Seaweed is one of the large and diverse ecosystems, playing an essential role in marine environment. It is mainly involved in global primary production, providing food and shelter for variety of organisms [1]. Additionally, the seaweed surface provides a suitable substratum for the settlement of microorganisms and also secretes various organic substances. These are excellent nutrients for microbial and the marine-derived endophytic fungi which are promising sources of novel interesting bioactive natural products with great pharmacological and agrochemical potentials [2, 3]. Kappaphycus alvarezii is an edible red seaweed imported to Vietnam in 1993. Recently, it is grown popular at the sea areas in the Central Vietnam. The K. alvarezii is a living environment of numerous endophytic microorganisms as well as the fungus Aspergillus micronesiensis. In our ongoing search for structurally novel and bioactive metabolites from marine-derived microorganism, we report herein the isolation, structural elucidation, and biological activity results of three novel dibenzospiroketals 1–3 from the culture broth of K. alvarezii-derived endophytic fungus A. micronesiensis.

Compound 1 was obtained as a yellowish amorphous powder. The molecular formula of 1 was determined as C19H18O10 by HR-ESI-MS with the exhibition of quasi-molecular ion peak at m/z 405.0833 [M-H]- (Cacld. for C19H17O10, 405.0822). The IR spectrum of 1 exhibited absorption band at 3360 and 1635 cm−1, characteristic for stretching vibration of hydroxy and carbonyl groups, respectively. The 1H NMR spectrum of 1 exhibited two methyl groups [δH 2.13 and 2.03 (3H, each, s)], one methoxy group [δH 3.60 (3H, s)], one oxygenated methine [δH 5.77 (1H, s)], and one oxygenated methylene [δH 5.06 and 5.24 (each, 1H, d, J = 15.5 Hz). The 13C-NMR spectra of 1 exhibited the signals of 19 carbon atoms, including two methyl groups (δC 10.2 and 12.0), one methoxy group (δC 55.9), one oxygenated methylene (δC 75.0), one oxygenated methine (δC 97.6), and 14 non-protonated carbons confirming by HSQC spectra. Among non-protonated carbons, one ketone group was identified at δC 195.4. 12 carbon signals in olefinic chemical shift region corresponded to two benzene rings. An acetal carbon was revealed at δC 109.7. The NMR data of 1 were close similarity to those of eleganketal A (4), a dibenzospiroketal isolated from the fungus Spicaria elegan [4]. The difference between them was the appearance of an oxygenated methine (δC 97.6) and a methoxy group (δC 55.9) in 1 instead of the oxygenated methylene (δC 60.4) in 4 (Fig. 1 and Table 1). In the HSQC spectrum, protons at δH 2.13, 2.03, 5.77, 3.60 had cross peaks with carbons at δC 10.2, 12.0, 97.6, and 55.9, respectively; protons at δH 5.06 and 5.24 had cross peaks with carbon at δC 75.0. The HMBC spectrum exhibited the HMBC correlations from H3-9 (δH 2.03) to C-2 (133.1)/C-3 (109.7)/C-4 (147.8), from H2-1 (δH 5.06 and 5.24) to C-2/C-3/C-7 (δC 108.5)/C-8 (δC 109.7) confirming the structure of the isobenzofuran system (rings A and B) (Fig. 2). Similarly, the HMBC correlations from H3-9ʹ (δH 2.13) to C-2ʹ (δC 131.1)/C-3ʹ (δC 115.8)/C-4ʹ (δC 153.8), from H-1ʹ (δH 5.77) to C-2ʹ/C-3ʹ/C-7ʹ (δC 107.3), and four bonding HMBC coupled of H-1ʹ/C-8ʹ (δC 195.4) confirming the structure of the isochroman-4-one system (rings C and D) [4]. The HMBC correlations between H2-1 (δH 5.06, 5.24)/H-1ʹ (δH 5.77) and C-8 (δC 109.7) suggested for the connection of isobenzofuran with isochroman-4-one system via the acetal carbon C-8. The methoxy group was attached to C-1ʹ confirming by the HMBC correlation from methoxy protons (δH 3.60) to C-1ʹ (δC 97.6). Due to containing two chiral carbon atoms (C-8 and C-1ʹ), absolute configuration of 1 was attempted to study by ECD spectral analysis. Four possible stereoisomers of 1 including 1a (8S,1ʹR), 1b (8S,1ʹS), 1c (8R,1ʹS), 1d (8R,1ʹR) were subjected to TD-DFT calculation their theoretical ECD spectra [5,6,7]. The experimental ECD spectrum of 1 showed negative Cotton effects at wavelengths of 255 nm (Δε: −8.63, −34.9% in relative Δε)/293 nm (Δε: −5.89, −23.8 % in relative Δε) and positive Cotton effects at wavelengths of 215 nm (Δε: + 18.45, + 74.5% in relative Δε)/358 nm (Δε: + 3.89, + 15.7% in relative Δε) which were well agreed with the theoretical calculated ECD spectrum of isomer 1a (8S,1ʹR) (Fig. 3). Thus, the structure of compound 1 was established and named as aspermicrone A.

Fig. 1
figure 1

Chemical structures of compounds 1–4

Full size image
Table 1 NMR data of compounds 1–4
Full size table
Fig. 2
figure 2

Key HMBC correlations (H → C) of compounds 1–3

Full size image
Fig. 3
figure 3

Experimental ECD spectra of compounds 1–3 and TD-DFT calculated ECD spectra of their possible stereoisomers

Full size image

Compound 2 was obtained as a yellowish amorphous powder. The HR-ESI-MS of 2 exhibited a quasi-molecular ion peak at m/z 405.0839 [M-H]− (calcd for C19H17O10, 405.0822), suggesting the molecular formula of 2 to be C19H18O10. The NMR spectral data of 2 were very similar to those of 1 indicating that these two compounds have the same dibenzospiroketal skeleton structure (Fig. 1). Major difference in the NMR data between compounds 1 and 2 were signals of oxygenated methine and oxygenated methylene groups. The shielded movement of oxygenated methylene carbon signal (δC 62.4 in 2 and δC 75.0 in 1) suggested the assignment of oxygenated methylene at C-1ʹ as that reported in compound 4 [4]. Meanwhile, the deshielded movement of oxygenated methine carbon signal (δC 108.3 in 2 and δC 97.6 in 1) expected for the presence of methoxy group at C-1. This deduction was further confirmed by HMBC correlation between methoxy protons (δH 3.37) and C-1 (δC 108.3). The both absolute configurations at C-1 and C-8 of compound 2 were determined to be R by comparison experimental ECD spectrum of 2 [217 nm (Δε: +8.85, +27.3% in relative Δε), 243 nm (Δε: −4.22, −13.0% in relative Δε), 347 nm (Δε: +2.31, +7.1% in relative Δε)] with those theoretical calculation ECD spectra for its possible stereoisomers (2a-2d, Fig. 3). Consequently, the structure of 2 was unambiguously established and named as aspermicrone B.

Compound 3 was obtained as yellowish amorphous powder. The molecular formula of 3 was also deduced as C19H18O10 by a quasi-molecular ion peak at m/z 405.0829 [M-H]- in the HR-ESI-MS (calcd for C19H17O10, 405.0822). Interestingly, the NMR spectral data of 3 was identical to those of 2 except slight difference in signals of oxygenated methine, ketone, and methoxy group (Table 1). Particularly, the signals of oxygenated methine (C-1) exhibited deshielding trend from δH-1 6.16/δC-1 108.3 in compound 2 to δH-1 6.42/δC-1 108.8 in compound 3. Meanwhile, the signals of ketone and methoxy groups displayed shielding trend from δC 194.1, δC 53.3, δH 3.37 in compound 2 to corresponding δC 193.5, δC 52.4, δH 3.20 in compound 3. On the other hand, the HMBC correlations from methoxy at δC 3.20 to C-1 (δC 108.8), from H-1 (δH 6.42) to C-2/C-3/C-7/C-8, as well as from methyl protons at δH 2.15 to C-2/C-3/C-4 were observed, confirming the position of methoxy group at C-2 and methyl group at C-3. The HMBC interactions from H-9ʹ (δH 2.06) to C-2ʹ (δC 132.9)/C-3ʹ (δC 112.3)/C-4ʹ (δC 153.5), from H-1ʹ (δH 4.89, 5.20) to C-8 (δC 106.9)/C-2ʹ/C-3ʹ/C-7ʹ (δC 107.9) further confirmed the position of methyl group at C-3ʹ. Moreover, NOESY cross peaks from H-1 (δH 6.42) to H-9 (δH 2.15) and from H-1ʹ (δH 4.89, 5.20) to H-9ʹ (δH 2.06) further indicated the close in proximity of H-1/H3-9 and H-1ʹ/H-9ʹ, confirming location of methyl groups at C-3 and C-3ʹ. From the above evidence, compound 3 was determined to be a stereoisomer of compound 2. In the ECD spectrum of 3, the negative Cotton effect at wavelength of 287 nm (Δε: −1.32, −9.7% in relative Δε) and the positive Cotton effect at 209 nm (Δε: +13.60, +100.0% in relative Δε)/351 nm (Δε: +1.19, +8.7% in relative Δε) which were well agreed with those TD-DFT calculated ECD spectrum of stereoisomer 2c (Fig. 3). Thus, absolute configurations at C-1 and C-8 of compound 3 were determined to be S and R, respectively. Compound 3 was also a novel dibenzospiroketal and named as aspermicrone C.

Dibenzospiroketals sharing carbon backbone of compounds 1-3 are very rare in the nature. To the best of our knowledge, to date, only one dibenzospiroketal, named eleganketal A (4), was isolated from modified culture broth of the fungus S. elegans [4]. Compounds 1–3 were then evaluated their cytotoxicity against HepG2 and LU-1 cancer cell lines, and Vero normal cell line by SRB assay [8, 9]. Interestingly, compound 2 displayed selective cytotoxic effect against HepG2 cells (IC50 = 9.9 µM) and did not show cytotoxic activities toward either LU-1 cancer cell or Vero normal cells in our experiments (IC50 > 50 µM). Compounds 1 and 3 were inactivity (IC50 > 50 µM). Doxorubicin was used as a positive control with IC50 values of 0.53 and 0.57 µM against HepG2 and LU-1 cancer cell lines, respectively. Additionally, compounds 1–3 were also evaluated their anti-microbial activity against microorganisms, Gram-positive bacteria (Bacillus subtilis and Staphylococcus aureus), Gram-negative bacteria (Escherichia coli and Pseudomonas aeruginosa), and yeast (Candida albicans), using dilution turbidimetric broth method as the standard protocols published by the Clinical and Laboratory Standard Institute [10]. Both of compounds 2 and 3 exhibited a MIC value of 123.2 µM toward Gram-positive S. aureus. But they did not inhibit the growth of others tested strains (MIC > 1000 µM). Streptomicin was used as positive control against S. aureus bacteria (MIC = 24.75 µM). Compound 1 did not show anti-microbial activity in our conditions (MIC > 1000 µM).

In conclusion, our results indicated that the fungus A. micronesiensis produced novel dibenzospiroketals (1–3). Compounds 2 exhibited selective cytotoxic effect toward HepG2 cell (IC50 = 9.9 µM). Additionally, among five tested strains, both of compound 2 and 3 displayed anti-microbial activity against S. aureus (MIC = 123.2 µM for each compound). The results warned that dibenzospiroketals such as compounds 2 and 3 would be potential anti-microbial agents. Compound 2 would be useful as a selective anti-cancer agent.