Hepatic fibrosis is characterized by massive deposition of extracelluler matrix components in the liver. There are differences on liver fibrogenesis due to etiological factors [1].Currently, there are no animal models that completely mimic the spectrum of the diseases seen in humans.

One of the most useful experimental fibrosis models is the "bile duct-ligated rats". There is substantial information on morphological changes in the liver of the rats after obstructive jaundice to date. In 1932, Cameron et al. described gross macroscopic features and light microscopical changes in bile duct-ligated rats by intervals from 1/2 hour to 9 weeks [2]. Trams et al. in 1957 and Cameron et al. in 1958 studied functional and structural disturbances in the bile duct-ligated rats' liver [3, 4]. Then, several investigators studied serum enzymes derived from liver cell fraction in the bile duct-ligated rat in 1972 [5]. Later studies investigated the effect of biliary obstruction on bile flow and bile acid secretion [6]. Kountauras et al. surveyed prolonged bile duct obstruction as a new experimental model for cirrhosis in the rat and proposed that as a good model for hepatic fibrosis studies in 1984 [7].

However, none of the previous studies investigated hepatic collagen content during ongoing and progressive hepatic fibrosis after bile duct ligation (BDL). Our aim was to investigate the quantitative hepatic collagen content by two different methods in bile duct-ligated rats within the course of time.

53 male Wistar rats, weighting 244.62 ± 5.03 (mean ± SE) were included in the study. Eight rats (Group 0) were used as a healthy control group without operation. Thirty-seven underwent BDL. These animals were sacrificed seven days later in group 1 (n = 10), 14 days later in group 2 (n = 9), 21 days later in group 3 (n = 9) and 28 days later in group 4 (n = 9).

Eight rats underwent sham-operation (Sham). This study was approved by the Marmara University, Animal Use and Care Committee. All the experimental procedure was performed in conformity with the Guiding Principles for Research Involving Animals. Sacrifice was performed by decapitation, and trunk blood was obtained for each. Serum samples were collected for aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP), total and direct bilirubin and gamma-glutamyl transpeptidase (GGT) measurements, and stored at -80°C.

Collagen content determination by biochemically

The collagen content of the liver was assayed by the colorimetric method described by Lopez de Leon and Rojkind [8]. The principle is the coloring of collagenous protein by Sirius red (36554-8, 2610-10-8; Aldrich Chemical, Deisenhofen, Germany) and non-collagenous proteins by fast green (14280; MERCK, Darmstadt, Germany) (Figure 1). Fifteen micrometer-thick liver slices were taken from each paraffin block and layered on glass slides. Slices were deparaffinized after being incubated by xylol, xylol: ethanol(1:1), ethanol, water: ethanol, and water. The slides were stained with a saturated solution of picric acid in distilled water containing 0.01% fast green. Each section was kept out of the light and incubated at room temperature for 15 minutes. Then, sections were stained by a saturated solution of picric acid in distilled water containing 0.04% fast green and 0.1% of sirius red and incubated in the dark room temperature for 30 minutes. Then, samples were rinsed and transferred to a test tube containing 1 ml of 0.1% NaOH in absolute methanol (1:1). The tubes were gently mixed until the color was eluted completely. Absorbance of the eluted color was read at 540 and 605 nm by a Jasco V 50 UV-VIS spectrophotometer. Maximal absorbance of fast green is at 630 nm and 540 nm for Sirius red. Fast green exhibites a small absorbance (7.78%) at 540 nm. This interference remains constant at different concentrations of fast green. By contrast, sirius red has no absorbance at 630 nm. Therefore, sirius red does not interfere with non-collagenous protein determination. Collagen content of tissue was calculated using the formula below. It was described as in microgram collagen per milligram protein [8].

Histopathological investigations

Five-micrometer liver sections were stained by hematoxylin-eosin and Masson's trichrom. We used Ishak fibrosis classification system for grading of necroinflammatory activity and staging of fibrosis [9]. We used a scoring system from 1 to 4 for grading of portal proliferation (Figure 2).

There is no animal fibrosis model that completely mimics the spectrum of the diseases seen in human so far [10]. Previously, many investigators used a 4-week bile duct ligated rat model in fibrosis studies [1–18]. Each model facilitated our understanding underlying cellular and molecular mechanisms of fibrogenesis and further assessment of the efficacy of various antifibrotic agents. Our aim was to investigate the quantitative hepatic collagen content with the levels of the serum liver enzymes during the different stages of hepatic fibrosis in bile duct-ligated rats.

In the presented study, we found collagen peak at the third week after bile duct ligation by utilizing both biochemical and histomorphometric methods of hepatic collagen measurement. Since these results were unexpected, we re-validated our results in both groups 3 and 4 by repating examinations. We re-measured the hepatic collagen concentration in revalidation study and found that the hepatic collagen of the third group was higher than the fourth group again. The correlation between liver collagen content of earlier group 3-4 and new group 3-4 was r: 0.92 (P < 0.001).

Serum ALP levels were increased in the first week, then they started to decrease. Finally, serum ALP level was the same as the control group at the fourth week. Similarly, serum GGT level reached its peaks value in the third week and at the end of the fourth week they were still higher. These results were reflecting the changes on the bile ducts in bile ductligated rat model and in a correlation with measured hepatic collegen concentrations. Having these results, we decided to survey the earlier studies to better elucidate who will use this experimental method. Bile duct ligation model has attracted many investigators and many studies in rats as well as other animals performed to clarify this model [1]. The most outstanding and comprehensive study in BDL was performed by Cameron and Oakley in 1932 [2]. They used 120 rats and investigated them from one hour to 9 weeks. They found increased hepatic fibroblastic proliferation starting by day 4 and perilobuler cirrhosis 47 days after bile duct ligation and further frank cirrhosis at the 61st day.

Cameron and Hassan evaluated the structural and functional disturbances in bile ductligated rat and observed them from 1/2 hour to 11 weeks. They found the same structural disturbances as their former study [4]. The serum bilirubin levels were increased to 10.5 mg at the first week which was less than what was observed by us. Then, it was decreased to 4. 8 mg at the 4th weeks, the serum bilirubin fluctuated from week 5 through week 11 in this previous study. Trams et al. evaluated the morphological and functional changes in the liver of rats both after ligation (without division) and excision of the common bile duct [3]. The rats were sacrificed after 1., 3., 14. and 28. days. Fibrous tissue was established around the duct at the 14th and 28th days in CBD excised rats. However, in bile duct-ligated (without division) rats they determined approximately 95% recanalization and normalization of the bile duct. Johnstone et al. reported a quantitave assessment of the structural changes in the rat's liver following bile duct ligation and investigated them from first day to 40 days [19]. Volume proportion of parenchyma, bile duct and supportive stroma were 97/2/1% in sham operated rats, respectively and 37/42/20% respectively at the 40 days after BDL without cirrhosis.

The most controversial area in BDL model is the degree or amount of hepatic fibrosis. Kountauaras et al sacrificed rats 5, 10, 15, and 28 days after BDL and found that all rats had cirrhosis at the 28th day. Several reported frank cirrhosis developed after 3 or 5 weeks [2–12]. On the other hand, there are several studies that did not show any evidence of cirrhosis even at the 40th day following bile duct ligation in the rat [5, 20–22]. We did not see cirrhosis in the presented study and our former studies after 28 days bile duct-ligation [23–28]. We observed grades 2 nd 3 fibrosis in our studies. Previously reported that cirrhosis in BDL rats occurs earliest 4-6 weeks after ligation [27, 28]. The reason for the differences is unclear and needs further clarification.

We have shown that fibrosis in bile duct-ligated rats is transient, i.e. reverses spontaneously after 3 weeks both biochemically and histomorphometrically. This contrasts any situation in patients where hepatic fibrosis is progressive and irreversible as countless studies performed by many investigators in the same animal model. In fact, it has been assumed to date, that hepatic fibrosis in patients and animal models progresses to cirrhosis, but does not reverse spontaneously as described in the manuscript.

Although markers of fibrosis and liver damage declined after the 3-week peak, they still remained elevated 4 weeks after ligation, compared to control groups in this study. Thus, this animal model may in fact be suitable to study fibrosis and experimental therapies, provided that correct untreated control groups are included.

Our results are fundamental because many investigators use four-week bile ductligated rats for hepatic fibrosis and antifibrotic therapy studies. To see the effect of antifibrotic therapy on maximum level of collagen is important. Understanding the dynamics of hepatic fibrogenesis we have to use three and four week bile duct-ligated rat models together. We need further studies to clarify the results of the study.