Cytotoxic properties of thiophenes from Echinops grijissi Hance.
The dichloromethane fraction of the crude ethanol extract of Echinops grijissi Hance roots exhibited different cytotoxicity against a panel of four human tumor cell lines, HepG2, K562, HL60 and MCF-7. By a bioassay-guided fractionation, eight thiophenes were isolated from the dichloromethane fraction, one of them was isolated from the plant for the first time. And they were assayed for their toxicity against the cell lines in order to compare their relative anti-tumor activity and find candidates of potential anti-tumor drugs. The dichloromethane fraction and isolated thiophenes showed different activity against the cell lines, and the anti-tumor activity of the dichloromethane fraction was also studied in vivo in S180 implanted ICR mice, however, it exhibited no anti-tumor activity at dosage of 20 or 40 mg/kg/d. There were also deaths recorded and the animals showed signs of toxicity when the dosage is 200 mg/kg/d.
[C] 2007 Elsevier GmbH. All rights reserved.
Keywords: Cytotoxicity: Echinops grijissi; Compositae; Thiophenes: Human tumor cell lines: Anti-tumor activity
The genus Echinops belongs to the family Compositae and comprises over 120 species, of which 17 are known to occur in China, E. grijissi is mainly distributed in the southeastern of the country (Shih, 1987). The root of E. grijissi (commercial Chinese name: Yuzhou Loulu) is listed in Chinese Pharmacopeia and is used to clear heat, expel miasma and stimulate milk secretion for a long history (China Pharmacopoeia Committee, 2005). Previous chemical investigation on the root of E. grijissi demonstrated the presence of thiophenes (Lu et al., 1989; Guo et al., 1992a, b; Koike et al., 1999; Lin et al., 1999; Liu et al., 2002) and essential oil (Guo et al., 1994).
Thiophenes, a type of natural products, have been proven to possess several activities of anti-tumor (Lambert et al., 1991; Marles et al., 1992), insecticide (Nivsarkar et al., 1991; Sharma and Goel, 1994), antivirus (Marles et al., 1992; Hudson et al., 1993) and antiinflammatory (Lin et al.,1992). [alpha]-Terthiophene and 15 synthetic derivations, exhibit extensive phototoxicity due to their unsaturated conjugated chain, high near-ultraviolet absorption and quantum yield of oxygen, are considered to be the major pharmacologically active components of E. grijissi against tumor (Marles et al., 1992). However, no further literatures about the relative anti-tumor activity of thiophenes isolated from the root of E. grijissi have been published.
The present paper describes the bioactivity-guided fractionation of the crude ethanol extract of E. grijissi roots, leading to the isolation of eight thiophenes, and compd. 7 is isolated from this plant for the first time. These isolated thiophenes exhibit different tumor inhibitory effects against the human tumor cell lines, respectively. And the dichloromethane fraction shows no anti-tumor activity in S180 implanted ICR mice, but there are deaths recorded and the animals show signs of toxicity during the experimental period.
Materials and methods
The roots of E. grijissi were collected in Bozhou, north of Anhui Province, People's Republic of China, in June, 2006. The plant material was identified by authors, and a voucher specimen (EGH060703) has been deposited in the herbarium of the Institute of Pharmaceutical Informatics, College of Pharmaceutical Sciences of Zhejiang University.
The (1) H-and (13) C-NMR, HMQC and HMBC spectra were recorded in CDC[1.sub.3] on Bruker Biospin AMX 400 MHz FT spectrometer. Silica gel (300-400 mesh; Qingdao Chemical Group; Qingdao, China) was employed for column chromatography, while chromatographic fractions as well as pure compounds were monitored through thin layer chromatography (TLC) (GF254, Qingdao Chemical Group; Qingdao, China), detected by TLC UV analysis Video System (Yoko-zs, Wuhan New Technique Company Ltd., Wuhan, China) at light 254 and 365 nm, and color reaction by spraying with a solution of [H.sub.2]S[O.sub.4]/EtOH (10%:90%) followed by 5min heating at ca. 100[degrees]C. The preparative HPLC experiments were performed on Agilent 1100 HPLC system (Agilent Technologies, Waldbronn, Germany) using column Zorbax SB-[C.sub.18] (21.2 mm x 250 mm, 10 [mirco]m). The colorimetric assay were performed with Microplate reader (EL x 800, BIO-TEK Instruments. Inc., USA).
Extraction and isolation
The air-dried roots (14.3 kg) were split to pieces, the split plant material was extracted with 95% ethanol for three times, each for 3h at 85[degrees]C. The extract was concentrated to give a brown residue after evaporation of solvent, which was suspended in the distilled water (41) and then partitioned with dichloromethane (3x21) and n-butanol (3x21) successively. The dichloromethane fraction (132.2 g) was subjected to column chromatography over a silica gel (5 x 50 cm, 300-400 mesh, 1.0kg), eluted with petroleum ether (60-90[degrees]C)/ EtOAc to obtain Frs. A--J (1:0, 200:1, 100:1, 50:1, 30:1, 20:1, 10:1, 5:1, 1:1, 0:1, each 31). Frs. A (24.6g) was separated on a silica gel column (5 x 50 cm, 300-400 mesh, 300g) eluted with petroleum ether/EtOAc (1:0, 200:1, 100:1, 50:1, 20:1, 10:1, each 11) to obtain two compounds: 5-(but-3-en-l-ynyl)-2,2'-bithiophene (1, 1.2g), [alpha]-terthiophene (2, 3.l g). Frs. C and D (total 35.0g) were combined according to the TLC analysis and separated on a silica gel column (5 x 50cm, 300-400 mesh, 300g) eluted with petroleum ether/EtOAc (100:1, 50:1, 20:1, 10:1, each 11), then CD-1 was separated by preparative HPLC using acetonitrile/[H.sub.2]O (90%/10%) as the elution, and obtain two compounds: isoardopatine (3, 2.1 mg), 5-(4-isovaleroyloxybut-l-ynyl)-2,2'-bithiophene (4, 5.6 mg). Frs. E (5g) was separated by preparative HPLC using acetonitrile/[H.sub.2]O (85%/15%) as the elution, obtain four compounds: (2-(3,4-dihy-droxybut-l-ynyl)-5-(prop-l-ynyl)-thiophene (5, 4.9 mg), 5-(3,4-diacetoxybut-l-ynyl)-2,2'-bithiophene (6, 7.6mg), 5-(penta-1,3- diynyl)-2-(3,4-diacetoxy-but-l-ynyl)-thiophene (7, 5.7 mg), 5-(3-acetoxy-4-diacetoxy-but-l-ynyl)-thiophine (7, 5.7 mg), 5-(3-acetoxy-4-isovalero-yloxybut-l-ynyl)-2,2'-bithiophene (8, 11.3mg).
The human hepatocarcinoma cell lines HepG2 and MFC-7 were cultured in DMEM medium (Gibco BRL) with 3.7g/1 sodium bicarbonate, supplemented with 10% fetal bovine serum (FBS), 1% non-essential amino acid, and seeding at a density no more than [10.sup.6] cells/ml. The human acute myeloid leukemia cell line HL-60 and human chronic myelogenous leukemia cell line K562 were maintained in RPMI-1640 medium (Gibco BRL) with 3.7g/l sodium bicarbonate, supplemented with 10% heat-inactivated FBS.
In the cytotoxicity assays, for HepG2, K562 and MCF-7 cell lines, adriamycin at concentration of 4 [micro]g/ml was choose as reference drug, while HL60 is platinol, with the final concentration is 4 [micro]g/ml, too. All the above cell lines were purchased from Shanghai Institute of Cell Biology, Chinese Academy of Sciences (Shanghai, China), seeded in 96-well tissue culture plates, and maintained in a humidified atmosphere of 5% C[O.sub.2] and 95% air at 37[degrees]C for 3-6 days before experimentation.
The plant extract, each isolated thiophene and reference drug were transferred to overnight cells. Non-treated cells were used as control. Cells were cultured overnight and were then subjected to colorimetric assay. Cytotoxicity was expressed as the cell inhibitory rate at different concentrations of samples. [IC.sub. 50] was calculated as the dose at which 50% cell death occurred relative to the untreated cells.
In the colorimetric assay, for HL-60 and K562, cell multiplication was evaluated by determining the [IC.sub.50] using a modification of the sulforhodamine B (SRB; Sigma, MI, USA) assay (Chen et al., 1997). Briefly, after removing the medium, the cultures (per well) were fixed to the plastic substratum by gently adding 100 [micro]l of a 40% trichloracetic acid (TCA) in 0.9% NaCl solution. The cultures were incubated for 1 h at 4[degrees]C, and then washed five times with water to remove the TCA. Plates were air-dried and then stained for 20 min with 0.4% SRB in 1.0% acetic acid (HPLC grade; TEDIA Company Inc., USA). SRB was then removed and the cultures were quickly rinsed five times with 1.0% acetic acid to remove unbound dye. The cultures were air-dried and the bound dye was solubilized with 150[micro]l of 10mmol/l unbuffered Tris base (pH 10.5). Absorbance was read at 490 nm using a microplate reader.
For HepG2 and MCF-7, cell proliferation was assessed on the 6th day using the MTT assay (Carmichael et al., 1987). Briefly, after each cell cultivation experiment, l00 [mu]l (5mg/ml) MTT solution was added per well and the cells were cultured for another 4h. When the incubation was ended, the supernatant fluid was discarded. The cells were dissolved with in 150 [micro]l DMSO per well, agitated for 10min and the optical density (OD) of the resultant solution was assayed with a microplate reader at a test wavelength of 550 nm.
Tumor control rate = Average tumor weight of control group--Average tumor weight of treatment group / Average tumor weight of control group x 100%.
Eight-weeks old male ICR mice, with an average weight of about 20g, were used in the study. The handling of the mice was according to the guidelines of the Public Health Service Policy on Humane Care and Use of Laboratory Animals and was approved by the Institutional Animal Care and Used Committee (IA-CUC) of Zhejiang Province, with ACUC number SCXK2003-0001. The mouse was killed by cervical dislocation just I week after tumor strain S180 was transplanted through intraperitoneal injection, then soaked in 0.1% bromogeramine for 5min, and placed on the ice plate. The ascites was drawn and diluted with three-fold of physiological saline, then the dilution was inoculated 0.2 ml per mouse subcutaneously. The inoculated mice were randomly assigned to the control and treatment group with 10 mice in each group. They were housed in the Animal Resource Unit, College of Pharmaceutical Sciences of Zhejiang University. The mice were housed individually per cage and kept for at least 5 days prior to the start of the study. They were maintained at an environmental condition of about 27[degrees]C, with 12 h of an alternate light and dark cycle. The mice were given unlimited supply of standard diet and water. The treatment group was given the dichloromethane fraction at three doses of 20, 40, 200 mg/kg BW while the control group received cyclophosphamide (CTX) as reference drug. The solvent for dissolution of the extract and CTX was Tween 80 and distilled water, both groups were administered orally using an intubation needle delivered in a 2 ml volume per day. The animals were observed systematically after the administration and daily thereafter for 10 days. Body weight was monitored at the same time of the day every 2 days throughout the test period. The mice were killed by cervical dislocation just 24h after the last administration, and weighed. The tumor massa was sliced and weighed immediately, tumor control rate is calculated by the equation below:
The difference in cell cytotoxicity and S180 tumor strain assay were compared using Student's t-test, p < 0.05 were considered as significant.
The cytotoxicity of the dichloromethane fraction and reference drug were evaluated in vitro against four human tumor lines, HepG2, K562, HL60 and MCF-7, respectively. Our cytotoxicity analysis of the dichloromethane fraction shows a direct dose-response result (Fig. 1). According to Fig. I, the dichloromethane fraction showed significant cell cytotoxicity against cell lines HL60, K562 and HepG2. For HL60, the highest cell inhibitory rate was 86% at the concentration 10 [micro]/ml, so the [IC.sub.50] was 5 [micro]g/ml, while the highest cell inhibitory rate of K562 was 86% at the concentration of 50 [micro]g/ml, with the [IC.sub.50] was 30 [micro]g/ml. And the highest cell inhibitory rate of HepG2 was 76% at the concentration of 50 [micro]g/ml, and its [IC.sub.50] was 38 [micro]g/ml. However, for MCF-7, the cell cytotoxicity of the dichloromethane fraction was not obvious compared with the above cell lines, the highest cell inhibitory rate was only 18% at the concentration of 50 [micro]g/ml.
[FIGURE 1 OMITTED]
According to the bioactivity-guided fractionation, the dichloromethane fraction was led to the isolation of eight thiophenes (Fig. 2). The structure of all thiophenes was determined by spectroscopic methods and confirmed by comparison with literature data (Selva et al., 1978; Lu et al., 1989; Menelaou et al., 1991; Guo et al., 1992a, b; Gutierrez and Herz, 1990; Lin et al., 1999).
[FIGURE 2 OMITTED]
Eight thiophenes were tested for anti-tumor activity against four human tumor cell lines, K562, HL60, HepG2, and MCF-7, respectively. Their [IC.sub.50] against different cell lines were calculated as the dose at which 50% cell death occurred relative to the untreated cells (Table 1). All the tested compounds except for compd. 3 show different cytotoxicity against the four cell lines. So the results of compd. 3 against the four cell lines are not described in the paper due to the cell inhibitory rate were not evident at all.
Table 1. [IC.sub.50] values for cytotoxicity of the dichloromethane fraction and seven thiophenes against different cell lines Cell DCM [IC.sub.50] lines * ([mu]g/ml) Compd. 1 Compd. 2 Compd. 4 ([mu]g/ml) ([mu]g/ml) ([mu]g/ml) HL60 5 36 - 10 K562 30 18 25 48 MCF-7 - - 50 50 HepG2 38 - 10 2 Cell [IC.sub.50] lines * Compd. 5 Compd. 6 Compd. 7 Compd. 8 ([mu]g/ml) ([mu]g/ml) ([mu]g/ml) ([mu]g/ml) HL60 8 12 41 45 K562 - 12 47 7 MCF-7 - - - - HepG2 - 18 50 1.8 "-" means 1[C.sub.50] > 50 [mu]g/ml. * For HL60, cell inhibitory rate of positive control (platinol) was 83% at [mu]g/ml. For K562, MCF-7 and HepG2, cell inhibitory rate of positive control (adriamycin) was 91%, 81%, and 90% at 4 [mu]g/ml, respectively.
In the cytotoxic analysis of HL60 (Fig. 3), compds. 4, 5 and 6 exhibit significant cell inhibitory effect, with the [IC.sub.50] is 10, 8 and 12 [micro]g/ml, respectively. Compds. 1, 7 and 8 show evident cytotoxicity against HL60 at high concentrations, so their [IC.sub.50] is 36, 41 and 45 [micro]g/ml, respectively. However, the cell inhibitory effect of compd. 2 against HL60 was not obvious.
[FIGURE 3 OMITTED]
For the cell line K562 (Fig. 4), compds. 1, 6 and 8 show significant inhibitory effect, with their [IC.sub.50] is 18, 12, and 7[micro]g/ml, correspondingly. And compd. 7 exhibits evident cytotoxicity against K562 at high concentration, so the [IC.sub.50] is 47 [micro]g/ml. Besides, compds. 2,4 and 5 display moderate toxicity ([IC.sub.50] > 50 [micro]g/ml), respectively.
[FIGURE 4 OMITTED]
For the cell line MCF-7 (Fig. 5), compds. 2 and 4 exhibit moderate cytotoxicity against MCF-7, correspondingly. However, compds. 1, 5, 6, 7 and 8 show no obvious cell inhibitory effect against MCF-7.
[FIGURE 5 OMITTED]
For the cell line HepG2 (Fig. 6), compds. 2, 6 and 8 display significant cytotoxicity against HepG2, with their [IC.sub.50] is 10, 18 and 1.8 [micro]g/ml, respectively. Compds. 1, 4 and 7 exhibit moderate toxicity, respectively. Besides, compd. 5 shows no evident toxicity against HepG2.
The dichloromethane fraction shows no evident antitumor activity in animal experiment (Table 2), but there were deaths recorded and the animals showed signs of toxicity when treated with dose of 200 mg/kg, in which 8 in 10 mice died during the experimental period.
[FIGURE 6 OMITTED]
Table 2. Tumor inhibitory effect of the dichloromethane fraction on the body weight and the incubated tumor weight of ICR mice Doses Mouse Group (mg/kg) amount Body weight (g) Before After experiment experiment Blank 0 10 23.2 [+ or -] 1.2 33.3 [+ or -] 3.8 CTX 100 10 22.4 [+ or -] 0.6 30.0 [+ or -] 2.2 Treatment 1 40 10 21.1 [+ or -] 0.9 34.8 [+ or -] 2.4 Treatment 2 20 10 21.5 [+ or -] 1.2 35.6 [+ or -] 2.9 Group Weight of tumor (g) Tumor inhibitory rate (%) Blank 0.828 [+ or -] 0.366 0 CTX 0.281 [+ or -] 0.127 66.1 Treatment 1 1.048 [+ or -] 0.588 -26.5 Treatment 2 0.815 [+ or -] 0.361 1.6
The structure-activity studies of photoactivated cytotoxic tricyclic thiophenes described that singlet oxidation speed is the determinative factor of antitumor. [alpha]-Terthiophene and 15 synthetic derivations show strong cell cytotoxicity under the radiation of ultraviolet, and only five thiophens exhibit obvious cytotoxicity without the ultraviolet (Marles et al., 1992).
Thiophenes are not stable due to its strong conjugate system in some conditions, such as the radiation of ultraviolet, high oxidation reagents (Jiang et al., 2000; Koopmans et al., 1995). In this study we isolate and identify eight thiophenes from the dichloromethane fraction of the crude ethanol extract of E. grijissi roots that show different cytotoxicity effect against four human tumor cell lines without the radiation of ultraviolet.
According to Table1, compds. 4, 5 and 6 show high cell inhibitory rate against HL60, with their [IC.sub.50] is 10, 8 and 12[mu]g/ml, respectively, while [IC.sub.50] of the dichloromethane fraction against HL60 is 5 [mu]g/ml. Therefore, these three thiophenes can be considered to be the major pharmacologically active components against HL60. In the cytotoxic analysis of K562, compds. 6 and 8, with [IC.sub.50] is 12 and 7 [micro]g/ml, respectively, while [IC.sub.50] of the dichloromethae fraction against K562 is 30 [micro]g/ml, do these two thiophenes can be considered to be the major pharmacologically active components against K562. In the cell inhibitory rate evaluation of HepG2, compd. 8 shows extremely low [IC.sub.50], while the [IC.sub.50] of compd. 2 is also lower than that of the dichloromethane fraction and other five thiophenes, so we believe compds. 8 and 2 are two main inhibitory influences on cytotoxicity against HepG2. Based on the discussion above, two acyl-substituted thiophenes, compds. 6 and 8 are promising thiophenes that may be candidates of potential anti-tumor lead compounds by further structure modification and pharmaceutical formulation design in the future study. Besides, we presume that the cytotoxicity of thiophenes is relative to the degree of thiophene polymerization and the substitution of side chain, because acyl-substituted bithiophenes exhibit stronger anti-tumor activity than [alpha]-terthiophene and monothiophenes without the radiation of ultraviolet in our experiment.
In our in vivo animal experiment, it is interesting that the dichloromethane fraction shows no significant antitumor activity. However, there are deaths recorded and the animals show signs of toxicity when treated with high dosage.
We thank Mr. Chen, Department of Chemistry, for the NMR measurements.
Carmichael, J., DeGraff, W.G., Gazadar, A.F., Minna, J.D., Mitchell, J.B., 1987. Evaluation of a tetrazolium-based semiautomated colorimetric assay: assessment of chemo-sensitivity testing. Cancer Res. 47, 943-946.
Chen, Z.P., Malapetsa, A., McQuillan, A., 1997. Evidence for nucleotide excision repair as a modifying factor of MGMT mediated innate chloroethylnitrosourea resistance in human tumor cell lines. Mol. Pharmacol. 52, 815-820.
Guo, D.A., Cui, Y.J., Lou, Z.C., Gao, C.Y., Huang, L.R., 1992a. The chemical constituents of Echinops grijissi Hance (I). Chin. Trad. Herbal Med. 23, 3-5.
Guo, D.A., Lou, Z.C., Gao, C.Y., Wang, D., 1992b. The chemical constituents of Echinops grijissi Hance (II). Chin. Trad. Herbal Med. 23, 512-514.
Guo, D.A., Lou, Z.C., Liu, Z.A., 1994, Studies on the naphthic constituents in the roots of Echinops grijisii. China J. Chin. Mater. Med. 19 (2), 100-101.
Gutierrez, A.B., Herz, W., 1990. Ambrosanolides and other constituents of two Rudbeckia species. Planta Med. 56, 295-297.
Hudson, J.B., Harris. L., Teeple, A., Towers, G.H., 1993. The anti-HIV activity of the phytochemical alpha-terthienyl Antiviral Res. 20. 33-43.
Jiang, A.G., Liu. Y.G., Huang, D.Y., 2000. AMI study of thiophene 1,1-dioxides compounds with large second-order polarizabilities and high thermal and photo stability. J. Mol. Struct. (Theochem) 499. 203-206.
Koike, K., Jia, Z.H., Nikaido, T., Liu, Y.. Zhao, Y.Y., Guo, DA., 1999. Echinothiophene, a novel benzothiophene glycoside from the roots of Echinops grijissi [J]. Org. Lett. 1 (2), 197-198.
Koopmans, M.P., Damate. J.S., Lewan, M.D., De Leeuw, J.W., 1995. Thermal stability of thiophene biomarkers as studied by hydrous pyrolysis. Org. Geochem. 23, 583-596.
Lambert, J.D., Campbell, G.. Amason, J.. Majak, W. 1991. Herbicidal properties of alpha-terthienyl, a naturally occurring phototoxin. Can. J. Plant Sci. 71, 215-218.
Lin, C.C., Lin, C.H., Chiu, H.F.. 1992. The pharmacological and pathological studies on Taiwan folk medicine(VII): the anti-inflamatory effect of Echinops grljisii. Am. J. Chin. Med. 20(2), 127 134.
Lin, Y.L., Huang, R.L., Kuo, Y.H., 1999. Thiophenes from Echinops grijissi. Hance Chin. Pharm. J. 51, 201-211.
Liu. Y.. Ye, M. Guo, H.Z., Zhao. Y.Y., Guo, D.A. 2002. New thiophenes from Echinops grijissi. J. Asian Nat. Prod. 4(3), 175-178.
Lu, H.C., Wang, S.X., Zhu, T.R., 1989. Studies on the chemical constituents of Radix echinopsis. Chin. Trad. Herbal Med. 20, 2-5.
Marles, R.J., Hudson, J.B., Graham, E.A., Arnason, J.T., 1992. Photochem. Photobiol. 56, 479-487.
Menelaou, M.A., Fronczek. F.R., Hjortso, M.A., Morrison, A.F., 1991. NMR spectral data of benzofurans and bithiophenes from hairy root cultures of Tagetes patula and the molecular structure of isoeuparin. Spectrosc. Lett 24(10), 1405-1413.
National Pharmacopoeia Committee, 2005. Pharmacopoeia of PRC. Chemical Industry Press, Beijing, p. 214.
Nivsarkar. M., Kumar. G.P., Laloraya, M., 1991. Super-oxidedismutase in the anal gills of the mosquito larvae of Aedes aegypti: its inhibition by [alpha]-tcrthienyl. Arch. Insect Biochem. Physiol. 16, 249-255.
Selva., A., Arnone, A., Mondelli. R., Vincenzo, S., 1978.Cardopatine and isocardopatine, two novel cycleobutane substances from Cardopatium corymbosum. Phytochemistry 17, 2097-2100.
Sharma, A., Goel, H.C, 1994. Some naturally occurring phytophoto-toxins for mosquito control. Indian J. Exp. Biol. 32, 745-751.
Shih, C, 1987. In: Flora Sinica. Scientific Press. Beijing. Vol. 78, pp. 1-3.
WenrongJin, QiangShi, ChengtaoHong, YiyuCheng, ZhongjunMa *, Haibin Qu *
Institute of Pharmaceutical Informatics, College of Pharmaceutical Sciences of Zhejiang University, Hangzhou, Chinu
* Corresponding authors. Tel: +8657188208427; fax: +86 571 88208428.
E-mail address: firstname.lastname@example.org (Z. Ma).
|Printer friendly Cite/link Email Feedback|
|Author:||Jin, Wenrong; Shi, Qiang; Hong, Chengtao; Cheng, Yiyu; Ma, Zhongjun; Qu, Haibin|
|Publication:||Phytomedicine: International Journal of Phytotherapy & Phytopharmacology|
|Article Type:||Clinical report|
|Date:||Sep 1, 2008|
|Previous Article:||Wound healing activity and docking of glycogen-synthase-kinase-3-[beta]-protein with isolated triterpenoid lupeol in rats.|
|Next Article:||Antiproliferative activity of triterpenoids from Eclipta prostrata on Hepatic stellate cells.|