Icetexane Diterpenoids from Perovskia atriplicifolia

Abstract

Five new icetexane diterpenoids, namely, perovskatones B–D (1, 3, 4), 1α-hydroxybrussonol (2), and 1α-hydroxypisiferanol (5), were isolated from Perovskia atriplicifolia, together with a new natural product (6) and two known compounds, przewalskin E (7) and brussonol (8). The structures of the new compounds were elucidated by detailed analyses of their MS, IR, 1D, and 2D NMR data. Compounds 1–8 were assayed for their inhibitory hepatitis B virus activities in the HepG 2.2.15 cell line. The results suggested that compounds 1 and 2 possessed noticeable anti-hepatitis B virus activity in vitro, suppressing the replication of hepatitis B virus DNA with selectivity index values of 154.3 and 137.7, respectively.

Key words

Perovskia atriplicifolia - Lamiaceae - icetexane diterpenoids - anti-HBV activity

Introduction

Perovskia atriplicifolia Benth., a perennial shrub belonging to the Lamiaceae family [1], is a folk medicine long been used as a parasiticide and analgesic in Tibet, China. Only a few investigations on this plant have been reported before [2], [3]. During our search of anti-hepatitis B virus (HBV) active constituents from a natural source, the 90 % ethanol extract of P. atriplicifolia was found to possess superior inhibitory HBV activity in vitro. A previous study on this plant had led to the isolation of ten compounds, of which perovskatone A was a novel C₂₃ terpenoid [4]. As a further phytochemical investigation of this folk medicine, eight icetexane diterpenoids were obtained from the 90 % ethanol extract. Based on MS, IR, 1D and 2D NMR data analyses, as well as comparison with the literature, the structures of compounds 1–8 were elucidated. Compounds 1–5 ([Fig. 1]) were new diterpenoids featuring a icetexane skeleton. Compound 6 [5], [6] was a new natural product. Compounds przewalskin E (7) [7] and brussonol (8) [8] were isolated from this plant for the first time. All isolates were evaluated for their inhibitory HBV potency in vitro. Compounds 1 and 2 showed noticeable anti-HBV activity in vitro in the HepG 2.2.15 cell line, suppressing the replication of HBV DNA with selectivity index (SI) values of 154.3 and 137.7, respectively. Herein we described the isolation, structural elucidation, and anti-HBV activities of compounds 1–8.

Fig. 1 Structures of compounds 1–8.

Results and Discussion

Compound 1 was obtained as a yellowish powder. It was assigned the molecular formula C₂₀H₂₆O₄ deduced by the positive HRESIMS at m/z 353.1721 (calcd. 353.1729 for C₂₀H₂₆O₄Na⁺). The IR spectrum exhibited characteristic absorption bands due to the ortho-quinone carbonyl groups at 1723 and 1680 cm⁻¹. In the ¹H NMR ([Table 1]) spectrum, an olefinic proton signal at δ _H 6.73 (1H, s, H-14), an oxygenated methine at δ _H 4.63 (1H, dd, J = 5.6, 2.0 Hz, H-7), and two singlet methyls at δ _H 0.98, 0.90 (each 3H, Me-19, 18) were observed, together with one isopropyl group at δ _H 2.89 (1H, sept. J = 6.8 Hz), 1.14 (3H, d, J = 6.8 Hz), and 1.12 (3H, d, J = 7.2 Hz). In addition, a pair of characteristic proton signals with a large coupling constant assignable to H-20 appeared at δ _H 2.94 (1H, d, J = 18.4 Hz) and 2.05 (1H, d, J = 18.4 Hz), suggesting that compound 1 possessed an icetexane-type diterpenoid skeleton, with a typical 6/7 carbon ring (rings A and B) [9], [10]. The ¹³C NMR spectrum ([Table 2]) displayed 20 carbon resonances, of which two conjugated carbonyl signals at δ _C 181.8 (s, C-12), 181.0 (s, C-11) were presented. A comparison of the NMR data of compound 1 with those of przewalskin E (7) [7] showed a high similarity except that compound 1 contained one more oxygenated methine than przewalskin E (7). Considering that compound 1 had 16 mass units more than przewalskin E, the occurrence of one more hydroxyl in compound 1 was proposed. In the HMBC spectrum ([Fig. 2]), long-range correlations from H-1 (δ _H 4.18) to C-10 (δ _C 84.7)and from H-2 (δ _H 1.52–1.61, 1.93–2.02), H-5 (δ _H 1.93–2.02), and H-20 (δ _H 2.05, 2.94) to C-1 were exhibited, suggesting that the additional hydroxyl should be linked at C-1. The ¹H-¹H COSY ([Fig. 2]) correlation between H-1 and H-2 further supported the above deduction. To establish the hydroxyl orientation at C-1, a ROESY experiment was conducted. The ROESY correlations H-1/H-19, H-1/H _β -2 demonstrated that OH-1 should be in the α-linkage. The other HSQC, HMBC, ¹H-¹H COSY ([Fig. 2]), and ROESY ([Fig. 3]) correlations allowed for the full assignments of proton and carbon signals. Consequently, the structure of compound 1 was determined as shown in [Fig. 1], and the compound was named perovskatone B (1).

Fig. 2 Key HMBC (→) and ¹H-¹H COSY (—) correlations of compounds 1–5.

Fig. 3 Key ROESY (←→) correlations of compounds 1–3, and 5.

Table 1 ¹H NMR (400 MHz, CD₃OD) data of compounds 1–6. δ in ppm, J in Hz.

Pos.	1	2	3	4	5	6
1α β	4.18 (1H, dd, 8.4, 6.0)	4.17 (1H, dd, 7.6, 6.0)	3.61 (1H, dd, 7.2, 6.8)	1.77 (1H, m) 1.50–1.57 (1H, m)	3.54 (1H, m)	1.75–1.83 (1H, m) 1.50–1.56 (1H, m)
2α β	1.52–1.61 (1H,overlapped) 1.93–2.02 (1H,overlapped)	1.58–1.65 (1H, m) 2.02–2.05 (1H, m)	1.44–1.50 (1H, m) 2.07–2.15 (1H, m)	1.82–1.86, (1H, overlapped) 1.36–1.41 (1H, overlapped)	1.45–1.51 (1H, m) 2.11–2.18 (1H, m)	1.82–1.88 (1H, m) 1.37–1.41(1H, overlapped)
3α β	1.52–1.61 (2H,overlapped)	1.46–1.55 (2H,m)	1.75–1.79 (1H, m) 1.07–1.12 (1H, m)	1.36–1.41 (1H, overlapped) 1.29–1.31 (1H, m)	1.79 (1H, td, 14.0, 3.6) 1.09 (1H, dt, 13.2, 3.2)	1.37–1.41 (1H, overlapped) 1.29–1.33 (1H, m)
5α	1.93–2.02 (1H, overlapped)	1.92 (1H, m)	1.69 (1H, dd, 12.2, 2.6)	1.33 (1H, overlapped)	1.68 (1H, dd, 12.4, 2.8)	1.33 (1H, overlapped)
6α β	2.10–2.13 (2H, m)	2.09–2.15 (1H, m) 1.82–1.85 (1H, m)	1.79–1.83 (1H, m) 1.28–1.34 (1H, m)	1.82–1.86 (1H, overlapped) 1.33 (1H, overlapped)	1.93 (1H, m) 1.34 (1H, m)	1.82–1.88 (1H, overlapped) 1.33 (1H, overlapped)
7α β	4.63 (1H, dd, 5.6, 2.0)	4.86 (1H, brd, 6.4)	3.30 (1H, m) 2.03 (1H, m)	3.33–3.37 (1H, m) 2.01–2.07(1H, m)	2.68 (2H, m)	2.65 (1H, m) 2.54 (1H, m)
11	–	–	–	–	6.53 (1H, s)	–
14	6.73 (1H, s)	6.42 (1H, s)	–	–	6.85 (1H, s)	6.77 (1H, s)
15	2.89 (1H, sept, 6.8)	3.23 (1H, sept, 6.8)	3.22 (1H, sept. 7.2)	3.22 (1H, sept. 7.2)	3.21 (1H, sept. 7.2)	2.87 (1H, sep, 6.8)
16	1.14 (3H, d, 6.8)	1.18 (3H, d, 7.2)	1.22 (3H, d, 7.2)	1.21 (3H, d, 6.8)	1.18 (3H, d, 6.8)	1.13 (3H, d, 7.2)
17	1.12 (3H, d, 7.2)	1.18 (3H, d, 7.2)	1.21 (3H, d, 7.2)	1.21 (3H, d, 6.8)	1.18 (3H, d, 6.8)	1.12 (3H, d, 7.2)
18α	0.90 (3H, s)	0.87 (3H, s)	0.92 (3H, s)	0.91 (3H, s)	0.93 (3H, s)	0.92 (3H, s)
19β	0.98 (3H, s)	0.98 (3H, s)	0.90 (3H, s)	0.89 (3H, s)	0.88 (3H, s)	0.92 (3H, s)
20α β	2.05 (1H, d, 18.4) 2.94 (1H, d, 18.4)	2.36 (1H, d, 16.8) 3.19 (1H, d, 16.8)	3.13 (1H, d, 14.4) 2.58 (1H, d, 14.4)	3.13 (1H, d, 14.4) 2.28 (1H, d, 14.4)	3.22 (1H, d, 14.4) 2.42 (1H, d, 14.4)	3.01 (1H, d, 14.4) 2.19 (1H, d, 14.4)

Table 2 ¹³C NMR (100 MHz) data of compounds 1–6 in CD₃OD.

	δ C (mult.)
Position	1	2	3	4	5	6
1	71.1 (d)	71.6 (d)	77.0 (d)	42.8 (t)	77.9 (d)	42.6 (t)
2	26.8 (t)	27.0 (t)	26.4 (t)	19.4 (t)	26.8 (t)	19.5 (t)
3	36.1 (t)	36.0 (t)	35.9 (t)	43.6 (t)	36.2 (t)	43.6 (t)
4	33.7 (s)	34.5 (s)	34.8 (s)	35.3 (s)	34.9 (s)	35.3 (s)
5	54.4 (d)	53.9 (d)	53.1 (d)	59.4 (d)	53.1 (s)	59.4 (d)
6	37.4 (t)	39.2 (t)	20.7 (t)	21.5 (t)	24.7 (t)	21.8 (t)
7	76.4 (d)	78.0 (d)	26.2 (t)	26.5 (t)	36.5 (t)	37.3 (d)
8	154.7 (s)	135.0 (s)	149.6 (s)	149.6 (s)	136.2 (s)	156.3 (s)
9	130.9 (s)	118.4 (s)	139.6 (s)	139.1 (s)	136.5 (s)	135.6 (s)
10	84.7 (s)	84.6 (s)	73.5 (s)	71.5 (s)	74.2 (s)	71.7 (s)
11	181.0 (s)	144.5 (s)	185.0 (s)	185.0 (s)	119.9 (d)	181.1 (s)
12	181.8 (s)	142.4 (s)	153.5 (s)	153.4 (s)	153.3 (s)	181.4 (s)
13	148.6 (s)	133.8 (s)	125.5 (s)	125.5 (s)	133.7 (s)	147.6 (s)
14	134.9 (d)	112.8 (d)	187.9 (s)	188.1 (s)	126.9 (d)	140.4 (d)
15	28.6 (d)	27.9 (d)	25.5 (d)	25.6 (d)	27.8 (d)	28.6 (d)
16	21.8 (q)	23.3 (q)	20.3 (q)	20.4 (q)	23.3 (q)	21.7 (q)
17	21.8 (q)	23.3 (q)	20.4 (q)	20.4 (q)	23.4 (q)	22.1 (q)
18	33.0 (q)	33.1 (q)	32.4 (q)	32.6 (q)	32.8 (q)	32.6 (q)
19	25.0 (q)	24.8 (q)	22.3 (q)	22.1 (q)	22.5 (q)	21.9 (q)
20	31.8 (t)	33.0 (t)	38.1 (t)	40.8 (t)	49.1 (t)	40.9 (t)

Compound 2 was obtained as a yellowish powder and had a molecular formula of C₂₀H₂₈O₄ deduced by the HRESIMS at m/z 355.1880 (calcd. 355.1885 for C₂₀H₂₈O₄Na⁺). The characteristic H-20 signals at δ _H 3.19 (1H, d, J = 16.8 Hz), 2.36 (1H, d, J = 16.8 Hz) in ¹H NMR spectrum suggested that compound 2 was also an icetexane diterpenoid. Comparison of its ¹³C NMR data ([Table 2]) with those of compound 1 demonstrated that both compounds had similar A and B rings. However, there was no carbonyl signal in the ¹³C NMR spectrum of compound 2. In addition, the carbon signals ascribable to ring C in compound 2 were identical to those of brussonol (8) [8], implying that both compounds had the same ring C. The above deduction was verified by the HMBC correlations ([Fig. 2]) from the oxygenated H-1 (δ _H 4.17) to C-2 (δ _c 27.0), C-3 (δ _c 36.0), C-5 (δ _c 53.9), C-10 (δ _c 84.6), and C-20 (δ _c 33.0), from H-20 to C-11, C-8, and from H-14 to C-9, C-12, C-15. The α-orientation of hydroxyl at C-1 was determined by the H-1/Me-19 correlation in the ROESY plot ([Fig. 3]). Accordingly, the structure of compound 2 was characterized as shown in [Fig. 1], and the compound was named 1α-hydroxybrussonol (2).

Compound 3 was obtained as a yellow powder and assigned the molecular formula C₂₀H₂₈O₅, in agreement with the HRESIMS at m/z 371.1825 (calcd. 371.1834 for C₂₀H₂₈O₅Na⁺). Compound 3 was deduced to have the icetexane diterpenoid skeleton by the typical H-20 signals at δ _H 3.13 (1H, d, J = 14.4 Hz), 2.58 (1H, d, J = 14.4 Hz) in the ¹H NMR spectrum ([Table 1]). The characteristic carbon shifts at δ _C 185.0 (C-11) and 187.9 (C-14) suggested that compound 3 contained a para-quinone ring C [11]. This was supported by the HMBC correlations from H-20 (δ _H 3.13, 2.58) to C-11 (δ _C 185.0), C-8 (δ _C 149.6), and C-9 (δ _C 139.6), as well as the correlations from H-7 (δ _H 3.30, 2.03) to C-14 (δ _C 187.9), C-8, and C-9. Additionally, the ¹H-¹H COSY correlation for H-1/H-2 ([Fig. 2]), together with the long-range HMBC correlations from H-1 to C-5, C-10, C-20, from H-5 to C-1, C-10, C-20, and from H-6 to C-10 indicated the presence of hydroxyls at C-1 and C-10. Biogenetically, the configurations of C-5 (α-oriented H-5) and C-10 (β-oriented OH-10) in the icetexane diterpenoids from the Perovskia genus were 5S and 10S, which has been extensively discussed and authenticated in some documents [10], [12], [13], [14], [15]. The α-orientation of H-5 was verified by the ROESY ([Fig. 3]) correlation of H-5/Me-18 (α-orientation with a carbon chemical shift larger than 30.0 ppm). The α-linked hydroxyl at C-1 could be established by the NOE correlation of H-1/Me-19 in the ROESY experiment ([Fig. 3]). Finally, the structure of compound 3 was determined as shown in [Fig. 1], and the compound was named perovskatone C (3).

Compound 4 was obtained as a yellow powder. It had the molecular formula C₂₀H₂₈O₄ deduced by the HRESIMS at m/z 355.1877 [M + Na]⁺ (calcd. for C₂₀H₂₈O₄Na⁺: 355.1885). The 1D and 2D NMR data of compound 4 were essentially identical to those of compound 3, implying that compounds 3 and 4 had a similar structure. Compound 4 differed from 3 mainly in ring A where there was no hydroxyl at C-1. This was ascertained by the ¹H-¹H COSY ([Fig. 2]) of H-1 (δ _H 1.77, 1.50–1.57)/H-2 (δ _H 1.82–1.86, 1.36–1.41) and HMBC correlations ([Fig. 2]) from H-1 to C-5, C-10, C-20. Consequently, the structure of compound 4 was elucidated as shown in [Fig. 1], and the compound was named perovskatone D (4).

Compound 5 was obtained as a yellowish powder. It was assigned the molecular formula C₂₀H₃₀O₃ by the positive HRESIMS at m/z 341.2083 [M + Na]⁺ (calcd. 341.2092 for C₂₀H₃₀O₃Na⁺). Detailed analyses of the 1D and 2D NMR data revealed that compound 5 was structurally similar to 1β-hydroxypisiferanol [16]. However, compound 5 included an α-oriented hydroxyl at C-1, which was definitely established by the ROESY correlations for H-1/H-19 and H-1/H _β -20 ([Fig. 3]). The other ¹H-¹H COSY and HMBC correlations ([Fig. 2]) further confirmed this deduction. Lastly, compound 5 was deduced as 1α-hydroxypisiferanol (5).

Compound 6, which had been previously synthesized by Moujir [5] and Majetich [6], was obtained as a new natural product in our experiment. The full assignments of proton and carbon signals ([Table 1] and [2]) were firstly performed based on the extensive analyses of 1D and 2D NMR data (see Supporting Information). The known compounds przewalskin E (7) and brussonol (8), whose structures were determined by comparing the NMR data with those in the literature [7], [8], were obtained from this plant for the first time.

All the isolates were tested for their anti-HBV activities in the HepG 2.2.15 cell line according to the method described in our previous report [17]. Results are summarized in [Table 3]. It was concluded that compounds 1, 2, 4, and 8 possessed moderate anti-HBV activity in vitro, suppressing the secretion of the hepatitis B surface antigen (HBsAg) with SI values ranging from 2.06 to 4.83. Compound 2 could also inhibit the secretion of HBsAg with an SI value of 2.0. Compounds 5–7 showed no anti-HBV activity in vitro. In addition, compounds 1 and 2 exhibited superior inhibitory HBV DNA replication activity with SI values of 154.3 and 137.7, respectively.

Table 3 In vitro anti-hepatitis B virus activities of compounds 1–8.

		HBsAg		HBeAg		Inhibiting HBV DNA replication
Compounds	CC50 (mM)a	IC50 (mM)a	SIb	IC50 (mM)	SI	IC50 (µM)	SI
All values are the mean of two independent experiments; a IC50: 50 % inhibitory concentration; CC50: 50 % cytotoxic concentration; b SI = CC50/IC50; c 3TC: Lamivudine, positive control; NT: not been tested for their trace amount; NO: IC50 value was not obtained at the highest tested concentration
1	> 2.13	1.03	> 2.06	1.97	> 1.08	13.8	154.3
2	2.85	0.59	4.83	1.42	2.00	20.7	137.7
3	2.13	1.54	1.38	3.67	–	NO	–
4	2.78	0.92	3.02	4.01	–	NT	NT
5	1.85	2.45	–	3.82	–	NT	NT
6	> 2.13	4.08	–	3.68	–	NO	–
7	1.44	2.23	–	1.72	–	NO	–
8	> 3.54	1.39	> 2.55	4.72	–	NO	–
3TCc	29.96	23.50	1.27	28.19	1.06	1.12	26 750.0

In conclusion, eight icetexane diterpenoids were isolated from the ethanol extract of P. atriplicifolia, of which compounds 1–5 were new ones and compound 6 was a new natural product. An in vitro anti-HBV bioassay suggested that compounds 1 and 2 could moderately inhibit HBV DNA replication in HepG 2.2.15 cells. To the best of our knowledge, icetexane-type diterpenoids were mainly found from the plants of Salvia and Chamaecyparis genera [12]. Our research illustrated that icetexane diterpenoids were also abundant in the Perovskia genus, which could be helpful for exploiting a new use of this medicinal plant.

Material and Methods

General experimental procedures

Optical rotations were measured using a Horiba SEPA-300 high-sensitive polarimeter. UV spectra were obtained on a Shimadzu UV-2401A spectrophotometer. IR spectra were carried out on a Bio-Rad FTS-135 spectrometer with KBr pellets, ν in cm⁻¹. 1D and 2D NMR spectra were recorded on a Bruker AV-400 (¹H/¹³C, 400 MHz/100 MHz) spectrometer with tetramethylsilane (TMS) as the internal standard. Chemical shifts (δ) are expressed in ppm with TMS as the internal reference. HRESIMS was performed on a VG Autospec-3000 spectrometer. HPLC was performed on an Agilent 1260 liquid chromatograph equipped with a Venusil XBP C18 (10 × 250 mm, 5 µm) column. Column chromatography (CC) was performed using silica gel (200–300 mesh, Qingdao Marine Chemical, Inc.), Lichroprep Rp-18 gel (40–63 µm, Merck), Sephadex LH-20 (Sigma-Aldrich Co.), or MCI gel (75–150 µm, Mitsubishi Chemical Corporation). Column fractions were monitored by TLC, and the spots were visualized by heating the plates after spraying with 10 % H₂SO₄ in EtOH. The positive control lamivudine (3-TC, purity > 99 %) was purchased from GlaxoSmithKline (Suzhou) Co., Ltd.

Plant material

The whole plant of P. atriplicifolia was collected in Tibet in September 2010, and identified as P. atriplicifolia Benth. by Prof. Dr. Li-Gong Lei from Kunming Institute of Botany, Chinese Academy of Sciences. A voucher specimen (TSYJ-201093) was deposited in the Key Laboratory of Chemistry in Ethnic Medicinal Resources, State Ethnic Affairs Commission & Ministry of Education, School of Chemistry and Biotechnology, Yunnan Minzu University.

Extraction and isolation

The air-dried whole P. atriplicifolia plants (9.0 kg) were powdered and extracted with 90 % ethanol (54 L) under reflux three times, 2 h each time. After being concentrated in vacuo, the extract was suspended in water and successively partitioned with petroleum ether, chloroform, and n-BuOH to give petroleum ether (A), chloroform (B), n-BuOH (C), and aqueous (D) fractions. The petroleum ether (A) (260 g) extract was subjected to silica gel (2.0 kg, 200–300 mesh; 10 × 170 cm) chromatography column (CC) and eluted with gradient petroleum ether/acetone (100 : 0, 98 : 2, 95 : 5, 90 : 10, 80 : 20; v/v; each 5.0 L; 1000 mL/flask; flow rate: 5 mL/min) to afford six fractions (Frs.A.1–6). These fractions were subjected to CC on silica gel, MCI, Rp-18, Sephadex LH-20, or HPLC to afford compounds 1 (23 mg), 2 (44 mg), 3 (16 mg), 4 (4.8 mg), 5 (3.2 mg), 6 (40 mg), 7 (72 mg), and 8 (113 mg). All the isolates had a degree of purity greater than 93 %, determined by HPLC.

Perovskatone B (1): Yellowish powder; [α]_D ^18.2 + 24.0 (c 0.50, MeOH); UV (MeOH) λ max(log ε) 272 (1.59), 238 (1.45); IR (KBr) ν _max 3447, 2967, 1723, 1680, 1635, 1455, 1368, 1246, 1066, 999, 860 cm⁻¹; ¹H and ¹³C NMR data, see [Tables 1] and [2]; ESIMS (neg.): m/z 329 [M – H]⁻; ESIMS (pos.): m/z 353 [M + Na]⁺; HRESIMS (pos.): m/z 353.1721 (calcd. for C₂₀H₂₆O₄Na⁺: 353.1729).

1α-Hydroxybrussonol (2): Yellowish powder; [α]_D ^18.4 − 18.9 (c 0.10, MeOH); UV (MeOH) λ max(log ε) 275 (1.02), 231 (2.15); IR (KBr) ν _max 3545, 2967, 1643, 1635, 1596, 1501, 1453, 1370, 1266, 1128, 1066, 957, 803 cm⁻¹; ¹H and ¹³C NMR data, see [Tables 1] and [2]; ESIMS (neg.): m/z 331 [M – H]⁻; ESIMS (pos.): m/z 355 [M + Na]⁺; HRESIMS (pos.): m/z 355.1880 [M + Na]⁺ (calcd. for C₂₀H₂₈O₄Na⁺: 355.1885).

Perovskatone C (3): Yellow powder; [α]_D ^18.0 + 15.8 (c 0.29, MeOH); UV (MeOH) λ max(log ε) 284 (3.09); IR (KBr) ν _max 3448, 2960, 1660, 1643, 1456, 1376, 1128, 1065 cm⁻¹; ¹H and ¹³C NMR data, see [Tables 1] and [2]; ESIMS (neg.): m/z 347 [M – H]⁻; ESIMS (pos.): m/z 371 [M + Na]⁺; HRESIMS (pos.): m/z 371.1825 [M + Na]⁺ (calcd. for C₂₀H₂₈O₅Na⁺: 371.1834).

Perovskatone D (4): Yellow powder; [α]_D ^16.8 + 65.0 (c 0.60, MeOH); UV (MeOH) λ max(log ε) 284 (3.02); IR (KBr) ν _max 3448, 2961, 1660, 1641, 1450, 1384, 1126, 1070 cm⁻¹; ¹H and ¹³C NMR data, see [Tables 1] and [2]; ESIMS (neg.): m/z 331 [M – H]⁻; ESIMS (pos.): m/z 355 [M + Na]⁺; HRESIMS (pos.): m/z 355.1877 [M + Na]⁺ (calcd. for C₂₀H₂₈O₄Na⁺: 355.1885).

1α-Hydroxypisiferanol (5): Yellowish powder; [α]_D ^17.9 + 115.0 (c 0.31, MeOH); UV (MeOH) λ max(log ε) 272 (1.13), 230 (1.98); IR (KBr) ν _max 3530, 2941, 1606, 1498, 1451, 1375, 1126, 1064 cm⁻¹; ¹H and ¹³C NMR data, see [Tables 1] and [2]; ESIMS (neg.): m/z 317 [M – H]⁻; ESIMS (pos.): m/z 341 [M + Na]⁺; HRESIMS (pos.): m/z 341.2083 [M + Na]⁺ (calcd. for C₂₀H₃₀O₃Na⁺: 341.2092).

Anti-hepatitis B virus assays

The inhibitory potency for the secretion of HBsAg and hepatitis B envelope antigen (HBeAg) was conducted according to the method described in our previous report [17]. All the evaluated compounds were dissolved in DMSO (Gibco Invitrogen). The concentration of DMSO in the media was maintained at less than 2.5 µL/mL to ensure that it did not affect the growth of HepG 2.2.15 cells.

HBV DNA extraction was also conducted in HepG 2.2.15 cells. Briefly, the HepG 2.2.15 cells were seeded in 24-well culture plates at a density of 5 × 10⁵ cells/mL. After two days, the culture medium was replaced with fresh medium supplemented with (or without) the tested compounds; this was repeated every other day for an additional five days. Cells were collected, and total DNA was extracted with a commercial kit (Qiagen) following the manufacturerʼs instructions. The real-time PCR assay was used to detect the HBV DNA according to the literature reported [18]. An antiviral agent, 3 TC [lamivudine, GlaxoSmithKline (Suzhou) Co., Ltd.], was used as a positive control.

Cytotoxicity assays

The toxicities of the compounds were assayed by a modified MTT method [19]. In brief, the test samples were prepared at different concentrations. After seeding HepG 2.2.15 cells in 96-well microplate for 4 h, the samples (20 µL) were placed in each well and incubated for three days at 37 °C, then 0.1 mL MTT (400 µg/mL) was added for 4 h. After removal of the MTT medium, DMSO (100 µL/well) was added to the microplate for 10 min. The formazan crystals were dissolved, and the absorbance was measured on a microplate reader at 490 nm.

Supporting information

The detailed isolation procedure and 1D and 2D NMR spectra of the new compounds 1–6 are available as Supporting Information.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (NSFC No. 21 162 041) and the Innovation Team project from the Education Department of Yunnan Province (IRTSYN, 2014).

Conflict of Interest

The authors declare no conflict of interest.

• References

• 1 Editorial Committee of Flora of China, Chinese Academy of Sciences. Flora of China, Vol. 17. Beijing: Science Press; 1994: 222-223
• 2 Alia MS, Saleema M, Erianb AW. A new acylated steroid glucoside from Perovskia atriplicifolia . Fitoterapia 2001; 72: 712-714
• 3 Sefidkon F, Ahmadi L, Mirza M. Volatile components of Perovskia atriplicifolia Benth. J Essent Oil Res 1997; 9: 101-103
• 4 Jiang ZY, Huang CG, Xiong HB, Tian K, Liu WX, Hu QF, Wang HB, Yang GY, Huang XZ. Perovskatone A: a novel C₂₃ terpenoid from Perovskia atriplicifolia . Tetrahedron Lett 2013; 54: 3886-3888
• 5 Moujir L, Gutiérrez-Navarro AM. Bioactive diterpenoids isolated from Salvia mellifera . Phytother Res 1996; 10: 172-174
• 6 Majetich G, Zou G. Total synthesis of (−)-barbatusol, (+)-demethylsalvicanol, (−)-brussonol, and (+)-grandione. Org Lett 2008; 10: 81-83
• 7 Xu G, Peng LY, Tu L, Li XL, Zhao Y, Zhang PT, Zhao QS. Three new diterpenoids from Salvia przewalskii Maxim. Helv Chem Acta 2009; 92: 409-413
• 8 Fraga BM, Díaz CE, Guadaňo A, González-Coloma A. Diterpenes from Salvia broussonetii transformed roots and their insecticidal activity. J Agric Food Chem 2005; 53: 5200-5206
• 9 Fraga BM, González AG, Herrera JR, Luis JG, Ravelo AG. Diterpenes from the roots of Salvia canariensis . Phytochemistry 1985; 25: 269-271
• 10 Gonzales AG, Andres LS, Luis JG, Brito I, Rodrfguez ML. Diterpenes from Salvia mellifera . Phytochemistry 1991; 30: 4067-4070
• 11 Wu SH, Zhang HJ, Lin ZW, Sun HD. Terpenoids from Isodon grandifolia var. atuntzensis . Phytochemistry 1993; 34: 1176-1178
• 12 Simmons EM, Sarpong R. Structure, biosynthetic relationships and chemical synthesis of the icetexane diterpenoids. Nat Prod Rep 2009; 26: 1195-1217
• 13 Aoyagi Y, Takahashi Y, Satake Y, Fukaya H, Takeya K, Aiyama R, Matsuzaki T, Hashimoto S, Shiina T, Kurihara T. Biomimetic synthesis of grandione from demethylsalvicanol via hetero-Diels-Alder type dimerization and structure revision of grandione. Tetrahedron Lett 2005; 46: 7885-7887
• 14 Simmons EM, Yen JR, Sarpong R. Reconciling icetexane biosynthetic connections with their chemical synthesis: total synthesis of (±) − 5, 6- dihydro − 6α-hydroxysalviasperanol, (±)- brussonol, and (±) – abrotanone. Org Lett 2007; 9: 2705-2708
• 15 Khaliq S, Volk FJ, Frahm AW. Phytochemical investigation of Perovskia abrotanoides . Planta Med 2007; 73: 77-83
• 16 Lin TC, Fang JM, Cheng YS. Terpenes and lignans from leaves of Chamaecyparis formosensis . Phytochemistry 1999; 51: 793-801
• 17 Jiang ZY, Zhang XM, Zhang FX, Liu N, Zhao F, Zhou J, Chen JJ. A new triterpene and anti-hepatitis B virus active compound from Alisma orientalis . Planta Med 2006; 72: 951-954
• 18 Zhang GL, Li YX, Zheng SQ, Liu M, Li X, Tang H. Suppression of hepatitis B virus replication by microRNA-199a-3p and microRNA-210. Antiviral Res 2010; 88: 169-175
• 19 Mosmann T. Rapid colorimetric assay for cellular growth and survival: application to proliferation and cytotoxicity assays. J Immunol Methods 1983; 65: 55-63