Liver Fibrosis Protects Mice From Acute ... - Gastroenterology

GASTROENTEROLOGY 2012;142:130 –139

Liver Fibrosis Protects Mice From Acute Hepatocellular Injury ÉRIC BOURBONNAIS,* VALÉRIE–ANN RAYMOND,* CHANTAL ÉTHIER,* BICH N. NGUYEN,‡ MARC SABA EL–LEIL,§ SYLVAIN MELOCHE,§ and MARC BILODEAU* *Laboratoire d’Hépatologie Cellulaire du Centre de Recherche du CHUM-Hôpital Saint-Luc, ‡Département de Pathologie du CHUM, and §Institut de Recherche en Immunologie et Cancérologie and Département de Pharmacologie de l’Université de Montréal, Montréal, Québec, Canada

BASIC AND TRANSLATIONAL LIVER

BACKGROUND & AIMS: Development of fibrosis is part of the pathophysiologic process of chronic liver disease. Although it is considered deleterious, it also represents a form of tissue repair. Deposition of extracellular matrix changes the cellular environment of the liver; we investigated whether it increases resistance to noxious stimuli and the role of changes in intracellular signaling to hepatocytes in mediating this effect. METHODS: Primary cultures of mouse hepatocytes were exposed to type I collagen (COL1); cell injury was assessed by morphologic and biochemical criteria. The expression of Bcl-2 family members was evaluated by immunoblot analyses. Activation of extracellular signal–regulated kinase (ERK) was assessed using phospho-specific antibodies. Liver fibrosis was induced by repeated administration of thioacetamide or carbon tetrachloride to mice; mice were then exposed to Fas antibodies. RESULTS: Hepatocytes exposed to COL1 were more resistant to a variety of hepatotoxins, in a dose-dependent manner, and had lower levels of Bad, Bid, and Bax proapoptotic proteins compared with control hepatocytes. Activation of ERK1/2 was stronger and quicker in hepatocytes exposed to COL1. The MEK1/2 inhibitors U0126 and PD98059 reversed the protective effects of COL1 and the decrease in proapoptotic proteins. Hepatocytes isolated from ERK1⫺/⫺ mice were insensitive to the protective effect of COL1. Fibrotic livers from wild-type mice had high levels of phospho-ERK1 and were resistant to Fas-induced cell death. ERK1⫺/⫺ mice lost this effect. CONCLUSIONS: Production of COL1 during liver fibrosis induces a hepatoprotective response that is mediated by activation of ERK1 signaling. Keywords: Type I Collagen; ECM; Signal Transduction; Apoptosis Induction.

L

iver fibrosis is a characteristic feature of almost every type of chronic liver injury. The synthesis, deposition, and accumulation of extracellular matrix (ECM) occur following the activation of hepatic stellate cells; this observation has been the object of intense research activities in the past decade.1 This process is considered a dynamic one because opposing pathways that drive the degradation of the ECM have also been described.2 The total amount of fibrosis at a given time point is therefore the result of the accumulation of matrix components that are constantly being remodeled by opposing forces over time.

The composition of the ECM is complex, as heralded by the identification of the presence of collagens (I, III, and IV), fibronectin, undulin, elastin, laminin, hyaluronan, and proteoglycans.3 COL1 represents the most abundant form of collagens in both normal and pathologic livers.4 It is mainly produced by activated stellate cells through increased collagen ␣1 messenger RNA stability.5 The localization of ECM deposits within the liver lobule is variable and depends on the etiology of the liver injury. However, in the majority of chronic liver diseases, the process is diffuse in that all areas of the liver are affected in opposition to the local accumulation of scar tissue following vascular disorders in other organs (eg, myocardial infarction). Therefore, the accumulation of ECM inside the liver represents a significant change in the liver microenvironment. Liver fibrosis is considered a wound-healing response to chronic liver injury.6 It is a normal response that occurs whenever a given tissue needs to repair or to protect itself. On the other hand, for clinicians, liver fibrosis is always considered pathologic, the severe accumulation of fibrosis being tied with the development of cirrhosis. Being interested in the mechanisms used by the liver to protect itself against injury, we hypothesized that liver fibrosis significantly modifies the way the liver responds to injury. We herein present evidence that liver fibrosis renders the liver more resistant to acute injury and that type I collagen (COL1) significantly protects hepatocytes against a variety of toxic stimuli via activation of the extracellular signal–regulated kinase (ERK) 1/2 mitogenactivated protein kinase (MAPK) signaling pathway.

Materials and Methods Animals and Materials Mice were obtained from Charles River Laboratories (Saint-Constant, Quebec, Canada). All procedures were performed in accordance with the Canadian Council on Animal Care and approved by the Comité Institutionnel de Protection Animale du CHUM. COL1 from rat tail was obtained from BD Biosciences (Mississauga, Ontario, Canada). Williams’ E medium was from Gibco BRL (Burlington, Ontario, Canada), and formalin was from Fisher (Nepean, Ontario, Canada). U0126 and Abbreviations used in this paper: COL1, type I collagen; ECM, extracellular matrix; ERK, extracellular signal–regulated kinase; MAPK, mitogen-activated protein kinase; TAA, thioacetamide. © 2012 by the AGA Institute 0016-5085/$36.00 doi:10.1053/j.gastro.2011.09.033

FIBROSIS PROTECTS AGAINST LIVER INJURY

PD98059 were from Calbiochem (Gibbstown, NJ). Fas JO2 antibody was purchased from BD Biosciences. All other products were from Sigma-Aldrich (Oakville, Ontario, Canada).

Hepatocyte Isolation and Culture Hepatocytes were isolated from Balb/c, CD1, Erk1⫺/⫺, Erk1⫹/⫺, Erk2⫹/⫺, and ERK1&2⫹/⫺ mice as described.7,8 After isolation, hepatocytes were either plated directly on plastic or on COL1; they were left 2 hours to attach. Cells were cultured at 26,000 cells/cm2 in Williams’ E medium without supplementation.

Morphologic Detection of Apoptosis Cells were fixed with 5% formalin and stained with 0.25 ␮g/mL Hoechst 33258. Apoptosis was evaluated by counting the ratio of nuclei showing predefined alterations.9 3H-Thymidine

Incorporation Assay

Balb/c mouse hepatocytes and Hepa 1-6 mouse cell line (American Type Culture Collection, Manassas, VA) were cultured on plastic or COL1 (13.9 ␮g/cm2). Medium was changed for Williams’ E supplemented with 1 ␮Ci/mL 3H-thymidine (MP Biomedicals, Solon, OH). After 24 and 48 hours, cells were washed with phosphate-buffered saline 1⫻, ice-cold trichloroacetic acid 10%, and ice-cold methanol. Cells were lysed in 0.5 mol/L NaOH, 1 mmol/L EDTA, and 1 ␮L/mL Triton X-100. The solution (200 ␮L) was mixed with 5 mL scintillation cocktail (GE Healthcare, Piscataway, NJ) and radioactivity measured using a TRI-Carb 2800TR analyzer (PerkinElmer, Woodbridge, Ontario, Canada). Protein content was measured according to Bradford.

Caspase Assays Cells were lysed in ice-cold buffer (10 mmol/L HEPES [pH 7.4], 5 mmol/L MgCl2, 42 mmol/L KCl, 0.1 mmol/L EDTA, 0.1% 3-[(3-cholamidopropyl) dimethylammonio]-2-hydroxy-1propanesulonate [Calbiochem; San Diego, CA], 0.1% Triton X-100, 1 mmol/L dithiothreitol, 1 mmol/L phenylmethylsulfonyl fluoride, 10 ␮g/mL leupeptin, 10 ␮g/mL aprotinin, 10 ␮g/mL soybean trypsin inhibitor, and 100 ␮mol/L benzamidine). Lysates were centrifuged at 13,000g. The fluorometric Ac-IETD-AMC cleavage assay was performed. Reaction mixture contained 40 ␮L lysates (100 ␮g for caspase-3 and 200 ␮g protein for caspase-8) and 50 ␮L assay buffer (2⫻; 100 mmol/L HEPES [pH 7.2], 200 mmol/L NaCl, 2 mmol/L EDTA, 20% sucrose, 0.2% 3-[(3-cholamidopropyl) dimethylammonio]-2-hydroxy-1-propanesulonate, 20 mmol/L dithiothreitol). The reaction was started by adding 10 ␮L of caspase-3 or caspase-8 fluorescent substrates: Ac-DEVD-AMC and Ac-IETD-AMC (100 ␮mol/L; Biosource International, Camarillo, CA). The cleavage activity of caspase-3 and caspase-8 was evaluated using 380 and 460 nm as excitation and emission wavelengths in a SPECTRAmax GEMINI microplate spectrofluorometer (Molecular Devices, Sunnyvale, CA). The maximal substrate cleavage rate (Vmax/s) was calculated by SoftMax Pro software (Molecular Devices). The activity of caspase-3 and caspase-8 was derived from a calibration curve relating Vmax/s to increasing units of activated recombinant caspase-3 or caspase-8.

Western Blotting Protein extracts were prepared, subjected to sodium dodecyl sulfate/polyacrylamide gel electrophoresis, and blotted as described.10 Membranes were probed with the following antibodies: anti–Bcl-xl (Transduction Laboratories, Lexington, KY),

131

anti-Bax (Santa Cruz Biotechnology, Santa Cruz, CA), anti-Bak (Upstate Cell Signaling Solutions, Charlottesville, VA), anti-Bid (R&D Systems, Minneapolis, MN), anti-Fas (Santa Cruz Biotechnology, Santa Cruz, CA), anti-Bad, anti-ERK1/2, anti–phosphospecific ERK1/2, anti-AKT, and phospho-AKT (Cell Signaling Technology, Danvers, CA), and anti-actin (Oncogene Research Products, Cambridge, MA). As secondary antibodies, horseradish peroxidase– conjugated anti-rabbit immunoglobulin G and antimouse immunoglobulin G were used (both from BD Pharmingen, San Diego, CA).

Induction of Liver Fibrosis To assess the sensitivity of mice with liver fibrosis to lethal intraperitoneal injections of Fas (0.5 ␮g/g body wt), mice were treated with intraperitoneal injections of saline or 2% thioacetamide (TAA) solution 3 times a week. Animals received 200 ␮g/g body wt of TAA at each injection (this dose had been determined to induce significant liver fibrosis after 12 weeks in preliminary experiments) or saline (vehicle). Five groups of animals were studied; group 1 received saline only for 12 weeks, group 2 received TAA for 12 weeks, group 3 received saline for 12 weeks followed by a 2-week recuperation period, group 4 received TAA for 12 weeks followed by a 2-week recuperation period, and group 5 received TAA for 4 weeks). In an additional experiment, ERK1⫺/⫺ and their controls were treated with TAA for 12 weeks with a 2-week increasing dose protocol (total of 14 weeks) to avoid animal demise. Liver fibrosis was also induced through the intraperitoneal administration of CCl4 (25% solution in sterile mineral oil, 5 ␮L/g body wt injected twice a week) for 8 weeks.

Measurement of Hydroxyproline The hydroxyproline content was expressed as micrograms per gram of wet weight as already reported.11 Briefly, 20 mg of freeze-dried liver samples was hydrolyzed in 6N HCl at 110°C in an autoclave at a pressure of 1.2 kgf/cm2 overnight. After centrifugation at 2000 rpm at 4°C for 5 minutes, 2 mL of supernatant was mixed with 50 mL of 1% phenolphthalein and 8N KOH to obtain a total volume of 5 mL liquid (pH 7– 8). Then, 2 mL of this solution was stirred with 2 g of KCl and 1 mL of 0.5 mol/L borate buffer (pH 8.2) for 15 minutes at room temperature and for another 15 minutes at 0°C, after which 0.2 mol/L chloramine T solution (1 mL) was added and stirred for 60 minutes at 0°C. After addition of 3.6 mol/L sodium thiosulfate (2 mL), the solution was incubated for 30 minutes at 120°C and stirred with toluene (3 mL) for 20 minutes. Next, Ehrlich’s solution (0.8 mL) was added to 2 mL of supernatant after centrifugation at 2000 rpm at 4°C and left for 30 minutes at room temperature. Absorbance was measured at 560 nm.

Histologic Studies Slices were obtained from formalin-fixed specimens of livers from the 5 treatment groups before Fas injection. Assessment of liver fibrosis was performed on Sirius red–stained sections by a single experienced hepatopathologist (B.N.N.) according to the Ishak score.12 Histologic evaluation of injury was performed on liver sections stained with Masson’s trichrome or hematoxylin-phloxin-safranin after Fas injection.

Statistical Analysis All data represent the values of at least 4 experiments, each from a different animal or cell isolation. Data are expressed


January 2012

132

BOURBONNAIS ET AL

GASTROENTEROLOGY Vol. 142, No. 1


Figure 1. Protective effect of COL1 on hepatocyte cultures. (A) Dose-response curve of COL1 (0.003 to 139 ␮g/cm2) on Balb/c mouse hepatocytes exposed to a single dose of Fas (250 ng/mL) for 24 hours. Hepatocytes plated on plastic dishes were used as control. (B) Time-response curve of a fixed dose (13.9 ␮g/cm2) of COL1 following Fas stimulation (250 ng/mL). (C) Catalytic activities of activator capase-3 and effector caspase-8 obtained from hepatocytes plated on plastic or COL1 following Fas stimulation. (D) Apoptotic response of hepatocytes exposed or not to COL1 following Fas stimulation for 8 hours or exposed to actinomycin D (10 nmol/L) for 30 minutes followed by actinomycin D (10 nmol/L) plus tumor necrosis factor ␣ (25 ng/mL) for 20 hours. Effect of (E) COL1 on viability and (F) ALT release of hepatocytes exposed to tertbutyl hydroperoxide for 1 hour. Results are expressed as mean ⫾ SEM from at least 4 different experiments. *P ⬍ .05, **P ⬍ .01, ***P ⬍ .001.

as means ⫾ SEM. Differences among groups were analyzed by using 1- and 2-way analysis of variance for repeated measures or Student t test. A P value less than .05 was considered significant.

Results To test the effect of COL1 on the sensitivity of hepatocytes toward receptor-mediated apoptosis, mouse hepatocytes were isolated and plated on increasing doses of COL1. After attachment, cells were exposed to a single concentration of Fas (250 ng/mL). Figure 1A shows the apoptotic index measured 24 hours after the addition of Fas. Results show that COL1 exposure significantly and dose-dependently decreases Fas receptor-dependent apoptosis by more than 70% at doses of up to 100 ␮g/cm2 COL1. There was no obvious change in hepatocyte morphology on exposure to the collagen monolayer (Fig-

ure 1B, inset). Figure 1B also shows the time response of a fixed (13.9 ␮g/cm2) dose of COL1 on controls and Fasstimulated cells; results show that no significant apoptosis was observed in controls and that the decrease in the sensitivity of hepatocytes was observed at all time points, indicating that the antiapoptotic response was not just a mere delay of response. The measure of alanine aminotransferase (ALT) release in the medium was also found to be decreased when hepatocytes exposed to Fas were plated on COL1 in comparison to those plated on plastic (not shown). Finally, COL1 significantly reduced the activation of both caspase-3 and caspase-8 following Fas stimulation (Figure 1C). Of note, hepatocytes exposed to increasing doses of collagen showed comparable proliferative activity as measured by 3H-thymidine incorporation (Supplementary Figure 1).

January 2012


133

The sensitivity of hepatocytes to other types of receptor-dependent apoptotic stimulation was evaluated by incubating cells in the presence of tumor necrosis factor ␣ together with actinomycin D; Figure 1D shows that, in these conditions, hepatocytes had a lower apoptotic response when exposed to COL1. We also tested the sensitivity of hepatocytes to a nonapoptotic/non–receptor-mediated mode of cell death. Tert-butyl hydroperoxide kills hepatocytes without causing apoptosis.13,14 Hepatocytes plated on COL1 displayed lower sensitivity toward tertbutyl hydroperoxide then those plated on plastic; both the MTT reduction assay (Figure 1E) as well as the ALT release in the medium (Figure 1F) were decreased in cultures plated on COL1. Finally, COL1 was also found to protect against acetaminophen-induced hepatocyte cell death (data not shown). To get insights into the mechanisms by which collagen might protect hepatocytes against apoptotic stimulation, we evaluated the levels of expression of regulatory proteins of the Bcl-2 family in hepatocyte cultures exposed over time to a protective dose of COL1. Figure 2 represents the cumulative data of Western blots for Bcl-xl, Bak, Bad, Bax, and Bid performed on cultures of hepatocytes plated on plastic and COL1 for 4 hours. There was no significant change in the steady-state levels of Bcl-xl over time and between the 2 culture conditions. On the other hand, hepatocytes cultured on COL1 showed a significant

decrease in the levels of Bid and Bax over time, reaching levels lower than 50% of the basal and plastic-control levels. A small decrease in the levels of the proapoptotic Bad protein was also observed when hepatocytes were cultured on COL1, whereas no change in Bak was noted. Altogether, these results show that, when hepatocytes are exposed to COL1, significant changes occur in the expression of Bcl-2 family cell death-regulatory proteins. We next investigated how COL1 transmits intracellular signals that have the potential to impact on the hepatocyte resistance to cell death. Many signal transduction pathways have been shown to mediate the effect of extracellular proteins on target cells.15–18 Because of their wellcharacterized role in hepatocyte behavior and cell survival signaling, we evaluated the contribution of ERK1/2 MAPK.19 –22 Quickly after plating on plastic and collagen, cells were collected and the levels and activating phosphorylation of the MAPKs ERK1/ERK2 were analyzed by Western blotting. Figure 3A shows that ERK1 and ERK2 were activated to a higher extent on cells plated on COL1 at 15 and 30 minutes following the release of inhibition in comparison to plastic. We also tested the expression and activation of AKT in hepatocytes in vitro and in TAA-treated mice; results show no significant AKT activation in hepatocytes exposed to COL1 but increased phospho-AKT levels in TAA-treated livers (Supplementary Figure 2).


Figure 2. Expression of proapoptotic and antiapoptotic proteins on COL1 exposure. Western blots performed on protein extracts of cultures of Balb/c mouse hepatocytes plated on plastic or COL1 (13.9 ␮g/cm2) for 4 hours after attachment. (A) Bcl-xL, (B) Bak, (C) Bad, (D) Bax, and (E) Bid. ␤-actin was used as control. Results are expressed as mean ⫾ SEM from at least 4 different experiments. *P ⬍ .05, ***P ⬍ .001.

134

BOURBONNAIS ET AL



Figure 3. ERK expression in hepatocytes exposed to COL1. Balb/c mouse hepatocytes were plated on plastic or COL1-coated dishes. After attachment in Williams’ E medium containing 10% foetal bovine serum, medium was changed and replaced with serum-free fresh medium overnight. (A) U0126 (25 ␮mol/L) was used to inhibit MAPK for 1 hour before the next change for fresh serum-free media and beginning of the kinetic analysis for 4 hours. (B) U0126 or PD98059 were incubated for 1 hour before the beginning of the experiment for 4 hours. Blots represent the levels of pERK and ERK activities in protein extracts from hepatocyte cultures. (C) Effect of MAPK pathway inhibition by U0126 for 1 hour on Bax and Bid protein expression from hepatocytes plated on COL1coated dishes for 4 hours. Plastic was used as control. (D) Apoptotic index of hepatocytes treated with U0126 or PD98059 for 1 hour before Fas stimulation for 6 hours on plastic or COL1-coated dishes. (E) Effect of collagen on Fas-induced apoptosis on hepatocytes isolated from control CD1 (ERK1&2⫹/⫹) mice and on ERK1&2⫹/⫺, ERK1⫹/⫺, and ERK1⫺/⫺ animals for 6 hours. Results are expressed as mean ⫾ SEM from at least 4 different experiments. *P ⬍ .05, **P ⬍ .01, ***P ⬍ .001.

To better define the role of the ERK1/2 pathway, we tested the effect of 2 chemical inhibitors of MEK1/2 (upstream activators of ERK1/2) on both the protection afforded by COL1 and the changes in the levels of expression of Bid and Bax. First, we showed that both U0126 (25 ␮mol/L) and PD98059 (20 ␮mol/L) efficiently blocked the activation of ERK1 and ERK2 following cell attachment to COL1 (Figure 3B). Treatment with U0126 antagonized the decrease in the levels of expression of Bid and Bax associated with the exposure of COL1 (Figure 3C). We next tested the sensitivity of hepatocytes plated on COL1 to Fas-induced apoptosis in the presence of U0126 and PD98059. As shown in Figure 3D, cells cultured in the presence of the 2 MEK1/2 inhibitors lost the protective effect of COL1. To further show the involvement of ERK1/2 MAPK in this response, we tested the sensitivity

of hepatocytes isolated from ERK1⫺/⫺ knockout animals to Fas stimulation with or without COL1. Results clearly show that ERK1-deficient cells had a similar sensitivity to Fas irrespective of the presence or not of COL1. This observation was not found in single heterozygote ERK

Table 1. Levels of Fibrosis Following 3-Month Treatment With Saline or Thioacetamide Conditions

Ishak score

P value

Saline Saline ⫹ 2-week recuperation period Thioacetamide Thioacetamide ⫹ 2-week recuperation period

0⫾0 0⫾0 3.6 ⫾ 0.4 1.8 ⫾ 0.2

NS NS ⬍.01 ⬍.01

January 2012


135

1⫹/⫺ and ERK2⫹/⫺ and double heterozygote ERK1&2⫹/⫺ cells (Figure 3E). Collagen is the main component of the liver ECM deposited during chronic liver injury. To test the relevance of our observations in vivo, mice were treated with thrice-weekly doses of TAA for 12 weeks to develop experimental liver fibrosis. With this protocol, no significant fibrosis was observed after 4 weeks but, on average, grade 3.6 ⫾ 0.4 fibrosis and a significant increase in the levels of hydroxyproline content were observed after a 12-week treatment (Table 1 and Figure 4A). Two weeks after the end of the TAA treatment, liver fibrosis significantly decreased to grade 1.8 ⫾ 0.2 (14-week time point). We then measured the activity of ERK1/2 in livers obtained from these animals at 4, 12, and 14 weeks. Figure 4B shows that both kinases were activated following the 12-week treatment with TAA, with ERK1 isoform activation more prominent than ERK2. Animals from these 3 different groups were then treated with a lethal injection of Fas (0.5 ␮g/g body wt)

given by intraperitoneal injection. Table 2 shows that animals from the 4-week and 14-week groups showed a similar sensitivity toward the lethal effects of Fas in comparison to controls. On the other hand, animals from the 12-week treatment group all survived. This was associated with a significant decrease in the levels of ALT released in the serum and a decrease in the levels of caspase activities observed in the liver following the injection of Fas (Figure 4C and D). The protective effect of fibrosis was first observed at 8 weeks of TAA protocol, at a time where hydroxyproline content was inTable 2. Survival Rates of Mice 6 Hours After Fas Injection Following TAA or Saline Treatment Conditions

Survival

Death

Rates (%)

1 mo TAA 3 mo saline 3 mo TAA Saline ⫹ 2-week recuperation period TAA ⫹ 2-week recuperation period

5 4 10 5 5

6 6 0 5 6

45 40 100 50 45


Figure 4. Protective effect of fibrosis against Fas-induced injury in vivo. (A) Measure of hydroxyproline content in livers of animals following 1 month, 3 months, and 3 months with a 2-week recuperation period of TAA (200 ␮g/g body wt, 3 times a week) or saline treatments. (B) Phospho-ERK1 and 2 protein expression following TAA or saline treatments for 3 months or 3 months with a 2-week recuperation period treatment. Representative blots of phospho- and total ERK1 and 2. (C) ALT release after 3 months of TAA or saline treatment followed, 48 hours after the last TAA or saline injection, by Fas injection (0.5 ␮g/g body wt) for 6 hours. (D) Catalytic activity of caspase-3 in livers of mice treated with TAA or saline for 3 months followed by Fas injection for 6 hours. (E) ALT release after 8 weeks of CCl4 treatment or vehicle followed, 48 hours after the last CCl4 or mineral oil injection, by a single dose of Fas (0.5 ␮g/g body wt). Animals were killed 6 hours after Fas injection. Controls were obtained from non-TAA, non-Fas–treated animals. (F) ALT release from control CD1 (ERK1&2⫹/⫹) mice and ERK1⫺/⫺ animals after 3 months of TAA or saline treatment followed, 48 hours after the last TAA or saline injection, by Fas injection (0.5 ␮g/g body wt) for 6 hours. Results are expressed as mean ⫾ SEM from at least 8 animals. *P ⬍ .05, **P ⬍ .01, ***P ⬍ .001.

136

BOURBONNAIS ET AL



Figure 5. Representative microphotographs of histologic sections of mice livers. (A and B) Masson-trichrome stain and (C–F) hematoxylin-phloxin-safranin stain. (A) Saline-treated animal. (B) Three-month TAAtreated animal. (C) CD1 salinetreated animal. (D) CD1 3-month TAA-treated animal. (E) ERK1⫺/⫺ animal with saline treatment. (F) ERK1⫺/⫺ animal with 3-month TAA treatment. All animals received a single dose of Fas and were killed 6 hours later. The arrows represent fibrotic tissue. Original magnification 400⫻.

creased (Supplementary Figure 3A). The decrease in liver enzyme release following Fas injection was observed both with ALT and aspartate aminotransferase (AST) (Supplementary Figure 3B and C). The protective effect of fibrosis was also observed when fibrosis was induced by CCL4 (Figure 4E). Liver injury was also significantly less intense when histologic analysis was performed on livers from animals of the TAA-treated group in comparison to controls (Figure 5A and B). No significant difference in the levels of Fas expression was observed between the 4 groups of animals (data not shown). When the expression of proteins from the Bcl-2 family was analyzed in fibrotic livers, we observed similar decreases in BID and BAX expressions as observed in vitro. On the other hand, Bcl-xL levels were higher in fibrotic livers (Supplementary Figure 4). An additional experiment was performed to assess the sensitivity of ERK1⫺/⫺ knockout animals with liver fibro-

sis to Fas in vivo; results show that, in comparison to controls, ERK1⫺/⫺ animals had serum ALT and AST levels that were similar following the injection of Fas, irrespective of the treatment with TAA or not (Figure 4F and Supplementary Figure 5). This was also reflected on histologic examination (Figure 5C–F); ERK1⫺/⫺ animals were shown to develop liver fibrosis following TAA treatment, but the livers of these animals showed no reduced hepatocellular injury following Fas treatment. ERK1⫺/⫺ animals had hydroxyproline levels identical to that of controls after the 12-week TAA protocol (Supplementary Figure 5A).

Discussion The main findings of this study are that liver fibrosis is associated with increased resistance of hepatocytes toward a wide variety of injurious stimuli and that

this protection is mediated at least in part through the MAPK signaling pathway. The role of the ECM on the response of cells to death stimuli has been studied for a long time and is complex.23 In some systems, ECM or its components have been shown to block cell death often through the maintenance of cell attachment and/or the architecture of the tissue,24 while in other systems ECM has been found to favor cell death.25,26 In certain systems, different components of the ECM may have an opposing effect on cell death.27 In the liver, studies have revealed a major role of the ECM in hepatocyte cell survival both in vitro and in vivo, a mechanism mediated by integrins.28,29 Depletion of integrin-linked kinase, a recognized effector of the transduction signal induced by ECM, increases apoptosis in primary cultures of mouse hepatocytes.30 The nature of the ECM to which hepatocytes are exposed has also been shown lately to impact on the sensitivity of these cells toward apoptotic stimuli and their degree of differentiation.31 Collagen is well known to have profound effects on hepatocyte behavior; this is particularly true if cells are exposed to gels of collagen or plated in between layers of collagen (the so-called “sandwich” type).32,33 In these settings, hepatocytes adopt a more cuboidal or rounded morphology, which is different from the one we observed (Figure 1B). Our results also suggest that the age of the ECM might have an impact on the sensitivity of hepatocytes toward death stimuli; newly formed collagen in vivo is associated with increased ERK1 activity, a phenomenon that is quickly lost when the stimulus for new collagen formation is not present anymore (as shown in Figure 4B). Recently, an indirect role of the ECM in affording protection in the setting of ischemia-reperfusion injury has been described through inhibition of matrix metalloproteinases.34 Adhesion to ECM is also believed to be important for the survival of hepatocellular carcinoma cells.35,36 The recognition of a protective effect of fibrosis on the liver is counterintuitive to the clinician’s perspective. However, this idea is consistent with the remarkable capacity of the liver to adapt through tissue repair.37 Experimental evidence has shown that the liver develops an acute phase response to injury and will engage in renewal of its cell mass through regeneration, that cell injury is accompanied by the production and deposition of a complex network of ECM, and that liver fibrosis is at least in part reversible following the cessation of the injury process. Then again, this process, albeit beneficial in the short-term setting, might become problematic in long-term situations. For example, while the increased survival of normal hepatocytes is a desirable goal, it is not the case for abnormal transformed hepatocytes. The increased resistance of normal hepatocytes to cell death might explain in part why patients with chronic liver disease are not that sensitive to the dele-


137

terious effects of toxic compounds; for example, patients with chronically elevated liver enzyme levels are not more sensitive to the hepatotoxicity of statins.38 Furthermore, patients with chronic liver disease have also been shown to tolerate acetaminophen at doses that have been demonstrated to increase liver enzyme levels in healthy individuals.39,40 Other experimental liver models have also confirmed the increased resistance of hepatocytes to certain stimuli; Osawa et al have shown that bile duct–ligated animals (which actually rapidly develop fibrosis) are more resistant to the lethal effects of Fas.41 Recently, Dechêne et al have shown that rapid fibrosis can also be observed in humans in the context of acute liver failure.42 The identification of ERK1, and to a lower extent ERK2, as a key effector signal involved in the establishment of the protective effect of fibrosis and COL1 on hepatocytes is original but is not without precedent. Indeed, Frémin et al have shown that ERK1 was responsible for the long-term survival of primary hepatocytes. ERK1/2 signaling has been mainly tied to the proliferative response of hepatocytes; Frémin et al elegantly showed that the pro-proliferative effect of ERK1/2 is conveyed by the ERK2 isoform.43 The liver is not the only system where a differential role of the ERK proteins has been proposed.44,45 Despite the clear evidence that ERK1 is involved in the process of hepatocyte protection, we do not exclude that other protein kinases are actually involved in this process. Adaptor molecules such as the integrins, focal adhesion kinases, integrin-linked kinase, PINCH, and others are likely to contribute to hepatocyte survival signaling. In that regard, we have evaluated the potential role of AKT in our experimental conditions. Supplementary Figure 2 shows little activation of phospho-AKT in vitro but a strong one in the livers of TAA-treated animals. One could also argue that the effects we observed were caused by the presence of growth factors that are closely embedded in the ECM.15 This is unlikely to be the case because (1) the effect was observed in vitro, (2) the effect was not observed in vivo after 4 weeks of repeated administration of TAA, a condition where liver regeneration does occur and growth factors are being produced, and (3) the most typical effect of growth factors on Bcl-2 family proteins in hepatocytes, which is the increase in the levels of Bcl-xl protein,9,46 was not observed when hepatocytes were exposed to COL1. On the other hand, in vivo, we did observe an increase in Bcl-xL expression in the livers subjected to the induction of fibrosis by TAA; this could be explained by an indirect adaptative response of hepatocytes to the toxic effect of TAA. Alternatively, the source of the increase in Bcl-xL might be originating from nonparenchymal cells. Our results also potentially bear implications for novel agents aimed at removing fibrosis in chronic liver conditions as well as for the MEK1/2 inhibitors that are currently being evaluated for cancer therapy. These


January 2012

138

BOURBONNAIS ET AL

agents may sensitize the liver to liver injury under certain conditions; cases of acute hepatotoxicity secondary to some of these agents have indeed been described lately.47,48

Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of Gastroenterology at www.gastrojournal.org, and at doi: 10.1053/j.gastro.2011.09.033.


20.

21.

22.

References


1. Friedman SL. Liver fibrosis—from bench to bedside. J Hepatol 2003;38(Suppl 1):S38 –S53. 2. Arthur MJ. Fibrogenesis II. Metalloproteinases and their inhibitors in liver fibrosis. Am J Physiol Gastrointest.Liver Physiol 2000;279: G245–G249. 3. Bataller R, Brenner DA. Liver fibrosis. J Clin Invest 2005;115: 209 –218. 4. Zender L, Hutker S, Mundt B, et al. NFkappaB-mediated upregulation of bcl-xl restrains TRAIL-mediated apoptosis in murine viral hepatitis. Hepatology 2005;41:280 –288. 5. Stefanovic B, Hellerbrand C, Holcik M, et al. Posttranscriptional regulation of collagen alpha1(I) mRNA in hepatic stellate cells. Mol Cell Biol 1997;17:5201–5209. 6. Albanis E, Friedman SL. Hepatic fibrosis. Pathogenesis and principles of therapy. Clin Liver Dis 2001;5:315–334. 7. Pages G, Guerin S, Grall D, et al. Defective thymocyte maturation in p44 MAP kinase (Erk 1) knockout mice. Science 1999;286: 1374 –1377. 8. Éthier C, Raymond VA, Musallam L, et al. The anti-apoptotic effect of Epidermal Growth Factor on mouse hepatocytes is associated with a down-regulation of the pro-apoptotic Bid protein. Am J Physiol Gastrointest Liver Physiol 2003;285:G298 –G308. 9. Musallam L, Ethier C, Haddad PS, et al. Role of EGF receptor tyrosine kinase activity in antiapoptotic effect of EGF on mouse hepatocytes. Am J Physiol Gastrointest Liver Physiol 2001;280: G1360 –G1369. 10. Éthier C, Houle R, Bilodeau M. Protective effect of EGF and other growth factors on Fas-induced apoptosis in mouse hepatocytes (abstr). Gastroenterology 1997;112:A339. 11. Sakaida I, Matsumura Y, Kubota M, et al. The prolyl 4-hydroxylase inhibitor HOE 077 prevents activation of Ito cells, reducing procollagen gene expression in rat liver fibrosis induced by cholinedeficient L-amino acid-defined diet. Hepatology 1996;23:755– 763. 12. Ishak K, Baptista A, Bianchi L, et al. Histological grading and staging of chronic hepatitis. J Hepatol 1995;22:696 – 699. 13. Nieminen AL, Saylor AK, Tesfai SA, et al. Contribution of the mitochondrial permeability transition to lethal injury after exposure of hepatocytes to t-butylhydroperoxide. Biochem J 1995; 307:99 –106. 14. Nieminen AL, Byrne AM, Herman B, et al. Mitochondrial permeability transition in hepatocytes induced by t-BuOOH: NAD(P)H and reactive oxygen species. Am J Physiol 1997;272:C1286 –C1294. 15. Hynes RO. The extracellular matrix: not just pretty fibrils. Science 2009;326:1216 –1219. 16. Legate KR, Wickstrom SA, Fassler R. Genetic and cell biological analysis of integrin outside-in signaling. Genes Dev 2009;23: 397– 418. 17. Berrier AL, Yamada KM. Cell-matrix adhesion. J Cell Physiol 2007; 213:565–573. 18. Hehlgans S, Haase M, Cordes N. Signalling via integrins: implications for cell survival and anticancer strategies. Biochim Biophys Acta 2007;1775:163–180. 19. Fremin C, Bessard A, Ezan F, et al. Multiple division cycles and long-term survival of hepatocytes are distinctly regulated by extra-

23. 24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37. 38.

39.

cellular signal-regulated kinases ERK1 and ERK2. Hepatology 2009;49:930 –939. Caja L, Sancho P, Bertran E, et al. Overactivation of the MEK/ERK pathway in liver tumor cells confers resistance to TGF-{beta}induced cell death through impairing up-regulation of the NADPH oxidase NOX4. Cancer Res 2009;69:7595–7602. Haga S, Ozaki M, Inoue H, et al. The survival pathways phosphatidylinositol-3 kinase (PI3-K)/phosphoinositide-dependent protein kinase 1 (PDK1)/Akt modulate liver regeneration through hepatocyte size rather than proliferation. Hepatology 2009;49:204 – 214. Kim HS, Loughran PA, Rao J, et al. Carbon monoxide activates NF-kappaB via ROS generation and Akt pathways to protect against cell death of hepatocytes. Am J Physiol Gastrointest Liver Physiol 2008;295:G146 –G152. Meredith JE Jr, Fazeli B, Schwartz MA. The extracellular matrix as a cell survival factor. Mol Biol Cell 1993;4:953–961. Kong J, Ringer DP. Quantitative analysis of changes in cell proliferation and apoptosis during preneoplastic and neoplastic stages of hepatocarcinogenesis in rat. Cancer Lett 1996;105:241–248. Kuzuya M, Satake S, Ramos MA, et al. Induction of apoptotic cell death in vascular endothelial cells cultured in three-dimensional collagen lattice. Exp Cell Res 1999;248:498 –508. Tang MJ, Hu JJ, Lin HH, et al. Collagen gel overlay induces apoptosis of polarized cells in cultures: disoriented cell death. Am J Physiol 1998;275:C921–C931. Fukai F, Mashimo M, Akiyama K, et al. Modulation of apoptotic cell death by extracellular matrix proteins and a fibronectin-derived antiadhesive peptide. Exp Cell Res 1998;242:92–99. Pinkse GG, Voorhoeve MP, Noteborn M, et al. Hepatocyte survival depends on beta1-integrin-mediated attachment of hepatocytes to hepatic extracellular matrix. Liver Int 2004;24:218 –226. Kumaran V, Joseph B, Benten D, et al. Integrin and extracellular matrix interactions regulate engraftment of transplanted hepatocytes in the rat liver. Gastroenterology 2005;129:1643–1653. Gkretsi V, Mars WM, Bowen WC, et al. Loss of integrin linked kinase from mouse hepatocytes in vitro and in vivo results in apoptosis and hepatitis. Hepatology 2007;45:1025–1034. Godoy P, Hengstler JG, Ilkavets I, et al. Extracellular matrix modulates sensitivity of hepatocytes to fibroblastoid dedifferentiation and transforming growth factor beta-induced apoptosis. Hepatology 2009;49:2031–2043. Hansen LK, Wilhelm J, Fassett JT. Regulation of hepatocyte cell cycle progression and differentiation by type I collagen structure. Curr Top Dev Biol 2006;72:205–236. Berthiaume F, Moghe PV, Toner M, et al. Effect of extracellular matrix topology on cell structure, function, and physiological responsiveness: hepatocytes cultured in a sandwich configuration. FASEB J 1996;10:1471–1484. Defamie V, Laurens M, Patrono D, et al. Matrix metalloproteinase inhibition protects rat livers from prolonged cold ischemia-warm reperfusion injury. Hepatology 2008;47:177–185. Hsu SL, Cheng CC, Shi YR, et al. Proteolysis of integrin alpha5 and beta1 subunits involved in retinoic acid-induced apoptosis in human hepatoma Hep3B cells. Cancer Lett 2001;167:193–204. Schaefer KL, Wada K, Takahashi H, et al. Peroxisome proliferatoractivated receptor gamma inhibition prevents adhesion to the extracellular matrix and induces anoikis in hepatocellular carcinoma cells. Cancer Res 2005;65:2251–2259. Mehendale HM. Tissue repair: an important determinant of final outcome of toxicant-induced injury. Toxicol Pathol 2005;33:41–51. Chalasani N, Aljadhey H, Kesterson J, et al. Patients with elevated liver enzymes are not at higher risk for statin hepatotoxicity. Gastroenterology 2004;126:1287–1292. Gelotte CK, Temple AR, Zimmerman BA, et al. Acetaminophen biotransformation at single and repeat maximum doses is similar between adults with and without chronic liver disease (abstr). Hepatology 2007;46:810A.

40. Watkins PB, Kaplowitz N, Slattery JT, et al. Aminotransferase elevations in healthy adults receiving 4 grams of acetaminophen daily: a randomized controlled trial. JAMA 2006;296:87–93. 41. Osawa Y, Hannun YA, Proia RL, et al. Roles of AKT and sphingosine kinase in the antiapoptotic effects of bile duct ligation in mouse liver. Hepatology 2005;42:1320 –1328. 42. Dechêne A, Sowa JP, Gieseler RK, et al. Acute liver failure is associated with elevated liver stiffness and hepatic stellate cell activation. Hepatology 2010;52:1008 –1016. 43. Frémin C, Ezan F, Guegan JP, et al. The complexity of ERK1 and ERK2 MAPKs in multiple hepatocyte fate responses. J Cell Physiol 2011 Mar 24 [Epub ahead of print]. 44. Owens DM, Keyse SM. Differential regulation of MAP kinase signalling by dual-specificity protein phosphatases. Oncogene 2007; 26:3203–3213. 45. Lips DJ, Bueno OF, Wilkins BJ, et al. MEK1-ERK2 signaling pathway protects myocardium from ischemic injury in vivo. Circulation 2004;109:1938 –1941. 46. Kosai K, Matsumoto K, Nagata S, et al. Abrogation of Fas-induced fulminant hepatic failure in mice by hepatocyte growth factor. Biochem Biophys Res Commun 1998;244:683– 690.


139

47. Ramanarayanan J, Scarpace SL. Acute drug induced hepatitis due to erlotinib. JOP. 2007;8:39 – 43. 48. Ramanarayanan J, Krishnan GS. Review: hepatotoxicity and EGFR inhibition. Clin Adv Hematol Oncol 2008;6:200 –201.

Received October 13, 2010. Accepted September 12, 2011. Reprint requests Address requests for reprints to: Marc Bilodeau, MD, Centre Hospitalier de l’Université de Montréal, Hôpital Saint-Luc, 264, Boul. René-Lévesque Est, Montréal, Québec, Canada, H2X 1P1. e-mail: [email protected]; fax: (514) 412-7314. Conﬂicts of interest The authors disclose no conﬂicts. Funding Supported by a Canadian Institutes of Health Research operating grant (EOP-67275 to M.B.) and by a National Cancer Institute of Canada grant (to S.M.).


January 2012

139.e1

BOURBONNAIS ET AL


Supplementary Figure 1. Proliferative status of mouse hepatocytes. 3H-Thymidine (1 ␮Ci/mL) incorporation was measured on Hepa 1-6 murine cell line and primary mouse Balb/c hepatocytes plated on plastic and on increasing doses of COL1 for 24 and 48 hours.

Supplementary Figure 2. AKT expression in vitro and in vivo. (Top) Western blots of phospho-AKT and AKT performed on protein extracts of cultured Balb/c mouse hepatocytes plated on plastic or COL1 (13.9 ␮g/cm2) for 4 hours after attachment. (Bottom) pAKT and AKT protein expression following TAA or saline treatments for 3 months. Representation blots of phospho-AKT and total AKT in the same conditions. **P ⬍ .01.

January 2012

Supplementary Figure 3. Kinetic of TAA-induced fibrosis. (A) Measurement of hydroxyproline content in livers of Balb/c animals after 4, 8, 10, 12, 14, and 12 weeks followed by a 2-week recuperation period of TAA (200 ␮g/g body wt 3 times a week) or saline treatment. (B) Serum ALT levels were measured from TAA- or saline-treated animals 48 hours after the last TAA or saline injection, followed by a single dose of Fas (0.5 ␮g/g body wt) for 6 hours. (C) Serum AST levels were measured from TAA- or saline-treated animals 48 hours after the last TAA or saline injection, followed by a single dose of Fas (0.5 ␮g/g body wt) for 6 hours.


139.e2

139.e3

BOURBONNAIS ET AL


Supplementary Figure 4. Expression of proapoptotic and antiapoptotic proteins on saline or fibrosis treatments. Western blots performed on livers of mice treated with thrice-weekly doses of saline or TAA for 3 months. (A) Bcl-xL, (B) Bak, (C) Bad, (D) Bax, and (E) Bid. ␤-actin was used as control. Results are expressed as mean ⫾ SEM from at least 8 different animals. *P ⬍ .05, **P ⬍ .01.

January 2012

Supplementary Figure 5. Hydroxyproline content and serum AST enzymes in ERK⫺/⫺ animals and their controls. (A) Measurement of hydroxyproline content in livers of CD1 control and ERK1⫺/⫺ animals following 12 weeks of TAA (200 ␮g/g body wt 3 times a week) or saline treatment. (B) Serum AST levels were measured from TAA- or salinetreated animals 48 hours after the last TAA or saline injection, followed by a single dose of Fas (0.5 ␮g/g body wt) for 6 hours.


139.e4