Cytotoxic furanosesquiterpenoids and steroids from Ircinia mutans sponges

Abstract Context Ircinia mutans Wilson (Irciniidae) is a sponge with antimicrobial and cytotoxic constituents. Objective Our objective was to characterise the cytotoxic constituents of two seasonal collections of I. mutans. Materials and methods The sponges were extracted in methanol-dichloromethane and their constituents were purified and characterised using column chromatography, GC-MS, 1 D and 2 D NMR. Anti-proliferative activities of the compounds, were evaluated using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) colorimetric assay (0.25–100 μg/mL, 72 h) against leukaemia (MOLT-4), breast (MCF-7) and colon cancer (HT-29) human cells. Results Three furanosesquiterpoids; furodysin (1), ent-furodysinin (2) and furoircin (3) and ten sterols were characterised in I. mutans, for the first time. Cholesterol (4), cholesta-5, 7-dien-3β-ol (5) and ergosterol (6) were determined in the sponge from the winter collections, while cholesta-5, 22-dien-3β-ol (7), 24-methyldesmosterol (8), campesterol (9), stigmasterol (10), γ-ergostenol (11), chondrillasterol (12) and γ-sitosterol (13) were detected in the summer samples. The steroids from the winter collection exhibited cytotoxic activity with IC50 values of 13.0 ± 0.9, 11.1 ± 1.7 and 1.1 ± 0.4 µg/mL, against the mentioned cancer cell lines, respectively, while those from the summer sample, showed greater activity, IC50 = 1.1 ± 0.2 μg/mL against MOLT-4. The purified steroids showed potent MOLT-4 cytotoxic activity, IC50 values = 2.3–7.8 µg/mL. Discussion and conclusion The present study suggests that I. mutans is a rich source of cytotoxic steroids, and introduces 3 as new natural product. Considering the high cytotoxic activity of the steroids, these structures could be candidates for anticancer drug development in future research.


Introduction
Sponges (Phylum Porifera) are the most primitive multicellular eukaryotic organisms. They are sessile without defense organs and therefore produce different secondary metabolites to protect themselves from predators and pathogens in their environment (Proksch 1994). The genus Ircinia Nardo (Irciniidae) comprises about 73 species which have been widely investigated for drug discovery. This genus of sponge is a rich source of a variety of steroids, terpenoids and other types of secondary metabolites with different biological activities (Faulkner 1973;Emura et al. 2006;Xu et al. 2008;Hahn et al. 2014;Yang et al. 2014).
Several researchers have investigated the effects of different ecological and biological factors on the production of secondary metabolite and announced that variations in the natural products production patterns can be explained by the effect of these biotic and abiotic factors. Sacrist an-Soriano et al. (2012), by examining the temporal trends in the secondary metabolite production of the sponge Aplysina aerophoba Nardo (Aplysinidae), showed that concentrations of bromo-alkaloid compounds in ectosome layer during the summer are more than those in the winter season and that the levels of secondary compounds from the outer layer of the sponge, were significantly correlated with water temperature. Abdo et al. (2007), studied the effect of temperature and spatiotemporal variation on production of salicylihalamide A in the sponge Haliclona sp. Grant (Chalinidae) and found environmental and physiological factors to be important in the production of this compound. They showed that "salicylihalamide A" concentration was significantly correlated with water temperature and was significantly higher in summer than in the winter collections in sponges from Bremer Bay. Turon et al. (1996) suggested that the greater variety of natural compounds in the warmer seasons of the year may have been a response to increased activity of competitors, while Duckworth and Battershill (2001), implicated fouling from other organisms on the sponge surface.
Previously, different solvent extracts of I. mutans were subjected to different bioassays such as antibacterial (Nazemi et al. 2014), antifungal (Nazemi et al. 2013), and cytotoxic against human cancerous cell lines (Jamebozorgi et al. 2018;Nazemi et al. 2020), but to the best of our knowledge there is no report on chemical characterisation and biological evaluation of the isolated compounds from this species in the literature. We report here, for the first time, the structure elucidation of cytotoxic steroids and furanosesquiterpenoids from I. mutans collected from the Persian Gulf during two seasons.

Instrumentation and reagents
NMR spectra of the purified compounds were recorded on a Bruker Avance 500 spectrometer (Bruker Biospin, Karlsruhe, Germany), operating at resonance frequencies of 500 MHz for 1 H and 125 MHz for 13 C, respectively. Standard Bruker pulse sequences were used for measuring 1 H NMR, 13 C APT, 1 H-1 H COSY, 1 H-13 C HSQC and 1 H-13 C HMBC spectra. Tetramethylsilane (TMS) was used as an internal standard for referencing 1 H and 13 C NMR spectra. Data acquisition and processing was accomplished using Bruker Topspin 2.1. EI-MS spectra were recorded on an Agilent 5975 C inert GC/MS instrument. For isolation and purification of the sponge's metabolites, different chromatographic separations were performed, including silica gel open column chromatography (CC: 0.063-0.200 mm particle size), flash column chromatography (FCC: 0.040-0.063 mm particle size) and TLC using silica gel 60 F 254 pre-coated plates (0.25 mm film thickness). The adsorbents were purchased from Merck, Darmstadt, Germany. For further purification of the fractions, reversed-phase (RP-18) HPLC analyses were performed using a Knauer semi-preparative HPLC with a K-1050 pump and a four wavelength K-2600 UV detector set at k 210 nm (Jassbi et al. 2014a). The semi-preparative HPLC column (Phenomenex RP-18, 250 Â 10 mm) was eluted with 95% acetonitrile (solvent B) and 5% in ultrapure water (solvent A). The flow rate of the mobile phase was set at 4.5 mL/min.

Collection of the sponges
Ircinia mutans was collected by scuba diving in January 2015, and June 2016 at a depth of 10-13 m near Larak Island in the Persian Gulf. Samples were placed immediately in plastic bags containing seawater and transferred to the laboratory on ice and then stored at À20 C after other organisms, such as brittle sea star, bivalves, barnacles, oligo-and poly-chaetes, were removed. One of us, M. Nazemi, the marine biologist, identified the sponge sample, based on scanning and optical microscope studies of skeletal slides and dissociated spicule mounts with the sponge taxonomic keys developed by John N.A. Hooper (2000). Parts of the sponge were stored in 70% ethanol and kept in Medicinal and Natural Products Chemistry Research Center, Shiraz University of Medical Science museum, as a voucher specimen; 93-7-1-1/1.

Extraction of the sponge
The sponge sample collected in winter (W) (652 g fresh weight) was cut into small pieces (approximately 1 cm) and extracted by methanol (2 Â 4 L) for 4 days followed by the same volumes of dichloromethane (DCM) continuously at room temperature in the dark to afford residues of MeOH (3.2 g) and DCM (1.4 g) extracts after evaporation of the solvents in reduced pressure at 40 C. The fresh sponge sample (3.29 kg) collected in summer (S) was extracted by MeOH and DCM as described earlier (Jamebozorgi et al. 2019). The solvent of the DCM (8.5 g) extract was evaporated under reduce pressure at 40 C. The methanol extract (500 mL) of the summer sample was subjected to liquidliquid extraction (LLE), using n-hexane (3 Â 250 mL), DCM (3 Â 250 mL), and 1-butanol (3 Â 200 mL) to afford three fractions: n-hexane (5.05 g), DCM (1.92 g), and 1-butanol after evaporation of the solvents in vacuum. On the basis of TLC examination, we choose the hexane fraction for further phytochemical investigation.

Isolation and purification of the chemical constituents
Isolation of the winter collected sponges' extract The MeOH and DCM extracts of the winter samples were mixed based on their similar patterns from the TLC analyses. The combined extracts were subjected to silica gel-FCC (50 Â 4 cm; 100 g). The column was eluted with n-hexane with stepwise increasing the polarity of the mobile phase to ethyl acetate (EtOAc) and then to pure methanol to afford 24 fractions. Compound 1 (240 mg) was obtained as a colourless oil and detected as a UV quenching spot (R f ¼ 0.5) on silica gel F 254 TLC using 100% n-hexane.

Isolation of the summer collected sponges' extract
The hexane layer of the methanol extract of the summer-sponge sample was subjected to FCC (50 Â 4.5 cm) over AgNO 3 -impregnated (7.5 g) silica gel (150 g). The column was eluted with different portions of n-hexane and DCM followed by MeOH to afford 32 fractions. Compound 1-3 were detected on TLC in the hexane rich fractions as described above and were further purified with RP18 semi preparative HPLC with the retention times of 7.48, 7.75 and 8.42 min, respectively ( Figure 1).
Repeated silica gel (45 g)-FCC (50 Â 3 cm) of fractions F6-F10 (300 mg) as described for that of the winter sample yielded a pale-yellow powder (FS6-10-F25-29) after removal of the eluting solvents (98 mg). The fraction was further analysed by GC-MS for characterisation of its constituents; 4-13 ( Figure 2). For further purification of this fraction, reversed-phase (RP-4) HPLC analyses were performed using the same HPLC system described earlier. The HPLC column (Eurospher-100-5 C4, 250 Â 4.6 mm, Knauer, Germany) was eluted with 95% acetonitrile and 5% in ultrapure water. The flow rate of the mobile phase was set at 1 mL/min. The major peaks were collected as FH2 (3.5 mg) FH3 (7.6 mg), FH4 (0.6 mg) and FH5 (0.8 mg) (Figure 3). These fractions appeared to be pure based on analytical HPLC chromatograms and were subjected to GC-MS for their identification and qualitative analyses ( Figure 4) and further cytotoxic tests.
Gas chromatography-mass spectrometry (GC-MS) GC-MS analyses were performed using an Agilent 7890 A GC coupled to HP-6890 mass spectrometer operating in EI mode at 70 eV. The GC was equipped with a DB-5 MS and HP-5 MS (J & W Scientific column, 30 m Â 0.25 mm i.d., 0.25 lm film thickness) for the winter and the summer samples, respectively. To analyse the steroids, the oven temperature was set at 265 C for 20 min, increased to 300 C at 5 C/min and kept for 10 min at the final temperature. Helium (He) was used as the carrier gas with a flow rate of 1 mL/min. The injector temperature was set at 300 C, in the split mode (1:10). The injection volume was 0.1 lL for all of the samples (Jassbi et al. 2013).

Derivitization of the steroids to trimethylsilyl derivatives
In order to quantify the steroid contents in the sub-fractions FW9-10-F51-80 of the winter sample by GC-MS, fraction constituents were transformed to trimethylsilyl derivatives. Briefly, 200 lL of the reagent, N-O-bis-(trimethylsilyl)-trifluoroacetamide (BSTFA) containing 1% trimethylsilyl chloride (TMCS) was added to 2 mg of F9,10 -F51-80 and vortexed well. Afterwards, the mixture was heated in a hot water bath (60 C) for 1 h. Samples were evaporated under a stream of nitrogen gas, redissolved in 200 lL DCM and the resulting solution was subjected to GC-MS as previously described (Jassbi et al. 2013).
In vitro cytotoxic activity of compounds was evaluated by using the MTT reduction assay. MTT is a rapid colorimetric assay that can be used to measure cytotoxicity, proliferation or activation of the cells (Mosmann 1983). The purified compounds were dissolved in DMSO and isopropanol, and then diluted in growth medium at least 400 times. Cells were seeded in 96-well plates at the density of 50,000 cells/mL (MOLT-4 cell line), 25,000 cells/mL (MCF-7 cell line) and 10,000 cells/mL (HT-29 cell line) in 100 lL medium and incubated for 24 h at 37 C. Then, four different concentrations of each fraction and cisplatin (as positive controls) were added to the wells of the plates in triplicate and they were incubated at 37 C for 72 h. Afterwards, the media (80 mL) of each well was replaced by 80 mL MTT solution diluted in RPMI without phenol red (concentration of 0.5 mg/mL) and incubated at 37 C. After 4 h incubation and formation of formazan crystals, the media was removed and 200 lL DMSO was added to each well to dissolve the crystals. Finally, absorbance was measured at a wavelength of 570 nm with background correction at 650 nm using a microplate reader (model 680, Bio-Rad, Japan) and IC 50 (concentration that results in 50% inhibition of cell viability) for each compound was calculated with Curve Expert statistical analyses (Jassbi et al. 2014b).  Table 1.

Spectroscopic data of the compounds
The spectroscopic data of previously described compounds are presented in the supplemental material.

Statistical analysis
The normality of data and the homogeneity of variances were surveyed with Kolmogorov-Smirnov and Leven's tests, respectively, and log10 transformation of some data was necessary to achieve analysis requirement. Then, the effect of each fraction on cell line were compared by means of one-way ANOVAs, and the mean comparison was conducted with the Duncan's test at 5% significant level for steroid's sub fractions. To compare the effect of steroids extracted from two seasons on MOLT-4 cell line, the independent sample t-test was used between the mean activity of winter and summer samples, where the assumption of equal variances was met. Significance levels in all tests was 5% (p < 0.05) and data were expressed as mean ± SD. Data were analysed using SPSS statistical software (USA, release 16).

Results
The DCM-MeOH extracts of the winter and summer sponge samples were subjected to repeated FCC on silica gel and HPLC to yield compounds 1-3 ( Figure 1) and 4-13 (Figure 2). The chemical structures of the compounds were elucidated using different spectroscopic data including 1 D and 2 D NMR, EI-MS, UV and GC-MS and comparing these data with those reported in the literature. Compound 3 was previously synthesised as a precursor of 1 (Vaillancourt et al. 1991), but it is a new natural product and its NMR spectral signals were not fully assigned.

Structure elucidation of compound 3
In the 1 H NMR, three signals at d 7.34 (brt, J ¼ 1.8 Hz), 6.24 (brd, J ¼ 1.7 Hz), and d 7.17 (s) suggested a mono-substituted furan ring in the molecule. An isopropenyl group represented by  two broad singlet signals at d 4.87 and 4.72 and an olefin methyl at d 1.77. The signals of a tri-substituted cyclohexene moiety were similar to those reported for furodysin represented by an olefin methyl group signal at d 1.63 which is coupled with an allylic methine at d 5.39 (br.dd, J ¼ 1.8, 3.8 Hz), together with two methylene at d 1.68 and 2.0 and two methine signals at d 2.41 and 2.24. A methylene signal at d 2.38 (dd, J ¼ 12.6, 3.9 Hz) and 2.07 (ddd, J ¼ 12.7, 12.7, 2.6 Hz) attached the cyclohexene ring to the furan ring. Using the APT 13 C NMR and HSQC spectra, 15 signals representing four CH 2 , six CH, two CH 3 and three quaternary carbon atoms suggested the structure of a sesquiterpene for compound 1. The connectivity of the proton signals to their respective attached carbon atoms were confirmed by an HSQC experiment. Three methine and one quaternary carbon signals at d 142.5, 111.5, 139.4 and 123.8, respectively, represented the mono-substituted furan ring, the signals of the isopropenyl resonated at d 22.8 (CH 3 ), 110.1 (CH 2 ) and 147.6 (C). In the HMBC spectrum the cross peaks between H-12 and C-3 and C-11 confirmed the connectivity of the furan and cyclohexene rings via C-12, and the cross peaks between H-6 and C-5 and H-15 and C-9 and C-10 determine the position of the methyl and isopropenyl group on the cyclohexene ring. Two small (J ¼ 3.8 Hz) and one big (J ¼ 11.9 Hz) coupling constants for H-6 at d 2.24 and a weak cross peak between H-14 and H-12 in the NOESY spectrum was compatible with the cis configuration at C-6 and C-11 substitution (Table 1).

Steroids from winter sample (compounds 4-6)
GC-MS and 13 C NMR analyses of sub fraction FW9-10-F51-80 of winter sample revealed that this sponge is a rich source of C 27 and C 28 sterols. The sub-fraction, despite its TLC analysis suggesting its purity, consisted of three major D 5 -steroids including; cholest-5-en-3b-ol (cholesterol) (4), cholesta-5, 7-dien-3b-ol (5) and ergosta-5, 7, 22-trien-3b-ol (ergosterol) (6) (NIST Chemistry Web Book, Pub Chem.ncbi and The Pherobase Tools). The structures of the steroids were elucidated by GC-MS and 13 C NMR spectral data analyses, which were confirmed by comparison to those previously reported in the literature for the authentic compounds. The presence of D 5 -en, D 5,7 -dien and D 5,7,22trien were confirmed by the 13 C-NMR spectra which exhibited signals in the sp 2 carbon region (d 116.25-141.40). The chemical shifts of the 13 C NMR signals of the sub-fraction FW9-10-F51-80 (4-6) are consistent with the spectroscopic data previously reported for the above mentioned compounds (Batta et al. 1995;Plouguern e et al. 2006;Zhao et al. 2012).
Steroids from the winter sample FW9-10-F51-80 showed potent cytotoxic activity against all three tested cell lines with IC 50 values of 13.0 ± 0.9, 11.1 ± 1.7 and 1.1 ± 0.4 lg/mL against the mentioned cell lines, respectively. Steroids from summer sample FS6-10-F25-29 just was tested against MOLT-4 cell lines and showed strong activity with an IC 50 value of 1.1 ± 0.2 lg/mL. The cytotoxic activity of the steroids purified from the summer sample, FH2, FH3, FH4 and FH5 were evaluated against MOLT-4 cell line and showed that all steroids are potent cytotoxic compounds with IC 50 values of 7.8 ± 0.2, 3.1 ± 0.7, 2.2 ± 0.3 and 2.3 ± 1.1 lg/mL, respectively ( Figure 6).

Discussion
Furodysin (1) was previously isolated from genus Dysidea and its structure was determined by single crystal x-ray crystallography (Kazlauskas et al. 1978). The full NMR spectroscopic data assignment of the ent-furodysin enantiomer, isolated from Dysidea fragilis, was performed by 1 D and 2 D NMR spectroscopy and found to be different in their 1 H and 13 C NMR d values and optical rotation with those reported for compound 1 (Su et al. 2010). The structure of compound 2 isolated from summer sample was determined based on 1 H and 13 C NMR spectra and comparing them with those reported previously from a China Sea sponge, Dysidea fragilis (Su et al. 2010). The 1 H and 13 C NMR data of 1 and 2 are reported in the supplementary file of the paper. Compound 3 has similar pattern of fragmentation in the webbook.nist.gov Ã Column type: VF-5MS; DB-5; HP-5 MS: 5% phenylmethyl polysiloxane, OV-1, DB1 ¼ 100% methylsiloxane Table 3. Sterol composition of the HPLC Purified fractions from FS6-10-F25-29, characterised by GC-MS.
Since Larak Island (26 51 0 12 00 N and 56 21 0 20 00 E) is located in a subtropical region, where the water temperature during winter (18 C) is substantially lower than of the summer (29 C), we can infer that the greater variety of isolated steroids and furanosesquiterpenes from the summer sample is related to water temperature, which of course is likely also correlated with great fouling and competition from other organisms.
In this research the steroids showed stronger cytotoxic activity than did the furanosesquiterpenes. Although steroids obtained from both seasons showed strong cytotoxic activity, it should be noted that the steroid fraction from the summer sample, which not only contains more diverse D 5 and D 5,7 steroids, but also comprises D 7 sterols, showed stronger cytotoxic activity (IC 50 values 1.1 ± 0.2 lg/mL) than the winter sample (IC 50 values 13.0 ± 0.9 lg/mL) against MOLT-4 cell lines (p < 0.05). Various researches showed that environmental variations can affect the bioactivity of natural compounds. Turon et al. (1996) found clear annual pattern in toxicity of Crambe crambe Schmidt (Crambeidae) and showed the highest biological active in late summer and autumn, similar to the results reported here. Cholesterol with IC 50 value 3.1 ± 0.7 lg/mL was one of the potent cytotoxic steroids which extracted from Persian Gulf sponge, I mutans. Kellner-Weibel et al. (1999), showed that free cholesterol generated by the hydrolysis of cytoplasmic cholesteryl esters in model macrophage foam cells, transported to the plasma membrane by acidic vesicles, and accumulation of this compound in the pool caused cell death by necrosis and apoptosis. c-Sitosterol, IC 50 value 2.3 ± 1.1 lg/mL was the other potent cytotoxic steroid which isolated from I. mutans. Sundarraj et al. (2012), investigated the inhibitory effect of Acacia nilotica Delile (Fabaceae) leaves extract and c-sitosterol on cell proliferation, apoptosis and cell cycle arrest in breast and lung cancer cells, and showed that c-sitosterol with induces G2/M cell cycle arrest and apoptosis through c-Myc suppression in MCF-7 and A549 cells was a potent anticancer agent.
Generally, we showed that steroids from I. mutans have strong cytotoxic activity on MOLT-4, MCF-7 and HT-29 cell lines with IC 50 value <10 mg/mL. The steroids composition of the sponge, I. mutans is similar to those of its neighbouring sponge, collected from the same location, A. sinoxea. Both fractions and purified compounds have the same range of activity against the mentioned cell lines with IC 50 values of 1.20-4.12 lg/ mL (Jamebozorgi et al. 2019). Therefore, our present results confirmed the previous findings that suggested steroids from sponges as potent cytotoxic agents.

Conclusions
The Persian Gulf sponge, I. mutans, is a rich source of valuable secondary metabolites. Steroids are one of the dominant groups of secondary metabolites found in I. mutans and they show strong cytotoxic activity against three human cancer cell lines. Therefore, they can be considered as candidates for further investigation as anticancer agents. Also, the study of the extracted metabolites of this sponge in the winter and summer season showed that both the diversity and cytotoxic activity of metabolite extracted in the summer was greater than winter season, However, additional sample collections and statistical analyses are needed to verify these differences in future research. Finally, compound 3 or its proper isomer may be considered as intermediates in the biosynthetic pathway of compounds 1 and 2, respectively. Based on statistical analysis, cytotoxic activity of FH2 has significant difference with other fractions (p < 0.05) while FH4 and FH5 were not significantly different compared to that of cisplatin as a positive control (p > 0.05).