English

• The [2+2]-type cyclobutane-containing natural products (CCNPs) are medicinally highly valuable and exist in various structural classes of specialized metabolites, such as alkaloids, flavonoids, stilbenes, terpenoids, and phenylpropanoids [1]. The CCNPs consist of over 300 members, all featuring a unique and highly strained cyclobutane core. These molecules are biogenetically derived from intermolecular or intramolecular [2+2] cycloaddition reactions between double bonds (generally in conjugation with chromophores) in various biogenetic precursors [1–7]. The [2+2] cycloaddition reactions enable the formation of two C–C bonds along with four new stereogenic centers in a single step with a high degree of regioselectivity and stereoselectivity, resulting in the framework variability and stereochemical diversity of the newly generated scaffolds. Compared to their biosynthetic precursors, CCNPs have a significantly expanded chemical space due to their more complex structures and multiple chiral centers [8–10]. Aside from their intriguing structural features, many CCNPs were reported to display stronger biological and pharmacological activities than their precursors [1,4].

  Due to the existence of various substituents and four contiguous stereocenters in a confined space, the structure elucidation and configuration assignment of [2+2]-type cyclobutane rings present a challenging and fallible task. For [2+2]-type cyclobutane-containing dimers, the connection modes (i.e., head-to-head vs. head-to-tail) of the two halves are commonly determined by analyzing the MS² fragment ions when suitable crystals for single-crystal X-ray diffraction (SCXRD) analysis are unavailable. Meanwhile, the relative and absolute configuration assignments of the cyclobutane rings are usually deduced through comprehensive analysis, including 2D nuclear Overhauser effect (NOE) correlation, ³J_H,H coupling constant, dihedral angle, electronic circular dichroism (ECD) curve, and density functional theory (DFT)-NMR calculation [4,11–14]. Despite the wide use of these methods, there have been misinterpretations in the stereostructures or connection modes for [2+2]-type cyclobutane-containing dimers, such as piperarborenine D [15,16], dipiperamide A [17,18], biyouyanagins A and B [19–22], and andamanicin [23,24], and they were eventually corrected through synthesis methods or SCXRD analysis. Furthermore, the stereochemistry of some reported cyclobutane-containing dimers (e.g., pipernigramide E, mooniines A and B) remains unresolved [25,26]. Overall, the structure elucidation of [2+2]-type cyclobutane derivatives is a complicated and time-consuming process that requires extreme caution. Therefore, it is highly desired for the development of rapid and accurate approaches to determine the connection modes and stereochemistry of [2+2]-type cyclobutane rings.

  Our group has long been focused on isolation, characterization, biomimetic synthesis and biological evaluation of natural products from Chinese medicinal plants [27–30]. In our continuing investigation, with the assistance of high performance liquid chromatography-ultraviolet detector-mass spectrometry (HPLC-UV-MS), three new [2+2] cyclobutane-containing iridoid glycoside dimers, plumerubradins A–C (1–3), together with their biogenetic precursor plumieride p-E-coumarate (4), were isolated from the flowers of medicinal plant Plumeria rubra Linn. cv. Acutifolia (Apocynaceae). Structurally, compounds 1–3 possess two rare plumeria-type iridoid glycoside units with four contiguous stereogenic centers at C-1, C-5, C-8, and C-9, as well as an isolated chiral center at C-13. The two iridoid glycoside units in compounds 1–3 were linked through different connection modes (head-to-head or head-to-tail) to generate phenylpropanoid-derived cyclobutane motifs with diversified stereochemistry. Despite the high complexity, the complete structures of compounds 1–3 (Fig. 1) including their absolute configurations were successfully established through a combination of extensive spectroscopic analysis, biomimetic semisynthesis, chemical degradation, as well as SCXRD analysis of their degradation products. In particular, the structure of compound 2 was further verified by microcrystal electron diffraction (MicroED) study at atomic resolution. Using quantum mechanics-based ¹H NMR iterative full spin analysis (QM-HiFSA), we gained a deeper understanding of signal patterns and coupling constants of the cyclobutane protons in synthetic phenylpropanoid-derived [2+2] dimers 7–10. This analysis helped establish the correlations between the proton signal patterns and the structural information of the cyclobutane ring, including [2+2] connection modes and relative configurations. Lastly, the anti-ulcerative colitis activities of 1–3 were evaluated using in vitro unicellular (HT-29 and Caco-2 cell lines) and multicellular (Caco-2/THP-1) models.

  Figure 1

  

  Figure 1.  Chemical structures of plumerubradins A–C (1–3).

  DownLoad: Full-Size Img PowerPoint

  In our initial attempts to determine the chemical structures of compounds 1–3, we used comprehensive spectroscopic analyses as we described in the previous investigation for the structure elucidation of [2+2] cyclobutane-containing stilbene glycoside dimers isolated from Polygonum multiflorum [30]. However, due to the particular structural features of 1–3, we were only able to identify their partial structures (see the Supporting information for details). Further identification was needed for the relative configuration of the isolated stereocenters (C-13 and C-13′), the [2+2] connection mode and relative configuration of the cyclobutane rings in 2 and 3, and the absolute configurations of all the compounds.

  Biogenetically, dimers 1–3 could be formed from two molecules of the putative biosynthetic precursor 4 via [2+2] cycloaddition. In order to investigate the structure further, we aimed to prepare sufficient samples of 1–3 using a biomimetic semisynthesis approach. We had over 3.5 g of 4, which was isolated from P. rubra in the same study. Inspired by the biosynthetic hypothesis, we initially conducted the [2+2] photocycloaddition of 4 in CH₃OH or CH₃CN solution with various light sources (Table S5 in Supporting information). However, we did not obtain the desired [2+2] dimers, and in most cases, only the starting material 4 and its double bond isomerized derivative 5 were detected in the reaction mixtures. Interestingly, after testing different solvents, we discovered that when 4 was slightly or poorly dissolved in solvents such as CH₂Cl₂, THF, toluene, H₂O, and cyclohexane, the [2+2] photocycloaddition proceeded smoothly to produce the expected products 1–3, along with 5. We further optimized the reaction conditions and identified the following optimum protocol (entry 10, Table S5): suspending 4 in cyclohexane and irradiating with visible light (75 W, LED, λ = 390 nm) at room temperature for 48 h (Scheme 1); then, compounds 1–3 can be isolated with yields ca. 5%, 8%, and 3%, respectively. It is worth noting that these [2+2] photocycloadditions could also yield the products in moderate yield when water was used as the solvent (entry 9, Table S5). This suggests that those iridoid glycoside [2+2] dimers could be generated from the abundant precursor 4 in plants. The spectroscopic data of synthetic 1–3 were in good agreement with those of the isolated ones (Tables S6–S8 in Supporting information). It should be noted that 4 was first reported in 1983, but the stereochemistry of C-13 in 4 was still uncertain [31,32]. In this study, we determined the absolute configuration of 4 (1S,5S,8R,9S,13S) for the first time using chemical degradation and SCXRD analysis (Scheme 1) [Flack parameter = 0.08(10)] (CCDC: 2333495). Considering the fact that the iridoid moieties of 1–3 were inherited from their biosynthetic precursor 4, therefore the absolute configurations of the two iridoid moieties in 1–3 were determined as 1S,5S,8R,9S,13S and 1′S,5′S,8′R,9′S,13′S, respectively.

  Scheme 1

  

  Scheme 1.  Biomimetic semisynthesis of 1–3 and chemical degradation of 4. Reagents and conditions: (a) 4 (0.97 mmol), cyclohexane (20 mL), N₂, room temperature, 75 W 390 nm LED, 48 h; (b) 4 (0.32 mmol), CH₃OH (25 mL), K₂CO₃ (2.60 mmol), room temperature, 5 h. LED = light-emitting diode.

  DownLoad: Full-Size Img PowerPoint

  To determine the [2+2] connection modes and relative configurations of cyclobutane motifs in 1–3, a chemical degradation strategy was employed. Unfortunately, only a limited amount of simplified phenylpropanoid-derived cyclobutane methyl ester dimers of 1–3 were obtained, which hindered further structure elucidation. To address this, a [2+2] photocycloaddition reaction was carried out using 6 as the substrate in cyclohexane to synthesize these simplified dimers and their analogues (Fig. 2). Theoretically, eleven possible cyclobutane-containing dimers could be generated from aryl-conjugated alkene monomers via [2+2] photodimerization reaction (structures a–k, Fig. S9 in Supporting information) [4,33]. However, liquid chromatography-mass spectrometry (LC-MS) analysis of the reaction mixture revealed the presence of six homodimers, and only four of them (7–10) were isolated in sufficient quantities by semi-preparative HPLC with yields ca. 9.7%, 16.8%, 5.1%, and 3.7%, respectively. With these samples, various crystal cultivation conditions were tested to obtain single crystals of 7–10 for SCXRD analysis. However, despite multiple attempts, only suitable crystals of 7 were obtained. Moving forward, we prepared the 4-bromobenzoate derivatives of 8–10 (8a–10a) through chemical modification (Fig. 2). Fortunately, high-quality crystals of 8a–10a were successfully obtained and SCXRD analysis allowed for the establishment of the structures of 7 and 8a–10a (Fig. 2). Consequently, the connection modes and relative configurations of dimers 7–10 were unambiguously determined as: (head-to-tail, syn-anti-syn-anti), (head-to-head, syn-anti-syn-anti), (head-to-head, all anti), and (head-to-tai, all anti) (CCDC for 7 and 8a–10a: 2333496, 2333494, 2333498, and 2333574), respectively. It is worth noting that the four dimeric forms represent the four predominant ones among the eleven possible structures based on the reported literatures [4,33] and the theoretical calculation results [34]. Subsequently, HPLC retention time comparison as well as co-injection analysis confirmed that 7–9 were the corresponding degradative cyclobutane methyl ester dimers of 1–3, respectively (Fig. S10 in Supporting information). Therefore, the [2+2] cycloaddition connection modes and relative configurations of cyclobutane ring in 1–3 were determined to be consistent with 7–9, respectively. Finally, the structures with absolute configurations of 1–3 were completely established.

  Figure 2

  

  Figure 2.  Syntheses of 7–10 and 4-bromobenzoate derivatives 8a, 9a, and 10a. THF = tetrahydrofuran.

  DownLoad: Full-Size Img PowerPoint

  Considering the potential risks of stereoscopic configuration inversion during the chemical degradation processes [11], we attempted to prepare single crystals of 1–3 for SCXRD analysis to determine their intact molecular structures. However, despite multiple attempts using various solvent systems and crystallization conditions, we did not obtain well-ordered single crystals with proper size for SCXRD analysis. Microcrystal electron diffraction (MicroED) is an emerging structure elucidation method that requires only a microcrystalline sample [35–38]. Fortunately, compound 2 was obtained as a yellow powder, which was confirmed as microcrystals through powder X-ray diffraction analysis. We then used MicroED to determine the 3D structure of compound 2. Initially, the resolution of the acquired data for compound 2 diffracted to >2 Å on a carbon transmission electron microscope (TEM) grid, which was too poor for ab initio structure determination. Through further optimization of crystallization conditions, we found that slow evaporation of binary solvent system (CH₃OH-H₂O, 9:1, v/v, 2 mL) at 4 ℃ ultimately led to the generation of qualified microcrystals of 2 that diffracted to 1.0 Å (complete crystallographic parameters are presented in Table S22 in Supporting information). The data of five crystals were merged to provide a 1.0 Å ab initio solution (R₁ = 12.25%) (CCDC: 2333627). The crystal structure of compound 2 was refined anisotropically, and the asymmetric unit was plotted in Fig. 3. Finally, the absolute configuration of the cyclobutane ring in 2 was determined by using the known chiral centers of glucose residues and iridoid units as chiral references. Thus, the structure with absolute configuration of compound 2 was further confirmed by MicroED analysis, verifying the validity and reliability of the combination of spectroscopic analysis, semisynthesis, chemical degradation, and SCXRD methods for our structure elucidation. To the best of our knowledge, compound 2 is the first small-molecule natural product with a molecular weight of more than 1200 identified by MicroED [39–43]. Interestingly, an unexpected intramolecular H-bonding between 4a-OH and C-15 carbonyl was observed in the microcrystalline structure of compound 2 (Fig. S11 in Supporting information), resulting in the spatial proximity of the benzene ring and the plumeria-type iridoid unit.

  Figure 3

  

  Figure 3.  Structure determination of 2 by MicroED method. (A) Corresponding electron density map superimposed on the refined structure (1.0 Å resolution). (B) Oak Ridge thermal ellipsoid plot (ORTEP) drawing of 2.

  DownLoad: Full-Size Img PowerPoint

  As discussed above, we have successfully determined the structures with absolute configurations of the natural [2+2]-type cyclobutane-containing dimers 1–3 and synthetic ones 7–10. Upon further analysis of their ¹H NMR spectra, it is interesting to point out the strong correlation between the signal patterns of the cyclobutane protons with the [2+2] connection modes and relative configurations of the cyclobutane rings (Fig. 4).

  Figure 4

  

  Figure 4.  Enlarged stacked ¹H NMR spectra of 1–3 and 7–9: (A) Comparison of 1 and 7 in CD₃OD; (B) Comparison of 2 and 8 in CD₃OD; (C) Comparison of 3 and 9 in DMSO-d₆. (D) Enlarged ¹H NMR spectrum of 10 in CD₃OD. DMSO = dimethylsulfoxide.

  DownLoad: Full-Size Img PowerPoint

  Owing to their magnetically equivalent peculiarity, the cyclobutane protons (H-7a/H-7b and H-8a/H-8b) in compound 10 exhibited a first-order spin system (A₂X₂) and appeared as two sets of triplets (t) in the ¹H NMR spectrum (Fig. 4D). However, in the ¹H NMR spectra of 7–9, the cyclobutane protons displayed rather complex signal patterns, which frequently were labeled as “multiplets”, suggesting the presence of higher-order effects as described in previous studies [12,13,44–46]. To investigate the precise spin-spin coupling patterns of the cyclobutane protons in compounds 7–10, a quantum mechanics-based ¹H iterative full spin analysis (QM-HiFSA) [47–49] was performed using Cosmic Truth (CT) software. For compound 7, the cyclobutane protons constituted an AA'XX' spin system [50]. The QM-HiFSA profile of 7 revealed that H-7a/H-7b were successively coupled to H-8a (³J = 10.52 Hz), H-8b (³J = 7.21 Hz), and the aromatic protons [H-2a(6a) and H-2b(6b) (⁴J = -0.60 Hz); H-3a(5a) and H-3b(5b) (⁵J = 0.24 Hz)] (Table 1). These four spin-spin couplings theoretically would result in a “ddtt” multiplicity for the signal of H-7a/H-7b. However, due to the overlapping peaks, a pseudo-doublet of doublets (dd) was observed instead for H-7a/H-7b signal in the ¹H NMR spectrum of compound 7 (Fig. 4a). Accordingly, H-8a/H-8b signal appeared with reduced multiplicity as a doublet of doublets, reflecting the ³J couplings to H-7a and H-7b. Similar signal patterns were observed in the ¹H NMR spectrum of compound 1 (Fig. 4a). The “ddtt” multiplicity for H-7a/H-7b in compounds 8 (an AA'XX' system, Fig. 4b) [50] and 9 (an AA'BB' system, Fig. 4c) [50] could be attributed to the vicinal couplings (³J) and the long-range couplings (≥⁴J) to aromatic protons (Table 1). For compounds 8 and 9, a doublet of doublets was also observed for H-8a/H-8b. Similar signal profiles were observed in the ¹H NMR spectra of 2 (Fig. 4b) and 3 (Fig. 4c). Additionally, the QM-HiFSA approach also provided vicinal coupling values of cyclobutane protons in compound 10, which were consistent with the conventional determination of J values (³J = 9.7 Hz) through manual measurement (Table 1). Thus, the QM-HiFSA analysis of cyclobutane protons in compounds 7–10 enabled the precise determination of their ¹H NMR splitting patterns and spin-spin coupling patterns, leading to an in-depth understanding of the observed “multiplets”. We summarized the correlations between ¹H NMR signal patterns and the structural information of cyclobutane rings in compounds 7–10 in Table 1.

  Table 1

  

  Table 1.  Correlations between the ¹H NMR signal patterns and the structural information of cyclobutane ring in [2+2]-type cyclobutane derivatives.

  DownLoad: CSV

  As shown in Table 1, the different connection modes and relative configurations of the cyclobutane rings in compounds 7–9 resulted in the distinctive multiplicities and coupling constants of the cyclobutane resonances in their ¹H NMR spectra. It is particularly noteworthy that the signal patterns were barely influenced by the NMR solvents (acetonitrile-d₃, chloroform-d, CD₃OD, or DMSO-d₆) and the magnetic field strength of the spectrometers (400, 500, or 600 MHz) (Fig. S12 in Supporting information). The above results strongly suggest that we may use the signal patterns of the cyclobutane protons as a diagnostic tool for the rapid structure elucidation of [2+2]-type cyclobutane-containing dimers. Particularly, the splitting patterns and coupling constants of cyclobutane protons can be accurately obtained through the online open-source QM-HiFSA tool (https://ctb.nmrsolutions.fi./).

  To further explore the applicability of the diagnostic signal patterns summarized in this study, we conducted a comprehensive literature review. As a result, the connection mode and stereochemistry of the cyclobutane rings of a series of reported phenylpropanoid-derived [2+2] dimers could be quickly identified by analyzing their ¹H NMR spectra using our summarized diagnostic signal patterns [12,13,33,51–53]. Additionally, we also observed similar diagnostic signal patterns in [2+2] homodimers derived from aryl-conjugated alkenes (styrenes, stilbenes, chalcones, etc.) [30,54–57]. Particularly, we found a strong correlation between the ¹H NMR profiles of cyclobutane protons in the revised structure of abrusamide A, a hydrocinnamamide-derived [2+2] homodimer with questionable original structure, and the corresponding summarized signal patterns [58,59]. These results demonstrate the feasibility of using this approach to elucidate the structures of [2+2]-type cyclobutane-containing natural products.

  According to the traditional use of P. rubra for inflammation-related diseases in Southern China, here, we established in vitro ulcerative colitis (UC) model based on unicellular (HT-29 and Caco-2 cell lines) and muticellular (Caco-2/THP-1) system to explore the potential therapeutic effect of compounds 1–3 on UC [60]. As demonstrated in Fig. S17 (Supporting information), compound 3 significantly decreased tumor necrosis factor-α (TNF-α) expression in HT-29 monolayer model compared to control group. Meanwhile, compounds 1 and 3 both significantly downregulated the expression of pro-inflammatory cytokines [TNF-α, interleukin-6 (IL-6), IL-1β, and IL-8] in lipopolysaccharide (LPS)-stimulated Caco-2 cells and Caco-2/THP-1 co-culture model. Collectively, these results suggest that compounds 1 and 3 exhibited anti-inflammatory activities and may serve as potential therapeutic agents for UC.

  In conclusion, three new iridoid glycoside [2+2] cyclobutane dimers (1–3) with unprecedented scaffolds were discovered from P. rubra. Compound 1 possesses a centrosymmetric phenylpropanoid-derived cyclobutane motif, compound 2 has a plane of symmetry in its phenylpropanoid-derived cyclobutane unit, and compound 3 contains a 2-fold (C₂) symmetry axis in its molecule, which posed a significant challenge for their structure elucidation. The planar structures and absolute configurations of compounds 1–3 were unequivocally determined by using a combination of multiple methods, including spectroscopic analysis, biomimetic semisynthesis, chemical degradation, and SCXRD analysis. In particular, the 3D structure of compound 2 was further verified using MicroED analysis, showcasing the potential of this emerging technique for elucidating the complete structure of complex natural products, including the ones with multiple stereogenic centers and highly substituted sugar moieties. Furthermore, the QM-HiFSA approach was employed to perform the complete ¹H NMR spectral analysis of the synthetic phenylpropanoid-derived [2+2] dimers 7–10. Notably, this analysis established correlations between the structures of cyclobutane motifs and the signal patterns of cyclobutane protons, which can serve as an empirical rule for the rapid and accurate assignment of dimerization mode and stereochemistry of the cyclobutane ring in phenylpropanoid-derived [2+2] dimers, as well as other types of [2+2]-type cyclobutane-containing derivatives. Remarkably, compounds 1 and 3 demonstrated efficient amelioration of mucosal inflammation in vitro, highlighting their potential as therapeutic options for UC treatment.

  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

  Yu Xiong: Writing – original draft, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Li-Jun Hu: Writing – review & editing, Writing – original draft, Conceptualization. Jian-Guo Song: Writing – review & editing, Formal analysis, Data curation. Di Zhang: Formal analysis, Data curation. Yi-Shuang Peng: Methodology, Investigation, Data curation. Xiao-Jun Huang: Writing – review & editing, Formal analysis, Data curation. Jian Hong: Writing – review & editing, Methodology. Bin Zhu: Writing – review & editing, Methodology, Data curation. Wen-Cai Ye: Writing – review & editing, Resources, Project administration. Ying Wang: Writing – review & editing, Resources, Project administration, Conceptualization.

  Acknowledgments

  This work was financially supported by the National Key R&D Program of China (No. 2023YFC3503902), the National Natural Science Foundation of China (Nos. 82293681(82293680) and 82321004), the Guangdong Basic and Applied Basic Research Foundation (Nos. 2022B1515120015 and 2021A1515111021), the Guangdong Major Project of Basic and Applied Basic Research (No. 2023B0303000026), and the Science and Technology Projects in Guangzhou (No. 202102070001). The authors are grateful to Mr. Matthias Niemitz, the founder of NMR Solutions Ltd., for granting access to Cosmic Truth software.

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110149.

  
  
• 1. [1]

     Y.Y. Fan, X.H. Gao, J.M. Yue, Sci. China Chem. 59 (2016) 1126–1141. doi: 10.1007/s11426-016-0233-1

  2. [2]

     M.R. van der Kolk, M.A.C.H. Janssen, F.P.J.T. Rutjes, D. Blanco-Ania, ChemMedChem 17 (2022) e202200020. doi: 10.1002/cmdc.202200020

  3. [3]

     T. Seiser, T. Saget, D.N. Tran, et al., Angew. Chem. Int. Ed. 50 (2011) 7740–7752. doi: 10.1002/anie.201101053

  4. [4]

     P.Y. Yang, Q. Jia, S.J. Song, et al., Nat. Prod. Rep. 40 (2023) 1094–1129. doi: 10.1039/D2NP00034B

  5. [5]

     V.M. Dembitsky, Phytomedicine 21 (2014) 1559–1581. doi: 10.1016/j.phymed.2014.07.005

  6. [6]

     B. Zhen, X.Y. Suo, J. Dang, et al., Chin. Chem. Lett. 32 (2021) 2338–2341. doi: 10.1016/j.cclet.2020.10.027

  7. [7]

     Y.Q. Du, L.G. Yao, X.W. Li, Y.W. Guo, Chin. Chem. Lett. 34 (2023) 107512. doi: 10.1016/j.cclet.2022.05.026

  8. [8]

     S. Poplata, A. Tröster, Y.Q. Zou, et al., Chem. Rev. 116 (2016) 9748–9815. doi: 10.1021/acs.chemrev.5b00723

  9. [9]

     D. Sarkar, N. Bera, S. Ghosh, Eur. J. Org. Chem. 10 (2020) 1310–1326.

  10. [10]

      W.R. Gutekunst, P.S. Baran, J. Org. Chem. 79 (2014) 2430–2452. doi: 10.1021/jo4027148

  11. [11]

      K.C. Nicolaou, S.A. Snyder, Angew. Chem. Int. Ed. 44 (2005) 1012–1044. doi: 10.1002/anie.200460864

  12. [12]

      Y.M. Ren, C.Q. Ke, C.P. Tang, et al., Tetrahedron Lett. 58 (2017) 2385–2388. doi: 10.1016/j.tetlet.2017.05.013

  13. [13]

      J. Li, L.H. Tan, H. Zou, et al., J. Nat. Prod. 83 (2020) 216–222. doi: 10.1021/acs.jnatprod.9b00470

  14. [14]

      X. Yuan, J.S. Jiang, Y.N. Yang, et al., Chin. Chem. Lett. 33 (2022) 2923–2927. doi: 10.1016/j.cclet.2021.10.085

  15. [15]

      I.L. Tsai, F.P. Lee, C.C. Wu, et al., Planta Med. 71 (2005) 535–542. doi: 10.1055/s-2005-864155

  16. [16]

      W.R. Gutekunst, P.S. Baran, J. Am. Chem. Soc. 133 (2011) 19076–19079. doi: 10.1021/ja209205x

  17. [17]

      S. Tsukamoto, B.C. Cha, T. Ohta, Tetrahedron 58 (2002) 1667–1671. doi: 10.1016/S0040-4020(02)00048-0

  18. [18]

      M. Takahashi, M. Ichikawa, S. Aoyag, C. Kibayashi, Tetrahedron Lett. 46 (2005) 57–59. doi: 10.1016/j.tetlet.2004.11.042

  19. [19]

      N. Tanaka, M. Okasaka, Y. Ishimaru, et al., Org. Lett. 7 (2005) 2997–2999. doi: 10.1021/ol050960w

  20. [20]

      K.C. Nicolaou, D. Sarlah, D.M. Shaw, Angew. Chem. Int. Ed. 46 (2007) 4708–4711. doi: 10.1002/anie.200701552

  21. [21]

      N. Tanaka, Y. Kashiwada, S.Y. Kim, et al., J. Nat. Prod. 72 (2009) 1447–1452. doi: 10.1021/np900109y

  22. [22]

      K.C. Nicolaou, S. Sanchini, T.R. Wu, D. Sarlah, Chem. Eur. J. 16 (2010) 7678–7682. doi: 10.1002/chem.201001474

  23. [23]

      S. Malhotra, S.K. Koul, S.C. Taneja, et al., Phytochemistry 29 (1990) 2733– 2734. doi: 10.1016/0031-9422(90)85233-6

  24. [24]

      R.N. Mahindru, S.C. Taneja, K.L. Dhar, R.T. Brown, Phytochemistry 32 (1993) 1073–1075. doi: 10.1016/0031-9422(93)85260-X

  25. [25]

      A.U. Rahman, K.F. Khattak, F. Nighat, et al., Phytochemistry 48 (1998) 377–383. doi: 10.1016/S0031-9422(97)01079-0

  26. [26]

      H.Y. Pei, L.L. Xue, M.H. Tang, et al., J. Agric. Food Chem. 68 (2020) 2406–2417. doi: 10.1021/acs.jafc.9b07754

  27. [27]

      J. Wang, J.G. Song, D.L. Zong, et al., Angew. Chem. Int. Ed. 62 (2023) e202312568. doi: 10.1002/anie.202312568

  28. [28]

      Q.F. He, Z.L. Wu, L.R. Li, et al., Angew. Chem. Int. Ed. 60 (2021) 19609–19613. doi: 10.1002/anie.202103878

  29. [29]

      L.M. Deng, L.J. Hu, Y.T.Z. Bai, et al., Org. Lett. 23 (2021) 4499–4504. doi: 10.1021/acs.orglett.1c01616

  30. [30]

      S.G. Li, X.J. Huang, M.M. Li, et al., J. Nat. Prod. 81 (2018) 254–263. doi: 10.1021/acs.jnatprod.7b00540

  31. [31]

      Y.Y. Xia, C.Z. Lin, X.J. Lu, et al., Phytochem. Lett. 25 (2018) 81–85. doi: 10.1016/j.phytol.2018.02.003

  32. [32]

      J.J.W. Coppen, A.L. Cobb, Phytochemistry 22 (1983) 125–128. doi: 10.1016/S0031-9422(00)80071-0

  33. [33]

      G.L. Ma, J. Xiong, G.X. Yang, et al., J. Nat. Prod. 79 (2016) 1354–1364. doi: 10.1021/acs.jnatprod.6b00061

  34. [34]

      A. Yaşar, N. Yaylı, A. Usta, N. Yaylı, J. Mol. Model. 16 (2010) 1347–1355. doi: 10.1007/s00894-009-0642-7

  35. [35]

      D. Shi, B.L. Nannenga, M.G. Iadanza, T. Gonen, eLife 2 (2013) e01345. doi: 10.7554/eLife.01345

  36. [36]

      B.L. Nannenga, D. Shi, A.G.W. Leslie, T. Gonen, Nat. Methods 11 (2014) 927– 930. doi: 10.1038/nmeth.3043

  37. [37]

      B.L. Nannenga, T. Gonen, Nat. Methods 16 (2019) 369–379. doi: 10.1038/s41592-019-0395-x

  38. [38]

      M.W. Martynowycz, M.T.B. Clabbers, J. Hattne, T. Gonen, Nat. Methods 19 (2022) 724–729. doi: 10.1038/s41592-022-01485-4

  39. [39]

      E. Danelius, S. Halaby, W.A. van der Donk, T. Gonen, Nat. Prod. Rep. 28 (2021) 421–423.

  40. [40]

      E. Danelius, K. Patel, B. Gonzalez, T. Gonen, Curr. Opin. Struc. Biol. 79 (2023) 102549. doi: 10.1016/j.sbi.2023.102549

  41. [41]

      T. Kunde, B.M. Schmidt, Angew. Chem. Int. Ed. 58 (2019) 666–668. doi: 10.1002/anie.201813215

  42. [42]

      L.J. Kim, M.Z. Xue, X. Li, et al., J. Am. Chem. Soc. 143 (2021) 6578–6585. doi: 10.1021/jacs.1c01729

  43. [43]

      J.D. Park, Y.C. Li, K. Moon, et al., Angew. Chem. Int. Ed. 61 (2022) e202114022. doi: 10.1002/anie.202114022

  44. [44]

      M. Kumar, P. Rawat, N. Rahuja, et al., Phytochemistry 70 (2009) 1448–1455. doi: 10.1016/j.phytochem.2009.07.029

  45. [45]

      Y.S. Wang, B.T. Li, S.X. Liu, et al., J. Nat. Prod. 80 (2017) 798–804. doi: 10.1021/acs.jnatprod.6b00391

  46. [46]

      H.J. Xie, X.Y. Chen, M.M. Li, et al., Tetrahedron Lett. 61 (2020) 151946. doi: 10.1016/j.tetlet.2020.151946

  47. [47]

      J.G. Napolitano, D.C. Lankin, J.B. McAlpine, et al., J. Org. Chem. 78 (2013) 9963–9968. doi: 10.1021/jo4011624

  48. [48]

      S.X. Jing, M. Reis, Y. Alania, et al., J. Agric. Food Chem. 70 (2022) 12456–12468. doi: 10.1021/acs.jafc.2c04288

  49. [49]

      Y. Tang, J.B. Friesen, D.C. Lankin, et al., J. Nat. Prod. 86 (2023) 256–263. doi: 10.1021/acs.jnatprod.2c00836

  50. [50]

      S. Sommerwerk, R. Kluge, D. Ströhl, et al., Tetrahedron 72 (2016) 1447–1454. doi: 10.1016/j.tet.2016.01.045

  51. [51]

      H.V. Stein, C.J. Berg, J.N. Maung, et al., Environ. Sci.: Processes Impacts 19 (2017) 851–860. doi: 10.1039/C7EM00059F

  52. [52]

      T.B. Nguyen, A. Al-Mourabit, Photochem. Photobiol. Sci. 15 (2016) 1115–1119. doi: 10.1039/c6pp00201c

  53. [53]

      R. Telmesani, S.H. Park, T. Lynch-Colameta, A.B. Beeler, Angew. Chem. Int. Ed. 54 (2015) 11521–11525. doi: 10.1002/anie.201504454

  54. [54]

      R.A. Davis, A.R. Carroll, S. Duffy, et al., J. Nat. Prod. 70 (2007) 1118–1121. doi: 10.1021/np070073x

  55. [55]

      Y. Liu, X.J. Zhang, N. Kelsang, et al., J. Nat. Prod. 81 (2018) 307–315. doi: 10.1021/acs.jnatprod.7b00736

  56. [56]

      L.W. Tian, Y.J. Feng, T.D. Tran, et al., Bioorg. Med. Chem. Lett. 27 (2017) 4007–4010. doi: 10.1016/j.bmcl.2017.07.062

  57. [57]

      M.J. Genzink, M.D. Rossler, H. Recendiz, T.P. Yoon, J. Am. Chem. Soc. 145 (2023) 19182–19188. doi: 10.1021/jacs.3c07132

  58. [58]

      X.J. Yuan, L. Lin, X.Q. Zhang, S.D. Deng, Phytochem. Lett. 7 (2014) 137–142. doi: 10.1016/j.phytol.2013.11.003

  59. [59]

      X.J. Yuan, Y.D. Liu, H. Zhao, et al., Phytochemistry 181 (2021) 112572. doi: 10.1016/j.phytochem.2020.112572

  60. [60]

      E.J. Wang, R. Han, M.Y. Wu, et al., Chin. Chem. Lett. 35 (2024) 108361. doi: 10.1016/j.cclet.2023.108361

• 

  Figure 1  Chemical structures of plumerubradins A–C (1–3).

  下载: 全尺寸图片 幻灯片
  

  Scheme 1  Biomimetic semisynthesis of 1–3 and chemical degradation of 4. Reagents and conditions: (a) 4 (0.97 mmol), cyclohexane (20 mL), N₂, room temperature, 75 W 390 nm LED, 48 h; (b) 4 (0.32 mmol), CH₃OH (25 mL), K₂CO₃ (2.60 mmol), room temperature, 5 h. LED = light-emitting diode.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Syntheses of 7–10 and 4-bromobenzoate derivatives 8a, 9a, and 10a. THF = tetrahydrofuran.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Structure determination of 2 by MicroED method. (A) Corresponding electron density map superimposed on the refined structure (1.0 Å resolution). (B) Oak Ridge thermal ellipsoid plot (ORTEP) drawing of 2.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Enlarged stacked ¹H NMR spectra of 1–3 and 7–9: (A) Comparison of 1 and 7 in CD₃OD; (B) Comparison of 2 and 8 in CD₃OD; (C) Comparison of 3 and 9 in DMSO-d₆. (D) Enlarged ¹H NMR spectrum of 10 in CD₃OD. DMSO = dimethylsulfoxide.

  下载: 全尺寸图片 幻灯片

  Table 1.  Correlations between the ¹H NMR signal patterns and the structural information of cyclobutane ring in [2+2]-type cyclobutane derivatives.

  下载: 导出CSV
•