Artlesp53 Induction as a Genotoxic Test for Twenty-Five Chemicals Undergoingin Vivo Carcinogenicity TestingPenelope J. Duerksen-Hughes, Jun Yang, and Ozan OzcanDepartment of Biology, Georgia State University, Atlanta, Georgia, USA

In vivo carcinogenicity testing is an expensive and time-consuming process, and as a result, only arelatively small fraction of new and existing chemicals has been tested in this manner. Therefore,the development and validation of alternative approaches is desirable. We previously developed amammalian in vitro assay for genotoxicity based on the ability of cells to increase their level of thetumor-suppressor protein p53 in response to DNA damage. Cultured cells are treated with van-ous amounts of the test substances, and at defined times following treatment, they are harvestedand lysed. The lysates are analyzed for p53 by Westem blot and/or enzyme-linked immunosor-bent assay analysis. An increase in cellular p53 following treatment is interpreted as evidence forDNA damage. To determine the ability of this p53-induction assay to predict carcinogiy inrodents and to compare such results with those obtained USing alternate approaches, we subjected25 chemicals from the predictive toxicology evaluation 2 list to analysis with this method. Fivesubstances (citral, cobalt sulhate heptahydrate, D&C Yellow No. 11, oymetholone, and t-butyl-hydroquinone) tested positive in this assay, and three substances (emodin, phenolphthalein, andsodium ylenesulfonate) tested as possibly positive. Comparisons between the results obtainedwith this assay and those obtained with the in vivo protocol, the Salmonella assay, and the Syrianhamster embryo (SHE) cell assay indicate that the p53-induction assay is an excellent predictor ofthe limited number of genotoxic carcinogens in this set, and that its accuracy is roughly equiva-lent to or better than the SalmonceUa and SHE assays for the complete set of chemicals. Key workr.assay development, carcinogenicity, genotoxicity, p53, PTE2. Environ Health Perspect107:805-812 (1999). [Online 1 September 1999]hbp://ehpnetl.niebs.ni.gov/ldocs/l999/107p805-812duerksen-hughes/abstracst.bnl

Because society has made human health andsafety an important consideration, consid-erable resources have been and are beingexpended in efforts to identify and classifyhuman carcinogens. Unfortunately, methodscurrently used, such as the Ames test (1) andin vivo animal testing, suffer from severalshortcomings. These include the need toextrapolate from prokaryotes to humans(Ames test) and the need to extrapolate fromrodents to humans, the high cost, and thelong period before results are known (in vivoanimal testing). With all current methodsthere is also limited predictivity, both whenexamining the agreement of various methodswith each other (2-6) and when applying lab-oratory results to actual human populations.

To aid in the development and valida-tion of improved ways to predict the car-cinogenicity of compounds, the NationalInstitute of Environmental Health Sciences(NIEHS; Research Triangle Park, NC) initi-ated the predictive toxicology evaluationproject (PTE). This project listed two sets ofchemicals undergoing in vivo carcinogenicitytesting; the first set of 44 chemicals waslisted in 1990 (PTE1) and the second set of30 chemicals in 1994 (PTE2) (7,8).Researchers were invited to submit predic-tions regarding the ability of these chemicalsto induce carcinogenesis in rodents, and thepredictions and actual in vivo results wereto be compared when the animal results

became available. The results from PTE1provided information regarding the featuresof chemicals most predictive of their abilityto act as carcinogens, and the results fromPTE2 will add to this dataset once all of thein vivo results are available for comparisonwith the predictions.

We previously developed a novel methodfor assessing the genotoxicity of substancesthat is simple, rapid, and cost-effective. Thismethod is based on the biologic response ofmammalian cells in culture to agents thatdamage their DNA (9). This assay is basedon the well-documented observation that thecellular level of a tumor-suppressor proteincalled p53 increases following DNA damage(10-15), primarily because of an increase inthe stability of the protein and a resultingdecrease in its rate of proteolysis (9,10). Thep53 then acts to prevent replication ofdamaged DNA, either by causing the cellto undergo a reversible growth arrest(11,16,17) or by initiating a cell's apoptoticpathway (18-21).

The assay itself is simple (9). Culturedmammalian cells (NCTC 929) are treatedwith various doses of the test agent, and atspecified time points following treatment, thecells are harvested and lysed. The level of p53in the lysates is measured by p53 enzyme-linked immunosorbent assay (ELISA) and/orWestern blot analysis, and compared to thelevel in untreated cells.

Our previous work (9) indicated thatboth Western blot and ELISA analyses yield-ed similar results. Therefore, our currentwork focused on the more quantitativeELISA assay. We also found that the NCTC929 cell line was a useful model system.NCTC 929 cells were initially selectedbecause previously published results (10)indicated that their levels of p53 rise follow-ing genotoxin treatment. We verified this,and found in addition that p53 in these cellscan be immunoprecipitated by a monoclonalantibody specific for wild-type p53. We alsofound that the level of p53 in these cells risessignificantly following treatment with indi-rect genotoxins such as aflatoxin BI and 2-acetylaminofluorene, indicating that theypossess significant amounts of the metabolicactivities necessary for biotransformation (9.

To determine the ability of this assay topredict carcinogenicity in rodents and tocompare its performance with other proposedalternatives, we subjected 25 of the 30 PTE2substances to analysis by this method. Fivesubstances (citral, cobalt sulfate heptahydrate,D&C Yellow No. 11, oxymetholone, and t-butylhydroquinone) tested positive in thisassay and three substances (emodin, phe-nolphthalein, and sodium xylenesulfonate)tested as possibly positive. Comparisonsbetween the results obtained with this assayand those obtained with the in vivo protocol,the Salmonella assay, and the Syrian hamsterembryo (SHE) cells indicate that the p53-induction assay is an excellent predictor of thelimited number of genotoxic carcinogens inthis set, and that its accuracy is roughly equiv-alent to or better than the Salmonella andSHE assays for the complete set of chemicals.

Materials and MethodsCell culture. These experiments were doneusing the NCTC 929 cell line derivedfrom mouse fibroblasts and obtained fromthe American Type Culture Collection

Address correspondence to P.J. Duerksen-Hughes,Department of Biology, Georgia State University,PO Box 4010, Atlanta, GA 30302 USA. Telephone:(404) 651-0987. Fax: (404) 651-2509. E-mail:biopjd@panther.gsu.eduWe acknowledge the assistance and advice of D.

Bristol and R. Tennant at the NIEHS, the help ofT. Poole and V. Rehder for the statistical analysis,and the contributions of M. Kale and M. Zaki dur-ing the early stages of this project. This work wassupported in part by NSF grant MCB 9513527 toP.J. Duerksen-Hughes.Received 7 April 1999; accepted 11 May 1999.

Environmental Health Perspectives * Volume 107, Number 10, October 1999 805

Articles * Duerksen-Hughes et al.

(Rockville, MD). Cells were cultured inmodified Eagle media (Gibco/Life Sciences,Gaithersburg, MD) supplemented with 10%fetal calf serum (Gibco) (MEM 10). Cellswere plated at a density of 0.5-1.5 x 106cells per 100 mm plate 1 day prior to treat-ment. pAB 122 hybridoma cells, alsoobtained from the American Type CultureCollection, were propagated in Dulbecco'smodified Eagle medium supplemented with10% fetal calf serum, and served as thesource for the primary anti-p53 antibodyused in the ELISA analysis.

Chemicals. Fourteen of the 25 substancestested were obtained from the NIEHS. Thesesubstances and their Chemical AbstractService (CAS) numbers are as follows: codeine(76-57-3); D&C Yellow No. 11 (8003-22-3);diethanolamine (111 -42-2); 1,2-dihydro-2,2,4-trimethylquinoline (147-47-7); eth-ylbenzene (100-41-4); ethylene glycolmonobutyl ether (111-76-2); furfuryl alcohol(98-00-0); gallium arsenide (1303-00-0);methyleugenol (93-15-2); oxymetholone(434-07-1); phenolphthalein (77-09-8); pri-maclone (125-33-7); pyridine (110-86-1);and scopolamine hydrobromide trihydrate(6533-68-2). Ten substances were obtainedfrom Sigma/Aldrich (St. Louis, MO):1-chloro-2-propanol (127-00-4), citral(5392-40-5), cobalt sulfate heptahydrate(10026-24-1), emodin (518-82-1), isobu-tyraldehyde (78-84-2), nitromethane (75-52-5), sodium nitrate (7632-00-0), sodiumxylenesulfonate (1300-72-7), t-butylhydro-quinone (1948-33-0), and tetrahydrofuran(109-99-9). Cinnamaldehyde (104-55-2)was obtained from Acros (Pittsburgh, PA).

Unless noted otherwise, 10 mg/mL stocksolutions of the test substances were pre-pared in either culture media (MEM 10) orin dimethylsulfoxide (DMSO). We havepreviously shown that DMSO alone at theconcentrations used does not induce p53 intreated cells (9).

Cell treatment. Volumes of the stocksolution corresponding to the doses testedwere added to a series of plates (approximate-ly 1 x 106 cells per 100 mm plate in a vol-ume of 10-mL MEM 10). Control plateswere treated with an amount of vehicle alone(media or DMSO) that corresponded to thehighest volume added to the experimentalplates. For each chemical, several dosesbetween 1 and 100 pg/mL of the test sub-stance were evaluated; the exact doses usedare listed in the sections for the separatechemicals. Treated and control cells wereincubated at 37°C and 5% CO2 until har-vest. One series was harvested approximately6 hr posttreatment, and a second series washarvested approximately 17 hr posttreatment.Each chemical was tested, at the severaldosages and at the two time points, between

two and four times. In some cases, specificdoses were repeated in additional trials.

To ensure that the p53 response of thesecells to both direct- and indirect-actinggenotoxic chemicals remained intact and rel-atively constant during these experiments,cultures of NCTC 929 cells from passagescurrently in use were periodically treatedwith known direct (N-methyl-N'-nitro-nitrosoguanidine) and indirect (mitomycinC) genotoxins (9) to verify that the cellscould adequately respond to both types ofchemicals. We found that the cellularresponse to both direct- and indirect-actingchemicals was stable throughout and beyondthe course of these experiments.

Cell lysis andp53 ELISA. Cells were har-vested, lysed, and analyzed for their proteinconcentration by the BCA assay (Pierce,Rockford, IL) and for their p53 levels byELISA, as described previously (9). Briefly,cells were removed from the plate bytrypsinization, concentrated by centrifuga-tion, and suspended in lysis buffer. After lysis,lysates were stored at -80°C for no more than1 week prior to analysis. The ELISA analysisused pAB 122 as the primary or capture anti-body, biotinylated anti-p53 (BoehringerMannheim, Annapolis, IN) as the detectionantibody, and GST-p53 (Santa CruzBiotechnology, Inc., Santa Cruz, CA) as astandard. Each point was measured in tripli-cate, and the results were normalized to theamount of protein present in each sample.

Analysis. Data were analyzed using one-way analysis of variance software (SuperAnova,Abacus Concepts, Berkeley, CA). The ScheffeS-test was used post hoc to identify significantdifferences between p53 levels in treated anduntreated cells. Test results were interpretedas positive if one or more doses yielded a sig-nificant increase (p < 0.001) in the level ofp53 in treated versus untreated cells, in twoor more separate experiments, at one or both

.5aCa.

._

2

20 '

14

12

1016sg

.420 .Z

0 1 10 20Dose (g/mL)

50 100

of the times tested (6 and 17 hr), and withno more than one of three or four experi-ments failing to show such an increase.Results were interpreted as possibly positiveusing the same criteria, except that the sig-nificance of the increase was less (p < 0.05).

Data regarding the Salmonella and invivo results (where available) were obtainedfrom the PTE2 website (8).

ResultsCobalt sulfate heptahydrate and D& CYellow No. 11 induce p53 strongly by 6 hr.Two substances, cobalt sulfate heptahydrateand D&C Yellow No. 11, caused increases incellular p53 by 6 hr. The increases for cobaltsulfate heptahydrate were significant at both50 and 100 pg/mL, and those for D&CYellow No. 11 were significant at 5, 10, and25 pg/mL. On the basis of these results, theyare classified as positive. Figure 1 depicts rep-resentative experiments for cells treated witheach of these substances and harvested at 6hr. For these two substances, the increase inthe cellular level of p53 in cells treated withthe statistically significant doses is robust-greater than a 4-fold increase over that seenin untreated cells.

Citral cobalt sulfate heptahydrate, D&CYellow No. 11, oxymetholone, andt-butylhy-droquinone inducep53 strongly by 17 hr. Inaddition to the two substances that inducedp53 by 6 hr, three additional substancescaused significant increases in the cellularlevel of p53 by 17 hr (citral, oxymetholone,and t-butylhydroquinone). The increase forcitral was significant at doses of 10, 15, 20,25, and 30 pg/mL; some cytotoxicity wasnoted at 25 and 30 pg/mL and increasedcytotoxicity at 50 pg/mL. The increases forcobalt sulfate heptahydrate were significant atdoses of 20 and 50 pg/mL; significant cyto-toxicity was noted in cells treated with 100ig/mL for 17 hr. Cells treated with 25

e0-0

._

0EE

30

25

20

15

10

5

5Dose (j.g/mL)

10 25

Figure 1. Chemicals that strongly increased cellular p53 levels by 6 hr. (A) Cobalt sulfate heptahydrate. (B)D&C Yellow No. 11. Cultured mammalian (NCTC 929) cells were treated with the indicated doses of thetest substances, then harvested and lysed at 6 hr posttreatment. The level of p53 present in the lysateswas quantitated by enzyme-linked immunosorbent assay, as described in "Materials and Methods." Errorbars represent standard error.

Volume 107, Number 10, October 1999 * Environmental Health Perspectives806

Articles * p53 induction results for 25 PTE2 chemicals

ig/mL D&C Yellow No. 11 for 17 hrshowed a significant increase in their cellularp53 level; cells treated with a higher level dis-played cytotoxicity. Oxymetholone causedsignificant increases in the p53 level at dosesof 10 and 25 pg/mL, with increasing cytotox-icity noted at doses of 25 and 50 pg/mL.Finally, t-butylhydroquinone significantlyincreased cellular p53 levels at doses of 1, 5,and 10 pg/mL, with significant cytotoxicitynoted at higher doses. On the basis of theseresults, these five substances are classified aspositive. Figure 2 depicts representativeexperiments for cells treated with each ofthese substances and harvested at 17 hr. Foreach of the substances that tested positive,the increase in the cellular level of p53 incells treated with the statistically significantdoses is robust-greater than a 4-foldincrease over that seen in untreated cells.

70

CL

20

Dose (tgIlmL)

30

. E 2.5

E _1020

5 _B

0 1 5 10 25 50Dose (gg/mL)

15 1

14

12

f10 -

CLA

a. o 5 5

a.~ ~ ~ Doe(#L

Emodin, phenolphthalein,xylenesulfonate inducep53 less s,hr. Emodin increased cellular pdose of 20 pg/mL, phenolphthaof 10 pg/mL by 17 hr, and snesulfonate at doses of 1 and 5 pcase of these three chemicals, thof the increase was less than thatfive chemicals previously discithey are dassified as possible po%3 depicts representative experimtreated with each of these substavested at 17 hr. The extent of tdp53 was reduced (approximatelyover that seen in untreated cells)to that of the previous five substa

Seventeen chemicals didcellular levels ofp53. The rem;teen chemicals (1-chloro-2cinnamaldehyde, codeine, diet

E

CD

Ea.

U 1U

Dose ({g/mL)30

'20020Ct

CD 15E

a.

DDose (jglmiL)

Figure 2. Chemicals thincreased cellular p53 levi(A) Citral. (B) Cobalt sulfadrate. (C) D&C YellowOxymetholone. (E) t-ButylhlCultured mammalian (NC1were treated with the indicEthe test substances, then hlysed at 17 hr posttreatmenlp53 present in the lysates ued by ELISA, as describedand Methods.' Error barstandard error.

I a

Dose (ig/mL)

and sodium 1,2-dihydro-2,2,4-trimethylquinoline, ethyl-trongly by 17 benzene, ethylene glycol monobutyl ether,53 levels at a furfuryl alcohol, gallium arsenide, isobu-lein at a dose tyraldehyde, methyleugenol, nitromethane,;odium xyle- primaclone, pyridine, scopolamine hydrobro-ig/mL. In the mide trihydrate, sodium nitrite, and tetrahy-e significance drofuran) were unable to reproducibly andnoted for the significantly increase cellular p53 levels at theussed; hence, times and doses tested. Based on thesesitives. Figure results, they are classified as negative.ients for cells Substances classified on the basis oftheirnces and har- ability to increase cellular p53. Table 1 listshe increase in the substances that tested positive (p < 0.001)K 2- to 3-fold or possibly positive (p < 0.05) in this assay,as compared along with the time(s) and dose(s) at whichmces. positive results were achieved. Five substancesnot increase (citral, cobalt sulfate heptahydrate, D&Caining seven- Yellow No. 11, oxymetholone, and t-butyl-2-propanol, hydroquinone) gave clearly positive signals.thanolamine, Cobalt sulfate heptahydrate and D&C

Yellow No. 11 were positive at both the earlyand late time points, whereas the remaining

'g=|ts three were positive only at the later timepoint. Three substances (emodin, phenolph-thalein and sodium xylensulfonate) werepossibly positive in our assay. That is, theydemonstrated the ability to reproducibly

-|i1X k induce p53 at at least one tested dose,although the increase was less significant thanseen with the previous five substances.

Summary. Table 2 shows the summarydata for the substances that we tested as well

_ as previously reported results from the20 50 Salmonella and SHE assays (22). It also lists a

summary of the in vivo test results where avail-able. In this table, the either- or single-speciesin vivo test results are listed as positive if thereis some or dear evidence of carcinogenicity inany of the four treatment groups; thetransspecies results are listed as positive only ifthere was some or dear evidence of carcino-genicity in each of the two tested species.

Specific notes regarding the individualsubstances tested follow. Classifications (pos-itive, possibly positive, and negative) refer tothe results of this study.

1-Chloro-2-propanol. A 10-mg/mLstock of 1 -chloro-2-propanol was prepared in

10 25 DMSO, and cells were treated with doses of1, 5, 10, 25, and 50 pglmL. Cells treatedwith 1-chloro-2-propanol did not display

eat strongly cytotoxicity at any of the times or doses

3te heptahy- tested, and 1-chloro-2-propanol did notNo. 11. (D) induce p53 at any of the doses or times test-ydroquinone ed. 1-Chloro-2-propanol is therefore classi-FC 929) cells fied as negative by this method; this contrastsated doses of with positive results in the Salmonella testiarvested and and the SHE assay. The in vivo test resultst. The level ofwas quantitat- were also negative.in 'Materials Cinnamaldehyde. A 10-mg/mL stock ofrs represent cinnamaldehyde was prepared in DMSO, and

cells were treated with doses of 1, 10, 20, and50 [ig/mL. Cells treated with cinnamaldehydedisplayed cytotoxicity at the highest dose

Environmental Health Perspectives * Volume 107, Number 1 0, October 1999

iu

807

Articles * Duerksen-Hughes et al.

tested (50 gg/mL) at both 6 and 17 hr.Cinnamaldehyde did not induce p53 at any ofthe doses or times tested, and is thereforeclassified as negative. The Salmonella resultswere weakly positive.

Citral. A 10-mg/mL stock of citral wasprepared in DMSO, and cells were treatedwith doses of 1, 5, 10, 15, 20, 25, 30, and 50pg/mL. Cells treated with citral displayedincreasing cytotoxicity at the highest dosestested (25, 30, and 50 mg/mL) by 6 and 17hr. Citral induced p53 by 17 hr, and is there-fore classified as positive. This contrasts withthe results from the Salmonella test, whichwas negative.

Cobalt sulfate heptahydrate. A 10-mg/mL stock of cobalt sulfate heptahydratewas prepared in H20, and cells were treatedwith doses of 1, 10, 20, 50, and 100 pig/mL.Cells treated with cobalt sulfate heptahydratedisplayed cytotoxicity when treated with thesubstance at 100 pg/mL for 17 hr. Cobalt sul-fate heptahydrate strongly induced p53 atboth the 6 and 17 hr time points, and istherefore dassified as positive. The Salmonellatest gave a weakly positive response, and theSHE assay was also positive. Animal tests haveshown some evidence for carcinogenicity inmale rats and dear evidence for carcinogenici-ty in female rats, male mice, and female mice.

Codeine. A 10-mg/mL stock of codeinewas prepared in DMSO, and cells were treat-ed with doses of 1, 5, 10, 25, and 50 Pg/mL.Cells treated with codeine did not displaycytotoxicity at any of the times or doses test-ed, and codeine did not induce p53 at any ofthe doses or times tested. Codeine is there-fore classified as negative. This is in agree-ment with the negative test results seen in theSalmonela and SHE assays, and a lack of evi-dence for carcinogenicity seen in the animaltests.D&C Yellow No. 11. A 5-mg/mL stock

of D&C Yellow No. 11 was prepared inDMSO, and cells were treated with doses of1, 5, 10, 25, and 50 pg/mL. Cells treated withD&C Yellow No. 11 displayed some cytotox-icity at 50 pg/mL by 17 hr. The actual dosesreceived by the cells were less than thoseadministered, as some precipitation of thesubstance was observed after its addition tothe media. Cells took up the dye; cell pelletswere yellow. This color partitioned with themembrane fraction during centrifugation ofthe lysate. D&C Yellow No. 11 stronglyinduced p53 at both 6 and 17 hr, and istherefore classified as positive. The Salmonellatest gave a wealdy positive result, the SHE testresults were positive, and in animal studies,male and female rats showed some evidenceof carcinogenicity.

Diethanolamine. A 10-mg/mL stock ofdiethanolamine was prepared in DMSO,and cells were treated with doses of 1, 5, 10,

25, and 50 pg/mL. Cells treated withdiethanolamine did not display cytotoxicityat any of the times or doses tested, anddiethanolamine did not induce p53 at any ofthe doses or times tested. Diethanolamine istherefore classified as negative. This is inagreement with a negative result in theSalmonella assay, and contrasts with a posi-tive result from the SHE test. During animaltesting, no evidence for carcinogenicity was

._

CD0,ECLCDE

3.5t(D 3.00

` 2.5

c 2.0CD

1.52 1.5

1.0

0.5

0 1

seen in male or female rats, whereas clear evi-dence was observed in male and female mice.

1 ,2-Dihydro-2,2,4-trimethylquinoline(DTQ). A 10-mg/mL stock ofDTQwas pre-pared in DMSO, and cells were treated withdoses of 1, 5, 10, 25, and 50 jig/mL. Cellstreated with DTQ did not display cytotoxici-ty at any of the times or doses tested, andDTQ did not induce p53 at any of the dosesor times tested. DTQ is therefore dassified as

3.0 A...........Q

; 2

E0.CLCD1.a.

Figure 3. Chemicals that increasedcellular p53 levels to a lesser extent.(A) Emodin. (B) Phenolphthalein. (C)Sodium xylenesulfonate. Culturedmammalian (NCTC 929) cells weretreated with the indicated doses ofthe test substances, then harvestedand lysed at 17 hr. The level of p53present in the lysates was quantitat-ed by enzyme-linked immunosorbentassay, as described in 'Materialsand Methods." Error bars representstandard error.

Dose (jg/mL)

Table 1. Positive and possibly positive substances.

Time(s) of Dose(s) of p-ValuePTE2 no. Chemical observed induction observed induction (exp 1, exp 2)Positive5 Citral 17 hr 10 pg/mL 0.0019, 0.0001

15 pg/mL 0.0001, 0.000120 pg/mL 0.0001, 0.000125 pg/mL 0.0001, 0.000130 pg/mL 0.0001, 0.0001

6 Cobalt sulfate 6 hr 50 pg/mL 0.0001, 0.0001heptahydrate 100 pg/mL 0.0001, 0.0001

17 hr 20 pg/mL 0.0001, 0.000150 pg/mL 0.0001, 0.0001

8 D&C Yellow No. 1 1 6 hr 5 pg/mL 0.0016, 0.000110 pg/mL 0.0001, 0.000125 pg/mL 0.0015, 0.0001

17 hr 25 pg/mL 0.0001, 0.048521 Oxymetholone 17 hr 10 pg/mL 0.0001, 0.0001

25 pg/mL 0.0001, 0.000128 t-Butylhydroquinone 17 hr 1 pg/mL 0.0001, 0.0001

5 pg/mL 0.0001, 0.000110 pg/mL 0.0001, 0.0001

Possibly positive1 1 Emodin 17 hr 20 pg/mL 0.0110, 0.043522 Phenolphthalein 17 hr 10 pg/mL 0.0387, 0.003927 Sodium xylensulfonate 17 hr 1 pg/mL 0.0035, 0.0001

5 pg/mL 0.0001, 0.0025Abbreviations: exp, experiment; PTE2, predictive toxicology evaluation 2.

Volume 107, Number 10, October 1999 * Environmental Health Perspectives

0.

0 0.5 1 10 20 0 1 0iDose (g/mL) Dose (gImL)

mU q

808

Articles * p53 induction results for 25 PTE2 chemicals

negative. This is in agreement with a negativeresult from the Salmonella assay, and contrastswith a positive result from the SHE test. Theanimal results list some evidence for carcino-genicity in male rats and no evidence infemale rats, male mice, and female mice.

Emodin. A 10-mg/mL stock of emodinwas prepared in DMSO, and cells were treatedwith doses of 0.5, 1, 10, 20, and 50 pg/mL.Cells treated with emodin displayed somecytotoxicity when treated with the substanceat 50 ,ug/mL for 17 hr. Emodin weaklyinduced p53 by 17 hr, and is therefore dassi-fied as possibly positive. The Salmonelal andSHE tests gave positive responses.

Ethylbenzene. A 10-mg/mL stock ofethylbenzene was prepared in DMSO, andcells were treated with doses of 1, 10, 20, and50 gg/mL. Cells treated with ethylbenzenedid not display significant cytotoxicity at anyof the times or doses tested, and ethylbenzenedid not induce p53 at any of the doses ortimes tested. Ethylbenzene is therefore dassi-fied as negative. This is in agreement with anegative result from the Salmonella assay andcontrasts with a positive result from the SHEtest. The animal results list dear evidence forcarcinogenicity in male rats, and someevidence in female rats, male mice, andfemale mice.

Ethylene glycol monobutyl ether(EGMBE). A 10-mg/mL stock of EGMBE

was prepared in MEM 10, and cells weretreated with doses of 1, 5, 10, 25, and 50pig/mL. Cells treated with EGMBE did notdisplay cytotoxicity at any of the times ordoses tested, and EGMBE did not inducep53 at any of the doses or times tested.EGMBE is therefore classified as negative.This is in agreement with a negative resultfrom the Salmonella assay and contrasts witha positive result from the SHE assay. Theanimal results list some evidence in male andfemale mice, no evidence in male rats, andequivocal evidence in female rats.

Furfuryl alcohol. A 10-mglmL stock offurfuryl alcohol was prepared in MEM 10,and cells were treated with doses of 1, 5, 10,25, and 50 pg/mL. Cells treated with furfurylalcohol did not display cytotoxicity at any ofthe times or doses tested, and furfuryl alcoholdid not induce p53 at any of the doses ortimes tested. Furfuryl alcohol is thereforeclassified as negative. This is in agreementwith negative results from the Salmonella andthe SHE assays. The animal results list equiv-ocal evidence for carcinogenicity in femalerats, some evidence in male rats and mice,and no evidence in female mice.

Gallium arsenide. A 1-mg/mL stock ofgallium arsenide was prepared in DMSO,and cells were treated with doses of 1, 10,20, and 50 pg/mL. The actual dose receivedby the cells is lower than that listed, as some

Table 2. Summary and comparisons.

PTE2 Salmonella p53 SHE In vivono. Name CAS no. results induction cell results resultsa1 Anthraquinone 84-65-1 + NT2 Chloroprene 126-99-8 - NT NT +/+3 1-Chloro-2-propanol 127-00-4 + - + -/-4 Cinnamaldehyde 104-55-2 +w - NT5 Citral 5392-40-5 + NT6 Cobalt sulfate heptahydrate 10026-24-1 +w + + +/+7 Codeine 76-57-3 - - - -/-8 D&C Yellow No. 11 8003-22-3 +w + + +/NT9 Diethanolamine 111-42-2 - - + +/-10 1,2-Dihydro-2,2,4-trimethylquinoline 147-47-7 - - + +/-11 Emodin 518-82-1 + Possibly + +12 Ethylbenzene 100-41-4 - - + +/+13 Ethylene glycol monobutyl ether 111-76-2 - - + +/-14 Furfuryl alcohol 98-00-0 - - - +/+15 Gallium arsenide 1303-00-0 - - +16 Isobutene 115-11-7 - NT NT +/-17 Isobutyraldehyde 78-84-2 ? - - -/-18 Methyleugenol 98-15-5 - - + +/+19 Molybdenum trioxide 1313-27-5 - NT + +/-20 Nitromethane 75-52-5 - - + +/+21 Oxymetholone 434-07-1 - + + +/NT22 Phenolphthalein 77-09-8 - Possibly + + +/+23 Primaclone 125-33-7 + - - +/-24 Pyridine 110-868-1 - - - +/+25 Scopolamine hydrobromide trihydrate 6533-68-2 - + -/-26 Sodium nitrite 7632-00-0 + - +27 Sodium xylenesulfonate 1300-72-7 - Possibly + - -/-28 t-Butylhydroquinone 1948-33-0 - + - -/-29 Tetrahydrofuran 109-99-9 - - - +/+30 Vanadium pentoxide 1314-62-1 - NT +

Abbreviations: +, positive; -, negative; CAS, Chemical Abstracts Registry; NT, not tested; PTE2, predictive toxicologyproject 2; SHE, Syrian hamster embryo; w, weak."Either species/transspecies.

precipitation could be observed followingthe addition of the substance to the media.Cells treated with gallium arsenide displayedsome cytotoxicity at the highest doses tested(20 and 50 pg/mL), and gallium arsenidedid not induce p53 at any of the doses ortimes tested. Gallium arsenide is thereforeclassified as negative. This is in agreementwith a negative result from the Salmonellaassay and contrasts with a positive resultfrom the SHE assay.

Isobutyraldehyde. A 10-mg/mL stock ofisobutyraldehyde was prepared in DMSO,and cells were treated with doses of 1, 10, 20,50, and 100 ,glmL. Cells treated with isobu-tyraldehyde did not display cytotoxicity at anyof the times or doses tested, and isobutyralde-hyde did not induce p53 at any of the dosesor times tested. Isobutyraldehyde is thereforeclassified as negative. The Salmonella resultswere incondusive, the SHE results were nega-tive, and no evidence for carcinogenicity wasobserved in male or female rats or in male orfemale mice.

Methyleugenol. A 10-mg/mL stock ofmethyleugenol was prepared in DMSO,and cells were treated with doses of 1, 5,10, 25, and 50 pg/mL. Cells treated withmethyleugenol did not display cytotoxicityat any of the times or doses tested, andmethyleugenol did not induce p53 at any ofthe doses or times tested. Methyleugenol istherefore classified as negative. This is inagreement with a negative result from theSalmonella assay and contrasts with a posi-tive result from the SHE assay. In vivo testresults showed clear evidence for carcino-genicity in male and female rats and maleand female mice.

Nitromethane. A 10-mg/mL stock ofnitromethane was prepared in DMSO, andcells were treated with doses of 1, 10, 20, and50 pg/mL. Cells treated with nitromethanedid not display cytotoxicity at any of thetimes or doses tested, and nitromethane didnot induce p53 at any of the doses or timestested. Nitromethane is therefore classified asnegative. This is in agreement with a negativeresult from the Salmonella assay, and contrastswith a positive result from the SHE assay.The animal results list no evidence for car-cinogenicity in male rats and clear evidence infemale rats, male mice, and female mice.

Oxymetholone. A 1 0-mg/mL stock ofoxymetholone was prepared in DMSO,and cells were treated with doses of 1, 5,10, 25, and 50 pg/mL. Cells treated withoxymetholone displayed cytotoxicity at 50pg/mL by 6 hr and at 25 and 50 pg/mL by17 hr. Oxymetholone strongly induced p53by 17 hr, and is therefore classified as posi-tive. This is in contrast to a negative resultfrom the Salmonella assay, but in agreementwith a positive result from the SHE assay.

Environmental Health Perspectives * Volume 107, Number 10, October 1999 809

Articles a Duerksen-Hughes et al.

Animal results showed equivocal results inmale rats and clear evidence in female rats.Mice were not tested.

Phenolphthalein. A 10-mglmL stock ofphenolphthalein was prepared in DMSO,and cells were treated with doses of 1, 10, 20,40, and 50 pg/mL. Cells treated with phe-nolphthalein did not display cytotoxicity atany of the times or doses tested, and phe-nolphthalein induced p53 weakly by 17 hr.Phenolphthalein is therefore classified as pos-sibly positive. This is in contrast to a negativeresult from the Salmonella assay, and inagreement with a positive result from theSHE assay. The animal results list clear evi-dence for carcinogenicity in male rats, malemice, and female mice, and some evidence infemale rats.

Primaclone. A 10-mg/mL stock ofprimaclone was prepared in DMSO, andcells were treated with doses of 1, 10, 20, and50 pg/mL. Cells treated with primaclone didnot display cytotoxicity at any of the times ordoses tested, and primaclone did not inducep53 at the doses and times tested. Primacloneis therefore classified as negative. This is incontrast with a positive result from theSalmonella assay, but in agreement with anegative result from the SHE assay. The ani-mal results list equivocal evidence for car-cinogenicity in male rats, no evidence infemale rats, and clear evidence in both maleand female mice.

Pyridine. A 10-mg/mL stock of pyridinewas prepared in MEM 10, and cells weretreated with doses of 1, 5, 10, 25, and 50pg/mL. Cells treated with pyridine did notdisplay cytotoxicity at any of the times ordoses tested, and pyridine did not inducep53 at any of the doses or times tested.Pyridine is therefore classified as negative.This is in agreement with negative resultsfrom the Salmonella and SHE assays. Theanimal results list some evidence for carcino-genicity in male rats, equivocal evidence infemale rats, and clear evidence for both maleand female mice.

Scopolamine hydrobromide trihydrate.A 10-mg/mL stock of scopolamine hydro-bromide trihydrate was prepared in DMSO,and cells were treated with doses of 1, 5, 10,25, and 50 jig/mL. Cells treated with scopo-lamine hydrobromide trihydrate did not dis-play cytotoxicity at any of the times or dosestested, and scopolamine hydrobromide trihy-drate did not induce p53 at any of the dosesor times tested. Scopolamine hydrobromidetrihydrate is therefore classified as negative.This is in agreement with a negative resultfrom the Salmonella assay and contrasts witha positive SHE result. The animal results listno evidence for carcinogenicity.

Sodium nitrate. A 10-mg/mL stock ofsodium nitrate was prepared in DMSO, and

cells were treated with doses of 1, 5, 10, 25,and 50 jig/mL. Cells treated with sodiumnitrate did not display cytotoxicity at any ofthe times or doses tested, and sodium nitratedid not induce p53 at any of the times ordoses tested. Sodium nitrite is therefore clas-sified as negative. The Salmonella and SHEresults were positive.

Sodium xylenesulfonate. A 10-mg/mLstock of sodium xylenesulfonate was pre-pared in DMSO, and cells were treated withdoses of 1, 5, 10, 25, and 50 pg/mL. Cellstreated with sodium xylenesulfonate did notdisplay cytotoxicity at any of the times ordoses tested, and sodium xylenesulfonateinduced p53 weakly by 17 hr. Sodiumxylenesulfonate is therefore classified as pos-sibly positive. The Salmonella and SHEresults were negative, and no evidence of car-cinogenicity was observed in male or femalerats or male or female mice.

t-Butylhydroquinone. A 1 0-mg/mLstock of 5-butylhydroquinone was preparedin DMSO, and cells were treated with dosesof 1, 5, 10, 25, and 50 pg/mL. Cells treatedwith t-butylhydroquinone displayed cyto-toxicity at 25 and 50 pg/mL by 17 hr, andt-butylhydroquinone induced p53 by 17 hr.t-Butylhydroquinone is therefore classifiedas positive. This contrasts with negativeSalmonella and SHE assay results. No evi-dence of carcinogenicity was observed inmale or female rats or male or female mice.

Tetrahydrofuiran. A 10-mg/mL stock oftetrahydrofuran was prepared in DMSO, andcells were treated with doses of 1, 5, 10, 20,25, 50, and 100 ,ug/mL. Cells treated withtetrahydrofiran did not display cytotoxicityat any of the times or doses tested, and didnot induce p53 at any of the times or dosestested. Tetrahydrofuran is therefore dassifiedas negative. The Salmonella and SHE resultswere negative. In animal studies, some evi-dence for carcinogenicity was found in malerats, no evidence in female rats and malemice, and clear evidence in female mice.

DiscussionThe purpose of this study was to determinethe ability of the newly developed p53-in-duction assay to identify genotoxic sub-stances and to predict carcinogenicity inrodents. To this end, we subjected 25 of the30 PTE2 substances to this test. Five sub-stances strongly and reproducibly increasedcellular p53 levels and therefore were dassi-fied as positive, three substances increasedcellular p53 levels, but to a less significantextent and therefore were identified as possi-bly positive, and seventeen substances didnot significantly increase cellular p53 levelsand were therefore classified as negative.

Five of the PTE2 substances were nottested. As currently configured, our assay

system is capable of testing solid or liquidsubstances that can be dissolved in culturemedia to the required concentrations andwhich will remain in solution for several hr at37°C. Anthraquinone, molybdenum triox-ide, and vanadium pentoxide were not testedbecause of the limited solubility of these sub-stances in water, culture media, and DMSO.It may be possible in the future to identifyconditions that will allow us to deliver thesubstances to the cells at the required concen-trations. Isobutene is a gas at room tempera-ture and was not tested; it may be possible inthe future to set up a delivery system thatwould deliver a constant concentration of thegas to cells in culture. Chloroprene was nottested because it cannot be shipped.

The level of p53 detected in untreatedcells varied between experiments from a lowof undetectable (0) to a high of approxi-mately 10 ng p53/mg total cellular protein.Therefore, each experiment was analyzedseparately and the results of replicate experi-ments compared with each other. We foundgood agreement between experiments; formost chemicals, all experiments demonstrat-ed the same increase or lack of increase, andwith the same or a similar statistical signifi-cance. For a few chemicals, one experimentof three or four either failed to show thesame trend, or did show the same trend butat a lower level of significance.

In this study we identified chemicalsthat induced p53 but to a lesser extent andto a lower level of significance as possiblypositive. It may be that weakly positivewould be a more appropriate classification.To distinguish between the two, it may behelpful to perform additional trials of thesesubstances to confirm the reproducibility ofthe observed increases.

The strength of this p53-induction assayis expected to be in identifying genotoxiccompounds, of which this PTE2 group con-tains only a few. We therefore drew compar-isons between results from the p53-inductionassay and the in vivo results for both the entireset of tested compounds and separately for thegenotoxic substances. Figure 4 summarizesthe comparisons between the Salmonel4, p53induction, SHE, and in vivo test results.Figure 4A compares the overall in vivo testresults with each of the three in vitro proto-cols separately, and with a combination ofpredictions from the three in vitro protocols.In this analysis, chemicals were dassified as invivo positives if some or dear evidence of car-cinogenicity was obtained in either or bothspecies. Chemicals were classified as in vitropositives if they tested either positive, possiblypositive, or weakly positive. Analyses includedonly those chemicals where both assays havebeen done; or, in the case of the "any of thethree" analysis, all four assays have been done.

Volume 107, Number 10, October 1999 * Environmental Health Perspectives810

Articles * p53 induction results for 25 PTE2 chemicals

This accounts for the differences in the totalsfor the induded analyses. Also, for the "any ofthe three" analysis, a chemical is scored as invitro positive if any of the three tests(Salmonella, p53 induction, or SHE) gave apositive result. By these criteria, theSalmonella test gave a concordance with thein vivo test results of 32%, the p53-inductionassay a concordance of 40%, the "any of thethree" results a concordance of 59%, and theSHE assay a concordance of 67%. Therefore,by this analysis, the SHE test yielded thehighest concordance with the in vivo results,followed by the composite, then the p53induction, and finally the Salmonella assays.These results also reflect the higher likelihoodof the SHE test to produce positive results forthis set of chemicals, most of which were alsopositive in in vivo testing.

A comparison of the result from the p53-induction assay and the in vivo results for onlythe genotoxic compounds in this group (thosethat tested positive in the Salmonella assay)shows that two substances were positive inboth (cobalt sulfate heptahydrate and D&CYellow No. 11), one substance was positive invivo but negative in the p53-induction assay(primaclone), and one substance was negativein both assays (1-chloro-2-propanol). Of thislimited number of genotoxic compounds pre-sent in the PTE2 list, the p53-induction assayhas a concordance of75%.

Because the primary motivation inperforming these types of studies is theprotection of human health, transspeciescarcinogens may be of greater concern thansingle-species carcinogens. Figure 4B comparesthe in vivo test results for trans-speciescarcinogens with each of the three in vitroprotocols separately, and with a combinationof predictions from the three in vitro proto-cols. In this analysis, chemicals were classifiedas in vivo positives only if some or clear evi-dence of carcinogenicity was obtained inboth tested species; otherwise, the analysiswas as described for Figure 4A. By these cri-teria, the "any of the three" results gave aconcordance of 35%, the Salmonella test aconcordance of 50%, the SHE assay a con-cordance of 53%, and the p53-inductionassay a concordance of 56%. Therefore, bythis analysis, the p53-induction assay yieldedthe highest concordance with the in vivoresults, followed by the SHE analysis, thenthe Salmonella, and finally the compositeassays. Because fewer of the in vivo results areclassified as positive by the transspecies crite-ria, the concordance of the two assays lesslikely to test positive for this set of chemicals(Salmonella and p53 induction) with the invivo test results was improved, and the con-cordance of the two data sets more likely totest positive for this set of chemicals (SHEand "any of the three") decreased.

A comparison of the p53-induction assayresults and the transspecies in vivo results foronly the genotoxic compounds in this group(those that tested positive in the Salmonellaassay) shows that one substance was positivein both assays (cobalt sulfate heptahydrate)and two substances were negative in bothassays (1-chloro-2-propanol and pri-maclone). Of this limited number of geno-toxic compounds present in the PTE2 list,the p53-induction assay has a concordancewith the in vivo results of 100%.

Figure 4C compares results from thethree in vitro assays with each other. In thisanalysis, only those chemicals for which bothlisted assays have been reported are induded.The concordance of Salmonella and SHEassays is 44%, that of the p53-induction andSHE assays is 48%, and the p53-inductionand Salmonella assays is 63%, indicating thatof the three, the p53 induction and theSalmonella assays are the most like eachother, although still significantly different.

Based on the biology underlying theSalmonella SHE, and p53-induction assays, itseems clear that the p53-induction assay sens-es a fundamentally different type of biologicalinformation than do the other methods. TheAmes test (1) is perhaps the most widely usedtest for genotoxicity; it is based on the ability

In vivw In vivo

Concordance: 7/22 = 32% 820 = 40%

fi In vivw

Concordance: 1WM = 50% 1U118 = 56%

I.Salmonoll

+ 1-

Concordance: 15/24 = 83%

CL

1il

to _

14j

& _=4.49

of mutagens to alter the phenotype of specificstrains of Salmonella by inducing heritablechanges in its genetic material. The SHE testlikewise relies on heritable changes in themorphology of treated cells (22). In contrast,the p53-induction assay takes advantage of anormal cellular response to DNA damage;namely, the increase of the tumor-suppressorprotein p53. It is therefore not dependent ona permanent change in DNA; rather, it worksby eavesdropping on the events that normallyfollow DNA damage.

An advantage of this assay, therefore, isthat it may be capable of detecting transitoryDNA damage that is repaired too quickly tobe detected in tests which require heritablechanges in the DNA. Diethylstilbestrol,which produces DNA lesions with a shortbiologic half-life, induced p53 [(9) and refer-ences therein] while giving a negative or weakresponse in the Salmonella assay. This featureis one possible explanation for substances(such as sodium xylenesulfate or t-butylhy-droquinone) testing positive or possibly posi-tive in this assay while testing negative in thein vivo tests; the substance may have causedDNA damage that was quickly repairedbefore DNA replication.

Another possible explanation for chemi-cals testing positive in this assay and negative

In vivo _ in wvo+ I- | I

21=67%X1 m9=59%

In vivo viwvo

J19= 53% 0117 = 35%

In vivo (genotoxic only)

324= 75%

i wvo(genotoxic only)+ 1-

33= 100%

SHE SHE+--. 1

i_=48% 11125=44%

Figure 4. Comparisons of assay results. Abbreviations: +, positive; -, negative; SHE, Syrian hamsterembryo. (A) In vitro results as compared to in vivo results (single- and transspecies carcinogens com-bined). The results of the three in vitro assays (Salmonella, p53 induction, and SHE) were comparedagainst the in vivo results. Chemicals were classified as in vivo positives if some or clear evidence of car-

cinogenicity was obtained in either or both species. For the last panel, only Salmonella positive chemicalswere included in the analysis. (B) In vitro results compared against in vivo results (transspecies carcino-gens only). The results of the three in vitro assays were compared against the in vivo results. In this case,chemicals were classified as in vivo positives only if some or clear evidence of carcinogenicity was

obtained in both tested species. For the last panel, only Salmonella positive chemicals were included inthe analysis. (C) Results from the three in vitro assays as compared to each other. For the "any of thethree" analyses in (A) and (B), a chemical is scored as in vitro positive if any of the three tests yielded a

positive result. Also, two chemicals (D&C Yellow No. 11 and oxymetholone) were tested only in one

species and are therefore included in the analysis in (A) but not in (B).

Environmental Health Perspectives * Volume 107, Number 10, October 1999

t811

Articles * Duerksen-Hughes et al.

in the in vivo protocol is that they couldcause increases in cellular p53 levels by amechanism independent of DNA damage(23). For example, a substance that couldblock proteosome function would be expect-ed to increase cellular p53 (24,25). We havefound, however, that the ability of sub-stances to increase p53 and to cause cellularcytotoxicity is not correlated; indeed, mostsubstances do not induce p53 even whenclearly inducing stress and cytotoxicity [(9)and this study]. Hence, it seems unlikelythat p53 induction is a general nonspecificresponse to chemical or cytotoxic stress.

The p53-induction assay used in thisstudy utilized NCTC 929 cells, which are amouse fibroblast cell line. Although we haveshown that these cells can respond to indirectgenotoxins such as aflatoxin B1 by increasingtheir level of p53, demonstrating the pres-ence of activating metabolic enzymes, it ispossible that differences in the array of meta-bolic activities in these cells as opposed toliver extract could create a different combina-tion of metabolites. This altered range ofactivities could then either increase ordecrease DNA damage. This could accountfor differences in the results noted with thep53-induction assay and those noted forother assays. It may be interesting to test asubset of these substances for their ability toinduce p53 in a different cell line, perhapsone derived from liver.

We have previously shown that theincrease in p53 is transient with respect totime and is maximal at certain substance-specific doses, and that a higher dose doesnot necessarily lead to a higher level of cellu-lar p53. In this study, only a limited numberof times and doses could be tested. Based onour previous work (9), direct-acting geno-toxic chemicals induced p53 at early times(2-8 hr), whereas indirect-acting genotoxicchemicals induced p53 at later times (12-24hr). Hence, we tested each of these 25 PTE2chemicals at both early (6 hr) and late (17hr) time points in an attempt to detect bothdirect and indirect genotoxins. We foundthat one chemical exerted significant effectsat more doses at the early time point than atthe latter (D&C Yellow No. 11), and thatothers showed an effect at only the lattertime point (citral, oxymetholone, t-butylhy-droquinone; compare Figure 2 and Figure1). It is possible, however, that a substancecapable of weakly inducing p53 could exertits maximal effect at a time other than thosetested, and that any induction seen at thetested times was not statistically significant.

Previously tested genotoxins yieldedmaximal induction at doses between 1 and100 pg/mL; therefore, for each of the 25substances analyzed, we tested several doses

between 0.5 and 100 gg/mL. We foundthat most of the chemicals that tested posi-tive did so only at one or a few of the testeddoses, suggesting that the dose range atwhich an effect can be observed may berather narrow. For a number of the sub-stances, the higher doses (25, 50, and 100j.g/mL) induced extensive cytotoxicity,compromising the validity of the p53 assayat those dosages. It is possible that a sub-stance testing negative in our assay wouldhave tested positive at some other untesteddose. Our results were analyzed with theScheffe S-test, which is a relatively conserva-tive post hoc test for differences.

The expected strength of this p53-induc-tion assay is in the identification of genotox-ic compounds. There were few of these com-pounds in this PTE2 group, and with thesefew compounds, the p53-induction assay didan excellent job of identifying them. Themost likely explanation for the inability ofthis assay to identify many of the carcino-genic compounds in this group is that theyare epigenetic substances that cause cancerby nongenotoxic mechanisms. Additionally,factors such as absorption, distribution,metabolism, and excretion, which operate inwhole animals but do not operate in in vitrotest systems, could play a role.

It has been suggested by Ashby andTennant (26) that carcinogenicity may resultfrom

a specific interaction between the chemical andthe tissue rather than it being an intrinsic andunique property of the chemical;

therefore, it may well be that no one chemi-cal feature or assay will provide the predictiv-ity desired. It may therefore be helpful toadd this new set of information to that con-sidered when evaluating the carcinogenicpotential of substances. For example, theresults of the p53-induction assay could beconsidered as an input in the induction ofrules for predicting chemical carcinogenicityin rodents (27).

REFERENCES AND NOTES

1. Maron DM, Ames BN. Revised methods for theSalmonella mutagenicity test. Mutat Res 113:173-215(1983).

2. Tennant RW, Margolin BH, Shelby MD, Zeiger E, HasemanJK, Spalding J, Caspary W, Resnick M, Stasiewicz S,Anderson B, et al. Prediction of chemical carcinogenicityin rodents from in vitro genetic toxicity assays. Science236:933-941 (1987).

3. Zeiger E, Haseman JK, Shelby MD, Margolin BH,Tennant RW. Evaluation of four in vitro genetic toxicitytests for predicting rodent carcinogenicity: confirmationof early results with 41 additional chemicals. Environ MolMutagen 16:1-14 (1990).

4. Tennant RW, Spalding JW, Stasiewicz S, Caspary WD,Mason JM, Resnick MA. Comparative evaluation ofgenetic toxicity patterns of carcinogens and noncarcino-gens: strategies for predictive use of short-term assays.Environ Health Perspect 75:87-95 (1987).

5. Tennant RW. The genetic toxicology database of theNational Toxicology Program: evaluation of the relation-ships between genetic toxicology and carcinogenicity.Environ Health Perspect 96:47-51 (1991).

6. Wachsman JT, Bristol DW, Spalding J, Shelby M,Tennant RW. Predicting chemical carcinogenesis inrodents. Environ Health Perspect 101:444-445 (1993).

7. Bristol DW, Wachsman JT, Greenwell A. The NIEHS pre-dictive-toxicology evaluation project. Environ HealthPerspect 104:1001-1010 (1996).

8. National Institute of Environmental Health Sciences.NIEHS Predictive-Toxicology Evaluation Project.Available: http://dir.niehs.nih.gov/dirlecm/pte2.htm [cited10 March 1999].

9. Yang J, Duerksen-Hughes PJ. A new approach to identi-fying genotoxic carcinogens: p53 induction as an indica-tor of genotoxic damage. Carcinogenesis 19:1117-1125(1998).

10. Fritsche M, Haessler C, Brandner G. Induction of nuclearaccumulation of the tumor-suppressor protein p53 byDNA-damaging agents. Oncogene 8:307-318 (1993).

11. Kastan MB, Onyekwere 0, Sidransky D, Vogelstein B, CraigRW. Participation of p53 protein in the cellular response toDNA damage. Cancer Res 51:6304-6311 (1991).

12. Nelson WG, Kastan MB. DNA strand breaks: the DNAtemplate alterations that trigger p53-dependent DNA dam-age response pathways. Mol Cell Biol 14:1815-1823 (1994).

13. Zhan Q, Carrier F, Fornace AJ Jr. Induction of cellularp53 activity by DNA-damaging agents and growth arrest.Mol Cell Biol 13:4242-4250 (1993).

14. Hickman AW, Jaramillo RJ, Lechner JF, Johnson NF. x-Particle-induced p53 protein expression in a rat lungepithelial cell strain. Cancer Res 54:5797-5800 (1994).

15. Hess R, Plaumann B, Lutum AS, Haessler C, Heinz B,Fritsche M, Brandner G. Nuclear accumulation of p53 inresponse to treatment with DNA-damaging agents.Toxicol Lett 72:43-52 (1994).

16. Kuerbitz SJ, Plunkett BS, Walsh WV, Kastan MB. Wild-type p53 is a cell cycle checkpoint determinant followingirradiation. Proc Nati Acad Sci USA 89:7491-7495 (1992).

17. Ginsberg D, Michael-Michalovitz D, Ginsberg D, Oren M.Induction of growth arrest by a temperature sensitivep53 mutant is correlated with increased nuclear localiza-tion and decreased stability of the protein. Mol Cell Biol11:582-585 (1991).

18. Yonish-Rouach E, Resnitzky D, Lotem J, Sachs L, KimchiA, Oren M. Wild-type p53 induces apoptosis of myeloidleukaemic cells that is inhibited by interleukin-6. Nature352:345-347 (1991).

19. Yonish-Rouach E, Grunwald D, Wilder S, Kimchi A, MayE, Lawrence J, May P, Orem M. p53-mediated cell death:relationship to cell cycle control. Mol Cell Biol13:1415-1423 (1993).

20. Clarke AR, Purdie CA, Harrison DJ, Morris RG, Bird CC,Hooper ML, Wyllie AH. Thymocyte apoptosis induced byp53-dependent and independent pathways. Nature362:849-852 (1993).

21. Ryan JJ, Danish R, Gottliebg CA, Clarke MF. Cell cycleanalysis of p53-induced cell death in murineerythroleukemia cells. Mol Cell Biol 13:711-719 (1993).

22. Kerckaert GA, Brauninger R, LeBoeuf RA, Isfort RJ. Useof the Syrian hamster embryo cell transformation assayfor carcinogenicity prediction of chemicals currentlybeing tested by the National Toxicology Program inrodent bioassays. Environ Health Perspect 104:1075-1084(1996).

23. An WG, Chuman Y, Fogo T, Blagosklonny MV. Inhibitorsof transcription, proteasome inhibitors, and DNA-damag-ing drugs differentially affect feedback of p53 degrada-tion. Exp Cell Res 244:54-60 (1998).

24. Shinohara K, Tomioka M, Nakano H, Tone S, Ito H,Kawashima S. Apoptosis induction resulting fromproteasome inhibition. Biochem J 317:385-388 (1996).

25. Lopes UG, Erhardt P, Yao R, Cooper GM. p53-dependentinduction of apoptosis by proteasome inhibitors. J BiolChem 272:12893-12896 (1997).

26. Ashby J, Tennant RW. Definitive relationships amongchemical structure, carcinogenicity and mutagenicity for301 chemicals tested by the U.S. NTP. Mutat Res257:229-306 (1991).

27. Bahler D, Bristol DW. The Induction of Rules forPredicting Chemical Carcinogenesis in Rodents. MenloPark, CA:AAAI/MIT Press, 1993.

812 Volume 107, Number 10, October 1999 - Environmental Health Perspectives