Aspergillus terreus, a prolific filamentous fungus, synthesizes an array of secondary metabolites with distinct biological functions. Lovastatin, a prominent metabolite, exhibits a unique hexahydronaphthalene ring and a 2-methylbutyric acid moiety, efficaciously inhibiting HMG-CoA reductase for cholesterol regulation (Huang et al., 2021). Itaconic acid, another metabolite, serves as an industrial precursor and exhibits potential anti-inflammatory properties (Diankristanti and Ng, 2023). Terrein, with its 4H-furan-3-one framework, demonstrates multifaceted biological activities, including antitumor and anti-inflammatory effects, and inhibits melanin synthesis, suggesting therapeutic applications in hyper-pigmentation (Zhao et al., 2016). The dihydroxyphenylalanine (DOPA) pigments, geodin and epi-geodin, are noted for their antibiotic activity and metal ion chelation capabilities (Boruta et al., 2021). This spectrum of metabolites, each with unique structural and pharmacological properties, underscores the importance of Aspergillus terreus in pharmaceutical and industrial applications. Ongoing research in this domain is crucial for advancing natural product chemistry and drug discovery, underscoring the significant role of fungi in producing bioactive compounds.

In exploring marine-derived bioactive compounds, a cytotoxic dihydrobenzofuran was extracted from the deep-sea fungus Aspergillus terreus CC-S06-18 (Wang et al., 2020). Following this, extensive large-scale fermentation enabled the identification and characterization of eighteen secondary metabolites. This included three novel dimeric nitrophenyl trans-epoxyamides (1–3) and an additional fifteen compounds (4–18) (Figure 1). Subsequently, each of these compounds underwent evaluation to determine their cytotoxic and antibacterial activities.

Compound 1, a white powder, was analyzed using HRESIMS and demonstrated a sodium adduct ion peak with m/z of 607.2017 [M + Na]⁺. This peak suggested the molecular formula of C₂₈H₃₂N₄O₁₀, indicative of 15 degrees of unsaturation. Examination of the ¹³C NMR and DEPT spectra revealed 14 carbon atoms: comprising two methyl groups (δ_C 19.2, C-13 and δ_C 15.3, C-14), a single methoxyl group (δ_C 54.9, C-15), three methines linked to oxygen or nitrogen (δ_C 86.4, C-1; δ_C 47.6, C-2; δ_C 62.7, C-6), an oxygen-bearing quaternary carbon (δ_C 65.0, C-5), four aromatic methines (δ_C 124.3, C-9/11 and δ_C 128.8, C-8/12), two quaternary aromatic carbons (δ_C 142.7, C-7 and δ_C 148.9, C-10), and a carbonyl carbon atom (δ_C 172.3, C-4) (Table 1). The ¹H NMR data supplemented these observations, showing signals for three methyl groups (δ_H 3.43, OCH₃-1; δ_H 1.31, H-13; δ_H 1.39, H-14), three methines related to oxygen or nitrogen (δ_H 4.46, H-6; δ_H 5.63, H-1; δ_H 4.66, H-2), along with four aromatic methines (δ_H 7.68, d, H-8/12; δ_H 8.33, d, J = 8.6 Hz, H-9/11) (Table 1). Considering the NMR data and its molecular formula, compound 1 was theorized to be a symmetrically structured dimer.

The planar configuration of compound 1 was established through a comprehensive analysis using both 1D and 2D NMR spectroscopy. The existence of a p-nitrophenyl group was initially inferred from HMBC spectral correlations, including signals from H-9/H-11 to C-7 (δ_C 142.7) and C-10 (δ_C 148.9), and from H-8/H-12 to C-10. These findings were further corroborated by COSY spectral correlations, specifically between H-9 and H-8, and H-11 and H-12 (Figure 2). Further, the structure was identified to contain a 2-methyl-2,3-epoxyamide group, evidenced by HMBC correlations linking H-6 to C-4 and C-5, and H₃-14 to C-4, C-5, and C-6, with the chemical shifts of C-5 (δ_C 65.0) and C-6 (δ_C 62.7) confirming the presence of an epoxy group (Figure 2). The isopropyl segment in the structure was pinpointed through COSY correlations from the methyl group H₃-13 to H-1 and H-2, and its attachment to N-3 was established by HMBC correlations from H-2 to C-4 (Figure 2). As compound 1 was presumed to be a dimer, the molecular formula of its monomer was deduced to be C₁₄H₁₆N₂O₅. This monomer comprised a 2-methyl-2,3-epoxyamide group, an isopropyl group, and a 1,4-disubstituted phenyl, along with a methoxy group attached to the C-1 position. Considering the remaining atomic composition of one nitrogen and two oxygens in the monomer, it was concluded that a nitro group was attached at the C-10 position. All these data led to the confirmation of subunit A in the structure, as shown in Figure 2. Additionally, the HMBC correlations between H-1 and C-4′, coupled with the chemical shifts of C-1/1′ (δ_C 86.4) (Figure 2), suggested that two subunit A units are linked via two C-N bonds, forming a dimer with a central symmetrical framework, as represented in Figure 2.

Key NOESY correlations, specifically from H-1,1′ to H₃-13/13′, as well as between H-2,2′ and H₃-15,15′, demonstrated that H-1/1′ and H₃-13/13′ resided on the same piperazine ring, suggesting an α-orientation. Further NOESY correlations between H-8,8′ and H₃-14,14′ suggested that H-6,6′ and H₃-14,14′ occupy different faces of the oxirane ring (Figure 2). This led to the deduction of two possible relative configurations of (1S*,1’S*,2S*,2’S*,5S*,5’S*,6R*,6’R*)-1 (1a) and (1S*,1’S*,2S*,2’S*,5R*,5’R*,6S*,6’S*)-1 (1b). A conformational search was conducted for these configurations, followed by geometry optimization. The lowest energy conformers identified were subjected to NMR calculations at the mPW1PW91-SCRF/6–31 + G(d,p) level using GIAO method. 1a had a DP4+ possibility of 100%, with its calculated ¹³C NMR data closely aligned with the experimental results, exhibiting R², MAE, and CMAE values of 0.9990, 1.8, and 1.2, respectively (Figure 3A). The absolute configuration was definitely identified as 1S,1’S,2S,2’S,5S,5’S,6R,6’R through comparative analysis of the experimental ECD spectrum and the CAM-B3LYP/6-311G(d) calculated spectra for both stereoisomers in MeOH. This identification was affirmed as the calculated ECD for (1S,1’S,2S,2’S,5S,5’S,6R,6’R) closely matched the experimental spectrum (Figure 3B).

Compound 2 has been established to possess the molecular formula C₂₆H₂₈N₄O₁₀, which was derived from HRESIMS data exhibiting a deprotonated ion peak at a m/z of 555.1711 [M − H]⁻. This indicated a reduction of 28 atomic mass units (amu) compared to compound 1, along with the presence of 15 degrees of unsaturation. A comparative analysis of NMR data between compounds 1 and 2 revealed that the primary variations are concentrated in the piperazine ring. Notably, the chemical shifts at C-1 and C-2 altered from δ_C 86.4 to δ_C 91.3, and from δ_C 47.6 to δ_C 58.9, respectively. The observation of 13 unique carbon resonances in the ¹³C NMR spectrum, in conjunction with its molecular weight, indicates that compound 2 shared the same symmetrical framework as compound 1, as detailed in Table 1. In the NOESY spectrum, correlations between H-1,1′ and H-2,2′ in compound 2 suggested that H₃-13/13′ and H-1/1′ were positioned on opposite sides of the piperazine ring. Furthermore, NOESY correlations between H-8,8′ and H₃-14,14′ implied that H-6,6′ and H₃-14,14′ were situated on different faces of the oxirane ring, paralleling the pattern observed in compound 1, as depicted in Figure 2. Considering the biosynthetic pathway, compound 2 is likely the 1,1′-epimer derivative of compound 1, characterized by a hydroxyl group attached to both C-1′ and C-1. Finally, the absolute configuration of compound 2 is designated as 1R,1’R,2S,2’S,5S,5’S,6R,6’R by ECD calculation at CAM-B3LYP/6-311G(d) (Figure 3B).

Compound 3, isolated as a white powder, had a molecular formula of C₂₇H₃₀N₄O₁₀. This was established from the HRESIMS data, which showed a sodium adduct ion peak at m/z 593.1859 [M + Na]⁺, indicating a reduction of 14 amu from 1 and 15 degrees of unsaturation. A comparison of the ¹H NMR and ¹³C NMR data revealed similarities between 1 and 3. The most significant difference was the substitution of the nitrogenated methine group (δ_C 86.4, C-1) in 1 with a formyl group (δ_C 198.3) in 3, as detailed in Table 1. This alteration was corroborated by COSY correlations of H₃-13/H-2/H-1 and HMBC correlations of H-1/C-13, H-2/C-1′, and H-2/C-4, as illustrated in Figure 2. The coupling constant ³J_{H1’ − H2} of 9.2 Hz indicated that H-1′ and H-2′ were positioned on opposite faces of the octatomic ring, which was formed by a hydrogen bond between the formyl and the amide group. In the NOESY spectrum, correlations between H₃-13/H-1′, H₃-13’/H-1′, and H-2/H-2′ indicated that H₃-13, H-1′, and H₃-13′ are located on the same side of the octatomic ring. Similarly, the positioning of H-6/6′ and H₃-14/14′ on opposite faces of the oxirane ring was evidenced by NOESY correlations between H-8/8′ and H₃-14/14′. Considering the biosynthetic pathway, the absolute configurations of C-2 and C-2′ were supposed to be S. Compound 3 was assigned the absolute configuration as 2S,5S,6R,1’S,2’S,5’S,6’R, which was corroborated by ECD calculations at the CAM-B3LYP/6-311G(d) level, as illustrated in Figure 3B.

The biosynthetic pathway, as proposed in Figure 4, began with L-Phenylalanine undergoing a sequence of transformations: deamination, oxidation, methylation, and aryl nitration, leading to the formation of intermediate a. Acylation of intermediate a with L-Alanine resulted in intermediate b. Two units of b underwent a condensation reaction by removing two molecules of water to give intermediate c, which was reduced to yield 2. Intermediate c underwent reduction and methylation to yield 1. Similarly, two units of b condensed by eliminating one molecule of water, followed by reduction and methylation steps, resulting in 3 (Figure 4).

A comparison of NMR data with previously reported data in the literature led to the identification of the remaining fifteen known compounds, numbered 4–18. They were questin (4) (Liu et al., 2015), ω-hydroxyemodin (5) (Khamthong et al., 2012), penicillixanthone (6) (Rukachaisirikul et al., 2014), isorhodoptilometrin (7) (Kim et al., 2022), emodin (8) (Khaliq et al., 2023), citrinin H1 (9) (Ngan et al., 2017), 3,5-dichloroasterric acid (10) (Lee et al., 2002), dihydrogeodin (11) (Hamed et al., 2019), dimethoxyphtalide (12) (Wei et al., 2020), anodendroic acid (13) (Prompanya et al., 2016), varioloid B (14) (Zhang et al., 2015), italicum (15) (Tomás-Barberán et al., 1990), 46-dichloro-5-methylbenzene-13-diol (16) (Slavov et al., 2010), 4-hydroxybenzaldehyde (17) (Hsu et al., 2009), and 2-phenylethanol (18) (Pehlivan et al., 2011).

We performed a CCK8 assay to investigate the effects of all compounds on the viability of two gastric cancer cell lines, MKN28 and MGC803. 1 exhibited a moderate inhibitory effect on the growth of MKN28 cells, with an IC₅₀ value of 7.4 μM. Similarly, 7 displayed a moderate inhibitory effect on MGC803 cells, with an IC₅₀ value of 2.5 μM. The remaining compounds did not achieve a 50% inhibition rate within the tested concentration range, indicating that their IC₅₀ values exceeded 10 μM. Paclitaxel served as the positive control, with IC₅₀ values of 0.32 and 0.7 nM for MKN28 and MGC803, respectively.

The in vitro antibacterial activities of compounds 1–18 were tested against six bacterial strains. Among them, Compound 16 showed potency against Vibrio parahaemolyticus ATCC 17802, exhibiting a MIC of 7.8 μg/mL. This potency was compared to the positive control, chloramphenicol, which had an MIC of 1.1 μg/mL.

This study details the characterization of three new compounds (1–3) and fifteen known compounds (4–18) from the deep-sea fungus Aspergillus terreus. Compound 1 was identified as a symmetrical dimer with a complex structure, including a p-nitrophenyl moiety, a 2-methyl-2,3-epoxyamide group, and an isopropyl fragment. Its absolute configuration was determined using DFT methods and ECD spectrum analysis. 2, structurally similar to 1 but differing in the piperazine ring, is considered an epimer of 1. Compound 3 shares a similar structure with 1, but with a formyl group replacing a nitrogenated methine group. The biogenetic pathways of these compounds are proposed, involving precursors like L-phenylalanine and L-alanine. The study also reports the biological activities of these compounds. Specifically, compounds 1 and 7 exhibit moderate growth inhibitory effects on gastric cancer cell lines, while compound 16 shows moderate antibacterial properties. The study emphasizes the need for extensive research on the biological activities and potential therapeutic applications of these compounds, particularly those showing promise in preliminary tests. Developing synthetic methods for these compounds could facilitate further studies and potential pharmaceutical applications. Investigating the mechanisms behind their observed biological activities could lead to the discovery of new drug targets or therapeutic strategies.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/Supplementary material.

Ethical approval was not required for the studies on humans in accordance with the local legislation and institutional requirements because only commercially available established cell lines were used. Ethical approval was not required for the studies on animals in accordance with the local legislation and institutional requirements because only commercially available established cell lines were used.

XH: Conceptualization, Formal analysis, Funding acquisition, Writing – original draft, Writing – review & editing, Supervision. YW: Formal analysis, Writing – review & editing, Methodology. GL: Writing – review & editing, Resources. ZS: Writing – review & editing, Project administration, Investigation. JX: Data curation, Methodology, Supervision, Validation, Writing – review & editing. J-JQ: Formal analysis, Writing – review & editing. WW: Conceptualization, Formal analysis, Funding acquisition, Project administration, Writing – original draft, Writing – review & editing.