Six new diterpenoids with anti-inflammatory and cytotoxic activity from Isodon serra

Abstract

Diterpenoids occupy an important slot of the natural products diversity space with wide ranges of bioactivities and complex structures, providing potential applications for the development of therapeutics. In this study, we reported four new abietane-type diterpenoids viroxocin B-E (1–4), a new totarane-type diterpenoid viroxocin F (5), and a new sempervirane-type diterpenoid viroxocin G (6) along with four known compounds (7–10), isolated and identified from a widely used Traditional Chinese Medicine, Isodon serra (I. serra). Their structures were established by spectroscopic data analysis, experimental and calculated electronic circular dichroism (ECD) data, as well as X-ray diffraction analysis. Compounds 2, 5, 7, 8 and 10 exhibited promising anti-inflammatory activities in lipopolysaccharide (LPS)-induced RAW 267.4 cells, and their inhibition rates on NO production were more than 60% at 10 μM. Compound 7 showed cytotoxicity against human renal cell carcinoma 769P at 20 μM, the inhibition rate was 52.66%.

1. Introduction

Plants of the Isodon genus have been an abundant source of terpene natural products, prized both for the diversity and complexity of their chemical structures, and for their extensive profile of biological activity [1]. Some plants from this genus have been used in Traditional Chinese Medicine, including I. eriocalyx [2], I. adenolomus [3], and I. serra [4].

I. serra, a perennial plant named “Xihuangcao” in China, is mainly distributed in Guangdong, Jiangxi, and Fujian provinces in China [4]. As a Chinese folk medicine, I. serra has been popularly used to treat arthritis, enteritis, jaundice, hepatitis, and acute cholecystitis [5]. Previous phytochemical investigations on this species afforded abundant bioactive diterpenoids, involving enmein, spirolactone, C-20 non-oxygenated and C-20 oxygenated types [6]. Some of these diterpenoids possess potential cytotoxic, anti-inflammatory, and antibacterial activities, such as rabdocoestin B, serrin F and shikokianin A [7,8].

To find new active substances, we have reinvestigated the chemical components of I. serra collected from the Qingyuan in Guangdong Province of China. Four new abietane-type diterpenoids (1–4), a new totarane-type diterpenoid (5), and a new sempervirane-type diterpenoid (6), along with four known compounds graciliflorin F (7) [9], graciliflorin A (8) [10], graciliflorin B (9) [10] and wulfenioidin G (10) [11] were isolated from the ethanol extract of I. serra (Fig. 1). Described herein were the structural elucidation using NMR, HR-ESI-MS, ECD, and X-ray diffraction data. Compounds 2, 5, 7, 8 and 10 exhibited promising anti-inflammatory activities in LPS-induced RAW 267.4 cells, and their inhibition rates on NO production were more than 60% at 10 μM. Compound 7 showed cytotoxicity against human renal cell carcinoma 769P at 20 μM, the inhibition rate was 52.66%.

Fig. 1.

Chemical structures of compounds 1–10.

2. Experimental sections

2.1. General experimental procedures

The ECD spectra were recorded on a JASCO-J1500 spectropolarimeter. The NMR spectra were obtained on a Bruker Avance NEO 600 MHz spectrometer. HR-ESI-MS data were obtained on an Agilent 6560 mass spectrometer. Semipreparative HPLCs equipped with Agilent ZORBAX Eclipse XDB-C18 column (9.4 × 250 nm, 5 μm) were performed using SEP LC-52 series HPLC instruments with corresponding detectors, fraction collectors, and software. Column chromatography (CC) was performed using silica gel (Yu-Ming-Yuan Silysia Chemical Ltd., Qingdao, China) and Sephadex LH-20 (Pharmacia Biotech AB, Uppsala, Sweden). TLC was conducted on precoated silica gel GF254 plates. Spots were visualized under UV light (254 nm). X-ray data were collected on an ROD, Synergy Custom system, HyPix diffractometer equipped with graphite-monochromatized Cu Kα radiation.

2.2. Plant material

Whole plants of I. serra were collected by Haibiao Guo (Hutchison Whampoa Guangzhou Baiyunshan Chinese Medicine Co., Ltd., Guangzhou, China) in August 2020 from the Qingyuan in Guangdong Province of China. The plant was identified by Dr. Shengpeng Wang (State Key Laboratory of Quality Research in Chinese Medicine, Institute of Chinese Medical Sciences, University of Macau, Macau, China).

2.3. Extraction and isolation

The dried whole plant of I. serra (15 kg) was extracted with EtOH (3 × 20 L) at room temperature. After filtration, the solvent was removed under vacuum to give a crude extract (650 g), which was suspended in H₂O (2 L) and partitioned with EtOAc (3 × 1.5 L). The EtOAc extract (650 g) was subjected to silica gel (200–300 mesh) CC with petroleum ether (PE)/EtOAc gradients (from 10:1 to 0:1, v/v) and CH₂Cl₂/MeOH gradients (from 10:1 to 0:1, v/v) to afford 15 fractions (A1-A15) monitored by TLC.

Fr. A8 (20.5 g) was obtained from initial Sephadex LH-20 CC [CH₂Cl₂/MeOH/H₂O (5:5:1, v/v/v)] and purified using semipreparative HPLC (MeOH/H₂O, 80:20, v/v, 3.0 mL/min) to yield compound 1 (5.0 mg, t_R = 21.0 min).

Fr. A9 (26.9 g) was separated to silica gel (200–300 mesh) CC with PE/EtOAc gradients (from 5:1 to 1:1, v/v) to afford 15 fractions (B1-B15) monitored by TLC. Fr. B2 (2.1 g) purified using semipreparative HPLC (MeOH/H₂O, 60:40, v/v, 3.0 mL/min) to yield compounds 8 (6 mg, t_R = 17.7 min) and 9 (6 mg, t_R = 19.5 min). Fr. B3 (1.2 g) purified using semipreparative HPLC (MeOH /H₂O, 80:20, v/v, 3.0 mL/min) to yield compounds 10 (5.5 mg, t_R = 15.9 min). Fr. B4 (1.8 g) was subjected to repeated silica gel CC and purified using semipreparative HPLC (MeOH/H₂O, 65:35, v/v, 3.0 mL/min) to yield compound 6 (3.0 mg, t_R = 50.1 min). Fr. B6 (1.6 g) was subjected to Sephadex LH-20 CC and further purified by semipreparative HPLC with MeOH/H₂O (60:40, v/v) to yield compound 4 (1.6 mg, t_R = 74.0 min).

Fr. A10 (24.3 g) was separated by an Sephadex LH-20 eluted with isocratic of MeOH/H₂O (80%, v/v) to afford 4 fractions (C1-C4) monitored by TLC. Fr. C2 (3.0 g) purified using semipreparative HPLC (MeOH/H₂O, 80:20, v/v, 3.0 mL/min) to yield compounds 2 (3.8 mg, t_R = 9.7 min). Fr. C3 (1.4 g) purified using semipreparative HPLC (MeOH/H₂O, 60:40, v/v, 3.0 mL/min) to yield compounds 7 (2 mg, t_R = 17.6 min). Fr. C4 (1.7 g) was separated by an Sephadex LH-20 eluted with isocratic of MeOH/H₂O (80%, v/v) and purified using semipreparative HPLC (MeOH/H₂O, 65:35, v/v, 3.0 mL/min) to yield compounds 3 (3.0 mg, t_R = 20.7 min).

Fr. A12 (17.3 g) was separated by repeated silica gel chromatography and purified using semipreparative HPLC (MeOH/H₂O, 80:20, v/v, 3.0 mL/min) to yield compound 5 (3 mg, t_R = 17.8 min).

2.4. Spectroscopic data of compounds

2.4.1. Viroxocin B (1)

yellow powder; IR (KBr) ν_max 3448, 2928, 2855, 2076, 1636, 661 cm⁻¹; ¹H and ¹³C NMR data, see Table 1; HR-ESI-MS m/z 297.1869 [M + H-H₂O]⁺ (calcd for C₂₀H₂₅O₂, 297.1854).

Table 1.

¹H (600 MHz) and ¹³C (150 MHz) NMR data of compound 1 and 2 in CDCl₃, 3 in methanol‑d₄ (δ in ppm, J in Hz).

Position	1		2		3
	δ H	δ C	δ H	δ C	δ H	δ C
1	a 3.91 (m)	26.5	a 3.40 (m)	26.7	4.88 (d, J = 1.8)	66.2
	b 2.93 (m)		b 1.90 (m)
2	a 2.18 (m)	23.0	2.77 (m)	33.3	5.32 (dd, J = 8.5, 1.8)	77.2
	b 1.64 (m)
3	a 1.59 (m)	33.3		213.6	4.82 (d, J = 8.5)	120.4
	b 1.25 (m)
4		85.4		49.6		137.9
5		127.5		173.5		132.8
6	7.12 (d, J = 8.3)	127.3	6.43 (s)	123.2	7.16 (d, J = 8.4)	126.4
7	7.49 (d, J = 8.3)	126.4		184.4	7.62 (d, J = 8.4)	127.5
8		130.3		123.0		125.8
9		131.8		134.2		120.6
10		133.5		40.5		124.3
11		135.2		141.9		134.5
12		146.6		146.4		139.9
13		133.6		128.1		135.5
14	7.46 (s)	120.5	7.63 (s)	116.5	7.39 (s)	115.2
15		72.6		80.7		73.2
16	1.73 (s)	29.3	1.65 (s)	26.3	1.69 (s)	28.5
17	1.70 (s)	29.5	1.66 (s)	26.1	1.69 (s)	28.6
18	1.16 (s)	27.5	1.41 (s)	29.7	1.87 (s)	17.3
19	1.64 (s)	26.3	1.47 (s)	26.6	1.57 (s)	24.4
20	2.43 (s)	19.9	1.49 (s)	20.2	2.49 (s)	16.4
OCH3			3.26 (s)	51.0

2.4.2. (10S)-Viroxocin C (2)

colorless crystals (from MeOH); [α²⁵_D] +58.065 (c 0.3, MeOH); IR (KBr) ν_max 3368, 2974, 2935, 1712, 1650, 1601, 1462, 1370, 1307, 1273, 1150, 1136, 1054, 1027, 914, 885, 680 cm⁻¹; ECD (MeOH) λ_max (Δε) 256 (+19.43), 294 (+1.46), 327 (−0.95), 353 (−0.99) nm; ¹H and ¹³C NMR data, see Table 1; HR-ESI-MS m/z 359.1850 [M + H]⁺ (calcd for C₂₁H₂₇O₅, 359.1858).

2.4.3. (1S, 2S)-Viroxocin D (3)

white powder; [α²⁵_D] −14.286 (c 0.15, MeOH); IR (KBr) ν_max 3855, 3362, 2925, 2852, 1710, 1597, 1415, 1378, 1258, 1127, 1030, 962, 910, 819, 791, 758, 681,607 cm⁻¹; ECD (MeOH) λ_max (Δε) 211 (+0.49), 251 (+1.32), 287 (−0.49), 330 (+0.20), 379 (+0.29) nm; ¹H and ¹³C NMR data, see Table 1; HR-ESI-MS m/z 351.1569 [M + Na]⁺ (calcd. For C₂₀H₂₄O₄Na, 351.1572).

2.4.4. (2S)-Viroxocin E (4)

yellow amorphous powder; IR (KBr) ν_max 3447, 1638, 1247, 1119, 618 cm⁻¹; ECD (MeOH) λ_max (Δε) 207 (−3.02), 222 (+7.20), 235 (−1.92), 268 (+2.55), 297 (−0.68), 313 (+0.69), 327 (−0.22) nm; ¹H and ¹³C NMR data, see.

Table 2; HR-ESI-MS m/z 327.1596 [M + H]⁺ (calcd for C₂₀H₂₃O₄, 327.1596).

Table 2.

¹H (600 MHz) and ¹³C (150 MHz) NMR data of compound 4–6 in CDCl₃ (δ in ppm, J in Hz).

Position	4		5		6
	δ H	δ C	δ H	δ C	δ H	δ C
1		192.0	a 2.19 (m)	39.4	a 2.51 (m)	37.7
			b 1.38 (m)		b 1.97 (m)
2	5.42 (d, J = 8.7)	78.8	a 2.49 (m)	21.9	a 2.74 (m)	34.6
			b 1.59 (m)		b 2.62 (m)
3	5.27 (d, J = 8.8)	117.6	a 1.72 (m)	40.7		216.5
			b 1.42 (m)
4		140.8		32.7		47.4
5		139.7	1.33 (m)	45.7	1.87 (dd, J = 12.1, 2.8)	50.6
6	7.18 (d, J = 8.5)	127.8	a 2.26 (m)	19.5	1.82 (m)	19.8
			b 1.77 (m)
7	7.80 (d, J = 8.4)	133.1	a 2.94 (m)	28.1	a 2.99 (m)	31.1
			b 2.61 (m)		b 2.87 (m)
8		124.9		136.2		145.4
9		122.4		134.5		138.9
10		119.3		40.6		36.9
11		133.5	6.92 (d, J = 8.4)	123.6	7.59 (s)	127.5
12		140.6	6.51 (d, J = 8.4)	114.2		118.6
13		134.1		152.1		159.7
14	7.36 (s)	116.0		130.9	6.68 (s)	117.4
15		72.9	3.34 (m)	27.2		203.9
16	1.70 (s)	28.9	1.36 (dd, J = 7.1, 4.7)	20.4	2.61 (s)	26.5
17	1.72 (s)	28.7	1.36 (dd, J = 7.1, 4.7)	20.5
18	1.82 (s)	18.0	a 3.75 (d, J = 11.1)	66.3	1.18 (s)	26.6
			b 3.30 (d, J = 11.0)
19	1.69 (s)	25.0	0.79 (s)	23.6	1.15 (s)	21.2
20	2.74 (s)	21.3	4.54 (s)	105.9	1.31 (s)	25.0
OCH3			3.09 (s)	55.3

2.4.5. (4S, 5S, 10R, 20R)-Viroxocin F (5)

yellow powder; [α²⁰_D] −10.0 (c 0.2, MeOH); IR (KBr) ν_max 3368, 2927, 2866, 2374, 1715, 1651, 1590, 1453, 1367, 1280, 1232, 1188, 1110, 1026, 954, 889, 859, 808, 691, 631 cm⁻¹; ECD (MeOH) λ_max (Δε) 203 (+11.86), 207 (+11.47), 234 (+14.60), 279 (−3.15), 285 (−3.30) nm; ¹H and ¹³C NMR data, see.

Table 2; HR-ESI-MS m/z 353.2090 [M + Na]⁺ (calcd for C₂₁H₃₀O₃Na, 353.2093).

2.4.6. (5R, 10S)-Viroxocin G (6)

colorless crystals (from MeOH); [α²⁵_D] +14.545 (c 0.8, MeOH); IR (KBr) ν_max 3424, 2955, 2924, 2853, 1642, 1600, 1459, 1382, 1265, 1187, 1126, 1038, 1011, 951 cm⁻¹; ECD (MeOH) λ_max (Δε) 219 (+8.46), 243 (+1.12), 265 (+2.31), 335 (−0.52) nm; ¹H and ¹³C NMR data, see.

Table 2; HR-ESI-MS m/z 301.1802 [M + H]⁺ (calcd for C₁₉H₂₅O₃, 301.1804).

2.5. X-ray crystallographic data

The crystal structure of compound 2 and 6 was solved by direct methods using SHELXS-97. Refinements were performed with SHELXL-2013 using full-matrix least-squares calculations on F with anisotropic displacement parameters for all the non-hydrogen atoms. The hydrogen atom positions were geometrically idealized and allowed to ride on their parent atoms.

2.6. Experimental ECD data

Experimental ECD of compound 2–6 were carried out on a JASCO J1500 (Japan). The data was collected by BioKine software at 200–400 nm. Compound 2 in MeOH, compound 5 in EtOH. All data analysis were carried out using the Excel.

2.7. Calculation of ECD data

The geometries of compounds 3 and 4 were optimized at the B3lyp/6–311 + g (2d, p) level of theory using the Gaussian 09 suite of programs [12]. The NMR calculations were carried out at the B3lyp/6–311 + g (2d, p) level of theory with GIAO method (in which the values of TMS at the B3LYP/6–311 + G (2d, p) level with GIAO method were used as a reference). The ECD spectra for the 3 and 4 were obtained using the time-dependent density functional theory (TDDFT) at the B3lyp/6–311 + g (2d, p) level of theory. The IEFPCM solvation model was employed for the CD spectrum and NMR calculations, and the ethanol and chloroform were used as the solvents which were in agreement with the experiment conditions, respectively.

The 2D structure of compound 5 were drawn in Maestro [13] and 3D-energy minimized at physiological pH 7.4 using the LigPrep [14] in the Schrödinger software, using the OPLS3e force field. The conformational search of the compound 5 was performed in the gas phase by deploying the MacroModel [15] program implemented in Schrödinger using the mixed torsional/low mode sampling method at an energy cutoff of 10 kcal/mol [16]. The resulting conformers were optimized by applying the mPW1PW91 functional and 6–311 + G (2d, p) basis set in EtOH using the polarizable continuum (PCM) solvation model [17] implemented in Gaussian 16 Rev. B.01 [18]. All of the optimized conformers of 5 were used for simulating the ECD spectra using time-dependent density functional theory (TDDFT) [19,20] at the level of mPW1PW91/6–311 + G (2d, p) in MeOH. ECD curves were generated by SpecDis version 1.71 [20,21] software using a Gaussian function with half bandwidth ranging from 0.16 to 0.29 eV, for comparison to experimental data. The calculated molecular orbitals (MOs) were analyzed with Avogadro 1.2.0 and the isosurface values employed were between 0.03 and 0.05 e⁻/au (Table 3) [22].

Table 3.

Key Transitions, Excitation Energies, Oscillator, and Rotatory Strengths that Contribute to the ECD Spectrum of the Lowest-energy Conformer 2 of 6 L at the PW1PW91/6–311 + G (2d, p) Level with PCM in EtOH.

Conformer	Transition	λ (nm)a	ΔE (eV)b	f c	R d vel	Coefficient
MO 90 (HOMO)	90➔91	251.71	4.9246	0.0448	−2.2006	0.63449
MO 91 (LUMO)	90➔92	231.86	5.3475	0.0034	3.6131	0.62455
MO 93 (LUMO+2)	90➔93	222.56	5.5708	0.0157	3.264	0.54972

2.8. Anti-inflammatory assay

2.8.1. Cell culture, viability assay

RAW 264.7 cells were purchased from the American Type Culture Collection (ATCC, Manassas, VA, USA). RAW 264.7 cells were cultured in DMEM supplemented with 12% FBS, 100 UmL⁻¹ penicillin, and 100 μgmL⁻¹ streptomycin in a humidified incubator containing 5% CO₂ and maintained at 37 °C. The cells were seeded in 96-well plates for 1 × 10⁵ cells per well. After cell attachment, the cells were treated with compounds (10 μM) or aspirin (5 μM), and LPS (1 μg/mL) for 24 h. The cell morphology was observed, and the cell vitality was detected using the MTT assay.

2.8.2. NO production bioassay

The nitrite concentration in the medium was measured according to the Griess reaction as an indicator of NO production. Briefly, RAW 264.7 cells were seeded into 96-well plates at a density of 1 × 10⁵ cells/well and stimulated with 1 μg/mL LPS in the presence or absence of test compounds. After incubation at 37 °C for 24 h, 50 μL of cell-free supernatant was mixed with 50 μL Griess reagent I and 50 μL Griess reagent II. The absorbance was measured in a microplate reader at 540 nm and compared with a calibration curve prepared using NaNO₂ standards.

2.9. Cytotoxicity assay

769P cells were purchased from the China Center for Type Culture Collection (CCTCC, Wuhan, Hubei, China). 769P cells were cultured in RPMI 1640 supplemented with 10% FBS, 100 UmL⁻¹ penicillin, and 100 μgmL⁻¹ streptomycin in a humidified incubator containing 5% CO₂ and maintained at 37 °C. The cells were seeded in 96-well plates for 5000 cells per well. After cell attachment, the cells were treated with compounds (20 μM) for 24 h. The cell morphology was observed, and the cell vitality was detected using the CCK8 assay.

3. Results and discussion

3.1. Structural elucidation of compounds 1–6

Compound 1 (Fig. 1) was obtained as a yellow powder. The molecular formula of 1 was determined as C₂₀H₂₆O₃ based on ¹³C NMR data and the protonated molecular ion at m/z 297.1869 [M + H-H₂O]⁺, (calcd for C₂₀H₂₅O₂, 297.1854) in the HR-ESI-MS, implying eight indices of hydrogen deficiency. The ¹H NMR spectrum exhibited signals characteristic of five methyl groups [δ_H 2.43 (3H, s, H-20), 1.73 (3H, s, H-16), 1.70 (3H, s, H-17), 1.64 (3H, s, H-19) and 1.16 (3H, s, H-18)]. The ¹³C NMR and HSQC spectrum suggested the presence of five methyls at δ_C 29.5, 29.3, 27.5, 26.3, and 19.9. In addition, there were three methylenes at δ_C 33.3, 26.5, and 23.0, three aromatic methines at δ_C 127.3, 126.4, and 120.5, and nine non-protonated carbons at δ_C 146.6, 135.2, 133.6, 133.5, 131.8, 130.3, 127.5, 85.4, and 72.6. The above NMR data were similar to those of a recently described natural product, 1-deoxyviroxocin (10-isopropyl-2, 2, 6-trimethyl-2, 3, 4, 5-tetrahydronaphtho[1, 8-bc] oxocin-11-ol) from Taxodium distichum [23]. Unlike the known structure, compound 1 has a hydroxy group at C-15 (δ_C 72.6), that was confirmed by the key HMBC correlations (Fig. 2) from H-14 (δ_H 7.46), H₃–16 (δ_H 1.73) and H₃–17 (δ_H 1.70) to C-15 (δ_C 72.6). Compound 1 was assigned as 10-isopropyl-2, 2, 6-trimethyl-2, 3, 4, 5-tetrahydronaphtho[1, 8-bc] oxocin-11, 15-diol, that we named viroxocin B.

Fig. 2.

Key HMBC (H → C) and COSY (H━H) Correlations of Compounds 1–6.

Compound 2 (Fig. 1) was isolated as colorless crystals (from MeOH). The molecular formula was established as C₂₁H₂₆O₅ with nine indices of hydrogen deficiency according to the HR-ESI-MS data (m/z 359.1850 ([M + H]⁺, calcd for C₂₁H₂₇O₅ 359.1858) and ¹³C NMR data. The ¹H NMR spectrum of 2 displayed signals of five methyls [δ_H 1.65 (3H, s, H-16), 1.66 (3H, s, H-17), 1.41 (3H, s, H-18), 1.47 (3H, s, H-19), 1.49 (3H, s, H-20)], an O-methyl singlet [δ_H 3.26 (3H, s, OMe)] and two aromatic protons [δ_H 7.63 (1H, s, H-14), 6.43 (1H, s, H-6)]. Moreover, analysis of the ¹³C NMR and HSQC spectra revealed the presence of five methyls (δ_C 29.7, 26.6, 26.3, 26.1, 20.2), two methylenes (δ_C 33.3, 26.7), two aromatic methines (δ_C 123.2, 116.5), an O-methyl (δ_C 51.0), nine non-protonated carbons (δ_C 173.5, 146.4, 141.9, 134.2, 128.1, 123.0, 80.7, 49.6, 40.5) and two carbonyl groups (δ_C 213.6, 184.4). Analysis of the 1D and 2D NMR spectroscopic data suggested 2 to be an abietane diterpenoid closely related to gerardianin C [24]. Compared with those of gerardianin C, the ¹³C NMR data of 2 exhibited an additional carbonyl group (δ_C 213.6). The HMBC correlations (Fig. 2) from H-1a (δ_H 3.40), H₃–18 (δ_H 1.41) and H₃–19 (δ_H 1.47) to C-3 (δ_C 213.6) established the additional carbonyl group in 2. Finally, the 2D structure and absolute configuration of 2 were further confirmed by the X-ray diffraction data analysis (Fig. 3). The final refinement of the Cu Kα data resulted in a small Flack parameter of 0.05 (12), allowing the assignment of the absolute configuration of 2. Therefore, the absolute configuration of 2 was defined as (10S). Compound 2 was assigned as (10S)-3, 7-one-5, 6-dehydro-11, 12-dihydroxy-15-O-methyl methoxyroyleanone, that we named viroxocin C.

Fig. 3.

X-ray ORTEP Drawing of the Structure of Compound 2 and 6.

Compound 3 (Fig. 1) was obtained as a white powder, and its molecular formula was determined to be C₂₀H₂₄O₄ based on the ¹³C NMR data and a sodium adduct ion at m/z 351.1569 ([M + Na]⁺, calcd for C₂₀H₂₄O₄Na, 351.1572) in its positive mode HR-ESI-MS, implying nine indices of hydrogen deficiency. In the ¹H NMR spectrum, five methyls [δ_H 2.49 (3H, s, H-20), 1.87 (3H, s, H-18), 1.69 (3H, s, H-17), 1.69 (3H, s, H-16) and 1.57 (3H, s, H-19)], two oxymethine protons [δ_H 5.32 (1H, dd, J = 8.5, 1.8 Hz, H-2), 4.88 (1H, d, J = 1.8 Hz, H-1)], and four olefinic protons [δ_H 7.62 (1H, d, J = 8.4 Hz, H-7), 7.39 (1H, s, H-14), 7.16 (1H, d, J = 8.4 Hz, H-6), 4.82 (1H, d, J = 8.5 Hz, H-3)] were observed. The ¹³C NMR and HSQC data of 3 showed 20 carbon resonances, consisting of five methyls at δ_C 28.6, 28.5, 24.4, 17.3, 16.4, six methines (two oxygenated at δ_C 77.2, 66.2 and four olefinic at δ_C 127.5, 126.4, 120.4, 115.2), and nine non-protonated carbons at δ_C 139.9, 137.9, 135.5, 134.5, 132.8, 125.8, 124.3, 120.6 and 73.2. The NMR data were similar to those of de-O-ethylsalvonitin isolated from Salvia prionitis Hance (Labiatae) [25]. Unlike this known structure, compound 3 has a hydroxy group at C-15 (δ_C 73.2), that was confirmed by the key HMBC correlations (Fig. 2) from H₃–16 (δ_H 1.69) and H-14 (δ_H 7.39) to C-15 (δ_C 73.2). The ³J_1,2 coupling constant of 1.8 Hz suggested that compound 3 possessed a 1, 2-cis relative configuration [25,26]. The absolute configuration of 3 was defined as (1S, 2S) by comparison of its experimental and calculated ECD spectra (Fig. 4). On the basis of the above findings, the structure of compound 3 was characterized as (1S, 2S)-15-hydroxy-12-de-O-ethylsalvonitin, that we named viroxocin D.

Fig. 4.

Experimental and Calculated ECD Data of Compounds 3 (a), 4 (b) in MeOH and 5 (c) in EtOH.

Compound 4 (Fig. 1) was obtained as a yellow amorphous powder. Its molecular formula was established to be C₂₀H₂₂O₄ by the positive ion mode HR-ESI-MS (m/z 327.1596 [M + H]⁺, calcd for C₂₀H₂₃O₄, 327.1596) and ¹³C NMR data. The ¹H NMR spectrum of 4 displayed resonance signals for five methyl singlets [δ_H 1.70 (3H, s, H-16), 1.72 (3H, s, H-17), 1.82 (3H, s, H-18), 1.69 (3H, s, H-19) and 2.74 (3H, s, H-20)], one oxymethine proton [δ_H 5.42 (1H, d, J = 8.7 Hz, H-2)], and four olefinic protons [δ_H 7.18 (1H, d, J = 8.5 Hz, H-6), 7.80 (1H, d, J = 8.4 Hz, H-7), 7.36 (1H, s, H-14) and 5.27 (1H, d, J = 8.8 Hz, H-3)]. The ¹³C NMR and HSQC spectra exhibited 20 carbon signals, including five methyls (δ_C 28.9, 28.7, 25.0, 21.3, 18.0), five methines (one oxygenated at δ_C 78.8 and four olefinic at δ_C 133.1, 127.8, 117.6, 116.0), ten non-protonated carbons (δ_C 140.8, 140.6, 139.7, 134.1, 133.5, 124.9, 122.4, 119.3, 72.9, including one carbonyl carbon δ_C 192.0). The 1D and 2D NMR spectrum of 4 suggested a 4, 5-seco-abietane substructure similar to that of prionoid B [27]. A detailed comparison of the NMR data of 4 with those of the known compound prionoid B revealed a high degree of similarity except for the presence of one additional hydroxy group in 4. This hydroxy group was placed at C-15, as inferred from the related HMBC correlation (Fig. 2) of C-15 (δ_C 72.9) with H₃–16 (δ_H 1.70) and H₃–17 (δ_H 1.72). Subsequently, the (2S) absolute configuration of 4 was determined by comparing the experimental and calculated ECD data (Fig. 4). Consequently, the structure of compound 4 was established as (2S)-15-hydroxyprionoid B, that we named viroxocin E.

Compound 5 (Fig. 1) was obtained as a yellow powder, and its molecular formula was determined to be C₂₁H₃₀O₃ based on the ¹³C NMR data and a sodium adduct ion at m/z 353.2090 [M + Na]⁺ (calcd for C₂₁H₃₀O₃Na, 353.2093) in its positive mode HR-ESI-MS, implying seven indices of hydrogen deficiency. The ¹H NMR spectrum of 5 displayed resonances for a methyl singlet [δ_H 0.79 (3H, s, H-19)], an isopropyl [δ_H 3.34 (1H, m, H-15), 1.36 (6H, dd, J = 7.1, 4.7 Hz, H-16, H-17)], an O-methyl [δ_H 3.09 (3H, s, OMe)], two oxymethylene protons [δ_H 3.75 (1H, d, J = 11.1 Hz, H-18a) and 3.30 (1H, d, J = 11.0 Hz, H-18b)], an oxymethine proton [δ_H 4.54 (1H, s, H-20)], and two aromatic protons [δ_H 6.92 (1H, d, J = 8.4 Hz, H-11) and 6.51 (1H, d, J = 8.4 Hz, H-12)]. The ¹³C NMR and HSQC spectra in CDCl₃ exhibited 21 carbon signals including three methyls (δ_C 23.6, 20.5, 20.4), an O-methyl (δ_C 55.3), six methylenes (δ_C 40.7, 39.4, 28.1, 21.9, 19.5 and one oxygenated at δ_C 66.3), five methines (δ_C 123.6, 114.2, 105.9, 45.7, 27.2), and six non-protonated carbons at δ_C 152.1, 136.2, 134.5, 130.9, 40.6 and δ_C 32.7. The above NMR data were similar to those of the (−)-(4S, 5S, 10R, 20R)-12, 18-dihydroxyabieta-8, 11, 13-trien-20-formyl 18, 20-methylacetal isolated from Fraxinus sieboldiana Blume (Oleaceae) [28]. Unlike this known structure, compound 5 was a totarane-type rather than an abietane-type diterpenoid, which was confirmed by the key COSY correlations from H-11 (δ_H 6.92) to H-12 (δ_H 6.51) and HMBC correlations (Fig. 2) from H₂–7 (δ_H 2.94, 2.61), H-15 (δ_H 3.34), and H₃–16/17 (δ_H 1.36) to C-14 (δ_C 130.9). The relative configuration of 5 was defined by NOESY data (Fig. 5). NOESY correlations between H-18a and H-6a, OCH₃ suggested those atoms to be assigned as β-oriented (Fig. 5). The NOESY cross-peaks between H-1a and H-20, H-2a indicated these protons to be assigned randomly as β-oriented (Fig. 5). NOESY correlations between H₃–19 and H-3b, H-5 suggested H-3b and H-5 to be assigned as α-oriented (Fig. 5). The absolute configuration of 5 was determined by comparing the experimental and computed ECD spectra (Fig. 4). The experimental CE at ca 280 nm, attributed to the¹L_b transition of the aromatic chromophore, represents the π → π* transition from MO90 (HOMO) to MO91 (LUMO) at 251.71 nm in the calculated ECD spectrum (Fig. 6). In addition, the experimental CE at ca 235 nm, attributed to the ¹L_a transition of the aromatic chromophore, represents the transition from MO90 (HOMO) to MO93 (LUMO+2)) at 235 nm in the calculated ECD spectrum. The close similarity of the experimental and calculated ECD spectra, thus, defines the (4S, 5S, 10R, 20R) absolute configuration of compound 5. On the basis of the above findings, the structure of compound 5 was characterized as (4S, 5S, 10R, 20R)-13, 18-dihydroxytotara-8, 11, 13-trien-20-formyl 18, 20-methyl acetal. It is a new diterpenoid with a rearranged totarane skeleton that we named viroxocin F.

Fig. 5.

Key NOESY (H↔H) Correlations of 5.

Fig. 6.

Molecular Orbitals Involved in Key Transitions in the Calculated ECD Spectrum of the Lowest-Energy Conformer 2 of 5 at the mPW1PW91/6–311 + G (2d, p) Level in EtOH. Orbitals are Plotted with 0.03 e⁻/au3 Isovalues.

Compound 6 (Fig. 1) was isolated as colorless crystals (from MeOH). The molecular formula was established as C₁₉H₂₄O₃ with eight indices of hydrogen deficiency according to the HR-ESI-MS data (m/z 301.1802 ([M + H]⁺, calcd for C₁₉H₂₅O₃, 301.1804) and ¹³C NMR data. The ¹H NMR spectrum of 3 displayed signals of four methyls [δ_H 2.61 (3H, s, H-16), 1.18 (3H, s, H-18), 1.15 (3H, s, H-19), 1.31 (3H, s, H-20)] and two aromatic protons [δ_H 7.59 (1H, s, H-11) and 6.68 (1H, s, H-14)]. Moreover, analysis of the ¹³C NMR and HSQC spectra revealed the presence of four methyls (δ_C 26.6, 26.5, 25.0, 21.2), four methylenes (δ_C 37.7, 34.6, 31.1, 19.8), three methines (one at δ_C 50.6 and two aromatic carbons at δ_C 127.5, 117.4), six non-protonated carbons (δ_C 159.7, 145.4, 138.9, 118.6, 47.4, 36.9) and two carbonyl groups (δ_C 216.5, 203.9). Analysis of the 1D and 2D NMR spectroscopic data suggested 6 to be a sempervirane-type diterpenoid closely related to hispidanol A [29]. Compared with those of hispidanol A, the ¹³C NMR data of 6 exhibited an additional carbonyl group (δ_C 216.5), instead of a hydroxy group at C-3 in hispidanol A. The HMBC correlations (Fig. 2) from H₂–1 (δ_H 2.51), H₃–18 (δ_H 1.18) and H₃–19 (δ_H 1.15) to C-3 (δ_C 216.5) established the additional carbonyl group in 6.

Finally, the 2D structure and absolute configuration of 6 were further confirmed by the X-ray diffraction data analysis (Fig. 3). The final refinement of the Cu Kα data resulted in a small Flack parameter of −0.06 (12), allowing the assignment of the absolute configuration of 6. Therefore, the absolute configuration of 6 was defined as (5R, 10S). Therefore 6 was determined as (5R, 10S)-3-one-15-oxo-17-norsempervirol, and named as viroxocin G.

3.2. Assay of biological activity

3.2.1. Anti-inflammatory activity

Because of the strong cytotoxic against RAW 267.4 cell lines of compounds 1, 3 and 4, only compounds 2, 5 and 7–10 were assessed to determine their anti-inflammatory effects on NO production in LPS-induced RAW 267.4 cells. Compounds 2, 5, 7, 8 and 10 exhibited promising anti-inflammatory activities, and their inhibition rates on NO production were more than 60% at 10 μM. Compounds 9 exhibited medium anti-inflammatory activities, and the inhibition rates on NO production were 57.63% at 10 μM (as shown in Table 4).

Table 4.

The effects of compounds 1–10 (10 μM) and aspirin (5 μM) on LPS-induced RAW264.7 Cells.

Group	Cell viability ($\bar{x}\pm \mathit{sd}$, n = 6)	Inhibition%
Control	1.000 ± 0.002	–
LPSa	0.961 ± 0.004	–
1	0.058 ± 0.004	–
2	1.022 ± 0.001	62.84
3	0.180 ± 0.032	–
4	0.304 ± 0.039	–
5	0.962 ± 0.001	60.59
6	0.144 ± 0.025	–
7	1.342 ± 0.127	65.38
8	1.003 ± 0.170	73.44
9	0.884 ± 0.064	57.63
10	0.914 ± 0.060	63.39
Aspirinb	0.878 ± 0.005	34.97

^aDose of LPS (1 μg/mL).

^bPositive control.

3.2.2. Cytotoxicity activity

Because of the small amount of 1 and 3–6, only other five compounds were used for in vitro cytotoxicity evaluation by the CCK8 assay (Table 5). Compound 7 displayed a medium cytotoxic effect against the 769P cells. At a concentration of 20 μM, the inhibition of compound 7 for 796P cells was 52.66%. Compound 2, 8 and 9 exhibited weak cytotoxicity. At a concentration of 20 μM, the inhibition rates of compound 2, 8 and 9 for 796P cells were 29.02%, 16.15% and 14.27%. It is the first time to report the cytotoxicity of diterpenoids 2, and 7–9. Compound 10 showed no cytotoxicity to 769P cells at 20 μM.

Table 5.

In vitro cytotoxic activity of compounds 2 and 7–10 (20 μM) against human renal cell carcinoma 769P.

Group	Cell viability ($\bar{x}\pm \mathit{sd}$, n = 6)	Inhibition (%)
Control	1 ± 0.076	–
2	0.710 ± 0.100	29.02
7	0.473 ± 0.021	52.66
8	0.838 ± 0.089	16.15
9	0.857 ± 0.077	14.27
10	1.008 ± 0.083	–

4. Conclusion

In a word, this work described six new diterpenoids (1–6) and four known compounds (7–10), which were isolated from the EtOH extract of I. serra. At a concentration of 10 μM, compounds 2, 5, 7, 8 and 10 exhibited promising inhibition rates on NO production, showing a good anti-inflammatory potential, while compound 7 showed a medium cytotoxicity against 769P cells at 20 μM. These findings enriched the structural diversity and bioactivities of secondary metabolites from I. serra.

Appendix A. supplementary data

¹H NMR, ¹³C NMR, HMBC, HSQC, ¹H—¹H COSY, NOESY, X-ray, ECD, IR and HR-ESI-MS data for compounds 1–6. (PDF).

Data availability

Data will be made available on request.