Chemical Constituents from Soft Coral Clavularia spp. Demonstrate Antiproliferative Effects on Oral Cancer Cells

Abstract

Five new eudensamane-type sesquiterpene lactones, clasamanes A–E (1–5), three new dolabellane-type diterpenes, clabellanes A–C (6–8), and fifteen known compounds (9–23) were isolated from the ethanolic extract of Taiwanese soft coral Clavularia spp. The structures of all undescribed components (1–8) were determined by analysis of IR, mass, NMR, and UV spectroscopic data. The absolute configuration of new compounds was determined by using circular dichroism and DP4+ calculations. The cytotoxic activities of all isolated marine natural products were evaluated. Compound 7 showed a significant cytotoxic effect against oral cancer cell line (Ca9-22) with an IC₅₀ value of 7.26 ± 0.17 μg/mL.

1. Introduction

In south–central Asia, people are susceptible to oral cancers because of the usage of areca nuts as chewing gum. Taiwan has one of the world’s highest incidences of oral cancer, which ranks fourth in the cause of cancer death among male Taiwanese [1]. In Taiwan, about 3000 deaths yearly are due to oral cancers. The treatments for oral cancer are usually combined surgery and chemotherapy; however, chemotherapy drugs sometimes produce adverse effects [2]. It is necessary to discover new anti-oral cancer drugs.

Marine sessile animals like sponges, soft corals, tunicates, and zoanthids are known to produce diverse secondary metabolites. Octacoral is one of the most abundant sources of bioactive marine natural products (MNPs) with unique backbones. Since 1977, the soft corals of the genus Clavularia have been found to have different kinds of MNPs, such as diterpenoids [3,4,5], sesquiterpenoids [6], prostanoids [7,8], and steroids [9]. Those MNPs usually demonstrate considerable cytotoxic effects against several cancer cell lines. For example, dolabellane-type diterpenes could significantly inhibit P-388 leukemia cells with an ED₅₀ value of 0.052 μg/mL [4]. Eudensamane-type sesquiterpene lactones were found to inhibit the growth of cancer cell lines. In our previous study on Taiwanese marine invertebrates, the methanol extract of Clavularia inflata exerts an apoptotic effect and DNA damage to oral cancer cells [10]. These findings propel us to conduct the natural product investigation of this coral extract.

2. Results

In this contribution, we describe the isolation, structural determination and cytotoxic evaluation from the bioactive coral extract of Clavularia spp. In total, eight new (1–8) and fifteen known compounds (9–23) were isolated from repeated column chromatography (Figure 1). Those known compounds were identified as atractylenolides III (9) [11], tubipolide A (10) [12], tubipolide C (11) [12], atractylenolactam (12) [13], clavinflol B (13) [5], clavinflol B monoacetate (14) [5], (1R,12R)-dolabella-4(16),7,10-triene-3,13-dione (15) [4], (1R*)-dolabella-4(16),7,11(12)-triene-3,13-dione (16) [4], (1R*,7R*,8S*,-12R*)-dolabella-4(16),10-diene-7,8-epoxy-3,13-dione (17) [4], (1R*,10R*,11S*,12R*)-dolabella-4(16),7-diene-10,11-epoxy-3,13-dione (18) [4], 2-((E)-(1S,3R,5R,12S)-1,5,9-trimethyl-4-oxa-tricyclo [10.3.0.0^3,5]pentadeca-8,13-dien-13-yl)-propan-2-ol (19) [14], 2-((E)-(1R,3R,12S,15S)-5-hydroxymethyl-12-methyl-9-methylene-2-oxa-tricyclo [10.3.0.0^1,3]pentadec-5-en-15-yl)-propan-2-ol (20) [15], stolonidiol (21) [16], stolonidiol monoacetate (22) [16], and clavinflol A (23) [5].

Clasamane A (1) was isolated as a colorless oil, and the molecular formula, C₁₇H₂₄O₃ (Δ = 6), was assigned from its HRESIMS data (m/z 299.1616 [M + Na]⁺, which was calculated for 299.1618). The UV maximum absorptions at λ_max 218 and 282 nm implied that compound 1 belongs to eudensamane-type sesquiterpene lactone [11], whereas the IR absorptions indicated the presence of lactone functionality (1751 cm⁻¹). The ¹H NMR spectrum (

Table 1. ¹H NMR (600 MHz) spectroscopic data of compounds 1–5 (δ in ppm, J value in Hz) ^a.

No.	1	2	3	4 b	5
1	1.23, dd (12.9, 5.9)	1.48, m	5.78, d (9.6)	5.83, d (3.1)	4.20, d (6.0)
	1.56, d (12.9)	1.63, m
2	1.63, m	1.48, m	5.85, d (9.6, 5.2)	5.83, d (3.1)	6.74, dd (8.4, 6.0)
		1.63, m
3	1.96, m	2.06, m	5.95, d (5.2)	5.95, m	6.39, d (8.4)
	2.36, d (13.7)	2.38, m
5	1.82, dd (12.8, 1.3)	2.74, m	1.86, dd (13.0, 4.2)	1.85, dd (12.5, 4.2)	2.53, dd (10.8, 2.9)
6	2.27, t (12.8)	2.44, m	2.17, t (13.0)	2.18, t (12.5)	2.61, m
	2.62, dd (12.8, 3.2)	2.64, dd (11.8, 8.2)	2.68, dd (13.0, 4.2)	2.67, dd (12.5, 4.2)	2.47, dd (16.4, 2.9)
9	1.41, d (13.7)	1.80, d (14.5)	1.55, d (14.2)	1.52, d (13.8)	1.40, d (13.9)
	2.36, d (13.7)	2.22, d (14.5)	2.52, d (14.2)	2.53, d (13.8)	2.26, d (13.9)
13	1.86, s	1.88, s	1.88, s	1.86, s	1.79, s
14	0.99, s	0.64, s	0.97, s	0.96, s	1.61, s
15	4.58, s	4.64, s	4.60, d (13.1)	4.59, d (13.0)	4.37, d (13.1)
		4.89, s	4.67, d (13.1)	4.66, d (13.0)	4.43, d (13.1)
1′	3.28, m	3.17, m	3.07, s	3.09, m	3.15, m
	3.46, m	3.41, m		3.38, m	3.46, m
2′	1.18, t (7.0)	1.18, t (7.0)		1.14, t (7.0)	1.21, t (7.0)
2″			2.11, s	2.10, s	2.17, s

^a Measured in CDCl3. ^b Measured at 400 MHz in CDCl3.

) displayed the signals for two methyls (δ 1.86 (s) (H₃-13) and δ 0.99 (s) (H₃-14)), an ethoxy group (δ 3.46 (m), δ 3.28 (m) (H₂-1′) and δ 1.18 (t, J = 7.0) (H₃-2′)), and an exomethylene (δ 4.86 (s) and δ 4.58 (s) (H₂-15)). The ¹³C NMR and DEPT spectra (

Table 2. ¹³C NMR (150 MHz) spectroscopic data of compounds 1–8 (δ in ppm) ^a.

No.	1	2	3	4 b	5	6	7 b	8
1	41.4 CH2	42.6 CH2	137.4 CH	137.6 CH	79.5 CH	44.6 C	44.6 C	53.1 C
2	22.3 CH2	23.2 CH2	120.2 CH	119.9 CH	135.8 CH	42.4 CH2	42.3 CH2	35.7 CH2
3	36.0 CH2	36.5 CH2	121.7 CH	121.7 CH	129.2 CH	25.1 CH2	25.0 CH2	29.9 CH2
4	148.7 C	148.6 C	133.6 C	133.5 C	79.8 C	147.4 C	147.1 C	150.3 C
5	51.8 CH	41.6 CH	44.9 CH	44.9 CH	42.5 CH	35.4 CH2	36.4 CH2	28.7 CH2
6	25.0 CH2	24.2 CH2	25.7 CH2	25.8 CH2	22.8 CH2	29.7 CH2	30.9 CH2	22.3 CH2
7	159.9 C	157.4 C	156.8 C	157.2 C	155.8 C	63.0 CH	46.0 CH	60.2 CH
8	106.2 C	106.9 C	105.7 C	105.7 C	105.3 C	75.0 C	75.1 C	63.8 C
9	50.2 CH2	50.7 CH2	48.6 CH2	48.8 CH2	43.8 CH2	33.3 CH2	32.1 CH2	37.8 CH2
10	36.8 C	34.9 C	36.5 C	36.5 C	38.7 C	54.7 CH	54.7 CH	65.0 CH
11	123.8 C	125.5 C	123.8 C	123.5 C	124.2 C	76.2 C	76.1 C	141.2 C
12	171.9 C	171.9 C	171.5 C	171.6 C	171.1 C	50.2 CH	50.2 CH	144.7 C
13	8.3 CH3	8.3 CH3	8.3 CH3	8.3 CH3	8.3 CH3	27.7 CH2	27.6 CH2	32.8 CH2
14	16.4 CH3	21.0 CH3	25.1 CH3	25.1 CH3	29.2 CH3	38.8 CH2	38.8 CH2	32.7 CH2
15	106.7 CH2	107.4 CH2	66.3 CH2	66.3 CH2	62.8 CH2	24.1 CH3	24.1 CH3	26.8 CH3
16						113.8 CH2	113.7 CH2	111.1 CH2
17						67.0 CH2	69.5 CH2	68.6 CH2
18						75.4 C	75.2 C	72.3 C
19						29.6 CH3	29.5 CH3	31.0 CH3
20						26.0 CH3	26.0 CH3	29.7 CH3
1′	58.7 CH2	58.8 CH2	50.3 CH3	50.3 CH3	59.1 CH3
2′	15.2 CH3	15.3 CH3			15.1 CH3
1″			170.8 C	170.8 C	170.7 C
2″			20.8 CH3	20.8 CH3	20.8 CH3

^a Measured at 150 MHz in CDCl₃. ^b Measured at 100 MHz in CDCl₃.

) displayed 17 carbons, which were assigned to one ester carbonyl (δ 171.9 (C-12)), three sp² non-protonated carbons (δ 159.9 (C-7), δ 148.7 (C-4), and δ 123.8 (C-11)), one exomethylene (δ 106.7 (C-15)), two quaternary carbons (δ 106.2 (C-8), and δ 36.8 (C-10)), six methylenes (δ 58.7 (C-1′), δ 50.2 (C-9), δ 41.4 (C-1), δ 36.0 (C-3), δ 25.0 (C-6), and 22.3 (C-2)), one methine (δ 51.8 (C-5)), and three methyls (δ 16.4 (C-14), δ 15.2 (C-2′), and δ 8.3 (C-13)). The aforementioned NMR data are similar to those of atractylenolides III (9), except for an additional ethoxy group that was found in the ¹H and ¹³C spectra. The planar structure of 1 was established by the COSY and HMBC experimental data (Figure 2). In the COSY spectrum, two proto sequences of H₂-1 (δ 1.56 and δ 1.23)/H₂-2 (δ 1.63 and δ 1.48)/H₂-3 (δ 2.36 and δ 1.96) and H-5 (δ 1.82)/H₂-6 (δ 2.62 and δ 2.27) were observed. These two proton sequences and the HMBC correlations from H₃-14 to C-1, C-5, C-9, and C-10; from H₂-15 to C-3, C-4, and C-5; from H₂-6 to C-7 and C-8; from H₃-13 to C-7, C-11, and C-12; and from H₂-9 (δ 2.36 and δ 1.41) to C-8 and C-7 can be used to construct the carbon skeleton of eudensamane-type sesquiterpene lactone. The attachment of an ethoxy group at C-8 was confirmed by virtue of the HMBC correlations from H₂-1′ to C-8. The relative configuration of 1 was determined by the NOESY correlations (Figure 2). NOE cross-peaks of H₃-14/H-9β (δ 2.36)/H-1′ (δ 3.46) suggested these protons are on the β-orientation. The NOESY correlations of H-6α (δ 2.27)/H-5/H-9α (δ 1.41) suggested they are α-oriented. The absence of NOESY correlation between H₃-14 and H-5 indicated the opposite side of these protons and a trans-decalin moiety in 1. The absolute configuration of 1 was determined to be (5S,8S,10R) by comparing the experimental ECD and NMR data of 1 with that of 9 (Figure S1). On the basis of the above spectroscopic data analysis, the structure of 1 was determined as shown.

Clasamane B (2) was an isomer of 1 because it possessed the same molecular formula as 1 and similar NMR spectrometric data. The most noticeable variations between 2 and 1 were the carbon chemical shifts of C-5 (δ 41.6 for 2 and δ 51.8 for 1) and C-14 (δ 21.0 for 2 and δ 16.4 for 1), which implied the configuration of these two positions might change. The absence of NOESY correlation (Figure S2) between H₃-14 (δ 0.64) and H-5 (δ 2.74) indicated the trans conformation of the decalin moiety, like that of 1. The configuration of C-8 was defined to be S due to the similar NMR data of α-methyl-α,β-unsaturated-γ-hydroxy-γ-lactone moiety between 2 and 1. The NOESY correlations of H-6β (δ 2.44)/H-5 (δ 2.74)/H-9β (δ 2.22) and H-5/H-1′ (δ 3.17) confirmed those protons located on the β-face. On the contrary, the NOESY correlations of H-9α (δ 1.80)/H₃-14/H-6α (δ 2.64) suggested they are on the α-face. Therefore, the configuration of 2 was unambiguously determined.

Clasamane C (3) possessed a molecular formula of C₁₈H₂₂O₅, which is consistent with its positive sodiated HRESIMS ion at m/z 341.1362. The ¹H and ¹³C NMR spectroscopic data of 3 were similar to those of 10, suggesting they are congeners. Comparison of the NMR spectra between 3 and 10 showed that 3 has an additional methoxy group (δ_H 3.07 (s) (H₃-1′); δ_C 50.3 (C-1′)). This methoxy group situated at C-8 was evidenced by the HMBC correlation from H₃-1′ to C-8 (δ 105.7). The NOESY correlation between H-5 (δ 1.86) and H₃-14 (δ 1.56) revealed they were on the same face (α-orientation) of the molecule. The absolute configuration of 3 was determined by ECD data analysis. Due to the consistency of ECD curves between 3 and 11 (Figure S3), the absolute configuration of 3 was defined as 5R,8S,10R.

Clasamane D (4) was isolated as a colorless oil and had the molecular formula of C₁₉H₂₄O₅ inferred from the sodiated HRESIMS ion at m/z 355.1518, which is 14 amu more than 3. The UV, IR and NMR data of 4 were quite similar to those of 3, except the methoxy group in 3 was replaced by an ethoxy group (δ_H 3.38 (m), 3.09 (m) (H₂-1′); δ_C 58.7 (C-1′); δ_H 1.14 (t) (H₃-2′); δ_C 15.1 (C-2′)) in 4. This speculation is consistent with the difference in mass spectrometry between 4 and 3, and it was confirmed by the HMBC correlation from H₂-1′ to C-8 (δ 105.7). The similar ECD trends of 4 and 3 suggested these two compounds share the same absolute configuration. Thus, the structure of 4 was determined as shown.

Clasamane E (5) was a colorless oil, and its HRESIMS data showed a [M + Na]⁺ ion at m/z 387.1414, suggesting the molecular formula of C₁₉H₂₄O₇ with seven indices of hydrogen deficiency. The ¹H and ¹³C data (Table 1 and Table 2) of 5 revealed typical signals of ethoxy (δ_H 3.46 (m), 3.15 (m) (H₂-1′); δ_C 59.1 (C-1′); δ_H 1.21 (t) (H₃-2′); δ_C 15.1 (C-2′)) and acetoxy (δ_C 170.7 (C-1″); δ_H 2.17 (s) (H₃-2″); δ_C 20.8 (C-2″)) groups. A detailed analysis of COSY and HMBC spectra (Figure 3) established the planar structure of 5. In the COSY spectrum, proton spin systems of H-1 (δ 4.20)/H-2 (δ 6.74)/H-3 (δ 6.39) and H-5 (δ 2.53)/H₂-6 (δ 2.61 and δ 2.47) were found. These two proton spin systems and the HMBC correlations from H₃-14 (δ 1.61) to C-1 (δ 79.5), C-5 (δ 42.5), C-9 (δ 43.8), and C-10 (δ 38.7), from H₂-15 (δ 4.43 and δ 4.37) to C-3 (δ 129.2), C-4 (δ 79.8), and C-5, from H₂-9 (δ 2.26 and δ 1.40) to C-8 (δ 105.3), from H₂-6 to C-7 (δ 155.8), C-8, and C-11 (δ 124.2), and from H₃-13 (δ 1.79) to C-7, C-11, and C-12 (δ 171.0) can be used to assemble the framework of eudensamane-type sesquiterpene lactone. The HMBC correlations from H₂-1′ to C-8 and from H₂-15 to C-1″ indicated the connection of ethoxy and acetoxy groups, respectively. The above findings accounted for six of the seven indices of hydrogen deficiency, which implied an additional ring should exist in 5. Considering the molecular formula, two oxygen atoms were not assigned yet, and the carbon chemical shift of C-1 and C-4 suggested that these two carbons are oxygen bearing. Thus, a peroxide bridge between C-1 and C-4 was allocated. This assignment was also confirmed by the down-field shifted signals of H-2 and H-3 [17]. The cis-decalin moiety of 5 was assured by the NOESY correlations between H₃-14 and H-5, which were assigned on the α-face. In addition, the NOESY correlations of H₂-15/H-5 and H₃-14/H-1 revealed the peroxide bridge was on the β-face of the molecule. The NOESY correlation of H-9β (δ 2.26)/H₂-1″ indicated the β-orientation of the ethoxy group. Therefore, the stereocenters of 5 could be temporarily assigned as 1S*,4R*,5S*,8S*,10R* or 1R*,4S*,5R*,8S*,10S* (Figure S45). The ¹H and ¹³C data of those two isomers were calculated by Gaussian 16, and the data were applied to DP4+ probability analysis. The analytic result indicated that the 1S*,4R*,5S*,8S*,10R* isomer has 100% possibility, so the configuration of 5 was determined.

Clabellane A (6) was obtained as a colorless oil with the molecular formula C₂₀H₃₃BrO₄ and four degrees of unsaturation based on the HRESIMS ion at m/z 439.1455 [M + Na]⁺. The presence of one bromine atom was confirmed by the equal intensity between [M + Na]⁺ and [M + 2 + Na]⁺ in the mass spectrum. The presence of hydroxy functionality was confirmed by the IR absorption at 3432 cm⁻¹. The ¹H NMR data (

Table 3. ¹H NMR (600 MHz) spectroscopic data of compounds 6–8 (δ in ppm, J value in Hz) ^a.

No.	6	7 b	8
2	1.96, m	1.95, m	1.74, m
	1.25, m	1.24, m	1.32, dd (7.3, 2.0)
3	2.11, m	2.10, m	2.04, m
	1.63, m	1.56, m	1.79, d (8.8)
5	2.42, td (8.9, 4.3)	2.37, m	2.46, m
	2.28, m	2.25, m	2.27, m
6	1.90, m	1.87, m	1.96, m
	2.11, m		1.49, m
7	4.04, d (11.8)	4.02, dd (7.6, 6.2)	3.13, t (6.7)
9	2.25, d (3.7)	2.22, m	2.58, dd (16.0, 8.0)
		2.32, m	1.96, m
10	2.89, d (6.1)	2.89, d (6.1)	4.37, d (8.0)
12	2.21, d (10.3)	2.22, m
13	1.90, m	1.90, m	2.35, m
	1.63, m	1.61, m	2.20, ddd (10.5, 6.2, 1.4)
14	1.76, m	1.76, m	1.43, ddd (12.4, 8.3, 2.0)
			1.74, m
15	0.86, s	0.85, s	1.05, s
16	4.84, s	4.82, s	4.67, s
	5.03, s	5.07, s	4.71, s
17	3.87, d (11.3)	3.87, d (11.4)	3.98, d (11.7)
	3.65, d (11.3)	3.67, d (11.4)	3.32, d (11.7)
19	1.21, s	1.22, s	1.35, s
20	1.27, s	1.27, s	1.38, s

^a Measured in CDCl₃ ^b Measured at 400 MHz in CDCl₃.

) of 6 demonstrated proton signals of three methyls (δ 1.27 (s) (H₃-20), δ 1.21 (s) (H₃-19), and δ 0.86 (s) (H₃-15)), an oxymethylene (δ 3.87 (d, J = 11.3), δ 3.65 (d, J = 11.3) (H₂-17)), and an exomethylene (δ 5.03 (s), δ 4.84 (s) (H₂-16)). The twenty carbon signals of 6 could be clearly separated into one exocyclic C=C double bond (δ 113.8 (C-16) and δ 147.4 (C-4)), one oxymethylene (δ 67.0 (C-17)), three oxygen-bearing quaternary carbons (δ 76.2 (C-11), δ 75.4 (C-18), and δ 75.0 (C-8)), one quaternary carbon (δ 44.6 (C-1)), seven methylenes (δ 42.4 (C-2), δ 38.8 (C-14), δ 35.4 (C-5), δ 33.3 (C-9), δ 29.7 (C-6), δ 27.7 (C-13), and δ 25.1 (C-3)), three methines (δ 63.0 (C-7), δ 54.7 (C-10) and δ 50.2 (C-12)), and three methyls (δ 29.6 (C-19), δ 26.0 (C-20) and δ 24.1 (C-15)) by using ¹³C NMR (Table 2) data together with DEPT-135 and HSQC spectra. Considering the above data and the reported compounds isolated from the genus Clavularia, 6 can be deduced as a dollabellane-type diterpenoid.

The planar structure of 6 was established by COSY and HMBC correlations (Figure 4). Three proton sequences of H₂-2 (δ 1.96 and δ 1.25)/H₂-3 (δ 2.11 and δ 1.63), H₂-5 (δ 2.42 and δ 2.28)/H₂-6 (δ 2.11 and δ 1.90)/H-7 (δ 4.04), and H₂-9 (δ 2.25)/H-10 (δ 2.86) were observed from the COSY spectrum. Those findings and the HMBC correlation from H₂-16 to C-3, C-4, and C-5; from H₂-17 to C-7, C-8, and C-9; from H₂-9 to C-8 and C-11; and from H₃-15 to C-1, C-2, and C-11 could build the cycloundecane moiety (C-1 to C-11) of 6, and it could confirm the presence of exomethylene (H₂-16) connecting at C-4, an oxymethylene attaching at C-8, and a methyl (H₃-15) connecting at C-1. In addition, the COSY correlations of H-12 (δ 2.21)/H₂-13 (δ 1.90 and δ 1.63)/H₂-14 (δ 1.76) together with the HMBC correlations from H-12 to C-11 and from H₃-15 to C-1, C-11, and C-14 could establish the cyclopentane ring of 6. The HMBC correlations from H-12 and the geminal methyls (H₃-19 and H₃-20) to C-3 constructed the isopropyl alcohol group. The planar structure of 6 was found to be similar to that of clavinflol B (13) except that the chlorine atom in the molecular formula (C₂₀H₃₃BrO₄) of 13 was replaced by a bromine atom in 6. The halogen atom of 6 was allocated at C-7, the same as that of 13, due to the downfield shifted carbon chemical shift at this position (δ 66.2 in 13 and δ 63.0 in 6). The relative configuration of 6 was determined through the NOESY spectrum (Figure 4). The bromine atom was assigned on the α-orientation to avoid steric interaction, while the NOESY correlations of H-7/H-10/H₃-15/H₃-19 indicated those protons were on the β–face of the molecule. On the other hand, the NOESY correlation between H₂-17 and H₂-9 implied that the hydroxy group should be β-oriented. Therefore, the configuration of 6 could be defined as 1R*,7R*,10S*,12R*. The configuration of C-8 was deduced to be 8R by the DP4+ probability measurement. The 1R*,7R*,8R*,10S*,12R* isomer showed an overwhelming possibility (100%) compared with the 1R*,7R*,8S*,10S*,12R* isomer (0%), which confirmed the configuration of 6.

Clabellane B (7) was found to possess the molecular formula C₂₀H₃₃IO₄ according to the sodium adduct ion at m/z 487.1318 (calcd for 487.1316) in the HRESIMS. The UV, IR, and NMR data (Figure 2 and Figure 3) were quite similar to that of 6, indicating they are close analogs. The major difference between 7 and 6 was found in the molecular formula; the bromine atom in 6 was replaced by an iodine atom in 7. The heavy atom effect of C-7 (δ 46.0 in 7 and δ 63.0 in 6) revealed that the bromine atom at C-7 in 6 was replaced by an iodine atom in 7. Since the specific rotation, ECD, and NOESY data of 7 also resembled those of 6, and the stereochemistry of 7 was thus assigned identically.

Clabellane C (8) has the molecular formula C₂₀H₃₂O₄ (IHD = 5), as deduced from HRESIMS and NMR spectrometric data. The IR spectrum of 8 indicated the presence of hydroxy (3378 cm⁻¹) and exomethylene (1643 cm⁻¹) functionalities. The ¹H and ¹³C NMR data of 8 were analogous to those of 6, suggesting it is also a dolabellane-type diterpenoid. Four proton sequences of H₂-2 (δ 1.74 and δ 1.32)/H₂-3 (δ 2.04 and δ 1.79), H₂-5 (δ 2.46 and δ 2.27)/H₂-6 (δ 1.96 and δ 1.49)/H-7 (δ 3.13), H₂-9 (δ 2.58 and δ 1.96)/H-10 (δ 4.37), and H₂-13 (δ 2.35 and δ 2.20)/H₂-14 (δ 1.74 and δ 1.43) were observed by the cross-peaks in the COSY spectrum. Moreover, the above finding and the HMBC correlations from H₂-16 (δ 4.71 and δ 4.67) to C-3 (δ 29.9), C-4 (δ 150.3), and C-5 (δ 28.7); from H₂-17 (δ 3.98 and δ 3.32) to C-7 (δ 60.2), C-8 (δ 63.8), and C-9 (δ 37.8); from H-10 to C-1 (δ 53.1), C-11 (δ 141.2), and C-12 (δ 144.7); from H₃-15 (δ 1.05) to C-1 (δ 53.1), C-2 (δ 35.7), C-11, and C-14 (δ 32.7); and from H₃-19 (δ 1.35) to C-12, C-18 (δ 72.3), and C-20 (δ 29.7) could establish the planar structure of 8. Three hydroxy groups allocated at C-10 (δ 65.0), C-17 (δ 68.6), and C-18 were assured by virtue of their downfield shifted carbon chemical shifts. The aforementioned data accounted for four of the five indices of hydrogen deficiency, suggesting an additional ring remained in 8. An epoxy group was assigned to C-7 and C-8 by their downfield shifted carbon chemical shifts. Therefore, the planar structure was established. The relative configuration of 8 was determined by interpretation of NOESY data (Figure 5) and DP4+ probability analysis. The NOESY correlation of H-10/H₃-15/H-2β (δ 1.74) indicated the β-orientation of these protons. The presence of NOESY cross-peaks between H-7 and H-2α (δ 1.32) indicated they are α-orientated. The configuration assignment was confirmed by the 100% possibility of DP4+ analysis. On the basis of the data described above, the configuration of 8 was established to be 1S*,7S*,8R*,10S*.

In our earlier investigation, we found that the methanol extract of Clavularia spp. had an apoptotic effect on oral cancer cells [10]. Hence, most of the isolated compounds were evaluated in vitro for their antiproliferative effect against oral cancer cells (Ca9-22) using cellular ATP assay. As shown in Table S22, new iodinated dolabellane 7 exhibited strong cytotoxic effects with an IC₅₀ value of 15.7 μM, while the eudensamane-type sesquiterpenes were less active. It is noted that the cytotoxic effect of compound 15 (IC₅₀ = 24.9 μM) was seven times higher than that of 16 (IC₅₀ = 166.7 μM), which implied the position of the C=C bond might change the bioactivity dramatically. For the halogenated dolabellanes 6, 7, and 13, the iodinated one showed the best cytotoxic activity and the chlorinated one was the weakest. Moreover, clasamane E (5) having a peroxide bridge showed a relatively good cytotoxic activity against the Ca9-22 cell among all isolated eudensamane-type sesquiterpene lactones.

3. Materials and Methods

3.1. General

Merck KGaA (Darmstadt, Germany) cellite 545 (0.02–0.1 mm) and silica gel 60 (0.015–0.040 mm) were used for dry sample and flash column chromatography, respectively. Phenomenex (Torrance, CA, USA) C₁₈, phenyl-hexyl, and biphenyl columns were used for high-performance liquid chromatography (HPLC). The Shimadzu (Kyoto, Japan) HPLC instrument consisted of an LC-40D solvent delivery module, DGU-405 degassing unit, CBM-40 system controller, CTO-40S column oven, SPD-M40 photo diode array detector, and FRC-10A fraction collector. A Jasco (Tokyo, Japan) V-650 spectrophotometer was used for measuring UV data. A Jasco FT/IR-4X spectrophotometer was chosen for measuring IR data. A Jasco J-815 CD spectrometer was used for recording the circular dichroism data. Specific optical rotation was measured by a Jasco P-2000 polarimeter. NMR spectra were obtained from Varian (Palo Alto, CA, USA) Mercury Plus 400 MHz and VNMRS 600 MHz FT-NMR spectrometers. A Bruker (Bremen, Germany) APEX II spectrometer was used for detecting HRSIMS.

3.2. Animal Material

The coral materials were collected in May 2021 off the coast of Green Island, Taiwan. Coral specimens were identified as Clavularia spp. by Dr. Yuan-Bin Cheng. A voucher specimen (code: CI2021) was given, and the specimens were deposited at the Department of Marine Biotechnology and Resources, National Sun Yat-sen University, Kaohsiung, Taiwan. It is noted that the coral materials were previously identified as Clavularia inflata [10]. However, the materials contained more than one species and can only be recognized as Clavularia spp.

3.3. Extraction and Isolation

Coral materials were lyophilized and immersed in EtOH at room temperature for three days (thrice) to provide an EtOH extract (124.6 g). This extract was partitioned between EtOAc and H₂O. The EtOAc soluble extract (34.8 g) was further partitioned with hexanes and 75% MeOH. The 75% MeOH layer (11.1 g) was first separated by a silica gel flash column and stepwise eluted with hexanes/EtOAc/MeOH (8/1/0 to 0/0/1) to give nine fractions (A–I). Further silica gel column chromatography of fraction B (622.8 mg), stepwise eluting from hexanes/acetone (40/1) to pure acetone, yielded nine subfractions (B1–B9). Subfraction B2 (55.8 mg) was purified by RP-HPLC (C₁₈ column) using 80% MeOH as eluent to give compounds 1 (3.3 mg) and 2 (1.9 mg). Subfraction B6 (373.7 mg) was isolated by a silica gel open column stepwise eluted with hexanes/EtOAc/MeOH (100/10/1 to 0/0/1) to give fractions B6A–B6D and compound 10 (4.2 mg). Fraction B6B (22.2 mg) was separated by RP-HPLC (phenyl hexyl column) eluting with 80% MeOH to yield compound 15 (8.0 mg). Fraction B6C (307.1 mg) was repurified by RP-HPLC (C₁₈ column) eluting with MeOH/H₂O (3/2 to 0/1), and compounds 17 (7.9 mg) and 18 (3.7 mg) were obtained. Fraction C (494.7 mg) was fractionated over a silica gel open column stepwise eluted with hexanes/CH₂Cl₂/acetone (80/20/0 to 0/0/1) to afford subfractions (C1–C7). Subfraction C4 (104.4 mg) was repurified by a silica gel open column and stepwise eluted by hexanes/EtOAC (15/1 to 0/1) to give fractions C4A–C4F. The RP-HPLC (phenyl hexyl column) separation of fraction C4D (29.9 mg) eluting with 65% MeOH produced compounds 3 (0.6 mg), 4 (5.8 mg), and 5 (1.0 mg). Subfraction C5 (95.0 mg) was applied to a silica gel open column stepwise eluted by hexanes/acetone (15/1 to 0/1) to give five fractions (C5A–C5E). Fraction C5A (18.1 mg) was purified by RP-HPLC (phenyl hexyl column) eluting with 75% MeOH to give compound 19 (5.6 mg). Fraction C5B (20.4 mg) was successively isolated by RP-HPLC (phenyl hexyl column) eluting with 75% MeOH to yield compound 16 (6.0 mg). Subfraction C6 (136.1 mg) was chromatographed on a silica gel column (stepwise eluted by hexanes/acetone/MeOH 6/1/0 to 0/0/1) to obtain fractions C6A–C6C. Fraction C6B (95.0 mg) was repeatedly purified by RP-HPLC (C₁₈ column) with decreasing polarity of MeOH to yield compound 9 (1.6 mg). Fraction D (2.7 g) was further fractionalized into six subfractions (D1–D6) by using a silica gel column stepwise eluted with hexanes/EtOAc/MeOH (100/10/1 to 0/0/1). Subfraction D3 (702.8 mg) was isolated by a silica gel column stepwise eluted with CH₂Cl₂/acetone/MeOH (120/1/0 to 0/0/1), and a diterpenoid-enriched fraction D3A was obtained. Fraction D3A (278.1 mg) was repurified by a silica gel column stepwise eluted with hexanes/acetone/MeOH (20/1/0 to 0/0/1) to afford compound 22 (100.6 mg). Subfraction D4 (986.2 mg) was fractionated by a silica gel column with a gradient of CH₂Cl₂/acetone/MeOH (120/1/0 to 0/0/1) to give six fractions (D4A–D4F). Fraction D4F (395.2 mg) was subjected to a C₁₈ column stepwise eluted with MeOH/H₂O (20/80 to 1/0) to yield eight fractions D4F1–D4F8. Compound 11 (4.1 mg) was purified from fraction D4F4 (11.0 mg) by RP-HPLC (phenyl hexyl column) eluted with MeCN/H₂O (45/55). Fraction D4F7 (99.8 mg) was repurified by a PR-HPLC (C₁₈ column) with a gradient of MeCN/H₂O (40/60 to 55/45) to afford compounds 12 (1.1 mg) and 14 (17.1 mg). Subfraction D5 (731.5 mg) was separated by a silica gel column and eluted with a gradient of hexanes/CH₂Cl₂/MeOH (80/10/1 to 0/0/1) to yield fraction D5A (382.4 mg). Fraction D5A was further separated by another silica gel column stepwise eluted with hexanes/acetone/MeOH (10/1/0 to 0/0/1), and the fraction D5A3 was obtained. Fraction D5A3 (89.9 mg) was then subjected to a C₁₈ column stepwise eluted with MeOH/H₂O (3/7 to 0/1) to give fraction D5A3F. Compound 20 (0.6 mg) was finally isolated from fraction D5A3F (22.3 mg) by RP-HPLC (biphenyl column) with a gradient of MeCN/H₂O (40/60 to 55/45). Subfraction E (1.1 g) was divided into seven fractions (E1–E7) by a silica gel column stepwise eluted with hexanes/CH₂Cl₂/MeOH (15/1/0 to 0/0/1). Fraction E2 (308.3 mg) was separated by another silica gel column stepwise eluted with hexanes/acetone/MeOH (15/1/0 to 0/0/1), and eight fractions (E2A–E2H) were produced. Fraction E2E (82.8 mg) was applied to RP-HPLC (C₁₈ column) eluted with MeCN/H₂O using a gradient from 25/75 to 60/40, resulting in the isolation of 7 (4.1 mg), 21 (43.0 mg) and a subfraction E2E10. Compounds 6 (0.8 mg) and 8 (0.9 mg) were purified from subfraction E2E10 (4.3 mg) by RP-HPLC (biphenyl column) using MeCN/H₂O (45/55) as eluent. Fraction E2F (73.3 mg) was separated by RP-HPLC (C₁₈ column) eluted with MeCN/H₂O (55/45) to yield compound 13 (26.2 mg) and the subfraction E2F4. Subfraction E2F4 (5.4 mg) was repurified by RP-HPLC (biphenyl column) eluted with MeCN/H₂O (35/65) to give compound 23 (1.4 mg).

3.4. Spectroscopic Data

Clasamane A (1): colorless oil; ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{25}$ +62 (c 0.05, MeOH); UV λmax (log ε) 282 (3.58), 218 (2.91) nm; IR (neat) v_max 2926, 1751, 1445, 1386, 1311, 1293, 1236, 1174, 1082, 1035 cm⁻¹; ECD λmax(Δε) 262 (+1.62), 225 (+161) nm; ¹H and ¹³C NMR data are presented in Table 1; HRESIMS m/z 299.1616 [M + Na]⁺ (calcd for C₁₇H₂₄NaO₃, 299.1618).

Clasamane B (2): colorless oil; ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{25}$ −9 (c 0.05, MeOH); IR (neat) v_max 2925, 1752, 1552, 1441, 1383, 1294, 1235, 1168, 1114, 1043 cm⁻¹; ECD λmax(Δε) 248 (−0.60), 214 (−1.26) nm; ¹H and ¹³C NMR data are presented in Table 1 and Table 2; HRESIMS m/z 299.1616 [M + Na]⁺ (calcd for C₁₇H₂₄NaO₃, 299.1618).

Clasamane C (3): colorless oil; ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{25}$ −315 (c 0.04, MeOH); UV λmax (log ε) 282 (3.24), 214 (3.74) nm; IR (neat) v_max 2925, 2858, 1758, 1448, 1373, 1228, 1175, 1125, 1083, 1025 cm⁻¹; ECD λmax(Δε) 274 (−4.58), 234 (+1.60), 209 (−1.59) nm; ¹H and ¹³C NMR data are presented in Table 1 and Table 2; HRESIMS m/z 341.1362 [M + Na]⁺ (calcd for C₁₈H₂₂NaO₅, 341.1359).

Clasamane D (4): colorless oil; ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{25}$ −263 (c 0.05, MeOH); UV λmax (log ε) 277 (3.55), 216 (3.95) nm; IR (neat) v_max 2975, 2928, 1763, 1443, 1378, 1320, 1233, 1175, 1088, 1026 cm⁻¹; ECD λmax(Δε) 274 (−10.61), 234 (+3.04), 211 (−3.36) nm; ¹H and ¹³C NMR data are presented in Table 1 and Table 2; HRESIMS m/z 355.1518 [M + Na]⁺ (calcd for C₁₉H₂₄NaO₅, 355.1516).

Clasamane E (5): colorless oil; ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{25}$ +14 (c 0.05, MeOH); UV λmax (log ε) 217 (3.72) nm; IR (neat) v_max 2928, 1758, 1448, 1378, 1295, 1237, 1171, 1113, 1042 cm⁻¹; ECD λmax(Δε) 290 (−0.08), 246 (+1.24), 207 (−1.81) nm; ¹H and ¹³C NMR data are presented in Table 1 and Table 2; HRESIMS m/z 387.1414 [M + Na]⁺ (calcd for C₁₉H₂₄NaO₇, 387.1414).

Clabellane A (6): colorless oil; ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{25}$ +36 (c 0.05, MeOH); UV λmax (log ε) 204 (3.85) nm; IR (neat) v_max 3432, 2966, 1725, 1646, 1447, 1383, 1252, 1169, 1131 cm⁻¹; ECD λmax(Δε) 204 (+1.38) nm; ¹H and ¹³C NMR data are presented in Table 2 and Table 3; HRESIMS m/z 439.1455 [M + Na]⁺ (calcd for C₂₀H₃₃BrNaO₄, 439.1454).

Clabellane B (7): colorless oil; ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{25}$ +5 (c 0.05, MeOH); UV λmax (log ε) 204 (3.97) nm; IR (neat) v_max 3412, 2963, 2930, 1726, 1645, 1451, 1381, 1243, 1171, 1126, 1045 cm⁻¹; ECD λmax(Δε) 208 (+1.68) nm; ¹H and ¹³C NMR data are presented in Table 2 and Table 3; HRESIMS m/z 487.1318 [M + Na]⁺ (calcd for C₂₀H₃₃INaO₄, 487.1316).

Clabellane C (8): colorless oil; ${\left[\mathsf{\alpha }\right]}_{\mathrm{D}}^{25}$ +50 (c 0.05, MeOH); UV λmax (log ε) 206 (3.85) nm; IR (neat) v_max 3412, 2963, 2930, 1726, 1645, 1451, 1381, 1243, 1171, 1126, 1045 cm⁻¹; ECD λmax(Δε) 203 (−2.81) nm; ¹H and ¹³C NMR data are presented in Table 2 and Table 3; HRESIMS m/z 359.2193 [M + Na]⁺ (calcd for C₂₀H₃₂NaO₄, 359.2193).

3.5. Cytotoxicity Assays

The cell viability (IC₅₀) at 72 h of oral cancer Ca9-22 cells (HSRRB, Ibaraki, Osaka, Japan) [18] was assessed by an ATP detection kit (PerkinElmer Life Sciences, Boston, MA, USA) [19,20] and measured by a luminometer (Berthold Technologies GmbH & Co., Bad Wildbad, Germany). The data are provided as means ± SD in three independent experiments.

4. Conclusions

Although the natural product investigation of octacoral Clavularia spp. started last century [3], new marine natural products have been successively identified from this genus to date [5,21,22]. In the current study, 23 marine natural products, including eight new compounds, were identified. In view of chemical structure, the isolates can be divided into eudensamane-type sesquiterpene lactones (1–5 and 9–12) and dolabellane-type diterpenes (6–8 and 13–22). In terms of bioactivity, the dolabellane-type diterpenes demonstrate better cytotoxic activity than eudensamane-type sesquiterpene lactones. Our findings support our previous investigation and prove that the marine soft coral of the genus Clavularia is a rich source for the identification of cytotoxic compounds.