Effect of Cordyceps militaris extract and active constituentson metabolic parameters of obesity induced by high-fat dietin C58BL/6J mice

Seon Beom Kim a, Byeongwoo Ahn b, Myounghwan Kimb, Hyeong-Jin Ji b,Sang-Kyung Shin b, In Pyo Hong c, Chul Young Kim d, Bang Yeon Hwang a, Mi Kyeong Lee a,n

a College of Pharmacy, Chungbuk National University, Cheongju, Chungbuk 361–763, Republic of Koreab Laboratory Animal Research Center, Chungbuk National University, Cheongju, Chungbuk 361-763, Republic of Koreac National Academy of Agricultural Science, Suwon, Gyunggi 441-707, Republic of Koread College of Pharmacy, Hanyang University, Ansan, Kyeonggi 426-791, Republic of Korea

a r t i c l e i n f o

Article history:Received 23 September 2013Received in revised form28 October 2013Accepted 28 October 2013Available online 11 November 2013

Keywords:Cordyceps militarisMetabolic disordersHigh-fat dietCordyrroles A and B

a b s t r a c t

Ethnopharmacological relevance: Cordyceps species which is well-known as ‘winter worm summer grass’has long been used as tonics and stimulants to enhance energy, exhibiting a potential for energymetabolism. Clinical trials have suggested their beneficial effect on lipid metabolic disorders such ashyperlipidemia.Materials and methods: The effect of Cordyceps militaris on metabolic parameters was investigated usingC58BL/6J mice induced by high-fat diet (HFD). The effect was first determined by assessing the body andorgan weight. For further investigation, sections of epididymal adipose tissue were stained withhematoxylin and eosin and the size of epididymal adipocyte was measured by Image analysis system.Fat accumulation in frozen liver sections was assessed by the Oil Red O staining and the plasmabiochemical parameters were also assessed. Active constituents were characterized using chromato-graphic and spectroscopic analysis.Results: The administration of Cordyceps militaris extract (CE) at the dose of 100 mg/kg and 300 mg/kgreduced body weight gain and food efficiency ratio induced by HFD. The amount of epididymal fat and sizeof adipocytes were also decreased by CE treatment. In addition, liver weight and fat deposition in liverwere dramatically reduced in CE-treated group. The treatment of CE also showed beneficial effects onplasma parameters related to lipid profiles. Further study for the characterization of active constituents ofCordyceps resulted in the isolation of two new compounds such as cordyrroles A (1) and B (7) togetherwith 12 known compounds including pyrrole alkaloids and nucleotide derivatives. Among the isolatedcompounds, cordyrrole A significantly inhibited adipocyte differentiation and pancreatic lipase activity,whereas cordyrrole B was more effective at inhibiting pancreatic lipase. Cordycepin, a characteristiccompound of Cordyceps militaris, decreased the rate of adipocyte differentiation.Conclusion: Treatment of CE inhibited HFD-induced metabolic disorders, mainly by improvement inmetabolic parameters. As active constituents, pyrrole alkaloids and nucleotide derivatives were character-ized. These results suggested that Cordyceps militaris might be beneficial for the treatment of metabolicdisorders obesity through the combined actions of diverse constituents.

& 2013 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Obesity stems from a prolonged imbalance between the levelsof energy intake and expenditure, with the surplus being stored asbody lipids. Metabolic disorders such as diabetes, atherosclerosis

and liver diseases are closely related to obesity (Kopelman, 2000).Fatty liver is the abnormal accumulation of triglycerides in thecytoplasm of hepatocytes and further progresses to steatohepatitisand advanced stages of liver diseases (Dai et al., 2013). Genetic andenvironmental factors play a role in the development of obesityand the diet is known to be one of the main environmental factors.Especially, increased fat intake, in part due to Western diet, ishighly associated with body weight gain which can lead to obesityand other related metabolic diseases. For the study of obesityand the development of anti-obesity therapeutics, diverse animal

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/jep

Journal of Ethnopharmacology

0378-8741/$ - see front matter & 2013 Elsevier Ireland Ltd. All rights reserved.http://dx.doi.org/10.1016/j.jep.2013.10.064

n Correspondence to: College of Pharmacy, Chungbuk National University, 410Seongbong-ro, Heungduk-gu, Cheongju 361-763, Korea. Tel.: þ82 43 261 2818;fax: þ82 43 268 2732.

E-mail address: mklee@chungbuk.ac.kr (M.K. Lee).

Journal of Ethnopharmacology 151 (2014) 478–484

model of obesity has been developed using physiologically,genetically and pharmacologically modification (Speakman et al.,2008). Among them, high-fat diet (HFD) feeding in rodent caninduce obesity and metabolic disorders that resemble the humanmetabolic syndrome (Buettner et al., 2007). Therefore, HFDinduced animal model is widely used for the development ofanti-obesity therapeutics.

Cordyceps species has its trivial name as ‘winter worm summergrass’ in Asian country, because of its different appearance in winterand summer. Its characteristic feature, a pathogenic complex of thefungus (Cordyceps militaris which belongs to Hypocreaceae Family)and the caterpillar (Hepialus armoricanus which belongs to Hepiali-dae), enable to produce diverse skeleton of biologically active com-pounds. Cordyceps has long been used as tonics and stimulants toenhance energy, which exhibited a potential for energy metabolism(Panda and Swain, 2011). Many recent studies and clinical trials havesuggested their beneficial effect on lipid metabolic disorders and itsrelated disorders such as hyperlipidemia and hyperglycemia (Paterson,2008). Many commercial products are available in the market asneutraceuticals.

Cordyceps is well known as a rich source of biologically activecomponents. Nucleosides and polysaccharides are reported asmajor characteristic constituents of Cordyceps species. Cordycepshave been used for treatment of several diseases such as cancer,fatigue, hyposexualities (Koh et al., 2003; Lee et al., 2006; Kimet al., 2010). Related to metabolic disorders, favorable roles ofCordyceps militaris in the regulation of obesity and diabetes havebeen reported (Yun et al., 2006; Shimada et al., 2008; Liu et al.,2011). In the present study, the effect of Cordyceps militaris extract(CE) was evaluated using high-fat diet (HFD)-induced obese mice.Characterization of active constituents was also accomplished bychromatographic and spectroscopic analysis.

2. Materials and methods

2.1. Materials

Dried Cordyceps militaris samples were identified and providedby Rural Development Administration in November 2009. Cordy-ceps militaris samples were produced as previously reported(Choi et al., 1999; Sung et al., 2002; Hong et al., 2010). Briefly,hyphal body suspension of Cordyceps militaris was cultured in ricebran medium. After incubation, fruity bodies from rice branmedium were harvested. Voucher specimen was deposited in theherbarium of College of Pharmacy at Chungbuk National Univer-sity (CBNU200911-CM). For the preparation of extract, Cordycepsmilitaris was pulverized and extracted with 50% ethanol for 2 days.The filtrate was evaporated in vacuo, which resulted in Cordycepsmilitaris extract (CE) for further investigation.

2.2. In vivo anti-obesity effect of Cordyceps extract

2.2.1. Animal treatmentMale C57BL/6J mice (n¼60, 4 weeks old, 15710% g) were

purchased from Central Lab. Animal Inc., Seoul, Korea. All micewere housed in a room with controlled temperature (21–23 1C),humidity (55–60%), and lighting (12 h light/dark cycle), and givenwater ad libitum. After acclimation for 1 week, mice wererandomly divided into four groups (n¼15/group), normal diet(ND), high-fat diet control (HFD), HFDþCE100 (100 mg/kg of CE),HFDþCE300 (300 mg/kg of CE). Normal diet (ND) group was fedwith normal diet, and HFD, HFDþCE100 and HFDþCE300 groupswere fed with high fat diet (Table 1) ad libitum. CE sample wasdiluted in distilled water (DW) and administered orally to mice toHFDþCE100 and HFDþCE300 groups, and DW was administered

orally to ND and HFD groups. At the end of experiment, mice wereanesthetized and blood, organs were collected. The protocol forthis study was approved by the Animal Care and Use Committee ofChungbuk National University (Approved no. CBNUA-370-11-01).

2.2.2. HistopathologyHistological photograph of adipose tissue was analyzed based

on the paraffin method using a light microscope. Epididymaladipose tissue was fixed with 10% neutral buffered formalin andembedded in paraffin block. Six μm sections were cut andmounted on glass slide. Paraffin was removed with xylen andalcohol. The sections were then stained with hematoxylin andeosin (H&E). After dehydration by alcohol, the photograph wastaken with light microscope. The size of epididymal adipocyte wascalculated by Image analysis system (IPKR-1003, Saramsoft Co.,Ltd., Korea). For the detection of lipid deposition in liver, liversection were prepared from frozen liver and stained with Oil Red Oas previously reported (Fowler and Greenspan, 1985).

2.2.3. Serum biochemical parametersAll mice were fasted for 12 h before sacrifice. Blood was

collected and serum was obtained by the centrifugation at3000 rpm for 10 min at 4 1C. The content of triglyceride (TG), totalcholesterol (TC), low-density lipoprotein cholesterol (LDL), high-density lipoprotein cholesterol (HDL), alanine transaminase (ALT),aspartate transaminase (AST) and alkaline phosphatase (ALP) weredetermined using Hitachi7080 analyzer.

2.3. Characterization of active constituents

2.3.1. Extraction and isolation of compounds.Cordyceps (1.5 kg) was extracted twice with 50% EtOH, which

yielded the ethanol extract (93.5 g). The ethanol extract wassuspended in H2O and partitioned successively with CH2Cl2,EtOAc, and n-BuOH. The CH2Cl2 fraction (18.9 g) was subjectedto medium pressure liquid chromatography (MPLC) over silica geland eluted with hexane-EtOAc–MeOH to give 22 subfractions(CORM1–CORM22). CORM16 was subjected to column chromato-graphy over Sephadex LH-20 and eluted with CH2Cl2–MeOHto yield four subfractions (CORM16A–CORM16D). Compounds 3(1.9 mg) and 4 (4.0 mg) were obtained from CORE6 and CORM8,

Table 1Composition of experimental diets.

Unit: g% ND HFD

FormulaProtein 20 24Carbohydrate 64 41Fat 7 24

IngredientCasein 20 23.3L-Cystein 0.3 0.3Corn starch 39.7 8.5Maltodextrin 13.2 11.7Sucrose 10 20.1Cellulose 5 5.8Soybean oil 7 2.9Lard 0 20.7Cholesterol 0 0.5Mineral mixture 3.5 1.2Dicalcium phosphate 0 1.5Calcium carbonate 0 0.6Potassium citrate 0 1.9Vitamin mixture 1 1.2Choline bitartrate 0.3 0.2

S.B. Kim et al. / Journal of Ethnopharmacology 151 (2014) 478–484 479

respectively, by semi-preparative high pressure liquid chromato-graphy (HPLC) and eluted with MeCN-water and MeOH-water.

The n-BuOH fraction (67.9 g) was subjected to HP-20 columnchromatography with mixture of MeOH-water (0, 20, 40, 60, 80,and 100% MeOH in water) to give six fractions (CORB1–CORB6).The CORB2 fraction was subjected to silica gel MPLC and elutedwith CH2Cl2–MeOH to yield ten subfractions (CORB2A–CORB2J).Compound 5 (3.0 mg) was obtained from CORB2C3 by semi-preparative HPLC and eluted with MeOH-water. Compounds 12(1.5 mg), 13 (2.1 mg), and 14 (2.5 mg) were obtained from COR2Gby column chromatography over Sephadex LH-20 eluted withMeOH, followed by semi-preparative HPLC eluted with MeOH-water. Compound 8 (2.4 mg) and 11 (3.4 mg) were obtained fromCORB2I by column chromatography over Sephadex LH-20 elutingwith MeOH. Compound 9 (1.4 mg) was obtained from CORB2J bysemi-preparative HPLC eluting with MeOH-water.

The CORB3 fraction was subjected to silica gel MPLC and elutedwith CH2Cl2–MeOH to yield 12 fractions (CORB3A–CORB3L). Com-pound 2 (1.7 mg) was obtained from CORB3A by column chromato-graphy over Sephadex LH-20 eluted with MeOH, followed bysemi-preparative HPLC eluted with MeOH-water. Compound 1(1.6 mg) was obtained from CORB3C by column chromatographyeluted with MeOH. Compound 10 (6.3 mg) was obtained fromCORB3D by column chromatography over Sephadex LH-20 elutedwith MeOH, followed by semi-preparative HPLC eluted withMeOH-water.

2.3.2. Spectroscopic data of the new compoundsCordyrrole A (1): Light brown amorphous gum; α½ �25D �177.01 (c

0.08, MeOH); UV (MeOH) λmax 294 nm; IRmax 3345, 1645 cm�1; 1HNMR (CD3OD, 400 MHz) δ 9.34 (1H, s, H-1), 7.07 (1H, d, J¼4.0 Hz,H-3), 6.29 (1H, d, J¼4.0 Hz, H-4), 5.06 (1H, dd, J¼11.2, 6.8 Hz,H-1ʹ), 4.66 (2H, d, J¼2.0 Hz, H-6), 3.65 (1H, m, H-4ʹa), 3.35 (1H, m,

H-4ʹb), 2.27 (2H, m, H-2ʹ), 2.01 (2H, m, H-3ʹ) ppm; 13C NMR(CD3OD, 100 MHz) δ 178.8 (C-1), 170.5 (C-6ʹ), 143.9 (C-2), 132.0(C-5), 125.8 (C-3), 109.8 (C-4), 56.5 (C-1ʹ), 55.4 (C-6), 41.5 (C-4ʹ),28.8 (C-2ʹ), 22.0 (C-3ʹ) ppm; ESIMS (negative mode) m/z: 221[M�H]�; HRESIMS (positive mode) m/z: 245.0911 (calc forC11H14N2O3 245.1004).

Cordyrrole B (7): White amorphous powder; UV (MeOH) λmax

265 nm; IRmax 3316, 2927, 1730, 1617 cm�1; 1H NMR (CD3OD,400 MHz) δ 8.28 (1H, s, H-7), 8.26 (1H, s, H-2), 5.98 (1H, d,J¼7.6 Hz, H-1″), 4.76 (1H, dd, J¼5.2, 6.4 Hz, H-2″), 4.35 (1H, m,H-3″), 4.32 (2H, t, J¼5.6 Hz, H-2ʹ), 4.19 (1H, dd, J¼2.5, 5.2 Hz, H-4″a), 3.91 (2H, m, H-1ʹ), 3.90 (1H, dd, J¼2.4, 12.8 Hz, H-5″), 3.76 (1H,dd, J¼2.4, 12.8 Hz, H-4″b), 2.05 (3H, s, COCH3) ppm; 13C NMR(CD3OD, 100 MHz) δ 152.0 (C-2), 148.2 (C-4), 115.6 (C-5), 154.9(C-6), 140.2 (C-7), 39.1 (C-1ʹ), 62.8 (C-2ʹ), 171.4 (COCH3), 19.3 (COCH3),89.8 (C-1″), 74.0 (C-2″), 71.2 (C-3″), 86.8 (C-4″), 62.1 (C-5″) ppm;ESIMS (positive mode) m/z: 354 [MþH]þ; HRESI-TOF-MS (positivemode) m/z: 354.1408 (calc for C14H19N5O6 354.1414).

2.4. In vitro activity of compounds

2.4.1. Pancreatic lipase activityPancreatic lipase inhibitory activity was evaluated using a

method reported previously (Kim et al., 2012). The enzymesolution was prepared by reconstituting porcine pancreatic lipase,and pancreatic lipase activity was determined by measuring thehydrolysis of p-NPB to p-nitrophenol at 405 nm using a microplatereader. Relative pancreatic lipase activity (%) was calculated as(activity of compound with substrate – negative control ofcompound without substrate)/(activity of without compoundand with substrate – negative control of without compound andsubstrate)�100.

Fig. 1. Effect of CE100 and CE300 on body weight (A), FER (C) and organ weight (C) in HFD-induced obese mice. FER was calculated as total weight gain/total food intake.Results are expressed as the mean7SE (n¼12–15). npo0.05 compared with ND group, #po0.05 compared with HFD group.

S.B. Kim et al. / Journal of Ethnopharmacology 151 (2014) 478–484480

2.4.2. Adipocyte differentiation3T3-L1 mouse embryo fibroblasts (ATCC CL-173) were obtained

from the American Type Culture Collection (Manassas, VA). The cellswere stimulated to differentiate with differentiation medium con-taining DMEM with 10% FBS, 0.5 mM 3-isobutyl-1-methyl-xanthine,1 μM insulin and 1 μM dexamethasone for 2 days (day 2). Cells werethen maintained in DMEM supplemented with 10% fetal bovineserum (FBS) and 1 μM insulin for another 2 days (day 4), followed byculturing with DMEM with 10% FBS for an additional 4 days (day 8).The cultures were treated with 100 μM isolated compounds for theentire culture period (days 0–8). Lipid droplets in cells were stainedwith Oil Red O. The Oil Red O stain was dissolved in isopropyl alcoholand optical density was measured at 520 nm using ELISA plate readerfor the quantitative analysis (Choi et al., 2011).

2.5. Statistical analysis

The evaluation of statistical significance was determined byLevene's test followed by one-way ANOVA or Turkey HSD-testwith a value of po0.05 considered to be statistically significant.

3. Results and discussion

3.1. In vivo effect of the CE on HFD-induced obesity

The effect of CE on obesity was investigated using HFD-inducedmale C58BL/6J mice. Diet-induced obesity in rodents has beenused as an animal model to investigate environmental effect. HFD-fed rodents become obese and show distinctive symptoms such asincrease of adipose tissue, disturbance of lipid metabolism, hyper-insulinemia and fatty liver, which are typically associated withhuman obesity (Sclafani and Springer, 1976). In our study, the body

weight of HFD group was significantly increased compared tothat of ND after 2 weeks diet. However, the body weight gains ofHFDþCE100 and HFDþCE300 groups were significantly reducedcompared to HFD groups, after 21 days and 10 days treatment,respectively (Fig. 1A). On day 42, at the end of experiment, thebody weight gains of CE100 [4.29 g] and CE300-treated groups[4.01 g] were reduced dramatically compared to HFD group[8.39 g]. The relative ratio of epididymal fat was also significantlyincreased in HFD group compared to ND group (Fig. 2A). Treat-ment of CE100 and CE300 significantly reduced the relative ratioof epididymal fat in dose related manner. The liver weight in HFDgroup was significantly increased compared to ND group in ourstudy. However, the increase of liver weight induced by highfat diet was significantly attenuated in CE100 and CE300 groupscompared to HFD group. The treatment of CE100 and CE300,however, did not affect the relative weight spleen and heartwhereas relative weight of kidney was slightly reduced (Fig. 1C).Food intake was higher in the ND group than in HFD group, but didnot differ significantly among the HFD, HFDþCE100 andHFDþCE300 groups. However, food efficiency ratio (FER) ofHFDþCE100 and HFDþCE300 groups were significantly reducedcompared to HFD group (Fig. 1B).

High ratio of fat consumption accompanies excessive growthof adipose tissue in both cell number and cell size, and conse-quently induces fat accumulation. As shown in Fig. 2A, the size ofadipocyte in epididymal tissue was greatly increased in HFDgroup, as analyzed by image analyzer after H&E staining. However,the size of adipocyte in CE100 and CE300-treated groups wasdramatically decreased compared to HFD (Fig. 2B). HFD diet alsoinduced fat accumulation in liver, which was easily observedby Oil Red O staining (Fig. 2C). However, treatment of CE100and CE300 greatly reduced fat accumulation in liver. Especially, fataccumulation in HFDþCE300 group was almost completely

Fig. 2. Effect of CE100 and CE300 on lipid parameters in HFD-induced obese mice. (A) Sections of epididymal adipose tissue stained with H&E (magnification, 200� ).(B) Quantitation of the size of epididymal adipocyte by Image analysis system. (C) Fat accumulation in frozen liver sections stained with Oil Red O. (D) Plasma biochemicalparameters. Data are expressed as mean7SD (n¼12–15 per group). ALP, alkaline phosphatase (IU/l); AST, aspartate transaminase (IU/l); ALT, alanine transaminase (IU/l);HDL, high-density lipoprotein (mg/dl); LDL, low-density lipoprotein (mg/dl); TC, total cholesterol (mg/dl); TG, triglyceride (mg/dl); npo0.05 compared with ND group,#po0.05 compared with HFD group.

S.B. Kim et al. / Journal of Ethnopharmacology 151 (2014) 478–484 481

reduced comparable to ND group. These result shows that CE100and CE300 efficiently inhibit fat accumulation in liver and epidi-dymal adipocyte tissues.

Persistent high fat and cholesterol diets cause damage to liver,which results in change of enzymes and lipid profiles. HFD causedelevation of plasma level of ALP, AST and ALT, which was recoveredin HFDþCE100 and HFDþCE300 groups. In addition, HFD-inducedincreases of TC, LDL and TG were significantly decreased in CE100and CE300-treated groups. HDL, however, was decreased in HFDgroup, which was significantly recovered in CE100 and CE300-treated group (Fig. 2D).

3.2. Structural elucidation

3.2.1. Structural determination of new compoundsCompound 1 was obtained as a pale brown amorphous powder

with a molecular formula of C11H14N2O3, determined by HRESI-MS(m/z 245.0910 [MþNa]þ; calc 222.1004). The IR spectrum showedthe presence of an amine group (3345 cm�1) and carbonyl group(1645 cm�1). The maximum UV absorption was observed at293 nm, which is characteristic of a pyrrole alkaloid (Chin et al.,2003). The presence of pyrrole ring was also supported by thecharacteristic J value of the pyrrole signals at δH 7.08 (1H, d,J¼4.0 Hz), 6.29 (1H, d, J¼4.0 Hz) together with four carbon signalsat δC 143.9, 132.0, 125.8, and 109.8 in the 1H- and 13C NMR spectra.The presences of aldehydes and hydroxymethyl moieties wereeasily deduced from the signals at [δH 9.34 (1H, s) and δC 178.8]and [δH 4.66 (2H, d, J¼2.0) and δC 55.4], respectively, in the HSQCspectrum. The HSQC Spectrum of 1 also showed a methine signalat [δH 5.06 (1H, dd, J¼11.2 6.8) and δC 56.5] and three methylenesignals at [δH 2.27 (2H, m) and δC 28.8], [δH 2.01 (2H, m) and δC22.0] and [δH 3.65 (1H, m), 3.35 (1H, m) and δC 41.5]. An additionalcarbonyl signal was observed at δC 170.5 in the 13C NMR spec-trum,. The HMBC correlations were observed between δH 3.65

(H-4′a) and δC 22.0 (C-3′), between 3.35 (H-4′b) and δC 28.8 (C-2′),170.5 (C-6′), and between δH 5.06 (H-1′) and δC 28.8 (C-2′), 170.5(C-6′), suggesting the presence of a 2-piperidone ring (Ghosh et al.,2009). The position of the 2-peperidone ring was determined to beN-1 by the correlation between at δH 5.06 (H-1′) and δC 132.0(C-5), 143.9 (C-2) in the HMBC spectrum. The positions of thealdehyde and hydroxymethyl moiety were C-2 and C-5, respec-tively, by the HMBC correlation between δH 9.34 (H-1) and δC143.9 (C-2), and between δH 4.66 (H-6) and δC 132.0 (C-5). Basedon these data, the structure of 1was determined as shown in Fig. 3and named cordyrrole A.

Compound 7 was obtained as a white amorphous powder witha molecular formula of C14H19N5O6, determined by HRESIMS (m/z354.1408 [MþH]þ; calcd 354.1414). The 1H- and 13C NMR spectraof 7 showed characteristic signals for adenosine, a major consti-tuent of Cordyceps (Furuya et al., 1983). The 1H and 13C NMR of 7also showed two methylene signals at [δH 3.91 (2H, m) and 4.32(2H, t, J¼5.6 Hz)] and [δc 39.1 and 62.8], which was very similar tothose of N6-(2-hydroxyethyl)adenosine.18However, additional sig-nals for an acetyl moiety were observed at [δH 2.05 (3H, s), δC 19.3and 171.4] in the HMBC spectrum, which suggested 7 as anacetylated form of N6-(2-hydroxyethyl)adenosine. The position ofan acetyl moiety was determined from the HMBC correlationbetween δC 171.4 (COCH3) and δH 4.32 (H-2ʹ). Taken together,the structure of 7 was determined as shown and was namedcordyrrole B.

3.2.2. Identification of known compoundsTwelve known compounds were identified, by the spectroscopic

data analysis and comparison with literature values including 5-(hydroxymethyl)-1-(2-oxopiperidin-3-yl)-1H-pyrrole-2-carbaldehyde(2), dihydrouracil (3), uracil (4), nicotinamide (5), N6-(2-hydroxyethyl)adenosine (6), cordycepin (8), adenosine (9), 2′-O-methyladenosine(10), xanthosine (11), 2′-deoxyuridine (12), uridine (13), thymine (14)

Fig. 3. (A) Structures of compounds 1–14 from Cordyceps militaris and (B) key HMBC correlations of new compounds, cordyrroles A (1) and B (7).

S.B. Kim et al. / Journal of Ethnopharmacology 151 (2014) 478–484482

(Furuya et al., 1983; Jiang, Gerwick, 1991; Beigelman et al., 2000;Jurczyk et al., 2000; Zhu et al., 2011; Zhang, Xu, 2011). (Fig. 3)

3.3. In vitro activity of isolated compounds

The effects of the isolated compounds on adipocyte differentiationwere investigated using 3T3-L1 preadipocytes as an assay system.Among the isolated compounds, compounds 1 and 8 significantlyreduced fat accumulation in differentiated 3T3-L1 cells (Fig. 4). Theinhibitory effects of isolated compounds on pancreatic lipase wereinvestigated employing porcine pancreatic lipase and p-nitrophenylbutyrate as the substrate. Among the isolated compounds, two newcompounds 1 and 7 showed the most significant inhibitory effects(Fig. 4). Other compounds also significantly inhibited pancreatic lipaseactivity.

3.4. Effect of Cordyceps and its compounds on lipid metabolism

In our present study, increase of fat accumulation and abnormallipid metabolismwere observed in HFD group, which was improved inHFDþCE100 and HFDþCE300 groups. Especially, the liver weight andfat accumulation in liver were greatly increased in HFD groups, whichwas reduced in HFDþCE100 and HFDþCE300 groups comparable toND group. Therefore, treatment of CE100 and CE300 inhibited fataccumulation into liver, which resulted in decrease of liver weight.HFD-induced increase of epididymal fat weight and adipocyte sizewere also improved in HFDþCE100 and HFDþCE300 groups. Inaddition, biochemical parameters related to lipid metabolism, suchas TC, HDL, LDL and TG were also recovered in HFDþCE100 andHFDþCE300 groups. Taken together, treatment of CE100 and CE300mainly act on inhibition of fat accumulation, which further resulted indecrease in liver weight and lipid profiles in blood. Therefore, CEmight be effective in liver dysfunction induced by HFD, which wassupported by the improvement of ALP, AST and ALT parameters inHFDþCE100 and HFDþCE300 groups.

Taken together, oral administration of CE at 100 and 300 mg/kgdramatically improved many parameters of HFD-induced obesity.However, food intake was not affected by the treatment of CE100and CE300 groups, which resulted in the decrease of FERs in theHFDþCE100 and HFDþCE300 groups (Fig. 2B). Therefore, thedecrease in body weight in the HFDþCE100 and HFDþCE300groups was achieved partially by a decrease in FER not by a loss ofappetite. Compounds isolated from Cordyceps, including two newcompounds, inhibited pancreatic lipase activity (Fig. 2). Therefore,treatment of CE reduced fat absorption by the inhibition ofpancreatic lipase. These compounds also reduced fat accumulation

in the adipocyte, which might have disturbed fat accumulation inthe liver induced by HFD. Collectively, inhibition of fat absorptionand fat accumulation by the constituents of CE are responsible forthe reduction of fat accumulation in liver and adipocyte, whichleads to recover liver function and lipid metabolism.

4. Conclusion

Conclusively, two new compounds, cordyrroles A and B, wereisolated from Cordyceps together with twelve known compounds.They were effectively inhibited pancreatic lipase activity andadipocyte differentiation. Cordyceps extract was also improvedmany metabolic parameters in high-fat diet animal model. Theseresults suggested that CE might be beneficial for the preventionand/or treatment of obesity and related metabolic disorders.

Acknowledgements

This study was supported by a grant from the Korean RuralDevelopment Administration (Agenda Program, PJ0067792010),and by Basic Science Research Program (2012-0008023) andMedical Research Center Program (2010-0029480) through theNational Research Foundation of Korea.

Appendix A. Supplementary material

Supplementary data associated with this article can be found inthe online version at http://dx.doi.org/10.1016/j.jep.2013.10.064.

References

Beigelman, L., Haeberli, P., Sweedler, D., Karpeisky, A., 2000. Improved syntheticapproaches toward 2mpO-methyl-adenosine and guanosine and their N-acylderivatives. Tetrahedron 56, 1047–1056.

Buettner, R., Schölmerich, J., Bolheimer, C., 2007. High-fat diets: modeling themetabolic disorders of human obesity in rodents. Obesity 15, 798–808.

Chin, Y.W., Lim, S.W., Kim, S.H., Shin, D.Y., Suh, Y.G., Kim, Y.B., Kim, Y.C., Kim, J., 2003.Hepatoprotective pyrrole derivatives of Lycium chinensis fruits. Bioorg. Med.Chem. Lett. 13, 79–81.

Choi, I.Y., Choi, J.S., Lee, W.H., Yu, Y.J., Joung, G.T., Ju, I.O., Choi, Y.K., 1999. Thecondition of production of artificial fruiting body of Cordyceps militaris. KoreanJ. Mycol. 27, 243–248.

Choi, K.M., Shin, E., Liu, Q., Yoo, H.S., Kim, Y.C., Sung, S.H., Hwang, B.Y., Lee, M.K.,2011. Hydroxyframoside B, a secoiridoid of Fraxinus rhynchophylla, inhibitsadipocyte differentiation in 3T3-L1 cells. Planta Med. 77, 1020–1023.

Dai, F.J., Hsu, W.H., Huang, J.J., Wu, S.C., 2013. Effect of pigeon pea (Cajanus cajan L.)on high-fat diet-induced hypercholesterolemia in hamsters. Food Chem.Toxicol. 53, 384–391.

Fowler, S.E., Greenspan, P., 1985. Application of Nile red, a fluorescent hydrophobicprobe, for the detection of neutral lipid deposits in tissue section: comparisonwith Oil Red O. J. Histochem. Cytochem. 33, 833–836.

Furuya, T., Hirotani, M., Matsuzawa, M., 1983. N6-(2-hydroxyethyl)adenosine, abiologically active compound from cultured mycelia of Cordyceps and Isariaspecies. Phytochemistry 22, 2509–2512.

Ghosh, S.C., Muthaiah, S., Zhang, Y., Xu, X., Hong, S.H., 2009. Direct amide synthesisfrom alcohols and amines by phosphine-free ruthenium catalyst systems. Adv.Synth. Catal. 351, 2643–2649.

Hong, I.P., Kang, P.D., Kim, K.Y., Nam, S.H., Lee, M.Y., Choi, Y.S., Kim, N.S., Kim, H.K.,Lee, K.G., Humber, R.A., 2010. Fruit body formation on silkworm by Cordycepsmilitaris. Mycobiology 38, 128–132.

Jiang, Z.D., Gerwick, W.H., 1991. Novel pyrroles from the oregon red algagracilariopsis lemaneiformis. J. Nat. Prod. 54, 403–407.

Jurczyk, S.C., Horlacher, J., Devined, K.G., Benner, S.A., Battersby, T.R., 2000.Synthesis and characterization of oligonucleotides containing 2′-deoxyxantho-sine using phosphoramidite chemistry. Helv. Chim. Acta 83, 1517–1524.

Kim, E.S., Lee, K.J., Oh, K.H., Ahn, J.H., Kim, S.B., Liu, Q., Hwang, B.Y., Lee, M.K., 2012.Inhibitory effects of marine algae extract on adipocyte differentiation andpancreatic lipase activity. Nat. Prod. Sci. 18, 153–157.

Kim, H.S., Kim, J.Y., Kang, J.S., Kim, H.M., Kim, Y.O., Hong, I.P., Lee, M.K., Hong, J.T.,Kim, Y., Han, S.-B., 2010. Cordlan polysaccharide isolated from mushroomCordyceps militaris induces dendritic cell maturation through toll-like receptor4 signaling. Food Chem. Toxicol. 48, 1926–1933.

Fig. 4. Effects of isolated compounds on adipocyte differentiation and pancreaticlipase. Activity was measured at 100 μM and expressed as mean7SD of threeindependent experiments each performed in triplicate. The optical density ofcontrols in adipocyte differentiation and pancreatic lipase were 0.93470.059and 0.47770.024, respectively. npo0.05 compared with control.

S.B. Kim et al. / Journal of Ethnopharmacology 151 (2014) 478–484 483

Koh, J.H., Kim, K.M., Kim, J.M., Song, J.C., Suh, H.J., 2003. Antifatigue and antistresseffect of the hot-water fraction from mycelia of Cordyceps sinensis. Biol. Pharm.Bull. 26, 691–694.

Kopelman, P.G., 2000. Obesity as a medical problem. Nature 404, 635–643.Lee, H., Kim, Y.J., Kim, H.W., Lee, D.H., Sung, M.K., Park, T., 2006. Induction of

apoptosis by Cordyceps militaris through activation of caspase-3 in leukemiaHL-60 cells. Biol. Pharm. Bull. 29, 670–674.

Liu, Q., Hong, I.P., Ahn, M.-J., Yoo, H.-S., Han, S.-B., Hwang, B.Y., Lee, M.K., 2011. Anti-adipogenic activity of Cordyceps militaris in 3T3-L1 cells. Nat. Prod. Commun. 6,1839–1841.

Panda, A.K., Swain, K.C., 2011. Traditional uses and medicinal potential of Cordycepssinensis of Sikkim. J. Ayurveda Integr. Med. 2, 9–13.

Paterson, R.R.M., 2008. Cordyceps—A traditional Chinese medicine and anotherfungal therapeutic biofactory? Phytochemisty 69, 1469–1495.

Sclafani, A., Springer, D., 1976. Dietary obesity in adult rats: similarities tohypothalamic and human obesity syndromes. Physiol. Behav. 17, 461–471.

Shimada, T., Hiramatsu, N., Kasai, A., Mukai, M., Okamura, M., Yao, J., Huang, T.,Tamai, M., Takahashi, S., Nakamura, T., Kitamura, M., 2008. Suppression of

adipocyte differentiation by Cordyceps militaris through activation of the arylhydrocarbon receptor. Am. J. Physiol. Endocrinol. Metab., Endocrinology andMetabolism 295, E859–E867.

Speakman, J., Hambly, C., Mitchell, S., Król, E., 2008. The contribution of animalmodels to the study of obesity. Lab. Anim. 42, 413–432.

Sung, J.M., Choi, Y.S., Shrestha, B., Park, Y.J., 2002. Investigation on artificial fruitingof Cordyceps militaris. Korean J. Mycol. 30, 6–10.

Yun, Y., Han, S., Lee, S., Ko, S., Lee, C., Ha, N., Kim, K., 2006. Anti-diabetic effects ofCCCA, cmESS, and cordycepin from Cordyceps militaris and the immuneresponses in streptozotocin-induced diabetic mice. Nat. Prod. Sci. 9, 291–298.

Zhang, X., Xu, Y.Z., 2011. NMR and UV Studies of 4-thio-2treptozotocin-inducitsderivatives. Molecules 16, 5655–5664.

Zhu, L., Liang, Y., Lao, D., Zhang, T., Itoc, Y., 2011. Preparative separation of high-purity cordycepin from Cordyceps militaris (L.) link by high-speed counter-current chromatography. J. Liq. Chromatogr. Related Technol. 37, 491–499.

S.B. Kim et al. / Journal of Ethnopharmacology 151 (2014) 478–484484