Mechanistic Study on Hepatocarcinogenesis of Piperonyl Butoxide in Mice

Abstract

To clarify the mechanism of piperonyl butoxide (PBO)-induced hepatocarcinogenesis in mice, male mice were subjected to a two-thirds partial hepatectomy, N-diethylnitrosamine (DEN) initiation, and a diet containing 0.6% PBO for eight weeks. The incidence of γ-glutamyl transpeptidase (GGT)-positive foci and PCNA-positive cells was significantly increased in the DEN + PBO group compared with the DEN-alone group. Real-time reverse transcription-polymerase chain reaction (RT-PCR) analysis showed up-regulation of genes related to metabolism, such as cytochrome P450 1A1 and 2B10, and metabolic stress, such as Por, Nqo1, Nrf2, abcc3, and abcc4. Early responsive genes downstream of mitogen-activated protein kinase (MAPK), such as c-fos, c-jun, c-myc, and activating transcription factor 3 (ATF3), were also up-regulated in this group. Positive immunohistochemical staining for ATF3 was diffusely observed in nonproliferating hepatocytes of the DEN + PBO group, but altered foci were negative or weakly positive for ATF3. The nuclei of hepatocytes within ATF3-negative foci were positive for cyclin D. Thus PBO can induce oxidative stress, activate the MAPK pathway, and increase ATF3 transcript levels in hepatocytes outside the altered foci during the early stage of PBO-induced hepatocarcinogenesis in mice.

Introduction

Piperonyl butoxide (PBO), [2-(2-butoxyethoxy) ethoxy]-4,5-methylenedioxy-2-propyltoluene, is a pesticide synergist that is widely used with pyrethroids for grain protection and as a domestic insecticide. PBO acts as a hepatocarcinogen in F344 rats fed a diet containing 1.2% PBO for two years (Takahashi et al. 1994) and induces histopathological changes, such as hepatocellular enlargement, anisonucleosis, and single-cell necrosis, in the livers of imprinting control region (ICR) mice administered a diet containing 0.9% PBO for twenty days (Fujitani et al. 1993). Although the mechanism of PBO-induced hepatocarcinogenesis is not completely known, this compound is classified as a nongenotoxic carcinogen (Beamand et al. 1996). In a previous study (Okamiya et al. 1998), both PBO and phenobarbital induced cytochrome P450 (CYP) isozymes, such as Cyp2b1, and inhibited intercellular communication through gap junctions by down-regulating connexin 32 in the liver. Thus the tumor-promoting mechanism of PBO is similar to that of phenobarbital in terms of hepatocarcinogenesis.

In our previous study, microarray analysis and real-time reverse transcription-polymerase chain reaction (RT-PCR) revealed that the oxidative and metabolic stress–related genes—Cyp1a1, Cyp2a5, Cyp2b9, Cyp2b10, and NADPH-cytochrome P450 oxidoreductase (Por)—were overexpressed in male ICR mice fed a diet containing 0.6% PBO for eight weeks (Muguruma et al. 2006). Thus, PBO metabolites could generate reactive oxygen species (ROS) via the metabolic pathway and induce oxidative stress. Reactive oxygen species are required for physiologic function and regulate cellular signaling in normal cells. However, excessive ROS can induce DNA damage that leads to genomic instability, which may contribute to cancer progression. Thus, ROS are thought to play multiple roles in tumor initiation, progression, and maintenance (Benhar et al. 2002). Meanwhile, ROS cause apoptosis through several mechanisms such as activation of JNK, disruption of mitochondrial membrane potential, and/or direct activation of caspase cascades (Tanaka et al. 2002).

Activating transcription factor 3 (ATF3) is a member of the ATF/CREB transcription factor family, with at least five naturally occurring isoforms derived from alternative splicing. ATF3 is a stress-inducible gene and its basal level of mRNA is low, but ATF3 expression is strongly induced by oxidative stresses such as ROS. Transforming growth factor (TGF) β, p53, and the JNK/MAPK pathway are involved in the induction of ATF3 by stress signals (Hai et al. 1999, Kang et al. 2003). Moreover, ATF3 isoform can heterodimerize with each other and with other transcription factors, such as c-jun, ATF2, Smad3, and Nrf2 (Brown et al. 2008; Okamoto et al. 2006). AFT3 functions can regulate cell proliferation both positively and negatively (Hai et al. 1999; Tamura et al. 2005). For example, ATF3 suppresses Ras-meditated tumorigenesis using ATF3^−/− and RAS/ATF3^−/− cells (Lu et al. 2006), and ATF3 enhances and reduces S-phase arrest through different modulation of the Chk1-Cdk2 pathway using knockdown of ATF3 and Fra1 cells (Hamdi et al. 2008). In our previous study (Kawai et al. 2008), ATF3 mRNA was up-regulated in the liver of DEN-initiated rasH2 mice treated with fenofibrate, and ATF3 was probably involved in the mechanism of hepatocarcinogenesis of rasH2 mice induced by fenofibrate. But it was unclear why ATF3 actually contributed to a promotion or a suppression of preneoplastic lesions of mice.

In this study, we investigated the molecular mechanism underlying the liver tumor–promoting activity of PBO in mice using high-density microarrays, real-time RT-PCR, ROS measurements, and immunohistochemistry in livers obtained from a two-stage hepatocarcinogenesis model.

Materials and Methods

Chemicals

PBO (CAS 51-03-6; technical grade; purity >90%) was obtained from Acros Organics (Morris Plains, NJ, USA), and N-diethylnitrosamine (DEN) was purchased from Nacalai Tesque Co. (Kyoto, Japan). All other chemicals were of analytical grade and obtained commercially.

Animals

Six-week-old male ICR mice were obtained from Japan, SLC Inc. (Shizuoka, Japan). They were housed in plastic cages (each cage contained five animals) with absorbent paper chip bedding in an animal room maintained under standard conditions (room temperature, 22°C ± 2°C; relative humidity, 55% ± 5%; and a twelve-hour light/dark cycle) and given free access to a powdered diet (Oriental MF; Oriental Yeast, Tokyo, Japan) and tap water. The animals were acclimatized for one week prior to beginning the experiment. The experiment was performed in accordance with the guidelines for animal experimentation of the Tokyo University of Agriculture and Technology.

Experimental Design

We used a short-term, two-stage liver carcinogenesis model (Moto et al. 2006) in ICR mice. To enhance hepatocellular proliferation, mice were subjected to a two-thirds partial hepatectomy. Twenty-four hours after the hepatectomy, mice were given a single intraperitoneal (i.p.) injection of DEN (20 mg/kg body weight) dissolved in saline to initiate hepatocarcinogenesis. One week after the injection, the animals were subdivided into two groups and given a powdered diet containing 0% or 0.6% PBO for eight weeks. Mice in the PBO-alone group were given an i.p. injection of saline (1 mL/kg body weight) and a diet containing 0.6% PBO. In our previous study, Muguruma et al. (2006) reported that 0.6%PBO alone for eight weeks was carcinogenic, and that exposure was used in this experiment. On completion of the treatment period, the mice were killed by exsanguination from the posterior vena cava under ether anesthesia, and livers were immersed in 4% paraformaldehyde solution for microscopy. Some of the livers were cut into small pieces, frozen in RNAlater (QIAGEN, Hilden, Germany), and stored at −80°C until analysis.

Histological, Histochemical, and Immunohistochemical Evaluations

After sacrifice, livers were fixed in 4% paraformaldehyde, embedded in paraffin, sectioned at 3 μm thickness, and stained with hematoxylin and eosin (H&E) for histopathological examination. Additionally, immunohistochemical staining for proliferating cell nuclear antigen (PCNA) (PC10; DakoCytomation, USA) and ATF3 was performed using the avidin-biotin-peroxidase complex method.

Histochemical staining of GGT was performed by modifying the methods of Rutenberg et al. (1969). The frozen tissues were cryosectioned and fixed using methanol. After air-drying, a freshly prepared solution containing the substrate, L-glutamic acid-γ-(4-metoxy-β-naphthylamide) (Sigma-Aldrich, St. Louis, MO, USA), and fast blue BBN (Wako Pure Chemical Industries, Osaka, Japan) in 0.1 M Tris-buffered saline (pH 7.4) was coated onto the section. Following incubation, the slides were transferred into a 0.1 M cupric sulfate solution. The sections were then stained with hematoxylin and mounted in 10% glycerol. The number of GGT-positive cells was calculated with positive cells in all lobes on the slide and from the total area in all lobes measured using a computer-assisted image analyzer (WinRoof version 5.7.2; Mitani Corporation, Tokyo, Japan 2006).

The anti-ATF3 (C-19) (rabbit polyclonal antibody); and anti-cyclin D (rabbit polyclonal antibody) were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA, USA) and Upstate Biotechnology (Lake Placid, NY, USA), respectively. The liver sections were deparaffinized in xylene and rehydrated in ethanol. Endogenous peroxidase activity was blocked by incubation for twenty minutes in 0.3% hydrogen peroxide. The slides were preincubated with normal goat serum (Vectastain elite ABC kit; Vector Laboratories, Burlingame, CA, USA) for twenty minutes and PBS containing 0.1% Tween-20 (PBST) for fifteen minutes, and then incubated overnight with the ATF3 antibody at a dilution of 1:100 at 4°C. For antigen retrieval, the sections were heated in 10 mM citrate buffer by autoclaving for ten minutes before incubation with the cyclin D antibody. Then the slides were incubated with a biotinylated secondary antibody for thirty minutes and the streptavidin–biotin peroxidase complex (ABC) for thirty minutes. Subsequently, 3,3′-diaminobenzidine (DAB; Dojindo Laboratories, Kumamoto, Japan) was applied as a chromogen. Finally, the sections were counterstained with hematoxylin, ATF3-negative foci were counted, and the results are expressed as the ATF3-negative and number of foci per mm².

PCNA (Dako, Glostrup, Denmark) immunohistochemistry was performed to determine cell proliferation. The sections were deparaffinized in xylene and rehydrated in ethanol. To enhance immunostaining, the sections were placed in citrate buffer (pH 6.0) and heated in a microwave oven for thirty minutes (low power) for antigen unmasking and then allowed to cool at room temperature for forty minutes. Endogenous peroxidase activity was blocked by incubation with 0.3% hydrogen peroxide in PBS for twenty minutes. After gradual quenching at room temperature, nonspecific binding sites were blocked with blocking solution A (Histofine; Nichirei, Tokyo, Japan), and the specimens were incubated overnight with the primary antibody at a dilution of 1:300 in 0.5%casein-PBS at 4°C. Each section was then incubated with dextran polymers conjugated with the secondary antibodies and peroxidase molecules (Envision; Nichirei, Tokyo, Japan). Subsequently, DAB was applied as a chromogen. Finally, the sections were counterstained with hematoxylin. For quantification of cell proliferation, sections from five mice in each group were analyzed. The number of cells that reacted positively with PCNA antibody was determined by counting at least 3,000 hepatocytes in each liver section, and the PCNA-positive index was calculated as the percentage of positive cells.

Measurement of Microsomal ROS Production in the Liver

The liver samples from each treatment group were homogenized with three volumes of ice-cold 0.25 mol/L sucrose and 0.05 mol/L Tris–HCl buffer (pH 7.4) using a glass-Teflon homogenizer. The homogenate was centrifuged at 700 × g for ten minutes, and the supernatant was centrifuged at 10,000 × g for twenty minutes. The resultant supernatant was further centrifuged at 105,000 × g for ninety minutes. Finally, the pellet was resuspended in 0.25 mol/L sucrose and 0.05 mol/L Tris–HCl buffer (pH 7.4) as the microsomal fraction and stored at −80°C. The microsomal protein concentrations were determined using a BCA Protein Assay Kit (Pierce, IL, USA).

2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA; Molecular Probes Inc., Eugene, OR, USA) was used as a sensitive intracellular probe for detecting ROS formation during the metabolic process by converting DCFH-DA to the highly fluorescent DCF through oxidation by ROS. Liver microsomes (0.05 mg protein) were incubated in the dark at 37°C in 40 mmol/L Tris buffer (pH 7.4) and DCFH-DA (5 mmol/L) for thirty minutes followed by further incubation for thirty minutes after addition of 0.6 mmol/L NADPH. The formation of ROS was detected by measurement of the fluorescent product with a fluorescent micro-plate reader (excitation, 485 nm; emission, 528 nm).

DNA Microarray Analysis

Total RNA was extracted from the DEN-alone group and DEN + PBO-treated mice with TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions and shipped to Kurabo Industries (Osaka, Japan) for oligonucleotide microarray analysis. Gene expression was analyzed using GeneChip Mouse Genome 430 2.0 arrays (Affymetrix, Inc., Santa Clara, CA) consisting of 45,000 probe sets to analyze the expression levels of over 39,000 transcripts and variants from over 34,000 well-characterized mouse genes. The changes were considered significant in the DEN + PBO group if the fold change was greater than 2.0 compared with the DEN-alone group. We selected the oxidative stress–response genes, ATF3 expression-related gene, and genes that changed in a previous study (Muguruma et al. 2006) (Table 1). Gene information was retrieved from the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov) Web site.

Real-time RT-PCR Analysis

The expression of selected genes (Table 1) was quantified using quantitative real-time RT-PCR (qRT-PCR) analysis. Briefly, the total RNA from six mice per treatment group was extracted using TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. The total RNA was reverse-transcribed using ThermoScript reverse transcriptase (SuperScript III First-Strand Synthesis System; Invitrogen). All PCR reactions were performed using SYBR Green I (Applied Bio-systems, Foster City, CA, USA) and performed on an ABI PRISM 7000 Sequence Detection System (Applied Biosystems) using the following conditions: one incubation at 50°C for two minutes followed by 95°C for ten minutes, and then forty-five incubations at 95°C for fifteen seconds and 60°C for one minute. The forward and reverse primers listed in Table 1 were designed using Primer Express 2.0 software following Applied Biosystems’s instructions for optimal primer design. The relative differences in gene expression were calculated using the cycle time (Ct) values that were first normalized to β-actin, the endogenous control in the same sample, and then relative to a control Ct value by a 2-ΔΔCt method described in the Applied Biosystems User Bulletin #2, “Relative quantification of gene expression.” The data represent the average fold changes with standard deviation.

Statistical Analysis

All results are presented as mean ± SD. Data in the DEN-alone group and the PBO-alone group or DEN + PBO group were analyzed by Dunnett’s multiple comparison test, followed by a test of the homogeneity of variance between the groups using Bartlett’s test. When the data were homogenous, Dunnett’s test was used, and when heterogeneous, Dunnett’s rank sum test was used. A p value of less than .05 was considered statistically significant.

Results

Body and Liver Weights and Macroscopic Examination

No deaths or clinical symptoms related to PBO treatment were observed in any of the groups, but all mice in the PBO-treated groups had liver enlargement at necropsy. The final body weights and absolute and relative liver weights are shown in Table 2. The absolute and relative liver weights of the PBO-treated groups were significantly higher than those from the DEN-alone group.

Formation of Hepatic Microsomal ROS in Liver DNA

Reactive oxygen species generation was significantly increased in liver microsomes obtained from mice treated with PBO for eight weeks compared with that of the DEN-alone group (Figure 1).

DNA Microarray Analysis

On the Mouse Genome 430 2.0 array, 6,227 genes showed a more than two-fold increase in the DEN + PBO group compared with the DEN-alone group (data not shown). We focused on the genes up-regulated by PBO treatment, including NAD(P)H dehydrogenase, quinone 1 (Nqo1), cytochrome P450 (Cyp1), family 1, subfamily a, polypeptide 1 (Cyp1a1), Cyp2b5, Cyp2a5, Cyp2b9, Cyp2b10, P450 (cytochrome) oxidoreductase (Por), Cyclin G1, Cyclin D1, x-ray repair complementing defective repair in Chinese hamster cells 5 (Xrcc5), activating transcription factor 2 (ATF2), and ATF3. These genes were also up-regulated in the livers of mice given PBO alone for eight weeks in our previous study (Muguruma et al. 2006).

Real-time RT-PCR Analysis of Genes that Increase with PBO

Using RT-PCR, we confirmed that PBO treatment significantly increased the expression of AhR, Cyp1a1, Cyp2a5, Cyp2b9, Cyp2b10, Nqo1, Nrf2, Por, abcc3, and abcc4 compared with the DEN-alone group. Abcc2 was significantly up-regulated only in the PBO-alone group. In contrast, ATF3, c-fos, c-myc, and c-jun were significantly increased only in the DEN + PBO group (Table 3).

Histopathological and Immunohistochemical Examination

PBO–treated mice showed centrilobular hypertrophy in hepatocytes. Moreover, the number and area of altered foci were significantly increased in the livers of mice from the DEN + PBO group compared with the DEN-alone group (Table 4, Figure 2).

The numbers of GGT-positive cells and PCNA-positive hepatocytes were significantly increased in the DEN + PBO group compared with the DEN-alone group (Table 2, Figure 2). ATF 3 was diffusely and weakly stained in hepatocytes of the DEN-alone and PBO-alone groups, but strongly and diffusely expressed in nonproliferative hepatocytes of the DEN + PBO group. Hepatocytes in altered foci were mostly negative for ATF3 in the DEN + PBO group, despite several small altered foci being diffusely positive for ATF3 (data not shown). Neither ATF3-negative nor positive foci were found in the PBO-alone group or the DEN-alone group. The number and area of ATF3-negative foci were significantly increased in the DEN + PBO groups compared with the DEN-alone group (Table 4, Figure 2). The results of immunostaining of liver lesions using serial sections are shown in Figure 3. The nuclei of hepatocytes within ATF3-negative foci were positive for cyclin D (Figure 3), but the surrounding hepatocytes were almost negative for cyclin D.

Discussion

In our previous studies, we confirmed that metabolized PBO produced ROS and induced oxidative stress, including oxidative DNA damage, in mice and rats (Muguruma et al. 2006; Muguruma et al. 2007). Here, DEN + PBO treatment increased GGT-positive foci and PCNA-positive cells compared with the DEN-alone group. In addition, real-time RT-PCR analysis revealed that genes related to metabolism, such as Cyp1a1, Cyp2b9, and Cyp2b10, and metabolic stress-related genes, such as Por, Nqo1, Nrf2, abcc3, and abcc4, were up-regulated after PBO treatment owing to ROS generation, which leads to hepatocellular tumor promotion. To estimate the cellular sources of oxidative stress, we examined the NADPH-dependent, microsomal ROS production. It is usually considered that the amount of ROS produced is determined by the mitochondria. But despite the fact that the most electron transfer processes in the cell are localized in the mitochondria, more than 60% of the electron-transfer hemoproteins and about 20%–30% of membrane-bound flavoproteins are localized in the smooth endoplasmic reticulum of live cells (Zangar et al. 2004). Moreover, it is generally accepted that microsomal CYP450s sequentially transfer two electrons to oxygen from microsomal NADPH-cytochrome P450 reductase, with the subsequent formation of an oxygenated substrate and water (Poulos and Raag 1992). Although electron transfer is normally a well-coupled process, superoxide and H₂O₂ may be released in the presence of CYP1A inducers that are poorly metabolized. Indeed, PCB increases CYP1A1-dependent microsomal ROS production in the livers of rats as well as scup (Stenotomus chrysops) (Schlezinger et al. 2006).

In the present study, the microsomes isolated from the livers of rats treated with PBO showed enhanced ROS production with a concomitant increase in Cyp1a1 mRNA expression. Therefore, PBO can induce CYP1A1 enzyme and subsequently enhance microsomal ROS production; however, other intracellular sources of ROS generation (e.g., mitochondria) might contribute to cellular oxidative stress responses. In addition, we observed increased gene expression levels of Nrf2-regulated, anti-oxidative stress genes (Nqo1) by qRT-PCR. These results suggest that PBO triggers oxidative stress responses, and that the gene expression levels of phase I and phase II enzymes are intrinsically induced to maintain the cellular redox balance. In our previous report (Muguruma et al. 2006), Xrcc5 overexpression after PBO treatment suggested that ROS production probably led to DNA damage and induced overexpression of DNA repair genes. However, Xrcc5 expression was not changed in this study.

qRT-PCR analysis in our study revealed changes in gene expression driven by oxidative stress. The stress response of ATF3 is mediated through the c-jun N-terminal kinase (JNK), mitogen-activated protein kinase (MAPK), a p53-dependent pathway, or TGFβ (Kang et al. 2003). In the present study, we confirmed the overexpression of certain downstream molecules of the Ras pathway, such as c-fos, c-myc, and c-jun, in the liver of DEN+PBO-treated mice, but TGFβ1 and p53 mRNAs were not changed. In addition, ATF3 can repress Nrf2-mediated signaling, and the formation of ATF3-Nrf2 heterodimers at antioxidant response elements located in the proximal promoter of Nrf2 target genes suppresses recruitment of cAMP-responsive element binding protein, which in turn represses the ability of Nrf2 to induce target gene expression (Brown et al. 2008). Nrf2 can regulate several members of the multidrug resistance–associated protein family, including abcc2, abcc3, and abcc4, in response to a wide variety of chemicals and conditions (Maher et al. 2007). Thus Nrf2 can regulate phase I detoxification enzymes, such as epoxide hydrolase and Nqo1, and efflux transporters, such as abcc2, abcc3, and abcc4. Furthermore, the regulation of Abcc transporters by Nrf2 may contribute to liver disease (Maher et al. 2007). Maher et al. (2007) implicated oxidative stress in the chemical toxicity of the liver and in the pathogenesis of numerous hepatic diseases, including cholestasis, fibrosis, tyrosinemia, and cancer. Moreover, Nrf2 targets and Abcc transporters occur in rodent models of each of these diseases, and Nrf2 coordinately regulates detoxification and transport pathways to mitigate cellular injury (Maher et al. 2007). These references and our results suggest that up-regulation of Nrf2 is deeply related to the injuries resulting from oxidative stress.

PBO treatment also increased expression of Por, a member of the flavoprotein family that transfers electrons from NADPH to cytochrome P450s located in the endoplasmic reticulum (Wang et al. 2005). PBO enhanced Cyp1a1 and Por expression in mouse liver (Muguruma et al. 2006), which would generate oxidative stress. PBO treatment also increased expression of the Cyp2b family members Cyp2b9, Cyp2b10, and Cyp2a5, resulting in increased oxidative stress (Muguruma et al. 2006). Moreover, Cyp2a5 is overexpressed in liver tumors in mice (Kobliakov et al. 1993). PBO increased Cyp2b9, Cyp2b10, and Cyp2a5 in rats given PBO for eight weeks after DEN initiation because of increased oxidative stress and the production of liver tumors (Muguruma et al. 2007).

Oxidative stress produces a rapid and transient increase in mRNA levels of the early response genes, such as c-fos, c-jun, and c-myc (Luna et al. 1994). Early response genes encode transcription factors and therefore play a role in the regulation of cellular responses following endogenous or exogenous stress (Luna et al. 1994). Moreover, these genes also affect cell division and transformation. Deregulated expression of the proto-oncogene c-myc and accumulation of ROS result in increased DNA damage and cell arrest (Sagun et al. 2006; Vafa et al. 2002). Vafa et al. (2002) suggested that c-myc can induce DNA damage and override damage controls, thereby accelerating tumor progression via genetic instability. ATF3 is induced by ROS (Okamoto et al. 2006), DNA damage, and carcinogens (Hai et al. 1999; Hai and Hartman 2001). Moreover, ATF3 promotes cell death and cell cycle arrest and suppresses Ras-mediated tumorigenesis (Lu et al. 2006). In our study, overexpression of c-fos, c-myc, and c-jun was observed in the liver of DEN+PBO-treated mice. It seems that ATF3 may well mediate through the MAPK.

Moreover, ATF3 and Fra1 may dimerize with c-jun to regulate DNA damage-responses, and c-jun controls expression of both pro- and anti-apoptotic genes and can protect cells against the accumulation of DNA damage (Hamdi et al. 2008). Here, PBO increased both ATF3 mRNA and immunostaining in non-proliferating liver cells, suggesting that ATF3 suppressed the growth of hepatocytes other than the altered foci, which are the predominant cell population in the liver, reflecting increased oxidative stress. ATF 3 also blocks cell division by slowing the transition from G1 to S phase in HeLa cells (Fan et al. 2002). Recent work has shown that ATF3 can control the levels and/or activities of p53, Nrf2 and cyclin D. In particular, ATF3 represses Cyclin D1 and A gene transcription during chondrocyte differentiation (James et al. 2006). In the present study, PBO+DEN treatment induced diffuse ATF3 staining, but this staining occurred mostly in hepatocytes outside altered foci. In contrast, PBO- or DEN- alone did not induce ATF3 overexpression or ATF3-negative foci. Thus a lack of functional ATF3 probably inhibits apoptosis and cell cycle arrest and may lead to preneoplastic lesions. In fact, in the present study, the nuclei of hepatocytes within ATF3-negative foci were positive for cyclin D, but it was unclear why ATF3 was down-regulated and cyclin D was up-regulated in the altered foci. Further studies are now in progress to clarify the roles of ATF3 and other factors in hepatocellular tumors induced by PBO treatment for twenty-six weeks after DEN initiation. We also observed several small ATF3-positive foci after DEN + PBO treatment, although staining intensity was similar to surrounding hepatocytes. We could not clarify its role in these ATF3-positive foci, as we focused on the molecular and morphological changes in early hepatocarcinogenesis induced by PBO.

In our previous study, we demonstrated that dicyclanil, a pyrimidine-derived insect growth regulator that inhibits the molting and development of insects, generates ROS because of CYP1A1 induction, and DNA damage causedby ROS is probably involved in hepatocarcinogenesis in mice (Moto et al. 2006). To clarify whether ROS-mediated hepatocarcinogenesis in mice given dicyclanil is inhibited by the treatment of the Siraitia grosvenorii extract (SGE), which has antioxidative properties, we employed a two-stage liver carcinogenesis model in partially hepatectomized male mice. As a result, SGE suppressed the induction of Cyp1a1, leading to inhibition of ROS generation and consequently inhibited hepatocarcinogenesis, probably owing to suppression of Ahr activity (Matsumoto et al. 2009). To clarify whether hepatocellular tumors are induced in mice given PBO owing to the generation ofROS, it is necessary to perform anadditional study as a future work in which ROS-mediated hepatocarcinogenesis in mice given PBO is inhibited by the treatment of ROS quenchers or antioxidants.

In conclusion, the results of our study suggest that PBO can induce oxidative stress, activate MAPKs, and increase ATF3 gene expression in hepatocytes outside the altered foci during the early stage of PBO-induced hepatocarcinogenesis in mice, but further studies are necessary to clarify why the decreased expression of ATF3 in altered foci results in increased cell proliferation.