Asperones A–E, five dimeric polyketides with new carbon skeletons from the fungus Aspergillus sp. AWG 1–15

Guo-Ping Yin, Ya-Rong Wu, Chao Han, Xiao-Bing Wang, Hong-Liang Gao, Yong Yin, Ling-Yi Kong* and Ming-Hua Yang*
Jiangsu Key Laboratory of Bioactive Natural Product Research and State Key Laboratory of Natural Medicines, China Pharmaceutical University, 24 Tong Jia Xiang, Nanjing 210009, People's Republic of China. E-mail: cpu_lykong@126.com; skeepjack@163.com

Received 21st January 2018 , Accepted 14th March 2018

First published on 15th March 2018


Asperones A–E (1–5), five dimeric polyketides with two distinct skeletons generated by the crucial [3 + 2] and [3 + 3] cycloadditions, were isolated from fungal metabolites of an Aspergillus sp. Their structures were established by spectroscopic data, chemical derivatization, single-crystal X-ray diffraction, and electronic circular dichroism calculations. Compounds 2–4 showed significant nitric oxide (NO) inhibition in lipopolysaccharide (LPS)-induced RAW 264.7 macrophage cells, and exhibited IC50 values of 16.0, 13.2, and 6.0 μM, respectively.


Introduction

Fungal secondary metabolites (FSMs) are one of the most attractive sources of natural products for the pharmaceutical industry. Their unique structures and potential biological activities have inspired chemists, pharmacologists, and biologists and stimulated new drug development.1 Aspergillus sp. metabolizes various natural products such as alkaloids, polyketides, peptides, and terpenes;2–4 thus, it is a vital source of FSMs. More impressively, numerous metabolites with novel and fascinating carbon skeletons have been reported from Aspergillus sp. A fantastic example is the recently reported series of cytochalasan complexes including asperchalasine A,5a epicochalasines A and B,5b and aspergilasines A–D.5c They not only have incredibly complicated structures but also show significant cytotoxic activities. However, these impressive dimers, trimers, and even tetramers were produced in very small quantities even in vast fermentation processes. Actually, the low content of the target compounds is a challenge often faced when uncovering novel metabolites from fungi, such as asperterpenes A and B from Aspergillus terreus,6a dalesconols A and B from mantis-associated fungus,6b and pyrazolofluostatins A–C from Micromonospora rosaria.6c Therefore, trace fungal metabolites require careful inspection in the discovery of bioactive natural products.

In our previous search for structurally intriguing and biologically valuable fungal metabolites, an Aspergillus sp. strain isolated from the intestines of centipedes was found to produce novel aromatic polyketides.7 Those polyketides were shown to have NO inhibitory effects on LPS-induced RAW 264.7 macrophage cells, indicating their potential as lead anti-inflammatory agents. In a further investigation, we detected two additional anti-inflammatory subfractions with trace constituents possessing similar molecular weights to the reported compounds (ESI, Fig. S1–S3). However, the low concentrations of these compounds prevented us from isolating and identifying them. In this work, extended fermentation as well as HPLC-MS- and UV-guided isolation were performed and resulted in the isolation of another five dimeric polyketides. Compared with previously reported dimers,7 the furanone derivatives isolated here are cyclized with two distinct aromatic polyketides to generate two [3 + 2] (1 and 2) and three [3 + 3] (3–5) cycloadducts, possessing the novel bicyclo[3.2.1]octane and spiro[bicyclo[3.3.1]nonane-2,2′-furan]-4′-one skeletons, respectively. Herein, the isolation, structural elucidation, and bioactivity as well as plausible biosynthetic pathways of compounds 1–5 are reported.

Results and discussion

Asperone A (1) was obtained as a colorless oil and was determined to have the molecular formula C26H32O8 by HR-ESI-MS. In its 1H NMR spectrum, six methyl signals were easily observed as two singlets, three doublets, and one triplet. In addition, there was a conjugated diene with four coupled olefinic protons along with seven additional aliphatic protons (Table S1). With the aid of HSQC and DEPT data, the corresponding proton-bearing carbons were identified in the 13C NMR spectra, and the remaining carbon resonances included three ketone, one carboxyl, four olefinic, and two aliphatic carbons. These NMR data resembled those of citrifurans A–D, which we reported previously,7 suggesting that 1 was also a polyketide dimer.

The same 7,8-dihydonivefuranone unit7,8 was confirmed by the 1H–1H COSY correlations of H-11′/H-12′/H-13′/H-14′/H2-15′/H3-16′ and H-7′/H-8′/H3-9′ as well as the HMBC correlations from H3-10′ to C-2′, C-3′ and C-11′ and from H-8′ to C-5′ (Fig. 2). The structure of the other polyketide unit was quite different from that of the citrinin9 unit in citrifurans A–D.7 A sec-butanol moiety was revealed by the 1H–1H COSY correlations of H3-9/H-8/H-7/H3-10. The remaining signals were assembled as a 6-methylcyclohex-4-ene-1,3-dione fragment mainly by the key HMBC correlations between H-3 and C-1/C-2/C-4/C-5 and between H3-11 and C-1/C-2/C-6. Meanwhile, the olefinic C-5 position was hydroxylated based on its downfield chemical shift at δC 147.4. The sec-butanol moiety was attached at C-6 based on the HMBC correlations from H-7 to C-5 and C-6 and from H3-10 to C-6, which completes the second polyketide unit as shown in Fig. 2. Moreover, one degree of unsaturation was still unaccounted for, which along with the HMBC correlation from H3-11 to C-7′ and the 1H–1H COSY correlation of H-8′/H-3, confirmed the C-3/C-8′ and C-1/C-7′ bonds between the two units. Thus, the planar structure of 1 was determined.


image file: c8qo00070k-f1.tif
Fig. 1 Structures of compounds 1–5, 1A and 2A.

image file: c8qo00070k-f2.tif
Fig. 2 Key 1H–1H COSY and HMBC correlations of 1–4.

Asperone B (2) had very similar 1D NMR data to 1, and its HMBC and 1H–1H COSY correlations unambiguously confirmed that they had the same planar structure (Fig. 2). However, several carbons, like C-7′ (Δδ −2.7) and C-8′ (Δδ −1.2), had slightly different chemical shifts, indicating the distinct configuration of 2. However, the limited NOE correlations were insufficient to distinguish their relative configurations.

Analysis of the methyl derivatives 1A and 2A (Fig. 1) was a simple but effective solution for confirming the configurations. In the ROESY spectrum of 1A, the correlations of 5-OCH3/H3-9′, 5-OCH3/H3-10, 5-OCH3/H-8, and H3-10/H-7′ revealed that they were in the same orientation, which was assigned as the α-orientation, and this was consistent with the coupling constant between H-7 and H-8 (J = 9.2 Hz). The correlations of H3-11/H-7, H3-11/H-8′, and H-3/H-8′ established that H-3, H-7, H3-11, and H-8′ were in the opposite β-orientation. Since NOE correlations were also observed from H3-11 to H3-10′ and from H-8′ to H-11′/H-12′, their close spatial relationship was determined as shown in Fig. 3. The ROESY spectrum of 2A was noticeably different from that of 1A. The 5-OCH3 signal was correlated with H3-9′ and H-8, while NOE correlations were seen as H-7′/H-7 and H3-10/H3-11. Therefore, the orientation of H-3, H3-11, H-7′, and H-8′ in 2A was opposite to what was seen in 1A. The stereochemistry of 1A and 2A was finally confirmed by X-ray diffraction using Cu Kα radiation (Fig. 4 and 5). The Flack parameter −0.02(4) of 1A allowed unambiguous assignment of the absolute configurations of 1 and 1A as 2′R, 7′R, 8′R, 1R, 3R, 7S, 8R. However, since the imperfect single-crystal data of 2A gave a Flack parameter of 0.0(3), the calculated ECD spectrum (Fig. S27) of 2 was recorded to confidently assign the configuration of 2′R, 7′S, 8′S, 1S, 3S, 7S, 8R for 2 and 2A.


image file: c8qo00070k-f3.tif
Fig. 3 Key ROESY correlations of 1A, 2A, 3 and 4.

image file: c8qo00070k-f4.tif
Fig. 4 X-ray crystallographic analysis of 1A.

image file: c8qo00070k-f5.tif
Fig. 5 X-ray crystallographic analysis of 2A.

Asperone C (3) was obtained as a white amorphous powder. Its molecular formula, C23H30O4, was determined by HR-ESI-MS data (m/z 393.2034 [M + Na]+, calcd for C23H30O4Na 393.2042) and indicated nine degrees of unsaturation. In its 1D NMR data (Table S3), the distinctive signals of a conjugated diene group strongly indicated the presence of a 7,8-dihydonivefuranone derivative that had been decarboxylated and reduced in 3. This deduction was fully supported by the 1H–1H COSY correlations of H-10′/H-11′/H-12′/H-13′/H2-14′/H3-15′ and H2-6′/H-7′/H3-8′ as well as the key HMBC correlations from H3-9′ to C-2′, C-3′, and C-10′ and from H2-4′ to C-2′, C-3′, C-5′, and C-6′. The remaining signals were assigned to be three methyl, a methine, two keto-carbonyl, a quaternary, and two olefinic carbons based on the HSQC spectrum. These groups formed the 2,4,5-trimethylcyclohex-4-ene-1,3-dione moiety since HMBC correlations were observed from H3-7 to C-3, C-4, and C-5, from H3-8 to C-1, C-2, and C-3, from H3-9 to C-1, C-2, and C-6, and from H-6 to C-4 and C-5. Moreover, the important HMBC correlations from H3-7 to C-7′, from H2-4′ to C-6, from H2-6′ to C-6, and from H-6 to C-4′, C-5′, and C-6′ unambiguously revealed the C-7′/C-4 and C-5′/C-6 bonds between the two units, forming the novel spiro[bicyclo[3.3.1]nonane-2,2′-furan]-4′-one skeleton. Hence, the planar structure of 3 was established (Fig. 2).

The bicyclo[3.3.1]nonane ring system of 3 required the bridgehead-substituted H-6 and H3-7 to be in the same orientation, which was arbitrarily assigned as the α-configuration. Since the upward H3-8 had an NOE (Fig. 3) with H3-8′ rather than H-7′, the β-orientation of H3-8′ was confirmed. The same β-configuration was also assigned for H-6′a (δH 1.64) due to its ROESY correlations with H3-8 and H3-9. This determination was consistent with the high coupling constant (J6′a/7′ = 13.7 Hz) caused by the large dihedral angle between H-7′ and H-6′a. Furthermore, the correlation of H3-9 with H2-4′ indicated that C-4′ was β-oriented, and the C-5′/O-1′ bond was in the opposite orientation. The NOE correlations from H-4′b to H-6 and H3-9′ revealed that they were on the same side of the dihydrofuranone ring. H-4′a had an NOE with H2-6′, while H-10′/H-11′ were correlated with H-6′/H-7′. These correlations indicated that these protons were on the other face of the dihydrofuranone ring. Therefore, the relative configuration of 3 was established as shown in Fig. 3. The absolute configuration was determined by TDDFT calculations of the ECD spectra (Fig. 7), which confirmed the 4R, 6S, 2′R, 5′S, 7′R configuration.

The similarities between the 1D NMR data of 4 and 3 indicated that they shared a planar structure, and this was further confirmed by comprehensive analysis of their 2D NMR data (Fig. 2 and Table S4). However, slight differences in some chemical shifts, especially those of C-4′, C-5′ and C-6′, suggested the opposite configuration of spiro C-5′ in 4. Direct evidence was found in the ROESY spectrum, which showed that H-10′ and H-11′ were near H-6 and H3-9 based on their mutual NOE correlations (Fig. 3). The ROESY correlations of H3-9′/H-6′a, H-6/H-4′a, H-4′b/H-7′, H-4′b/H-6′a, H3-9/H-6′b, H3-8/H-6′b, and H3-8/H3-8′ determined their spatial proximity, namely, C-4, C-6, C-7′, and C-2′ were in the same configurations as were seen in 3. The configuration was further confirmed by the calculated ECD spectra, in which the calculated spectrum of the 4R, 6S, 2′R, 5′R, 7′R configuration was a good match with the experimental curve (Fig. 7). Thus, 4 was determined to be the C-5′ spiro-epimer of 3.


image file: c8qo00070k-f6.tif
Fig. 6 Key 1H–1H COSY, HMBC, and key ROESY correlations of 5.

image file: c8qo00070k-f7.tif
Fig. 7 Calculated and experimental ECD spectra of 3 and 4.

Asperone E (5), obtained as a colorless oil, was established to have a molecular formula C23H32O4 by its HR-ESI-MS data, requiring eight degrees of unsaturation. By comparing their molecular formula and NMR data, 5 was found to closely resemble 4. However, the keto-carbonyl C-5 (δC 206.3) in 4 was reduced to an oxygenated methine (δH 3.96, brs; δC 76.0) in 5. This assignment was confirmed by the HMBC correlations (Fig. 6) since the methine at δH 3.96 was correlated with C-1, C-3, C-4, C-6, C-5′, C-7′, and C-7 and the carbon at δC 76.0 was correlated with H3-7, H-6, and H-7′. Additionally, H-5 was in the same orientation as H3-8/H3-9 due to the observed ROESY correlations between H-5 and H3-8/H3-9. Comprehensive 2D NMR data further confirmed its planar structure and the relative configurations as drawn (Fig. 6). Finally, the absolute configuration of 5 was determined to be 4R, 6R, 2′R, 5′R, 7′R, 5R by comparing the calculated and experimental ECD spectra (Fig. S61).

Plausible biosynthetic pathways for 1–5 were postulated in which the known polyketides 6[thin space (1/6-em)]7, 7[thin space (1/6-em)]10 and 8[thin space (1/6-em)]11a were presumed to be precursors (Scheme 1). For compounds 1 and 2, the oxidation11b of 6 led to intermediate 9, and precursor 8 underwent oxidation and deprotonation to give enolate anion 14. Then, 14 attacked 9 to construct the [3 + 2] cycloadducts 1 and 2 through intermediates 12 and 13, respectively.12a Regarding compounds 3–5, one of their monomers (10) was generated from the further decarbonylation of 9, and the other was enolate anion 15 produced from the deprotonation of 7. A subsequent [3 + 3] cycloaddition12b occurred between 10 and 15 to construct the novel spiro[bicyclo[3.3.1]nonane-2,2′-furan]-4′-one structure of 3 and 4 by forming intermediate 16. Additionally, further reduction at the C-5 carbonyl of 4 produced its derivative 5.


image file: c8qo00070k-s1.tif
Scheme 1 Plausible biosynthetic pathways of 1–5.

Compounds 1–5, 1A and 2A were tested for their inhibitory effects on LPS-induced NO production along with their cytotoxicities in RAW 264.7 macrophage cells. No obvious cytotoxicities were found at 50 μM, and the inhibition results are shown in Table 1. Compound 4 showed the most significant NO inhibitory activity even compared to previously isolated compounds. All compounds were also assessed for their antibacterial, antifungal, and antitumor activities. But only 2 showed weak antibacterial activity toward Pseudomonas aeruginosa (MIC50 = 71.6 μg mL−1); the positive control, streptomycin, showed an MIC50 value of 0.34 μg mL−1.

Table 1 Inhibitory effects of compounds 1–5, 1A and 2A on LPS-induced NO production in macrophage RAW 264.7 cells
Compound IC50[thin space (1/6-em)]a (μM) Compound IC50[thin space (1/6-em)]a (μM)
a Results are expressed as the mean IC ± SD values in (μM) from triplicate experiments.b Positive control.
1 >50 5 >50
2 16.0 ± 0.5 1A >50
3 13.2 ± 0.3 2A 42.9 ± 0.5
4 6.0 ± 0.5 L-NMMAb 41.9 ± 0.9


Conclusions

In summary, five dimeric polyketides (1–5) with two unprecedented skeletons were identified from the fungus Aspergillus sp. AWG 1–15. [3 + 2] and [3 + 3] cycloadditions were proposed as the crucial steps in their biogenetic pathway, through which furanone derivatives conjugated with two aromatic polyketides respectively to construct their novel skeletons. Among them, compounds 2–4 showed significant NO inhibitory activities on LPS-induced RAW 264.7 macrophage cells. These findings may provide a new template for chemists and biologists for the development of new anti-inflammatory drugs.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was funded by the National Natural Science Foundation of China (81503218 and 81602985), the Program for Changjiang Scholars and Innovative Research Team in University (IRT15R63), the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), and the Fundamental Research Funds for the Central Universities (2016ZZD010).

Notes and references

  1. (a) T. Rodrigues, D. Reker, P. Schneider and G. Schneider, Nat. Chem., 2016, 8, 531–541 CrossRef CAS PubMed; (b) L. Du, A. J. Robles, J. B. King, D. R. Powell, A. N. Miller, S. L. Mooberry and R. H. Cichewicz, Angew. Chem., Int. Ed., 2014, 53, 804–809 CrossRef CAS PubMed.
  2. (a) J. F. Sanchez, A. D. Somoza, N. P. Keller and C. C. C. Wang, Nat. Prod. Rep., 2012, 29, 351–371 RSC; (b) F. P. Miao, X. D. Li, X. H. Liu, R. H. Cichewicz and N. Y. Ji, Mar. Drugs, 2012, 10, 131–139 CrossRef CAS PubMed; (c) L. Liu, L. Wang, L. Bao, J. W. Ren, B. B. Basnet, R. X. Liu, L. W. He, J. J. Han, W. B. Yin and H. W. Liu, Org. Lett., 2017, 19, 942–945 CrossRef CAS PubMed.
  3. (a) T. S. Lin, Y. M. Chiang and C. C. C. Wang, Org. Lett., 2016, 18, 1366–1369 CrossRef CAS PubMed; (b) Z. M. Liu, Y. Chen, S. H. Chen, Y. Y. Liu, Y. J. Lu, D. N. Chen, Y. C. Lin, X. S. Huang and Z. G. She, Org. Lett., 2016, 18, 1406–1409 CrossRef CAS PubMed.
  4. H. Kato, T. Nakahara, K. Sugimoto, K. Matsuo, I. Kagiyama, J. C. Frisvad, D. H. Sherman, R. M. Williams and S. Tsukamoto, Org. Lett., 2015, 17, 700–703 CrossRef CAS PubMed.
  5. (a) H. C. Zhu, C. M. Chen, Y. B. Xue, Q. Y. Tong, X. N. Li, X. T. Chen, J. P. Wang, G. M. Yao, Z. W. Luo and Y. H. Zhang, Angew. Chem., Int. Ed., 2015, 54, 13374–13378 CrossRef CAS PubMed; (b) H. C. Zhu, C. M. Chen, Q. Y. Tong, X. N. Li, J. Yang, Y. B. Xue, Z. W. Luo, J. P. Wang, G. M. Yao and Y. H. Zhang, Angew. Chem., Int. Ed., 2016, 55, 3486–3490 CrossRef CAS PubMed; (c) G. Z. Wei, C. M. Chen, Q. Y. Tong, J. F. Huang, W. J. Wang, Z. D. Wu, J. Yang, J. J. Liu, Y. B. Xue, Z. W. Luo, J. P. Wang, H. C. Zhu and Y. H. Zhang, Org. Lett., 2017, 19, 4399–4402 CrossRef CAS PubMed.
  6. (a) C. X. Qi, J. Bao, J. P. Wang, H. C. Zhu, Y. B. Xue, X. C. Wang, H. Li, W. G. Sun, W. X. Gao, Y. J. Lai, J. G. Chen and Y. H. Zhang, Chem. Sci., 2016, 7, 6563–6572 RSC; (b) Y. L. Zhang, H. M. Ge, W. Zhao, H. Dong, Q. Xu, S. H. Li, J. Li, J. Zhang, Y. C. Song and R. X. Tan, Angew. Chem., Int. Ed., 2008, 120, 5907–5910 CrossRef; (c) W. J. Zhang, C. F. Yang, C. S. Huang, L. P. Zhang, H. B. Zhang, Q. B. Zhang, C. S. Yuan, Y. G. Zhu and C. S. Zhang, Org. Lett., 2017, 19, 592–595 CrossRef CAS PubMed.
  7. G. P. Yin, Y. R. Wu, M. H. Yang, T. X. Li, X. B. Wang, M. M. Zhou, J. L. Lei and L. Y. Kong, Org. Lett., 2017, 19, 4058–4061 CrossRef CAS PubMed.
  8. Y. Shiono, T. Hatakeyama, T. Murayama and T. Koseki, Nat. Prod. Commun., 2012, 7, 1065–1068 CAS.
  9. C. M. C. Dos Santos, G. L. Da Costa and J. D. Figueroa-Villar, Nat. Prod. Res., 2012, 26, 2316–2322 CrossRef PubMed.
  10. Z. Y. Lu, Z. J. Lin, W. L. Wang, L. Du, T. J. Zhu, Y. C. Fang, Q. Q. Gu and W. M. Zhu, J. Nat. Prod., 2008, 71, 543–546 CrossRef CAS PubMed.
  11. (a) Y. He and R. J. Cox, Chem. Sci., 2016, 7, 2119–2127 RSC; (b) T. D. H. Bugg and C. J. Winfield, Nat. Prod. Rep., 1998, 15, 513–530 RSC.
  12. (a) K. Okamoto, E. Tamura and K. Ohe, Angew. Chem., Int. Ed., 2014, 53, 1–6 CrossRef; (b) C. L. Cao, X. L. Sun, Y. B. Kang and Y. Tang, Org. Lett., 2007, 9, 4151–4154 CrossRef CAS PubMed.

Footnote

Electronic supplementary information (ESI) available. CCDC 1573433 and 1548345. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c8qo00070k

This journal is © the Partner Organisations 2018