An aqueous extract of Rubus chingii fruits protects primary rat hepatocytes against tert-butyl hydroperoxide induced oxidative stress

Life Sciences 72 (2002) 329 – 338 www.elsevier.com/locate/lifescie

An aqueous extract of Rubus chingii fruits protects primary rat hepatocytes against tert-butyl hydroperoxide induced oxidative stress Ming-Hon Yau a, Chun-Tao Che b, Song-Ming Liang b, Yun-Cheung Kong b, Wing-Ping Fong a,* a b

Department of Biochemistry, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong, China School of Chinese Medicine, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong, China Received 23 April 2002; accepted 30 August 2002

Abstract Different in vitro free radical generating systems were used to assess the antioxidative activity of aqueous extracts of the five herbal components of Wu-zi-yan-zong-wan, a traditional Chinese medicinal formula with a long history of use for tonic effects. Fructus Rubi [Rubus chingii (Rosaceae) fruits] was found to be the most potent. It was further investigated using the primary rat hepatocyte system. tert-Butyl hydroperoxide (t-BHP) was used to induce oxidative stress. Being a short chain analog of lipid hydroperoxide, t-BHP is metabolized into free radical intermediates by the cytochrome P450 system in hepatocytes, which in turn, initiate lipid peroxidation, glutathione depletion and cell damage. Pre-treatment of hepatocytes with Fructus Rubi extract (50 Ag/ml to 200 Ag/ ml) for 24 h significantly reversed t-BHP-induced cell viability loss, lactate dehydrogenase leakage and the associated glutathione depletion and lipid peroxidation. The amount of reactive oxygen species formed was also decreased as visualized by the fluorescence probe 2V,7V-dichlorofluorescin diacetate. These results suggested that Fructus Rubi was useful in protecting against t-BHP-induced oxidative damage and may also be capable of attenuating cytotoxicity of other oxidants. D 2002 Elsevier Science Inc. All rights reserved. Keywords: Rubus chingii fruits; Antioxidant; Free radical; Hepatocytes; Oxidative stress; tert-butyl hydroperoxide; Traditional Chinese medicine

* Corresponding author. Tel.: +852-2609-6868; fax: +852-2603-5123. E-mail address: [email protected] (W.-P. Fong). 0024-3205/02/$ - see front matter D 2002 Elsevier Science Inc. All rights reserved. PII: S 0 0 2 4 - 3 2 0 5 ( 0 2 ) 0 2 2 3 9 - 7

330

M.-H. Yau et al. / Life Sciences 72 (2002) 329–338

Introduction Wu-zi-yan-zong-wan (WZ) is a multi-component Chinese medicinal formula with a long history of use for tonic effects [1]. The composition of the WZ recipe includes Fructus Lycii (FL) [Lycium barbarum or L. chinensis (Solanaceae) fruits], Fructus Rubi (FR) [Rubus chingii (Rosaceae) fruits], Fructus Schisandrae (FS) [Schisandra chinensis (Magnoliaceae) fruits], Semen Cuscutae (SC) [Cuscata chinensis (Convolvulaceae) seeds], and Semen Plantaginis (SP) [Plantago asiatica or P. depressa (Plantaginaceae) seeds]. The traditional WZ prescription is generally indicated for male ‘‘kidney deficiency’’ manifested by sexual dysfunction [1]. It has also been reported to improve the undesirable effects of aging and to decrease alcohol-induced hepatotoxicity [2,3]. These effects can be attributed, at least partially, to the reduction of peroxidation reactions [3], suggesting that some components of the WZ preparation may afford protection against reactive oxygen species (ROS) [4]. To explore the free radical scavenging activity of traditional herbal medicine, a number of in vitro assays, generating different kinds of free radicals, are commonly used. For more detailed studies involving mechanism, it is preferable to use the primary rat hepatocyte system. tert-Butyl hydroperoxide (t-BHP), an analog of lipid peroxide, is widely used to induce oxidative stress in rat hepatocytes [5–7]. It is metabolized by the microsomal cytochrome P450 system to ROS [8,9] which subsequently initiates lipid peroxidation [10] and depletes cellular reduced glutathione (GSH) content [11]. In the present study, the aqueous extracts of the five components of the WZ formula were analyzed individually for their free radical scavenging activity in different in vitro systems. The antioxidative activity of the most potent extract, namely, that from FR, was further examined in primary rat hepatocyte culture subjected to t-BHP-induced oxidative stress.

Methods Materials Phenazine methosulfate (PMS), nitroblue tetrazolium (NBT), 2,2V-azobis(2-amidinopropane) dihydrochloride (AAPH), bleomycin sulfate, William’s Medium E, Hank’s balanced salt solution (HBSS), type VII collagen, type IV collagenase, insulin (bovine pancreas), dexamethasone, t-BHP, thiobarbituric acid (TBA), GSH, 3-[4,5-dimethylthiazol-2-yl]-2,5-dephenyl tetrazolium bromide (MTT) and 5,5Vdithiobis-(2-nitrobenzoic acid) (DTNB) were purchased from Sigma. Cell culture materials were purchased from GIBCO-BRL. 2V,7V-Dichlorofluorescin diacetate (DCFDA) was from Molecular Probe. All other chemicals used were of the highest grade available commercially. Preparation of extracts The five components of WZ formula, namely, FL, FR, FS, SC and SP, were purchased from a local vendor and their taxonomic identities were established by using pharmacopoeial standards. Each individual herb was ground into powder and boiled in distilled water (1 g/10 ml) for 3 h under reflux. The filtrate was freeze-dried, and the resulting powder was used as the crude aqueous extract of the herb. In selected experiments, FR extract was pre-treated as follows [5]: (i) Acidification with HCl to pH 2.3

M.-H. Yau et al. / Life Sciences 72 (2002) 329–338

331

or alkalinization with NaOH to pH 12.0, followed by neutralization back to pH 6.0. (ii) Digestion with 1 mg/ml trypsin at 37 jC for 30 min, followed by the addition of 1 mg/ml soybean trypsin inhibitor. (iii) Heating at 121 jC under high pressure, followed by cooling. (iv) Ultrafiltration through membrane with a molecular weight cutoff of 1000 (Amicon YM1 membrane). Determination of in vitro free radical scavenging activity The scavenging effect of the crude extracts against superoxide radical was studied in a PMS-NADH system by following the reduction of NBT [12]. The inhibition on lipid peroxidation [13] was examined by using liver microsomal preparation. The ability to inhibit erythrocyte hemolysis induced by thermal degradation of AAPH was investigated according to Zhang et al. [14]. The pro-oxidant activity of the extracts was assessed by the bleomycin-dependent DNA damage assay [15]. Isolation and cultivation of hepatocytes Male Sprague Dawley rats (body weight f 200 g) were obtained from the Laboratory Animal Services Centre of our University. Hepatocytes were isolated by two-stage collagenase liver perfusion in situ as described in [16]. Only preparations with cell viability greater than 90%, as determined by the trypan blue exclusion test, were used for subsequent experiments. Cells seeded onto collagen-precoated plates were cultured in William’s Medium E supplemented with 0.3 AM insulin, 0.1 AM dexamethasone, 100 U/ml penicillin, 100 Ag/ml streptomycin and 10% fetal bovine serum at 37 jC under 95% humidity and 5% CO2. The cells were either seeded onto 24-well plates at 2  105 cells/well for cytotoxicity studies [lactate dehydrogenase (LDH) leakage and MTT assay], or 6-well plates at 1  106 cells/well for malondialdehyde (MDA) and GSH measurements. All treatments were performed 24 h after cell attachment to allow monolayer formation. Hepatotoxicity assay After incubation with FR extract for a certain period of time, the hepatocytes were washed with HBSS to remove unabsorbed extract. Oxidative stress was then initiated by incubating hepatocytes with 500 AM t-BHP for 3 h. For the control group, neither FR nor t-BHP was added. The medium was removed and assayed for the leakage of LDH activity by following the absorbance change at 340 nm in an assay medium containing 1 mM pyruvate, 0.15 mM NADH and 0.1 M sodium phosphate (pH 8.0). Cell viability was measured by the MTT assay [17]. A volume of 400 Al of 0.5 mg/ml MTT in RPMI 1640 medium was added to each well and incubated for 2 h at 37 jC (95% humidity, 5% CO2). Formazan crystals formed in actively metabolizing cells were extracted with 10% SDS (in 10 mM HCl) and the absorbance at 540 nm was recorded. Biochemical assays MDA released into the culture medium was assayed spectrophotometrically at 532 nm after reaction with TBA [18]. For the determination of GSH content, cells were trypsinized and sonicated before being treated with DTNB. The absorbance at 412 nm was measured after 10 min [19]. Protein content was determined by Lowry’s method, using bovine serum albumin as the standard. The fluorescence probe

332

M.-H. Yau et al. / Life Sciences 72 (2002) 329–338

Table 1 Antioxidative activity of aqueous extracts of different components of WZ formula Free radical assays Superoxide generation Erythrocyte hemolysis Lipid peroxidation

IC50 (Ag/ml) FR

FL

FS

SC

SP

8.7 63.0 12.5

>500 >500 >500

158 >500 >500

>500 >500 >500

56.0 >500 35.0

The inhibition was studied at different concentrations (1 Ag/ml to 500 Ag/ml) of individual aqueous extract. IC50 refers to the concentration where 50% inhibition was observed.

DCFDA was used to evaluate the amount of t-BHP-induced ROS formation [20]. Hepatocytes seeded onto 24-well culture plates were incubated with FR extract and t-BHP as aforementioned, except cells were treated with 20 AM DCFDA for 30 min prior to t-BHP treatment. The fluorescence intensity was measured with excitation and emission wavelengths of 485 nm and 530 nm, respectively. Statistical analysis Results obtained are expressed as mean F SD. Statistical analysis was performed according to Student’s t-test by one-way analysis of variance. Significant difference was taken as p < 0.05.

Results In vitro free radical scavenging activity of aqueous extracts The aqueous extract of FR was found to be the most potent superoxide scavenger. The effect was dose-dependent with an IC50 value of 8.7 Ag/ml. The FR extract also exhibited potent antioxidative activity in both lipid-peroxidation (IC50 = 12.5 Ag/ml) and erythrocyte hemolysis (IC50 = 63.0 Ag/ml) assays, that measure the generation of hydroxyl and peroxyl radicals, respectively (Table 1). At a concentration of 100 Ag/ml, the FR extract could inhibit the generation of these free radicals by at least 90%. For the other extracts, SP showed some scavenging activity against superoxide and hydroxyl radicals while FS was slightly active towards superoxide. The extracts of SC and FL, at concentrations Table 2 Cytotoxicity of FR extract on rat hepatocytes FR (Ag/ml)

% viability of control 24 h

50 100 200 500

102.5 98.5 92.2 86.5

48 h F F F F

3.1 1.6 1.4** 2.2**

100.3 95.9 86.8 88.0

F F F F

1.8 1.9 2.5** 2.0**

Cells were treated with FR extract for 24 h or 48 h in William’s Medium E under conditions as described in Methods. Cell viability was determined by MTT assay. Results are expressed as mean F SD (n = 4). ** p < 0.01, compared with cells cultured in the absence of FR extract.

M.-H. Yau et al. / Life Sciences 72 (2002) 329–338

333

below 500 Ag/ml, did not show any scavenging effect in all the three free radical generating systems (Table 1). On the other hand, all the aqueous extracts did not exert a pro-oxidant effect as reflected by the failure to induce DNA damage in a bleomycin-dependent system (data not shown). Cytotoxicity of FR aqueous extract on rat hepatocytes The most potent extract, FR, was further evaluated using the rat hepatocyte system. Prior to testing the antioxidant potential of FR, the cytotoxicity was checked by incubating the cells with the extract for periods of up to 48 h. Significant cytotoxicity was observed at FR concentrations above 200 Ag/ml. For example, at a dose of 500 Ag/ml, FR caused cell viability to drop by 13%. At 200 Ag/ml FR, cytotoxicity was mild as cell viability could still be maintained at above 90% after 24 h of incubation (Table 2). Hence, the maximum dose of FR extract used in subsequent experiments was 200 Ag/ml. Effect of FR pre-treatment on t-BHP-induced hepatocytes damage The oxidative stress induced by 500 AM t-BHP caused over 80% cells to die after 3 h of incubation, together with a 10-fold increase in LDH leakage (Fig. 1). Pre-treating the cells with FR extract for 24 h

Fig. 1. Concentration-dependent effect of FR pre-treatment on t-BHP-induced cell damage in rat hepatocytes. Cells were treated with different doses of FR extract for 24 h before being challenged with 500 AM t-BHP for 3 h. Control group was treated with culture medium only. Cell damage was evaluated by (upper) MTT assay, and (lower) LDH leakage. Values are expressed as mean F SD (n = 4). *p < 0.05, **p < 0.01, compared with cells treated with t-BHP only.

334

M.-H. Yau et al. / Life Sciences 72 (2002) 329–338

protected the hepatocytes against the cytotoxicity of t-BHP in a dose-dependent manner. At a concentration of 200 Ag/ml, FR extract revived cell viability from 20% in cell treated with t-BHP only to 90%, and LDH leakage was also strongly suppressed. Fig. 2 shows the time-dependent inhibitory effect of FR pre-treatment on t-BHP-induced cell damage. Significant protection was observed with 3 h of pre-treatment. Effect of FR on t-BHP- induced lipid peroxidation, GSH depletion and ROS generation Hepatocytes treated with t-BHP alone showed a 7-fold increase in the amount of MDA released into the medium. However, pre-treating the cells with FR at doses ranging from 50 Ag/ml to 200 Ag/ml significantly inhibited MDA formation (Fig. 3). Furthermore, intracellular GSH content was also depleted under oxidative stress. Pre-treating cells with FR extract did not show any direct effect on cellular GSH content (data not shown), but it could reverse t-BHP-induced GSH depletion in a dosedependent manner (Fig. 3). Metabolism of t-BHP by hepatocytes leads to ROS generation as shown by the increased fluorescence of DCF, which was quenched by pre-treatment with the FR extract in a dosedependent pattern (Fig. 4).

Fig. 2. Time-dependent effect of FR pre-treatment on t-BHP-induced cell damage in rat hepatocytes. Cells were pre-incubated with 200 Ag/ml of FR extract for different time periods prior to t-BHP treatment. Control group received neither FR nor t-BHP treatment. Cell damage was evaluated by (upper) MTT assay, and (lower) LDH leakage. Values are expressed as mean F SD (n = 4). *p < 0.05, **p < 0.01, compared with cells treated with t-BHP only.

M.-H. Yau et al. / Life Sciences 72 (2002) 329–338

335

Fig. 3. Effect of FR pre-treatment on t-BHP-induced (left) lipid peroxidation and (right) GSH depletion in rat hepatocytes. Cells were treated with 500 AM t-BHP for 3 h after pre-incubation with FR extract for 24 h. MDA leakage into culture medium was determined by TBARS assay. Cells were trypsinized and homogenized. Protein thiols were removed by TCA precipitation and GSH content in the supernatant was determined spectrophotometrically by DTNB. Values are expressed as mean F SD (n = 4). **p < 0.01, compared with cells treated with t-BHP only.

Characteristics of the protective activity of FR extract Table 3 shows the effects of different pre-treatments on the antioxidative activity of FR extract against t-BHP-induced hepatocyte damage. A sub-optimal dose of 50 Ag/ml FR extract was chosen for the investigation. Such a dose showed f 50% inhibitory effect on the t-BHP-induced cell damage (Fig. 1), making any change in protective activity after the pre-treatments more visible. It was found that neither

Fig. 4. Effect of FR pre-treatment on t-BHP-induced ROS formation. Cells were loaded with 20 AM of DCFDA (in DMSO, final concentration 0.2%) for 30 min prior to t-BHP treatment. Change in fluorescence intensity which reflects ROS content was measured fluorimetrically. Results are expressed as mean F SD (n = 4). *p < 0.05, **p < 0.01, compared with cells treated with t-BHP only.

336

M.-H. Yau et al. / Life Sciences 72 (2002) 329–338

Table 3 Effect of different pre-treatments on the protective activity of FR extract in rat hepatocytes against t-BHP-induced oxidative damage Treatment

% inhibition of LDH leakage

% viability of control

t-BHP only FR (untreated) FR (acidified) FR (alkalinized) FR (trypsinized) FR (ultrafiltered) FR (boiled)

0 71.4 78.3 78.4 73.4 57.8 72.8

20.6 71.9 78.9 81.3 77.3 67.5 79.4

F F F F F F F

7.2 3.2** 2.6** 2.4** 1.4** 1.6** 2.8**

F F F F F F F

2.7 3.6** 0.7** 5.7** 2.9** 3.3** 4.2**

A concentrated FR extract (500 Ag/ml) was pre-treated under various conditions as described in Methods. Afterwards, it was diluted and added to hepatocytes at a final concentration of 50 Ag/ml. After incubation for 24 h, cells were treated with 500 AM t-BHP for 3 h. Results are expressed as mean F SD (n = 4). ** p < 0.01, compared with cells treated with t-BHP only.

acid, alkaline, trypsin nor high temperature treatments affected the antioxidative activity of FR extract. Over 80% of activity was found in the ultrafiltrate, showing that most of the active principles were small molecules with molecular weight below 1000.

Discussion In vitro free radical generating systems were used to assess the individual herbal components in WZ formula for their antioxidative activities. The aqueous extract of FR was found to be the most effective free radical scavenger among the aqueous extracts. It scavenged against different types of radicals, including superoxide, hydroxyl and peroxyl radicals. Such activity was obvious even at concentration as low as 10 Ag/ml. FR showed no pro-oxidant activity at concentrations below 200 Ag/ml in the presence of free transition metals. Thus, the aqueous extract of FR represented a source of potentially useful antioxidants worthy of further investigation. Nevertheless, whether the in vitro antioxidative activity is valid in living systems remains a question, since its cytotoxicity and bioavailability must be taken into consideration. In the present study, FR extract showed insignificant cytotoxicity at doses found to be effective in scavenging free radicals. FR extract, at a concentration of 50 Ag/ml, was capable of inhibiting t-BHPinduced cell damage, MDA formation, GSH depletion and ROS generation. This dose was 10-fold lower than the toxic dose of 500 Ag/ml in which over 85% cell viability can still be observed. Being non-toxic, the bioavailability of the active components also governs the overall activity. The effectiveness of pre-treatment and subsequent removal of the extract prior to the oxidative stress indicated that the active principles accumulated in hepatocytes in a time-dependent manner. Hence, the FR extract exerted its protective effect intracellularly, rather than extracellularly reacting with t-BHP in the culture medium. The use of the fluorescence probe also indicated that the scavenging occurs intracellularly. DCFDA allows sensitive, in situ monitoring of ROS production in different cellular systems. It is deacetylated to DCF by intracellular esterase and only shows characteristic fluorescence after being oxidized by ROS. FR treatment reversed t-BHP-induced ROS formation, indicating that the

M.-H. Yau et al. / Life Sciences 72 (2002) 329–338

337

active constituent uptaken into the cells was in close proximity to the site of ROS generation. Taking the observations from in vitro free radical scavenging studies and hepatocyte culture experiments together, aqueous extract of FR obviously inhibited t-BHP-induced cell damage through scavenging the ROS formed. At present, the active principle(s) in FR responsible for its antioxidative activity remains unclear. Nevertheless, the activity was resistant to heat, trypsinization and extreme pH conditions. Ultrafiltration study indicated that the activity was associated with relatively low molecular weight species, allowing them to have rapid access into cells. The FR extract contained 36% by weight of tannin. Preliminary results showed that the antioxidative activity of the FR extract remained associated with the tanninenriched fraction after passing the aqueous extract through the polyamide CC6 resin (data not shown). Incubation with bovine serum albumin, which is known to precipitate tannin [21], resulted in a significant decrease of the antioxidative activity (data not shown). These results suggested that tannin might be the active ingredient in the aqueous extract of FR, especially when it is known that tannins from other sources possess antioxidant activities [22]. Only limited information has been documented on the chemical composition of FR [23]. Fupenzic acid [24] and some diterpene glycosides [25,26] were identified from FR. Two tannins, namely lambertianins C and D, have been isolated from the leaves of Rubus chingii Hu [27]. However, the biological activities of these compounds remain undetermined. Compared with the aqueous extracts of other herbal components of the WZ formula, FR possessed the most potent antioxidative activity. Using methanolic extract of the same herbs might yield different results because of the different constituents contained in the aqueous and methanolic extracts. Further studies using the methanolic extracts of the herbal components may lead to the identification of other less water-soluble antioxidants. An aqueous extract of FR has been reported to decrease the levels of luteinizing hormone, follicle stimulating hormone and estradiol in rats, whereas the levels of luteinizing hormone-releasing hormone and testosterone were increased in the blood [28]. However, its antioxidative activities have not been reported. The present study showed that the aqueous extract of FR possessed potent antioxidative activity in both in vitro systems, including primary rat hepatocytes subjected to t-BHP-induced oxidative stress. Further studies on isolating, identifying and characterizing the active antioxidant constituent(s) may provide useful leads in understanding the pharmacological effects of this herbal medicine.

Acknowledgements The authors would like to thank Dr. S.P. Ip, School of Chinese Medicine, The Chinese University of Hong Kong for helpful discussion; Ms. L. Ang, Ms. N.Y. Huen and Mr. W.T. Tsang for their technical assistance.

References [1] Chinese Pharmacopoeia 2000 edition (in Chinese). Guangzhou: Guangdong Science and Technology Press; 2000. [2] Wang XM, Xie ZF. A clinical study of the effect of wuzi yanzong solution in retarding aging process. Zhongguo Zhong Xi Yi Jie He Za Zhi (in Chinese) 1992;12:23 – 5. [3] Li Y, Wu Q, Liang Q, Han J, Liu Z. Effect of wuzi yanzong wan on serum glucose and lipid peroxide of serum and liver in alcohol-induced liver injury in rats. Journal of Guangzhou College of TCM (in Chinese) 1992;9:143 – 6.

338

M.-H. Yau et al. / Life Sciences 72 (2002) 329–338

[4] Huang G, Luo J, Wei Y. Leukogenic effect and antioxygen radicals function of fufang wuzi yanzong pills. Zhongguo Zhong Yao Za Zhi (in Chinese) 1991;16:414 – 6, 447. [5] Gebhardt R. Antioxidative and protective properties of extracts from leaves of the artichoke (Cynara scolymus L.) against hydroperoxide-induced oxidative stress in cultured rat hepatocytes. Toxicology and Applied Pharmacology 1997; 144:279 – 86. [6] Wang CJ, Wang JM, Lin WL, Chu CY, Chou FP, Tseng TH. Protective effect of Hibiscus anthocyanins against tert-butyl hydroperoxide-induced hepatic toxicity in rats. Food and Chemical Toxicology 2000;38:411 – 6. [7] Lin WL, Wang CJ, Tsai YY, Liu CL, Hwang JM, Tseng TH. Inhibitory effect of esculetin on oxidative damage induced by t-butyl hydroperoxide in rat liver. Archives of Toxicology 2000;74:467 – 72. [8] Davies MJ. Detection of peroxyl and alkoxyl radicals produced by reaction of hydroperoxide with rat liver microsomal fraction. Biochemical Journal 1989;257:603 – 6. [9] Minotti G, Borrello S, Palombini G, Galeotti T. Cytochrome P-450 deficiency and resistance to t-butyl hydroperoxide of hepatoma microsomal lipid peroxidation. Biochimica et Biophysica Acta 1986;876:220 – 5. [10] Naohiko M, Marlene EL, Serroni A, Farber JL. Mitochondrial damage as a mechanism of cell injury in the killing of cultured hepatocytes by tert-butyl hydroperoxide. Archives of Biochemistry and Biophysics 1989;270:672 – 80. [11] Martin C, Martinez R, Navarro R, Ruiz-Sanz JI, Lacort M, Ruiz-Larrea MB. tert-Butyl hydroperoxide-induced lipid signaling in hepatocytes: involvement of glutathione and free radicals. Biochemical Pharmacology 2001;62:705 – 12. [12] Ponti V, Dianzani MU, Cheeseman K, Slater TF. Studies on the reduction of nitroblue tetrazolium chloride mediated through the action of NADH and phenazine methosulphate. Chemico – Biological Interactions 1978;23:281 – 91. [13] Zhou YC, Zheng RL. Phenolic compounds and an analog as superoxide anion scavengers and antioxidants. Biochemical Pharmacology 1991;42:1177 – 9. [14] Zhang A, Zhu QY, Luk YS, Ho KY, Fung KP, Chen ZY. Inhibitory effects of jasmine green tea epicatechin isomers on free radical-induced lysis of red blood cells. Life Sciences 1997;61:383 – 94. [15] Chan TY, Tang PL. Characterization of the antioxidant effects of melatonin and related indoleamines in vitro. Journal of Pineal Research 1996;20:187 – 91. [16] Seglen PO. Preparation of isolated rat liver cells. In: Prescott DM, editor. Methods in Cell Biology, vol. 19. New York: Academic Press; 1976. p. 29 – 83. [17] Alpini G, Phillips JO, Vroman B, LaRusso NF. Recent advances in the isolation of liver cells. Hepatology 1994;20: 494 – 514. [18] Wey HE, Pyron L, Woolery M. Essential fatty acid deficiency in cultured human keratinocytes attenuates toxicity due to lipid peroxidation. Toxicology and Applied Pharmacology 1993;120:72 – 9. [19] Brehe J, Burch H. Enzymatic assay for glutathione. Analytical Biochemistry 1976;74:189 – 97. [20] Shen HM, Shi CY, Shen Y, Ong CN. Detection of elevated reactive oxygen species level in cultured rat hepatocytes treated with aflatoxin B1. Free Radical Biology and Medicine 1996;21:139 – 46. [21] Rider PJ, Der-Marderosian A, Poter JR. Evaluation of total tannins and relative astringency in teas. In: Ho CT, Lee CY, Huang MT, editors. Phenolic compounds in food and their effects on health I. Analysis, occurrence and chemistry. Washington, DC: American Chemical Society; 1992. p. 93 – 9. [22] Wei T, Sun H, Zhao X, Hou J, Hou A, Zhao Q, Xin W. Scavenging of reactive oxygen species and prevention of oxidative neuronal cell damage by a novel gallotannin, pistafolia A. Life Sciences 2002;70:1889 – 99. [23] Xu Z, Zhao J. Chemical constituents of Fu-pen-zi (Rubus chingii). Zhongcaoyao (in Chinese) 1981;12:19. [24] Hattori M, Kuo KP, Shu YZ, Tezuka Y, Kikuchi T, Nambe T. A triterpene from the fruits of Rubus chingii. Phytochemistry 1988;27:3975 – 6. [25] Chou WH, Toshihiko O, Fujiko K, Kenji M, Chen FH, Osamu T. Diterpene glycosides from leaves of Chinese Rubus chingii and fruits of R. suavissimus, and identification of the source plant of the Chinese folk medicine ‘‘Fu-pen-zi’’. Chemical and Pharmaceutical Bulletin 1987;35:3021 – 4. [26] Tanaka T, Kawamura K, Kitahara T, Kohda H, Tanaka O. ent-Labdane-type diterpene glycosides from the leaves of Rubus chingii. Phytochemistry 1984;23:615 – 21. [27] Tanaka T, Tachibana H, Nonaka G, Kishioka I, Hsu FL, Kohda H, Tanaka O. Tannins and related compounds. CXXII. New dimeric, trimeric and tetrameric ellagitannins, lambertannins A – D from Rubus lambertianus. Chemical and Pharmaceutical Bulletins 1993;41:1214 – 20. [28] Chen K, Fang J, Kuang X, Mo Q. Effects of the fruit of Rubus chingii Hu on hypothalamus-pituitary-sex gland axis in rats. Zhongguo Zhong Yao Za Zhi (in Chinese) 1996;21:560 – 2.