DNA hypomethylation induced by non-genotoxic carcinogens in mouse and rat colon

The ability of non-genotoxic colon carcinogens to induce DNA hypomethylation was evaluated. Administering 0, 0.2 and 0.4 mg/kg of 5-aza-2'-deoxycytidine to female mice for 5 days resulted in a dose-related decrease in 5-methylcytosine in colon DNA. Rutin (3.0 mg/kg) and five bile acids (4.0 mg/kg) were administered in the diet to male F344 rats for 14 days. Rutin and four bile acids that promote colon cancer, deoxycholic acid, chenodeoxycholic acid, cholic acid and lithocholic acid caused DNA hypomethylation, while ursodeoxycholic acid that prevents colon cancer did not. Bromodichloromethane (BDCM) was administered to male F344 rats and B6C3F1 mice by gavage at 0, 50 and 100 mg/kg or in their drinking water at 0, 350 and 700 mg/l for up to 28 days. In rats, BDCM decreased DNA methylation, being more effective when administered by gavage, correlating to its greater carcinogenic potency by this route. In mice, BDCM did not decrease DNA methylation, corresponding to its lack of carcinogenic activity in the colon of this species. In summary, the ability of non-genotoxic colon carcinogens to cause DNA hypomethylation correlated with their carcinogenic activity in the colon.     

Discrimination of genotoxic from non-genotoxic carcinogens by gene expression profiling

Carcinogenesis, 25,     

Discrimination of genotoxic from non-genotoxic carcinogens by gene expression profiling

Carcinogenesis, 25,     

Discrimination of genotoxic from non-genotoxic carcinogens by gene expression profiling

Two general mechanisms are implicated in chemical carcinogenesis. The first involves direct damage to DNA, referred to as genotoxic (GTX), to which the cell responds by repair of the damages, arrest of the cell cycle or induction of apoptosis. The second is non-DNA damaging, non-genotoxic (NGTX), in which a wide variety of cellular processes may be involved. Therefore, it can be hypothesized that modulation of the underlying gene expression patterns is profoundly distinct between GTX and NGTX carcinogens, and thus that expression profiling is applicable for classification of chemical carcinogens as GTX or NGTX. We investigated this hypothesis by analysing modulation of gene expression profiles induced by 20 chemical carcinogens in HepG2 cells with application of cDNA microarrays that contain 597 toxicologically relevant genes. In total, 22 treatments were included, divided in two sets. The training set consisted of 16 treatments (nine genotoxins and seven non-genotoxins) and the validation set of six treatments (three and three). Class discrimination models based on Pearson correlation analyses for the 20 most discriminating genes were developed with data from the training set, where after the models were tested with all data. Using all data, the correctness for classification of the carcinogens from the training set was clearly better than that for the validation set, namely 81 and 33%, respectively. Exclusion of the treatments that had only marginal effects on the expression profiles, improved the discrimination for the training and validation sets to 92 and 100% correctness, respectively. Exclusion of the gene expression signals that were hardly altered also improved classification, namely to 94 and 80%. Therefore, our study proves the principle that gene expression profiling can discriminate carcinogens with major differences in their mode of actions, namely genotoxins versus non-genotoxins.     

Induction of minisatellite DNA rearrrangements by genotoxic carcinogens in mouse liver tumors

We Investigated whether somatic rearrangements in minisatelliteDNA are more frequent in chemically induced mouse liver tumorsthan they are in spontaneous tumors. CD-1 mouse liver tumorswere induced by either a single dose or 15 consecutive dailydoses of 7,12-dimethylbenz[a]anthracene, 4-aminoazobenzene,N-hydroxy-2-acetyl-aminofluorene or diethylnitrosamine (DEN).Using DNA fingerprinting analysis, we found that the single-and multiple-dose carcinogen treatments caused a 2- to 5-foldhigher frequency of minisatellite DNA rearrangements comparedwith that found in spontaneous tumors-with the exception ofsingle-dose DEN tumors, which showed no increase in rearrangements.Our results suggest that DNA fingerprinting may be a valuahleassay for differentiating certain chemically induced tumorsfrom spontaneous tumors.     

Physicochemical alterations in the conformation of rat liver chromatin induced by carcinogens in vivo.

Administration of methylating carcinogens such as methyl methanesulfonate (120 mg/kg), dimethylnitrosamine (5 mg/kg), or methylnitrosourea (80 mg/kg) to rats resulted in an increased ellipticity in circular dichroism spectra and in an enhanced ability to bind ethidium bromide in the liver chromatin. Although shearing of the chromatin preparations increased both the ellipticity and number of binding sites for ethidium bromide, the carcinogen-induced effects were noticeable whether or not chromatin was sheared. Although the doses of the 3 carcinogens used in these studies are equivalent in their ability to induce strand breaks in liver DNA at 4 hr, their effects on the induction of conformational changes in liver chromatin are different. For example, methyl methanesulfonate induced the minimum conformational changes in liver chromatin at 4 hr, whereas methylnitrosourea induced the maximum changes at 4 hr. Methyl methanesulfonate and dimethylnitrosamine, on the other hand, induced maximum changes at 3 days. The conformational changes induced by methyl methanesulfonate and methylnitrosourea, and not by dimethylnitrosamine, tend to be repaired by 14 days.     

The possible mechanism of enhanced carcinogenesis induced by genotoxic carcinogens in rasH2 mice

Microarray and RT-PCR analyses were performed for the transgene and Ras-related genes in forestomach squamous cell carcinomas (SCCs) induced by 7,12-dimethylbenz[a]anthracene (DMBA) in rasH2 mice; these results were compared with our previous molecular data of N-ethyl-N-nitrosourea-induced forestomach SCCs and urethane-induced lung adenomas in rasH2 mice. Overexpression of the transgene was detected in the DMBA-induced SCCs, suggesting that the transgene plays an important role in enhanced carcinogenesis in rasH2 mice. In addition, the mouse endogenous ras genes were up-regulated in the DMBA-induced SCCs, and are probably involved in the tumorigenesis of forestomach SCCs. Genes such as osteopontin, Cks1b, Tpm1, Reck, gelsolin, and amphiregulin that were commonly altered in these three different carcinogen-induced tumors may contribute to the development of tumors in rasH2 mice.     

The colonic response to genotoxic carcinogens in the rat: regulation by dietary fibre

The apoptotic response to DNA damage appears to be an innate biological mechanism for protection against tumourigenesis. It is possible that agents that protect against colorectal cancer act by enhancing the apoptotic deletion of cells suffering DNA damage, with consequent removal of those with tumourigenic mutations. We examined the acute apoptotic response to genotoxic carcinogens ("AARGC") in colonic epithelium and the possibility that dietary fibres of different fermentability might regulate AARGC. To fully define the time-course and nature of AARGC in response to the carcinogen azoxymethane (AOM), a single injection of AOM (10 mg/kg) was given to rats and apoptosis monitored in the colon by light microscopy and terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labelling staining over a 72 h period. Having defined the site and time of maximum response, two groups of eight rats were fed diets containing 10% wheat bran fibre (WB; fermentable) or 10% methylcellulose (MC; poorly fermentable) for 4 weeks. Colonic AARGC was compared by light microscopy; lumenal short chain fatty acids (SCFAs) and pH were measured as indicators of the fermentative environment. AOM-induced AARGC was maximal at 8 h and greater in distal compared with proximal colon. Apoptotic cells were situated predominantly in the lower half of the crypt, with the median at position 9 indicating involvement of daughter as well as stem cells. There was no "second wave" of apoptosis within 72 h as follows irradiation in small intestine. Distal colonic AARGC in rats fed WB was twice that in rats fed MC (P < 0.01). Compared with MC, WB significantly lowered lumenal pH and increased all SCFAs including butyrate, while proliferation did not differ between the fibres. Certainly, dietary fibres can regulate AARGC and further studies are warranted to determine if this biological effect is the way in which dietary factors regulate tumourigenesis. Lumenal generation of butyrate may enhance AARGC as butyrate is proapoptotic in vitro.     

Rat and mouse forestomach tumors induced by chronic oral administration of styrene oxide

Styrene oxide (CAS: 96-09-3) was administered in corn oil by gavage three times a week at two dose levels to groups of 52 male and 52 female F344 rats and 52 male and 52 female B6C3F1 mice for 2 years, after which the surviving animals were killed and examined histopathologically. The doses given to rats were 550 and 275 mg/kg (body wt) per treatment; in mice the two doses were 750 and 375 mg/kg (body wt) per treatment. The main pathologic findings were high incidences of squamous cell carcinomas or papillomas of the forestomach in both sexes of both rats and mice. These neoplasms were virtually absent from the 52 controls of either sex of both species given parallel treatment with corn oil alone. There was a statistically significant increase in the incidence of hepatocellular neoplasms in male mice receiving 375 mg styrene oxide/kg. Styrene oxide is carcinogenic to both sexes of F344 rats and B6C3F1 mice when administered orally in corn oil.     

Changes in global gene and protein expression during early mouse liver carcinogenesis induced by non-genotoxic model carcinogens oxazepam and Wyeth-14,643

We hypothesized that the mouse liver tumor response to non-genotoxic carcinogens would involve some common early gene and protein expression changes that could ultimately be used to predict chemical hepatocarcinogenesis. In order to identify a panel of genes to test, we analyzed global differences in gene and protein expression in livers from B6C3F1 mice following dietary treatment with two rodent carcinogens, the benzodiazepine anti-anxiety drug oxazepam (2500 p.p.m.) and the hypolipidemic agent Wyeth (Wy)-14,643 (500 p.p.m.) compared with livers from untreated mice. Male mice were exposed for 2 weeks and 1, 3 or 6 months to oxazepam or Wy-14,643 in an age-matched study design. By histopathological evaluation, no liver preneoplastic foci or tumors were detected at 6 months in treated or control groups. By cDNA microarray analysis [NIEHS Mouse Chip (8700 genes); n = 3 individual livers/group, four hybridizations/sample], expression of 36 genes or 220 genes were changed relative to control livers following 6 months of oxazepam or Wy-14,643 treatment, respectively. To obtain a more comprehensive picture of gene/protein expression changes, we also conducted a proteomics study by 2D-gel electrophoresis followed by matrix assisted laser desorption/ionization-mass spectrometry on cytoplasmic, nuclear, and microsomal subcellular fractions of the same liver samples utilized for the cDNA microarray analysis. Real-time PCR, western blot analysis and immunohistochemistry were utilized for validation and to expand the results to other time points. Cyp2b20, growth arrest- and damage-inducible gene beta (Gadd45beta), tumor necrosis factor alpha-induced protein 2 and insulin-like growth factor binding protein 1 (Igfbp5) genes and proteins were upregulated by oxazepam, and Cyp2b20, Cyclin D1, proliferating cell nuclear antigen, Igfbp5, Gadd45beta and cell death-inducing DNA fragmentation factor alpha subunit-like effector A exhibited higher expression after Wy-14,643 treatment. Most of these genes/proteins were also deregulated at 2 weeks. There appeared to be more distinct than common changes in the expression of carcinogenesis-related genes/proteins between the two compounds, suggesting that the major carcinogenic pathways are different for these compounds and may be distinct for different chemical classes.     

5-Azacytidine potentiates initiation induced by carcinogens in rat liver

To test the validity of the hypothesis that hypomethylationof DNA plays an important role in the initiation of carcinogenicprocess, 5-azacytidine (5-AzC) (10 mg/kg), an inhibitor of DNAmethylation, was given to rats during the phase of repair synthesisinduced by the three carcinogens, benzo[a]-pyrene (200 mg/kg),N-methyl-N-nltrosourea (60 mg/kg) and 1,2-dimethylhydrazine(1,2-DMH) (100 mg/kg). The initiated hepatocytes in the liverwere assayed as the -glutamyltransferase (-GT) positive fociformed following a 2-week selection regimen consisting of dietary0.02% 2-acetylaminofluorene coupled with a necrogenic dose ofCCl₄. The results obtained indicate that with all three carcinogens,administration of 5-AzC during repair synthesis increased theincidence of initiated hepatocytes, for example 10–20foci/cm² in 5-AzC and carcinogen-treated rats compared with3–5 foci/cm² in rats treated with carcinogen only. Administrationof [³H]-5-azadeoxycytidine during the repair synthesis inducedby 1,2-DMH further showed that 0.019 mol% of cytosine residuesin DNA were substituted by the analogue, indicating that incorporationof 5-AzC occurs during repair synthesis. In the absence of thecarcinogen, 5-AzC given after a two thirds partial hepatectomy,when its incorporation should be maximum, failed to induce any-GT positive foci. The results suggest that hypomethylationof DNA per se may not be sufficient for initiation. Perhapstwo events might be necessary for initiation, the first causedby the carcinogen and a second involving hypomethylation ofDNA.     

32P-postlabeling analysis of DNA adducts in rat livers after treatment with genotoxic and non-genotoxic 4-aminoazobenzene derivatives

Formation of hepatic DNA adducts was studied in rats following intraperitoneal administration of a hepatocarcinogen, 3-methoxy-4-aminoazobenzene (3-MeO-AAB) and a non-hepatocarcinogen, 2-methoxy-4-aminoazobenzene (2-MeO-AAB). The 32P-post-labeling assay revealed 3-MeO-AAB to give more than 20-fold higher amounts of DNA adducts than did 2-MeO-AAB. Furthermore, five adducts, one of which accounted for over 70% of the total modified bases, were found in DNA from 3-MeO-AAB-treated rats, whereas only one adduct was apparent in 2-MeO-AAB-treated DNA. Our data thus suggested that the difference in hepatocarcinogenic activity between 3-MeO-AAB and 2-MeO-AAB might be, at least in part, dependent on quantitative and qualitative differences in their azo dye-DNA adduct formation in the rat liver.     

Failure of genotoxic carcinogens to produce tumors in human skin xenografts transplanted to SCID mice

Chemical carcinogenesis of human skin was investigated usinghuman skin xenografts (16 full thickness and 48 split thicknessskin grafts) transplanted to CB-17-scid (SCID) mice. Topicalapplication of a carcinogen, i.e. 7, 12-dimethyl-benz[a]anthracene(DMBA), benzo[a]pyrene, methylchol-anthrene or N-methyl-N'-nitro-N-nitrosoguanidine,to the human skin xenografts once a week for 25–30 weeksfailed to produce skin tumors. Both DMBA application plus UV-Birradiation and alternate applications of the above four carcinogensin combination with UV-B irradiation also failed to producetumors. All of these treatments induced skin papillomas in skinsof host SCID mice. DMBA induced skin papillomas in allogenicCD-1 mouse skin grafts transplanted to SCID mice. These resultsindicate that susceptibility of human skin to these carcinogenicstimuli is much lower than that of mouse skin.     

Absence of p53 mutations and various frequencies of ki-ras exon 1 mutations in rat hepatic tumors induced by different carcinogens

Mutations of p53 and Ki-ras exon 1 were investigated in rat hepatic lesions induced by four kinds of hepatocarcinogenic protocols: continuous feeding of 3′-methyl-4-dimethylaminoazobenzene (3′-Me-DAB), daily intraperitoneal injection of aflatoxin B, (AFB,), and the Soft and Farber regimen (Nature 236:701–703, 1976), in which diethylnitrosamine (DEN) or nitrosomethylurea (NMU) was used as initiating agents. DNA from microdissected tissue sections was amplified by the polymerase chain reaction (PCR) directly using primers for p53 exons 5–8 and Ki-ras exon 1 and analyzed for mutations by denaturing gradient gel electrophoresis (DGGE) or constant denaturant gel electrophoresis (CDGE). One or both of the p53 PCR primers were located within introns to prevent amplifying the p53 processed pseudogenes. In a total of 59 hepatocellular carcinomas (HCCs), no p53 aberrations were detected, indicating that p53 mutations are not very important in rat hepatic carcinogenesis. On the other hand, Ki-ras codon 12 mutations were found at low frequency in HCCs, hyperplastic foci, and cholangiof ibroses induced by 3′-Me-DAB and by AFB! but not in the lesions induced by the Solt and Farber regimen. Although Ki-ras codon 12 mutations are generally infrequent in rat hepatic tumors, their incidence thus appears to vary depending on the carcinogen used for their induction. © 1994 Wiley-Liss, Inc.     

The histogenesis of mammary tumours induced in the rat by chemical carcinogens

126 Wistar female rats were given carcinogenic compounds, 2-acetylaminofluorene (2-AAF), 20-methylcholanthrene (MC) and 7,12-dimethylbenz(a)anthracene (DMBA), for induction of breast tumors to determine the site of origin and biological behavior of mammary carcinoma in the rat. 18 virgin rats served as control material of a corresponding age to the tumor-bearing rats. A dissecting microscope was used to identify early neoplastic lesions in the breast glands for histological study. 17 weeks after carcinogen treatment, breast gland tumors, especially adenocarcinomas became evident by palpation. Mammary carcinoma was found to be multifocal in origin, arising almost exclusively in ducts, ductules, and end-buds without the intervention of any recognizable lesion preceding the appearance of tumor cells. Fibroadenomas and adenomas occurred in rats with generally well-developed breast glands, while mammary carcinomas occurred in younger rats with less developed breast glands. For successful tumor induction, the important factors to consider are the architectural state of the mammary gland, the action of hormonal factors, genetically-determined susceptibility, and the carcinogenic agent and its dosage. Mast cells did not play a significant role in tumor formation, nor were significant myoepithelial cell contributions found in the early carcinomas and fibroadenomas.     

Carcinogenicity of dimethylarsinic acid in male F344 rats and genetic alterations in induced urinary bladder tumors

Arsenic is a well-documented human carcinogen, and contamination with this heavy metal is of global concern, presenting a major issue in environmental health. However, the mechanism by which arsenic induces cancer is unknown, in large part due to the lack of an appropriate animal model. In the present set of experiments, we focused on dimethylarsinic acid (DMA), a major metabolite of arsenic in most mammals including humans. We provide, for the first time, the full data, including detailed pathology, of the carcinogenicity of DMA in male F344 rats in a 2-year bioassay, along with the first assessment of the genetic alteration patterns in the induced rat urinary bladder tumors. Additionally, to test the hypothesis that reactive oxygen species (ROS) may play a role in DMA carcinogenesis, 8-hydroxy-2'-deoxyguanosine (8-OHdG) formation in urinary bladder was examined. In experiment 1, a total of 144 male F344 rats at 10 weeks of age were randomly divided into four groups that received DMA at concentrations of 0, 12.5, 50 and 200 p.p.m. in the drinking water, respectively, for 104 weeks. From weeks 97-104, urinary bladder tumors were observed in 8 of 31 and 12 of 31 rats in groups treated with 50 and 200 p.p.m. DMA, respectively, and the preneoplastic lesion, papillary or nodular hyperplasias (PN hyperplasia), was noted in 12 and 14 rats, respectively. DMA treatment did not cause tumors in other organs and no urinary bladder tumors or preneoplastic lesions were evident in the 0 and 12.5 p.p.m.-treated groups. Urinary levels of arsenicals increased significantly in a dose-responsive manner except for arsenobetaine (AsBe). DMA and trimethylarsine oxide (TMAO) were the major compounds detected in the urine, with small amounts of monomethylarsonic acid (MMA) and tetramethylarsonium (TeMa) also detected. Significantly increased 5-bromo-2'-deoxyuridine (BrdU) labeling indices were observed in the morphologically normal epithelium of the groups treated with 50 and 200 p.p.m. DMA. Mutation analysis showed that DMA-induced rat urinary bladder tumors had a low rate of H-ras mutations (2 of 20, 10%). No alterations of the p53, K-ras or beta-catenin genes were detected. Only one TCC (6%) demonstrated nuclear accumulation of p53 protein by immunohistochemistry. In 16 of 18 (89%) of the TTCs and 3 of 4 (75%) of the papillomas, decreased p27(kip1) expression could be demonstrated. Cyclin D1 overexpression was observed in 26 of 47 (55%) PN hyperplasias, 3 of 4 (75%) papillomas, and 10 of 18 (56%) TCCs. As a molecular marker of oxidative stress, increased COX-2 expression was noted in 17 of 18 (94%) TCCs, 4 of 4 (100%) papillomas, and 39 of 47 (83%) PN hyperplasias. In experiment 2, 8-OHdG formation in urinary bladder was significantly increased after treatment with 200 p.p.m. DMA in the drinking water for 2 weeks compared with the controls. The studies demonstrated DMA to be a carcinogen for the rat urinary bladder and suggested that DMA exposure may be relevant to the carcinogenic risk of inorganic arsenic in humans. Diverse genetic alterations observed in DMA-induced urinary bladder tumors imply that multiple genes are involved in stages of DMA-induced tumor development. Furthermore, generation of ROS is likely to play an important role in the early stages of DMA carcinogenesis.