C Basic & Clinical Pharmacology & Toxicology 2005, 96, 375–380.Printed in Denmark . All rights reserved

Copyright C

ISSN 1742-7835

Resolution of Liver Fibrosis in Chronic CCl4 Administrationin the Rat after Discontinuation of Treatment:Effect of Silymarin, Silibinin, Colchicine and

Trimethylcolchicinic AcidPablo Muriel, Mario G. Moreno, Marıa del C. Hernandez, Enrique Chavez and Lidia K. Alcantar

External Section of Pharmacology, Cinvestav-I.P.N., P.O. Box 14-740, Mexico 07000, D.F., Mexico

(Received September 9, 2004; Accepted November 10, 2004)

Abstract: The purpose of this work was to obtain a suitable model of fibrosis, in which spontaneous reversion wasminimal, to study the ability of silymarin, silibinin, colchicine and trimethylcolchicinic acid (TMCA) to reverse it. Reversalof liver fibrosis was studied in male Wistar rats after one, two or three months of CCl4 administration (0.4 g/kg intraperito-neally, three times per week), by discontinuation of the toxin for 2 months. Silymarin (50 mg/kg), silibinin (50 mg/kg),colchicine (10 mg/rat) and trimethylcolchicinic acid (100 mg/rat) were administered daily, by gavage, after 3 months of CCl4administration. Collagen content was determined by measuring hydroxyproline in liver samples; glycogen, was determinedutilizing the anthrone reagent; Mallory’s trichromic stains of liver sections were performed. The best scheme of treatmentwas obtained when CCl4 was administered during three months (collagen increased 6 times). Discontinuation of the toxinfor two months produced a significant but relative small reduction of fibrosis (collagen was still 4.5 times over control).Colchicine, TMCA, silymarin or silibinin treatment showed no significant fibrolitic effect. This scheme of treatment maybe an excellent tool to study the ability of drugs to reverse fibrosis. The hepatoprotective properties of silymarin, silibinin,colchicine and trimethylcolchinic acid may be irrelevant to reverse established cirrhosis.

Cirrhosis is characterized by increased deposition and alter-ed composition of extracellular matrix, so that there is anexcess of collagens I, III and IV (Friedman 1993). Whenadvanced, the liver architecture is distorted by dense bandsof collagen that link vascular structures and surround is-lands of regenerating parenchymal cells. Advanced fibrosisand cirrhosis are generally considered to be irreversible con-ditions even after removal of the injurious agent. Thus, thedisease is characterized by the incapacity of the injured liverto remodel the fibrotic matrix (Rojkind & Greenwel 2001).In contrast, experimental models of liver cirrhosis in ani-mals have shown spontaneous resolution of fibrosis oncethe toxic agent or cell insult is removed (Varga et al., 1966;Perez-Tamayo 1979; Abdel-Aziz et al. 1990). Therefore, ex-perimental models of cirrhosis have only been useful to testthe effectiveness of some drugs to prevent fibrosis, but havefailed to study the ability of these compounds to reverse theestablished disease.

The aim of the present work was to induce fibrosis bychronic intoxication with CCl4 in the rat during increasingperiods of time until spontaneous resolution of fibrosis wasminimal. We found that after 3 months of chronic CCl4administration (3 times per week), collagen increased about6 times and that fibrosis remained significantly elevatedeven after discontinuation of the toxin administration for

Author for correspondence: Pablo Muriel, External Section ofPharmacology, CINVESTAV-I. P. N., P.O. Box 14-740, Mexico07000, D. F., Mexico (fax π5255 5061 3394, e-mail pamuriel/mail-.cinvestav.mx).

two months. The present results suggest that this scheme oftreatment constitutes an excellent tool to study the fibroliticeffect of a given compound in the absence of the noxiousstimuli, but in the presence of an established fibrosis. Thisscheme of treatment prompted us to evaluate the ability ofsilymarin, silibinin, colchicine and trimethylcolchicinic acidto reverse fibrosis previously established by prolongedchronic CCl4 administration. Interestingly, despite that wehave proved previously the ability of these compounds toprevent fibrosis (Mourelle et al. 1989; Muriel & Mourelle1990; Castro & Muriel 1996; Cedillo et al. 1996; Muriel1997; Muriel et al. 1997), they failed to reverse it signifi-cantly when compared with an appropriate vehicle control.

Materials and Methods

Materials. Silymarin, silibinin, colchicine, trimethylcolchicinic acid,chloramine-T, sodium thiosulfate, p-dimethylaminobenzaldehyde,anthrone, thiobarbituric acid, p-nitrophenyl phosphate and bovineserum albumin were obtained from Sigma Chemical Co. (St. Louis,MO, USA). Citric acid, CCl4, sodium acetate, sodium hydroxide,glacial acetic acid, sodium chloride, toluene, sulfuric acid, iodine,ethanol, xylene, potassium hydroxide and formaldehyde were ob-tained from J. T. Baker (Xalostoc, Mexico).

Treatments of animals. Male Wistar rats (weighing initially 100–120g) from our own breeding facilities were used in this study. All ani-mals received human care and the study complied with the insti-tution’s guidelines and the Mexican official regulation regardingtechnical specifications for production, care and use of laboratoryanimals (NOM-062-ZOO-1999). Cirrhosis was produced by intra-peritoneal administration of CCl4 (0.4 g/kg of body weight) dis-

PABLO MURIEL ET AL.376

solved in mineral oil three times per week. The animals in groups 1to 4 received CCl4 for four, eight, twelve or sixteen weeks, respec-tively. These animals were sacrificed 72 hr after the last dose ofCCl4. The rats in groups 5 to 7 received CCl4 for eight, twelve orsixteen weeks, respectively. Then CCl4 was discontinued for twomonths. Group 8, the control group, received mineral oil (the ve-hicle of CCl4) for sixteen weeks. The number of rats at the begin-ning (Nb) and at the end (Ne) of the treatments are shown in fig.1. Animals were anaesthetized with ether and submitted to exsangu-ination by cardiac puncture at the end of treatments. This bloodwas used to determine enzyme markers of liver damage. The liverwas immediately removed and, finally, animals were sacrificed by anexcess of anesthesia. All samples were kept on ice until analysis.

To evaluate the fibrolitic effect of the compounds studied herein,six additional groups were performed. Animals in groups 1 to 4(NbΩ25, each) received CCl4 (as previously described) for 12 weeks,then CCl4 was discontinued and silymarin (50 mg/kg; NeΩ17), silib-inin (50 mg/kg; NeΩ19), colchicine (10 mg/rat; NeΩ16) and trime-thylcolchicinic acid (100 mg/rat; NeΩ18) were administered daily bygavage. Groups 5 and 6 were appropriate controls and receivedCCl4 for 12 weeks and then the vehicle of the drugs (CMC 0.7%)for 8 weeks (NbΩ25; NeΩ18) or mineral oil for 12 weeks and CMCfor 8 weeks (NbΩ8; NeΩ8), respectively.

Serum enzyme activities and hepatic glycogen determination. Bloodwas obtained by cardiac puncture and samples were kept on ice forclotting, which occurred after two hours. Then, samples were centri-fuged at 10,000 rpm for 15 min. to obtain the serum. Serum was usedfor determination of alkaline phosphatase (ALP) (Bergmeyer et al.1983) and alanine aminotransferase (ALT) (Reitman & Frankel 1957)activities. Small liver pieces (0.5 g) were separated for glycogen deter-mination using the anthrone reagent (Seifter et al. 1950).

Collagen quantification. Collagen concentrations were determinedby measuring hydroxyproline-containing liver samples, after diges-tion with acid as previously described (Muriel 1998). Briefly, theprocedure was as follows: Fresh liver samples (100 mg) were placedin ampules, 2 ml of 6 N HCl was added, and then the samples weresealed and hydrolyzed at 100 æ for 48 hr. Next, the samples wereevaporated at 50 æ for 24 hr and resuspended in 3 ml of sodiumacetate-citric buffer, pH 6.0; 0.5 g of activated charcoal was added,the mixture was stirred vigorously, and then it was centrifuged at5,000 ¿g for 10 min. The mixture was kept for 20 min. at roomtemperature and the reaction was stopped by the addition of 2 Msodium thiosulfate and 1 N sodium hydroxide. The aqueous layerwas transferred into test tubes. The oxidation product from hy-droxyproline was converted to a pyrrole by boiling the samples. The

Fig. 1. Percentage of mortality in the groups treated with CCl4 dur-ing 1, 2, 3 or 4 months (CCl4-1, 2, 3, 4 m) and in those where CCl4was discontinued for 2 months (CCl4 2, 3, 4 m πD2 m). ‘‘Nb’’means number of rats at the beginning of treatment and ‘‘Ne’’means number of rats at the end of treatment.

pyrrole-containing samples were incubated with Ehrlich’s reagentfor 30 min., and the absorbance was read at 560 nm. Recovery ofknown amounts of standards was carried out on similar liversamples to provide quantification.

Histology. Samples of liver were taken from all the animals andfixed with 10% formaldehyde in PBS for 24 hr. Those tissue pieceswere washed with tap water, dehydrated in alcohol, and embeddedin paraffin. Sections of 6–7 microns were mounted in glass slidesand covered with silane. Mallory’s trichromic stains were performedin each slide.

Statistics. For statistical analysis, ANOVA with the Tukey test wasused to compare experimental groups. In all cases a difference wasconsidered significant when P0.05.

Results

Mortality was assessed in all groups (fig. 1). The controlgroup, treated chronically with mineral oil only, showed nomortality. After one month of CCl4 intoxication, 100% ofrats survived. However, mortality appeared after 2 monthsof intoxication (5.55%), increased by 3 months (26.6%) andreached 55.5% at month 4 of chronic intoxication withCCl4. Discontinuation of treatment for 2 months did notstop mortality but increased it progressively. The longesttreatment, four months of chronic intoxication with CCl4followed by a period of two months in which rats were notfurther injured, showed the highest mortality (97.5%; onlyone out of forty rats survived). This group was not furtherstudied because one rat was not considered representativeand selection of the less damaged rat was obvious.

The degree of fibrosis was assessed by using two indepen-dent methodologies: the biochemical quantification of col-

Fig. 2. Time-course of liver collagen content in rats, estimated ashydroxyproline, after chronic CCl4 administration and discontinu-ation (vehicle) of the CCl4 treatment. Capital letters in circles indi-cate the panel in fig. 3 of a representative micrograph of a liversample from the same group. Results are expressed as themean∫S.E.M. *Means statistically different versus control (time 0).**Means statistically different versus the group receiving CCl4 dur-ing two months. ***Means statistically different versus the groupreceiving CCl4 during three months.

377LIVER CIRRHOSIS ESTABLISHED BY CCl4

Fig. 3. Representative Mallory’s trichromic stains of liver sections obtained from rats treated with vehicle (panel A), CCl4 for 2 (panel B), 3(panel C) or 4 (panel D) months, and rats treated with CCl4 during 2 (panel E) or 3 (panel F) months and then 2 months with the vehicleonly prior to sacrifice. Collagen can be visualized by the blue color in stains. The mean quantitative amount of collagen, determined bymeasuring liver hydroxyproline, of the same groups is shown in fig. 2 where capital letters correspond to the panels of this figure.

lagen by measuring the hydroxyproline content in liversamples (fig. 2) and by the histopathological analysis underlight microscopy (fig. 3). Fig. 2 shows the hepatic contentof collagen and capital letters in circles indicate the panelin fig. 3 in which collagen deposition can be observed byblue staining in representative micrographs of liver sections.Two months of chronic CCl4 intoxication increased the col-lagen content about 3.5 times over control, P0.05; threemonths of treatment increased fibrosis more than 6 times,and intoxication for four months about 8 times. Discon-tinuation of CCl4 treatment during two months induced apartial but significant reversion of fibrosis in both the twoand three months intoxication groups. This effect was moreimportant in the two months intoxication group, since finalvalues of collagen were only about 2.5 times over control,while in the three months intoxication group fibrosis re-

mained near to 4.5 times over control after discontinuationof CCl4 for two months.

Fibrosis was also evaluated by a histological approach(fig. 3). In general, chronic CCl4 intoxication produced amarked increase in collagen deposition around the portaltriad, also nodules of hepatocytes surrounded by variablesize collagen band (blue stained) were frequently observed,the normal architecture was lost and extended necroticareas were present. These effects were qualitatively similarin the groups receiving CCl4 for two (panel B), three (panelC) or four months (panel D) and absent in the controlgroup (panel A). However, the magnitude of the effect cor-related well with the duration of the intoxication period.Discontinuation of the treatment for two months resultedin the absence of necrotic areas and in thinner bands ofcollagen. In the group treated with CCl4 for two months

PABLO MURIEL ET AL.378

and two months of discontinuation, the nodules of collagenwere incomplete (panel E). However, in the group treatedfor 3 months followed by discontinuation of intoxicationfor two months, the nodules of collagen surroundinghepatocytes were still present in the hepatic parenchyma(panel F). It is worth noting that the histopathologicalanalysis revealed similar degree of fibrosis than the bio-chemical analysis of hepatic hydroxyproline (fig. 2).

Glycogen, the main source of energy in the liver was de-pleted by two, three or four months of chronic intoxicationwith CCl4 (fig. 4). The decrement in the hepatic glycogenstores was similar in the three groups (about 4 times). Dis-continuation of CCl4 treatment in the two or three monthsCCl4-treatment groups led to a complete restoration of thecontent of glycogen in the liver.

Alanine aminotransferase (ALT) and alkaline phospha-tase (ALP) serum enzyme activities are depicted in fig. 5,upper and lower panels, respectively. The serum indicatorof necrosis, ALT increased about 3 times in the groupstreated with CCl4 during two or three months. This resultis in agreement with the findings observed in the histo-pathological analysis where necrosis was clearly identified(fig. 3, panels B, C and D, respectively). As expected, ALP,an indicator of cholestasis, increased moderately in thegroups receiving CCl4 during two or three months (fig. 5,lower panel), probably because this is not a model of liverdamage characterized by cholestasis. Interestingly, the fourmonths CCl4-treated group, showed both ALT and ALPenzyme activities similar or lower to those of the controlgroup. In addition, discontinuation of CCl4 reversed to nor-mal or below normal values those enzyme activities. Thisbiochemical results, also correlated with histopathologicalfindings, where necrosis was not observed in groups not re-ceiving CCl4 for the last 2 months prior to sacrifice (fig. 3,panels E and F).

Fig. 4. Time-course of liver glycogen content in rats after chronicCCl4 administration and discontinuation (vehicle) of the CCl4 treat-ment. Results are expressed as the mean∫S.E.M. *Means statisti-cally different versus control (time 0). **Means statistically differentversus the group receiving CCl4 during two months. ***Means stat-istically different versus the group receiving CCl4 during threemonths.

Previously we have described that 3 months of CCl4treatment lead to important fibrosis (about 6 times increasein collagen content) and that discontinuation of the toxinfor 2 months did not reverse fibrosis completely but a 4.5times increase in collagen content was still present. There-fore, we decided to evaluate the effect of some known anti-fibrotic compounds in this scheme of treatment. Silymarin,silibinin, colchicine, TMCA and vehicle treatments beganafter 3 months of chronic CCl4 intoxication and the toxinwas discontinued thereafter. As you can see in fig. 6, spon-taneous resolution of fibrosis (vehicle group) was partialand similar to the previous experiment (fig. 2), indicatingreproducibility of this scheme of treatment. Unfortunately,none of the pharmacological treatments evaluated hereinshowed statistical difference when compared with the ve-hicle group, thus being the same as spontaneous reversionof fibrosis. Mortality and the histopathology of the groupstreated with these drugs (not shown) were similar to thegroup treated with the vehicle only and shown in fig. 1 and3, respectively.

Fig. 5. Time-course of enzymatic activities of alanine aminotrans-ferase (upper panel) and alkaline phosphatase (lower panel), deter-mined in plasma from rats after chronic CCl4 administration anddiscontinuation (vehicle) of the CCl4 treatment. Enzymatic activityis expressed as mmol of product per liter per minute (mmol/l min.).Results are shown as the mean∫S.E.M. *Means statistically differ-ent versus control (time 0). **Means statistically different versusthe group receiving CCl4 during two months. ***Means statisticallydifferent versus the group receiving CCl4 during three months.

379LIVER CIRRHOSIS ESTABLISHED BY CCl4

Fig. 6. Effect of silymarin, silibinin, colchicine and trimethylcolchic-inic acid (TMCA) on the fibrosis induced by chronic CCl4 adminis-tration during 3 months. After this time, CCl4 was discontinuedand drugs or vehicle were administered daily for 2 months. Resultsare expressed as the mean∫S.E.M. *Means statistically differentversus control (time 0). **Means statistically different versus thegroup receiving CCl4 during three months.

Discussion

There is a discrepancy between the amount of experimentalcompounds with proven beneficial effects in preventionmodels of cirrhosis and the lack of effective drugs to treatadvanced cirrhosis in humans. The scheme of treatmentpresented herein may be an excellent tool to study the abil-ity of drugs to reverse fibrosis. In this model of fibrosis,colchicine, TMCA, silymarin or silibinin treatment showedno significant fibrolitic effect despite its well known anti-fibrogenic properties (Mourelle et al. 1989; Muriel & Mour-elle 1990; Castro & Muriel 1996; Cedillo et al. 1996; Muriel1997; Muriel et al. 1997).

Human liver cirrhosis is generally considered as an irre-versibly process. In animal models, on the contrary, liverfibrosis induced by several toxins or bile duct ligation hasbeen shown by morphological studies to regress after thetoxic agents were removed or bile duct reconstructed (Vargaet al. 1966; Perez-Tamayo 1979; Abdel-Aziz et al. 1990). Inthe present work, reversal of liver fibrosis was studied afterone, two or three months of CCl4 administration, by discon-tinuation of the toxin for 2 months. The purpose was toobtain a suitable model of cirrhosis, in which spontaneousreversion of fibrosis was minimal in order to study the abil-ity of drugs to reverse it. Treatment with CCl4 for 2 monthsproduced only mild fibrosis; thus, we considered that thisscheme was not suitable to study fibrolitic drugs. On theother hand, four months of CCl4 treatment plus twomonths of vehicle produced 97.5% mortality. Therefore, thebest scheme of treatment to study pharmacological rever-sion of fibrosis was obtained when CCl4 was administeredduring three months, which produced a 6 times incrementin collagen content. In this case, discontinuation of thetoxin for two months produced a significant but relatively

small reduction of fibrosis (collagen was still 4.5 times overcontrol).

Lee et al. (2001a) showed that reversibility of liver fibrosisin mice would be gradually lost as liver injuries were pro-longed. Interestingly, in the study of Lee et al. (2001b), amore prolonged CCl4 treatment did not lead to a higherdegree of fibrosis, but to a reduction of the expression ofmatrix metalloproteinase. In contrast, the present resultsshow that the degree of fibrosis increased linearly with timeof CCl4 treatment, but spontaneous regression of fibrosiswas similar after 2 or 3 months of chronic intoxication. Themain difference between both studies is that Lee et al.2001a&b utilized mice and that doses and periods of treat-ment were considerable different.

Interestingly, animals intoxicated during four monthsshowed normal or under normal values of markers of liverdamage. Probably, livers from rats treated with CCl4 during4 months were so damaged (fig. 3D) that were no longerable to produce and release enzymes to plasma. It is alsoworth noting that discontinuation of CCl4 resulted in res-toration of glycogen and plasma enzyme, but not collagen,normal values. It is possible that enzymes released toplasma were eliminated during the discontinuation periodand that necrosis stops in the absence of CCl4 (as shown byhistology). In this situation, carbohydrate metabolism maybe stabilized. However, gradual loss of the expression ofmatrix metalloproteinases (Lee et al. 2001b) may be respon-sible for the loss of fibrosis reversibility.

Most of the compounds with beneficial effects on fibrosishave been studied by administering both the toxin and thedrug simultaneously with the risk that many of these com-pounds may interfere with the action mechanism of the nox-ious stimuli. Thus, prevention of experimental cirrhosis hasbeen demonstrated for a variety of compounds (Muriel1997). However, therapeutic interventions that reverse fib-rosis are needed. In this work we decided to test some drugspreviously reported to have the ability in order to preventthe disease.

We have previously shown that silymarin co-treatmentpartially prevents CCl4-induced cirrhosis (Mourelle et al.1989), probably because silymarin is a good free radicaltrapping agent as we have also demonstrated (Pascual et al.1993). It seems that this antioxidant property is no longerimportant when the compound is administered once CCl4is discontinued and free radicals are no longer produced.This may explain why silymarin is capable of preventingfibrosis (Mourelle et al. 1989; Muriel & Mourelle 1990) butfailed to reverse the disease (the present work).

We have shown that colchicine and a less toxic derivative,TMCA, are capable of preventing CCl4 (Cedillo et al. 1996)and BDL (Muriel et al. 1997) induced cirrhosis in the rat.Therefore, in this work we evaluated the ability of thesecompounds to reverse fibrosis induced by CCl4. Colchicineand TMCA were unable to reverse fibrosis significantlywhen compared with vehicle control. It is possible that thesecompounds are acting by inhibiting the mechanism of toxic-ity of the etiological agent. For example there is evidence

PABLO MURIEL ET AL.380

that colchicine inhibits CYP450IIE1 (Martinez et al. 1995),responsible for the bioactivation of CCl4 (Muriel 1997) andthat both colchicine and TMCA prevent lipid peroxidationinduced by CCl4 (Cedillo et al. 1996) and BDL (Muriel etal. 1997). These sorts of mechanisms, very useful to preventdamage (Muriel 1997), are irrelevant when damage is estab-lished and the toxic stimuli removed.

AcknowledgementsThe authors express their gratitude to Dr. Victor Tsutsu-

mi and Dr. Mineko Shibayama for performing the histo-pathological analysis and to Mr. Alfredo Padilla for prepar-ing the figures, Mrs. Patricia Gonzalez, Mr. Ramon Her-nandez and Mr. Benjamın Salinas for their excellenttechnical assistance. This study was supported in part bygrant 34394M from Conacyt, Mexico.

References

Abdel-Aziz, G., G. Lebeau, P.Y. Rescan, B. Clement, M. Rissel, Y.Deugnier, J. P. Campion & A. Guillouzo: Reversibility of hepaticfibrosis in experimentally induced cholestasis in rat. Amer. J.Pathol. 1990, 137, 1333–1342.

Bergmeyer, H. U., M. Grabl & H. E. Walter: Enzymes. In: Methodsof enzymatic analysis. Eds.: J. Bergmeyer & M. Grabl. Verlag-Chemie, Weinheim, 1983, pp. 269–270.

Castro, V. & P. Muriel: Comparative study of colchicine and trime-thylcolchicinic acid on prolonged bile duct obstruction in the rat.J. Appl. Toxicol. 1996, 16, 269–275.

Cedillo, A., M. Mourelle & P. Muriel: Effect of colchicine and tri-methylcolchicinic acid on CCl4-induced cirrhosis in the rat. Phar-macology. Toxicology. 1996, 79, 241–246.

Friedman, S. L.: The cellular basis of hepatic fibrosis: mechanismsand treatment strategies. New Engl. J. Med. 1993, 328, 1828–1835.

Lee, H. S., G. T. Huang, C. H. Chen, L. L. Chiou, C. C. Lee, P. M.Yang, D. S. Chen & J. C. Sheu: Less reversal of liver fibrosis afterprolonged carbon tetrachloride injection. Hepato-Gastroenterol.2001a, 48, 1312–1315.

Lee, H. S., G. T. Huang, C. H. Chen, L. L. Chiou, C. C. Lee, P. M.

Yang, D. S. Chen & J. C. Sheu: Expression of matrix metallopro-teinases in spontaneous regression of liver fibrosis. Hepato-Gastroenterol. 2001b, 48, 1114–1117.

Martınez, M., M. Mourelle & P. Muriel: Protective effect of colchi-cine on acute liver damage induced by CCl4. Role of cytochromeP450. J. Appl. Toxicol. 1995, 15, 49–52.

Mourelle, M., P. Muriel, L. Favari & T. Franco: Prevention of CCl4induced liver cirrhosis by silymarin. Fund. Clin. Pharmacol. 1989,3, 183–191.

Muriel, P.: Peroxidation of lipids and liver damage. In: Oxidants,antioxidants and free radicals. Eds. S. I. Baskin & H. Salem. Tayl-or & Francis, Washington, USA, 1997, pp. 237–257.

Muriel, P.: Nitric oxide protection of rat liver from lipid peroxi-dation, collagen accumulation, and liver damage induced by car-bon tetrachloride. Biochem. Pharmacol. 1998, 56, 773–779.

Muriel, P. & M. Mourelle: The role of membrane composition inATPases activities of cirrhotic rats. Effect of silymarin. J. Appl.Toxicol. 1990, 10, 281–284.

Muriel, P., P. V. Ostoa-Saloma, J. A. Reyes-Esparsa & L. Rodrıguez-Fragoso: Trimethylcolchicinic acid decreases liver fibrosis andcholestasis induced by prolonged biliary obstruction in the rat.J. Appl. Toxicol. 1997, 17, 145–151.

Okasaki, I., T. Watanabe, S. Hozawa, M. Arai & K. Maruyama:Molecular mechanism of the reversibility of hepatic fibrosis: withspecial reference to the role of matrix metalloproteinases. J.Gastroenterol. Hepatol. 2000, 15S, 26–32.

Pascual, C., R. Gonzalez, J. Armesto & P. Muriel: Effect of silymar-in and silybinin on oxygen radicals. Drug Dev. Res. 1993, 29, 73–77.

Perez-Tamayo, R.: Cirrhosis of the liver: a reversible disease? Pathol.Annu. 1979, 2, 183–213.

Reitman, S. & S. Frankel: A colorimetric method for determinationof serum oxaloacetic and glutamic pyruvic transaminases. Amer.J. Clin. Pathol. 1957, 28, 56–63.

Rojkind, M. & P. Greenwel: Pathophysilogy of liver fibrosis. In: Theliver biology and pathobiology. Eds.: I. M. Arias, J. L. Boyer, F.V. Chisari, N. Fausto, D. Schachter, D. A. Shafritz. LippincottWilliams & Wilkins, Philadelphia, USA, 2001, pp. 721–738.

Seifter, S., S. Dayton, B. Novic & E. Muntwyler: The estimation ofglycogen with the anthrone reagent. Arch. Biochem. 1950, 25,191–200.

Varga, F., G. Mehes & Z. Molnar: Reversibility of hepatic fibrosisinduced by carbon tetrachloride in the rat. Acta Physiol. Acad.Sci. Hung. 1966, 29, 69–74.