Winolides A–C, bioactive sesquiterpene lactones with unusual 5,6-secoeudesmane frameworks from Inula wissmanniana

Abstract

Winolides A–C (1–3), three bioactive sesquiterpene lactones with irregular 5,6-secoeudesmane frameworks, which probably arise from different dienol-benzene rearrangements, were isolated from Inula wissmanniana. Their structures and absolute configurations were established by extensive analysis of spectroscopic data, computational methods, and asymmetric synthesis.

---

Sesquiterpene lactones (SLs), a large group of structurally diverse natural products, are prevalently produced in plants from the Asteraceae family.¹ The isolation and synthesis of SLs have attracted considerable endeavors, due to their intriguing skeletons and biological activities.² Notably, the presence of an α-methylene-γ-butyrolactone unit in most SLs, which plays a pivotal role as Michael acceptor for biological nucleophiles, has been profiled as a structural prerequisite for their anti-cancer, anti-inflammatory, and anti-malarial properties.^1,2 For instance, SL parthenolide (Fig. 1) exhibits exceptional anti-cancer and anti-inflammatory activities, which has resulted in its involvement in cancer clinical trials.¹

Fig. 1 Examples of naturally occurring sesquiterpenoids.

Most α-methylene-γ-butyrolactone derivatives, either naturally occurring or tailored synthetically, are characterized by a disubstituted 3-methylenedihydrofuran-2-one ring, whereas only a few are in mono-substitution patterns (Fig. 1), exemplified by linear SLs from the Anthemis genus with undetermined configuration.³ Plants of the Inula genus (Asteraceae family) are such pharmacologically important resources with a plethora of bioactive SLs on which numerous phytochemical investigations have been carried out to afford ca. 360 SLs in disubstitution patterns.^2b Our previous phytochemical investigation on Inula wissmanniana, a plant exclusively distributed in China and popularly used in medicines of the Yi ethnic group for promoting digestion and replenishing energy, have led to the isolation of novel bioactive sesquiterpenic acids and SLs.⁴ In continuation of the investigation, winolides A–C (1–3, Fig. 2), three potential anti-inflammatory SLs in mono-substitution patterns with unusual 5,6-secoeudesmane frameworks, which are probably generated via different dienol-benzene rearrangements were obtained. In this communication, we present the structural elucidation, probable biogenetic pathway, and potential anti-inflammatory activities of 1–3. Moreover, this work demonstrates that the absolute configuration of the mono-substituted α-methylene-γ-butyrolactone could be quickly determined by means of a CD technique.

Fig. 2 Structures of winolides A–C (1–3).

Winolide A (1)§ was obtained as colorless oil with optical activity ([α]²⁰_D −91.2, c = 0.1, CH₂Cl₂). Its molecular formula of C₁₅H₁₈O₂, with seven degrees of unsaturation, was established by means of HRESIMS (m/z 253.1206, [M + Na]⁺) and NMR spectroscopy (Table 1). In the ¹H NMR spectrum, methyl (δ_H 2.31, 6H, s), oxymethylene (δ_H 4.55, dd, J = 8.8, 8.4 Hz; 4.09, dd, J = 8.8, 5.2 Hz), olefinic methylene (δ_H 6.33 and 5.67, d, J = 2.6 Hz), and aromatic (δ_H 7.02, 3H, m) protons were clearly apparent. The ¹³C NMR and DEPT spectra exhibited 15 carbon resonances, including two methyls, four methylenes, four methines, and five quaternary carbons, further indicating the presence of a mono-substituted α-methylene-γ-butyrolactone unit and a symmetrical tri-substituted benzene ring. All protons were ascribed to their corresponding carbons by a HSQC experiment. The HMBC correlations from the methyl singlet at δ_H 2.31 to carbon resonances at δ_C 128.4, 135.7, and 137.5 established a symmetrical 2,6-dimethylphenyl unit. The spin system (H₂-9/H₂-8/H-7/H₂-6, CH₂CH₂CHCH₂) established by a ¹H–¹H COSY experiment coupled with the HMBC correlations from H₂-9 to C-4 and C-10, from H₂-6 to C-11 and C-12, and from H₂-13 to C-7 and C-12 (Table 1), gave rise to the planar structure of 1 as an irregular SL in a mono-substitution pattern. Since single crystals of 1 cannot be obtained despite many attempts, a computational modeling study of the optical rotation value and ECD spectrum of 1 was performed using a time-dependent density functional theory (DFT) method in the Gaussian 03 program to determine the absolute configuration of the only chiral center (C-7).⁵

Table 1 ¹H and ¹³C NMR data and HMBC correlations for 1^a

No.	δC	δH (mult, J in Hz)	HMBC (H → C)
a Data were recorded in CDCl3 at 400 MHz for ^1H NMR and 100 MHz for ^13C NMR.
1	128.4 d	7.02 m (overlap)	C-3, C-10, C-5, C-14
2	126.2 d	7.02 m (overlap)	C-1, C-3, C-4, C-10
3	128.4 d	7.02 m (overlap)	C-1, C-4, C-5, C-15
4	135.7 s
5	137.5 s
6	71.0 t	4.55 dd (8.8, 8.4)	C-8, C-11, C-12
		4.09 dd (8.8, 5.2)
7	39.2 d	3.18 m	C-12, C-13
8	32.9 t	1.84 m; 1.70 m	C-5, C-6, C-11
9	26.4 t	2.67 m	C-4, C-10, C-7
10	135.7 s
11	138.1 s
12	170.6 s
13	122.2 t	6.33 d (2.6)	C-7, C-11, C-12
		5.67 d (2.6)
14	19.8 q	2.31 s	C-1, C-10, C-5
15	19.8 q	2.31 s	C-3, C-4, C-5

---

A conformational search was carried out using molecular mechanics methods with the MMFF94 force field in Maestro7.5. The obtained conformations were optimized using ab initio DFT at the B3LYP/6-31G** level to give stable conformers of the enantiomers (S-1, R-1), three of which were further determined as mainly populated conformers for each enantiomer (S-1a–S-1c, R-1a–R-1c, see Tables S1 and S2, ESI†) by conformational analysis. The optical rotation values of S-1a–S-1c and R-1a–R-1c were calculated at the B3LYP-SCRF/6-31G(d)//B3LYP/6-311++G(2d, p) level with the PCM model in CH₂Cl₂ solution, which were further weighted by the Boltzmann population to obtain the rotation values of S-1 and R-1 (−78.2 and +78.2, see Table S3, ESI†). The comparison between the experimental and calculated optical rotation values suggested the consistency of 1 and S-1 in this profile.

The ECD spectra of S-1a–S-1c and R-1a–R-1c were calculated at the B3LYP-SCRF/6-31G*//B3LYP/6-31G* level with the COSMO model in MeOH solution, further affording the weighted ECD spectra of S-1 and R-1, as shown in Fig. 3. According to the calculated spectra, the ECD spectrum generated by S-1 was consistent with the experimental one. Molecular orbital (MO) analysis of conformer S-1b at the B3LYP-SCRF/6-31G*//B3LYP/6-31G* level with the COSMO model in MeOH solution, interpreted the origin of the ECD spectrum of S-1 at molecular levels. The diagnostic rotatory strength at 212 nm was mainly contributed by the electronic transitions from HOMO62 to LUMO63, involving the electrons of α,β-unsaturated ketone (Fig. 4), which is in good accordance with the strong negative Cotton effect at 223 nm of the experimental spectrum of 1. Therefore, the absolute configuration of 1 was indicated to be 7S.

Fig. 3 Calculated ECD spectra of R/S-1 and experimental CD spectra of compounds 1, 3, R/S-6, and R/S-8 in MeOH. The calculated ECD spectra of 1 represents the Boltzmann population-weighted averages of 1a–1c.

Fig. 4 Some molecular orbitals involved in the key transitions in the ECD of conformer S-1b at the B3LYP-SCRF/6-31G*//B3LYP/6-31G* level with the COSMO model in MeOH solution.

The carbon skeleton of 1 was such an irregular framework abhorrent to the isoprene rule that no precedent was found from nature. For the unambiguous structural and configurational determination of 1, asymmetric synthesis of a series of mono-substituted α-methylene-γ-butyrolactones (5–9) was further achieved according to the route shown in Scheme 1. Enantiomers of lactones 5 and 6 were prepared by an asymmetric metal catalysed Alder-ene reaction using [Rh((R/S)-BINAP)]SbF₆ exactly according to the method previously reported.^2d Enantiomers 6 were then reacted with different nucleophiles to facilitate the following sp³-C–sp²-C cross-coupling, where alkyl bromides 7 were found optimal for a Hiyama cross-coupling with arylsilanes to afford the demethyl analogues 8. The absolute configurations of lactones 5–8 were determined by X-ray crystallographic analysis of the p-bromophenyl carbamates (9) as well as their optical rotation values (Table S3, ESI†) and CD curves (Fig. 3). The diagnostic Cotton effects at 212 nm in the CD spectra of 6 and 8 confirmed the computational conclusion and suggested that the CD technique allows a quick determination of mono-substituted α-methylene-γ-butyrolactones. Consequently, the absolute stereochemistry of 1 was clearly determined to be 7S.

Scheme 1 Synthesis of lactones 5–9 in mono-substitution patterns and their configurational determination achieved by X-ray crystallographic analysis of 9 (CuKα). Reagents and conditions: (i) DCC, DMAP, DCM, −30 °C rt, 14 h; (ii) [Rh(cod)Cl]₂, AgSbF₆, R/S-BINAP, DCE, rt, 21 h; (iii) BH₃·THF, rt, 10 min; (iv) PPh₃, CBr₄, DCM, rt, 2 h; (v) PdBr₂, [HP(t-Bu)₂Me]BF₄, TBAF, (MeO)₃SiPh, THF, rt, 12 h; (vi) Et₃N, p-bromophenyl carbamate, DCM, rt, 1 h.

Winolide B (2) and C (3) were isolated from I. wissmanniana as well. The NMR data of 2 clearly revealed it to be a reduced derivative of 1, with a typical α-methyl-γ-butyrolactone unit confirmed by ¹H–¹H COSY and HMBC experiments (Fig. S5, ESI†). The relative configuration of 2 was established by a NOESY experiment using deuterated pyridine. The NOESY correlations of H-7 with H-6b and H₃-13 suggested their α orientation, while those of H-11 with H-6a suggested their β orientation (Fig. S6, ESI†). Thus, the structure of 2 was elucidated as depicted in Fig. 1. Winolide C (3) is a structural isomer of 1, as evidenced by their same molecular formula and similar NMR data. The ¹H and ¹³C NMR data of 3 revealed the presence of an asymmetric 2,3-dimethylphenyl rather than a 2,6-dimethylphenyl, which was further confirmed by a HMBC experiment (Fig. S5, ESI†). Likewise, the negative Cotton effect at 218 nm in the CD spectrum of 3 (Fig. 3) revealed its 7S configuration. With regard to tridensenal (Fig. 1), a sesquiterpene aldehyde bearing a non-aromatic carbon skeleton isolated from the Taiwanese liverwort Bazzania tridens,⁶ compound 3 is the second instance of its skeleton from nature.

A possible biogenetic pathway for 1–3 was proposed as shown in Scheme 2. Similar to the dienone-phenol rearrangements of santonin,⁷ cyclohexadienol 10 might rearrange to produce various aromatic products. The cyclohexadienone precursors and rearrangement products of CH₃-14 to C-1 have been isolated from I. wissmanniana in our previous study,^4a whereas 1–3 featured the rare rearrangements of CH₂-9 or CH₃-14 to C-5 and the subsequent cleavage of C-5 and C-6. It seems probable that the eudesmane intermediate 11 would rearrange to give a spiro-intermediate 12 as the reported rearrangement of eremophilane intermediate 13.^4b Intermediate 12 could then undergo a nucleophilic reaction and cleavage to yield 1. To date, there is no biological or chemical report about the disruption of C-5–C-6 bond in eudesmanes. Therefore, the formation of 1–3 would provide a new insight into the structural transformation of eudesmanes.

Scheme 2 Plausible biogenetic pathway of 1–3.

Inducible nitric oxide synthase-derived nitric oxide (NO) plays an important role in inflammatory conditions, and inhibitors of NO production are considered as potential anti-inflammatory agents.⁸ Previously, we have reported eudesmane and germacrane SLs in disubstitution patterns from I. wissmanniana with IC₅₀ values of NO inhibitory effects ranging from 0.65 to 10.28 μM.⁴ In the present study, the natural and synthetic lactones in mono-substitution patterns (Table 2) were evaluated for their inhibitory effects against lipopolysaccharide-induced NO production in RAW264.7 macrophages according to a previously reported method.⁴ Compound 1 exhibited significant inhibition against NO release with an IC₅₀ value of 0.38 μM, whereas those of 2 and 15 enormously diminished (IC₅₀ > 20 μM), supporting the α-methylene-γ-butyrolactone core to be a structural prerequisite for their biological activities. Moreover, the configuration of the α-methylene-γ-butyrolactone core greatly affected its NO inhibitory effects, as can be observed from the comparisons of enantiomers 5, 6, 8, and 9 (Table 2). This indicated that the S-configuration of the mono-substituted α-methyl-γ-butyrolactone unit would lead to a stronger NO inhibitory effect. Note that compound 1 presented much stronger NO inhibitory effect than 3 and S-8. This was a result of the structural differences in the phenyl unit, which indicated that chemical modification on the aromatic core would improve the biological activities of these unusual SLs.

Table 2 Inhibitory effects of the natural and synthetic lactones against NO production in RAW264.7 macrophages (n = 4)

Compounds	IC50 values (μM)
1	0.38 ± 0.04
2	>20
3	6.09 ± 0.51
R-5	10.01 ± 0.86
S-5	7.94 ± 0.46
R-6	6.08 ± 0.69
S-6	5.14 ± 0.52
R-8	5.02 ± 0.41
S-8	1.16 ± 0.12
R-9	0.72 ± 0.09
S-9	0.35 ± 0.05
15	>20
Aminoguanidine	0.37 ± 0.06

---

In summary, we have discovered three irregular SLs named winolides A–C (1–3) that possess unusual rearranged 5,6-secoeudesmane frameworks from I. wissmanniana. To the best of our knowledge, 1–3 represent the first cyclic SLs in mono-substitution patterns from nature. Moreover, the determination of absolute configuration of mono-substituted α-methylene-γ-butyrolactones is firstly achieved and simplified. It is noteworthy that compound 1 exhibited much stronger NO inhibitory activity than any SL in disubstitution patterns from I. wissmanniana.⁴ It is obvious that this new type of cyclic SLs in mono-substitution patterns will be an interesting target for synthetic chemistry and biochemical pharmacology.

Acknowledgements

This work was supported by the NSFC (81230090), the Global Research Network for Medicinal Plants (GRNMP), the Shanghai Leading Academic Discipline Project (B906), the National Key Technology R&D Program of China (2012BAI29B06), the 12th Five-Year Plan for Science and Technology Development (2012BAD33B05), and the Natural Science Foundation of Jiangsu Province, China (BK20140147).

Notes and references

1. (a) A. Ghantous, H. Gali-Muhtasib, H. Vuorela, N. A. Saliba and N. Darwiche, Drug Discovery Today, 2010, 15, 668–678; (b) A. Ghantous, A. Sinjab, Z. Herceg and N. Darwiche, Curr. Drug Targets, 2013, 18, 894–905.
2. (a) D. Chaturvedi, Opportunity, challenge and scope of natural products in medicinal chemistry, Research Signpost, Kerala, India, 2011, pp. 313–334; (b) G. W. Wang, J. J. Qin, X. R. Cheng, Y. H. Shen, L. Shan, H. Z. Jin and W. D. Zhang, Expert Opin. Invest. Drugs, 2014, 23, 317–345; (c) X. R. Cheng, Q. Zeng, J. Ren, J. J. Qin, S. D. Zhang, Y. H. Shen, J. X. Zhu, F. Zhang, R. J. Chang, Y. Zhu, W. D. Zhang and H. Z. Jin, Eur. J. Med. Chem., 2011, 46, 5408–5415; (d) D. M. Hodgson, E. P. A. Talbot and B. P. Clark, Org. Lett., 2011, 13, 5751–5753; (e) F. Zhang, Y. Yang, L. Xie and X. Xu, Chem. Commun., 2013, 49, 4697–4699.
3. (a) J. D. Staneva, M. N. Todorova and L. N. Evstatieva, Biochem. Syst. Ecol., 2005, 33, 97–102; (b) M. Todorova, J. Staneva, P. Denkova and L. Evstatieva, Nat. Prod. Res., 2008, 22, 907–914.
4. (a) X. R. Cheng, S. D. Zhang, C. H. Wang, J. Ren, J. J. Qin, X. Tang, Y. H. Shen, S. K. Yan, H. Z. Jin and W. D. Zhang, Phytochemistry, 2013, 96, 214–222; (b) X. R. Cheng, C. H. Wang, P. L. Wei, X. F. Zhang, Q. Zeng, S. K. Yan, H. Z. Jin and W. D. Zhang, Fitoterapia, 2014, 95, 139–146.
5. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr, T. Vreven, K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, C. Gonzalez and J. A. Pople, GAUSSIAN 03, revision B.05, Gaussian Inc, Wallingford, CT, 2004.
6. C. L. Wu, S. J. Chang, M. Tori, H. Furuta, A. Sumida and Y. Asakawa, J. Chin. Chem. Soc., 1990, 37, 387–391.
7. (a) S. M. Bloom, J. Am. Chem. Soc., 1958, 80, 6280–6283; (b) B. Miller, Acc. Chem. Res., 1975, 8, 245–256.
8. J. N. Sharma, A. Al-Omran and S. S. Parvathy, Inflammopharmacology, 2007, 15, 252–259.

---

Footnotes

† Electronic supplementary information (ESI) available: 1D and 2D NMR, MS, UV, IR spectra and data for 1–9, crystallographic data for S-9 and R-9. CCDC 934198 and 934199. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c4ra05131a

‡ These three authors have contributed equally to this work.

§ Winolide A (1): colorless oil; C₁₅H₁₈O₂; [α]²⁰_D −91.2 (c 0.10, CH₂Cl₂); UV (MeOH) λ_max (log ε): 220 (3.15) nm; IR (KBr) ν_max 2952, 1765, 1661, 1489, 1405, 1267, 1165, 1115, 1013, 947 cm⁻¹; ESIMS m/z 253.1 [M + Na]⁺, m/z 229.1 [M − H]⁻; HR-ESI-MS m/z 253.1206 [M + Na]⁺ (calcd for C₁₅H₁₈O₂Na, 253.1204); ¹H and ¹³C NMR data, see Table 1.

---