Effect of 1,4-Dioxane Solvent on β-Glucuronidation Using Methyl 1,2,3,4-Tetra-O-acetyl-β-D-glucuronate as the Glycosyl Donor

Abstract

A facile and selective β-D-glucuronidation of alcohols, such as (−)-menthol, cholestanol, (+)- and (−)-borneols, and 2-adamantanol, using commercially available methyl 1,2,3,4-tetra-O-acetyl-β-D-glucuronate as the glycosyl donor and trimethylsilyl bis(trifluoromethanesulfonyl)imide (Tf₂NTMS) (0.5 equivalent) as the activator in 1,4-dioxane at 60 °C gave products in moderate yields. The addition of MS4A increased the β : α ratios of D-glucuronides when cholestanol, (+)-borneol, and 2-adamantanol were used as the acceptor substrate.

Introduction

We have reported an α-glucuronidation reaction that employs methyl 1,2,3,4-tetra-O-acetyl-β-D-glucuronate (1) and bis(trifluoromethanesulfonyl)imide (Tf₂NH) in dichloromethane (DCM).¹⁾ However, β-D-glucuronidation is a much more useful reaction than α-D-glucuronidation for the following reasons: D-glucuronates are usually present in nature as β-anomers, such as heparin, heparan sulfate, and glycyrrhizin, and D-glucuronides produced in the liver as metabolites of administered medicines have a β-configuration. Strategies to obtain the desired β-glucuronides include the use of 2-O-acetyl-6,3-lactone of D-glucuronic acid as the glycosyl donor²⁾ and the method of pre-glucosylation followed by oxidation,^3–7) which require multiple steps with tedious isolation procedures for each step. Thus, we set out to develop a facile β-D-glucuronidation reaction using commercially available 1 as the glycosyl donor by scrutinizing the reaction conditions.

As reported previously,¹⁾ when 1 is activated with Tf₂NH (1.5 equivalent (equiv.)) in DCM at room temperature, neighboring group participation of the methyl ester at C–6 occurs more efficiently than that of the acetyl group at C–2 to reduce the energy level of the transition state, resulting in the formation of α-glucuronide. As a test, the reaction of (−)-menthol (2a) and 1 was performed in the presence of Tf₂NH (6 equiv.) at 0 to 5 °C in DCM, and only a 1 : 1 mixture of β- and α-glucuronides (2b and 2c) was obtained in 55% yield (Table 1, entry 1).

Table 1. Conditions and Results of β-Glucuronidation Reaction

Entry	Substrate (1 : 2a)	Acid (equiv.)	Solvent	Temp	Yield (%)	Ratio (2b : 2c)
1	3 : 1	Tf2NH (6 equiv.)	DCM	0–5 °C	55	1 : 1
2	5 : 1	Tf2NH (10 equiv.)	1,4-dioxane	10 °C	10	9 : 1
3	1 : 1.5	Tf2NH (2 equiv.)	THP	45 °C	16	5 : 3
4	1 : 1.5	Tf2NTMS (0.5 equiv.)	DMF	60 °C	0	—
5	1 : 1.5	Tf2NTMS (0.5 equiv.)	CH3CN	60 °C	15	0 : 1
6	1 : 1.5	Tf2NTMS (0.5 equiv.)	DME	60 °C	35	3 : 1
7	1 : 1.5	Tf2NTMS (0.5 equiv.)	THP	60 °C	34	7 : 2
8	1 : 1.5	Tf2NTMS (0.5 equiv.)	1,4-dioxane	60 °C	54	9 : 1

Accordingly, we hypothesized that aprotic polar solvents may facilitate the formation of a solvent-coordinated intermediate at the anomeric position, thereby blocking the α-face from the nucleophilic attack of alcohols^8–11) (Chart 1).

Chart 1. Plausible Mechanisms of α- and β-Glucuronidation Reactions Using 1 as the Glycosyl Donor

Results and Discussion

First, several reactions of 1 and 2a were performed using acetonitrile, tetrahydrofuran, dimethylformamide (DMF), 1,4-dioxane, or tetrahydropyran (THP) as the solvent in the presence of Tf₂NH at room temperature. Whereas almost no desired product was obtained in the reactions using acetonitrile, tetrahydrofuran, or DMF, the reaction using 1,4-dioxane with Tf₂NH (10 equiv.) at 10 °C gave β-glucuronide (2b) as the predominant product, albeit in low yield (9%) (Table 1, entry 2). The reaction using THP with Tf₂NH (2 equiv.) at 45 °C yielded 2b (10%) and 2c (6%) (Table 1, entry 3).

Thus, instead of Tf₂NH, trimethylsilyl bis(trifluoromethanesulfonyl)imide (Tf₂NTMS) was used as the activator, and the reaction was performed at the substrate ratio and the temperature of 1 : 1.5 and 60 °C, respectively, with only the solvent varied. The combination of Tf₂NTMS (0.5 equiv.) as the activator and 1,4-dioxane as the solvent was found to provide the best result, yielding predominantly β-D-glucuronide at a reaction temperature of 60 °C (Table 1, entry 8). Neither the chemical yield nor the anomer ratio was improved when polar solvents having boiling points higher than 60 °C were used in the reaction performed under the same conditions as those shown in Table 1, entry 8 (Table 1, entries 4–7).

Next, the reactions with other acceptor substrates, i.e., cholestanol (3a), (+)-borneol (4a), (−)-borneol (5a), and 2-adamantanol (6a), were performed under the same conditions as those used for Table 1, entry 8, and the results are shown in Table 2. Unfortunately, the ratios of the β-isomer to the α-isomer (β : α ratios) were not as high as that obtained in the reaction with 2a (Table 1, entry 8) when cholestanol, (+)-borneol, and 2-adamantanol were used as the acceptor substrate. (Table 2, entries 1, 2, and 4). Despite performing molecular model studies, we could not explain why the α/β anomer ratio of D-glucuronides in the reaction of 4a differed from that of 5a. MS4A is known to suppress the undesirable effects of Brønsted acids generated from Lewis acids with a trimethylsilyl group in the presence of a small amount of water.^12,13) We also considered that Tf₂NH, an excellent α-D-glucuronidation activator,¹⁾ might accelerate the production of the undesired α-isomer. The reaction of 1.5 equiv. of 4a with the same amount of 1 and 0.5 equiv. of Tf₂NTMS in the presence of MS4A gave only β-isomer 4b in a very low yield (8%) (Data not shown as a Table). It was reported that MS4A releases inorganic bases such as potassium carbonate in the presence of a polar solvent.¹⁴⁾ Tf₂NTMS might be consumed by the base, thereby suppressing its ability to sufficiently activate donor substrate 1. This hypothesis was supported by the reaction of 4a with 1 (1.5 equiv. each) and two times the amount of Tf₂NTMS (1.0 equiv.) in the presence of MS4A, which afforded a mixture of 4b and 4c (2 : 1) in an improved yield (26%) (Data not shown as a Table). After several attempts, we found that the reaction using excess amounts of the acceptor substrate (3 equiv.) and Tf₂NTMS (3 equiv.) relative to 1 in the presence of MS4A improved the β : α ratios compared with the reactions in Table 2 entries 1, 2, and 4 (Table 3).

Table 2. β-D-Glucuronidation of Terpenoid Alcohols in 1,4-Dioxane at 60 °C

Table 3. β-D-Glucuronidation of Terpenoid Alcohols in the Presence of MS4A

The reaction mechanism underlying the high β-selectivity can be explained by the anomeric orientation of 1,4-dioxane to form an intermediate, as shown in Chart 2. Specifically, after the intermediate is formed by the activation of 1 with Tf₂NTMS, where the cation at the anomeric carbon is stabilized by the methyl ester at C-6 via neighboring group participation, the oxygen atom of 1,4-dioxane attacks from the α-face. Here, the anomeric effect assists the axial orientation of polar 1,4-dioxane and hinders the nucleophilic attack of the hydroxyl group of alcohols from the α-face, thereby allowing β-D-glucuronidation to predominantly take place.

Chart 2. Plausible Mechanism of D-Glucuronidation with a High β:α Ratio

Conclusion

A facile and selective β-D-glucuronidation reaction using commercially available methyl 1,2,3,4-tetra-O-acetyl-β-D-glucuronate (1) as the glycosyl donor and Tf₂NTMS as the activator was achieved by employing 1,4-dioxane as the solvent at 60 °C. The addition of MS4A afforded better results when cholestanol, (+)-borneol, and 2-adamantanol were used as the acceptor substrate. The present method is useful and practical for the synthesis of biologically active, naturally occurring β-D-glucuronides, such as heparin, heparan sulfate, and glycyrrhizin, as well as artificial β-D-glucuronides used in drug metabolism studies, albeit with moderate yields.

Experimental

All ¹H- and ¹³C-NMR spectra of the products were measured in CDCl₃ by spectrometers operated at 400 MHz (100 MHz for ¹³C-NMR) at 20 °C. ¹H-NMR chemical shifts were recorded in parts per million (ppm, δ) relative to tetramethylsilane (δ = 0.00 ppm) as the internal standard. ¹³C-NMR chemical shifts were reported in ppm with the solvent as the reference peak (CDCl₃: δ = 77.0 ppm). Mass spectra were recorded on an electrospray ionization time-of-flight mass spectrometer (Micromass LCT). Column chromatography was performed using Merck Silica Gel 60 (60–230 mesh) with n-hexane and ethyl acetate as eluents for the isolation of the products. Analytical TLC was performed using Merck Silica Gel F254 plates (0.25 mm). Spots were detected by UV irradiation (254 nm) and/or staining with 5% phosphomolybdic acid followed by heating. Melting points were measured with a Yanaco Micro Melting Point Apparatus without any corrections.

Materials: Methyl 1,2,3,4-tetra-O-acetyl-β-D-glucuronate (1), Tf₂NH, and Tf₂NTMS are commercially available and used as received. All other starting materials were purchased and used without further purification. Organic solvents used in the reaction, including 1,4-dioxane, THP, and DCM, were freshly distilled over CaH₂ under a nitrogen atmosphere each time. MS4A was used after heating at 120 °C under reduced pressure of <1 mmHg for at least 3 h and lowering the temperature to room temperature.

Methyl 1-O-(−)-Menthyl 2,3,4-tri-O-acetyl-β-D-glucuronate (2b) [Reaction in the Absence of MS4A]

Tf₂NTMS (15 µL, 0.07 mmol) was added to a solution of 1 (52.8 mg, 0.14 mmol) and (−)-menthol (2a) (34.7 mg, 0.22 mmol) in 1,4-dioxane (3.0 mL) at room temperature, and the reaction mixture was stirred for 15 min at 60 °C. The reaction mixture was extracted with chloroform, and the organic layer was dried over anhydrous Na₂SO₄ and condensed in vacuo. The residue was purified by silica gel column chromatography (n-hexane : ethyl acetate = 4 : 1) to afford desired β-glucuronide (2b) (32.3 mg, 49%) as colorless crystals (recrystallized from ethyl acetate and n-hexane) along with α-glucuronide (2c) (3.6 mg, 5%). 2b: M.p. 128.1–128.5 °C. [α]²⁶_D=−70.1 (c 1.27, CHCl₃). ¹H-NMR (400 MHz, CDCl₃) δ: 0.73 (3H, d, J = 6.8 Hz, MeCH), 0.77–0.84 (2H, m), 0.87, 0.91 (each 3H, d, J = 6.8 Hz, MeCH), 0.92–1.05 (2H, m), 1.17–1.40 (2H, m), 1.60–1.69 (2H, m), 1.91–1.99 (1H, m), 2.020, 2.024, 2.036 (each 3H, s, 3 × Ac), 2.24 (1H, m), 3.42 (1H, td, J = 4.4, 10.8 Hz), 3.74 (3H, s, OMe), 4.00 (1H, d, J = 10.0 Hz, H-5′), 4.60 (1H, t, J = 8.0 Hz, H-2′), 5.20–5.29 (2H, m, H-3′, 4′). ¹³C-NMR (100 MHz, CDCl₃) δ: 15.2, 20.5, 20.59, 20.62, 20.8, 22.2, 24.9, 31.3, 34.1, 40.4, 47.4, 52.6, 69.4 (C-4′), 71.2 (C-2′), 72.3 (2C) (C-3′, 5′), 78.7, 98.4 (C-1′), 167.1, 169.2, 169.4, 170.2. High resolution (HR)-MS electrospray ionization (ESI): Calcd for C₂₃H₃₆O₁₀Na [M + Na]⁺ 495.2206. Found 495.2206.

Methyl 1-O-Cholestanyl 2,3,4-Tri-O-acetyl-β-D-glucuronate (3b) [Reaction in the Presence of MS4A]

Tf₂NTMS (60 µL, 0.26 mmol) was added to a solution of 1 (117.4 mg, 0.31 mmol), cholestanol (3a) (39.7 mg, 0.10 mmol), and MS4A (390 mg) in 1,4-dioxane (6.0 mL) at room temperature, and the reaction mixture was stirred for 30 min at 60 °C. The reaction mixture was extracted with chloroform, and the organic layer was dried over anhydrous Na₂SO₄ and condensed in vacuo. The residue was purified by silica gel column chromatography (n-hexane : ethyl acetate = 5 : 1) to afford desired β-glucuronide (3b) (20.2 mg, 27%) as colorless crystals (recrystallized from ethyl acetate and n-hexane) along with α-glucuronide (3c) (2.3 mg, 3%). 3b: M.p. 176.0–176.3 °C. [α]²⁶_D=−5.3 (c 1.13, CHCl₃). ¹H-NMR (400 MHz, CDCl₃) δ: 0.59 (1H, m), 0.64, 0.77 (each 3H, s, 2 × CH₃), 0.85, 0.86, 0.89 (each 3H, d, J = 6.6 Hz), 0.92-1.99 (30H, m), 2.02 (6H, s, 2 × Ac), 2.04 (3H, s, Ac), 3.58 (1H, m), 3.75 (3H, s, OMe), 4.01 (1H, d, J = 9.0 Hz, H-5′), 4.65 (1H, d, J = 8.0 Hz, H-1′), 4.96 (1H, t, J = 8.0 Hz, H-2′), 5.21 (1H, t, J = 9.0 Hz, H-4′), 5.25 (1H, t, J = 8.5 Hz, H-3′). ¹³C-NMR (100 MHz, CDCl₃) δ: 12.0, 12.2, 18.6, 20.5, 20.6, 20.7, 21.2, 22.5, 22.8, 23.8, 24.2, 28.0, 28.2, 28.8, 29.1, 32.0, 34.3, 35.4, 35.5, 35.7, 36.1, 36.9, 39.5, 39.9, 42.5, 44.6, 52.8, 54.3, 56.2, 56.4, 69.4 (C-4′), 71.4 (C-2′), 72.1 (C-3′), 72.5 (C-5′), 79.5, 99.2 (C-1′), 167.3, 169.2, 169.4, 170.2. HR-MS (ESI): Calcd for C₄₀H₆₄O₁₀Na [M + Na]⁺ 727.4397. Found 727.4398.

Methyl 1-O-(+)-Bornyl 2,3,4-Tri-O-acetyl-β-D-glucuronate (4b)

M.p. 136.6–136.9 °C. [α]²⁶_D=−10.7 (c 1.10, CHCl₃). ¹H-NMR (400 MHz, CDCl₃) δ: 0.82, 0.83, 0.84 (each 3H, s), 0.84–0.91 (1H, m), 1.10–1.32 (2H, m), 1.57–1.72 (2H, m), 1.80–1.89 (1H, m), 2.02, 2.03, 2.04 (each 3H, s, Ac), 2.12–2.24 (1H, m), 3.76 (3H, s, OMe), 3.79 (1H, m), 4.00 (1H, d, J = 10.0 Hz, H-5′), 4.05 (1H, J = 8.0 Hz, H-1′), 5.07 (1H, t, J = 8.0 Hz, H-2′), 5.21 (1H, dd, J = 8.0 Hz, 10.0 Hz, H-4′), 5.25 (1H, t, J = 8.0 Hz, H-3′). ¹³C-NMR (100 MHz, CDCl₃) δ: 13.6, 18.7, 19.6, 20.5, 20.6 (2C), 26.4, 27.8, 36.9, 44.9, 47.5, 49.4, 52.8, 69.5 (C-4′), 71.3 (C-2′), 72.0 (C-3′), 72.5 (C-5′), 87.8, 102.4 (C-1′), 167.3, 169.1, 169.4, 170.2. HR-MS (ESI): Calcd for C₂₃H₃₄O₁₀Na [M + Na]⁺ 493.2050. Found 493.2050.

Methyl 1-O-(−)-Bornyl 2,3,4-Tri-O-acetyl-β-D-glucuronate (5b)

M.p. 144.3–144.6 °C. [α]²⁶_D = −51.6 (c 1.06, CHCl₃). ¹H-NMR (400 MHz, CDCl₃) δ: 0.83, 0.84, 0.86 (each 3H, s, 3 x CH₃), 0.92 (1H, dd, J = 3.2, 13.2 Hz), 1.10 (1H, m), 1.19 (1H, m), 1.65 (1H, t, J = 4.4 Hz), 1.71 (2H, m), 1.83 (1H, ddd, J = 4.4, 9.6, 13.2 Hz), 2.02, 2.03, 2.06 (each 3H, s, 3 × Ac), 2.13 (1H, m), 3.76 (3H, s, OMe), 4.00 (1H, d, J = 8.0 Hz, H-5′), 4.52 (1H, d, J = 8.0 Hz, H-1′), 5.03 (1H, t, J = 8.0 Hz, H-2′), 5.23 (1H, t, J = 8.0 Hz, H-4′), 5.26 (1H, t, J = 8.0 Hz, H-3′). ¹³C-NMR (100 MHz, CDCl₃) δ: 13.2, 18.9, 19.6, 20.5, 20.58, 20.63, 26.3, 28.3, 35.9, 44.8, 47.9, 49.1, 52.8, 69.6 (C-4′), 71.1 (C-2′), 72.0 (C-3′), 72.6 (C-5′), 83.6, 99.4 (C-1′), 167.3, 169.0, 169.4, 170.2. HR-MS (ESI): Calcd for C₂₃H₃₄O₁₀Na [M + Na]⁺ 493.2050. Found 493.2050.

Methyl 1-O-(2-Adamantyl)-2,3,4-tri-O-acetyl-β-D-glucuronate (6b)

M.p. 184.3–184.7 °C. [α]²⁶_D=−39.5 (c 0.97, CHCl₃). ¹H-NMR (400 MHz, CDCl₃) δ: 1.45 (2H, br d, J = 12.4 Hz), 1.62-2.02 (12H, m), 2.025, 2.031, 2.04 (each 3H, s, 3 x CH₃), 3.76 (3H, s, OMe), 3.84 (1H, br s), 4.02 (1H, d, J = 9.0 Hz, H-5′), 4.66 (1H, d, J = 8.0 Hz, H-1′), 5.07 (1H, t, J = 8.0 Hz, H-2′), 5.23 (1H, t J = 9.0 Hz, H-4′), 5.27 (1H, t, J = 8.5 Hz, H-3′). ¹³C-NMR (100 MHz, CDCl₃) δ: 20.5, 20.66, 20.67, 27.0, 27.2, 30.8, 31.0, 31.3, 33.1, 36.1, 36.4, 37.3, 52.8, 69.6 (C-4′), 71.3 (C-2′), 72.1 (C-3′), 72.6 (C-5′), 81.4, 98.5 (C-1′), 167.3, 169.2, 169.4, 170.2. HR-MS (ESI): Calcd for C₂₃H₃₂O₁₀Na [M + Na]⁺ 491.1893. Found 491.1893.

Acknowledgments

This research was financially supported in part by a Grant from Kobayashi Foundation.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

References

• 1) Du T., Kaneko K., Sakai Y., Tsunoi M., Yoshitake T., Kotera A., Matsushima Y., Miura T., Kajimoto T., Chem. Pharm. Bull., 70, 589–593 (2022).
• 2) Furukawa T., Hinou H., Shimawaki K., Nishimura S.-I., Tetrahedron Lett., 52, 5567–5570 (2011).
• 3) Liptak A., Fekte A., Gyergyoi K., Kover K., Bajza I., Carbohydr. Res., 341, 1312–1321 (2006).
• 4) Herczeg M., Mezo E., Eszenyi D., Antus S., Borbas A., Tetrahedron, 70, 2919–2927 (2014).
• 5) Macchione G., de Paz J., Nieto L., Carbohydr. Res., 394, 17–25 (2014).
• 6) Xion C., Feng S., Qiao Y., Guo Z., Gu G., Chem. Eur. J., 24, 8205–8216 (2018).
• 7) Supria D., Lo H.-Y., Wong C.-H., J. Am. Chem. Soc., 141, 10309–10314 (2019).
• 8) Kanie O., Kiso M., Hasegawa A., J. Carbohydr. Chem., 7, 501–506 (1988).
• 9) Murase T., Ishida H., Kiso M., Hasegawa A., Carbohydr. Res., 184, c1–c4 (1988).
• 10) Hasegawa A., Ohki H., Nagahama T., Ishida H., Kiso M., Carbohydr. Res., 212, 277–281 (1991).
• 11) Hasegawa A., Nagahama T., Ohki H., Hotta K., Ishida H., Kiso M., J. Carbohydr. Chem., 10, 493–498 (1991).
• 12) Nielsen M. M., Pedersen C. M., Chem. Rev., 118, 8285–8358 (2018).
• 13) Paulsen H., Paal M., Carbohydr. Res., 135, 53–69 (1984).
• 14) Mizuno M., Kobayashi K., Nakajima H., Koya M., Inazu T., Synth. Commun., 32, 1665–1670 (2002).