Concurrent Administration of Ascorbic Acid Enhances Liver Tumor

The Journal of Toxicological Sciences (J. Toxicol. Sci.) 127 Vol.33, No.2, 127-140, 2008

Original Article

Concurrent administration of ascorbic acid enhances liver tumor-promoting activity of kojic acid in rats Masayoshi Takabatake, Makoto Shibutani, Yasuaki Dewa, Jihei Nishimura, Hironobu Yasuno, Meilan Jin, Masako Muguruma, Taichi Kono and Kunitoshi Mitsumori

Laboratory of Veterinary Pathology, Tokyo University of Agriculture and Technology, 3-5-8 Saiwai-cho, Fuchu, Tokyo 183-8509, Japan

(Received November 19, 2007; Accepted December 11, 2007)

ABSTRACT — We previously found that administration of ascorbic acid (AA) enhances the liver tumor- promoting activity of kojic acid (KA) in mice. To examine the reproducibility of these results in rats and the underlying mechanism of this effect, we employed a two-stage liver carcinogenesis model using male F344 rats. Two weeks after initiation with diethylnitrosamine (DEN), the animals received a diet containing 2% KA and drinking water with or without 5,000 ppm AA for a period of 7 weeks. A DEN- alone group was also established as a control. One week after the commencement of the administration, the animals were subjected to two-thirds partial hepatectomy. At the end of the experiment, the livers were analyzed immunohistochemically, and the mRNA expression level and extent of lipid peroxidation were measured. AA treatment enhanced the KA-induced tumor-promoting activity in terms of the number and area of liver cell foci that were positive for glutathione-S-transferase placental form. AA coadminis- tration increased the number of hepatocytes positive for proliferating cell nuclear antigen and inversely decreased the number of TUNEL-positive cells. However, the increased level of thiobarbituric acid reac- tive substances resulting from KA treatment was suppressed by coadministration of AA. Gene expression analyses using low-density microarrays and real-time RT-PCR showed that coadministration of AA result- ed in upregulation of genes related to cell proliferation and downregulation of those involved in apopto- sis and/or cell cycle arrest. These results indicate that the concerted effects of AA on cell proliferation and apoptosis/cell cycle arrest probably through its antioxidant activity are involved in this enhancement.

Key words: Apoptosis, ascorbic acid, kojic acid, cell proliferation, hepatocarcinogenesis, reactive oxygen species (ROS)

INTRODUCTION cosmetics and dermatological preparations because it can also inhibit human melanocyte tyrosinase (Maeda et al., Kojic acid (KA: 5-hydroxy-2-(hydroxymethyl)-4- 1991). pyrone) is a metabolic product of glucose that is gen- With regard to the genotoxicity of KA, several in vit- erated by various species of fungi or bacteria, such as ro studies, including those on mutations in Salmonel- Aspergillus oryzae, Penicillium, and Acetobacter species la typhimurium strains TA98 and TA100, sister chroma- (Parrish et al., 1966). It is known that KA inhibits tyrosi- tid exchanges and chromosomal aberrations in Chinese nase, which catalyzes the conversion of tyrosine to mel- hamster ovary cells, have shown positive results (Wei et anin in mushrooms, potatoes, apples, and crustaceans al., 1991). However, negative results have been obtained (Chen et al., 1991). Additionally, KA has antimicrobial in hprt gene mutation assays in Chinese hamster V79 effects on several bacterial strains (Nohynek et al., 2004). cells and mouse L5178 cells and in an in vitro micronu- KA has been used as a food additive for preventing enzy- cleus assay using human keratinocytes and hepatocytes matic browning and maintaining the freshness of vegeta- (Nohynek et al., 2004). bles, raw crabs, and shrimps for many years. In addition KA is carcinogenic in the thyroid and liver of rodents. to this, KA has been used as a skin-lightening agent in Dietary-administered KA can induce thyroid follicular

Correspondence: Makoto Shibutani (E-mail: [email protected])

Vol. 33 No. 2 128

M. Takabatake et al.

cell adenomas in B6C3F1 mice (Fujimoto et al., 1998) purpose, we utilized a medium-term liver bioassay and and can also exert a tumor-promoting effect on thyroid DVVHVVHGWKHPRGL¿FDWLRQDIWHUZHHNVRIFKHPLFDOWUHDW- follicular cells in F344 rats (Mitsumori et al., 1999). KA ments (KA + AA) during the promotion phase. We also decreases the levels of serum thyroxine (T4) and 3,5,3-tri- examined gene expression changes related to cell prolif- iodothyronine (T3) and elevates those of thyroid stimulat- eration, cell cycle, and apoptosis to determine the corre- ing hormone (TSH) (Mitsumori et al., 1999) by inhibit- sponding molecular mechanism in case a tumor-promo- LQJWK\URLGLRGLQHXSWDNHDQGLWVRUJDQL¿FDWLRQ )XMLPRWR tion modifying effect existed. et al., 1999; Tamura et al., 1999). Consequently, thyroid carcinogenicity is strongly suggested to be due to nega- MATERIALS AND METHODS tive feedback of the pituitary-thyroid axis. With regard to hepatocarcinogenesis, B6C3F1 mice receiving 3% KA Chemicals and preparation of test diet and in the diet were found to develop hepatocellular tumors drinking water (Fujimoto et al., 1998). It was reported that 26-week KA (purity: 100%), kindly provided by Alps GLHWDU\DGPLQLVWUDWLRQRI.$DWFRQFHQWUDWLRQV• Pharmaceutical Industry (Gifu, Japan), was admixed induced hepatocellular adenomas in heterozygous p53- into powdered basal diet (MF; Oriental Yeast Co., Ltd., knockout CBA mice (Takizawa et al., 2003). In our pre- Tokyo, Japan) at 3%. AA in the form of sodium ascor- vious study using CBA wild-type mice, dietary adminis- bate was purchased from Wako Pure Chemicals (Tokyo, tration of 1% KA for 26 weeks increased the incidence of Japan) and mixed in the drinking water at a concentra- proliferative lesions in the liver (Watanabe et al., 2005). tion of 5,000 ppm. Diet and drinking water containing A significant increment in the number and area of glu- KA and AA, respectively, were prepared once a week and tathione-S-transferase placental form (GST-P)-positive stored at 4°C until use. DEN was purchased from Nacalai liver cell foci, which is a well-established early mark- Tesque (Kyoto, Japan). er for hepatocarcinogenic activity in the rat liver (Shirai, 1997; Ito et al., 2000), was reported in a two-stage hepa- Animals tocarcinogenesis study using N-bis(2-hydroxypropyl) nit- A total of 27 male F344 rats (4 weeks old) were pur- rosamine (DHPN) or N-diethylnitrosamine (DEN) as chased from Japan SLC (Shizuoka, Japan) and used after initiators (Takizawa et al., 2004). However, liver tumor- being subjected to an acclimatization period of 2 weeks. initiating activity was absent in ICR mice and F344 rats They were housed in stainless steel cages with four ani- (Moto et al., 2006a; Higa et al., 2007). These results sug- mals per cage. Each animal had free access to diet and tap gest that KA exerts only promotion activity in the liver water. All the animals were handled under standard con- of mice and rats even though it has an in vitro genotoxic ditions (room temperature, 23 ± 3°C; relative humidity, potential, as reported earlier (Nohynek et al., 2004). 55 ± 15%; 12-hr light and dark cycle). Ascorbic acid (AA) is a well-known water-soluble vitamin that can exert antioxidative activities by scav- Experimental design enging reactive oxygen and nitrogen species (Frei et al., Animals were treated with test chemicals during the 1989). The antimutagenic, anticlastogenic, and anticarci- post-initiation phase in the two-stage hepatocarcino- nogenic effects of AA (Cameron et al., 1979) have been genesis model that utilizes a medium-term rat liver bio- reported. On the other hand, enhancement of the tumor- assay (Fig. 1; Shirai, 1997; Ito et al., 2000). At the age promoting activity of sodium nitrite has been reported in of 6 weeks, the animals, weighing 114.7 ± 12.1 g each, the forestomach of rats (Okazaki et al., 2006). In addi- were divided into 3 groups (Group 1: DEN-alone; Group tion, AA enhanced the promoting effect of NaHCO3 in a 2: DEN + KA; and Group 3: DEN + KA + AA) and sub- rat two-stage urinary bladder carcinogenesis model (Iwata jected to a single i.p. injection of DEN (200 mg/kg body et al., 1997). Light-induced skin tumors in hairless mice weight) dissolved in saline in order to initiate hepato- were also increased by AA (D’Agostini et al., 2005a). carcinogenesis. Two weeks after the injection, the ani- Furthermore, we recently found that AA can enhance the mals were fed KA at a concentration of 0% (group 1) proliferation activity of KA in mice hepatocytes (Kono et or 2% (groups 2 and 3) for 7 weeks. Animals in group al., 2007). 3 were also given 5,000 ppm AA in the drinking water In the present study, we examined the role of AA in for 7 weeks. In order to enhance hepatocellular prolifera- enhancing KA-induced tumor promotion in hepatocytes. tion, animals were subjected to two-thirds partial hepate- A two-stage carcinogenesis model using rats and with ctomy 1 week after the commencement of KA and/or AA GST-P as the endpoint marker was employed. For this administration. Nine animals died after partial hepatecto-

Vol. 33 No. 2 129

AA enhanced liver tumor promotion activity of KA in rats

slipped for microscopic examination. The number and area of GST-P-positive foci larger than 0.2 mm in diame- ter and the total area of the liver sections were measured using a computer-assisted image analyzer (Scion image; Scion Corp., Frederick, MD, USA) to obtain values per square centimeter of liver section. With regard to cellu- lar proliferation, the number of PCNA-positive liver cells on each slide was randomly counted in 10 different are- DVXQGHUîPDJQL¿FDWLRQWRREWDLQWKH3&1$SRVLWLYH cell index (calculated as the value per 100 cells). Final- ly, more than 2,500 hepatocellular nuclei were counted in Fig. 1. Experimental design. any one slide from each animal. With regard to quantitative measurement of apoptosis, IRUPDOLQ¿[HGVHFWLRQVWKDWZHUH—PLQWKLFNQHVVZHUH my, while 18 animals survived (4 in group 1, 6 in group analyzed by the terminal deoxynucleotidyl transferase- 2, and 8 in group 3). At the end of the 7-week adminis- mediated dUTP-biotin nick end labeling (TUNEL) meth- WUDWLRQSHULRGDOOVXUYLYLQJDQLPDOVZHUHVDFUL¿FHGXQGHU od using the ApopTag in situ detection kit (Chemicon, deep anesthesia with ether. Their livers were excised and Temecula, CA, USA), according to the manufacturer’s weighed, and the right, right medial, and caudate lobes instructions. On each slide, the number of TUNEL-pos- were sliced and fixed in neutral buffered formalin (pH itive liver cells per 3,000 liver cells was counted from 7.4) or in methacarn solution (methanol:chloroform:ace- UDQGRPO\VHOHFWHGDUHDVXQGHUîPDJQL¿FDWLRQWR WLFDFLG  DQGHPEHGGHGLQSDUDI¿QIRUKLVWRORJL- obtain the apoptotic index (AI). The AI was expressed as cal and immunohistochemical examinations. Portions of the apoptotic cells scored per 100 cells. the right lobe were also cut into small pieces to immerse in RNAlater (QIAGEN, Hilden, Germany) overnight and Measurement of lipid peroxidation by formation stored at -80°C until further analysis. The experiment was of TBARS performed in accordance with the guidelines for animal Lipid peroxidation was measured by the formation of experimentation of the Faculty of Agriculture, Tokyo Uni- thiobarbituric acid (TBA) reactive substances (TBARS). versity of Agriculture and Technology. The animal proto- Liver samples (300 mg) were homogenized in 1 ml of col was reviewed and approved by the Animal Care and 1.15% KCl solution. Sodium dodecyl sulfate (SDS) solu- Use Committee of the University. WLRQ —O EXIIHUHGDFHWLFDFLGVROXWLRQ ȝO EXW\ODWHGK\GUR[\WROXHQHVROXWLRQ  Immunohistochemical evaluation ȝO DQG7%$VROXWLRQ ȝO ZHUHDGGHGWR 6HFWLRQV ȝPLQWKLFNQHVV RIPHWKDFDUQ¿[HGOLYHU ȝORIWKHKRPRJHQL]HGPL[WXUHLQDPOPLFURWXEHDQG WLVVXHVHPEHGGHGLQSDUDI¿QZHUHSUHSDUHGDQGVWDLQHG vortexed. After incubation at 4°C for 60 min, the reac- with hematoxylin and eosin (HE) for histological exam- tion mixture was incubated at 95°C for 30 min, and the ination. Immunohistochemical analysis was performed reaction was stopped by placing the tubes in tap water. on these sections to estimate the enzyme-altered liver cell TBARS were subsequently extracted by adding 2.5 ml of foci that were positive for GST-P and to examine prolif- a pyridine-butanol mixture (1:15), and the aqueous phase erating cells exhibiting proliferating cell nuclear anti- was separated by centrifugation in order to measure the gen (PCNA). Rabbit anti-rat GST-P antibodies (1:1000; optical densities at 532 nm. Malondialdehyde (MDA), Medical & Biological Laboratories Co., Ltd., Nagoya, obtained by acid hydrolysis of 1,1,3,3-tetra-ethoxy-pro- Japan) and mouse monoclonal anti-PCNA antibod- pane, was used as the standard for quantifying TBARS. ies (clone PC10; 1:1000; Dako Japan, Kyoto, Japan) Data were expressed as nmol MDA per g liver. were used for the former and latter, respectively. Immu- nodetection was carried out by the avidin-biotin-per- cDNA microarray analysis oxidase complex method in accordance with the man- Global gene expression analysis was performed ufacturer’s instructions (Nichirei, Tokyo, Japan). The with one liver sample from each group using the Oligo peroxidase-binding sites were demonstrated using 3,3’- GEArray® Rat Toxicology & Drug Resistance Microarray diaminobenzidine/H2O2 as the chromogen. Sections (ORN-401; SuperArray Bioscience Corp., Frederick, were then counterstained with hematoxylin and cover- MD, USA). Total RNA was isolated using the TRI-

Vol. 33 No. 2 130

M. Takabatake et al. zol reagent, according to the manufacturer’s protocol expression changes by using real-time RT-PCR: caspase 8 (Invitrogen, Carlsbad, CA, USA). cRNA was synthe- (Casp8), cyclin-dependent kinase inhibitor 1B (Cdkn1b), VL]HGIURPȝJWRWDO51$XVLQJWKH7UXH/DEHOLQJ$03 CHK2 checkpoint homolog (Schizosaccharomyces kit (SuperArray Bioscience Corp.) with conversion of the pombe) (Chek2), DNA-damage inducible transcript 3 total RNA biotin-labeled cRNA probe through a cDNA (Ddit3), growth arrest and DNA-damage-inducible 45 synthesis step. The array filters were hybridized over- beta (Gadd45b), O-6-methylguanine-DNA methyltrans- QLJKWDWƒ&ZLWKWKHELRWLQODEHOHGSUREHV7KH¿OWHUV ferase (Mgmt), tumor necrosis factor (TNF superfamily, were then washed twice with 2× saline-sodium citrate member 2) (Tnf), tumor protein p53 (Tp53), TNFRSF1A- buffer (SSC)/1% SDS and then twice with 0.1× SSC/1% associated via death domain (Tradd), and peroxiredoxin SDS at 60°C for 15 min each. Chemiluminescent detec- 1 (Prdx1). Several genes related to apoptosis or the cell WLRQZDVFDUULHGRXWE\VXEVHTXHQWLQFXEDWLRQRIWKH¿O- cycle, such as those encoding caspase 4 (Casp4), mitogen- ters with alkaline phosphatase-conjugated streptavi- activated protein kinase 14 (Mapk14), transforming din and CDP-Star substrate, and the signal was detected JURZWKIDFWRUȕ 7JIE DQGF\FOLQ* &FQJ ZHUH using a Bio Imaging System (UVP BioImaging System additionally examined for changes in mRNA expression Corp., Upland, CA, USA) with a cooled integrating mon- induced by the treatment. The genes and sequences of ochrome CCD camera, according to the manufacturer’s primers for real-time RT-PCR analysis are shown in Table 1. protocol. The image data obtained from GEArray® were analyzed using the GEArray® Expression Analysis Suite Statistical analyses software (SuperArray Bioscience Corp.). The image data All data except that obtained from the real-time RT- were normalized to the expression value of a housekeep- PCR analysis were analyzed by the Tukey-Kramer mul- ing gene, i.e., peptidylprolyl isomerase A (Ppia; Acces- ticomparison test when the variance was proven to be sion No.: NM_017101), which was reported to be a better homogeneous among the groups by using the Bartlett test internal control than beta-actin or GAPDH (Feroze et al., IRUHTXDOYDULDQFH,IDVLJQL¿FDQWGLIIHUHQFHLQWKHYDUL- 2002). For each spot, the ratio of the intensities between ance was observed, the Steel-Dwass test was performed. groups 1 and 2 and between groups 2 and 3 was ana- With regard to the real-time RT-PCR data, comparisons lyzed. were made by the Student’s t-test between groups 1 and 2 and between groups 2 and 3 when the variance was prov- Real-time RT-PCR en to be homogeneous among the groups by the test for Quantitative real-time RT-PCR was performed with HTXDOYDULDQFH,IDVLJQL¿FDQWGLIIHUHQFHLQWKHYDULDQFH the SYBR green PCR master mix (Applied Biosystems, was observed, Welch’s t-test was performed. Differences Foster, CA, USA) using the ABI PRISM 7000 Sequence ZHUHFRQVLGHUHGVLJQL¿FDQWZKHQWKHp value was <0.05. Detection System (Applied Biosystems) to validate the microarray analysis results. Three liver samples from RESULTS each group were used. For reverse transcription, cDNA ZDVV\QWKHVL]HGIURPȝJRIWRWDO51$XVLQJ'77 Body and liver weights dNTP, random primers, RNaseOUT, and SuperScript III Data on the final body and liver weights are shown UHYHUVHWUDQVFULSWDVH ,QYLWURJHQ&RUS LQȝORIWKH in Table 2. Although the food consumption was slightly reaction mixture. The real-time PCR reaction was then higher in the DEN + KA and DEN + KA + AA groups performed according to the SYBR green PCR master than in the DEN-alone group (data not shown), a signif- mix protocol (http://docs.appliedbiosystems.com/cms/ icant decrease in the body weight was observed in the groups/mcb_support/documents/generaldocuments/cms_ DEN + KA + AA group as compared with the DEN-alone 041053.pdf). The oligonucleotide primers for PCR were group. There was no major difference in the absolute liver designed using the Primer Express software (version ZHLJKW7KHUHODWLYHOLYHUZHLJKWZDVVLJQL¿FDQWO\KLJK- 2.0.0, Applied Biosystems). The expression value of the er in the DEN + KA and DEN + KA + AA groups than in WDUJHWJHQHWUDQVFULSWVZDVFDOFXODWHGE\WKHǻǻ&WPHWK- WKH'(1DORQHJURXSEXWQRVLJQL¿FDQWGLIIHUHQFHVZHUH ods (Livak et al., 2001) after normalization with the level observed between the DEN + KA and DEN + KA + AA of the housekeeping gene, i.e., myosin Ib (Myo1b; Acces- groups. sion No.: NM_053986) (Cai et al., 2004), and the values were expressed relative to the control level. Based on the Pathological and immunohistochemical changes gene expression changes in the DEN + KA + AA group, in the liver the following 10 genes were selected for validation of the Histologically, the vacuolation of hepatocytes, devel-

Vol. 33 No. 2 Table 1. List of primer sequences used in real-time RT-PCR analysis Accession no. Symbol Gene name Forward primer Reverse primer NM_022277 Casp8 Caspase 8 GCCAGGAGAGCAAGAGAGTGA AGACAGTACCCCCGAGGTTTG NM_053736 Casp4 Caspase 4 CGAGACAAACCCAAAGTCATCAT TCCTGAAGACTCTCTGATCCACACT NM_012923 Ccng1 Cyclin G1 GGAGAAGACGTGGCTGTCAA TCCCGGAGTCTTGCAGTC AA enhancedlivertumorpromotionactivityofKA inrats NM_031762 Cdkn1b Cyclin-dependent kinase AACAAAAGGGCCAACAGAACA GGGCGTCTGCTCCACAGT inhibitor 1B NM_053677 Chek2 CHK2 checkpoint homolog AAGAGACGAATACATCATGTCAAAAACT GGCCACTTTCTTACACGTTTTCC (S. pombe) NM_024134 Ddit3 DNA-damage inducible CAGCGACAGAGCCAAAATAACA GAGACTCAGCTGCCATGACTGTA transcript 3 NM_001008321 Gadd45b Growth arrest and ACTTCACCCTGATCCAATCGTT GCCTTTGCATGCCTGATACC DNA-damage-inducible 45 beta NM_012861 Mgmt O-6-methylguanine-DNA AGGAGCGATGAGGAGCAATC GGCACCGTCACTGCGAAT methyltransferase NM_057114 Prdx1 Peroxiredoxin 1 CACGGTTGGTTCTGTTTGTGA CCCAATTTTTGCATTTCCTGAA NM_021578 Tgfb1 Transforming growth factor beta-1 GAGGAGTGGGAGGAGGGACGAG GCGGGGGACTCAAGAAGCAG NM_012675 Tnf Tumor necrosis factor ACAAGGCTGCCCCGACTAT CTCCTGGTATGAAGTGGCAAATC (TNF superfamily, member 2) NM_030989 Tp53 Tumor protein p53 CATGAGCGTTGCTCTGATGGT GATTTCCTTCCACCCGGATAA XM_341671 Tradd TNFRSF1A-associated CCCCGACGTACTGCAGATACTC GCAGAAACGCAACTGAACGA via death domain NM_031020 Mapk14 Mitogen activated protein kinase 14 CGAAGATGAACTTCGCAAATGTA ATCCGAGTCCAAAACCAGCAT NM_053986 Myo1b Myosin Ib GCAGGAGAAAGTTTCAACCACAT AACCGGCTGTAGAGGTTTTTAGC Vol. 33No.2 131 132

M. Takabatake et al.

Table 2. Body and liver weights of animals Group DEN-alone DEN+KA DEN+KA+AA No. of animals 4 6 8 Body weight (g) 273.6 ± 16.8a 243.5 ± 18.1 247.5 ± 4.6* Liver weight Absolute (g) 6.63 ± 0.50 6.92 ± 0.33 7.09 ± 0.35 Relative (%) 2.42 ± 0.13 2.85 ± 0.16** 2.86 ± 0.12**

*P < 0.05, **P < 0.01 vs. DEN-alone. a Mean ± S.D. Abbreviations: DEN, diethylnitrosamine; KA, kojic acid; AA, ascorbic acid. opment of microgranuloma, and proliferation of small bile ducts were observed in all groups, but no specif- ic pathologic abnormalities were detected in the DEN + KA and DEN + KA + AA groups in comparison with the DEN-alone group. Immunohistochemically, the area of GST-P-positive foci was significantly higher in the DEN + KA group than in the DEN-alone group and in the DEN + KA + AA group than in the DEN-alone and DEN + KA groups (Fig. 2A). Although statistically non- VLJQL¿FDQWWKHQXPEHURI*673SRVLWLYHIRFLZDVKLJKHU in the DEN + KA group than in the DEN-alone group. On WKHRWKHUKDQGWKHQXPEHURIWKHVHIRFLZDVVLJQL¿FDQW- ly higher in the DEN +KA + AA group than in the DEN- alone and DEN + KA groups (Fig. 2B). The number of 3&1$SRVLWLYHOLYHUFHOOVZDVVLJQL¿FDQWO\KLJKHULQWKH DEN + KA + AA group than in the DEN-alone and DEN + KA groups (Fig. 3A). The number of TUNEL-positive OLYHUFHOOQXFOHLZDVVLJQL¿FDQWO\KLJKHULQWKH'(1.$ JURXSWKDQLQWKH'(1DORQHJURXSDQGVLJQL¿FDQWO\ORZ- er in the DEN + KA + AA group than in the DEN + KA group (Fig. 3B).

Lipid peroxidation level The TBARS levels were significantly higher in the DEN + KA and DEN + KA + AA groups than in the DEN- alone group and lower in the DEN + KA + AA group than in the DEN + KA group (Fig. 4).

Gene expression data Microarray analysis of the livers using one animal sam- ple in each group revealed that 32 genes were upregulated (>1.5-fold) and 8 genes were downregulated (< 0.67-fold) Fig. 2. The area (panel A) and number (panel B) of GST-P- in the DEN + KA group in comparison with the DEN- positive foci. Columns represent the mean ± S.D. of alone group (Table 3). In the DEN + KA + AA group, 2 the animals (DEN-alone group: n = 4; DEN + KA group: n = 6; and DEN + KA + AA group: n = 8). *, genes were upregulated (>1.5-fold) and 79 genes were **Significantly different from the DEN-alone group downregulated (<0.67-fold) in comparison with the DEN (*P < 0.05, **P < 0.01). #6LJQL¿FDQWO\GLIIHUHQWIURP + KA group (Table 4). Data for real-time RT-PCR are the DEN + KA group (P < 0.05).

Vol. 33 No. 2 133

AA enhanced liver tumor promotion activity of KA in rats

Fig 4. Lipid peroxidation level measured by TBARS forma- tion. Columns represent the mean ± S.D. of animals (DEN-alone group: n = 4; DEN + KA group: n = 5; DQG'(1.$$$JURXSQ    6LJQL¿FDQWO\ different from the DEN-alone group (*P < 0.05, **P < 0.01). #Significantly different from the DEN + KA group (P < 0.05).

DISCUSSION

In the present study, enhancement in the number and area of GST-P-positive foci was evident when AA was coadministered with KA in the rat hepatocarcinogenesis model. While KA alone did not apparently increase the PCNA-positive proliferating cells, concurrent adminis- Fig. 3. Number of PCNA (panel A) and TUNEL (panel B)-pos- tration of AA increased liver cell proliferation in this car- itive nuclei in hepatocytes. Columns represent the mean cinogenesis model. While the endpoint marker for early ± S.D. of the animals (DEN-alone group: n = 4; DEN + KA group: n = 5; and DEN + KA + AA group: n = 5). preneoplastic proliferation is different from that in rats, a  6LJQL¿FDQWO\GLIIHUHQWIURPWKH'(1DORQHJURXS similar enhancement by AA has already been shown in a (*P < 0.05, **P < 0.01). #6LJQL¿FDQWO\GLIIHUHQWIURP mouse hepatocarcinogenesis model (Kono et al., 2007). the DEN + KA group (P < 0.05). Apart from its well-known antioxidant activity, AA has been shown to function as a prooxidant (Podmore et al., shown in Fig. 5. The expression levels of Mgmt, Chek2, 1998). In this study, we measured TBARS levels to deter- DQG&DVSP51$VZHUHVLJQL¿FDQWO\KLJKHULQWKH'(1 mine whether AA has an antioxidant or prooxidant effect + KA group than in the DEN-alone group. In compari- on KA-induced tumor promotion. The results showed that son with the DEN + KA group, in the DEN + KA + AA in comparison with the administration of KA alone, con- JURXSWKHH[SUHVVLRQRI&FQJP51$ZDVVLJQL¿FDQW- current administration of AA with KA decreased the lip- ly higher, while that of Gadd45b, Chek2, Ddit3, Casp8, id peroxidation levels. Additionally, expression of the Casp11, Prdx1, Cdkn1b, and Tgfb1 mRNAs was signif- mRNA of Prdx1, an antioxidant enzyme (Wood et al., icantly lower. The transcript levels of Sod2, Tnf, Tp53,  ZDVVLJQL¿FDQWO\GHFUHDVHGE\FRDGPLQLVWUDWLRQRI and Tradd, all of them exhibiting decreases by microar- AA with KA, suggesting that AA might exert an antioxi- ray analysis in the DEN + KA + AA group in compari- dant effect in the present study. The presence of transi- VRQZLWKWKH'(1.$JURXSVKRZHGQRVLJQL¿FDQWGLI- tion metals may be important for the prooxidant activity ferences by real-time RT-PCR, although similar trends for of AA (Samuni et al., 1983). Since KA is known to be a decreases in the microarray analysis were found in these metal ion chelator (Katoh et al., 1992), it can be hypothe- genes. sized that the administered KA would trap metal ions, and this would allow AA to exert its antioxidant effects in the present study.

Vol. 33 No. 2 134

M. Takabatake et al.

Oxidative stress induced by reactive oxygen species dicyclanil shown earlier in our laboratory (Muguruma et (ROS) generated in the metabolic pathway of chemicals al., 2007; Moto et al., 2006b). KA in the present study is suggested to be involved in chemical-induced hepato- increased the lipid peroxidation level as with the tumor- carcinogenesis, as in the case of piperonyl butoxide and promoting activity. On the other hand, coadministra-

Table 3. List of genes showing altered expressions in a low-density microarray (DEN+KA/DEN) Relative expression Accession no. Gene symbol Description (DEN+KA/DEN) Up-regulated genes (>1.5-fold ) Apoptosis NM_030826 Gpx1 Glutathione peroxidase 1 1.81 Cell cycle or cell growth and proliferation NM_080782 Cdkn1a Cyclin-dependent kinase inhibitor 1A 1.99 NM_031051 Mif Macrophage migration inhibitory factor 1.95 XM_343065 Nfkbia Nuclear factor of kappa light chain gene 1.77 enhancer in B-cells inhibitor, alpha NM_053677 Chek2 CHK2 checkpoint homolog (S. pombe) 1.56 Chaperon NM_022936 Ephx2 Epoxide hydrolase 2, cytoplasmic 1.66 NM_032079 Dnaja2 DnaJ (Hsp40) homolog, subfamily A, member 2 1.54 Drug metabolizing Enzymes NM_031509 Gsta3 Glutathione S-transferase A3 2.33 NM_012792 Fmo1 Flavin containing monooxygenase 1 2.31 NM_017013 Gsta2 Glutathione-S-transferase, alpha type 2 2.29 NM_031971 Hspa1a Heat shock 70kD protein 1A 2.21 XM_217195 Gsta4 Glutathione S-transferase, alpha 4 2.16 NM_017014 Gstm1 Glutathione S-transferase, mu 1 2.01 NM_019303 Cyp2f2 Cytochrome P450, family 2, subfamily f, polypeptide 2 2.00 NM_012844 Ephx1 Epoxide hydrolase 1, microsomal 1.97 NM_138515 Cyp2d22 Cytochrome P450, family 2, subfamily d, polypeptide 22 1.97 NM_177426 Gstm2 Glutathione S-transferase, mu 2 1.78 NM_031576 Por P450 (cytochrome) oxidoreductase 1.75 NM_138877 Cyb5r3 Cytochrome b5 reductase 3 1.74 NM_134349 Mgst1 Microsomal glutathione S-transferase 1 1.68 NM_013198 Maob Monoamine oxidase B 1.66 NM_001004082 Hspcb Heat shock 90kDa protein 1, beta 1.65 NM_017156 Cyp2b15 Cytochrome P450, family 2, subfamily b, 1.60 polypeptide 15 NM_001004086 Pon3 Paraoxonase 3 1.60 NM_022513 Sult1b1 Sulfotransferase family 1B, member 1 1.58 NM_031027 Dpyd Dihydropyrimidine dehydrogenase 1.57 NM_031565 Ces1 Carboxylesterase 1 1.56 NM_031154 Gstm3 Glutathione S-transferase, mu type 3 1.54 NM_031834 Sult1a1 Sulfotransferase family 1A, phenol-preferring, member 1 1.53 NM_175837 Cyp4a22 Cytochrome P450, family 4, subfamily A, polypeptide 22 1.51 Transcription factors and regulators NM_022941 Nr1i3 Nuclear receptor subfamily 1, group I, member 3 1.93 XM_223843 Rarb Retinoic acid receptor, beta 1.72

Vol. 33 No. 2 135

AA enhanced liver tumor promotion activity of KA in rats tion of AA enhanced KA-induced tumor promotion and gest that apoptosis and cell cycle arrest were suppressed, exerted an antioxidant effect. These results suggest that and cell proliferation was accelerated when AA was con- KA-induced oxidative stress may not be the mechanism currently administered during KA-induced tumor promo- responsible for the tumor-promoting activity. Increased tion. liver cell apoptosis, as observed here by KA treatment, In the present study, tumor-promoted liver by KA might be the outcome of the oxidative stress response, exhibited an increase of lipid peroxidation level as well as discussed later. Since accumulation of cholesterin-like as an increase in the number of TUNEL-positive apop- FU\VWDOVDQGGHJHQHUDWLYHDQGLQÀDPPDWRU\FKDQJHVZHUH totic liver cells. It is widely known that ROS possesses reported in rats fed 2% KA in their diet (Takizawa et al., “second messenger” characteristics that link these mol- 2004), it is likely that the cytotoxicity induced by these ecules to a wide variety of biological activities includ- changes can be the primary reason for the tumor promo- ing apoptosis (Nathan, 2003). It is reported that cellular tion activity. exposure to ROS can directly activate apoptosis signaling We analyzed the results of mRNA expression obtained (Slater et al., 1995). On the other hand, coadministra- from low-density microarrays by real-time RT-PCR tion of AA in the present study reduced the lipid peroxi- focusing on the genes related to apoptosis, cell cycle, dation level and apoptosis. It is reported that programmed and growth/proliferation that may possibly be involved cell death triggered by a variety of stimuli can be inhib- in the enlargement of preneoplastic foci. We observed ited by AA (D’Agostini et al., 2005b). For instance, AA that genes related to apoptosis (Casp8, Casp4, Ddit3, inhibited TRAIL-induced apoptosis in cancer cell lines by and Tgfb1: Weishaupt et al., 2003; Hur et al., 2001; suppressing caspase 8 activity (Perez-Cruz et al., 2007). Lovat et al., 2002; Herrera et al., 2004), cell cycle arrest In this study, we also observed a reduction in the Casp8 (Cdkn1b, Chek2, and Gadd45b: Polyak et al., 1994; mRNA level due to coadministration of AA. Further- Bartek et al., 2003; Higashi et al., 2006), DNA repair more, several lines of evidence indicate that transforming

(Mgmt: Sekiguchi et al., 1996), and antioxidant enzymes JURZWKIDFWRUȕ1 7*)ȕ1) induces ROS production that (Prdx1: Wood et al., 2003) were downregulated in the activates apoptosis (Franklin et al., 2003). In rat preneo- DEN + KA + AA group in comparison with the DEN + plastic hepatocytes, cytokine-induced ROS generation can

KA group. Chek2 and Gadd45b are also involved in apop- WULJJHUKHSDWRF\WLF7*)ȕ1 induction leading to reduction tosis (Bartek et al., 2003; Higashi et al., 2006). Only one of the antioxidant defenses and induction of apoptosis of the genes examined here, i.e., Ccng1, which is related (Quiroga et al., 2007). Additionally, AA and anti-TGF- to cell proliferation (Tamura et al., 1993), was found to be ȕ1 antibody can reduce the number of apoptotic cells and upregulated by AA coadministration. In these animals, a ROS production induced by cytokines (Quiroga et al., marked decrease and increase were observed for TUNEL 2007). We also observed a reduction in the mRNA lev- and PCNA-positive cells, respectively. These results sug- el of Tgfb1 due to coadministration of AA and also not-

Table 3. Continued. Relative expression Accession no. Gene symbol Description (DEN+KA/DEN) Down-regulated genes (<0.67-fold ) Apoptosis XM_341671 Tradd TNFRSF1A-associated via death domain 0.66 Cell cycle or cell growth and proliferation NM_171992 Ccnd1 Cyclin D1 0.66 NM_017019 Il1a Interleukin 1 alpha 0.58 NM_031512 Il1b Interleukin 1 beta 0.43 Drug metabolizing Enzymes NM_031641 Sult4a1 Sulfotransferase family 4A, member 1 0.66 NM_012687 Tbxas1 Thromboxane A synthase 1 0.55 Drug related transporters NM_031013 Abcc6 ATP-binding cassette, sub-family C (CFTR/MRP), member 6 0.44 NM_012833 Abcc2 ATP-binding cassette, sub-family C (CFTR/MRP), member 2 0.38

Vol. 33 No. 2 136

M. Takabatake et al. ed a decrease in the levels of other genes related to apop- tion of cell proliferation-related Ccng1 because it is well tosis and/or cell cycle arrest. It is known that ROS induce known that Cdkn1b, also known as p27, negatively reg- cell cycle arrest by upregulating Cdkn1b (Wolf, 2005). In ulates other cyclins (Polyak et al., 1994). These results addition, downregulation of Cdkn1b may cause upregula- indicate that KA-induced ROS generation might be

Table 4. List of genes showing altered expressions in a low-density microarray (DEN+KA+AA/DEN+KA) Relative expression Accession no. Gene symbol Description (DEN+KA+AA/ DEN+KA) Up-regulated genes (>1.5-fold ) Transcription factors and regulators NM_012551 Egr1 Early growth response 1 2.40 Drug related transporters NM_012833 Abcc2 ATP-binding cassette, sub-family C (CFTR/MRP), member 2 1.92 Down-regulated genes (<0.67-fold ) Apoptosis NM_022277 Casp8 Caspase 8 0.55 NM_057114 Prdx1 Peroxiredoxin 1 0.55 NM_145681 Tnfsf10 Tumor necrosis factor (ligand) superfamily, member 10 0.56 NM_030826 Gpx1 Glutathione peroxidase 1 0.56 NM_053021 Clu Clusterin 0.56 NM_001008321 Gadd45b Growth arrest and DNA-damage-inducible 45 beta 0.57 NM_012675 Tnf Tumor necrosis factor (TNF superfamily, member 2) 0.58 NM_030989 Tp53 Tumor protein p53 0.59 NM_024134 Ddit3 DNA-damage inducible transcript 3 0.59 NM_012861 Mgmt O-6-methylguanine-DNA methyltransferase 0.60 XM_341671 Tradd TNFRSF1A-associated via death domain 0.66 NM_017169 Prdx2 Peroxiredoxin 2 0.66 Cell cycle or cell growth and proliferation XM_343065 Nfkbia Nuclear factor of kappa light chain gene enhancer in 0.42 B-cells inhibitor, alpha NM_031051 Mif Macrophage migration inhibitory factor 0.54 XM_235169 Mdm2 Transformed mouse 3T3 cell double minute 2 0.55 NM_031762 Cdkn1b Cyclin-dependent kinase inhibitor 1B 0.55 NM_012756 Igf2r Insulin-like growth factor 2 receptor 0.62 NM_053677 Chek2 CHK2 checkpoint homolog (S. pombe) 0.66 Chaperon NM_031970 Hspb1 Heat shock 27kDa protein 1 0.65 NM_022934 Dnaja1 DnaJ (Hsp40) homolog, subfamily A, member 1 0.63 XM_214583 Hspa9a_predicted Heat shock 70kDa protein 9A (predicted) 0.59 NM_031122 St13 Suppression of tumorigenicity 13 0.59 NM_001004082 Hspcb Heat shock 90kDa protein 1, beta 0.52 NM_012966 Hspe1 Heat shock 10 kDa protein 1 (chaperonin 10) 0.49 NM_175761 Hspca Heat shock protein 1, alpha 0.48 NM_032079 Dnaja2 DnaJ (Hsp40) homolog, subfamily A, member 2 0.48 NM_022229 Hspd1 Heat shock protein 1 (chaperonin) 0.48 NM_031971 Hspa1a Heat shock 70kD protein 1A 0.19 Drug metabolizing Enzymes NM_022715 Mvp Major vault protein 0.31 NM_017319 Pdia3 3URWHLQGLVXO¿GHLVRPHUDVHDVVRFLDWHG 0.39 NM_138867 Hyou1 Hypoxia up-regulated 1 0.42 NM_134349 Mgst1 Microsomal glutathione S-transferase 1 0.43 NM_022936 Ephx2 Epoxide hydrolase 2, cytoplasmic 0.43 NM_031834 Sult1a1 Sulfotransferase family 1A, phenol-preferring, member 1 0.43 NM_012883 Ste Sulfotransferase, estrogen preferring 0.45 NM_175837 Cyp4a22 Cytochrome P450, family 4, subfamily A, polypeptide 22 0.45 NM_017050 Sod1 Superoxide dismutase 1 0.46

Vol. 33 No. 2 137

AA enhanced liver tumor promotion activity of KA in rats responsible for the increase in liver cell apoptosis and cell ble for the enhancement of KA-induced tumor-promotion cycle arrest that can be blocked by coadministration of by AA in the present study. AA through suppression of Tgfb1-signaling and a result- The enhancement of KA-induced tumor promotion by ing reduction in ROS generation. This phenomenon along AA was rather strong, but interestingly, AA coadministra- with increased cell proliferation activity may be responsi- tion did not result in any accompanying histopathological

Table 4. Continued. Relative expression Accession no. Gene symbol Description (DEN+KA+AA/ DEN+KA) NM_019184 Cyp2c Cytochrome P450, subfamily IIC (mephenytoin 4-hydroxylase) 0.46 NM_138515 Cyp2d22 Cytochrome P450, family 2, subfamily d, polypeptide 22 0.47 NM_138514 Cyp2c13 Cytochrome P450 2c13 0.48 NM_017158 Cyp2c7 Cytochrome P450, family 2, subfamily c, polypeptide 7 0.48 NM_012844 Ephx1 Epoxide hydrolase 1, microsomal 0.49 NM_017156 Cyp2b15 Cytochrome P450, family 2, subfamily b, polypeptide 15 0.50 NM_031241 Cyp8b1 Cytochrome P450, family 8, subfamily b, polypeptide 1 0.51 NM_001004086 Pon3 Paraoxonase 3 0.51 NM_022513 Sult1b1 Sulfotransferase family 1B, member 1 0.51 NM_177426 Gstm2 Glutathione S-transferase, mu 2 0.52 NM_031154 Gstm3 Glutathione S-transferase, mu type 3 0.52 NM_138512 Cyp2c70 Cytochrome P450, family 2, subfamily c, polypeptide 70 0.53 NM_017051 Sod2 Superoxide dismutase 2, mitochondrial 0.53 NM_031543 Cyp2e1 Cytochrome P450, family 2, subfamily e, polypeptide 1 0.54 NM_017014 Gstm1 Glutathione S-transferase, mu 1 0.54 NM_012520 Cat Catalase 0.54 NM_024387 Hmox2 Heme oxygenase (decycling) 2 0.54 NM_031732 Sult1c1 Sulfotransferase family, cytosolic, 1C, member 1 0.55 NM_031641 Sult4a1 Sulfotransferase family 4A, member 1 0.55 NM_013198 Maob Monoamine oxidase B 0.55 NM_133547 Sult1c2 Sulfotransferase family, cytosolic, 1C, member 2 0.57 NM_016999 Cyp4b1 Cytochrome P450, family 4, subfamily b, polypeptide 1 0.57 NM_057105 Ugt1a6 UDP glycosyltransferase 1 family, polypeptide A6 0.57 NM_138877 Cyb5r3 Cytochrome b5 reductase 3 0.58 NM_019623 Cyp4f2 Cytochrome P450, family 4, subfamily F, polypeptide 2 0.58 NM_012792 Fmo1 Flavin containing monooxygenase 1 0.60 XM_343764 Maoa Monoamine oxidase A 0.61 NM_012531 Comt Catechol-O-methyltransferase 0.62 XM_217138 Nnmt_predicted Nicotinamide N-methyltransferase (predicted) 0.63 NM_173093 Cyp2d13 Cytochrome P450, family 2, subfamily d, polypeptide 13 0.63 NM_019303 Cyp2f2 Cytochrome P450, family 2, subfamily f, polypeptide 2 0.63 NM_031576 Por P450 (cytochrome) oxidoreductase 0.63 NM_031509 Gsta3 Glutathione S-transferase A3 0.64 NM_013083 Hspa5 Heat shock 70kDa protein 5 (glucose-regulated protein) 0.64 NM_017013 Gsta2 Glutathione-S-transferase, alpha type2 0.64 DNA repair genes NM_001013190 Rad23a RAD23a homolog (S. cerevisiae) 0.54 NM_001031644 Ercc3 ([FLVLRQUHSDLUFURVVFRPSOHPHQWLQJURGHQWUHSDLUGH¿FLHQF\ 0.65 complementation group 3 Drug related transporters NM_022399 Calr Calreticulin 0.58 Transcription factors and regulators NM_052980 Nr1i2 Nuclear receptor subfamily 1, group I, member 2 0.51 NM_031528 Rara Retinoic acid receptor, alpha 0.53 NM_022941 Nr1i3 Nuclear receptor subfamily 1, group I, member 3 0.60 XM_223843 Rarb Retinoic acid receptor, beta 0.65 Abbreviations: DEN, diethylnitrosamine; KA, kojic acid; AA, ascorbic acid.

Vol. 33 No. 2 138

M. Takabatake et al.

Fig. 5. Relative levels of mRNA expression determined by real-time RT-PCR. Expression values were normalized based on the levels of myosin Ib. The values are expressed as group mean (n = 3) fold change over the '(1DORQHJURXS 6LJQL¿FDQWO\GLIIHUHQWIURPWKH'(1DORQHJURXS (P < 0.05). #, ##6LJQL¿FDQWO\GLIIHUHQWIURPWKH'(1.$JURXS #P < 0.05, ##P < 0.01). alterations and had no effect on the liver weight. In liv- increase in the number of glucose transporters was report- er cells, AA is transported in an oxidized form via sev- ed in dimethylaminoazobenzene-induced rat hepatocellu- eral glucose transporters (Bánhegyi et al., 1998). It has lar tumors (Kim et al., 1990), suggesting the possibility been reported that cancer cells show higher intracellular that the capacity of glucose transport is higher in preneo- AA levels than normal tissues (Liebes et al., 1981) prob- plastic liver cell foci than in the surrounding normal cells ably due to the higher capacity of cancer cells to uptake/ and that this may lead to the selective uptake of adminis- transport glucose (Dang et al., 1999). Additionally, an tered AA in the foci.

Vol. 33 No. 2 139

AA enhanced liver tumor promotion activity of KA in rats

In conclusion, data obtained in the present study Fujimoto, N., Onodera, H., Mitsumori, K., Tamura, T., Maruyama, revealed that KA-induced liver tumor promoting activi- S. and Ito, A. (1999) : Changes in thyroid function during devel- opment of thyroid hyperplasia induced by kojic acid in F344 ty was enhanced by concurrent administration of AA in rats. Carcinogenesis, 20, 1567-1571. rats, as observed in our previous study using mice. Fur- Herrera, B., Murillo, M.M., Alvarez-Barrientos, A., Beltràn, J., thermore, our results indicate that the concerted effects Fernàndez, M. and Fabregat, I. (2004) : Source of early reactive of AA on cell proliferation and apoptosis/cell cycle oxygen species in the apoptosis induced by transforming growth arrest, possibly enhanced by a common signaling path- IDFWRUȕLQIHWDOUDWKHSDWRF\WHV)UHH5DGLF%LRO0HG36, 16- 26. way, was involved in this enhancement. The mechanism Higa, Y., Kawabe, M., Nabae, K., Toda, Y., Kitamoto, S., Hara, T., may involve an antioxidant activity although the mecha- Tanaka, N., Kariya, K. and Takahashi, M. (2007) : Kojic acid - nism by which the liver tumor-promoting activity of KA absence of tumor-initiating activity in rat liver, and of carcino- is associated with the generated ROS remains unclear. genic and photo-genotoxic potential in mouse skin. J. Toxicol. Further studies are required to address the association of Sci., 32, 143-159. Higashi, H., Vallböhmer, D., Warnecke-Eberz, U., Hokita, S., Xi, H., the key mechanism of this enhancement with antioxidant Brabender, J., Metzger, R., Baldus, S.E., Natsugoe, S., Aikou, T., activity as well as the putative promotion mechanism of Hölscher, A.H. and Schneider, P.M. (2006) : Down-regulation of KA. Gadd45 expression is associated with tumor differentiation in non-small cell lung cancer. Anticancer Res., 26, 2143-2147. REFERENCE Hur, J., Kim, S.Y., Kim, H., Cha, S., Lee, M.S. and Suk, K. (2001) ,QGXFWLRQRIFDVSDVHE\LQÀDPPDWRU\VWLPXOLLQUDWDVWUR- cytes: lipopolysaccharide induction through p38 mitogen-acti- Bánhegyi, G., Marcolongo, P., Puskás, F., Fulceri, R., Mandl, J. and vated protein kinase pathway. FEBS Lett., 507, 157-162. Benedetti, A. (1998) : Dehydroascorbate and ascorbate transport Ito, N., Imaida, K., Asamoto, M. and Shirai, T. (2000) : Early detec- in rat liver microsomal vesicles. J. Biol. Chem., 273, 2758-2762. tion of carcinogenic substances and modifiers in rats. Mutat. Bartek, J. and Lukas, J. (2003) : Chk1 and Chk2 kinases in check- Res., 462, 209-217. point control and cancer. Cancer Cell, 3, 421-429. Iwata, H., Yamamoto, S., Yano, Y., Ohtani, S. and Fukushima, S. Cai, Q., Ferraris, J.D. and Burg, M.B. (2004) : Greater tolerance of (1997) : Dose-dependent amplification by L-ascorbic acid of renal medullary cells for a slow increase in osmolality is associ- NaHCO3 promotion of rat urinary bladder carcinogenesis. Tox- ated with enhanced expression of HSP70 and other osmoprotec- icol. Pathol., 25, 284-290. tive genes. Am. J. Physiol. Renal Physiol., 286, 58-67. Katoh, S., Toyama, J., Kodama, I., Kamiya, K., Akita, T. and Abe, Cameron, E., Pauling, L. and Leibovitz, B. (1979) : Ascorbic acid T. (1992) : Protective action of iron-chelating agents (catechol, and cancer: a review. Cancer Res., 39, 663-681. mimosine, deferoxamine, and kojic acid) against ischemia-reper- Chen, J.S., Wei, C.I., Rolle, R.S., Otwell, W.S., Balaban, M.O. and fusion injury of isolated neonatal rabbit hearts. Eur. Surg. Res., Marshall, M.R. (1991) : Inhibitory effect of kojic acid on some 24, 349-355. plant and crustacean polyphenol oxidase. J. Agr. Food Chem., Kim, Y.S., Paik, E.M., Lee, M.S., Kim, K.S., Chai, K.J. and Ahn, 39, 1396-1401. Y.H. (1990) : A study on the regulation of translocation of glu- D’Agostini, F., Balansky, R.M., Camoirano, A. and De Flora, S. cose transporters during hepatocarcinogenesis induced by 3’-Me (2005a) : Modulation of light-induced skin tumors by N-acetyl- DAB. Yonsei Med. J., 31, 315-324. cysteine and/or ascorbic acid in hairless mice. Carcinogenesis, Kono, T., Moto, M., Muguruma, M., Takahashi, M., Jin, M., 26, 657-664. Kenmochi, Y., Yokouchi, Y. and Mitsumori, K. (2007) : Enhance- D’Agostini, F., Izzotti, A., Balansky, R.M., Bennicelli, C. and De ment of hepatocellular proliferative activity of kojic acid in mice Flora, S. (2005b) : Modulation of apoptosis by cancer chemo- by a simultaneous administration of ascorbic acid. J. Vet. Sci., preventive agents. Mutat. Res., 591, 173-186. 69, 899-908. Dang, C.V. and Semenza, G.L. (1999) : Oncogenic alterations of Liebes, L., Krigel, R., Kuo, S., Nevrla, D., Pelle, E. and Silber, R. metabolism. Trends Biochem. Sci., 24, 68-72. (1981) : Increased ascorbic acid content in chronic lymphocytic Feroze, M.F., Berquin, I.M., Dey, J. and Chen, Y.Q. (2002) : Pepti- leukemia B lymphocytes. Proc. Natl. Acad. Sci. USA, 78, 6481- dylprolyl isomerase A (PPIA) as a preferred internal control over 6484. GAPDH and beta-actin in quantitative RNA analyses. Biotech- Livak, K.J. and Schmittgen, T.D. (2001) : Analysis of relative gene niques, 32, 776-778, 780, 782. expression data using real-time quantitative PCR and the 2íǻǻ&T Franklin, C.C., Rosenfeld-Franklin, M.E., White, C., Kavanagh, method. Methods, 25, 402-408. 7-DQG)DXVWR1  7*)ȕ1-induced suppression of glu- Lovat, P.E., Oliverio, S., Ranalli, M., Corazzari, M., Rodolfo, C., tathione antioxidant defenses in hepatocytes: caspase-dependent Bernassola, F., Aughton, K., Maccarrone, M., Hewson, Q.D., post-translational and caspase-independent transcriptional regu- Pearson, A.D., Melino, G., Piacentini, M. and Redfern, C.P. latory mechanisms. FASEB J., 17, 1535-1537. (2002) : GADD153 and 12-lipoxygenase mediate fenretinide- Frei, B., England, L. and Ames, B.N. (1989) : Ascorbate is an out- induced apoptosis of neuroblastoma. Cancer Res., 62, 5158- standing antioxidant in human blood plasma. Proc. Natl. Acad. 5167. Sci. USA, 86, 6377-6381. Maeda, K. and Fukuda, M. (1991) : In vitro effectiveness of several Fujimoto, N., Watanabe, H., Nakatani, T., Roy, G. and Ito, A. (1998) whitening cosmetic components in human melanocytes. J. Soc. : Induction of thyroid tumors in (C57BL/6N×C3H/N) F1 mice by Cosmet. Chem., 42, 361-368. oral administration of kojic acid. Food Chem. Toxicol., 36, 697- Mitsumori, K., Onodera, H., Takahashi, M., Funakoshi, T., Tamura, 703. T., Yasuhara, K., Takegawa, K. and Takahashi, M. (1999) : Pro-

Vol. 33 No. 2 140

M. Takabatake et al.

moting effects of kojic acid due to serum TSH elevation result- Pharmacol., 73, 1776-1785. ing from reduced serum thyroid hormone levels on development Samuni, A., Aronovitch, J., Godinger, D., Chevion, M. and Czapski, of thyroid proliferative lesions in rats initiated with N-bis(2-hydr G. (1983) : On the cytotoxicity of vitamin C and metal ions. A oxypropyl)nitrosamine. Carcinogenesis, 20, 173-176. VLWHVSHFL¿F)HQWRQPHFKDQLVP(XU-%LRFKHP137, 119-124. Moto, M., Mori, T., Okamura, M., Kashida, Y. and Mitsumori, K. Sekiguchi, M., Nakabeppu, Y., Sakumi, K. and Tuzuki, T. (1996) (2006a). Absence of liver tumor-initiating activity of kojic acid : DNA-repair methyltransferase as a molecular device for pre- in mice. Arch. Toxicol., 80, 299-304. venting mutation and cancer. J. Cancer Res. Clin. Oncol., 122, Moto, M., Umemura, T., Okamura, M., Muguruma, M., Ito, T., Jin, 199-206. M., Kashida, Y. and Mitsumori, K. (2006b) : Possible involve- Shirai, T. (1997) : A medium-term rat liver bioassay as a rapid in ment of oxidative stress in dicyclanil-induced hepatocarcinogen- vivo test for carcinogenic potential, a historical review of mod- esis in mice. Arch. Toxicol., 80, 694-702. el development and summary of results from 291 tests. Toxicol. Muguruma, M., Unami, A., Kanki, M., Kuroiwa, Y., Nishimura, J., Pathol., 25, 453-460. Dewa, Y., Umemura, T., Oishi, Y. and Mitsumori, K. (2007) : Slater, A.F., Nobel, C.S. and Orrenius, S. (1995) : The role of intrac- Possible involvement of oxidative stress in piperonyl butoxide ellular oxidants in apoptosis. Biochim. Biophys. Acta., 1271, 59- induced hepatocarcinogenesis in rats. Toxicology, 236, 61-75. 62. 1DWKDQ&  6SHFL¿FLW\RIDWKLUGNLQGUHDFWLYHR[\JHQDQG Takizawa, T., Imai, T., Onose, J., Ueda, M., Tamura, T., Mitsumori, nitrogen intermediates in cell signaling. J. Clin. Invest., 111, K., Izumi, K. and Hirose, M. (2004) : Enhancement of hepato- 769-778. carcinogenesis by kojic acid in rat two-stage models after initia- Nohynek, G.J., Kirkland, D., Marzin, D., Toutain, H., Leclerc-Ribaud, tion with N-bis(2-hydroxypropyl)nitrosamine or N-diethylnitro- C. and Jinnai, H. (2004) : An assessment of the genotoxicity samine. Toxicol. Sci., 81, 43-49. and human health risk of topical use of kojic acid (5-hydroxy- Takizawa, T., Mitsumori, K., Tamura, T., Nasu, M., Ueda, M., Imai, 2-(hydroxymethyl)-4H-pyran-4-one). Food Chem. Toxicol., 42, T. and Hirose, M. (2003) : Hepatocellular tumor induction in het- 93-105. HUR]\JRXVSGH¿FLHQW&%$PLFHE\DZHHNGLHWDU\DGPLQ- Okazaki, K., Ishii, Y., Kitamura, Y., Maruyama, S., Umemura, T., istration of kojic acid. Toxicol. Sci., 73, 287-293. Miyauchi, M., Yamagishi, M., Imazawa, T., Nishikawa, A., Tamura, K., Kanaoka, Y., Jinno, S., Nagata, A., Ogiso, Y., Shimizu, Yoshimura, Y., Nakazawa, H. and Hirose, M. (2006) : Dose- K., Hayakawa, T., Nojima, H. and Okayama, H. (1993) : Cyclin dependent promotion of rat forestomach carcinogenesis by com- G: a new mammalian cyclin with homology to fission yeast bined treatment with sodium nitrite and ascorbic acid after ini- Cig1. Oncogene, 8, 2113-2118. tiation with N-methyl-N’-nitro-N-nitrosoguanidine: Possible Tamura,T., Mitsumori, K., Onodera, H., Fujimoto, N., Yasuhara, K., contribution of nitric oxide-associated oxidative DNA damage. Takegawa, K. and Takahashi, M. (1999) : Inhibition of thyroid Cancer Sci., 97, 175-182. LRGLQHXSWDNHDQGRUJDQL¿FDWLRQLQUDWVWUHDWHGZLWKNRMLFDFLG Parrish, F.W., Wiley, B.J., Simmons, E.G. and Long Jr., L. (1966) : Toxicol. Sci., 47, 170-175. 3URGXFWLRQRIDÀDWR[LQVDQGNRMLFDFLGE\VSHFLHVRI$VSHUJLOOXV Watanabe, T., Mori, T., Okamura, M., Kashida, Y. and Mitsumori, and Penicillium. Appl. Microbiol., 14, 139. K. (2005) : Induction of hepatocellular proliferative lesions in Perez-Cruz, I., Cárcamo, J.M. and Golde, D.W. (2007) : Caspase-8 CBA mice by a 26-week dietary administration of kojic acid. J. dependent TRAIL-induced apoptosis in cancer cell lines is inhib- Toxicol. Pathol., 18, 159-165. ited by vitamin C and catalase. Apoptosis, 12, 225-234. Wei, C.I., Huang, T.S., Fernando, S.Y. and Chung, K.T. (1991) : 3RGPRUH,'*ULI¿WKV+5+HUEHUW.(0LVWU\10LVWU\3 Mutagenicity studies of kojic acid. Toxicol. Lett., 59, 213-220. and Lunec, J. (1998) : Vitamin C exhibits pro-oxidant properties. Weishaupt, J.H., Diem, R., Kermer, P., Krajewski, S., Reed, J.C. Nature, 392, 559. and Bähr, M. (2003) : Contribution of caspase-8 to apoptosis of Polyak, K., Lee, M.H., Erdjument-Bromage, H., Koff, A., Roberts, axotomized rat retinal ganglion cells in vivo. Neurobiol. Dis., 13, J.M., Tempst, P. and Massagué, J. (1994) : Cloning of p27Kip1, 124-135. a cyclin-dependent kinase inhibitor and a potential mediator of Wolf, G. (2005) : Role of reactive oxygen species in angiotensin II- extracellular antimitogenic signals. Cell, 78, 59-66. mediated renal growth, differentiation, and apoptosis. Antioxid. Quiroga, A.D., Alvarez, M.D., Parody, J.P., Ronco, M.T., Francés, Redox Signal., 7, 1337-1345. D.E., Pisani, G.B., Carnovale, C.E. and Carrillo, M.C. (2007) : Wood, Z.A., Poole, L.B. and Karplus, P.A. (2003) : Peroxiredox- Involvement of reactive oxygen species on the apoptotic mech- in evolution and the regulation of hydrogen peroxide signaling. DQLVPLQGXFHGE\,)1ĮELQUDWSUHQHRSODVWLFOLYHU%LRFKHP Science, 300, 650-653.

Vol. 33 No. 2