Initially, a diastereoselective synthesis of CBD (1) was planned by a coupling involving a trans:cis mixture of isopiperitenol (9a) and olivetol (8) revisited by our research group (vide infra). The synthesis of mixture 9a was idealized by reduction of isopiperitenone (16), which would be obtained by selective oxidation of (R)‐(+)‐limonene (17)^{[ 43 ]} (route A) or by diastereoselective reactions starting with (R)‐(−)‐carvone (15a)^{[ 37 ]} (route B) (Scheme  3a).

Following Route A shown in the retrosynthetic analysis for CBD (1) (Scheme 3a), (R)‐(+)‐limonene (17) was subjected to oxidation with CrO₃ (4 equiv) and tert‐butanol (10 equiv) in CCl₄ under reflux for 3 h,^{[ 43 ]} but isopiperitenone (16) could not be isolated by column chromatography due to purification issues. The oxidation of (R)‐(+)‐limonene (17) to isopiperitenone (16) or directly to isopiperitenol (9a) was extensively studied using different oxidants (see Table S1 in the Supporting Information). However, the best result obtained was the oxidation of (R)‐(+)‐limonene (17) in the presence of an excess of Collins reagent (CrO₃ • 2 Py) (15 equiv) in CH₂Cl₂ at room temperature for 24 h, which afforded isopiperitenone (16) in 14% isolated yield.^{[ 44 ]} When the same reaction was carried out in the presence of 5 equiv of Collins reagent, isopiperitenone (16) was isolated in 7% yield (Scheme 3b). These unpromising results led us to abandon Route A for the synthesis of CBD (1) (Scheme 3a). Considering Route B (Scheme 3a), the synthesis of CBD (1) was initiated by the three‐step conversion of (R)‐(−)‐carvone (15a) to intermediate 18a in 47% yield as a trans:cis mixture with a ratio of ≈2:1. The diastereoisomeric mixture 18a was subjected to condensation with p‐tosylhydrazide (19) in the presence of AcOH and 37% HCl to give intermediate 20a in 60% yield as a trans:cis mixture with a ratio of ≈3:1. A one‐pot reduction/rearrangement of mixture 20a was performed using catecholborane (21) and AcONa to afford isopiperitenol (9a) in 53% yield as a trans:cis mixture with a ratio of ≈8:1.^{[ 37 ]} In order to succeed in the key step of the synthesis, which consists of an acid‐catalyzed Friedel–Crafts reaction between mixture 9a and olivetol (8), experiments were carried out using cyclohex‐2‐enol as a model compound and olivetol (8) in the presence of different acid catalysts. The best Brønsted acid was CSA (10 mol%) (see Table S2 in the Supporting Information), which allowed the coupling between the trans:cis mixture 9a and olivetol (8) to afford CBD (1) and abn‐CBD (13) in yields of 34% and 22%, respectively (Scheme  4 ). Furthermore, the most effective Lewis acids tested during the optimization were TMSOTf and AgOTf. The third‐best performance was obtained with FeCl₃·6H₂O, which suggests a potential limitation of the model compound used, given that this iron catalyst has previously proven effective in CBD synthesis^{[ 45 ]} (see Table S2 in the Supporting Information). Thus, when the coupling between the trans:cis mixture 9a and olivetol (8) was performed with TMSOTf (20 mol%), only abn‐CBD (abn‐13) could be isolated in 10% yield (result not shown). As an alternative to TMSOTf, we used 20 mol% of AgOTf for the Friedel–Crafts reaction and obtained CBD (1) and abn‐CBD (13) in 30% yield each (see Supporting Information). In the presence of an appropriate acid, both trans‐9a and cis‐9a diastereoisomers generate the same carbocation intermediate, which reacts diastereoselectively with olivetol (8), leading exclusively to the formation of CBD (1) with a trans substitution pattern on the cyclohexene ring. The reactions with CSA and AgOTf did not produce dialkylated products (14a,b) or THC (2), possibly due to their short times, but we could not avoid the formation of abn‐13.

To optimize the α‐hydroxylation of carvone shown in Scheme 4, we performed several experiments to convert (S)‐(+)‐carvone (15b) to α‐hydroxycarvone (18b) by a Davis oxidation targeting ent‐CBD (ent‐1)^{[ 46 ]} (see Table S3 in the Supporting Information). The best result for the transformation allowed the exclusive formation of trans‐α‐hydroxycarvone (trans‐18b) in 38% isolated yield. This was achieved when the enolate of compound 15b was generated with 1 equiv of LDA, initially at 0 °C, and then at room temperature for 30 min, followed by reaction with 1.5 equiv of oxaziridine 22 (Davis reagent) at −78 °C for 60 min (Scheme  5 ). Although the transformation shown in Scheme 5 is highly diastereoselective and leads exclusively to the formation of compound trans‐18b, we used the synthetic route shown in Scheme 4 from (S)‐(+)‐carvone (15b) to synthesize ent‐CBD (ent‐1).

Pharmacological effects of CBD (1) and ent‐CBD (ent‐1) in young and aged male mice were conducted using PM‐DAT (Figure  1 ) to assess memory, anxiety, spontaneous motor activity, and their respective interactions.^{[ 39 ]} Despite recommendations to include both male and female mice in behavioral experiments,^{[ 47 ]} intrinsic genetic and hormonal differences, as well as age‐related sex‐specific variations, can significantly affect responses to cannabinoid treatments such as CBD (1),^{[ 48 ]} which seems to increase anxiety in female mice.^{[ 49 ]} Therefore, to characterize the effects of CBD (1) on anxiety‐like responses, spontaneous motor activity, and memory, this study focused on male mice. We administered synthesized cannabinoids in doses ranging from 3 to 20 mg kg⁻¹, which is within the ideal range for anxiety studies with CBD (1), as noted by Guimarães and co‐workers.^{[ 38} , ^{50 ]} This dosage range is within the inverted U‐shaped dose‐response curve observed in behavioral experiments with CBD (1) in rats, indicating that doses below 2.5 mg kg⁻¹ and above 20 mg kg⁻¹ are ineffective.^{[ 50 ]}

The anxiolytic‐like effect was assessed by the percentage of time each animal spent in open arms (OA) during the training session, as mice naturally prefer protected, unlit spaces and have an aversion to unprotected, lit spaces.^{[ 39 ]} Figure 1A shows that young adult male mice (3–6 months old) treated with 1 mg kg⁻¹ of diazepam (b) spent more time in OA than those treated with vehicle (a), ent‐CBD (ent‐1) at doses of 3 mg kg⁻¹ (c) and 10 mg kg⁻¹ (d), and CBD (1) at doses of 3 mg kg⁻¹ (f) and 20 mg kg⁻¹ (g), demonstrating the anxiolytic effect of diazepam. In addition, treatment with ent‐1 at a dose of 20 mg kg⁻¹ (e) was not significantly different from vehicle (a) and diazepam (1 mg kg⁻¹) (b) [F_(6,41) = 6.399; P < 0.0001]. In Figure 1B, aged male mice (15–18 months old) treated with 20 mg kg⁻¹ of ent‐CBD (ent‐1) spent considerable time in OA, supporting the anxiolytic properties of ent‐1 (e). No significant difference was observed between the aged mice treated with ent‐1 at a dose of 20 mg kg⁻¹ (e) and those treated with diazepam (1 mg kg⁻¹) (b). Further, no significant difference was found between vehicle‐treated mice (a) and those treated with ent‐1 at doses of 3 mg kg⁻¹ (c) and 10 mg kg⁻¹ (d), and with CBD (1) at doses of 3 mg kg⁻¹ (f) and 20 mg kg⁻¹ (g) [F_(6,42) = 2.440; P = 0.0409]. The results indicate that the anxiolytic effect of ent‐1 at the dose of 20 mg kg⁻¹ (e) is more pronounced in older male mice compared to younger male mice [F_(3,26) = 25.53; P ≤ 0.0001] [the corresponding vehicles (a) were included in the analysis]. The findings related to ent‐1 suggest that further experiments, including molecular analysis of brain structures underlying anxiety and long‐term memory formation, as well as toxicological studies, are required before ent‐1 can be considered as a potential drug candidate. Furthermore, mice treated with CBD (1) showed no considerable anxiolytic‐like effect (Figure 1A,B). The lack of anxiolytic‐like effects for CBD (1) is in contrast to the literature,^{[ 15 ]} but different animal models, doses, and treatment times were used [stress‐induced hyperthermia (SIH) (doses: 1, 3, and 10 mg kg⁻¹)^{[ 15 ]} versus PM‐DAT (doses: 3, 10, and 20 mg kg⁻¹)], both using intraperitoneal (IP) injections to administrate the drugs.

The effects of ent‐CBD (ent‐1) at doses of 3 mg kg⁻¹ (c), 10 mg kg⁻¹ (d), and 20 mg kg⁻¹ (e) and CBD (1) at doses of 3 mg kg⁻¹ (f) and 20 mg kg⁻¹ (g) on the long‐term discriminative avoidance memory of young male mice (Figure 1C) and aged male mice (Figure 1D) were evaluated by the percentage of time spent in the aversive enclosed arm (AEA) during the test session.^{[ 39 ]} No effects of treatments were found in both young [F_(6,43) = 1.512; P = 0.973] and aged [F_(6,41) = 0.6622; P = 0.6835] mice, suggesting that ent‐1 and CBD (1) did not impair long‐term memory formation at all doses evaluated. The number of entries in the nonaversive enclosed arm (NAEA) was used as a measure of spontaneous motor activity and was evaluated during the training session under the effect of the drugs. ent‐CBD (ent‐1) at a dose of 20 mg kg⁻¹ increased the number of entries (exploratory behavior) of young mice compared to all groups [F_(6,43) = 7.590; P < 0.0001]. Moreover, aged mice treated with 1 mg kg⁻¹ of diazepam (b) exhibited a significantly reduced number of entries in the NAEA compared to aged mice treated with ent‐1 at doses of 3 mg kg⁻¹ and 10 mg kg⁻¹ [F_(6,41) = 3.223; P = 0.0110]. Aged mice treated with ent‐1 at a dose of 20 mg kg⁻¹ did not differ in the number of entries in the NAEA from the aged mice treated with vehicle (a) (results not shown). In addition, experiments showed that CBD (1) at a dose of 3 mg kg⁻¹ (f) had a sedative effect in more than half of the young male mice (Figure 1A,C). Animals that did not explore the maze were excluded from data analysis.