The herbaceous plant Cannabis sativa has been used in medicine for centuries and still attracts significant interest due to the biological and pharmaceutical activity of many of its metabolites.1 More than 60 compounds, known as cannabinoids (a group of C₂₁ terpenophenolic compounds), are exclusively found in Cannabis sativa.2 Owing to the development of synthetic cannabinoids,3, 4 the unique components of Cannabis sativa are known as phytocannabinoids. The most abundant compound is Δ⁹‐tetrahydrocannabinol (THC, 1; Figure 1), which shows interesting pharmacological activity as an analgesic, antiemetic, and appetite stimulant, among others, besides its well‐known psychotropic effects.5 Several total syntheses of 1 have been accomplished to date.6 Cannabidiol (CBD, 2) is another important phytocannabinoid with great potential as a drug7 since it modulates the undesired effects of THC when they are administrated together.8

A structurally different cannabinoid named cannabimovone (3) has recently been isolated by the groups of Taglialatela‐Scafati and Appendino from a nonpsychotropic variety of hemp (Cannabis sativa L.; Figure 1).9 In their attempt at preparing 3 from CBD (2) through an intramolecular aldol reaction of keto aldehyde 4 under mild acidic conditions, the product of dehydration (5) was formed instead (Scheme 1). Under basic conditions, the novel cannabinoid anhydrocannabimovone (6) was directly formed through an intramolecular oxy‐Michael addition of one of the phenol groups to the intermediate enone. Synthetic 6 was found to be active against metabotropic and ionotropic cannabinoid receptors, showing a similar biological profile to THC, whereas cannabimovone (3) has affinity only for ionotropic receptors.9

The unprecedented abeo‐menthane terpenoid structure of cannabimovone (3) includes a densely functionalized cyclopentane with four contiguous stereocenters. The novel structure of 3, coupled with its lability towards dehydration under acidic or basic conditions and the interesting biological profiles of both 3 and 6, inspired us to develop a total synthesis that could allow access to a wide variety of analogues. Herein, we report the first total synthesis of enantiopure cannabimovone (3) and also revise the structure originally assigned to anhydrocannabimovone from trans‐fused 6′ to cis‐tetrahydro‐1H‐cyclopenta[b]benzofuran 6. Our approach to the synthesis of these compounds relies on a gold(I)‐catalyzed cycloisomerization10, 11, 12, 13, 14, 15 of aryl‐substituted 1,5‐enyne 7, which could be obtained in a few steps from commercially available (+)‐methyl (S)‐3‐hydroxybutyrate (9; Scheme 2).

The synthesis commenced with alkylation of the lithium enolate of 9 with prenyl bromide to provide known compound 10 with excellent diastereoselectivity (98:2) by following a slight modification of the reported procedure16 (Scheme 3). Protection of the alcohol of 10 as a silyl ether, conversion of the ester into an aldehyde by a two‐step procedure (DIBAL reduction/Swern oxidation), and subsequent homologation with the Ohira–Bestmann reagent led to 1,5‐enyne 11 (31 % over 5 steps). Sonogashira coupling of 11 with iodo arene 12, prepared in two steps from olivetol, gave 7 in 83 % yield on a multi‐gram scale. The gold(I)‐catalyzed cyclization of 1,5‐enyne 7 was highly solvent dependent. Exposing 7 to the cationic gold(I) complex [(JohnPhos)Au(MeCN)]SbF₆ in CH₂Cl₂ led to bicyclic compound 13 (49 %). A similar result was obtained using other solvents such as Et₂O or toluene. Reaction in MeOH afforded methyl ether 14 (93 %). However, when the reaction was performed in DMSO, cyclopentene 8 was obtained in excellent yield (88 %). This reaction was performed up to a 2.1 g scale. A similar result was observed when the reaction was performed in DMF (79 %). Presumably, the initial intermediate of the gold(I)‐catalyzed cyclization (Int) undergoes proton elimination assisted by the solvent to give 8 after protodeauration. Notably, the gold‐catalyzed cyclization led exclusively to the product with the correct relative configuration, thereby setting two of the final four stereocenters.

Although deprotection of the TBS group of 8 followed by oxidation of the alcohol to the methyl ketone could be carried out uneventfully, isomerization to form the α,β‐unsaturated ketone failed under all the conditions we examined with this and with other intermediates with different phenol protecting groups. Fortunately, the desired functionality in the five‐membered ring could be introduced by epoxidation with m‐CPBA and NaHCO₃ to exclusively form 15, followed by Meinwald rearrangement with stoichiometric BF₃⋅Et₂O to give ketone 16 (2,3‐cis). Epimerization and cleavage of the silyl ether was achieved with aqueous HCl to give 17 (Scheme 4). The relative configuration of 17 was determined by NMR studies and was confirmed by preparation of the same compound by a different route, namely cleavage of the TBS group of 8 followed by epoxidation to give 18, which underwent Meinwald rearrangement to provide 17. Diastereoselective reduction of β‐hydroxy ketone 17 by Saksena–Evans reaction with NaBH(OAc)₃ in CH₂Cl₂ afforded diol 19. Protection with AllocCl, which proceeded with moderate selectivity, followed by Dess–Martin oxidation gave protected cannabimovone 20. Cleavage of the MOM groups using MgBr₂ and BnSH,17 followed by Pd⁰ deprotection of the allyl carbonate provided 3 (53 % over 2 steps). The spectral data and optical rotation of the synthetic cannabimovone (3) matched those reported for the natural compound.18

Whereas the synthesis of 3 fully supported the assigned configuration for all of the intermediates, a crystalline oxabicycle intermediate 21 was obtained through treatment of epoxide 18 with a Brønsted acid, which led to opening of the epoxide and trapping of the benzylic carbocation by the free alcohol (Scheme 5). The molecular structure of 21 was determined by X‐ray diffraction, which confirmed the relative configuration between the isopropenyl and the hydroxyethyl substituents.19 Treatment of 21 with ZnI₂ effected a pinacol rearrangement to give ketone 17.

The synthesis of anhydrocannabimovone (6) was carried out from acetate 22 by MOM cleavage followed by treatment with K₂CO₃ to promote the oxy‐Michael addition. Under these conditions, two separable epimers (6 and 6′′) were formed in a 4:1 ratio (Scheme 6). Surprisingly, although ¹H NMR of the major isomer 6 was identical to that reported for anhydrocannabimovone, very significant differences were observed in the ¹³C NMR spectrum.20 Furthermore, the optical rotation of 6 ([α]22D =+40.6° (c=0.29, CHCl₃)) was very different to that reported ([α]22D =−17° (c=0.02, CHCl₃)).9 The structure of anhydrocannabimovone (6) was finally confirmed by X‐ray diffraction19 and its absolute configuration was assigned on the basis of the X‐ray structure of anhydrocannabimovone 2‐bromobenzoate (23; Figure 2).19

In order to clarify the discrepancy between our structural assignment and that originally reported,9 we performed DFT calculations to study the oxy‐Michael cyclization. Under basic conditions, oxy‐Michael cyclization of the phenolate anion leads to cis fusion, which is more favored than the trans addition by 19.8 Kcal mol⁻¹ (Figure 3).21 Furthermore, DFT calculations were employed to predict the expected ¹³C NMR chemical shifts of the different possible products.22 Our data were in better agreement with the cis‐tetrahydro‐1H‐cyclopenta[b]benzofuran structure for anhydrocannabimovone (6).20

In conclusion, we have accomplished the first total synthesis of cannabimovone (3). The four stereogenic centers of the target molecule were set up starting from the stereogenic center present in commercially available (+)‐methyl (S)‐3‐hydroxybutyrate (9) and by using a fully diastereoselective gold(I)‐catalyzed cyclization. Interestingly, this is the first example of the cycloisomerization of simple 1,5‐enynes into 3‐vinylcyclopent‐1‐enes catalyzed by gold in the context of natural product synthesis. We also synthesized anhydrocannabimovone (6) and revised the stereochemistry at the ring fusion by X‐ray crystallography and DFT calculations. This synthetic endeavor provides ready access to 3 and 6, as well as other synthetic cannabinoids, for biological testing.

Dedicated to Professor Miquel A. Pericàs on the occasion of his 65th birthday