Genes and Environment, Vol. 36, No. 2 pp. 54–64 (2014)

Regular article Time-course Comparison of Expression Proˆles Induced by the Genotoxic Hepatocarcinogen, Chrysene, in the Mouse Liver

Mikiya Sakurai1, Takashi Watanabe1, Takayoshi Suzuki2 and Chie Furihata1,2,3 1Functional Genomics Laboratory, School of Science and Engineering, Aoyama Gakuin University, Sagamihara, Kana- gawa, Japan 2Division of Cellular and Gene Therapy Products, National Institute of Health Sciences, Tokyo, Japan

Received January 14, 2014; Revised March 5, 2014; Accepted March 18, 2014 J-STAGE Advance published date: April 1, 2014

Changes in proˆle in rodent liver at the Introduction acute stage within 48 h after administration of a Kerns and Bushel wrote that the era of toxicogenom- hepatocarcinogen have not been extensively reported. In ics has introduced a new way of monitoring the eŠect of the present study we examined changes in gene expres- environmental stressors and toxicants on biological sion in mouse liver within 48 h induced by chrysene, a systems through the quantiˆcation of changes in gene polycyclic aromatic hydrocarbon and genotoxic hepatocar- expression. Since the liver is one of the major organs for cinogen, by quantitative real-time PCR (qPCR). We quanti- the synthesis and secretion of substances that metabo- ˆed 50 candidate which discriminated genotoxic hepatocarcinogens from non-genotoxic hepatocarcino- lize endogenous and exogenous materials, much gens as determined from our previous DNA microarray stu- research eŠort has aimed to elucidate the predictive and dies. Chrysene (100 mg/kg bw) was injected in- mechanistic genomic markers of hepatotoxicity (1). traperitoneally into male 9-week-old B6C3F1 mice, and at Although DNA microarray and qPCR are often used 4, 16, 20, 24 and 48 h after chrysene administration, for this purpose (2), qPCR is the method of choice for livers were dissected and processed for gene expression. the speciˆc and sensitive quantiˆcation of nucleic acid A total of 35 genes exhibited statistically signiˆcant in- expression (3). creases at least once within 48 h after chrysene adminis- According to the IARC (International Agency for tration. Cyp1a1 and Cyp1a2 showed remarkably consis- Research on ) carcinogen classiˆcation, chrysene tent increases in gene expression during 4 to 48 h. Fifteen (CASRN 218-01-9), a polycyclic aromatic hydrocarbon genes (Bhlhe40, Btg2, Casp4, Ccng2, Cdkn1a, Crp, (PAH), is a 2B class chemical (possibly carcinogenic to Cyp1a1, Cyp1a2, Fkbp5, Gadd45b, Gadd45g, Hmox1, Igfbp1, Lcn2 and Ly6a) at 4 h, 6 genes at 16 h, 7 genes at humans), i.e., limited evidence of carcinogenicity in 20 h, 7 genes at 24 h, and 10 genes (Bhlhe40, Ccnf, humans and less than su‹cient evidence of carcinogen- Cyp1a1, Cyp1a2, Ephx1, Hhex, Hmox1, Rcan1, Tubb2a icity in experimental animals (4). According to the and Tubb4b) at 48 h exhibited statistically signiˆcant in- IARC monograph, human exposure to chrysene occurs creases of more than two-fold. At 4 h, 10 of 15 expres- primarily through the smoking of tobacco, inhalation of sion-increased genes were associated with DNA damage, polluted air and by ingestion of food and water con- DNA repair, cell cycle, cell proliferation and apoptosis. The taminated by combustion eŒuent. Chrysene was tested expression-increased genes at 16 to 48 h were associated for carcinogenicity in several studies by skin application with a variety of biological processes. In conclusion three to mice and produced skin tumors. Local tumors were time-dependent patterns in gene expression were ob- observed following its subcutaneous injection in mice. served within 48 h after chrysene administration in mouse Perinatal administration of chrysene to mice by sub- liver: Cyp1a1 and Cyp1a2 exhibited consistent increases; the highest number of genes (15 genes) increased in ex- cutaneous or intraperitoneal injection increased the pression at 4 h; and 6 diŠerent genes expressed at 4 h in- incidence of liver tumors. Chrysene was mutagenic to creased at 48 h. 3Correspondence to: Chie Furihata, Division of Cellular and Gene Key words: chrysene, gene expression proˆle, mouse Therapy Products, National Institute of Health Sciences, Kamiyoga liver, quantitative real-time PCR 1-18-1, Setagaya-ku, Tokyo 158-8501, Japan. Tel: ＋81-3-3700-1926, Fax: ＋81-3-3700-1926, E-mail: chiefurihata＠gmail.com This work was the ˆfth collaborative study of Toxicogenomics/ JEMS･MMS. doi: org/10.3123/jemsge.2014.005

54  The Japanese Environmental Mutagen Society Gene Expression Proˆle of Chrysene

Salmonella typhimurium in the presence of an exogene- qPCR with 33 candidate genes and statistical analysis ous metabolic system (5). The in vivo mutagenicity of via PCA successfully diŠerentiate the genotoxic chrysene was observed in six organs (liver, spleen, lung, hepatocarcinogens (DEN and 2,6-dinitrotoluene) from kidney, bone marrow, and colon) of lacZ transgenic the non-genotoxic hepatocarcinogen (DEHP) and the mouse (Mutatrade markMouse) (6). non-genotoxic non-hepatocarcinogen (phenacetin) (10). Previously, in our Toxicogenomics/JEMS･MMS col- The present study was a preceding study to determine laborative studies, we examined by using DNA microar- the optimal examination time within 48 h of the ad- rays the diŠerential gene expression induced by the ministration of the genotoxic hepatocarcinogen chry- application of 13 diŠerent chemicals: 8 genotoxic he- sene for the clariˆcation of biomarker genes; to this patocarcinogens [o-aminoazotoluene, chrysene, diben- end, we examined the gene expression proˆles at 4, 16, zo[a,l]pyrene, diethylnitrosamine (DEN), 7,12-dimethy- 20, 24, and 48 h after the administration of this chemi- lbenz[a]anthracene, dimethylnitrosamine, dipropyl- cal. We supposed that 4 h was the time for DNA nitrosamine (DPN) and ethylnitrosourea (ENU)], 4 damage and repair, 16 h the time for DNA repair for non-genotoxic hepatocarcinogens [carbon tetrachlo- some chemicals (11), 20 h the beginning of cell prolifer- ride, di(2-ethylhexyl)phthalate, phenobarbital (PB) and ation, and 24 and 48 h the time of cell proliferation trichloroethylene], and a non-genotoxic non-hepatocar- (12,13). cinogen [ethanol]. The DNA microarray analysis was In a preliminary study, we examined these time- conducted on 9-week-old male B6C3F1 mouse liver sam- course changes in the gene expression proˆle in the plesstarting4handupto28daysafterasinglein- mouse liver 4 to 48 h after the administration of chry- traperitoneal administration. Dozens of candidate genes sene, using home-prepared DNA microarrays (glass were found to discriminate the genotoxic hepatocar- slide: Gene Slide, Toyo Kohan Co. Ltd., Tokyo, Japan; cinogens from the non-genotoxic hepatocarcinogens; Gene Arrayer: Kaken Geneqs Inc., Chiba, Japan; these results with home-prepared DNA microarray 40-47-mer unmodiˆed oligonucleotides for 140 genes: (selected 268 genes) were reported in part for DEN, Invitrogen Japan K.K., Tokyo, Japan). The nucleotide DPN, PB, and ethanol (7) and those with AŠymetrix sequences were applied to Japan Patent No. GeneChip Mu74AV2 (12,000 genes) were registered in 2005–99843. The experimental procedures were similar the GEO database (GEO accession GSE33248 by Suzuki to those described previously (7). We extracted 50 candi- T. and Yang L.). Consequently we selected from previ- date genes from the DNA microarray study (results not ous studies the present 50 candidate genes which dis- shown), and in the present study, we determined the criminate genotoxic hepatocarcinogens from non-geno- gene expression proˆle of these 50 genes by qPCR, toxic hepatocarcinogens in mouse liver. Our previous which is a powerful tool for the measurement of mRNA results showed that most of candidate genes exhibited expression and is the gold standard for microarray gene increases in gene expression and few genes such as Egfr, expression data conˆrmation (14). Lrp1 and Mbd1 exhibited decreases after treatment with genotoxic hepatocarcinogens. Additionally, dose- Materials and Methods dependent alterations in the gene expression of 31 out of Animal treatment: Thirty male 8-week-old B6C3F1 51 of the examined candidate genes were demonstrated mice were obtained from Charles River Japan, Inc. 4 h and 28 days after the administration of DEN (3, 9, (Yokohama, Japan). The animals were maintained in 27 and 80 mg/kg bw, 1/40-1/2 of 50z of lethal dose plastic cages with wood chips as bedding and given food

(LD50)) and ENU (6, 17, 50 and 150 mg/kg bw, (Oriental MF, Oriental Yeast Co., Tokyo) and water ad

1/80-1/3 of the LD50), as determined by qPCR (8). We libitum in an air-conditioned room (12-h light, 12-h recently successfully discriminated between 8 genotoxic dark; 23±29C; 55±5z humidity). All of the animal mouse hepatocarcinogens (2-acetylamino‰uorene, 2,4- experiments were conducted in accordance with the diaminotoluene, diisopropanolnitrosamine, 4-dimethy- NIH Guide for Care and Use of Laboratory Animals laminoazobenzen, 4-(methylnitrosamino)-1-(3-pyridyl)- and approved by the Animal Care and Use Committee 1-butanone, N-nitrosomorpholine, quinoline and at Aoyama Gakuin University. Groups of three 9-week- urethane) and 4 non-genotoxic hepatocarcinogens (1,4- old mice were injected intraperitoneally (i.p.) with 0.2 dichlorobenzene, dichlorodiphenyltrichloroethane, mL (each mouse) of chrysene (100 mg/ kg bw; 1/3 of DEHP and furan) using qPCR with 35 candidate genes 50z lethal dose; Supleco, Bellefonte, PA, USA; CAS and principal component analysis (PCA) for 12 selected 218-01-9) suspended in olive oil. The control animals genes associated with Trp53-mediated signaling path- received olive oil. At 4, 16, 20, 24 and 48 h after treat- way involved in the DNA damage response. This dis- ment, the animals were sacriˆced, and the liver was col- crimination was demonstrated at 4 and 48 h after a sin- lected, frozen on dry ice, and stored at －809C until use. gle administration of the chemicals (9). More recently, RNA isolation: The total RNA was extracted from we succeeded in applying a similar study in rat liver. approximately 30 mg from each liver using Micro

5555 Mikiya Sakurai et al.

Table 1. Fifty genes quantiˆed in the present study

No. Symbol Gene name Accession no. 1 Aen apoptosis enhancing nuclease NM_026531 2 Aldh3a2 aldehyde dehydrogenase family 3, subfamily A2 NM_007437 3 Bax BCL2-associated X NM_007527 4 Bhlhe40 basic helix-loop-helix domain containing, class B2 NM_011498 5 Btg2 B cell translocation gene 2, anti-proliferative NM_007570 6 Casp1 caspase 1 NM_009807 7 Casp4 caspase 4, apoptosis-related cysteine peptidase NM_007609 8CcnfcyclinF NM_007634 9 Ccng1 cyclin G1 NM_009831 10 Ccng2 cyclin G2 NM_007635 11 Cdkn1a cyclin-dependent kinase inhibitor 1A (P21) NM_007669 12 Crp C-reactive protein, pentraxin-related NM_007768 13 Cyp1a1 cytochrome P450, family 1, subfamily a, polypeptide 1 NM_009992 14 Cyp1a2 cytochrome P450, family 1, subfamily a, polypeptide 2 NM_009993 15 Cyp2a5 cytochrome P450, family 2, subfamily a, polypeptide 5 NM_007812 16 Cyp4a10 cytochrome P450, family 4, subfamily a, polypeptide 10 NM_010011 17 Ddit4 DNA-damage-inducible transcript 4 NM_029083 18 Ddit4l DNA-damage-inducible transcript 4-like NM_030143 19 Egfr epidermal growth factor receptor NM_207655 20 Ephx1 epoxide 1, microsomal NM_010145 21 Fkbp5 FK506 binding protein 5 NM_010220 22 Gadd45b growth arrest and DNA-damage-inducible 45 beta NM_008655 23 Gadd45g growth arrest and DNA-damage-inducible 45 gamma NM_011817 24 Gapdh glyceraldehyde-3-phosphate dehydrogenase NM_008084 25 Gdf15 growth diŠerentiation factor 15 NM_011819 26 Hhex hematopoietically expressed homeobox NM_008245 27 Hist1h1c histone cluster 1, H1c NM_015786 28 Hmox1 heme oxygenase (decycling) 1 NM_010442 29 Hprt hypoxanthine guanine phosphoribosyl NM_013556 30 Hspb1 heat shock protein 1 NM_013560 31 Igfbp1 insulin-like growth factor binding protein 1 NM_008341 32 Jun jun proto-oncogene NM_010591 33 Lcn2 lopocalin 2 NM_008491 34 Lpp LIM domain containing preferred translocation partner in lipoma NM_178665 35 Lrp1 low density lipoprotein receptor-related protein 1 NM_008512 36 Ly6a lymphocyte antigen 6 complex, locus A NM_010738 37 Mbd1 methyl-CpG binding domain protein 1 NM_013594 38 transformed mouse 3T3 cell double minute 2 NM_010786 39 Phlda3 pleckstrin homology-like domain, family A, member 3 NM_013750 40 Plk2 polo-like kinase 2 NM_152804 41 Pml promyelocytic leukemia NM_008884 42 Pmm1 phosphomannomutase 1 NM_013872 43 Ppp1r3c protein phosphatase 1, regulatory (inhibitor) subunit 3C NM_016854 44 Psma3 proteasome (prosome, macropain) subunit, alpha type 3 NM_011184 45 Rad52 RAD52 homolog (S. cerevisiae)NM_011236 46 Rcan1 regulator of 1 NM_001081549 47 St3gal5 ST3 beta-galactoside alpha–2,3-sialyltransferase 5 NM_001035228 48 Trp53 transformation-related protein 53 NM_011640 49 Tubb2a tubulin, beta 2A class IIA NM_009450 50 Tubb4b tubulin, beta 4B class IVB NM_146116

56 Gene Expression Proˆle of Chrysene

Table 2. Primer sequences of the 50 genes examined in the present study

No. Symbol Left primer Right primer 1 Aen TTGAAGGGCAAGGTGGTGGTG GAGCAGGTTTGGGACATAAGTG 2 Aldh3a2 CTGTTAATTCTCTGCCCTTTGG CTGAGCTTGTTGACGCTTTCC 3 Bax CCAGGATGCGTCCACCAAGAAG GGAGTCCGTGTCCACGTCAGC 4 Bhlhe40 CCAGGCCTCAACACCTCAGCTG CCGAAGAGTCGAGGGACGAATG 5 Btg2 ACGGGAAGAGAACCGACATGC ATGATCGGTCAGTGCGTCCTG 6 Casp1 GTCTTGGAGACATCCTGTCAGG GCATCTGTAGCCTAAATTCTGG 7 Casp4 GAAGACTTAGGCTACGATGTGGTG TGTCTGATGTCTGGTGTTCTGAG 8 Ccnf AGCACAAAGCCTTGCCACCATC AAGCCAGGTGCATGTCCTTGTC 9 Ccng1 TGGCCGAGATTTGACCTTCTGG GTGCTTCAGTTGCCGTGCAGTG 10 CCng2 GCCATCAAGCTAGGACTGTTAG CACTTATCAACTCCATTCCCTG 11 Cdkn1a TCCCGTGGACAGTGAGCAGTTG CGTCTCCGTGACGAAGTCAAAG 12 Crp TGGGATTACATTCTCTCCCTTG TGCTCTAGTGCTGAGGACCAC 13 Cyp1a1 TGGCCGATCGGAGGTCTTTC AAGTGTTCACAGCGGGCGTG 14 Cyp1a2 AGGTGATGCTCTTCGGCTTGG CACACCCGGTGGCACACTAAAC 15 Cyp2a5 CCCAAAGCACTTCCTAGATGAC AGGAAGAGGAAGAGTTCCATCC 16 Cyp4a10 ACCCGAAGGTGTGGCCAAATCC GCAGTAGTTCGAAGCGGAGCAG 17 Ddit4 GCACCTGTGTGCCAACCTGATG TGTATGCCAGGCGCAGGAGTTC 18 Ddit4l ACCAGCTTGGCTGGGACAAATG CGTGCTCATTGGAACAGTGATG 19 Egfr ATGGGAGCTGCCGTGTCAAAG TGGCCTCTTGGGAACAGATTGG 20 Ephx1 GATGGAGGCCTGGAGAGGAAG CCAGTGGGCACAAAGACCTTC 21 Fkbp5 GACTGACTGGCCTGCTAAGAAC ACTCAAGAGTCTGCGAAAGGAC 22 Gadd45b CGACAACGCGGTTCAGAAGATG ATCAGTTTGGCCGCCTCGTAC 23 Gadd45g GGAAAGCACAGCCAGGATGCAG ATTCAGGACTTTGGCGGACTCG 24 Gapdh GCTCTCAATGACAACTTTGTCAAG TCCTTGGAGGCCATGTAGGC 25 Gdf15 AACACGCATGCGCAGATCAAAG CCACCAGGTCATCATAAGTCTG 26 Hhex GGA CAGTTTGGAC ACTTCCTGTG GGTCGGAATC CTCTGAGATC TC 27 Hist1h1c GGCTGTGACCAAGAAAGTGGCC CTTCTTGGCCTTGGCAACCTTG 28 Hmox1 AAGACCGCCTTCCTGCTCAAC CGAAGTGACGCCATCTGTGAGG 29 Hprt CTTGCTCGAGATGTCATGAAGGAG TAATCCAGCAGGTCAGCAAAGAAC 30 Hspb1 CGGTGCTTCACCCGGAAATAC GCTGACTGCGTGACTGCTTTGG 31 Igfbp1 GATCAGCCCATCCTGTGGAACG TTCTCGTTGGCAGGGCTCCTTC 32 Jun GCCAAGAACTCGGACCTTCTC AGTGGTGATGTGCCCATTGCTG 33 Lcn2 CACCACGGACTACAACCAGTTC GCTCCTTGGTTCTTCCATACAG 34 Lpp CCGTGATTTC CATGTGCACT GC CT TGGCCGTCAA GACCCTGATG 35 Lrp1 GGGCCATGAATGTGGAAATTGG GTGGCATACACTGGGTTGGTG 36 Ly6a CTTGTGGCCCTACTGTGTGCAG GGGCAGGTAATTGATGGGCAAG 37 Mbd1 GGATCCTGACACTCAAGAATGG GTTTGGGCTAACACAGGAAGAG 38 Mdm2 TTGATCCGAGCCTGGGTCTGTG AAGATCCTGATGCGAGGGCGTC 39 Phlda3 TGGCTGGAACGCTCAGATCAC TTAGGACACAAGGGTCCCAGTC 40 Plk2 CTGTTGAGAGCGTCTTCAGTTG CCATAGTTCACAGTTAAGCAGC 41 Pml GGCAAGAAGCGTCCTTACCTTC GGACAGCAACAGCAGTTCAGTC 42 Pmm1 TCTCCCGAGGAGGCATGATAAG CAAAGTCATTCCCGCCAGGAC 43 Ppp1r3c TGGAAACCTGACGGAGTGCAG GCAAGCCTTGGACTGCCAAAG 44 Psma3 GATCGACCCATCAGGTGTTTC CACGGCAAGTCATTTCCTTCATCTG 45 Rad52 TGACGCCACTCACCAGAGGAAG GCTGGAAGTACCGCATGCTTGG 46 Rcan1 GGTCCACGTGTGTGAGAGTG TGGATGGGTGTGTACTCCGG 47 St3gal5 GCAGGTCATGCACAATGTGACC CTGGGTGAGGTTTGCCGTGTTC 48 Trp53 TTGGACCCTGGCACCTACAATG GCAGACAGGCTTTGCAGAATGG 49 Tubb2a CGAGCTGACCCAGCAGATGTTC CACGGCCGTCTTGACGTTGTTG 50 Tubb4b TTGGCAACAGCACCGCTATTC TCGGACACCAGGTCGTTCATG

57 Mikiya Sakurai et al.

Smash MS-100 (Tomy Digital Biology Co., Ltd., lyzed by Welch's t-test in Microsoft O‹ce Excel 2007

Tokyo, Japan) and QuickGene-800 (Fujiˆlm, Tokyo, after the logarithmic (log2) transformation of the data. Japan) as described previously (8,9). A p-value of less than 0.05 was recognized as statistical- Quantiˆcation by qPCR: The complementary ly signiˆcant in the present study. DNA (cDNA) was obtained from the total RNA using the SuperScript First strand synthesis system for RT- Results PCR kit (Invitrogen Corp., Carlsbad, CA, USA) (8). Time-dependent alteration of the gene expression We quantiˆed 50 candidate genes based on our previous proˆle induced in the mouse liver with chrysene: DNA microarray and qPCR studies. The symbols, gene The gene expression ratio (experimental group/control names and accession numbers of the 50 genes are sum- group; exp/cont) was calculated individually for each marized in Table 1. The qPCR ampliˆcations were per- group(3miceintriplicateassays)fromtheqPCR formed in triplicate assays using the SYBR Green I as- results. The mean±SD was determined, and the results say in an Opticon II instrument (MJ Research, Inc., were analyzed by Welch's t-test, unsupervised hierarchi- Waltham, MA, USA). Reactions were conducted in a cal clustering and unsupervised k-means clustering 96-well plate, and each reaction was performed in a (Table 3). Total of 35 genes (Cyp1a1, Cyp1a2, Ly6a, volume of 20-ml containing 2× SYBR Green Master Gadd45g, Igfbp1, Lcn2, Cyp2a5, Aldh3a2, Casp1, Mix (Applied Biosystems, CA, USA), 2 pmol of each Casp4, Cdkn1a, Btg2, Gdf15, Phlda3, St3gal5, forward and reverse primer and a cDNA template corre- Bhlhe40, Egfr, Ccng2, Mdm2, Jun, Bax, Fkbp5, Crp, sponding to 400 pg of total RNA. Each primer sequence Hmox1, Gadd45b, Pmm1, Tubb2a, Ephx1, Lpp, is shown in Table 2. The SYBR Green PCR conditions Ppp1r3c, Hhex, Ccnf, Rcan1, Ccng1 and Tubb4b)ex- were 959C for 10 minutes and 45 cycles of 959Cfor10s, hibited statistically signiˆcant increases at least once 589Cfor50sand729C for 20 s. Each 96-well reaction within 48 h, as computed by Welch's t-test. Of these plate included six standard samples of the liver cDNA, genes, 25 genes (Cyp1a1, Cyp1a2, Ly6a, Gadd45g, which was diluted to 1/5, 1/25, 1/125, 1/625 and Igfbp1, Lcn2, Cyp2a5, Aldh3a2, Casp4, Cdkn1a, Btg2, 1/3125. Finally, the relative quantitative values of each Bhlhe40, Ccng2, Fkbp5, Crp, Hmox1, Gadd45b, sample were normalized with those found for the Gapdh Pmm1, Tubb2a, Ephx1, Lpp, Hhex, Ccnf, Rcan1 and gene (8). Tubb4b) exhibited statistically signiˆcant increases of Data analysis and clustering algorithm: To use the more than two-fold. Cyp1a1 and Cyp1a2 exhibited cluster analysis program, we performed a logarithmic remarkably consistent increases in gene expression dur-

(log2) transformation of the data to stabilize the vari- ing4to48h.At4h,15genes(Cyp1a1, Cyp1a2, Ly6a, ance and the gene expression proˆle of each chrysene Gadd45g, Igfbp1, Lcn2, Casp4, Cdkn1a, Btg2, and control sample, which eŠectively normalized the Bhlhe40, Ccng2, Fkbp5, Crp, Hmox1 and Gadd45b) data to the median gene expression level of the entire showed statistically signiˆcant increase more than two- sample set. Each dataset obtained from the time points fold.At16h,6genes(Cyp1a1, Cyp1a2, Lcn2, Fkbp5, 4, 16, 20, 24 and 48 h after chrysene administration was Gadd45b and Pmm1) exhibited statistically signiˆcant subjected to both unsupervised hierarchical and un- increases of more than two-fold; 9 expression-increased supervised k-means clustering using the GENESIS genes at 4 h returned to the control level, 1 gene (Ly6a) software (http://genome.tugraz.at/) (15). decreased and 1 gene (Pmm1)increased.At20h,7 , pathways, and network analysis: genes (Cyp1a1, Cyp1a2, Ly6a, Cyp2a5, Aldh3a2, The gene ontology analysis was performed using The Gadd45b and Pmm1) exhibited statistically signiˆcant Gene Ontology (http://www.geneontology.org/), Can- increases of more than two-fold; 9 expression-increased cer Gene census (http://cancer.sanger.ac.uk/cancer genes at 4 h returned to the control level, 2 genes genome/projects/census/) and Tumor suppressor gene (Gadd45g and Lcn2) decreased and 3 genes (Cyp2a5, database (http://bioinfo.mc.vanderbilt.edu/TSGene/). Aldh3a2 and Pmm1)increased.At24h,7genes The gene pathways and networks were generated with (Cyp1a1, Cyp1a2, Igfbp1, Fkbp1, Tubb2a, Ephx1 and GeneSpringGX11.0.1 and Ingenuity Pathway Analysis Lpp) exhibited statistically signiˆcant increases of more (www.Ingenuity.com) as described previously (8). The than two-fold; 8 expression-increased genes at 4 h score in Table 4 indicates the likelihood of the focus returned to the control level, 1 gene (Gadd45g) genes in a given network being found together due to decreased and 3 genes (Tubb2a, Ephx1 and Lpp)in- random chance. A score of À2 indicates that there is a creased. At 48 h, 10 genes (Cyp1a1, Cyp1a2, Bhlhe40, º1 in 100 chance that the focus genes were assembled Hmox1, Tubb2a, Ephx1, Hhex, Ccnf, Rcan1 and randomly into a network (8). The results were conˆrmed Tubb4b) exhibited statistically signiˆcant increases of with references in PubMed (http://www.ncbi.nlm. more than two-fold; 5 expression-increased genes at 4 h nih.gov/pubmed). returned to the control level and 6 genes (Tubb2a, Statistical analysis: The data were statistically ana- Ephx1, Hhex, Ccnf, Rcan1 and Tubb4b)increased.By

5858 Gene Expression Proˆle of Chrysene

Table 3. Gene expression ratio (Exp/Cont) and Welch's t-test after chrysene administration

The total RNA was extracted from the individual liver and used to prepare the cDNA. The expression of the 50 genes was quantiˆed by qPCR, and the gene expression ratio (exp/cont) of each gene was calculated. The results were analyzed by Welch's t-test (boldface with ** indicates signiˆcant at Pº0.01; boldface with * indicates signiˆcant at Pº0.05). The clusters in the table were sorted through unsupervised hierarchical clustering. The dark pink color shows the values that are higher than 2, and the light pink color shows the values that are higher than 1.5.

59 Mikiya Sakurai et al. comparing the results, we noted the biggest diŠerences Ccng1, Ccng2, Cdkn1a, Gadd45b, Gadd45g, Gapdh, between 4 and 48 h: 11 genes (Ly6a, Gadd45g, Igfbp1, Hhex, Hist1h1c, Mdm2, Pml, Psma3, Tubb2a and Lcn2, Casp4, Cdkn1a, Btg2, Ccng2, Fkbp5, Crp and Tubb4b) were associated with network 1 and its ``top Gadd45b) exhibited statistically signiˆcant increase diseases and functions'' were ``cancer, hematological more than two-fold only at 4 h and 6 genes (Tubb2a, disease and cell cycle''. Thirteen genes (Bhlhe40, Casp4, Ephx1, Hhex, Ccnf, Rcan1 and Tubb4b) exhibited Crp, Cyp1a1, Cyp1a2, Cyp2a5, Ddit4, Ephx1, Fkbp5, statistically signiˆcant increase more than two-fold only Gdf15, Hmox1, Lcn2 and Plk2)wereassociatedwith at 48 h. network 2, and its top functions were ``drug Unsupervised hierarchical clustering sorted all the metabolism, small molecule biochemistry, and endo- above genes into four groups (clusters), as shown in crine system development and function''. Eight genes Table 3. Cyp1a1 and Cyp1a2 in the cluster denoted (Aen, Aldh3a2, Hprt, Jun, Lrp1, Phlda3, Pmm1 and ``others'' showed statistically signiˆcant remarkable in- Trp53)wereassociatedwithnetwork3,andthetop creases at 4–48 h. The genes in clusters 1 and 2 showed functions of this network were ``cell death and survival an increasing tendency in their gene expression at 4 h, as and cellular growth and proliferation''. conˆrmed by k-means clustering: 3 genes in cluster 1 (Gadd45g, Igfbp1 and Lcn2) and 5 genes in cluster 2 Discussion (Casp4, Cdkn1a Btg2, Bhlhe40 and Ccng2) showed In the present study we examined by qPCR the changes statistically signiˆcant increases of more than two-fold. in gene expression in the mouse liver within 48 h induced The genes in cluster 3 showed increases in their gene ex- by chrysene, a polycyclic aromatic hydrocarbon and pression at 48 h, as conˆrmed by k-means clustering: 7 genotoxic hepatocarcinogen. As a result, we observed genes in cluster 3 (Hmox1, Tubb2a, Ephx1, Hhex, three time-dependent patterns. The ˆrst one in cluster 1 Ccnf, Rcan1 and Tubb4b) showed statistically sig- and 2 in Table 3 showed that 11 genes (Gadd45g, Igfbp1 niˆcant increases of more than two-fold at 48 h. In and Lcn2 in cluster 1 and Casp4, Cdkn1a, Btg2, Phlda3, Table 3, the results found for Gapdh, a housekeeping Bhlhhe40, Egfr, Ccng2 and Mdm2 in cluster 2) exhibit- gene, are shown. This gene was used to normalize the ed statistically signiˆcant increases in their gene expres- gene expression ratio, as this gene did not show changes sion at 4 h. The biological processes to which these in its expression. genes are associated or classiˆed are DNA damage, The biologically relevant gene networks: The numeri- DNA repair, apoptosis, cell cycle, cell proliferation and cal data of all 50 examined genes were analyzed by IPA, oncogene. The second one in cluster 3 in Table 3 showed and 7 gene networks were extracted for 4, 16, 20, 24 and that 8 genes (Hmox1, Tubb2a, Ephx1, Lpp, Hhex, 48 h time points. Of these 7 networks, the 3 major gene Ccnf, Rcan1 and Tubb4b) exhibited statistically sig- networks at 4 h are summarized in Table 4 as a niˆcant increases in their gene expression at 48 h. The representative result. Similar gene networks were ex- biological processes to which these genes are associated tracted at 16, 20, 24 and 48 h. Fourteen genes (Btg2, are DNA damage, apoptosis, cell cycle, cell prolifera-

Table 4. Gene networks and their pathway functions 4 h after chrysene administration

Focus Networks Molecules in network Score molecule Top diseases and functions

1 14-3-3, 26s, Proteasome, Akt, Alpha tubulin, Beta tublin, BTG2,Calcine- 29 14 Cancer, rin protein(s), calpain, Cbp/p300, CCNG1, CCNG2,Cdc2,CDKN1A, Cell cycle, Hematological disease Cyclin A, Cyclin E, E2f, GADD45, GADD45B, GADD45G, GAPDH, Hdac, HHEX, HIST1H1C, Histone, Histone H1, MDM2,MEF2,Pak, PML,Ppp2c,PSMA3, Rb, TIP60, TUBB2A, TUBB4B

2 20s proteasome, ADRB, Ahr-aryl hydrocarbon-Arnt, BHLHE40, C/ebp, 28 13 Drug metabolism, CASP4, creatine kinase, CRP, CYP, CYP1A1, CYP1A2, CYP2A13, Small molecule biochemistry, DDIT4, EPHX1, ERK1/2, FKBP5, GDF15, GOT, Growth hormone, Endocrine system development hemoglobin, HMOX1, LCN2, Ldh, N-cor, Nos, Nuclear factor 1, p70 and function S6k, PDGF BB, P13K, Pias, PLK2, PP2A, Stat3-Stat3, thyroid hormone receptor, tyrosine kinase

3 2-methoxyestradiol, 7-kethocholesterol, ADRA1A, ADRA2A, AEN, 15 8 Cell death and survival, ALDH3A2,CD3,CD3E,CDKN2D,Cg,CGB7,Ck2,FSH,GSTA1,Hi- Cellar growth and proliferation stone h4, HPRT1, Jnk, Jnk dimer, JUN, Jun-AFT2, Jun-GABP, LRP1, Cellular development MT1H, PHLDA3, PMM1, PTP4A1, Ras, RNA polymerase II, Sapk, Smad, STAT, TP53, Tublin, tyrosine kinase, Ubiquitin

The gene networks and their pathway functions were analyzed by IPA. The focus molecules (boldface) are the genes that were quantiˆed in the present study. The gene names are written as the names of the human genes (as in IPA). The score is explained in the section titled ``Gene ontology, pathways, and network analysis'' in Materials and Methods. CYP2A13 is the human orthologue of mouse Cyp2a5.

60 Gene Expression Proˆle of Chrysene tion, oxidative stress, cell adhesion, aromatic com- to 48 h. In our previous studies (7–10), the gene expres- pound metabolic process and microtubule-based proc- sion of Trp53 scarcely changed in the mouse and rat ess. The third one in the cluster ``others'' in Table 3 liver 4 and 48 h and 28 days after the administration of showed that 2 genes (Cyp1a1 and Cyp1a2) exhibited a genotoxic and non-genotoxic hepatocarcinogens. The continuous gene expression increase from 4 to 48 h. The basal gene expression of Trp53 in the control animals biological processes to which these 2 genes are associ- may already be su‹cient for DNA damage under the ated are the toxin metabolism process and the ox- present experimental conditions. Recently, it was rev- idation-reduction process. It was reported that carcino- iewed that post-translational modiˆcations of in- genic PAHs induced gene expression of Cyp1a1 and cluding methylation, phosphorylation, acetylation, and Cyp1a2 (16). Our results conˆrmed this. As in our previ- ubiquitination aŠected the functions of p53 in response ous advanced qPCR studies (8–10) we did not examine to cellular stress (37). As we described in the previous PAH, and we did not observe remarkable increases in studies (9,10) the major, biologically relevant gene path- these genes (8–10). It was reported that Cyp1a1 and way suggested was the DNA damage response pathway, Cyp1a2 were regulated by arylhydrocarbon receptor resulting from signal transduction by a p53-class media- (AhR) (16). However, AhR was not extracted as a can- tor leading to the induction of apoptosis. Little is didate marker gene in our previous DNA microarray known about the acute expression changes of Trp53 in studies, and gene expression of AhR was not determined the rodent liver after exposure to hepatocarcinogens; in in the present study. The present results revealed that fact, only a few reports have suggested the activation of marked gene expression changes can be observed in the a Trp53-mediated signaling pathway after the adminis- mouse liver at 4 h and at 48 h after chrysene administra- tration of hepatocarcinogens (38). tion. These gene expression changes induced by chry- Although most of the 50 genes evaluated in the sene were mostly increases in the examined 50 candidate present study were also examined in our previous studies genes; however, some of the genes, such as Ly6a, in the mouse liver (7–9), the results obtained for 9 of the Gadd45g and Lcn2, exhibited bimodal gene expression. genes (Lcn2 in cluster 1; Cyp2a5, Aldh3a2 and Casp4 in These bimodal phenomena have been scarcely reported. cluster 2; and Fkbp5, Crp, Tubb2a, Lpp and Hhex in Therefore, further studies will be required to elucidate cluster 3) were found only in the present study. In the the roles of bimodal gene expression of these genes. present study, all of these genes exhibited statistically It has been reported that DNA damage and repair are signiˆcant increases in their gene expression of more observed a few hours (2–4 h) after carcinogen adminis- than two-fold at least once within 4 to 48 h after ad- tration in the mouse liver, as determined by a UDS assay ministration. These 9 genes were also selected from our (11) and adduct formation (17). DNA replication and previous DNA microarray results. Although 12 genes cell proliferation have been observed approximately 48 (Aen, Cyp4a10, Ddit4, Ddit4l, Hprt, Hspb1, Lrp1, h after carcinogen administration in the mouse liver, as Mbd1, Plk2, Pml, Psma3 and Rad52) did not show determined by RDS assay (18). Before our study, we statistically signiˆcant changes in their gene expression hypothesized that the genes which exhibited changes at 4 in the present study, they showed statistically signiˆcant hmayberelatedtoDNAdamageandrepairandthat changes in their gene expression in response to the the genes that exhibited changes at 48 h may be related administration of 15 genotoxic hepatocarcinogens (7 to cell proliferation. In the present study changes in N-nitroso compounds, namely diethylnitrosamine, di- gene expression associated with DNA damage and isopropanolnitrosamine, dimethylnitrosamine, dipro- repair were observed at 4 h as we expected, while pylnitrosamine, ethylnitrosourea, 4-(methylnitrosa- changes in gene expression related to cell proliferation mino)-1-(3-pyridyl)-1-butanone, and N-nitrosomorpho- were not concentrated at 48 h. line; 2-acetylamino‰uorene, o-aminoazotoluene, 2,4-di- Twenty of the genes quantiˆed in this study have been aminotoluene, dibenzo[a,l]pyrene, 4-dimethylaminoa- reported to be directly or indirectly associated with zobenzene, 7,12-dimethylbenz[a]anthracene, quinoline Trp53 (Aen (19), Bax (20), Btg2 (21), Casp1 (22), Ccng1 and urethane) in our previous studies (11–13). Further- (23), Ccng2 (24), Cdkn1a (20), Crp (25), Ddit4 (26), more, Cyp1a1 and Cyp1a2 seem to be speciˆc markers Ephx1 (27), Gft15 (28), Hmox1 (29), Hspb1 (30), Igfbp1 for carcinogenic PAH and Ccng1 seems to be a useful (31), Lcn2 (32), Mdm2 (20), Phlda3 (33), Plk2 (34), Pml marker for carcinogenic N-nitroso compounds. (35) and Pmm1 (36)). Chrysene exhibited statistically Although in our previous studies a considerable number signiˆcant increases of gene expression in 15 genes (Bax, of genotoxic hepatocarcinogens exhibited decreased-ex- Btg2, Casp1, Ccng1, Ccng2, Cdkn1a, Crp, Ephx1, pression in Egfr, Lrp1 and Mbd1 but instead a sig- Gdf15, Hmox1, Igfbp1, Lcn2, Mdm2, Phlda3 and niˆcant increase in expression of these genes in the Pmm1) and increase tendencies in 3 genes (Ddit4, present study. Hspb1 and Pml) among these genes. However, the gene Although the number of published studies is limited, expression of Trp53 itself did not show changes from 4 gene expression proˆling of the eŠects of hepatocarcino-

6161 Mikiya Sakurai et al. gens in the experimental animal liver has been per- Acknowledgments: This work was partly supported by formed using DNA microarray and qPCR with diŠerent KAKENHI (18310047) (C. Furihata, T. Watanabe, and time courses (38–41). Ellinger-Ziegelbauer et al.exam- T. Suzuki) and a Grant-in-Aid for The Private School ined the eŠects of four genotoxic hepatocarcinogens, High-tech Research Center Program of The Ministry of namely dimethylnitrosamine, 2-nitro‰uorene, a‰atoxin Education, Culture, Sports, Science, and Technology of B1 and 4-(methylnitrosamino)1-(3-pyridyl)-1-butanone, Japan. in the rat liver 1, 3, 7 and 14 days after administration using an AŠymetrix RG_U34 microarray (38). Iida et al. Con‰icts of interest: None of the authors have any examined the eŠects of the non-genotoxic carcinogens con‰icts of interest associated with this study. oxazepam and Wyeth-14643 in the mouse liver 2 weeks after administration using real-time PCR and References oligonucleotide microarrays (39). Kang et al.examined 1 Kerns RT, Bushel PR. The impact of classiˆcation of in- the eŠects of the genotoxic hepatocarcinogen MeIQx at terest on predictive toxicogenomics. Front Genet. 2012; weeks4,16and102intheratliverusinganA‹metrix 3: 14. GeneChip Rat Genome 230 2.0 Array, observed no 2 OvandoBJ,EllisonCA,VezinaCM,OlsonJR.Tox- major diŠerences after 4 and 16 weeks administration, icogenomic analysis of exposure to TCDD, PCB126 and PCB153: identiˆcation of genomic biomarkers of ex- and found a few diŠerentially expressed genes in tumors posure to AhR ligands. BMC Genomics. 2010; 11: 583. at 102 weeks (40). Woods et al. examined the eŠects of a 3 Bar T, Kubista M, Tichopad A. Validation of kinetics non-genotoxic carcinogen, namely the peroxisome similarity in qPCR. Nucleic Acids Res. 2012; 40: proliferator Wy-14,643, at 8, 24 and 72 h and 1 and 4 1395–406. weeks in the mouse liver using an Agilent Mouse Oligo 4 IARC (International Agency for Research on Cancer). Microarray (41). Each of these researchers performed IARC Monographs on the Evaluation of Carcinogenic diŠerential gene expression proˆling at diŠerent time Risks to Humans. http://monographs.iarc.fr/ENG/ points in the mouse or rat liver. However, there is scarce Preamble/currentb6evalrationale0706.php information on the acute gene expression changes (wi- 5 IARC (International Agency for Research on Cancer). thin 48 h) in the mouse or rat liver. The present results Monographs on the Evaluation of the Carcinogenic Risk provide an example of the acute gene expression of Chemicals for Humans. Polynuclear Aromatic Com- pounds. Part 1. Chemical, Environmental and Ex- changes induced by a polycyclic aromatic organic car- perimental Data. World Health Organization. 1983; 32: cinogenic compound in the mouse liver. 247–61. In conclusion, in the present study, we examined by 6 Yamada K, Suzuki T, Kohara A, Kato TA, Hayashi M, qPCR the time-dependent changes in the gene expres- Mizutani T, et al. Nitrogen-substitution eŠect on in vivo sion proˆles in the mouse liver during a very early mutagenicity of chrysene. Mutat Res. 2005; 586: 1–17. period (4, 16, 20, 24 and 48 h) after chrysene adminis- 7 Watanabe T, Tobe K, Nakachi Y, Kondoh Y, Nakajima tration. We selected 50 candidate genes which discrimi- M, Hamada S, et al. DiŠerential Gene Expression In- nated genotoxic hepatocarcinogens from non-genotoxic duced by Two Genotoxic N-nitroso Carcinogens, hepatocarcinogens based on our previous DNA Phenobarbital and Ethanol in Mouse Liver Examined microarray studies, and we used these to determine the with Oligonucleotide Microarray and Quantitative Real- optimal examination time within a short period after the time PCR. Gene Env. 2007; 29: 115–27. administration of the genotoxic hepatocarcinogen. 8 Watanabe T, Tanaka G, Hamada S, Namiki C, Suzuki T, Nakajima M, et al. Dose-dependent alterations in gene Thirty-ˆve genes exhibited statistically signiˆcant in- expression in mouse liver induced by diethylnitrosamine creases at least once within the ˆrst 48 h after adminis- and ethylnitrosourea and determined by quantitative real- tration. Cyp1a1 and Cyp1a2 exhibited remarkably con- time PCR. Mutat Res. 2009; 673: 9–20. sistent increases in gene expression during 4 to 48 h. At 9 Watanabe T, Suzuki T, Natsume M, Nakajima M, Naru- the 4-h and 48-h time points, 15 genes (Bhlhe40, Btg2, mi K, Hamada S, et al. Discrimination of genotoxic and Casp4, Ccng2, Cdkn1a, Crp, Cyp1a1, Cyp1a2, Fkbp5, non-genotoxic hepatocarcinogens by statistical analysis Gadd45b, Gadd45g, Hmox1, Igfbp1, Lcn2 and Ly6a) based on gene expression proˆling in the mouse liver as and 10 genes (Bhlhe40, Ccnf, Cyp1a1, Cyp1a2, Ephx1, determined by quantitative real-time PCR. Mutat Res. Hhex, Hmox1, Rcan1, Tubb2a and Tubb4b), respec- 2012; 747: 164–75. tively, showed statistically signiˆcant increases of more 10 Suenaga K, Takasawa H, Watanabe T, Wako Y, Suzuki T, Hamada S, et al. DiŠerential gene expression proˆling than two-fold. Consequently, three time-dependent pat- between genotoxic and non-genotoxic hepatocarcinogens terns in gene expression were observed within 48 h after in young rat liver determined by quantitative real-time chrysene administration in mouse liver: Cyp1a1 and PCR and principal component analysis. Mutat Res. 2013; Cyp1a2 exhibited consistent increases; the highest num- 751: 73–83. ber of genes (15 genes) increased at 4 h; and 6 diŠerent 11 Maddle S, Dean SW, Andrae U, Brambilla G, Burlinson genes expressed at 4 h increased at 48 h.

6262 Gene Expression Proˆle of Chrysene

B, Doolittle DJ et al. Recommendatios for the perfor- reactive protein induces p53-mediated cell cycle arrest in mance of UDS tests in vitro and in vivo. Mutat Res. 1994; H9c2 cardiac myocytes. Biochem Biophys Res Commun. 312: 263–85. 2011; 410: 525–30. 12 Sawada S, Furihata C, Matsushima T. In vivo short-term 26 Zemskova M, Lilly MB, Lin YW, Song JH, Kraft AS. assays of repair and replication of rat liver DNA. J Can- p53-dependent induction of prostate cancer cell senes- cer Res Clin Oncol. 1989; 115: 345–50. cence by the PIM1 protein kinase. Mol. Cancer Res. 13 Miyagawa M, Takasawa H, Sugiyama A, Inoue Y, Mura- 2010; 8: 1126–41. ta T, Uno Y, et al. The in vivo-in vitro replication DNA 27 Chauhan PS, Ihsan R, Mishra AK, Yadav DS, Saluja S, synthesis (RDS) test with hepatocytes prepared from male Mittal V, et al. High order interactions of xenobiotic B6C3F1 mice as an early prediction assay for putative metabolizing genes and P53 codon 72 polymorphisms in nongenotoxic (Ames-negative) mouse hepatocaqrcino- acute leukemia. Environ Mol Mutagen. 2012; 53: 619–30. gens. Mutat Res. 1995; 343: 157–83. 28 Zhou Y, Zhong Y, Wang Y, Zhang X, Batista DL, Gej- 14 Dymond JS. Explanatory chapter: quantitative PCR. man R, et al. Activation of p53 by MEG3 non-coding Methods Enzymol. 2013; 529: 279–89. RNA. J Biol Chem. 2007; 282: 24731–42. 15 Sturn A, Quackenbush J, Trajanoski Z. Genesis: Cluster 29 Meiller A, Alvarez S, Dranáe P, Lallemand C, Blanchard analysis of microarray data. Bioinformatics 2002; 18: B, Tovey M, et al. p53 dependent stimulation of redox- 207–8. related genes in the lymphoid organs of gamma-ir- 16 Shimada T, Inoue K, Suzuki Y, Kawai T, Azuma E, radiated—mice identiˆcation of Haeme-oxygenase 1 as a Nakajima T, et al. Arylhydrocarbon receptor-dependent direct p53 target gene. Mol Cancer Res. 2010; 8: 1126–41. induction of liver and lung cytochromes P450 1A1, 1A2, 30 Vedam K, Nishijima Y, Druhan LJ, Khan M, Moldovan and 1B1 by polycyclic aromatic hydrocarbons and poly- NI, Zweier JL, et al. Role of heat shock factor–1 activa- chlorinated biphenyls in genetically engineered C57BL/6J tion in the doxorubicin-induced heart failure in mice. Am mice. Carcinogenesis 2002; 23: 1199–207. J Physiol Heart Circ Physiol. 2010; 298: H1832–41. 17 Mientjes EJ, Luiten-Schuite A, van der Wolf E, Bor- 31 Leu JI, George DL. Hepatic IGFBP1 is a prosurvival fac- sboomY,BergmansA,BerndsF,etal.DNAadducts, tor that binds to BAK, protects the liver from apoptosis, mutant frequencies, and mutation spectra in various and antagonizes the proapoptotic actions of p53 at organs of lamda lacZ mice exposed to ethylating agents. mitochondria. Genes Dev. 2007; 21: 3095–109. Environ Mol Mutagen. 1998; 31: 18–31. 32 Wang PH, Yang SF, Tseng CJ, Ying TH, Ko JL, Lin LY. 18 Sawada S, Asakura S, Daimon H, Furihata C. Compari- The role of lipocalin 2 and its concernment with human son of autoradiography, liquid scintillation counting and nonmetastatic clone 23 type 1 and p53 in carcinogenesis immunoenzymatic staining of 5-bromo-2?-deoxyuridine of uterine cervix. Reprod Sci. 2011; 18: 447–55. for measurement of unscheduled DNA synthesis and 33 Kawase T, Ohki R, Shibata T, Tsutsumi S, Kamimura N, replicative DNA synthesis in rat liver. Mutat Res. 1995; Inazawa J, et al. PH domain-only protein PHLDA3 is a 344: 109–16. p53-regulated repressor of Akt. Cell 2009; 136: 535–50. 19 Kawase T, Ichikawa H, Ohta T, Nozaki N, Tashiro F, 34 Strebhardt K. Multifaceted polo-like kinases: drug targets Ohki R, et al. p53 target gene AEN is a nuclear ex- and antitargets for cancer therapy. Nat Rev Drug Discov. onuclease required for p53-dependent apoptosis. On- 2010; 9: 643–60. cogene. 2008; 27: 3797–810. 35 Shen H, Maki CG. p53 and p21(Waf1) are recruited to 20 Lanza F, Bi S. Role of p53 in leukemogenesis of chronic distinct PML-containing nuclear foci in irradiated and myeloid leukemia. Stem Cells. 1995; 13: 445–52. Nutlin-3a-treated U2OS cells. J Cell Biochem. 2010; 111: 21 Rouault JP, Falette N, Guáehenneux F, Guillot C, 1280–90. Rimokh R, Wang Q, et al. Identiˆcation of BTG2, an an- 36 Jacobs SBR, Basak S, Murray JI, Pathak N, Attardi LD. tiproliferative p53-dependent component of the DNA Siva is an apoptosis-selective p53 target gene important damage cellular response pathway. Nat Genet. 1996; 14: for neuronal cell death. Cell Death DiŠer. 2007; 14: 482–6. 1374–85. 22 Gupta S, Radha V, Furukawa Y, Swarup G. Direct tran- 37 Taira N, Yoshida K. Post-translational modiˆcations of scriptional activation of human caspase-1 by tumor sup- p53 tumor suppressor: determinants of its functional tar- pressor p53. J Biol Chem. 2001; 276: 10585–8. gets. Histol Histopathol. 2012; 27: 437–43. 23 Zhao Z, Samuels T, Winckler S, Korgaonkar C, Tom- 38 Ellinger-Ziegelbauer H, Stuart B, Wahle B, Bomann W, pkins V, Horne MC, et al. Cyclin G1 has growth inhibito- Ahr HJ. Characteristic expression proˆles induced by ry activity linked to the ARF-Mdm2-p53 and pRb tumor genotoxic carcinogens in rat liver. Toxicol Sci. 2004; 77: suppressor pathways. Mol Cancer Res. 2003; 1: 195–206. 19–34. 24 Zimmermann M, Arachchige-Don AS, Donaldson MS, 39 Iida M, Anna CH, Holliday WM, Collins JB, Cunnin- Dallapiazza RF, Cowan CE, Horne MC. Elevated cyclin gham ML, Sills RC, et al. Unique patterns of gene expres- G2 expression intersects with DNA damage checkpoint sion changes in liver after treatment of mice for 2 weeks signaling and is required for a potent G2/M checkpoint with diŠerent known carcinogens and non-carcinogens. arrest response to doxorubicin. J Biol Chem. 2012; 287: Carcinogenesis. 2005; 26: 689–99. 22838–53. 40 Kang JS, Wanibuchi H, Murai T, Morimura K, Kinoshita 25 Choi JW, Lee KH, Kim SH, Jin T, Lee BS, Oh J, et al. C- A, Fukushima S. Analysis of gene expression in diŠerent

6363 Mikiya Sakurai et al.

stages of MeIQx-induced rat hepatocarcinogenesis. Oncol of PPARalpha- and KupŠer cell-dependent eŠects of Rep. 2007; 17: 747–52. WY–14,643 in mouse liver using microarray gene expres- 41 Woods CG, Kosyk O, Bradford BU, Ross PK, Burns sion. Toxicol Appl Pharmacol. 2007; 225: 267–77. AM, Cunningham ML, et al. Time course investigation

6464