Uncommon Capnosane Diterpenes with Neuroprotective Potential from South China Sea Soft Coral Sarcophyton boettgeri

The first investigation of the South China Sea soft coral Sarcophyton boettgeri afforded five new capnosane diterpenes, sarboettgerins A–E (1–5), together with one known related compound, pavidolide D (6). Their structures, including absolute configurations, were elucidated by the extensive spectroscopic analysis, 13C NMR calculations, and X-ray diffraction. Among them, new compounds 1–5 were featured by the rarely encountered Z-geometry double bond Δ1 within the 5/11-fused bicyclic capnosane carbon framework. Plausible biogenetic relationships of all isolates were proposed, and they might give an insight into future biomimetic synthesis of these novel compounds. In an in vitro bioassay, compound 5 displayed potent anti-neuroinflammatory activity against LPS-induced NO release in BV-2 microglial cells, which might be developed as a new type of potential neuroprotective agent in future.

Soft corals of the genus Sarcophyton are well-known as a reservoir of diverse terpenes, including capnosanes [3]. As part of our ongoing project searching for bioactive secondary metabolites from Chinese marine cnidarians [41,42] and their predators [43], a recent investigation on the chemical constituents of Sarcophyton boettgeri collected from Weizhou Island, South China Sea, has afforded five new capnosane-type diterpenes, sarboettgerins A-E (1-5), along with one related known analog (6) (Figure 1). A survey of the literature revealed that there is no report regarding the chemical constituents of this species before this study. Herein, we describe the isolation, structural elucidation, bioactivity evaluation, and plausible biosynthetic pathway of the new compounds.

Results and Discussion
The frozen bodies of title animals were cut into pieces and extracted exhaustively with acetone. The Et2O-soluble portion of the acetone extract was repeatedly column chromatographed over silica gel, Sephadex LH-20, and RP-HPLC to yield five new compounds, i.e., compounds 1-5, along with one known related compound, compound 6 ( Figure 1). The known compound 6 was readily identified as pavidolide D, a capnosane

Results and Discussion
The frozen bodies of title animals were cut into pieces and extracted exhaustively with acetone. The Et 2 O-soluble portion of the acetone extract was repeatedly column chromatographed over silica gel, Sephadex LH-20, and RP-HPLC to yield five new compounds, i.e., compounds 1-5, along with one known related compound, compound 6 ( Figure 1). The known compound 6 was readily identified as pavidolide D, a capnosane diterpene Mar. Drugs 2022, 20, 602 3 of 14 isolated from the soft coral S. pavida, by comparison of the NMR data and optical rotation with those reported in the literature [30].
Compound 1 was obtained as a colorless crystal and had the molecular formula C 20 H 32 O 2 on the basis of its quasi-molecular ion peak at m/z 327.2299 ([M + Na] + , calcd. for C 20 H 32 O 2 Na, 327.2295), requiring four degrees of unsaturation. The IR absorption bands at 3452 and 1640 cm −1 revealed the existence of the hydroxyl and olefinic groups, which were in agreement with the presence of a tertiary alcohol functionality [δ C 82.0 (qC, C-4)] and a terminal double bond [δ H 4.92 (1H, s, H-19a) and 4.96 (1H, s, H-19b), δ C 108.6 (CH 2 , C-19) and 149.1 (qC, C-8)], as indicated by the 1 H and 13 C NMR data (Table 1). Its 13 C NMR spectrum exhibited twenty carbon resonances assigned to four methyls (δ C 21.3, 20.8, 24.6, and 15.4), seven methylenes (including one olefinic, δ C 108.6, and six aliphatic, δ C 40. 6, 26.7, 32.5, 24.9, 33.8, and 25.3), five methines (including one olefinic, δ C 120.9; an oxygenated, δ C 67.0; and three aliphatic, δ C 28.9, 50.3, and 47.0), and four quaternary carbons (including two olefinic, δ C 143.6 and 149.1; and two oxygenated, δ C 82.0 and 59.3) (Table 1), which were deduced from DEPT and HSQC experiments. In addition, its 1 H NMR spectrum displayed signals of an epoxide moiety at δ H 3.15 (1H, dd, J = 10.7, 3.4 Hz, H-11), which corresponded to the 13 C chemical shifts at δ C 67.0 (CH, C-11) and 59.3 (qC, C-12) ( Table 1). As revealed by the 1 H and 13 C NMR data, there were one double bond and one epoxide, accounting for two degrees of unsaturation. The remaining two degrees of unsaturation were due to the presence of two rings in the molecule. Considering the co-isolated secondary metabolite pavidolide D, compound 1 was likely a capnosane-type diterpene. Through a detailed literature review on diterpenes from soft corals, the abovementioned structural features were reminiscent of previously reported lobophytrol C (7) [35], a capnosane from South China Sea soft coral Lobophytum sp. Our interpretation of the 2D NMR spectra ( Figure 2) suggested they shared the same gross structure. Indeed, the major differences were observed around C-1 (δ C 143.6 for 1 vs. δ C 148.1 for 7), C-2 (δ C 120.9 for 1 vs. δ C 122.7 for 7), and its neighboring carbons, C-3 (δ C 50.3 for 1 vs. δ C 51.8 for 7), and C-7 (δ C 47.0 for 1 vs. δ C 54.4 for 7), disclosing they were a pair of stereoisomers. The relative configuration of 1 was deduced from the analysis of its NOESY spectrum. As shown in Figure 3, the absence of NOE correlation between H-2 and H-15 indicated that the double-bond ∆ 1 in 1 took Z-geometry. Additionally, the key NOE cross-peaks of H-3/H-7, H-7/H-9β, H-9β/H-10β, and H-10β/H 3 -20, along with the lack of NOE interactions between H-3 and H 3 -18 and between H-11 and H 3 -20, revealed that H-3, H-7, and H 3 -20 were oriented on the same side of the ring system and tentatively assigned as β, while H-7, H-11, and H 3 -18 were α-oriented. In order to establish the absolute configuration of 1, a suitable single crystal was obtained from methanol after many attempts. The successful X-ray diffraction experiments with Cu Ka (λ= 1.54178 Å) radiation allowed us to confirm the proposed structure and to determine its absolute configuration as 3R,4R,7R,11S,12S [Flack parameter: -0.06(7)] ( Figure 4). Hereto, the structure determination of 1 was completely accomplished. H3-18 were α-oriented. In order to establish the absolute configuration of 1, a suitable sin gle crystal was obtained from methanol after many attempts. The successful X-ray diffrac tion experiments with Cu Ka (λ= 1.54178 Å) radiation allowed us to confirm the proposed structure and to determine its absolute configuration as 3R,4R,7R,11S,12S [Flack parame ter: -0.06 (7)] ( Figure 4). Hereto, the structure determination of 1 was completely accom plished.

Figure 4.
Perspective ORTEP drawing of X-ray structure of 1 (displacement ellipsoids are drawn a the 50% probability level).      resembled those of 1, and the 2D NMR ( 1 H-1 H COSY, HSQC, and HMBC) analysis revealed that both 2 and 1 had the same gross structure (Figure 2). The main difference between 2 and 1 was the significantly upfield-shifted H-7 (δ H 2.19, m) in 2 in comparison with that (δ H 2.88, m) of 1, in addition to the substantially downfield shifted C-7 (δ C 55.4) in 2, indicating that 2 was a C-7 epimer of 1. This hypothesis was ascertained by the examination of its NOESY data ( Figure 3). The characteristic NOE cross-peak of H 3 -18/H-7 without the observation of H 3 -18/H-3 or H-7/H-3 revealed that H 3 -18 and H-7 were co-facial, while the orientations of H-3 and H-7 were trans. To further distinguish the epimer relationship between compounds 1 and 2, quantum NMR calculations using DP4+ probability analysis were performed, which was a powerful tool for the structural determinations of natural products [43][44][45]. As the stereochemistry of 1 was corroborated by X-ray diffraction, the correctness of the prediction for 1 from using the DP4+ method was checked at first. Because the main concern of the stereochemistry was the cis orientation for H-3 and H-7, (3R*,4R*,7R*,11S*,12S*)-1a and (3R*,4R*,7S*,11S*, 12S*)-1b were selected as two initiative possible conformers and then subjected to quantum NMR calculations. As a result, the experimental NMR data of compound 1 best matched those of 1a with 100% probability (Supplementary Figure S46), which agreed with that derived from X-ray diffraction. This result perfectly confirmed the reliability of the application of this method for capnosane-type diterpenes. The subsequent calculations of two possible isomers, (3R*,4R*,7R*,11S*,12S*)-2a and (3R*,4R*,7S*,11S*,12S*)-2b, were conducted. The result (Supplementary Figure S47) indicated that the experimental NMR data of (3R*,4R*,7S*,11S*,12S*)-2b were most consistent with the experimental data with DP4+ probability 100%, which confirmed our conclusion. Based on the biogenetic consideration and their nearly identical NMR chemical shifts, the absolute configuration of 2 could be assigned as 3R,4R,7S,11S,12S.
Compound 3 was obtained as a colorless oil. Its molecular formula was established as C 20 H 32 O 2 by HRESIMS from the quasi-molecular ion peak at m/z 327.2293 ([M + Na] + , calcd. for C 20 H 32 O 2 Na, 327.2295), which was the same as that of 1. A detailed analysis of the NMR data (Table 1) revealed the considerable structural similarity between 3 and 1. They only differed in the migration of the double bond, in which the terminal ∆ 8 (19) in 1 was replaced by the endo ∆ 7 in 3. This was supported by the HMBC correlations between H 3 -19 and C-7, C-8, and C-9 ( Figure 2). As shown in its NOESY spectrum (Figure 3), H-2 exhibited NOE correlations with H 3 -18 and H-11, indicating that these protons were all co-facial. Meanwhile, the absence of NOE interaction between H 3 -18 and H-3 clearly clarified that their orientations were opposite. The Z-geometry of the 1,2-double bond was assigned due to the lack of NOE cross-peak of H-2/H-15. The trans orientations for H-11 and H 3 -20 were deduced from the 13 C NMR shift value of C-19 (δ C 16.1 < 20 ppm), which was confirmed by the absence of correlation between H-11 and H 3 -20. In light of these observations, the full structure of compound 3 was unambiguously established, as depicted in Figure 1.
The quasi-molecular ion peak at m/z 327.2291 ([M + Na] + , calcd. for C 20 H 32 O 2 Na, 327.2295) displayed in the HRESIMS spectrum revealed that compound 4 had the same molecular formula, C 20 H 32 O 2 , as that of the model compound 1. A detailed analysis of NMR data (Table 1) of 4 suggested that 4 and 1 possessed the common moieties, including one trisubstituted double bond, one terminal bond, one tertiary alcohol functionality, and one epoxide group. Meanwhile, the extensive analyses of its 1 H-1 H COSY, HSQC, and HMBC spectra (Figure 2) disclosed that the positions of the terminal bond and the tertiary alcohol group were exchanged as compared to 1. This was corroborated by the HMBC correlations between H 2 -18 and C-3 and C-5 and between H 3 -19 and C-7 and C-9. The geometry of double-bond ∆ 1 , as well as its relative configurations at C-3, C-7, C-11, and C-12, was proven to be the same as that of 1 on the basis of a NOESY experiment (Figure 3 ambiguity, quantum NMR calculations of two possible isomers, 8S*-4a and 8R*-4b, using DP4+ probability analysis, were performed. As a result, the experimental NMR data of compound 4 gave the best match for 4b, with 99.96% probability ( Figure 5). Based on the above analysis, the structure of 4 was shown as depicted in Figure 1.
correlations between H2-18 and C-3 and C-5 and between H3-19 and C-7 and C-9. The geometry of double-bond Δ 1 , as well as its relative configurations at C-3, C-7, C-11, and C-12, was proven to be the same as that of 1 on the basis of a NOESY experiment ( Figure  3). The absence of the NOE interaction from H-2 to H-15 clearly demonstrated the 1Z geometry. The NOESY correlations between H-3 and H-7 implied the cis-configuration of H-3 and H-7. The trans geometry of the epoxide group was confirmed by the absence of correlation between H-11 and H3-20. However, the NOESY experiment did not afford any useful correlations for the elucidation of the configuration of the chiral center C-8. To resolve this ambiguity, quantum NMR calculations of two possible isomers, 8S*-4a and 8R*-4b, using DP4+ probability analysis, were performed. As a result, the experimental NMR data of compound 4 gave the best match for 4b, with 99.96% probability ( Figure 5). Based on the above analysis, the structure of 4 was shown as depicted in Figure 1.   Figure 3). Meanwhile, the absence of the NOE cross-peak of H-11/H 3 -20 suggested the 11E geometry ( Figure 3). Owing to the overlapped signals, the orientations of H-3, H-7, and H 3 -19 could not be determined via the NOESY experiment. Thus, quantum NMR calculations of the four possible conformers (3R*,7R*,8S*-5a, 3R*,7R*,8R*-5b, 3R*,7S*,8S*-5c, and 3R*,7S*,8R*-5d) were subjected to the GIAO method, at the mPW1PW91/6-31+G(d,p) level, with the PCM model in chloroform solvent ( Figure 6). The result indicated that the experimental NMR data of (3R*,7R*,8R*)-5b were consistent with the experimental data, with a correlation coefficient R 2 = 0.9918 and DP4+ probability of 100%. On the basis of the above analysis, compound 5 was depicted as shown in Figure 1.
Compounds 1-6 shared the same 5/11-fused bicyclic carbon framework, namely capnosane. It is worth pointing out that the structural similarities between them and their co-occurrence suggested that they should originate via the same biogenetic pathway. Therefore, a plausible biosynthetic connection from the known compound 6 to all the new compounds was proposed based on their structural features (Scheme 1). The origin of 1 was proposed to be multiple possible biosynthetic pathways, including ∆ 1 isomerization, C-11/C-12 epoxidation, and a elimination of water from 6. Compound 2 was thus derived from compound 1 by C-7 isomerization, which could undergo double-bond migration to afford compound 3. Compound 5 was similarly generated by an elimination of water from 6, followed by ∆ 1 isomerization. Subsequently, C-11/C-12 epoxidation on compound 5 yielded compound 4.
Considering the interesting neuroprotective potential exhibited by some marine diterpenoids [3], the anti-neuroinflammatory activity of these isolates was assessed by employing lipopolysaccharide (LPS)-stimulated BV-2 microglial cells by measuring the release of nitric oxide (NO) levels in the supernatants. As shown in Table 2, compound 5 exhibited a statistically significant inhibitory effect on LPS-induced NO release in BV-2 microglial cells, and the level of NO decreased to 53.57 ± 1.24% at 20 µM and 13.30 ± 3.21% at 40 µM, respectively, comparable to that of the positive control, resveratrol (49.93 ± 6.66%), at 20 µM (NO data were normalized by the value of each LPS group). Compounds 3 and 4 (40 µM) displayed moderate anti-neuroinflammatory activities in LPS-stimulated BV-2 microglial cells when the level of NO decreased to 67.33 ± 4.11% and 39.03 ± 2.31%, respectively. In addition, all the isolates at 20 µM and 40 µM did not show marked cytotoxicity on BV-2 cells, suggesting that the anti-neuroinflammatory activity of the abovementioned active compounds was not attributed to their cytotoxicity. The excessive microglial inflammation induced by a variety of pathological insults could drive the progressive loss of neurons that occurs in multiple neurodegenerative diseases [46][47][48], and the abovementioned active isolates that potently suppress neuroinflammation are be expected to play a critical role in preventing and treating neuronal damages and considered to have neuroprotective potential.

General Experimental Procedures
The IR spectrum was recorded on a Nicolet iS50 spectrometer (Thermo Fisher Scientific, Madison, WI, USA). Optical rotations were measured on a PerkinElmer 241MC polarimeter. 1 H and 13 C NMR spectra were acquired on a Bruker AVANCE III 600 MHz spectrometer. Chemical shifts are reported with the residual CHCl 3 (δ H 7.26 ppm; δ C 77.16 ppm) as the internal standard for 1 H and 13 C NMR spectra. HRESIMS spectra were recorded on Agilent G6250 Q-TOF (Agilent, Santa Clara, CA, USA). Commercial silica gel (Qingdao Haiyang Chemical Co., Ltd., Qingdao, China, 200-300 mesh, 300-400 mesh) was used for column chromatography, and precoated silica gel GF254 plates (Sinopharm Chemical Reagent Co., Shanghai, China) were used for analytical TLC. Sephadex LH-20 (Amersham Biosciences, Piscataway, NJ, USA) was also used for column chromatography. Reverse-phase (RP) HPLC was performed on an Agilent 1260 series liquid chromatography equipped with a DAD G1315D detector at 210 nm (Agilent, Santa Clara, CA, USA). An Agilent semi-preparative XDB-C18 column (5 µm, 250 × 9.4 mm) was employed for the purification. 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.

Animal Material
The soft coral Sarcophyton boettgeri was collected off the coast of Weizhou Island, Hainan Province, China, in 2011, at a depth of −20 m. The animal material was identified by Prof. Xiu-Bao Li from Hainan University. A voucher specimen (No. 11WZ-34) is available for inspection at the Shanghai Institute of Materia Medica.

Crystal Structure Analysis of 1
Block crystals of compound 1 were obtained from solvent methanol at room temperature. Single-crystal X-ray diffraction of 1 (0.18 × 0.11 × 0.05 mm 3 ) was performed on a Bruker Apex II CCD diffractometer (Bruker AXS Inc., Karlsruhe, Germany), using Cu Kα radiation (λ = 1.54178 Å) at 170 K. The collected data integration and reduction were processed with SAINT software (V8.40A, 2016, Bruker AXS Inc., Karlsruhe, Germany), and multi-scan absorption corrections were performed by using the SADABS program (V2014/7, 2014, Bruker AXS Inc., Karlsruhe, Germany). The structures were solved by direct methods, using SHELXTL (V2018/3, 2018, George M. Sheldrick), and were refined on F 2 by the full-matrix least-squares technique, using the SHELXL program package. The absolute configuration was determined on the basis of a Flack parameter of −0.06(7) (Supplementary Table S1). Crystallographic data for 1 in this article were deposited at the Cambridge Crystallographic Data Centre (deposition number CCDC 2203997). Copies of these data can be obtained free of charge via www.ccdc.cam.ac.uk/conts/retrieving.html (accessed on 29 August 2022) or from the Cambridge Crystallographic Data Centre (12 Union Road, Cambridge CB21EZ, UK. Fax: (+44) 1223-336-033. E-mail: deposit@ccdc.cam.ac.uk).

Computational Details
Conformational search was performed by using the torsional sampling (MCMM) approach and the OPLS_2005 force field with the conformational search in an energy window of 21 kJ/mol. Conformers above 1% Boltzmann populations were reoptimized at the B3LYP/6-311G(d,p) level with the IEFPCM solvent model for chloroform. A frequency analysis was also carried out to confirm that the reoptimized geometries were at the energy minima. Subsequently, NMR calculations were performed at the PCM/ mPW1PW91/6-31+G** level, as recommended for DP4+. NMR shielding constants were calculated by using the GIAO method. Finally, shielding constants were averaged over the Boltzmann distribution obtained for each stereoisomer and correlated with the experimental data.

In Vitro Anti-Neuroinflammatory Activity Assay
BV-2 microglial cells were cultured in 96-well plates at a density of 2×10 5 cells/well and incubated in a 37 • C incubator containing 5% CO 2 for 24 h. BV-2 microglial cells were preincubated with 20 µM or 40 µM of test compounds for 2 h, followed by 100 ng/mL of LPS exposure for 24 h. Then the culture medium of each well (50 µL) was collected to react with the same volume of Griess buffer (Sigma-Aldrich, St. Louis, MO, USA) in a 96-well plate for 15 min, in the dark, at room temperature. The absorbance, at the wavelength of 540 nm, was detected with a microplate reader, and the level of nitrite was calculated according to a sodium nitrite standard curve.

Cell Viability Determination
To measure the influence of isolates on cell viabilities, BV-2 microglial cells were seeded in 96-well plates at a density of 1.25×10 4 cells/well and incubated in a 37 • C incubator containing 5% CO 2 for 24 hours. BV-2 microglial cells were incubated with 20 µM and 40 µM of test compounds for 24 hours and then treated with 0.5 mg/mL MTT for 3 hours. After the removal of the culture medium, the cells were then incubated with 100% DMSO to dissolve formazan. The absorbance at the wavelength of 490 nm was detected with a microplate reader, and the cell viability was calculated.

Conclusions
The chemical investigation of the South China Sea soft coral S. boettgeri led to the discovery of six capnosane diterpenes 1-6, of which five new ones furnished the uncommon Z-geometry olefinic bond, ∆ 1 . Their structures, including the absolute stereochemistry, were elucidated through extensive spectroscopic analysis, quantum mechanical-NMR calculations with DP4+ probability analysis, and X-ray diffraction analysis. Although many cembranoids were reported from soft corals, capnosane-type diterpenes characterized by a 5/11-fused bicyclic carbon skeleton are infrequently encountered. To the best of our knowledge, this is the first report of chemical investigation on the soft coral S. boettgeri, which demonstrated that S. boettgeri was a potential source of structurally diverse capnosanes. In the bioassay, compound 5 displayed significant anti-neuroinflammatory activity against LPS-induced NO release in BV-2 microglial cells. Although the detailed mechanism of action is still undefined for the neuroprotective effect of compound 5, this research identified this uncommon capnosane diterpene as new chemotype of potential neuroprotective agents against neuroinflammation in microglial cells, which might give an insight for the development of novel drug candidates.