[CANCER RESEARCH 55, 1649-1654, April 15, 1995]

Advances in Brief

Disruption of p53 Function Sensitizes Breast Cancer MCF-7 Cells to Cisplatin

and Pentoxifylline

Saijun Fan, Martin L. Smith, Dennis J. Rivet II,1 Diane Duba, Qimin Zhan, Kurt W. Kohn, Albert J. Fornace, Jr.,and Patrick M. O'Connor2

Laboratory of Molecular Pharmacology, Developmental Therapeutics Program, Division of Cancer Treatment, National Cancer Institute, NIH, Bethesda, Maryland 20892

Abstract

The possibility that appropriately designed chemotherapy could actselectively against p53-defective tumor cells was explored in MCF-7 hu

man breast cancer cells. These cells were chosen because they have normalp53 function but are representative of a tumor cell type that does notreadily undergo p53-dependent apoptosis. Two sublines (MCF-7/E6 andMCF-7/mu-p53) were established in which pS3 function was disrupted bytransfection with either the human papillomavirus type-16 E6 gene or adominant-negative mutant p53 gene. p53 function in MCF-7/E6 and MCF-7/mu-p53 cells was defective relative to control cells in that there were noincreases in p53 or p21Wafl/Clp*protein levels and no G, arrest following

exposure to ionizing radiation. Survival assays showed that p53 disruptionsensitized MCF-7 cells to cisplatin (CDDP) but not to several otherDNA-damaging agents. CDDP sensitization was not limited to MCF-7 cells

since p53 disruption in human colon carcinoma RKO cells also enhancedsensitivity to CDDP. Contrary to the other DNA-damaging agents tested,CDDP-induced DNA lesions are repaired extensively by nucleotide excision, and in agreement with a defect in this process, MCF-7/E6 andMCF-7/mu-p53 cells exhibited a reduced ability to repair a CDDP-dam-aged chloramphenicol acetyltransferase-reporter plasmid transfected into

the cells. Therefore, we attributed the increased CDDP sensitivity ofMCF-7 cells with disrupted p53 to defects in G, checkpoint control,nucleotide excision repair, or both. The G2 checkpoint inhibitor pentoxi-fylline exhibited synergism with CDDP in killing MCF-7/E6 cells but did

not affect sensitivity of the control cells. Moreover, pentoxifylline inhibitedG, checkpoint function to a greater extent in MCF-7/E6 than in the

parental cells. These results suggested that, in the absence of p53 function,cancer cells are more vulnerable to G2 checkpoint abrogators. Our resultsshow that a combination of CDDP and pentoxifylline is capable of syner-gistic and preferential killing of p53-defective tumor cells that do not

readily undergo apoptosis.

Introduction

Mammalian cells demonstrate complex cellular responses to DNAdamage, including activation of genes involved in cell cycle arrest,DNA repair, and apoptosis (1-4). Cell cycle arrest following DNA

damage is mediated by a series of negative feedback control systemscalled checkpoints (1-6). Checkpoints monitoring DNA damage op

erate in late G1 and G2 phases and possibly during S phase. Arrest atthese cell cycle junctions allows an extended time for DNA repair totake place before progression through critical phases of the cell cycle:S and M phases.

A critical component of the Gj checkpoint is the p53 tumor suppressor (4-6), which is also the most frequently mutated gene in

human cancer (7, 8). The p53 gene encodes a nuclear phosphoprotein

Received 1/17/95; accepted 3/6/95.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1 D. J. R. is a Howard Hughes Medical Institute Research Scholar at the NIH.2 To whom requests for reprints should be addressed, at Laboratory of Molecular

Pharmacology, Developmental Therapeutics Program, Room 5C-19, Bldg. 37, National

Cancer Institute, NIH, Bethesda, MD 20892.

which normally activates Gj cell cycle arrest in response to at leastsome forms of DNA damage (6, 9, 10). p53 thereby extends the timeavailable for DNA repair before S phase entry and so perhaps functions as a "guardian of the genome" (11). This scenario suggests that

p53-defective tumors should have increased susceptibility to someDNA-damaging anticancer drugs (1). p53 can, however, also activate

an apoptotic response to DNA damage, especially in hematopoieticand lymphoid cells (10, 12-15), which often overrides the G, check

point response. In cell types programmed for apoptosis, loss of p53function decreases sensitivity to a wide variety of DNA-damagingagents (10, 12-18), while in cell types of some solid tumors not

inherently programmed for apoptosis, a clear relationship betweenp53 gene status and radiosensitivity or chemosensitivity has beenmore difficult to establish (19, 20).

Our recent investigations into the function of the p53-regulated

gene product, Gadd45, have revealed that this protein stimulatesnucleotide excision repair in vitro (21). This and other results we havegathered suggested that cells lacking p53 function might be moresensitive to anticancer drugs that inflict DNA damage of the typerepaired primarily through nucleotide excision. We explored thispossibility in the breast carcinoma MCF-7 cell line. Although this linecontains functionally intact wild-type p53 genes, it represents a cell

type that does not readily undergo apoptosis following DNA damage(22). Successful transfection of p53 into MCF-7 cells has been re

ported as have some of the biological properties of the p53 transgene(23). p53 function in MCF-7 cells was disrupted by constitutive highlevel expression of the human papillomavirus type-16 E6 gene or adominant-negative, mutant p53 gene transfected into the cells (24).Cells were treated with CDDP,3 which induces DNA lesions repaired

primarily through nucleotide excision (25-27), MMS, whose DNA

lesions are repaired through base excision (26, 27), or with thetopoisomerase II inhibitors, ADR and VP16, or 7-irradiation which

induce DNA strand breaks (9, 10). We also investigated the effects ofPENT (28, 29) to determine whether cells lacking p53 function wouldbe more sensitive to the effects of G2 checkpoint abrogation.

Our studies revealed that p53 disruption sensitized MCF-7 cells toCDDP and PENT and suggested that some p53-defective tumors

might be vulnerable to: (a) anticancer drugs that induce DNA damageof the type repaired through nucleotide excision; and (b) G2 checkpoint abrogators.

Materials and Methods

Cell Culture and Stable Transfection. The breast carcinoma MCF-7 and

colon carcinoma RKO cell lines were cultured in RPMI 1640 (Mediatech,Washington, DC) containing 15% fetal bovine serum (Intergen, Purchase,NY), 2 mM L-glutamine, 50 units penicillin, and 50 /xg/ml streptomycin.

3 The abbreviations used are: CDDP, cisplatin; MMS, methylmethanesulfonate; ADR,Adriamycin; MTT, 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide; CAT,chloramphenicol acetyltransferase; CMV, cytomegalovirus; mu, mutant; PENT, pentoxifylline; VP16, etoposide.

1649

on May 7, 2018. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

P53 AND CHEMOSENSITIVITY

Synchronization of MCF-7 cells in G,-S phase was performed using aphidi-colin (1 fig/ml; 24 h). For stable transfection, 1 X IO6 cells were seeded in

10-cm diameter Petri dishes 24 h before transfection. The human papilloma-virus type-16 E6 gene (24) or a dominant-negative mutant p53 (143 Ala to Val)

cloned into a pCMV plasmid (graciously provided by Bert Vogelstein, Johns

Hopkins University, Baltimore, MD), which also contained a neomycin gene,were introduced into cells by calcium phosphate precipitation as describedpreviously (30). Approximately 14 days later, G418-resistant colonies were

selected, and after expansion, colonies demonstrating high level expression of

E6 mRNA or mutant p53 protein were analyzed for their inability to G, arrestand accumulate p53 and p2l°P1/Wafl following -y-irradiation. The generation

and characterization of colon carcinoma RKO cells expressing E6 or mutantp53 transgenes has been described previously (24, 31).

Irradiation and Drug Treatment. Irradiation was performed using a137Cs source delivering 7-rays at a dose rate of 3.46 Gy/min. ADR, VP16,

MMS, or CDDP were obtained from the National Cancer Institute. Stocksolutions of ADR and MMS were prepared in deionized water and stored at-20°C until use. VP16 was prepared in DMSO. CDDP and PENT (Sigma

Chemical Co.) were dissolved in PBS. Unless stated, cells were irradiated atroom temperature or treated with chemical agents for l h at 37°C.PENT was

added for 24 h following CDDP exposure or -y-irradiation.

Survival Assays. Cytotoxicity was evaluated using clonogenic and MTTsurvival assays. For clonogenic survival, cells were plated in replicate 10-cm

diameter Petri dishes for 24 h, rinsed with fresh medium, and then irradiatedor treated with various chemical agents for 1 h. Cells were then washed twicewith fresh medium and incubated for 10-14 days; the number of colonies

formed (>50 cells) were counted after staining with Giemsa. The platingefficiency of the untreated cells was between 50 and 75%. PENT (2 mM; 24 h)reduced clonogenic survival of the MCF-7 cell lines by less than 15%, and this

was taken into account when analyzing clonogenic survival of the cisplatin andPENT combination. For the MTT assay, 250 cells in 200 /xl of completemedium were placed into each well of a 96-well plate. Twenty-four h later,

CDDP was added and left in the culture for 6 days. At this time, viability wasassessed by the ability of cells to convert the soluble salt of MTT into an

insoluble formazan precipitate which was quantitated spectrophotometricallyfollowing solubilization in DMSO, as described previously (32).

Flow Cytometry. Samples were prepared for flow cytometry as describedpreviously (16). Briefly, cells were fixed in ice-cold 70% ethanol, washed withPBS, treated with RNase (Sigma; 500 units/ml) at 37°Cfor 15 min, and stained

with propidium iodide (Sigma; 50 /xg/ml). Cell cycle analysis was performedusing a Becton Dickinson fluorescence-activated cell analyzer. 15,000 cells

were analyzed for each point, and quantitation of cell cycle distribution was

performed using the SOBR model program provided by the manufacturer.Gel Electrophoresis and Western Blot Analysis. Samples were lysed on

ice for 30 min in 1% NP40 prepared in PBS that contained 10 /¿g/mlleupeptin,10 /xg/ml aprotinin, 2 mM 4-(2-aminoethyl)benzenesulfonyl fluoride, 1 mM

sodium o-vanadate, 10 mM sodium fluoride, and 5 mM sodium PP(. Solubleprotein was then boiled for 5 min in a SDS-loading buffer and 100 fLgof total

cell protein loaded onto 15% SDS-polyacrylamide gels. Proteins were then

transferred electrophoretically to Immobilen membranes (Millipore, Bedford,MA) and then blocked for 30 min in 5% nonfat milk at room temperature.Immunodetection of p53 protein was performed with a monoclonal pAblSOlantibody (Oncogene Science), and for p2lWafl/c'P1 determination, a polyclonal

antibody (PharMingen) was used. Antibody reaction was revealed using enhanced chemiluminescence (Amersham Corp., Arlington Heights, IL).

Host Cell (CAT) Reactivation Assay. The nonreplicating pSV2CAT plasmid, encoding bacterial CAT, was treated with 200 /XMcisplatin in 1 mMTris-HCl (pH 7.8), 10 mM NaCl, and 1 mM EDTA for 10 min at room

temperature. The MCF-7 cell lines were then transfected with 5 /xg of thecisplatin-damaged or undamaged pSV2CAT plasmid using lipofectamine rea

gent (GIBCO-BRL). As an internal control, 0.5 fig of pRSV ß-galactosidase

was included in the transfections. Transient C4Tgene expression was assayed72 h after transfections. Quantitation of [14C]chloramphenicol conversion was

performed by counting on a Betascope Analyzer (Belagen). Values werenormalized to the ß-galactosidase internal standard, and values were expressed

relative to the CAT activity obtained for the undamaged CAT plasmid in eachcell line (33, 34).

Results

Human Papillomavirus Type-16 E6 and mu-p53 TransfectedMCF-7 Cells with Impaired p53 Function. The human papilloma-virus type-16 E6 gene product, which stimulates degradation of p53through a ubiquitin pathway (24, 35), and a dominant-negative mutantp53 transgene were used as a means to inhibit p53 function in MCF-7

cells. These transgenes under the control of the CMV promoter weretransfected into MCF-7 cells by calcium phosphate precipitation, andG418-resistant colonies were selected and then expanded for analysis.

Four clones expressing high levels of E6 mRNA and six clonesexpressing high levels of mu-p53 protein exhibited p53 disruption,as indicated by impaired responses to y-ray-induced p53 andp2i°pi/wafi protein accumulation and Gt arrest. Fig. 1 illustrates the

impaired responses of the MCF-7/E6 and MCF-7/mu-p53 clones,

which were used in all subsequent studies. In addition to preventingthe -y-ray-induced increases in p53 and p2iClpl/Wafl protein, E6 reduced the basal levels, especially of p21ciP1/Wafl (Fig. 1A). Thereduced level of p21ciP1/Wafl in exponentially growing MCF-7/E6

cells could stem from inactivation of the basal transcriptional activityof p53 (31); however, we cannot exclude the possibility that E6 couldalso stimulate p2lc'P1/Wafl degradation. Impairment of G, arrest is

shown in the panels labeled 6.3 Gy in Fig. IB. The mitotic inhibitornocodazole was included to determine the ability of the cells to arrestin G2 (16). The lack of effect of nocodazole on the cell cycle patterns

MCF-7/CMV MCF-7/E6 MCF-7/mu-p53

B MCF-7/CMV

MCF-7/E6

MCF-7/mu-pS3

CONTROL NOCODAZOLE 6.3 Gy NOCODAZOLE

Fig. 1. Functional characterization of G, and G2 checkpoints in exponentially growingMCF-7 cells transfected with control (CMV), E6-, or mu-p53-containing plasmids. A,immunoblot analysis of p53 and p21 levels in control exponentially growing (—)andy-irradiated (+) cultures. Cells were assayed 4 h following 6.3 Gy. B, cell cycledistribution 16 h following addition of nocodazole (0.4 jxg/ml), exposure to 6.3 Gy ofy-rays, or both. The X-axis shows propidium iodide fluorescence intensity used to assessDNA content of the cells. The similarity of the cell cycle profiles following y-irradiationwith or without nocodazole indicates that E6 or mu-p53 did not prevent MCF-7 cells fromarresting in G2. The difference between the cell cycle profiles of CMV-, E6-, or mutant-p53-transfected cells shows that E6 and mutant-p53 abrogated the G| checkpoint responseto y-irradiation.

1650

on May 7, 2018. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

P53 AND CHEMOSENSmVITY

MCF-7/CMV

B.eaai5 18 IS

CISPLATIN (i/M)

B

CISPLATIN (flM)

Fig. 2. Effec!of p53 disruption on the sensitivity of MCF-7 and RKO cells to cisplatin.A, MCF-7 cells transfected with control-CMV or Ee-containing plasmids were treatedwith cisplatin for 1 h, and survival was measured in clonogenic assays. Values shown arethe means of at least four data sets; bars, SD. B, cisplatin was added to MCF-7 or RKOcells transfected with control-CMV, E6-, or mu-p53 genes and left for 6 days before cellviability was determined using the MTT assay as described in "Materials and Methods."

Values shown are the means of at least three determinations from two independentexperiments; bars, SD.

shows that G2 arrest occurs equally well in the control and E6 ormu-p53 transfected cells. Furthermore, these results illustrate that inthe absence of p53 function cells are still capable of G-, arrest.

p53 Disruption Sensitizes MCF-7 and RKO Cells to Cisplatin.The sensitivity of MCF-7 and MCF-7/E6 cells to several types ofDNA-damaging agents was assessed in clonogenic survival assays(Table 1). Of the agents tested, only cisplatin exhibited a p53-depend-

ent sensitivity difference. Colony survival as a function of cisplatinconcentration showed a significantly increased sensitivity of the E6-

transfected cells (Fig. 2A). For example, at a survival level of 10% forMCF-7 cells, the survival of MCF-7/E6 cells was 25-fold less. The

dose modification factor for equivalent cell killing was approximately2 (P < 0.01). The enhanced sensitivity of MCF-7/E6 cells to cisplatinwas also observed when MCF-7 cells were transfected with a mu-p53

transgene (Fig. 2B), arguing against a unique property of the E6 geneproduct in this sensitization. Furthermore, cisplatin sensitization wasnot limited to MCF-7 cells since p53 disruption in human colon

carcinoma RKO cells also enhanced sensitivity to cisplatin (Fig. 2B).The pCMV control plasmid did not affect cisplatin sensitivity compared to parental cells (data not shown). These results are consistentwith the recent findings that p53 disruption sensitizes colon carcinomaRKO cells to 254 nm UV (31), an agent like cisplatin that inducesDNA lesions which are repaired primarily through nucleotide excision(25-27).

p53 Disruption Suppresses Repair of a Cisplatin-damaged

Transfection Plasmid. Since the enhanced UV sensitivity of RKOcells lacking p53 function was previously associated with reducedDNA repair capacity (31), we investigated whether MCF-7 cells withdisrupted p53 exhibited a deficiency in repair of cisplatin-induced

DNA lesions. The assay consisted of damaging a CAT reporterplasmid with cisplatin in vitro, transfecting the damaged plasmid intohost cells, and then measuring CAT activity relative to a controltransfection with an undamaged plasmid. Host cells having greaterDNA repair activity should reactivate the damaged plasmid to agreater extent than cells having reduced DNA repair activity (33, 34).The results of a representative host cell reactivation assay are shownin Fig. 3A, and data compiled from two to three independent experiments are shown in Fig. 3B. Using the host cell reactivation assay, wefound that pCMV transfected MCF-7 cells were able to recover CATactivity from the CDDP-damaged plasmid to approximately 34% of

that achievable with the undamaged plasmid (Fig. 35). When thesame CDDP-damaged CAT plasmid was transfected into MCF-7/E6or MCF-7/mu-p53 cells, however, we only recovered approximately

14 and 23% of control CAT activity, respectively, indicating thatMCF-7/mu-p53 and MCF-7/E6 cells were approximately 1.5- to 2.5-fold less competent at reactivating the CDDP-damaged CAT plasmidcompared to the control cells. Therefore, the repair of CDDP-inducedDNA lesions in the transfected CAT-plasmid is greater when p53

function remains intact. Furthermore, the differences in reactivation ofthe CDDP-damaged plasmid in these lines was similar to the dose

modification factor for equivalent cell killing amongst the differentMCF-7 lines (Fig. 2).

PENT Synergizes with Cisplatin to Kill MCF-7/E6 Cells. Lossof p53 function could impair the ability of MCF-7 cells to survive

CDDP treatment because of loss of the G, checkpoint response and/orbecause of reduced DNA repair capacity. Loss of p53 function might,however, be partially compensated by activation of the G2 checkpoint,which was found to be operational in MCF-7 with disrupted p53 (Fig.1). The G2 checkpoint would delay entry of DNA-damaged cells into

mitosis, thereby allowing more time for constitutive DNA repairprocesses to remove DNA lesions before the critical chromosomesegregation process (1-5). Abrogation of G2 checkpoint functioncould, therefore, be especially deleterious in DNA-damaged cells

with defective p53. We investigated this possibility using PENT, a

Table 1 Effects of y-rays, ADR, VPI6, CDDP. and MMS on clonogenic survival of MCF-7/CMVand MCF-7/E6 cells

Shown is the mean and SD of the dose of each agent required to inhibit clonogenic survival by 50 and 90%. At least four data sets were used for the analysis.

Treatment7-RAYS

(Gy)ADR (IÕM)VP16 (/UM)CDDP (UM)MMS (ng/ml)MCF-7/

CMV3.5

±0.81.88 ±0.24.7 ±0.1

3.38 ±0.216.3 ±0.2ID.,,,MCF-7/E63.2

±0.61.88±0.44.7 ±0.2

1.69 ±0.316.3 ±0.1MCF-7/CMVMCF-7/E61.09

1.01.02.0*

1.0MCF-7/

CMV10

±0.47.9 ±0.224 ±0.2°

13.5 ±0.170 + 0.1ID90MCF-7/CMVMCF-7/E6

MCF-7/E68.8

±0.17.5 ±0.1

23.7 ±0.2"

6.9 ±0.166 ±0.1.1

.05

.01

.96''

.06

1651

on May 7, 2018. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

P53 AND CHEMOSENSmVITY

Fig. 3. Effect of E6 and mu-p53 on the ability of MCF-7cells to reactivate a CDDP-damaged CAT-plasmid. TheCAT-plasmid was either untreated ( —)or treated for 10 minwith 200 UM cisplatin (+) before transfection into MCF-7/CMV, MCF-7/E6, or MCF-7/mu-p53 cells. Seventy-two h

following transfection, the cells were lysed, and CAT activity was measured on a Betascope as described in "Materials and Methods." CAT values shown are the relative

conversion of chloramphenicol (Cm) to the acetylated species (Ac-Cm) relative to an undamaged plasmid in each cellline. All values shown were normalized to a ß-galactosidaseinternal control transfected along with the CAT-plasmid. A,

results from a representative experiment in which duplicatesamples were taken, ß.compiled data showing the meanand SD of CAT activity from two to three independentexperiments.

MCF-7/CMV MCF-7/E6 MCF-7/mu-p53

Ac-Cm

Cm —

PLASMID DAMAGE -++ -+ + - ++CONVERSION

'•• 8'51 e'55 '•" 8'17 8'19 '••B'22*•"MEAN

8.33 8.188.23BCELL

LINESMCF-7/CMVMCF-7/E6MCF-7/mu-p53%CONTROL ACTIVITY RELATIVE TOMCF-7/CMV34.1

± 5.52 (n=7)1.81

3.8 ±4.89 (n=7)8.422.5

± 1.91 (n=7) B.66

n = Sample Numbers

caffeine-derivative known to interfere with G2 checkpoint control (28,

29). Efforts were focused on the interaction of PENT with cisplatin inthe MCF-7/E6 line, which had already been found to be more sensi

tive to cisplatin compared to control transfected cells (Table 1; Fig. 2).Clonogenic survival was measured for MCF-7 and MCF-7/E6 cells

treated with cisplatin for l h and then incubated in PENT for 24 h. Fig.4 shows the results of our analysis. PENT did not affect the survivalof control MCF-7/CMV cells but strikingly enhanced CDDP-inducedcell killing of MCF-7/E6 cells. PENT reduced the surviving fractionof MCF-7/E6 cells treated with 5 and 10 JLLMCDDP by approximately

I

5 IB

CISPLATIN (/¿M)

15

O CISPLATIN• CISPLATINtPENT

D CISPLATIN• CISPLATINtPENT

MCF-7/CMV

H MCF-7/E6

Fig. 4. Effect of PENT on the cytotoxicity of cisplatin in MCF-7/CMV and MCF-7/E6

cells. Cells were treated with CDDP for l h and then incubated for 24 h with or without2 mM PENT. Clonogenic survival was measured as described in "Materials and Methods."

Values shown are the means of at least four data sets. Bars, SD.

2.3- and 30-fold, respectively (P < 0.01). PENT, therefore, alsoamplified the difference in CDDP sensitivity between MCF-7/CMVand MCF-7/E6 cells; at 5 JUMCDDP, PENT amplified a 2.4-folddifference into a 5.5-fold difference, and at 10 (UMcisplatin, PENTamplified a 7-fold difference into a 200-fold difference. The dose modification factor for equivalent cell killing in MCF-7 and MCF-7/E6 cells

for the combination of CDDP and PENT was approximately 3.G2 Checkpoint Abrogation by PENT May Depend on p53

Function. The effect of PENT on the G2 checkpoint of MCF-7/CMVand MCF-7/E6 cells was evaluated using ionizing radiation as ameans of introducing DNA damage with well-defined timing (Fig. 5).The effect of PENT was first examined in random-phase exponential

cultures (Fig. 5A) in an experiment that was otherwise similar to thatshown in Fig. 1. The results suggest that PENT interferes with theradiation-induced G2 arrest in MCF-7/E6 cells but not appreciably inMCF-7/CMV cells. This can be observed by comparing the G2-Mpopulation in the panels marked 6.3 Gy with the G2-M population in

the panels marked 6.3 Gy + PENT (Fig. 5A).In order to more clearly evaluate the effects of PENT on the G2

checkpoint of MCF-7/CMV and MCF-7/E6 cells, a similar experiment was performed using cells that were synchronized in G,-S with

the DNA polymerase inhibitor, aphidicolin (Fig. 55). Nocodazolecontrols (Fig. 5B, first panel) show that washing the cells free ofaphidicolin released the cells from the G,-S block and allowed them

to progress and arrest in mitosis. Without nocodazole present, themajority of cells entered G, of the next cycle (Fig. 5ß,second panel).Cells that were irradiated (panels marked 6.3 Gy) became arrestedprimarily in G2 phase. There was little opportunity to arrest in G,,because aphidicolin blocks cells past the G1 checkpoint.4 The appear

ance of some irradiated MCF-7/E6 cells in Gt at 24 h suggests that

4 P. M. O'Connor, unpublished observations.

1652

on May 7, 2018. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

P53 AND CHEMOSENSmVITY

EUlm

uiU

EUlm

i3

MCF-7/CMV

MCF-7/E6

J6.3 Gy 6.3 Gy >

PENTOXIFVLLINE

BEnim

G2/M

MCF-7/CMV

OCuiO

p

MCF-7/E6

L, J 6.3 Gy 6.3 Gy «

PENTOXIFVLLINE

Fig. 5. Effect of PENT on G, checkpoint function of MCF-7/CMV and MCF-7/E6cells exposed to 6.3 -y-rays. A, exponentially growing cells were irradiated with 6.3 "y-rays

and then incubated in the presence or absence of PENT (2 mM) for 16 h. Also shown isthe effect of nocodazole (0.4 /¿g/ml)given for 16 h to unirradiated cultures. B, cells weresynchronized in G,-S with aphidicolin (1 fig/ml; 24 h). Upon release, the cells weretreated as described in (A ), and samples were recovered for analysis 24 h later. The X-axis

shows propidium iodide fluorescence intensity used to assess DNA content of the cells.

these cells might arrest in G2 for a shorter period of time when p53function is disabled. Addition of PENT following the irradiationdecreased the population of MCF-7/E6 cells that arrested in G2 andallowed these cells to reenter G, (compare 6.3 Gy panel with panelmarked 6.3 Gy + PENT). PENT had much less effect on irradiatedMCF-7/CMV cells. We infer from these results that PENT can inhibitthe G2 checkpoint when p53 function is deficient but that it has areduced effect when p53 function is intact.

Discussion

In the present study, we investigated the effect of disrupting p53function on the sensitivity of human tumor cell lines not inherentlyprogrammed for apoptosis. The breast cancer MCF-7 and colon cancer RKO cell lines were suitable for this purpose, because they do notreadily undergo apoptosis following DNA damage, despite havingnormal p53 genes (22). In order to disrupt p53 function in MCF-7cells, we transfected the cells with the human papillomavirus type-16E6 gene, the product of which stimulates destruction of the p53protein (24, 35) or a dominant-negative mutant p53 gene. p53 functionwas tested by means of ionizing radiation which can be administeredin a controlled dose over a well-defined brief period of time and the

effects of which have been extensively characterized. We showed thatthe E6 and mu-p53 transfected cells were defective with respect to thenormal p53 responses to ionizing radiation: an increase in p53 proteinlevel, an induction of p2lWafl/c'P1,and arrest in G, phase of the cell

cycle (Fig. 1).Several types of DNA-damaging agents (7-rays, ADR, VP16,

CDDP, and MMS) were tested for differential killing of MCF-7/E6cells relative to control transfected cells using clonogenic survivalassays. Of these agents, only cisplatin exhibited a significant differential killing (Table 1; Fig. 2). Indeed, 90% of the E6-transfected cellswere killed at a dose approximately 2-fold lower than that required tokill the same proportion of control MCF-7/CMV cells (Fig. 2). Similarresults were obtained for MCF-7 cells expressing a mu-p53 transgene,indicating that the enhanced sensitivity was not a unique property ofthe E6 gene product. Furthermore, CDDP sensitization was not limited to MCF-7 cells since p53 disruption in human colon carcinomaRKO cells (present study) and the CDDP-resistant wild-type p53ovarian cancer cell line A2780/cp70 (36) also enhanced sensitivity tocisplatin.

The lack of dependence of radiosensitivity on p53 function inMCF-7 cells (Table 1) is in agreement with previous studies in coloncarcinoma RKO cells and murine fibroblasts (19) and studies in headand neck cancer cell lines (20) but contrary to studies in some othercell types (10, 12-18). These differences appear related to the apop-totic potential of the cell lines used in these different studies, althoughother possibilities are emerging (37). Taken together these studieshighlight the importance of cellular context to the evaluation of thebiological properties of the p53 tumor suppressor.

Cisplatin differs from the other agents tested in our study in that itcauses DNA damage that is extensively repaired through nucleotideexcision (25-27). The possibility that nucleotide excision repair has ap53-dependent aspect was suggested in recent studies on the role ofthe Gadd45 protein in the p53 pathway (21) and in a recent report thatdescribed sensitization of colon carcinoma RKO cells to UV lightfollowing p53 disruption (31). In order to test directly for an effect ofp53 function on the ability of cells to repair CDDP-induced DNAdamage, we damaged a reporter plasmid with CDDP and measuredthe ability of host cells to reactivate the transfected reporter gene (Fig.3). We found that expression of E6 or mu-p53 indeed reduced theability of cells to carry out this reactivation. The result agreed qualitatively and quantitatively with the previously reported effect of p53function in aiding the repair of a UV-damaged reporter plasmid inRKO cells (31). Thus, p53 appears to play a role in the repair of DNAdamaged by cisplatin as well as by UV light. The p53 regulated geneproduct Gadd45 appears to be one component of the p53 pathwaycontributing to this DNA repair system (21), although additionalcomponents are also likely (38). At the present time, we cannotexclude the possibility that p53 disruption affects other cellular processes that contribute to altered cisplatin sensitivity. For example,transfection of MCF-7 cells with a mu-p53 transgene has recentlybeen reported to suppress Bcl2 levels (23). We are, therefore, continuing to evaluate these cell lines.

Cell cycle checkpoints in G, and G2 phases protect DNA-damagedcells by delaying entry into S phase and into mitosis, respectively.Since these checkpoints provide more time for DNA repair, wehypothesized that when the p53-dependent G, checkpoint is inactivated, the protective role of the G2checkpoint should become increasingly important. We tested this hypothesis by inhibiting G2 checkpointfunction with PENT and expected to find that the combination ofCDDP and PENT would synergistically and selectively kill MCF-7cells with disrupted p53. We found this to be the case. We also foundthat PENT inhibited the G2 checkpoint more effectively in E6-trans-fected cells than in control cells. This was observed in assays of

1653

on May 7, 2018. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

P53 AND CHEMOSENSmVITY

random phase cultures or in cells previously synchronized in Gt-S

with aphidicolin (Fig. 5). Similar observations have been made withcaffeine by Powell et al. (39) and Russell et al. (40) and, takentogether, indicate that cells with defective p53 are more susceptible toG2 checkpoint inhibitors. The molecular basis for enhanced G2 checkpoint abrogation in cells with disrupted p53 remains to be determinedbut could involve an altered regulation of the cdc2 kinase (29, 41) incells with intact versus disrupted p53. If this is so, it would point to aninvolvement of the p53 tumor suppressor in G2 checkpoint regulation.

Although PENT is undergoing clinical trials following observationsthat this drug synergizes with DNA-damaging agents (28, 42), the

xanthines are limited by their interaction with a broad range of cellularprocesses (43). The development of more specific inhibitors of the G2checkpoint could overcome these limitations and offer a more precisemeans of exploiting differences in G2 checkpoint control betweennormal and tumor cells with defective p53 (1).

In summary, we have obtained evidence that disruption of p53function sensitizes human breast cancer MCF-7 and colon cancer

RKO cells to CDDP and may do so by: (a) preventing the G,checkpoint response to DNA damage; and/or (b) impairing nucleotideexcision repair of CDDP-induced DNA lesions. In addition, we pre

sent evidence that PENT enhances the selectivity of CDDP againstp53-defective cells by inhibiting the G2 checkpoint in such cells. We

are presently exploring the molecular basis for the differential effectsof PENT on the G2 checkpoint of cells with normal versus defectivep53 with a focus on cdc2 kinase regulation. The involvement of thep53 tumor suppressor in cell cycle checkpoint regulation and DNArepair processes could be exploited in the development of new che-

motherapeutic stratagems for cancer treatment.

References1. O'Connor, P. M., and Kohn, K. W. A fundamental role for cell cycle regulation in the

chemosensitivity of cancer cells? Semin. Cancer Biol., 3: 409-416, 1992.

2. Smith, M. L., and Fornace, Jr., A. J. Genomic instability and the role of p53 mutationsin cancer cells. Curr. Opin. Oncol., 7: 69-75, 1995.

3. Hartwell. L. H., and Weinert, T. A. Checkpoints: controls that ensure the order of cellcycle events. Science (Washington DC), 246: 629-634, 1989.

4. Hartwell, L. H., and Kastan, M. B. Cell cycle control and cancer. Science (Washington DC), 266: 1821-1825, 1994.

5. Murray, A. W. Creative blocks: cell-cycle checkpoints and feedback controls. Nature(Lond.), 359: 599-604, 1992.

6. Kuerbitz, S. J., Plunkett, B. S., Walsh, W. V., and Kastan, M. B. Wild-type p53 is acell cycle checkpoint determinant following irradiation. Proc. Nati. Acad. Sci. USA,89: 7491-7495, 1992.

7. Hollstein, M., Sidransky, D.. Vogelstein, B., and Harris, C. C. pS3 mutations inhuman cancers. Science (Washington DC), 253: 49-52, 1991.

8. Levine, A., Momand, J., and Finlay, C. A. The p53 tumor suppressor gene. Nature(Lond.), 351: 453-456, 1991.

9. Nelson, W. G., and Kastan, M. B. DNA strand breaks: the DNA template alterationsthat trigger p53 dependent DNA damage response pathways. Mol. Cell. Biol., 14:1815-1823, 1994.

10. Fan, S., El-Deiry, W. S., Bae, I., Freeman, J., Jondle, D., Bhatia, K., Fornace, A. J.,Jr., Magrath, I., Kohn, K. W., and O'Connor, P. M. p53 gene mutations are associated

with decreased sensitivity of human lymphoma cells to DNA damaging agents.Cancer Res., 54: 5824-5830, 1994.

11. Lane, D. P. p53: guardian of genome. Nature (Lond.), 358: 15-16, 1992.

12. Clarke, A. R., Purdie, C. A., Harrison, D. J., Morris, R. G., Bird, C. C, Hooper, M. L.,and Wyllie, A. H. p53 is required for radiation-induced apoptosis in mouse thymo-cytes. Nature (Lond.), 362: 847-849, 1993.

13. Lowe, S. W., Schmitt, E. M., Smith, S. W., Osborne, B. A., and Jacks, T. Thymocyteapoptosis induced by /753-dependent and independent pathways. Nature (Lond.), 362:849-852, 1993.

14. Lowe, S. W., Ruley, H. E., Jacks, T., and Houseman, D. E. p53-dependent apoptosismodulates the cytotoxicity of anticancer agents. Cell, 74: 957-967, 1993.

15. Radford, I. R., Murphy, J. M., Radley, J. M., and Ellis, S. L. Radiation response ofmouse lymphoid and myeloid cell lines. Part II. Apoptotic death is shown by all linesexamined. Int. J. Radial. Biol., 65: 217-227, 1994.

16. O'Connor, P., Jackman, J., Jondle, D., Bhatia, K., Magrath, I., and Kohn, K. W. Roleof the p53 tumor suppressor gene in cell cycle arrest and radiosensitivity of Burkitt's

lymphoma cell lines. Cancer Res., 53: 4776-4780, 1993.

17. Lee, J. M., and Bernstein, A. p53 mutations increase resistance to ionizing radiation.Proc. Nati. Acad. Sci. USA, 90: 5742-5746, 1993.

18. Mcllwrath, A. J., Vasey, P. A., Ross, G. M., and Brown, R. Cell cycle arrests andradiosensitivity of human tumor cell lines: dependence on wild-type p53 for radio-sensitivity. Cancer Res., 54: 3718-3722, 1994.

19. Slichenmyer, W. J., Nelson, W. G., Slebos, R. J., and Kastan, M. B. Loss of ap5J-associated G, checkpoint does not decrease cell survival following DNA damage. Cancer Res., 53: 4164-4168, 1993.

20. Brachman, D. G., Beckett, M., Graves, D., Haraf, D., Yokes, E., and Weichselbaum,R. R. p53 mutation does not correlate with radiosensitivity in head and neck cancercell lines. Cancer Res., 53: 3666-3669, 1993.

21. Smith, M. L., Chen, I., Zhan, Q., Bae, I., Chen, C., Gilmer, T., Kastan, M. B.,O'Connor, P. M., and Fornace, A. J., Jr. Interaction of the p53-regulated protein

Gadd45 with proliferating cell nuclear antigen. Science (Washington DC), 266:1376-1380, 1994.

22. Zhan, Q., Fan, S., Bae, I., Guillouf, C., Liebermann, D. A., O'Connor, P. M., and

Fornace, A. J., Jr. Induction of bax by genotoxic stress in human cells correlates withnormal p53 status and apoptosis. Oncogene, 9: 3743-3751, 1994.

23. Haldar, S., Negrini, N., Monne, M., Sabbinoni, S., and Croce, C. M. Down-regulationof bcl2 by p53 in breast cancer cells. Cancer Res., 54: 2095-2097, 1994.

24. Kessis, T. D., Slebos, R. J., Nelson, W. G., Kastan, M. B., Plunkett, B. S., Han, S. M.,Lorincz, A. T., Hedrick, L., and Cho, K. R. Human papillomavirus 16 E6 expressiondisrupts the p53-mediated cellular response to DNA damage. Proc. Nati. Acad. Sci.USA, 90: 3988-3992, 1993.

25. Ullah, S., Husain, L, Carlton, W., and Sanear, A. Human nucleotide excision repairin vitro: repair of pyrimidine dimers, psoralen and cisplatin adducts by HeLa cell-freeextract. Nucleic Acids Res., 17: 4471-4484, 1989.

26. Sanear, A., and Sanear, G. A. DNA repair enzymes. Annu. Rev. Biochem., 57:29-67, 1988.

27. Sanear, A. Mechanisms of DNA excision repair. Science (Washington DC), 266:1954-1956, 1994.

28. Fingert, H. J., Chang, J. D., and Pardee, A. B. Cytotoxic, cell cycle and chromosomaleffects of methylxanthines in human tumor cells treated with alkylating agents.Cancer Res., 46: 2463-2467, 1986.

29. O'Connor, P. M., Ferris, D. K., Pagano, M., Draetta, G., Pines, J., Hunter, T., Longo,

D. L., and Kohn, K. W. G2 delay induced by nitrogen mustard in human cells affectscyclin A/Cdk2 and cyclin Bl/Cdc2 kinase complexes differently. J. Biol. Chem., 268:8298-8308, 1993.

30. Chen, C., and Okayama, H. High efficiency transformation of mammalian cells byplasmid DNA. Mol. Cell. Biol., 7: 2745-2752, 1987.

31. Smith, M. L., Chen, I. T., Zhan, Q., O'Connor, P. M., and Fornace, A. J., Jr.

Involvement of the p53 tumor suppressor in repair of UV-type DNA damage.Oncogene, ¡npress, 1995.

32. Alley, M. C., Scudieco, D. A., Monks, A., el al. Feasibility of drug screening withpanels of human tumor cell lines using a microculture tetrazolium assay. Cancer Res.,48: 589-601, 1988.

33. Protic, M., Roilides, E., Levine, A. S., and Dixon, K. Enhancement of DNA repaircapacity of mammalian cells by carcinogen treatment. Somat. Cell Mol. Genet., 14:351-357, 1988.

34. Dabholkar, M., Eastman, A., and Reed, E. Host-cell reactivation of cisplatin-damaged pRSVcat in a human lymphoid cell line. Carcinogenesis (Lond.), 11:1761-1764, 1990.

35. Scheffner, M., Huibregtse, J. M., and Howley, P. M. Identification of a humanubiquitin-conjugating enzyme that mediates the E6-AP-dependent ubiquitination ofp53. Proc. Nati. Acad. Sci. USA, 91: 8797-8801, 1994.

36. Brown, R., Clugston, C., Burns, P., Edlin, A., Vasey, P., Vojtesek, B., and Kaye, S.Increased accumulation of p53 protein in cisplatin-resistant ovarian cell lines. Int. J.Cancer, 55: 678-684, 1993.

37. Di Leonardo, A., Linke, S. P., Clarkin, K., and Wahl, G. M. DNA damage triggers aprolonged p53-dependent G, arrest and long-term induction of Cipl in normal humanfibroblasts. Genes Dev., 8: 2540-2551, 1994.

38. Wang, X. W., Forrester, K., Yeh, H., Feitelson, M. A., Gu, J. R., and Harris, C. C.Hepatitis B virus X protein inhibits p53 sequence-specific DNA binding, transcrip-tional activity, and association with transcription factor ERCC3. Proc. Nati. Acad.Sci. USA, 91: 2230-2234, 1994.

39. Powell, S. N., DeFrank, J. S., Connell, P., Eogan, M., Preffer, F., Dombkowski, D.,Tang, W., and Friend, S. Differential sensitivity of p53- and p53+ cells to caffeine-induced radiosensitization and override of G2 delay. Cancer Res., 55: 1643-1648,1995.

40. Russell, K. J., Wiens, L. W., Galloway, D. A., and Groudine, M. Abrogation of theG2 checkpoint results in differential radiosensitization of G, checkpoint-deficient andcompetent cells. Cancer Res., 55: 1639-1642, 1995.

41. O'Connor, P. M., Ferris, D. K., Hoffmann, I., Jackman, J., Draetta, G., and Kohn,

K. W. Role of the cdc25C phosphatase in G2 arrest induced by nitrogen mustard.Proc. Nati. Acad. Sci. USA, 91: 9480-9484, 1994.

42. Fingert, H. J., Pu, A. T., Chen, Z., Googe, P. B., Alley, M. C, and Pardee, A. B. Invivo and in vitro enhanced antitumor effects by pentoxifylline in human cancer cellstreated with thiotepa. Cancer Res., 48: 4375-4381, 1988.

43. Rail, T. W. Drugs used in the treatment of asthma. In: A. G. Oilman, T. W. Rail, A. S.Nies, and P. Taylor (eds.), Goodman and Oilman's Pharmacological Basis of Ther

apeutics, Ed. 8, pp. 618-637. New York: Pergamon Press, 1985.

1654

on May 7, 2018. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1995;55:1649-1654. Cancer Res   Saijun Fan, Martin L. Smith, Dennis J. Rivert II, et al.   Cells to Cisplatin and PentoxifyllineDisruption of p53 Function Sensitizes Breast Cancer MCF-7

  Updated version

  http://cancerres.aacrjournals.org/content/55/8/1649

Access the most recent version of this article at:

   

   

   

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

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

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

  Permissions

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

.http://cancerres.aacrjournals.org/content/55/8/1649To request permission to re-use all or part of this article, use this link

on May 7, 2018. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from