English

• Polycyclic polyprenylated acylphloroglucinols (PPAPs) are a special secondary metabolites from the genus Hypericum. Acylphloroglucinol core decorated with prenyl and isoprenyl groups can further be oxidized and cyclized to build the complexity and diversity of PPAPs [1, 2]. So far > 1000 members of the PPAPs have been reported, of which the majority is the typical [3, 3, 1]nonane-2, 4, 9-trione core PPAPs [3]. Biosynthetically, bicyclic PPAPs can be further cyclized to caged PPAPs or be oxidized to open the ring to generate seco-PPAPs [2]. The seco-PPAPs were considered to be formed by the cleavage of chemical bonds at different positions, including 1, 2-seco, 2, 3-seco, 3, 4-seco, 1, 9-seco, 1, 2;4, 5-diseco [2, 4-6]. Around 40 intriguing architectures of seco-PPAPs were characterized [7].

  Recently, diverse carbon skeletons of PPAPs were continuously reported from H. perforatum (St John's wort), an antidepressive traditional chinese medicine, and other hypericum plants, such as hyperfol A, hyperforen A, hymoins A–C and monosescinol A [8-13]. As part of our continuing search for novel bioactive substances from Hypericum plants [14-18], chemical investigation on the EtOH extract of the aerial parts of H. perforatum led to the discovery of hyperforatone A (1), as well as two known precursors, 3-hydroxyhyperforin-3, 9-hemiketal (2) [19] and hyphenrone A (3) [20]. Compound 1 was a 1,8-seco rearranged PPAP bearing a rare bicyclo[5.4.0]undecane carbon skeleton (Fig. 1). In preparing process for this paper, we found that hyperforatone A (1) was the same structure as hyperlanin A [21], a PPAP originally isolated from a distinct plant, Hypericum lancasteri. Noteworthily, we performed an accurate identification of the relative and absolute configurations using residual dipolar couplings (RDCs), nuclear magnetic resonance spectroscopy (NMR) and electronic circular dichroism (ECD) calculations due to the presence of multiple quaternary stereocenters in compound 1. Moreover, density functional theory (DFT) calculations were used to reveal a reasonable formation mechanism for carbocation process of 1 compared to the radical reaction. In addition, neuroinflammation is one of the crucial factors in the development of depression. Compound 1 displayed a significant anti-neuroinflammatory effect in lipopolysaccharide (LPS)-induced BV-2 microglial cells and zebrafish model, thereby providing new additional evidence for the antidepressant effect of H. perforatum. Herein, we describe the structural elucidation, hypothetical biogenetic pathway, and biological activity of the isolation (1).

  Figure 1

  

  Figure 1.  Chemical structure of 1.

  DownLoad: Full-Size Img PowerPoint

  Hyperforatone A (1) was isolated as colorless crystals. Its molecular formula was assigned as C₃₅H₅₂O₅ according to pseudomolecular ion peak at m/z 553.3884 [M + H]⁺ (calcd. for C₃₅H₅₃O₅, 553.3887) in positive high resolution electrospray ionization mass spectroscopy (HR-ESI-MS) spectrum, corresponding to 10 degrees of unsaturation. Three olefinic protons [δ_H 4.96 (1H, overlap), 4.93 (1H, overlap) and 4.77 (1H, t, J = 7.8 Hz)], two characteristic doublet methyls of isopropyl [δ_H 1.16 (d, J = 6.6 Hz), 1.13 (d, J = 6.6 Hz)], and nine singlet methyls (δ_H 1.68–1.04) were presented in the ¹H NMR spectrum. In the ¹³C NMR and distortionless enhancement by polarization transfer (DEPT) 135 spectra, 35 carbon signals were assigned, including 11 methyls, 6 methylenes, 6 methines, and 12 quaternary carbons. Among them, a ketone (δ_C 205.8), an ester carbonyl (δ_C 174.3), three isoprenyl groups (δ_C 138.0, 136.2, 133.3, 123.3, 117.8, 115.9, 31.2, 31.2, 29.3, 26.4, 26.2, 26.0, 18.1, 18.1, 18.0), two olefinic carbons (δ_C 148.5, 110.0), three oxygenated carbons (δ_C 93.2, 90.0, 81.3), and an isopropyl group (δ_C 31.8, 20.5, 20.4) were distinguished (Table 1). Since 1 had 10 degrees of unsaturation, while a ketone, an ester carbonyl, and four olefinic bonds simply accounted for 6 degrees of unsaturation, 1 should bear another four rings. The above analysis showed that 1 was a tetracyclic PPAP-derivative.

  Table 1

  

  Table 1.  ¹H (500 MHz) and ¹³C NMR (125 MHz) data of 1 in CDCl₃.

  DownLoad: CSV

  The completed assignment for protons and carbons of 1 was addressed by 2D NMR experiments (Fig. 2). From ¹H–¹H correlation spectroscopy (COSY) experiment, five fragments, a (H₂–27/H-28), b (H₂–22/H-23), c (H₂–17/H-18), d (H₃–12/H-11/H₃–13) and e (H₂–33/H₂–34/H-35), were observed. The heteronuclear multiple bond correlation (HMBC) correlations from H₂–17 to C-3/C-4, from H₂–22 to C-4/C-5/C-6, from H₂–27 to C-6/C-7, and from H₂–6 to C-4/C-5/C-7/C-8 undoubtedly confirmed that the connectivity of C-3/C-4/C-5/C-6/C-7/C-8. The 2-hydroxyl-3-methyl-1, 1-disubstuted-1-butenyl group at C-1 position was confirmed by the HMBC correlations from the proton of 10–OH to C-1/C-10/C-11 and H-11 to C-1/C-10. The two methyls (δ_C 30.4, 22.2; δ_H 1.22, 1.32) were located at oxygenated C-36 by the HMBC correlations of H₃–38 to C-36/C-37 and H₃–37 to C-35. The HMBC correlations from H₃–32 to C-33, from H₂–33 to C-2, and from H-35 to C-1/C-2/C-10/C-33/C-34 conclusively indicated that the connections of C-1/C-2/C-8/C-33/C-34/C-35. The fused two rings A and C through the bridge of C-2/C-8 bond were confirmed by critical the HMBC correlations from H_b-17 to C-2 and H₃–32 to C-2/C-7/C-8. An internal ester bridge between C-3 and C-5 (ring B) was deduced by the HMBC correlation from H₂–22 to C-9, along with the characteristic downfield chemical shifts of C-3 at δ_C 93.2 and C-5 at δ_C 53.2. In addition, ring D was formed via an ether bond which connected C-2 at δ_C 90.0 with C-36 at δ_C 81.3.

  Figure 2

  

  Figure 2.  Key 2D NMR correlations for 1.

  DownLoad: Full-Size Img PowerPoint

  The relative configuration of 1 was determined by the analysis of its nuclear overhauser effect spectroscopy (NOESY) spectrum (Fig. 2). The correlations of H₃–32/H_a-27/H_a-33, H_a-33/H_a-34/H-35, H_a-27/H_a-6 and H-35/H₃–37 suggested that H_a-27, H₃–32, H_a-33, H_a-34, H-35 and H₃–37 were in the same orientation. In addition, the Z configuration of Δ¹⁽¹⁰⁾ double bond was determined by the NOE correlation of H-35/H-11. RDCs were a powerful tool for the relative configuration analysis of complicated natural products [22]. To further validate the above assignment for 1, RDCs were successfully applied to determine its relative configuration. Total couplings (¹T_CH), scalar couplings (¹J_CH), and residual dipolar couplings (¹D_CH) were calculated in a 2.9% oligopeptide amphiphile (OPA) aligned medium, and the corresponding Q factors were used to evaluate the possibility of relative configurations. The eight possible diastereomeric configurations were obtained. The (2R*, 3R*, 5S*, 7R*, 8S*, 35S*)−1a had the best fitting with the lowest Q factor of 0.012 (Fig. 3). Moreover, the calculated ¹³C NMR chemical shifts of (2R*, 3R*, 5S*, 7R*, 8S*, 35S*)-truncation of 1 was quite consistent with the experimental chemical shifts, with a good linear correlation coefficient (R², 0.9983), a lower mean absolute error (MAE, 1.7) and a lower corrected mean absolute error (CMAE, 1.7) (Tables S4 and S5 in Supporting information).

  Figure 3

  

  Figure 3.  Corresponding Q factors of the RDCs for the eight possible diastereomeric configurations of 1.

  DownLoad: Full-Size Img PowerPoint

  The absolute configuration was further confirmed by the ECD calculation. The calculated ECD spectrum of (2S, 3S, 5R, 7S, 8R, 35R)−1 was fully matched with the experimental ECD spectrum with positive Cotton effect around 200 and 225 nm and negative Cotton effect around 310 nm (Fig. 4). Finally, the orthorhombic crystals of 1 were obtained from acetone, and the single-crystal X-ray diffraction analysis (Cu Kα) [Flack parameter = −0.06(6)] unequivocally confirmed the structural assignments (Fig. 5). Thus, the absolute configuration of 1 was clarified as 2S, 3S, 5R, 7S, 8R, 35R.

  Figure 4

  

  Figure 4.  Calculated and experimental ECD spectra of 1.

  DownLoad: Full-Size Img PowerPoint

  Figure 5

  

  Figure 5.  Oak ridge thermal ellipsoid plot (ORTEP) drawing of 1.

  DownLoad: Full-Size Img PowerPoint

  To the best of our knowledge, two biosynthetic pathways, cationic and radical cyclization, are involved in the carbon skeleton rearrangement of PPAPs [7, 23-25]. To assess the feasibility of 5/7/6/5 ring system formation in the biosynthesis of 1, the chemical computation was performed.

  Based on the calculated results of the cationic process (Scheme 1 and Fig. 6), 1 might derive from hyperforin (HF), the antidepressant constituent of St John's wort, via a crucial intermediate hemiketal 2. Then, 2 cleaved the C-1/C-9 bond via a retro-aldol mechanism to form a five-membered lactone fragment (B ring) of 3 with free energy of −22.47 kcal/mol. The carbonyl carbon (C-2) occurred to protonation. Then the key intermediate (i) of the seven-membered A ring was produced by the heterolytic cleavage of the C-1/C-8 bond and the formation of the C-2/C-8 bond which the tert-alkyl anion of C-8 attacked to the carbonyl cation of C-2 [26, 27]. The process needed to overcome the activation barrier (ΔG = 32.78 kcal/mol). It was easier to form the C ring through the pair of electrons of the olefinic bond at C-35 attacking to the cation of C-1 to afford the intermediate ii by transition structure (TS_8) with an activation free energy barrier of 1.97 kcal/mol. It tended to form the D ring from intermediate ii to intermediate iii, because of the distance (3.65 Å) between the oxygen at C-2 and the carbon of C-36, a barrier-free and exergonic step (ΔG = −4.20 kcal/mol). Finally, 1 was obtained by the ketone/alcohol conversion, which the process included protonation of the carbonyl O under acidic condition (TS_9a with a free energy of −37.35 kcal/mol) and deprivation of the proton at C-1 position (TS_9b with a free energy of −7.68 kcal/mol).

  Scheme 1

  

  Scheme 1.  Plausible biosynthetic pathway of 1.

  DownLoad: Full-Size Img PowerPoint

  Figure 6

  

  Figure 6.  The calculated energy profile (kcal/mol) with DFT for the carbocation process of compound 1.

  DownLoad: Full-Size Img PowerPoint

  A radical process of 1 was also speculated and calculated with the DFT (Scheme S1 and Fig. S17 in Supporting information). The radical transition state (TS_13) might be produced under irradiation conditions, which triggered the radical reaction chain. As we all know, the radical intermediate might undergo β-fragmentation (1,8-seco) to produce a tert-alkyl radical and an enol anion substituent [28]. Then, the tert-alkyl radical (C-8) might add to the α-carbon radical (C-2) of the enol anion group to produce the seven-membered A ring (intermediate iv) [29, 30]. However, the steps from 3 to intermediate iv proceed through overcoming activation free energy of 78.34 kcal/mol. The formation of the C ring (intermediate v) could be produced by two steps of the radical cyclization via transition structure (TS_14) with barrier of 20.35 kcal/mol. The formation of D ring (intermediate vi) was a highly exergonic step (ΔG = −72.60 kcal/mol). For the ketone/alcohol conversion, the protonation of TS_9a was barrier-free and exergonic, and the proton at C-1 position was deprived through an activation barrier (ΔG = 45.03 kcal/mol). At this level of calculation, it was concluded that the biosynthetic proposal for the formation of 1 through the cationic cyclization reaction was feasible compared to the radical reaction.

  We evaluated the anti-neuroinflammatory activities of compounds 1–3 and HF using LPS-induced BV-2 microglial cells model and S-methylisothiourea (SMT) as a positive control. The results showed that compound 1 and HF exhibited moderate inhibition of nitric oxide (NO) production with half maximal inhibitory concentration (IC₅₀) values of 31.00 and 21.68 µmol/L, respectively. The IC₅₀ value of SMT was 5.93 µmol/L. Compounds 2 and 3 showed no significant NO inhibition with IC₅₀ values greater than 50 µmol/L (Table S12 in Supporting information). None of these tested drugs showed significant cytotoxicity against BV-2 cells at the tested concentrations (Fig. S20 in Supporting information). In addition, we tested the effect of each drug at 40 µmol/L on LPS-induced NO production and the number of apoptotic nerves in the head using zebrafish larvae. The results showed that compound 1, HF and SMT significantly inhibited LPS-induced NO production and markedly reduced the number of neuronal apoptosis in the head of zebrafish larvae. Compounds 2 and 3 did not show significant inhibitory activity at this concentration (Figs. S22 and S23 in Supporting information). This is basically consistent with the results of intracellular activity.

  Inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2) play key roles in the production of pro-inflammatory mediators. Western blot analysis revealed that 1 downregulated the expression of iNOS and COX-2 in LPS-stimulated BV-2 microglial cells (Figs. S24B–D in Supporting information). In addition, we measured NO production and neuronal apoptosis in LPS-induced zebrafish. As shown in Figs. 7A and B, NO production was significantly increased when zebrafish larvae were stimulated with LPS. Conversely, NO level was notably reduced in a dose-dependent manner after treatment with compound 1. Furthermore, 1 significantly reduced the neuronal apoptosis in the head of zebrafish after LPS stimulation (Figs. 7C and D). The results in zebrafish demonstrated that 1 exerted anti-neuroinflammatory effects by inhibiting NO production and neuronal apoptosis.

  Figure 7

  

  Figure 7.  Effects of compound 1 on NO production and neuronal apoptosis in LPS-induced zebrafish. (A, C) NO production and neuronal apoptosis were imaged using fluorescence microscopy. (B, D) Mean fluorescence intensity and the number of apoptotic cells were quantified using Image J. ^####P < 0.0001 vs. control (DMSO) group; P < 0.05, ****P < 0.0001 vs. LPS group (ns, no significance).

  DownLoad: Full-Size Img PowerPoint

  In summary, a novel PPAP, hyperforatone A (1) was isolated from the aerial parts of H. perforatum. Compound 1 was a bicyclo[5.4.0]undecane architecture by 1,8-seco rearranged reactions. The cationic cyclization was favored to construct rare carbon skeleton by DFT calculations. Its structure may attract particular attention of organic synthesis and phytochemists. In addition, 1 showed a moderate anti-neuroinflammatory effect in LPS-induced BV-2 microglial cells and zebrafish.

  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.

  CRediT authorship contribution statement

  Wu-Yang Liu: Writing – original draft, Visualization, Software, Methodology, Formal analysis, Data curation, Conceptualization. Xin-Xiang Lei: Software, Methodology, Data curation. Wen-Ji Wang: Software, Methodology, Investigation, Data curation. Jun-Mian Tian: Writing – review & editing, Supervision, Methodology, Data curation. Yu-Qi Gao: Writing – review & editing, Validation, Supervision. Jin-Ming Gao: Writing – review & editing, Supervision, Project administration, Funding acquisition.

  Acknowledgments

  This work was financially supported by the National Natural Science Foundation of China (Nos. 22077102, 32201249) and the Science and Technology Program of Gansu Province (No. 22YF7GA046). We also greatly acknowledge high performance computing (HPC) of Northwest A & F University for the density functional theory calculations carried out in this work.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110478.

  
  
• 1. [1]

     R. Ciochina, R.B. Grossman. Chem. Rev. 106 (2006) 3963–3986. doi: 10.1021/cr0500582

  2. [2]

     X.W. Yang, R.B. Grossman, G. Xu, Chem. Rev. 118 (2018) 3508–3558. doi: 10.1021/acs.chemrev.7b00551

  3. [3]

     Z.H. Xu, R.B. Grossman, Y.F. Qiu, et al., J. Nat. Prod. 85 (2022) 2845–2855. doi: 10.1021/acs.jnatprod.2c00888

  4. [4]

     W.J. Lu, Y.Q. Zhang, Y.W. Li, et al., Chin. Chem. Lett. 33 (2022) 4121–4125. doi: 10.1016/j.cclet.2021.11.011

  5. [5]

     R.D. Liu, Y.L. Su, J.B. Yang, et al., Phytochemistry 142 (2017) 38–50. doi: 10.1016/j.phytochem.2017.06.011

  6. [6]

     S.S. Xie, C.X. Qi, Y.L. Duan, et al., Org. Chem. Front. 7 (2020) 1349–1357. doi: 10.1039/d0qo00259c

  7. [7]

     Y.L. Hu, K. Hu, L.M. Kong, et al., Org. Lett. 21 (2019) 1007–1010. doi: 10.1021/acs.orglett.8b04022

  8. [8]

     Y. Guo, N. Zhang, C. Chen, et al., J. Nat. Prod. 80 (2017) 1493–1504. doi: 10.1021/acs.jnatprod.6b01178

  9. [9]

     H. Bridi, G. de. C. Meirelles, G.L. von Poser, Phytochemistry 155 (2018) 203–232. doi: 10.1016/j.phytochem.2018.08.002

  10. [10]

      H.Y. Lou, F.W. Ma, P. Yi, et al., J. Arab. J. Chem. 15 (2022) 104057. doi: 10.1016/j.arabjc.2022.104057

  11. [11]

      H.Y. Lou, Y.N. Li, P. Yi, et al., Org. Lett. 22 (2020) 6903–6906. doi: 10.1021/acs.orglett.0c02434

  12. [12]

      C.M. Yuan, Y.R. Zeng, L. Huang, et al., Chin. Chem. Lett. (2024), https://doi.org/10.1016/j.cclet.2024.109859. doi: 10.1016/j.cclet.2024.109859

  13. [13]

      Z.Y. Shi, J. Yin, Y. Xiao, et al., Chin. Chem. Lett. 35 (2024) 109458. doi: 10.1016/j.cclet.2023.109458

  14. [14]

      X.T. Yan, Z. An, D. Tang, et al. RSC Adv. 8 (2018) 26646–26655. doi: 10.1039/c8ra05322g

  15. [15]

      X.T. Yan, Z. An, Y. Huangfu, et al., Phytochemistry 159 (2019) 65–74. doi: 10.3390/membranes9050065

  16. [16]

      J.Y. Xie, P.F. Li, X.T. Yan, et al. Commun. Chem. 7 (2024) 1. doi: 10.1038/s42004-023-01091-1

  17. [17]

      X.T. Yan, J.X. Chen; Z.X. Wang, et al., J. Nat. Prod. 86 (2023) 119–130. doi: 10.1021/acs.jnatprod.2c00810

  18. [18]

      J.Y. Xie, Z.X. Wang, W.Y. Liu, et al., J. Nat. Prod. 86 (2023) 1910–1918. doi: 10.1021/acs.jnatprod.3c00226

  19. [19]

      L. Verotta, G. Appendino, J. Jakupovic, et al., J. Nat. Prod. 63 (2000) 412–415. doi: 10.1021/np9903752

  20. [20]

      X.W. Yang, Y. Ding, J.J. Zhang, et al., Org. Lett. 16 (2014) 2434–2437. doi: 10.1021/ol500808p

  21. [21]

      J.Q. You, Y.Q. Jin, M.Y. Li, et al., Chin. J. Chem. 42 (2024) 2180-2186. doi: 10.1002/cjoc.202400248

  22. [22]

      W.B. Han, Y.J. Zhai, R. Zhang, et al., Angew. Chem. Int. Edit. 62 (2023) e202300773. doi: 10.1002/anie.202300773

  23. [23]

      W.J. Xu, P.F. Tang, W.J. Lu, et al., Org. Lett. 21 (2019) 8558–8562. doi: 10.1021/acs.orglett.9b03098

  24. [24]

      N.N. Jiang, Y.S. Ye, X. Liu, et al., Org. Lett. 25 (2024) 8965–8969.

  25. [25]

      X.Y. Jia, Y.M. Wu, C. Lei, et al., Chin. Chem. Lett. 31 (2020) 1263–1266. doi: 10.1016/j.cclet.2019.10.014

  26. [26]

      J.F. Zong, Z. Hu, Y.Y. Shao, et al., Org. Lett. 22 (2022) 2797–2800.

  27. [27]

      Y.L. Duan, Y. Guo, Y.F. Deng, et al., J. Org. Chem. 87 (2022) 6824–6831. doi: 10.1021/acs.joc.2c00503

  28. [28]

      C. Janine, Photoinduced electron tranfer in radical reactions, Vol. 1, in: R. Philippe, P.S. Mukund (Eds.), Radicals in Organic Synthesis, Wiley-Vch Press: Strauss Offsetdruck Gmbh Press, Mörlenbach, 2001, pp. 245.

  29. [29]

      L. Canonica, B. Danieli, G. Lesma, et al., Helv. Chim. Acta 70 (1987) 701–716. doi: 10.1002/hlca.19870700320

  30. [30]

      K. Cen, J.J. Bao, X.D. Wang, et al., J. Am. Chem. Soc. 146 (2024) 6481–6486. doi: 10.1021/jacs.4c01016

• 

  Figure 1  Chemical structure of 1.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Key 2D NMR correlations for 1.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Corresponding Q factors of the RDCs for the eight possible diastereomeric configurations of 1.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Calculated and experimental ECD spectra of 1.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Oak ridge thermal ellipsoid plot (ORTEP) drawing of 1.

  下载: 全尺寸图片 幻灯片
  

  Scheme 1  Plausible biosynthetic pathway of 1.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  The calculated energy profile (kcal/mol) with DFT for the carbocation process of compound 1.

  下载: 全尺寸图片 幻灯片
  

  Figure 7  Effects of compound 1 on NO production and neuronal apoptosis in LPS-induced zebrafish. (A, C) NO production and neuronal apoptosis were imaged using fluorescence microscopy. (B, D) Mean fluorescence intensity and the number of apoptotic cells were quantified using Image J. ^####P < 0.0001 vs. control (DMSO) group; P < 0.05, ****P < 0.0001 vs. LPS group (ns, no significance).

  下载: 全尺寸图片 幻灯片

  Table 1.  ¹H (500 MHz) and ¹³C NMR (125 MHz) data of 1 in CDCl₃.

  下载: 导出CSV
•