English

• Sesquiterpene lactones, such as eudesmanolide, guaianolide, and germacranolide, have been widely reported from Compositae plants, with the α-methylene-γ-lactone moiety connected to other building blocks to form diverse structurally intriguing and biologically active adducts [1-7]. Although the isolated sesquiterpene lactone oligomers present a dramatic increase [8], hybrids of sesquiterpene lactones especially eudesmanolide hybrids, are extraordinary rare. So far, only two types of eudesmanolide hybrids, eudesmanolide−furan [4] and eudesmanolide...alkaloid [9], have been isolated from natural resources. Structurally, both of them possessed the furan or alkaloid moiety linkage to eudesmanolide skeleton only via the methyl at C-11. Due to the fascinating structures, eudesmanolide hybrids have attracted broad interest from natural products and synthetic chemists since 2019 [6, 9].

  A recently research published in Cell revealed that neutrophil elastase (NE) was a major anticancer protein released by human neutrophils, which specifically activated the cell death pathway in cancer cells. Researchers observed that NE initiated a complex cancer killing program that inhibited cell survival pathways, induced DNA damage, increased the production of mitochondrial reactive oxygen species, and ultimately activated programmed cell death, that is so-called apoptosis [10]. It made NE became a hot topic in various scientific communities. However, apart from these inspiring reports, the presence and value of natural products possessing effect on NE remain largely unexplored.

  The genus Atractylodes (Compositae) is represented by a relatively limited number of species and mainly distributed in eastern region of Asia. Atractylodes macrocephala Koidz, is a perennial herb belonging to the Atractylodes genus, which is widely distributed in China, Korea, and Japan [11]. The rhizomes of A. macrocephala are the certified plant source for "Bai zhu", a common traditional Chinese medicine that is often used for the treatment of splenic hypofunction with inappetence, edema, spontaneous sweating, etc. [12]. Nowadays, "Bai zhu" is also one of the ingredients in Qingfei Paidu decoction, a key traditional Chinese medicine (TCM) formula used for treatment of COVID-19 in China [13]. Extensive phytochemical studies from A. macrocephala revealed that sesquiterpenes were the main chemical constituents [14]. However, there was remained without a significant breakthrough for this plant.

  Our research group has already reported a number of sesquiterpenes from A. macrocephala [15]. Continuing the search for structurally diverse and biologically interesting metabolites from this plant, we turned our focus toward eudesmanolide hybrids. This effort resulted in the isolation of three eudesmanolide sesquiterpene-phenol hybrids with extraordinary C—C linkage, named atramacronoids A−C (1−3). Remarkably, compounds 1−3 represent the first example of naturally occurring unusual hybrids constructed from eudesmanolide sesquiterpene and phenol, possessing pentacyclic 6/6/5/5/6 skeleton by formation of C-8−C-16 and C-7−O−C-17. Compound 2 was 18-demethyl derivative of 1, while 3 was a C-11 epimer of 1 (Fig. 1). Herein, the isolation, structure elucidation, possible biosynthetic pathways of these isolates, as well as preliminary mechanism of induced apoptosis through promoting the synthesis of NE, are described. As representative pioneering works, the new linkage pattern of these compounds is not only crucial for the chemical diversity and biosynthesis of sesquiterpene hybrids, but also for the pharmacological studies on anti-cancer.

  Figure 1

  

  Figure 1.  Structures of compounds 1–3.

  DownLoad: Full-Size Img PowerPoint

  Atramacronoid A (1) was obtained as colorless needle crystals. Its molecular formula was deduced as C₂₂H₂₆O₄ from the protonated molecular ion peak at m/z 355.1900 [M + H]⁺ (calcd. for C₂₂H₂₇O₄, 355.1904) in HRESIMS, which corresponded to 10 indices of hydrogen deficiency. The IR spectrum clearly exhibited absorption bands of hydroxyl (3421 cm^–1), γ-lactone (1778 cm^–1), and unsubstituted double bond (1644 cm^–1) functional groups. The ¹H NMR data (Table S1 in Supporting information) displayed the diagnostic signals of an exocyclic double bond at δ_H 4.86 (1H, br d, J = 1.6 Hz, H_a-15) and 4.63 (1H, br d, J = 1.6 Hz, H_b-15), three methyl protons at δ_H 2.13 (3H, s, H-22), 1.25 (3H, d, J = 7.1 Hz, H-13) and 0.76 (3H, s, H-14), two downshielded aromatic protons at δ_H 6.56 (1H, d, J = 2.5 Hz, H-19) and 6.54 (1H, d, J = 2.5 Hz, H-21), indicating the existence of a 1, 2, 3, 5-tetrasubstituted phenyl ring. In accordance with the molecular formula, analysis of the ¹³C NMR and DEPT data (Table S1) displayed 22 carbon resonances that could be attributed to three methyl groups at δ_C 17.0 (C-14), 15.1 (C-22) and 8.2 (C-13), five aliphatic methylene groups at δ_C 49.4 (C-9), 43.0 (C-1), 38.0 (C-3), 26.7 (C-6) and 24.0 (C-2), one exocyclic methylene at δ_C 107.5 (C-15), two aliphatic methines at δ_C 45.8 (C-11) and 43.6 (C-5), two aromatic methines at δ_C 119.3 (C-19) and 108.1 (C-21), five olefinic quaternary carbons at δ_C 153.9 (C-20), 150.3 (C-4), 149.0 (C-17), 132.9 (C-16) and 124.0 (C-18), one aliphatic quaternary carbon at δ_C 35.0 (C-10), one ester carbonyl group at δ_C 177.7 (C-12), and two oxygenated quaternary carbons at δ_C 92.2 (C-7) and 90.4 (C-8). The assignments of all hydrogen and carbon signals were further achieved by its HSQC spectrum. Further analysis of the ¹H-¹H COSY and HMBC spectra (Fig. 2) led to the elucidation of the planar structure of 1. The ¹H-¹H COSY spectrum revealed three spin systems, including H₂-1/H₂-2/H₂-3, H-5/H₂-6 and H-11/H₃-13. The three coupling networks were connected by use of HMBC correlations of H₃-14/C-1, C-5, C-9, C-10; H₂-15/C-3, C-4, C-5; H-5/C-7, C-14; H₂-6/ C-5, C-7, C-8, C-10, C-11; H₃-13/C-7, C-11, C-12; H₂-9/C-1, C-7, C-8. Above evidence made for the construction of the eudesmanolide moiety. The position of the remaining methyl group was located at C-18, as confirmed by HMBC correlation from H₃-22 to C-17, C-18 and C-19. Correlations from H-21 to C-16, C-17 and C-19, and from H-19 to C-17, C-18 and C-21 also resulted the assignment of substituent position of methyl on phenyl ring. In addition, key correlation from H-21 to C-8 unequivocally indicated the C—C connection between C-8 and C-16. To satisfy molecular formula of unsaturation, it was deduced that C-7 was joined with C-17 by sharing oxygen atom, forming a pentacyclic 6/6/5/5/6 skeleton. Hence, the planar structure of 1 was determined as shown in Fig. 1.

  Figure 2

  

  Figure 2.  The ¹H–¹H COSY (thick lines) and key HMBC correlations (pink arrows, from ¹H to ¹³C) of compounds 1−3.

  DownLoad: Full-Size Img PowerPoint

  The relative configuration of 1 was established by ROESY spectrum (Fig. 3). The CH₃-14 was arbitrarily assigned as β-orientation. The key correlation of H₃-14/H-6b indicated the junction of the eudesmane rings was trans-fused and H-5 was α-oriented. Meanwhile, the β-orientation of CH₃-13 was deduced from the cross-peak between H₃-13 and H-6b. Furthermore, the relative configurations of the chiral centers of C-7 and C-8 were deduced because of the rigid structure of 1. Consequently, the established relative configuration of 1 was determined as shown in Fig. 3. The calculated ECD curve for 1 matched well with the experimental one (Fig. 4), suggesting its absolute configuration to be 5S, 7S, 8S, 10R, 11S. Subsequent X-ray diffraction using Cu Kα radiation with the Flack parameter of 0.02(7) (Fig. 5) verified the planner structure as well as the absolute configuration of compound 1.

  Figure 3

  

  Figure 3.  Main NOE correlations (pink dashed double arrows) of 1–3.

  DownLoad: Full-Size Img PowerPoint

  Figure 4

  

  Figure 4.  The overlaid experimental ECD (full lines) and calculated ECD (dash lines) spectra of compounds 1–3.

  DownLoad: Full-Size Img PowerPoint

  Figure 5

  

  Figure 5.  The X-ray ORTEP diagrams of 1 (left) and 3 (right). The thermal ellipsoid is scaled to the 30% probability level.

  DownLoad: Full-Size Img PowerPoint

  Atramacronoid B (2) was isolated as a yellow oil and had the molecular formula C₂₁H₂₄O₄, with 14 mass unit lower than that of 1, which was deduced from HRESIMS at m/z 341.1746 [M + H]⁺ (calcd. for C₂₁H₂₅O₄, 341.1747). The NMR spectroscopic data of 2 closely resembled those of 1 (Table S1), except for the absence of resonances for the methyl unit, which suggested that 2 was 18-demethyl derivative of 1. The deduction was verified by 2D NMR experimental data analysis (Fig. 2), especially by the ¹H-¹H COSY cross-peak of H-18/H-19 and the HMBC correlations from H-21 to C-8, C-17 and C-19, from H-19 to C-17 and C-21, and from H-18 to C-16 and C-20 as well as by the chemical shifts of these hydrogen and carbon resonances. Finally, the intact planner structure of 2 was determined. Compared with 1, similar NOE correlations (Fig. 3) of H₃-14/H-6b and H₃-13/H-6b of 2 indicated that both compounds shared identical relative configurations. Thus, the absolute configuration of 2 was determined as 5S, 7S, 8S, 10R, 11S by comparison of the experimentally measured ECD and theoretically calculated ECD spectra.

  Atramacronoid C (3) was isolated as colorless needle crystals, and possessed the same molecular formula as 1 by the HRESIMS ions at m/z 355.1900 [M + H]⁺ (calcd. for C₂₂H₂₇O₄, 355.1904). Compound 3 was an isomer of 1, which could be preliminary concluded through analyzing detailed 1D NMR (Table S1). In ROESY spectrum of 3, the correlation between H₃-14 and H-6b implied that CH₃-14 and H-5 were β- and α-oriented, respectively. In addition, correlation of H-11 with H-6b (Fig. 3) suggested that 3 was 11-epimer of 1. Taken together, the relative configuration of 3 should be fixed as 5S*, 7S*, 8S*, 10R*, 11R*. To further consolidate the above deduction, the ¹³C NMR chemical shifts of two isomers (5S*, 7S*, 8S*, 10R*, 11R*)-3 (3a) and (5S*, 7S*, 8S*, 10R*, 11S*)-3 (3b) were calculated. Comparison between the experimental and the calculated ¹³C NMR data allowed the determination of relative configuration of 3 to be 5S*, 7S*, 8S*, 10R*, 11R*, with a DP4⁺ probability of approximately 100% (Fig. 6). Based on the agreement between the experimental and calculated ECD spectra (Fig. 4), finally, the absolute configuration of 3 was assigned as 5S, 7S, 8S, 10R, 11R. After repeated recrystallization in different solvent systems, we successfully obtained a crystal of 3 at CH₂Cl_2-CH₃CN (1:1). Single crystal X-ray diffraction (Fig. 5) using Cu Kα radiation with the Flack parameter of 0.04 (8) unequivocally confirmed the absolute configuration of 3.

  Figure 6

  

  Figure 6.  The ¹³C NMR calculation results of two plausible stereoisomers of compound 3. (a) Linear correlation plots of calculated vs. experimental ¹³C NMR chemical shift values of 3a and 3b. (b) Relative errors between the calculated ¹³C NMR chemical shifts the recorded data and DP4⁺ probability analysis.

  DownLoad: Full-Size Img PowerPoint

  To the best of our knowledge, atramacronoids A−C (1−3) are the first example of eudesmanolide sesquiterpene with phenol forming C—C bond via cyclization. The hypothetical biosynthetic pathways for 1−3 are proposed as shown in Scheme 1. Biogenetically, their biogenetic precursor is traced to be atractylolone, which is biosynthesized through mevalonate pathway and abundantly occurs in this plant. Atractylolone undergoes hydration, dehydration and oxidation to produce atractylenolide II. Afterwards, sequential oxidation and dehydration reactions to afford atractylenolide III and atractylenolide I. Atractylenolide I then yield i and ii through addition and elimination reaction, which is followed by the intermolecular addition of the C—H bond and intramolecular addition of the phenol O—H bond [16] involving the formation of C—C bond between C-8 and C-16 and C—O bond between C-7 and C-17, respectively. The cyclization of i with hydroquinone or methylhydroquinone finally give rise to compounds 2 and 1, respectively. In addition, compound 2 also can be converted to 1 by Friedel−Crafts alkylation, while ii is cyclized with methylhydroquinone to produce compound 3.

  Scheme 1

  

  Scheme 1.  The plausible biosynthetic pathways of compounds 1−3.

  DownLoad: Full-Size Img PowerPoint

  In the bioactivity assays, compound 1 exhibited stronger growth inhibitory effects on SGC-7901 cells than other human cancer lines of A549, HCT-8, HepG2, and MCF-7 (Fig. 7a), with the IC₅₀ value of 13 µmol/L. Thus, SGC-7901 cells were used for further study. Morphological analysis is one of the considerations in process of apoptosis [17]. To determine whether 1 could induce cell death by apoptosis, the AO/EB straining assay was performed. As shown in Fig. 7b, with increasing the concentration of 1, more apoptotic cells exhibited apoptotic characteristic such as nuclear shrinkage and chromatin condensation.

  Figure 7

  

  Figure 7.  (a) Effects of 1 on the growth of various cell lines (48 h), with the IC₅₀ values ranging from 13 µmol/L to 25 µmol/L, etoposide as positive. (b) 1 induced SGC-7901 cells apoptosis detected by AO/EB staining. (c, d) Representative histograms depicting cell-cycle distribution as analyzed by flow cytometry in SGC-7901 cells treated with indicated concentrations of 1 for 24 h. Counts of G2/M phase cells increased remarkably in the treated cell in a concentration-dependent manner. (e, f) SGC-7901 cells were incubated with 1 at concentrations of 0, 15, and 30 µmol/L for 24 h. Apoptosis was analyzed by annexin V-FITC/TO-PRO-3 staining. (g) NE synthesis concentration in conditioned medium treated with 1. (h) NE secretion concentration in lysate treated with 1. Values are presented as mean ± SD for three individual experiments. ^*P < 0.05, ^***P < 0.001, ^****P < 0.0001 vs. control.

  DownLoad: Full-Size Img PowerPoint

  The propidium iodide (PI) and annexin V-FITC/TO-PRO-3 double staining were applied further to examine cell cycle distribution and cell apoptosis induced by 1, respectively. Cell cycle arrest is considered a critical control point for the management of cancer cell growth [18]. As shown in Figs. 7c and d, 1 caused G2/M cell cycle arrest in SGC-7901 cells. In the apoptosis assay (Figs. 7e and f), the increase of early and late apoptosis was significantly observed in SGC-7901 cells treated with 1 when compared to the control, with the percentage running up to 59.6%, and the rate of apoptosis increased in a dose-dependent manner, indicating that compound 1 could induce apoptosis in SGC-7901 cells. Amazingly, we found the synthesis and secretion of NE increased after treating cell with 1 (Figs. 7g and h). This suggested that 1 could promote inflammation. From another perspective, we preliminarily inferred 1 induced apoptosis by promoting the synthesis of NE. While in contrast to 1, compound 2 showed no cytotoxicity against SGC-7901 cells at 50 µmol/L, and 3 showed moderate cytotoxicity against SGC-7901 cells with the IC₅₀ value of 27 µmol/L. However, surprisingly, compound 2 inhibited NE synthesis (Fig. S6 in Supporting information), which further backup our aforementioned result for 1. Compound 3 could also inhibit the synthesis of NE, but the effect was weaker than 2 (Fig. S6). Basing on this, we speculated that 3 showing moderate cytotoxic activity may ascribe to another mechanism, which needed to be further researched.

  In summary, the first discovery of atramacronoids A−C (1−3), featuring unprecedented eudesmanolide sesquiterpene-phenol skeleton furnished by forming an unexpected C-8−C-16 linkage, represents a milestone work after decades of the research on A. macrocephala. Predictably, the discovery of fascinating compounds 1−3 enriches the structural diversity of sesquiterpene hybrids and would attract much interest of synthetic chemists for total synthetic and biosynthetic purposes. In addition, mechanistic study revealed compound 1 showed cytotoxicity against SGC-7901 cells by inducing cell apoptosis, which might through promoting the synthesis of NE. These results could provide a new insight for further pharmacological investigation.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This work was supported by the National Natural Science Foundation of China (No. 82073992) and the CAMS Innovation Fund for Medical Sciences (CIFMS, No. 2021-I2M-1-071).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.107743.

  
  
• 1. [1]

     J.W. Wu, C.P. Tang, Y.Y. Cai, et al., Chin. Chem. Lett. 28 (2017) 927–930. doi: 10.1016/j.cclet.2016.11.011

  2. [2]

     J.X. Zhu, J.J. Qin, H.Z. Jin, et al., Fitoterapia 84 (2013) 40–46. doi: 10.1016/j.fitote.2012.09.026

  3. [3]

     T.Z. Li, X.T. Yang, J.P. Wang, et al., Org. Lett. 23 (2021) 8380–8384. doi: 10.1021/acs.orglett.1c03120

  4. [4]

     Y.X. Bai, J. Tan, F.L. Yan, et al., Chin. Chem. Lett. 24 (2013) 55–56. doi: 10.1016/j.cclet.2012.12.015

  5. [5]

     T. Warashina, T. Miyase, Chem. Pharm. Bull. 56 (2008) 1445–1451. doi: 10.1248/cpb.56.1445

  6. [6]

     Y.S. Cai, Z. Wu, J.R. Wang, et al., Org. Lett. 21 (2019) 9478–9482. doi: 10.1021/acs.orglett.9b03676

  7. [7]

     S.H. Tian, X.Y. Chai, K. Zan, et al., Chin. Chem. Lett. 24 (2013) 797–800. doi: 10.1016/j.cclet.2013.05.020

  8. [8]

     W.Y. Zhao, J.J. Yan, T.T. Liu, et al., Eur. J. Med. Chem. 203 (2020) 112622–112644. doi: 10.1016/j.ejmech.2020.112622

  9. [9]

     Y.S. Cai, Z. Wu, X.Q. Zheng, et al., Org. Chem. Front. 7 (2020) 303–309. doi: 10.1039/c9qo01248f

  10. [10]

      C. Cui, K. Chakraborty, X.A. Tang, et al., Cell 25 (2021) 3163–3177.

  11. [11]

      Editor Committee of Flora of China, Flora of China, Science Press, Beijing, 2015, pp. 40–41.

  12. [12]

      C.M. Yao, X.W. Yang, J. Ethnopharmacol 151 (2014) 791–799. doi: 10.1016/j.jep.2013.10.005

  13. [13]

      Editorial Committee of Chinese Pharmacopoeia, Chinese Pharmacopoeia, China Medical Science and Technology Press, 2015, pp. 746–1134.

  14. [14]

      W.J. Zhang, Z.Y. Zhao, L.K. Chang, et al., J. Ethnopharmacol. 266 (2021) 113415–113447. doi: 10.1016/j.jep.2020.113415

  15. [15]

      J.G. Si, H.X. Zhang, M. Yu, et al., J. Funct. Foods 83 (2021) 104541–104550. doi: 10.1016/j.jff.2021.104541

  16. [16]

      R.V. Nguyen, X. Yao, C.J. Li, Org. Lett. 8 (2006) 2397–2399. doi: 10.1021/ol0607692

  17. [17]

      J.A. Hickman, Cancer Metastasis Rev. 11 (1992) 121–139. doi: 10.1007/BF00048059

  18. [18]

      X. Yuan, Z.Y. Li, Z.M. Feng, et al., Chin. Chem. Lett. 32 (2021) 405–4062. doi: 10.1016/j.cclet.2020.02.056

• 

  Figure 1  Structures of compounds 1–3.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  The ¹H–¹H COSY (thick lines) and key HMBC correlations (pink arrows, from ¹H to ¹³C) of compounds 1−3.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Main NOE correlations (pink dashed double arrows) of 1–3.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  The overlaid experimental ECD (full lines) and calculated ECD (dash lines) spectra of compounds 1–3.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  The X-ray ORTEP diagrams of 1 (left) and 3 (right). The thermal ellipsoid is scaled to the 30% probability level.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  The ¹³C NMR calculation results of two plausible stereoisomers of compound 3. (a) Linear correlation plots of calculated vs. experimental ¹³C NMR chemical shift values of 3a and 3b. (b) Relative errors between the calculated ¹³C NMR chemical shifts the recorded data and DP4⁺ probability analysis.

  下载: 全尺寸图片 幻灯片
  

  Scheme 1  The plausible biosynthetic pathways of compounds 1−3.

  下载: 全尺寸图片 幻灯片
  

  Figure 7  (a) Effects of 1 on the growth of various cell lines (48 h), with the IC₅₀ values ranging from 13 µmol/L to 25 µmol/L, etoposide as positive. (b) 1 induced SGC-7901 cells apoptosis detected by AO/EB staining. (c, d) Representative histograms depicting cell-cycle distribution as analyzed by flow cytometry in SGC-7901 cells treated with indicated concentrations of 1 for 24 h. Counts of G2/M phase cells increased remarkably in the treated cell in a concentration-dependent manner. (e, f) SGC-7901 cells were incubated with 1 at concentrations of 0, 15, and 30 µmol/L for 24 h. Apoptosis was analyzed by annexin V-FITC/TO-PRO-3 staining. (g) NE synthesis concentration in conditioned medium treated with 1. (h) NE secretion concentration in lysate treated with 1. Values are presented as mean ± SD for three individual experiments. ^*P < 0.05, ^***P < 0.001, ^****P < 0.0001 vs. control.

  下载: 全尺寸图片 幻灯片
•