Environmental Carcinogens and p53 Tumor-SuppressorGene Interactions in a Transgenic Mouse Model for

Mammary Carcinogenesis

Daniel Medina,1* Robert Ullrich,2 Raymond Meyn,3 Roger Wiseman,4

and Larry Donehower1,51Department of Molecular and Cellular Biology, Baylor College of Medicine,

Houston, Texas2Radiological Health Sciences, Colorado State University, Fort Collins, Colorado

3Department of Experimental Radiation Oncology, University of Texas, M.D.Anderson Cancer Center, Houston, Texas

4National Institute of Environmental Health Sciences, Research Triangle Park,North Carolina

5Department of Molecular Virology and Microbiology, Baylor College of Medicine,Houston, Texas

Mouse mammary tumorigenesis is greatly influ-enced by a variety of exogenous agents, such asMMTV, chemical carcinogens (i.e., polycyclic aro-matic hydrocarbons), and radiation, as well as byendogenous/physiological factors, such as steroidhormones, tumor-suppressor genes (i.e., Brca1/2,p53), and gene products of modifier genes. In themouse model, the most frequently used chemicalcarcinogen has been 7,12-dimethylbenz[a]anthra-cene (DMBA), which activates the Ha-ras gene butdoes not alter the p53 tumor-suppressor gene.However, on an existing background of p53 genealteration, low doses of DMBA are strongly cocar-cinogenic. Using a transgenic model system, inwhich the p53 gene was deleted in the mammarygland, we examined the carcinogenic effects of avariety of external agents and internal factorsgiven at either low doses or physiological doses.

These agents/factors included DMBA, �-radiation,Brca2 heterozygosity, and steroid hormones. Allagents/factors increased the tumorigenic responseof the p53 null mammary cells, even under condi-tions where no tumorigenic response was observedin the p53 wildtype mammary cell. The strongestcocarcinogenic effect was observed with the ste-roid hormone progesterone. The majority of tumorswere highly aneuploid and composed of nuclearigh-grade cells. The mechanism for the aneuploidyand secondary events associated with high tumorige-nicity were examined using array technology. Theseresults demonstrate that, on a background of underly-ing genetic instability, very low doses of environmen-tal mutagens and mitogens can produce strong cocar-cinogenic effects. Environ. Mol. Mutagen. 39:178–183, 2002. © 2002 Wiley-Liss, Inc.

Key words: p53; mammary; environmental agents; hormones

INTRODUCTION

Mouse mammary carcinogenesis is strongly influencedby a variety of environmental agents as well as internalfactors. The environmental factors include traditional chem-ical carcinogens, such as polycyclic aromatic hydrocarbons;alkylating agents, such as nitrosureas and urethane; nitro-samines, fluorenylacetamides, and naphthylamines [re-viewed in Medina, 2000]. Radiation is also a weak butimportant carcinogen for the mouse mammary gland [Ull-rich et al., 1996]. The mouse mammary tumor virus(MMTV) is unique to mice and can be transmitted bothgenetically and vertically via the milk [Medina, 2000].Physiological factors include the pituitary peptide hor-mones, prolactin and growth hormones, steroid hormones,estrogen and progesterone, as well as numerous growthfactors, such as EGF, IGF, and FGF [Medina, 2000; Muller

and Neville, 2001]. The roles of many of these factors havebeen examined using transgenic and gene-deletion mice[Cardiff et al., 2000]. The effects of factors, singly and incombination, determine the final tumor incidence in anymouse.

It has been known for many years that different mousestrains exhibit varying susceptibility to mammary carcino-

Grant sponsor: National Institutes of Health; Grant numbers: U01CA84243and R01CA84320.

*Correspondence to: Daniel Medina, Department of Molecular and Cellu-lar Biology, Baylor College of Medicine, One Baylor Plaza, Houston, TX77030.

Received 25 September 2001; and accepted 5 November 2001

Published online 28 February 2002 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/em.10064

Environmental and Molecular Mutagenesis 39:178–183 (2002)

© 2002 Wiley-Liss, Inc.

gens. This is true for MMTV, chemical carcinogens, and ingenetically engineered mice. For example, virgin mice ofstrains DBA/2f and BALB/c are more susceptible thanC3Hf, C57BL, and FVB to chemical carcinogen-inducedmammary tumorigenesis [Liebelt and Liebelt, 1967; Me-dina, 2000]. In genetically engineered mice, the tumorigeniceffect of p53 haploidy is markedly greater on a BALB/cgenetic background than on a 129SV or C57BL geneticbackground [Kuperwasser et al., 2000].

In mice, 7,12-dimethylbenz[a]anthracene (DMBA) hasbeen the carcinogen most frequently studied. In mice, unlikerats, the virgin mammary gland has a relatively low degreeof susceptibility to chemical carcinogens. However, suscep-tibility is increased by hormone stimulation [Medina, 2000].Hormones act both to increase initiation frequency and aspromoting agents. Also, as in the rat mammary gland,Ha-ras is frequently activated; however, p53 is not mutated[Cardiff et al., 1988; Kumar et al., 1990; Jerry et al., 1994].The mouse models have been used frequently to examinethe responsiveness of the initiated gland to very low dosesof DMBA. Thus, mammary hyperplasias that arise sponta-neously have a marked sensitivity to very low doses (0.5–1.0 g DMBA) [Medina and Kittrell, 1987]. This increasedsensitivity to DMBA can also be observed in mammaryglands of transgenic mice generated by specific geneticalterations (e.g., TGF� activation; p53 dominant negativemutation) [Coffey et al., 1994; Li et al., 1998].

The above-mentioned results led us to examine the ef-fects of low doses of specific environmental agents andmodifiers of DNA repair on the tumorigenic response of themammary epithelium in which the tumor-suppressor genep53 was deleted by homologous recombination. Loss of p53function results in increased cell proliferation [Jones et al.,1997] and genomic instability [Attardi and Jacks, 1999;Goepfert et al., 2000] in mammary epithelial cells as well asincreased tumorigenic capability [Jerry et al., 2000]. Theresults presented herein demonstrate the increased suscep-tibility of the mammary epithelial cell deficient in p53 genefunction to several environmental and physiological factors.

MATERIALS AND METHODS

Mice

All mice were bred and maintained at Baylor College of Medicine in aconventional mouse facility with food and water provided ad libitum. Theroom temperature was set at 70°F in the animal facility that is AALACaccredited. The BALB/c heterozygous mice were provided by Dr. RogerWiseman and bred with BALB/c p53 mice at Baylor College of Medicine.

Transplantation

The transplantation protocol was described previously [Jerry et al.,2000]. A normal-appearing mammary duct from p53 null or heterozygousmammary gland was transplanted into the cleared mammary fat pads ofBALB/c wildtype mice. This was necessary because p53 null BALB/c

mice die of lymphosarcomas by 4 months of age; therefore, to studymammary carcinogenesis in a p53 null background, we developed thisprotocol. The donor mice were 8–10 weeks old and the recipient mice werefollowed for up to 15 months for tumor incidence. Tumors were detectedas a result of weekly palpation.

Carcinogens

DMBA was dissolved in cottonseed oil and administered by oral gavageat a dose of 1.0 mg in 0.2 ml, once a week for 4 weeks. Irradiation wasperformed as previously described [Medina et al., 1998]. BALB/c micebearing the p53 null and p53 wildtype mammary transplants were irradi-ated with a whole-body dose of 1 Gy using a 137Cs animal irradiator. Themammary transplants had completely filled the mammary fat pads at 10weeks after transplantation (the time of irradiation) and cell proliferationwas reduced to a steady-state rate compared with the rate observed duringthe preceding period when the transplants were expanding in the mammaryfat pads.

Hormones

Progesterone (20 mg) was administered by silastic tubing [Guzman etal., 1999]. The tubes were implanted subcutaneously in the upper back andreplaced at 6-week intervals.

Statistics

The tumor incidences were evaluated by �2 tests [Peto, 1974] and tumorlatency times by life-table procedures described by Mantel [1996]. Theresults were considered significantly different at P � 0.05.

RESULTS

DMBA–p53 Interaction

The tumorigenic effects of DMBA were investigated intwo experiments. In experiment 1, a total dose of 4 mg wasgiven starting at 5 weeks after transplantation, a time whenthe transplants were actively proliferating to completelyoccupy the mammary fat pad. In this experiment, both p53null and p53 wildtype mammary epithelia provided separatetarget cell populations for DMBA. Figure 1 shows theresults. Very few tumors [Medina, 2000] arose in the p53wildtype transplants with DMBA treatment and no tumorsin the untreated wildtype transplants. In the p53 null trans-plants, the untreated cell population produced 30% tumors(8/26) over a period of 40 weeks. In contrast, DMBA-treated transplants produced 60% (15/25), with a 50% tu-mor endpoint reached at 35 weeks. Thus, DMBA had asignificant effect on tumor latency and tumor incidencecompared to that of the untreated p53 null transplants.

In the second experiment, we slightly changed the exper-imental design. Previous experiments indicated that theeffects of genetic instability increased with chronologicalage of the p53 null mammary epithelial cell [Goepfert et al.,2000]. To see a more dramatic effect of DMBA, the mam-mary epithelial transplants were treated with DMBA at11–13 weeks after transplantation, a time when the trans-

Mammary Carcinogenesis in Transgenic Mice 179

plants had filled the mammary fat pad but had aged suffi-ciently to exhibit signs of genetic instability in the form oflow levels of aneuploidy. The results are shown in Figure 2.Two independent donor mice were used for these experi-ments. The results indicated that the tumorigenic risk inboth donor transplants was low at 40 weeks after transplan-tation (2 tumors/24 transplants in each group), whereasDMBA treatment significantly increased the tumorigenicresponse (8/24 and 13/24, respectively). Given that theresponse to DMBA of wildtype mammary epithelium underthese conditions is very low, the presence of underlyinggenetic instability [Goepfert et al., 2000] greatly increased

the tumorigenic capability of DMBA. This experiment dem-onstrated a marked interaction between these two events.

Irradiation–p53 Interaction

DMBA is a model carcinogen but not known to be acarcinogen for the human breast epithelium. Therefore, weapplied the same idea of testing a low dose of an oncogenicagent on a background of genetic instability but using aknown human carcinogen, that of radiation. The results areshown in Figure 3. This dose had absolutely no effect ontumorigenicity of p53 wildtype mammary epithelium butincreased the tumor incidence of p53 null mammary epithe-lium. Mammary tumors were examined for ploidy statususing metaphase preparations. All four tumors were highlyaneuploid with mean ploidy numbers of 60, 67, 68, and 76and percentage aneuploid cells of 98, 76, 98, and 98, re-spectively.

Progesterone–p53 Interaction

Previous studies have shown that moderate levels ofprogesterone can induce aneuploidy in morphologically ap-pearing normal p53 null mammary epithelial cells [Goepfertet al., 2000]. The tumorigenic effects of progesterone wereevaluated over long-term exposure, the results of which areshown in Figure 4. Whereas only one tumor appeared in 24transplants in the untreated p53 null transplants, 21 tumorsappeared in 22 transplants in the progesterone treated miceby 43 weeks after transplantation. The tumors appearing inthe progesterone-treated transplants were highly aneuploid,with a mean chromosome number of 71, 66, and 57, respec-tively, in the three tumors analyzed. A comparison of thethree treatments (DMBA, irradiation, progesterone) shows

Fig. 1. The effect of 7,12-dimethylbenz[a]anthracene (DMBA) on tu-morigenesis in p53 null mammary epithelial cells. The DMBA was ad-ministered starting at 5 weeks after transplantation, a time period when thetransplants were actively proliferating. The p53 wildtype transplants arenot plotted on the graph because they developed only one tumor during theassay period.

Fig. 2. The effect of DMBA on tumorigenesis in two separate donorpopulations of p53 null mammary epithelial cells. In both cases, DMBAwas administered starting at time periods after the transplants had filled themammary fat pad. In set A, DMBA was started at 11 weeks and in set B,DMBA was started at 13 weeks after transplantation.

Fig. 3. The effect of 1 Gy �-radiation on tumorigenesis in p53 nullmammary epithelial cells. The transplants were irradiated once at 10 weeksafter transplantation, a time period when the transplants had filled themammary fat pad.

180 Medina et al.

that the most dramatic and strongest effect was exhibited byprogesterone.

Brca2–p53 Interaction

Alterations in Brca2 represent an important genetic de-terminant in human breast cancer that is translated as ageneral decrease in DNA repair capabilities in the cell. Toexamine potential interactions among these two differentgene families, we bred Brca2 heterozygous (het) mice withp53 null and p53 heterozygous (het) mice to generate off-spring with either a p53 null/Brca2 het background or adouble het background. The results are shown in Figure 5.The double heterozygote produced 1 tumor (20 transplants),whereas the p53 null/Brca2 wildtype produced eight tumors(of 20 transplants). In contrast, Brca2 haploinsufficiency

increased the tumorigenic response in the p53 null mam-mary cells. The effect of Brca2 heterozygosity was notobserved on initial tumor latency or on accelerated appear-ance during the first year, as in the case of DMBA orprogesterone, but on an increase in tumor incidence duringthe second year. The effects were modest but significant.

All the experiments are summarized in Table I. Thistabulation provides a comparison of the different treatmentgroups. An examination of the tumorigenic potential of theuntreated p53 null mammary cells in the six experimentsillustrates the marked heterogeneity in tumorigenic potentialof the individual p53 null mammary cell. We interpret thisheterogeneity to reflect the stochastic occurrence of second-ary events necessary for tumorigenesis that is driven by theunderlying genetic instability attributed to the absence ofp53 gene function.

DISCUSSION

It is virtually impossible to define the periods and extentof exposure to environmental breast carcinogens in thehuman female. Most likely, exposure occurs over a longtime period, as is the case with supplemental hormones. Inmodel systems, one has the advantage of defining timeperiods, exposure levels, and genetic alterations, thus allow-ing an examination of specific interactions of environmentalfactors and genetic alterations under defined conditions. Inthe experiments reported herein, we were interested in thecarcinogenic effects of low doses on environmental muta-

Fig. 4. The effect of continuous progesterone stimulation on tumorigen-esis in p53 null mammary epithelial cells. The silastic tubing containing 20mg progesterone was replaced at 6-week intervals.

Fig. 5. The effect of Brca2 haploidy on tumorigenesis in p53 null mam-mary epithelial cells. The different genotypes were the offspring of cross-ing BALB/c p53 and BALB/c Brca2 mice.

TABLE I. Mammary Carcinogenesis in p53 Null TransgenicMammary Epithelial Cells

Treatmenta No. tumors/no. transplants (%) TE50 (weeks)b

Untreated 8/26 (31) �40DMBA (1)c 15/25 (60)* 35

Untreated 2/24 (8) �40DMBA (2A)c 8/24 (33)* �40

Untreated 2/24 (8) �40DMBA (2B)c 13/24 (54)* 38

Untreated 6/20 (30) �65�-Radiation 14/20 (70)* 56

Untreated 1/22 (5) �44Progesterone 21/22 (95)* 35

p53 �/� Brca2 �/� 8/20 (40) �60p53 �/� Brca2 �/� 13/20 (65)* 52p53 �/� Brca2 �/� 1/20 (5) �60

aAll mice contain p53 null mammary transplants except the p53/Brcacrosses.bTE50 � 50% tumor endpoint.cNumber in brackets refers to groups identified in Figures 1 and 2.*P � 0.05 compared to untreated p53 null mammary cells.

Mammary Carcinogenesis in Transgenic Mice 181

gens and physiological mitogens operating on a cellularbackground of underlying genetic instability.

Alterations in the tumor-suppressor p53 gene are frequentin breast cancer and include changes that result in both lossand gain of gene function. In both cases, genetic instabilityis prevalent, although the underlying mechanisms and man-ifestations may be different [Attardi and Jacks, 1999; Go-epfert et al., 2000; Murphy et al., 2000]. The three types ofexposures examined in these experiments are different instructure, mechanism, and consequence. Irradiation is aclassical DNA-damaging agent inducing strand breaks;DMBA causes gene mutations as well as alterations in thehost hormonal and immunological responses, whereas pro-gesterone at these levels induces primarily mitogenesis,although alterations in gene expression cannot be ruled outat this time. The most interesting result was that, in theabsence of p53 gene function, progesterone had the mostprofound effect on mammary tumorigenesis. Consideringthat progesterone is the only physiological agent of the threetreatment modalities, the results are somewhat surprising. Itwas recently demonstrated that progesterone/estrogen hor-mone combinations impart a greater breast cancer risk thanestrogen hormones alone, when both are administered longterm as hormone replacements in postmenopausal women[Ross et al., 2000]. In p53 null mammary cells, it has alsobeen recently demonstrated that these low doses of proges-terone, but not estrogen, for short time periods (5 weeks)induce aneuploidy in morphologically appearing normalcells [Goepfert et al., 2000]. These results emphasize themarked effects of modest physiological doses of a naturallyoccurring hormone on tumorigenic process in cells with apredisposing genetic instability. It should be emphasizedthat these same hormone conditions do not induce tumori-genesis in p53 wildtype mammary epithelial cells.

The mechanism behind this marked stimulation of tumor-igenesis in the p53 null mammary cells is presently unclear,although it is evident that aneuploidy is markedly stimu-lated. Previous experiments have ruled out alterations incentrosome alterations [Goepfert et al., 2000] and have notobtained evidence for DNA damage as measured by chro-mosome breaks or gaps, increased sister chromatid ex-change, or increased oxidative damage (Medina, unpub-lished observations). Instead, it appears that progesterone isstrongly mitogenic. It is possible that the absence of p53results in altered transcriptional regulation of genes that areimportant for ensuring proper chromosome segregation andthe increased aneuploidy results from an accumulation ofchromosome missegregants that are not deleted by a G1checkpoint. This speculation remains to be tested.

A second result of interest is that low does of environ-mental agents can have serious consequences for tumori-genic risk if combined with an underlying genetic instabil-ity. The doses of radiation and chemical carcinogen used inthese experiments had little impact on tumorigenesis in thegenetically stable (p53 wildtype) mammary epithelial cell.

However, with the genetic instability imposed by the ab-sence of p53, the risk was discernable and significant. It isa tenet in chemical carcinogenesis that the frequency ofcarcinogen-induced transformation is a function of the levelof cell proliferation at the time of carcinogen/target cellinteraction. For instance, the tumorigenic response of theactively proliferating mammary epithelium at 6–10 weeksof age is much greater than that of the quiescent gland at12–16 weeks of age [Medina, 1971]. The results of theDMBA experiments demonstrate that the genomic instabil-ity present in the mature but relatively quiescent mammaryepithelium 11–13 weeks after transplantation imparts anequal, if not greater, susceptibility to carcinogen-inducedtransformation than does increased cell proliferation. In allcases, the tumors were aneuploid and metastatic.

This experimental approach would seem a reasonableway to examine the potential tumorigenic risk of low dosesof other suspected environmental mutagens or carcinogens.For instance, the tumorigenic risk of environmental estro-gens or pesticides in this model or in other mammarymodels where a relevant gene has been dysregulated mightprovide insight into the relevant risk factors for humanbreast cancer. Several model systems that might be relevantare mice that overexpress cyclin D1 [Wang et al., 1994],c-myc [D’Cruz et al., 2001], p53 127arg-his [Li et al., 1998],and erbb2 [Andrechek and Muller, 2000], and mice that arehaploid for Brca1 [Xu et al., 1999] and Brca2 [Bennett etal., 2000].

In summary, the availability of genetically engineeredmice with alterations in genes relevant to human breastcancer provides new and important model systems to revisitthe issue of environmental risk factors for human breastcancer.

ACKNOWLEDGMENTS

The authors acknowledge the technical assistance ofValerie Lutes, Maureen McCarthy, Anne Shepard, KathyKey, and Frances Kittrell.

REFERENCES

Andrechek ER, Muller WJ. 2000. Tyrosine kinase signalling in breastcancer: tyrosine kinase-mediated signal transduction in transgenicmouse models of human breast cancer. Breast Cancer Res 2:149–153.

Attardi LD, Jacks T. 1999. The role of p53 in tumor suppression: lessonsfrom mouse models. Cell Mol Life Sci 55:48–63.

Bennett LM, McAllister KA, Malphurs J, Ward T, Collins NK, Seely JC,Gowen LC, Koller BH, Davis BJ, Wiseman RW. 2000. Miceheterozygous for a Brca1 or Brca2 mutation display distinct mam-mary gland and ovarian phenotypes in response to diethylstilbes-trol. Cancer Res 60:3461–3469.

Cardiff RD, Gumerlock PH, Soong MM, Dandekar S, Barry PA, YoungLJT, Myers FJ. 1988. c-H-ras-1 expression in 7,12-dimethylbenza-thracene-induced BALB/c mouse mammary hyperplasias and theirtumors. Oncogene 3:205–213.

182 Medina et al.

Cardiff RD, Moghanaki D, Jensen RA. 2000. Genetically engineeredmouse models of mammary intraepithelial neoplasia. J MammaryGland Biol Neoplasia 5:421–437.

Coffey RJ Jr, Meise KS, Matsui Y, Hogan BL, Dempsey PJ, Halter SA.1994. Acceleration of mammary neoplasia in transforming growthfactor alpha transgenic mice by 7,12-dimethylbenzanthracene. Can-cer Res 54:1678–1683.

D’Cruz CM, Gunther EJ, Boxer RB, Hartman JL, Sintasath L, Moody SE,Cox JD, Ha SI, Belka GK, Golant A, Cardiff RD, Chodosh LA.2001. c-MYC induces mammary tumorigenesis by means of apreferred pathway involving spontaneous Kras2 mutations. NatMed 7:235–239.

Goepfert TM, McCarthy M, Kittrell FS, Stephens C, Ullrich RL, BrinkleyBR, Medina D. 2000. Progesterone facilitates chromosome insta-bility (aneuploidy) in p53 null normal mammary epithelial cells.FASEB J 14:2221–2229.

Guzman RC, Yang J, Rajkumar L, Thordarson G, Chen X, Nandi S. 1999.Hormonal prevention of breast cancer: mimicking the protectiveeffect of pregnancy. Proc Natl Acad Sci USA 96:2520–2525.

Jerry DJ, Butel JS, Donehower LA, Paulson EJ, Cochran C, Wiseman RW,Medina D. 1994. p53 mutations occur infrequently in 7,12-dimeth-ylbenzanthracene-induced mammary tumors in BALB/c and hemi-zygous p53 mice. Mol Carcinog 9:175–183.

Jerry DJ, Kittrell FS, Kuperwasser C, Laucirica R, Dickinson ES, BonillaPJ, Butel JS, Medina D. 2000. A mammary-specific model dem-onstrates the role of the p53 tumor suppressor gene in tumordevelopment. Oncogene 19:1052–1058.

Jones JM, Attardi L, Godley LA, Laucirica R, Medina D, Jacks T, VarmusHE, Donehower LA. 1997. Absence of p53 in a mouse mammarytumor model promotes tumor cell proliferation without affectingapoptosis. Cell Growth Differ 8:829–838.

Kumar R, Medina D, Sukumar S. 1990. Activation of H-ras oncogenes inpreneoplastic mouse mammary tissues. Oncogene 5:1271–1277.

Kuperwasser C, Hurlburt GD, Kittrell FS, Dickinson ES, Laucirica R,Medina D, Naber SP, Jerry DJ. 2000. Development of spontaneousmammary tumors in BALB/c p53 heterozygous mice. Am J Pathol157:2151–2159.

Li B, Murphy KL, Laucirica R, Kittrell F, Medina D, Rosen JM. 1998. Atransgenic mouse model for mammary carcinogenesis. Oncogene16:997–1007.

Liebelt AG, Liebelt RA. 1967. Chemical factors in mammary tumorigen-esis. In: Carcinogenesis: a broad critique. Presented at Annual

Symposium of Fundamental Cancer Research. Baltimore: Williams& Wilkins. p 315–345.

Mantel N. 1966. Evaluation of survival data and two new rank orderstatistics arising in its consideration. Cancer Chemother Rep 50:163–170.

Medina D. 1971. The effects of hormone stimulation, dose and time ofadministration on carcinogen-induced tumors from preneoplasticnodule outgrowths. J Natl Cancer Inst 46:909–916.

Medina D. 2000. Mouse models for mammary cancer. In: Ip MM, AschBB, editors. Methods in mammary gland biology and breast cancerresearch. New York: Kluwer Academic. p 3–17.

Medina D, Kittrell FS. 1987. Enhancement of tumorigenicity with mor-phological progression in a BALB/c preneoplastic outgrowth line.J Natl Cancer Inst 79:569–576.

Medina D, Stephens LC, Bonilla PJ, Hollmann CA, Schwahn D, Kuper-wasser C, Jerry DJ, Butel JS, Meyn RE. 1998. Radiation-inducedtumorigenesis in preneoplastic mouse mammary glands in vivo;significance of p53 status and apoptosis. Mol Carcinog 22:199–207.

Muller WJ, Neville MC. 2001. Signal transduction in mammary develop-ment and tumorigenesis. J Mammary Gland Biol Neoplasia 6:1–150.

Murphy KL, Dennis AP, Rosen JM. 2000. A gain of function p53 mutantpromotes both genomic instability and cell survival in a novelp53-null mammary epithelial cell model. FASEB J 14:2291–2302.

Peto R. 1974. Guidelines in the analysis of tumor rates and death rates inexperimental animals. Br J Cancer 29:101–105.

Ross RK, Paganini-Hill A, Wan PC, Pike MC. 2000. Effect of hormonereplacement therapy on breast cancer risk: estrogen versus estrogenplus progestin. J Natl Cancer Inst 92:328–332.

Ullrich RL, Bowles ND, Satterfield LC, Davis CM. 1996. Strain-dependentsusceptibility to radiation-induced mammary cancer is a result ofdifferences in epithelial cell sensitivity to transformation. RadiatRes 146:353–355.

Wang TC, Cardiff RD, Zukerberg L, Lees E, Arnold A, Schmidt EV. 1994.Mammary hyperplasia and carcinoma in MMTV-cyclin D1 trans-genic mice. Nature 369:669–671.

Xu X, Wagner KU, Larson D, Weaver Z, Li C, Ried T, Hennighausen L,Wynshaw-Boris A, Deng CX. 1999. Conditional mutation of Brca1in mammary epithelial cells results in blunted ductal morphogene-sis and tumour formation. Nat Genet 22:10.

Mammary Carcinogenesis in Transgenic Mice 183