Vibralactone derivatives isolated from co-cultures of the basidiomycetes Stereum hirsutum and Boreostereum vibrans

1 Introduction

Natural products are a vital reservoir of compounds with intriguing structures and diverse bioactivity [1-6]. Microbes are particularly advantageous for natural product discovery due to their easy accessibility and amenability to genetic manipulation [7]. Secondary metabolites of fungal origin represent a crucial subgroup of natural products, exhibiting structural and biological diversity distinct from those found in plants and animals [8]. Several effective strategies are commonly employed to fully explore and expand the chemodiversity of microbial natural products. These include culturing in the presence of epigenetic modifiers [9] or in large-scale [10], co-culturing fungi with other fungi or bacteria [11], overexpressing key transcription factors within gene clusters of interest [12], and heterologous expression of biosynthetic gene clusters [13].

Our research group has long been dedicated to the chemical and biological study of specialized metabolites from higher fungi [14-18]. In previous work, vibralactone, a bicyclic β-lactone-containing meroterpene isolated from Boreostereum vibrans, a basidiomycetous fungus [19], was found to exhibit significant pancreatic lipase inhibitory activity (IC₅₀ 0.4 μg/mL). Subsequent studies on the cultures of B. vibrans resulted in the isolation of numerous vibralactone derivatives [20-23]. Both the total synthesis and biosynthesis of vibralactone have been successfully achieved [24-28]. Beyond our work with B. vibrans, we also focus on the chemodiversity of Stereum hirsutum HFG27, the type strain of the genus Stereum. Stereum hirsutum is a prolific producer of meroterpenoids [29], steroids [30], and sesquiterpenoids [31], and has also been found to produce trace amount of vibralactone (about 100 μg/L). Genetic analysis of these two fungi revealed that they both harbor the key biosynthetic genes for vibralactone and extensive homologous genes [26], but they also display distinct biosynthetic potential for other types of natural products.

Co-culturing microorganisms often involves combining two phylogenetically distinct fungi or a fungus with a bacterium to significantly stimulate the production of secondary metabolites during microbial interactions [32]. While only a few studies have focused on the co-culture of two pairs of basidiomycetous fungi from different families, such as co-culture Trametes versicolor with Ganoderma applanatum, Pleurotus ostreatus with Trametes robiniophila. As a result, eleven compounds, including phenolic glycosides and terpenoids, have been reported in total [33-35]. Boreostereum vibrans and Stereum hirsutum, two closely related fungal species with great potential of producing secondary metabolites, were co-cultured. Mixed cultures of these two fungi may increase the probability of producing a diverse range of compounds, potentially influenced by homologous biosynthetic enzymes. In this study, these two fungi were co-cultured to investigate their chemical profiles. This work details the isolation, structural characterization, and biological assessment of the resulting compounds.

2 Results and discussion

2.1 Structural elucidation of the previously undescribed compounds

Hirsutavibrin A (1, Fig. 1) was isolated as a colorless oil. High-resolution positive electrospray ionization mass spectrometric [( +)-HRESIMS] analysis of 1 returned a protonated ion peak at m/z 241.10703 [M + H]⁺, corresponding to the molecular formula of C₁₂H₁₆O₅ (mass error 0.08201 ppm) with five degrees of unsaturation. The ¹H NMR spectrum of 1 (Table 1) revealed two olefinic protons at δ_H 6.72 (1H, s) and 5.12 (1H, t, J = 7.6 Hz), in addition to two methyl groups at δ_H 1.70 (3H, s) and 1.61 (3H, s). The ¹³C NMR (Table 2) and Distortionless Enhancement by Polarization Transfer (DEPT) spectroscopic data of 1 presented 12 carbon signals, including two CH₃ at δ_C 18.2 (C-11) and 26.3 (C-12), two CH₂ at δ_C 36.0 (C-8) and 40.8 (C-4), three methine carbons at δ_C 78.4 (C-5), 145.7 (C-2), and 120.6 (C-9), three proton-deficiency carbons at δ_C 67.4 (C-1), 135.9 (C-10), and 136.6 (C-3), and two carbonyl groups at δ_C 177.3 (C-7) and 169.1 (C-13). The ¹H-¹H COSY correlations of H-4/H-5, along with the HMBC correlations from H-2 to C-1, C-3, C-4, C-5, C-13 (Fig. 2) revealed the existence of a five-membered ring in 1, which, together with the two carbonyl groups and two double bonds mentioned above, satisfied the unsaturation degree of compound 1. The ¹H–¹H COSY correlation of H₂-8/H-9, and the HMBC correlations from H₃-12 to C-9, C-10, C-11, and from H-8 to C-1 indicated a prenyl group attach to C-1. In addition, the HMBC correlations from H-2 to C-13, from H-8 and H-5 to C-7 enabled the assignment of two carboxylic groups at C-1 and C-3, respectively. Therefore, the planar structure of 1 was established (Fig. 1), which highly resembled vibralactone E [21], except that the CH₂OH group (C-13) was replaced by a COOH group in 1.

Open image in new window

Open image in new window

The conclusion of the relative configuration of 1 arrived by analyzing the ROESY spectrum (Fig. 3). The key correlations of H-5 (δ_H 4.22)/H-8a (δ_H 2.30), H-5/H-4α (δ_H 2.87), and H-8a/H-4α helped to locate H-5 and the prenyl group orienting the same side of the five-membered ring. Besides, almost all the prenyl group at C-1 from the previously reported vibralactone and its derivatives are β configuration, thus the relative configuration of 1 was determined to be 1R*, 5S*. The absolute configuration of 1 was determined by ECD calculations (Supplementary Material, Fig. 4). As a result, the calculated ECD of the stereoisomer 1S, 5R was most consistent with the experimental CD data of 1, while the calculated ECD data of 1R, 5S stereoisomer showed mirror Cotton effects at specific wavelengths (Fig. 4B). Therefore, the absolute configuration of 1 was assigned as depicted in Fig. 1, which was different from the previously reported vibralactone derivatives.

Open image in new window

Open image in new window

Hirsutavibrin B (2, Fig. 1) was obtained as a colorless oil. Its molecular formula was determined to be C₁₂H₁₆O₅ based on (+)-HRESIMS, which presented an [M + H]⁺ ion peak at m/z 241.10707 (calcd for C₁₂H₁₇O₅, 241.10705). The ¹³C NMR and DEPT spectroscopic data of 2 (Table 2) showed two CH₃ at δ_C 18.0 and 26.1, two CH₂ at δ_C 30.9 and 39.9, three CH at δ_C 76.9, 145.1, and 120.7, three non-protonated carbons at δ_C 64.6, 135.5, and 135.4, and two carbonyl carbons at δ_C 178.0 and 168.5. The 1D NMR data showed high similarity with those of 1. By interpretating of the 2D NMR spectra, including HSQC, ¹H–¹H COSY, and HMBC spectra (Fig. 2), 2 was determined to have a planar structure identical to 1. The weak correlation signal of H-5 (δ_H 4.59)/H-8a (δ_H 2.60) in the ROESY spectrum indicating that H-5 and the prenyl group resided the opposite sides of the five-membered ring, which helped to establish the relative configuration of 2 as 1R*, 5R*. An ECD calculation workflow was applied to determine the absolute configuration of 2. As shown in Fig. 4C, the calculated ECD of the 1R, 5R stereoisomer matched well with the experimental CD data. Therefore, the evidence allowed the complete assignment of 2D structure and absolute configuration of 2 (Fig. 1).

Hirsutavibrin C (3, Fig. 1) was obtained as a colorless oil. The molecular formula of 3 was determined as C₁₄H₁₈O₆ according to the HRESIMS data (m/z 283.11765 [M + H]⁺, calcd for C₁₄H₁₉O₆, 283.11761), indicating six degrees of unsaturation. The ¹H and ¹³C NMR data of 3 (Tables 1 and 2) were closely similar to the data of 2, indicating the structure of 3 closely resembled that of 2, with the primary difference being associated with an extra acetoxyl group in 3. The methyl singlet at δ_H 1.98, the carbonyl carbon at δ_C 172.1, together with the key HMBC correlation from H-5 (δ_H 5.27) to δ_C 172.1 (Fig. 2) of 3 enabled the assignment of the acetoxyl group attaching to C-5. Although in the ¹³C NMR spectrum, the signal for C-13 was not detected, which is common for carboxylic groups [36], the HRESIMS result and the chemical shifts of C-2, C-3 and C-4 confirmed that a carboxylic acid group attached to C-3. The crucial and significant correlation of H-5/H-8 (δ_H 2.60, 2.35) in the ROESY spectrum (Fig. 3) assigned the H-5 and the prenyl group locating at the same side of the five-membered ring (Fig. 2). In addition, the calculated ECD spectrum of (1R, 5S)-3 showed similar adsorption trend with the experimental CD spectrum of 3. Therefore, the absolute configuration of 3 was established as 1R, 5S (Figs. 1 and 5).

Open image in new window

The colorless oil hirsutavibrin D (4, Fig. 1) had a molecular formula of C₁₄H₂₀O₅ which returned from the (+)-HRESIMS analysis (m/z 269.13834 [M + H]⁺), revealing five degrees of unsaturation. In the ¹³C NMR spectrum of 4, 14 carbon signals were displayed (Table 2), which showed similarity to 3, except that an oxygenated methylene (C-13, δ_C 61.6) in 4 replaced the carbonyl group signal of 3. The key HMBC correlations from H-2 (δ_H 5.72) and H-13 (δ_H 4.11) to C-3 (δ_C 144.4), and from H-2 to C-13 (Fig. 2) supported the conclusion that the C-13 of 4 was a hydroxymethyl group instead of being a carboxylic acid group as in 3. The key ROESY correlation between H-5 and H-8 (Fig. 3) suggested that the prenyl group and H-5 were in the same plane of the five-membered ring. Therefore, compound 4 was identified as a C-13 reduced product of 3.

The molecular formula of hirsutavibrin E (5, Fig. 1) was determined as C₁₄H₂₀O₅ according to the (+)-HRESIMS data (m/z 269.13834 [M + H]⁺, calcd. for C₁₄H₂₁O₅). The 1D NMR spectroscopic data suggested that 5 (Table 2) was a congener of 4, except that the acetoxyl group was changed to locate at C-13 according to the ³J-HMBC correlations from the proton at δ_H 4.59 (H-13) to δ_C 170.9, and the ⁴J-HMBC correlations from methyl at δ_H 2.08 (13-OCOCH₃) to C-13 (δ_C 62.8) (Fig. 2). The key correlation between H-5 and H-8 (Fig. 3) was absent in the ROESY spectrum, while between H-5 and H-4α was seen. Therefore, the relative configuration of 5 was speculated to be 1R*, 5R*. The calculated ECD curve of (1R, 5R)-5 showed a Cotton effect similar to the experimental one (Fig. 5), thus establishing the absolute configuration of 5 to be 1R, 5R.

Compound 6 (Fig. 1) was isolated as a colorless oil. The molecular formula was assigned as C₁₆H₂₂O₇ based on the positive HRESIMS ion peak at m/z 327.14359 [M + H]⁺ (calcd for C₁₆H₂₃O₇, 327.14383). Compared to the NMR data of 5 (Tables 1 and 2), the signals for acetyl group were absent in compound 6. In addition, four additional carbon resonances which were ascribable to two carbonyl groups (δ_C 174.1, 176.7) and two methylenes (δ_C 30.2, 30.3) (Table 2) were presented in 6. By interpreting the 2D NMR spectra, the four carbons were assigned to be a succinic acid moiety, which was determined by the correlations from H-2' (δ_H 2.58) and H-3' (δ_H 2.62) to C-1' (δ_C 174.1) and C-4' (δ_C 176.7), from H-13 (δ_H 4.61) to C-1' and C-3 (δ_C 138.3) (Fig. 2) in the HMBC spectrum. The correlation between H-5 and H-8 was not seen in the ROESY spectrum, suggesting a 1R*, 5R* configuration of 6. With the help of ECD calculations, an absolute configuration of 1R, 5R was assigned to 6 based on the highly similar Cotton effect between the experimental CD diagram and the calculated ECD data of (1R, 5R)-6 (Fig. 5). Therefore, compound 6 was trivially named hirsutavibrin F.

The ¹H and ¹³C NMR data of 7 (Tables 2 and 3) displayed twelve carbon resonances which were classified into two CH₃, four CH₂ (one oxygenated), two sp³ CH (one oxygenated), one sp² CH, one sp³ qC, and two sp² qC (one carbonyl and one olefinic carbon). The NMR data of 7 displayed similarity to those of vibralactone E, except that the C-2–C-3 double bond was reduced to a single bond, which was corroborated by the key ¹H-¹H COSY correlations between H-2 (δ_H 2.31) and H-3 (δ_H 2.21) (Fig. 2). These assignments reached the conclusion that the molecular formula of 7 was C₁₂H₂₀O₄, which was consistent with the HRESIMS analysis results (m/z 229.14349 [M + H]⁺, calcd. for C₁₂H₂₁O₄, 229.14344). The key correlations of H-2β/H-3/H-4β/H-5 (δ_H 4.27)/H-8a (δ_H 2.51), H-2α/H-13/H-4α, and H-5/H-9 in the ROESY spectrum (Fig. 3) indicated the relative configurations of 7 to be 1R*, 3S*, 5S*. The ECD calculation steps were applied to deduce the absolute stereochemistry of 7. Comparative analysis of the measured and theoretically calculated ECD curves (Fig. 5) revealed a 1R, 3S, 5S configuration of 7. Compound 7 was named hirsutavibrin G.

The molecular formula of compound 8 (Fig. 1) was deduced to be C₁₂H₂₀O₄, which was identical with 7. Detailed interpretation of the NMR of 8 revealed the same planar structure as 7 (Tables 2 and 3). The lack of key ROESY signals between H-5 and H-8 indicated a 1R*, 3S*, 5R* configuration of 8 (Fig. 3). Considering the same biosynthetic pathways of 7 and 8, the absolute configurations of 8 were determined to be 1R, 3S, 5R. Compound 8 was trivially named hirsutavibrin H.

Hirsutavibrin I (9, Fig. 1) was isolated as a colorless oil. The (+)-HRESIMS sodium adduct ion peak at m/z 265.10452 [M + Na]⁺ (calcd for C₁₂H₁₈O₅Na, 265.10519) revealed a chemical formula of C₁₂H₁₈O₅ of 9. The ¹³C NMR data of 9 (Tables 2 and 3) displayed two methyls, three methylenes (one oxygenated), two oxygenated sp³ methines (δ_C 63.9, 75.0), two sp³ quaternary carbons (δ_C 59.7, 66.8), etc. These data are different from those of compounds 1–6, while are similar to those of compound 7, and are highly resemble to those of vibralactone B [20, 37]. Compound 9 was 18 Da larger than vibralactone B, suggesting that 9 was the lactone ring-opening product of vibralactone B. The crucial ROESY correlations of H-2 (δ_H 3.38)/H-8 (δ_H 2.31, 2.01)/H-5 (δ_H 3.76), H-2/H-9 (δ_H 5.08) indicated the 1S, 2S, 3S, 5S configuration of 9. Hence compound 9 was elucidated as depicted in Fig. 1.

The chemical formula of 10, obtained as a colorless oil, was determined to be C₁₄H₂₂O₇ (m/z 303.14383 [M + H]⁺, Δ 0.00048 ppm) by (+)-HRESIMS. The 1D NMR data between 10 (Tables 2 and 3) and 5 were quite similar, except that the C-2/C-3 double bond between in 5 was replaced by a vicinal diol in 10, according to the key HMBC correlations from H-2 (δ_H 3.45) to C-1 (δ_C 54.3), C-3 (δ_C 76.9), C-5 (δ_C 70.6), from H-8 (δ_H 2.29) to C-2 (δ_C 83.7), and from H-13 (δ_H 3.97) to C-2, C-3, C-4 (δ_C 45.0), and from 3-OH (δ_H 4.60) to C-2, C-3, C-4, and C-13 (Fig. 2). The crucial ROESY correlations of H-2 (δ_H 3.45)/H-8/H-5 (δ_H 3.57)/3-OH (Fig. 3) indicated the 1S, 2S, 3R, 5S stereochemistry of 10. Therefore, the structure of compound 10 was established and named hirsutavibrin J.

( +)-HRESIMS spectrometric analysis established the molecular formula of 11 (Fig. 1) as C₁₂H₁₈O₅Na (m/z 265.10461 [M + Na]⁺, mass error 0.11536 ppm). The 1D NMR spectroscopic data of 11 was similar to those of compound 7, except with the existence of a new carbonyl carbon (δ_C 181.0) in 11 while absence of the hydroxymethyl group in 7. This evidence indicated that compound 11 differs from 7 by the oxygenated status of C-13. The carbonyl group at δ_C 181.0 in compound 11 was assigned to a carboxylic group (C-13). The correlation from H-3 (δ_H 2.86) to C-13 in the HMBC spectrum (Fig. 2) as well as the molecular formula corroborated the above assignments. The vital ROESY correlations of H-8/H-5/H-3 suggested a 1R, 3S, 5S configuration of compound 11 (Fig. 3). Therefore, the structure of compound 11 was established and named hirsutavibrin K.

2.2 Biological activity evaluation of 1–11

The isolated compounds with abundant yield (1, 2, 4, 6) were subjected to a panel of biological activity screening, including the cytotoxicity against A549, a human lung cancer cell line, and anti-NO (nitric oxide) activity in murine monocytic RAW 264.7 macrophages. As a result, compounds 1 and 2 showed weak cytotoxicity activity toward A549 with the IC₅₀ 39.7 and 34.3 μM, respectively (positive control cisplatin, IC₅₀ 5.09 μM). Compound 4 displayed anti-NO activity with IC₅₀ of 26.6, which were more significant than the positive control L-N^G-monomethyl arginine (IC₅₀ 51.2 μM).

3 Conclusion

In conclusion, eleven new vibralactone derivatives (1 − 11) were obtained from the co-culture broth of the two basidiomycetous fungi Stereum hirsutum and Boreostereum vibrans. Comprehensive analyses of 1D & 2D NMR spectroscopic data and computational calculations allowed the establishment of the structures of all the isolates. Most of the isolates are structurally similar to vibralactone but display distinct configurations.

Co-culturing microbes typically involves ascomycetes or ascomycetes with bacteria, with few studies focusing on secondary metabolite discovery from co-cultures of two basidiomycetes. Our research demonstrates that co-culturing phylogenetically related fungal species is an effective approach for discovering natural products. For instance, previous studies on Boreostereum vibrans cultures primarily yielded vibralactone derivatives with the same absolute configuration as vibralactone. However, when Boreostereum vibrans was co-cultured with Stereum hirsutum, vibralactone derivatives with distinct configurations and additional modifications were discovered. Notably, none of the isolated compounds have been found from the individual cultures of each fungus. While further evidence is needed to confirm the involvement of new biosynthetic pathways in the modification of vibralactone, this study underscores that co-culturing two basidiomycetes is a promising strategy for natural product discovery. This work not only expands the members of vibralactone derivatives with different configurations but also opens a new avenue for fungal co-culturing study between congeneric fungi.

Notes

Acknowledgement

This work was financially supported by the National Natural Science Foundation of China (Grant number 82473810) and the Fundamental Research Funds for the Central Universities, South-Central Minzu University (Grant No. CPT22033). The authors thank the Analytical & amp; Measuring Center, School of Pharmaceutical Sciences, South-Central Minzu University for the spectra test, and the Bioactivity Screening Center of Natural Products, Kunming Institute of Botany, CAS for biological activity screening.

Author contributions

He-Ping Chen and Ji-Kai Liu: Conceptualization and writing—reviewing and editing, funding acquisition, methodology project administration; Jinjuan Wei, Zhe-Xi Li, and Gao-Ke Peng, investigation, formal analysis, writing-original draft preparation; Xinyang Li: Writing—reviewing and editing. All authors approved the final version for publication.

Availability of data and materials

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Ji-Kai Liu is the Editor-in-Chief of this journal but was not involved in the peer review or decision-making process in this article. The other 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.