Ocellatuspyrones A‒G, new antibacterial polypropionates from the Chinese mollusk Placobranchus ocellatus

Marine invertebrates serve as rich sources of secondary metabolites with intriguing chemical diversities and a wide spectrum of biological activities. Particularly, marine shell-less sacoglossan mollusks have attracted much attentions due to the fact that mollusks apply complex metabolites as chemical defense agents against to their predators. With the purpose of discovering bioactive secondary metabolites to develop marine-derived medicines from the South China Sea, we have conducted a chemical study on the photosynthetic mollusk Placobranchus ocellatus. As a result, seven new γ-pyrone polypropionates, namely ( ±)-ocellatuspyrone A (1), ( ±)-ocellatuspyrone B (2), and ocellatuspyrones C−G (5, 9−12), along with five known polypropionates, have been isolated and characterized from the South China Sea photosynthetic mollusk Placobranchus ocellatus. Extensive spectroscopic analysis, single crystal X-ray diffraction analysis, modified Mosher’s method, ECD comparison, CD exciton chirality method, TDDFT-ECD calculation, and chemical conversion were used to determine the structures and absolute configurations of the new compounds and the stereochemistry of undefined known compounds 4, 6 and 7. All these isolated polypropionates were evaluated in bioassays for their biological activities, including antibacterial, neuroprotective effect, anti-inflammatory, PTP1B inhibitory, and antiviral activities. Compounds 7, 8 and 11 were found for the first time to show antibacterial activity against fish pathogenic bacteria Streptococcus parauberis (the main pathogen causing fish streptococcal infections and acute death) with MIC values of 35.8, 34.2, and 37.4 μg/mL, respectively, which might be potential novel antibacterial agents for the treatment of fish infectious diseases. Supplementary Information The online version contains supplementary material available at 10.1007/s42995-023-00179-w.


Introduction
Fish farming has a significant economic market in Asian countries. However, the acute death of fish affected by numerous bacterial (such as Streptococcus parauberis) infections are the most important reasons for the economic losses in fish farming business (Park et al. 2018). The prevention and treatment of fish infections mainly rely on antibacterial drugs, such as ampicillin, oxytetracycline, etc., while the frequent use of antibiotics has led to the development of drug resistance in pathogenic bacteria. Therefore, studies related to the exploring novel Edited by Chengchao Chen.
* Hong Wang hongw@zjut.edu.cn * Xueting Liu liuxueting@ecust.edu.cn antibacterial drugs of fish infectious disease are of great value. Sacoglossans are small colorful marine mollusks, and most are characterized by reduced shell or exposed mantle (Cimino 1999). The degeneration of the protective shell forced mollusks to secrete large amounts of possibly toxic mucous to defend against the attack of predators (Bornancin et al. 2017). From this interesting point of view, marine sacoglossan mollusks have long been investigated both chemically and biologically. Many bioactive marine natural products have been isolated from animals of this order (Bornancin et al. 2017;Cimino 1999). Among them, metabolites derived from the condensation of propionate units are characteristic chemical defense compounds of mollusks, which have shown a wide spectrum of biological activities ranging from cytotoxic, ichthyotoxic, antibiotic, antifungal to antiviral properties (Davies-Coleman and Garson 1998; Gavagnin et al. 1994a, b;Liu et al. 2020). In addition, terpenes and alkaloids with feeding deterrent activities and/or color deterrence are also regarded as important defensive chemicals of mollusks (Gavagnin et al. 1994a, b;Mudianta et al. 2014).
Polypropionates are among the most structurally complex polyketides, which represent a large family of typical bioactive natural products, and have been isolated from many species of the order sacoglossa, such as Elysia chlorotica (Dawe and Wright 1986), E. timida, and E. viridis (Marin and Ros 2004). Examples include simple acyclic polypropionates and complicated polypropionates containing furanone, 2-pyrone, and 4-pyrone (γ-pyrone) rings (Davies-Coleman and Garson 1998). Previous reports have indicated that polypropionatederived secondary metabolites containing the γ-pyrone moiety exhibited promising in vitro growth-inhibitory activity against human cancer cell lines (Carbone et al. 2013a, b;Rodriguez et al. 1992a, b;Zhou et al. 2018). Moreover, pyrone polypropionates are likely functioning as a natural sunscreen against photochemical damages (Faulkner 2000;Ireland and Scheuer 1979;Manzo et al. 2005;Zuidema et al. 2005) and/or as a scavenger of harmful reactive oxygen species (ROS) (Powell et al. 2018) due to the presence of a polyunsaturated side chain.
More recently, significant attention has been directed towards studying the biosynthesis of these sacolossan polypropionates. Feeding experiments in sacoglossans have shown that mollusks can utilize C 3 units for the biosynthesis of pyrone polypropionates (Cutignano et al. 2009(Cutignano et al. , 2012Di Marzo et al. 1991). In addition, some sacoglossans have been proven to prey on algae, digesting the cells but maintaining functional chloroplasts, and incorporating the fixed carbon obtained from de novo chloroplast photosynthesis into polypropionate metabolites using fatty acid synthase-(FAS)-like polyketide synthase (PKS) proteins, which open the door to understanding the metabolic patterns of marine mollusk metabolites (Torres et al. 2020).
While continuing our research with the purpose of discovering bioactive secondary metabolites from marine mollusks in the South China Sea, we have very recently conducted a chemical study on the photosynthetic mollusk Placobranchus ocellatus (phylum Mollusca, class Gastropoda, subclass Opisthobranchia, order Sacoglossa) collected off Ximao Island, Hainan Province, China. We have isolated a series of racemic rearranged γ-pyrone polypropionates with novel skeletons, named ocellatusones (Wu et al. 2020), and several racemic endoperoxide-bridged γ-pyrone polypropionates, named ocellatuperoxides (Li et al. 2022). In the present study, we describe the results of the following chemical investigation of the titled animals, intending to expand structural diversity and biological activity of the polypropionates, leading to the discovery of seven new γ-pyrone polypropionates, including two pairs of enantiomers with a bicyclohexane fragment, namely ( ±)-ocellatuspyrone A (1) and ( ±)-ocellatuspyrone B (2), and five optically active compounds ocellatuspyrones C−G (5, 9−12) (Fig. 1). Herein, we report the isolation, structural elucidation, and biological activity evaluation of these new compounds.
A detailed comparison of the NMR data of 1 with those of the co-occurring known polypropionate photodeoxytridachione (3) and the model compound phototridachiapyrone J, a known compound isolated previously from the Argentina sacoglossan Elysia patagonica (Carbone et al. 2013a, b), revealed that they were structural analogues with the only difference being the presence of a hydroxyl group at C-13 in 1, in agreement with the mass data and the chemical shift of C-13 at δ C 76.8. The hydroxyl group was connected to C-13, as evidenced by the observation of 1 H− 1 H COSY cross peaks of H-13 (δ H 4.08) with both H 2 -14 (δ H 1.68) and the HMBC correlation (Fig. 2) from H 3 -15 (δ H 0.96) to C-13. The geometry of the double bond at Δ 9(10) was assigned to be Z by the NOESY correlation of H 3 -20 (δ H 1.64) with H-9 (δ H 5.42). The relative configurations of C-6, C-7, C-8, and C-11 in 1 were proven to be the same as those of 3 and phototridachiapyrone J due to the similar carbon chemical shifts, which were further supported by the observed NOESY correlations (Fig. 2) of H 3 -18 (δ H 1.14) with H-11 (δ H 2.90) and of H 3 -19 (δ H 1.20) with H-7 (δ H 1.25). Considering that compound 1 contains a chiral center C-13 at the rotatable side chain, QM-NMR calculations (Lodewyk et al. 2012) were applied to assign the relative configuration of all the chiral carbons on this molecule to give two possible diastereoisomers ( Supplementary Fig. S15a). Geometrical optimization at the DFT level was performed following the DP4 + protocol (Grimblat et al. 2015) using the B3LYP functional with the 6-31G* basis set, followed by NMR calculations at the PCM/mPW1PW91/6-31G(d) level.
( ±)-Ocellatuspyrone B (2), which was isolated as a colorless oil, gave the molecular formula C 21 H 30 O 3 on the basis of the HR-ESIMS ion peak at m/z 331.2272 [M + H] + (calcd. for C 21 H 31 O 3 , 331.2268), requiring seven degrees of unsaturation. The IR spectrum (KBr) displayed the obvious absorption at 1662 and 1602 cm −1 , consistent with the presence of an unsaturated carbonyl group. The 1 H and 13 C NMR data (Tables 1 and 2) of 2 were nearly identical to those of 3, with exception of an isobutyl group at C-11 in 2 instead of the 1-methyl-1-butenyl group in 3. This replacement caused the 13 C NMR resonance of C-11 to be shifted upfield from δ C 58.5 to δ C 46.7. The position of the isobutyl group at C-11 was further confirmed by the 1 H− 1 H COSY cross peaks ( Fig. 2) from H-11 (δ H 2.19) to H 3 -15 (δ H 0.95). A detailed analysis of the HMBC correlations from CH 3 -15 to C-12, C-13, and C-14; CH 3 -16 to C-1, C-2, and C-3; CH 3 -17 to C-3, C-4, and C-5; CH 3 -18 to C-5, C-6, and C-8; CH 3 -19 to C-6, C-8, and C-9; CH 3 -20 to C-9, C-10, and C-11; −OCH 3 to C-1, established the planar structure of 2 (Fig. 2). The geometry of the double bond at Δ 9(10) was assigned to be Z by the NOESY correlation of H 3 -20 (δ H 1.68) with H-9 (δ H 5.22). The similar NOESY correlation patterns of 2 and 1, indicated that they have the same relative configuration. Unfortunately, the absolute configurations of ( ±)-2 were only tentatively assumed to be the same as ( ±)-1 base on biogenetic grounds due to the failed attempts to separate ( ±)-2 by chiral HPLC. Ocellatuspyrone C (5)  .2166), suggesting that 5 possessed eight degrees of unsaturation. Its IR spectrum exhibited a broad absorption at 3332 cm −1 (−OH) and strong absorptions at 1651 and 1574 cm −1 , consistent with the presence of an unsaturated carbonyl group. Similarly, the typical tetrasubstituted γ-pyrone moiety in 5 was also identified by the characteristic NMR data of a ketone carbonyl (δ C 182.3, qC), two tetrasubstituted double bonds (δ C 99.4, qC, δ C 162.0, qC; δ C 121.3, qC, 160.0, qC), two methyls (δ C 7.1, CH 3 -16; δ C 12.6, CH 3 -17), and one methoxyl (δ C 55.7, OMe), which was also supported by its UV spectrum (MeOH, λ max = 257 nm, logε = 2.8). The 1 H NMR data (Table 1) (3H, s, H 3 -20) and two olefinic protons appearing at δ H 5.47 (1H, s, H-7) and 5.73 (1H, s, H-9), which were attributed to two trisubstituted double bonds. In addition, proton signals were also observed for one terminal double bond at δ H 5.20 (1H, s, H-21) and 5.17 (1H, s, H-21), one oxymethine at d H 3.58 (1H, br s, H-13), and one triplet methyl at δ H 0.88 (3H, t, J = 7.3 Hz, H 3 -15) in the 1 H NMR spectrum. Further analysis of the 13 C NMR data (Table 2) and HSQC experiment of 5, disclosed 22 signals for seven methyls, one sp 3 methylene, two sp 3 methines, one sp 3 quaternary carbon, one sp 2 methylene, two sp 2 methines, and eight sp 2 quaternary carbons. The NMR signal of an oxygenated carbon was observed at δ C 88.1, implying that the remaining two oxygen atoms were involved in a hydroperoxyl group. Finally, the 1 H− 1 H COSY cross peaks readily assigned the only one spin system from H-13 to H 3 -15, along with the significant HMBC correlations from the protons of seven methyls to the carbon signals, via two and three bonds, elucidated the planar structure of 5 as depicted in Fig. 2.
The detailed comparison of the NMR data of 5 with those of the co-occurring known compound, tridachiapyrone J (4), revealed that the two compounds are stereoisomers with different configurations at C-13. Notably, tridachiapyrone J (4) together with the hydroxyl analogues tridachiapyrone G (6) and H (7) were first isolated by Schmitz's group with the stereochemistry unassigned (Fu et al. 2000). The absolute configuration of 4 was unambiguously determined to be 6R, 11R, 13R (as shown in Fig. 1) by X-ray diffraction analysis using Cu Kα radiation (λ = 1.54178 Å) (Fig. 4, CCDC deposition number 1953946) with the absolute structure parameter as − 0.07(6) in this study. Therefore, the absolute configuration of 5 was indirectly determined to be 6R, 11R, 13S. The chemical conversion from 4 to 6 with the triphenylphosphine (PPh 3 ) as reductant was accomplished (Fig. 3B), which assigned the stereochemistry of 6 and 7. Moreover, a simple protonic acid (TFA) catalyzed reaction of 6 was further completed, giving a high yield epoxidation product 6a.
Ocellatuspyrone E (10) was also isolated as an optically active colorless oil ([α] + 96.7 (c 0.17 CHCl 3 )), with the same molecular formula of C 22 H 30 O 5 as 9 by the HR-ESIMS ion peak at m/z 375.2166 [M + H] + (calcd. for C 22 H 31 O 5 , 375.2166). Compound 10 had the same IR and UV absorptions with those for 9. Comparison of 1 H and 13 C NMR data of 10 and 9 demonstrated that almost all of these data were virtually identical except the chemical shift of H 3 -15 (δ H 0.89, d, J = 6.3 Hz) in 10 instead of the H 3 -15 (δ H 1.09, d, J = 6.4 Hz) in 9, suggesting that they are 14-epimers. The geometry of the double bond at Δ 12(13) was also assigned to be E by the NOESY correlation of H 3 -21 (δ H 1.46) with H-14 (δ H 4.57), and the relative configuration of C-6 and C-11 was determined to be the same 6R*, 11R* as those of 9 on the basis of the NOESY experiment (Fig. 2). Furthermore, the relative configurations at C-14 in 9 and 10 were deduced to be 14R* and 14S*, respectively, by the analysis of the vicinal proton-proton coupling constants (Fig. 3C) based on the well-known Karplus-equation (Haasnoot et al. 1980).
The absolute configurations of all the stereogenic centers in 9 and 10 were defined by ECD experiments. The ECD spectrum of 9 showed a positive Cotton effect at 272 nm (Δε = + 13.6) and a negative Cotton effect at 217 nm (Δε = − 4.2), while the ECD spectrum of 10 showed a positive Cotton effect at 272 nm (Δε = + 16.9) and a negative Cotton effect at 218 nm (Δε = − 1.8), consistent with those of 4−8 (Fig. 5), indicative of (6R, 11R, 14R)-and (6R, 11R, 14S)-configurations, respectively. Accordingly, the structures of 9 and 10 were characterized as shown in Fig. 1 (Tables 1 and 2) of 11 showed typical signals for a α-methoxy-β-methyl-γ-pyrone polypropionate nucleus with eight methyls, one sp 3 methylene, one sp 3 methine, one sp 3 quaternary carbon, two sp 2 methines, and nine sp 2 quaternary carbons, which was also supported by a UV absorption at 252 nm (logε = 2.9) and IR band at 1652 and 1580 cm −1 . The remaining two oxygen atoms were assigned to be a hydroperoxyl group, which was confirmed by the presence of a proton signal at δ H 4.47 (1H, s, H-9) and an oxygenated methine carbon at δ C 84.0 (CH, C-9), signaling that 11 was also the isomer of 9. In fact, the NMR data of 11 were nearly identical to those of 9, with the exception of the hydroperoxyl group substituted at C-9 in 11 instead of the C-14 in 9 and the migration of Δ 9(10) double bond in 9 to Δ 10(11) [δ C 125.3 (qC, C-10), 143.3 (qC, C-11)] in 11. This replacement led the 13 C resonance of C-9, C-21 to be shifted upfield and C-11 downfield, respectively. The position of the hydroperoxyl group at C-9 was further secured by the HMBC correlations (Fig. 2) from both H 3 -19 (δ H 1.90) and H 3 -20 (δ H 1.79) to C-9. The geometry of the Δ 12(13) double bond in 11 was determined to be E by the NOESY cross peaks of H 3 -21 (δ H 1.57) and H 2 -14 (δ H 1.96). The determination of relative configurations of C-6 and C-9 of 11 were further determined by QM-NMR calculations following the DP4 + protocol. Two possible candidate isomers of 11 were built (11a and 11b, Supplementary Fig. S15b). Next, the conformational searches followed by DFT optimization and calculation of NMR parameters at PCM/mPW1PW91/6-31G(d) level of theory were undertaken. As a result, the candidate structure with the relative configuration 6R*, 9S* gave a best match between experimental and calculated data with over 99% probability (Supplementary Fig. S15d).
To determine the absolute configurations of 11 and 12, the CD exciton chirality method (Zhang et al. 2013) was applied. As shown in Fig. 6, the ECD spectra of 11 and 12 showed similar positive chirality owing to the exciton coupling between the chromophores γ-pyrone and conjugated double bonds, suggested the C-6 connected γ-pyrone as the main chromophore to impact the ECD spectrum of such compounds. Thus, the transition dipole moments of the two chromophores oriented in a clockwise manner determined the 6R absolute configurations of 11 and 12. Moreover, from a biogenetic point of view, the absolute configurations at C-6 of 11 and 12 were also suggested to be the R-configuration as in the co-occurring compounds 4−10. Meanwhile, for further confirmation of the absolute configurations of compounds 11 and 12, the time-dependent density functional theory electronic circular dichroism (TDDFT-ECD) calculations (Ye et al. 2017) were performed. Finally, the calculated ECD spectra of (6R, 9S)-11 and (6R)-12 appeared to be similar to positive Cotton effect curves and highly matched to those experimental ones (Fig. 6). Therefore, the absolute configurations of ocellatuspyrone F (11) and ocellatuspyrone G (12) were assigned to be (6R, 9S) and (6R), respectively.
In in vitro bioassays, the isolates were tested for antibacterial activities against Staphylococcus aureus, Streptococcus parauberis, Lactococcus garviea, Aeromonas salmonicida, Pseudomonas aeruginosa, and Photobacterium halotolerans, and only three compounds (7, 8, and 11) showed weak antibacterial activities against Streptococcus parauberis with the MIC value of 35.8, 34.2, and 37.4 μg/ mL, respectively. Comparing the active compounds and the inactive compounds 1−3, we found that, the bicyclo[3.1.0] hexene fragment was not essential for the activity. Moreover, a comparison of active 11 and the inactive 12 indicated that the introduction of a hydroxyl group at C-9 is beneficial for the antibacterial activity. In addition, the anti-inflammatory effect in lipopolysaccharide (LPS) induced RAW264.7 cell inflammation, the neuroprotective effect on hydrogen peroxide (H 2 O 2 ) induced SH-SY5Y cell damage, the PTP1B inhibitory activity, and the antiviral activity against 2019-nCoV of these metabolites were also screened, and all these compounds were inactive at the 20 μmol/L level (see the details in Supplementary Tables S2−S6).
In this study, starting from the α-methoxyl-γ-pyronyl triene precursor, compound 2 could be directly formed via a π4a + π2a electrocyclization (Fig. 7 path C). As for path D, using α-methoxyl-γ-pyronyl tetraene as the precursor, the common branched compound 8 is generated through a photochemical 6π conrotatory electrocyclization on intermediate d. Subsequent σ2a + π2a electrocyclization leads to the formation of 3, which might be further oxidized to give 1. Peroxidation occurred on different carbon positions of 8, offering a series of downstream metabolites 4, 5, and 9−11.
Chemically, we performed a photochemical reaction on 6 and 8, respectively, with the treatment of methylene blue and rose Bengal in CH 2 Cl 2 followed with exposure to sunlight for serval hours. As a result, neither compound 1 nor 3 was detected accordingly as previously reported (Supplementary Figs. S17a and S17b) (Zuidema et al. 2005). Therefore, it is speculated that the selection of appropriate photosensitizer and the light radiation energy play the important role in this reaction. We further treated compound 8 with TFA and/or silica gel in CH 2 Cl 2 at room temperature overnight and a new constituent was not formed (Supplementary Fig. S17c). Thus, we can conclude that bicyclo[3.1.0]hexene containing pyrones are not artifacts from the purification process.

Conclusions
In summary, further chemical investigation of the South China Sea sacoglossan P. ocellatus has resulted in the isolation and characterization of seven new γ-pyrone polypropionates, namely ( ±)-ocellatuspyrone A (1) and Fig. 6 The electric transition dipole of the main chromophores for compounds 11 and 12, and the assignment of the absolute configurations of 11 and 12 by comparing TDDFT-ECD calculated and the experimental spectra. CE cotton effect ( ±)-ocellatuspyrone B (2) and ocellatuspyrones C−G (5, 9−12), along with five known compounds (3, 4, 6−8). Our continuous chemical study on this animal has not only enriched the chemical diversity of γ-pyrone polypropionates but also expanded the total number of compounds within the polypropionate family. In particular, compounds ( ±)-1 and ( ±)-2, two pairs of enantiomers, represent new examples of γ-pyrone polypropionate featuring an unprecedented bicyclo[3.1.0]hexene scaffold. Moreover, the structures of stereochemistry undefined known compounds 4, 6 and 7 were determined for the first time in the present work. It is interesting to note that ( ±)-photodeoxytridachione (3) was first reported to be the photochemical rearranged products of 9,10-deoxytridachione (8) upon exposure to sunlight, and the 14 C-labeling experiments indicated that this photoisomerization conversion was a non-enzymatic light-catalyzed reaction occurring in vivo (Ireland and Scheuer 1979). The impressive structures with their photochemical properties attracted much attention from natural products chemists and synthetic chemists. Several types of mechanistic and synthetic evidence were also reported previously to support this biosynthesis isomerization process (Bruckner et al. 2003;Eade et al. 2008;Jeffery and Perkins 2004;Zuidema et al. 2005). Similarly, molecules ( ±)-1 were proposed to be the rearranged products of tridachiapyrone G (6) by the same photochemical conversion mechanism. In bioassays, compounds 7, 8, and 11 showed weak antibacterial activities against S. parauberis with MIC values of 35.8, 34.2, and 37.4 μg/mL, respectively. These active γ-pyrones provide promising starting points for further functional optimization. In addition, further studies could be conducted to complete the characterization of the proposed biosynthetic pathway. At the same time, because of the unique ecological and living environment of marine mollusks, many silent genes may need to be activated by external stimuli, leading to promising future prospects of even greater chemical diversity and biological activity.

General experimental procedures
Melting points were measured on an X-4 digital micromelting point apparatus. IR spectrum was recorded on a Nicolet 6700 spectrometer (Thermo Scientific, Waltham, MA, USA), peaks are reported in cm -1 . UV and ECD spectra were recorded on a Jasco J-815 spectropolarimeter (JASCO, Japan) at ambient temperature using chromatographic grade CH 3 OH and CH 3 CN as solvents. Optical rotations were measured on a PerkinElmer 241MC polarimeter (PerkinElmer, Fremont, CA, USA). 1 H and 13 C NMR spectra were acquired on a Bruker DRX 400, 500 and Avance 600 MHz NMR spectrometers (Bruker Biospin AG, Fallanden, Germany). Chemical shifts are reported in parts per million (δ) in CDCl 3 with the residual CHCl 3 (δ H 7.26 × 10 -6 ) as the internal standard for 1 H NMR spectrometry and CDCl 3 (δ C 77.16 × 10 -6 ) for 13 C NMR spectrometry. The HR-ESIMS spectra were recorded on an Agilent G6250 Q-TOF (Agilent, Santa Clara, CA, USA). Semi-preparative HPLC was performed on an Agilent 1260 series liquid chromatography system equipped with a DAD G1315D detector at 210 and 254 nm (Agilent, Santa Clara, CA, USA), and an Agilent semi-preparative XDB-C18 column (5 μm, 250 mm × 9.4 mm) was employed for the purification. Commercial silica gel (200˗300 mesh and 300˗400 mesh; Qingdao Haiyang Chemical Co., Ltd., Qingdao, China) was used for column chromatography, and precoated silica gel GF254 plates (Sinopharm Chemical Reagent Co., Shanghai, China) were used for analytical TLC. Spots were detected on TLC under UV light or by heating after spraying with anisaldehyde H 2 SO 4 reagent. Sephadex LH-20 (Pharmacia, USA) was also used for column chromatography. 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., CA, USA), respectively.

Biological material
The mollusk P. ocellatus (500 specimens) was collected off the shallow water area, Ximao Island, Hainan Province, China, in 2017. The biological samples and high-definition pictures of the mollusk were sent to marine biologist Dr. Christiane Waldrich and identified as P. ocellatus. The voucher specimen (No. 17XD-12) is available for inspection at the Shanghai Institute of Materia Medica, CAS.

Preparation of the (S)-and (R)-MTPA ester derivatives of compounds ( ±)-1
Compound ( +)-1 (1.4 mg) was dissolved in 1.0 mL of pyridine-d 5 , and the solution was divided into two equal parts and transferred into two NMR tubes. To initiate the reaction, 15 μL of (R)-and (S)-MTPA-Cl were added to the two NMR tubes, respectively. The reactions were found to be completed in 30 min by the monitoring of 1 H NMR, and the reaction residues were purified by RP-HPLC, yielding the mono (S)-and (R)-MTPA ester derivatives of ( +)-1, respectively. Analogously, 1.2 mg of compound (−)-1 was reacted in two NMR tubes with 15 μL of (R)-and (S)-MTPA-Cl for 30 min, to afford the mono (S)-and (R)-MTPA ester derivatives of (−)-1, respectively.

Chemical conversion of compounds 4 and 6
A total of 5.0 mg of compound 4 was dissolved in 5.0 mL of CH 2 Cl 2 and added to excess PPh 3 (10.0 mg) and stirred for 2 h at room temperature. The residue obtained after the reaction was purified by Silica-gel CC eluting with PE-Et 2 O (50:50) to afford compound 6 (3.6 mg, 72% yield). The NMR data of the obtained compound are consistent with that of isolated natural compound 6 ( Supplementary Fig. S8).

QM-NMR calculation of compounds 1 and 11
All calculations followed the general protocols previously described for DP4 + (Grimblat et al. 2015). Conformational searches were performed in the gas phase applying the MMFF94S force field. NMR calculations were undertaken on structures above 1% Boltzmann population. A theory level of B3LYP/6-31G* was used to optimize structures. The GIAO methodology was used to calculate magnetic shielding constants (σ) at the mPW1PW91/6-31G(d) level of theory as recommended for DP4 + .

TDDFT-ECD calculation of compounds 11 and 12
Torsional sampling (MCMM) conformational searches using MMFF94S force field were carried out by the means of conformational search module in Macro model 9.9.223 software (Schrodinger, http:// www. schro dinger. com/ Macro Model), applying an energy window of 21 kJ/mol (5.02 kcal/mol) for saving structures. The dominant conformers with over 1% Boltzmann population were used for re-optimization and the following TDDFT-ECD calculation. The re-optimizations and the TDDFT-ECD calculations were performed with Gaussian 09 (Gaussian, http:// www. Gauss ian. com) at the same B3LYP/6-311G(d,p) level with the IEFPCM solvent model for acetonitrile. Finally, the SpecDis 1.62 software (Bruhn et al. 2013) was used to obtain the calculated ECD spectrum and visualize the results. compound purification and structural elucidation. D-DY and M-ZS contributed to biological assays. L-GY provided biological material. S-WL wrote the manuscript and Y-WG and X-TL revised it. All authors discussed the results and approved the manuscript.

Data availability
The data that supports the findings of this study are included in this published article (and its supplementary information file).

Conflict of interest
The authors declare that they have no conflict of interest.
Animal and human rights statement This article does not contain any studies with human participants or animals performed by any of the authors.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.