Computation-Guided Total Synthesis of Vitisinol G

Abstract

Computational chemistry is useful in synthetic organic chemistry, as it can be used not only to analyze reaction mechanisms, but also to calculate biosynthetic pathways and to plan and evaluate strategies for total syntheses. Here we report the computation-guided total synthesis of vitisinol G, a resveratrol dimer.

Introduction

Resveratrol (1) is a causative agent of the French paradox.^1,2) Resveratrol oligomers (ROs) are natural products starting from resveratrol that possess beneficial pharmacological activities for human health.^3,4) Oxidative multimerization of resveratrol has produced many ROs, which have greatly attracted biologists and synthetic organic chemists as research targets.^5–11)

Computational analysis has been used to simulate biosynthetic pathways. Computational simulations enable organic chemists to evaluate biomimetic synthetic plans, propose undiscovered natural products and intermediates, and elucidate chemical structures.^12–16) We recently reported the asymmetric total syntheses of various resveratrol dimers derived from calculating biosynthetic transformations of resveratrol (1).¹⁷⁾ In that report, generation of the core skeleton of acuminatol (6) was calculated from epoxy-ε-viniferin (3) via 5: epoxide ring-opening followed by the Friedel–Crafts reaction.

Further calculations revealed a different transition state from the cationic intermediate: a semi-pinacol rearrangement of the hydrobenzofuran moiety to afford aldehyde (7) (Chart 1). Friedel–Crafts reaction applied to the aldehyde (7) led to the construction of a seven-membered ring with the positions of phenol and alcohol reversed. We found that the skeleton (8) was the same as that of vitisinol G (9), which was isolated with other resveratrol dimers by Huang in 2009.¹⁸⁾ These calculations guided us to a synthetic strategy for vitisinol G (9), and we report its total synthesis herein.

Chart 1. Calculated Biosynthetic Transformation of Acuminatol and Vitisinol G

Results and Discussion

We first performed computational analysis of the Friedel–Crafts reaction and semi-pinacol rearrangement. The structure was optimized with the Gaussian 16 program (Revision C.01)¹⁹⁾ at 298.15 K, using the ωB97X-D²⁰⁾ functional with an ultrafine grid and the 6-31G(d,p) basis sets, which were found to be optimal in our benchmark study.¹⁷⁾ The energy was corrected by single-point energy calculations using 6-311++G(d,p) basis sets with the same functional. All calculations were performed in water, methanol or dichloromethane solvent using the solvation model based on density (SMD²¹⁾). Harmonic vibrational frequencies were computed with the same level of theory to confirm that no imaginary vibration was observed for the optimized structure, and only one imaginary vibration was observed for the transition state. The intrinsic reaction coordinate (IRC²²⁾) method was used to track minimum energy paths from the transition structures to the corresponding local minima. Initial structures for the calculations were based on our previous report.¹⁷⁾ In our calculation of the biosynthetic transformation, the activation energy for the semi-pinacol rearrangement was 10.3 kcal/mol, while the competing Friedel–Crafts reaction was 5.6 kcal/mol (Chart 1). Therefore, we considered that it would be difficult to prioritize semi-pinacol rearrangement with unprotected phenol in the total synthesis. Therefore, we changed the protecting group on phenol and again compared the activation energies of these two reactions (Fig. 1). We selected the acetyl group for simulation of the reaction, which revealed that the activation energy of the Friedel–Crafts reaction was 14.7 kcal/mol (TS2_Ac)—significantly higher than that of the unprotected substrate (TS2_H). On the other hand, the activation energy of the semi-pinacol rearrangement was 8.1 kcal/mol (TS1_Ac)—lower than that of the unprotected substrate (TS1_H). This result was attributed to the difference in the reaction mechanism. The Friedel–Crafts reaction is a nucleophilic addition by π-electrons, and the introduction of an acetyl group decreases the electron density of the aromatic ring, thereby increasing the activation energy. On the other hand, in the semi-pinacol rearrangement, the σ bond perpendicular to the π orbital reacts rather than the π electrons, resulting in almost no change in the activation energy. Comparison of the transition state structures revealed that the interatomic distance between the newly formed bonds is 0.15 Å shorter in TS2_Ac than in TS2_H, indicating that the orbital coefficient of the π-electron in TS2_Ac is reduced by the acetyl group, thereby leading to a late transition state. On the other hand, the interatomic distance in TS1_Ac is 0.08 Å longer than in TS1_H, indicating that the change in the orbital coefficient of the π-electron has a smaller effect on the transition state. Moreover, it has been reported that deacetylation of SM_Ac under basic conditions leads to the Friedel–Crafts reaction.²³⁾ Thus, our analysis suggested that the use of acetyl-protected substrates would afford the aldehyde under acidic conditions.

Fig. 1. Energy Diagram for the Friedel–Crafts Reaction and Semi-pinacol Rearrangement

Based on the above calculations, we began our synthetic studies toward the total synthesis of vitisinol G (Chart 2). We synthesized epoxide (11) by dimerization of resveratrol (1) followed by acetylation and epoxidation.^23,24) We then attempted the semi-pinacol rearrangement. When 11 was subjected to various Lewis acids, such as In(OTf)₃, Bi(OTf)₃, Yb(OTf)₃, or Gd(OTf)₃, the desired semi-pinacol rearrangement proceeded and generation of the aldehyde (12) was observed by ¹H-NMR analysis. Isolating the aldehyde (12), was difficult, however, because of its instability. Thus, we decided to lead the aldehyde to vitisinol G (9) by a one-pot procedure without isolating the aldehyde (12). Because the disappearance of the epoxide was only confirmed with Gd(OTf)₃, we attempted the one-pot reaction using Gd(OTf)₃.

Chart 2. Attempted Semi-pinacol Rearrangement

After confirming by TLC analysis that the epoxide (11) was consumed, we added hydrogen chloride in methanol and heated the reaction mixture, which resulted in vitisinol G (12) in 40% via deacetylation, Friedel–Crafts reaction, and dehydration (Table 1, entry 1). To improve the yield, we next attempted p-toluenesulfonic acid monohydrate, which successfully yielded vitisinol G in 60% yield (Table 1, entry 2). According to our calculations, this reaction should have proceeded even under Brønsted acid conditions. When using hydrochloric acid or p-toluenesulfonic acid monohydrate instead of gadolinium triflate, however, no reaction occurred at room temperature and complex mixture was given via deacetylation under heating conditions (Table 1, entries 3, 4). Bronsted acids can be modeled as protons, but it is not easy to model Lewis acids because of their huge variety. Therefore, even if a modeled Brønsted acid is assumed in the calculation, we think Lewis acids should also be examined.

Table 1. Total Synthesis of Vitisinol G and Screening of the Reaction Condition

Entry	Acid	Yield
1	HCl	40%
2	TsOH·H2O	60%
3a)	HCl	0%
4a)	TsOH·H2O	0%

a) An acid in the table was used in the first step instead of Gd(OTf)₃.

Interestingly, when cis-10 was applied to the one-pot reaction after the epoxidation, vitisinol G (9) was obtained and no epi-vitisinol G (13) was observed, indicating that epimerization proceeded during the process²⁵⁾ (Chart 3).

Chart 3. Total Synthesis of Vitisinol G via the Epimerization

Thus, we performed density functional theory (DFT) calculations to analyze the isomerization pathways of epi-vitisinol G and vitisinol G (Fig. 2). The results revealed that the isomerization proceeds via ring-opening by cleavage of the C–O bond of hydrobenzofuran (TS3) followed by C–C bond rotation (TS4) and ring closing to regenerate the hydrobenzofuran (TS5). The activation energies for each step are less than 20 kcal/mol, indicating that the epimerization easily proceeds under acidic and heating conditions. Vitisinol G is 4.2 kcal/mol more stable than epi-vitisinol G, suggesting that vitisinol G is thermodynamically produced.

Fig. 2. Energy Diagram for the Epimerization

Conclusion

We focused on the semi-pinacol rearrangement derived from calculations of the biosynthetic pathway of a resveratrol dimer and conducted a total synthetic study of vitisinol G. DFT calculations were used to calculate the biosynthetic pathway, develop a synthetic strategy, and elucidate the reaction mechanism, and the total synthesis of vitisinol G at 7.8% was achieved in 4 steps. With advances in computer technology, methods for simulating experimental results will become mainstream. We hope that such computation-guided methods will become the global standard for synthetic organic chemists.

Acknowledgments

This work was supported by the Uehara Memorial Foundation and JSPS KAKENHI (Grant Number: 22H02741). Numerical calculations were carried out on SR24000 computer at the Institute of Management and Information Technologies, Chiba University of Japan.

Conflict of Interest

The authors declare no conflict of interest.

Supplementary Materials

This article contains supplementary materials.

References

• 1) Jang M., Cai L., Udeani G. O., Slowing K. V., Thomas C. F., Beecher C. W. W., Fong H. H. S., Farnsworth N. R., Kinghorn A. D., Mehta R. G., Moon R. C., Pezzuto J. M., Science, 275, 218–220 (1997).
• 2) Fremont L., Life Sci., 66, 663–673 (2000).
• 3) Qiao H., Chen X., Xu L., Wang J., Zhao G., Hou Y., Ge H. M., Tan R.-X., Li E., FASEB J., 27, 4561–4571 (2013).
• 4) Shen J., Zhou Q., Li P., Wang Z., Liu S., He C., Zhang C., Xiao P., Molecules, 22, 2050 (2017).
• 5) Snyder S. A., Zografos A. L., Lin Y., Angew. Chem. Int. Ed., 46, 8186–8191 (2007).
• 6) Nicolaou K. C., Wu T. R., Kang Q., Chen D. Y. K., Angew. Chem. Int. Ed., 48, 3440–3443 (2009).
• 7) Nicolaou K. C., Kang Q., Wu T. R., Lim C. S., Chen D. Y. K., J. Am. Chem. Soc., 132, 7540–7548 (2010).
• 8) Snyder S. A., Gollner A., Chiriac M. I., Nature (London), 474, 461–466 (2011).
• 9) Snyder S. A., Thomas S. B., Mayer A. C., Breazzano S. P., Angew. Chem. Int. Ed., 51, 4080–4084 (2012).
• 10) Jepsen T. H., Thomas S. B., Lin Y., Stathakis C. I., de Miguel I., Snyder S. A., Angew. Chem. Int. Ed., 53, 6747–6751 (2014).
• 11) Keylor M. H., Matsuura B. S., Griesser M., Chauvin J.-P. R., Harding R. A., Kirillova M. S., Zhu X., Fischer O. J., Pratt D. A., Stephenson C. R. J., Science, 354, 1260–1265 (2016).
• 12) Tantillo D. J., Nat. Prod. Rep., 28, 1035–1053 (2011).
• 13) Sato H., Hashishin T., Kanazawa J., Miyamoto K., Uchiyama M., J. Am. Chem. Soc., 142, 19830–19834 (2020).
• 14) Tantillo D. J., Chem. Soc. Rev., 47, 7845–7850 (2018).
• 15) Elkin M., Newhouse T. R., Chem. Soc. Rev., 47, 7830–7844 (2018).
• 16) Bemis C. Y., Ungarean C. N., Shved A. S., Jamieson C. S., Hwang T., Lee K. S., Houk K. N., Sarlah D., J. Am. Chem. Soc., 143, 6006–6017 (2021).
• 17) Nakajima M., Adachi Y., Nemoto T., Nat. Commun., 13, 152 (2022).
• 18) Chiou W.-F., Shen C.-C., Chen C.-C., Lin C.-H., Huang Y.-L., Planta Med., 75, 856–859 (2009).
• 19) Frisch M. J., Trucks G. W., Schlegel H. B., et al., Gaussian 16, Rev. C.01 Gaussian, Inc.: Wallingford, CT, 2016.
• 20) Chai J.-D., Head-Gordon M., Phys. Chem. Chem. Phys., 10, 6615–6620 (2008).
• 21) Marenich A. V., Cramer C. J., Truhlar D. G., J. Phys. Chem. B, 113, 6378–6396 (2009).
• 22) Fukui K., Acc. Chem. Res., 14, 363–368 (1981).
• 23) Takaya Y., Yan K.-X., Terashima K., Ito J., Niwa M., Tetrahedron, 58, 7259–7265 (2002).
• 24) Yao C.-S., Lin M., Wang Y.-H., Chin. J. Chem., 22, 1350–1355 (2004).
• 25) Natori Y., Tsutsui H., Sato N., Nakamura S., Nambu H., Shiro M., Hashimoto S., J. Org. Chem., 74, 4418–4421 (2009).