Az

YIS

a

ARRAA

KODSAM

1

iodih

AGfie

0d

Journal of Ethnopharmacology 126 (2009) 543–550

Contents lists available at ScienceDirect

Journal of Ethnopharmacology

journa l homepage: www.e lsev ier .com/ locate / je thpharm

cute toxicity and sub-chronic toxicity of steroidal saponins from Dioscoreaingiberensis C.H.Wright in rodents

uan Qin1, Xiaohua Wu1, Wen Huang ∗, Guohua Gong, Dan Li, Yang He, Yinlan Zhaonstitute for Nanobiomedical Technology and Membrane Biology, State Key Lab of Biotherapy of Human Diseases, Cancer Center, West China Hospital, West China Medical School,ichuan University, Keyuan 4 Road No. 1, Gaopeng Avenue, Gaoxin District, Chengdu 610041, China

r t i c l e i n f o

rticle history:eceived 4 May 2009eceived in revised form 27 August 2009ccepted 30 August 2009vailable online 6 September 2009

eywords:ral toxicityioscorea zingiberensis C.H.Wrightteroidal saponinscute and sub-chronic dosingetabolite

a b s t r a c t

Ethnopharmacological relevance: Steroidal saponins from Dioscorea zingiberensis are widely used in Chinafor curing cardiovascular diseases. However, there was little toxicological information available on them.Aim of the study: The study evaluated potential toxicity of the steroidal saponins and analyzed themetabolites in rats.Materials and methods: For the acute study, the steroidal saponins were administered to kunming micein single doses of 112.5–9000 mg/kg given by gavage. General behavior adverse effects, mortality andliver histopathological changes were examined. For the sub-chronic toxicity study, Sprague–Dawley ratswere gavaged at doses of 127.5, 255 and 510 mg/kg/day for 30 days, then examined the biochemical andhematological parameters. Metabolites in serum were analyzed by HPLC–MS.Results: The steroidal saponins caused dose-dependent general behavior adverse effects, mortalityand liver histopathological changes in the acute toxicity study. In the sub-chronic toxicity study,510 mg/kg/day of steroidal saponins increased total bilirubin (TBIL) in serum and decreased proteincontent in liver significantly. The metabolic process of TBIL in liver includes TBIL intaking, conjugatedbilirubin forming, conjugated bilirubin excreting to biliary passage. Treatment with high dose of thesteroidal saponins in vivo may lead to vacuolization of the cytoplasm of hepatocytes and canalicularcholestasis. In all doses used in the experiment, the steroidal saponins decreased aspartate aminotrans-ferase (GOT), alanine aminotransferase (ALT) and alkaline phosphatase (AKP) in serum and increased

reduced glutathione hormone (GSH), glutathione reductase (GR) and glutathione S-transferases (GST) inliver. Diosgenin was the main metabolite in serum.Conclusions: The steroidal saponins did not show any sign of toxicity up to oral dose of 562.5 mg/kg in mice.No significant changes of biochemical and hematological parameters in rats (except at 510 mg/kg/day), it

teroid

was concluded that the suses.

. Introduction

Dioscorea zingiberensis, belonged to Dioscoreaceae, Dioscorea L.,s widely distributed in China (Yang et al., 2007). The rhizome

f Dioscorea zingiberensis is extensively used for the extraction ofiosgenin sapogenin and its glycoside dioscin, for use as steroid

ntermediate in the pharmaceutical industry to synthesize steroidormone such as cortical hormone, sex hormone and progestogen

Abbreviations: GOT, aspartate aminotransferase; ALT, alanine aminotransferase;KP, alkaline phosphatase; TBIL, total bilirubin; GSH, reduced glutathione hormone;R, glutathione reductase; GST, glutathione S-transferases; GSSG, glutathione disul-de; NOAEL, no-observed adverse effect value; LOAEL, lowest-observed adverseffect value; DX, Di-ao-xin-xue-kang.∗ Corresponding author. Tel.: +86 028 85164076; fax: +86 028 85164073.

E-mail address: huangwen@scu.edu.cn (W. Huang).1 Both authors contribute equally to this work.

378-8741/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved.oi:10.1016/j.jep.2009.08.047

al saponins did not appear to have significant toxicity in their traditional

© 2009 Elsevier Ireland Ltd. All rights reserved.

(Nie et al., 2004; Zan et al., 2005; Wang et al., 2006). Steroidalsaponins are widely spread throughout the plant kingdom, suchas Dioscorea zingiberensis, which was first documented in Pharma-copoeia of People’s Republic of China in the 2000 edition (China,2000).

Steroidal saponins have been reported a variety of medicinaluses. Clinical research showed that steroidal saponins could obvi-ously increase coronary flow (MBF), nutrite the cardiac muscle,improve peripheral circulation and depress platelet aggregation,as well as decrease cholesterol and triglyceride in blood (Sun et al.,2002). Di-ao-xin-xue-kang capsule containing 35% total steroidalsaponins, a commercial production, which was prepared from

Dioscorea nipponica Makino, has been used to cure coronary heartdisease for more than 10 years in China (Liu et al., 2004).

Studies on metabolite of steroidal saponins in vivo werereported recently. Shen et al. had studied metabolites of Ophio-pogon japonicus saponin in vivo and detected diosgenin in blood by

5 pharm

H(Ibms

ahaccewhgotii

2

2

fato

2d

prptoce

(fahwddwcwtvc(aArFwt

z

poration, Milford, USA) was equipped with quaternary pump,

44 Y. Qin et al. / Journal of Ethno

PLC–MS (Shen et al., 2006). Metabolites of Di-ao-xin-xue-kangDX) by intestinal bacteria in vivo and vitro had been studied.n vitro, DX was decomposed easily by rat and human intestinalacteria and various metabolites were found. In vivo, the mainetabolite with molecular weight of 415.3 was shown to be corre-

ponding to diosgenin in rat’s blood (Ma et al., 2002).Notwithstanding the widespread use of traditional Chinese drug

bundant of steroidal saponins, few studies on their toxic effectsave been reported. Oral administration of Di-ao-xin-xue-kang forlong time could cause serious acute icteric hepatitis in its clini-

al applications (Zhou et al., 1999). Meanwhile, steroidal saponinsan also induce adverse reactions such as cardiopalmus and gen-ral weakness and so on (Zhang and Zuo, 2007). The present studyas carried out to determine the biochemical, hematological andistopathological toxicity of steroidal saponins from Dioscorea zin-iberensis rhizome after acute oral dosing in mice and sub-chronicral administration in rats, with the aim to obtain information onhe safety of Dioscorea zingiberensis and provide guidance for select-ng a safe dose of steroidal saponins from Dioscorea zingiberensis ints use.

. Materials and methods

.1. Plant material

The fresh rhizomes of Dioscorea zingiberensis were purchasedrom Tianhe Pharmaceutical Company (Hubei, China). They wereuthenticated by the Hubei Province Institute of Drug Identifica-ion. The voucher specimens have been deposited at the Herbariumf the Jiangshu botanical garden.

.2. Preparation of the steroidal saponins and diosgenin contentetermination

The fresh rhizomes of Dioscorea zingiberensis were cut into smallieces and ground in a grinding machine. The air-dried crushedhizomes (600 g) were soaked for 12 h in ethanol (2 L) at room tem-erature. The residue was extracted with hot EtOH under reflux 3imes (each 1200 ml) after vacuum filtration. All solvent was evap-rated under vacuum and extract was kept in glass bottles. Theoncentrate was suspended into H2O for freeze drying got ethanolxtracts (Zhang and Lu, 2006).

Diosgenin was determined as described by Baccou et al.1977), with some modification. Standard diosgenin was purchasedrom Sigma. The diosgenin level was determined by measuringbsorbance at 454 nm, based on the color reaction with anisalde-yde glacial acetic acid and perchloric acid. In brief, 2.0 g extractas taken to get total diosgenin by above method. Diosgenin wasissolved with 25 ml chloroform and taken out of 20 �l volumeiosgenin solutions to new color comparison tube. The chloroformas evaporated under 70 ◦C; 0.2 ml of 5% vanillin and 0.8 ml of per-

hloric acid were added to the tube. The test tube was placed in aater bath maintained at 70 ◦C for 15 min to develop color fully,

hen allowed to cool for 2 min in 0 ◦C water bath and meteredolume to 25 ml with glacial acetic acid. The absorbance of theolor developed solution was measured in a spectrophotometerDU800, BECKMEN, USA) after stabilized for 30 min. Glacial aceticcid was used as a control for the measurement of absorbance.s a reagent blank, 0.2 ml of 5% vanillin and 0.8 ml of perchlo-

ic acid were placed in a tube and assayed in similar manner.or the calibration curve, 1–40 �g standard diosgenin in 1 mlas used. Each sample was repeated thrice and the average was

aken.The diosgenin content of the steroidal saponins from Dioscorea

ingiberensis was 28.34% (w/w) of lyophilized power.

acology 126 (2009) 543–550

2.3. Treatment of animals

For the acute toxicity study, male and female adult kunmingmice had an average weight of 23 ± 2 g. They were obtained fromthe animal colony of department, which was approved by theSichuan Animal Care and Use Committee. Upon arrival at the lab-oratory, the animals were acclimatized in cages under standardenvironmental conditions of light/dark cycles (12 h/12 h), temper-ature (23 ± 1 ◦C), with 55 ± 5% of humidity and air changes. Animalshad free access to tap water and standard pellet diet. They were feda week before experiment. Mice were randomly divided into eightgroups (10 mice/group). The steroidal saponins were administeredto mice by gavage (n = 10 in each group) at doses of 112.5, 225,562.5, 1125, 2250, 4500 and 9000 mg/kg, while the control groupreceived the water only.

For the chronic study, adult SD rats (250–300 g) were ran-domly divided into four groups (I–IV) of 10 rats each (5 femalesand 5 males). The animals were separated by gender and housedfive in each cage under the same conditions as mentioned abovefor the mice. While the first group was maintained as control,groups (II–IV) were administered 127.5, 255 and 510 mg/kg/dayfor 30 days by gavage. The doses selected for sub-chronic toxi-city study were based on the result of acute toxicity study. Thesub-chronic dose gradient was designed by 1/20, 1/10 and 1/5of rats LD50, which kept rats alive (Wang and Yuan, 1997). Thestock concentration of the steroidal saponins was 12.75, 25.50 and51.00 mg/ml, respectively. In the 30th day, each rat was to anes-thetize with pentobarbital, then to draw blood and take out theliver tissue. Blood samples were collected from the abdominalaorta for separation of serum, which for the assays of GOT, ALT,AKP and TBIL. The livers were excised from the rats and washed.One gram liver tissue was added with 10 ml of chilled normalsaline (pH = 7.0). Homogenizer was used to homogenate the tis-sue (200 rpm and amplitude 10 times). The homogenate solutionwas centrifuged at 12,000 rpm for 10 min at 4 ◦C. The upper layerwas collected and stored at −80 ◦C for assaying GSH, GR, GST andprotein content. Livers of mice were taken out for pathologicalhistology.

2.4. Metabolite studies in vivo

After an overnight fast, a group of male SD rats (n = 5) was orallydosed at 255 mg/kg of the steroidal saponins. Freeze-dried pow-der of the steroidal saponins was dissolved with distilled wateras 25.5 mg/ml. The above solution was orally administered to rats(1 ml/100 g). After 2 h, the rats were anaesthetized by intraperi-toneal injection of 1% pentobarbital sodium (0.15 ml/100 g). Theblood was collected from the hepatic portal vein and then cen-trifuged at 4000 rpm for 10 min at 4 ◦C. The supernatant wasobtained and stored at −20 ◦C, and thawed before analysis. Bloodplasma was extracted with acetic ether and 0.1% orthophosphoricacid, repeated 3 times. Extraction was merged and dried under N2at 40 ◦C. The residues were re-dissolved in methanol and filteredthrough a 0.22-�m filter, the filtrate was used as HPLC–MS sample.

2.4.1. Chromatographic conditionWaters AcquityTM Ultra Performance LC system (Waters Cor-

vacuum degasser, autosampler, diode-array detector. The chro-matographic condition was as follows: UPLCTM BEH C18 column(1.7 �m, 2.1 mm × 50 mm); mobile phase: a linear gradient systemof A (HCOOH:H2O = 0.1:100) and B (CH3CN), the gradient programis shown in Table 1; flow rate: 0.2 ml/min; column temperature:25 ◦C; detecting wavelength: 204 nm; injection volume: 5 �l.

Y. Qin et al. / Journal of Ethnopharm

Table 1Solvent gradient program of HPLC analysis.

Time (min) Flow (ml/min) A (%) B (%)

0 0.2 40.0 60.015 0.2 5.0 95.0

2

eTotfdscm

2

saot

2

2

la1trmn2

2

TBa

2

tw

2

t2

ap

f42t

retention time and mass spectra. By comparing the chromatogramsof plasma containing drug with control plasma, two special peakswere found in plasma containing drug (Fig. 1 b). Mass spectrumresult had showed that diosgenin was the main metabolite in blood.The quasi-molecular ion split characteristic fragments of standard

18 0.2 5.0 95.020 0.2 40.0 60.0

.4.2. Mass spectra conditionWaters Micromass Q-TOF-microTM (Manchester, UK) was

quipped with an electrospray ion source operating in positive ion.he ion source temperature was set at 100 ◦C with a cone gas flowf 40 L h−1, a desolvation gas temperature of 300 ◦C and a desolva-ion gas flow of 600 L h−1. The capillary voltage was set at 3.0 kVor positive ion mode, the cone voltage was up to 40 V. Systematicata was collected in centroid mode, the lockspray frequency waset at 5 s and the lock-mass data was averaged over a 10 s scan fororrection. The mass spectrometric data was collected in full scanode, the m/z were from 200 to 1000 in positive ion.

.5. Acute toxicity studies in mice

Karber method was used to investigate acute toxicity of theteroidal saponins on mice (Klassen, 1991). Behavioral changesnd signs of toxicity and/or death, and the latency of death werebserved in 7 days. The LD50 value was determined according tohe method of Litchfield and Wilcoxon (Katayama et al., 2007).

.6. Sub-chronic toxicity studies on liver in rats

.6.1. Hepatic injury study of liverDetermining GOT, ALT and AKP in serum is to judge hepatic cel-

ular membrane injury of the steroidal saponins. Blood were placedt room temperature for 1 h, and then centrifuged at 1000 × g for0 min to obtain serum. GOT and ALT were measured accordingo the method of Reitman and Frankel using DNPH as coloringeagent (Dang et al., 2008; Liu et al., 2009). AKP activity waseasured according to the method of Walter and Schutt with p-

itrophenylphosphate (pNPP) as a substrate standard (Taylor et al.,005).

.6.2. Protein synthesis function study of liverLiver plays a very important role in synthesizing protein in vivo.

he protein content was measured according to the method ofradford using the bovine serum albumin as a standard (Kilkowskind Gross, 1999).

.6.3. TBIL excretion function study of liverBlood were placed at room temperature for 1 h, and then cen-

rifuged at 1000 × g for 10 min to obtain serum. The TBIL in serumas measured using the colorimetric method (Liu et al., 2007).

.6.4. Antioxidation study of liverThe GSH level of the liver was examined by using spectropho-

ometric diagnostic kits based on the method of Jollow (Gu et al.,004). Results of GSH were expressed as mg GSH/g protein.

The GR activity was dependent on the NADPH decrease inbsorbance at 340 nm. GR activity results were expressed in U/grotein (KumTatt et al., 1975).

◦

Glutathione S-transferase activity was measured at 37 C byollowing the conjugation of glutathione with either 1, 2-dichloro--nitrobenzene or 1-chloro-2,4-dinitrobenzene (Samiec et al.,000). The reaction was monitored at 340 nm in a spectrophotome-er using CDNB as the substrate. Assays were performed at 30 ◦C in

acology 126 (2009) 543–550 545

0.1 M phosphate buffer, pH 6.5, in the presence of 5 mM GSH and1 mM CDNB.

2.7. Histopathological examination of liver in mice

The liver of mice was sampled, immobilized in 10% formalde-hyde solution, wrapped with wax, and cut into slices with 4–5 �mthick. Formalin-fixed specimens were embedded in paraffin andstained with hematoxylin and eosin (H£E) staining kit for 15 min.After being washed in turn with dimethylbenzene (5 min, twice),100% ethanol (2 min), 95% ethanol (1 min), 80% ethanol (1 min), 75%ethanol (1 min) and distillated water (1 min), the liver slices weredetected under a light microscope for conventional morphologicalevaluation (Bustos et al., 2003).

2.8. Statistical analysis

All data are presented as mean ± S.E.M. Data were analyzed byone-way ANOVA followed by post hoc LSD’s test using SPSS soft-ware package Version 12.0 for Windows (SPSS, USA). All values ofp < 0.05 were considered as significantly different.

3. Results

3.1. HPLC–MS analysis of plasma sample after oraladministration

The typical chromatograms of samples were shown in Fig. 1a–c.The constituents in rats’ plasma after oral administration thesteroidal saponins were separated and identified by using their

Fig. 1. Chromatograms of reference compounds by HPLC–MS at 204 nm. (a) Stan-dard diosgenin, (b) plasma after oral administration of the steroidal saponins, (c)control plasma + standard diosgenin and (d) control plasma.

546 Y. Qin et al. / Journal of Ethnopharmacology 126 (2009) 543–550

F sing ESI-MS scanning in positive ion mode. (a) Plasma after oral administration of thes

dFispa

3

saT1hstmdow

TA

Tot(tt

ig. 2. Analysis of parent ion spectra and product ion spectra (peak 1 in Fig. 1b) uteroidal saponins and (b) control plasma.

iosgenin were m/z415, m/z397 and m/z271 (Fig. 3a). Peak 2 inig. 1b showed mass spectrum as Fig. 3b that diosgenin character-stic fragments had also been found. Peak 1 in Fig. 1b showed masspectrum as Fig. 2a, m/z806 and m/z703 were distinct matters com-ared to control. But nowadays there was no standard substancebout them so we did not define them at present in Fig. 3.

.2. Acute toxicity studies in mice

The no-observed adverse effect level (NOAEL) for the steroidalaponins in this study for mice was 562.5 mg/kg, which was equiv-lent to 5 times of the normal human dose in clinical prescription.he lowest-observed adverse effect level (LOAEL) in mice was125 mg/kg, which was equivalent to 10 times of the normaluman dose in clinical prescription. The calculated LD50 of theteroidal saponins was 3653 mg/kg in mice, which was equivalent

o 32.5 times of the normal human dose in clinical prescription. The

ortality and general behavior adverse effects changed in a dose-ependent manner (Table 2). The main behavioral signs of toxicitybserved were asthenia, piloerection, ataxia, anorexia, syncope andeight loss.

able 2cute toxicity effects of the steroidal saponins in mice.

Dose of steroidal saponins(mg/kg of bw)

Mortality Toxic symptoms

D/T Latency (h)

0 0/10 – None112.5 0/10 – None225 0/10 – None562.5 0/10 – None

1125 2/10 >24, <72 Asthenia, piloerection,weight loss

2250 4/10 <12 Asthenia, piloerection,ataxia, anorexia,syncope, weight loss

4500 6/10 <12 Asthenia, piloerection,ataxia, anorexia,syncope, weight loss

9000 10/10 <8 Asthenia, piloerection,ataxia, anorexia,syncope, weight loss

he lyophilized steroidal saponins, dissolved in distilled water, were administeredrally; each dose was administered to groups of 10 mice (5 males, 5 females). Allhe treated animals were carefully examined for 7 days for any signs of toxicitybehavioral changes and mortality). D/T: dead/treated mice; none: no toxic symp-oms were seen during the observation period; latency: time to death (in days) afterhe dose.

Fig. 3. Analysis of parent ion spectra and product ion spectra (peak 2 in Fig. 1b)using ESI-MS scanning in positive ion mode. (a) Standard diosgenin, (b) controlplasma + standard diosgenin, (c) plasma after oral administration of the steroidalsaponins and (d) control plasma.

Y. Qin et al. / Journal of Ethnopharmacology 126 (2009) 543–550 547

F l, (b) t( n 200×

w1eawd

3

ot

ig. 4. Effects of the steroidal saponins on liver histomorphology in mice. (a) Controe) treated with 4500 mg/kg and (f) treated with 9000 mg/kg. Original magnificatio

The lyophilized steroidal saponins, dissolved in distilled water,ere administered orally; each dose was administered to groups of

0 mice (5 males, 5 females). All the treated animals were carefullyxamined for 7 days for any signs of toxicity (behavioral changesnd mortality). D/T: dead/treated mice; none: no toxic symptomsere seen during the observation period; latency: time to death (inays) after the dose.

.3. Effects on histomorphology of liver in mice

Microscopic observations showed a normal liver histomorphol-gy in control mice (Fig. 4a). On the whole, present results showhat the treatment mice at the dose of 562.5 mg/kg (Fig. 4b) did not

Table 3Effects of the steroidal saponins on GOT, ALT and AKP in serum of rats.

Groups GOT (100 U/L)

Control 1.0740 ± 0.160127.5 mg/kg of bw 0.596 ± 0.123**

255 mg/kg of bw 0.839 ± 0.145*

510 mg/kg of bw 0.830 ± 0.120*

Data are expressed as mean ± S.E.M; n = 8.* P < 0.05, versus the control group.

** P < 0.01, versus the control group.

reated with 562.5 mg/kg, (c) treated with 1125 mg/kg, (d) treated with 2250 mg/kg,.

cause any adverse effect on the histomorphology change of hepa-tocytes compared with control. But high doses at 1125, 2250, 4500and 9000 mg/kg showed a regular presence of cytoplasmatic vac-uoles with a light vascular congestion and sporadic focal necrosisin the liver (Fig. 4c–f).

3.4. Sub-chronic toxicity studies on liver in rats

3.4.1. Effects on serum biochemical parameters GOT, ALT and AKPSeveral hepatic enzymes in serum were used for the biochemical

markers to understand the early hepatic injury such as GOT, ALT andAKP. Table 3 shows the effects of the steroidal saponins on GOT, ALTand AKP in serum for 30 days. Rats at the doses of 127.5, 255 and

ALT (100 U/L) AKP (U/g protein)

0.404 ± 0.102 32.532 ± 12.3210.290 ± 0.114* 30.675 ± 19.6650.355 ± 0.124 20.476 ± 8.1360.308 ± 0.079 32.352 ± 12.321

548 Y. Qin et al. / Journal of Ethnopharmacology 126 (2009) 543–550

Fig. 5. Effect of the steroidal saponins on TBIL in serum of rats. Data are expressedas mean ±S.E.M; n = 8; **P < 0.01, versus the control group.

Fe

5c

3

Twc

3

gpc

3

stcwf

4

sizot

ig. 6. Effect of the steroidal saponins on protein content in liver of rats. Data arexpressed as mean ± S.E.M; n = 8; **P < 0.01, versus the control group.

10 mg/kg/day decreased the GOT, ALT and AKP levels in serumompared with the control group (*P < 0.05, **P < 0.01).

.4.2. Effect on serum TBILFig. 5 shows the effect of the steroidal saponins on TBIL in serum.

BIL in serum increased in each treatment group, especially thereas a significant increase (**P < 0.01) in 510 mg/kg/day group when

ompared with the control group (Fig. 5).

.4.3. Effect on protein content in liverProtein content decreased in liver was observed on experiment

roup compared to the control. There was an obvious decrease ofrotein content in liver at 255 and 510 mg/kg/day doses groups inomparison to the control group (**P < 0.01) (Fig. 6).

.4.4. Effects on GSH, GR and GST activities in liverGSH constitutes the first line of defense against free radicals. As

hown in Fig. 7, there was a significant increase of GSH content inhe liver of treatment groups (**P < 0.01). GR level in liver signifi-antly increased at dose of 127.5 mg/kg/day (*P < 0.05). Treatmentith the steroidal saponins at dose of 127.5, 255 and 510 mg/kg/day

or 30 days increased GST.

. Discussion

Ethanol extract of Dioscorea zingiberensis was the steroidal

aponin fractions including water-soluble fractions and water-nsoluble fraction. Water-soluble fractions include zingiberenin A,ingiberenin B, protozingiberenin A, protozingiberenin B and son. Water-insoluble steroidal saponins include trillin, gracillin, pro-ogracillin, deltonin and so on (Xu et al., 2007). Steroidal saponins

Fig. 7. Effects of the steroidal saponins on GSH (a), GR (b) and GSTs (c) in liver ofrats. Data are expressed as mean ± S.E.M; n = 8; *P < 0.05, **P < 0.01, versus the controlgroup.

are glycosides consisting of an aglycone (diosgenin) and several gly-cosyl moieties. The most common sugars encountered in saponinsare pentoses (arabinose, xylose, etc.), hexoses (glucose, galac-tose, etc.) and 6-deoxyhexose (rhamnose, etc.) (Yuan et al., 2005;Lin et al., 2007). Diosgenin was the main metabolite of steroidalsaponins from Dioscorea zingiberensis in serum by HPLC–MS in thisresearch, which provides helpful chemical information for toxicol-ogy and further pharmacology research on Dioscorea zingiberensis.Saponins, when added in vitro induce lysis of red blood cells, withsteroid saponins being more potent lytic agents than the triter-penoid saponins (Baumann et al., 2000). Any extract containingsaponins has the potential to cause hemolytic anemia. It has also

been shown that cholesterol inhibits the lytic activity of saponins(Takechi et al., 1992). So it is likely that the steroidal saponins fromDioscorea zingiberensis could cause lysis of red cells in hepatic portalvein in the acute toxicity study.

pharm

fiis2wffswtt13tozdamgks2istf

onpt1ih

tvmGwmanag

itstlitaHgibbfadts

Y. Qin et al. / Journal of Ethno

Acute toxicity studies with a range of doses have to be conductedrst to select proper dose(s) for chronic and sub-chronic stud-

es; the doses selected for chronic and sub-chronic toxicity studieshould be at and above the suggested human dose (Rhiouani et al.,008). The dose of steroidal saponin that has been used traditionallyas 10 mg/kg/day, which contain 3.5 mg diosgenin in it. There-

ore, we quantified the doses of the steroidal saponin extractedrom Dioscorea zingiberensis based on diosgenin content. The doseselected for acute toxicity study were 112.5–9000 mg/kg, whichere equivalent to 1 time, 2 times, 5 times, 10 times, 20 times, 40

imes and 80 times of the normal human dose in clinical prescrip-ion, respectively. The doses selected for sub-chronic toxicity were27.5, 255, 510 mg/kg/day, which were equivalent to 1.7 times,.4 times and 6.8 times of the human dose in clinical use, respec-ively. In the acute toxicity study, mice administered at oral dosef 562.5 mg/kg (NOAEL) of the steroidal saponins from Dioscoreaingiberensis did not exhibit any signs of adverse effects, which atoses of 1125 mg/kg (LOAEL) and higher exhibited adverse effectsnd mortality in dose-dependent manners. The calculated LD50 inice after a single oral dose was 3653 mg/kg. The sub-chronic dose

radient was designed by 1/20, 1/10 and 1/5 of rats LD50, whichept rats alive. According to Horn, falls in the category of non-toxicubstances, since, plants or plant products with LD50 higher than–3 g/kg are considered free of any toxicity (Horn, 1956). A product

s considered non-toxic if no deaths are registered and no clinicaligns of toxicity are observed at doses at or below 5 g/kg in acuteoxicity studies (Brock et al., 1995). Above all, steroidal saponinsrom Dioscorea zingiberensis do not have significant toxicity.

Microscopic observations showed a normal liver histomorphol-gy in control rats (Fig. 4a). Liver histomorphological change wasormal in mice at the dose of 562.5 mg/kg (Fig. 4b). But a regularresence of cytoplasmatic vacuoles with a light vascular conges-ion and sporadic focal necrosis was showed in rats treated with125–9000 mg/kg (Fig. 4c–f). Necropsy of the highest dose group

n acute experiment exhibited atrophy of liver accompanied byemorrhages and pallid spots in liver boundaries.

Analysis of blood parameters is relevant to risk evaluation ashe changes in the hematological system have a higher predictivealue for human toxicity, when the data are translated from ani-al study (Olson et al., 2000). In the present study, we determinedOT, ALT and AKP in serum, the major markers of hepatic injury,hich were to assay toxicity of the steroidal saponins on hepatocyteembrane. The overall results suggested that GOT, ALT and AKP

ctivities were decreased, which showed the steroidal saponins didot have toxicity on hepatocyte membrane, and could inhibit hep-tic cytolemma lipid peroxidation and prevent hepatocyte enzymeo in blood contradictory inversely.

TBIL increased in serum of each treated group, especially signif-cantly increased at the dose of 510 mg/kg when compared withhe control group in the sub-chronic toxicity study. The resultshowed that treatment of high dose of the steroidal saponins ledo the release of high levels of serum bilirubin, which may increaseethality. The metabolic process of TBIL in liver includes TBIL intak-ng, conjugated bilirubin forming, conjugated bilirubin excretingo biliary passage. Any course blocked that might induce bilirubinccumulation in blood, which caused jaundice at last (Yao, 2004).igh dose of the steroidal saponins in vivo may damage liver andall system. The consequence was that too much bilirubin siltatedn liver, which makes bilirubin difficult to excrete from liver. Biliru-in is a metabolic breakdown product of blood heme with greatiological and diagnostic values. Abnormal bilirubin concentrations

ound in human serum or plasma usually signify the presence of

variety of diseases with liver dysfunctions, ranging from jaun-ice to infectious hepatitis (Wu et al., 1992). Our results provedhe report that long-term and high dose administration of steroidalaponins induced acute icteric hepatitist in their clinic use (Zhou

acology 126 (2009) 543–550 549

et al., 1999). Biliverdin is the end product of heme metabolism; butin most mammals biliverdin is converted to bilirubin before excre-tion. This reaction is catalysed by the cytosolic enzyme biliverdinreductase, which requires NADH or NADPH for activity. Bilirubinis toxic and needs conjugation before excretion. This mechanism isperformed by the enzyme UDP-glucuronosyltransferase, catalysingthe transfer of the glucuronosyl moiety of UDP-glucuronic acid tobilirubin (Dutton, 1966; Colleran and O’Carra, 1977). It has beenreported that bilirubin infusion in pigs leads to vacuolization of thecytoplasm of hepatocytes and canalicular cholestasis (Veel et al.,1991). So we should strengthen TBIL monitoring in the long timeuse of steroid saponins for curing cardiovascular disease.

In the liver, bilirubin affects the protein synthesis (Majumdarand Greenfield, 1974). There was an obvious decrease of proteincontent in liver at doses of 255 and 510 mg/kg/day in comparisonto the control group. Only 127.5 mg/kg/day group had protein levelequivalent to the control group. The most commonly used conver-sion factor assumes a cost of 5 ATP equivalents per peptide bond,i.e., 2 ATP to activate each amino acid, one for peptide bond syn-thesis, one for translocation, and one to take into account variousunknown costs such as errors, the production of RNA, and aminoacid transport. Energy from mitochondrion promotes protein syn-thesis and secretion in endoplasmic reticulum (Aoyagi et al., 1988).Treatment of high dose of saponins induced hepatocyte injury,which may changed structure and function of organelle in hepa-tocyte such as endoplasmic reticulum, mitochondrion and so on,which hampered energy metabolism and decreased protein syn-thesis seriously. This article first showed that treated with steroidalsaponins for a long time and a high dose could weak protein synthe-sizing ability of liver. Under normal conditions, circulating bilirubinis bound to albumin, which protects cells against the potential tox-icity of bilirubin. Several human diseases have been reported tobe closely related to abnormally increased levels of bilirubin. In theCrigler–Najjar syndrome, the bilirubin levels are increased up to 10times due to the absence of bilirubin UDP-glucuronosyltransferaseactivity (Odell, 1959). Ribosome binding on endoplasmic reticulumin hepatocyte might be damaged. So the decreased protein contentin liver and increased total bilirubin in serum both represented theinjury of liver under the high dose.

We investigated hepatotoxicity of the steroidal saponins bydetermining GSH, GR and GST in liver and evaluated the toxicity onliver antioxidation. GSH acts as an antioxidant both intracellularlyand extracellulary in conjunction with various enzymatic processesthat reduced hydrogen peroxide and hydroperoxide by oxidizingGSH to GSSG and other mixed disulfides (Prakash et al., 2001). Inaddition, the GSH antioxidant system plays a fundamental role incellular defense against reactive free radicals and other oxidantspecies. Furthermore, induction of the hepatic GSH antioxidant sys-tem by chemopreventive agents has been reported (Velmuruganet al., 2001). Rat liver tissue content GSH concentration is about7–8 mM (Halliwell and Gutteridge, 1998). Our results showed thatGSH, GR and GST in liver were increased in the rats at doses of127.5–510 mg/kg/day (Fig. 7). This study revealed that steroidalsaponins from Dioscorea zingiberensis caused no toxicity on liverantioxidative system after administered for a long time and demon-strated the steroidal saponins abilities on scavenging free radicaland counteracting oxidative stress, which should be investigatedfurther for other possible therapeutic implications.

5. Conclusion

Because of the relatively high NOAEL values in the acute study inmice, and lack of mortality or significant changes in the biologicaland hematological parameters in rats after 30 days of daily dosing,it may be concluded that the steroidal saponins does not appearto have significant toxicity (except at high doses). This study also

5 pharm

rahm

A

tCF“(eF

R

A

B

B

B

B

C

D

D

G

H

HK

K

K

K

L

L

L

Zhang, J., Lu, J., 2006. Advances in study on extraction and content determination ofsaponin. Modern Chinese Medicine 8, 26–28.

50 Y. Qin et al. / Journal of Ethno

evealed the steroidal saponins abilities on scavenging free radicalnd counteracting oxidative stress. Metabolite research providedelpful chemical information for the toxicology and further phar-acology research on Dioscorea zingiberensis.

cknowledgement

We thank Mr Wang for valuable help in the pathological his-ology examination and the Animal Experimental Center of Westhina Hospital, Sichuan University for the daily feed of mice.inancial supports of this research were from the China National11.5” Foundation (No. 2006BAB04A14), National 863 ProjectsNos. 2006AA03Z356 and 2007AA021800), National Natural Sci-nce Foundation of China (No. 20972105) and Sichuan Provinceoundation (No. 2008SZ 0024).

eferences

oyagi, Y., Tasaki, I., Okumura, J.-I., Muramatsu, T., 1988. Energy cost of wholebody protein synthesis measured in vivo in chicks. Journal of Physiology andBiochemistry 91, 765–768.

accou, J., Lambert, F., Sanvaire, Y., 1977. Spectrophotometric method for the deter-mination of total steroidal sapogenin. Analyst 102, 458–466.

aumann, E., Stoya, G., Volkner, A., Richter, W., Lemke, C., Linss, W., 2000. Hemol-ysis of human erythrocytes with saponin affects the membrane structure. ActaHistochemica 102, 21–35.

rock, W.J., Trochimowicz, H.J., Millischer, R.J., Farr, C., Kawano, T., Rusch, G.M., 1995.Acute and sub-chronic toxicity of 1,1-dichloro-1-fluoroethane (HCFC-141b).Food and Chemical Toxicology 33, 483–490.

ustos, M., Beraza, N., Lasarte, J., Baixeras, E., Alzuguren, P., Bordet, T., Prieto, J., 2003.Protection against liver damage by cardiotrophin-1: a hepatocyte survival factorup-regulated in the regenerating liver in rats. Gastroenterology 125, 192–201.

olleran, E., O’Carra, P., 1977. Enzymology and comparative physiology of biliverdinreduction. In: Berk, P.D., Berlin, N.I. (Eds.), International Symposium on Chem-istry and Physiology of Bile Pigments. Department of Health, Education andWelfare, Washington, DC, pp. 69–85.

ang, S., Zhang, X., Jia, X., Cheng, Y., Song, P., Liu, E., He, Q., Li, Z., 2008. Protectiveeffects of emodin and astragalus polysaccharides on chronic hepatic injury inrats. Chinese Medical Journal 121, 1010–1014.

utton, G., 1966. Variations in glucuronide formation by perinatal liver. BiochemicalPharmacology 15, 947–951.

u, J., Cui, H., Behr, M., Zhang, L., Zhang, Q., Yang, W., Jack, A.H., Ding, X., 2004. In vivomechanisms of tissue-selective drug toxicity: effects of liver-specific knockoutof the NADPH-cytochrome P450 reductase gene on acetaminophen toxicity inkidney, lung, and nasal mucosa. Molecular Pharmacology 67, 623–624.

alliwell, B., Gutteridge, J.M.C., 1998. Antioxidant Defense: Free Radicals in Biologyand Medicine. Oxford University Press Inc., New York, pp. 105–245.

orn, H.J., 1956. Simplified LD50 (ED50) calculation. Biometrics 12, 312–322.atayama, S., Irifune, M., Kikuchi, N., Takarada, T., Shimizu, Y., Endo, C., Takata, T.,

Dohi, T., Sato, T., Kawahara, M., 2007. Increased {gamma}-aminobutyric acid lev-els in mouse brain induce loss of righting reflex, but not immobility, in responseto noxious stimulation. Anesthesia and Analgesia 104, 1422–1429.

ilkowski, W.J., Gross, G.G., 1999. Color reaction of hydrolyzable tannins with Brad-ford reagent, Coomassie brilliant blue. Phytochemistry 51, 363–366.

lassen, C., 1991. Principles of toxicology. In: Gilman, A.G., Tall, T.W., Nies, A.S.,Taylor, P. (Eds.), Pharmacological Basis of Therapeutics. McGraw-Hill, pp. 49–61.

umTatt, L., Tan, I., Seet, A., 1975. A new colorimetric method for the determina-tion of NADH/NADPH dependent glutathione reductase in erythrocytes and inplasma. Clinica Chimica Acta 58, 101–108.

in, S., Wang, D., Yang, D., 2007. Characterization of steroidal saponins in crudeextract from Dioscorea nipponica Makino by liquid chromatography tandemmulti-stage mass spectrometry. Analytica Chimica Acta 599, 98–106.

iu, J., Tan, H., Sun, Y., Zhou, S., Cao, J., Wang, F., 2009. The preventive effectsof heparin–superoxide dismutase on carbon tetrachloride-induced acute liverfailure and hepatic fibrosis in mice. Molecular and Cellular Biochemistry 327,219–228.

iu, M., Wang, B., Zhao, X., Wang, G., Zhou, H., 2007. Induction of heme oxygenase-1improves cold preservation effect of liver graft. Biochemistry 72, 545–551.

acology 126 (2009) 543–550

Liu, Z., Zou, W., Wang, R., Zhou, Z., 2004. Clinical application of Di’ao Xin Xue Kangcapsule for ten years in China. China Journal of Traditional Chinese Medicineand Pharmacy 19, 620–622.

Ma, H., Zhou, Q., Wang, B., 2002. Studies on the metabolism of DX by intestinalbacteria and the absorbed components in serum. China Pharmacy 13, 204–205.

Majumdar, A.P.N., Greenfield, S., 1974. Evidence of defective protein synthesis inliver in rats with congenital hyperbilirubinemia. Biochimica et Biophysica Acta335, 250–257.

Nie, L., Lin, S., Ning, Z., 2004. Diosgenin in species of genus dioscorea research inprogress. Chinese Journal of Biochemical Pharmaceutics 25, 318–320.

Odell, G.B., 1959. The dissociation of bilirubin from albumin and its clinical implica-tions. Journal of Pediatrics 55, 268–273.

Olson, H., Betton, G., Robinson, D., Thomas, K., Monro, A., Kolaja, G., Lilly, P., Sanders,J., Sipes, G., Bracken, W., Dorato, M., Deun, K.V., Smith, P., Berger, B., Heller, A.,2000. Concordance of toxicity of pharmaceuticals in humans and in animals.Regulatory Toxicology and Pharmacology 32, 56–67.

Prakash, J., Gupta, S.K., Kochupillai, V., Singh, N., Gupta, Y.K., Joshi, S., 2001.Chemopreventive activity of Withania somnifera in experimentally inducedfibrosarcoma tumours in swiss albion mice. Phytotherapy Research 15,240–244.

Rhiouani, H., El-Hilaly, J., Israili, Z.H., Lyoussi, B., 2008. Acute and sub-chronic toxicityof an aqueous extract of the leaves of Herniaria glabra in rodents. Journal ofEthnopharmacology 118, 378–386.

Samiec, P.S., Dahm, L.J., Jones, D.P., 2000. Glutathione S-transferase in mucus of ratsmall intestine. Toxicological Sciences 54, 52–59.

Shen, L., Fang, Y., Xu, D., Liu, Y., Lai, Y., 2006. Determining metabolite diosgen in ofOphiopogon japonicus saponin in vivo by HPLC–MS. Chinese Traditional PatentMedicine 28, 1178–1180.

State Pharmacopoeia Committee of the People’s Republic of China, 2000. Pharma-copoeia of the People’s Republics of China [S]. Part I, 2000 ed. Chemical IndustryPress Inc., Beijing, Appendix pp. 53–63.

Sun, Q., Ju, Y., Zhao, Y., 2002. Steroid saponins with biological activities. ChineseTraditional and Herbal Drugs 33, 276–280.

Takechi, M., Shimada, S., Tanaka, Y., 1992. Time course and inhibition of saponin-induced hemolysis. Planta Medica 58, 128–130.

Taylor, R.H., Fournier, S.M., Simons, B.L., Kaplan, H., Hefford, M.A., 2005. Covalentprotein immobilization on glass surfaces: application to alkaline phosphatase.Journal of Biotechnology 118, 265–269.

Veel, T., Villanger, O., Holthe, M., Skjorten, F., Raeder, M., 1991. Intravenous bilirubininfusion causes vacuolization of the cytoplasm of hepatocytes and canalicularcholestasis. Acta Physiol Scand 143, 421–429.

Velmurugan, B., Bhuvaneswari, V., Balasenthil, S., Nagini, S., 2001. Lycopene, anantioxidant carotenoid modulates glutathione-dependent hepatic biotransfor-mation enzymes during experimental gastric carcinogenesis. Nutrition Research21, 1117–1124.

Wang, A., Hao, X., Zhao, L., 2006. New technology study of saponin extract fromDioscorea zingiberensis C.H.Wright. Chemical Industry and Engineering 23,559–561.

Wang, Q., Yuan, B., 1997. Preclinical Safety Evaluation and Practice of New Drug.Military Medical Science Press Inc, p. 47.

Wu, N., Wang, T., RA, H., CW, H., 1992. Separation of serum bilirubin species bymicellar electrokinetic chromatography with direct sample injection. Journal ofChromatography 582, 77.

Xu, D., Hu, C., Wei, L., Pang, Z., 2007. Isolation and structure determination ofsteroidal saponin from Dioscorea zingiberensis. Acta Pharmaceutica Sinica 42,1162–1165.

Yang, R., Tang, S., Pan, F., Zhao, A., Pang, Z., 2007. Advances in Study of Dioscoreazingiberensis. Chinese Wild Plant Resources 26, 1–5.

Yao, G., 2004. Bilirubin metabolism. Chinese Journal of Hepatology 9, 17.Zan, L., Sun, W., Zhang, K., 2005. New technology study of saponin extract from

Dioscorea zingiberensis C.H.Wright. Chemical Industry and Engineering 20,138–139.

Yuan, S., Yan, Y., Lin, H., 2005. Plant regeneration through somatic embryogene-sis from callus cultures of Dioscorea zingiberensis. Plant Cell Tissue and OrganCulture 80, 157–161.

Zhang, Y., Zuo, G., 2007. Adverse reaction induced by dioscin. Chinese Journal ofPharmacoepidemiology 16, 124.

Zhou, Y., Lu, Y., Che, W., 1999. Two examples of hepatic injury induced by Di-ao-xin-xue-kang. Chinese Pharmaceutical Affairs 13, 132–133.