Introduction

Nowadays, poor eating habits, particularly those involving high dietary fat and cholesterol intake, have led to a rapid increase in the incidence of health problems, such as hyperlipidemia and obesity (Bray and Popkin 1998; Formiguera and Canton 2004). Chronic consumption of the diet described alters the plasma and tissue levels of cholesterol and triglyceride and results in a higher risk for liver disease (Pendino et al. 2005; Wouters et al. 2008; Ferre et al. 2009). Although several types of evidence suggest that hyperlipidemia may not appear to impair liver function, it may affect the severity of tissue injury by increasing the vulnerability of the liver to the harmful effects of cytokines and oxidative agents (Jou et al. 2008; Farrell and Larter 2006). Previous research in experimental models of hyperlipidemia has shown that inflammatory gene expression in liver is induced by increased plasma cholesterol and modified lipoprotein levels (Pendino et al. 2005; Wouters et al. 2008; Ferre et al. 2009). It has been demonstrated that nuclear factor kappa-B (NFκB) may play a role in the development of hyperlipidemia-related chronic diseases by modulating the expression of various cytokines and adhesion molecules (Valen et al. 2001). In addition to NFκB, mitogen-activated protein kinases (MAPKs) are a group of signaling molecules that are involved in inflammatory cytokines production (Pearson et al. 2001; Ajizian et al. 1999). Three MAPK cascades are well described: (1) extracellular signal-regulated kinase (ERK) 1/2, (2) p38, (3) c-Jun N-terminal kinase (JNK)/stress-activated protein kinase (SAPK), which regulates immune responses, including inflammatory cytokine production (Boulton et al. 1991; Lee et al. 1994; Kyriakis et al. 1994). Furthermore, NFκB and MAPK cascades have been associated with the control of cell apoptosis through transcriptional regulation of adhesion molecules and cytokines (Chen and Greene 2004). Therefore, it is considered to be of benefit to reduce the risk of hepatic injury in hyperlipidemia by blunting the inflammatory response and impeding apoptosis in liver.

Panax ginseng C.A. Meyer (Ginseng) has been used for several 1,000 years to prolong longevity in Asian countries (Liu and Xiao 1992). It has been used as a “cure-all” in Asia and as an adaptogenic herb or a food supplement in Western countries (Xie et al. 2005; Mayr et al. 2004). Ginsenoside in ginseng, namely triterpenoidal dammarane saponin, which is uniquely present in the Panax species, is divided into protopanaxadiols (PDs) and protopanaxatriols (PTs). The PDs, including ginsenoside Rb1, Rb2, Rc, and Rd, have an anti-obesity effect (Liu et al. 2010; Kim et al. 2009) and exhibit anti-inflammatory activity (Lee et al. 2005), whereas the PTs, such as ginsenoside Re, Rf, and Rg1, have efficacy on the prevention of oxidative stress (Kwok et al. 2010) and the modulation of inflammatory processes (Oh et al. 2004). These reports show that each ginsenoside has different biological effects, but their synergetic properties in hyperlipidemia have not been examined so far. Although ginsenosides might attenuate hyperlipidemia-related chronic diseases, such as atherosclerosis, via an anti-inflammatory action and regulation of the blood lipid profiles (Liu et al. 2009, 2010; Wan et al. 2009), there are no studies on the effects of the PDs and PTs ginsenosides combination on liver inflammation in hyperlipidemia. The purposes of this work were to evaluate the protective effects of a combination of PDs and PTs ginsenosides on liver inflammation and apoptosis in an in vivo hyperlipidemic model, apoE null (KO) mice that were fed a high-cholesterol diet.

Materials and methods

Animals and treatment

Male apoE KO mice with wild-type C57BL/6 were purchased from Japan SLC Incorporation (Tokyo, Japan), which were kept both for 12 weeks on a high-cholesterol diet (1.25% cholesterol, 7.5% cocoa butter, 0.5% sodium cholate, DYET#102068; Dyets Inc., Bethelehem, PA, USA) (Guo et al. 2005). On the 8th week post induction of hyperlipidemia, the mice were intraperitoneally (i.p.) injected with ginseng extracts (100 mg/kg/day) for 4 weeks. Experiments were performed in accordance with the guidelines of the Committee on Animal Care at Sungshin Women’s University. At the end of the study period, the animals were euthanized, and plasma samples were obtained along with the different tissues.

Preparation of ginseng extracts

Ginseng extracts were obtained from the Korean Food Research Institute (Gyeonggi-do, Republic of Korea) and prepared using a modification of a previously published method (Shibata et al. 1966). Four-year-old ginseng plant roots (Anseong Nonghyup, Republic of Korea) were separated into the main body, lateral, and fine parts. R1 samples (PDs:PTs = 1:1) were made with the main-to-lateral ratio of 5:1, while R2 (PDs:PTs = 2:1) samples were made with the lateral-to-fine ratio of 1:5. The same amounts of different parts of the roots (2 g) were heat-dried for 7 days at 60°C, ground in a Gold HM-5000 (Mixer, 30 mesh) (Hyundae Co., Incheon, Republic of Korea), and extracted with 100 ml of 80% ethanol. During extraction, the samples twice underwent the process using a circulating condenser for 3 h each, at 80°C (Korea Food Research Institute). Extracts were concentrated under reduced pressure and used over the 65 brix. HPLC analysis was used to identify the ginsenosides in R1 and R2, showing that R1 contained the richest amounts of Rg1 (PT) and Rb1 (PD), while R2 contained the richest amounts of Re (PT) and Rd (PD) (Fig. 1a, b).

Fig. 1
figure 1

The distribution of ginsenosides in R1 and R2 ginseng extracts identified by HPLC (a). The inserted table in Fig. 1 a provides the HPLC conditions used. R1 contained higher contents of Rb1 and Rg1, while R2 contained higher levels of Rd and Re (b). The fourteen standards of ginsenosides included compounds K, Rh2, Rh1, Rg5, Rk1, Rg2, Rg3, Rg1, Rf, Re, Rd, Rb2, Rc, and Rb2, which were provided by Daedeck Science Town, Daejeon, Republic of Korea

Full size image

Measurement of lipid profiles

Lipid profiles for total cholesterol (TC) and triacylglyceride (TG) were measured in plasma and liver tissue. TC and TG determinations were measured with enzymatic colorimetric methods (ASSEL S.r.l., Guidna, Italy) (Vassault et al. 1981). High-density lipoprotein cholesterol (HDL) levels were also measured with an enzymatic colorimetric method (Daiichi Pure Chemicals Co., Ltd. Tokyo, Japan) (Gordon et al. 1977).

Oxidative stress indicators

Lipid peroxides (LPO) and malondialdehyde (MDA) were measured as representative metabolites derived from the peroxidation of dietary poly-unsaturated fatty acids (PUFA). LPO and MDA levels were measured using a microplate assay kit (# fr22; Oxford Biochemical Research, Oxford, MI, USA) as previously reported by Lee and Bae (2007).

Western blot analysis

Liver samples were homogenized in lysis buffer (pH 7.4, 10% v:v) containing 1 M Tris, 5 M NaCl, and 0.1 M EDTA. The homogenates were then centrifuged at 15,000 rpm for 45 min at 4°C. The cytosol fraction was boiled for 5 min at 95°C with a sodium dodecyl sulfate-loading buffer. Proteins were fractioned by SDS–polyacrylamide gel electrophoresis and transferred to nitrocellulose membranes (#401296; Schleicher & Schuell, Dassel, Germany). The membrane was incubated with a primary antibody (Cell signaling, Beverly, MA) that reacted with a horseradish peroxidase-conjugated secondary antibody (Cell signaling, Beverly, MA). The primary antibodies were rabbit anti-COX-2, rabbit anti-IκB-α, rabbit anti-phospho-ERK1/2, rabbit anti-SAPK/JNK, rabbit anti-phospho-c-Jun (Ser73), rabbit anti-Bcl-2, rabbit anti-caspase-8/-9, and rabbit anti-cleaved PARP. Antibody bindings were visualized by chemiluminescence with ECL kit (#SC-2048; Santa Cruz Biotechology, Inc., Santa Cruz, CA, USA) and exposed to X-ray film (Kodak XAR) to identify the target protein.

Statistical analysis

All values are shown as means ± standard deviation (SD). Statistical significances were assessed using one-way analysis of variance (ANOVA) with Duncan’s test. A P-value of 0.05 was considered significant.

Results

Lipid profiles and atherogenic indices

As shown in Table 1, the plasma TG concentration in the hyperlipidemic mice was significantly increased as compared to the wild-type control mice by 6-fold, but it was significantly decreased by the R1 and R2 treatments. The hepatic level of TG was not significantly different between the wild-type mice and the hyperlipidemic mice, but the R2 treatment showed a significant decrease in this level. The increase in plasma TC concentration was significantly attenuated by R1 treatment compared with the hyperlipidemic mice. Hepatic TC levels were not affected by the high-cholesterol diet or ginsenosides combination. While plasma HDL level was significantly increased in the hyperlipidemic mice compared to the wild-type control mice, this level was increased by 1.4-fold in both the R1 and R2 treatment groups. Moreover, the atherogenic indices significantly decreased in the groups in which R1 and R2 were administered, as compared to the hyperlipidemic mice.

Table 1 Alterations of lipid profiles and atherogenic indices in plasma and liver tissues of apoE KO mice administrated with the R1 and R2 ginseng extracts
Full size table

Effect of PDs and PTs ginsenosides combination on oxidative stress

As shown in Table 2, the plasma LPO level of the hyperlipidemic mice was significantly increased as compared with that of the wild-type control mice by 3.7-fold. However, it was decreased by 0.8-fold in the R1 treatment group. Hepatic LPO content in the R1 treatment mice decreased by 0.6-fold compared with the hyperlipidemic mice. Furthermore, the MDA level of plasma was markedly higher in the hyperlipidemic mice compared with the wild-type control mice (160.7 vs. 711.9 μM/μl). However, the administration of R1 and R2 significantly lowered the level of plasma by 0.9- and 0.8-fold, respectively, compared to the hyperlipidemic mice. On the other hand, there was no significant difference in the MDA level in the liver between the hyperlipidemic mice and ginsenosides combination groups.

Table 2 The changes in production of lipid peroxides, LPO, and MDA in apoE KO mice treated with R1 and R2 ginseng extracts
Full size table

Effect of PDs and PTs ginsenosides combination on liver inflammation

The effects of ginsenosides combination on protein levels concerned with the inflammatory state of the liver in the hyperlipidemic mice were evaluated by Western blot analysis (Fig. 2). The hepatic protein level of IκB-α showed a significant increase in R1 and R2 treatment groups compared with that of the hyperlipidemic mice. The protein level of COX-2 in the liver was decreased in the groups that were administered the R1 and R2 ginsenosides combination by 0.6- and 0.5-fold, respectively, compared with the hyperlipidemic mice. Although the hepatic phospho-ERK 1/2 protein level was not different between the wild-type control mice and the hyperlipidemic mice, the protein level of the R1 and R2 ginsenoside combination groups was significantly decreased as compared to that of the hyperlipidemic mice. Moreover, the hepatic phospho-c-JUN protein level in the hyperlipidemic mice was increased compared with the wild-type mice by 2.1-fold. However, the hyperlipidemic mice that were administered 100 mg/kg of R2 showed a significant decrease in the hepatic phospho-c-JUN protein level by 0.7-fold. There was no significant difference among all the groups in terms of the hepatic phospho-SAPK/JNK protein level.

Fig. 2
figure 2figure 2

Changes in inflammatory response in hepatic cytosol of hyperlipidemic mice after treatment with R1 and R2 ginseng extracts having varying PD/PT ratios. Differences between means were analyzed by one-way analysis of variance followed by Duncan’s test (P < 0.05). The bars indicate the means ± SD

Full size image

Effect of PDs and PTs ginsenosides combination on liver apoptosis

Figure 3 shows the effect of the ginsenosides combination on apoptosis under a high-cholesterol diet. Bcl-2, the anti-apoptotic protein, in both the wild-type control mice and the hyperlipidemic mice was not different, whereas the R1 and R2 ginsenosides combination led to a significant increase by 2.1- and 1.5-fold, respectively. Furthermore, the R1 and R2 ginsenosides combination attenuated an increased caspase-8, -9, and cleaved-PARP protein expression shown in the hyperlipidemic mice.

Fig. 3
figure 3figure 3

Regulation of PD/PT ratio suppresses induction of apoptosis in hepatic cytosol by atherogenic diet in hyperlipidemic mice. Differences between means were analyzed by one-way analysis of variance followed by Duncan’s test (P < 0.05). The bars indicate the means ± SD

Full size image

Discussion

Ginseng is considered a valuable natural herb in the East as well as the West. Although most research has shown that there were no side effects of ginseng extracts, several studies reported that various saponins extracted from herbal plants facilitate hemolysis or exert hepatic toxicity (Lee et al. 2009). In our study, the administration of 100 mg/kg of ginseng extract did not demonstrate any hepatic damages as well as other studies. After an intraperitoneal injection with ginseng powder (50–100 mg/kg) to the animals in Benzo(α)pyrene induced liver damage, serum levels of aspartate aminotransferase (AST) and alanine transaminase (ALT) were significantly decreased (Kim et al. 2007). The red ginseng extract (50–100 mg/kg) also reduced the activity of AST and ALT in carbon tetrachloride induced liver injury (Lee et al. 2003). We have even found that an R1 ginsenosides combination decreased the plasma levels of AST and ALT in hyperlipidemia mice (Jang 2008). Therefore, we determined the standard levels of injection by the amount of injection when AST/ALT activity was improved. We have now extended our ginseng studies to the apoE KO mouse model that exhibits advanced atherosclerosis, increased tissue oxidation, and altered immune system.

R1 contained richer amounts of ginsenoside Rb1, Rb2, and Rc than R2. It was confirmed that the R1 exerted a protective effect against hepatic injury prompted by H2O2 treatment in previous studies (Lee et al. 2009). Also, previous studies reported that PD types of ginsenosides were effective not only in improving conditions of obesity, fatty liver, and hypertriglyceridemia in rodent models fed with a high-fat diet (Liu et al. 2010; Kim et al. 2009) but also in regulating inflammatory responses by inducing heme oxygenase (HO)-1, inhibiting inducible nitric oxide synthase (iNOS), and suppressing the tumor necrosis factor (TNF)-α expression in vitro (Lee et al. 2005; Cho et al. 2006). Moreover, ginsenoside PTs inhibit energy gain and normalize hypothalamic neuropeptides associated with the control of obesity in rats (Kim et al. 2009) and have a cytoprotective effect through their action against oxidative stress by blocking an increase in LPS-induced iNOS and COX-2 expressions via control of NFκB, which may be responsible for the chemoprevention of inflammatory diseases (Kwok et al. 2010; Oh et al. 2004). Thus, these data suggest that the ginsenosides combination with different ratios may affect hyperlipidemia and its related inflammatory response in liver.

Cell death plays an important role in immunogenesis, during inflammation and in the resolution of inflammatory reactions. The mechanisms and mediators involved have not yet been fully elucidated, but there is evidence that surrounding cells adjacent to inflammation kill themselves by apoptosis, increasing the damage caused by inflammatory reactions. One of the major pro-inflammatory factors, TNF-α, is involved in insulin resistance, inflammatory reactions, and cell apoptosis via various pathways. It also activates JNK and NF-κB (Zhang et al. 2008; Hotamisligil et al. 1993; Feinstein et al. 1993). MAPKs such as ERK1/2, JNK, p38 MAP kinase, PI3 K/PKB, and SAPK regulate inflammatory reactions, proliferation, differentiation, apoptosis, and survival of cells (Lin et al. 2007). PDs that have higher levels of Rb1 and Rd are a TNF-α antagonist, and they lower the synthesis of TNF-α through the inhibition of COX-2, phospho-ERK, and c-Jun expression (Widmann et al. 1999). In this study, the R2 ginsenosides combination was significantly effective in blocking the MAPK signaling pathway, such as c-Jun phosphorylation, and in inhibiting the expression of COX-2, even though both R1 and R2 treatments inhibited IκB-α degradation. Caspase-8 is the most proximal caspase to become activated upon ligation of the Fas molecule. Caspase-8 also cleaves Bid to produce tBid fragments, which then relocates to the mitochondria to further alter the mitochondrial membrane. The mitochondria-mediated pathway results in the activation of caspase-3 and caspase-9, which then activates downstream caspase-8 (Oh and Lee 2004). Our data showed that an R1 and R2 ginsenosides combination increased the expression of Bcl-2, whereas those treatments suppressed the expression of apoptotic signals, caspase-8/-9, and cleaved PARP. Therefore, R1 and R2 had potential anti-proinflammatory and anti-apoptotic effects in the liver of hyperlipidemic mice.

The present study showed that the plasma lipid profiles and AI were higher in hyperlipidemic mice that were fed a high-cholesterol diet, whereas the administration of an R1 and R2 ginsenosides combination selectively attenuated the changes. We found that ginseng extract (PD/PT = 1:1) had anti-apoptotic effect on TUNEL-induced apoptotic cardiomyocyte of high-cholesterol fed apo E KO mice. And TC- and TG-fed cholesterol plus PD/PT = 1:1 were also decreased in heart tissues by more than 200% compared to apoE KO-fed high cholesterol (Jang 2008). However, this study was performed for the comparison of 1:1 versus 2:1 of PD/PT to prevent cholesterol-derived fatty liver in apoE KO mice with mechanism of systemic inflammation. Inflammation causes alteration of lipid metabolism that, in turn, makes worse the inflammatory response by leading to a malicious cycle (Hotamisligil 2006; Khovidhunkit et al. 2004). Improvement of lipid metabolism disorders was beneficial to recovering from inflammation, and, in turn, the anti-inflammatory effect of Panax notoginseng saponins maintained a balanced lipid metabolism (Liu et al. 2010). The PDs and PTs ginsenosides combination obstructed the vicious cycle via regulation of lipid profiles in the plasma and liver as well as the anti-inflammatory response. Moreover, while the R2 ginsenosides combination decreased the hepatic TG level in hyperlipidemic mice, the R1 treatment was effective on an increased plasma TC level in the hyperlipidemic mice. In previous studies, the oral administration of Re, which is an abundant ginsenoside in R2, has shown to induce an anti-hypercholesterolemic effect in an STZ-induced diabetic rat (Cho et al. 2006). The i.p. injection of Rb1 (10 mg/kg), in which R1 is plentiful, displayed a tendency to lower the AI-2 (TG/HDL), but it did not affect the serum lipid profiles in rats (Park et al. 2002). Therefore, the effect of the R1 and R2 ginsenosides combination on the lipid profile may be dependent on its composition of ginsenosides and tissue applied. This is also connected with other study that a close interrelation between regulators of lipid metabolism and systemic inflammation. It showed the amount of atherosclerosis is associated with macrophages and T-cells within plaques and adipose tissue and correlates with plasma cholesterol values (Lohmann et al. 2009).

MDA, the final product of LPO, is highly cytotoxic and exerts oxidative injury to the cell. Previous research has recently shown that dietary cholesterol contributes to the development of hepatic inflammation (Wouters et al. 2008; Mari et al. 2006; Vergnes et al. 2003). Increased levels of oxidized cholesterol products have been shown to induce oxidative stress (Wouters et al. 2008; Ferre et al. 2009). In our study, while an R1 and R2 ginsenosides combination decreased the plasma MDA level in hyperlipidemic mice, an R1 treatment reduced the LPO level in both the plasma and liver, suggesting that the ginsenosides combination, either R1 or R2, showed a tissue-specific effect toward reducing oxidative stress. In previous studies, the MDA levels in the kidney and eye were significantly reduced in diabetic rats treated with 20 mg/kg ginsenoside Re, the main constituent in R2, whereas the MDA level in the aorta was not significantly altered (Cho et al. 2006). The study also demonstrated that administration of Re is needed to reduce the activity of ROS in diabetic rats and to induce the protective effect against oxidative stress (Cho et al. 2006).

In conclusion, the administration of a combination of PDs and PTs ginsenosides demonstrated a preventive effect on liver inflammation and apoptosis in hyperlipidemic apo E KO mice through the modulation of oxidative stress.