In recent years, marine sponges have emerged as one of the most prolific sources for discovery of novel secondary metabolites [1,2,3]. Some of these have been found to possess several kinds of biological activities, such as cytotoxic [4,5], antimicrobial [6], and antiviral properties [7]. 4-Methylenesterols were first reported in 1981 as being obtained from two marine sponges, Theonella conica and Theonella swinhoei, from the Red Sea [8]. Marine sponges of the genus Theonella are now known to be a rich source of novel 4-methylene sterols [8,9,10,11,12,13,14,15,16,17,18,19,20,21]. Some of these sterols display a variety of biological activities, and pharmacological evaluation has shown them to be modulators of two well-known nuclear receptors, FXR and PXR [16,17,18,19,20,21], and cytotoxic activity [9]. Furthermore, a recently published report indicates that theonellasterol (4) is a highly selective FXR antagonist that protects against cholestatic liver injury [21]. Research into the pharmacological properties of this class of natural products is of particular interest. Our investigation of the chemical constituents of the sponge Theonella swinhoei (Figure 1) yielded three new 4-methylene sterols, theonellasterol K (1), acetyltheonellasterol (2) and acetyldehydroconicasterol (3), along with two known sterols, theonellasterol (4) [8] and theonellasterone (5) [10] (Chart 1). The structures of 1–5 were established by detailed spectroscopic analysis, including extensive examination of 2D NMR (¹H-¹H COSY, HMQC and HMBC) correlations. The cytotoxicity of metabolites 1–5 against human colorectal carcinoma (DLD-1), human hormone-dependent breast cancer (T-47D), human colon adenocarcinoma (HCT-116), human breast adenocarcinoma (MCF-7 and MDA-MB-231), human chronic myelogenous leukemia (K562) and human T lymphoblast, acute lymphoblastic leukemia (Molt 4) was studied in order to discover bioactive compounds.

The EtOAc extract of the freeze-dried specimen was fractionated by silica gel column chromatography and the eluted fractions were further separated utilizing normal phase HPLC to yield metabolites 1–5. The new compounds were given the trivial names theonellasterol K (1), acetyltheonellasterol (2) and acetyldehydroconicasterol (3). The known compounds were identified as theonellasterol (4) and theonellasterone (5).

Theonellasterol K (1) was obtained as a white amorphous solid. The HRESIMS spectrum of 1 exhibited a pseudomolecular ion peak at m/z 481.3659 [M + Na]⁺, which established a molecular formula of C₃₀H₅₀O₃, implying six degrees of unsaturation. IR absorptions were observed at 3362 and 3251 cm⁻¹, suggesting the presence of hydroxy groups in 1. The structure of this compound was deduced from its ¹³C NMR and DEPT spectroscopic data (Table 1), which showed that the compound has 30 carbons, including six methyls, ten sp³ methylenes, one sp² methylene, seven sp³ methines (including one oxymethine), one sp² methine, three sp³ quaternary carbons and two sp² quaternary carbons. From the ¹H NMR spectrum of 1, resonances of one olefinic methine proton (δ 5.68, d, J = 5.5 Hz), two olefinic methylene protons (δ 5.18 and 4.74, each s) and one oxygenated methine (δ 4.01, dd, J = 11.0, 5.0 Hz) were observed. Moreover, the ¹H NMR spectrum revealed the presence of one hydroperoxy proton resonating as a broad singlet at δ_H 6.79. The planar structure and all of the ¹H and ¹³C chemical shifts of 1 were elucidated by 2D NMR spectroscopic analysis, in particular ¹H-¹H COSY and HMBC experiments (Figure 2). From the ¹H-¹H COSY correlations, it was found that one ring-juncture methine proton H-5 (δ 1.97) and one oxymethine proton H-3 (δ 4.01) exhibited allylic correlations with the exomethylene protons at C-30 (δ_H 4.74 and 5.18). In addition, ¹H-¹H COSY spectral analysis established five partial structures of consecutive proton spin systems (Figure 2). Further analysis of the HMBC correlations was employed successfully to establish the gross structure of 1 (Figure 2). Thus, 1 was found to possess two double bonds at C-7/C-8 and C-4/C-30, one hydroxy group at C-3 and one hydroperoxy group at C-14.

The relative configuration of 1, elucidated mainly from the NOESY spectrum, was compatible with that of 1 ascertained using molecular mechanics calculations (MM2), which suggested the most stable conformations, as shown in Figure 3. In the NOESY spectrum of 1, the NOE correlations between one of the methylene protons at C-11 (δ_H 1.47) and both methyls (H₃-18 and H₃-19); H₃-18 and H-20 as well as between H₃-18, H₃-19 and H-20 indicated that these protons adapt a β-orientation. H-5 was found to interact with H-3, but not with H₃-19, revealing the β orientation of the hydroxy group at C-3. Moreover, the α-orientation of 14-OOH was further confirmed by the lower field chemical shift of H-17 (δ_H 1.85, m). Furthermore, the chemical shifts of the side chain from C-20 to C-29 in 1 were nearly identical to those of 4 and 5. Thus, the structure of steroid 1 was established. After determining the structure of 1, we discovered that its molecular framework has been obtained as a known 4-methylenesterol conicasterol H, which was isolated previously from sponge Theonella swinhoei [18].

Acetyltheonellasterol (2) was isolated as a white powder with the molecular formula C₃₂H₅₂O₂, which possesses seven degrees of unsaturation, as indicated by HRESIMS (m/z 491.3862, [M + Na]⁺) and NMR spectroscopic data (Table 1). By comparison of the NMR data of 2 with those of 4, it was found that the ¹H and ¹³C NMR data of 2 were very similar to those of 4, with the difference that 2 contains one more acetyl group relative to 4. The chemical shift of H-3 in 4 (δ_H 4.01) was shifted downfield (δ_H 5.15) in 2, suggesting that 2 is the 3-acetyl derivative of 4. We observed further that acetylation of 4 gave a product which was found to be identical to 2 by comparison of the physical and spectroscopic data. Thus, compound 2 was established as the 3-acetyl derivative of 4.

The new metabolite acetyldehydroconicasterol (3) was obtained as a white powder and possessed the molecular formula C₃₁H₄₈O₂, as established from the HRESIMS and NMR data, implying eight degrees of unsaturation. Both the ¹H and ¹³C NMR signals of 3 were found to be very closely related to those of compound 2, suggesting a very similar steroidal skeleton. By comparison of the NMR data of 3 with those of 2 (Table 1), it was found that an ethyl protons signal [δ_H 0.86 t, J = 7.5 Hz (H₃-29); 1.16 m and 1.32 m (H₂-28)] in 2 was replaced by two exomethylene proton signals (δ_H 4.73 and 4.67, each s) in 3. The only difference observed was that the 24-ethyl side chain of 2 was replaced by two exomethylene proton signals in 3. This was further confirmed by the HMBC correlations from H₂-28 to C-23, C-24, and C-25. On the basis of the above analysis, the structure of 3 was established.

Finally, we used a 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay to examine the cytotoxic activities of compounds 1–5 against DLD-1, T47D, HCT-116, MCF-7, MDA-MB-231, K562 and Molt 4 cancer cells. Cells were treated with different concentrations of 1–5 for 72 h. The results showed that compound 1, the most potent of compounds 1–5, exhibited cytotoxicity against DLD-1, T47D, HCT-116, MDA-MB-231, K562 and Molt 4, with IC₅₀s of 12.9, 12.0, 6.3, 11.5, 11.4, 4.3, and 6.3 μg/mL, respectively. Furthermore, compound 2 exhibited weak cytotoxic activity against K562 and Molt 4 cancer cell lines (the IC₅₀ values were 13.7 and 17.8 μg/mL for K562 and Molt 4, respectively). The other tested compounds were not cytotoxic (IC₅₀ > 20 μg/mL) towards the above seven cancer cell lines (Table 2).

A series of new 4-methylenesterols were isolated from the sponge Theonella swinhoei [8,9,10,11,12,13,14,15,16,17,18,19,20,21]. Our continued investigation of the chemical constituents of sponge T. swinhoei has again led to the isolation of three new 4-methylenesterols (1–3) and two known 4-methylenesterols (4 and 5). Compound 1 exhibited significant cytotoxicity against HCT-116, K562 and Molt 4 cells, and moderate to weak cytotoxicity against DLD-1, MCF-7, and MDA-MB-231 cells. In addition, compound 2 exhibited weak cytotoxicity toward K562 and Molt 4 cell lines. However, the new compound 3 and the other known compounds had no significant activity. According to our research, metabolite 1 exhibited significant cancer cells inhibitory activity. This result suggests that 1 should be subjected to further biomedical investigation.