Six Undescribed Capnosane-Type Macrocyclic Diterpenoids from South China Sea Soft Coral Sarcophyton crassocaule: Structural Determination and Biological Evaluation

Six undescribed capnosane-type macrocyclic diterpenes sarcocrassolins A–F (1–6) and one related known analog pavidolide D (7) were isolated from Sarcophyton crassocaule, a soft coral collected off the Nansha Islands, in the South China Sea. Their complete structures, relative configurations and absolute configurations were established through comprehensive spectroscopic analysis, quantum mechanical nuclear magnetic resonance (QM-NMR) and single-crystal X-ray diffraction. Sarcocrassolins D (4) and E (5) showed inhibitory activity against lipopolysaccharide (LPS)-stimulated inflammatory responses in RAW264.7 cells with IC50 values of 76.8 ± 8.0 μM and 93.0 ± 3.8 μM, respectively.

Soft corals of genus Sarcophyton are widely distributed in shallow waters of the Mediterranean Sea, Red Sea, Arctic and Indo-Pacific region [22].Besides, the genus Sarcophyton, which includes many species, is a rich source of bioactive natural metabolites.Most of these species of the genus Sarcophyton have undergone thorough chemical investigations [23], including the species Sarcophyton crassocaule.Despite the fact that many diterpenoids were isolated from S. crassocaule, capnosane-type macrocyclic diterpenes have not been encountered.In our continuous search for pharmaceutical leads from structurally diverse secondary metabolites of soft corals [24], six undescribed capnosane-type macrocyclic diterpenes named sarcocrassolins A-F (1-6) and one related known analogue, pavidolide D (7), (Figure 1) were isolated from S. crassocaule, a soft coral collected off the Nansha Islands.This is the first report of capnosane-type macrocyclic diterpenes in the soft coral S. crassocaule.Herein, we discuss the isolation, structural elucidation, biological activity and plausible biosynthetic pathway of these capnosane-type macrocyclic diterpenes.

Structure Elucidation of New Compounds 1-6
Sarcocrassolin A (1) was isolated as a white powder.The HRESIMS data at m/z 391.2820 (calculated as 391.2819 for C22H40O4Na, [M + Na] + ) revealed the molecular formula of 1 to be C22H40O4 with three indexes of hydrogen deficit, which was validated by the 13 C NMR and DEPT data.The presence of hydroxy and olefinic groups was suggested by the IR absorption bands at 3418 and 1638 cm −1 , respectively.Five methyl groups (δH 0.99, 1.07, 1.08, 1.10, 1.26), two methoxy groups (δH 3.19, 3.22), one oxymethine (δH 3.59) and one olefinic proton (δH 5.02) were detected in the 1 H NMR spectrum of 1 (Table 1).The  2).A double bond in 1 accounted for one degree of unsaturation and indicated that the structure had two rings as a result of the remaining two indexes of hydrogen deficit.Because 1 and pavidolide D (7) were co-isolated secondary metabolites, their similar NMR data suggested that 1 was most likely a diterpene of

Structure Elucidation of New Compounds 1-6
Sarcocrassolin A (1) was isolated as a white powder.The HRESIMS data at m/z 391.2820 (calculated as 391.2819 for C 22 H 40 O 4 Na, [M + Na] + ) revealed the molecular formula of 1 to be C 22 H 40 O 4 with three indexes of hydrogen deficit, which was validated by the 13 C NMR and DEPT data.The presence of hydroxy and olefinic groups was suggested by the IR absorption bands at 3418 and 1638 cm −1 , respectively.Five methyl groups (δ H 0.99, 1.07, 1.08, 1.10, 1.26), two methoxy groups (δ H 3.19, 3.22), one oxymethine (δ H 3.59) and one olefinic proton (δ H 5.02) were detected in the 1 H NMR spectrum of 1 (Table 1).The 13   2).A double bond in 1 accounted for one degree of unsaturation and indicated that the structure had two rings as a result of the remaining two indexes of hydrogen deficit.Because 1 and pavidolide D (7) were co-isolated secondary metabolites, their similar NMR data suggested that 1 was most likely a diterpene of the capnosane-type.Following that, the planar structure of 1 was ascertained by the 1 H-1 H COSY and HMBC spectra.First of all, four fragments a-d (Figure 2) were established by comprehensive analysis of the   3.22) to C-12 (δ C 79.4) were utilized to infer the position of two methoxy groups, one at C-8 and the other at C-12.Thus, as can be seen in Figure 1, the complete planar structure of 1 was determined.the capnosane-type.Following that, the planar structure of 1 was ascertained by the 1 H-1 H COSY and HMBC spectra.First of all, four fragments a-d (Figure 2) were established by comprehensive analysis of the     The significant ROESY cross peaks between H-2 (δ H 5.02) and H 3 -16 (δ H 1.08) indicated the double-bond ∆ 1 in 1 in E-geometry.The ROESY cross peaks between H-3 (δ H 2.49)/H-7 (δ H 2.88), H-2 (δ H 5.02)/H 3 -18 (δ H 1.10), H-2 (δ H 5.02)/H 3 -19 (δ H 1.07) and the absence of ROESY correlation between H-3 (δ H 2.49)/H 3 -18 (δ H 1.10), H-2 (δ H 5.02)/H-7 (δ H 2.88), H-2 (δ H 5.02)/H-3 (δ H 2.49) indicated that H-3 and H-7 were at the same side and arbitrarily assigned as being in α-configuration, whereas H 3 -18 and H 3 -19 were assigned as being in β-configuration (Figure 3).The ROESY experiment did not allow for determining the relative configuration of C-8, C-11 and C-12 due to the presence of a highly flexible macrocycle.As a result, QM-NMR employing DP4+ probability analysis was carried out.This is a potent tool whose application for capnosane-type diterpenes has been proven to be reliable [3].QM-NMR calculations on eight potential isomers 1a-1h (Figure 4) were then performed.According to the experimental results, the NMR data of compound 1 offered the highest degree of matching for 1g (3S*, 4S*, 7S*, 8S*, 11S*, 12S*) (DP4+ probability 99.81%).Therefore, the 8S* configuration was consistent with the correlations observed in the ROESY spectrum.Finally, the structure of compound 1 was determined by the analysis presented above and represented in Figure 1.
Sarcocrassolin D (4) was obtained as a colorless oil.Its molecular formula was established as C 21 H 38 O 4 by HRESIMS data at m/z 731.5432 (calculated as 731.5432 for C 42 H 76 O 8 Na, [2M + Na] + ).The analysis of the NMR data of 4 showed that 4 mostly resembled 3. The major difference between 4 and 3 was at the C-8 position, where the methoxy group in 3 was substituted by a hydroxy group in 4.This alteration was further validated by their 14-mass unit difference.The double-bond ∆ 1 in 4 was proposed to have E-geometry, in accordance with the ROESY cross peak between H-2 (δ H 5.21) and H 3 -16 (δ H 1.11).Furthermore, the relative configuration of C-3, C-4 and C-7 could be tentatively established as 3S*, 4S* and 7R*, respectively, based on the similar crucial ROESY correlation between 3 and 4 (Figure 3).Finally, eight potential isomers 4a-4h (Figure S58) were calculated to explicate the spatial arrangement of C-11, C-12 and C-8.The correlation coefficient R 2 = 0.9835 (DP4+ probability, 100%) suggested that 4e (3S*, 4S*, 7R*, 8S*, 11R*, 12R*) was more compatible with the experimental NMR results.Consequently, the relative configuration of 4 was defined as shown in Figure 1.
The HRESIMS ion peak at m/z 345.2398 (calculated as 345.2400 for C 20 H 34 O 3 Na, [M + Na] + ) revealed the molecular formula of sarcocrassolin E (5) to be C 20 H 34 O 3 with four indexes of hydrogen deficit.The existence of hydroxy and olefinic groups could be determined by the characteristic infrared absorption bands at 3393 and 1650 cm −1 , respectively.The 13 C NMR spectrum revealed twenty carbon signals, which were classified by DEPT and HSQC spectra and then attributed to four quaternary carbons, five methines, six methylenes and five methyls.One double bond and one epoxy ring were composed of two methines, C-2 (δ C 120.4, δ H 5.31) and C-11 (δ C 61.7, δ H 3.16), as well as two quaternary carbons, C-1 (δ C 146.5) and C-12 (δ C 60.1), which contributed to the two indexes of hydrogen deficit.The remaining two indexes of hydrogen deficit thus revealed that the molecule contained two rings.Following that, the entire planar structure of 5 could be constructed using the 2D NMR data as depicted in Figure 2. When the NMR data were further analyzed, it became clear that compound 5 and sarcophytrol B [7], a capnosane-type macrocyclic diterpene that had previously been identified, were highly similar.The major difference was the configuration of double-bond ∆ 1 .In fact, that had an E configuration in 5 and Z configuration in sarcophytrol B, which was established by the ROESY cross peaks between H-2 and H 3 -16 in compound 5.It was also possible to establish the relative configurations of C-3, C-4 and C-7 as 3S*, 4S* and 7R*, respectively, by the similar key ROESY cross peaks observed in compounds 4 and 5. Consequently, four possible isomers 5a-5d (Figure S59) were calculated.Based on the experimental results, the NMR data of compound 5 provided the highest degree of matching for 5d (3S*, 4S*, 7R*, 8S*, 11S*, 12R*), with 100% probabilities.Consequently, the relative configuration of 5 was defined as 3S*, 4S*, 7R*, 8S*, 11S* and 12R* (Figure 1).
Sarcocrassolin F (6), isolated as a colorless oil, had a molecular formula C 21 H 36 O 3 based on the HRESIMS data at m/z 359.2557 (calculated as 359.2557 for C 21 H 36 O 3 Na, [M + Na] + ).The analysis of its 1 H and 13 C NMR spectral characteristics data showed that 6 displayed extensive resemblance to 1, except for the existence of an exocyclic double bond at C-8, C-19 in 6 instead of a methoxy and a methyl at C-8 in 1, which was validated by key HMBC cross peaks from H-19 to C-7/C-8/C-9.According to the noticeable ROESY cross peak of H-2/H 3 -16, the double-bond ∆ 1 in 6 was proven to be of the E-geometry.The ROESY cross peak between H-2 and H 3 -18, along with the lack of cross peaks of H-2/H-3 and H-3/H 3 -18, indicated that H-3 and H 3 -18 were oriented on the opposite side of the molecule.Due to the insufficient signals, ROESY correlations were unable to ascertain the relative configuration of C-7.However, the nearly identical coupling constant of J H-3/H-7 (5.9 Hz) in 6 and J H-3/H-7 (5.3 Hz) in 1 revealed that H-3 and H-7 had the same orientation in both molecules.Finally, the QM-NMR using DP4+ probability analysis was carried out for the four potential isomers 6a-6d (Figure S60) to explicate the spatial arrangement of C-11 and C-12.The experimental data indicated that 6d (3S*, 4S*, 7S*, 11S*, 12R*) was more consistent with the correlation coefficient R 2 = 0.9975 (DP4 + probability, 100%).Consequently, the relative configuration of 6 was depicted as drawn in Figure 1.
Compound 7 was identified as pavidolide D on the basis of the MS and NMR data analyses and comparison with the NMR data from a previously reported article [14].
Because all compounds 1-7 possessed the same 5/11-fused bicyclic carbon skeleton and were obtained from the same animal material, it was reasonable to assume that they originated from the same biosynthetic pathway.According to their structural characteristics, a potential biosynthetic pathway of compounds 1-7 has been postulated (Scheme 1).The precursor geranylgeranyl diphosphate (GGPP) via 14,1 cyclization and dehydrogenation generates intermediate cembrene A. Subsequently, the 5/11-fused bicyclic carbon skeleton was generated by epoxidation and transannular cyclization from cembrene A. Finally, all compounds 1-7 were obtained after undergoing a series of reactions (such as methylation, epoxidation and dehydrogenation).

In Vitro Biological Assay
In biological assays, all the abovementioned compounds were tested for their cytotoxic and anti-inflammatory activities.In cytotoxicity screening assay, all compounds were assayed against five cell lines: SGC-7901 (human gastric cells), K562 (human myeloid leukemia cells), BEL-7402 (human hepatocellular carcinoma cells), HeLa (human cervical carcinoma cells) and A549 (human lung cancer cells).Unfortunately, none of these com-

In Vitro Biological Assay
In biological assays, all the abovementioned compounds were tested for their cytotoxic and anti-inflammatory activities.In cytotoxicity screening assay, all compounds were assayed against five cell lines: SGC-7901 (human gastric cells), K562 (human myeloid leukemia cells), BEL-7402 (human hepatocellular carcinoma cells), HeLa (human cervical carcinoma cells) and A549 (human lung cancer cells).Unfortunately, none of these compounds displayed obvious cytotoxic activity at a concentration of 20.0 µg/mL.In antiinflammatory testing, compounds 4 and 5 exhibited weak inhibitory activity of NO release in LPS-stimulated RAW264.7 cells, with the IC 50 values of 76.8 ± 8.0 µM and 93.0 ± 3.8 µM, respectively (Table 3).The positive control used was quercetin, which had an IC 50 value of 12.7 ± 2.4 µM.The cell viability of RAW264.7 cells upon treatment with the compounds was also determined by the MTT method, and the result indicated that compounds 3 and 4 had no cytotoxicity against RAW264.7 cells at a concentration of 100.0 µM.

General Experimental Procedures
Optical rotation was performed on a Modular Circular Polarimeter (Anton Paar, Graz, Austria).The IR spectra were recorded on a Nicolet 380 infrared spectrometer (Thermo Electron Corporation, Madison, WI, USA).UV and CD data were collected on a MOS-500 spectrometer (Biologic, Seyssinet-Pariset, France).NMR spectra were obtained on a Bruker AV-500 NMR spectrometer (Bruker, Bremen, Germany), using the TMS as an internal standard.HRESIMS spectra were measured on an API QSTAR Pulsar mass spectrometer (Bruker, Bremen, Germany).Analytical HPLC was recorded on an Agilent Technologies 1260 Infinity II with an Agilent DAD G1315D detector (Agilent, Palo Alto, CA, USA).Semipreparative HPLC was performed on an ODS column (COSMOSIL-packed C 18 , 5 mm, 10 mm × 250 mm).Column chromatography was performed on a Sephadex LH-20 (Merck, Darmstadt, Germany) and Silica gel (60-80, 200-300 and 300-400 mesh, Qingdao Marine Chemical Co., Ltd., Qingdao, China).Analytical TLC was conducted on precoated silica gel GF254 plates (Qingdao Marine Chemical Co., Ltd., Qingdao, China), and the spots were detected by spraying with 10% H 2 SO 4 in EtOH and subsequent heating.

Animal Materials
The soft coral S. crassocaule was collected in October 2018 at a depth of −20 m off the Nansha Islands, Hainan Province, China.The fresh sample was immediately frozen.The animal material was identified by Prof. X.-B.Li (Hainan University).A voucher specimen (No. 18-NS-10) was deposited at the Institute of Tropical Bioscience and Biotechnology, Chinese Academy of Tropical Agricultural Sciences.

X-ray Crystallographic Analysis
The solvent methanol/H 2 O (20:1) was used to create an acceptable crystal of compound 2 (0.15 × 0.12 × 0.1 mm 3 ) at ambient temperature.The crystallographic data were collected at 170 K on a diffractometer Rigaku Oxford Diffraction Supernova Dual Source, Cu at Zero, equipped with an AtlasS2 CCD using Cu Kα radiation (λ = 1.54184Å) by using a w scan mode.CrysAlisPro (version: 1.171.38.41) was used to process the data, and the structures were solved using direct methods with the SHELXT structure solution program via intrinsic phasing algorithm using Olex2 software (version: 1.5).The nonhydrogen atoms in the trial structure were detected and then refined anisotropically with SHELXL-2018 using a full-matrix least-squares program based on F 2 ; F 2 was used to obtain the weighted R factor, wR and goodness-of-fit S values.The placements of the hydrogen atoms were fixed geometrically at the calculated distances and permitted to ride on the parent atoms.Detailed crystallographic data are provided in the Supplementary Materials (Table S11) and have been deposited at the Cambridge Crystallographic Data Centre for inspection (deposition numbers: CCDC 2291468 for 2).

Computational Details
The torsional sampling (MCMM) method and the OPLS_2005 force field were used for carrying out a conformational search inside a 21 kJ/mol energy window.The conformers were reoptimized using the IEFPCM solvent model for chloroform at the B3LYP/6-31G(d) level for conformers over 1% Boltzmann populations.The reoptimized geometries' position at the energy minima was further verified by frequency analysis.Then, as advised for DP4+, NMR calculations were carried out at the PCM/mPW1PW91/6-311+G(d,p) level.The GIAO method was used to compute NMR shielding constants.After obtaining a Boltzmann distribution for each stereoisomer, shielding constants were averaged and correlated with the experimental results.

Cytotoxic Detection
The cytotoxic activity bioassay of all the abovementioned compounds was carried out as described in our previous paper [25].Cisplatin was used as a positive control.

Anti-Inflammatory Assay
In the bioassay for anti-inflammation, all the abovementioned compounds were assayed for the inhibition of NO released in lipopolysaccharide-stimulated RAW264.7 cells (bought from the Stem Cell Bank of the Chinese Academy of Sciences) by the Griess method [26].The mouse macrophage cells were cultured in 96-well plates at a concentration of 5 × 10 4 cells per mL (100 µL each well) in a humidified incubator of 5% CO 2 at 37 • C for 24 h.The transfected cells were then stimulated using LPS (50 µL, 500 ng/mL) for 24 h after being treated with 50 µL solution of the tested compound at different concentrations (100.0,50.0, 25.0, 12.5 µM) for 1 h.Then, 100 µL of supernatants were taken from each well, and 100 µL of Griess reagent (40 mg/mL) was added into a new 96-well plate.Finally, using a microplate reader, the absorbance of each well was measured at the wavelength of 540 nm.The IC 50 was calculated using GraphPad Prism 9.0.The positive control employed was quercetin.The cytotoxicity of compounds against RAW264.7 cells was determined by the MTT method.

Conclusions
In conclusion, six uncommon capnosane-type macrocyclic diterpenes, sarcocrassolins A-F (1-6), and one related known analog pavidolide D (7) were isolated from the soft coral S. crassocaule collected off the Nansha Islands.Due to the presence of a highly flexible macrocycle, the configuration assignments of capnosane-type macrocyclic diterpenes are still confronted with huge challenges.In our continuing efforts, the absolute configuration of compound 2 was assigned based on an X-ray crystallography study, and the relative configurations of the remaining compounds were established through QM-NMR using DP4+ probability analysis.Despite the fact that the soft coral S. crassocaule has been acknowledged as a major source of cembrane-type diterpenes, the capnosane-type macrocyclic diterpenes have not been previously reported.This is the first time that capnosane-type macrocyclic diterpenes were described in the soft coral S. crassocaule.Consequently, the soft coral S. crassocaule could be a potential source of structurally diverse capnosane-type macrocyclic diterpenes.In anti-inflammatory testing, all isolated compounds were assayed for the inhibitory activity on NO release in LPS-induced RAW264.7 cells.Sarcocrassolin D (4) and sarcocrassolin E (5) exhibited weak inhibitory effects.Furthermore, all compounds were tested for cytotoxicity.Unfortunately, none of the compounds demonstrated notable activity.

Figure 4 .
Figure 4. Structures of isomers of compound 1.Sarcocrassolin B (2), isolated as a colorless crystal, had a molecular formula of C 21 H 36 O 3 according to its HRESIMS data at m/z 695.5211 (calculated as 695.5221 for C 42 H 72 O 6 Na, [2M + Na] + ), with four degrees of unsaturation.Additionally, the characteristic infrared absorption bands at 3421 and 1602 cm −1 provided evidence for the presence of hydroxyl and olefinic groups.The 1 H NMR data revealed the signals of five methyl groups (δ H 1.06, 1.07, 1.08, 1.15, 1.69), one methoxy group (δ H 3.12) and two olefinic protons (δ H 5.00, 5.38).The 13 C NMR data showed the presence of 21 signals, which comprised two pairs of olefinic carbons, five methyl signals, five methylene signals, four sp 3 methines, two sp 3 quaternary carbons and one methoxy carbon.Based on a comprehensive analysis of its NMR spectra, compound 2 almost resembled trocheliophol G, a capnosane-type

Figure 5 .
Figure 5. Single-crystal X-ray structure of 2 (the ellipsoids are shown at a 30% probability level).