Yu-Kan Liu, Wei Shen

Yu-Kan Liu, Wei Shen, Department of Gastroenterology, the Second affiliated hospital, Chongqing University of Medical Sciences, Chongqing 400010, China
Correspondence to: Wei Shen, Department of Gastroenterology, the Second affiliated hospital, Chongqing University of Medical Sciences, Chongqing 400010, China. shenwei2002932@sohu.com
Telephone: +86-23-63849076-2323 Fax: +86-23-63849076
Received: 2002-07-08 Accepted: 2002-10-21

Abstract
AIM: To investigate the inhibitive effect and its possible mechanism of Cordyceps Sinensis (CS) on CCl₄-plus ethanol-induced hepatic fibrogenesis in experimental rats.

METHODS: Rats were randomly allocated into a normal control group, a model control group and a CS group. The latter two groups were administered with CCl₄ and ethanol solution at the beginning of the experiment to induce hepatic fibrosis. The CS group was also treated with CS 10 days after the beginning of CCl₄ and ethanol administration. All control groups were given corresponding placebo at the same time. At the end of the 9th week, rats in each group were humanely sacrificed. Blood and tissue specimens were taken. Biochemical, radioimmunological, immunohistochemical and molecular biological examinations were used to determine the level change of ALT, AST, HA, LN content in serum and TGFb₁, PDGF, collagen I and III expression in tissue at either protein or mRNA level or both of them.

RESULTS: As compared with the model control group, serum ALT, AST, HA, and LN content levels were markedly dropped in CS group (86.0±34.4 vs 224.3±178.9, 146.7±60.2 vs 272.6±130.1, 202.0±79.3 vs 316.5±94.1 and 50.4±3.0 vs 59.7±9.8, respectively, P<0.05). Tissue expression of TGFb₁ and its mRNA, collagen I mRNA were also markedly decreased (0.2±0.14 vs 1.73±1.40, 1.68±0.47 vs 3.17±1.17, 1.10±0.84 vs 2.64±1.40, respectively, P<0.05). More dramatical drop could be seen in PDGF expression (0.87±0.43 vs 1.91±0.74, P<0.01). Although there was no statistical significance, it was still strongly suggested that collagen III mRNA expression was also decreased in CS group as compared with model control group (0.36±0.27 vs 0.95±0.65, P=0.0615). In this experiment, no significant change could be found in PDGF mRNA expression between two groups (0.35±0.34 vs 0.70±0.46, P>0.05).

CONCLUSION: Cordyceps sinensis could inhibit hepatic fibrogenesis derived from chronic liver injury, retard the development of cirrhosis, and notably ameliorate the liver function. Its possible mechanism involves inhibiting TGFb₁ expression, and thereby, down regulating PDGF expression, preventing HSC activation and deposition of procollagen I and III.

Liu YK, Shen W. Inhibitive effect of cordyceps sinensis on experimental hepatic fibrosis and its possible mechanism. World J Gastroenterol 2003; 9(3): 529-533

http://www.wjgnet.com/1007-9327/9/529.htm

INTRODUCTION
The incidence rate of chronic in China is high, which afflicts the patients by progressively developing into irreversible cirrhosis^[1,2]. Hepatic fibrosis is the intermediate and crucial stage of this process, characterized by reversibility. If treated properly in this stage, cirrhosis could be successfully prevented^[3].
   Clinical observation and experimental data suggested that liver fibrosis could be reabsorbed under certain conditions. Chinese herbs, well known for their definite effectiveness, cheap prices and negligible side effects, have particular advantages in therapeutic research of hepatic fibrogenesis. Several herbs were suggested recently by some reports to have preventive effect on hepatic fibrosis^[4-12], and cordycep sinensis (CS) is one of them^[4,5]. However, its exact effectiveness and detailed mechanisms have not been elaborated. In this study, we established the animal model of chronic liver injury-hepatic fibrosis-cirrhosis, intervened with CS, and observed its inhibitive effect. An array of indexes in protein and mRNA levels was established in order to thoroughly investigate its possible mechanism.

MATERIALS AND METHODS
Animals
Male Wistar rats weighing between 200 g and 300 g were obtained from Experimental Animal Center of ChongQing University of Medical Science, China. The rats were housed 3 or 4 per cage and subjected to 12-day/12-night cycle with unrestrictive access to basic food. All animals were treated humanely according to the national guideline for the care of animals in the country.

Preparation for CS suspension
CS was purchased from Bao Ding Pharmaceutical Company, China.
The CS and double-distilled water were mixed in proportion of 1:3 and subject to full vibration.

Reagents
TGFb₁, PDGF, procollagen I and III RNA probe and detection kit for in-situ hybridization were purchased from Boster Biologic Technology Company, China. Anti- TGFb₁ monoclonal antibody, anti-PDGF multiclonal antibody and its detection kit for immunohistochemical assay were purchased from Santa Cluz biologic technology Company, USA. Serum ALT, AST, HA, and LN examinations were performed by the Laboratory Department of Chong Qing University of Medical Sciences, China.

Establishment of animal model: carbon tetrachloride (CCL4)-plus-ethanol induced hepatic fibrosis
Sixty-six male Wista rats were randomly assigned to a normal control group, a model control group and a CS group. At the beginning of the experiment, rats in model control group and CS group were subjected to hypodermic injection of (40 % in bean oil) at a dose of 0.3 ml/100 g of body weight twice a week. Besides, rats in these two groups also received 5 % ethanol solution as the only fluid to drink. Rats in normal control group received hypodermic injection of bean oil at the same dose and frequency as the other two groups and received water ad libitum. Ten days after the CCL₄ administration (for 3 times), CS group was given CS suspension orally at a dose of 1 ml/100 g body weight daily. In the meantime, three rats in model control group were randomly sacrificed to evaluate the liver histological change at this moment while other rats along with rats in normal control group were given saline orally at a dose of 1 ml/100 g body weight daily. All the administrations lasted 9 weeks.

Collection of specimens
At the end of the 9th week, rats in each group were humanely sacrificed by amobarbital sodium anesthesia. Midline laparotomy was performed. Livers were excised and blood was collected through cardiopuncture.

Histological grading
Liver tissues were fixed in formalin and embedded in paraffin blocks according to standard procedures (glass slide was cleaned with 95 % ethanol, treated with APES solution and air dried.) four to six micron thick tissue sections were cut using microtome and applied to slides; And deparafinized in xylenes using three changes for 5 minutes each. Hydrate sections gradually passed through graded alcohols: washing in 100 % ethanol twice, then 95 % ethanol twice for 10 minutes each; and washing in deodorized water for 1 minute. Hemotoxylin and eosin (HE) staining was performed according to the standard procedure.
   Fibro-proliferation degree of liver sections were graded numerically based on the criterion described below: grade 0: normal liver; grade 1: few collagen fibrils extend from the central vein and portal tract; grade 2: collagen fibrils extension are obvious but do not encompass the whole lobule; grade 3: collagen fibrils extend and encompass the whole lobule; grade 4: collagen fibrils extend and separate lobule into pseudo-lobule, but mainly shaped in square form; and grade 5: pseudo-lobule shaped mainly in circle form.
   Two pathologists who had no knowledge of their sources and each other's assessment examined stained slide independently.

Immunohistochemistry
Liver tissues were fixed in formalin and embedded in paraffin blocks according to standard procedures. Glass slide was cleaned with 95 % ethanol, treated with subbing solution and air-dried. Tissue sections of 4-6 micron thick were cut using microtome and applied to slides; and deparafinized in xylenes using three changes for 5 minutes each. Hydrate sections gradually through graded alcohols: washing in 100 % ethanol twice, then 95 % ethanol twice for 10 minutes each; and washing in deironized water for 1 minutes. Incubate sections for 15 minutes in 0.1 % pepsin at room temperature to expose the antigen masked by formalin fixation and paraffin embedding. Incubate sections with 3 % H₂O₂ and normal non-immunal goat serum for 10 minutes respectively to inactivate endogeneous peroxidase and biotin; incubate sections with primary antibody overnight at 4 ℃. Optimal antibody concentration was determined previously. Wash with three changes of PBS for 5 minutes each. Incubate for 20 minutes with biotin-conjugated secondary antibody and avidin biotin enzyme reagent respectively. Wash with three changes of PBS for 5 minutes each. Incubate in peroxidase substrate for 5 minutes, after that dehydrate through alcohols and xylenes. Immediately add 1-2 drops of permanent mounting medium and covered with glass coverlip.

In situ hybridization
The sequence of RNA probe is as follows:
1. TGFb₁mRNA
(1) 5'CGTTT CACCA GCTCC ATGTC GATGG TGTTG CAGGT-3'
(2) 5'CTTGA TTTTA ATCTC TGCAA GCGCA GCTCT GCACG-3'
(3) 5'TTGGT ATCCA GGGCT CTCCG GTGCC GTGAG CTGTG-3'
2. PDGF mRNA
(1) 5'CTCGG CTTCC TCGGC CAGAA CATGG GCGAG GTATC-3'
(2) 5'AACCT CACCT GGACT TCTTT CAATT TTGGC TTCTT-3'
(3) 5'TTGCA CTCGG CGATC ATGGC CGGCT CAGCA ATGGT-3'
(4) 5'GGCTC CAAGG GTCTC CTTCA GTGCC GTCTT GTCAT-3'
3. Col-I mRNA
(1) 5'CACAG ATCAC GTCAT CGCAC AACAC CTTGC CGTTG-3'
(2) 5'AGCTT CACCG GGACG ACCAG CTTCA CCAGG AGATC-3'
(3) 5'TCACT CCTTC TACAT TATAT TCAAA CTGGC PGCCA-3'
4. Col-III mRNA
(1) 5'ATTAA CAGAC TTGAG TGAAG TCATA ATCTC ATCGG-3'
(2) 5'AGAAT ACAAT CTGTG TTTCT GACCA GGTGA GGTAG-3'
(3) 5'GAAGG AGGAG AATCC CGTGG CATGC CAATG AATCT-3'

   In situ hybridization was performed as described elsewhere. Briefly, formalin-fixed and paraffin-embedded liver section slides were pretreated by incubation with 0.1 % pepsin, 3 % H₂O₂ and normal non-immunal bovine serum for 10-15 minutes respectively to expose signals masked by formalin fixation and paraffin embedding and inactivate endogeneous peroxidase and biotin. Antisense RNA probe was then added to the sections and incubated together in humidified chamber overnight at 37 ℃. After washing with three changes of 2×standard saline citrate (SSC) and 0.2×SSC for 10 minutes each, the sections were subject to incubation with biotin-conjugated secondary antibody and avidin biotin enzyme reagent respectively. Wash with three changes of PBS for 5 minutes each, incubate in peroxidase substrate for 20 minutes, and dehydrate through alcohols and xylenes. One to two drops of permanent mounting medium were immediately added and covered with glass coverlip.

Semiquantitative image analysis
Computer morphometry (CM-2000B Medical Image Analysis Software, Beijing Medical Software Company, China) was used to quantify the optical density of the signal generated by the immunohistochemical and in situ hybridization examination. The exact method is described as follows: The video image was generated with a video camera and digitalized for image analysis at 256 gray levels. An optical threshold and filter combination was set to select positive stains. The structure of interest was discriminated automatically by computer and measured for its optical intensity and total area. Staining index was obtained by multiplying the quantified number of optical intensity and total area.

Statistical analysis
Data were analyzed with SAS software. Quantitative data were presented as means ±SD and were compared using t test procedure. Frequency data were compared using NPAR1WAY procedure.

RESULTS
CS inhibits fibril deposition and ameliorates liver function of chronic hepatitis
After 10 days (3 times) of CCL4 administration, rats suffered hepatocyte lipoid degeneration, narcosis, and inflammatory cells infiltration, which fulfilled the diagnostic standard for chronic hepatitis.
   Specimens from model control group showed apparent formation of fibrotic septa, encompassing regenerated hepatocytes into pseudo-lobules. Regenerated hepatocytes underwent severe lipraoid degeneration. Specimens from CS group showed only slight fibrogenesis without pseudo-lobule formation.
   Although statistical analysis failed to present any significant disparity between quantitative data of histological grading of the two groups, it still indicated that fibrogenesis of CS group was much less severe than that of model control group (Table 1). Compared with model control group, serum contents of HA and LN in CS group were markedly decreased (P<0.05), which indicates from another perspective that CS could inhibit hepatic fibrogenesis (Table 2).
   Serum contents of ALT and AST in model control group were significantly elevated compared with both normal control and CS groups. However, there was no significant statistical difference between normal control and CS groups (P>0.05), which showed that CS could notably ameliorate liver function(Table 2).

CS reduces procollagen I and III mRNA synthesis
Procollagen I and III mRNA synthesis level in liver tissues were determined by in situ hybridization and quantified by computer image analyzing system. Positive staining could only be seen at central vein and periportal areas in normal control group. As the model control and CS groups, positive stains were distributed mainly along fibrotic septa other than central vein and periportal areas.
   Compared with model control group, staining index of CS group remarkably dropped (P<0.05), which suggested that procollagen I mRNA synthesis was strongly inhibited by CS treatment. On the part of procollagen III, despite that there was no significant difference between the two groups, it was also reasonably suggested that CS could inhibit the synthesis of procollagen III mRNA on the consideration of P value (0.0695) (Table 3).

CS reduces TGFb1 expression
TGFb₁ expression in protein and mRNA level were determined by immunohistochemistry and in situ hybridization, respectively. Positive stains were quantified with computer image analyzing system. For the sections of normal control group, positive staining of TGFb₁ and its mRNA was found at central vein and periportal areas. In the sections of model control group and CS group, positive staining could be seen at interstitial cells, inflammatory cells, impaired hepatocytes as well as normal hepatocytes. Fibrotic septa were only slightly stained. TGFb₁ mRNA positive stain distribution was not completely consistent with that of TGFb₁. More stain was found at fibrotic septa and less at impaired and normal hepatocytes.
   Compared with model control group, the staining index of both TGFb₁ and TGFb₁mRNA in CS group was markedly decreased (P<0.05, respectively), indicating that CS could inhibit TGFb₁ mRNA transcription and, thereby, reducing the TGFb₁ expression (Table 3).

CS reduces PDGF expression
PDGF expression in protein and mRNA level was determined, like TGFb₁, by immunohistochemistry and in situ hybridization respectively. Positive staining was quantified with computer image analyzing system. For the sections of normal control group, positive stains of PDGF and its mRNA could be seen at central vein and periportal areas. In the sections of model control group and CS group, positive stains mainly appeared at fibrotic septa.
   PDGF expression in CS group dropped dramatically compared with model control group (P<0.01). But for PDGF mRNA, there was no significant difference by statistical analysis. In spite of this, the means of two groups still suggested difference. To make sure of this, further studies are needed (Table 3).

Table 1 Histological grading

^aP<0.05 vs normal control group, ^bP<0.05 vs model control group.

Table 2 Serum content of HA, LN, ALT and AST (x±s)

^aP<0.05 vs normal control group, ^bP<0.05 vs model control group.

Table 3 Staining index of procollagen I, III mRNA, TGFb₁, TGFb₁ mRNA, PDGF and PDGF mRNA in liver tissues (x±s)

^aP<0.05, ^bP<0.01, and ^cP>0.05, vs model control group.

DISCUSSION
In this study, we demonstrated that CS could inhibit hepatic fibrogenesis and retard the development of cirrhosis by evaluating histological grading and serum contents of HA and LN. Its possible mechanism involves inhibiting the synthesis of TGFb₁mRNA and thereby downregulating the expression of TGFb₁ and PDGF, reducing the deposition of collagen I and III.
   CS is a type of traditional Chinese tonic that has already been demonstrated by modern pharmacological researches to have extensively positive effect on a few systems of human body^[13-19]. Recently, some reports suggested that this herb might also have preventive effect on hepatic fibrosis^[4,5]. But the real effect and mechanisms have not been elaborated. This study was designed to evaluate its exact effect on hepatic fibrosis and to investigate its possible mechanism.
   Various kinds of chronic liver injury widely spread all over the world and afflicted patients greatly. Effective ways to inhibit fibrogenesis and prevent the development of cirrhosis are of great clinical and academic significance. Although many new agents were being tested, no satisfactory agent with ascertained effectiveness and negligible side effects has appeared as yet. Traditional Chinese herbs were well known for their cheap prices and negligible side effects. Exploration in this area is promising.
   We started to treat the rats with CS after 10 days of CCL4 administration. Pathological evaluation showed that rats suffered chronic liver injury in this moment. CS treatment based on this disorder presented its inhibitive effect on preventing the development of cirrhosis. HA and LN are good serum markers of hepatic fibrogenesis^[20]. In this study, serum contents of HA and LN in CS group markedly dropped compared with model control group, which indicates that CS could successfully prevent hepatic fibrogenesis. Histological grading also supported this conclusion.
   To address the ways in which this herb yielded in a significant reduction in fibrosis, we investigated the effect of CS treatment in the expression of TGFb₁ as well as its mRNA. Overexpression of this cytokine was associated closely with fibrogenesis in many ways^[21-25]. It can promote HSC to synthesize collagen I and III, and simultaneously inhibit their decomposition by upregulating the expression of Tissue Inhibitor of Metalproteinase (TIMP), which neutralize the activity of Matrix Metalproteinase (MMP)^[26-28], an important degrading enzyme of collagen I and III. In addition, TGFb₁ could also indirectly promote the HSC proliferation by enhancing the expression of PDGF and its receptor^[21]. One strategy in the development of antifibrotic drug is the exploration of TGFb₁ inhibitors^[21]. Because TGFb₁ expression was regulated by diverse factors in transcription, post-transcription, secretion and releasing levels, the expression of its protein and mRNA varied considerably^[22]. Consequently, analyses in two levels were indispensable. In this study, we determined the expression of this cytokine by immunohistochemistry and in situ hybridization to investigate the effect of CS on TGFb₁ expression in the both levels. The results showed that both TGFb₁ and its mRNA expression remarkably decreased in CS group, indicating that CS could downregulate the expression of this important cytokine, which possibly contributed to the reduction of fibrosis.
   PDGF is another important cytokine that influences the development of fibrosis. According to the previous reports^[29-33], it is the most potent HSC-proliferation promoter, which plays an important role in fibrogenesis. In spite of its earlier identification and isolation, few pharmacological studies observed the effect of the potential agents on this cytokine. In this study, we initially observed the change of this cytokine responding to CS treatment. Compared with model control group, PDGF expression level in protein of CS group dramatically dropped, indicating that CS could inhibit the PDGF expression. Statistical analysis showed no significant difference of PDGFmRNA expression between the two groups. Whether CS exerts inhibitive effect on PDGF expression in mRNA or directly in protein still remains unclear. Further studies are needed to elucidate the detailed mechanisms.
   Pathological feature of hepatic fibrosis is the excessive deposition of ECM components^[34-36]. As the medium of parenchyma cells, constancy of ECM component is essential to the maintenance of liver function. Changes of proportion of ECM components, and the change of their quantities, cause the damage of hepatocytes and the deterioration of liver function. Syntheses of Collagen I and III increase greatly when fibrogenesis occurs, which are mainly responsible for the adverse effects brought about by ECM. As a result, Collagen I and III overshadow other components and become the most important ECM in the development of fibrosis. After nine weeks of CS treatment, expression of procollagen I and III decreased. On one hand, this result further manifests that CS has inhibitive effects on fibrosis; on the other hand, it might be one of the possible explanations for the amelioration of liver function.
   CS is a cheap and widely available herb that is well tolerated and has been used for centuries in traditional Chinese medicine without any side effect reported. In this study, we demonstrated that, administered at the stage of chronic hepatitis, CS could successfully inhibit hepatic fibrogenesis and retard the development of cirrhosis. Moreover, it can strikingly ameliorate the liver function. Therefore, we suggest that this herb should serve as a promising antifibrotic agent and deserves further investigations.

REFERENCES
1    Lamireau T, Desmouliere A, Bioulac-Sage P, Rosenbaum J. Mechanisms of hepatic fibrogenesis.
      Arch Pediatr 2002; 9: 392-405
2    Brenner DA, Waterboer T, Sung Kyu Choi, Lindquist JN, Stefanovic B, Burchardt E, Yamauchi M, Gillan A, Rippe RA.
      New aspects of hepatic fibrosis. J Hepatol 2000; 32 (Suppl 1): 32-38
3    Riley TR 3rd, Bhatti AM. Preventive strategies in chronic liver disease: part II. Cirrhosis. Am Fam Physician
      2001; 64: 1735-1740
4    Liu YK, Shen W. Effect of Cordyceps sinensis on hepatocytic proliferation of experimental hepatic fibrosis in rats. Shijie
      Huaren Xiaohua Zazhi 2002; 10: 388-391
5    Ma X, Qiu DK, Xu J, Zeng MD. Effects of Cordyceps polysaccharides in patients with chronic hepatitis C. Huaren Xiaohua
      Zazhi 1998; 6: 582-584
6    Yang Q, Yan YC, Gao YX. Inhibitory effect of Quxianruangan Capsulae on liver fibrosis in rats and chronic hepatitis
      patients. Shijie Huaren Xiaohua Zazhi 2001; 9: 1246-1249
7    You H, Wang B, Wang T. Proliferation and apoptosis of hepatic stellate cells and effects of compound 861 on liver
      fibrosis. Zhonghua Ganzangbing Zazhi 2000; 8: 78-80
8    Nan JX, Park EJ, Kim YC, Ko G, Sohn DH. Scutellaria baicalensis inhibits liver fibrosis induced by bile duct ligation or
      carbon tetrachloride in rats. J Pharm Pharmacol 2002; 54: 555-563
9    Wang QC, Shen DL, Zhang CD, Xu LZ, Nie QH, Xie YM, Zhou YX. Effect of Rangansuopiwan in expression of tissue inhibitor
      of metalloproteinase-1/2 in rat liver fibrosis. Shijie Huaren Xiaohua Zazhi 2001; 9: 379-382
10  Shen M, Chen Y, Qiu DK, Xiong WJ. Effects of recombinant augmenter of liver regeneration protein, danshen and
      oxymatrine on rat fibroblasts. Shijie Huaren Xiaohua Zazhi 2001; 9: 1129-1133
11  Wang XL, Liu P, Liu CH, Liu C. Effects of coordination of FZHY decoction on functions of hepatocytes and hepatic satellite
      cells. Shijie Huaren Xiaohua Zazhi 1999; 7: 663-665
12  Yao XX, Tang YW, Yao DM, Xiu HM. Effect of yigan decoction on the expression of type I, III collagen proteins in
      experimental hepatic fibrosis in rats. Shijie Huaren Xiaohua Zazhi 2001; 9: 263-267
13  Nakamura K, Yamaguchi Y, Kagota S, Shinozuka K, Kunitomo M. Activation of in vivo Kupffer cell function by oral
      administration of Cordyceps sinensis in rats. Jpn J Pharmacol 1999; 79: 505-508
14  Huang BM, Hsu CC, Tsai SJ, Sheu CC, Leu SF. Effects of Cordyceps sinensis on testosterone production in normal
      mouse Leydig cells. Life Sci 2001; 69: 2593-2602
15  Chiou WF, Chang PC, Chou CJ, Chen CF. Protein constituent contributes to the hypotensive and vasorelaxant activities
      of Cordyceps sinensis. Life Sci 2000; 66: 1369-1376
16  Yamaguchi Y, Kagota S, Nakamura K, Shinozuka K, Kunitomo M. Antioxidant activity of the extracts from fruiting bodies
      of cultured Cordyceps sinensis. Phytother Res 2000; 14: 647-649
17  Huang BM, Ju SY, Wu CS, Chuang WJ, Sheu CC, Leu SF. Cordyceps sinensis and its fractions stimulate MA-10 mouse
      Leydig tumor cell steroidogenesis. J Androl 2001; 22: 831-837
18  Dai G, Bao T, Xu C, Cooper R, Zhu JS. CordyMax Cs-4 improves steady-state bioenergy status in mouse liver. J
      Altern Complement Med 2001; 7: 231-240
19  Yang LY, Chen A, Kuo YC, Lin CY. Efficacy of a pure compound H1-A extracted from Cordyceps sinensis on
      autoimmune disease of MRL lpr/lpr mice. J Lab Clin Med 1999; 134: 492-500
20  Li BS, Wang J, Zhen YJ, Liu JX, Wei MX, Sun SQ, Wang SQ. Experimental study on serum fibrosis markers and liver
      tissue pathology and hepatic fibrosis in immuno-damaged rats. Shijie Huaren Xiaohua Zazhi 1999; 7: 1031-1034
21  Bissell DM. Chronic liver injury, TGF-beta, and cancer. Exp Mol Med 2001; 33: 179-190
22  Friedman SL. Cytokines and fibrogenesis. Semin Liver Dis 1999; 19: 129-140
23  Chen WX, Li YM, Yu CH, Cai WM, Zheng M, Chen F. Quantitative analysis of transforming growth factor beta 1 mRNA
      in patients with alcoholic liver disease. World J Gastroenterol 2002; 8:379-381
24  Gressner AM, Weiskirchen R, Breitkopf K, Dooley S. Roles of TGF-beta in hepatic fibrosis. Front Biosci 2002; 7: d793-807
25  Lewindon PJ, Pereira TN, Hoskins AC, Bridle KR, Williamson RM, Shepherd RW, Ramm GA. The role of hepatic stellate
      cells and transforming growth factor-beta (1) in cystic fibrosis liver disease. Am J Pathol 2002; 160: 1705-1715
26  Eichler W, Friedrichs U, Thies A, Tratz C, Wiedemann P. Modulation of Matrix Metalloproteinase and TIMP-1 Expression
      by Cytokines in Human RPE Cells. Invest Ophthalmol Vis Sci 2002; 43: 2767-2773
27  Britton RS, Bacon BR. Intracellular signaling pathways in stellate cell activation. Alcohol Clin Exp Res 1999; 23: 922-925
28  Li D, Friedman SL. Liver fibrogenesis and the role of hepatic stellate cells: new insights and prospects for therapy.
      J Gastroenterol Hepatol 1999; 14: 618-633
29  Kuter DJ. Future directions with platelet growth factors. Semin Hematol 2000; 37 (2 Suppl 4): 41-49
30  Pinzani M. Pdgf and signal transduction in hepatic stellate cells. Front Biosci 2002; 7: D1720-1726
31  Isbrucker RA, Peterson TC. Platelet-derived growth factor and pentoxifylline modulation of collagen synthesis
      in myofibroblasts. Toxicol Appl Pharmacol 1998; 149: 120-126
32  Fibbi G, Pucci M, Grappone C, Pellegrini G, Salzano R, Casini A, Milani S, Del Rosso M. Functions of the fibrinolytic system
      in human Ito cells and its control by basic fibroblast and platelet-derived growth factor. Hepatology 1999; 29: 868-878
33  Yuan N, Wang P, Wang X, Wang Z. Expression and significance of platelet derived growth factor and its receptor in
      liver tissues of patients with liver fibrosis. Zhonghua Ganzangbing Zazhi 2002; 10: 58-60
34  Benyon RC, Arthur MJ. Extracellular matrix degradation and the role of hepatic stellate cells. Semin Liver
      Dis 2001; 21: 373-384
35  Neubauer K, Saile B, Ramadori G. Liver fibrosis and altered matrix synthesis. Can J Gastroenterol 2001; 15: 187-193
36  Schuppan D, Ruehl M, Somasundaram R, Hahn EG. Matrix as a modulator of hepatic fibrogenesis. Semin Liver
      Dis 2001; 21: 351-72

Edited by Ma JY