Four Novel Dammarane-Type Triterpenoids from Pearl Knots of Panax ginseng Meyer cv. Silvatica

Panax ginseng Meyer cv. Silvatica (PGS), which is also known as “Lin-Xia-Shan-Shen” or “Zi-Hai” in China, is grown in forests and mountains by broadcasting the seeds of ginseng and is harvested at the cultivation age of 15–20 years. In this study, four new dammarane-type triterpenoids, ginsengenin-S1 (1), ginsengenin-S2 (2), ginsenoside-S3 (3), ginsenoside-S4 (4), along with one known compound were isolated from pearl knots of PGS. Ginsengenin-S2 significantly alleviated oxidative damage when A549 cells were exposed to cigarette smoke (CS) extract. In addition, ginsengenin-S2 could inhibit the CS-induced inflammatory reaction in A549 cells. Protective effects of ginsengenin-S2 against CS-mediated oxidative stress and the inflammatory response in A549 cells may involve the Nrf2 and HDAC2 pathways.


Introduction
Ginseng, the root of Panax ginseng Meyer, a kind of widely used folk medicine in many countries including in China for thousands of years, belongs to the Araliaceae family and mainly grows in Korea, China and Japan [1]. Panax ginseng Meyer cv. Silvatica (PGS), which is also called "Lin-Xia-Shan-Shen" or "Zi-Hai" in China according to the Chinese Pharmacopoeia, is grown naturally in forests and mountains by broadcasting seeds of ginseng with no human intervention and is harvested at the cultivation age of 15-20 years [2]. In contrast with garden cultivated ginseng (GCG), which is collected after 4-6 years of artificial cultivation in the garden, PGS can imitate the growing environment of wild ginseng [3]. There are obvious verrucous warts on the fibrous roots of PGS, which are called pearl knots (Zhen-zhu-Ge-da) in China. It was reported that the biological nature of pearl knots is the foundation of the seasonal absorbing root of ginseng, in other words, more pearl knots on the PGS indicate longer growth time, while there are scarcely any pearl knots on the GCG [2]. According to the previous report, PGS is of better quality than GCG and exhibited greater anticancer activity than GCG in lung cancer and other cancers [4,5]. However, chemical research of the pearl knots of PGS remains poorly understood. Ginsenosides, the main active principle of Panax ginseng, possesses various pharmacological activities including antioxidant [6], immunomodulatory [7], neuroprotective [8], anti-depressive effects [9,10] and renal protective effect [11].
Chronic obstructive pulmonary disease (COPD) is "the third killer" in the worldwide public health problem characterized by progressive airflow limitation, progressive lung inflammation and increases in the levels of some inflammatory mediators such as MMP-9, TNF-α and IL-8 [12]. Cigarette to CS can drive the severe inflammation of neutrophils and macrophages to the lung and induce apoptosis of epithelial and endothelial cells [13]. It was reported that ginsenoside Rg1 could ameliorate CS-induced airway fibrosis and pulmonary epithelial-mesenchymal transition [14,15], while ginsenoside Rb3 exerts protective properties against CS extract-induced cell injury in fibroblasts and epithelial cells [16].
In this study, four new triterpenoids and one known triterpenoid were isolated from the pearl knots of PGS and we used the CS-stimulated human lung epithelial cells to test their antioxidant activity. Among these five compounds, ginsengenin-S2 exhibited a significant protective effect against CS-induced oxidative stress; then, its protective effects and potential mechanisms against the CS stimulation were preliminarily investigated.

Structure Elucidation of Compounds 1-5
Compound 1, white powder: its molecular formula was established as C30H52O4 by HRESIMS. The 1 H-NMR spectrum of 1 showed eight methyl singlet signals at δH 0.81, 1.01, 1.11, 1.29, 1.39, 1.62, 1.65, 1.88. The 13 C-NMR spectrum of 1 showed signals of eight methyl carbons, six quaternary, eight methane and eight methylene among which two carbon signals (δ 126.6 and 131.1) showed the presence of a C=C bond. The chemical shifts of 1 resembled those of protopanaxatriol except the chemical shifts on A ring [17]. The proton and carbon signals of 1 were fully assigned on the basis of 2D-NMR spectra (Table 1). According to the literature, as compared with its 3β-epimers, the C-1~C-5 chemical shifts of 3α-dammarane-type triperpenes are in relatively higher nuclear magnetic fields, while carbon chemical shifts of C-28~C-29 are in relatively lower nuclear magnetic fields. For example, the C-1~C-5 and C-28~C- 29 Figure 1). Therefore, 1 was elucidated as dammar-24-ene-3α,6α,12β,20S-tetraol (ginsengenin-S1, Figure 2).  Compound 2, white powder: its molecular formula was established as C 30 H 52 O 4 by HRESIMS. The NMR data of 2 ( Table 1) resembled those of 1. The 3α configuration of 2 was determined by its NOEs the same way as 1. Compared with 1, the carbon NMR data of 2 showed the absence of a methyl carbon signal and an additional signal of methylene (δ C 39.3) and a pair of typical signals of a terminal double bond (δ C 110.4 and 146.8) in place of (δ C 126.6 and 131.1), therefore, briefly, the obvious differences between 1 and 2 were the carbon chemical shifts of C-24~C-27. The 1 H-NMR data of 2 also showed the absence of a methyl hydrogen signal and two additional hydrogen signals of the terminal double bond (δ H 4.80 s, 4.84 s). Thus, 2 was deduced to be dammar-25-ene-3α,6α,12β,20S-tetraol (ginsengenin-S2, Figure 2). Compound 2, white powder: its molecular formula was established as C30H52O4 by HRESIMS. The NMR data of 2 ( Table 1) resembled those of 1. The 3α configuration of 2 was determined by its NOEs the same way as 1. Compared with 1, the carbon NMR data of 2 showed the absence of a methyl carbon signal and an additional signal of methylene (δC 39.3) and a pair of typical signals of a terminal double bond (δC 110.4 and 146.8) in place of (δC 126.6 and 131.1), therefore, briefly, the obvious differences between 1 and 2 were the carbon chemical shifts of C-24~C-27. The 1 H-NMR data of 2 also showed the absence of a methyl hydrogen signal and two additional hydrogen signals of the terminal double bond (δH 4.80 s, 4.84 s). Thus, 2 was deduced to be dammar-25-ene-3α,6α,12β,20Stetraol (ginsengenin-S2, Figure 2).    Delta is ppm of chemical shifts which are reported in parts per million (δ), and coupling constants (J) are expressed in Hertz.  Table 2). The configuration of the C-24 (δ C 80.2) of 4 was indicated to be 24R according to the C-24 configuration analysis of 3 and 5, thus the structure of 4 was inferred as Figure 2). The MTT result showed that cell viability of A549 cells was substantially affected by CSE at a concentration 40% ( Figure 3A). Therefore, 30% CSE was used as stimulation in subsequent experiments. As shown in Figure 3B, the cell viability of A549 cells was not affected by 1-5. Then we tested the antioxidant effects of 1-5 on CS-stimulated A549 cells. Delta is ppm of chemical shifts which are reported in parts per million (δ), and coupling constants (J) are expressed in Hertz.

Cytotoxicity of Cigarette Smoke Extract (CSE) and Compounds 1-5 on A549 Cells
The MTT result showed that cell viability of A549 cells was substantially affected by CSE at a concentration ≧ 40% ( Figure 3A). Therefore, 30% CSE was used as stimulation in subsequent experiments. As shown in Figure 3B, the cell viability of A549 cells was not affected by 1-5. Then we tested the antioxidant effects of 1-5 on CS-stimulated A549 cells.

Antioxidant Activity of Compounds 1-5
As shown in Figure 4, CS exposure significantly increased the intracellular malondialdehyde (MDA) level in human lung epithelial cells. Among these five compounds, ginsengenin-S2 induced a dose-dependent reduction of MDA level. . Effect of compounds 1-5 on the intracellular malondialdehyde (MDA) level in cigarette smoke (CS)-exposed A549 cells. All data were expressed as mean ± S.D., n = 6. ** p < 0.01, compared with control; # p < 0.05, ## p < 0.01, compared with the CSE group.
Tobacco smoke is involved in the increased synthesis of reactive oxygen species (ROS) [27]. In our study, CS exposure induced significantly higher ROS level in A549 cells than the control group ( Figure 5A). Interestingly, ginsengenin-S2 down-regulated the ROS level in CS exposed-A549 cells.

Antioxidant Activity of Compounds 1-5
As shown in Figure 4, CS exposure significantly increased the intracellular malondialdehyde (MDA) level in human lung epithelial cells. Among these five compounds, ginsengenin-S2 induced a dose-dependent reduction of MDA level.
Delta is ppm of chemical shifts which are reported in parts per million (δ), and coupling constants (J) are expressed in Hertz.

Cytotoxicity of Cigarette Smoke Extract (CSE) and Compounds 1-5 on A549 Cells
The MTT result showed that cell viability of A549 cells was substantially affected by CSE at a concentration ≧ 40% ( Figure 3A). Therefore, 30% CSE was used as stimulation in subsequent experiments. As shown in Figure 3B, the cell viability of A549 cells was not affected by 1-5. Then we tested the antioxidant effects of 1-5 on CS-stimulated A549 cells.

Antioxidant Activity of Compounds 1-5
As shown in Figure 4, CS exposure significantly increased the intracellular malondialdehyde (MDA) level in human lung epithelial cells. Among these five compounds, ginsengenin-S2 induced a dose-dependent reduction of MDA level. . Effect of compounds 1-5 on the intracellular malondialdehyde (MDA) level in cigarette smoke (CS)-exposed A549 cells. All data were expressed as mean ± S.D., n = 6. ** p < 0.01, compared with control; # p < 0.05, ## p < 0.01, compared with the CSE group.
Tobacco smoke is involved in the increased synthesis of reactive oxygen species (ROS) [27]. In our study, CS exposure induced significantly higher ROS level in A549 cells than the control group ( Figure 5A). Interestingly, ginsengenin-S2 down-regulated the ROS level in CS exposed-A549 cells. . Effect of compounds 1-5 on the intracellular malondialdehyde (MDA) level in cigarette smoke (CS)-exposed A549 cells. All data were expressed as mean ± S.D., n = 6. ** p < 0.01, compared with control; # p < 0.05, ## p < 0.01, compared with the CSE group.
Tobacco smoke is involved in the increased synthesis of reactive oxygen species (ROS) [27]. In our study, CS exposure induced significantly higher ROS level in A549 cells than the control group ( Figure 5A). Interestingly, ginsengenin-S2 down-regulated the ROS level in CS exposed-A549 cells. Then, we investigated the role of ginsengenin-S2 on the activities of two first line defense antioxidants-superoxide dismutase (SOD) and glutathione (GSH) in CS-exposed human lung epithelial cells. As shown in Figure 5B,C, the CS challenge caused the decrease of intracellular GSH content and SOD activity (p < 0.01), with a parallel of increased MDA level which demonstrates that the CS challenge induced oxidative stress in the in vitro cell model. Further, we observed that pretreatment with ginsengenin-S2 could attenuate CS-induced oxidative stress by clearly up-regulating the content of SOD and GSH to a relatively basal level.
Then, we investigated the role of ginsengenin-S2 on the activities of two first line defense antioxidants-superoxide dismutase (SOD) and glutathione (GSH) in CS-exposed human lung epithelial cells. As shown in Figure 5B,C, the CS challenge caused the decrease of intracellular GSH content and SOD activity (p < 0.01), with a parallel of increased MDA level which demonstrates that the CS challenge induced oxidative stress in the in vitro cell model. Further, we observed that pretreatment with ginsengenin-S2 could attenuate CS-induced oxidative stress by clearly upregulating the content of SOD and GSH to a relatively basal level. Figure 5. Effects of ginsengenin-S2 (100 μM) on (A) the reactive oxygen species (ROS) level, (B) intracellular dismutase (SOD) activity and (C) glutathione (GSH) level in CS-exposed A549 cells. All data were expressed as mean ± S.D., n = 6. ** p < 0.01, compared with control; ## p < 0.01, compared with the CSE group.

Effect of Ginsengenin-S2 on CS-Induced IL-8 Levels In Vitro
CS is also a leading risk factor for the development of an inflammatory condition characterized by the release of proinflammatory mediators such as interleukin-8 (IL-8). We then investigated whether ginsengenin-S2 inhibited the release of chemokine IL-8 from CS-stimulated lung epithelial cells. As shown in Figure 6, IL-8 production markedly rose in the cell treated with CS (p < 0.01), however pretreatment with ginsengenin-S2 at 100 μM resulted in a decreased IL-8 level (p < 0.05). All data were expressed as mean ± S.D., n = 6. ** p < 0.01, compared with control; ## p < 0.01, compared with the CSE group.

Effect of Ginsengenin-S2 on CS-Induced IL-8 Levels In Vitro
CS is also a leading risk factor for the development of an inflammatory condition characterized by the release of proinflammatory mediators such as interleukin-8 (IL-8). We then investigated whether ginsengenin-S2 inhibited the release of chemokine IL-8 from CS-stimulated lung epithelial cells. As shown in Figure 6, IL-8 production markedly rose in the cell treated with CS (p < 0.01), however pretreatment with ginsengenin-S2 at 100 µM resulted in a decreased IL-8 level (p < 0.05).
antioxidants-superoxide dismutase (SOD) and glutathione (GSH) in CS-exposed human lung epithelial cells. As shown in Figure 5B,C, the CS challenge caused the decrease of intracellular GSH content and SOD activity (p < 0.01), with a parallel of increased MDA level which demonstrates that the CS challenge induced oxidative stress in the in vitro cell model. Further, we observed that pretreatment with ginsengenin-S2 could attenuate CS-induced oxidative stress by clearly upregulating the content of SOD and GSH to a relatively basal level. Figure 5. Effects of ginsengenin-S2 (100 μM) on (A) the reactive oxygen species (ROS) level, (B) intracellular dismutase (SOD) activity and (C) glutathione (GSH) level in CS-exposed A549 cells. All data were expressed as mean ± S.D., n = 6. ** p < 0.01, compared with control; ## p < 0.01, compared with the CSE group.

Effect of Ginsengenin-S2 on CS-Induced IL-8 Levels In Vitro
CS is also a leading risk factor for the development of an inflammatory condition characterized by the release of proinflammatory mediators such as interleukin-8 (IL-8). We then investigated whether ginsengenin-S2 inhibited the release of chemokine IL-8 from CS-stimulated lung epithelial cells. As shown in Figure 6, IL-8 production markedly rose in the cell treated with CS (p < 0.01), however pretreatment with ginsengenin-S2 at 100 μM resulted in a decreased IL-8 level (p < 0.05). Figure 6. Anti-inflammatory effect of ginsengenin-S2 (100 µM) on the inflammatory cytokine interleukin-8 (IL-8) in CS-exposed A549 cells. All data were expressed as mean ± S.D., n = 4. ** p < 0.01, compared with control; # p < 0.05, compared with the CSE group.

Effect of Ginsengenin-S2 on CS-Mediated Protein Expression of Nrf2 and HDAC2 In Vitro
To explore the underlying mechanism of the antioxidant activity and anti-inflammatory property of ginsengenin-S2, we examined the effect of ginsengenin-S2 on the protein expression of nuclear-related factor 2 (Nrf2) and histone deacetylase 2 (HDAC2) in CS-exposed A549 cells. In our study, we observed that the expression level of Nrf2 increased and the level of HDAC2 decreased after the CS challenge (p < 0.05). Interestingly, treatment with ginsengenin-S2 significantly activated the expression of Nrf2 (p < 0.01) as compared to the control group (Figure 7), which suggested that Nrf2 may participate in the meditative effect of ginsengenin-S2 in CS-induced oxidative stress. In addition, ginsengenin-S2 up-regulated the expression of HDAC2 in CS-exposed lung epithelial cells (p < 0.05).

Effect of Ginsengenin-S2 on CS-Mediated Protein Expression of Nrf2 and HDAC2 In Vitro
To explore the underlying mechanism of the antioxidant activity and anti-inflammatory property of ginsengenin-S2, we examined the effect of ginsengenin-S2 on the protein expression of nuclear-related factor 2 (Nrf2) and histone deacetylase 2 (HDAC2) in CS-exposed A549 cells. In our study, we observed that the expression level of Nrf2 increased and the level of HDAC2 decreased after the CS challenge (p < 0.05). Interestingly, treatment with ginsengenin-S2 significantly activated the expression of Nrf2 (p < 0.01) as compared to the control group ( Figure  7), which suggested that Nrf2 may participate in the meditative effect of ginsengenin-S2 in CSinduced oxidative stress. In addition, ginsengenin-S2 up-regulated the expression of HDAC2 in CS-exposed lung epithelial cells (p < 0.05). Oxidative stress is involved in COPD pathogenesis and could cause the enrichment of ROS and even leads to organ tissue damage [28,29]. CS is one of the major risk contributors to COPD which could increase ROS production. As an important role in anti-oxidation, GSH could help reduce the production of lipid peroxide and prevent cell damage [30]. SOD is an enzyme which could scavenge superoxide radicals. It is reported that SOD and GSH-associated enzymes in the lungs of COPD patients are significantly affected by CS [31]. MDA, an important marker for monitoring the process of membrane lipid peroxidation and damage degrees, appears to have a different level between healthy smokers and COPD patients [32]. The current data suggested that ginsengenin-S2, a novel sapogenin, noticeably attenuated CS-induced oxidative damage and inflammatory action via upregulating the content of intracellular SOD and MDA and inhibiting the level of MDA and IL-8 release. Nrf2, an important antioxidant transcription factor against oxidative stress, is released from the Keap1-Nrf2 complex and translocates from the cytoplasm when cells are exposed to oxidative stress. Increasing studies show that Nrf2 mediates the protective effects of natural products on CS Figure 7. (A) Effect of ginsengenin-S2 (100 µM) on the protein expression of nuclear-related factor 2 (Nrf2) and histone deacetylase 2 (HDAC2) in the CS-exposed A549 cells was examined by western blot analysis; (B) quantitative analyses of Nrf2 in each group. Data are expressed as mean ± SD (n = 3). * p < 0.05, ** p < 0.01 as compared to the control group; # p < 0.05 as compared to the CSE treated group.
Oxidative stress is involved in COPD pathogenesis and could cause the enrichment of ROS and even leads to organ tissue damage [28,29]. CS is one of the major risk contributors to COPD which could increase ROS production. As an important role in anti-oxidation, GSH could help reduce the production of lipid peroxide and prevent cell damage [30]. SOD is an enzyme which could scavenge superoxide radicals. It is reported that SOD and GSH-associated enzymes in the lungs of COPD patients are significantly affected by CS [31]. MDA, an important marker for monitoring the process of membrane lipid peroxidation and damage degrees, appears to have a different level between healthy smokers and COPD patients [32]. The current data suggested that ginsengenin-S2, a novel sapogenin, noticeably attenuated CS-induced oxidative damage and inflammatory action via up-regulating the content of intracellular SOD and MDA and inhibiting the level of MDA and IL-8 release. Nrf2, an important antioxidant transcription factor against oxidative stress, is released from the Keap1-Nrf2 complex and translocates from the cytoplasm when cells are exposed to oxidative stress. Increasing studies show that Nrf2 mediates the protective effects of natural products on CS exposure caused oxidative stress and inflammatory responses [16,33]. CS-induced lung inflammation involves the reduction of HDAC2 which is related to steroid resistance in patients with COPD who smoke cigarettes [34]. In this study, we found that ginsengenin-S2 activated the Nrf2 and enhanced the HDAC2 activity, suggesting that the protective effect of ginsengenin-S2 may involve the Nrf2 and HDAC2 pathways. To better utilize the medicinal resource of Panax ginseng Meyer cv. Silvatica, it is necessary to study the protective effect of its root extract against the cigarette smoke-induced COPD or airway fibrosis in vivo in the future. triterpenoid structure; 1 H and 13 C-NMR: see Table 1

Preparation of Cigarette Smoke Extract
Xiongshi cigarette (China Tobacoo Zhejiang Industrial Co., Ltd, Hangzhou, China; each cigarette contained 0.7 mg of nicotine, 8 mg of tar and 10 mg of carbon monoxide) were used to prepare the cigarette smoke extract (CSE) according to the literature [35]. In brief, one Xiongshi cigarette was bubbled through 25 mL of culture medium to generate 100% concentration CSE, then the solution was filtered through the 0.22 µm pore filter. The CSE (pH 7.2-7.4) was freshly prepared for each experiment and was then diluted to the desired concentration and used within 30 min.

Cell Viability Assay
The triterpenoids were dissolved in DMSO as stock solutions and diluted with DMEM to the needed concentrations. Final DMSO concentration in the sample was less than 0.1%. Cell culture and MTT assay were done as we described previously to evaluate 24 h-incubation of triterpenoids or the 18 h-incubation of CSE on the cell viability of A549 cells [6].

Drug Treatment
A549 cells were cultured in 6-well plates at a density of 5 × 10 5 cells per 1 mL and were pretreated with triterpenoids for 6 h. After that, the cells were treated with both CSE and triterpenoids for another 18 h of incubation.

Determination of Oxidative Stress
After 18 h of stimulation of CSE, the levels of intracellular MDA, GSH and SOD in the lysates were measured using commercial MDA, GSH and SOD kits (Jiancheng Bioengineering Institute, Nanjing, China) according to the manufacturer's protocol.

Reactive Oxygen Species (ROS) Assay
A549 cells were cultured in 96-well plates (5 × 10 3 cells per well). Drug treatment and CS stimulation were carried out as described in Sections 3.5.3 and 3.5.4. The cells were washed with warm PBS and the ROS detection was performed with DCFH-DA probe based ROS Assay Kit (Beyotime Biotechnology, Shanghai, China) at 37 • C. The ROS level was expressed as the fluorescence intensity.

Enzyme-Linked Immunosorbent Assay
After 18 h of stimulation of CSE, IL-8 content in the cell culture supernatant was measured using IL-8 ELISA kit (R&D Systems, Minneapolis, MN, USA) according to the manufacturer's protocol.

Statistical Analysis
Data analysis was performed using GraphPad Prism 6.0 software (GraphPad Software, San Diego, USA). All the data are presented as mean ± S.D. and a two tailed test or a one-way analysis of variance (ANOVA) was used to determine the statistical significance, and p-value <0.05 was considered as significant.

Conclusions
In our study, four new dammarane-type triterpenoids, elucidated as ginsengenin-S1 (1), ginsengenin-S2 (2), ginsenoside-S3 (3) and ginsenoside-S4 (4) were isolated from pearl knots of PGS and their antioxidant activities were investigated using CS-stimulated A549 cells. Among these compounds, ginsengenin-S2 exhibited protective activity against CS-induced oxidative damage and ameliorated inflammatory reaction. Further, its protective effect against CS may involve the Nrf2 and HDAC2 pathways. This study revealed the evidence that an active compound from the root of Panax ginseng Meyer cv. Silvatica-ginsengenin-S2 is beneficial to reduce injury from cigarette smoke in vitro and provides some theoretical basis for the development and utilization of the resource Panax ginseng Meyer cv. Silvatica. It reminded us that it is worthwhile to study the in vivo protective effect of the root of Panax ginseng Meyer cv. Silvatica against cigarette smoke using animal models.