Numerosol A-D, new cembranoid diterpenes from the soft coral Sinularia numerosa.

Four new cembrane-type diterpenes; numerosol A-D (1-4); along with a known steroid; gibberoketosterol (5); were isolated from the Taiwanese soft coral Sinularia numerosa. The structures of these metabolites were determined by extensive analysis of spectroscopic data. Gibberoketosterol (5) exhibited cytotoxicity against P-388 (mouse lymphocytic leukemia) cell line with an ED50 of 6.9 μM.

1. Introduction

Soft corals belonging to the genus Sinularia have proven to be a rich source of diterpenes, sesquiterpenes, steroids, steroidal glycosides, sphingosine derivatives, glycolipids, and spermidine derivatives [1,2]. We have previously discovered a series of farnesyl quinols [3], sesquiterpenoids [4], norditerpenoids [5], diterpenoids [6,7], biscembranoid [8], and furanosesquiterpenoids [9] from Sinularia sp. Our recent investigation of natural metabolites from the soft coral Sinularia numerosa (Tixier-Durivault, 1970) (Figure 1), has led to the isolation of four new cembranoids, numerosols A–D (1–4) (Figure 2) and a known steroid, gibberoketosterol (5) [10]. The structures of numerosols A–D (1–4) were established by extensive spectroscopic analysis. The anti-human cytomegalovirus (anti-HCMV) activity of compounds 1–5 and its cytotoxicity against P-388 (mouse lymphocytic leukemia), HT-29 (human colon adenocarcinoma), and A-549 (human lung carcinoma) cancer cell lines were evaluated in vitro.

Figure 1

Soft coral Sinularia numerosa.

Figure 2

Structures of Metabolites 1–5 *.

* Structures 2–4 are relative structures.

2. Results and Discussion

Numerosol A (1) was isolated as a colorless oil, [α]D25 −28.2 (c 0.4, CHCl3). HRESIMS, 13C-NMR, and DEPT spectroscopic data established the molecular formula of 1 as C20H34O4. The IR spectrum of 1 indicated the presence of the hydroxy functionality (νmax 3413 cm−1). The 1H-NMR spectrum of 1 also showed signals for three olefinic protons at δ 5.94 (d, J = 15.6 Hz, H-3), 5.63 (d, J = 15.6 Hz, H-2), and 5.20 (d, J = 9.6 Hz, H-7) ppm; one oxymethine proton at δ 3.56 (d, J = 9.6 Hz, H-11); and an olefinic methyl group at δ 1.69 (s, H3-19). The spectral data of 1 indicated some similarities to those of (2E,7E)-4,11-dihydroxy-1,12-oxidocembra-2,7-diene [11], except for the data correspoding to C-15. The 1H–1H COSY spectrum exhibited correlations from H-2 to H-3, H2-5 to H-7, H2-9 to H-11, and H2-13 to H2-14. These spectroscopic findings and the requirement for four degrees of unsaturations indicated that 1 was a 14-membered cembrane-type diterpene skeleton with an ether ring.

The HMBC correlations are shown in Figure 3. NOESY correlations between H-7 and H2-9, and the chemical shift values at δC 16.2 (C-19) disclosed the E configuration for the trisubstituted olefin [10]. The J values for both H-2 and H-3 (15.6 Hz) further confirmed the presence of a trans 1,2-disubstituted double bond at C-2. NOESY correlations (Figure 4) observed between H-2 and H3-17, H3-17 and H-14a, H-14a and H3-20, H-3 and H3-16/H3-18, H3-18 and H-6a, H3-19 and H-6a/H-10, and H-7 and H-5b/H-6b/H-11 indicated the relative configurations for the 14-membered ring carbons, which were identical to those of (2E,7E)-4,11-dihydroxy-1,12-oxidocembra-2,7-diene. Analysis of the Δδ values (Figure 5) according to the Mosher model indicated an R configuration for C-11 of 1 based on the deshielded nature of H-10, H-9, and H-7 of the (S)-MTPA ester 1a. Therefore, the absolute stereochemistry of Numerosol A (1) was established as (1R,4R,11R,12S,2E,7E)-4,11,15-trihydroxy-1,12-oxidocembra-2,7-diene.

Figure 3

Selected 1H−1H COSY (▬) and HMBC (→) correlations of 1–4.

Figure 4

Key NOESY Correlations for 1 and 2.

Figure 5

Absolute stereochemistry of 1: Δδ values in ppm for MTPA esters 1a and 1b.

HRESIMS of Numerosol B (2) exhibited a pseudo-molecular ion peak at m/z 385.2353 [M + Na]+, consistent with the molecular formula of C22H34O4, requiring six degrees of unsaturation. The IR spectrum of 2 revealed the presence of hydroxy (νmax 3440 cm−1) and carbonyl (νmax 1730 cm−1) moieties. The 13C-NMR spectrum of 2 (Table 1) displayed 22 carbon signals, and a DEPT experiment confirmed the presence of six methyls, five methylenes, five methines, and six quaternary carbons. The presence of six carbon signals at δ 150.3 (qC), 138.8 (qC), 134.7 (qC), 126.2 (CH), 125.7 (qC), and 115.3 (CH) and the three proton signals at δ 5.83 (1H, dd, J = 7.5, 2.5 Hz), δ 5.35 (1H, d, J = 10.5 Hz), and δ 5.24 (1H, dd, J = 11.0, 3.0 Hz), were attributable to three trisubstituted double bonds. Moreover, an acetoxy group was determined based on NMR signals at δC 171.7 (qC), 21.3 (CH3), and δH 2.10 (3H, s). The 1H-NMR spectrum also displayed two hydroxy-containing methine signals (δ 5.39, d, J = 9.0 Hz; δ 5.01, d, J = 10.5 Hz). Thus, the bicyclic structure of 2 was revealed. From the 1H–1H COSY spectrum (Figure 3), it was possible to identify four different structural units (Figure 3). Key HMBC correlations of H-2 to C-1, C-3, C-4, C-12, and C-15; H-3 to C-5; H2-5 to C-6 and C-18; H-6 to C-8; H-7 to C-9; H-11 to C-9, C-10, C-12, C-13, C-20, and 11-OAc; H2-13 to C-1, C-11 and C-20; H-14 to C-2 and C-15; H3-16 to C-1, C-15, and C-17; H3-17 to C-1, C-15, and C-16; H3-18 to C-3, C-4, and C-5; H3-19 to C-7, C-8, and C-9; H3-20 to C-11, C-12, and C-13; H3-OAc to 11-OAc permitted the establishment of the cembrane-type skeleton of 2. The E geometry of the double bonds at C-3/C-4, C-7/C-8, and C-14/C-1 was supported by NOE correlations between H-3 and H-5a (δ 2.08, m), between H-7 and H-9a (δ 1.82, m), and between H-14 and H3-17 (Figure 4). The chemical shift values at δC 15.2 (C-18) and δC 16.4 (C-19) further supported the E configuration at C-3/C-4, C-7/C-8 [10]. The β-orientation of H-2 was established from NOE correlations observed between H-2 and H3-18, H3-18 and H-6b (δ 2.43, dd, J = 11.0, 3.0 Hz), and H-6b and H3-19. The NOE correlations observed between H-6a (δ 2.08, m) and H-7, H-7 and H-11, H-11 and H-13a (δ 2.22, m), reflected the α-orientation of H-11. Also, the NOE correlations observed between H-13b (δ 2.14, m) and H3-20, reflected the β-orientation of H3-20. According to the above NOE correlations, and the others shown in Figure 4, the relative configurations at C-2, C-11, and C-12 for 2 were determined as 2R*,11S*, and 12R*, respectively.

Table 1

1H and 13C-NMR Spectroscopic Data for compounds 1 and 2.

Position	1		2
	δH a	δC b	δH c		δC d
1		91.3, qC f		150.3, qC
2	5.63 d (15.6) e	128.2, CH	5.01 d (10.5)	69.2, CH
3	5.94 d (15.6)	138.4, CH	5.35 d (10.5)	126.2, CH
4		74.5, qC		138.8, qC
5	a:1.85 m; b:1.59 m	43.9, CH2	a:2.08 m; b:2.20 m	39.7, CH2
6	a:2.27 m; b:2.17 m	24.35, CH2	a:2.08 m; b:2.43 dd (11.0, 3.0);	26.1, CH2
7	5.20 d (9.6)	129.2, CH	5.24 dd (11.0, 3.0)	125.7, CH
8		133.5, qC		134.7, qC
9	2.10 m	35.4, CH2	a:1.82 t (3.0); b:1.94 m	35.4, CH2
10	a:1.91 dd (14.7, 9.6);b:1.36 d (14.7)	29.4, CH2	a:1.82 m; b:1.44 m	26.5, CH2
11	3.56 d (9.6)	76.4, CH	5.39 d (9.0)	78.8, CH
12		85.4, qC		73.6, qC
13	1.79 m	36.6, CH2	a:2.22 m; b:2.14 m	32.4, CH2
14	a:2.43 m; b:1.69 m	31.0, CH2	5.83 dd (7.5, 2.5)	115.3, CH
15		72.4, qC		71.8, qC
16	1.07 s	25.9, CH3	1.31 s	28.8, CH3
17	1.14 s	24.41, CH3	1.31 s	28.9, CH3
18	1.28 s	28.4, CH3	1.74 s	15.2, CH3
19	1.69 s	16.2, CH3	1.59 s	16.4, CH3
20	1.12 s	19.4, CH3	1.05 s	23.3, CH3
OAc-11			2.10 s	21.3, CH3
				171.7, qC

a Spectra recorded at 400 MHz in CDCl3; b Spectra recorded at 100 MHz in CDCl3; c Spectra recorded at 500 MHz in CDCl3; d Spectra recorded at 125 MHz in CDCl3; e J values (in Hz) are in parentheses; f Carbon types are deduced by HSQC and DEPT experiments.

Numerosol C (3) was obtained as colorless oil. Its HRESIMS (m/z 327.2299 [M + Na]+) established the molecular formula C20H32O2, requiring five degrees of unsaturation. The IR spectrum of 3 revealed the presence of a hydroxyl (νmax 3404 cm−1) moiety. Analysis of the 1H–1H COSY and HMBC correlations (Figure 3) were diagnostic in determining that the planar framework of numerosol C, having a 14-membered ring, was proposed as 3. The observed COSY correlation between H-10 and H-11 and the key HMBC correlations from H-11 to C-10 confirmed the location of the hydroxyl group. The NMR data (Table 2) of 3 were similar to those of gibberosene G [12]. However, resonances for the methane proton at C-15 in gibberosene G were absent from the 1H-NMR spectrum of 3. In addition, the methine carbon at δ 34.4 in gibberosene G was downfield-shifted to δ 74.1 in 3. Thus, the methine proton in gibberosene G should be replaced by a hydroxyl group in 3. The E-geometries of the four double bonds at C-1/C-2, C-3/C-4, C-7/C-8, and C-12/C-13 were determined by the NOE interactions (Figure 6) displayed by the methyl protons at H3-16 with H-2, H-2 with H3-18, H-7 with H2-9a, and H-11 with H-13a, respectively. The chemical shift values at δC 18.0 (C-18), δC 16.7 (C-19), and δC 15.9 (C-20) also supported the E configuration at C-3/C-4, C-7/C-8, and C-11/C-12 [10]. The relative structure of numerosol C (3) was established as (−)-(1E,3E,7E,12E)-10-hydroxycembra-1,3,7,12-tetraene. Due to the decomposition of compound 3, the stereo center present in compounds 3 could not be conclusively assigned.

Table 2

1H and 13C-NMR Spectroscopic Data for compounds 3 and 4.

Position	3		4
	δH a	δC b	δH a	δC b
1		147.1, qC d		145.5, qC
2	6.39 d (11.6) c	118.7, CH	6.37 d (10.4)	118.7, CH
3	5.78 d (11.6)	120.0, CH	5.71 d (10.4)	121.3, CH
4		138.3, qC		137.4, qC
5	2.18 m	38.4, CH2	2.15 m	38.0, CH2
6	a:2.24 m; b:2.18 m;	24.5, CH2	2.16 m	24.5, CH2
7	5.02 d (5.2)	127.8, CH	4.95 brs	126.0, CH
8		131.2, qC		133.4, qC
9	a:2.10 dd (12.4, 10.8); b:2.46 m	48.0, CH2	a:2.13 m; b:2.00 m	35.9, CH2
10	4.55 ddd (14.8, 9.6, 4.0)	66.2, CH	a:1.85 m; b:1.74 m	29.4, CH2
11	5.13 d (9.6)	128.4, CH	3.93 dd (10.4, 3.6)	79.1, CH
12		140.2, qC		135.5, qC
13	a:1.92 m; b:2.42 m	42.4, CH2	5.20 t (5.2)	130.4, CH
14	a:2.22 mb:2.40 m	26.0, CH2	a:2.83 dd (16.8, 5.2)b:3.00 dd (16.8, 5.2)	26.7, CH2
15		74.1, qC		73.9, qC
16	1.38 s	29.9, CH3	1.37 s	28.8, CH3
17	1.38 s	29.8, CH3	1.37 s	28.9, CH3
18	1.79 s	18.0, CH3	1.76 s	17.6, CH3
19	1.60 s	16.7, CH3	1.56 s	15.2, CH3
20	1.74 s	15.9, CH3	1.65 s	11.0, CH3

a Spectra recorded at 400 MHz in CDCl3; b Spectra recorded at 100 MHz in CDCl3; c J values (in Hz) are in parentheses; d Multiplicities are deduced by HSQC and DEPT experiments.

Figure 6

Key NOESY Correlations for 3 and 4.

Numerosol D (4) was found to have the molecular formula C20H32O2, as indicated by the HRESIMS (m/z found 327.2299, [M + Na]+), requiring five degrees of unsaturation. The 13C-NMR signals (Table 2) at δC 145.5 (qC, C-1), 118.7 (CH, C-2), 121.3 (CH, C-3), 137.4 (qC, C-4), 126.0 (CH, C-7), 133.4 (qC, C-8), 135.5 (qC, C-12) and 130.4 (CH, C-13) revealed four trisubstituted double bonds in 4. The oxygen atoms were attributable to two hydroxy groups (IR 3420 cm−1). The first hydroxymethine proton (δ 3.93, dd, J = 10.4, 3.6 Hz) was located at C-11 based on COSY correlations with a set of methylene protons (δ 1.74, m and 1.85, m, H2-10) and a HMBC correlation from H-11 to H3-20 (δ 1.65, s). HMBC correlations (Figure 3) from H3-16 and H3-17 to C-15 (δ 73.9, qC) confirmed the second hydroxymethine to be located at C-15. The E-geometries of the four double bonds at C-1/C-2, C-3/C-4, C-7/C-8, and C-12/C-13 were determined by the following NOE interactions (Figure 6): H-15 with H-2, H-2 with H3-18, H-7 with H2-9, and H-11 with H-13, respectively. The chemical shift values at δC 17.6 (C-18), δC 15.2 (C-19), and δC 11.0 (C-20) also supported the E configuration at C-3/C-4, C-7/C-8, and C-2/C-3 [10]. The relative structure of numerosol D (4) was established as (+)-(1E,3E,7E,12E)-11-hydroxycembra-1,3,7,12-tetraene. Due to the decomposition of compound 4, the stereo center present in compounds 4 could not be determined.

Metabolites 1–5 in our study were evaluated for cytotoxicity against P-388, A549, and HT-29 cancer cell lines as well as antiviral activity against human cytomegalovirus. Preliminary cytotoxic screening revealed that 5 exhibited cytotoxicity against P-388 (mouse lymphocytic leukemia) cell line with an ED50 of 6.9 μM.

3. Experimental Section

3.1. General Experimental Procedures

The NMR spectra were recorded on a Varian Unity INOVA 500 FT-NMR spectrometer at 500 MHz for 1H and 125 MHz for 13C or on a Varian MR 400 FT-NMR at 400 MHz for 1H and 100 MHz for 13C, respectively. 1H-NMR chemical shifts are expressed in δ referring to the solvent peak δH 7.27 for CDCl3, and coupling constants are expressed in Hz. 13C-NMR chemical shifts are expressed in δ referring to the solvent peak δC 77.0 for CDCl3. Optical rotations were determined with a JASCO P1020 digital polarimeter. IR spectra were recorded on a JASCO FT/IR4100 infrared spectrophotometer. LRMS and HRMS were obtained by ESI on a Bruker APEX ΙΙ mass spectrometer. Silica gel 60 (Merck, Darmstadt, Germany, 230–400 mesh) and LiChroprep RP-18 (Merck, 40–63 μm) were used for column chromatography. Precoated silica gel plates (Merck, Kieselgel 60 F254, 0.25 mm) and precoated RP-18 F254s plates (Merck) were used for thin-layer chromatography analysis. High-performance liquid chromatography was carried out using a Hitachi L-7100 pump equipped with a Hitachi L-7400 UV detector at 220 nm together with a semi-preparative reversed-phase column (Merck, Hibar LiChrospher RP-18e, 5 μm, 250 × 25 mm).

3.2. Animal Material

The octocoral S. numerosa was collected by hand using scuba at Sansiantai, Taitong County, Taiwan, in August 2008, at a depth of 6–8 m and stored in a freezer until extraction. This soft coral was identified by Prof. Chang-Fong Dai, Institute of Oceanography, National Taiwan University. A voucher specimen (SST-12) was deposite in the Department of Marine Biotechnology and Resources, National Sun Yat-sen University.

3.3. Extraction and Separation

A specimen of S. numerosa (4.0 kg, wet weight) was chopped into small pieces and extracted with acetone (3L × 3) at room temperature. The combined acetone extract was then partitioned with EtOAc and H2O. The EtOAc-soluble residue (27.5 g) was subjected to Si 60 CC using n-hexane–EtOAc–MeOH mixtures of increasing polarity for elution. Fraction 19, eluted with n-hexane–EtOAc (1:2), was purified by reverse-phase HPLC (MeOH–H2O, 70:30) to afford 2 (1.2 mg, 0.00003 w/w%). Fraction 23, eluted with n-hexane–EtOAc (1:6), was purified by reverse-phase HPLC (MeOH–H2O, 70:30) to afford 3 (1.3 mg, 0.000032 w/w%) and 4 (2.4 mg, 0.00006 w/w%). Fraction 24, eluted with n-hexane–EtOAc (1:8), was purified by reverse-phase HPLC (MeOH–H2O, 75:25) to afford 1 (1.9 mg, 0.000047 w/w%).

Numerosol A (1): Colorless oil; [α]D25 = −28 (c 0.4, CHCl3); IR (neat) νmax 3413, 2925, 2857, 1456, 1378, 1166, 1074, 952, 756, 665 cm−1; 1H and 13C-NMR data, see Table 1; ESIMS m/z 361 [M + Na]+; HRESIMS m/z 361.2357 (calcd for C20H34O4Na, 361.2355) (Supplementary Figures S1–S9).

Numerosol B (2): Colorless oil; [α]D25 = −60 (c 0.2, CHCl3); IR (neat) νmax 3440, 2919, 2851, 1730, 1575, 1371, 1240, 1017 cm−1; 1H and 13C-NMR data, see Table 1; ESIMS m/z 385 [M + Na]+; HRESIMS m/z 385.2353 (calcd for C22H34O4Na, 385.2355) (Supplementary Figures S10–S15).

Numerosol C (3): Colorless oil; [α]D25 = −100 (c 0.3, CHCl3); IR (neat) νmax 3404, 2922, 2853, 1564, 1417, 1255, 1016, 771 cm−1; 1H and 13C-NMR data, see Table 2; ESIMS m/z 327 [M + Na]+; HRESIMS m/z 327.2299 (calcd for C20H32O2Na, 327.2300) (Supplementary Figures S16–S21).

Numerosol D (4): Colorless oil; [α]D25 = +8 (c 0.4, CHCl3); IR (neat) νmax 3420, 2972, 2931, 1564, 1419, 1381, 998, 765 cm−1; 1H and 13C-NMR data, see Table 2; ESIMS m/z 327 [M + Na]+; HRESIMS m/z 327.2299 (calcd for C20H32O2Na, 327.2300) (Supplementary Figures S22–S27).

3.4. Cytotoxicity Assay

Cytotoxicity was determined on P-388 (mouse lymphocytic leukemia), HT-29 (human colon adenocarcinoma), and A-549 (human lung epithelial carcinoma) tumor cells using a modification of the MTT colorimetric method according to a previously described procedure [13,14]. The provision of the P-388 cell line was supported by J.M. Pezzuto, formerly of the Department of Medicinal Chemistry and Pharmacognosy, University of Illinois at Chicago. HT-29 and A-549 cell lines were purchased from the American Type Culture Collection. To measure the cytotoxic activities of the tested compounds, five concentrations with three replications were performed on each cell line. Mithramycin was used as a positive control.

3.5. Anti-HCMV Assay

To determine the effects of natural products upon HCMV cytopathic effect (CPE), confluent human embryonic lung (HEL) cells grown in 24-well plates were incubated for 1 h in the presence or absence of various concentrations of tested natural products with three replications. Ganciclovir was used as a positive control. Then, cells were infected with HCMV at an input of 1000 pfu (plaque forming units) per well of a 24-well dish. Antiviral activity was expressed as IC50 (50% inhibitory concentration), or compound concentration required to reduce virus induced CPE by 50% after 7 days as compared with the untreated control. To monitor the cell growth upon treatment with natural products, an MTT-colorimetric assay was employed [15].

4. Conclusions

In previous studies, gibberoketosterol (5) showed moderate cytotoxicity and selectivity against the growth of Hepa59T/VGH cancer cells was evaluated [12]. This investigation of soft coral S. munerosa collected at San-Hsian-Tai (Taitong County, Taiwan) has led to the isolation of four new cembrane-type diterpenoids, Numerosol A–D (1–4) and the known steroid gibberoketosterol (5). Compound 5 displayed moderate cytotoxicity and selectivity against P-388, with ED50 of 6.9 μM.