Transgenic tumor models for carcinogen identification: the heterozygous Trp53-deficient and RasH2 mouse lines.

Genetically altered mouse models (GAMM) for human cancers have been critical to the investigation and characterization of oncogene and tumor suppressor gene expression and function and the associated cancer phenotype. Similarly, several of the mouse models with defined genetic alterations have shown promise for identification of potential human carcinogens and investigation of mechanisms of carcinogen-gene interactions and tumorigenesis. In particular, both the B6.129N5-Trp53 mouse, heterozygous for a p53 null allele, and the CB6F1-RasH2 mouse, hemizygous for the human H-ras transgene, have been extensively investigated. Using 26-week exposure protocols at or approaching the maximum tolerated dose, the summary results to date indicate the potential for GAMM to identify and, possibly, classify chemicals of potential risk to humans using short-term carcinogenicity experiments. This IWGT session focused on: (1) the development of recommendations for genetic/molecular characterization required in animals, tissues, and tumors before and after treatment for identification of presumptive human carcinogens based on the current state of knowledge, (2) identification of data gaps in our current state of knowledge, and (3) development of recommendations for research strategies for further development of our knowledge base of these particular models. By optimization of protocols and identification of significant outcomes and responses to chemical exposure in appropriate short-term mechanism-based genetically altered rodent models, strategies for prevention and intervention may be developed and employed to the benefit of public health.

Detection of DNA adducts using a quantitative long PCR technique and the fluorogenic 5' nuclease assay (TaqMan).

The detection of DNA adducts is an important component in assessing the mutagenic potential of exogenous and endogenous compounds. Here, we report an in vitro quantitative long PCR (XL-PCR) assay to measure DNA adducts in human genomic DNA based on their ability to block and inhibit PCR amplification. Human genomic DNA was exposed to test compounds and then a target sequence was amplified by XL-PCR. The amplified sequence was then quantified using fluorogenic 5' nuclease PCR (TaqMan) and normalized to a solvent-treated control. The extent of DNA adduction was determined based on the reduction in amplification of the target sequence in the treated sample. A 17.7kb beta-globin fragment was chosen as the target sequence for these studies, since preliminary experiments revealed a two-fold increased sensitivity of this target compared to a 10.4kb HPRT fragment for detecting hydrogen peroxide-induced DNA damage. Validation of the XL-PCR assay with various compounds demonstrated the versatility of the assay for detecting a wide range of adducts formed by direct acting or S9-activated mutagens. The same DNA samples were also analyzed using 32P-postlabeling techniques (thin-layer chromatography or high-performance liquid chromatography) to confirm the presence of DNA adducts and estimate their levels. Whereas 32P-postlabeling with nuclease P(1) enrichment was more sensitive for detecting bulky adducts induced by the compounds benzo[a]pyrene, dimethylbenzanthracene, 3-methylindole, indole 3-carbinol, or 2-acetylaminofluorene, the XL-PCR procedure was more sensitive for detecting smaller or labile DNA adducts formed by the compounds methyl methanesulfonate, diethyl nitrosamine, ethylnitrosourea, diepoxybutane, ICR-191, styrene oxide, or aflatoxin B(1). Compounds not expected to form adducts in DNA, such as clofibrate, phenobarbital, chloroform or acetone, did not produce a positive response in the XL-PCR assay. Thus, quantitative XL-PCR provides a rapid, high-throughput assay for detecting DNA damage that complements the existing 32P-postlabeling assay with nuclease P(1) enrichment.

The nature of the heterozygous Trp53 knockout model for identification of mutagenic carcinogens.

The heterozygous Trp53 null allele C57BL/6 (N5) mouse is susceptible to the rapid development of neoplasia by mutagenic carcinogens relative to control strains. This mouse model of chemical carcinogenesis demonstrates 1) dose-related rapid induction of tumors (26 wks), 2) multiple sites of carcinogen-specific tissue susceptibility, and 3) carcinogen-induced loss of heterozygosity involving the Trp53 wild-type allele or a p53 haploinsufficiency permitting mutation of other critical protooncogenes and/or inactivation of tumor suppressor genes driving tumorigenesis. Demonstration of mutation or loss of heterozygosity involving the Trp53 locus is consistent with a common finding in human cancers and supports extrapolation between rodents and humans. Using diverse experimental protocols, almost all mutagenic rodent carcinogens (including all mutagens that are carcinogenic to humans), but not nonmutagenic rodent carcinogens, induce tumors within 26 weeks of continuous exposure. These characteristics and results indicate that the mouse heterozygous for the Trp53 null allele may be of significant use for the prospective identification of mutagenic carcinogens of potential risk to human health.

P53+/- hemizygous knockout mouse: overview of available data.

The performance of the p53-/- transgenic (knockout) mouse model was evaluated through review of the data from 31 short-term carcinogenicity studies with 21 compounds tested as part of the International Life Sciences Institute's (ILSI) Alternatives to Carcinogenicity Testing (ACT) project, together with data from other studies which used comparable protocols. As expected based on the hypothesis for the model, a significant number (12/16 or 75%) of the genotoxic human and/or rodent carcinogens tested were positive and the positive control, p-cresidine, gave reproducible responses across laboratories (18/19 studies positive in bladder). An immunosuppressive human carcinogen, cyclosporin A, was positive for lymphomas but produced a similar response in wild type mice. Two hormones that are human tumorigens, diethylstilbestrol and 17beta-estradiol, gave positive and equivocal results, respectively, in the pituitary with p53-deficient mice showing a greater incidence of proliferative lesions than wild type. None of the 22 nongenotoxic rodent carcinogens that have been tested produced a positive response but 2 compounds in this category, chloroform and diethylhexylphthalate, were judged equivocal based on effects in liver and kidney respectively. Four genotoxic noncarcinogens and 6 nongenotoxic, noncarcinogens were also negative. In total (excluding compounds with equivocal results), 42 of 48 compounds or 88% gave results that were concordant with expectations. The technical lessons learned from the ILSI ACT-sponsored testing in the p53+/- model are discussed.

Current status and use of short/medium term models for carcinogenicity testing of pharmaceuticals--scientific perspective.

Short- and medium-term rodent bioassays have been proposed under ICH guidelines for use in testing for the carcinogenic potential of pharmaceuticals. Further evaluation of these models is needed urgently and coordinated efforts are in progress worldwide to expand the available database. Models currently being investigated include transgenic mice (Tg-rasH2, Tg.AC, p53(+/-), XPA(-/-)) and neonatal mice. As more data become available on the performance of these assays, regulatory and industry scientists will be faced with the difficult challenge of determining how the performance (accuracy) of each assay will be measured and deciding which assays have value in the risk assessment process.

Occupational hazard evaluation of p-bromobenzyl bromide from tests for genotoxicity.

As part of an occupational hazard evaluation, p-bromobenzyl bromide (p-BBB) was evaluated for genotoxic activity in the Ames microbial mutagenicity assay, the alkaline elution assay for DNA strand breaks in rat hepatocytes and the in vitro chromosome aberration assay in Chinese hamster ovary cells. The compound produced equivocal results in the microbial mutagenicity assay but was negative in the alkaline elution assay for DNA strand breaks in rat hepatocytes. The compound produced weakly positive results in the in vitro chromosome aberration assay. There was substantial cytotoxicity in all three assays. It is concluded that p-BBB is weakly genotoxic.

Evaluation of the Emu-pim-1 transgenic mouse model for short-term carcinogenicity testing.

The value of the chronic rodent carcinogenicity assay in adequately predicting cancer risk in humans has become a matter of debate over the past few years. Therefore, more rapid and accurate alternative tests are urgently needed. Transgenic mouse models, those harboring genetic changes that are relevant to the multistage cancer process, may provide such alternative tests. Transgenic Emu-pim-1 mice, developed by Berns and coworkers in 1989, contain the pimn-1 oncogene, which is expressed at elevated levels in their lymphoid compartments. As a result, these mice are predisposed to the development of T-cell lymphomas. Because of the low incidence of spontaneous tumors and the increased sensitivity to N-ethyl-N-nitrosourea-induced carcinogenesis, Emu-pim-1 mice were suggested to be one of the first potential and attractive candidates to be used in short-term carcinogenicity testing. In the present article, the results from 2 recent short-term assays (with mitomycin C and x-rays) are briefly presented, together with a review of all 11 performed bioassays and their corresponding histopathologic and molecular data. The overall results allow the first evaluation of the Emu-pim-1 mouse model with regard to its usefulness in short-term carcinogenicity testing. It has been shown that the model is primarily suitable as a sensitive short-term assay for genotoxic carcinogens that not only induce (at least) gene mutations and/or large deletions and rearrangements but that also sufficiently target the lymphoid system. However, the Emu-pim-1 mice lack sufficient sensitivity to justify their routine use in short-term carcinogenicity testing in general.

The mouse lymphoma L5178Y Tk+/- cell line is heterozygous for a codon 170 mutation in the p53 tumor suppressor gene.

The p53 tumor suppressor protein plays an important role in regulating the cellular response to DNA damage, including cell cycle arrest and apoptosis induction. Normal p53 function is critical for the maintenance of genomic stability. The mouse lymphoma L5178Y/TK(+/-)-3.7.2C cell line is widely used in genetic toxicology for mutagenesis and clastogenesis testing. A related line L5178Y-R, has previously been shown to react with antibodies specific for mutant as well as wild-type p53 protein and to exhibit delayed cell death after radiation. For this reason, as well as the mouse lymphoma assay's reputation for high sensitivity of detection for genotoxic agents but low specificity, we examined several clones of L5178Y cells for mutations in the conserved core domain (exons 5-8) of the p53 gene. Using single-strand conformational polymorphism analysis, we found evidence for the same mutation in exon 5 of p53 in L5178Y-R, L5178Y-S and L5178Y/TK(+/+)-3.7.2C cells. The mutation was identified by sequencing of exon 5 as a TGC (Cys) to CGC (Arg) transition in codon 170 (= codon 176 in humans). Sequencing showed approximately equivalent signals for the mutant and normal alleles for all 3 lines. The mutation in codon 170 is adjacent to a mutation hotspot of the human p53 gene (codon 175) and eliminates a critical zinc-coordinating cysteine residue such that the mutant protein is likely to be denatured and have a dominant negative effect on normal p53 function. Western blots showed approximately 100-fold higher levels of p53 protein in unirradiated L5178Y cells as compared to induced levels of p53 in normal mouse splenocytes 4 h after 5 Gy of gamma radiation. The high levels of p53 protein in L5178Y cells were not further inducible by radiation, whereas an 11-fold induction was seen in the irradiated splenocytes. These results indicate that p53 protein in L5178Y cells is dysfunctional and suggest that this line may therefore be abnormally susceptible to the induction of genetic alterations.

Detection of DNA damage induced by human carcinogens in acellular assays: potential application for determining genotoxic mechanisms.

Positive outcomes of in vitro genotoxicity tests may not always occur as a consequence of direct reaction of a compound or a metabolite with DNA. To follow-up positive responses in in vitro tests, we developed two supplemental, cell-free assays to examine the potential of compounds and metabolites to directly damage DNA. Calf thymus DNA was used as the target for the direct detection of adducts by 32P-postlabeling/TLC and electrochemical detection, and alkaline gel electrophoresis was used to detect single-strand breakage of bacteriophage lambda DNA. To show that these assays would detect damage from relevant compounds, we examined nine human carcinogens (aflatoxin B1, busulfan, chlorambucil, cyclophosphamide, diethylstilbestrol, melphalan, 2-naphthylamine, phenacetin and potassium chromate). Each of the nine compounds produced a positive result for one or both endpoints. Using multifraction contact-transfer TLC, we detected 32P-labeled DNA adducts produced by aflatoxin B1, chlorambucil, diethylstilbestrol, melphalan, 2-naphthylamine, and potassium chromate (plus hydrogen peroxide). Aflatoxin B1, diethylstilbestrol and 2-naphthylamine required metabolic activation (induced rat liver S9) to generate DNA adducts. Although potassium chromate alone induced a slight increase in the content of 8-hydroxydeoxyguanosine (a promutagenic adduct produced by reactive oxygen species), addition of hydrogen peroxide greatly increased 8-hydroxydeoxyguanosine levels. The damage to lambda DNA by each human carcinogen (or metabolites), except diethylstilbestrol, was sufficient to generate single-strand breaks after neutral thermal hydrolysis at 70 degrees C. Chromate was a weak inducer of DNA fragmentation, but adding hydrogen peroxide to the reaction mixtures dramatically increased the DNA strand breakage. Our data suggest that these non-routine, acellular tests for determining direct DNA damage may provide valuable mechanistic insight for positive responses in cell-based genetic toxicology tests.

Revalidation of the in vitro alkaline elution/rat hepatocyte assay for DNA damage: improved criteria for assessment of cytotoxicity and genotoxicity and results for 81 compounds.

The in vitro alkaline elution/rat hepatocyte assay is a sensitive assay for genotoxicity, measured as DNA strand breaks induced in primary cultures of rat hepatocytes after 3-h treatments with test compounds. Since DNA degradation can be rapid and extensive in dead and/or dying cells, the original criteria for a positive result in the assay were that a compound induce a 3.0-fold or greater increase in the elution slope (for the terminal phase of alkaline elution from 3 to 9 h) in the absence of significant cytotoxicity (defined as relative cell viability of less than 70% by trypan blue dye exclusion; TBDE). Recently we have shown that false-positive results can still be obtained due to cytotoxicity when loss of membrane integrity is a late event in toxic cell death relative to the induction of endonucleolytic DNA degradation. To improve the ability of the assay to discriminate between genotoxic vs. cytotoxic effects of chemicals, we have evaluated additional assays of cytotoxicity including cell adenosine triphosphate (ATP) and potassium (K+) content, tetrazolium dye reduction (MTT), TBDE after a further 3-h recovery incubation without test chemicals (delayed toxicity), cell blebbing and endonucleolytic DNA degradation (double-strand breaks; DSBs) assessed by pulsed-field gel electrophoresis (PFGE). We have also evaluated 2 parameters derived from the elution data which can indicate extensive, cytotoxicity-induced DNA degradation: the fraction of the DNA recovered in the neutral lysis/rinse fraction and the gamma-intercept of the extrapolation of the 3-9-h segment of the elution curve. Twenty-eight rodent non-carcinogens that are negative (or inconclusive) in the Ames assay with no, or limited, other evidence of genotoxicity, and 33 genotoxins, most of which are also carcinogens, were evaluated. The results showed that DNA degradation as measured by a 1-h PACE (Programmed Autonomously Controlled Electrodes)/PFGE assay was a sensitive indicator of cytotoxicity which correlated well with results of the other cytotoxicity indicators. The delayed TBDE (after a 3-h recovery), intracellular potassium and ATP assays as well as the gamma-intercept parameter were also shown to be sensitive and in some cases complementary measures of cytotoxicity. Using new criteria based on these data of an induced slope (treatment slope-negative control slope) of 0.020 for the 3- to 9-h elution period and cytotoxicity limits of 70% relative viability for the delayed TBDE assay and 50% for intracellular ATP content, the assay scores the genotoxicity of these 61 reference compounds with an overall accuracy of 92%. Test results using these new criteria are provided for an additional 20 compounds (5 non-genotoxic carcinogens and 15 compounds whose genotoxic and carcinogenic potential are unknown or equivocal).

Rapid induction of hyperplasia in vitro in rat bladder explants by elevated sodium ion concentrations and alkaline pH.

The effects of alkaline pH and elevated sodium concentrations in culture medium on rat bladder explants for 1, 2, and 3 weeks were investigated by continuous BrdU labeling and histopathology. Increasing the sodium chloride concentration of normal medium by 50 or 100 mM caused slight urothelial hyperplasia with statistically significant increases in labeling in week 2 (50 mM) and at all time points with 100 mM NaCl. Cytotoxicity was seen in the high salt group. Increasing the pH from 7.2 to 7.8 and 8.2 also caused a slight hyperplastic response with significant increases in labeling and cytotoxicity at pH 8.2. However, bladder explants treated at pH 7.8 or 8.2 with excess sodium concentrations of 50 to 100 mM had a more marked hyperplastic response with evidence of cytotoxicity as well. There were significant increases in the labeling index (6.4- to 15.0-fold relative to control) after 1 week, with the maximum response at 100 mM sodium/pH 8.2. These results suggest that alkaline pH and elevated sodium concentration have a direct mitogenic effect on rat urothelial cells with some cytotoxicity-induced regenerative cell proliferation as well. These in vitro results in an organ culture system are in agreement with in vivo studies that have shown an important role for elevated urinary cation concentrations and pH in the stimulation of DNA synthesis, induction of hyperplasia, and tumor promotion in rat bladder epithelium.

Extent and persistence of streptozotocin-induced DNA damage and cell proliferation in rat kidney as determined by in vivo alkaline elution and BrdUrd labeling assays.

The extent of DNA damage and cellular proliferation induced in rat kidneys by single doses of the diabetogenic alkylating agent streptozotocin (STZ) and the time course of repair of that damage were evaluated using an in vivo alkaline elution assay for DNA strand breaks and a bromodeoxyuridine (BrdUrd) labeling assay for cell replication. Male Sprague-Dawley rats were given iv injections of 0.25 to 60 mg/kg STZ and kidneys were harvested 3 hr later for alkaline elution. A dose of 2.5 mg/kg STZ was the lowest dose to induce detectable DNA strand breaks and extensive damage was produced by the commonly used diabetogenic dose of 60 mg/kg. To characterize the repair of the drug-induced DNA damage, kidneys were harvested from a 60 mg/kg group of animals 3 hr to 27 days after dosing. BrdUrd-labeled kidney sections were also evaluated to assess any cellular proliferative response associated with STZ administration. Significant DNA damage was detected up to 14 days after dosing with return to near background levels by 20 days. Similarly, treatment with 60 mg/kg STZ was associated with increases in BrdUrd labeling indices 4 and 9 days after treatment with resolution by 27 days. These results indicate that the cellular and molecular repair responses to a single diabetogenic dose of STZ are prolonged, requiring up to 3 weeks to complete. Thus, to avoid potential additive or synergistic effects on STZ-induced nephrotoxicity and/or genotoxicity, a delay in the start of experimental therapies in this model (other than insulin) should be considered.

Short-term carcinogenesis bioassay of genotoxic procarcinogens in PIM transgenic mice.

E mu-pim-1 transgenic mice, which overexpress the pim-1 oncogene in lymphoid tissues, have shown increased susceptibility to induction of T cell lymphomas by N-ethyl-N-nitrosourea, a direct-acting chemical carcinogen (Nature, 340, 61-63, 1989). We sought to further evaluate E mu-pim-1 transgenic mice as a potential test animal for a short-term carcinogenesis bioassay. We chose to test four genotoxic procarcinogens; 2-acetylaminofluorene (2-AAF), N-nitro-sodiethylamine (NDEA), 1,2-dichloroethane (1,2-DCE) and benzene (BEN). These compounds require metabolic activation and, with the exception of benzene, are not mouse lymphomagens. Compounds were administered by gavage daily for 38 (NDEA and 2-AAF) or 40 (BEN and 1,2-DCE) weeks to groups of 25-29 male and female PIM mice at 1 and 3 mg/kg for NDEA, 50 and 100 mg/kg for BEN, 25-100 mg/kg for 2-AAF and 100-300 mg/kg for 1,2-DCE. Small but statistically significant increases in the incidence of malignant lymphoma were seen for three of the four carcinogens tested; in high dose males treated with 2-AAF, high and low dose females treated with NDEA and high dose females treated with 1,2-DCE. Results for BEN, the only mouse lymphomagen tested, did not show a statistically significant increase in the incidence of malignant lymphomas in transgenic mice within the 40 week duration of the study. NDEA also produced a high incidence (> 70%) of hepatic hemangiosarcomas in both sexes at the low and high dose levels. These results demonstrate that over-expression of the pim-1 oncogene in lymphoid tissue can confer susceptibility of this tissue to chemical carcinogenesis by genotoxic procarcinogens. However, whereas potent genotoxic carcinogens produced only small increases in the incidence of lymphoma and since BEN, a mouse lymphomagen, was negative, PIM transgenic mice may lack sufficient sensitivity to established carcinogens to justify their routine use in a short-term carcinogenesis screening assay.

Activation of H-ras is prevalent in 1,3-butadiene-induced and spontaneously occurring murine Harderian gland tumors.

Treatment of B6C3F1 mice with concentrations of 62.5-625 p.p.m. 1,3-butadiene by inhalation for up to 2 years causes a significantly increased incidence of Harderian gland (HG) neoplasms over untreated controls (Melnick,R., Huff,J., Chou,B.J. and Miller,R.A. Cancer Res., 50, 6592-6599, 1990). Since a specific K-ras mutation (codon 13 GGC-->CGC) had previously been described in lung and liver tumors from 1,3-butadiene-treated B6C3F1 mice, we analyzed 23 adenomas and six adenocarcinomas of the HG from mice exposed to 1,3-butadiene for this mutation and mutations in the H-ras gene. We also examined ras activation in 16 spontaneously occurring HG adenomas and one adenocarcinoma. DNA samples were prepared from paraffin-embedded tissues and analyzed by PCR followed by direct sequencing methods. Only one 1,3-butadiene-induced HG tumor contained the K-ras codon 13 mutation previously detected in lung and liver tumors. However, 16/29 HG tumors from the treated B6C3F1 mice contained H-ras codon 61 mutations. The mutations detected were: 12 CAA-->CGA transitions, two CAA-->CTA and two CAA-->AAA transversions. Eleven of 17 spontaneous HG tumors contained mutations in H-ras codon 61: five CAA-->CGA transitions, two CAA-->CTA transversions and four CAA-->AAA transversions. While the spectrum of ras mutations did not differ between the spontaneously occurring and chemically induced tumors, these data indicate that activation of H-ras contributes to the process of HG tumorigenesis in both groups of these neoplasms.

Co-purification of gastric mucoproteins with DNA: an explanation for the reported 'interaction' of omeprazole with DNA in rat tissues.

Recently, Phillips et al. reported that small amounts of radioactivity derived from [14C]omeprazole were 'associated' with DNA purified from gastrointestinal tissues of treated rats (Mutagenesis 7, 277-283, 1992). We hypothesized that this radioactivity arose from omeprazole bound to contaminating protein in the DNA fraction (Mutagenesis 7, 395-396, 1992). Using rats injected with 35S-labeled amino acids, we found significant protein contamination (0.06 microgram of protein per microgram of DNA) in DNA purified from gastrointestinal tissues. Gastric mucous proteins represent likely candidates for binding of omeprazole in the rat model used by Phillips et al. To investigate this, we partially purified proteins from gastric mucus, incubated them with [14C]omeprazole, and then added these radiolabeled mucoproteins to homogenates of rat colon and duodenum before starting the DNA purification. Detectable amounts of the added mucoproteins remained in the DNA fraction, but none of the control protein, bovine serum albumin, remained with the DNA. Further characterization of the mucoproteins by hydroxyapatite chromatography indicated that a certain population of these proteins survived the DNA purification procedures. These data indicate that the association of omeprazole with DNA reported by Phillips et al. most probably is explained by binding of omeprazole to mucous glycoproteins (or other proteins present in the GI tract) that selectively survive DNA purification protocols.

Rapid DNA degradation in primary rat hepatocytes treated with diverse cytotoxic chemicals: analysis by pulsed field gel electrophoresis and implications for alkaline elution assays.

The use of genetic toxicology tests for hazard identification is complicated by the fact that some in vitro tests using cultured mammalian cells are subject to potential artifacts that can make it difficult to distinguish between direct, chemically-induced genotoxicity, and DNA damage that occurs secondary to chemically-induced cytotoxicity (e.g., mediated by endogenous nucleases). Recently, we demonstrated that cytotoxicity-induced DNA double strand breaks (dsb) can produce artifacts in the in vitro alkaline elution/rat hepatocyte assay [Elia et al., 1993]. To explore this further, we used pulsed field gel/DNA dsb assays to characterize the relationship between chemically-induced cytotoxicity and the degradation of genomic DNA to high molecular weight fragments. Two sets of compounds were tested: 17 cytotoxic agents judged to be neither genotoxic nor carcinogenic, and 10 known genotoxic carcinogens. We found a close correlation between chemically-induced cytotoxicity and the rapid degradation of DNA to high molecular weight, double-stranded fragments. In contrast, the classic genotoxic chemicals tested generally did not trigger DNA dsb fragmentation at doses that were genotoxic but not immediately cytotoxic. These data indicate that pulsed field gel/DNA dsb assays can be used with in vitro genetic toxicology assays to help distinguish between genotoxic and cytotoxic mechanisms of DNA damage.

Cytotoxicity as measured by trypan blue as a potentially confounding variable in the in vitro alkaline elution/rat hepatocyte assay.

Rat hepatocytes treated in vitro with A2RA, an angiotensin II receptor antagonist, displayed an increased level of DNA-strand breaks as determined by alkaline elution, without an appreciable increase in cytotoxicity as determined by a trypan blue dye exclusion assay at harvest. The alkaline elution profile appeared to have two components: a rapidly eluting component detected in the first fraction collected (often associated with DNA from dead or dying cells), followed by a more slowly eluting component detected in the subsequent fractions. Further analysis of hepatocytes treated with A2RA by pulsed-field gel electrophoresis and neutral elution revealed significant levels of DNA double-strand breaks. Electron microscopy (EM) showed pronounced damage to mitochondria; although cell blebbing was seen using both EM and light microscopy, the plasma and nuclear membranes appeared intact when examined by EM. Cellular ATP levels decreased precipitously with increasing doses of A2RA, falling to less than 10% of control values at a dose of 0.213 mM A2RA, a concentration showing 100% relative viability by trypan blue at harvest. Thus, whereas in our experience trypan blue dye exclusion accurately reflects cytotoxicity induced by the majority of test agents, in this rather unusual case, trypan blue did not accurately reflect compound-induced cytotoxicity at harvest since there was no concurrent loss of membrane integrity. However, when hepatocytes treated with A2RA were incubated for either 3 h or 20 h in the absence of compound, a sharp, dose-dependent decline in viability was observed using trypan blue dye exclusion. Together with the initial, dose-dependent drop in the alkaline elution curve, these data suggest that the observed DNA double-strand breaks arose as a consequence of endonucleolytic DNA degradation associated with cytotoxicity, rather than by a direct compound-DNA interaction. Since DNA double-strand breaks behave under alkaline denaturing conditions as two single-strand breaks and can therefore produce increases in the alkaline-elution slope values, a necessary criteria for a valid positive result in this assay is that cytotoxicity by trypan blue dye exclusion will not be greater than 30%. Our data, however, indicate that interpretation of the elution assay as a test for genotoxicity can still be confounded by the failure of the trypan blue dye exclusion assay to reflect cytotoxicity in the unusual instance when there is no concurrent, immediate loss of membrane integrity.

Activation of the Ha-, Ki-, and N-ras genes in chemically induced liver tumors from CD-1 mice.

We compared the profile of ras gene mutations in spontaneous CD-1 mouse liver tumors with that found in liver tumors that were induced by a single i.p. injection of either 7,12-dimethylbenz(a)anthracene (DMBA), 4-aminoazobenzene, N-hydroxy-2-acetylaminofluorene, or N-nitrosodiethylamine. By direct sequencing of polymerase chain reaction-amplified tumor DNA, the carcinogen-induced tumors were found to have much higher frequencies of ras gene activation than spontaneous tumors. Furthermore, each carcinogen caused specific types of ras mutations not detected in spontaneous tumors, including several novel mutations not previously associated with either the carcinogen or mouse hepatocarcinogenesis. For example, the model compound DMBA is known to cause predominantly A to T transversions in Ha-ras codon 61 in mouse skin and mammary tumors, consistent with the ability of DMBA to form bulky adducts with adenosine. Our results demonstrate that the predominant mutation caused by DMBA in mouse liver tumors is a G to C transversion in Ki-ras codon 13 (DMBA is also known to form guanosine adducts), illustrating the influence of both chemical- and tissue-specific factors in determining the type of ras gene mutations in a tumor. 4-Aminoazobenzene and N-hydroxy-2-acetylaminofluorene also caused the Ki-ras codon 13 mutation. In addition, we found that N-nitrosodiethylamine, 4-aminoazobenzene, and N-hydroxy-2-acetylaminofluorene all caused G to T transversions in the N-ras gene (codons 12 or 13). This is the first demonstration of N-ras mutations in mouse liver tumors, establishing a role for the N-ras gene in mouse liver carcinogenesis. Finally, comparison of the ras mutations detected in the direct tumor analysis with those detected after NIH3T3 cell transfection indicates that spontaneous ras mutations (in Ha-ras codon 61) are often present in only a small fraction of the tumor cells, raising the possibility that they may sometimes occur as a late event in CD-1 mouse hepatocarcinogenesis.

Polymerase chain reaction/sequencing analysis of ras mutations in paraffin-embedded tissues as compared with 3T3 transfection and polymerase chain reaction/sequencing of frozen tumor deoxyribonucleic acids.

The efficiency of detection of H- and K-ras mutations in 27 CD-1 mouse liver tumors by direct sequencing of polymerase chain reaction (PCR)-amplified DNA isolated from formalin-fixed and paraffin-embedded tissues was compared with that after assay by both NIH 3T3 transfection (followed by sequencing of amplified transformant DNA) and direct sequencing of PCR-amplified DNA isolated from frozen tumors. Some tumor samples were chosen for comparison because they contained ras mutations that were detected by either NIH 3T3 transfection or sequencing of PCR-amplified DNA derived from frozen tumors, but were not detected by both techniques. The efficiency of detecting K-ras mutations was similar for sequencing of amplified fragments derived from both paraffin-embedded tissues and from frozen tumors. However, these two techniques differed in their efficacy for detection of H-ras codon 61 mutations. Furthermore, this difference appeared to be mutation-specific: the sequencing of amplified products from paraffin-embedded tumor tissues allowed increased detection of CAA to AAA mutations but decreased detection of CAA to CTA mutations relative to sequencing of amplified fragments derived from frozen tumor DNA. Direct sequencing of PCR products from paraffin-embedded sections was more sensitive than NIH 3T3 transfection for detection of activated K-ras genes containing codon 13 mutations but less sensitive for detection of activated H-ras genes containing codon 61 mutations. In summary, direct sequencing of amplified DNA from either frozen tumors or formalin-fixed, paraffin-embedded tissues can be more sensitive than NIH 3T3 transfection for detection of codon 13-activated K-ras genes. However, it appears to be less sensitive than NIH 3T3 transfection for detection of certain activating H-ras mutations. Depending upon the questions being asked of the data, each of the methods can provide useful information about ras gene mutations in tumor samples. The apparent differences in sensitivities between the methods is not yet understood, but such differences should be considered in the analysis of data obtained when only one method is used.