The EtOAc extracts of Aspergillus flavus were analyzed using the LC − MS/MS molecular networking. The data were then screened on the Global Natural Product Social Molecular Networking (GNPS) library, with some sample-specific clusters identified (Fig. 1). Based on this observation, the specific fractions were traced through mass spectrometry for separation. This process led to the isolation of four new secondary metabolites (1–4) and three known analogues (5–7) (Fig. 2).

Molecular networks for secondary metabolites of Aspergillus flavus

Structures of compounds 1–7

Aspergiflone A (1), was obtained as a colorless oil. The molecular formula was deduced as C29H40O11 based on the HRESIMS ion peak at m/z 565.2638 ([M + H]+, calcd for 565.2643), indicating ten degrees of unsaturation. The 1H NMR data (Table S1) indicated the presence of ten methyl groups, and one olefinic proton. The 13C NMR data (Table S1) and HSQC spectrum revealed 29 carbon signals, assigned to ten methyl carbons (two oxygenated and eight sp3-hybridized), three methylene carbons (all sp3 hybridized), three methine carbons (one olefinic and two sp3-hybridized), and thirteen non-hydrogenated carbons (three olefinic, four six sp3-hybridized and six carbonyls). These data collectively suggest that compound 1 possesses a bicyclic framework.

Analysis of the 1D and 2D NMR data (Table S1 and Fig. 3) revealed that the structure of compound 1 is similar to that of compound 7, a known cyclohexenone previously isolated from the fungus Paecilomyces lilacinus [23]. The primary difference is that in compound 1, two cyclohexenone residues are connected at C-15 through the C − C single bond, forming a cyclohexenone dimer. This structural variation is evident by the presence of a methylene carbon (δC 31.4) at C-15, which connects rings A and B at C-6 (δC 115.6) and C-4' (δC 53.0), feature absent in compound 7. These differences are further supported by the HMBC correlations from H2−15 to C-1 (δC 191.9), C-5 (δC 168.5), C-6, C-4', C-3' (δC 204.7), and C-8' (δC 21.3).

Key HMBC, 1H-1H COSY and NOESY correlations of 1

The chemical shifts around the chiral centers at C-2, C-4, C-10, C-2', and C-10' in compound 1 were nearly identical to those of compound 7 (Table S5), for which the absolute configuration was determined as 2R4R8S by single-crystal X-ray diffraction analysis [Flack parameter −0.12(7)] (Fig. 4). This suggested that the relative configuration of C-2, C-4, C-10, C-2', and C-10' in 1 was same as that of 7. In the NOESY spectrum, correlations between H3−7'/H-15a (δH 3.27) suggested the α-orientation of H3−8' and the β-orientation of H3−7' (Fig. 3). To elucidate the configuration of the chiral centers within the bicyclic ring system of compound 1, ECD calculation and DP4+ probability analyses were performed. The ECD spectra of 2R4R10S2'R4'R10'S-1a, and 2R4R10S2'S4'S10'R -1b were calculated using the time-dependent density functional theory (TDDFT) method at the CAM-B3LYP/6-311G (d, p) level. The experimental ECD spectrum of compound 1 matched the calculated ECD spectrum for the 2R4R10S2'R4'R10'S (Fig. 5A). Furthermore, DP4+ probability analysis (100% for 2R4R10S2'R4'R10'S configurations, Fig. 5B, Table S6 and Fig. S2) based on the gauge-independent atomic orbital (GIAO) method at the PCM/mPW1PW91/6–31 + G (d, p) level supported this relative configuration assignment. Considering the biosynthetic pathway, compound 7 could undergo the methyltransferases and polymerization to form aspergiflone A (1) [31,32,33], further supporting the accuracy of the configuration assignment of 1 (In Sect. 5 of the Supporting Information).

Perspective ORTEP drawings of the X-ray structures of 7 (displacement ellipsoids are drawn at the 50% probability level)

A Experimental and calculated ECD spectra of 1 and 7 B Results of DP4+ analyses for 2R4R10S2'R4'R10'S-1a and 2R4R10S2'S4'S10'R-1b

Aspergiflone B (2), isolated as a colorless oil, was determined to have the molecular formula C13H19NO4 base on the HRESIMS ion peak at m/z 254.1387 ([M + H]+, calcd. 254.1387), indicating five degrees of unsaturation. The 1D and 2D NMR data (Table S2 and Fig. 6) of compound 2 closely resembled those of compound 5 [19], with the only difference being the absence of a methyl group at C-8 in 2, as supported by the 1H-1H COSY correlations of H-8/H-9/H-10. Finally, the absolute configuration of 2 was defined as 2R in the TDDFT/ECD calculations (Fig. 7).

Key HMBC, 1H-1H COSY, and NOESY correlations of 2–4

Experimental and calculated ECD spectra of 2–4

Aspergiflone C (3) was isolated as a colorless oil. The molecular formula C14H21NO4 was determined by the HRESIMS ion peak at m/z 268.1548 ([M + H]+, calcd. 268.1543), implying five degrees of unsaturation. The 1H NMR data (Table S3) indicated the presence of four methyls groups, one methoxyl group, and one olefinic proton. The 13C NMR data (Table S3) and HSQC spectrum revealed the presence of 14 carbons signals, which were classified as five methyl carbons (one oxygenated, one olefinic and three sp3-hybridized), one methylene carbons (one sp3-hybridized), two methine carbons (one olefinic and one sp3-hybridized), and six non-hydrogenated carbons (three olefinic, one sp3-hybridized and two carbonyls).

The HMBC correlations (Fig. 6) from H-8 to C-3 (δC 200.0), C-4 (δC 105.6), and C-5 (δC 169.7), as well as from H3−14 to C-5, indicated the presence of a carbonyl group at C-3 and a methoxy group at C-5, respectively. Combined with the HMBC correlations from H3−7 to C-1 (δC 150.1), C-2 (δC 74.7), and C-3, and from H-6 to C-1, C-5, a cyclohexenone skeleton was established. The presence of a 2-methylbutanoic acid group was inferred from the 1H–1H COSY correlations of H-12/H-11/H-10/H-13 (Fig. 6), together with the HMBC correlations from H3−13 to C-9 (δC 176.2), C-10 (δC 44.3), C-11 (δC 27.4). Based on the molecular formula, the presence of an amino group in compound 3 was confirmed. A literature survey revealed that the chemical shifts was ranging from δC 160.8 to δC 175.0 when an amino group located at a double bond [13, 19]. Accordingly, the highfield chemical shift of C-1 (δC 150.1) indicated that the amino group was located at C-2 (δC 74.7), rather than at C-1. Thus, the 2-methylbutyric acid group was determined to be attached to C-1 (δC 150.1), which was further supported by its molecular formula and the required degrees of unsaturation.

To elucidate the absolute configuration of C-10, a hydrolysis reaction was performed. Acid hydrolysis of compound 3 yielded 2-methylbutanoic acid, which exhibited a positive optical rotation (\([\alpha]^_}\) + 9.0, c 0.05, MeOH) (Fig. S5). By comparing the optical rotation of the hydrolysate with those reference compounds [14], (+) and (−)−2-methylbutanoic acids, the absolute configuration at C-10 is defined as 10S. The DP4+ probability analysis suggested that the 2R*10S*configuration of 3 was in good correlation with the experimental data (Table S7 and Fig. S3). Finally, the absolute configuration of 3 was defined as 2R10S in the TDDFT/ECD calculations (Fig. 7).

Aspergiflone D (4), was obtained as a colorless oil. The molecular formula was deduced as C13H18O6 based on the HRESIMS ion peak at m/z 271.1174 ([M + H]+, calcd. 271.1176), implying five degrees of unsaturation. The 1D and 2D NMR data (Table S4 and Fig. 6) revealed that compound 4 closely resembled compound 7 [23]. The only difference was that the absence of a methyl at C-10 in 4 compared to 7, as confirmed by the 1H-1H COSY correlations of H-10/H-11/H-12 (Fig. 6). The nearly identical NMR data for the chiral centers in 4 and 7 suggested that the relative configuration of 4 is the same as that of 7, which was further supported by DP4 + probability analysis (Table S7 and Fig. S3). The absolute configuration of 4 was identified as 2R4R, rather than 2S4S, by the TDDFT/ECD calculations (Fig. 7).

Foodborne pathogens are important source of food safety issues and have attracted increasing attention from governments and the public worldwide [34]. The antimicrobial activities of compounds 1−6 were evaluated against four foodborne pathogens, including Staphylococcus aureus, Escherichia coli, Proteus sp, and Shigella flexneri. Compounds 1−6 exhibited antibacterial activity against Staphylococcus aureus, Escherichia coli, and Proteus sp with MIC values ranging from 1.0 to 64 μg/mL (Table 1). Among them, compound 1 exhibited antibacterial activity against Escherichia coli equivalent to that of the positive control, ciprofloxacin.

To investigate the antibacterial mechanism of compound 1, scanning electron microscope (SEM) was utilized, with dimethyl sulfoxide (DMSO) serving as the blank control. The surface morphology of cell structure was observed and analyzed under a microscope. As shown in Fig. 8A, the cell membrane of Escherichia coli exhibited a neat arrangement, plump structure, intact shape, and smooth surface. However, following treatment with compound 1, the hyphal surface appeared contracted, rough, and irregularly intertwined, indicating that compound 1 can damage the ultrastructure of fungal hyphae.

A SEM images of Escherichia coli hyphae treated by DMSO and compound 1. Scale bars, 10 μm and 2 μm, respectively. B The outer membrane of Escherichia coli by measuring fluorescence intensity of NPN treated with increasing concentrations of compound 1. C The inner membrane of Escherichia coli by measuring fluorescence intensity of PI treated with increasing concentrations of compound 1. D Intracellular changes of K+ probed with the K+-sensitive dye PBFI treated with increasing concentrations of compound 1. Data represented as mean ± standard deviation (SD). *P < 0.05, **P < 0.01, ***P < 0.01. compared to the model group, respectively

Cell membrane, as an important barrier, retains cellular constituents while removing unwanted substances [35]. Many antimicrobial compounds have been reported to increase cell membrane permeability [36, 37]. Therefore, the outer membrane (OM) permeability and integral membrane (IM) permeability of bacteria were assessed using 1-N-phenylnaphthylamine (NPN) and propidium iodide (PI) fluorescence probe analysis, respectively. NPN, a hydrophobic fluorescent probe that emitted fluorescence upon interacting interacts with the hydrophobic regions of the phospholipid bilayer, was used to monitor the permeability of the OM [38]. PI, a cell impermeable nucleic acid intercalating dye that can penetrate only dead or damaged cells, was employed to assess integrity of the bacterial inner membrane (IM) [38].

As shown in Fig. 8B and C, compound 1 dose-dependently increased fluorescence intensity in both probe analyses, indicating that compound 1 could enhance the permeability of both the OM and IM of the bacterium. To further investigate the effect of compound 1 on cell membrane function, a potassium (K+) indicator fluorescent dye was used to assess K+ flux after treatment with different concentrations of compound 1 [39]. As shown in Fig. 8D, compound 1 dose-dependently decreased intracellular potassium levels, indicating that compound 1 disrupts the cell membrane by affecting K+ ions flux.

In order to better understand the potential antibacterial mechanism of compound 1 against Escherichia coli in gene expression, strand-specific prokaryotic transcriptome sequencing was performed [40]. By comparing the transcriptomic analysis of the control and treatment groups, 248 differentially expressed genes (DEGs) were identified in samples after FCS treated (|log2 (fold change)|> 1, FDR < 0.05), consisting of 230 upregulated genes and 18 downregulated genes (Fig. 9A and Table S10). The higher proportion of DEGs in Escherichia coli indicated its more cellular response to compound 1, and this result was consistent with the high antibacterial activity of compound 1 against Escherichia coli. Additionally, the cluster analysis of DEGs after FCS treatment indicated that the gene expression patterns in samples were also mostly up-regulated (Fig. 9B), indicating that the treatment with compound 1 might target specific biological processes or pathways within Escherichia coli [41].

A Transcriptome analysis of Escherichia coli after treatment with compound 1. A Volcano plot of expression of genes in Escherichia coli after treatment with compound 1. Blue dots indicate downregulated genes, and red dots indicate upregulated genes. B Clustered heatmap of DEGs, with red cluster representing up-regulated and blue cluster representing down-regulated. The color from blue to red indicates higher gene expression

Gene Ontology (GO) analysis, an internationally standardized system for classifying gene function, provides a controlled vocabulary and strictly defined concepts to describe gene functions and gene products in any organism [42]. As shown in Fig. 10A, DEGs annotated by GO enrichment analysis were further characterized into three categories, including biological process (BP), cellular component (CC) and molecular function (MF). Regarding the category of BP, the DEGs were mainly distributed in cellular process, biological regulation, response to stimulus, and metabolic process. For the category of CC, the DEGs were mainly distributed in cellular anatomical entity and protein-containing complex. With the category of MF, the main groups that DEGs distributed were binding, catalytic activity, transporter activity, and transcription regulator activity. The results revealed that many genes of Escherichia coli. involved in biological process, cellular component, and molecular function were altered, especially key functional modules in cellular process, transportation, binding, catalytic activity, and response to stimulus up-regulated, suggesting that exposure to compound 1 affected the metabolism and activity of Escherichia coli.

A Go enrichment analysis of DEGs annotated in three main categories: molecular function (MF), biological process (BP), and cellular component (CC). B KEGG pathway enrichment analysis of DEGs in Escherichia coli after treatment with compound 1

Systematic research on biological pathways is important for understanding and facilitating genomics research. Kyoto Encyclopedia of Genes and Genome (KEGG) enrichment analysis was further performed to categorize DEGs into different pathways, which could reveal the functional information and relationship at the molecular, cellular, and organism levels [43]. As shown in Fig. 10B, the results indicated that ABC transporters were the most enriched pathways in DEGs. The amino acid biosynthesis pathways including arginine biosynthesis, alanine, aspartate and glutamate metabolism, glycine, serine and threonine metabolism, sulfur metabolism were enriched. Other major pathways related to energy metabolism and cofactor synthesis involving in thiamine metabolism, biosynthesis of cofactors, phosphonate and phosphinate metabolism were enriched, together with pathways of DNA damage, repair involving in DNA replication, mismatch repair, homologous recombination enriched. The results above suggested that these pathways played important roles in antibacterial processes of E. coli.