Potential conflict of interest: Nothing to report.
The p8 protein is a transcription factor that regulates the expression of genes involved in cell defense against the adverse effects of stress. Its expression is strongly, rapidly, and transiently induced in most cells on exposure to various stress agents. This study assessed the role of p8 in the response of the liver to CCl4-induced injury. We found that p8 was indeed overexpressed in the liver after CCl4 administration. Hepatic injury following CCl4 injection was monitored in wild-type and p8−/− mice. Serum alanine and aspartate aminotransferase activities were higher and peaked earlier in p8−/− mice than in wild-type mice, which is in agreement with the observation of significantly larger areas of necrosis in p8−/− liver. Absence of p8 expression is therefore associated with increased liver sensitivity to CCl4. In fact, CCl4 toxicity is mediated by derivatives generated by its conversion by the enzyme CYP2E1. It is known that CYP2E1 is downregulated in the liver during the first hours following CCl4 administration as part of a self-defense mechanism. We found that CYP2E1 downregulation was significantly delayed in p8−/− liver compared with wild-type liver, allowing increased production of toxic CCl4 derivatives. In conclusion, inactivation of the p8 gene increases liver sensitivity to CCl4, as it appears to delay the triggering of CYP2E1 downregulation. The p8 protein is therefore an important element of hepatocyte stress response. (HEPATOLOGY 2005;42:176–182.)
The p8 protein was first identified as a new stress-induced protein that is strongly and rapidly but also transiently activated in pancreatic acinar cells during the acute phase of pancreatitis.1 Further experiments also demonstrated that p8 mRNA expression is strongly activated in response to several stresses such as systemic lipopolysaccharide (LPS) administration.2 Moreover, minimal stresses such as routine change of the culture medium also induce p8 gene expression.3 Activation is not restricted to pancreatic cells but occurs also in the liver, kidney, brain, and intestine.2 Therefore, p8 is a ubiquitous protein expressed in response to cellular stress induction. Several functions—some of which are difficult to reconcile—have been attributed to p8, such as growth promotion1, 4 or inhibition5, 6 and promotion of apoptosis.5 In addition, transforming growth factor β-1 activates p8 expression, which in turn enhances the Smad-transactivating function responsible for transforming growth factor β-1 activity.7 Finally, p8 was also shown to promote tumor growth.8
The messenger RNA (mRNA) of p8 comprises about 600 nucleotides and shows a single open reading frame encoding a protein of 80 amino acids. Analysis of p8 primary structure showed presence of a canonical bipartite nuclear localization signal sequence expected to promote nuclear targeting. The protein was indeed localized to the nucleus—though not exclusively, because some labeling was observed in the cytoplasm.4 The p8 protein shares many features of the HMG-I(Y) protein,9 which modulates gene expression by inducing changes in DNA conformation.10–13 On that basis, p8 was defined as a transcription factor that regulates gene expression to improve cell resistance to stress.14 More recently, we compared by DNA microarray analysis the modifications in the pattern of gene expression induced by LPS in the livers of p8+/+ and p8−/− mice. Significant differences were observed for several genes involved in many cellular pathways. Such differences probably account for the increased noxiousness of LPS in p8−/− mice, which underscores the importance of p8 in the defense mechanisms of the liver.15 The aim of the present study was to check whether such function would extend, beyond LPS aggression, to xenobiotic-induced hepatotoxicity. The well-defined model of carbon tetrachloride–induced hepatitis in rodents was chosen.16 Contrary to other hepatotoxic agents, CCl4 is not toxic per se but through generation by cytochrome P450 of a secondary highly reactive agent (trichloromethyl radical) responsible for lipid peroxidation and eventual cellular damage.17, 18 CCl4 is mainly but not exclusively metabolized by the 2E1 isoform of cytochrome P450 (CYP2E1). It was shown that CYP2E1 activity was transiently decreased during the first hours following CCl4 administration. This result was interpreted as a self-defense reaction against CCl4 hepatotoxicity, although the mechanism involved remains unknown. We report here that inactivation of the p8 gene increases CCl4-induced damages to the liver, apparently because transient downregulation of CYP2E1 is delayed. Therefore, p8 seems to be involved in the first line of defense against CCl4 hepatotoxicity.
Three month-old p8−/− and wild-type C57/BL6 mice were used in this study. The p8−/− C57/BL6 mice were generated as previously described.5 After fasting for 18 hours with access to water ad libitum, CCl4 (Sigma, St. Louis, MO) was administered intraperitoneally at 2 μL/g body weight (V/V solution in corn oil) to 64 mice (32 p8+/+, 32 p8−/−). Controls for each group were injected with the corn oil vehicle only. All studies were performed according to the American National Institute of Health guidelines for animal care.
Biochemical Assay and Tissue Samples.
Mice were sacrificed at 0, 12, 18, 24, 48, 72, 96, and 120 hours following CCl4 administration (4 p8+/+ mice and 4 p8−/− mice at each time). For enzyme analyses, blood was collected from the abdominal aorta in anesthetized mice and centrifuged at 4°C, and sera were stored at −80°C until use. The liver was removed on ice, weighed, immediately frozen in liquid nitrogen, and stored at −80°C until use for RNA and protein extractions—except for the left lobe, which was used for histological analysis. Blood plasma levels of alanine aminotransferase (ALT) and aspartate aminotransferase (AST) were measured on the same run by a multiple-point rate test using a Vitros 950AT (Ortho-Clinical Diagnostics) apparatus, according to the manufacturer's instructions.
Liver Histology and Immunohistochemical Analysis.
Liver samples were fixed in 4% formalin and embedded in paraffin. Five-micrometer-thick sections were stained with hematoxilin-eosin for standard examination. For necrosis scoring, 10 randomized fields were selected (original magnification ×100), and necrotic areas were contoured manually and measured via image analysis. Results were expressed as a percentage of the total surface. Images were obtained with an Axiophot microscope (Zeiss, Le Pecq, France) and a 3CCD Camera (Sony, Paris, France). They were processed with an image analysis system (SAMBA 2005; Alcatel TITN, Grenoble, France). Immunohistochemical studies on proliferative cell nuclear antigen (PCNA) were performed using a modification of the avidin-biotin complex method.19 The purpose of the modification was to reduce the background staining due to the binding of the secondary goat anti-mouse antibody to endogenous immunoglobulins. Briefly, complexes of primary PCNA monoclonal mouse antibody (clone PC10, DakoCytomation) and biotinylated secondary antibodies were first generated. Then, sites still available for mouse IgG binding were blocked by incubating the complexes with normal mouse serum before use on tissue sections. Tissue-bound complexes were visualized using an avidin-biotin detection system. Only the PCNA-positive hepatocytes in nonnecrotic areas were considered. For PCNA-positive scoring, 10 randomized fields were selected (original magnification ×400); images were obtained with the Axiophot microscope and processed using the SAMBA 2005 image analysis system. The PCNA-positive nuclei index was calculated as the percentage of PCNA-positive cells per total number of cells counted.
Terminal Deoxynucleotidyl Transferase-Mediated dUTP Nick End Labeling Assay.
Apoptotic cells in liver sections were determined via terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay using the Cell Death Detection Kit (Roche Diagnostics, Meylan, France) following the manufacturer's instructions. Briefly, sections were digested with proteinase K (20 μg/mL) for 15 minutes at room temperature and rinsed with double-distilled water. Slides were then quenched by 2% H2O2 for 5 minutes at room temperature. Slides were then incubated with terminal deoxynucleotidyl transferase (TdT) buffer (30 mmol/L Trizma base [pH 7.2], 140 mmol/L sodium cacodylate, 1 mmol/L cobalt chloride), followed by terminal deoxynucleotidyl transferase reaction solution containing terminal deoxynucleotidyl transferase and dUTP for 90 minutes at 37°C, then washed with 2× standard saline citrate to stop the reaction for 10 minutes at room temperature. The slides were then washed and incubated with antidigoxigenin peroxidase for 30 minutes at room temperature. Color was developed using 0.05% 3,3'-dimethyl aminobenzene in 0.01% H2O2 and then lightly counterstained with hematoxylin. Sections were then washed, dehydrated, and mounted. Apoptotic cells were identified by a brown stain over the nuclei, and the apoptotic index was calculated as the percentage of TUNEL-positive cells per total number of cells counted.
Determination of p8 and CYP2E1 mRNA Levels by Semiquantitative Reverse-Transcriptase Polymerase Chain Reaction.
Determinations of p8, CYP2E1 (cytochrome P450, family 2, subfamily E, polypeptide 1), and control GAPDH mRNA levels were performed on RNA extracted from tissue samples. RNA was extracted using Trizol (Life Technologies, Cergy Pontoise, France) and a 2-mL Lysing Matrix D tube. Total RNA (1 μg) was analyzed via semiquantitative reverse-transcriptase polymerase chain reaction (RT-PCR) using the SuperScript One-Step RT-PCR with Platinum Taq (Invitrogen, Cergy Pontoise, France; Life Technologies) and according to the manufacturer's protocol. For p8, the forward primer was 5′-GCCACCTTGCCACCAACAGCC-3′ and the reverse primer was 5′-GCGCCAGGCTTTTTTCC-3′; for CYP2E1, the forward primer was 5′-CAGGACCTTTCCCAATTCCT-3′ and the reverse primer was 5′-TGACTTTTCTGTGGCTTCCA-3′; and for GAPDH, the forward primer was 5′-ACCACAGTCCATGCCATCAC-3′ and the reverse primer was 5′-TCCACCACCCTGTTGCTGTA-3′. RT-PCR was performed using different numbers of cycles to verify that the conditions chosen were within the linear range. Reverse transcription was carried out for 45 minutes at 45°C followed by 25 (p8), 24 (GAPDH), and 29 (CYP2E1) cycles of PCR, each cycle consisting in a denaturing step for 10 seconds at 95°C, an annealing step for 1 minute at 57°C, and a polymerization step for 1 minute at 72°C. PCR products were separated via electrophoresis on 2% agarose gels.
The CYP2E1 protein level was estimated via Western blotting as previously described.8 Briefly, proteins were extracted in a 2-mL Lysing Matrix D tube containing the following lysis buffer: 0.5% sodium deoxycholate, 50 mmol/L Tris-HCl (pH 8.0), 0.1% SDS, 1% Triton X-100, 150 mmol/L NaCl, and a cocktail of protease and phosphatase inhibitors (Sigma). Mouse monoclonal anti-CYP2E1 was applied overnight at a 1:100 concentration (clone 2-106-12, kindly provided by Kristopher Krausz from the National Institutes of Health,Bethesda, MD). Western blotting was also performed for PCNA using the MAb (1:500) (Santa Cruz Inc., Santa Cruz, CA). ImageJ 1.32 (http://rsbweb.nih.gov/ij/download.html) was used to quantify intensities in Western blots and RT-PCR.
Statistical evaluation was performed using the unpaired Student t test, or ANOVA when multiple comparisons were made. A P value less than .05 was considered statistically significant.
Results and Discussion
The p8 Protein Is Transiently Induced in the Liver After CCl4 Exposure.
Following CCl4 injection, the liver goes through several well-characterized stages: necrosis, inflammatory infiltration, hepatic regeneration, cell proliferation, and deposition of connective tissue (successively). We explored p8 expression in the liver of wild-type mice following CCl4 administration. As shown in Fig. 1, p8 mRNA was barely detectable before treatment. Expression was rapidly and strongly increased after CCl4 injection, reaching at 12 hours a value more than 20 times higher than in control animals. A further increase to 50 times the control value was observed at 18 hours. This was the climax, because p8 mRNA concentration at 24 hours was back to values observed at 12 hours and had almost returned to basal value at 48 hours. Hence, p8 induction by CCl4 is transient. Because p8 is a transcription factor required to promote hepatic defense against LPS,15 we made the hypothesis that the early induction of a strong and transient expression of p8 following CCl4 injection would also help limit liver injury.
Necrotic Injury Is Increased in p8−/− Mice.
We intended to define the role of the early peak of p8 expression that follows CCl4 injection. A single low (nonlethal) dose of CCl4 that causes significant acute liver injury was used to monitor the course of liver damage and repair. Serum ALT and AST activities were used as markers of liver injury (Fig. 2A-B). Maximum levels of serum ALT and AST were observed at 36 hours post–CCl4 injection in wild-type mice, but at 24 hours in p8−/− mice. In addition, those peak values were approximately 1.5-fold higher in p8−/− mice than in wild-type mice. Unexpectedly, p8−/− mice showed an additional peak of serum ALT and AST at 48 hours, less intense than the first one. This observation might reveal the existence of two consecutive necrotic processes after CCl4 treatment, which could be evidenced in p8−/− livers only because, in that case, the first of them is brought forward. In p8+/+ livers, the two processes would have similar time courses and would thus appear as a single process. In p8+/+ mice, ALT and AST values declined from 36 hours post–CCl4 injection, reaching at 72 hours the basal value observed in the untreated animals. However, values at 48 and 60 hours remained higher in p8−/− than in wild-type mice. These data show that in the absence of p8 activity, CCl4 hepatotoxicity occurs earlier and is more severe. Livers from both groups of animals had similar weights (data not shown). We compared the extent of necrosis in hematoxylin-eosin–stained liver sections from p8−/− and wild-type mice. In control animals (corn oil–treated), normal liver architecture was observed in wild-type and p8−/− mice (Fig. 3). The necrotic process started between 12 hours and 36 hours post–CCl4 injection, as indicated by the occurrence of isolated foci of necrotic hepatocytes. Before 36 hours, areas of necrosis represented less than 1% of the total and could not be adequately quantified. From 36 hours to 60 hours, necrotic areas progressed. They exhibited a centrilobular distribution, invading in some cases a large part of the hepatic lobules but always sparing periportal areas. From 72 hours to 96 hours, necrosis was progressively replaced by inflammatory cells. Liver histology returned to normal at 120 hours in both groups. At 36, 48, and 60 hours post–CCl4 injection, p8−/− mice exhibited significantly larger necrotic areas compared with wild-type mice (3.8, 1.8, and 2 times respectively) (Fig. 4). At 72 hours, p8−/− livers presented with persistent centrilobular lesions, contrary to livers from wild-type mice. These data confirm ALT/AST data that CCl4-induced necrosis is more severe in the absence of p8. In other words, our findings strongly suggest that p8 overexpression in damaged liver is involved in tissue protection. One possibility is that p8 is required for efficient apoptosis. In the absence of p8 activity, apoptosis would be inhibited and the only issue for damaged cells would be necrosis. However, that possibility was ruled out by experiments showing the extent of apoptosis, as measured via TUNEL, was similar in p8+/+ and p8−/− livers (data not shown).
Hepatocytic Proliferation Is Similar in Wild-Type and p8−/− Mice.
Because evaluating an area of necrosis at a given time is a snapshot, it reflects a balance between death and regeneration but does not tell about the relative contribution of cells dying through necrosis and of daughter cells resulting from proliferation. To obtain part of the answer, we compared the expression of PCNA, a marker of liver regeneration, in p8+/+ and p8−/− CCl4-treated livers. Both Western blotting (Fig. 5A) and immunohistochemistry (Fig. 5B) revealed that PCNA expression was similar in both groups. This observation suggests that p8 is not involved in hepatic regeneration but rather in a mechanism that takes place in the early stages of the disease.
CYP2E1 Downregulation Is Delayed in p8−/− Mice.
The enzyme CYP2E1 is responsible for most of CCl4 metabolism and, because derivatives generated during that metabolism are hepatotoxic, it plays a major role in the modulation of CCl4-induced liver injury.20, 21 For example, transient downregulation of CYP2E1, always observed in liver following CCl4 administration, is considered an adaptive mechanism that limits toxicity. If, on the contrary, CYP2E1 synthesis is increased22 or the enzyme is stabilized,23 CCl4-induced liver injury is more severe. That mechanism accounts for ethanol potentiation of CCl4 hepatotoxicity.24, 25 Finally, CYP2E1 knockout mice are resistant to hepatoxicity induced by CCl4.21 We asked whether p8 could interfere with the regulation of CYP2E1 gene expression. It does not control basal expression of the gene, because, in untreated mice, CYP2E1 expression was the same as in wild-type and p8−/− mice. The situation was different after CCl4 treatment. Twelve hours after injection, the hepatic level of CYP2E1 mRNA dropped by 50% in wild-type mice as expected (Fig. 6A,C). By contrast, CYP2E1 mRNA expression was maintained at 12 hours in p8−/− livers and showed only a modest decrease after 18 hours. Similar results were found when monitoring the level of the CYP2E1 protein (Fig. 6B,D). It is important to note that in the rodent liver, the CYP2E1 protein has a half-life of only 6 to 7 hours.23 Therefore, the CYP2E1 mRNA downregulation that follows CCl4 treatment is quickly passed on to the protein level, explaining why the decrease in CYP2E1 protein level is rapid in p8+/+ livers and delayed in p8−/− livers (Fig. 6). As a result, the duration of sustained CYP2E1 expression is significantly longer in the absence of p8 activity. A much larger production of noxious CCl4 metabolites is therefore expected. We did not monitor such accumulation but, if confirmed, it could very well account for the increased necrosis observed in p8−/− livers. These results indicate that p8 is involved in the regulation of gene expression that takes place in liver in response to CCl4 administration. Involvement of p8 in protecting a tissue against necrosis was already reported in the pancreas during acute pancreatitis.26 However, p8 expression could reduce—but not prevent—the occurrence of necrosis in liver. Also, it is noteworthy that p8 expression following CCl4 injection and inhibition of CYP2E1 are not concomitant, because maximal p8 expression occurred when the CYP2E1 level had returned to normal values. In addition, CYP2E1 was delayed but not suppressed in p8−/− animals, indicating that factors other than p8 can downregulate the enzyme. Additional functions of the p8 protein in the mechanisms of liver response to CCl4 are therefore expected.
In conclusion, the present study demonstrates that p8 is involved in the mechanisms protecting the mouse liver submitted to CCl4 treatment. As in other tissues submitted to a stress, p8 is rapidly, strongly, and transiently activated in the liver in response to CCl4 injection. Activation of p8 allows rapid downregulation of CYP2E1 expression, which is known to limit the amount of hepatotoxic CCl4-derived metabolites and consequently the extension of liver injury. These findings open up the possibility that p8-associated pathways are involved in the regulation of other CYP450 isoforms, which, if confirmed, would provide new therapeutic options for xenobiotic-induced acute liver injury.27
We thank S. Lotersztajn for helpful discussions and insightful comments.