New Polyene Pregnane Steroids From the Stems of Nerium oleander: Structural Characterization and In Vitro Anti‐Inflammatory Evaluation

ABSTRACT

Phytochemical investigation on the stems of Nerium oleander led to the isolation of four previously undescribed polyene pregnane steroids, namely neristerone A‐D (1–4), together with six known analogues (5–10). Their structures were fully elucidated by extensive spectroscopic analysis (UV, IR, HR‐ESI‐MS, and NMR) and electronic circular dichroism (ECD) calculations. All isolates were evaluated for the inhibitory effects on nitric oxide (NO) production in lipopolysaccharide (LPS)‐induced RAW264.7 macrophages. Notably, compounds 1, 4, and 10 exhibited significant inhibitory activity with IC₅₀ values ranging from 11.27 ± 0.81 to 18.13 ± 1.25 µmol/L, which were lower than that of the positive control dexamethasone (21.48 ± 0.83 µmol/L). A preliminary structure–activity relationship analysis was also discussed. This study expands the chemical diversity of polyene pregnane steroids from N. oleander and provides new insights into their anti‑inflammatory potency.

Four previously undescribed polyene pregnane steroids (1–4) and six known analogues (5–10) were isolated from the dried stems of Nerium oleander.

1. Introduction

Nerium oleander Linn. (Apocynaceae) is a perennial medium‐sized evergreen flowering tree or shrub, typically reaching a height of 2–5 m at maturity. Owing to its elegant floral morphology and strong adaptability to diverse environments, it has been extensively introduced and cultivated worldwide, with concentrated distribution in tropical and subtropical regions. Beyond its high ornamental value in landscape greening, this species is well‐recognized for its intense toxicity, mainly due to abundant cardiac glycosides with a narrow therapeutic index, whereby the effective dose is close to the toxic dose [1, 2]. It is categorized as a cardiac herb in traditional Chinese medicine, described as toxic, bitter, and cold, and linked to the heart meridian [3]. Recent reviews indicate that the flowers and leaves of N. oleander are commonly used in folk medicine for various conditions. The most extensively reported pharmacological effects include anticancer [4, 5, 6], antimicrobial [7, 8, 9], anti‐inflammatory [10, 11, 12], antidiabetic [13, 14], and hepato‐ and reno‐protective activities [15, 16]. Additional reported effects encompass antioxidant [17, 18, 19], gastrointestinal [20], and reproductive activities [21], as well as traditional uses for leprosy and indigestion [22, 23]. Chemical investigations have demonstrated that N. oleander contains a diverse array of chemical components. More than 140 compounds have been isolated from its different parts (roots, leaves, and branches) during the last decade, with cardiac glycosides being the most abundant [3]. Additionally, polyene pregnane steroids also stand out as a distinctive class of specialized chemical components, yet only a few representatives of this compound class have been reported in existing studies [24, 25, 26, 27, 28]. Their unique polyene‐conjugated structure endows them with significant potential biological activities, among which anti‐inflammatory activity has been widely verified in previous research [29, 30, 31].

Macrophages are the key effector cells in inflammatory responses, and excessive nitric oxide (NO) production by LPS‐activated macrophages is a hallmark of acute inflammation. The LPS‐induced RAW264.7 cell model has therefore become the most widely used in vitro system for evaluating the anti‐inflammatory potential of natural products [32, 33]. Given the limited reports on this class and their promising anti‐inflammatory properties, we focused on isolating and identifying polyene pregnane steroids from N. oleander to fill the research gap and explore novel active compounds. To discover novel and anti‐inflammatory active pregnanes from N. oleander, 10 polyene pregnane steroids (1–10), including four new compounds (1–4) (Figure 1) were successfully obtained from its dried stems. Herein, we present the isolation and structural characterization of these pregnane steroids, along with an evaluation of their anti‐inflammatory activities.

FIGURE 1.

Chemical structures of compounds 1−10.

2. Results and Discussion

2.1. Structure Characterization

Compound 1 was obtained as a white amorphous powder. Its molecular formula was established to be C₂₃H₂₈O₄ based on (−)‐HR‐ESI‐MS at m/z 367.1905 [M − H]⁻ (calcd for C₂₃H₂₇O₄, 367.1909) (Figures S1 and S2), accounting for ten degrees of unsaturation. Conjugated carbonyl functional groups were detected in IR spectral data, as shown by distinctive absorption peaks at 1612 and 1635 cm⁻¹ (Figure S1‐1). Full assignments of the ¹H and ¹³C NMR data were achieved using 1D and 2D NMR experiments (Table 1). The ¹H and ¹³C NMR spectra (Figures S1‐3 and S1‐4) revealed the presence of four tertiary methyl groups, with corresponding signals at [δ _H 1.15 (3H, s, H₃‐18), 1.22 (3H, s, H₃‐19), 2.27 (3H, s, H₃‐21), and 2.10 (3H, s, H₃‐2′); δ _C 12.7 (C‐18), 16.4 (C‐19), 28.2 (C‐21), and 21.7 (C‐2′)], six olefinic carbons at δ _C 124.5 (C‐4), 166.1 (C‐5), 129.7 (C‐6), 141.3 (C‐7), 146.8 (C‐16), and 155.7 (C‐17), and two carbonyl carbons at δ _C 199.4 (C‐3) and 202.1 (C‐20). Along with HMBC correlations from H₃‐19 to C‐1 (δ _C 34.6), C‐9 (δ _C 52.9), C‐10 (δ _C 37.4), and C‐5 (δ _C 166.1), from H₃‐18 to C‐12 (δ _C 76.1), C‐13 (δ _C 52.9), C‐14 (δ _C 50.5), and C‐17(δ _C 155.7), from H₃‐21 to C‐20 (δ _C 202.1), from H‐6 (δ _H 6.26, dd, J = 9.7, 2.1 Hz) to C‐4 (δ _C 124.5), C‐5, and C‐7 (δ _C 141.3), and from H‐16 (δ _H 6.92, dd, J = 3.5, 1.8 Hz) to C‐13, C‐14, and C‐15 (δ _C 32.0), as well as ¹H‐¹H COSY correlations of H₂‐1/H₂‐2, H‐6/H‐7/H‐8/H‐9/H₂‐11/H‐12, and H‐8/H‐14/H₂‐15/H‐16 (Figures S1‐5 and S1‐7). The aforementioned information suggested that 1 was a pregnane C₂₁‐steroid with two carbonyls (C‐3 and C‐20) and three alkenes (Δ⁴, Δ⁶, and Δ¹⁶) groups [28]. Comparison of the ¹H and ¹³C NMR data of 1 with those of 12β‐hydroxy pregna‐4,6,16‐triene‐3,20‐dione (5) revealed that they share the same planar structure, with the main difference lying in the substituents at C‐12, and the presence of an additional acetyl substitution [δ _H 2.10 (3H, s, H₃‐2′); δ _C 173.1 (C‐1′) and 21.7 (C‐2′)] in 1. That acetyl group was confirmed to be located at C‐12 in 1 by the HMBC correlation (Figure 2) from H‐12 (δ _H 5.07, dd, J = 10.7, 5.4 Hz) to C‐1′. The NOESY correlations (Figure S1‐8) of H‐8 with H₃‐18 and H₃‐19 indicated that these protons are co‐facial and were tentatively assigned a β‐orientation. In contrast, the correlations of H‐9 with H‐12 and H‐14 revealed an α‐orientation for these protons (Figure 3). The absolute configuration of 1 was established as 8R, 9S, 10R, 12R, 13S, 14S by comparing the measured ECD spectrum with the theoretically calculated ECD spectra of its enantiomers (Figure 4). Consequently, the structure of 1 was established as 12β‐oxyacetylpregna‐4,6,16‐triene‐3,20‐dione and named neristerone A.

TABLE 1.

The ¹H and ¹³C NMR data for compounds 1−4 (¹H for 400 MHz and ¹³C for 100 MHz).

No.	1 b		2 a		3 a		4 a
	δ C	δ H (J in Hz)	δ C	δ H (J in Hz)	δ C	δ H (J in Hz)	δ C	δ H (J in Hz)
1	34.6	1.77, 1H, m 2.01, 1H, m	35.6	1.70, 1H, m 1.96, 1H, m	34.0	1.76, 1H, m 2.43, 1H, m	33.6	1.74, 1H, m 1.87, 1H, m
2	34.6	2.39, 1H, m 2.66, 1H, m	32.7	2.35, 1H, m 2.44, 1H, m	34.1	2.05, 1H, m 2.55, 1H, m	33.8	2.47, 1H, m 2.56, 1H, m
3	199.4		197.3		199.5		199.2
4	124.5	5.72, 1H, s	124.6	5.75, 1H, s	124.1	5.70, 1H, s	124.7	5.71, 1H, s
5	166.1		169.6		163.1		161.9
6	129.7	6.26, 1H, dd (9.7, 2.1)	33.9	2.38, 1H, m 2.42, 1H, m	128.5	6.12, 1H, br d (11.4)	129.1	6.16, 1H, dd (9.7, 2.6)
7	141.3	6.30, 1H, dd (9.7, 1.5)	31.2	1.13, 1H, m 1.88, 1H, m	140.2	6.09, 1H, d (9.8)	138.6	6.10, 1H, dd (9.7, 2.0)
8	36.0	2.48, 1H, m	32.7	1.72, 1H, m	37.1	2.22, 1H, m	37.0	2.64, 1H, m
9	52.9	1.72, 1H, m	52.8	1.47, 1H, m	52.8	1.31, 1H, m	52.1	1.60, 1H, m
10	37.4		38.6		36.3		36.6
11	28.4	1.71, 1H, m 1.79, 1H, m	27.7	1.58, 1H, m 1.72, 1H, m	24.8	1.26, 1H, m 2.02, 1H, m	37.8	2.22, 1H, m 2.62, 1H, m
12	76.1	5.07, 1H, dd (10.7, 5.4)	74.4	5.07, 1H, dd (11.0, 5.3)	87.8	3.14, 1H, m	210.9
13	52.9		50.9		51.6		57.7
14	50.5	1.47, 1H, m	53.4	1.13, 1H, m	48.9	1.26, 1H, m	54.4	1.56, 1H, m
15	32.0	2.34, 1H, m 2.63, 1H, m	31.5	2.08, 1H, m 2.35, 1H, m	24.0	1.56, 1H, m 1.90, 1H, m	23.6	1.56, 1H, m 1.97, 1H, m
16	146.8	6.92, 1H, dd (3.5, 1.8)	144.0	6.64, 1H, dd (3.3, 1.8)	25.4	1.77, 1H, m 2.18, 1H, m	25.1	1.37, 1H, m 1.89, 1H, m
17	155.7		154.9		61.8	2.73, 1H, t (9.2)	46.7	2.40, 1H, m
18	12.7	1.15, 3H, s	12.5	1.12, 3H, s	8.7	0.76, 3H, s	12.2	1.10, 3H, s
19	16.4	1.22, 3H, s	17.2	1.26, 3H, s	16.4	1.13, 3H, s	16.1	1.18, 3H, s
20	202.1		199.2		210.8		71.1	4.77, 1H, m
21	28.2	2.27, 3H, s	28.2	1.72, 3H, s	32.4	2.23, 3H, s	19.7	1.22, 3H, s
1′	173.1		171.5		56.2	3.35, 3H, s	170.9
2′	21.7	2.10, 3H, s	21.8	2.13, 3H, s			21.6	1.98, 3H, s

^a Recorded in CDCl₃.

^b Recorded in CD₃OD.

FIGURE 2.

¹H‐¹H COSY and key HMBC correlations of compound 1−4.

FIGURE 3.

The key NOESY correlations of 1−4.

FIGURE 4.

Experimental and calculated ECD spectra of compounds 1−4 in methanol.

Compound 2 was obtained as a white amorphous powder. It was assigned a molecular formula of C₂₃H₃₀O₄, which was established by its HR‐ESI‐MS at m/z 369.2066 [M − H]⁻ (calcd for C₂₃H₂₉O₄, 369.2066) (Figure S2‐2) and accounted for nine degrees of unsaturation. The ¹H and ¹³C NMR signals for 2 were assigned using 1D and 2D NMR experiments (Table 1). A comparison of the ¹H and ¹³C NMR data of 2 (Figures S2‐3 and S2‐4) with those of 12β‐hydroxy‐4,16‐pregnadien‐3,20‐dione (6) [34] revealed that they shared the same skeleton, with the key discrepancy being the carbon chemical shift at C‐12 and the presence of an additional acetyl substitution [δ _H 2.13 (3H, s, H₃‐2′); δ _C 171.5 (C‐1′) and 21.8 (C‐2′)] in 2. The HMBC correlation (Figures 2 and S2‐7) confirmed that the acetyl group is linked at C‐12, as evidenced by the correlation from H‐12 (δ _H 5.07, dd, J = 11.0, 5.3) to C‐1′ (δ _C 171.5). The NOESY correlations (Figures 3 and S2‐8) suggested that 2 and 6 have the same relative configurations, as evidenced by the key correlations of H‐8 (δ _H 1.72, m) with H₃‐18 (δ _H 1.12, s) and H₃‐19 (δ _H 1.26, s), and of H‐9 (δ _H 1.47, m) with H‐12 (δ _H 5.07, dd, J = 11.0, 5.3 Hz) and H‐14 (δ _H 1.13, m). The absolute configuration of 2 was determined to be 8R, 9S, 10R, 12R, 13S, 14S by comparing the calculated ECD spectra of its two possible enantiomers with the experimental data (Figure 4). Thus, the structure of 2 was established as 12β‐oxyacetylpregna‐4,16‐dien‐3,20‐dione and named neristerone B.

Compound 3 was obtained as a white amorphous powder. It was assigned a molecular formula of C₂₂H₃₀O₃, which was established by the HR‐ESI‐MS at m/z 387.2173 [M + HCOO]⁻ (calcd for C₂₃H₃₁O₅, 387.2171) (Figure S3‐2), and accounted for 8 degrees of unsaturation. Analysis of the 1D and 2D NMR data (Table 1) indicated that 3 was structurally related with 12β‐hydroxypregna‐4,6‐diene‐3,20‐dione (7) [35], and the main difference lies in the presence of a methoxyl substituent [δ _H 3.35 (3H, s, H₃‐1′); δ _C 56.2 (C‐1′)] at C‐12, which further supported by the HMBC correlation (Figures 2 and S3‐7) from H₃‐1′ to C‐12 (δ _C 87.8). The NOESY correlations (Figures 3 and S3‐8) of H‐14 (δ _H 1.26, m) with H‐9 (δ _H 1.31, m), H‐12 (δ _H 3.14, m), and H‐17 (δ _H 2.73, t, J = 9.2 Hz), as well as those of H‐8 (δ _H 2.22, m) with H₃‐18 (δ _H 0.76, s) and H₃‐19 (δ _H 1.13, s), suggested that 3 and 7 shared the same relative configurations. Combined with the ECD Cotton effects and calculation results (Figure 4), the absolute configuration of 3 was determined to be 8R, 9S, 10R, 12R, 13S,14S,17S as shown in Figure 1. Hence, the structure of 3 was established to be 12β‐methoxylpregna‐4,6‐diene‐3,20‐dione and named neristerone C.

Compound 4 was obtained as a white amorphous powder. It shared the same molecular formula, C₂₃H₃₀O₄, with compound 2, as established by the HR‐ESI‐MS at m/z 369.2069 [M − H] ⁻ (calcd for C₂₃H₂₉O₄, 369.2066) (Figure S4‐2)and accounted for 9 degrees of unsaturation. The ¹H and ¹³C NMR signals for 4 were assigned by using 1D and 2D NMR experiments (Table 1). Comparison of the ¹H and ¹³C NMR data of 4 (Figures S4‐3 and S4‐4) with those of 20R‐hydroxypregna‐4,6‐diene‐3,12‐dione (8) [24] revealed that they had the same skeleton and the main difference lay in the functional group of C‐20, which was confirmed as an acetyl group [δ _H 1.98 (3H, s, H₃‐2′); δ _C 170.9 (C‐1′) and 21.6 (C‐2′)]. The acetyl group was located at C‐20 based on the HMBC correlation (Figures 2 and S4‐7) from H‐20 (δ _H 4.77, m) to C‐1′. The NOESY correlations (Figures 3 and S4‐8) of H‐8 (δ _H 2.64, m) with H₃‐18 (δ _H 1.10, s) and H₃‐19 (δ _H 1.18, s), of H‐14 (δ _H 1.56, m) with H‐9 (δ _H 1.60, m)and H‐17 (δ _H 2.40, m), of H‐20 with H₃‐18 and H‐16α (δ _H 1.37, m), and of H‐17 with H‐16β (δ _H 1.89, m), suggested that 4 and 8 shared the same relative configuration. Furthermore, the good agreement of experimental and calculated ECD spectra (Figure 4) for 8R, 9S, 10R, 13S, 14S, 17S, 20R confirmed the absolute configurations of 4 as shown in Figure 1. Therefore, the structure of 4 was established as 20R‐oxyacetylpregna‐4,6‐diene‐3,12‐dione and named neristerone D.

2.2. In Vitro Anti‐Inflammatory Evaluation

The viability of RAW264.7 cells treated with these compounds (1–10) were assessed using the MTT assay. Compounds 5 and 6 exhibited significant cytotoxicity (p < 0.001), whereas the remaining compounds (1–4 and 7–10) maintained cell viability above 90% at a concentration of 50 µmol/L (Figure 5A). At the concentration (50 µmol/L), dexamethasone was used as the positive control. We evaluated the inhibitory effects of compounds 1–4 and 7–10 on LPS‐induced NO production at an LPS concentration of 1 µg/mL. All tested compounds exhibited potent NO inhibitory activity (Figure 5B). Compounds 1, 4, 7, and 10, which exhibited promising NO inhibition at 50 µmol/L, were further subjected to IC₅₀ determination. As shown in Figure 6, compounds 1 (IC₅₀ = 16.02 ± 0.95 µmol/L) and 4 (IC₅₀ = 18.13 ± 1.25 µmol/L) showed lower IC₅₀ values than dexamethasone (IC₅₀ = 21.48 ± 0.83 µmol/L), with significant differences observed (p < 0.05). Compound 10 (IC₅₀ = 11.27 ± 0.81 µmol/L) exhibited the strongest inhibitory activity, with an IC₅₀ significantly lower than that of dexamethasone (p < 0.01). Compound 7 (IC₅₀ = 21.87 ± 1.13 µmol/L) presented comparable anti‐inflammatory potency to dexamethasone (p < 0.05). Overall, the tested compounds showed anti‐inflammatory effects. Notably, compounds 1, 4, and 10 displayed particularly strong activity, with IC₅₀ values lower than that of the positive control dexamethasone.

FIGURE 5.

(A) Effects of compounds 1−10 on the viability of RAW264.7 cells; (B) Compounds 1−4 and 7−10 inhibit LPS‐induced NO production in RAW264.7 cells. ^*** p < 0.001 compared with the model group. ^### p < 0.001 for the model group vs. the control group.

FIGURE 6.

IC₅₀ values of compounds 1, 4, 7, and 10, and dexamethasone for inhibition of NO production in RAW264.7 cells. The data were calculated from three independent experiments and expressed as mean ± SD. IC₅₀ values were determined by non‐linear regression analysis. Statistical comparisons between each compound and dexamethasone were performed using one‐way ANOVA followed by Dunnett's post hoc test.

Based on the structural and pharmacological results, conduct a preliminary structure–activity relationship analysis of these polyene pregnane steroids. Comparing 1 with 5, it was found that the C‐12 acetyl group evidently correlates with improved anti‐inflammatory potency. Compound 1 bearing a C‐12 acetyl moiety showed more potent NO inhibitory effect than its C‐12 hydroxyl analogue (5). Notably, compound 5 exhibited obvious cytotoxicity toward RAW264.7 cells, which further limited its anti‐inflammatory application, whereas C‐12 acetylation effectively improved both bioactivity and cellular safety. Meanwhile, after methylation of the hydroxy group at C‐12 of 7 to afford 3, its anti‐inflammatory activity decreased compared to before. For the C‐20 position, compared with 8, 4 with a C‐20 acetyl group displayed reduced anti‐inflammatory activity, suggesting that free hydroxyl group at C‐20 is more conducive to maintaining favorable inhibitory capacity. Consistent with previous literature records, 8 possesses good anti‐inflammatory potential [24]. Overall substituent modification at C‐12 and C‐20 positions can obviously regulate the anti‐inflammatory activity of polyene pregnane skeletons.

3. Conclusions

Phytochemical exploration of the dried stems of N. oleander afforded four undescribed polyene pregnane steroids (1–4), along with six known related congeners (5–10). These results further expand the structural diversity of this steroid subclass and enrich the phytochemical resource library of the genus Nerium. Pharmacological evaluation assays demonstrated that compounds 1, 4, and 10 possess prominent anti‐inflammatory potential. Further compound expansion and molecular docking verification are still required to confirm the definite action rules and underlying mechanisms. Preliminary SAR analysis revealed that acetylation substitution exerts differential modulatory effects on anti‐inflammatory efficacy. Collectively, Steroids represent another major class of bioactive constituents of this plant, which exhibit prominent anti‐inflammatory effects.

4. Experimental Section

4.1. General Experimental Procedures

Thin‐layer chromatography (TLC) was performed using pre‐coated silica gel GF254 plates (Qingdao Haiyang Chemical Group Corp, Qingdao, China). Open column chromatography (CC) was performed using silica gel (200–300 mesh, Qingdao Haiyang Chemical Group Corp, Qingdao, China). HPLC was performed using the Agilent Technologies 1260 infinity (Germany), and the column used was an Epic C₁₈ column (4.6 × 250 mm, 5 µm, USA). Optical rotations were obtained using the P‐1020 digital polarimeter (JASCO Corporation, Tokyo, Japan). Semi‐preparative HPLC was performed using Agilent Technologies 1200 Series (Germany), and a semi‐preparative column (Neo Novi C₁₈ column,10 × 250 mm, 5 µm, China). UV spectra were measured using the JASCO V‐550 UV/vis spectrophotometer (JASCO Corporation, Tokyo, Japan). HR‐ESI‐MS and ESI–MS data were respectively obtained using the Micromass Q‐TOF and SCIEX Triple Quad 4500 mass spectrometer (USA). IR spectra were recorded using the QF‐510A FT‐IR spectrometer (Beijing Beifen‐Ruili Analytical Instrument Group Co., Ltd., Beijing, China). 1D and 2D NMR data were recorded using the Bruker AV 400 MHz instruments.

4.2. Plant Material

The fresh stems of N. oleander were collected from Guigang City, Guangxi Zhuang Autonomous Region, in June, 2022 and authenticated by Prof. Z. P. Gou (School of Pharmacy, Guangdong Medical University). The fresh stems of N. oleander were cut into small pieces, air‐dried at room temperature (25 ∼ 30°C) for 14 days, and then ground into a coarse powder before extraction. The voucher specimen (No. HH202206) had been deposited at the School of Ocean and Tropical Medicine, Guangdong Medical University, Zhanjiang 524023, P.R. China

4.3. Extraction and Isolation

The dried stems (50 kg) of N. oleander were reflux‐extracted twice with 75% ethanol (EtOH, 5 × 30 L) for 2 h per extraction. The crude extracts were concentrated under reduced pressure at 45°C to remove ethanol, then suspended in water and sequentially partitioned with ethyl acetate and n‐butanol. After concentration, the EtOAc fraction (612 g) and a n‐BuOH fraction (946 g) were obtained. Two soluble fractions (total 1558 g) were combined and subjected to silica gel column chromatography (CC) eluted with a DCM‐MeOH gradient (100:0 → 80:1 → 60:1 → 40:1 → 20:1 → 10:1 → 5:1, v/v, four column volumes for each gradient) to yield an extract rich in cardiac glycosides (256 g). An extract enriched with cardiac glycosides was subjected to silica gel column chromatography (CC) eluted with a PE‐EtOAc gradient (4:1 → 3:1→ 2:1→ 1:1, v/v, four column volumes for each gradient) to yield 9 fractions (Fr. 1 ∼ Fr. 9). Fr. 2 (55 g) was subjected to silica gel CC (PE/EtOAc, 4:1 → 3:1 → 2:1 → 1:1 → 1:2, v/v, five column volumes for each gradient) to give 10 subfractions (Fr. 2‐5∼Fr. 2‐7). Fr. 2‐5 (2.2 g) was applied to semi‐preparative RP‐HPLC (60% MeOH‐H₂O, a flow rate of 3.0 mL/min, 220 nm) to obtain 9 (t _R = 15.0 min, 9.9 mg, yield: 0.00064%), 5 (t _R = 18.1 min, 24.0 mg, yield: 0.00154%), 1 (t _R = 20.2 min, 9.9 mg, yield: 0.00064%). Fr. 2‐6 (1.7 g) was applied to semi‐preparative RP‐HPLC (58% MeOH‐H₂O, a flow rate of 3.0 mL/min, 220 nm) to obtain 10 (t _R = 9.5 min, 14.8 mg, yield: 0.00095%), and 7 (t _R = 25.0 min, 28.9 mg, yield: 0.00186%), 3 (t _R = 24.2 min, 11.8 mg, yield: 0.00076%). Fr. 2‐7 (1.5 g) was applied to semi‐preparative RP‐HPLC (56% MeOH‐H₂O, a flow rate of 3.0 mL/min, 220 nm) to obtain 6 (t _R = 30.5 min, 4.9 mg, yield: 0.00032%), 2 (t _R = 29.2 min, 1.5 mg, yield: 0.000096%), 8 (t _R = 28.2 min, 21.5 mg, yield: 0.00138%), 4 (t _R = 27.2 min, 11.2 mg, yield: 0.00072%).

Neristerone A (1). White amorphous powder; ${[\alpha ]}_D^{23}$ +71.5 (c = 0.2, MeOH); UV (MeOH) λ_max (log ε) 236 (2.45) nm, 286 (2.86) nm; IR (KBr) ν _max 3404, 3300, 2939, 2858, 1747, 1635, 1612, 1579 cm⁻¹; ECD (MeOH) λmax (Δε) 218 (− 0.25), 240 (− 3.15), 260 (+ 0.12), 290 (− 2.30), 338 (+ 3.87); (+)‐ESI‐MS m/z 369.2 [M + H]⁺; (−)‐HR‐ESI‐MS m/z 367.1905 [M − H]⁻ (calcd for C₂₃H₂₇O₄, 367.1909); ¹H and ¹³C NMR data, see Table 1.

Neristerone B (2). White amorphous powder; ${[\alpha ]}_D^{23}$ +86.5 (c = 0.2, MeOH); UV (MeOH) λ_max (log ε) 242 (2.89) nm; IR (KBr) ν _max 3394, 2943, 2854, 1670, 1647 cm⁻¹; ECD (MeOH) λmax (Δε) 218 (+ 37.25), 228 (+11.06), 248 (+ 66.72), 270 (+ 2.90), 300 (+ 5.24); (+)‐ESI‐MS m/z 371.2 [M + H]⁺; (−)‐HR‐ESI‐MS m/z 369.2066 [M − H]⁻ (calcd for C₂₃H₂₉O₄, 369.2066); ¹H and ¹³C NMR data, see Table 1.

Neristerone C (3). White amorphous powder; ${[\alpha ]}_D^{23}$ +88.6 (c = 0.4, MeOH); UV (MeOH) λ_max (log ε) 288 (3.16) nm; IR (KBr) ν _max 3388, 2941, 2872, 1683, 1658, 1647, 1612 cm⁻¹; ECD (MeOH) λ_max (Δε) 226 (− 14.72), 274 (− 52.62), 316 (+ 44.80); (+)‐ESI‐MS m/z 387.2 [M + HCOO]⁻; (+)‐HR‐ESI‐MS m/z 387.2173 [M + HCOO]⁻ (calcd for C₂₃H₃₁O₅, 387.2171); ¹H and ¹³C NMR data, see Table 1.

Neristerone D (4). White amorphous powder; ${[\alpha ]}_D^{23}$ +85.3 (c = 0.3, MeOH); UV (CHCl₃) λ_max (log ε) 288 (3.23) nm; IR (KBr) ν _max 3423, 2964, 2873, 1687, 1660, 1614 cm⁻¹; ECD (MeOH) λ_max (Δε) 220 (+ 1.30), 258 (− 2.04), 288 (− 7.42), 352 (+ 4.50); (+)‐ESI–MS m/z 371.2 [M + H]⁺; (−)‐HR‐ESI‐MS m/z 369.2069 [M − H]⁻ (calcd for C₂₃H₂₉O₄, 369.2066); ¹H and ¹³C NMR data, see Table 1.

12β‐Hydroxy pregna‐4,6,16‐triene‐3,20‐dione (5). The structure of 5 was confirmed by the ¹H and ¹³C NMR analysis, and its physical and spectroscopic data were consistent with those reported in the literature [28]. ¹³C NMR (CDCl₃, 125 MHz) δ 33.6 (C‐1), 33.9 (C‐2), 199.4 (C‐3), 124.3 (C‐4), 162.7 (C‐5), 129.1 (C‐6), 138.6 (C‐7), 34.8 (C‐8), 49.0 (C‐9), 36.1 (C‐10), 28.6 (C‐11), 73.2 (C12), 52.8 (C‐13), 51.6 (C‐14), 31.7 (C‐15), 149.0 (C‐16), 155.2 (C‐17), 11.5 (C‐18), 16.2 (C‐19), 198.9 (C‐20), 26.8 (C‐21).

12β‐Hydroxy‐4,16‐pregnadien‐3,20‐dione (6). The structure of 6 was confirmed by the ¹H and ¹³C NMR analysis, and its physical and spectroscopic data were consistent with those reported in the literature [34]. ¹³C NMR (CDCl₃, 75 MHz) δ 32.6 (C‐1), 33.9 (C‐2), 199.1 (C‐3), 124.4 (C‐4), 170.0 (C‐5), 32.1 (C‐6), 31.1 (C‐7), 26.7 (C‐8), 52.6 (C‐9), 38.6 (C‐10), 35.5 (C‐11), 73.3 (C12), 52.0 (C‐13), 53.4 (C‐14), 28.9 (C‐15), 149.4 (C‐16), 155.2 (C‐17), 11.7 (C‐18), 17.1 (C‐19), 199.0 (C‐20), 32.7 (C‐21).

12β‐Hydroxypregna‐4,6‐diene‐3,20‐dione (7). The structure of 7 was confirmed by the ¹H and ¹³C NMR analysis, and its physical and spectroscopic data were consistent with those reported in the literature [36]. ¹³C NMR (Pyridine‐d₅ , 125 MHz) δ 34.2 (C‐1), 34.0 (C‐2),198.1 (C‐3), 140.2 (C‐4), 163.0 (C‐5), 124.1 (C‐6), 128.3 (C‐7), 36.8 (C‐8), 52.1 (C‐9), 36.1 (C‐10), 30.6 (C‐11), 77.9 (C12), 51.2 (C‐13), 49.1 (C‐14), 23.8 (C‐15), 25.3 (C‐16), 62.4 (C‐17), 8.1 (C‐18), 16.1 (C‐19), 211.8 (C‐20), 31.8 (C‐21).

20R‐Hydroxypregna‐4,6‐diene‐3,12‐dione (8). The structure of 8 was confirmed by the ¹H and ¹³C NMR analysis, and its physical and spectroscopic data were consistent with those reported in the literature [24]. ¹³C NMR (CDCl₃, 125 MHz) δ 33.5 (C‐1), 33.6 (C‐2), 198.8 (C‐3), 124.7 (C‐4), 161.4 (C‐5), 129.0 (C‐6), 138.1 (C‐7), 36.3 (C‐8), 51.0 (C‐9), 36.2 (C‐10), 37.4 (C‐11), 216.6 (C12), 57.7 (C‐13), 52.7 (C‐14), 22.9 (C‐15), 24.8 (C‐16), 50.7 (C‐17), 12.1 (C‐18), 15.9 (C‐19), 68.0 (C‐20), 23.1 (C‐21).

16β,17β‐Epoxy‐12β‐hydroxypregna‐4,6‐diene‐3,20‐dione (9). The structure of 9 was confirmed by the ¹H and ¹³C NMR analysis, and its physical and spectroscopic data were consistent with those reported in the literature [24]. ¹³C NMR (CDCl₃, 125 MHz) δ 33.5 (C‐1), 33.9 (C‐2), 199.3 (C‐3), 124.3 (C‐4), 162.4 (C‐5), 129.1 (C‐6), 138.1 (C‐7), 34.9 (C‐8), 48.6 (C‐9), 36.0 (C‐10), 28.7 (C‐11), 72.2 (C12), 49.7 (C‐13), 58.2 (C‐14), 28.4 (C‐15), 67.1 (C‐16), 74.8 (C‐17), 10.1 (C‐18), 16.0 (C‐19), 208.2 (C‐20), 25.3 (C‐21).

12β‐hydroxyandrosta‐1,4,6‐triene‐3,17‐dione (10). The structure of 10 was confirmed by the ¹H and ¹³C NMR analysis, and its physical and spectroscopic data were consistent with those reported in the literature [35]. ¹³C NMR (CDCl₃, 125 MHz) δ 152.1 (C‐1), 128.5 (C‐2), 186.1 (C‐3), 124.5 (C‐4), 161.2 (C‐5), 128.9 (C‐6), 134.8 (C‐7), 36.6 (C‐8), 46.1 (C‐9), 40.9 (C‐10), 28.8 (C‐11), 72.1 (C12), 52.0 (C‐13), 46.7 (C‐14), 21.3 (C‐15), 35.5 (C‐16), 220.8 (C‐17), 8.3 (C‐18), 20.8 (C‐19), 68.0 (C‐20), 23.1 (C‐21).

4.4. ECD Calculations

Briefly, conformational analyses of compounds 1–4 were carried out via random conformational searches using the MMFF force field in Spartan'14. Conformers with Boltzmann populations over 5% were selected for further geometry optimization. Ground‐state geometries were optimized at the B3LYP/6‐31G(d) level in methanol using the IEFPCM solvation model with Gaussian 09 software. The optimized stable conformers were selected for further ECD calculations at the B3LYP/6‐311G (d,p) level in methanol. All quantum computations were performed on a high‐performance computing platform. The calculated ECD spectra were simulated by overlapping Gaussian functions (half bandwidth at 1/e peak height, 0.3 eV) with UV correction, and the final ECD curves and enantiomeric ECD curves were generated using SpecDis 1.70.1 software by Boltzmann weighting according to the calculated contribution of each conformer. Computational details are provided in the Supporting Information.

4.5. Cell Culture and Cytotoxicity Assay

RAW264.7 murine macrophage cells (RRID: CVCL_0493, purchased from ATCC, Rockville, MD, USA) were cultured in high‐glucose DMEM supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin at 37°C in a 5% CO₂ incubator. Cell viability was assessed using the MTT assay. Cells in the logarithmic growth phase were seeded into 96‐well plates at a density of 100,000 cells per well in 100 µL of medium, with three replicate wells for the blank control, positive control (dexamethasone), and each experimental compound. After 24 h of incubation, the cells were treated with different monomer compounds at a concentration of 50 µmol/L for another 24 h. Subsequently, 10 µL of MTT solution (5 mg/mL) was added to each well, followed by incubation for 4 h. The medium was carefully removed, and 150 µL of dimethyl sulfoxide (DMSO) was added to each well to dissolve the formazan crystals. After thorough mixing, the absorbance at 490 nm was measured using a microplate reader.

4.6. NO Production Bioassay

At noncytotoxic concentrations, the inhibitory effects of monomer compounds on nitric oxide (NO) production in LPS‐stimulated RAW264.7 cells were assessed using the Griess method. Briefly, cells in the logarithmic growth phase were seeded into 96‐well plates at a density of 100,000 cells per well in 100 µL of medium. The experimental design included a blank control group, a model group (LPS alone), a positive control group (dexamethasone), and compound treatment groups. All experiments were performed in three independent biological replicates (n = 3), with each replicate consisting of three technical replicates per condition. After 24 h of culture (37°C, 5% CO₂), cells were pretreated with 50 µmol/L of the test compounds or dexamethasone for 2 h, followed by co‐stimulation with LPS at a final concentration of 1 µg/mL for an additional 24 h. For NO measurement, 50 µL of cell culture supernatant was collected from each well and mixed with 50 µL of Griess reagent I. After incubation at room temperature for 5 min, 50 µL of Griess reagent II was added. The mixture was mixed thoroughly, and the absorbance at 540 nm was measured using a microplate reader. The NO concentration was determined by interpolation from a sodium nitrite standard curve. For compounds that exhibited significant inhibition, the half‐maximal inhibitory concentration (IC₅₀) against NO release was calculated.

Author Contributions

Tong‐Xin Zheng: writing – original draft, data curation. Xin‐Hong Wang: data curation. He‐Hui Zhan: data curation. Zi‐Xin Lin: writing – original draft, data curation. Cheng‐Jun Huang: data curation. Xiao‐San Li: project administration, writing – review and editing, resources, investigation. Li Liu: investigation, supervision. Hui Luo: project administration, resources, writing – review and editing. Yong‐Mei Huang: data curation, investigation. Xing‐Wei Li: data curation.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: cbdv71446‐sup‐0001‐SuppMat.pdf

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.