New cytotoxic constituents from the Red Sea soft coral Nephthea sp.

Nephthea are soft coral species rich in sesquiterpenoids and steroids. An organic extract of Nephthea sp. resulted in the isolation of a new steroid (1), as well as several previously reported metabolites (2–9). Structures were elucidated by employing NMR and HR-EI-MS analyses. The total extract, fractions and purified compounds exhibited differential cytotoxicity against the breast cancer MCF-7 cell line.


Introduction
The Red Sea which is arguably the world's warmest (up to 358C in summer) and most saline marine habitat (ca. 40 psu in the northern Red Sea) contains extensive reef formation (Edwards & Head 1987). Despite the sea's size and diversity of reef-associated inhabitants, marine invertebrates associated with this ecosystem remain poorly studied compared to other coral reef systems around the world. The genus Nephthea (Acyonaceae, subfamily Nephtheidae) is present in the Red Sea and is a rich source of sesquiterpenoids and steroids. In addition, the genus produces metabolites that have been shown to possess diverse biological properties including cytotoxic (Duh et al. 1999;Januar et al. 2010;Liang et al. 2010), antimicrobial and antiinflammatory activities (Cheng et al. 2009). As part of the project to chemically and biologically characterise metabolites from the Red Sea soft corals, steroids (1-6), sesquiterpenes (7 and 8) and a ceramide (9) were isolated and identified from an organic extract of Nephthea sp. with isolated compounds chemically characterised by spectroscopic methods. Compound purification was guided by a breast cancer toxicity assay using cell line MCF-7 since a previous study has shown Nephthea compound efficacy against this cancer line.

Results and discussion
A dichloromethane -methanol (1:1) extract of freshly collected Nephthea sp. was subjected to normal and reversed phase chromatography to afford nine metabolites with 1 being a newly reported natural product (Figure 1 1H, m) and one oxymethylene at d H 3.54 (1H, d, J ¼ 11.7 Hz) and 3.66 (1H, d, J ¼ 11.7 Hz). The 13 C NMR and DEPT spectra also exhibited six methyl at d C 12.6(q), 19.9(q), 18.8(q), 19.1(q), 16.8(q) and 14.4(q); two olefinic at at d C 135.4(d), and 132.1(d), and three OH bearing carbon at d C 76.0(d), 71.5(s), and 62.1(t), which was confirmed by HR-ESI-FT-MS analysis. NMR comparisons with nebrosteroid-C, reported from Nephthea chabroli (Huang et al. 2008), indicated compound similarities with 1 except for a disappearance of an acetate group as well as a combined keto group loss/exomethylene to an endo-olefin shift. A loss of downfield H-18 signals originally observed in nebrosteroid-C occurred with a concurrent appearance of oxygenated methylene protons at d H 3.54 (d, J ¼ 11.7 Hz) and 3.66 (d, J ¼ 11.7 Hz); these methylene protons correlated with d C 38.6 (C-12), 46.5 (C-13) and 57.0 (C-17) in heteronuclear multiple bond coherence (HMBC), indicating an acetate/hydroxyl group substitution in 1.
Postion of the secondary hydroxyl group was based on HMBC correlation between doublet signal at d H 0.93 (J ¼ 6.2 Hz, H-29) and the oxygenated carbon at d C 76.0, allowing for the assignment of C-3. Postion of the third hydroxyl group was based on correlation of multiplet signals d H 1.20 and 1.80 (H 2 -7) with quaternary oxygenated signal at d C 71.5, allowing for the assignment of C-8. Complete NMR assignments (Table S1) were determined by analyses of DQF-COSY, HMQC and HMBC data ( Figure S1), establishing the structure of 1.
The relative stereochemistry of 1 was deduced from NOESY data in which identified alpha protons as 1 and 5 correlated with H-3 allowing for the beta hydroxyl assignment at C-3. In addition, correlations between H 3 -19/H-4 and H 3 -19/H-6b indicated the b configuration of H 3 -19 and a configuration of H 3 -29. The coupling constant between H-22 and H-23 (J ¼ 15.1 Hz) suggested the double bond to have E configuration. From all available data, 1 was deduced to be 4a,24-dimethyl-5a-cholest-8b,18-dihydroxy, 22E-en-3b-ol (1).
Compound 2 was obtained as a colourless oil with an optical rotation ([a ] D 25 þ 44.0 in MeOH). HR-ESI-FT-MS analysis showed a molecular ion peak at m/z 444.3635 [M] þ corresponding to a molecular formula of C 29 H 50 O 3 (calcd. 444.3539).
The 13 C NMR and DEPT spectra revealed the presence of 6 methyls, 10 methylenes (one of them olefinic), 8 methines (one of them oxygenated, d C 76.0) and 5 quaternary carbons (one of them oxygenated, d C 72.7, one olefinic d C 155.9, one keto d C 203.3). NMR comparison of 2 with those of nebrosteroid-C, published by Huang et al. (2008), indicated that the disappearance of downfield oxygenated methylene protons of nebrosteroid-C and the appearance of additional methyle group in 2.
HMBC correlations were observed between the carbonyl carbon signal C-23 (  , suggesting a spin system containing H22 -H20 -H21. On the other hand, location of CH 3 -29 group was assaigned to be at C-4 which was supported by HMBC correlations between d H 0.92 (H 3 -29) and d C 76.0 (C-3)/51.9 (C-5). Correlation between H 3 -18 and H 3 -19 with signal at d C 72.7 allowed the assignment quaternary hydroxyl group to be at C-8. From all the above data, 2 was deduced to be 4a-methyl-3b,8b-dihydroxy-5a-ergost-24(28)-en-23-one with 13 C NMR data recorded here for the first time (Bortolotto et al. 1976).
From previous studies about cytotoxicity against breast cancer MCF-7 cell line have been studies only for purified diterpenes by Januar et al. ( 2010), with a non-selective activity towards the MCF-7 cell line (GI50 . 100 mM), which encourage us to study the total extract, fractions and purified compounds against breast cancer MCF-7 cell line. As shown in Figure 2, fraction N4 produced the most potent cytotoxicity against MCF-7 cell recording the lowest IC 50 ¼ 37 mg/mL among other tested fractions/compounds. The order of the activity of the rest of the samples is 3 . 4 . 8 . 2 . 1 (Table S1). It can be concluded that compounds 3 and 4 isolated from fraction N-6 have anatagonitic effect when present together as a mixture in fraction N-6, as their individual activities against the MCF-7 cell viability are more potent than the activity displayed by the N-6 fraction. It can also be concluded that compound 8 may synergise with other non-isolated compound(s) in fraction N-4. Similar weak effects (higher IC 50 values) on MCF-7 cytotoxicity were produced by the rest of the tested fractions/compounds (Table S1).

Experimental
3.1. General experimental procedures 1 H and 13 C NMR spectra were recorded in CDCl 3 or CD 3 OD on a JEOL ECA-600 spectrometer (600 MHz for 1 H and 150 MHz for 13 C). All chemical shifts (d) are given in ppm units with reference to TMS as an internal standard and coupling constants (J) are reported in Hz. HR-ESI-FT-MS experiments were performed on HR-ESI-FT-MS: Thermo Fisher Scientific LTQ Orbitrap XL mass spectrometer in m/z. Purification was run on a Agilent HPLC system equipped with an Agilent-G1314 variable wavelength UV detector at 254 nm and compound separation was performed on YMC-Pack ODS-A (250 £ 4.6 mm i.d.) and (250 £ 20 mm i.d.) columns for analytical and preparative separation, respectively. Normal-phase chromatography separation included silica gel 60 (230 -400 mesh, Merck, Darmstadt, Germany). Pre-coated silica gel plates (Merck, Darmstadt, Germany, Kieselgel 60 F 254 , 0.25 mm) were used for TLC analyses. Spots were visualised by heating after spraying with 10% H 2 SO 4 .

Animal material
Soft coral Nephthea sp. was collected from the Egyptian Red Sea off the coast of Hurghada in March 2013.The soft coral was identified by one of the authors Alhammady and a voucher specimen (03RS38) has been deposited in the National Institute of Oceanography and Fisheries, Marine Biological Station, Hurghada, Egypt.

Extraction and separation
Frozen soft coral (6.0 kg, total wet weight) was chopped into small pieces and extracted with dichloromethane -methanol (1:1) at room temperature (4 L £ 5). The combined dichloromethane-methanol extracts were concentrated in vacuo to a brown gum. The dried material (350 g) was subjected to gravity chromatography in a silica gel column (6 £ 120 cm) eluting with n-hexane (4 L) followed by a gradient of n-hexane -CH 2 Cl 2 up to 100% CH 2 Cl 2 and CH 2 Cl 2 -MeOH up to 50% MeOH (4 L each of the solvent mixture) to afford nine major fractions (N-1 to N-9).

Anti-tumour activity
Compound cytotoxicity against MCF-7 cell lines was assayed via a 3-[4,5-dimethylthiazole-2yl]-2,5-diphenyl tetrazolium bromide (MTT) spectrophotometric cell-viability assay. The MTT assay is based on the ability of active mitochondrial dehydrogenase enzyme of living cells to cleave the tetrazolium ring of the yellow MTT to form dark-blue insoluble formazan crystals that are largely impermeable to cell membranes, resulting in its accumulation within healthy cells. With loss of cell viability, formazan crystals leak into the medium. Cell viability is monitored and quantified spectrophotometrically from formazan crystals extracted from the cells (Hansen et al. 1989). In brief, cells are diluted in serum-free media to 5.0 £ 104 cells/well based on turbidity measurements, plated into flat-bottomed 96-well microplates, and incubated for 48 h with 20 mL of test sample. After incubation, the medium is removed and 40 mL MTT (0.5 mg MTT/mL 0.9% aqueous NaCl) is added into each well and then incubated for an additional 4 h. MTT crystals are solubilised by adding 180 mL of acidified isopropanol to each well and plates are agitated at room temperature, followed by absorbance measurements at 570 nm using a microplate ELISA reader (BMG LABTECH, Ortenberg, Germany). Triplicate repeats were performed for each concentration and the average was calculated. Data are expressed as the percentage of relative viability compared with the untreated cells and with the vehicle control. Cytotoxicity was reported as a relative viability compared with the untreated control. Percentare relative viability was determined based on cell viability: (absorbance of treated cells/absorbance of control cells) £ 100.

Statistical analysis
Half-maximal inhibitory concentrations (IC 50 ) of the compounds, fractions and extracts were calculated using regression analysis from the corresponding mean of %viability versus concentration. r 2 values (goodness of fit) of the regression fit are presented in supplementary data.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This project was supported financially by the Science and Technology Development Fund (STDF), Egypt [grant number 1102].