A new approach to identifying genotoxic carcinogens: p53 induction ...

Carcinogenesis vol.19 no.6 pp.1117–1125, 1998

A new approach to identifying genotoxic carcinogens: p53 induction as an indicator of genotoxic damage

Jun Yang and Penelope Duerksen-Hughes1 Department of Biology, Georgia State University, Atlanta, GA 30303, USA 1To

whom correspondence should be addressed Email: [email protected]

The tumor suppressor gene p53 encodes a nuclear phosphoprotein which is critical for cell cycle control and prevention of uncontrolled cell proliferation that can lead to cancer. Previous studies have shown that cells respond to DNA damage by increasing their levels of p53, which then acts to prevent replication of damaged DNA. This study examined the effects on p53 protein levels of several different categories of chemical carcinogens. N-MethylN9-nitro-nitrosoguanidine and N-ethyl-N-nitrosourea, two direct-acting genotoxic (DNA-reactive) carcinogens, caused p53 induction as early as 2 h following treatment, with peak increases within 4–12 h. Aflatoxin B1 and 2-acetylaminofluorene, indirect-acting genotoxic carcinogens, caused a later induction of p53, with the peak increase appearing between 16 and 24 h following treatment. These observations demonstrate a correlation between p53 induction pattern and DNA damaging mechanism of genotoxins. Phenol, diethylstilbestrol and ethylacrylate also induced increases in cellular p53. The half-life of p53 protein was increased in cells treated with genotoxic agents. On the other hand, the epigenetic (non-DNA-reactive) carcinogens azathioprine and saccharin, as well as two substances generally considered to be non-carcinogens, dimethylsulfoxide and benzethonium chloride, had no effect on p53 protein levels of treated cells. Measurement of the cytotoxic effects of each of these chemicals led to the conclusion that p53 protein induction is not a general, non-specific consequence of the cytotoxic effect of these genotoxins. These results suggest that measurement of p53 protein induction may be an effective tool to identify environmental genotoxins. Introduction It has long been known that exposure to certain chemicals is associated with the development of specific human cancers. Examples include the associations between amine dyes and bladder cancer, benzene and leukemia, vinyl chloride and hepatic cancer and cigarette smoking and lung cancer (1). Some naturally occurring mycotoxins, such as aflatoxins, ochratoxin A and stergmatocystin, are also found to be associated with an increased incidence of human cancer (2). Such chemicals are defined as carcinogens and, based on their *Abbreviations: ENU, N-ethyl-N-nitrosourea; MNNG, N-methyl-N-nitro-Nnitrosoguanidine; AFB1, aflatoxin B1; AFB2, aflatoxin B2; AAF, 2acetylaminofluorene; FBS, fetal bovine serum; DMEM, Dulbecco’s modified Eagle’s medium; DMSO, dimethylsulfoxide; DES, diethylstilbestrol; PBS, phosphate-buffered saline; ABTS, 2,29-azino-di(3-ethylbenzthiazolin sulfonate). © Oxford University Press

chemical and biological properties, they are divided into two major categories: genotoxic (DNA-reactive) and epigenetic (non-genotoxic or non-DNA-reactive) carcinogens (1,3). Genotoxic agents are those chemicals which can produce alterations in the genetic material of the host and they can be further subdivided into direct-acting genotoxins and indirect-acting genotoxins. Direct-acting agents are intrinsically reactive and do not require metabolic activation by cellular enzymes to interact with DNA. N-EthylN-nitrosourea (ENU*), N-methyl-N-nitro-N-nitrosoguanidine (MNNG), methyl methanesulfonate and the nitrogen and sulfur mustards are in this category (3). Indirect-acting agents, on the other hand, require metabolic activation by cellular enzymes to form the DNA-reactive metabolite. Examples include aflatoxin B1 (AFB1), aflatoxin B2 (AFB2), benzo[a]pyrene, 2-acetylaminofluorene (AAF) and benzidine. The enzymes responsible for activation are usually members of the cytochrome P450 family (3). Radiation and inorganic agents such as arsenic, chromium and nickel are also considered genotoxic agents (1). The second category, epigenetic carcinogens, consists of those substances which promote cancer in ways other than direct DNA damage. They include promoters, hormone modifying agents, peroxisome proliferators, cytotoxic agents and immunosuppressors. These compounds do not change the primary sequence of DNA but alter the expression or repression of certain genes and cellular events related to proliferation and differentiation. Organochlorine pesticides, saccharin, estrogen, cyclosporin A and azathioprine are examples of this class (1,3). However, it should be noted that the division between genotoxic and epigenetic carcinogens is not absolute. Genotoxic agents or their metabolites may also have epigenetic effects and epigenetic agents can be indirectly genotoxic. Therefore, the classification of a specific chemical is dependent upon which effect is considered to be the major effect of the chemical and which effect is secondary (1,3). Because society has made human health and safety an important consideration, considerable resources have been and are being expended in efforts to identify and classify human carcinogens. Unfortunately, methods currently used, such as the Ames test (4) and in vivo animal testing, suffer from several shortcomings, including a limited predictivity. In the Ames test one must estimate human effects from effects seen in bacteria and in the animal models high costs, long testing periods and rodent-to-human extrapolation are required. Other methods have also been used, however, their predictivity is also quite limited. One report by Tennant et al. (5) determined the ability of test results from four in vitro assays (mutagenesis in Salmonella, mutagenesis in mouse lymphoma cells, chromosome aberrations and sister chromatid exchanges in Chinese hamster ovary cells) to predict carcinogenicity in rodents of selected chemicals. The authors found that the carcinogenicity results and the results from each individual in vitro test had a 1117

Jun Yang and P.Duerksen-Hughes

concordance of ~60% and that no combination of these tests could improve substantially on the performance of the Salmonella mutagenicity assay. A number of additional reports confirmed this lack of predictivity (6–8). More recently a set of 44 chemicals being tested by the National Toxicology Program was defined and groups were invited to publish predictions of carcinogenicity. Eight groups did so, with mixed results (9). For 14 of the chemicals there was overall agreement among the predictions and the results of a 2 year bioassay. For 15 chemicals the results of the bioassay were poorly correlated with many of the predictive methods. Predictions have also been published for another group of 30 chemicals (10,11). While there is an increasing interest in replacing animal bioassays with some combination of in vitro testing procedures (12,13), thus far the available alternatives leave something to be desired in terms of predictivity. Hence, a more accurate and cost-effective alternative is desirable. Recently there has been increasing interest in the use of approaches based on gene induction to identify carcinogens (see for example ref. 14). Some sort of analysis based on p53 expression would seem to be a promising candidate for a low cost predictive assay of genotoxic damage. The tumor suppressor protein p53 has been the focus of much interest in recent years and has been carefully characterized (reviewed in 15–19). It is the most commonly mutated gene reported in human tumors (20–22) and the status of the p53 gene can serve as a prognostic indicator in human malignancies (23,24), at least for breast cancer (25) and squamous cell carcinoma of the head and neck (26). Previous work has shown that p53 is a DNA binding protein which can act both as a positive and negative modulator of transcription. Binding of p53 to its cognate response element can activate expression of target genes, several of which are involved in cell cycle regulation. These include p21CIP1/WAF1 (27), cyclin G (28) and human proliferating cell nuclear antigen (29). p53 also activates its negative modulator, MDM2 (30,31), the mouse muscle creatine kinase gene (32), GADD45 (involved in DNA repair) (33,34) and Bax (important in cell death) (35–37). p53 can also repress transcription by a direct interaction with other transcriptional factors, such as the TATA box binding protein (38,39), CCAAT binding factor (40) and Sp1 (41). The role for p53 in cell cycle control has also been well studied. One important cellular response to genotoxic stress is an increase in p53 protein levels (42–48). This protein then functions as a transcriptional modulator to activate or repress specific gene expression. One possible outcome is G1/S arrest, which allows repair of damaged DNA before replication, while a second possible outcome is apoptosis (programed cell death). In either case replication of damaged DNA is prevented. In an effort to determine which type of DNA damage is responsible for the increase in p53 protein levels Nelson and Kastan (45) analyzed a diverse collection of different DNA damaging agents and found that DNA strand breaks may be the critical signal to trigger p53 induction. p53 then acts on several downstream genes. The p21WAF1/CIP1 gene product contributes to growth arrest; it functions as an inhibitor of several cyclin-dependent kinases whose activity is required for a cell to enter S phase (27,49,50). Gadd45 is important in DNA repair (34) and bax functions in the apoptotic pathway (35–37). Since the majority of human carcinogens are genotoxins (1,3), it seemed likely that cells would respond to such 1118

substances by increasing their p53 protein levels. Such an increase could thus be useful as an indicator of DNA damage. Some genotoxic agents have been tested for their effects on cellular p53 levels. However, this work has generally been done with agents used clinically, such as those used in cancer therapy (e.g. mitomycin C, cyclophosphamide, cisplatin and etoposide) and UV, ionizing and α-radiation, rather than with chemicals which might be found in the environment (see for example 42,47). Hence, the role of p53 in responding to such environmentally relevant agents has not been demonstrated. Furthermore, reported results have not always been consistent. For example, Fritsche et al. (42) found that mitomycin C could induce p53 in NCTC 929 cells while in the experiments conducted by Nelson and Kastan (45) no significant p53 induction was discernible in LNCaP cells at early time points after treatment with mitomycin C. To test the utility of p53 induction in cultured cells as an indicator of genotoxic damage we chose NCTC 929 cells as our model system and assessed the effects on p53 levels of several environmentally relevant chemicals. NCTC 929 cells were treated with four categories of chemicals: direct-acting genotoxins, indirect-acting genotoxins, epigenetic carcinogens and non-carcinogens. The time- and dose-dependent effects of the chemicals on cellular p53 protein levels and on cytotoxicity were analyzed and a close correlation between the DNA damaging mechanism and p53 induction pattern is reported here. In addition, there was no clear correlation between p53 induction and the cytotoxic effects of genotoxins. Materials and methods Cell lines and culture NCTC 929 mouse fibroblast cells were cultivated in minimum essential medium (Gibco/Life Sciences, Gaithersburg, MD) supplemented with 10% fetal bovine serum (FBS) (Gibco). SVHTI cells are SV40 T antigentransformed mouse fibroblasts (generously provided by Linda R.Gooding, Emory University School of Medicine, Atlanta, GA) which were cultivated in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco/Life Sciences), supplemented with 10% FBS. Because the T antigen stabilizes p53 (51), lysates of these cells were used as a positive control for p53 in Western blot analysis. Pab122 hybridoma cells (ATCC), which produce anti-p53, were propagated in DMEM with 10% FBS. Genotoxin treatment MNNG (70-25-7) and ethylacrylate (140-88-5) were obtained from Aldrich Chemical Co., dimethylsulfoxide (DMSO) (67-68-5) was acquired from J.T.Baker Inc. and AFB1 (1162-65-8), AAF (53-96-3), ENU (759-73-9), azathioprine (446-86-6), benzethonium chloride (121-54-0), saccharin (12844-9), phenol (108-95-2) and diethylstilbestrol (DES) (56-53-1) were all obtained from Sigma Chemical Co. All chemicals were dissolved in DMSO and added directly to NCTC 929 cell culture plates. DMSO alone did not induce p53 (see Results). For time–response experiments cells were exposed to a given dose of the agent, then plates were harvested at various time points for immunoblot analysis. For dose–response experiments cells were exposed to different concentrations of the chemical and harvested at 6–7 and/ or 16–18 h following treatment. Immunoblot analysis of p53 protein levels After exposure to different concentrations of chemicals or for different times cells were removed from the plates by trypsin treatment (0.25 mg/ml) and pelleted by centrifugation at 1000 r.p.m. for 10 min. The cell pellet was then treated with lysis buffer (20 mM Tris, 150 mM NaCl, 1% NP-40, pH 8.0, 1% aproteinin) for 20 min on ice and membrane fragments were removed by centrifugation at 14 000 r.p.m. for 15 min. The protein concentration of the cleared lysate was determined by the bicinchoninic acid assay (Pierce). Equivalent amounts of cell extract (by protein concentrations) were electrophoresed on a SDS–12% polyacrylamide gel and then transferred to a nitrocellulose membrane (Micron Separations Inc.) using a semi-dry transfer cell (BioRad). p53 protein was detected by Western blot using the mixture of

p53 induction as an indicator of genotoxic damage anti-p53 antibodies pAb 122 (purified from hybridoma culture media on protein A–Sepharose; Pharmacia) and Ab-3 (Oncogene Science) (at a ratio of 10:1) and visualized with the chemiluminescent horseradish peroxidase system (Kirkegaard & Perry Laboratories Inc.). The absorbance of the bands was measured by densitometry (PDI). p53 ELISA The capture antibody, pAb 122, was added to the wells of a 96-well immunoassay plate [4.0 µg/ml diluted in coating buffer (0.1 M NaHCO3, pH 8.2), 50 µl/well] and allowed to adhere overnight at 4°C. The plate was then washed twice in phosphate-buffered saline (PBS)/Tween (PBS plus 0.05% Tween-20) and 200 µl of a blocking solution (PBS plus 10% newborn calf serum, PBS 10) was added and allowed to incubate for 2 h at room temperature. The p53–glutathione S-transferase fusion protein obtained from Santa Cruz Biotechnology Inc. was used as the standard and solutions of 10.0, 5.0, 2.0, 1.0 and 0.5 ng/ml p53 diluted in PBS 10 were prepared. Samples were prepared by adding 35 µl cell lysate to 315 µl PBS 10. After blocking the plates were washed twice in PBS/Tween and the standards and samples were added at 100 µl/well. Each data point was measured in triplicate. The plates were then incubated overnight at 4°C and washed four times in PBS/Tween. Aliquots of 100 µl secondary antibody (polyclonal biotinylated anti-p53 at a concentration of 4 µg/ml diluted in PBS 10; Boehringer Mannheim) were added to each well and allowed to adhere for 45 min at room temperature. The plates were washed six times in PBS/Tween, then avidin–peroxidase (2.5 µg/ml in PBS 10; Sigma) was added (100 µl/well) and allowed to incubate at room temperature for 30 min. The plates were washed eight times in PBS/Tween, then 100 µl substrate were added to each well [150 mg 2,29-azino-di-(3-ethylbenzthiazolin sulfonate) (ABTS) dissolved in 500 ml 0.1 M citric acid, pH 4.35, and stored in 11 ml aliquots at –20°C, 1 µl 30% H2O2/ml substrate solution added just before use] and allowed to incubate at room temperature for 30 min. The absorbance at 405 nm was then read using a BioRad 3550-UV microplate reader and the results analyzed using the MicroPlate Manager software supplied with the plate reader. Cytotoxicity measurement For each experiment cells were examined microscopically for morphological evidence of cell death. Cell death was also measured by ELISA. This assay was performed using a commercially available kit (Boehringer Mannheim) according to the manufacturer’s directions. Lysates prepared as described above were added to wells to which the anti-histone antibody was fixed adsorptively. Nucleosomes in the sample bound via their histone components to the immobilized anti-histone antibody. Anti-DNA peroxidase, which binds to the DNA portion of the nucleosomes, was then added. The amount of bound peroxidase was determined photometrically by measuring absorbance at 405 nm following addition of ABTS as substrate. Immunoprecipitation Cells (13106 per plate) were incubated in methionine-free RPMI 1640 (Gibco) with 5% dialyzed FBS for 1 h before labeling with 100 µCi/ml [35S]methionine (SJ.204; Amersham) for 1 h. After labeling cells were either harvested immediately or washed and treated with unlabeled medium. At defined times thereafter cells were lysed in PBS/TDS (50 mM Na2HPO4, 17 mM NaH2PO4, 68 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 0.1% SDS, pH 7.4) and equal amounts of cell lysate were immunoprecipitated with Ab-1 (Oncogene Sciences) bound to protein A–agarose (Oncogene Sciences), using the procedure recommended by the manufacturer. Immunoprecipitated proteins were separated by SDS–PAGE (15%) and the gel was then dried and autoradiographed.

Results Direct-acting genotoxins rapidly increase p53 levels in a timeand dose-dependent manner In previous studies induction of cellular p53 protein was detected at a selected time, either 2–3 h (for example ref. 45) or 16 h (42) after exposure to DNA damaging agents and after exposure to a given dose of that agent. To obtain a more complete picture of the effects of direct-acting environmentally relevant genotoxins on cellular p53 levels the time– and dose– response characteristics of NCTC 929 cells to two well-known direct-acting genotoxins, MNNG and ENU, were studied. Both chemicals are used broadly in different bioassays for mutagenesis and carcinogenesis (for example 52–55). Western blotting and ELISA analyses allowed us to measure the relative

levels of p53 in each sample. Concentrations ranging from 1 to 100 µg/ml were applied and cells were exposed to the chemicals for 6 h. For MNNG p53 induction was apparent at 1 µg/ml and was strongest at 10 µg/ml. For ENU induction was apparent between doses of 20 and 100 µg/ml, with 40 µg/ml yielding maximal induction (data not shown). At a concentration which can induce p53 (10 µg/ml MNNG and 40 µg/ml ENU) the time course measured by both Western blot and ELISA (Figure 1) showed that both direct-acting genotoxins clearly elevated p53 protein levels within 2 h following treatment, that induction reached a peak around 4–12 h and that p53 levels then decreased. Figure 1C shows cytotoxicity data for MNNG and ENU. MNNG induced significant cell death at 6 h, at which point induction of p53 was also significant. However, cell death became more significant with time, while the p53 protein level diminished. For ENU cell death became significant at ~16 h, at which point the p53 level had already begun to drop. This indicates that the genotoxins affect levels of p53 protein and cytotoxicity differently. Indirect-acting genotoxins induce p53 at later times Similarly, the time– and dose–response characteristics of cells exposed to indirect-acting genotoxins were studied. AFB1 is produced by Aspergillus flavis and can be found in such food products as nuts, oil seeds and grains. It is highly mutagenic after biological transformation (2,3,56,57). AAF is used as a positive control by toxicologists to study the carcinogenicity and mutagenicity of aromatic amines (58). After exposure to different doses of AAF or AFB1 for 16 h cells were harvested and the p53 protein levels measured. At low concentrations (1 and 5 µg/ml) AFB1 did not affect p53 levels, while between 10 and 100 µg/ml p53 induction was apparent. A dose of 20 µg/ml AFB1 yielded maximal induction, with induction decreasing at concentrations .20 µg/ml. A similar dose– response effect was observed in AAF-treated cells, with induction detectable at 20 and 40 µg/ml yielding maximal induction (data not shown). Thus the dose–response characteristics are similar to those seen with the direct-acting genotoxins. However, the two groups differed noticeably in the kinetics of p53 induction. For both AAF and AFB1 peak induction appeared after 16 h treatment, as measured by Western blot and p53 ELISA (Figure 2A and 2B). In contrast, cells treated with the direct-acting genotoxins MNNG and ENU had peak p53 levels much earlier. This difference reflects the different DNA damaging mechanisms of the two groups. Similar results were obtained using AFB2 and benzidine, two other indirectacting genotoxins (data not shown). Figure 2C shows cytotoxicity data for AAF and AFB1. AAF produced detectable cell death beginning ~12 h after treatment. However, AFB1, which strongly induced p53 at 10 µg/ml, did not induce significant cell death compared with the control at any of the time points measured. This indicates that p53 can indeed be induced in the absence of significant cytolysis. Epigenetic carcinogens do not affect cellular p53 protein levels Since the signal for p53 induction is considered to be damage of DNA, epigenetic carcinogens and non-carcinogens are not expected to influence cellular p53 levels. Patients treated with azathioprine, an immunosuppressor, sometimes develop cancers such as leukemias (59–61). Saccharin has also been related to an increased incidence of bladder cancer in rats; the mechanism is thought to be ionic and pH imbalances due to 1119

Jun Yang and P.Duerksen-Hughes

Fig. 1. Time–response characteristics of p53 induction and cell death in NCTC 929 cells by direct-acting genotoxins. p53 levels were analyzed by both ELISA and Western blot. Each experiment was done at least twice; the results from one representative experiment are shown. (A) Western blot and p53 ELISA for cells treated with 10 µg/ml MNNG for the times indicated. (B) Western blot and p53 ELISA for cells treated with 40 µg/ml ENU for the times indicated. (C) Cell death for cells treated with MNNG or ENU respectively.

Fig. 2. Time–response characteristics of p53 induction in NCTC 929 cells by indirect-acting genotoxins. p53 levels were analyzed by both ELISA and Western blot. Each experiment was done at least twice; the results from one representative experiment are shown. (A) Western blot and p53 ELISA for cells treated with 10 µg/ml AFB1 for the times indicated. (B) Western blot and p53 ELISA for cells treated with 40 µg/ml AAF for the times indicated. (C) Cell death for cells treated with AFB1 or AAF respectively.

the huge doses of sodium saccharide administered (62). The effects on p53 levels of these two epigenetic carcinogens were evaluated at peak induction times for both direct-acting and indirect-acting genotoxins (6 and 16 h respectively). At all concentrations tested (1–100 µg/ml) neither caused a detectable increase in p53 levels at either time. A complete time course (0–24 h) for each chemical also gave negative results by both Western blot and ELISA analysis (data not shown). Even higher concentrations of saccharin (up to 20 mg/ml) did not increase p53 levels detectably (data not shown). 1120

Azathioprine was relatively non-cytotoxic to cells at most times and doses tested. However, in cells treated with the highest dose (100 µg/ml) for the longest time (16 h) (Figure 3A) some cytotoxicity was observed. Saccharin was relatively non-cytotoxic to cells at most of the doses and times tested (up to and including 1 mg/ml), however, at a dose of 20 mg/ ml significant cytotoxicity was observed. The fact that p53 induction was not observed in cells undergoing this process indicates that p53 induction is not a non-specific response to generalized cellular cytotoxicity.

p53 induction as an indicator of genotoxic damage

Fig. 3. The different cytotoxic effects of epigenetic carcinogens and nongenotoxic non-carcinogens. (A) Cell death assay for cells treated with azathioprine (Aza) or saccharin (Sac) for the times and doses indicated. (B) Cell death assay for cells treated with benzethonium chloride (BCI) or DMSO for the times and doses indicated.

Non-genotoxic non-carcinogens do not affect cellular p53 protein levels The same set of experiments were done with benzethonium chloride, used in cosmetics for its antimicrobial and cationic surfactant properties and reported to be a non-genotoxic noncarcinogen (63), and with DMSO, the solvent used in our experiments. For benzethonium chloride some cytotoxicity was observed as early as 2 h and at doses as low as 1 µg/ml (Figure 3B). However, under these conditions as well as at higher doses and longer times (up to 100 µg/ml and 16 h) no p53 accumulation was observed. In the case of DMSO no p53 induction and no cytotoxicity was observed for any of the doses (1–100 µg/ml) or times (up to 24 h) tested. In conclusion, those chemicals which do not cause DNA damage do not cause increases in p53 protein levels and this

lack of induction is apparent both in the presence and absence of cellular cytotoxicity. Phenol, DES and ethylacrylate cause increases in cellular p53 levels Phenol has been reported in the literature both to possess and to lack weak genotoxic activity, depending on the system studied and the assay used (see for example 64–66). Likewise, both positive and negative results have been found when DES was tested for genotoxicity (67–69, and references therein). A possible explanation for these discrepant results may be that while DES metabolites can form DNA adducts, the adducts are unstable and are removed or repaired quickly (biological half-life of ~14 h) (70,71). Likewise, the genotoxicity data for ethylacrylate is somewhat equivocal and contradictory (72–75, and references therein). Our results (Figure 4) show that each of these substances can increase cellular levels of p53 by 16 h. Phenol induced cytotoxicity at high doses (.150 µg/ml), DES induced cytotoxicity at doses .20 µg/ml, while EA induced lower levels of cytotoxicity (Figure 4D). This p53-based method of screening substances for carcinogenicity, therefore, has the potential to detect even transitory DNA damage and may be capable of identifying DNA damaging agents which could escape existing screens. The half-life of p53 protein is increased in genotoxin-treated cells The level of p53 is thought to be regulated mainly by posttranslational modification, although some control may be exerted at the transcriptional level (76). Previous work, for example, has shown that an increase in p53 protein levels occurred following treatment with DNA damaging agents without a corresponding change in the p53 mRNA level (42,77). Here we used conventional [35S]methionine pulsechase labeling to determine the half-life of p53 protein in both untreated and genotoxin-treated cells. MNNG (10 µg/ml) was added to the cells at the same time as the radioactive label. In untreated cells the labeled p53 protein was nearly gone 4 h after initial labeling (t1/2 ~ 3 h). However, in MNNG-treated cells the labeled p53 protein was still present after 16 h (t1/2 ~ 9 h) (Figure 5). This indicates that environmental genotoxins can stabilize p53. However, the mechanism responsible for the increased half-life is still unknown. Discussion In this study we observed changes in cellular p53 protein levels resulting from treatment of mouse fibroblast NCTC 929 cells with several environmentally relevant chemicals. The results demonstrated that an assay designed to measure p53 levels has the potential to discriminate between genotoxins and non-genotoxins, as both direct-acting and indirect-acting genotoxins increased cellular p53 levels, while non-carcinogens and epigenetic carcinogens did not. Furthermore, the results of cytotoxicity assays for these chemicals revealed that induction of p53 protein is not a general consequence of the cytotoxic effects of genotoxins. This is in agreement with and extends previous work showing that genotoxic stress causes p53 induction (42–48). In particular, we have newly shown that this assay can identify genotoxic agents found in the environment, that it may be able to identify and distinguish between direct and indirect agents and that it can detect transient DNA damage. For each genotoxin tested we were able to define an effective dose 1121

Jun Yang and P.Duerksen-Hughes

Fig. 4. Phenol, DES and ethylacrylate (EA) all induce increased p53 levels in NCTC 929 cells (A–C) Western blot and p53 ELISA for cells treated with various concentrations of the three agents for 16 h before harvest. (D) Cell death assay for cells treated with the three agents.

range. Some of the genotoxins appear to be cytotoxic, especially at the higher doses. This may partially explain the lower levels of p53 observed at these doses. We also observed that direct-acting genotoxins increased cellular p53 levels rapidly, while the indirect-acting ones did so only after a lag of several hours. Because DNA damage caused by indirect-acting genotoxins is dependent on metabolic 1122

activation, in which the chemicals are metabolized to DNAactive intermediates, this longer time period is anticipated. Different chemicals are often converted by different P450 isozymes, thus the characteristics of a specific set of enzymes, such as substrate affinity and conversion efficiency, may be determining factors for the length of this time lag. A close correlation therefore appears between the DNA

p53 induction as an indicator of genotoxic damage

Fig. 5. MNNG treatment extends the half-life of p53. NCTC 929 cells were pulse labeled with [35S]methionine, then chased with unlabeled medium for the indicated times before harvest, immunoprecipitation, electrophoresis and autoradiography.

damaging mechanism and the p53 induction pattern of a given chemical. This could explain some discrepancies noted in previous studies: looking for p53 induction by a direct-acting agent at a later time or by an indirect-acting agent at an earlier time would give false negative results. We tested three chemicals for which the literature had given inconclusive results regarding their genotoxic potential: phenol, DES and ethylacrylate. In each case our assay was capable of detecting p53 increases, indicative of DNA damage. Therefore, this p53-based assay is capable of identifying DNA damaging agents which might be missed by other methods. In particular, we were able to measure a response following treatment with DES, which produces only transitory genotoxic damage. Bacteria- and rodent-based assays, which measure a heritable change in the genome leading to a phenotypic change, are dependent on the DNA damage remaining through at least one cycle of replication. In contrast, the phenomenon measured with this method (increases in cellular p53) is observable even if the damage is repaired quickly, before DNA replication occurs. We found NCTC 929 cells to be a useful model for these studies. We chose to use this cell line because previous reports have shown that its level of p53 can be increased by treatment with genotoxic agents (42) and because we found that an antibody specific for wild-type protein (Ab-1; Oncogene Science) can immunoprecipitate p53 from these cell lysates (data not shown). p53 levels increase in treated NCTC 929 cells in a reproducible and predictable manner and these cells appear to be able to perform the metabolism necessary to transform the tested indirect genotoxins into DNA damaging agents. It is likely that other cells (such as those from hepatocyte-derived lines) will possess a higher metabolic capability and be able to metabolize indirect-acting genotoxins more quickly, thus decreasing the observed difference between direct- and indirect-acting genotoxins with respect to induction time. If one wishes to use this assay to distinguish between direct- and indirect-acting substances, therefore, it may be helpful to use a cell line with a more moderate metabolic capacity, such as NCTC 929. As this system is further developed it will be desirable to characterize other cell lines, including those from humans, for their behavior in this assay. It is likely that other cells may display different characteristics, which may reflect the vulnerability of the source to genotoxic damage. For example, cells with a different complement of P450 isozymes may deal with indirect genotoxins such that the kinetics or type of DNA damage differ from that seen in the NCTC 929 system. It should be noted that many, and perhaps the majority, of cells

in culture have acquired p53 mutations and may not be suitable models for this system. Hence, the selection of an appropriate cell model (or models) is not trivial and must take into consideration numerous factors, including species differences, organ specificities and individual variations, as well as cellular function. Using both the Western blot and ELISA procedures we found that the patterns reported here with respect to the presence or absence of p53 induction and the time at which that induction occurred were very reproducible. The strength of the Western blot protocol is that one can specifically observe an increase in p53 even if cross-reactive proteins are present in the cell lysate; one can be certain that the changes being measured occur in a protein with a molecular weight of 53 kDa. However, the exact magnitude of the baseline level of p53 and of induction as measured by densitometry was somewhat more variable from blot to blot, making the Western blot analysis more useful as a qualitative rather than a quantitative measurement of p53 induction. The strengths of the ELISA method are that it is more cost- and time-effective and yields more easily quantifiable measurements. This will be important if the p53-based approach is to be used for widespread screening. The half-life of p53 has been measured under a number of conditions. During normal metabolism p53 has a half-life of 20–40 min in most cells tested (78,79) and 3 h in human mammary epithelial cells (80). Consistent with previous reports, we also show that DNA damage caused an increase in p53 half-life. Cancer is a major threat to human life and chemical carcinogens are responsible for a portion of the occurrence of the disease. One preventive approach is to identify agents that cause or are likely to cause cancer in humans and to minimize exposure, thereby reducing the risk of cancer. Many methods have been developed to screen for carcinogenic substances. Among them, the Ames test (4) and the rodent bioassay, also known as ‘the gold standard’ (1), are the most widely used. However, there are some important limitations to these methods. In the Ames test a special strain of Salmonella is used and the applicability of results in this system to humans or rodents is therefore limited. Hundreds of rodents are required for the rodent bioassay, the cost is so high and the duration so long (2 years) that the number of chemicals that can be assayed each year is limited. Furthermore, in order to obtain statistically meaningful results high toxic doses are often used in the rodent bioassay. Because of this toxicity, cell proliferation is induced and some substances will generate false positive results. There is, therefore, a continuing need to develop new methods which can more accurately, quickly and efficiently identify carcinogens. As the role of p53 in cell cycle control and tumor prevention has become more clear, some researchers have tried to incorporate this information into such testing protocols. Using a p531/– transgenic mouse model Tennant et al. (81) found that genotoxic carcinogens can be identified with increased sensitivity and specificity and within a time frame of only 6 months, rather than 2 years. The assay described here would fit broadly into the category of assays that measure changes in gene or protein expression in an effort to identify carcinogens, although it should be noted that the increase in p53 appears to be based primarily on an increase in stability rather than on increased transcription. p53 may be a particularly good protein to measure, as the 1123

Jun Yang and P.Duerksen-Hughes

available evidence indicates that it is directly involved in the chain of biochemical events that follows genotoxic damage, rather than having its levels regulated by events secondary to the actual damage. In this report we show that cellular p53 responses, as detected in a tissue culture system, can discriminate between genotoxins and non-genotoxins. Compared with the Ames and rodent bioassays the advantages of this method are significant. First, we are using mammalian cells, so it is expected to be a more accurate reflection of what occurs in humans. In fact, it should be possible to use human cells from various tissues directly; this may be helpful in identifying and explaining differences between species and between tissues regarding their sensitivity to particular genotoxins. Secondly, no live animals are used and therefore the cost should be much less than that required for the rodent bioassay. Also, we will be able to test lower less toxic doses. Finally, results are available in a matter of days, rather than years. With our assay it should also be possible to obtain information which is either unavailable or more difficult and costly to achieve using rodent bioassays. For example, this method may identify a substance as either a direct or an indirect genotoxin. Also, the cost differential makes it possible to test many more substances, as well as more doses of each substance, than is currently feasible. It should now also be practical to test mixtures of substances in various formulations and look for interactions, such as synergy and antagonism. Some limitations exist with this method. One is that it is not expected to identify non-genotoxic carcinogens (epigenetic carcinogens). However, because most human carcinogens are genotoxins, this test should identify the vast majority of human carcinogens. Ultimately, the usefulness of this method will depend on how well p53 induction can predict carcinogenicity. Chemical carcinogenesis is, of course, a complex process that involves many factors, including metabolic activation, detoxification and DNA damage and repair. To adequately address the issue of accuracy and the frequency of false positives and negatives in this assay it will be necessary to test a large number of chemicals using this method and to compare the results with both those obtained from in vivo testing and from the alternative in vitro procedures, such as mutagenesis in Salmonella and in mouse lymphoma cells, chromosome aberrations and sister chromatid exchange. This type of data will allow an assessment of the usefulness of this assay as compared with currently available alternatives. Acknowledgement This work was supported in part by grant MCB 9513527 from the National Science Foundation.

References 1. Smart,R.C. (1994) Carcinogenesis. Appleton & Lange, Norwalk, CT. 2. Chu,F.S. (1991) Mycotoxins: food contamination, mechanism, carcinogenic potential and preventive measures. Mutat. Res., 259, 291–306. 3. Williams,G.M. and Weisburger,J.H. (1991) Chemical Carcinogenesis. Pergamon Press, New York, NY. 4. Maron,D.M. and Ames,B.N. (1983) Revised methods for the Salmonella mutagenicity test. Mutat. Res, 113, 173–215. 5. Tennant,R.W. et al. (1987) Prediction of chemical carcinogenicity in rodents from in vitro genetic toxicity assays. Science, 236, 933–941. 6. Zeiger,E., Haseman,J.K., Shelby,M.D., Margolin,B.H. and Tennant R.W. (1990) Evaluation of four in vitro genetic toxicity tests for predicting

1124

rodent carcinogenicity: confirmation of early results with 41 additional chemicals. Environ. Mol. Muatagen., 16, 1–14. 7. Tennant,R.W., Spalding,J.W., Stasiewicz,S., Caspary,W.D., Mason,J.M. and Resnick,M.A. (1987) Comparative evaluation of genetic toxicity patterns of carcinogens and noncarcinogens: strategies for predictive use of shortterm assays. Environ. Hlth Perspect., 75, 87–96. 8. Tennant,R.W. (1991) The genetic toxicology database of the National Toxicology Program: evaluation of the relationships between genetic toxicology and carcinogenicity. Environ. Hlth Perspect., 96, 47–51. 9. Wachsman,J.T., Bristol,D.W., Spalding,J., Shelby,M. and Tennant,R.W. (1993) Predicting chemical carcinogenesis in rodents. Environ. Hlth Perspect., 101, 444–445. 10. Ashby,J. (1997) Comparison of 17 methods of predicting the carcinogenicity of 30 chemicals. Environ. Hlth Perspect., 105, 466–467. 11. Ashby,J. (1996) Prediction of rodent carcinogenicity for 30 chemicals. Environ. Hlth Perspect., 104, 1101–1104. 12. Ashby,J. and Morrod,R.S. (1991) Detection of human carcinogens. Nature, 352, 185–186. 13. Ashby,J. (1996) Alternatives to the 2-species bioassay for the identification of potential human carcinogens. Hum. Exp. Toxicol., 15, 183–202. 14. Todd,M.D., Lee,M.J., Williams,J.L., Nalezny,J.M., Gee,P., Benjamin ,M.B. and Farr,S.B. (1995) The CAT-Tox (L) assay: a sensitive and specific measure of stress-induced transcription in transformed human liver cells. Fundam. Appl. Toxicol., 28, 118–128. 15. Donehower,L.A. and Bradley,A. (1993) The tumor suppressor p53. Biochim. Biophys. Acta, 1155, 181–205. 16. Haffner,R. and Oren,M. (1995) Biochemical properties and biological effects of p53. Curr. Opin. Genet. Dev., 5, 84–90. 17. Lee,J.M. and Bernstein,A. (1995) Apoptosis, cancer and the p53 tumor suppressor gene. Cancer Metastasis Rev., 14, 149–161. 18. Levine,A.J. (1993) The tumor suppressor genes. Annu. Rev. Biochem., 62, 623–651. 19. Selter,H. and Montenarh,M. (1994) The emerging picture of p53. Int. J. Biochem., 26, 145–154. 20. Greenblat,M.S., Bennet,W.P., Hollstein,M. and Harris,C.C. (1994) Mutations in the p53 tumor suppressor gene: clues to cancer etiology and molecular pathogenesis. Cancer Res., 54, 4855–4878. 21. Hollstein,M., Sidransky,D., Vogelstein,G. and Harris,C.C. (1991) p53 mutations in human cancer. Science, 253, 49–53. 22. Vogelstein,B. and Kinzler,K.W. (1992) p53 function and dysfunction. Cell, 70, 523–526. 23. Fisher,D.E. (1994) Apoptosis in cancer therapy: crossing the threshold. Cell, 78, 539–542. 24. DeVita,V.T. and Deisseroth,A.B. (1996) Oncology: resistance to chemotherapy. J. Am. Med. Ass., 275, 1833–1834. 25. Kovach,J.S., Hartmann,A., Blaszyk,H., Cunningham,J., Schaid,D. and Sommer,S.S. (1996) Mutation detection by highly sensitive methods indicates that p53 gene mutations in breast cancer can have important prognostic value. Proc. Natl Acad. Sci. USA, 93, 1093–1096. 26. Brennan,J.A., Mao,L., Hruban,R.H., Bohle,J.O., Eby,Y.J., Koch,W.M., Goodman,S.N. and Sidransky,D. (1995) Molecular assessment of histopathological staging in squamous-cell carcinoma of the head and neck. New Engl. J. Med., 332, 429–453. 27. El-Deiry,W.S. et al. (1993) WAF1, a potential mediator of p53 tumor suppression. Cell, 75, 817–825. 28. Okamoto,K. and Beach,D. (1994) Cyclin G is a transcriptional target of the p53 tumor suppressor protein. EMBO J., 13, 4816–4822. 29. Shivakumar,C.V., Brown,D.R., Deb,S. and Deb,S.P. (1995) Wild-type human p53 transactivates the human proliferating cell nuclear antigen promotor. Mol. Cell. Biol., 15, 6785–6793. 30. Momand,J., Zambetti,G.P., Olson,D.C., George,D. and Levine,A.J. (1992) The mdm2 oncogene product forms a complex with the p53 protein and inhibits p53 mediated transactivation. Cell, 69, 1237–1245. 31. Barak,Y., Juven,T., Haffner,R. and Oren,M. (1993) mdm2 expression is induced by wild type p53 activity. EMBO J., 12, 461–468. 32. Weintraub,H., Hauschka,S. and Tapscott,S.J. (1991) The MCK enhancer contains a p53 responsive element. Proc. Natl Acad. Sci. USA, 88, 4570–4571. 33. Kastan,M.B., Zhan,Q., El-Deiry,W.S., Carrier,F., Jacks,T., Walsh,W.V., Plunkett,B.S., Vogelstein,B. and Fornace,A.J. (1992) A mammalian cell cycle checkpoint pathway utilizing p53 and GADD45 is defective in ataxia-telangiectasia. Cell, 71, 587–597. 34. Smith,M.L., Chen,I.-T., Zhan,Q., Bae,I., Chen,C.-Y., Gilmer,T.M., Kastan,M.B., O’Connor,P.M. and Fornace,A.J.,Jr (1994) Interaction of the p53-regulated protein GADD45 with proliferation cell nuclear antigen. Science, 266, 1367–1380.

p53 induction as an indicator of genotoxic damage 35. Miyashita,T. and Reed,J.C. (1995) Tumor suppressor p53 is a direct transcriptional activator of the human bax gene. Cell, 80, 293–299. 36. Miyashita,T., Krajewski,S., Krejewska,M., Wang,H.G., Lin,H.K., Liebermann,D.A., Hoffman,B. and Reed,J.C. (1994) Tumor suppressor p53 is a regulator of bcl-2 and bax gene expression in vitro and in vivo. Oncogene, 9, 1799–1805. 37. Han,J., Sabbatini,P., Perez,D., Rao,L., Modha,D. and White,E. (1996) The E1B 19K protein blocks apoptosis by interacting with and inhibiting the p53-inducible and death-promoting Bax protein. Genes Dev., 10, 461–477. 38. Truant,R., Xiao,H., Ingles,C.J. and Greenblatt,J. (1993) Direct interaction between the transcriptional activation domain of human p53 and the TATA box-binding protein. J. Biol. Chem., 268, 2284–2287. 39. Seto,E., Usheva,A., Zambetti,G.P., Momand,J., Horikoshi,N., Weinmann,R., Levine,A.J. and Shenk,T. (1992) Wild-type p53 binds to the TATA-binding protein and represses transcription. Proc. Natl Acad. Sci. USA, 89, 12028–12032. 40. Agoff,S.N., Hou,J., Linzer,D.I.H. and Wu,B. (1993) Regulation of the human hsp70 promotor by p53. Science, 259, 84–87. 41. Borellini,F. and Glazer,R.I. (1993) Induction of Sp1–p53 DNA-binding heterocomplexes during granulocyte/macrophage colony-stimulating factor-dependent proliferation in human erythroleukemia cell line TF-1. J. Biol. Chem., 268, 7923–7928. 42. Fritsche,M., Haessler,C. and Brandner,G. (1993) Induction of nuclear accumulation of the tumor-suppressing protein p53 by DNA-damaging agents. Oncogene, 8, 307–318. 43. Kastan,M.B., Onyekwere,O., Sidransky,D., Vogelstein,N. and Craig,R.W. (1991) Participation of p53 protein in the cellular response to DNA damage. Cancer Res., 51, 6304–6311. 44. Kuerbitz,S.J., Plunkett,B.S., Walsh,W.V. and Kastan,M.B. (1992) Wildtype p53 is a cell cycle checkpoint determinant following irradiation. Proc. Natl Acad. Sci. USA, 89, 7491–7495. 45. Nelson,W.G. and Kastan,M.B. (1994) DNA strand breaks: the DNA template alterations that trigger p53-dependent DNA damage response pathways. Mol. Cell. Biol., 14, 1815–1823. 46. Zhan,Q., Carrier,F. and Fornace,A.J.,Jr (1993) Induction of cellular p53 activity by DNA-damaging agents and growth arrest. Mol. Cell. Biol., 13, 4242–4250. 47. Hickman,A.W., Jaramillo,R.J., Lechner,J.F. and Johnson,N.F. (1994) αParticle-induced p53 protein expression in a rat lung epithelial cell strain. Cancer Res., 54, 5797–5800. 48. Hess,R., Plaumann,B., Lutum,A.S., Haessler,C., Heinz,B., Fritsche,M. and Brandner,G. (1994) Nuclear accumulation of p53 in response to treatment with DNA-damaging agents. Toxicol. Lett., 72, 43–52. 49. El-Deiry,W.S. et al. (1994) WAF1/CIP1 is induced in p53-mediated G1 arrest and apoptosis. Cancer Res., 54, 1169–1174. 50. Canman,C.E., Gilmer,T.M., Coutts,S.B. and Kastan,M.B. (1995) Growth factor modulation of p53-mediated growth arrest versus apoptosis. Genes Dev., 9, 600–611. 51. Lane,D.P. and Crawford,L.V. (1979) T antigen is bound to a host protein in SV40-transformed cells. Nature, 278, 261–263. 52. Narayan,S., Beard,W.A. and Wilson,S.H. (1995) DNA damage-induced transcriptional activation of a human DNA polymerase β chimeric promoter: recruitment of preinitiation complex in vitro by ATF/CREB. Biochemistry, 34, 73–80. 53. Benford,D.J., Price,S.C., Lawrence,J.N., Grasso,P. and Bremmer,J.N. (1994) Investigations of the genotoxicity and cell proliferative activity of dichlorvos in mouse forestomach. Toxicology, 92, 203–215. 54. Bean,C.L., Bradt,C.I., Hill,R., Johnson,T.E., Stallworth,M. and Galloway,S.M. (1994) Chromosome aberrations: persistence of alkylation damage and modulation by O6-alkylguanine-DNA alkyltransferase. Mutat. Res., 307, 67–81. 55. Aidoo,A., Lyn-Cook,L.E., Lensing,S. and Wamer,W. (1994) Ascorbic acid (vitamin C) modulates the mutagenic effects produced by an alkylating agent in vivo. Environ. Mol. Mutagen., 24, 220–228. 56. Massey,T.E., Stewart,R.K., Daniels,J.M. and Liu,L. (1995) Biochemical and molecular aspects of mammalian susceptibility to aflatoxin B1 carcinogenicity. Proc. Soc. Exp. Biol. Med., 208, 213–227. 57. Stoloff,L. (1977) Aflatoxins—An Overview. Pathtox Publishers, Park Forest South, IL. 58. National Toxicology Program (1991) Sixth Annual Report on Carcinogens: 1991 Summary. US Department of Health and Human Services, Washington, DC. 59. Penn,I. (1988) Tumors of the immunocompromised patient. Annu. Rev. Med., 39, 63–73. 60. Silman,A.J., Petrie,J., Hazleman,B. and Evans,S.J.W. (1988) Lymphoproliferative cancer and other malignancy in patients with

rheumatoid arthritis treated with azathioprine: a 20 year follow up study. Annls Rheum. Dis., 47, 988–992. 61. Confavreux,C., Saddier,P., Grimaud,J., Moreau,T., Adeleine,P. and Aimard,G. (1996) Risk of cancer from azathioprine therapy in multiple sclerosis: a case–control study. Neurology, 46, 1607–1612. 62. Gaylor,D.W., Kadlubar,F.F. and West,R.W. (1988) Estimates of the risk of bladder tumor promotion by saccharin in rats. Regulat. Toxicol. Pharmacol., 8, 467–470. 63. National Toxicology Program (1995) Toxicology and Carcinogenesis Studies of Benzethonium Chloride (CAS 121-54-0) in F344/N Rats and B6C3F1 Mice (Dermal Studies), NTP technical report 438. US Department of Health and Human Services, Washington, DC. 64. Kolachana,P., Subrahmanyam,V.V., Meyer,K.B., Zhang,L. and Smith,M.T. (1993) Benzene and its phenolic metabolites produce oxidative DNA damage in HL60 cells in vitro and in the bone marrow in vivo. Cancer Res., 53, 1023–1026. 65. Yager,J.W., Eastmond,D.A., Robertson,M.L., Paradisin,W.M. and Smith,M.T. (1990) Characterization of micronuclei induced in human lymphocytes by benzene metabolites. Cancer Res., 50, 393–399. 66. Barale,R., Marrazzini,A., Betti,C., Vangelisti,V., Loprieno,N. and Barrai,I. (1990) Genotoxicity of two metabolites of benzene: phenol and hydroquinone show strong synergistic effects in vivo. Mutat. Res., 244, 15–20. 67. Hellmer,L. and Bolcsfoldi,G. (1992) An evaluation of the E. coli K-12 uvrB/recA DNA repair host-mediated assay. I. In vitro sensitivity of the bacteria to 61 compounds. Mutat. Res., 272, 145–160. 68. Batiste-Alentorn,M., Xamena,N., Creus,A. and Marcos,R. (1994) Further studies with the somatic white-ivory system of Drosophila melanogaster: genotoxicity testing of ten carcinogens. Environ. Mol. Mutagen., 24, 143–147. 69. Adams,S.P., Laws,G.M., Storer,R.D., DeLuca,J.G. and Nichols,W.W. (1996) Detection of DNA damage induced by human carcinogens in acellular assays: potential application for determining genotoxic mechanisms. Mutat. Res., 368, 235–248. 70. Liehr,J.G. (1990) Genotoxic effects of estrogens. Mutat. Res., 238, 269–276. 71. Gladek,A. and Liehr,J.G. (1989) Mechanism of genotoxicity of diethylstilbestrol in vivo. J. Biol. Chem., 264, 16847–16852. 72. Moore,M.M., Parker,L., Huston,J., Harrington-Brock,K. and Dearfield,K.L. (1991) Comparison of mutagenicity results for nine compounds evaluated at the hgprt locus in the standard and suspension CHO assays. Mutagenesis, 6, 77–85. 73. Przybojewska,B., Dziubaltowska,E. and Kowalski,Z. (1984) Genotoxic effects of ethyl acrylate and methyl acrylate in the mouse evaluated by the micronucleus test. Mutat. Res., 135, 189–191. 74. Moore,M.M., Amtower,A., Doerr,C.L., Brock,K.H. and Dearfield,K.L. (1988) Genotoxicity of acrylic acid, methyl acrylate, ethyl acrylate, methyl methacrylate, and ethyl methacrylate in L5178Y mouse lymphoma cells. Environ. Mol. Mutagen., 11, 49–63. 75. Wiegand,H.J., Schiffmann,D. and Henschler,D. (1989) Non-genotoxicity of acrylic acid and n-butyl acrylate in a mammalian cell system (SHE cells). Arch. Toxicol., 63, 250–251. 76. Wu,H. and Lozano,G. (1994) NF-κB activation of p53. J. Biol. Chem., 269, 20067–20074. 77. Kessis,T.D., Slebos,R.J., Nelson,W.G., Kastan,M.B., Plunkett,B.S., Han,S.M., Lorincz,A.T., Hedrick,L. and Cho,K.R. (1993) Human papillomavirus 16 E6 expression disrupts the p53-mediated cellular response to DNA damage. Proc. Natl Acad. Sci. USA, 90, 3988–3992. 78. Reich,N.C. and Levine,A.J. (1984) Growth regulation of a cellular tumor antigen, p53, in nontransformed cells. Nature, 308, 199–201. 79. Gronostajski,R.M., Goldberg,A.L. and Pardee,A.B. (1984) Energy requirement for degradation of tumor-association protein p53. Mol. Cell. Biol., 4, 442–448. 80. Band,V., Dalal,S., Delmolino,L. and Androphy,E.J. (1993) Enhanced degradation of p53 protein in HPV-6 and BPV-1 E6-immortalized human mmamary epithelial cells. EMBO J., 12, 1847–1852. 81. Tennant,R.W., French,J.E. and Spalding,J.W. (1995) Identifying chemical carcinogens and assessing potential risk in short-term bioassays using transgenic mouse models. Environ. Hlth Perspect., 103, 942–950. Received on September 5, 1997; revised on January 8, 1998; accepted on January 20, 1998

1125