Multiple Stress Signals Activate Mutant p53 In Vivo · cellular stress signals, allowing for transcriptional modulation of multiple genes that play important roles in controlling

Molecular and Cellular Pathobiology

Multiple Stress Signals Activate Mutant p53 In Vivo

Young-Ah Suh1, Sean M. Post1,2, Ana C. Elizondo-Fraire1, Daniela R. Maccio1, James G. Jackson1,Adel K. El-Naggar3, Carolyn Van Pelt4, Tamara Terzian5, and Guillermina Lozano1

Abstractp53 levels are tightly regulated in normal cells, and thus, the wild-type p53 protein is nearly undetectable until

stimulated through a variety of stresses. In response to stress, p53 is released from its negative regulators, mainlymurine double minute 2 (Mdm2), allowing p53 to be stabilized to activate cell-cycle arrest, senescence, andapoptosis programs. Many of the upstream signals that regulate wild-type p53 are known; however, limitedinformation for the regulation of mutant p53 exists. Previously, we showed that wild-type and mutant p53R172Hare regulated in a similar manner in the absence ofMdm2 or p16. In addition, this stabilization of mutant p53 isresponsible for the gain-of-functionmetastatic phenotype observed in themouse. In this report, we examined therole of oncogenes, DNA damage, and reactive oxygen species, signals that stabilize wild-type p53, on thestabilization of mutant p53 in vivo and the consequences of this expression on tumor formation and survival.These factors stabilized mutant p53 protein which oftentimes contributed to exacerbated tumor phenotypes.These findings, coupled with the fact that patients carry p53 mutations without stabilization of p53, suggest thatpersonalized therapeutic schemes may be needed for individual patients depending on their p53 status. CancerRes; 71(23); 7168–75. �2011 AACR.

Introduction

The p53 pathway is impaired in most human cancers. Morethan 50% of human tumors carry p53 mutations, and most ofthese are missense mutations that result not only in loss oftumor-suppressing activities but also acquisition of oncogenicactivities, defined as gain of function. Specifically, mutant p53enhances proliferation and survival in cells and tumorigenesisinmice when comparedwith cells ormice that are deficient forp53 (1).

Wild-type p53 normally exists in a latent state but becomesstabilized and activated in response to various genotoxic andcellular stress signals, allowing for transcriptional modulationof multiple genes that play important roles in controlling cell-cycle progression, senescence, and apoptosis (2). The regula-tion of wild-type p53 is mediated mainly by protein turnover.Primarily, its stability is regulated by murine double minute 2(Mdm2), an E3 ubiquitin ligase that targets p53 for proteolyticdegradation (3–7). Recently, several other ubiquitin ligaseshave been identified that also regulate p53 stability and include

COP1 (8), Pirh2 (9), ARF-BP1/Mule (10), and Trim24 (11),although their roles in vivo are less clear.

The importance of DNA damage in p53 signaling has beenextensively studied. In particular, exposure to chemotherapeu-tic drugs results in stabilization of wild-type p53 throughposttranslational modifications in the amino terminus (2).These modifications disrupt the ability of Mdm2 to interactwith p53 (12). The generation of reactive oxygen species (ROS),cytokines, and g-radiation also play roles as a potent activatorsof p53 (13). Mechanistically, ROS activates p53 through directdamage to DNA. In addition, ROS contributes to cross-talk andactivates other signaling pathways such as p38, c-jun-NH2-kinase (JNK), or NF-kB signaling resulting in synergy of p53activation through phosphorylation, stabilization, and activa-tion (14).

Two independent tumor suppressors, p16INK4a and p19ARF

(p14ARF in humans) also impact the p53 signaling pathwaythrough differentmechanisms. p16INK4a loss, as occurs in sometumors, allows cyclin D/CDK4 kinase activity to phosphorylateretinoblastoma (Rb) and results in its dissociation from E2F.Released E2F then transcriptionally activates p19ARF whichbinds Mdm2 and thereby affects p53 stabilization (15, 16). Incancers, p19ARF is also upregulated following oncogene acti-vation (17–19). Mutant Ras or increased levels of c-Myc, forexample, stimulate the ARF–Mdm2 complex formation(18, 20–22) which in turn causes a sequestration of Mdm2and subsequent stabilization of p53 (23).

Although the signaling pathways targeting wild-type p53 arewell documented, the regulatory signals that control mutantp53 levels are less well understood. Recently, we and othersshowed that mutant p53 is inherently unstable in normaltissues and that some of the factors that regulate wild-type

Authors' Affiliations: Departments of 1Genetics, 2Leukemia, 3Pathology,and 4Veterinary Medicine, The University of Texas MD Anderson CancerCenter, Houston, Texas; and 5Department of Dermatology/Stem CellBiology, University of Colorado Denver, Denver, Aurora, Colorado

Y.-A. Suh and S.M. Post contributed equally to this work.

Corresponding Author: Guillermina (Gigi) Lozano, Department of Genet-ics, The University of Texas MD Anderson Cancer Center, 1515 HolcombeBlvd, Houston, TX 77030. Phone: 713-834-6386; Fax: 713-834-6380;E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-11-0459

�2011 American Association for Cancer Research.

CancerResearch

Cancer Res; 71(23) December 1, 20117168

on March 6, 2021. © 2011 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst October 7, 2011; DOI: 10.1158/0008-5472.CAN-11-0459

http://cancerres.aacrjournals.org/

p53 are also responsible for the stabilization of mutant p53protein (24–26). Loss ofMdm2, for example, stabilizes mutantp53 in many normal tissues (26). In these experiments, stabi-lization of mutant p53 led to gain-of-function phenotypesmanifested as increased tumor incidence and metastasis.Disruption of the Rb pathway via p16INK4a deletion also sta-bilizedmutant p53 in vivo (26). We have now explored whetherother signals that stabilize wild-type p53 likewise affectmutantp53 stabilization. Because p53 is mutated in the majority ofhuman tumors and expression of mutant p53 is often associ-ated with poor outcomes, we examined a variety of cellularstimuli that may potentially stabilize mutant p53 in vivo andmay thus lead to an enhanced tumor phenotype.We found thatactivation of oncogenes stabilized mutant p53 resulting inmore potent tumor phenotypes than mice that only harboredthe p53 mutation; however, these types of stimuli did notimpact overall survival. On the other hand, when we examinedthe effect of doxorubicin, a chemotherapeutic DNA-damagingagent, on mutant p53 stability, we observed stabilization ofmutant p53 as well as decreased survival compared with theuntreated p53 homozygous mutant mice. Furthermore, weanalyzed the effect of a ROS scavenger on mutant p53 proteinstability and also observed a gain-of-function phenotype. Inmost cases, stabilization of mutant p53 led to worse pheno-types than the absence of p53. These results led us to proposethat multiple cellular stress pathways that regulate wild-typep53 also act to increase mutant p53 levels yielding gain-of-function phenotypes.

Materials and Methods

Generation of Em-myc and K-rasLA1 cohortsp53515A/þ, p53�/�, K-rasLA1, and p53515A/515A mice were

maintained on more than a 95% C57BL/6 background(24, 27, 28). Em-myc mice (29) were crossed to p53515A/515A togenerate p53515A/þEm-myc. The background ofEm-myc, Em-mycp53515A/þ, or Em-myc p53þ/� mice were 75% C57BL/6 and 25%129Sv. Tails from Em-myc, Em-myc p53515A/þ, or Em-myc p53þ/�

mice were genotyped using primers previously described (30).K-rasLA1 mice (27) were crossed to p53515A/515A to generatep53515A/þ K-rasLA1 mice. p53515A/þ K-rasLA1 mice were furthercrossed with p53515A/515A to generate p53515A/515A K-rasLA1

mice. To determine mouse genotypes, PCR analysis was con-ducted on tail DNA using published primer sets for the p53�,p53515A, and K-rasLA1 alleles (24, 27, 28, 31). The animal studieswere conducted according to the MD Anderson and IACUCguidelines.

StatisticsSurvival curves were plotted by the Kaplan–Meier method

by GraphPad Prism to assess statistical differences. A factorwas considered statistically significant if it had a 2-sided Pvalue less than 0.05.

Immunohistochemical analysisImmunohistochemical (IHC) analysis was conducted as

previously described (32). Protein expression was analyzedusing rabbit a-p53 (CM5) antibodies (Vector Laboratories) at

a 1:200 dilution for 2 hours at 37�C, and visualized using ABCand DAB kits from (Vector Laboratories). Slides were counter-stained with Nuclear Fast Red. Hematoxylin and eosin stainingwas used for pathologic analysis of tumors.

Western blottingProtein lysates were prepared from either the tissues or

tumors ofmice. Fiftymicrogramof total proteinswere resolvedby SDS-PAGE and transferred to nitrocellulose membranes(GE Bioscience). After blocking with 5% skimmilk in PBS–0.1%Tween 20 (PBS-T) for 1 hour at room temperature, membraneswere incubated with a-p53 (CM5; Vector Laboratories, 1:1,000dilution) antibodies at 4�C overnight. Membranes were thenwashedwith PBS-T and incubatedwith anti-rabbit horseradishperoxidase–conjugated secondary antibody and visualizedwith ECL plus (GEBioscience). Anti-b-actin or vinculin (Sigma;1:5,000) antibodies were used as a loading control. All anti-bodies were diluted in blocking buffer.

Treatment of miceTwo-day-old homozygous pups were treated with 2 Gray

(Gy) of g-radiation. For doxorubicin treatment, pups wereinjected with either PBS or 2 mg of doxorubicin suspended inPBS per gram of body weight at days 5 and 6. Five hours afterthe last injection, pups were sacrificed. Protein expression intissue was measured by Western blot analysis as describedearlier. For ROS experiments, parent mice were treated withthe ROS scavenger N-acetyl-cysteine (NAC) for 2 weeks beforemating through drinking water and continuedwhile pupswerenursing. At day 5, homozygous mutant pups were irradiatedwith 2 Gy. Four hours after irradiation, mice were injected witheither PBS or 2 mg of NAC in PBS per gram of body weight. Onehour later, pups were sacrificed and protein expression wasexamined.

Measurement of intracellular ROSThe dye 5-(and-6)-chloromethyl-20,70-dichlorodihydro-

fluorescein diacetate (CM-H2DCFDA; Molecular Probes) wasused to detect intracellular ROS. The fluorescence intensityof intracellular DCFDA is a linear indicator of ROS in stainedcells (33). To compare intracellular ROS in thymocytes fromp53515A/515A mice, freshly isolated thymocytes were incubat-ed with 10 mmol/L DCFDA in the culture medium (RPMI-1640) for 30 minutes at 37�C. The cells were harvested andthe DCFDA fluorescence profiles determined by flow cyto-metric analysis with a Coulter Epics software program,version 4.02.

LOHLOH at the p53 locus in Em-myc p53515A/þ and Em-myc

p53þ/� lymphomas was determined by PCR amplificationusing primers 50-TACTCTCCTCCCCTCAATAAGCTATTC-30

(exon 5) and 50-AGTCTAGGCTGGAGTCAACTGTCT-30

(intron 6). PCR amplicons were separated by gel electro-phoresis followed by DNA purification of the correct pro-ducts as previously described (34). Sequencing was carriedout using the exon 5 primer and analyzed with Chromassoftware.

Activation of Mutant p53 In Vivo

www.aacrjournals.org Cancer Res; 71(23) December 1, 2011 7169




Results

K-ras activation stabilized the p53R172H mutantand exacerbated tumor progression

Wild-type p53 is stabilized and activated in response tohyperproliferative signals and thus protects cells from aber-rant growth signals. Tissue culture stabilizes wild-type andmutant p53 and is, thus, not an optimal system to studythese effects (24). Therefore, we examined the effect ofoncogenes on mutant p53 protein stabilization in vivo bytaking advantage of 2 tumor prone mouse models that aredriven by hyperproliferative signals. The first is the mousecarrying an oncogenic allele of K-rasLA1 (27). Following aspontaneous recombination event in the Ras allele, therecombined cells express activated Ras and thus drive tumorprogression. The second is the Em-myc transgenic mousemodel that expresses high levels of c-Myc in B cells andrapidly develops B-cell lymphomas (29). Myc overexpressionelevates p53 activity (30, 35). These were crossed to a p53mutant mouse model inheriting the p53515A missense muta-tion, encoding the mutant p53R172H protein, and to p53-null mice for comparison of stabilization of mutant proteinand tumor phenotypes. We generated homozygous p53mutant mice that carry the K-rasLA1 allele to avoid wild-type p53 activation by K-ras. The latent K-rasLA1 allele isspontaneously recombined and expresses an active rasprotein harboring a glycine to aspartic acid alteration atamino acid 12 (27). We examined whether activated K-rasexpression in the lungs affected p53 stability in 4-week-oldmice, prior to overt tumor development. By immunohis-tochemistry, we observed that mutant p53 is visible onlyin hyperplastic lesions in lungs from young K-rasLA1

p53515A/515A but not in the lungs from young K-rasLA1 micewith wild-type p53 (Fig. 1A). The level of p53 proteinexpression in different genotypes was also analyzed byWestern blotting. p53 levels were higher in the lungs fromK-rasLA1 p53515A/515A mice than wild-type, K-rasLA1, orp53515A/515A mice at 4 weeks of age (Fig. 1B). The absenceof wild-type p53 staining may be due to the fact that cellsactivating wild-type p53 initiate senescence or apoptoticprograms, quickly rendering p53 undetectable. Stabilizationof mutant p53 was also examined in tumors from K-rasLA1

p53515A/515A mice. Immunohistochemistry revealed thatmutant p53 was expressed in all tumors (Fig. 1C), whichcontrasts to the 75% detection rate in tumors that sponta-neously arise in p53515A/515A mice (ref. 26; and this study).Thus, activation of K-ras stabilizes mutant p53 in vivo.

Despite stabilization of p53R172H, the survival of K-rasLA1

p53515A/515A and K-rasLA1 p53�/� was similar (Fig. 1D). How-ever, the K-rasLA1 p53515A/515A mice developed more advancedand metastatic lung adenocarcinomas than K-rasLA1 p53�/�

mice (Table 1). This fact is highlighted by the observation that55.4% of the tumors from K-rasLA1 p53515A/515A mice wereadenocarcinomas, whereas only 27.8% of K-rasLA1 p53�/�micedeveloped adenocarcinomas. Taken altogether, these datasuggest the shift to a more aggressive tumor phenotype is theresult of stabilized p53R172H gain-of-function activities fol-lowing oncogenic K-ras stimulation.

c-Myc stabilized p53R172H mutant protein but did notaccelerate tumor progression

To expand the generality of these findings to other onco-genes, we next assayed how overexpression of c-Myc influ-enced the stability of p53R172H in vivo. We generated hetero-zygous p53 mutant mice harboring the Em-myc transgenedue to prenatal lethality of homozygous mutant p53 mice withEm-myc (30). Western blotting and IHC analysis of Em-mycp53515A/þ lymphomas showed that all Em-myc p53515A/þ spleniclymphomas overexpressed p53 as compared with wild-typespleens (Fig. 2A and B). On the other hand, Em-myc p53þ/�

lymphomas, which delete the one wild-type p53 allele, failed toexpress p53 (Fig. 2B; ref. 35). To explore whether the expressed

A K-ras K-ras p53515A/515A

K-ras p53515A/515A

K-rasWt K-ras p53515A/Ap53515A/A

K-ras

K-ras p53515A/515A

K-ras p53–/–

p53–/–

p53515A/515A

p53

p53

p53

Vin

Lung

Lung adenocarcinoma

Days

Perc

ent surv

ival

0 100

100

50

0200 300

B

C

D

Figure 1. K-rasLA1 stabilizes mutant p53R172H and results in a moreaggressive tumor phenotype. A, p53 levels were examined byimmunohistochemistry. Paraffin-embedded lungs from 4-week-oldK-rasLA1 (K-ras) or K-rasLA1 p53515A/515A mice were stained with a p53antibody. B, Western blot analysis of p53 in lungs from 4-week-oldwild-type (WT), K-rasLA1, p53515A/515A, or K-rasLA1 p53515A/515A mice.Vinculin (Vin) serves as a loading control. C, IHC analysis was conductedto analyze p53 expression in paraffin-embedded lung adenocarcinomasfrom K-rasLA1 p53515A/515A mice. D, Kaplan–Meier curves indicatesurvival rates of K-rasLA1 (K-ras), p53515A/515A, p53�/�, K-rasLA1 p53�/�,and K-rasLA1 p53515A/515A mice.

Suh et al.

Cancer Res; 71(23) December 1, 2011 Cancer Research7170




p53 in lymphomas from Em-myc p53515A/þ is wild-type ormutant protein, we analyzed the p53 loci for LOH. Lymphomasfrom the Em-myc p53515A/þ mice retained a p53 allele, whereasEm-myc p53þ/� mice lost their single wild-type p53 allele (Fig.2C). Sequencing revealed that 100% (15 of 15) Em-myc p53515A/þ

lymphomas lost the wild-type p53 allele (Fig. 2C). Thus,increased expression of c-Myc stabilized p53R172H in vivo.We also monitored the survival of Em-myc p53515A/þ and Em-myc p53þ/� mice. The Em-myc p53515A/þ mice succumbed tolymphomas at a similar rate as the Em-myc p53þ/� miceshowing mean survival of 40 days and 38 days, respectively(Fig. 2D). Thus, while the mutant p53 allele was retained andthe proteinwas stable, no alteration in tumor developmentwasobserved.

Mutant p53 stabilized by p16INK4a loss does not affectoverall survivalGiven the fact that oncogenic activation resulted in sta-

bilization of mutant p53, we next examined how loss of asecond tumor suppressor, p16Ink4a, affected mutant p53stability and tumor progression. Previously, we showed thatmutant protein is stabilized in a variety of tissues fromyoung mice in the absence of p16INK4a (26). To this end, weanalyzed the effect of stabilized mutant protein in a cohortof p16INK4a�/� p53515A/515A mice with stable p53R172H.Western blotting and IHC analysis revealed that mutantp53 protein was detectible in all (11 of 11) p16INK4a�/�

p53515A/515A tumors at varying levels (Fig. 3A and B). Thus,p16INK4a loss also resulted in stabilization of p53R172H in alltumors. Although the survival of p53 homozygous mutantand p53�/� mice in a p16INK4a-null background was notstatistically significant (Fig. 3C), loss of p16 enhanced themutant p53 gain-of-function phenotype. We observed that 5of 17 (29.4%) of the p16INK4a�/� p53515A/515A mice developedmetastatic sarcomas in contrast to the lack of metastaticdisease in the p16INK4a�/� and p53515A/515A mice (26).

ROS stabilized mutant p53 protein in vivoIt is widely appreciated that the p53 pathway is activated in

response to gains (oncogenes) or losses (tumor suppressors) ofgenetic material. However, exogenous factors have also beenimplicated in activation of wild-type p53. One factor, oxidativestress, is caused by ROS and is known to stabilize and activatewild-type p53. High levels of ROS are often observed in solidtumors and are elevated following radiation treatment incancer patients (36). To test the role of ROS in stabilizationof mutant p53 in vivo, we treated pups with g-radiation to firstinduce ROS and then treated with the ROS scavenger NAC. Wefirst measured ROS induction in the thymus of p53515A/515A

pups following irradiation. As shown in Fig. 4A, thymi fromp53515A/515A mice had 3-fold higher levels of ROS after irradi-ation. We next irradiated additional pups to examine the levelsof p53R172H. Four hours after treatment with ionizing radi-ation, the pups were then intraperitoneally injected with NACor PBS. Protein expression was analyzed one hour after NAC orPBS treatment. For this experiment, we treated 2mice for eachcondition. Mutant p53 protein was increased in the spleen andthymus after irradiation; however, mutant p53 stabilizationwas dampened following treatment with NAC in both thethymus and spleen (Fig. 4B). This result indicates thatincreased ROS levels resulting from g-radiation also stabilizemutant p53.

Doxorubicin stabilized mutant p53 protein in vivoAlthough we have shown that both ionizing radiation and

ROS result in stabilization of mutant p53 in vivo, other DNA-damaging agents are also often used in clinical settings. Theimpact of these chemotherapeutic agents on the stabiliza-tion of mutant p53 is currently unknown. Doxorubicin isused as a treatment strategy for multiple tumor types. Itinduces DNA damage resulting in wild-type p53 activation.To determine whether doxorubicin treatment resulted in thestabilization of mutant p53 in vivo, we injected doxorubicin

Table 1. Tumor spectrum of K-rasLA1 p53515A/515A and K-rasLA1 p53�/� mice

Tumor type K-ras LA1 p53515A/515A (%) K-ras LA1 p53�/� (%)

Adenocarcinoma 31 (55.4) 5 (27.8)Adenocarcinoma (multifocal) 18 (32.1) 4 (22.2)Early adenocarcinoma (multifocal) 7 (12.5) 0Early adenocarcinoma (multifocal) þ Lymphoma 2 (3.6) 1 (5.6)Metastatic adenocarcinoma 2 (3.6) 0Poorly differentiated adenocarcinoma 2 (3.6) 0

Sarcoma 14 (25) 8 (44.4)Osteosarcoma 2 (3.6) 0Angiosarcoma (mutifoocal) 2 (3.6) 2 (11.1)Spindle cell sarcoma (high grade) 8 (3.6) 4 (22.2)Sarcoma in soft tissue (high grade) 2 (3.6) 1 (5.6)Metastatic sarcoma in lung 0 1 (5.6)

Lymphoma 8 (14.3) 5 (27.8)Carcinoma (teratocarcinoma) 1 (1.8) 0Others (hemorrhage liver or arm) 2 (3.6) 0

Total tumor:total mice (# of tumors per mouse) 56:26 (2.15) 18:8 (2.25)






into 5- and 6-day-old pups. Western blot analysis revealedthat p53 levels increased after doxorubicin treatmentnot only in the thymi of homozygous mutant mice but alsoin wild-type and p53 heterozygous mice (Fig. 5A). Homozy-gous p53 mutant pups treated with doxorubicin once a dayfor 2 consecutive days starting at day 5 were monitored fortumor formation and survival. Doxorubicin-treated mutantmice showed no difference in survival as compared withdoxorubicin-treated p53-null mice (Fig. 5B). Interestingly,treated p53 homozygous mutant mice died significantlyearlier due to increased tumor burden than untreated p53mutant mice, suggesting that timing or mechanisms bywhich the mutant p53 protein is stabilized may impacttumor onset.

Stabilization of p53R172H by g-radiation alters survivalPreviously, we observed that g-irradiation of 4-week-old p53

homozygous mutant mice resulted in a longer half-life of themutant protein over a 15-hour time period than the wild-typeprotein (26). To further explore the effects of mutant p53stabilization, we treated pups with a low dose of radiation as

A

p53515A/515A p16 –/– tumors

p53515A/515A p16 –/–

p53515A/515A

p53515A/515A p16 –/–

p16 –/–

p53 –/– p16 –/–

p53 –/–

p53515A/515A

p16 –/–

Tu

WTSp

515A515A+IR

p53

p53

β-Actin

Days

Pe

rce

nt

su

rviv

al

0 100

100

75

50

25

0200 300

B

C

Figure 3. p16Ink4a loss stabilizes mutant p53R172H but does not affectsurvival rate. A, stabilized p53 protein in tumors from p53515A/515A p16�/�

mice. Western blot analysis of p53515A/515A p16�/� tumors blotted with ap53 antibody. b-Actin serves as a loading control. B, stabilization ofmutant p53 protein in tumors from p53515A/515A p16�/� byimmunohistochemistry. C, survival rate of p53515A/515A p16�/� wascompared with p16�/�, p53515A/515A, p53�/�, and p16�/� p53�/� byKaplan–Meier curves. Tu, tumor; Sp, spleen.

A

B

C

D

p53515A/515A Eµ-myc p53515A/+

Eµ-myc p53515A/+

Eµ-myc p53515A/+ Eµ-myc p53 +/ –

Eµ-myc p53 +/ –

Eµ-myc p53515A/+

Eµ-myc p53 +/ –

Eµ-myc

p53

IR

(B-cell lymphoma)

Days

Perc

ent surv

ival

0 100

100

80

60

40

20

0200 300

Normal spleen Tumors

p53

*5 6 7

Spleens

– + + 1 2 3 4 5 6 7 8 9 10 WT WTTu

β-Actin

Figure 2. c-Myc expression in B cells harboring one p53515A allele resultsin loss of wild-type p53 and stabilization of mutant p53. A, Western blotanalysis of p53 in lymphomas from Em-myc p53515A/þ mice. Wild-type(WT) and irradiated (IR) spleens were used for comparison of c-Myclevels. All spleens were lysed by tissue homogenization in NP-40 lysisbuffer, followed by centrifugation to remove insoluble proteins anddebris. b-Actin serves as a loading control. B, immunohistochemistryof Em-myc p53515A/þ and Em-myc p53þ/� lymphomas using a p53antibody. C, PCR analysis of the p53 locus in lymphomas from Em-mycp53515A/þand Em-myc p53þ/� mice. The PCR primers were designed toamplify p53 from exon 5 to intron 6. The asterisk denoted the p53515A

mutation in exon5. LOHat thep53 locus inEm-mycp53515A/þ lymphomaswas analyzed by sequencing of PCR product. The tumors showed LOHas the 515 nucleotide (arrow) had only onemutant peak and lost the wild-type peak. D, Kaplan–Meier curves indicating survival rates of Em-mycp53515A/þ, Em-myc p53þ/�, and Em-myc mice. Tu, tumor.

Suh et al.





p53�/� pups treated with g-radiation show significant differ-ences in survival as compared with untreated mice (37).Irradiated p53515A/515A pups died significantly sooner thanirradiated p53�/� mice and nontreated control mice (Fig.5C; P ¼ 0.036, P < 0.0001, respectively). Importantly, 100% ofthe tumors from irradiated mice expressed stable mutant p53(Fig. 5D). This result is contrasted by the fact that only 75% ofspontaneous tumors from p53515A/515A mice expressed stablemutant p53 (ref. 26, and this study). Together, these resultssuggest that radiation-induced stability of mutant p53 resultsin a deleterious gain-of-function phenotype manifested bydevelopment of multiple tumors and decreased survival ascompared with irradiated p53�/� mice.

Discussion

Mutant p53, like wild-type p53, is inherently unstable (26).However, p53R172H becomes stabilized in the absence of

Mdm2 and p16Ink4a and also has an extended half-life ascompared with wild-type p53 in response to DNA damagesignals (26). These data indicate that mutant p53 may bestabilized by mechanisms that also stabilize wild-type p53,implying that chemotherapeutic strategies that aim to activatewild-type p53 will also stabilize mutant p53. Because p53 ismutated inmore than 50% of tumors and expression ofmutantp53 is associated with poor outcomes, we asked whether otherfactors; such as those that contribute to human tumor forma-tion (oncogenic activation) and clinical therapeutics (irradi-ated and doxorubicin), stabilize mutant p53.

Two potent oncogenes, K-ras and c-Myc, were used todetermine their impact on mutant p53 stability in in vivo

A

B

p53515A/515A

Spleen

NAC

IR

Thymus

p53

β-Actin

+IR−IR

Flu

ore

sce

nce

2,000

1,500

1,000

500

0

Figure 4. ROS stabilizes p53R172H. A, ROS were measured inthymocytes from p53515A/515A mice after 2 Gy g-irradiation. Thymocyteswere prepared from 3 individualmicewith orwithout radiation. Two hoursposttreatment, cells were incubated with 10 mmol/L DCFDA, and theirfluorescence was measured by fluorescence-activated cell sorting(FACS). B, four pups were treated with radiation. After 4 hours, 2 werethen treated with NAC, an ROS scavenger, intraperitoneally. Westernblot analysis was conducted using thymus and spleen samples fromp53515A/515A mice with (þ) or without (�) NAC 1 hour later. Spleen andthymus samples from untreated mice were used as control.

A

p53515A/515Ap53515A/+p53 +/+

p53

Doxorubicin

Thymus

β-Actin

p53515A/515A

p53515A/515A/Doxop53–/– /Doxo

p53 –/–

Days

Perc

ent surv

ival

0 100

100

75

50

25

0200 300

P = 0.537

B

p53515A/515A

p53515A/515A/IR (2G)p53–/– /IR (2G)

p53 –/–

Perc

ent surv

ival

100

75

50

25

0

Days

0 100 200 300

P = 0.036

C

p53515A/515A tumors

p53

1 2 3 4 5 6 7 8 9 10

β-Actin

D

Figure 5. Doxorubicin or radiation stabilizes mutant p53. A, p53 levelswere analyzed byWestern blot analysis of thymi fromwild-type, p53515A/þ, or p53515A/515A pups 24 hours after the last injection of doxorubicin.B and C, Kaplan–Meier curves indicating survival rate of doxorubicin-treated (B) or g-irradiated mice of different genotypes (C). D, p53 levels intumors from g-irradiated p53515A/515A mice.






mouse models. Expression of each oncogene resulted in thestabilization of mutant p53 but did not dramatically shortenlife span compared with p53-null mice. Significantly, theK-rasLA1 p53515A/515A displayed a more aggressive tumor phe-notype than the K-rasLA1 p53�/� mice; 55% and 28% adeno-carcinoms, respectively. These results confirm that the gain-of-function ability of p53R172H is associated with stabilizedmutant protein and indicate that p53 mutations are moredetrimental than p53 loss. However, this does not occur in Em-Myc p53515A/þ mice, suggesting either the timing of stabiliza-tion or tissue specific differences impact manifestation of thegain-of-function phenotypes.

Many chemotherapeutic strategies aim to activate the p53tumor suppressor signaling pathway. Here, we tested whetherdifferent DNA-damaging agents such as doxorubicin, g-radi-ation, and ROS affect mutant p53 protein levels in vivo.Doxorubicin and g-radiation are standard therapeutic agentsused to treat various cancers as they induce a powerful DNA-damaging response which elicits a p53 response. Our findingsshow that these therapeutic strategies stabilize p53 expressionregardless of mutation status. As a result, g-radiation yielded again-of-function phenotype by decreasing survival of mutantmice. Given that irradiated p53515A/515A mice die sooner thanirradiated p53�/�mice, it is tempting to speculate that harmfuloutcomes may occur in human patients who harbor mutatedp53 following therapeutic treatment regimens. Furthermore,the median survival for doxorubicin-treated p53515A/515A andp53�/� mice was 99 and 126 days, respectively. Given thepotential clinical importance of these preliminary findings, itwill be imperative going forward to compare the outcome andsurvival of patients harboring wild-type or mutant p53.

In addition to chemotherapeutic treatment, oxidative stress,mainly caused by ROS, is an important factor leading to theactivation of the p53 pathway (38, 39). Accumulating datasuggest that tumors treated with radiation contain highamount of ROS (40). Studies in p53�/� mice have shown thatthe antioxidant function of p53 may directly contribute to theprevention of tumor development (41). This may have impor-tant implications for p53 in the regulation of redox-sensitivesurvival pathways. Our results showed that decreasing the levelof ROS resulted in decreased mutant p53 protein expression.This result indicates the gain-of-function phenotype resultingfrom stabilized mutant p53 may be overcome by inhibiting

DNA damage caused by ROS and suggests thatmanagement ofROS levels in patients with mutant p53 may be warranted.

The survival of mutant mice was clearly affected in responseto certain stresses but not others. Irradiated mutant miceshowed a significant decreased latency in tumor formationas compared with p53�/� mice. Interestingly, however, thesurvival of p53 mutant mice that are also p16�/� or haveactivated Ras was not different from p53�/� mice in theserespective backgrounds, even though there were changes intumor spectrum and metastatic potential. Therefore, othercooperating events, timing of insult, or tissue specificitymay allcontribute to outcome.

In conclusion, our data indicate that p53R172H is regulatedby many of the same signals that regulate wild-type p53. Theimportance of this study cannot be overemphasized, especiallywith regards to tumor treatment. The molecular mechanismsofmutant p53 stabilization present a fundamental conundrumin therapeutic intervention, not only for patients withLi-Fraumeni syndrome but also for cancer patients with spon-taneous p53 mutations. These data suggest that direct knowl-edge of the p53 status of a patient may be critical in preventingunintended consequences when determining therapeuticstrategies. This study also emphasizes the need for individuallytailored treatment for cancer patients depending upon theirp53 mutation status.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors thank members of the Lozano laboratory for helpful discussionand technical advice.

Grant Support

This study was supported by NIH grants CA46392 and CA34936 (G. Lozano).DNA sequencing and veterinary core facilities were supported by NCI CancerCenter Support grant CA16672.

The costs of publication of this article were defrayed in part by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received February 8, 2011; revised August 11, 2011; accepted September 19,2011; published OnlineFirst October 7, 2011.

References1. Brosh R, Rotter V. When mutants gain new powers: news from the

mutant p53 field. Nat Rev Cancer 2009;9:701–13.2. Appella E, Anderson CW. Post-translational modifications and acti-

vation of p53 by genotoxic stresses. Eur J Biochem 2001;268:2764–72.

3. Freedman DA, Levine AJ. Nuclear export is required for degradation ofendogenous p53 by MDM2 and human papillomavirus E6. Mol CellBiol 1998;18:7288–93.

4. Haupt Y, Maya R, Kazaz A, Oren M. Mdm2 promotes the rapiddegradation of p53. Nature 1997;387:296–9.

5. HondaR, TanakaH, YasudaH.OncoproteinMDM2 is a ubiquitin ligaseE3 for tumor suppressor p53. FEBS Lett 1997;420:25–7.

6. Kubbutat MH, Jones SN, Vousden KH. Regulation of p53 stability byMdm2. Nature 1997;387:299–303.

7. Roth J, Dobbelstein M, Freedman DA, Shenk T, Levine AJ. Nucleo-cytoplasmic shuttling of the hdm2 oncoprotein regulates the levels ofthe p53 protein via a pathway used by the human immunodeficiencyvirus rev protein. EMBO J 1998;17:554–64.

8. Dornan D, Wertz I, Shimizu H, Arnott D, Frantz GD, Dowd P, et al. Theubiquitin ligase COP1 is a critical negative regulator of p53. Nature2004;429:86–92.

9. Leng RP, Lin Y, MaW,Wu H, Lemmers B, Chung S, et al. Pirh2, a p53-induced ubiquitin-protein ligase, promotes p53 degradation. Cell2003;112:779–91.

Suh et al.





10. Chen D, Kon N, Li M, Zhang W, Qin J, Gu W. ARF-BP1/Mule is acritical mediator of the ARF tumor suppressor. Cell 2005;121:1071–83.

11. Allton K, Jain AK, Herz HM, Tsai WW, Jung SY, Qin J, et al. Trim24targets endogenous p53 for degradation. Proc Natl Acad Sci U S A2009;106:11612–6.

12. Ho JW, Song JZ, Leung YK. Activation of p53 by specific agents inpotential cancer therapy. Curr Med Chem Anticancer Agents2005;5:131–5.

13. Soussi T, Lozano G. p53 mutation heterogeneity in cancer. BiochemBiophys Res Commun 2005;331:834–42.

14. Temkin V, Karin M. From death receptor to reactive oxygen speciesand c-Jun N-terminal protein kinase: the receptor-interacting protein 1odyssey. Immunol Rev 2007;220:8–21.

15. Hahn WC, Weinberg RA. Modelling the molecular circuitry of cancer.Nat Rev Cancer 2002;2:331–41.

16. Sherr CJ. The INK4a/ARF network in tumour suppression. Nat RevMolCell Biol 2001;2:731–7.

17. de Stanchina E, McCurrach ME, Zindy F, Shieh SY, Ferbeyre G,Samuelson AV, et al. E1A signaling to p53 involves the p19(ARF) tumorsuppressor. Genes Dev 1998;12:2434–42.

18. Pomerantz J, Schreiber-Agus N, Li�egeois NJ, Silverman A, Alland L,Chin L, et al. The Ink4a Tumor Suppressor Gene Product, p19Arf,Interacts with MDM2 and Neutralizes MDM2's Inhibition of p53. Cell1998;92:713–23.

19. Zindy F, Eischen CM, Randle DH, Kamijo T, Cleveland JL, Sherr CJ,et al. Myc signaling via the ARF tumor suppressor regulates p53-dependent apoptosis and immortalization. Genes Dev 1998;12:2424–33.

20. Kamijo T, Weber JD, Zambetti G, Zindy F, Roussel MF, Sherr CJ.Functional and physical interactions of the ARF tumor sup-pressor with p53 and Mdm2. Proc Natl Acad Sci U S A 1998;95:8292–7.

21. Stott FJ, Bates S, James MC, McConnell BB, Starborg M, Brookes S,et al. The alternative product from the human CDKN2A locus, p14(ARF), participates in a regulatory feedback loop with p53 and MDM2.Embo J 1998;17:5001–14.

22. Zhang Y, Xiong Y, Yarbrough WG. ARF promotes MDM2 degradationand stabilizes p53: ARF-INK4a locus deletion impairs both the Rb andp53 tumor suppression pathways. Cell 1998;92:725–34.

23. Korgaonkar C, Zhao L,ModestouM,Quelle DE. ARF function does notrequire p53 stabilization or Mdm2 relocalization. Mol Cell Biol2002;22:196–206.

24. Lang GA, Iwakuma T, Suh YA, Liu G, Rao VA, Parant JM, et al. Gain offunction of a p53 hot spot mutation in a mouse model of Li-Fraumenisyndrome. Cell 2004;119:861–72.

25. Olive KP, Tuveson DA, Ruhe ZC, Yin B, Willis NA, Bronson RT, et al.Mutant p53 gain of function in two mouse models of Li-Fraumenisyndrome. Cell 2004;119:847–60.

26. Terzian T, Suh YA, Iwakuma T, Post SM, Neumann M, Lang GA, et al.The inherent instability of mutant p53 is alleviated by Mdm2 orp16INK4a loss. Genes Dev 2008;22:1337–44.

27. Johnson L,Mercer K, GreenbaumD, Bronson RT, Crowley D, TuvesonDA, et al. Somatic activation of the K-ras oncogene causes early onsetlung cancer in mice. Nature 2001;410:1111–6.

28. Jacks T, Remington L,Williams BO, Schmitt EM, Halachmi S, BronsonRT, et al. Tumor spectrum analysis in p53-mutant mice. Curr Biol1994;4:1–7.

29. Adams JM, Harris AW, Pinkert CA, Corcoran LM, Alexander WS, CoryS, et al. The c-myc oncogene driven by immunoglobulin enhancersinduces lymphoid malignancy in transgenic mice. Nature 1985;318:533–8.

30. PostSM,Quintas-CardamaA, Terzian T,SmithC,EischenCM,LozanoG. p53-dependent senescence delays Emu-myc-induced B-cell lym-phomagenesis. Oncogene 2010;29:1260–9.

31. Zheng S, El-Naggar AK, KimES, Kurie JM, LozanoG. A geneticmousemodel for metastatic lung cancer with gender differences in survival.Oncogene 2007;26:6896–904.

32. Evans SC, Viswanathan M, Grier JD, Narayana M, El-Naggar AK,Lozano G. An alternatively spliced HDM2 product increases p53activity by inhibiting HDM2. Oncogene 2001;20:4041–9.

33. Qiang W, Cahill JM, Liu J, Kuang X, Liu N, Scofield VL, et al.Activation of transcription factor Nrf-2 and its downstream targetsin response to moloney murine leukemia virus ts1-induced thioldepletion and oxidative stress in astrocytes. J Virol 2004;78:11926–38.

34. Post SM, Quintas-Cardama A, Pant V, Iwakuma T, Hamir A, JacksonJG, et al. A high-frequency regulatory polymorphism in the p53 path-way accelerates tumor development. Cancer Cell 2010;18:220–30.

35. Eischen CM, Weber JD, Roussel MF, Sherr CJ, Cleveland JL. Disrup-tion of theARF-Mdm2-p53 tumor suppressor pathway inMyc-inducedlymphomagenesis. Genes Dev 1999;13:2658–69.

36. Martindale JL, Holbrook NJ. Cellular response to oxidative stress:signaling for suicide and survival. J Cell Physiol 2002;192:1–15.

37. Kemp CJ, Wheldon T, Balmain A. p53-deficient mice are extremelysusceptible to radiation-induced tumorigenesis. Nat Genet 1994;8:66–9.

38. Kastan MB, Onyekwere O, Sidransky D, Vogelstein B, Craig RW.Participation of p53 protein in the cellular response to DNA damage.Cancer Res 1991;51:6304–11.

39. Li JM, Fan LM, George VT, Brooks G. Nox2 regulates endothelial cellcycle arrest and apoptosis via p21cip1 and p53. Free Radic Biol Med2007;43:976–86.

40. Shackelford RE, Kaufmann WK, Paules RS. Oxidative stress and cellcycle checkpoint function. Free Radic Biol Med 2000;28:1387–404.

41. Sablina AA, Budanov AV, Ilyinskaya GV, Agapova LS, Kravchenko JE,Chumakov PM. The antioxidant function of the p53 tumor suppressor.Nat Med 2005;11:1306–13.






2011;71:7168-7175. Published OnlineFirst October 7, 2011.Cancer Res Young-Ah Suh, Sean M. Post, Ana C. Elizondo-Fraire, et al.

In VivoMultiple Stress Signals Activate Mutant p53

Updated version

10.1158/0008-5472.CAN-11-0459doi:

Access the most recent version of this article at:

Cited articles

http://cancerres.aacrjournals.org/content/71/23/7168.full#ref-list-1

This article cites 41 articles, 12 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/71/23/7168.full#related-urls

This article has been cited by 20 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/71/23/7168To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-11-0459

http://cancerres.aacrjournals.org/content/71/23/7168.full#ref-list-1

http://cancerres.aacrjournals.org/content/71/23/7168.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/71/23/7168


Documents

Multiple Stress Signals Activate Mutant p53 In Vivo · cellular stress signals, allowing for transcriptional modulation of multiple genes that play important roles in controlling