Liver fibrosis has been defined as a disease in which excessive collagen is deposited in the liver leading to dysfunction of the liver. Hepatic stellate cells are the major liver cells involved in the development of liver fibrosis. They comprise about 15% of the total liver cells and are located in the sub endothelial space of Disse between the hepatocytes and the endothelial cells [1]. These cells are usually quiescent and they store lipid droplets. Upon activation by other factors they change into myofibroblast like cells and lose Vitamin A droplets and become proliferative, fibrogenic and contractile and act as major sources of extracellular matrix components [2]. HSC activation usually involves 2 steps i.e. initiation and perpetuation. HSC initiation is due to the stimuli released by the surrounding cells such as hepatocytes, endothelial cells, Kupffer cells, and platelets. The stimuli, which initiate these cells, are growth factors, inflammatory cytokines and reactive oxygen species or lipid peroxides [2]. As previously indicated, pancreatic fibrosis shows remarkable similarity to hepatic fibrosis [3].

Aryl Hydrocarbon Receptor (AHR) knockout mice were developed by the homologous recombination in the embryonic stem cells [4]. These mice have decreased ability to catabolize retinoic acid due to Ah receptor deficiency and this leads to liver fibrosis [5]. AHR knockout mice have genes coding for the Aryl hydrocarbon receptor protein deleted and the mice develop fibrosis at about 6-10 months of age [6]. Yellow phosphorous is a well-known toxic compound that produces degenerative changes characteristic of fibrosis and cirrhosis [7,8]. The animal model demonstrating these biochemical and histological changes characteristic of hepatic fibrosis was first established in our lab, liver fibrosis was induced in pigs by 8 weeks YP treatment (0.6 mg/kg) [8].

In the current study there was a significant increase in collagen content as the mice developed from week 5 to week 24 when they showed marked fibrosis, indicating that spontaneous fibrosis occurs in untreated AHR knockout mice over the 7 month time period. The results also indicate that one week of YP treatment increased the rate of fibrosis development in the AHR knockout mice such that fibrosis was evident in 5-week-old AHR knockout mice.

This increase in liver collagen content in 5 weeks old AHR knockout mice treated for one week with YP was equivalent to spontaneously elevated collagen levels in 24-week-old untreated AHR knockout mice. In this study the molecular changes occurring in this mouse model are investigated and antisense technology is used to intervene.

Animals and treatments

Mice (age 5 weeks) were treated with yellow phosphorous (0.6 mg/ kg i.p.) for 1 week and compared to age matched controls. The mice were sacrificed and 15 μm sections of liver were stained by Sirius Red/Fast Green collagen staining technique to assess liver collagen content as an index of fibrosis. AHR knockout mice treated with yellow phosphorous were compared to untreated age matched AHR knockout mice and C57BL/6 mice controls (with and without treatment with yellow phosphorous). Collagen content was also assessed in AHR knockout mice aged 5-8 weeks, 10-12 weeks, 14-16 weeks and 24 weeks to demonstrate the spontaneous development of fibrosis in AHR knockout mice. In the experiments utilizing antisense, the phosphothioated c-jun antisense (5’ GCAGTCATAGAACAGTCCGTCACTTCACGT-3’ was administered (10 mg/kg) to a subgroup of AHR knockout mice, on three separate days via the tail vein, during the week of yellow phosphorous treatment.

Liver collagen estimation

The mice were sacrificed by injecting somnitol i.p. (0.05 ml). The liver was removed and 1cm x 1 cm piece of the liver was placed in 10% buffered formaldehyde and 4 % paraformaldehyde solution. The fixed liver tissue was embedded in paraffin and 5 (15 μm) sections were cut. The collagen content was evaluated by the Sirius Red/Fast Green staining [5]. The paraffin embedded 15 μm sections were deparaffinized by treatment with xylenes and ethanol. Then these tissue samples were incubated with Fast green FCF 0.04% (Gurr, BDH Chemicals, Canada) for 20 minutes and after subsequent washings they were again incubated with Fast green FCF 0.2% and Sirius Red F3B 0.08% (Gurr, BDH Chemicals, Canada) for 30 minutes and were then subsequently washed. Samples were eluted with 0.05N NaOH /50% methanol and were read in a spectrophotometer at 530 nm (Sirius red) and 605 nm (Fast Green). The average absorbance values were quantitated using their respective colour equivalencies [6,9] i.e. 38.4 for Sirius red and 2.08 for Fast green. The non-collagenous protein (mg) was determined as absorbance 605nm/2.08 and the collagen (μg) protein was calculated as absorbance 530 nm- 0.26 absorbance 605 nm/38.4. Finally the collagen content (μg/mg total protein) was determined as μg collagen/ μg collagen + mg non –collagenous protein [6].

Western analysis

Liver tissue lysates were prepared by pulverizing frozen tissue with a liquid nitrogen-cooled mortar and pestle. Ice-cold RIPA buffer (1% Igepal CA 630, 0.5% sodium deoxycholate, 0.1% SDS in PBS, pH 7.4) containing standard protease inhibitors (PMSF, aprotinin, sodium orthovanadate) was added at a ratio of 3uL RIPA/mg tissue. Suspensions were homogenized followed by sonication and were incubated on ice for 30 minutes with additional PMSF (0.03 uL/mg tissue) before centrifugation (10 min, 10,000 rpm, 4°C). Supernatants (lysates) were stored at -86°C. Protein estimation was done using the QuantiPro BCA Assay Kit (Sigma-Aldrich). 50 μg of protein were run on 10% Tris- HCl gels (Bio-Rad) and transferred to PVDF membranes overnight. Proteins were probed with primary antibody, e.g., c-Jun (1:1000; New England Biolabs); secondary antibody (Anti-rabbit; 1:2000) was applied and protein bands were visualized with LumiGlo (New England Biolabs). Densitometry was done using Scion Image software. Results are expressed as Relative Density Units (RDU).

Statistical analysis

An unpaired Student’s t-test was used to compare two variables and an analysis of variance Student-Newman-Keul’s test was used while comparing more than two variables [10].

Collagen content in untreated AHR knockout mice over 24 weeks

The collagen content in liver sections taken from AHR knockout mice (AHR-/-) of various ages is shown in Figure 1. Collagen content was determined in liver sections of AHR knockout mice (5-8 weeks) and compared with other age groups and was expressed as μg/mg protein. There was a significant increase (*p<0.001, when compared to 5-8 weeks) in collagen content as the mice developed, which indicated that fibrosis developed spontaneously in AHR knockout mice between 5 weeks and 24 weeks of age.

Figure 1: Quantitation of collagen content in liver sections taken from AHR knockout mice (AHR -/-) mice of various ages. Collagen content is shown in liver sections of AHR knockout mice (5-8 weeks) and compared to other age groups and is expressed as μg/mg protein. *p<0.001, significantly elevated when compared to 5-8 weeks.

Collagen content in AHR knockout mice treated with yellow phosphorous

Liver collagen content was determined in 5 week old AHR-/- mice with and without treatment with YP and contrasted to the collagen content in C57BL/6 mice with and without YP treatment. There was no significant difference in collagen content (ug/mg protein) in livers of C57BL/6 mice with and without treatment for 1 week with YP (Figure 2). In contrast, following 1 week of YP treatment in AHR knockout mice, there was a significant increase (p<0.001) in collagen content compared to age matched controls, representing a 46% increase in the collagen content when compared to untreated AHR knockout mice. This increase in collagen content following YP (1 week) treatment in 5 week old AHR knockout mice was equivalent to collagen levels in untreated 24 week old AHR knockout mice.

Figure 2: Collagen content in livers of C57BL/6 mice with and without treatment with YP for 1 week in contrast to the effect of Yellow phosphorous (1 wk) on the development of liver fibrosis in AHR knockout mice (AHR -/-). *p<0.001, significantly different from corresponding age-matched controls.

Effect of yellow phosphorous treatment of c-Jun in AHR knockout mice

Treatment of the AHR-/- mice with yellow phosphorous for only 4 days significantly increased c-Jun two fold compared to untreated AHR-/- mice (Figure 3). Based on these results we determined the optimum time to treat mice with c-jun antisense in subsequent experiments.

Figure 3: Effect of YP treatment on c-Jun in livers of AHR-/- mice at several time points after treatment. A representative Western (Panel A) showing the effect of YP treatment on c-Jun in livers of AHR-/- mice compared to control. The liver lysates samples were C57 (lane 1), AHR-/- (lane 2), AHR-/- treated for 2 days with YP (lanes 3 and 4), AHR-/- treated for 4 days with YP (lanes 5 and 6), AHR-/- treated for 7 days with YP (lanes 7 and 8). Equal protein loading on lanes was verified by anti- β actin antibody. *p<0.05, significantly different compared to control and other groups.

Effect of yellow phosphorous treatment on collagen -1 in AHR knockout mice

The results shown in Figure 4 indicate that treatment of the AHR-/- mice with yellow phosphorous for 2 weeks significantly elevated collagen type 1 compared to untreated AHR-/- mice. It is interesting to note that no change was observed in collagen type 3 in AHR-/- mice treated with yellow phosphorous for 2 weeks.

Figure 4: Collagen -1 and collagen-3 in livers of YP treated AHR-/- mice compared to untreated AHR/- mice. *p<0.05, significantly different compared to untreated AHR-/- mice.

Demonstration of c-jun fluorescence in liver sections following tail vein injection

Following injection of c-jun antisense via the tail vein, the c-jun antisense effectively reached the liver, shown as positive fluorescein labelled c-jun antisense by fluorescence microscopy of the liver (Figure 5).

Figure 5: Fluorescence microscopy of the liver following injection of fluorescein labelled c-jun antisense via the tail vein.

Effect of c-jun antisense on c-Jun and collagen -1 in liver of AHR knockout mice

The results shown in Figure 6 illustrate the effect of c-jun antisense on c-jun (Figure 6A) and coll-1 (Figure 6B) in the AHR-/- mice which had also been treated with yellow phosphorous. The c-jun antisense significantly reduced the c-Jun and coll-1 in livers of AHR-/- mice treated with yellow phosphorous; this is in contrast to the lack of effect that occurred using missense.

Figure 6: The effect of c-jun antisense on c-Jun (Figure 6A) and coll-1 (Figure 6B) in the AHR-/- mice which had also been treated with yellow phosphorous. *p<0.05, significantly different compared to control C57 mice. **p<0.05 compared to YP treated AHR-/- mice.

Liver fibrosis is characterized by a significant elevation in hepatic collagen content. Sirius red/Fast green staining allows us to quantitate the collagen deposition in liver sections [6,8,9]. Our results established that there was a significant spontaneous increase in the liver collagen content as the AHR knockout mice developed from 5 weeks to 24 weeks of age. The results also indicate that one week of YP treatment increased hepatic fibrosis in 5-week-old AHR-/- mice and this increase in liver collagen content was equivalent to spontaneously elevated collagen levels in fibrotic 24 week old AHR-/- mice. This is in contrast to C57BL/6 mice where the hepatic collagen content was not changed following one week treatment with YP. Hence the knockout of the AHR results in hepatic fibrosis which can be accelerated with the addition of another known fibrogenic agent, in this case, yellow phosphorous. Interestingly, a recent report indicates the protective role of AHR in acute pancreatitis induced by cerulein and further suggests this occurs through regulation of pancreatic IL-22 [11]. The AHR knockout mouse will spontaneously develop hepatic fibrosis and eventually develops hepatocellular carcinoma. In a similar fashion, it is well established that chronic pancreatitis can contribute to pancreatic cancer which is characterized by an extensive collagen deposition i.e., fibrosis [12].

The yellow phosphorous treatment of AHR-/- mice represents a novel way to accelerate liver fibrosis in an animal model and such a model may prove useful in other forms of fibrosis including pancreatic fibrosis. This fibrosis model is more rapid than the AHR-/- model or YP model alone. We characterized several key molecular changes in this YP induced liver fibrosis model in AHR-/- mice with Western analysis and our results indicate that c-jun and collagen type -1 are elevated. Using this information we intervened with c-jun antisense technology. In contrast to a lack of effect of missense, the c-jun antisense significantly reduced the c-jun and coll-1 in livers of AHR-/-mice treated with yellow phosphorous.

We have previously shown that the drug pentoxifylline (PTX) [13] blocks fibro proliferation and collagen synthesis by inhibiting phosphorylation of c-jun and thus blocks c-jun mediated signalling including signalling via the PDGF receptor. PTX also prevents both the biochemical and histological changes associated with hepatic fibrosis in the animal model of liver disease [8,14]; PTX inhibits neutrophil function [15-17], which may be important since neutrophils play a major role in mediating inflammatory conditions [18-20] and PTX prevents the fibro proliferative effect of PDGF in vitro [21-25]. PTX will reduce fibrosis in cirrhotic patients who undergo resections and decrease collagen deposition in colitis [26,27] and prevent fibrosis in an animal model [23,28,29]. PTX can also reduce fibro proliferation in early stages of injury in the bile duct ligated rat model [28] and down regulate c-jun gene expression known to be involved in growth and proliferation of fibrogenic cells [13,30]. Our studies indicated that PTX alters hepatic stellate cell function and other fibrotic parameters [31]. A recent report suggests that PTX modulates oxidative and nitrosative stress in acute pancreatitis [32], further suggesting the potential therapeutic role for PTX in acute pancreatitis.

The activation of c-jun is important in the up regulated expression of growth and proliferation genes [33] and phosphorylated-c-jun is known to stimulate expression of proliferative genes that promote HSC proliferation during liver fibrosis [13,34]. PTX, normally used to treat vascular disorders [35], decreases fibroblast [36,37] and HSC proliferation [38], decreases PDGF stimulated fibro proliferation and collagen synthesis [23,39] and inhibits IL-18 production [40], NF?B activation [41] and can also down regulate the expression of c-jun in HSCs [13]. HSCs are activated in liver fibrosis [2,42] and express α-smooth muscle actin (α-SMA) [43,44]. We have reported that sera obtained fibrotic HCV patients stimulated HSC proliferation and phosphorylation of c-jun on ser 73, which was decreased by PTX [45]. Recent results suggest that PTX can protect mitochondria in cerulein induced pancreatitis [46].

In this study the technique of Neurath et al. [47] was used to deliver antisense.

Our initial studies using the c-jun antisense involved dose-finding and duration-finding experiments. Our data indicates that we can visualize the c-jun antisense in the liver. Tail vein administration of c-jun antisense downregulated c-jun expression in the livers and this was verified using immunoblots to verify a decrease in c-jun protein levels in livers.

These studies were done in the accelerated model of fibrosis that was developed by administering yellow phosphorous to the AHR-/- mouse. A role for signalling molecules in experimental fibrosis was established using this novel accelerated model of fibrosis which we developed. The yellow phosphorous treatment of AHR-/- knockout mice caused fibrosis within 7 days and provided a narrow window in which to characterize the up regulation of signalling molecules and in this study an up regulation of c-jun was documented. The accelerated model allowed the identification of targets and also provided a reasonable treatment period to inhibit these target molecules. These results together with this accelerated fibrosis model may have significant application in pancreatic fibrosis where there appears to be significant overlap in target cells, mechanism and cytokine involvement [4]. Very recently, a report suggests that JNK signaling may play a significant role in production of reactive oxygen species in cerulein-induced acute pancreatitis [48].

In summary, this study characterized the early stimulation of c-jun and collagen type 1 in the accelerated fibrotic mouse model and employed antisense technology to block the increase in c-Jun and prevent the subsequent increase in collagen, thus blocking fibrosis. These findings could play a significant role in understanding the underlying mechanism in pancreatic fibrosis and have implications in the development of new therapeutics for the treatment of both hepatic and pancreatic fibrosis.

This project was supported by a grant from Canadian Institute of Health Research.