0-3.0 equivalents of Ti(OEt)4 along with a slight excess of tert-butanesulfinamide.
The reaction using Ti(OEt)4 yielded imine 105 effectively, though it was slow and incomplete with a stoichiometric amount, while excessive reagent led to complications during aqueous workup due to titanium oxide forming a hard cake Ellman and colleagues found that only the E isomers were present in N-tert-butanesulfinyl ketimines derived from methylphenyl and n-butylphenyl ketones, with the E assignment based on analogies to p-toluenesulfinyl imines, whose conformations were confirmed through X-ray crystal structures The 1H NMR and 13C NMR spectra further supported these findings.
In CDCl3, ketimine 105 exhibited a single geometric isomer The nucleophilic addition of a titanium enolate, generated in situ through the transmetalation of the lithium enolate of ethyl acetate with TiCl(Oi-Pr)3, to N-tert-butanesulfinyl imine 105 yielded ester 106 with an 81% yield and a diastereomeric ratio of 9:1 The stereochemical outcome was explained by a Zimmerman-Traxler-type six-membered chair-like transition state, stabilized by a four-membered metallocycle, a concept initially proposed by David and colleagues for the asymmetric addition of lithium enolates to N-p-toluenesulfinyl imines, and later supported by the Ellman group for titanium enolate addition to N-tert-butanesulfinyl imines.
Scheme 23 Rationalization of the diastereoselectivity in asymmetric enolate addition to imine
Transmetalation of lithium enolate to a more covalent titanium enolate is crucial for achieving high diastereoselectivity in the reaction with sulfinimine Ellman's research indicates that when the lithium enolate of methyl acetate is added to the N-tert-butylsulfinyl aldimine from benzaldehyde in THF, it results in moderate diastereoselectivity (dr = 83:17) However, employing 2.0 equivalents of ClTi(Oi-Pr)3 for the transmetalation process significantly enhances diastereoselectivity, yielding excellent results (dr = 98:2).
The addition of titanium enolates to N-tert-butylsulfinyl ketimines shows that using 2.0 equivalents or more of ClTi(Oi-Pr)3 enhances the formation of the titanium enolate, which exists in equilibrium with a lithium enolate and a lithium-titanium-ate complex, as 1.0 equivalent was insufficient for increased diastereoselectivity An intriguing study by Fujisawa and colleagues demonstrated that the addition of lithium enolate to p-toluenesulfinimine in THF with HMPA negated the counterion effect, resulting in a product via a non-chelation transition state In contrast, chelation with titanium enolate led to a Zimmerman-Traxler-type transition state, yielding a different diastereomer through a six-membered chair-like structure with a metallocycle.
Scheme 24 Diastereoselectivity in the addition to imine via non chelation and chelation-control,
Applying these concepts to the reaction of N-tert-butanesulfinyl imine 105, the titanium enolate of ethyl acetate was prepared by treatment of the lithium enolate with ca 2.0 equivalents of
TiCl(Oi-Pr)3 was synthesized using TiCl4 and Ti(Oi-Pr)4 in toluene An in situ lithium enolate was generated from freshly prepared LDA and ethyl acetate in THF at -78 °C for 30 minutes The addition of imine 105 as a THF solution at -78 °C to approximately 2.0-3.0 equivalents of titanium enolate for 1 hour yielded ester 106 with a 70-81% yield Notably, the reaction exhibited intriguing characteristics under these conditions.
In a Claisen condensation reaction utilizing approximately 6.0 equivalents of excess titanium enolate, a byproduct was observed with a yield of around 30% The formation of this byproduct increased when the reaction time was extended or when higher temperatures were applied in the presence of the excess titanium enolate.
Scheme 25 Claisen reaction of ester 104 with titanium enolate of ethyl acetate
The cleavage of the tert-butylsulfinyl group from ester 106 using a methanolic solution and hydrochloric acid (4 N in dioxane) at 10 °C resulted in the formation of β-amino ester 107 with an impressive yield of 85% It is important to highlight that both temperature and reaction time significantly influenced the outcome of the acidic methanolysis Specifically, the ratio of starting material 106 to the desired product 107 and an undesired byproduct varied with conditions: at 0 °C after 4 hours, the ratio was approximately 2/3/0, while at room temperature after 1 hour, it shifted to 1/3/1 Despite these variations, a satisfactory yield of 107 was consistently achieved at 10 °C after 2 hours.
106 can be easily obtained by the same approach using the enantiomeric chiral auxiliary (S)-()- tert-butanesulfinamide
Receptor Binding Studies
Our collaborators at Northeastern University, led by Professor Makriyannis, assessed the affinities for CB1 and CB2 receptors They measured CB2 receptor-ligand binding affinities in both mouse and human receptors due to species variation, while CB1 receptor-ligand binding affinities were evaluated only in rats, as no significant differences between rat and human receptors have been found The ligand affinities (Ki) of C1'-azacycloalkyl 9β-hydroxy hexahydrocannabinoids are summarized in Table 1.
Table 1 Ligand affinities (Ki) of C1'-azacycloalkyl 9-hydroxy hexahydrocannabinoids
109 in progress in progress in progress
Structure − activity relationships of C1'-azacycloalkyl 9-hydroxy hexahydrocannabinoids, especially on C3- side chain with its C1'-azacycloalkyl substituents can be summarized as follows:
The 2,2-disubstituted N-methyl pyrrolidine cannabinoid exhibits significantly greater affinities for CB1 and CB2 receptors compared to secondary pyrrolidine cannabinoids Additionally, the salts derived from these amines maintain their binding affinities, suggesting that they could enhance pharmaceutical potency due to their improved water solubility over the parent amines.
3,3-Disubstituted N-methyl azetidine and 2,2-disubstituted N-methyl pyrrolidine cannabinoids exhibit comparable affinities for CB1 and CB2 receptors, with binding affinities similar to (–)-Δ9-THC (hCB1 = 40.7 nM, hCB2 = 36.4 nM) However, their binding affinities are slightly lower than those of the corresponding C1'-cyclobutyl, C1'-cyclopentyl cannabinoids, and the 1',1'-dithiolane analog This observation indicates that the presence of a hydrophilic nitrogen atom in the C1'-cycloalkyl position may reduce binding affinities for both receptors, supporting Professor Makriyannis' earlier hypothesis that introducing a hydrophobic element in this position enhances binding affinities.
The 2,2-disubstituted N-methyl azetidine cannabinoid was deemed too unstable for binding affinity evaluation Currently, the binding affinities of the diastereomeric 2,2-disubstituted azetidinone cannabinoid are under investigation, with expectations that this compound will yield significant insights.
A valuable probe for assessing the deactivation of cannabinergic ligands is anticipated to undergo enzymatic hydrolysis by plasma esterases Consequently, the hydrolyzed form of this probe would exhibit altered behavior in receptor binding assays.
1H NMR and 13C NMR spectra were obtained at frequencies of 500 MHz and 126 MHz, respectively Chemical shifts are expressed in parts per million (δ), referencing the solvent CDCl3 at 7.26/77.0 The multiplicities are denoted as broadened (br), singlet (s), doublet (d), triplet (t), quartet (q), pentet (p), or multiplet (m) Coupling constants (J) are measured in Hertz (Hz) Thin layer chromatography (TLC) was conducted on glass plates with a thickness of 250 µm and a particle size of 5 µm.
Flash column chromatography was conducted using silica gel with a pore size of 60 Å and mesh sizes of 200-400 or premium silica gel with mesh sizes of 40-75 Å All moisture-sensitive reactions were carried out under a nitrogen or argon atmosphere using oven-dried or flame-dried glassware The purity and homogeneity of all materials were confirmed to be at least 95% through techniques such as TLC, 1H NMR, 13C NMR, and LC-MS Optical rotations were measured using a JASCO digital polarimeter with a 0.1 dL cell.
2-((1R,2R,5R)-6,6-dimethyl-4-oxobicyclo[3.1.1]heptan-2-yl)benzene-1,3,5-triyl triacetate
Anhydrous benzene-1,3,5-triol (3.78 g, 0.03 mol) was suspended in CH2Cl2 (250 mL) under a nitrogen atmosphere at 0 °C, to which Et3N (16.7 mL, 0.12 mol) was added slowly, followed by dropwise addition of TMSCl (15.0 mL, 0.12 mol) It is important to note that using a higher concentration with less CH2Cl2 can complicate stirring After the addition, the mixture was stirred at 0 °C for 20 minutes and then at room temperature for 2 hours The resulting solid was filtered out using a Celite pad, and the filtrate was washed with ice-cold water (100 mL × 3), dried over Na2SO4, filtered, and concentrated under reduced pressure at 22 °C, yielding 1,3,5-tris((trimethylsilyl)oxy)benzene 72 as a dark pink oil.
CHCl3/acetone (300 mL, 4/1) under a nitrogen atmosphere at 0 o C was added a solution of diacetates 67 (2.98 g, 12.5 mmol; 4.80 g of material containing 62% pure diacetates 67) and p-
A solution of TsOH·H2O (3.08 g, 16.2 mmol) in a mixture of CHCl3 and acetone (100 mL, ratio 4:1) was added dropwise using an addition funnel over a period of 1.5 to 2 hours Following the addition, the reaction mixture was gradually heated to room temperature and stirred for an additional 4 hours The reaction was then quenched by adding a minimal amount of saturated aqueous NaHCO3 until the pH reached approximately 8, and the mixture was stirred for 45 minutes under a nitrogen atmosphere.
(Note Too much NaHCO 3 led to difficulty in later extraction of product from aqueous layer)
The organic layer was isolated, and the aqueous layer was extracted using ethyl acetate (EtOAc) until thin-layer chromatography (TLC) indicated that most of the product had been removed from the aqueous layer The resulting organic layer was then combined for further analysis.
Crude phenol 73 was dried over MgSO4 and concentrated under reduced pressure A mixture of crude 73 and DMAP in CH2Cl2 was prepared under a nitrogen atmosphere at 0 °C, followed by the slow addition of pyridine and Ac2O, resulting in a homogeneous solution that was stirred for 12 hours The reaction was quenched with ice-cold water, and the organic material was washed with 1 M aqueous HCl and brine, then dried over MgSO4, filtered, and concentrated The residue was purified using silica gel column chromatography with EtOAc/hexane as the eluent, yielding triacetate 74 as a white solid with a 67% yield.
(1R,4R,5R)-6,6-dimethyl-4-(2,4,6-trihydroxyphenyl)bicyclo[3.1.1]heptan-2-one (73)
A solution of triacetate 74 (6.76 g, 17.4 mmol) in methanol (CH3OH) was prepared at 0°C, to which KOH (3.42 g, 60.9 mmol) was added all at once under a nitrogen atmosphere The mixture was stirred for 2 hours at this low temperature, and the reaction was subsequently quenched with 1M aqueous HCl under nitrogen until reaching the desired pH.
The solution was concentrated under reduced pressure and organic material was extracted using ethyl acetate (EtOAc) The combined organic layer underwent a brine wash, was dried with magnesium sulfate (MgSO4), filtered, and then concentrated The resulting residue was purified via short column chromatography with a mixture of EtOAc and hexane (from 1:1 to 4:1), yielding 73 (4.39 g, 96% yield) as an off-white foam, typically containing 10-20% ethyl acetate.
(6aR,10aR)-1,3-dihydroxy-6,6-dimethyl-7,8,10,10a-tetrahydro-6H-benzo[c]chromen-
A solution of 73 (589 mg, 2.25 mmol; 693 mg of material with 85% purity of 73 and 15% EtOAc) was prepared in CH3NO2 (150 mL) under a nitrogen atmosphere at 0 °C TMSOTf (1.0 mL, 5.63 mmol) was added slowly over 20 minutes, and the mixture was stirred at 0 °C for an additional 2.5 hours The reaction was then quenched with solid K2CO3, and the resulting heterogeneous mixture was stirred.
The reaction was conducted for 45 minutes under nitrogen at room temperature, followed by filtration of the solids The resulting solution was concentrated under reduced pressure, and the residue underwent purification using a short silica gel column chromatography with a 1:1 mixture of ethyl acetate and hexane as the eluent This process yielded 489 mg of compound 75, achieving an 83% yield, which appeared as a white foam typically containing 10-20% ethyl acetate.
(6aR,10aR)-1-hydroxy-6,6-dimethyl-9-oxo-6a,7,8,9,10,10a-hexahydro-6H- benzo[c]chromen-3-yl trifluoromethanesulfonate (76)
To a solution of phenol 75 (2.62 g, 0.01 mol; 3.25 g of material containing 81% pure 75 and