New ψ-Santonin Derivatives from Crossostephium chinense and Their Anti-Proliferative Activities against Leishmania major and Human Cancer Cells A549

Previously, we reported two cytotoxic ψ-santonin–amino acid conjugates isolated from the EtOAc layer of Crossostephium chinense. However, a further phytochemical investigation seems to be required because of the few reports of similar derivatives. In this study, we targeted the 1-BuOH layer, which resulted in the isolation of seven new ψ-santonin derivatives (1–7) together with ten known compounds (8–17). The structures of 1–7 were elucidated based on spectroscopic methods, including 1D and 2D NMR experiments (1H, 13C, DEPT, COSY, HSQC, and HMBC), IR spectrum, and high-resolution electrospray ionization–mass spectrometry (HR-ESI–MS). The stereochemistry of new compounds was confirmed by NOESY and ECD calculations. All isolated compounds were evaluated by in vitro experiments for their anti-proliferative activities against Leishmania major, human lung cancer cell line A549, and Vero cells. As a result, most of the ψ-santonin derivatives, especially 1–5, showed significant cytotoxicity against L. major with a lower IC50 than the positive control we used (miltefosine).


Introduction
Leishmaniasis is a tropical disease caused by at least 20 species of protozoan parasites from the Family trypanosomatidae and is transmitted by the bite of infected female sandflies.According to the reports from WHO, more than 350 million people are at risk, about 12 million of whom are infected, with an average annual incidence of one million new cases, mainly from low-income developing countries [1].To accelerate global drug development, WHO has designated leishmaniasis as one of the neglected tropical diseases (NTDs) and set a new roadmap for 2021-2030 to drive progress towards a world free of NTDs by 2030.Leishmaniasis has various clinical manifestations, which can be classified into cutaneous leishmaniasis (CL), visceral leishmaniasis (VL), and mucocutaneous leishmaniasis (MCL) [2,3].Leishmania major (L.major) is one of the species that causes CL, which is not fatal but is responsible for disfiguring lesions and social stigma [4].The symptoms after the sandfly bite begin from a small edema to a papule and finally develop into a nodule that grows into necrosis and ulceration [5].There are about 600,000 to 1 million new cases of CL annually, 75% of which are diagnosed in Afghanistan, Brazil, Iran, Iraq, and Syria [6].Many anticancer drugs have been developed and used clinically for years.However, cancer is still a leading cause of death in many countries.Among various cancers, lung cancer causes the most deaths worldwide [7].A549 is a cell line derived from human lung cancer and is frequently used in anticancer drug screening.
Crossostephium chinense is an evergreen shrub belonging to the Asteraceae family and distributed in Japan, China, and other Asian countries.The Family Asteraceae is famous for its promising source of bioactive compounds, such as pyrethrin, artemisinin, and santonin from the plant Tanacetum cinerariifolium, Artemisia annua, and Artemisia maritima, respectively.Our research team focuses on discovering bioactive constituents against Leishmania major from natural resources.C. chinense was selected based on the primary screening results, and then, two ψ-santonin-amino acid conjugates that showed potent cytotoxicity against L. major named crossoseamine A and B were previously discovered from the EtOAc layer [7].However, a further phytochemical investigation of C. chinense seems necessary because of the few reports of ψ-santonin derivatives and their potential bioactivities.Therefore, we targeted the 1-BuOH layer of C. chinense, aiming to isolate more novel ψ-santonin derivatives in this study.As a result, seven previously undescribed compounds (1-7) (Figure 1) together with ten known compounds (8)(9)(10)(11)(12)(13)(14)(15)(16)(17) were isolated (Figure S1), of which the new ψ-santonin derivatives 1-5 showed significant cytotoxic activities against L. major.In addition, the A549 cancer and Vero cell lines were used to evaluate the cytotoxicity of all isolated compounds.les 2023, 28, x FOR PEER REVIEW 2 of 15 However, cancer is still a leading cause of death in many countries.Among various cancers, lung cancer causes the most deaths worldwide [7].A549 is a cell line derived from human lung cancer and is frequently used in anticancer drug screening.
Crossostephium chinense is an evergreen shrub belonging to the Asteraceae family and distributed in Japan, China, and other Asian countries.The Family Asteraceae is famous for its promising source of bioactive compounds, such as pyrethrin, artemisinin, and santonin from the plant Tanacetum cinerariifolium, Artemisia annua, and Artemisia maritima, respectively.Our research team focuses on discovering bioactive constituents against Leishmania major from natural resources.C. chinense was selected based on the primary screening results, and then, two ψ-santonin-amino acid conjugates that showed potent cytotoxicity against L. major named crossoseamine A and B were previously discovered from the EtOAc layer [7].However, a further phytochemical investigation of C. chinense seems necessary because of the few reports of ψ-santonin derivatives and their potential bioactivities.Therefore, we targeted the 1-BuOH layer of C. chinense, aiming to isolate more novel ψ-santonin derivatives in this study.As a result, seven previously undescribed compounds (1-7) (Figure 1) together with ten known compounds (8)(9)(10)(11)(12)(13)(14)(15)(16)(17) were isolated (Figure S1), of which the new ψ-santonin derivatives 1-5 showed significant cytotoxic activities against L. major.In addition, the A549 cancer and Vero cell lines were used to evaluate the cytotoxicity of all isolated compounds.

Structure of Crossostenin A (1)
Crossostenin A (1) was isolated as a white amorphous powder with a molecular formula C20H24O5 established by HR-ESI-MS data, which showed a positive ion [M + Na] + at m/z 367.1515 (calculated for C20H24O5Na: 367.1516).The IR spectrum contained absorption signals at 1765 and 1710 cm −1 due to several carbonyl functions, of which 1765 cm −1 suggested a γ-lactone functional group and 1710 cm −1 due to ester/lactone/ketone carbonyl functions.The 13 C NMR (Table 1) spectrum revealed the existence of three carbonyl groups (δC 166.9, 171.9, and 216.4) and three double bonds (δC 116.2, 125.8, 128.1, 139.1, 141.1, and 160.1), which provided six indices of hydrogen deficiency.According to the other signals, such as four methine carbons (including two oxygenated carbons at δC 72.7 and 77.5 and one olefinic at δC 116.2), four methylene (including one olefinic carbon at δC 125.8) and four methyl carbons (δC 20.0, 20.5, 26.2 and 27.6), the remaining three unsaturation degrees were ascribed to a three-ring skeleton, which is similar to the   1) spectrum revealed the existence of three carbonyl groups (δ C 166.9, 171.9, and 216.4) and three double bonds (δ C 116.2, 125.8, 128.1, 139.1, 141.1, and 160.1), which provided six indices of hydrogen deficiency.According to the other signals, such as four methine carbons (including two oxygenated carbons at δ C 72.7 and 77.5 and one olefinic at δ C 116.2), four methylene (including one olefinic carbon at δ C 125.8) and four methyl carbons (δ C 20.0, 20.5, 26.2 and 27.6), the remaining three unsaturation degrees were ascribed to a three-ring skeleton, which is similar to the crossoseamine B previously isolated from the titled plant in our previous study [7].Correspondingly, there were four olefinic protons (δ H 5.49, 1H, sept, J = 1.4 Hz, H-17; δ H 5.90, 1H, d, J = 2.0 Hz, H-13a; δ H 6.34, 1H, d, J = 2.0 Hz, H-13b) and two oxygenated methine protons (δ H 5.68, 1H, d, J = 7.1 Hz, H-6; δ H 4.81, 1H, m, H-8) detected in the 1 H NMR spectrum (Table 1).The characteristic olefinic proton (δ H 5.49, 1H, sept, J = 1.4 Hz, H-17) coupled with two vinyl methyl protons (δ H 1.88, 3H, d, J = 1.4 Hz, H-19; δ H 2.12, 3H, d, J = 1.4 Hz, H-20) suggested a senecioic acid side chain [8].The low-field shifted chemical shift value caused by the deshielding effect of the carboxyl group from H-8 (δ H 3.89) in crossoseamine B to H-8 (δ H 4.81) in 1 indicated the attachment of a senecioic acid moiety at C-8 (δ C 72.7).The HMBC correlations (Figure 2) from H-14 (δ H 1.21) to C-10 (δ C 46.8) and C-1 (δ C 216.4); from H-2 (δ H 2.34 and 2.60) to C-1 and C-3 (δ C 30.3); from H-3 (δ H 2.37 and 2.75) to C-1, C-4 and C-5 (δ C 141.1) indicated the position of carbonyl group at C-1.The planar structure of 1 was also analyzed by COSY (Figure 2) correlations, such as the spin-spin coupling network from H-6 to H-9, and finally confirmed, as shown in Figure 1.The relative configuration of 1 was deduced by analysis of the 1D NMR spectrum (especially coupling constants and chemical shifts) and NOESY data (Figure 3).The following differences were observed between crossoseamine B and compound 1.First, a smaller coupling constant (J H-6/H-7 = 7.1 Hz) in compound 1 was detected instead of a large one (J H-6/H-7 = 11.4Hz) in crossoseamine B. Second, the chemical shifts of C-4 (δ C 128.1) and C-5 (δ C 141.1) in compound 1 were significantly different from C-4 (δ C 126.1) and C-5 (δ C 128.6) in crossoseamine B. Finally, the long-range coupling between H-6 and vinyl methyl group H 3 -15, which was detected in crossoseamine B, disappeared in compound 1.According to the above differences, compound 1 was deduced to have the opposite stereochemistry at C-6 to crossoseamine B. Therefore, NOESY analysis was conducted to confirm it.As a result, the correlations from H-6 to H-15 and H-7; from H-14 to H-3β, H-9β and H-8 were observed, which supported the above hypothesis, and other NOESY correlations indicated the relative stereochemistry as shown in Figure 3.The relative configuration of 1 was deduced by analysis of the 1D NMR spectrum (especially coupling constants and chemical shifts) and NOESY data (Figure 3).The following differences were observed between crossoseamine B and compound 1.First, a smaller coupling constant (JH-6/H-7 = 7.1 Hz) in compound 1 was detected instead of a large one (JH-6/H-7 = 11.4Hz) in crossoseamine B. Second, the chemical shifts of C-4 (δC 128.1) and C-5 (δC 141.1) in compound 1 were significantly different from C-4 (δC 126.1) and C-5 (δC 128.6) in crossoseamine B. Finally, the long-range coupling between H-6 and vinyl methyl group H3-15, which was detected in crossoseamine B, disappeared in compound 1.According to the above differences, compound 1 was deduced to have the opposite stereochemistry at C-6 to crossoseamine B. Therefore, NOESY analysis was conducted to confirm it.As a result, the correlations from H-6 to H-15 and H-7; from H-14 to H-3β, H-9β and H-8 were observed, which supported the above hypothesis, and other NOESY correlations indicated the relative stereochemistry as shown in Figure 3. Finally, the absolute configuration of 1 was determined by the application of the octant rule and ECD calculations.The n→π* absorption around 280-300 nm for the carbonyl group (C-1) showed a negative sign (Figure 4), which indicated the absolute stereochemistry as shown in Figure 3 by the application of the octant rule [9].Additionally, the experimental CD and UV curves were matched to the calculated spectra of 1a (deacyl part of 1 at C-8 to remove the unstable part for accurate calculation) (Figure 4).Considering the above results, the absolute configuration of 1 was concluded to be 6R, 7R, 8S, and 10R and designated as crossostenin A (1).The relative configuration of 1 was deduced by analysis of the 1D NMR spectrum (especially coupling constants and chemical shifts) and NOESY data (Figure 3).The following differences were observed between crossoseamine B and compound 1.First, a smaller coupling constant (JH-6/H-7 = 7.1 Hz) in compound 1 was detected instead of a large one (JH-6/H-7 = 11.4Hz) in crossoseamine B. Second, the chemical shifts of C-4 (δC 128.1) and C-5 (δC 141.1) in compound 1 were significantly different from C-4 (δC 126.1) and C-5 (δC 128.6) in crossoseamine B. Finally, the long-range coupling between H-6 and vinyl methyl group H3-15, which was detected in crossoseamine B, disappeared in compound 1.According to the above differences, compound 1 was deduced to have the opposite stereochemistry at C-6 to crossoseamine B. Therefore, NOESY analysis was conducted to confirm it.As a result, the correlations from H-6 to H-15 and H-7; from H-14 to H-3β, H-9β and H-8 were observed, which supported the above hypothesis, and other NOESY correlations indicated the relative stereochemistry as shown in Figure 3. Finally, the absolute configuration of 1 was determined by the application of the octant rule and ECD calculations.The n→π* absorption around 280-300 nm for the carbonyl group (C-1) showed a negative sign (Figure 4), which indicated the absolute stereochemistry as shown in Figure 3 by the application of the octant rule [9].Additionally, the experimental CD and UV curves were matched to the calculated spectra of 1a (deacyl part of 1 at C-8 to remove the unstable part for accurate calculation) (Figure 4).Considering the above results, the absolute configuration of 1 was concluded to be 6R, 7R, 8S, and 10R and designated as crossostenin A (1).Finally, the absolute configuration of 1 was determined by the application of the octant rule and ECD calculations.The n→π* absorption around 280-300 nm for the carbonyl group (C-1) showed a negative sign (Figure 4), which indicated the absolute stereochemistry as shown in Figure 3 by the application of the octant rule [9].Additionally, the experimental CD and UV curves were matched to the calculated spectra of 1a (deacyl part of 1 at C-8 to remove the unstable part for accurate calculation) (Figure 4).Considering the above results, the absolute configuration of 1 was concluded to be 6R, 7R, 8S, and 10R and designated as crossostenin A (1). Crossostenin B (2) was isolated as a white amorphous powder with a molecular formula determined as C20H24O5 from the HR-ESI-MS data, which displayed a positive ion [M + Na] + at m/z 367.1520 (calculated for C20H24O5Na: 367.1516).In the 13 C NMR spectrum (Table 1), the carbon signals for the senecioic moiety in 1 were changed to δC 168.6 (C-16),

Structure of Crossostenin B (2)
Crossostenin B (2) was isolated as a white amorphous powder with a molecular formula determined as C 20 H 24 O 5 from the HR-ESI-MS data, which displayed a positive ion [M + Na] + at m/z 367.1520 (calculated for C 20 H 24 O 5 Na: 367.1516).In the 13 C NMR spectrum (Table 1), the carbon signals for the senecioic moiety in 1 were changed to δ C 168.6 (C-16), 127.8 (C-17), 139.5 (C-18), 14.5 (C-19) and 12.2 (C-20), which indicated the substitution of tiglic acid moiety in 2 [10].The multiplicity of the olefinic proton (δ H 6.67, 1H, qq, J = 7.0, 1.4 Hz, H-18) and the vinyl methyl protons (δ H 1.77, 3H, dq, J = 7.0, 1.4 Hz, H-19; δ H 1.74, 3H, quint-like, J = 1.4 Hz, H-20), together with the COSY correlations from H-18 (δ H 6.67) to H 3 -19 (δ H 1.77) and the HMBC correlations from H-20 (δ H 1.74) to C-16 (δ C 168.6), 17 (δ C 127.8), and 18 (δ C 139.5) also validated the existence of the tiglic acid (Figure 2).The low-field shifted chemical shift value of H-8 (δ H 4.83) confirmed the connection between tiglic acid and ψ-santonin skeleton at C-8.The relative configuration was deduced as the same as 1 by chemical shift similarity, coupling constant-based configuration analysis, and NOESY correlations (Figure 3).Specifically, NOESY correlations from H-6 to H-7 and H-15 from H-8 to H-14 and H-9β revealed the same relative configurations between 1 and 2. Based on these spectroscopic data and similar Cotton effects (Figure S20) with compound 1, the structure of crossostenin B (2) was ultimately established, as shown in Figure 1.Considering the existence of a methoxy carbon (δ C 52.4) in 13 C NMR and the remaining two indices of hydrogen deficiency, the skeleton of compound 3 should be a non-lactone-type sesquiterpene structure, unlike compounds 1 and 2. Further detailed analysis of the 13 C NMR spectrum also revealed the presence of a tiglic acid moiety.The HMBC correlations (Figure 5) from methoxy protons (δ H 3.75) to C-12 (δ C 168.5) confirmed the linkage of the methoxy group and the deshielded H-8 (δ H 5.44) chemical shift and the HMBC correlation from H-8 to C-16 detected by 700 MHz NMR with cryoprobe (Figure S27) revealed the connection of the tiglic acid at C-8.This HMBC correlation from H-8 to C-16 was only detected using 700 MHz NMR with a cryoprobe, not by a standard inverse probe, which is likely because of the weakening of the signal intensity by the multiple spin-spin coupling, an instability of conformation around C-8 methylene by flip-flop movement, and rotation of acyl function.The success of the detection also depended on the isolation amount (3, 4.7 mg), which is higher than compounds 1 (1.1 mg) and 2 (1.4 mg).Finally, the COSY coupling system (Figure 5) from H-6 (δ H 4.90) to H-9 (δ H 1.51 and 2.15) validated the planar structure of 3 as shown.The relative configuration of compound 3 was determined by the coupling constant and NOESY correlations (Figure 6).A large coupling constant between H-6 and H-7 (JH-6/H-7 = 10.1 Hz) suggested that the stereochemistry of C-6 was different from compounds 1 and 2. In the NOESY spectrum, the correlations from H-6 to H-14 and H-8; from H-8 to H-9β and H-14; and from H-7 to H-9α were detected and consistent with the conclusion from The relative configuration of compound 3 was determined by the coupling constant and NOESY correlations (Figure 6).A large coupling constant between H-6 and H-7 (J H-6/H-7 = 10.1 Hz) suggested that the stereochemistry of C-6 was different from compounds 1 and 2. In the NOESY spectrum, the correlations from H-6 to H-14 and H-8; from H-8 to H-9β and H-14; and from H-7 to H-9α were detected and consistent with the conclusion from coupling constant analysis.The relative configuration of compound 3 was determined by the coupling constant and NOESY correlations (Figure 6).A large coupling constant between H-6 and H-7 (JH-6/H-7 = 10.1 Hz) suggested that the stereochemistry of C-6 was different from compounds 1 and 2. In the NOESY spectrum, the correlations from H-6 to H-14 and H-8; from H-8 to H-9β and H-14; and from H-7 to H-9α were detected and consistent with the conclusion from coupling constant analysis.
The absolute configuration of 3 was concluded by the application of the octant rule and the calculation of the ECD spectrum (Figure 7).The characteristic negative Cotton effect at around 220 nm and 300 nm was detected both in ECD and experimental CD spectra, which led to the assignment of the absolute configuration of 3 as 6S, 7S, 8S, and 10R, and named crossostenin C (3).The absolute configuration of 3 was concluded by the application of the octant rule and the calculation of the ECD spectrum (Figure 7).The characteristic negative Cotton effect at around 220 nm and 300 nm was detected both in ECD and experimental CD spectra, which led to the assignment of the absolute configuration of 3 as 6S, 7S, 8S, and 10R, and named crossostenin C (3).The relative configuration of compound 3 was determined by the coupling constant and NOESY correlations (Figure 6).A large coupling constant between H-6 and H-7 (JH-6/H-7 = 10.1 Hz) suggested that the stereochemistry of C-6 was different from compounds 1 and 2. In the NOESY spectrum, the correlations from H-6 to H-14 and H-8; from H-8 to H-9β and H-14; and from H-7 to H-9α were detected and consistent with the conclusion from coupling constant analysis.
Considering the structure-activity relationship, the lactone ring formation is important for the potent anti-Leishmania major activity of 1 and 2. The hydrolysis of lactone and the introduction of proline moiety reduced the activity.However, the stereochemistry at C-6 (6S) and the presence of the acyl function at C-8 as compound 5 showed the highest activity among proline conjugates 5-7.2) showed three carbonyl groups (δ C 213.6, 177.9, and 174.3), one double bond (δ C 127.2 and 139.1, both are quaternary carbons), and two methyl groups (δ C 23.9 and 19.2).The 1 H NMR spectrum displayed an oxygenated methine proton with a small coupling constant (δ H 5.56, 1H, d, J = 5.6 Hz, H-6), which indicated cis configuration to H-7.Furthermore, 1 H and 13 C NMR data of compound 6 suggested a similar structure to crossoseamine A with a difference in the configuration at C-6.The HMBC and COSY correlations (Figure 8) were also consistent with this assumption.In the PS-NOESY spectrum (Figure 9), key correlations from H-6 to H-15/H-7/H-13a, from H-7 to H-9α suggested the same direction of H-6 and H-7 as α; correlations from H-14 to H-9β/H-8, from H-9β to H-8, from H-8 to H-11 revealed a β configuration of H-14, H-8 and H-11.The absolute configuration was analyzed by comparing the CD spectra (Figure 10).The Cotton effects of compound 6 were detected at 229 nm (+5.41) and 298 nm (−5.46) in CH 3 CN, which showed an opposite sign at 229 nm to the 6S derivatives, crossoseamine C (5), A, and B, which indicated that compound 6 has a 6R configuration.Additionally, the negative Cotton effects at 298 nm (−5.46) belonging to the n-π* transition of the 1ketone functional group also supported the 6R configuration by applying the octant rule.According to the above results, the absolute configuration of 6 was determined to be 6R, 7S, 10R, 11R, and 2 ′ S, as shown in Figure 1.

Structure of Crossoseamine E (7)
Crossoseamine E (7) was purified as a white amorphous powder with a molecular formula C 20 H 27 O 6 N determined by HR-ESI-MS, which revealed a protonated molecule [M + H] + at m/z 378.1912, calculated for C 20 H 28 O 6 N: 378.1911.Comparing the 13 C NMR with compound 6, an additional oxymethine carbon at δ C 65.2 was observed, which suggested the presence of a hydroxyl group at C-8 like crossoseamine B and C (5).The COSY spin-spin coupling network from H-6 to H-9 and H-13, along with the HMBC correlations (Figure 8), confirmed the planar structure of 7. The small coupling constant (J H-6/H-7 = 5.8 Hz) and the NOESY correlations from H-8 to H-9β/H-14/H-11, from H-7 to H-6/H-9α revealed the same relative configuration as 6.To determine the absolute configuration, CD spectrum was measured (Figure 10).The positive Cotton effect in 7 at 230 nm (+5.53) and negative Cotton effect at 297 nm (−5.51) suggested the absolute configuration of 7 as 6R, 7S, 10R, 11R, and 2 ′ S.
Considering the structure-activity relationship, the lactone ring formation is important for the potent anti-Leishmania major activity of 1 and 2. The hydrolysis of lactone and the introduction of proline moiety reduced the activity.However, the stereochemistry at C-6 (6S) and the presence of the acyl function at C-8 as compound 5 showed the highest activity among proline conjugates 5-7.
1 H and 13 C NMR spectra were measured on an Avance III HD spectrometer (Bruker, Billerica, MA, USA) at 500 and 125 MHz, respectively, with the residual solvent signal as references.An Avance NEO 700 MHz spectrometer with a cryoprobe (Bruker, Billerica, MA, USA) was also used for the HMBC measurement.Positive-and negative-ion HR-ESI-MS were recorded on an LTQ Orbitrap XL spectrometer (Thermo Fisher Scientific, Waltham, MA, USA).A P-1030 spectropolarimeter (JASCO, Tokyo, Japan) was used for the measurement of specific optical rotations.IR and UV spectra were measured on FT-720 (HORIBA, Kyoto, Japan) and V-520 UV/Vis spectrophotometers (JASCO, Japan), respectively.The CD spectrum was measured on the J-720 spectropolarimeter (JASCO, Japan) and J-1500 spectropolarimeter (JASCO, Japan).

Plant Material
The aerial parts of C. chinense were collected in July 2008 in Okinawa, Japan, and a voucher specimen was deposited in the Herbarium of the Department of Pharmacognosy, Graduate School of Biomedical Sciences, Hiroshima (deposition number: 08-CC-Okinawa-0708).

Extraction and Isolation
Air-dried aerial parts of C. chinense (3.5 kg) were extracted with MeOH (3 × 10 L) at room temperature.The methanol extract was evaporated to 1.5 L and then partitioned with an equal volume of n-hexane to obtain an n-hexane soluble layer (27.7 g).The remaining layer was concentrated and resuspended in 1.5 L of water and then extracted with 1.5 L EtOAc and 1.5 L 1-BuOH successively to give EtOAc (74.The known compounds were identified by comparison of their physicochemical data ([α] D , IR, MS, 1 H and 13 C NMR) with the reported data.

Computational Calculations
Conformational analyses for 1a were performed using Spartan'20 V1.1.2. with default conformational search program (Wavefunction, Inc., Irvine, CA, USA).Stable conformers up to 40 kcal/mol for 1a were initially searched using the Merck molecular force field (MMFF) method.Then, the stable conformers obtained from the initial search were further optimized using the Hartree-Fock (HF)/3-21G and ωB97XD/6-31G* programs.The resulting conformers were subjected to ECD calculation, and the ECD calculations for these conformers were performed with Gaussian 16 (Revision A.03 by Gaussian) [23] on the ChemPark cloud system.The dominant conformers of 1a capable of covering 100% (conformer 1: 83.1%; conformer 2: 16.9%) of the population according to Boltzmann's distribution were selected.Time-dependent density functional theory calculations were conducted at the CAM-B3LYP/aug-cc-pVDZ level for these conformers.The resulting rotational strength data were converted to Gaussian curves (bandwidth sigma = 0.5 eV) to obtain the ECD spectra of each conformer, and the spectra were combined after Boltzmann weighting according to their population contributions.The wavelength of the spectra was corrected (+10 nm) based on the absorptions of about 220 nm (referring to the experimental and calculated UV spectra) to give the corresponding theoretical ECD spectrum.

Anti-Proliferative Activities
The anti-Leishmania major activities of the isolated compounds were evaluated using an MTT assay according to an established protocol [24].In brief, gradient concentrations of sample solutions in dimethyl sulfoxide (DMSO) and L. major (1 × 10 5 parasite/well) in 100 µL of M199 medium were added to a 96-well plate and then incubated for 72 h at 25 • C.An amount of 100 µL MTT solution was then replaced and incubated overnight.The absorbance of formazan solution in DMSO was measured using a microplate reader at 550 nm.Miltefosine was used as a positive control.The human lung cancer cell line A549 was cultured in 10% FCS-supplemented DMEM.The cytotoxicity assay of isolated compounds was also performed in 96-well plates using the MTT method.Various concentrations of samples in DMSO and A549 (5 × 10 3 cells/well) were cultured in a CO 2 incubator for 72 h.The medium was then replaced with 100 µL MTT solution and incubated for 1.5 h in the same condition.The viability was calculated by the absorbance of formazan at 550 nm using a microplate reader.The cytotoxicity assay against the Vero cell line was conducted as in A549, which is mentioned above.Doxorubicin was used as a positive control.The concentration gradient of each activity was essentially set as 100, 50, 25, 12.5, and 6.25 µg/mL (Figures S67-S69).

Conclusions
Due to the strong anti-proliferative activities of the previously reported compounds crossoseamine A and B, a more detailed phytochemical investigation of C. chinense was conducted, leading to the isolation of seven new ψ-santonin derivatives (1-7) in this work.All new compounds were evaluated for their cytotoxic activities against Leishmania major and A549.As a result, compounds 1-5 showed more significant activity against L. major than the positive control miltefosine, which was also the first time the anti-proliferative potentials of ψ-santonin derivatives were revealed.Moreover, the high selectivity against L. major of 1 and 2 was revealed by comparing the cytotoxicity to the Vero cells, which suggests the possibility of being seed compounds.However, a further and comprehensive structure-activity relationship investigation seems to be required in the future.

2 . 1 .
Structure Determination of New Compounds 1-7 2.1.1.Structure of Crossostenin A (1) Crossostenin A (1) was isolated as a white amorphous powder with a molecular formula C 20 H 24 O 5 established by HR-ESI-MS data, which showed a positive ion [M + Na] + at m/z 367.1515 (calculated for C 20 H 24 O 5 Na: 367.1516).The IR spectrum contained absorption signals at 1765 and 1710 cm −1 due to several carbonyl functions, of which 1765 cm −1 suggested a γ-lactone functional group and 1710 cm −1 due to ester/lactone/ketone carbonyl functions.The 13 C NMR (Table

Figure 4 .
Figure 4. Experimental and calculated ECD spectra of 1 and 1a in CH 3 CN.

Figure 7 .Figure 7 .
Figure 7. Experimental and calculated ECD spectra of 3 and 4 in CH3CN.2.1.4.Structure of Crossostenin D (4) Figure 7. Experimental and calculated ECD spectra of 3 and 4 in CH 3 CN.2.1.4.Structure of Crossostenin D (4) Crossostenin D (4) was obtained as a white amorphous powder with the same molecular formula as 3, which was established by HR-ESI-MS in positive ion mode.A different side chain, instead of acetic acid, senecioic acid, and tiglic acid observed in compounds 1-3, was detected in the 13 C NMR spectrum, which showed a carbonyl group (δ C 168.8), a double bond (δ C 129.3 and 138.8) and two vinyl methyl groups (δ C 16.1 and 20.8), and then ascribed to the angelic acid[10].The NOESY analysis (Figure6) and ECD calculation (Figure7) confirmed the relative and absolute configurations, respectively.Hence, the structure of 4 was determined, as shown in Figure1.

2. 1 . 5 .
Structure of Crossoseamine C (5) Crossoseamine C (5) was purified as a white amorphous powder and found to have a nitrogen function based on TLC examination using Dragendorff's reagent.The molecular formula was determined as C 22 H 29 O 7 N by HR-ESI-MS data, which displayed a positive ion [M + Na] + at m/z 442.1839 (calculated for C 22 H 29 O 7 NNa: 442.1836).The 13 C NMR spectrum (Table

Figure 10 .
Figure 10.Experimental CD spectra of Crossoseamine A-E in CH 3 CN.2.1.6.Structure of Crossoseamine D (6) Crossoseamine D (6) was obtained as a white amorphous powder with the same molecular formula (C 20 H 27 O 5 N, established by HR-ESI-MS positive peaks [M + Na] + at m/z 384.1780, calculated for C 20 H 27 O 5 NNa: 384.1781) to crossoseamine A [7].The 13 C NMR spectrum (Table2) showed three carbonyl groups (δ C 213.6, 177.9, and 174.3), one double bond (δ C 127.2 and 139.1, both are quaternary carbons), and two methyl groups (δ C 23.9 and 19.2).The 1 H NMR spectrum displayed an oxygenated methine proton with a small coupling constant (δ H 5.56, 1H, d, J = 5.6 Hz, H-6), which indicated cis configuration to H-7.Furthermore, 1 H and 13 C NMR data of compound 6 suggested a similar structure to crossoseamine A with a difference in the configuration at C-6.The HMBC and COSY correlations (Figure8) were also consistent with this assumption.In the PS-NOESY spectrum (Figure9), key correlations from H-6 to H-15/H-7/H-13a, from H-7 to H-9α suggested the same direction of H-6 and H-7 as α; correlations from H-14 to H-9β/H-8, from H-9β to H-8, from H-8 to H-11 revealed a β configuration of H-14, H-8 and H-11.The absolute configuration was analyzed by comparing the CD spectra (Figure10).The Cotton effects of compound 6 were detected at 229 nm (+5.41) and 298 nm (−5.46) in CH 3 CN, which showed an opposite sign at 229 nm to the 6S derivatives, crossoseamine C (5), A, and B, which indicated that compound 6 has a 6R configuration.Additionally, the negative Cotton effects at 298 nm (−5.46) belonging to the n-π* transition of the 1ketone functional group also supported the 6R configuration by applying the octant rule.According to the above results, the absolute configuration of 6 was determined to be 6R, 7S, 10R, 11R, and 2 ′ S, as shown in Figure1.
: 1D, 2D NMR spectra, MS and CD data of compounds 1-7.Author Contributions: Conceptualization, K.M. and Z.W.; Methodology, S.S. and S.K.; Software, Y.Y. and S.K.; Validation, K.M. and S.S.; Formal analysis, K.M. and H.O.; Investigation, Z.W. and S.K.; Resources, K.M. and H.O.; Data curation, Z.W. and S.K.; Writing-original draft preparation, Z.W.; Editing, K.M. and G.A.A.-H.; Visualization, Z.W.; Supervision, K.M.; Project administration, K.M.; Funding acquisition, K.M. and G.A.A.-H.All authors have read and agreed to the published version of the manuscript.Funding: This work was supported by JSPS KAKENHI Grant Number 22K06678.This work was also partly supported by JST SPRING, Grant Number JPMJSP2132 and the Researchers Supporting Project, number RSPD2023R878, King Saud University, Riyadh, Saudi Arabia.This work was also the result of using research equipment shared in the MEXT Project to promote public utilization of advanced research infrastructure (Program for Supporting Construction of Core Facilities) Grant Number JPMXS0441300023.Institutional Review Board Statement: Not applicable.Informed Consent Statement: Not applicable.Data Availability Statement: Data are contained within the article or Supplementary Material.
δ H in ppm (multiplicity, J in Hz), δ C in ppm, m-multiplet or overlapped signals.