An epipolythiodioxopiperazine alkaloid and diversified aromatic polyketides with cytotoxicity from the Beibu Gulf coral-derived fungus Emericella nidulans GXIMD 02509

Marine microorganisms, especially marine fungi, have historically proven their value as a prolific source for structurally novel and pharmacologically active secondary metabolites (Deshmukh et al., 2018; Carroll et al., 2022). The corals constitute a dominant part of reefs with the highest biodiversity, and harbor highly diverse and abundant microbial symbionts in their tissue, skeleton, and mucus layer, with species-specific core members that are spatially partitioned across coral microhabitats (Wang WQ et al., 2022). The coral-associated fungi were very recently found to be vital producers of structurally diverse compounds, terpenes, alkaloids, peptides, aromatics, lactones, and steroids. They demonstrate a wide range of bioactivity such as anticancer, antimicrobial, and antifouling activity (Chen et al., 2022). The genetically powerful genus Emericella (Ascomycota), which has marine and terrestrial sources, includes over 30 species and is distributed worldwide. It is considered a rich source of diverse secondary metabolites with antimicrobial activity or cytotoxicity (Alburae et al., 2020). Notably, Emericella nidulans, the sexual state of a classic biosynthetic strain Aspergillus nidulans, was recently reported as an important source of highly methylated polyketides (Li et al., 2019) and isoindolone-containing meroterpenoids (Zhou et al., 2016) with unusual skeletons.

The Beibu Gulf is a semi-enclosed gulf in the northwest of the South China Sea, and harbors tremendous underexplored biodiversity in terms of both marine organisms and microorganisms; these are rich in diversified secondary metabolites (Huang et al., 2022). In continuation of our efforts to discover interesting lead compounds from Beibu Gulf coral-derived marine fungi, a plethora of structurally novel secondary metabolites with remarkable biological activity have been recently obtained, including anti-tumor ascochlorins (Guo et al., 2021; Luo et al., 2021) and cytochalasans (Luo et al., 2020), anti-osteoclastogenic chlorinated polyketides (Zhang et al., 2022), phenolic derivatives (Lu et al., 2022), and cyclopiazonic acid alkaloids (Wang JM et al., 2022). In this study, a fungus Emericella nidulans GXIMD 02509 endemic to Weizhou coral reefs attracted our attention owing to the intriguing high-performance liquid chromatography (HPLC)-ultraviolet (UV) profiles of its extract. Subsequent chemical investigation led to the isolation of nine diverse aromatic polyketides, an epipolythiodioxopiperazine alkaloid, and a farnesylated phthalide derivative (Fig. 1). Several of these compounds showed cytotoxicity against three human cancer-cell lines (786-O, SW1990, and SW480). Here, the process of isolation and structural determination, as well as the cytotoxicity results, are described in detail.

Fig. 1. Chemical structures of compounds 1‒11.

Compound 1 was isolated as a bright-yellow solid and was deduced with the molecular formula C₂₁H₂₂O₆ based on the high resolution-electrospray ionization-mass spectroscopy (HR-ESI-MS) data [M+H]⁺ ion peak at m/z 371.1499 (calcd for C₂₁H₂₃O₆, 371.1495). The UV spectrum revealed the presence of benzene chromophores with absorption bands at 203, 255, and 320 nm. The ¹H nuclear magnetic resonance (NMR) (Table 1) and heteronuclear single quantum coherence (HSQC) data for 1 displayed a series of signals, including: two hydroxyl groups attributable to 1-OH (δ _H 10.93, s) and 7-OH (δ _H 11.38, s); one aldehyde group, H-11 (δ _H 9.98, s); four aromatic or olefinic protons, H-3 (δ _H 7.12, d, J=8.5 Hz), H-4 (δ _H 6.12, d, J=8.5 Hz), H-5 (δ _H 7.16, s), and H-2' (δ _H 5.25, t, J=7.0 Hz); one methylene, H-1' (δ _H 3.28, d, J=7.0 Hz); three singlet methyls, 6-Me (δ _H 2.24, s), H₃-4' (δ _H 1.72, s), and H₃-5' (δ _H 2.24, s); and one methoxyl, 4a-OMe (δ _H 4.03, s). Aside from these ten corresponding hydrogen-bearing carbons, ten aromatic or olefinic (four oxygenated) carbons and a carbonyl (δ _C 198.3) remained in the ¹³C NMR spectrum.

Table 1.

¹H (500 MHz) and ¹³C (125 MHz) NMR spectroscopic data for 4a-O-methoxyarugosin H (1) (CDCl₃)

Position	δ C, type	δ H (J (Hz))	HMBC
1	160.4, C
2	125.2, C
3	134.9, CH	7.12, d (8.5)	1, 1'
4	105.5, CH	6.12, d (8.5)	2, 4a, 9a
4a	170.5, C
4a-OMe	53.2, CH3	4.03, s	4a
5	128.0, CH	7.16, s	6, 6-Me, 7, 10a
6	139.8, C
6-Me	15.2, CH3	2.24, s	5, 6, 7
7	154.8, C
8	113.7, C
8a	125.3, C
9	198.3, C
9a	103.5, C
10a	140.2, C
11	194.1, CH	9.98, s	7, 8, 8a
1'	27.8, CH2	3.28, d (7.0)	1, 2, 3, 2', 3'
2'	121.5, CH	5.25, t (7.0)	2, 1', 4', 5'
3'	133.7, C
4'	25.9, CH3	1.72, s	2', 3', 5'
5'	17.9, CH3	2.24, s	2', 3', 4'
1-OH		10.93, s	1, 2, 9a
7-OH		11.38, s	6, 7, 8

NMR: nuclear magnetic resonance; HMBC: heteronuclear multiple bond correlation.

This information revealed that structurally, 1 was closely related to arugosin H, which was also obtained from the marine-derived fungus Emericella nidulans var. acristata (Kraljet al., 2006). The main difference was the appearance of a methoxyl group (δ _H/C 4.03/53.2) at C-4a (δ _C 170.5) in 1 instead of the hydroxyl group that appears in arugosin H. This deduction was verified by the heteronuclear multiple bond correlation (HMBC) correlation from 4a-OCH ₃ to C-4a (Fig. 2). Based on these findings, we determined that the structure of 1 was a methyl derivative of arugosin H, and accordingly assigned it a trival name: 4a-O-methoxyarugosin H (Figs. S1‍–‍S9). Compound 1 was probably an artifact produced during the isolation procedure when methanol was used as the main solvent (Capon, 2020).

Fig. 2. Key ¹H-¹H COSY, HMBC, and NOESY correlations of 4a-O-methoxyarugosin H (1). COSY: correlation spectroscopy; HMBC: heteronuclear multiple bond correlation; NOESY: nuclear overhauser effect spectroscopy.

We were able to pinpoint known compounds by comparing the physicochemical data of known compounds 2‍–‍11 (supplementary information) with data from the literature. We identified pre-shamixanthone (2) (Wu et al., 2015a), cycloisoemericellin (3) (Kawahara et al., 1988), sterigmatocystin (4) (Zhu and Lin, 2007), dihydrosterigmatocystin (5) (Zhu and Lin, 2007), dehydromicroperfuranone (6) (Kralj et al., 2006; Roux et al., 2020), varioxiranol I (7) (Wu et al., 2015b), arugosin G (8) (Kralj et al., 2006), arugosin C (9) (Kawahara et al., 1988; El-Kashef et al., 2021), emestrin J (10) (Li et al., 2016), and farnesylemefuranone D (11) (Chi et al., 2020). Interestingly, emestrin J (10) harbors with an uncommon disulfide moiety, which was biosynthesized by a peptide cyclization pathway along with additional ring-expansion and macrocyclization (Li et al., 2016).

During the course of our search for anti-tumor lead compounds from marine natural products (Zhouet al., 2019; Luo et al., 2021), all obtained compounds were evaluated for cytotoxicity against three human cancer cell lines, i.e., 786-O (human renal carcinoma cell), SW1990 (human pancreatic cancer cell), and SW480 (human colorectal cancer cell), along with the normal human liver cell line LO2 (Table 2). Among them, compounds 1‍-‍5, 7, and 10 showed inhibitory activity against these cell lines, with the half maximal inhibitory concentration (IC₅₀) values ranging from 4.3 to 33.4 μmol/L. Notably, emestrin J (10) exhibited the strongest activity against these cancer cell lines, especially for 786-O (4.3 μmol/L), and was at least as potent as the positive control, cisplatin. Interestingly, two xanthone derivatives (4 and 5) displayed antiproliferative activity, with IC₅₀ values of 18.3 and 24.7 μmol/L against 786-O, and 19.6 and >40 μmol/L against SW1990 cells, respectively. This revealed that the Δ ¹⁶ double bond in 4 probably promoted cytotoxic activity.

Table 2.

Cytotoxicity of compounds 1-11

Compound	IC50 (μmol/L)
	786-O	SW1990	SW480	LO2
1	–	–	32.6±1.2	–
2	–	33.2±5.1	–	–
3	–	24.7±2.2	30.2±4.8	–
4	18.3±1.9	19.6±2.8	–	–
5	24.7±1.8	–	–	–
6	–	–	–	–
7	33.4±1.8	30.4±5.1	27.2±6.5	–
8	–	–	–	–
9	–	–	–	–
10	4.3±0.7	14.1±1.8	4.9±0.3	11.1±2.4
11	–	–	–	–
Cisplatin	2.5±0.6	33.8±4.7	26.7±4.6	5.2±0.6

All data shown above are mean±SD of three independent experiments. IC₅₀: half maximal inhibitory concentration; SD: standard deviation. "‒": >40 μmol/L.

To further evaluate the potential anti-tumor activity of 10, we next investigated its activity against 786-O cells in cell colony and scratch wound assays. Consistent with the above-mentioned antiproliferative activity, compound 10 significantly reduced cell colony formation of 786-O cells at concentrations of 0.5 and 1.0 μmol/L (Figs. 3a and 3b). Also, compared with the vehicle group, compound 10 significantly suppressed migration of 786-O cells in a time- and dose‐dependent manner (Figs. 3c and 3d). To explore whether the antiproliferative activity of 10 was related to apoptosis, we further evaluated the compound for its effect on cell apoptosis and cell-cycle arrest in 786-O cells. The results were analyzed by flow cytometry and showed that the total apoptotic cells (early and late) induced by 10 at 24 h rose by 21.4% (2 μmol/L) and 29.5% (4 μmol/L), suggesting that 10 induced 786-O cell apoptosis in a dose-dependent manner (Figs. 4a and 4b). Meanwhile, the cell-cycle distribution results revealed that 10 primarily blocked the cell cycle during the G2/M phase, resulting in an inability of cells to proliferate (Figs. 4c and 4d). Therefore, it was clear that 10 could suppress the proliferation, colony formation, and migration of 786-O cells, and induce apoptosis, acting as a potential anti-tumor compound.

Fig. 3. Suppressive effects of emestrin J (10) on colony formation and migration of 786-O cellsin vitro. Representative wells (a) and quantitative results (b) of the colony-formation assay. Representative images (c) and quantitative results (d) of the scratch wound assay. 786-O cells were treated with vehicle (DMSO, Cont) or 10, as indicated. All data shown above are mean±SD of three independent experiments. ^** P<0.01, ^*** P<0.001 vs. Cont. DMSO: dimethylsulfoxide; Cont: control; SD: standard deviation.

Fig. 4. Effects of emestrin J (10) on cell apoptosis and cell-cycle arrest in 786-O cells. Emestrin J (10) induced apoptosis of 786-O cells (a, b) and arrested the cell cycle (c, d) in the G2/M phase. 786-O cells were treated with vehicle (DMSO, Cont) or 10 (2 and 4 μmol/L) for 24 h, as indicated. All data shown above are mean±SD of three independent experiments. ^** P<0.01, ^*** P<0.001 vs. Cont. DMSO: dimethylsulfoxide; Cont: control; SD: standard deviation; FITC: fluorescein isothiocyanate; PI: propidium iodide; PI-A: PI-area.

In conclusion, nine aromatic polyketides, including a new one, 4a-O-methoxyarugosin H (1), along with an epipolythiodioxopiperazine alkaloid, emestrin J (10), and a farnesylated phthalide derivative, were obtained from the Beibu Gulf coral-associated fungus Emericella nidulans GXIMD 02509. We determined their structures by spectral data analysis, as well as comparison with reported data. Several of the compounds exhibit cytotoxicity against three human cancer cell lines, 786-O, SW1990, and SW480. The most potent one, emestrin J (10), has an uncommon disulfide bond and suppresses proliferation, colony formation, and migration of 786-O cells, as well as inducing apoptosis. Our findings provide a basis for further development and utilization of emestrin derivatives as sources of potential anti-tumor chemotherapy agents.

Materials and methods

Detailed methods are provided in the electronic supplementary materials of this paper.

Author contributions

Xiaowei LUO and Yonghong LIU conceived the research and designed experiments. Miaoping LIN, Zhenzhou TANG, Jiaxi WANG, Humu LU, Chenwei WANG, Yanting ZHANG, Xinming LIU, and Chenghai GAO performed the experiments and analysis. Xiaowei LUO, Miaoping LIN, and Zhenzhou TANG interpreted the data and wrote the paper. All authors have read and approved the final manuscript, and therefore, have full access to all the data in the study and take responsibility for the integrity and security of the data.

Compliance with ethics guidelines

Miaoping LIN, Zhenzhou TANG, Jiaxi WANG, Humu LU, Chenwei WANG, Yanting ZHANG, Xinming LIU, Chenghai GAO, Yonghong LIU, and Xiaowei LUO declare that they have no conflict of interest.

This article does not contain any studies with human or animal subjects performed by any of the authors.