Anti-obesity effects of Lysimachia foenum-graecum characterized by decreased adipogenesis and regulated lipid metabolism

Lysimachia foenum-graecum has been used as an oriental medicine with anti-inflammatory effect. The anti-obesity effect of L. foenum-graecum extract (LFE) was first discovered in our screening of natural product extract library against adipogenesis. To characterize its anti-obesity effects and to evaluate its potential as an anti-obesity drug, we performed various obesity-related experiments in vitro and in vivo. In adipogenesis assay, LFE blocked the differentiation of 3T3-L1 preadipocyte in a dose-dependent manner with an IC50 of 2.5 µg/ml. In addition, LFE suppressed the expression of lipogenic genes, while increasing the expression of lipolytic genes in vitro at 10 µg/ml and in vivo at 100 mg/kg/day. The anti-adipogenic and anti-lipogenic effect of LFE seems to be mediated by the inhibition of PPARγ and C/EBPα expression as shown in in vitro and in vivo, and the suppression of PPARγ activity in vitro. Moreover, LFE stimulated fatty acid oxidation in an AMPK-dependent manner. In high-fat diet (HFD)-induced obese mice (n = 8/group), oral administration of LFE at 30, 100, and 300 mg/kg/day decreased total body weight gain significantly in all doses tested. No difference in food intake was observed between vehicle- and LFE-treated HFD mice. The weight of white adipose tissues including abdominal subcutaneous, epididymal, and perirenal adipose tissue was reduced markedly in LFE-treated HFD mice in a dose-dependent manner. Treatment of LFE also greatly improved serum levels of obesity-related biomarkers such as glucose, triglycerides, and adipocytokines leptin, adiponectin, and resistin. All together, these results showed anti-obesity effects of LFE on adipogenesis and lipid metabolism in vitro and in vivo and raised a possibility of developing LFE as anti-obesity therapeutics.


Introduction
Obesity has become a global health problem due to its association with various metabolic disorders such as type 2 diabetes, cardiovascular diseases, and certain types of cancer (Kopelman, 2000;Spiegelman and Flier, 2001;Yanovski and Yanovski, 2002;Hong and Park, 2010). Current medications used to treat obesity can be grouped into two categories (Bray and Tartaglia, 2000;Korner and Aronne, 2004). The first includes appetite suppressants or anorexics that control food intake by modulating the central nervous system. Sibutramine, a serotonin and noradrenaline re-uptake inhibitor, and rimonabant, a cannabinoid receptor antagonist, belong to this category. However, sibutramine is known to increase blood pressure, and may cause cardiac arrhythmias, constipation, and headache with minimum body weight loss (Luque and Rey, 1999). Furthermore, the use of rimonabant has been suspended due to its serious psychiatric side-effects (Leite et al., 2009). Drugs in the second category inhibit the absorption of specific nutrients in food. For example, orlistat reduces fat intake by inhibiting pancreatic lipase, but often causes steatorrhea and is associated with a risk of developing deficiencies in lipidsoluble vitamins and essential fatty acids (Heck et al., 2000).
Due to the undesirable side-effects associated with the currently available anti-obesity medications and limited efficacy, much attention has been focused on developing drugs that directly modulate energy metabolism without affecting the central nervous system. Some natural products such as berberine, resveratrol, and curcumin are known to have anti-obesity effects Zang et al., 2006;Lee et al., 2009). These natural compounds ameliorate obesity either by increasing energy expenditure or by inhibiting adipocyte differentiation.
Lysimachia foenum-graecum, an herbal plant, has been primarily used as a spice, insectifuge, and pest repellent. It has also been used as a traditional oriental medicine for the treatment of cold, rheum, headache, toothache, and diseases of the digestive system (Shen et al., 2005;Li et al., 2009). Although the chemical profile of the L. foenum-graecum has not been published yet to the best of our knowledge, triterpene saponins are found in abundance from the aerial part of L. foenum-graecum. Ten out of eleven triterpene saponins isolated from L. foenum-graecum were found to have novel structures and have been named as foenumoside A-E (Shen et al., 2005), lysimachiagenoside A, C-F (Li et al., 2009a(Li et al., , 2009b. Anti-inflammatory effect of foenumoside E (Shen et al., 2005), and anti-oxidant effects of the methanol extract of L. foenum-graecum were reported recently (Li et al., 2009), but anti-obesity effect of L. foenum-graecum has not been investigated yet.
In our high-throughput screening of natural product extract library against 3T3-L1 preadipocyte differentiation, we found a potent anti-adipogenic effect of L. foenum-graecum extract (LFE). To further characterize the anti-adipogenesis activity in detail, we conducted a series of in vitro and in vivo obesity-related assays. In this report, we demonstrate that LFE suppresses adipogenesis and shifts lipid metabolism towards lipid breakdown and suggest the potential use of LFE in treating obesity.

LFE inhibits adipogenesis in vitro
To develop an anti-obesity drug that primarily regulates lipid metabolism without affecting the central nervous system, we performed a highthroughput adipocyte differentiation assay using 3T3L1 cells. 3T3L1 cells were seeded in 96-well plates and differentiation of preadipocyte was induced as described previously (Seo et al., 2004) in the absence or presence of natural plant extract at 10 μg/ml. Six days later, cells were stained with Nile Red, a lipophilic fluorescent dye, and intracellular lipid content was measured using EnVision Multilabel Plate Reader. LFE was selected from 2,000 natural plant extracts for its potent antiadipogenic activity (data not shown). To confirm the anti-adipogenic effect of LFE, preadipocyte 3T3-L1 cells were differentiated into mature adipocytes in the presence of various concentrations of LFE, and lipid accumulation was measured at day 6 by staining with Nile Red. As shown in Figure 1, LFE suppressed lipid accumulation with an IC50 of 2.5 μg/ml ( Figure 1A), and reduced the number and size of cytosolic lipid droplets ( Figure 1B). Cell viability was determined using MTT assay, and we observed no difference in LFE-treated and untreated cells, even at concentrations as high as 10 μg/ml LFE (data not shown).
To investigate whether the reduced lipid accumulation in LFE-treated cells was due to the inhibition of adipocyte differentiation, we examined the expression of adipogenic markers in these cells. The expression of peroxisome proliferator-activated receptor γ (PPARγ) and CCAAT/enhancer binding protein α (C/EBPα), master regulators of adipogenesis, was markedly suppressed by the addition of LFE ( Figure 1C). In consistent with this result, expression of aP2 and adiponectin, downstream targets of PPARγ and C/EBPα, was also reduced significantly in LFE-treated cells as revealed by qPCR ( Figure 1D). In addition, LFE efficiently suppressed the induction of other mature adipocyte markers fatty acid synthase (FAS), acetyl-CoA carboxylase 1 (ACC1), and stearoyl-CoA desaturase 1 (SCD1). These data confirmed the inhibitory effect of LFE on adipocyte differentiation.

LFE regulates the expression of PPARγ and C/EBPα and the activity of PPARγ in mature adipocytes
Since PPARγ and C/EBPα are also key regulators of lipogenesis, we asked whether LFE could also influence the expression of these proteins in differentiated adipocytes. Mature 3T3-L1 adipocytes were treated with LFE for 24 h and the expression level (C) Cells were harvested at day 6 and the expression of PPARγ and C/EBPα was analyzed by Western blotting. (D) Total RNA was extracted at day 6 and the expression level of PPARγ, FAS, ACC1, SCD1, aP2, and adiponectin mRNA was analyzed by qPCR. Each bar represents mean ± SD of triplicate PCR reactions. Similar results were obtained from two independent experiments. *P ＜ 0.05, **P ＜ 0.01, and ***P ＜ 0.001.
of PPARγ and C/EBPα was determined by Western blotting. LFE at 10 μg/ml was sufficient to repress the expression of PPARγ and C/EBPα in these cells ( Figure 2A). To address if LFE could also modulate the activity of PPARγ directly, we transfected HEK 293 cells with PPARγ and its dimerization partner retinoid X receptor α (RXRα). The activity of PPARγ/RXRα was increased by the treatment of PPARγ agonist rosiglitazone and abrogated by the co-treatment of PPARγ antagonist bisphenol A diglycidyl ether (BADGE) ( Figure 2B) (Bishop-Bailey et al., 2000). Interestingly, LFE abolished the activation of PPARγ/RXRα by rosiglitazone as efficiently as BADGE did, indicating the suppressive effect of LFE on PPARγ activation. Taken together, these results demonstrate the inhibitory effect of LFE on the expression of PPARγ and C/EBPα and activation of PPARγ, and suggest LFE prohibits lipogenesis by regulating PPARγ and C/EBPα.

LFE decreases the expression of lipogenic genes while increasing that of lipolytic genes
We next asked if LFE could modulate the expression of genes involved in lipid metabolism in mature adipocytes and muscles. When adipocytes were treated with 10 μg/ml of LFE, the expression of lipogenic genes such as adipocyte differentiation and determination factor 1 (ADD1)/sterol regulatory element binding protein 1c (SREBP1c), FAS, ACC1, and SCD1 were significantly reduced ( Figure  3A). Since these are essential enzymes for fatty acid synthesis, it is likely that treatment of LFE   Differentiated 3T3-L1 adipocytes were treated with DMSO, LFE (2 and 10 μg/ml) or berberine (30 μg/ml) for 30 min after starving cells in DMEM containing 0.1% BSA for 16 h. Berberine was used as a positive control for activating AMPK. Total cell lysates were subjected to Western blot analysis using antibodies specific for phospho-AMPK, phospho-ACC, total AMPK, total ACC, and β-actin. (B) Differentiated C2C12 cells were treated with or without LFE (10 μg/ml) or compound C (10 μM) and fatty acid oxidation assays were performed. Each bar represents mean ± SD from three independent experiments. *P ＜ 0.05 vs. DMSO, # P ＜ 0.05 vs. LFE.
directly impairs the ability of adipocytes to synthesize fat.
To analyze the effect of LFE on the expression of lipolytic genes, the amount of PPARα and PPARδ, major transcription regulators of fatty acid beta-oxidation, was measured in C2C12 myotubes in the absence or presence of 10 μg/ml of LFE. The expressions of PPARα and PPARδ, as well as their downstream targets, acyl-CoA oxidase (ACO) and carnitine palmitoyltransferase 1 (CPT1), were significantly increased in these cells ( Figure 3B). These data imply a role of LFE on stimulating lipid consumption in muscles.

LFE promotes fatty acid oxidation by activating AMPK
Activation of AMPK and subsequent inhibition of ACC is known to play a crucial role in fatty acid oxidation by regulating mitochondrial availability of fatty acids (Winder et al., 1997;Viollet et al., 2007). When differentiated 3T3-L1 cells were treated with LFE for 30 min, phosphorylation of both AMPK and its substrate ACC was increased in a dosedependent manner, without affecting the protein levels of total AMPK and ACC ( Figure 4A). The rapid phosphorylation of AMPK by LFE suggests a direct regulatory role of LFE on AMPK activation in adipocytes.
To examine if the LFE-induced activation of AMPK and concomitant inhibition of ACC correlates with fatty acid oxidation, C2C12 myotubes were treated with LFE in the presence or absence of specific AMPK inhibitor compound C. Addition of LFE alone increased fatty acid oxidation significantly, but the increase was abrogated completely by compound C ( Figure 4B). These results show that LFE promotes lipid breakdown in an AMPK-dependent manner.

LFE reduces body weight gain and fat mass in HFD-induced obese mice
To investigate anti-obesity effects of LFE in vivo, mice were fed a HFD for 6 weeks in the presence of various amounts of LFE and body weight gain was monitored twice a week. Oral administration of LFE significantly reduced body weight gain by HFD in all doses of LFE tested ( Figure 5A). Interestingly, LFE at 100 mg/kg or higher suppressed the body weight gain in HFD mice similar to the level in vehicle-treated ND mice. Importantly, no significant difference in food intake was observed between the vehicle-and LFE-treated HFD groups ( Figure  5B), confirming that the reduction in body weight gain in LFE-treated mice was not due to reduced caloric intake in these animals. To examine if the reduced body weight gain in LFE-treated group is related to decreased fat accumulation, weight of the abdominal subcutaneous, epididymal, and perirenal white adipose tissues and the sizes of epididymal adipocytes were examined. HFD-induced fat accumulation was significantly ameliorated in all of the white adipose tissues examined ( Figure 5C) and adipocyte size was greatly reduced in LFEtreated group ( Figure 5D), showing that LFEinduced body weight loss is mainly due to decreased fat accumulation in white adipose tissue.

LFE modulates metabolic gene expression and stimulates AMPK activation in vivo
We showed that LFE suppresses adipogenesis and lipogenesis by inhibiting the expression of PPARγ and C/EBPα in undifferentiated and differentiated 3T3L1 adipocytes ( Figures 1C and 2A). In consistent with these findings, the expression of C/EBPα and PPARγ was reduced markedly in the epididymal WATs obtained from LFE treated HFD mice compared with those from vehicle-treated HFD mice ( Figure 5F).
To check if LFE could also regulate metabolic gene expression in vivo, the expression of ADD1/ SREBP1c, FAS, ACO, and CPT1 in epididymal tissue was measured by qPCR. Consistent with our observation in vitro (Figure 3), LFE inhibited the expression of lipogenic genes ADD1/SREBP1c and FAS, and promoted the expression of lipolytic genes ACO and CPT1 in vivo ( Figure 5E). It has been known that HFD suppresses AMPK activation (Kim et al., 2008;Pang et al., 2008). Interestingly, LFE treatment restored AMPK phosphorylation in HFD mice ( Figure 5G), further suggesting that anti-obesity effect of LFE partly relies on the activation of AMPK in vivo.

LFE improves the levels of triglyceride, glucose, and adipocytokines in plasma
Next, we measured plasma levels of obesityrelated biomarkers in these mice. Forty five % reduction in circulating triglycerides and 35% reduction in glucose was observed in mice treated with LFE compared with mice fed a HFD alone ( Figure 6A). Plasma levels of the adipocytokines leptin and resistin were also reduced significantly in LFE-treated group, comparable to those in ND group ( Figure 6B). In addition, treatment of LFE partially restored the plasma level of adiponectin in HFD group ( Figure 6C). It has been known that plasma levels of leptin and resistin are proportional to body fat mass (Friedman and Halaas, 1998;Steppan et al., 2001), whereas the level of adiponectin is inversely correlated with body fat percentage (Yamauchi et al., 2001). The decrease in leptin and resistin and the increase in adiponectin in LFE-treated mice, together with reduced triglycerides and glucose level in plasma, strongly support the anti-obesity effects of LFE in HFD-fed mice.

Discussion
Although L. foenum-graecum has been used as an oriental medicine to treat a variety of diseases, its effect on obesity or metabolic diseases has not been known so far. In this report, we investigated anti-obesity effect of LFE and showed that LFE efficiently blocks adipogenesis and lipogenesis in 3T3L1 adipocytes, promotes lipid catabolism in cultured cells of different tissue types, and inhibits body and adipose tissue weight gain in HFDinduced obese mouse model. LFE was first identified in a high-throughput screening for natural product extracts with antiadipogenic effect. The effect was further confirmed by the reduced expression of various adipogenesisrelated genes ( Figure 1D). Interestingly, LFE inhibited the expression of PPARγ and C/EBPα in differentiating adipocytes ( Figure 1C), mature adipocytes (Figure 2A), and the epididymal WATs obtained from LFE-treated HFD mice ( Figure 5F). In addition, LFE efficiently blocked rosiglitazoneinduced activation of PPARγ ( Figure 2B). In agreement with these results, the expression of PPARγ target genes aP2 and adiponectin was also significantly decreased in LFE-treated adipocytes ( Figure 1D). PPARγ and C/EBPα are known to be the most crucial transcription factors in adipogenesis and lipogenesis (Tontonoz et al., 1994a(Tontonoz et al., , 1994b. Ectopic expression of PPARγ is sufficient to induce C/EBPα and to fulfill the differentiation of the preadipocytes to mature adipocytes. Although the initial induction of C/EBPα is dependent on PPARγ, but C/EBPα, in turn, is known to reinforce the expression of PPARγ and many specific genes required for adipocyte differentiation and lipogenesis. This positive feedback loop between PPARγ and C/EBPα is believed to be essential for the maintenance of adipocytes in the terminally differentiated state (Brun et al., 1996;Rosen et al., 2000). Considering the importance of these factors, it seems likely that LFE exerts its anti-adipogenic and anti-lipogenic effects by primarily suppressing PPARγ and C/EBPα in vitro and in vivo.
Our data showed an activation of AMPK and AMPK-dependent fatty acid oxidation by the treatment of LFE in vitro and in vivo (Figures 4 and  5G). As a master regulator of glucose and lipid metabolism, AMPK orchestrates a cohort of cellular responses upon nutritional stresses (Hardie et al., 2003(Hardie et al., , 2006Kahn et al., 2005). Activation of AMPK is known to switch anabolic pathways to catabolic pathways by inducing the expression of lipolytic transcription factor PPARγ and suppressing the expression of lipogenic transcription factor ADD1/ SREBP1c (Zhou et al., 2001;Yoon et al., 2006). Here we showed that LEF treatment causes an induction of PPARγ and its downstream targets CPT1 and ACO, and a reduction of ADD1/SREBP1c and its target FAS, ACC1, and SCD1 (Figure 3). Considering the stimulating effect of LFE on AMPK activation, it is plausible that LFE regulates the expression of the key metabolic regulators PPARγ and ADD1/SREBP1c through activating AMPK.
It is intriguing to think the underline mechanism of PPARγ, C/EBPα, and ADD1/SREBP1c suppression by LFE. It was reported that AMPK agonist aminoimidazole carboxamide ribonucleotide (AICAR) inhibits adipocyte differentiation by down-regulating PPARγ, C/EBPα, and ADD1/SREBP1c (Giri et al., 2006), suggesting that LFE inhibits the expression of these transcription factors indirectly through AMPK activation as observed in mature adipocytes. It has been also known that AICAR induces KLF2 expression, a repressor of PPARγ activity, in mouse embryonic stem cells and fibroblasts (Adamo et al., 2009), along with the fact that AMPK acts as an upstream positive regulator of p38 MAPK (Xi et al., 2001) and 38 MAPK represses PPARγ transcriptional activity by phosphorylating PPARγ Diradourian et al., 2005). These reports suggest a possible link between LFE-induced activation of AMPK and AMPK-mediated suppression of PPARγ transcriptional activity.
High level of blood leptin in obese individual is known to elicit glucose tolerance in type II diabetes and hepatic steatosis by acting at pancreas and liver, respectively (Maffei et al., 1995;Friedman and Halaas, 1998;Petersen et al., 2002;Javor et al., 2005). On the other hand, adiponectin is known to play a role in the suppression of type II diabetes and obesity (Yamauchi et al., 2001;Ukkola and Santaniemi, 2002;Diez and Iglesias, 2003). Treatment of LFE not only blocked obesity, but also restored normal levels of blood adipokines leptin, resistin, and adiponectin as well as triglyceride and glucose in HFD-induced obesity model (Figure 6), suggesting a beneficial effect of LFE in other metabolic diseases. In accordance with this, we also found LFE significantly relieved fatty liver symptoms in HFD mice (data not shown). Interestingly, LFE has been shown to have anti-oxidant and anti-inflammatory activities (Shen et al., 2005;Li et al., 2009). Since chronic diseases such as metabolic diseases and cardiovascular diseases are known to be associated with oxidative stress and inflammation (Lusis, 2000;Hansel et al., 2004), the anti-oxidant and anti-inflammatory activities of LFE together with its anti-obesity effect may further contribute to modify metabolic diseases such as obesity, diabetes, and hepatic steatosis.
In conclusion, this study shows anti-obesity effects of L. foenum-graecum for the first time. LFE efficiently blocked adipogenesis and lipogenesis in vitro, and total body and adipose tissue weight gain in HFD-induced obesity model. The mechanism of LFE action involves, at least in part, the activation of AMPK and suppression of PPARγ and C/EBPα. Although the precise mechanisms and active components in LFE need to be elucidated further, LFE provides novel therapeutic potential against obesity and possibly other metabolic diseases.

Plant material and preparation of LFE
Dried whole plant of L. foenum-graecum was purchased from Jinheung pharmacy (Kyoungdong Oriental Herbal Market, Seoul, Korea) and identified from their external appearance according to Hu and Kelso (Hu and Kelso, 1996). Sequencing analysis was conducted as previously described (Hao et al., 2004) and the sequences of the internal transcribed spacer region of nuclear ribosomal DNA and the chloroplast trnL-F region from the purchased plant specimen were found to be identical with the known sequences of L. foenum-graecum (AF547771 and AF547716, data not shown), further confirmed the identity of L. foenum-graecum. A voucher specimen is deposited in our laboratory (R&D center, BRN Science Co., Ltd., Seoul, Korea). Plant material (200 g) was extracted two times with 95% ethanol at room temperature for 2 d and was subsequently filtered. The combined filtrate was concentrated under vacuum at 60 o C, and completely dried by freeze drying. The yield was 8% and LFE powder was dissolved in DMSO at 20 mg/ml for in vitro studies and in 0.5% methylcellulose for in vivo study. LFE was identified by foenumoside B, α-linolenic acid, and linoleic acid from HPLC elution profile (Supplemental Figure 1). In brief, an HPLC system (Agilant 1200) equipped with Agilent ZORBAX Eclipse XDB-C18 column (4.6×150 mm) was used for analysis. Acetonitrile gradient (0% to 100%) was used as a mobile phase with a flow rate of 1 ml/min. The absorption peaks were detected at 205 nm and compared with authentic standards.

Nile red staining
Cells were washed with phosphate-buffered saline (PBS) three times and fixed with 3.7% formaldehyde at room temperature for 1 h. Cells were washed two times with PBS and were stained with 0.5 μg/ml Nile Red (Sigma, St. Louis, MO) for 10 min. After washing with PBS, lipid content was measured using an EnVision Multilabel Plate Reader (Perkin Elmer, Wellesley, MA). Cells were subsequently stained with Hoechst 33342 (Invitrogen, Carlsbad, CA) and photographed using an IN Cell Analyzer 1000 (GE Healthcare, Buckinghamshire, UK).

Fatty acid oxidation assay
Fatty acid oxidation assays were performed as previously described, with modifications (Kim et al., 2009). Briefly, cells were incubated in α-minimal essential medium (α-MEM; Hyclone) containing 0.1 mM of 9,10-[ 3 H] palmitate (5 μCi/ml, Perkin Elmer) and 2% bovine serum albumin for 24 h. After incubation, the medium was precipitated with an equal volume of 10% trichloroacetic acid (Sigma). Samples were centrifuged and supernatants were transferred to 1.5 ml tubes placed in a scintillation vial containing 0.5 ml of water, and incubated at 55 o C for 12 h. The tubes were removed and the 3 H2O content was measured in a scintillation counter.

Animal treatment
All experimental procedures were approved by the Seoul National University Animal Experiment Ethics Committee. Forty C57BL/6 mice (7-week-old males) were obtained from Central Laboratory Animals (Seoul, Korea). All mice were housed (four mice/cage) and given water ad libitum, with a 12 h light-dark cycle beginning at 7:00 a.m. After acclimation for one week, 8 mice were randomly assigned to one of the five treatment groups with equal mean body weight between groups. Mice in a normal diet group (ND + Vehicle) were fed a standard chow diet and mice in high-fat diet (HFD) groups were fed a HFD (60% of calories derived from fat; Research Diets Inc, New Brunswick, NJ). LFE was dissolved in 0.5% methylcellulose, and administrated to HFD + LFE groups daily at doses of 30, 100, or 300 mg/kg/day by oral gavage for 6 weeks. Mice in ND + Vehicle and HFD + Vehicle groups were given an equal volume of 0.5% methylcellulose. Food intake was monitored weekly and body weight was measured twice per week. After 6 weeks, the mice were fasted for 16 h, and euthanized using CO2. Blood was collected from abdominal vena cava, immediately mixed with EDTA, and stored on ice. Plasma was isolated by centrifugation at 3,000 rpm for 10 minutes. Plasma levels of glucose and triglycerides were analyzed on an Olympus AU400 Chemistry Auto Analyzer (Olympus, Japan) using Infinity Glucose Liquid Stable Reagent and Infinity Triglycerides Liquid Stable Reagent (Thermo Electron). The concentration of adipocytokines in the plasma was analyzed by ELISA (R&D systems, Minneapolis, MN). Organs were harvested and frozen immediately in liquid nitrogen and stored at -80 o C.

Histological analysis
The epididymal adipose tissues were fixed with 4% paraformaldehyde, embedded in optimal cutting temperature (OCT) compound (Bright Instruments, Huntingdon, UK), and frozen with cryospray (MEDITE, Germany). Fifty μm sections were obtained with a cryostat microtome and stained with hematoxylin and eosin.

Statistical analysis
Data are expressed as mean ± SD for at least two independent experiments in triplicates. Variances in different groups were calculated by Student-Newman-Keuls posthoc ANOVA analysis. P ＜ 0.05 was interpreted as statistically significant.

Supplemental data
Supplemental Data include a figure and can be found with this article online at http://e-emm.or.kr/article/article_files/ SP-43-5-5.pdf.