Introduction

Ubiquitous fungi, hailed as high-performance creators of natural occurrences, produce overwhelming second metabolites with diverse original scaffolds and versatile biological activities.1 These attractive characteristics render fungi as an integral part of mining groundbreaking drug candidates and novel small-molecule probes.2 Higher fungi, which are typically spore-bearing fruiting bodies of fungi, are a paradigm of fabricating useful natural products for the upstream of drug development.3 The genus Stereum is noted for producing a variety of biologically active second metabolites, including sesquiterpenes,4, 5, 6, 7, 8 dimeric sesquiterpenes,9, 10, 11 isoindolinone alkaloids,12, 13 and vibralactone derivatives.14, 15, 16, 17, 18 Moreover, the isolates of this genus have gained organic chemists extraordinary interests. Elegant total synthesis of several isolates has been achieved.19, 20, 21, 22 In addition, the biosynthetic pathway for vibralactone, a pancreatic lipase inhibitor from S. vibrans has been deciphered and a monooxygenase from S. vibrans is also identified.23, 24

Previous chemical investigations of the genus Stereum mainly centered on culture broth, which were capable of producing diverse second metabolites by scale-up fermentation or using different culture media. Our follow-up search for bioactive natural products from higher fungi, twelve new chemical entities, sterenoids A–L (112, Figure 1), were isolated from fruiting bodies of Stereum sp., which is a wood decaying fungus dwelling at Xishuangbanna Tropical Botanical Garden. Compounds 112 are rare 14(13→12)abeo-lanostane-type triterpenoids featuring distinctive 13R configurations that are incompatible with previously covered counterparts.25, 26 To the best of our knowledge, the triterpenoid, possessing this 14(13→12)abeo-lanostane-type-6/6/5/6 ring core skeleton, was first synthesized occasionally and then was isolated from the medical plant Kadsura heteroclite.26, 27 Intriguingly, triterpenoids with this tetracyclic rearranged scaffold also originated from mushrooms Tyromyces fissilis and Ganoderma lucidum.28, 29, 30 All isolated compounds are evaluated for their cytotoxicity in vitro against five human tumor cell lines. Herein, the isolation, structure elucidation and biological evaluation of new compounds 112 are discussed.

Figure 1
figure 1

Chemical structures of isolated compounds 112.

Results and discussion

Structure elucidation

Compound 1, amorphous powder, had a molecular formula C30H46O3 with eight double bond equivalents as unraveled by the sodium adduct (+)-HR–ESI–MS ion at m/z 477.3350 [M+Na]+ (calcd for 477.3339) and the 13C NMR data. The 1H NMR spectrum (Table 1) of 1 revealed typical resonances for one secondary methyl at δH 0.99 (d, J=6.8 Hz, H3-21), seven tertiary methyls at δH 1.05 (H3-29), 1.11 (H3-28), 1.14 (H3-19), 1.21 (H3-18), 1.23 (H3-30), 1.60 (H3-27) and 1.65 (H3-26), one olefinic proton at δH 5.12 (t, J=7.2 Hz, H-24). A thorough analysis of the 13C NMR data (Table 2), with the aid of distortionless enhancement by polarization transfer (DEPT) and HSQC spectra, unlocked 30 carbon signals, including two carbonyls (δC 208.4 and δC 216.0), 2 double bonds, 8 methyls, 8 sp3 methylenes, 4 sp3 methines and 4 sp3 quaternary carbons (1 oxygenated). One proton resonance at δH 3.56 showed no correlations with any carbons in the HSQC spectrum and thus was designated to the hydroxy group. The aforementioned functionalities carbonyls and double bonds accounted for four of the eight degrees of unsaturation and the remaining four degrees of unsaturation exactly constructed four rings in the molecule.

Table 1 1H NMR spectroscopic data of compounds 16
Table 2 13C NMR spectroscopic data of compounds 16

The planar structure of 1 was established by detailed deciphering of 2D NMR spectra. 1H-1H COSY spectrum indicated eighteen proton-bearing fragments as delineated in bold bonds (Figure 2). Quaternary carbons were attached to these fragments to form the scaffold of 1 by the HMBC correlations. The multiple HMBC correlations of H3-28/C-3, C-4 and C-5; H3-29/C-3, C-4 and C-5; H3-19/C-1, C-5, C-9 and C-10, Ha-1/C-3 and H-7/C-8, C-9 along with COSY correlations of Ha-1/Ha-2, H-5/Hb-6 and Hb-6/H-7 constructed the two six-membered carbon rings A and B. The chemical shifts of 179.7 (C-8), 145.0 (C-9) and 208.4 (C-11) suggested the presence of an α,β-unsaturated ketone, which was appended by one methine (C-12) and one quaternary carbon (C-14) to form the five-membered ring C via HMBC correlations of H3-30/C-8, C-14, C-15 and H-12/C-8, C-11. Furthermore, the 1H-1H COSY revealed the connection of C-15, C-16, C-17, C-20, C-22, C-23 and C-24. The key HMBC correlations of H3-18/C-12, C-13 and C-17 were responsible for the linkage of the six-membered ring D. The hydroxy group at δH 3.56 was fixed at C-13 via the HMBC correlations of the hydroxy proton signal to C-12, C-13 and C-17. Therefore, the tetracyclic triterpenoid scaffold was established. The COSY spin systems between H-17 (δH 1.63) and H-20 (δH 1.81), as well as HMBC correlations of the methyl protons H3-21 (δH 0.99) to C-17 (δC 50.2), C-20 (δC 33.3) and C-22 (δC 34.3) offered solid evidence that the side chain containing eight carbons was fixed to C-17. Overall, the analysis strongly hinted that 1 had a rare 14(13→12)abeo-lanostane-type triterpenoid, which was similar to neokadsuranic acid A.26 The ROESY correlations of H-5 with H3-28 and H-12 with H-17, H3-30, HO-13, and H-17 with H3-21 revealed that H3-28, H-5, H-12, HO-13, H3-30, H-17 and H3-21 were cofacial and were assigned to be β-oriented (Figure 2). The ROESY cross-peaks of H3-19/H3-29 and H3-18/H-20 showed that they had α-orientations.31 Intriguingly, the absolute configuration of C-13 was R, which discriminated from the previously covered 14(13→12)abeo-lanostane-type triterpenoids.25, 32 This arbitrary assumption was confirmed by ROESY and comparison of the experimental electronic CD with quantum chemical calculated electronic CD spectra as shown in Figure 3. Hence, the absolute configuration of 1 was determined as 5R, 10s, 12R, 13R, 14R, 17R, 21R. Taken together, the structure of 1, sterenoid A, was unambiguously characterized as 24(E)-3,11-dioxo-13α-hydroxy-14(13→12)abeo-lanosta-8,24-dien.

Figure 2
figure 2

1H−1H COSY and selected HMBC (a) and key ROESY (b) correlations of 1.

Figure 3
figure 3

Experimental and calculated electronic CD (ECD) spectra of 1 (full line, experimentally recorded in methanol; dashed line, calculated for 5R, 10S, 12R, 13R, 14R, 17R and 21R configuration in methanol).

Compound 2 and 3 were isolated as optically active, white amorphous solid, which had identical molecular formulas C30H48O3 determined by HR–ESI–MS measurements of the sodium adduct ion at m/z 479.3488 [M+Na]+ and 479.3490 [M+Na]+ (calcd for 479.3496). Analysis of the NMR data of 2 (Tables 1 and 2) suggested that it was a derivative of 1, except for the absence of the carbonyl group for C-11 and an additional hydroxy group at C-7. The 1H-1H COSY correlation of H-7 (δH 4.24) with Hb-6 (δH 1.70) and the HMBC cross-peaks of H-7 with C-8 and C-9 supported the above deduction. The α-oriented hydroxy group at C-7 was defined by the ROESY spectrum via the correlations between H-7/H3-29 and H3-28/H-5. The remaining ROESY correlations suggested that 2 shared the same configuration with that of 1, except for the configuration of hydroxy group for C-7. In parallel, compound 3 was defined as the C-7 epimer of 2 and the configuration of hydroxy group for C-7 was thus assigned to be β-oriented. This assignment was supported by 2D NMR spectra analysis, especially the ROESY correlations of H-7 with H-5 and H3-30. Reinspection of the 13C NMR data of 2 and 3 uncovered that the chemical shift of C-5 (δC 46.4) in 2 was major difference (ΔδC 4.7 p.p.m.) relative to that of 3 (δC 50.2), suggesting the very existence of γ-gauche effects on the 13C NMR chemical shifts. The HO-7 and H-5 of 2 were 1,3-diaxially bonded and the HO-7 was mainly responsible for steric interactions with the H-5. On the contrary, the HO-7 of 3 was equatorial orientation and thus gave less steric hinderance. Consequently, the chemical shift of C-5 (δC 46.4) in 2 was relatively upfield in comparison with that of C-5 (δC 50.2) in 3. The above-discussed key differences in chemical shifts allowed a clear assignment of α or β steric position of the 7-substitutent on the basis of γ-gauche effects.33, 34 The structures of 2 and 3, namely sterenoids B and C, were thus established as 24(E)-3-oxo-7α,13α-dihydroxy-14(13→12)abeo-lanosta-8,24-dien and 24(E)-3-oxo-7β,13α-dihydroxy-14(13→12)abeo-lanosta-8,24-dien, respectively.

Compound 4 gave a molecular formula of C30H46O3, as established on the basis of 13C NMR and HR–ESI–MS spectra, indicating compound 4 was two less hydrogen atoms than that of 2. Their 1H and 13C NMR data (Tables 1 and 2) were similar. The major differences were that the resonances assigned to the hydroxy group in 2 replaced by a carbonyl group, together with the shift of the signals corresponding to C-7 from δH 4.24, δC 63.4 in 2 to δC 196.8 in 4. This change was verified by the HBMC correlations of H-5 to C-7 (δC 196.8), Ha-6 (δH 2.30) to C-7 and Hb-6 (δC 2.49) to C-7. The ROESY correlation of H3-18 with H-20 revealed that the OH-13 was α-oriented. 2D NMR data analysis substantiated the one-dimensional NMR data, relative configuration and regiochemical assignments. Accordingly, the structure of 4, sterenoid D, was deduced as 24(E)-3,7-dioxo-13α-hydroxy-14(13→12)abeo-lanosta-8,24-dien, a congener of 2.

Compound 5 was assigned as the molecular formula C32H50O4 by the HRESIMS ion at m/z 521.3606 [M+Na]+ (calcd for 521.3601). Its 1H and 13C NMR data (Tables 1 and 2) were closely related to those of 1, with the main difference occurring for the signals of the 11-substituent and 21-substituent. The 11-substituent in 5 was shifted as a methylene (δH 1.31, 2.30; δC 29.5) from the carbonyl (δC 208.4) in 1 and the 21-substituent in 5 comprised the resonances of an ester carbonyl (δC 171.6) and a tertiary methyl (δH 1.58; δC 21.2). Further analysis of the HMBC spectrum confirmed an acetoxy group for the 21-substituent. Based on the ROESY spectrum and similar NMR patterns, the relative configuration of all the stereogenic centers was assigned to be identical with those of 1. The structure of 5, sterenoid E, was thus established as 24(E)-3-oxo-13α-hydroxy-21-acetoxy-14(13→12)abeo-lanosta-8,24-dien.

Compound 6 was obtained as a white, amorphous solid. Its molecular formula C30H48O3, with seven indices of hydrogen deficiency, was established from the HR–ESI–MS sodium adduct ion at m/z 479.3492 [M+Na]+ (calcd for 479.3496). The 1H and 13C NMR data (Tables 1 and 2) highly resembled those of 1, suggesting that these two compounds should be homologous carbocyclic skeletons and substitution patterns, except for the existence of diagnostic resonances of one sp3 methylene (δC 28.4), two oxygenated quaternary carbons C-8 (δC 75.3) and C-9 (δC 73.9) replacing those of the α,β-conjugated carbonyl group (δC 208.4, 179.7 and 145.0) of 1 (Tables 1 and 2) in the B/C-ring, respectively. This deduction was further illustrated by complete examination of 2D NMR spectra. The relative configuration of 6 was assigned via ROESY data in comparison with the counterpart of stereogenic centers in 1, with the exception of 8,9-oxirane moiety. The 13C NMR calculations with quantum-based methods pinpointed the relative configuration of epoxide ring motif in 6 as previously reviewed.35 The density functional theory (DFT) calculations of 13C NMR data of the two possible stereoisomers of 6a and 6b were performed (Supplementary Figure S100; for details, see the NMR calculations for compound 6 in the Supplementary Information). The calculated NMR data of the isomer 6b were much closer to the experimental data of 6 as weighed by the linear correlation coefficients (R2) and root-mean-square deviations, suggesting that the epoxide ring motif was α-oriented. The compound of 6, sterenoid F, was elucidated as 24(E)-3-oxo-8(9)-epoxy-13α-hydroxy-14(13→12)abeo-lanosta-24-en.

Compound 7 was isolated as a white amorphous solid. It gave a molecular formula C30H50O2 based on the HR–ESI–MS ion at m/z 465.3705 [M+Na]+ (calcd for 465.3703), corresponding to six double bond equivalents. A comprehensive analysis of 1H NMR and 13C NMR data (Tables 3 and 4) revealed that 7 shared a common A–D ring system with that of 1, occurred an oxygenated methine (δC 76.2) and an additional sp3 methylene (δC 29.7), and disappeared two carbonyls. One carbonyl at C-3 was shifted to a hydroxy carbon (δC 76.2) and the other at C-11 was interchanged by the sp3 methylene, respectively. This plausible hypothesis was verified by the HMBC correlations from H3-28/H3-29 to C-3 (δC 76.2), C-4 (δC 37.4), C-5 (δC 45.9) and from Ha-11 (δH 2.24) to C-8 (δC 139.0), C-9 (δC 142.6), C-12 (δC 58.1) and C-14 (δC 48.7). The ROESY correlations of H3-19/Ha-2, Ha-2/H-3 and H-3/H3-29 showed that they were cofacial, indicating the hydroxy group at C-3 was α-oriented. Taken together, compound 7, sterenoid G, was thus characterized as 24(E)-3α,13α-dihydroxy-14(13→12)abeo-lanosta-8,24-dien.

Table 3 1H NMR spectroscopic data of compounds 712
Table 4 13C NMR spectroscopic data of compounds 712

The molecular formula C30H50O3 was assigned to 8 with six indices of hydrogen deficiency by the 13C NMR data and the HR–ESI–MS ion at m/z 481.3653 [M+Na]+ (calcd for 481.3625), which was more 16 mass units attributable to oxygenated motif than that of 7. The NMR data (Tables 3 and 4) of 8 were highly consistent with those of 7, except for the side-chain moiety. The emerging chemical shifts of 1H and 13C NMR spectra in 8 were assignable to three methyls (δH 0.95, d, J=6.9 Hz; δH 1.30, s and δH 1.30, s), one methylene, one methine, one persubstituted double bond (δH 5.59, m; δH 5.90, d, J=18.1 Hz) and one oxygenated quaternary carbon (δC 70.7). The 1H-1H COSY correlations of H-17/H-20/Ha-22/H-23 and HMBC cross-peaks from H3-21 to C-17, C-20 and C-22, from H3-26 and H3-27 to C-24 and C-25 established the side chain as depicted. The geometry of the Δ23, 24 double bond was assigned as E based on coupling constant (18.1 Hz). A thorough analysis of the ROESY spectrum and NMR patterns revealed that the hydroxy group at C-3 was α-oriented and the other stereogenic centers were assigned to be identical with those of 1. Compound 8, sterenoid H, was thereby elucidated as 24(E)-3α,13α,25-trihydroxy-14(13→12)abeo-lanosta-8,23-dien.

Analysis of HR–ESI–MS and 13C NMR data indicated that compound 9 had the same molecular formula with that of 8. The 1H and 13C NMR data (Tables 3 and 4) of 9 were highly analogous to those of 8, except for minor variations at the A ring, suggesting that 9 should be the C-3 epimer of 8 and HO-3 was determined to be β-oriented.25 This deduction was further confirmed by the ROESY correlation of H-3 with H-5. Therefore, compound 9, 24(E)-3β,13α,25-trihydroxy-14(13→12)abeo-lanosta-8,23-dien, was given trivial name sterenoid I.

The HR–ESI–MS ion at m/z 479.3493 [M+Na]+ (calcd for 479.3496) and 13C NMR data of compound 10 revealed the molecular formula C30H48O3 with seven indices of hydrogen deficiency. NMR data (Tables 3 and 4) showed that 10 was structurally related to 8 with the discrepancy of replacing HO-3 via the carbonyl group. The HMBC correlations from H3-28 (δH 1.10) and H3-29 (δH 1.08) to C-3 (δC 218.1), C-4 (δC 47.0) and C-5 (δC 51.7) supported this hypothesis. In addition, the ROESY spectrum uncovered that all stereogenic centers were agreement with those of 1. Compound 10, sterenoid J, was thus deducted as 23(E)-3-oxo-13α,25-dihydroxy-14(13→12)abeo-lanosta-8,23-dien.

Compounds 11 and 12 exhibited the identical molecular formula C30H50O4 as deduced from the HRESIMS ion at m/z 497.3603 and 497.3600 [M+Na]+ (calcd for 497.3601), respectively. Comparison of the one- and two-dimensional NMR data of 11 and 12 showed similarities, except for the subtle variations of the chemical shifts of C-22 (ΔδC 2.7), C-23 (ΔδC 0.7) and C-24 (ΔδC 1.2) in the side-chain motif, indicating that they were a pair of C-24 epimers. The general features of the 1H and 13C NMR spectra of 11 closely resembled those of 1, except that the Δ23, 24 double bond was subject to the direct hydration. This change was supported by 1H-1H COSY correlations between Hb-23 (δH 1.15) and H-24 (δH 3.28), as well as the HMBC correlations from H3-26 (δH 1.20) and H3-27 (δH 1.15) to C-24 (δC 79.5) and C-25 (δC 73.2). Thus, the structure of 11 was thus deducted as shown (Figure 1). It was not reliable to distinguish the stereogenic centers at C-24 between 11 and 12 on the basis of available NMR data. Thus, the absolute configuration of C-24 was defined by utilizing the Mo2(OAc)4-induced CD experiment for vicinal diols.36 As a result, compound 11 displayed a positive Cotton effect at 313 nm, indicating the 24S configuration for 11. Accordingly, the 24R configuration for 12 was postulated. Therefore, 11 and 12 were named sterenoids K and L, and were assigned as 3-oxo-13α,24S,25-trihydroxy-14(13→12)abeo-lanosta-8-en and 3-oxo-13α,24R,25-trihydroxy-14(13→12)abeo-lanosta-8-en, respectively.

Biogenetically, the conversion of intact lanostane-type triterpenoids to these 14(13→12)abeo-lanostane-type-6/6/5/6 ring core triterpenoids is likely to undergo Wagner–Meerwein rearrangement with the carbocation intermediate. The hydroxy group is subsequently embedded in the carbocation to form the 13R or 13S configurations. From the standpoint of physicochemical stabilization, sterenoids A–L (112) with 13R configurations, which mean that H-12/HO-13 or H-17/HO-13 are syn-coplanar, are more stable than those of 13S configuration counterparts. These 13S configuration counterparts embody intrinsic reactivity of E2 elimination, which requires that the leaving group and the hydrogen are anti-coplanar. Quantum chemistry calculations for natural products of structure verification are exemplified via compounds 1 and 6, suggesting that synergistic combination of experimental NMR data and DFT chemical shift predictions is an unbiased way for structure elucidation.

Cytotoxic activities

Compounds 1–12 were screened for cytotoxicity against five cancer cell lines HL-60 (human promyelocytic leukemia), SMMC-7721 (hepatic cancer), A-549 (lung cancer), MCF-7 (breast cancer) and SW-480 (colon cancer) using the MTS (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxy-phenyl)-2-(4-sulfophenyl)-2H-tetrazolium) method, with cisplatin and paclitaxel as positive controls. These cytotoxicity results (Table 5) disclosed compound 5 displayed potent cytotoxicity against HL-60 acute leukemia and SMMC-7721 hepatic tumor cell lines with IC50 values of 4.7 and 7.6 μM, respectively. Although the remaining compounds (23 and 79) showed weak cytotoxicity against five tumor cell lines and compounds (1, 6 and 1012) were inactive (IC50>40 μM). Comparison of the cytotoxic compounds 15, 79 against HL-60 tumor cell line with 6, 1012, brief structure–activity relationship concludes that Δ8, 9 and Δ24, 25 double bonds are essential for cytotoxicity.

Table 5 Cytotoxicity IC50 values (μM) of compounds 112 against human tumor cell lines

Experimental procedures

General experimental procedures

Optical rotations were recorded on a JASCO P-1020 digital polarimeter (Horiba, Kyoto, Japan). UV/Vis spectra were obtained using a Shimadzu UV2401PC spectrometer (Shimadzu, Kyoto, Japan). CD spectra were tested on an Applied Photophysics Chirascan Circular Dichroism Spectrometer (Applied Photophysics Limited, Leatherhead, Surrey, UK). IR spectra were obtained using a Bruker Tensor 27 FT-IR spectrometer (Bruker Optics, Inc., Billerica, MA, USA) with KBr pellets. One- and two-dimensional NMR spectra were measured on a Bruker Avance III 500 MHz, Bruker Avance III 600 MHz and Bruker Accend 800 MHz spectrometers (Bruker Biospin GmbH, Karlsruhe, Germany). HR–ESI–MS were recorded on an Agilent 6200 Q-TOF MS system (Agilent Technologies, Santa Clara, CA, USA). Silica gel (200–300 mesh, Qingdao Haiyang Chemical Co., Ltd, Qingdao City, China) and Sephadex LH-20 (Amersham Biosciences, Upplasa City, Sweden) were used for column chromatography (CC). Medium-pressure LC was performed on a Büchi Sepacore System equipped with pump manager C-615, pump modules C-605 and fraction collector C-660 (Büchi Labortechnik AG, Fällanden, Switzerland), and columns packed with Chromatorex C-18 (40–75 mm, Fuji Silysia Chemical Ltd, Kasugai, Japan). Preparative HPLC was performed on an Agilent 1260 LC system equipped with two types of Zorbax SB-C18 columns (9.4 mm × 150 mm and 21.2 mm × 150 mm, particle size 5 mm).

Fungal material

The fruiting bodies of Stereum sp. were collected in October 2013 from Xishuangbanna Tropical Botanical Garden Chinese Academy of Sciences and identified by Professor Yu-Cheng Dai (Beijing Forestry University). A voucher specimen (deposition no.: HPC 20131022) has been deposited at the Herbarium of Kunming Institute of Botany, Chinese Academy of Sciences.

Extraction and isolation

The air-dried and powdered fruiting bodies of Stereum sp. (848 g) were macerated four times with 95% ethanol to afford crude exact (191 g). The crude extract was suspended in water (800 ml) and partitioned with EtOAc three times to obtain the EtOAc fraction (45 g). The EtOAc fraction was subject to medium-pressure LC with a stepwise gradient elution of MeOH/H2O (v/v 40:100–100:0) to afford eight fractions (A–H). Fraction F (12 g) was applied to medium-pressure LC with isocratic elution (MeOH/H2O, 80:20) to obtain seven subfactions (F1–F7) based on TLC analysis. Subfraction F2 was separated by Sephadex LH-20 (methanol) CC to give two fractions (F2a and F2b). Fraction F2a was separated repeatedly by semipreparative HPLC (CH3CN/H2O, 40:60 to 70:30, 30 min, 7 ml min−1) to yield 8 (2.5 mg), 10 (9 mg), 11 (15.8 mg) and 12 (3 mg). Subfraction F3 was separated by Sephadex LH-20 (methanol) CC to give four fractions (F3a–F3d) based on TLC analysis. Fraction F3a was purified by Sephadex LH-20 (acetone) CC and then was separated by semipreparative HPLC (CH3CN/H2O, 45:55 to 69:31, 30 min, 7 ml min−1) to afford 9 (3.1 mg). Fraction F3b was purified by Sephadex LH-20 (acetone) CC and then was separated by semipreparative HPLC (CH3CN/H2O, 50:50 to 71:29, 30 min, 7 ml min−1) to afford 2 (6.2 mg), 3 (4 mg) and 4 (2.3 mg). Subfraction F4 was separated by Sephadex LH-20 (acetone) CC to yield two fractions (F4a and F4b) based on TLC analysis. F4b was subject to a silica gel column with petroleum ether-acetone gradient solvent system (v/v, 10:1 to 2:1) to afford four fractions (F4b1–F4b4). Fraction F4b2 exhibited interesting spots in the TLC, which reacted with anisaldehyde-sulfuric acid and next was separated by preparative HPLC (CH3CN/H2O, 62:38 to 86:16, 30 min, 18 ml min−1), further purified by semipreparative HPLC (CH3CN/H2O, 62:38 to 86:14, 30 min, 7 ml min−1) to afford 1 (3.6 mg) and 6 (2.1 mg). Subfraction F5 was purified by Sephadex LH-20 (acetone) CC and then was separated by semipreparative HPLC (CH3CN/H2O, 50:50 to 71:29, 30 min, 7 ml min−1) to afford 5 (4.5 mg). Subfraction F7 was subject to column chromatography over silica gel, eluting with isocratic petroleum ether-acetone (9:1) to give two fractions (F7a and F7b). Fractions F7b were purified by semipreparative HPLC (CH3CN/H2O, 70:30 to 100:0, 30 min, 7 ml min−1) to obtain 7 (3.5 mg).

Sterenoid A ( 1): White solid; [α]D20 +99.8 (c 0.18, MeOH); UV (MeOH) λmax (log ɛ) 202 (2.89), 240 (3.63) nm; electronic CD (MeOH) λ (Δɛ) 216 (−3.8), 240 (+11.5) 324 (−0.6) nm; IR (KBr) vmax 3437, 2962, 2877,1696, 1630, 1460, 1383, 1012 cm−1; 1H and 13C NMR data see Tables 1 and 2; (+)-HRESIMS m/z 477.3350 [M+Na]+ (calcd for C30H46O3Na, 477.3339).

Sterenoid B ( 2): White solid; [α]D20 +57.3 (c 0.31, MeOH); UV (MeOH) λmax (log ɛ) 202 (3.92), 240 (1.17) nm; IR (KBr) vmax 3445, 2957, 2931, 2867, 1702, 1458, 1376, 1251, 1111, 1034 cm−1; 1H and 13C NMR data see Tables 1 and 2; (+)-HR–ESI–MS m/z 479.3488 [M+Na]+ (calcd for C30H48O3Na, 479.3496).

Sterenoid C ( 3): White solid; [α]D20 +49.6 (c 0.08, MeOH); UV (MeOH) λmax (log ɛ) 204 (5.24), 253 (1.24) nm; IR (KBr) vmax 3445, 2958, 2930, 2867, 1701, 1457, 1380, 1251, 1111, 1036 cm−1; 1H and 13C NMR data see Tables 1 and 2; (+)-HR–ESI–MS m/z 479.3488 [M+Na]+ (calcd for C30H48O3Na, 479.3496).

Sterenoid D ( 4): White solid; [α]D20 +11.5 (c 0.20, MeOH); UV (MeOH) λmax (log ɛ) 202 (2.43), 252 (2.14) nm; IR (KBr) vmax 3435, 2964, 2874, 1707, 1665, 1458, 1381, 1247, 1113, 1034 cm−1; 1H and 13C NMR data see Tables 1 and 2; (+)-HR–ESI–MS m/z 455.3527 [M+H]+ (calcd for C30H47O3, 455.3520).

Sterenoid E ( 5): White solid; [α]D20 +44.4 (c 0.27, MeOH); UV (MeOH) λmax (log ɛ) 203 (3.64) nm; IR (KBr) vmax 3431, 3440, 295, 293, 2868, 1702, 1635, 1460, 1382, 1281, 1125, 1079 cm−1; 1H and 13C NMR data see Tables 1 and 2; (+)-HR–ESI–MS m/z 521.3606 [M+Na]+ (calcd for C32H50O4Na, 521.3601).

Sterenoid F ( 6): White solid; [α]D20 +70.7 (c 0.03, MeOH); UV (MeOH) λmax (log ɛ) 203 (2.27) nm; IR (KBr) vmax 3443, 2955, 2933, 2873, 1706, 1632, 1459, 1382, 1118, 1084 cm−1; 1H and 13C NMR data see Tables 1 and 2; (+)-HR–ESI–MS m/z 479.3492 [M+Na]+ (calcd for C30H48O3Na, 479.3496).

Sterenoid G ( 7): White solid; [α]D20 +14.4 (c 0.26, MeOH); UV (MeOH) λmax (log ɛ) 203 (2.23) nm; IR (KBr) vmax 3430, 2955, 2930, 2867, 1707, 1630, 1455, 1383, 1273, 1062 cm−1; 1H and 13C NMR data see Tables 3 and 4; (+)-HR–ESI–MS m/z 465.3705 [M+Na]+ (calcd for C30H50O2Na, 465.3703).

Sterenoid H ( 8): White solid; [α]D20 +15.5 (c 0.42, MeOH); UV (MeOH) λmax (log ɛ) 203 (1.93) nm; IR (KBr) vmax 3440, 2957, 2936, 1708, 1630, 1457, 1380, 1158, 1062 cm−1; 1H and 13C NMR data see Tables 3 and 4; (+)-HR–ESI–MS m/z 481.3653 [M+Na]+ (calcd for C30H50O3Na, 481.3652).

Sterenoid I ( 9): White solid; [α]D20 +22.7 (c 0.31, MeOH); UV (MeOH) λmax (log ɛ) 203 (2.70), 240 (1.11) nm; IR (KBr) vmax 3428, 2962, 2932, 2868, 1716, 1631, 1456, 1374, 1151, 1030 cm−1; 1H and 13C NMR data see Tables 3 and 4; (+)-HR–ESI–MS m/z 497.3403 [M+K]+ (calcd for C30H50O3K, 497.3403).

Sterenoid J ( 10): White solid; [α]D20 +6.7 (c 0.41, MeOH); UV (MeOH) λmax (log ɛ) 202 (3.92) nm; IR (KBr) vmax 3436, 2962, 2931, 2868, 1703 1631, 1459, 1380,1251, 1149, 1031 cm−1; 1H and 13C NMR data see Tables 3 and 4; (+)-HR–ESI–MS m/z 479.3493 [M+Na]+ (calcd for C30H48O3Na, 479.3496).

Sterenoid K ( 11): White solid; [α]D20 +13.0 (c 0.40, MeOH); UV (MeOH) λmax (log ɛ) 203 (3.92), 240 (1.17) nm; IR (KBr) vmax 3440, 2956, 2933, 2868, 1701, 1633, 1460, 1381, 1281, 1125, 1078 cm−1; 1H and 13C NMR data see Tables 3 and 4; (+)-HR–ESI–MS m/z 497.3601 [M+Na]+ (calcd for C30H50O4Na, 497.3603).

Sterenoid L ( 12): White solid; [α]D20 + 34.7 (c 0.45, MeOH); UV (MeOH) λmax (log ɛ) 203 (1.76), 240 (1.21) nm; IR (KBr) vmax 3443, 2956, 2932, 2866, 1700, 1632, 1461, 1382, 1281, 1124, 1077 cm−1; 1H and 13C NMR data see Tables 3 and 4; (+)-HR–ESI–MS m/z 497.3600 [M+Na]+ (calcd for C30H50O4Na, 497.3601).

Cytotoxicity assays

The human tumor cell lines HL-60, SMMC-7721, A-549, MCF-7 and SW-480 were used in the cytotoxic assay. These cell lines were obtained from ATCC (Manassas, VA, USA). Cells were cultured in RMPI-1640 or DMEM medium (Biological Industries, Kibbutz Beit-Haemek, Israel) supplemented with 10% fetal bovine serum (Biological Industries) at 37 °C in a humidified atmosphere with 5% CO2. The cytotoxicity assay was evaluated by theMTS, inner salt (Promega, Madison, WI, USA) assay.37 Briefly, cells were seeded into each well of a 96-well cell culture plate. After 12 h of incubation at 37 °C, the test compound (40 μM) was added. After incubated for 48 h, cells were subjected to the MTS assay. Compounds with a growth inhibition rate of 50% were further evaluated at concentrations of 0.064, 0.32, 1.6, 8 and 40 μM in triplicate, with cisplatin and paclitaxel (Sigma, St. Louis, MO, USA) as positive controls. The IC50 value of each compound was calculated with the method of Reed and Muench.38