Molecular Expression Analysis of β-Naphthoflavone-induced Hepatocellular Tumors in Rats

Abstract

The present study was performed to characterize molecular expression levels of preneoplastic and neoplastic lesions induced by β-naphthoflavone (BNF), an aryl hydrocarbon receptor (AhR) agonist in rat hepatocarcinogenesis. Male F344 rats were initiated with an intraperitoneal injection of 200 mg/kg N-diethylnitrosamine, and two weeks later, they were fed a diet containing 0% or 1% BNF for twenty-eight weeks. All animals were subjected to a two-thirds partial hepatectomy at week 3 and sacrificed at week 30. Histopathologically, BNF increased the incidence and multiplicity of altered foci (1.7-fold and 3.3-fold) and hepatocellular adenomas (HCAs) (4.0-fold and 4.7-fold). Immunohistochemically, BNF increased the number of proliferating cell nuclear antigen (PCNA)-positive cells in altered foci (2.3-fold) and HCAs (6.7-fold) compared with the surrounding tissue and decreased the staining of cell cycle regulators (P21, C/EBPα). In addition, loss of reactivity for AhR-regulated (CYP1A1, CYP1B1) molecules and increased reactivity of Nrf-2-regulated (AKR7, GPX2) molecules were also observed in proliferative lesions. Furthermore, increased staining of histone deacetylase (HDAC1) in the nucleus was prominent in HCAs. The differential expression patterns were confirmed at mRNA levels by real-time reverse transcription-polymerase chain reaction (RT-PCR) analysis. These results suggest that enhanced cell proliferation and protection against oxidative stress play an important role in BNF-induced hepatocarcinogenesis in rats.

Introduction

β-Naphthoflavone (BNF), which is also called 5,6-benzofla-vone, is a synthetic derivative of a naturally occurring flavonoid that has the potential to strongly induce cytochrome P450 1A enzymes via aryl hydrocarbon receptor (AhR) activation (Guengerich and Liebler 1985; Prochaska and Talalay 1988). Based on this feature, BNF has been used extensively as an enzyme inducer in assays for metabolic and mutagenic activity; however, there is very little information on the carcinogenic as well as the genotoxic potential of BNF. β-Naphthoflavone itself did not produce mutagenicity in the Salmonella/micro-some assay with or without metabolic activation induced by Aroclor-1254 (Brown and Dietrich 1979). To date, the carcinogenicity of BNF in the liver or other organs has not been clarified because there have so far been no reports of a two-year carcinogenicity study in the available literature.

We have demonstrated that BNF exerts a hepatocellular tumor-promoting activity in rats initiated with N-diethylnitrosamine (DEN) (Dewa et al. 2008; Shoda et al. 2000). The number and area of preneoplastic foci positive for glutathione S-transferase placental form (GST-P) were increased with enhanced oxidative stress responses (e.g., microsomal reactive oxygen species [ROS] production, 8-hydroxydeoxyguanosine [8-OHdG], and thiobarbituric acid-reactive substance [TBARS] contents) and cell proliferation (e.g., proliferating cell nuclear antigen [PCNA]-positive cells) in rats treated with 0.5 to 1% BNF in a medium-term liver carcinogenesis assay. In addition, microarray and real-time RT-PCR analyses revealed that mRNA expressions of AhR-dependent (Cyp1a1, Cyp1b1) and Nrf-2-dependent (Afar, Gpx2) genes were up-regulated, whereas that of cyclin-dependent kinase (CDK) inhibitor p21 was down-regulated in the whole liver. These results indicate that BNF treatment enhances oxidative stress responses and cell proliferation, which may contribute to its hepatocellular tumor–promoting potential in rats; however, the mechanistic role of these molecules in BNF-induced hepatocarcinogenesis has not yet been elucidated.

To gain a better understanding of the mechanism of BNF-induced hepatocarcinogenesis, we compared the differential expression of several relevant genes in proliferative lesions compared to normal hepatocytes. In addition, we also examined the expression levels of CCAAT/enhancer-binding protein α (C/EBPα) and histone deacetylase 1 (HDAC1). A basic leucine zipper protein, C/EBPα cooperates with p21 to inhibit cyclin-dependent kinase-2 activity and induces growth arrest (Harris et al. 2001), and HDAC1 binds to the Cyp1a1 promoter and blocks histone acetylation associated with AhR-mediated transactivation (Schnekenburger et al. 2007).

Materials and Methods

Chemicals

β-Naphthoflavone (CAS No. 6051-87-2) and DEN (CAS No. 55-18-5) were purchased from Wako Pure Chemical Industries (Osaka, Japan) and Tokyo Kasei Kogyo (Tokyo, Japan) with purities of >98% and >99%, respectively. All other reagents were commercially available and of analytical grade.

Animals and Experimental Design

Five-week-old F344/N rats were purchased from Japan SLC, Inc. (Shizuoka, Japan), maintained in an air-conditioned room with a twelve-hour light/dark cycle (room temperature, 23°C ± 3°C; relative humidity, 55% ± 15%), and given free access to a powdered basal diet (Oriental MF; Oriental Yeast, Tokyo, Japan) and tap water. Animals received humane care in accordance with the Guide for Animal Experimentation of the Tokyo University of Agriculture and Technology.

After a one-week acclimatization period, a medium-term liver carcinogenesis bioassay (Ito et al. 2003) was performed according to the following procedure. All animals were initiated with an intraperitoneal injection of DEN (200 mg/kg body weight, dissolved in saline), and given a diet containing 0% (basal diet) or 1% BNF for twenty-eight weeks starting two weeks after DEN initiation. The rationale for dosage was determined based on the results of our previous study, in which 1% BNF for six weeks significantly increased the preneoplastic foci in the livers of rats (Dewa et al. 2008). To enhance hepatocellular proliferation, animals were subjected to a two-thirds partial hepatectomy three weeks after DEN initiation. Body weight and food consumption were measured once a week. At the end of the experiment, rats were euthanized by exsanguination under ether anesthesia, and the livers were excised and weighed. The sliced liver samples were fixed in either 10% phosphate-buffered formalin for histopathological and immunohistochemical evaluations or methacarn solution for differential mRNA expression analysis.

Histopathology and Immunohistochemistry

After formalin fixation, the tissues were dehydrated in graded ethanol and embedded in paraffin. Serial sections were mounted onto glass slides and stained with hematoxylin and eosin (H&E) or immunohistochemically.

All antibodies and immunohistochemical staining conditions examined are detailed in Table 1. Briefly, the sections were deparaffinized, hydrated, quenched for endogenous peroxidase with hydrogen peroxide, and blocked with normal serum. For some antibodies, heat-induced epitope retrieval (HIER) was performed. Primary antibodies were incubated overnight at 4°C and detected using the Vectastain Elite ABC kit (Vector Laboratories, Burlingame, CA, USA) with 3,3′-diaminobenzidine (DAB) as a chromogen. Double staining for GST-P and PCNA were also performed using the Vectastain Elite ABC kit and Vectastain ABC-AP kit with DAB (for GST-P) and BCIP/NBT (for PCNA). After staining, slides were lightly counterstained with hematoxylin. The numbers and areas of GST-P-positive foci (>0.2 mm in diameter) and the total areas of the liver sections as well as PCNA-positive hepatocytes were quantified using WinRoof software (ver. 5.7.2, released on 24^th, Nov., 2006; Mitani Corp., Fukui, Japan).

Differential mRNA Expression by Real-time RT-PCR Analysis

The livers were fixed in a freshly prepared methacarn solution consisting of 60% (v/v) absolute methanol, 30% chloroform, and 10% glacial acetic acid at 4°C for four hours with gentle agitation for laser microdissection, based on the report of Shibutani et al. (2000). Fixed tissues were dehydrated in 99.5% ethanol at 4°C for three successive one-hour cycles, immersed in xylene once for one hour and three successive thirty-minute cycles at room temperature, and then immersed in hot paraffin (60°C) for three successive one-hour cycles. Paraffin-embedded tissues were stored at 4°C until use. Eight to fifteen serial sections (10 μm thick) were prepared from HCAs and the surrounding tissues (BNF-treated group) or normal tissues (DEN control group), mounted on PEN-foil film (Leica Microsystems, Tokyo, Japan), and then dried in an incubator overnight at 37°C. The sections for laser microdissection were deparaffinized, air-dried, and microdissected according to the method reported previously (Nishimura et al. 2008). These mounted tissues from five animals of each group were stained with H&E and anti-GST-P antibody to pathologically confirm the lesions.

Total RNA extraction from the mounted tissues was performed using RNAqueous-Micro (Ambion, Austin, TX) according to the manufacturer’s instructions. Total RNA was quantified using the Quant-iT RNA Assay Kit (Invitrogen, Carlsbad, CA, USA), and then cDNA were synthesized using 100 ng of total RNA as a template and ThermoScript reverse transcriptase (SuperScript III First-Strand Synthesis System, Invitrogen). All PCR reactions were performed using SYBR Green I chemistry (Applied Biosystems, Foster City, CA, USA) and were carried out under the following conditions using an ABI PRISM 7000 Sequence Detection System (Applied Biosystems): incubation at 50°C for two minutes followed by 95°C for ten minutes and forty-five cycles at 95°C for fifteen seconds and 60°C for one minute. The forward and reverse primers listed in Table 2 were designed using Primer Express 2.0 software following the instructions from Applied Biosystems for optimal primer design. The relative differences in gene expression were calculated using threshold cycle (Ct) values that were first normalized to those of the hypoxanthine-guanine phosphoribosyl-transferase (Hprt) gene, the endogenous control in the same sample, and then relative to a control Ct value by the 2-ddCt method (Livak and Schmittgen 2001). The data represent the average fold changes with standard deviation.

Statistical Analysis

All data are expressed as means with their standard deviations. The statistical significance of the difference between the DEN control and BNF-treated group was determined by the F test for homogeneity of variance. If the variance was homogeneous, a Student t test was applied for comparison, and if it was heterogeneous, the Aspin-Welch t test was used. The incidences and multiplicities of altered foci and HCAs observed in the DEN-control and BNF-treated group were analyzed using a Wilcoxon test.

Results

Body and Liver Weights, Food Consumption, and Estimated BNF Intake

β-Naphthoflavone suppressed body weight gain throughout the tumor promotion period (Figure 1), and final body weight was significantly decreased as compared with the DEN control group. However, food consumption was less affected by the treatment (Table 3). The absolute and relative liver weights were significantly increased in the BNF-treated rats, which were approximately 1.5-fold higher than those of the DEN alone group (Table 3).

Histopathological Examinations of Preneoplastic and Neoplastic Lesions

Representative microscopic findings in the livers are shown in Figure 2, and the results are presented in Table 4. β-Naphthoflavone induced diffuse hepatocellular hypertrophy with eosinophilic cytoplasm, altered foci composed of eosinophilic or clear hepatocytes, and HCAs consisting of large, pale eosinophilic and clear hepatocytes with cellular and nuclear atypia compressing adjacent hepatic cell cords. β-Naphthoflavone significantly increased not only the number and the area of GST-P-positive foci, but also the incidence and multiplicity of altered foci (1.7- and 3.3-fold) and HCAs (4.0- and 4.7-fold) compared with the DEN controls. In addition, the effect of BNF on cell proliferation was compared using immunohistochemistry for PCNA. In comparison with the liver in the DEN control, BNF induced slight but significant increases in the ratio of PCNA-positive cells (Table 5). In the livers of BNF-treated animals, the PCNA-positive ratio was significantly increased in the altered foci (2.3-fold) and in the HCAs (6.7-fold) as compared with the surrounding hepatocytes (Figure 2).

Immunohistochemical Examinations of Preneoplastic and Neoplastic Lesions

Since we confirmed in detail the proliferative potential in altered foci and HCAs induced by BNF, the protein expressions of p21 and C/EBPα were examined immunohistochemically (Figure 3). There were decreases in nuclear staining for p21 and C/EBPα in altered foci and HCAs. In particular, the nuclear intensity of C/EBPα was prominently decreased.

We also examined immunohistochemically the expressions of several molecules whose mRNA expression fluctuated in the livers of rats treated with BNF for six weeks in the previous study (Figure 4). Surprisingly, the staining intensity of CYP1A1 and CYP1B1, surrogate markers for AhR activation, were mostly absent in altered foci and HCAs. On the other hand, the Nrf2-regulated, xenobiotic, and oxidative stress-related molecules AKR7 and GPX2 enhanced the staining in the cytoplasm of altered foci and HCAs. Furthermore, it was apparent that the nuclei of HCAs were positive for histone deacetylase 1 (HDAC1) (Figure 5).

Differential mRNA Expression by Real-time RT-PCR Analysis

Differential mRNA expression levels were examined between HCAs and the surrounding hepatocytes by real-time RT-PCR (Figure 6). β-Naphthoflavone significantly decreased mRNA expression of Cdkn1a (p21) and Cebpa (C/EBPα) in the surrounding tissues as compared with the livers from the DEN controls. However, the expression levels in HCAs were not changed as compared with the surrounding tissues. mRNA expression levels of both Cyp1a1 and Cyp1b1 in the surrounding tissues were increased by BNF treatment, but their expression was mostly absent in HCAs. On the other hand, increased mRNA expressions of Afar and Gpx2 in the livers of BNF-treated animals were absolutely enhanced in adenomatous lesions.

Discussion

In this study, we have demonstrated that BNF enhances the incidence and multiplicity of altered foci and HCAs, suggesting that BNF is a potent tumor-promoting agent for hepatocarcinogenesis. There are several lines of evidence for the involvement of AhR in receptor-mediated hepatocarcinogenesis. Novel AhR agonists such as polychlorinated biphenyls (PCBs) and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) are well known to promote rodent hepatocarcinogenesis in two-stage initiation-promotion experiments (Buchmann et al. 1986; Pitot et al. 1980). In addition, Moennikes et al. (2004) demonstrated the oncogenic potential of activated AhR using a transgenic mouse line expressing a constitutively active AhR (CA-AhR). Male CA-AhR-transgenic B6C3F1-mice, which were initiated with DEN and then left untreated for thirty-five weeks, showed considerable increases in liver tumor prevalence and multiplicity as compared with untreated controls. Consequently, it is likely that AhR agonists have the potential to promote hepatocarcinogenesis.

Although the requirement of AhR activation to promote liver tumors is well established, the precise mechanism of AhR-mediated hepatocarcinogenesis is still unclear. Therefore, we have attempted to characterize BNF-induced altered foci and HCAs using immunohistochemical and real-time RT-PCR methods. In this study, altered foci and HCAs induced by BNF contained an increased number of PCNA-positive cells and showed decreased nuclear staining of p21 and C/EBPα as compared with the surrounding tissues. Timchenko et al. (1996) demonstrated that C/EBPα inhibits cell proliferation through post-translational stabilization of p21 protein levels. Low levels of C/EBPα lead to a decrease in p21 protein level, which fails to suppress PCNA activity and block DNA synthesis. Recently, Tan et al. (2005) reported that C/EBPα knock-in mice had reduced susceptibility to DEN-initiated hepatocarcinogenesis. Therefore, altered foci and HCAs were considered to acquire a phenotype that would allow proliferation during continuous tumor promotion by BNF and thereby could contribute to the progression of more cancerous lesions.

Subsequently, we examined the mRNA and protein expressions of AhR-regulated phase I (CYP1A1 and CYP1B1) and Nrf-2-regulated phase II (AFAR and GPX2) enzymes. This was because our recent study confirmed mRNA inductions of not only Cyp1a1 and Cyp1b1, but also of Nrf-2-regulated, detoxifying enzymes such as Akr7, Gpx2, Nqo1, Yc2, and Gstm1 accompanied by prominent oxidative stress in the early stages of hepatocarcinogenesis induced by BNF in rats (Dewa et al. 2008). Both mRNA expression levels and immunohistochemical staining of CYP1A1 and CYP1B1 were reduced, whereas those of AKR7 and GPX2 were increased in altered foci and/or HCAs as compared with the surrounding tissue; much greater levels of mRNA expression of these phase I and phase II enzymes in the surrounding tissues of BNF-treated livers were observed as compared with normal regions of the DEN controls.

Several lines of evidence support the results for the expression of AhR-regulated phase I and Nrf2-regulated phase II enzymes. Gudas and Hankinson (1987) confirmed that differentiated variants of rat hepatoma cells (H4IIEC3) failed to induce cytochrome P450c mRNA and aryl hydrocarbon hydroxylase (AHH; CYP1A1) activity following treatment with polycyclic aromatic hydrocarbons or TCDD. In general, chemically induced hepatocellular nodules were known to express lower levels of xenobiotic enzymes such as cytochrome P450 and mixed-function oxygenases and higher levels of detoxifying enzymes (Farber 1990). In addition, Nishimura et al. (2008) showed that Gpx2 is prominently expressed in GST-P-positive hepatocellular tumors induced in rats by fenofibrate, a peroxisome proliferator. Moreover, Ireland et al. (1998) demonstrated that Afar protein was over-expressed in the preneoplastic lesions and hepatomas in rats treated with AFB1. Collectively, the possibility that these liver preneoplastic and neoplastic lesions would escape from AhR-mediated oxidative stress by the constitutive expression of Nrf2-regulated enzymes can be speculated, whereas in the surrounding hepatocytes, the expression of Nrf2-regulated enzymes was dependent on AhR activation and the subsequent oxidative metabolism of BNF.

Although our results indicate that BNF activates AhR signaling and has a tumor-promoting effect on the livers of rats initiated with DEN, it is not clear how the clonal expansion of cells develops into AhR-independent altered foci and HCAs during the continuous activation of AhR by the treatment of BNF. Interestingly, Hines et al. (2001) suggested a working hypothesis that sustained exposure to the AhR agonist TCDD represented a potent selective pressure favoring mutations that promote cell proliferation by interfering with AhR-mediated cell cycle regulation directly or indirectly. Furthermore, Bock and Köhle (2005, 2006) have also postulated that TCDD-mediated liver tumor promotion may be due to multiple mechanisms favoring the growth advantage of cell clones or a strong selection pressure leading to the outgrowth of clones evading apoptosis and growth arrest in genotoxin-treated livers. Our data might be consistent with these hypotheses, but further study will be needed to clarify the detailed mechanism of AhR-mediated hepatocarcinogenesis.

Increased evidence to suggest that epigenetic alterations including DNA methylation and histone modifications play a critical role in carcinogenesis has accumulated in recent years (Jones and Baylin 2007). For example, an inappropriate acetylation state of histones is known to cause abnormal outgrowth and an altered pattern of cell death, which leads to neoplastic transformation (Kim et al. 2003). Steady-state levels of histone acetylation result from the balance between the opposing activities of histone acetyltransferase and histone deacetylase (HDAC), and HDAC has several subtypes, grouped into four families (class I, class II, Sirtuins, and class IV) (Rasheed et al. 2007). In this study, we have attempted to investigate the expression of histone deacetylase 1 (HDAC1), which belongs to class I HDAC, and nuclear accumulation of HDAC1 in immunostaining was observed in HCAs as compared with the surrounding tissue.

There are several lines of evidence that might be accountable for the epigenetic regulation of gene expression examined in this study. Svechnikova et al. (2007) showed that p21 does mediate the apoptotic effects of the HDAC inhibitors (HDACi) 4-phenylbutyrate and trichostatin A (TSA) on the hepatocellular carcinoma Hep3B cell line using a p21 antisense construct. It was confirmed that TSA treatment induced substantial amounts of p21 or C/EBPα mRNA in hepatoma cell lines (Chiba et al. 2004; Yamashita et al. 2003). In addition, an inhibitor of DNA methyltransferases and other HDACi, 5-aza-cytidine and sodium butyrate, were shown to reactivate a functional Ah receptor in the differentiated hepatoma cell line Fao, indicating that a requisite gene had been silenced by an epigenetic mechanism (Gudas and Hankinson 1987). Furthermore, Garrison et al. (2000) reported that both n-butyrate and TSA increased the constitutive activity of the murine AhR gene promoter in cell lines and was correlated with an increase in endogenous AhR activity in an AhR-deficient cell line. In this study, the expression of p21, C/EBPα, CYP1A1, and CYP1B1 in HCAs was decreased as compared with the surrounding hepatocytes at mRNA or protein levels accompanied by apparent accumulation of HDAC1 in the nuclei of HCAs, indicating that an epigenetic mechanism might be responsible for the regulation of their expression in HCAs.

The effect of HDACi on liver tumors has been explored in in vivo experimental animal models. 4-Phenylbutylate is a derivative of the short-chain fatty acid with HDACi activity, and its administration results in regression of xenografts derived from a liver carcinoma cell line Hep3B in in vivo models (Svechnikova et al. 2003). Recently, Lu et al. (2007) demonstrated that the oral treatment of OSU-HDAC42, a phenylbutyrate-derived HDACi, suppresses the growth of orthotopic and subcutaneous hepatocellular carcinoma xenograft in athymic nude mice. Furthermore, there are many cases in which clinical trials using HDACi for tumors except the liver have been conducted, and some of these studies have also demonstrated clear biologic activity in solid and hematopoietic tumors (Rasheed et al. 2007). Collectively, a potential HDAC inhibitor(s) may prove useful for the treatment of hepatocellular tumors.

In summary, we have characterized the differential expressions of several genes in altered foci and HCAs. The genes influenced include those guiding cell cycle regulation, xenobiotic metabolism, anti-oxidative defense, and histone deacetylation. These results suggest that enhanced cell proliferation and protection against oxidative stress play an important role in BNF-induced hepatocarcinogenesis in rats, which might be partly regulated through an epigenetic mechanism. Future work in this area should help to unravel mechanisms of AhR-mediated hepatocarcinogenesis in rats.