Bromination of Cannabinoids from Cannabis sativa L.: A Comparative Study Using an Enzymatic Catalysis Approach

Abstract

An in-depth study on bromination of six typical cannabinoids from Cannabis sativa L. is reported using three different methods: micellar-assisted enzymatic catalysis, NBS, and Oxone. Under the micelle-enabled CiVCPO-catalyzed conditions, selective dibromination of cannabinoids was achieved in up to 72% isolated yield for CBN and 67% for CBL with complete substrate conversion. Notably, the bromination level could be tuned by varying enzyme concentration (300–500 nM), enabling controlled access to mono- or dibrominated products. Compared to conventional methods, the enzymatic system demonstrated comparable or superior efficiency while avoiding overoxidation and enabling improved selectivity. This work highlights a mild and selective strategy for the late-stage functionalization of sensitive natural products, providing a promising platform for the development of bioactive cannabinoid derivatives.

Enzymatic halogenation has gained significant attention because of its benign and environmentally friendly nature. In contrast to conventional methods using stoichiometric metal-based reagents or oxidants in hazardous organic solvents, − enzymes act as efficient catalysts at low concentration and enable reactions in aqueous media. They also exhibit mild reactivity and improved safety for sensitive substrates. − Among these, vanadium-dependent haloperoxidases such as CiVCPO (Curvularia inaequalis chloroperoxidase) are widely studied. − CiVCPO shows high stability toward H₂O₂ and in various solvents, along with excellent catalytic activity and broad substrate tolerance, − making it suitable for halogenating sensitive natural products.

Natural products from plants and marine sources have long contributed to bioactive diversity and drug candidates. Brominated derivatives show various biological activities, , and bromine incorporation into complex structures is an effective late-stage drug modification. Here, we focus on cannabinoids from Cannabis sativa L., which exhibit diverse biological activities such as analgesic, antidiabetic, antioxidant, anti-inflammatory, and antimicrobial effects. , However, their modification, especially halogenation, has rarely been reported. Existing studies mainly address marijuana detection or dimer formation, − suggesting susceptibility to oxidative conditions. Although oxidative halogenation is common, its application to cannabinoids remains underexplored. Additionally, limited natural availability of minor compounds often hinders postfunctionalization. Therefore, developing a mild and effective oxidative bromination for sensitive substrates such as cannabinoids is needed.

In our previous report on micelle-enabled enzymatic bromination of α-oxo ketene dithioacetals, we developed a mild, environmentally friendly strategy. We investigated on-water and in-water CiVCPO-catalyzed bromination using TPGS-750-M surfactant in buffered aqueous media. This enzymatic pathway offers a benign and efficient route utilizing a biocatalyst, suitable for sensitive compounds such as cannabinoids. Here, we extend this approach to brominate representative cannabinoidscannabinol (CBN) 1, cannabidiol (CBD) 2, cannabicyclol (CBL) 3, cannabichromene (CBC) 4, cannabigerol (CBG) 5, and Δ⁹-tetrahydrocannabinol (THC) 6via CiVCPO catalysis. We also compare it with N-bromosuccinimide (NBS) and Oxone to evaluate efficiency and selectivity (Figure ). Comprehensive studies on cannabinoid bromination with different agents remain underexplored. We anticipate that the resulting brominated derivatives will exhibit enhanced biological activity over their precursors, offering a promising functionalization strategy for this class of natural product.

1.

Micellar-assisted enzymatic versus conventional bromination of cannabinoids.

To investigate this enzymatic bromination of cannabinoids, we selected CBN 1 as the model substrate. This compound was chosen due to its stability and available quantity. We used 20 mM CBN 1 in a DMSO/citrate buffer mixture (1:5, v/v) as the initial solvent system. This “on-water” condition was inspired by our previous study. We selected the engineered CiVCPO triple mutant L241V/T343A/P395D (CiVCPO_{L241V/T343A/P395D}), previously reported by the Wever group, to generate hypobromite at a 6-fold higher rate than the wild-type enzyme, for this study. Under CiVCPO_{L241V/T343A/P395D} catalysis with KBr and H₂O₂, 4-monobrominated CBN 1a was isolated in 18% yield after 24 h, with 60% starting material recovered (Table , entry 1). The product was fully characterized by 1D and 2D NMR, confirming bromination on the phenolic aromatic ring. Alternatively, micelle-enabled enzymatic bromination based on our previous work was examined (Table , entry 2). Interestingly, this condition yielded solely 2,4-dibrominated CBN 1b, along with 62% recovered starting material. Notably, a significant shift in the methylene (−CH₂−) signal of the n-pentyl chain was observed among CBN 1 and its brominated derivatives: 2.47 ppm (CBN 1), 2.67 ppm (1a), and 2.95 ppm (1b). This distinction enables NMR monitoring of reaction outcomes using CH₂Br₂ as an internal standard. Due to low conversion under both conditions, the substrate concentration was reduced to 10 mM (Table S1, entry 1). This led to complete consumption of CBN 1, and 1b was isolated in 65% yield. Although efficiency improved, the reaction scale remained limited. We hypothesized that the organic solvent content affected substrate dispersion in micellar media. Therefore, to maintain a 20 mM scale, acetone was increased to 30% of the total volume and KBr equivalents were varied (Table , entries 3 and 4). Although complete consumption of 1 was achieved, only 47% (NMR yield) of 1b was obtained, along with overoxidation products at 160 mM KBr, likely quinone dimers. This suggests that higher organic solvent improves solubility, but cannabinoid sensitivity to oxidation requires careful control of reagent equivalents. Accordingly, optimal conditions were established using 80 mM KBr and 300 nM CiVCPO_{L241V/T343A/P395D} in micellar media (Table , entry 5). This gave 1b as the sole product with complete consumption of 1, and it was isolated in 72% yield. Further variations, including KBr/CBN ratios and higher substrate concentrations, led to mixtures of 1a, 1b, and unreacted 1 (Table S1, entries 2–4). Control experiments confirmed no reaction in the absence of any component, highlighting the necessity of the KBr–H₂O₂–CiVCPO_{L241V/T343A/P395D} system (Table , entry 6). Switching surfactant from TPGS-750-M to SDS significantly reduced efficiency (Table , entry 7). We also attempted to favor 4-monobrominated CBN 1a by varying solvent composition and enzyme concentration under “on-water” conditions (Table S2), isolating 1a in 54% yield with minor 1b formation. However, extending this to other cannabinoids resulted in low conversion and mixtures of 4-mono and 2,4-dibrominated products, indicating lower efficiency compared to micellar “in-water” conditions. Thus, the optimal system was defined as 80 mM KBr, 170 mM H₂O₂, and 300 nM CiVCPO_{L241V/T343A/P395D} in 30% acetone/2 wt % TPGS-750-M in citrate buffer. This condition favored dibromination, which is advantageous for isolation since mono- and dibrominated products have similar polarity, complicating purification.

1. Optimization of the Reaction Condition.

		Isolated yield (%)
Entry	Deviation from standard condition ,	1a	1b
1	None (media A)	18	-
2	None (media B)	-	19
3	None (media C)	-	47
4	KBr 80 mM (media C)	33	60
5	KBr 80 mM, C i VCPO 300 nM(media C)	-	85 (72)
6	No KBr or H2O2 or enzyme (media C)	NR
7	KBr 80 mM, CiVCPO 300 nM (media D)	21	-

^a Media A: 20% DMSO/citrate buffer (pH 5.0); Media B: 10% acetone/2 wt % TPGS-750-M in citrate buffer (pH 5.0); Media C: 30% acetone/2 wt % TPGS-750-M in citrate buffer (pH 5.0); Media D: 30% acetone/2 wt % SDS in citrate buffer (pH 5.0).

^bReaction conditions: CBN 1 (20 mM–0.1 mmol), H₂O₂ (170 mM) in media, rt, 24 h.

^cNMR yield using CH₂Br₂ as an internal standard.

To evaluate this enzymatic strategy, parallel oxidative bromination of CBN 1 was performed using NBS, H₂O₂–CH₃COOH, and Oxone­(Table ). Overall, the enzymatic method gave comparable yields to conventional approaches, particularly Oxone, often considered a green oxidant. , However, Oxone required excess oxidizing reagent, while NBS and H₂O₂–CH₃COOH produced overoxidation products, indicating these methods are less mild and environmentally friendly than the enzymatic system. Moreover, the enzyme provided better selectivity by tuning its concentration, a nonchemical parameter, which helps reduce waste formation.

2. Bromination of CBN 1 Using Conventional Brominating Agents.

		Yield (%)
Entry	Conventional methods	1a	1b	CBN 1 recovery
1	CBN 1:NBS 1:1	28	24	50
2	CBN 1:NBS 1:3	-	75	-
3	CBN 1:NaBr:Oxone 1:3:3	-	91	5
4	CBN 1:NaBr:Oxone 1:2:1	9	24	60
5	CBN 1:NaBr:H2O2 1:3:3	-	34	-
6	CBN 1:NaBr:H2O2 1:2:1	50	11	45

^aIsolated yield.

^bCBN 1 (0.1 mmol), DMF (0.6 mL), rt, 24 h.

^cCBN 1 (0.1 mmol), acetone:H₂O 5:1 (0.6 mL), rt, 24 h.

^dCBN 1 (0.1 mmol), CH₃COOH (0.6 mL), rt, 3 h.

^eCBN 1 (0.1 mmol), CH₃COOH (0.6 mL), rt, 24 h.

To assess substrate scope, bromination was performed on cannabidiol (CBD) 2, cannabicyclol (CBL) 3, cannabichromene (CBC) 4, cannabigerol (CBG) 5, and Δ⁹-tetrahydrocannabinol (Δ⁹-THC) 6. For each, three conditionsmicellar enzymatic catalysis, NBS, and Oxonewere applied to evaluate efficiency and sensitivity. Each substrate shows distinct behavior, which will be discussed in the following sections.

For cannabidiol (CBD) 2, under optimal conditions, complete consumption was observed and monobrominated 2a was isolated in 36% yield (Scheme ). Increasing the enzyme concentration to 500 nM improved both conversion and selectivity for dibrominated 2b, resulting in a 60% isolated yield. The methylene proton of the n-pentyl chain in ¹H NMR showed a trend similar to CBN 1: 2.36 ppm (CBD 2), 2.57 ppm (2a), and 2.91 ppm (2b). Compared with NBS and Oxone, these methods gave lower conversion and multiple brominated product 2c. Compound 2c arose from addition to the endocyclic double bond followed by intramolecular nucleophilic cyclization, while the acyclic double bond remained unchanged. Notably, enzymatic conditions left the unsaturated moiety intact, showing selectivity for aromatic bromination. These results demonstrate that bromination degree can be controlled under enzymatic conditions.

1. Bromination of Cannabidiol (CBD) 2 .

Similarly, for cannabicyclol (CBL) 3, micelle-enabled enzymatic conditions gave good conversion and excellent selectivity toward the 2,4-dibrominated product 3a, isolated in 67% yield (Scheme ). In contrast, NBS and Oxone showed lower conversion and mixtures of 4-mono and 2,4-dibrominated CBL. As noted, similar polarities of brominated derivatives complicate purification; thus, high selectivity for 3b over 3a is advantageous. A consistent shift of the methylene signal was also observed: 2.44–2.48 ppm (3), 2.56–2.62 ppm (3a), and 2.90–2.98 ppm (3b). This clear NMR distinction enables rapid monitoring of bromination on the aromatic ring.

2. Bromination of Cannabicyclol (CBL) 3 .

The reaction outcome became more complex when the substrate contained an acyclic double bond, as in cannabichromene (CBC) 4 (Scheme ). Both conventional and enzymatic conditions gave similar products: aromatic substitution products 4a and 4b and substitution–cyclization products 4c and 4d. The enzymatic method still showed good conversion compared to NBS and Oxone, although product ratios were consistent across all methods. Compound 4c was fully characterized by 1D/2D NMR and confirmed by X-ray diffraction. Similar spectra of 4c and 4d confirmed a shared skeleton, differing only in bromination level on the aromatic ring. Formation of 4c likely involves bromination of the acyclic double bond followed by cyclization to a fused system, then further electrophilic aromatic bromination to give 4d as a minor product (see the Supporting information). This suggests cannabinoids with long-chain alkenes are highly sensitive to oxidants such as hypobromite or bromonium ions, leading to complicated structural diversification rather than simple electrophilic substitution.

3. Bromination of Cannabichromene (CBC) 4 .

For CBG 5, a limited amount of starting material allowed only enzymatic screening. Under optimal conditions, complete consumption was achieved, but the dibrominated product 5a was isolated in only 12% yield (Scheme ). Given its structural similarity to CBC 4, the low yield was attributed to the long-chain alkenyl moiety, likely increasing the susceptibility to oxidative conditions. This suggests that cannabinoid derivatives with long alkyl chain acyclic double bonds are highly sensitive to oxidants, limiting their functionalization.

4. Bromination of Cannabigerol (CBG) 5 .

Another low-stability substrate is Δ⁹-tetrahydrocannabinol (THC) 6, reported to possibly convert to CBN 1 under thermal conditions. Under micellar enzymatic conditions, no simple THC brominated products were observed; instead, 2,4-dibrominated CBD 2b was formed as the major product (11% isolated yield) with full substrate consumption (Scheme ). Based on results observed from CBC 4 and CBG 5, the enzyme concentration was reduced to 200 nM. THC 6 was again fully consumed, giving 2b as the main product in 20% isolated yield. In contrast, NBS and Oxone showed poor control over bromination and skeleton rearrangement. Notably, in the case of NBS, 38% of 6 recovered, while using Oxone led to full consumption. This highlights the high sensitivity of THC, limiting functionalization despite good enzymatic selectivity. Formation of 6a and 6b likely involves ring opening of THC 6 to CBD 2, followed by cyclization and electrophilic bromination (see Supporting Information). However, such products were observed only with Oxone, indicating uncontrolled generation of Br⁺ under strong oxidative conditions.

5. Bromination of Δ⁹-Tetrahydrocannabinol (THC) 6 .

To evaluate the practicability of the enzymatic bromination, we performed a 1.0 mmol scale synthesis of CBN 1 (Scheme a). To maintain reaction efficiency, the concentration of CiVCPO_{L241V/T343A/P395D} was increased to 400 nM and the reaction time was extended to 40 h. Although a mixture of products 1a and 1b was obtained, 1b was the major product with a 61% isolated yield, while 1a was formed in 15% yield. Following the success of micelle-enabled enzymatic bromination, we investigated the chlorination of CBN 1 (Scheme b). However, the enzyme showed poor chlorination activity, with 70% of the starting material recovered and only an 18% isolated yield of monochlorinated product 1c. Interestingly, chlorine substitution on the aromatic ring occurred at a different position than that in the monobrominated product, as confirmed by HMBC correlations. These results indicate that although CiVCPO can catalyze chlorination, it is more efficient for bromination.

6. a) Scale-up Bromination of CBN 1 and b) Chlorination of CBN 1 .

After evaluating the bromination of these cannabinoid derivatives, we tested the α-glucosidase inhibition and DPPH scavenging activity of all synthesized compounds (Figures S2, S3). In general, aromatic bromination products showed improved activity compared to their corresponding precursors, while modifications to the core skeleton had negative effects on the evaluated bioactivities. For α-glucosidase inhibition, brominated CBN, CBD, and CBL exhibited improved activity compared to their precursors, confirming the beneficial effect of bromination. Conversely, incorporating a chlorine atom into the cannabinoid structure such as 1c demonstrated 2-fold reduction in inhibitory effect. Such strikingly observed results suggested a partially predictable inhibitory effect on introducing different halogen atoms into the cannabinoid motif. The increase in biological activity from starting materials to aromatic brominated products was even more pronounced in the DPPH assay. Most brominated derivatives (1–6) showed superior activity, except 2c, 4c, and 6a, which involved skeletal modifications. In addition, dibrominated compounds displayed higher activity than monobrominated ones. Thus, our enzymatic strategy provides good selectivity toward dibrominated derivatives with enhanced efficacy.

In this research, we conducted an in-depth study of micelle-assisted enzymatic bromination of a class of sensitive natural products, cannabinoids. Compared with conventional methods, our approach offers several advantages: 1) the use of benign biocatalysts for bromination in aqueous media, 2) good selectivity toward dibrominated products with high efficiency, and 3) tunable reaction outcomes by adjusting the enzyme concentration rather than chemical reagents. Furthermore, the resulting brominated products showed improved bioactivities, suggesting a potentially beneficial effect of bromine incorporation on the biological activity. Future research will focus on expanding the substrate scope to other natural products and further optimizing the system for large-scale production.

The data underlying this study are available in the published article and its Supporting Information.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.orglett.6c01788.

• Additional solvent screening, experimental procedures, NMR yield calculation, characterization data of all synthesized compounds, and NMR spectra (PDF)

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.