Structural diversification of phenylspirodrimane lactams by employing a biosynthetic intermediate

Jimei Liu Zhenfei Wang Jun Wu Yaotian Han Fei Ye Tiantai Zhang Haibo Yu Zhengshun Wen Jungui Dai

Citation:  Jimei Liu, Zhenfei Wang, Jun Wu, Yaotian Han, Fei Ye, Tiantai Zhang, Haibo Yu, Zhengshun Wen, Jungui Dai. Structural diversification of phenylspirodrimane lactams by employing a biosynthetic intermediate[J]. Chinese Chemical Letters, 2023, 34(12): 108363. doi: 10.1016/j.cclet.2023.108363 shu

Structural diversification of phenylspirodrimane lactams by employing a biosynthetic intermediate

English

  • Phenylspirodrimanes, a type of meroterpenoid that are derived from polyketide–terpenoid hybrid pathway [1,2]. The pharmacological studies revealed that this kind of natural products displayed a broad spectrum of biological activities, including endothelin receptor antagonism, anticomplement, antiviral, antibacterial, anti-inflammatory, neuroprotective, and cytotoxic activities [310]. In our previous investigation on the metabolites from Stachybotrys chartarum CGMCC 3.5365, a series of phenylspirodrimanes [710] were isolated, among which the dimer stachybocin A (1) was present only at a very low level. Pharmacological evaluation reported that 1 was an endothelin receptor antagonist [3,11], inspiring Kende and his co-workers to the first enantioselective total syntheses from (+)-Wieland-Miescher ketone, through 24 steps but only in 1.1% overall yield [12]. In addition, our recent bioactive evaluation revealed that this phenylspirodrimane lactam dimer also displayed potent N-methyl-D-aspartate (NMDA) receptor antagonistic activity (Fig. S1 in Supporting information). However, its trace amount in nature and the difficulty to access and structurally modify by chemical synthesis limit the further drug discovery.

    Natural product biosynthesis often underwent multi-step combination of enzymatic and non-enzymatic conversions to produce a variety of final natural products [2,13,14]. Especially, a growing number of examples of natural products that involve non-enzymatic step in their formation, such as discoipyrroles [15], chimedermycins [16], pyonitrins [17], dibohemamines [18], jadomycins [19], and oxazinin A [20]. In the exploration and elucidation of biosynthesis of natural products, the multipotent biosynthetic intermediates often form various branch pathways to produce different scaffolds. Furthermore, the structural motif of biosynthetic intermediates with highly reactive property would exhibit great potential for use as starting material in diversity-oriented synthesis [21]. Carefully analysis of intermediate behavior at the structural level in the elucidation of biosynthetic pathway of natural products might result in the discovery of multipotent intermediates use for exploring the chemical diversity of pseudo-natural compounds.

    In this study, we utilized the supposed biosynthetic intermediate stachbotrydial (2) with high-reactivity o-phthalaldehyde unit and various amines including natural/unnatural amino acids to yield a number of structurally different phenylspirodrimane lactams (1, 325; Fig. 1) by non-enzymatic reaction. Among them, 18 products (36, 1119, 2125) were new compounds, and the yield of the bioactive dimer 1 reached 18.7 mg/g of cell dry weight. Herein, we report the microbial diversification, structural elucidation, and biological activities of these compounds.

    Figure 1

    Figure 1.  Structural diversity of phenylspirodrimane lactams from reactive biosynthetic intermediate 2 and different amines.

    Since the trace amount of 1 in S. chartarum CGMCC 3.5365, we initially attempted to identify all the genes/enzymes responsible for the biosynthesis of 1 and constructed all biosynthetic genes into Aspergillus oryzae NSAR1 to provide a platform to produce 1. Unfortunately, we haven't identified all the genes for the biosynthesis of 1 yet. However, we observed that adding L-lysine into the cultures of this fungal strain dramatically increased the production of 1, which indicated the presence of a key precursor for the biosynthesis of 1. Based on the structurally analysis, we presumed that the dimer 1 might be synthesized by nonenzymatic reaction between L-lysine and the highly reactive stachbotrydial (2) via Schiff-base formation, reduction, and lactam ring formation (Fig. S2 in Supporting information) [22]. This hypothesis was confirmed by testing 2 for reaction with L-lysine in buffer (50 mmol/L sodium phosphate, pH 7.0), acetonitrile, and H2O, respectively (Fig. 2), in which 1 was generated in a high yield. This result suggested that 2 was a direct precursor of 1 and might be a suitable starting material for chemical diversification because of the presence of high reactive o-phthalaldehyde unit. Subsequently, four pairs of phenylspirodrimane lactams (3/7, 4/8, 5/9, and 6/10) were yielded by chemical reaction between 2 and four natural amino acids (L-valine, L-alanine, L-glycine, and L-leucine) under reflux in CH3CN, respectively, further supporting the above deduction. In addition, the carbonyl of lactam ring of these products preferred to locate at C-22 rather than at C-23.

    Figure 2

    Figure 2.  Nonenzymatic formation of stachybocin A (1). (A) HPLC analysis of the nonenzymatic reaction which were performed in 100 µL reaction mixture containing 0.4 mmol/L of 2 and 0.2 mmol/L of L-lysine at 30 ℃ for 12 h (B) UV and MS spectra of 1. (C) UV and MS spectra of 2.

    To obtain a large amount of bioactive lactam dimer 1 and structurally diverse analogues for further biological evaluation and structure-activity relationship investigation, the high-yielding culture media of S. chartarum for 2 were screened. The results showed that 2 was accumulated with much higher yield in Czapeks medium (7.0 mg/g of cell dry weight) monitored by HPLC analysis (Fig. S3 in Supporting information). In addition, the reaction conditions of the microbial conversion of 2 to 1 were investigated and optimized (Fig. S4 in Supporting information). The results suggested that incubating in Czapeks medium at pH 5.0 for 4 days with the addition of 1 mg/mL of L-lysine on 0 day led to the optimal conversion. Subsequently, the feeding experiments were performed, in which various amines were added to Czapeks medium with S. chartarum cultures, respectively, and a series of natural and pseudo-natural phenylspirodrimane lactams (1125, Fig. 1 and Fig. S5 in Supporting information) were generated. When mono-amines including natural amino acids (L-methionine, L-aspartic acid, L-tyrosine, L-tryptophan, and L-phenylalanine), un-natural amino acid (D-phenylalanine), and other amines (racemate 3-aminopiperdine-2,6–dione) were fed, phenylspirodrimane lactam monomers (1123) were yielded. While, feeding diamines including L-lysine, D,L-2,6-diaminopimelic acid and trans-cyclohexane-1,4-diamine led to the production of phenylspirodrimane lactam dimers (1, 24, and 25). In addition, by comparison with chemical conversion, microbial conversion of 2 by feeding amino acids (L-valine, L-alanine, L-glycine, and L-leucine), only C-22 carbonyl products were generated (Figs. S5J-M), however, when feeding L-tryptophan, the C-22 and C-23 carbonyl products were obtained (Fig. S5E). These results indicated that the synthesis of phenylspirodrimane lactams can be designed and structurally diversified based on the structures of fed-amines and conversion approaches.

    As described above, totally 24 phenylspirodrimane lactams (1, 325) including 18 new compounds were synthesized. The structures of the new products (36, 1119, 2125) were elucidated by extensive spectroscopic data, single-crystal X-ray diffraction (Cu Kα), and calculated electronic circular dichroism (ECD) analyses. Taken two monomers (11 and 12) and one dimer (24) as examples, their structures were characterized as follows.

    Stachybotrylactam E (11) was the microbial accumulation product by adding L-methionine in the cell cultures. It was obtained as colorless columnar crystals (MeOH), and its molecular formula was determined as C28H39O6NS according to the HRESIMS peak at m/z 518.2571 [M + H]+ (calcd. for C28H40O6NS, 518.2571) with ten degrees of unsaturation. The IR spectrum indicated the presence of hydroxy (3412 cm-1) and carbonyl (1667 cm-1) groups. The UV spectrum displayed maximum absorptions at 218, 265, and 304 nm. By comparison with NMR spectroscopic data of 2, a set of typical phenylspirodrimane signals were observed in the 1H NMR spectrum (Table S2 in Supporting information) [δH 6.73 (1H, s, H-18); δH 3.34 (1H, d, J = 2.9 Hz, H-3); δH 3.26 (1H, d, J = 16.9 Hz, H-11a)/2.88 (1H, d, J = 16.9 Hz, H-11b); δH 1.04 (3H, s, H3-15); δH 0.98 (3H, s, H3-13); δH 0.87 (3H, s, H3-14); δH 0.72 (3H, d, J = 6.6 Hz, H3-12)]. In addition, the presence of one nonequivalent methylene [δH 4.43 (1H, d, J = 16.0 Hz, H-23a)/4.36 (1H, d, J = 16.0 Hz, H-23b)], a methine [δH 5.13 (1H, dd, J = 10.2, 5.1 Hz, H-1′)], and a methyl [δH 2.10 (3H, s, H-4′)]. The 13C NMR and DEPT spectroscopic data (Table S2) indicated 28 carbon resonances, including five methyl carbons (δC 29.0, 22.9, 16.5, 16.0, and 15.2), eight methylene carbons (δC 45.1, 32.7, 31.9, 31.4, 29.9, 26.0, 25.0, and 21.7), five methine carbons (δC 102.1, 75.4, 53.6, 40.6, and 37.9, including one aromatic, and one oxygenated), one carbonyl carbon (δC 169.6), one carboxyl carbon (δC 172.8), five nonprotonated aromatic carbons (δC 157.4, 154.7, 134.8, 118.1, and 114.2, including two oxygenated), one oxygenated tertiary carbon (δC 99.3), and two quaternary carbons (δC 43.1 and 38.4). These data suggested compound 11 had a typical phenylspirodrimane skeleton bearing an isoindolinone unit. The 1H-1H COSY and HMBC data (Fig. 3) established a methionine moiety, which was connected to the nitrogen of isoindolinone unit, as evident from the methine proton H-1′ (δH 5.13, dd, J = 10.2, 5.1 Hz) correlated to C-22 and C-23. On the basis of reaction mechanism considerations [22], compound 11 probably had the same absolute configuration as that of 2, and the intact incorporation the L-methionine moiety should maintain its natural absolute configuration. Thus, 11 might has a 3R, 5S, 8R, 9R, 10S, and 1′R configuration, which was further supported by single-crystal X-ray diffraction analysis using the anomalous scattering of Cu Kα radiation (Fig. 4).

    Figure 3

    Figure 3.  1H−1H COSY (▬) and key HMBC (→) correlations of 11, 12, and 24.

    Figure 4

    Figure 4.  Single-crystal X-ray diffractions of 11 and 12.

    Stachybotrylactam F (12) was the microbial accumulation product by supplying L-aspartic acid. It was obtained as colorless columnar crystals (MeOH), and its molecular formula was determined as C27H35O8N according to the HRESIMS peak at m/z 502.2435 [M + H]+ (calcd. for C27H36O8N, 502.2435). Analyses of the 2D NMR (1H-1H COSY, HSQC, and HMBC) data revealed that 12 possessed a phenylspirodrimane lactam framework same as 11 except for the N-linked side chain. The HMBC correlations of the side chain resonances established an aspartic acid group. This moiety was linked to the isoindolinone unit through N atom, according to the HMBC interactions from H-1′ (δH 5.24, dd, J = 8.6, 5.8 Hz) to C-22 (δC 171.4) and C-23 (δC 46.9) (Fig. 3). Single-crystal X-ray single diffraction analysis (Cu Kα) (Fig. 4) further confirm the above deduction and the stereochemistry of 12 was unambiguously characterized as 3R, 5S, 8R, 9R, 10S, and 1′R configuration, indicating that the absolute configurations of the phenylspirodrimane nucleus and aspartic acid group remained naturally.

    Stachybotrylactam Q (24) was the microbial conversion product by feeding d, L-2,6-diaminopimelic acid. It has a molecular formula of C53H70O12N2 as determined by HRESIMS ion peak at m/z 927.5017 [M + H]+ (calcd. for C53H71O12N2, [M + H]+ 927.5002). From the 13C NMR spectrum (Table S6 in Supporting information), most of spectroscopic data appeared in pairs and had the similar chemical shifts. In the 1H NMR, two sets of the typical phenylspirodrimane signals were observed, and most of had identical chemical shifts and coupling constants except for those at δH 3.18 (H-3, brs)/3.15 (H-3′, brs), δH 0.64 (H3-12, d, J = 7.9 Hz)/0.62 (H3-12′, d, J = 6.4 Hz), δH 0.95 (H3-15, s)/0.94 (H3-15′, s), δH 6.56 (H-18, s)/δH 6.57 (H-18′, s), δH 4.30 (H-23a, d, J = 16.4 Hz) and 4.19 (H-23b, d, J = 16.4 Hz)/δH 4.33 (H-23′a, d, J = 16.4 Hz) and 4.20 (H-23′b, d, J = 16.4 Hz). These data strongly indicated 24 to be a symmetrical structure bearing two identical phenylspirodrimane units. Further analysis of its 2D NMR spectroscopic data confirmed the two highly identical monomers were linked by 2,6-diaminopimelic acid through formed two isoindolinone units. Careful comparison of the chemical shifts and coupling constants of their protons with those reported previously [710], together with consideration of reaction mechanism, suggested that 24 has 3R, 5S, 8R, 9R, 10S, 3′R, 5′S, 8′R, 9′R, 10′S, 1′′R, and 5′′R configuration.

    On the basis of the HRESIMS and 2D NMR spectroscopic analyses (Tables S1–S6), the structures of stachybotrylactams A–D (36), G–P (1319, 2123), and stachybotrylactam R (25) were assigned as new phenylspirodrimane lactams. The absolute configurations of stachybotrylactams O (22) and P (23) were elucidated by comparison of the calculated ECD with their experimental ECD data (Fig. S6 in Supporting information). In addition, five known analogues were identified as O-demethylstachartin C (7), N-(2-propanoic acid) stachybotrylactam (8) [23,24], stachybotrin H (9) [23], N-[2-(4-methylpentanoic acid)] stachybotrylactam (10) [23,24], and N-(2-benzenpropanoic acid) stachybotrylactam (20) [23,24]. Furthermore, the NMR data of 7 was reported for the first time (Supporting information).

    In these effective structural diversification process, the yields of the products were different, possibly due to the different cellular uptake ability of amines and the effects of added amines on the growth of this strain. For example, we observed that 3-aminopiperdine-2,6–dione inhibited the growth of this strain. Nevertheless, most of the compounds were generated in high yields, which enable all the compounds in an adequate amount for a wide of bioactivity test. In the present study, these compounds were bio-assayed for hNaV 1.2 channels inhibitory, protein tyrosine phosphatase 1B (PTP1B) inhibitory, and anti-inflammatory activities. The effects of these compounds on hNaV1.2 channels mediated currents were recorded using the patch-clamp whole-cell configuration at a concentration of 10 µmol/L, respectively (Fig. 5), among which 1, 17, and 24 exerted potent inhibitory effects on the inactivated state of hNaV 1.2 channels with IC50 values of 0.22, 2.08, and 0.53 µmol/L, respectively (Table 1), of which activities were much stronger than the positive control carbamazepine. Analyzing the structure-activity relationship revealed that phenylspirodrimane lactam dimers exhibited higher inhibitory effect than monomers, especially the presence of flexible linkage unit between two monomers, and a N-linked indole moiety in the phenylspirodrimane lactam monomer might benefit for inhibitory activity. What's more, the inhibition effects of these compounds on the inactivated state were higher than those on the resting state (Fig. 5), suggesting their potential for new antiepilepsy agents.

    Figure 5

    Figure 5.  The inhibition of the test compounds (1, 3–14, 16–25) on hNaV 1.2 channels. The blue and red columns represent the inhibition level of test compounds at 10 µmol/L on the resting and inactivated state of hNaV 1.2 channels, respectively.

    Table 1

    Table 1.  Inhibition and state selectivity of compounds 1, 17, and 24 on hNaV 1.2 channels.
    DownLoad: CSV

    In addition, compound 1 displayed PTP1B inhibitory activity with an IC50 value of 1.32 µmol/L (CCF06240 as the positive control, IC50 = 1.70 µmol/L), and potent anti-inflammatory activity by suppressing LPS-induced TNF-α production in RAW264.7 cells with an IC50 value of 3.86 µmol/L (dexamethasone as positive control, IC50 of 0.048 µmol/L).

    In summary, inspired by the nonenzymatic reaction between amines and the highly reactive biosynthetic intermediate 2, we have developed a diversity-oriented microbial transformation and chemical reaction process with structurally diverse amines to efficiently access 24 phenylspirodrimane lactams including 18 new analogues. Bioassay results showed that compounds 1, 17, and 24 displayed significant inhibitory effects on the inactivated state of hNaV 1.2 channels with the potential for new antiepilepsy agents. Compound 1 exhibited various and potent activities, including PTP1B inhibitory and anti-inflammatory activities. The present approach is undoubtedly available to generate new analogues of phenylspirodrimane lactams. What's more, this work would suggest that multipotent intermediates in the biosynthesis of natural products could be employed for non-enzymatically structural diversification in drug discovery.

    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.

    This work was financially supported by National Natural Science Foundation of China (No. 81803403), CAMS Innovation Fund for Medical Sciences (Nos. CIFMS-2022-I2M-JB-011 and CIFMS-2021-I2M-1-029).

    Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2023.108363.


    1. [1]

      R. Geris, T.J. Simpson, Nat. Prod. Rep. 26 (2009) 1063–1094. doi: 10.1039/b820413f

    2. [2]

      Y. Matsuda, I. Abe, Nat. Prod. Rep. 33 (2016) 26–53. doi: 10.1039/C5NP00090D

    3. [3]

      K. Ogawa, M. Nakamura, M. Hayashi, et al., J. Antibiot. 48 (1995) 1396–1400. doi: 10.7164/antibiotics.48.1396

    4. [4]

      B. Wu, V. Oseker, J. Wiese, et al., Mar. Drugs 12 (2014) 1924–1938. doi: 10.3390/md12041924

    5. [5]

      A. Wang, Y. Xu, Y. Gao, et al., Phytochem. Rev. 14 (2015) 623–655. doi: 10.1007/s11101-014-9365-1

    6. [6]

      P. Zhang, Y. Li, C. Jia, et al., RSC Adv. 7 (2017) 49910–49916. doi: 10.1039/C7RA09859F

    7. [7]

      J. Zhao, J. Feng, Z. Tan, et al., J. Nat. Prod. 80 (2017) 1819–1826. doi: 10.1021/acs.jnatprod.7b00014

    8. [8]

      J. Zhao, J. Feng, Z. Tan, et al., Bioorg. Med. Chem. Lett. 28 (2018) 355–359. doi: 10.1016/j.bmcl.2017.12.039

    9. [9]

      M. Zhang, J. Feng, X. Jia, et al., Chin. Chem. Lett. 30 (2019) 435–438. doi: 10.1016/j.cclet.2018.04.031

    10. [10]

      J. Liu, X. Jia, J. Zhao, et al., Org. Chem. Front. 7 (2020) 531–542. doi: 10.1039/c9qo01284b

    11. [11]

      M. Nakamura, Y. Ito, K. Ogawa, et al., J. Antibiot. 48 (1995) 1389–1395. doi: 10.7164/antibiotics.48.1389

    12. [12]

      W.P. Deng, M. Zhong, X.C. Guo, A.S. Kende, J. Org. Chem. 68 (2003) 7422–7427. doi: 10.1021/jo030202o

    13. [13]

      J. Fan, G. Liao, F. Kindinger, et al., J. Am. Chem. Soc. 141 (2019) 4225–4229. doi: 10.1021/jacs.9b00110

    14. [14]

      L.M. Bouthillette, V. Aniebok, D.A. Colosimo, D. Brumley, J.B. MacMillan, Chem. Rev. 122 (2022) 14815–14841. doi: 10.1021/acs.chemrev.2c00306

    15. [15]

      D.A. Colosimo, J.B. MacMillan, J. Am. Chem. Soc. 138 (2016) 2383–2388. doi: 10.1021/jacs.5b13320

    16. [16]

      S. Yin, Z. Liu, J. Shen, et al., Nat. Commun. 13 (2022) 5169.

    17. [17]

      R.D. Shingare, V. Aniebok, H.W. Lee, J.B. MacMillan, Org. Lett. 22 (2020) 1516–1519. doi: 10.1021/acs.orglett.0c00098

    18. [18]

      P. Fu, A. Legako, S. La, J.B. MacMillan, Chem. Eur. J. 22 (2016) 3491–3495. doi: 10.1002/chem.201600024

    19. [19]

      A.W. Robertson, C.F. Martinez-Farina, D.A. Smithen, et al., J. Am. Chem. Soc. 137 (2015) 3271–3275. doi: 10.1021/ja5114672

    20. [20]

      Z. Lin, M. Koch, M.H. Abdel Aziz, et al., Org. Lett. 16 (2014) 4774–4777. doi: 10.1021/ol502227x

    21. [21]

      T. Asai, K. Tsukada, S. Ise, et al., Nat. Chem. 7 (2015) 737–743. doi: 10.1038/nchem.2308

    22. [22]

      S.M. Allin, C.C. Hodkinson, N. Taj, Synlett 8 (1996) 781–782.

    23. [23]

      X. Ma, W. Zheng, K. Sun, et al., Nat. Prod. Res. 33 (2019) 386–392. doi: 10.1080/14786419.2018.1455041

    24. [24]

      E. Claudia, K. Michael, T. Luigi, DE 10258650A1. 2004-06-24.

  • Figure 1  Structural diversity of phenylspirodrimane lactams from reactive biosynthetic intermediate 2 and different amines.

    Figure 2  Nonenzymatic formation of stachybocin A (1). (A) HPLC analysis of the nonenzymatic reaction which were performed in 100 µL reaction mixture containing 0.4 mmol/L of 2 and 0.2 mmol/L of L-lysine at 30 ℃ for 12 h (B) UV and MS spectra of 1. (C) UV and MS spectra of 2.

    Figure 3  1H−1H COSY (▬) and key HMBC (→) correlations of 11, 12, and 24.

    Figure 4  Single-crystal X-ray diffractions of 11 and 12.

    Figure 5  The inhibition of the test compounds (1, 3–14, 16–25) on hNaV 1.2 channels. The blue and red columns represent the inhibition level of test compounds at 10 µmol/L on the resting and inactivated state of hNaV 1.2 channels, respectively.

    Table 1.  Inhibition and state selectivity of compounds 1, 17, and 24 on hNaV 1.2 channels.

    下载: 导出CSV
  • 加载中
计量
  • PDF下载量:  3
  • 文章访问数:  559
  • HTML全文浏览量:  13
文章相关
  • 发布日期:  2023-12-15
  • 收稿日期:  2023-01-15
  • 接受日期:  2023-03-20
  • 修回日期:  2023-03-07
  • 网络出版日期:  2023-03-24
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

/

返回文章