Cembrane Diterpenes Possessing Nonaromatic Oxacycles from the Hainan Soft Coral Sarcophyton mililatensis

Documents on the chemical composition of the soft coral Sarcophyton mililatensis are sparse. The present investigation of the Hainan soft coral S. mililatensis resulted in the discovery of six new cembrane diterpenes, sarcoxacyclols A–F (1–6) and four known analogs (7–10). Their structures were elucidated by extensive spectroscopic analysis along with a comparison with the data in current literature. The nonaromatic oxacycles in their structures were the rarely found tetrahydrofuran ether across C-1 and C-12 and tetrahydropyran ether across C-1 and C-11, respectively. Moreover, the absolute configuration of compound 4 was established unambiguously by X-ray diffraction analysis using Ga Kα radiation (λ = 1.34139 Å). Based on the biogenetical consideration, the absolute configurations of other five new compounds were tentatively assumed. Assessment of the bioactivity for these secondary metabolites revealed that compound 1 exhibited significant tumor necrosis factor (TNF)-α inhibitory activity (IC50 = 9.5 μmol/L), similar to the positive control dexamethasone (IC50 = 8.7 μmol/L), but no obvious cytotoxicity towards RAW264.7 cells (CC50 > 50 μmol/L). The preliminary molecular docking suggested the crucial roles of the hydroxyl and acetoxyl groups in the computational prediction of the binding mode between the diterpene and the protein.

The assignment of absolute configuration to highly flexible cembrane diterpenes containing diverse functional groups remain challenging. Usually, X-ray crystallography techniques [16] and chemical derivatization approaches using chiral agents [17] are the primary choices [4]. However, the scarcity of isolates hampered the use of the two methods mentioned above in some cases. Instead, chiroptical properties of cembrane diterpenes, including electronic circular dichroism (ECD) aided by quantum chemical calculations, provided alternative approaches to establish the absolute configurations of this type of natural macrocyclic compounds [18]. Among the various calculation methods, the timedependent density functional theory (TDDFT) ECD calculation is a powerful tool that has been frequently used. Till now, the efficiency of TDDFT ECD calculation has been demonstrated in the numerous stereochemical studies of conformationally flexible cembrane diterpenes [19][20][21] and their dimers [22][23][24].
In the course of our ongoing research aiming for structurally novel and pharmacologically active secondary metabolites from South China Sea marine fauna [25], we encountered the soft coral Sarcophyton mililatensis, which was collected from Xigu Island, Hainan Province, China. As far as we know, there are only six reports focused on the chemical constituents of S. mililatensis [26][27][28][29][30][31], indicating documents on chemical composition of this species are sparse. Our previous chemical investigations on the title animals resulted in the discovery of an array of diterpenes with four carbon frameworks [26][27][28][29], involving an unprecedented tricyclo[11.3.0.0 2,16 ]hexadecane carbon skeleton [28]. These studies disclosed the productivity of this species and inspired us to conduct further research on the title species, especially for structurally intriguing diterpenes. In the present study, ten cembrane diterpenes possessed different types of nonaromatic oxacycles. Among them, sarcoxacyclols A-F (1-6) were new compounds ( Figure 1). As follows are the descriptions of the isolation, structural elucidation, and bioactivity screening of these secondary metabolites.
The assignment of absolute configuration to highly flexible cembrane diterpene containing diverse functional groups remain challenging. Usually, X-ray crystallography techniques [16] and chemical derivatization approaches using chiral agents [17] are th primary choices [4]. However, the scarcity of isolates hampered the use of the two methods mentioned above in some cases. Instead, chiroptical properties of cembran diterpenes, including electronic circular dichroism (ECD) aided by quantum chemica calculations, provided alternative approaches to establish the absolute configurations o this type of natural macrocyclic compounds [18]. Among the various calculation meth ods, the time-dependent density functional theory (TDDFT) ECD calculation is a pow erful tool that has been frequently used. Till now, the efficiency of TDDFT ECD calcula tion has been demonstrated in the numerous stereochemical studies of conformationally flexible cembrane diterpenes [19][20][21] and their dimers [22][23][24].
In the course of our ongoing research aiming for structurally novel and pharmaco logically active secondary metabolites from South China Sea marine fauna [25], we en countered the soft coral Sarcophyton mililatensis, which was collected from Xigu Island Hainan Province, China. As far as we know, there are only six reports focused on th chemical constituents of S. mililatensis [26][27][28][29][30][31], indicating documents on chemical com position of this species are sparse. Our previous chemical investigations on the title ani mals resulted in the discovery of an array of diterpenes with four carbon framework [26][27][28][29], involving an unprecedented tricyclo[11.3.0.0 2,16 ]hexadecane carbon skeleton [28] These studies disclosed the productivity of this species and inspired us to conduct fur ther research on the title species, especially for structurally intriguing diterpenes. In th present study, ten cembrane diterpenes possessed different types of nonaromatic oxacy cles . Among them, sarcoxacyclols A-F (1-6) were new compounds ( Figure 1). As fol lows are the descriptions of the isolation, structural elucidation, and bioactivity screen ing of these secondary metabolites.
In our previous work, the absolute configuration of sarcophytrol M (7), a cembrane diterpene possessing a furan ether isolated from the South China Sea soft coral Sarcophyton trocheliophorum, was determined as 1S,4S,11S,12R by the modified Mosher's method [7]. Based on the C-4 epimeric relationship between sarcophytrols M (7) and P (8), the absolute configuration of compound 8 could be assigned as 1S,4R,11S,12R, which was further confirmed by the same Cotton effects at ca. 208 nm in their experimental ECD spectra ( Figure 2) in the present study.
In our previous work, the absolute configuration of sarcophytrol M (7), a cembrane diterpene possessing a furan ether isolated from the South China Sea soft coral Sar cophyton trocheliophorum, was determined as 1S,4S,11S,12R by the modified Mosher's method [7]. Based on the C-4 epimeric relationship between sarcophytrols M (7) and P (8), the absolute configuration of compound 8 could be assigned as 1S,4R,11S,12R, which was further confirmed by the same Cotton effects at ca. 208 nm in their experimenta ECD spectra (Figure 2) in the present study. Sarcoxacyclol A (1), which was obtained as a colorless oil, possessed the molecula formula C22H36O5 on the basis of its HRESIMS ion peak at m/z 403.2448 ([M+Na] + , calcd for C22H36O4Na, 403.2455), requiring five degrees of unsaturation. As revealed by the 1 H and 13 C NMR data (Table 1) 3 , CH 3 -OAc), 171.2 (qC, CO-OAc)] in the molecule of compound 1. These functionalities accounted for three degrees of unsaturation. Thus, the remaining two degrees of unsaturation suggested the bicyclic framework of this compound. In fact, the NMR data of compound 1 closely resembled those of the co-occurring sarcophytrol P (8) [7], which was a cembrane diterpene that possessed a tetrahydrofuran ether across C-1 and C-12 from the South China Sea soft coral S. trocheliophorum, revealing that they were structural analogs. Careful analysis of their NMR data revealed the only difference was the presence of an acetoxyl group at C-11 in compound 1 instead of the hydroxyl in compound 8. was the 11-acetylation derivative of compound 8, which was consistent with their 42 mass unit difference. Considering their structural relationship, the absolute configuration of compound 1 was tentatively supposed to be the same as that of compound 8 ( Figure 1).     Sarcoxacyclol B (2) was isolated as a colorless oil and had a molecular fo C22H34O4 as established by the pseudo-molecular ion peak at m/z 385.2361 ( calcd. for C22H34O4Na, 385.2349) in its HRESIMS spectrum, requiring six degre saturation. The 1 H and 13 C NMR spectra (Table 1)  above-mentioned functionalities accounted for four degrees of unsaturation; the indicated that compound 2 was a bicyclic diterpene. Indeed, the NMR data of co 2 were almost superimposed on those of compound 1, suggesting the structura ity of both compounds. Further comparison of their NMR data revealed that th structures differed at the substituent attached to C-1, which was a propen-2-yl compound 2 whereas a peculiar isopropyl alcohol moiety in compound 1. T pen-2-yl group was assigned to C-1 based on the HMBC correlations from H3-17 to C-1 (δC 88.1), C-15 (δC 149.5) and C-16 (δC 109.4) ( Figure 3). Thus, the planar of compound 2 was determined as the dehydration derivative of compoun cis-orientation of the propen-2-yl group and CH3-20 was deduced from the char NOE cross-peak of H-16b (δH 4.96)/H3-20 (δH 1.18). The strong NOE correlations H-2 (δH 5.68) and H3-17 (δH 1.71) and between H-3 (δH 5.92) H3-18 (δH 1.28) imp H3-17 and H3-18 were trans-oriented. These NOE correlations ( Figure 4) indica compound 2 and the co-isolated diterpene sarcophytrol M (7) [7] shared the sa figurations for the four chiral centers C-1, C-4, C-11, and C-12. Tentatively, the configuration of compound 2 was assumed to be 1S,4S,11S,12R.
Sarcoxacyclol C (3) was purified as a colorless oil with a molecular for C22H34O3, which was inferred from the HREIMS data (m/z 346.2507, M + , c C22H34O3, 346.2508). Its 1 H NMR data (Table 1)   All the above-mentioned functionalities accounted for four degrees of unsaturation; the left two indicated that compound 2 was a bicyclic diterpene. Indeed, the NMR data of compound 2 were almost superimposed on those of compound 1, suggesting the structural similarity of both compounds. Further comparison of their NMR data revealed that their gross structures differed at the substituent attached to C-1, which was a propen-2-yl group in compound 2 whereas a peculiar isopropyl alcohol moiety in compound 1. This propen-2-yl group was assigned to C-1 based on the HMBC correlations from H 3 -17 (δ H 1.71) to C-1 (δ C 88.1), C-15 (δ C 149.5) and C-16 (δ C 109.4) ( Figure 3). Thus, the planar structure of compound 2 was determined as the dehydration derivative of compound 1. The cis-orientation of the propen-2-yl group and CH 3 -20 was deduced from the characteristic NOE cross-peak of H-16b (δ H 4.96)/H 3 -20 (δ H 1.18). The strong NOE correlations between H-2 (δ H 5.68) and H 3 -17 (δ H 1.71) and between H-3 (δ H 5.92) H 3 -18 (δ H 1.28) implied that H 3 -17 and H 3 -18 were trans-oriented. These NOE correlations (Figure 4) indicated that compound 2 and the co-isolated diterpene sarcophytrol M (7) [7] shared the same configurations for the four chiral centers C-1, C-4, C-11, and C-12. Tentatively, the absolute configuration of compound 2 was assumed to be 1S,4S,11S,12R.
Sarcoxacyclol C (3) was purified as a colorless oil with a molecular formula of C 22   .0, and one olefinic δ C 110.1), three methines (including one oxygenated δ C 75.7, two olefinic δ C 119.9, and 126.2), and six quaternary carbons (including two oxygenated δ C 83.5, 89.6, three olefinic δ C 132.9, 136.6, 150.2, and one carbonyl δ C 171.3) (Table 1). These abovementioned structural features of compound 3 were reminiscent of the co-isolated compound 2. According to the interpretation of the 2D NMR spectra (Figures 3 and 4), compounds 3 and 2 differed by the double bond at C-2/C-3 in compound 2 shifting to C-3/C-4 in compound 3 accompanied with the disappearance of the hydroxyl at C-4. This alteration was supported by the consecutive spinning proton system from H 2 -2 (δ H 2.14 m, 2.35 m) to H-3 (δ H 5.12) along with the key HMBC correlations of H 2 -2/C-1 (δ C 89.6), H 3 -18 (δ H 1.56)/C-3 (δ C 119.9), H 3 -18/C-4 (δ C 136.6), and H 3 -18/C-5 (δ C 38.0) (Figure 3). The NOESY spectra of diterpenes 2 and 3 ( Figure 4) showed similar distributions of NOE cross-peaks, implying they shared the same configurations for the three chiral carbons C-1, C-11, and C-12. Based on the biogenetical consideration, the absolute configuration of compound 3 was tentatively supposed to be 1S,11S,12R, and its structure was shown as depicted in Figure  , which accounted for two degrees of unsaturation. The remaining two degrees of unsaturation strongly indicated one bicyclic structure incorporating a nonaromatic oxacycle for 4. Indeed, compound 4 was identical in all respects with the co-occurring sarcophytrol Q (9), which was a diterpene possessing a tetrahydropyran ether across C-1 and C-11 previously isolated from the South China Sea soft coral S. trocheliophorum [7]. Their structural difference was in the reversed configuration of C-4, which was deduced from the chemical shift of C-4 upfield, shifting from δ C 74.3 in compound 9 to δ C 72.7 in compound 4. This hypothesis was supported by the NOE correlations between H-2 (δ H 5.20) and both H 3 -16 (δ H 1. 16) and H 3 -18 (δ H 1.37) implying that H 3 -16 and H 3 -18 were co-facial ( Figure 4). After many attempts, suitable single crystals of compound 4 were successfully obtained from the diethyl ether-acetone solvent system to elucidate its absolute configuration. The X-ray diffraction experiment with Ga Kα radiation (λ = 1.34139 Å) unambiguously established that the absolute configuration of compound 4 was 1S,4R,11R,12S (Flack parameter: −0.06 (9), Figure 5). Based on their epimeric relationship, the absolute configuration of compound 9 could be tentatively assumed to be 1S,4S,11R,12S, suggesting the revision of configurations of the C-11 and C-12 of compound 9 reported in the previous work [7].  -15), 81.1 (qC, C-1)], counted for two degrees of unsaturation. The remaining two degrees of un strongly indicated one bicyclic structure incorporating a nonaromatic oxacycl deed, compound 4 was identical in all respects with the co-occurring sarcoph which was a diterpene possessing a tetrahydropyran ether across C-1 and C ously isolated from the South China Sea soft coral S. trocheliophorum [7]. Their difference was in the reversed configuration of C-4, which was deduced from ical shift of C-4 upfield, shifting from δC 74.3 in compound 9 to δC 72.7 in com This hypothesis was supported by the NOE correlations between H-2 (δH 5.20 H3-16 (δH 1.16) and H3-18 (δH 1.37) implying that H3-16 and H3-18 were co-fac 4). After many attempts, suitable single crystals of compound 4 were succe tained from the diethyl ether-acetone solvent system to elucidate its absolute tion. The X-ray diffraction experiment with Ga Kα radiation (λ = 1.34139 Å ) u ously established that the absolute configuration of compound 4 was 1S,4 (Flack parameter: −0.06 (9), Figure 5). Based on their epimeric relationship, th configuration of compound 9 could be tentatively assumed to be 1S,4S,11R gesting the revision of configurations of the C-11 and C-12 of compound 9 r the previous work [7].  3 -19 (δ H 1.38)/C-7 (δ C 66.8), H 3 -19/C-8 (δ C 61.9) and H 3 -19/C-9 (δ C 35.5) revealed the linkage of partial fragments b and c through C-8, along with the fixation of the epoxide group at C-7/C-8. The subunit d was found to be linked to subunits c and a by the oxygenated carbons C-12 and C-1, as deduced from the significant HMBC correlations of H 3 -20 (δ H 1.17) with C-11 (δ C 75.0), C-12 (δ C 71.2), and C-13 (δ C 37.3), and of both H-2 (δ H 5.44) and H-14 (δ H 2.22) with C-1 (δ C 81.8), respectively. A peculiar isopropyl alcohol group was recognized by the mutual HMBC correlations from H 3 -16 (δ H 1.16) and H 3 -17 (δ H 1.11) to C-15 (δ C 74.6). This group was attached to C-1, as indicated by the key HMBC cross-peaks of H 3 -16/C-1 and H 3 -17/C-1. Thus, the carbon framework of compound 5 was established as a cembrane skeleton. Indeed, the 13 C chemical shifts of the C-11-C-12-C-13-C-14-C-1 segment of compound 5 were nearly identical to those of the co-isolate 4, suggesting the hydroxyl group bonded to C-12 and a tetrahydropyran ring bridged C-1 and C-11. Thus, the planar structure of compound 5 was established as an epoxidation derivative of compound 4. The above-mentioned almost identical chemical shifts of the C-11-C-12-C-13-C-14-C-1 segment suggested the same configurations for the three chiral centers C-1, C-11, and C-12 in compounds 4 and 5. The NOE correlations ( Figure 4 Table 2) revealed that the planar structure of compound 6 should be the same as that of compound 5, which was further supported by the extensive analysis of its 2D NMR spectra (Figure 3). The clear NOE cross-peak of H 3 -19 (δ H 1.27)/H-11 (δ H 3.49) in the NOESY spectrum of compound 6 ( Figure 4) suggested the same orientation for CH 3 -19 and H-11 in compound 6, which was different from that of compound 5. The trans geometry of the epoxide ring in compound 6 was indicated by the key NOE crosspeaks of H-7 (δ H 2.95)/H-9a (δ H 1.41) and H 3 -19/H-9b (δ H 2.03). Consequently, compound 6 was established as the C-7/C-8 stereoisomer of compound 5, namely another epoxidation derivative of compound 4, and the absolute configuration was tentatively supposed as 1S,4S,7R,8R,11R,12S.
The origin of compounds 2, 5, and 6 is a matter needing discussion. To determine if these three diterpenes were natural products or derived from their co-occurring precursors during the isolation process, we re-checked the crude extract of the soft coral S. mililatensis by comparisons of the R f values of secondary metabolites in the crude extract with that of pure samples 2, 5, and 6 on co-plate TLC as well as comparisons of their retention times on HPLC. We detected the compounds 2, 5, and 6 present in the Et 2 O-soluble portion of the acetone extract of the specimen. Evidence from these experiments excluded the possibility that the dehydration and epoxidation derivatives 2, 5, and 6 were artifacts obtained during the work-up.
Moreover, several epimeric epoxy analogs were reported as co-occurring isolates in the soft corals ( Figure S50). For instance, epimeric epoxy diterpenes sarcocrassolides A (S5) and B (S6) with close structural similarity coexisted in the South China Sea soft coral S. crassocaule, as shown in Xu et al.'s work [34]. It might be worthy to point out that the relative configuration of sarcocrassolide B (S6) was confirmed by the X-ray diffraction experiment. Sheu et al.'s study on the Formosan soft coral S. crassocaule also displayed the co-occurrence of epimeric epoxy diterpenes named crassocolides K-M (S7-S9) [35]. These five cembrane diterpenes are representatives of two groups of diterpenes with epimeric epoxides at C-3/C-4. Seifert et al. have reported the discovery of diterpenes with epimeric epoxides at C-7/C-8, such as dihydrocembrene C (S10) and sarcophytoxide (S11) from the Indonesian soft coral Sarcophyton ehrenbergi [36]. These findings of co-occurring natural epimeric epoxy analogs indicate that there is possibly more than one enzyme to catalyze the epoxidation reaction in the metabolic process in the soft corals, resulting in the seemingly non-stereospecific natural products.
In general, the discovery of the dehydration and epoxidation derivatives in this study inspired us to explore the terpene biosynthetic gene clusters of this species in future, which may be in high demand in addition to the chemo-and bio-investigations.
There have been no reports assessing the tumor necrosis factor (TNF)-α inhibitory bioactivity of cembrane diterpenes possessing the tetrahydrofuran and tetrahydropyran ethers. Therefore, all the cembrane diterpenes 1−10 were subjected to this bioassay using lipopolysaccharide (LPS)-induced TNF-α protein release in RAW264.7 macrophages. As a result, they exhibited different levels of inhibition ratios, ranging from 5.5% to 52.4% at a concentration of 20 µmol/L. Among them, only the new secondary metabolite 1 was judged an active compound (inhibition ratio 52.4%). Further evaluation showed an IC 50 value of 9.5 µmol/L for compound 1, which was similar to the positive control dexamethasone (IC 50 = 8.7 µmol/L). Notably, diterpene 1 displayed no obvious cytotoxic activity towards RAW264.7 cells (CC 50 > 50 µmol/L). Given the structural differences between compound 1 and its analog compound 2, it appeared that the isopropyl alcohol moiety at C-1 in the bicyclic cembrane structure played an important role in TNF-α inhibitory activity.
Then, a preliminary molecular docking experiment was performed using the highly resolved TNF-α crystal structure (PDB: 2AZ5 with a resolution of 2.10 Å) [37]. As shown in Figure 6, the preliminary computation clearly revealed the hydrogen bonds formed between the hydroxyl group at C-15 and the residue Leu120, between the hydroxyl group at C-4 and the residue Ile58, and between the acetoxyl group at C-11 and the residues Gly121 and Tyr151, which were lying in the active site. Moreover, compound 1 occupied the hydrophobic pocket, where methyl groups at C-7 and C-12 generate π-alkyl stacked interactions with the residues Leu57 and Tyr119. The low binding affinity of compound 1 (−8.02 kcal/mol), which was slightly lower than that of the previously reported ligand 6,7-dimethyl-3-{[methyl-(2-{methyl-[1-(3-trifluoromethyl-phenyl)-1H-indol-3ylmethyl]-amino}-ethyl)-amino]-methyl}-chromen-4-one [37] in the TNF-α crystal structure in this study (−5.31 kcal/mol), indicated that compound 1 was a potential TNF-α inhibitor.
The study for the elaborated mechanism is in the plan.

General Experimental Procedures
Optical rotations were recorded on a Perkin-Elmer 241MC polarimeter. IR spectra were obtained on a Nicolet 6700 spectrometer (Thermo Scientific, Waltham, MA, USA). CD spectra were measured on a JASCO J-810 instrument. NMR spectra were measured on a Bruker DRX-500 or Bruker DRX-600 spectrometer (Bruker Biospin AG, Fällanden, Germany). Chemical shifts (δ) were reported in ppm with reference to the solvent signals, and coupling constants (J) were in Hz. ESIMS spectra were obtained on a Finngan-MAT-95 mass spectrometer. HRESIMS spectra were measured on an Agilent 1290-6545 UHPLC-QTOF mass spectrometer. Commercial silica gel (Qingdao Haiyang Chemical Group Co., Ltd., Qingdao, China, 200-300 and 400-600 mesh), Sephadex LH-20 gel (Amersham Biosciences, Piscataway, NJ, USA) were used for column chromatography, and precoated silica gel plates (Yan Tai Zi Fu Chemical Group Co., Yantai, China, G60 F-254) were used for analytical TLC. Reversed-phase (RP) HPLC was performed on an Agilent 1260 series liquid chromatography system equipped with a DAD G1315D detector at 210 and 254 nm. A semi-preparative ODS-HG-5 column [5 µm, 250 × 9.4 mm] was employed for the purifications. All solvents used for column chromatography and HPLC were of analytical grade (Shanghai Chemical Reagents Co., Ltd., Shanghai, China) and chromatographic grade (Dikma Technologies Inc., Foothill Ranch, CA, USA), respectively.

Biological Material
The soft corals of Sarcophyton mililatensis were collected at a depth of -20 m by SCUBA diving from the coast of Xigu Island, Hainan Province, China, in May 2014. They were frozen immediately at -20 • C after collection and identified by Prof. X.-B. Li from Hainan University. A voucher specimen (No. 14S-80) is available for inspection at the Shanghai Institute of Materia Medica, Chinese Academy of Sciences.

Extraction and Isolation
The frozen animals (400 g dry weight) were cut into pieces and thoroughly extracted with acetone at room temperature (3 × 1.5 L). The organic extract was evaporated to give a dark brown residue that was partitioned between Et 2 O and H 2 O. The upper layer was concentrated under reduced pressure to give an Et 2 O portion (13.5 g). The Et 2 O extract was separated into twenty-one fractions (A-U) by gradient silica gel column chromatography [0 → 100% Et 2 O (EE) in petroleum ether (PE)]. Fraction G was further purified by Sephadex (V2016/6, 2016, George M. Sheldrick) and refined on F 2 by the full-matrix least-squares technique using the SHELXT-2015 program package. Crystallographic data for compound 4 in this article (Table S1)

TNF-α Inhibitory Activity Bioassay
A murine macrophage cell line, RAW264.7 cells, was obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). In the bioassay for anti-inflammation, cells were cultured in DMEM containing 10% FBS, 2 mmol/L L-glutamine, 100 µg/mL streptomycin, and 100 U/mL penicillin in a humidified incubator of 5% CO 2 at 37 • C. For the cytotoxicity part, RAW264.7 cells were incubated with the compounds or the media (0.125% DMSO in DMEM containing 10% FBS) for 24 h, respectively. CCK-8 reagents (20 µL per well) were added, and the OD values were collected after 1h incubation at 450 nm (650 nm calibration) by a microplate reader (Molecular Devices, Sunnyvale, CA, USA). For the anti-inflammatory activity assay, RAW264.7 cells were incubated with compounds or the media (0.125% DMSO in DMEM containing 10% FBS), and then cells were primed with LPS (1 µg/mL) for 24 h. The supernatants were centrifuged and then measured using the mouse TNF-α ELISA kit. The CC 50 and IC 50 were estimated using the log (inhibitor) vs. normalized response non-linear fit (Graph Pad Prism 6.0, GraphPad Software, San Diego, CA, USA). Dexamethasone was used as a positive control.

ECD Computational Protocol
Torsional sampling (MCMM) conformational searches using the OPLS_2005 force field were carried out by means of the conformational search module in the Macromodel 9.9.223 software [38], applying an energy window of 21 kJ/mol. Conformers above 1% Boltzmann populations were reoptimized with Gaussian 09 [39] at the B3LYP/6-311G(d,p) level with the IEFPCM solvent model for chloroform. Frequency analysis was also carried out to confirm that the reoptimized geometries were at the energy minima. Finally, the SpecDis1.62 software [40] was applied to obtain the Boltzmann-averaged ECD spectra and visualize the results.

Molecular Docking
The docking study was performed using the Operation Environment (ADT 1.5.7) software (https://autodock.scripps.edu/, accessed on 20 August 2022) between the compound and TNF-α (PDB: 2AZ5). The structure of the natural product was optimized by energy minimization using the MM2 method and converted to a readable format at the ADT interface. Replication of the experimental binding posed by molecular docking confirmed the suitability of the docking algorithm (RMSD <2 Å). The outcomes of the results were analyzed using the Discovery Studio Visualizer software, which reveals close contact, hydrogen bond, and hydrophobic interactions. The binding affinity of the known ligand 6,7-dimethyl-3-{[methyl-(2-{methyl-[1-(3-trifluoromethyl-phenyl)-1Hindol-3-ylmethyl]-amino}-ethyl)-amino]-methyl}-chromen-4-one [37] in the TNF-α crystal structure was −5.31 kcal/mol.

Conclusions
The present study provided the intriguing discovery of ten cembrane diterpenes from the Hainan soft coral S. mililatensis, which was rarely documented in literature. Among these secondary metabolites, six were new compounds. Their structures were elucidated by extensive spectroscopic analysis along with a comparison with existing data in the literature. Different types of nonaromatic oxacycles were found in their structures, which were identified as tetrahydrofuran and tetrahydropyran ethers that rarely occurred between C-1 and C-12 and between C-1 and C-11, respectively. Additionally, the absolute configuration of compound 4 was established unambiguously by an X-Ray diffraction experiment. The absolute configuration of 7 was assigned by the modified Mosher's method in our previous work [7]. Based on the comparison of ECD spectra, the absolute configuration of compound 8 was revealed in the present study. Subsequently, the absolute configurations of the other five new compounds were tentatively assumed using the diterpenes 4, 7 and 8 as model compounds. In the bioassay, the new secondary metabolite, compound 1, exhibited significant TNF-α inhibitory activity with an IC 50 value of 9.5 µmol/L and was not cytotoxic towards RAW264.7 cells (CC 50 > 50 µmol/L). A molecular docking experiment gave insight into the key binding actions between the active compound 1 and TNF-α, which indicated the important roles of the hydroxyl groups at C-4 and C-15 and the acetoxyl group at C-11. Although the elaborated mechanism is still under investigation, the present preliminary pharmacological result revealed that the natural product, compound 1, could be developed as a potential lead compound or drug candidate of a new chemotype of nontoxic TNF-α inhibitors. This study not only extends the chemical and biological diversities of cembrane diterpenes possessing nonaromatic oxacycles but also establishes the productivity of the title species.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/ijms24031979/s1, Figures S1-S48: MS, NMR, and IR data of compounds 1-6; Figure S49: Examples of the coexistence of dehydration derivatives and their related alcohol precursors from different soft corals; Figure S50: Examples of the coexistence of epimeric epoxy analogs from different soft corals; Table S1: X-ray crystallographic data for compound 4.