Emergence of p53 Mutant Cisplatin-resistant OvarianCarcinoma Cells following Drug Exposure:Spontaneously Mutant Selection1

Sabina C. Righetti, Paola Perego, Elisabetta Corna,Marco A. Pierotti, and Franco Zunino2

Istituto Nazionale per lo Studio e la Cura dei Tumori, 20133 Milan, Italy

AbstractWe have previously shown that p53 mutations areassociated with cisplatin resistance in ovarian carcinomaIGROV-1/Pt 1 cells. The relationship between p53 statusand the development of resistance has not beencompletely elucidated; in particular, the biologicalmechanisms behind the acquired drug-resistant p53-mutant phenotype were not clearly explained. Thus, inthis study, we investigated whether the p53 mutationsfound in IGROV-1/Pt 1 cells (270 and 282 codons)resulted from selection, under the selective pressure ofthe cytotoxic treatment, of a spontaneously mutant cellpopulation preexistent in the cisplatin-sensitive parentalcell line (IGROV-1) or were induced by drug (genotoxic)treatment. For this purpose, an allele-specific PCRapproach was used. Primers carrying the desiredmutations (T3A codon 270, C3T codon 282) in the 3*

terminus, and the corresponding wild-type primers wereused to amplify genomic DNA from the original IGROV-1cell line used to select the mutant IGROV-1/Pt 1. Toincrease sensitivity, we hybridized blots of the PCRs withthe radiolabeled PCR fragment from IGROV-1/Pt 1.Amplification was obtained for IGROV-1 DNA with themutated allele-specific primers, indicating thepreexistence of a mutated population in the IGROV-1 cellline. Titration experiments suggested that the frequencyof the mutated alleles was <0.1%. Single-strandconformation polymorphism and allele-specific PCRanalysis of the IGROV-1/Pt 0.1 cells, which are lessresistant to cisplatin than IGROV-1/Pt 1 cells and whichcarry both mutant and wild-type p53 alleles with a wild-type predominance, suggested a progressive selection ofthe mutant population by cisplatin treatment. This is thefirst observation that indicates that a subpopulation ofp53 mutant cells can occasionally be selected bycisplatin treatment. Thus, considering the susceptibility to

spontaneous mutations of the p53 gene in advancedovarian carcinoma, the selection process resulting inemergence of p53 mutant tumors is a possible origin ofresistance of ovarian carcinoma to DNA-damagingagents. The survival advantage of p53 mutant cells in thepresence of genotoxic agents could be related to a lossof susceptibility to p53-dependent apoptosis and todefects in checkpoints pathways, resulting in genomicinstability.

IntroductionCisplatin is among the most effective agents that are clini-cally available for the treatment of a variety of solid tumors,including ovarian carcinoma. However, the development ofdrug resistance is a common problem for the efficacy of thepharmacological treatment (1, 2). Different mechanisms maycontribute to define the resistant phenotype, including alter-ations in drug-target interactions, expression of defenseand/or detoxification mechanisms, and cellular responses toDNA damage (3–5). In particular, it has been proposed thatreduction of the apoptotic response is a critical determinantof cisplatin efficacy (6, 7). Wild-type p53 is an importantcomponent of the pathway leading from DNA damage toapoptosis because p53 protein is implicated in multiplefunctions that include control of cell cycle, DNA repair, cellsenescence, genomic stability, and stress responses (8,9). Mutations of the p53 gene are common alterationsfound in a variety of human tumors (10). Although loss ofnormal p53 function can confer resistance to DNA-dam-aging agents as a consequence of a reduced cell suscep-tibility to apoptosis, the relevance of p53 mutations inchemosensitivity remains controversial (11–13). Severalstudies indicate that p53 can be inactivated in cisplatin-resistant cell systems (14, 15). Clinical studies support acorrelation between missense mutations and resistance toplatinum drug therapy (16).

On the basis of these observations, the aim of this studywas to clarify whether, in an ovarian carcinoma p53 mutantcisplatin-resistant variant, IGROV-1/Pt 1, the presence ofp53 mutations was a consequence of the genotoxic treat-ment or resulted from a selection process. For our purpose,we used allelic-specific gene amplification by PCR. Theseresults support that the p53 allele carrying a mutation atcodon 270 preexisted in the IGROV-1 parental cell line witha frequency ,0.1% and that the cell population exhibitingthe mutation can occasionally be selected by cisplatin treat-ment.

ResultsCellular Sensitivity to Cisplatin. Fig. 1 shows cellular sen-sitivity to cisplatin of IGROV-1 cells and cisplatin-resistant

Received 2/4/99; revised 4/29/99; accepted 5/24/99.The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indi-cate this fact.1 This work was partially supported by the Consiglio Nazionale delleRicerche, Finalized Project ACRO, (Rome), by the Associazione Italianaper la Ricerca sul Cancro (Milan) and by the Ministero della Sanita (Rome).2 To whom requests for reprints should be addressed, at Istituto Nazio-nale Tumori, Via Venezian 1, 20133 Milan, Italy. Phone: 39-02-2390267;Fax: 39-02-2390692; E-mail: zunino@istitutotumori.mi.it.

473Vol. 10, 473–478, July 1999 Cell Growth & Differentiation

sublines, including the p53 mutant IGROV-1/Pt 1 andIGROV-1/Pt 0.1 cells. IGROV-1/Pt 0.1 cells, which representan early step in the process of selection of the IGROV-1/Pt 1variant, exhibited a degree of resistance of 4. The IGROV-1/Pt 1 variant, in addition to carrying mutations in the p53gene at codons 270 and 282, exhibits a reduced expressionof Bax (14) and increased expression of Bcl-2 (17). Themolecular features of the resistant sublines is consistent witha reduced susceptibility to cisplatin-induced apoptosis (14).

p53 Gene Analysis of IGROV-1/Pt 0.1 Cells. A molecularanalysis of p53 status of IGROV-1/Pt 0.1 cells by SSCP3

analysis of exon 8 revealed the presence of two bands withaltered mobility in comparison to control DNA (Fig. 2). Suchbands corresponded to mutant alleles representing muta-tions at codons 270 and 282, because they comigrated withthe mutant alleles previously detected in the IGROV-1/Pt 1variant (14). The frequency of the mutant alleles in IGROV-1/Pt 0.1 cells was lower than in the IGROV-1/Pt 1 subline, asindicated by the weakness of the bands with altered mobility,showing that, in the IGROV-1/Pt 0.1 subline, mutant alleleswith a predominance of wild-type alleles were concomitantlypresent.

Allele-specific PCR. We have shown earlier that theIGROV-1/Pt 1 cell line carries two mutant alleles at exon 8 ofthe p53 gene, involving codons 270 (T3A) and 282 (C3T;Ref. 14). To determine whether a small population of p53mutant cells existed in the parental IGROV-1 cell line beforeselection with cisplatin, primers carrying the desired muta-tions in the 39 terminus and the corresponding wild-typeprimers were used in allele-specific PCR experiments. Asshown in Fig. 3, mutation-specific primers specifically am-

plified the p53 mutagenized sequences contained in thepC53-M plasmids (Lane 4), thus ruling out the possibility ofnonspecific amplification of wild-type sequence. Similarly,amplification was specific also with wild-type p53 sequenceprimers (Lane 1). When the wild-type primers were used toamplify genomic DNA from the cisplatin-sensitive IGROV-1cells and the two cisplatin-resistant variants IGROV-1/Pt 0.1and IGROV-1/Pt 1, amplification of all of the templates wasobtained (Fig. 4). Such primers were expected to amplify alsogenomic DNA of the resistant sublines because the wild-typeallele is still present in these cells (Fig. 2). With the mutatedprimers amplification was observed for all of the ovariancarcinoma cell lines (Fig. 4). However, the intensity of theamplified band was different, the highest being observed for

3 The abbreviations used are: SSCP, single-strand conformation polymor-phism; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bro-mide.

Fig. 1. Dose-response curvesfor the antiproliferative effects ofcisplatin against ovarian carci-noma IGROV-1 cells and cispla-tin-resistant sublines. Sensitivityto cisplatin was assessed by tet-razolium dye (MTT) assay after96 h exposure. IC50, drug con-centrations required for 50% in-hibition of cell growth. RI, resist-ance index (ratio of IC50 ofparental sensitive cells to IC50 ofresistant cells).

Fig. 2. p53 gene status in the IGROV-1 cell systems. p53 gene statuswas evaluated by SSCP analysis. Exon 8 analysis is reported.

474 Cisplatin Resistance and Mutant p53 Cell Selection

IGROV-1/Pt 1 cells. Amplification was intermediate betweenIGROV-1 and IGROV-1/Pt 1 cells for the IGROV-1/Pt 0.1 cellvariant. Thus, the intensity of the amplified band paralleledthe amount of p53 mutation present based on SSCP analysisin cisplatin-resistant cells, suggesting that the mutation ex-isted before drug selection in the IGROV-1 cell line. Titrationexperiments were performed to estimate the proportion ofp53 mutated cells preexistent in the IGROV-1 cell line.Genomic DNA from IGROV-1/Pt 1 was mixed with the DNAof wild-type p53 cells, at different ratios ranging from 1:10 to1:10,000. As shown in Fig. 5, allele-specific PCR of thesetemplates revealed bands with an intensity corresponding tothe amount of the mutated DNA present in the mixture, witha detection limit of 0.001. The intensity of the IGROV-1 band,amplified with both the mutation-specific primers, corre-sponded to the signal of the band amplified when DNA frommutant and wild-type cells were mixed at a ratio between1:10 and 1:100 (Fig. 5). False-positive reactions caused bycontamination of DNA or nonspecific amplification of wild-type sequences were ruled out by appropriate controls. Al-though these results were consistently reproduced, allele-specific PCR revealed no amplification using as templateother independent DNA extractions from IGROV-1 cell line.Therefore, to further increase detection sensitivity, we per-formed blots of allele-specific PCR. At the annealing tem-

peratures used for non-hot PCR, weak positive signals wereobserved also with DNA from cell line carrying wild-type p53,suggesting that, at these temperatures, the primers were notcompletely destabilized. As shown in Fig. 6A, with the 270mutated primer higher annealing temperature (68°C) couldeliminate amplification on different wild-type p53 DNA tem-plates. However, under this condition, amplification occurredfor four independently prepared IGROV-1 templates (Fig. 6B;IGROV-1 I–IV), the highest intensity being confirmed forIGROV-1 I. Titration experiments confirmed that the intensityof the IGROV-1 I band corresponded to a ratio between 1:10and 1:100 of the reference mixtures, whereas for the otherthree extractions, it was lower than 0.001, indicating that,0.1% of the sample represented preexistent populationwith p53 mutation. Blot analysis of allelic-specific PCR ex-periments, performed with the 282 mutated primer, at its ownannealing temperature, also revealed weak amplification onwild-type templates, including negative controls, althoughthe amplification was lower than that obtained with theIGROV-1 DNA. However, under this extreme condition,higher annealing temperatures also completely abolishedsignals on the IGROV-1 templates (data not shown).

Discussionp53 is one of the most frequently mutated genes in humantumors (18), including advanced ovarian carcinoma. An as-sociation between p53 mutations and resistance to DNA-damaging agents has been documented in both preclinicaland clinical studies (14, 16, 19). Because induction or selec-tion of mutations is a controversial aspect of development ofdrug resistance, a problem that has remained unsolved since1979 (20), in this study, we took advantage of a previouslyselected cisplatin-resistant cellular system IGROV-1/Pt 1,characterized by mutant p53, to address the issue of whethercisplatin resistance results from a selection of a small frac-tion of p53 mutant cells preexisting in the original drug-

Fig. 3. Allelic-specific PCR on plasmid templates. A, PCR for the 270mutation (primers 6As/p53-270w or p53-270m); B, PCR for the 282 mu-tation (primers 6As/p53-282w or p53-282m). Ten ng of plasmid DNA wereamplified (30 cycles). Sample loading was as follows: Lane 1, wild-typep53 plasmid 1 wild-type primers; Lane 2, mutated p53 plasmid 1 wild-type primers; Lane 3, wild-type p53 plasmid 1 mutated primers; Lane 4,mutated p53 plasmid 1 mutated primers: Lane 5, no DNA 1 wild-typeprimers; Lane 6, no DNA 1 mutated primers; Lane M, markers.

Fig. 4. Allelic-specific PCR analysis of the p53 270 mutation on genomicDNA. Lane M, markers; Lanes 1, primers p53-8.3/p53-270w; Lanes 2,primers p53-8.3/p53-270m; Lane 2, negative control (no DNA).

Fig. 5. Allelic-specific PCR analysis: titration experiments. The genomicDNA of IGROV-1/Pt 1 was mixed with DNA from wild-type p53 cells atvarious ratios. A, primers p53-8.3/p53-270m; B, primers p53-8.3(-7)/p53-282m. Lane M, markers; Lane 2, no DNA.

475Cell Growth & Differentiation

sensitive cell line prior exposure to increasing drug concen-trations. For this purpose, sensitive procedures that are ableto discriminate between the wild-type and the mutated pop-ulations were used. In particular, allelic-specific gene ampli-fication by PCR was chosen because it allows detection ofmutations below the detection threshold of PCR-SSCP. Un-fortunately, the success of this technique is tightly depend-ent on the type of mutation, and appropriate conditionscannot be worked out for all of the mutations. These resultsare consistent with the interpretation that the p53 mutationsfound in the cisplatin-resistant IGROV-1/Pt 1 cell line preex-isted in the parental IGROV-1 cells and were selected bydrug exposure. In fact, data obtained by SSCP analysisindicated that the IGROV-1/Pt 0.1 cells, which represent anearly step during the process of selection of IGROV-1/Pt 1,carry both the wild-type and mutant p53 alleles with a wild-type predominance, thus suggesting the occurrence of aprogressive selection of the mutated population operated bycisplatin. We hypothesized that SSCP and sequence analy-sis of IGROV-1 genomic DNA could have been unable toreveal the presence of the hypothetical fraction of p53 mu-tated cells because of their detection threshold. Thus, allelic-specific gene amplification, having a greater theoretical limitof sensitivity, was performed. Experiments with plasmidicDNA demonstrated that primers carrying the specific muta-tions at the 39 end exclusively amplified the mutant DNA,thus supporting the specificity of sequence amplification.Using wild-type primers, we obtained amplification ofgenomic DNA from IGROV-1 and the two cisplatin-resistantvariants IGROV-1/Pt 0.1 and IGROV-1/Pt 1, as expected,based on the presence of not only mutant but also wild-typealleles detected in these cells in PCR-SSCP experiments.Mutant primers amplified all of the three cell lines and theintensity of the bands obtained from the two resistant cellsystems corresponded to the amount of p53 mutations pres-ent, based on SSCP analysis. The weak band amplified fromIGROV-1 indicated that mutant cells existed before drugselection. The fact that, in non-hot allele-specific PCR ex-periments, amplification was found only in one of four DNAextractions from IGROV-1 cell line suggests that, in inde-pendently grown cell cultures, spontaneous enrichment ofp53 mutant cells can occur. Blots of the 270 PCRs confirmedthe presence of the mutated clone in the other three extrac-tions of IGROV-1 cells, at a lower level than that observed forthe first extraction (0.1%). Thus, the fraction of cells contain-ing the p53 mutations at codon 270 is variable in indepen-dently cultured cell populations. The major problem in usingthe labeling procedure was that, at the working primer an-nealing temperatures, a weak band was revealed also with

genomic DNAs from cell lines with wild-type p53 sequence.With the 270 mutated primer, using higher annealing tem-perature, it was possible to eliminate the amplification in thenegative control. This was not the case of the 282 primer, forwhich nonspecific annealing was consistently observed(data not shown). Such a finding, which is probably related tothe nature of the primer, does not allow definitive conclu-sions on the preexistence of the mutation at codon 282because the two mutations found in IGROV-1/Pt 1 cells arelocalized in different alleles (14). It is unlikely that the muta-tions belong to distinct mutant clones because they havebeen selected at the same frequency in two cisplatin-resist-ant sublines (IGROV-1/Pt 0.5 and IGROV-1/Pt 1; Ref. 14).Thus, a plausible explanation of the occasional selection ofresistant mutant cells produced during drug treatment is thatit is the result of both selective pressure of the drug treatmentand induction of an additional mutation, produced by thegenotoxic stress of the drug itself. The presence of twoconcomitant mutations may be a favorable event allowingthe emergence of mutant resistant cells. It is likely that theselection mechanism involves a reduced susceptibility toapoptosis as a consequence of inactivation of p53 gene. Thecapability of cisplatin to efficiently kill ovarian carcinoma cellswith wild-type p53 is expected to result in a survival advan-tage for mutant cells. Indeed, cisplatin-induced apoptosishas been shown to be favored in cell lines carrying wild-typep53 (21). In our model system, cisplatin-induced apoptosishas been postulated to be p53 dependent because a re-duced apoptotic response of resistant p53 mutant cells isassociated with loss of p53 function (14). A further reason forthe development of a resistant phenotype in p53 mutant cellsis that loss of p53 function contributes to genomic instabilityas a consequence of defects in checkpoint pathways andgain of novel functions, including overexpression of defensefactors (22). This interpretation is consistent with a prelimi-nary observation that the introduction of a mutant p53(codon 282) in wild-type IGROV-1 cells conferred a low levelof resistance to cisplatin (data not shown).

Despite the technical difficulty to document the sequenceof the events involved in the development of this type ofresistance (i.e., p53-mediated), our model provides indirectevidence that selection of spontaneous or induced mutantcells is a process involved in the development of clinicalresistance. A high degree of resistance to DNA-damagingagents may arise readily after short periods of treatment orgradually as a result of treatment over prolonged periods.The known tumor heterogeneity in terms of p53 mutationsmay account for the pattern of response of ovarian carci-noma to conventional treatments, based on DNA-damaging

Fig. 6. Allelic-specific PCR anal-ysis of the p53 270 mutation ongenomic DNA: blot with labeledDNA fragment from IGROV-1/Pt1. A, negative controls (wild-typep53 cells) and IGROV-1 cells. B,genomic DNA of IGROV-1/Pt 1cells was mixed at different ratiowith DNA from wild-type p53 cells(H460).

476 Cisplatin Resistance and Mutant p53 Cell Selection

agents (i.e., cisplatin and alkylating agents; Ref. 16). Indeed,intrinsic drug resistance to cisplatin-based first line therapywas found to be correlated with missense mutations of p53gene (16). In partial responses, resistance may arise quicklyand relatively easily if mutant resistant cells constitute anappreciable fraction of tumor cell population at the time oftherapy.

Materials and MethodsCell Lines and Culture Conditions. The IGROV-1 ovarian carcinoma cellline originally obtained from Dr. J. Benard (Institut Gustave Roussy, Ville-juif, France) was cultured in RPMI 1640 with 10% fetal bovine serum (LifeTechnologies, Inc., Gaithersburg, MD). The cisplatin-resistant variantIGROV-1/Pt 1 was generated by continuous exposure of the IGROV-1 cellline to increasing concentration of cisplatin starting from 0.1 up to 1 mg/ml(23). IGROV-1/Pt 0.1 cells represents an early passage during the processof selection of IGROV-1/Pt 1 cells and were cultured in 0.1 mg/ml cisplatin.

Cytotoxicity Studies. Cytotoxicity after a 96-h exposure to cisplatinwas assessed by tetrazolium dye (MTT) assay (24). This antiproliferativeassay was used because the colony-forming assay was not suitable in ourcell system as a consequence of a low plating efficiency. However, apreliminary evaluation of the cytotoxic effects using the colony formingassay supports a comparable resistance index of the IGROV-1/Pt 1 foundwith the MTT method. Preliminary experiments were performed to deter-mine the appropriate seeding number of cells (4000 cells/well) after con-firming the linear relationship between the absorbance and the number ofcells in the growth curve of each cell line. The IC50 is defined as theinhibitory drug concentration causing a 50% reduction of A550 nm

(decrease of cell growth) over that of untreated control.Allele-specific DNA Amplification. Allele-specific DNA amplification

was carried out in PCR experiments to amplify 0.35 mg of genomic DNAextracted from IGROV-1, IGROV-1/Pt 1, and IGROV-1/Pt 0.1 cell lines,according to standard techniques (25). For IGROV-1 cells, four indepen-dent DNA extractions (from cells at comparable passages but from dif-ferent thawings, named I, II, III, and IV) of the original cell line used toselect the IGROV-1/Pt 1 subline were used. Primers carrying the desiredmutations (T3A codon 270, C3T codon 282; Table 1) in the 39 terminusand the corresponding wild-type primers were used. The oligonucleotidethat was complementary to primers p53-270m or p53-270w was the sameas that used for PCR-SSCP analysis (p53-8.3; amplified band, 193 bp),whereas that used with primers p53-282m and p53-282w was 7 nucleo-tides upstream [designated p53–8.3(-7); amplified band, 152 bp] formatching the melting temperature. All of the reactions were performedwith KlenTaq1 DNA polymerase (BioNova s.r.l., Bologna, Italy), an enzymecharacterized by a high heat stability, in a volume of 50 ml containing

template DNA, primers (0.5 mM each), dNTPs (200 mM each), and reactionbuffer. The mutant allele was specifically amplified at the annealing tem-perature at which no amplification was revealed with genomic DNA fromcell lines used as negative controls (H460, A549, and N592). For the 270mutation, reactions were carried out for 30 cycles at 94°C for 1 min, 68°C(58°C for the wild-type allele) for 40 s, and 72°C for 1 min. For the 282mutation, reactions were: 94°C for 1 min, 72°C (62°C for the wild-typeallele) for 30 s, and 72°C for 1 min (30 cycles). An aliquot (25 ml) of eachPCR was run in a 2% agarose gel, transferred to nitrocellulose membrane,and hybridized with the corresponding radiolabeled PCR mutated frag-ment from IGROV-1/Pt 1 (Megaprime DNA labeling system, Amersham,Little Chalfont, United Kingdom). To estimate the proportion of mutantalleles, genomic DNA from IGROV-1/Pt 1 was mixed at various ratios withgenomic DNA from wild-type p53 cells (from 1:10 to 1:10000) and sub-jected to allele-specific PCR. All of the experiments were repeated at leastthree times.

Allele-specific PCR analysis was used for amplification of plasmid DNA.Ten ng of the pC53 plasmid carrying wild-type (pC53-SN3) or mutated(pC53-M) sequences were amplified using the AmpliTaq DNA polymerase(Perkin-Elmer Corp., Branchburg, NJ) with the same PCR condition de-scribed above. The downstream primer (p53–6As) was specific for cDNAsequence and was used with all of the upstream primers (expected bandsof 281 bp for 270 mutation and 245 bp for the 282 mutation).

SSCP and Sequencing Analysis. SSCP analysis for the detection ofp53 gene mutation has been described previously (26). Primers used toamplify exon 8 were p53-8.3 and p53-8.5 (Table 1). PCR-amplified exonswere subjected to direct DNA sequencing with an AmpliCycle Sequencingkit (Perkin-Elmer Corp.). To sequence p53 cDNAs inserted into pC53plasmids, we used primer 5S.

Plasmid Mutagenesis. The pC53-SN3 plasmid (27) carrying wild-type p53 cDNA was mutagenized at sites 270 and 282 using theQuikChange site-directed mutagenesis kit (Stratagene, San Diego, CA).Briefly, for each site, two synthetic oligonucleotide primers containing thedesired mutation (p53-270As/p53-270S and p53-282As/p53-282S; Table1) were used to amplify the plasmid template with Pfu DNA polymerase,which replicates both plasmid strands with high fidelity and without dis-placing the mutant primer. The PCR product was then treated with theDpnI endonuclease to digest the parental DNA template and to select formutation-containing synthesized DNA. The nicked vector incorporatingthe mutation was then transformed into Epicurian Coli XL-1-Blue super-competent cells, and mutagenized plasmids were subjected to directsequencing.

AcknowledgmentsWe thank Dr. M. Asada for helpful discussion and Laura Zanesi for

editorial assistance.

References1. Weiss, R. B., and Christian, M. C. New cisplatin analogues in devel-opment. A review. Drugs, 46: 360–377, 1993.

2. Thigpen, T., Vance, R., Puneky, L., and Khansur, T. Chemotherapy inadvanced ovarian carcinoma: current standards of care based on ran-domized trials. Gynecol. Oncol., 55: 97–107, 1994.

3. Kelland, L. R. The molecular basis of cisplatin sensitivity/resistance.Eur. J. Cancer, 30A: 725–727, 1994.

4. Chu, G. Cellular responses to cisplatin. J. Biol. Chem., 269: 787–790,1994.

5. Dancey, J., and Le Chevalier, T. Non-small cell lung cancer: an over-view of current management. Eur. J. Cancer, 33: S2–S7, 1997.

6. Zunino, F., Perego, P., Pilotti, S., Pratesi, G., Supino, R., and Ar-camone, F. Role of apoptotic response in cellular resistance to cytotoxicagents. Pharmacol. Ther., 76: 177–185, 1997.

7. Hickman, J. A. Apoptosis and chemotherapy resistance. Eur. J. Can-cer, 32A: 921–926, 1996.

8. Mowat, M. R. A. p53 in tumor progression: life, death and everything.Adv. Cancer Res., 74: 25–48, 1998.

9. Steele, R. J., Thompson, A. M., Hall, P. A., and Lane, D. P. The p53tumor suppressor gene. Br. J. Surg., 85: 1460–1467, 1998.

Table 1 Allele-specific primers and primers used for mutagenesisa

Primer Sequence

Primers for mutagenesisp53-270S 59-CGGAACAGCTTAGAGGTGCGTGT-39p53-270As 59-ACACGCACCTCTAAGCTGTTCCG-39p53-282S 59-CTGGGAGAGACTGGCGCACAGAG-39p53-282As 59-CTCTGTGCGCCAGTCTCTCCCAG-39

Mutation allele-specific primersp53-282w 59-TGCCTGTCCTGGGAGAGACC-39p53-282m 59-TGCCTGTCCTGGGAGAGACT-3 9p53-270w 59-TACTGGGACGGAACAGCTTT-39p53-270m 59-TACTGGGACGGAACAGCTTA-3 9

Primers for wild-type sequencep53-8.3 59-AAGTGAATCTGAGGCATAAC-39p53-8.3(-7) 59-TCTGAGGCATAACTGCACCC-39p53-8.5 59-TATCCTGAGTAGTGGTAATC-39p53-5S 59-TGCCCTATGAGCCGCCTGAG-39p53-6As 59-CTTCCCAGCCTGGGCATCCT-39

aS, sense; As, antisense; m, mutated; w, wild-type. Boldface type indi-cates nucleotide substitution.

477Cell Growth & Differentiation

10. Hartmann, A., Blaszyk, H., Kovach, J. S., and Sommer, S. S. Themolecular epidemiology of p53 gene mutations in human breast cancer.Trends Genet., 13: 27–33, 1997.

11. Eliopoulos, A. G., Kerr, D. J., Herod, J., Hodgking, L., Krajewski, S.,Reed, J. C., and Young, L. S. The control of apoptosis and drug resistancein ovarian cancer: influence of p53 and Bcl-2. Oncogene, 11: 1217–1228,1995.

12. Milner, J. DNA damage, p53 and anticancer therapies. Nat. Med., 1:879–880, 1995.

13. Fan, S., El-Deiry, W. S., Bae, I., Freeman, J., Jondle, D., Bathia, K.,Fornace, A. J., Magrath, I., Kohn, K. W., and O’Connor, M. p53 genemutations are associated with decreased sensitivity of human lymphomacells to DNA-damaging agents. Cancer Res., 54: 5824–5830, 1994.

14. Perego, P., Giarola, M., Righetti, S. C., Supino, R., Caserini, C., Delia,D., Pierotti, M. A., Miyashita, T., Reed, J. C., and Zunino F. Associationbetween cisplatin resistance and mutation of p53 gene and reduced Baxexpression in ovarian carcinoma cell systems. Cancer Res., 56: 556–562,1996.

15. Gallagher, W. M., Cairney, M., Schott, B., Roninson, I. B., and Brown,R. Identification of p53 genetic suppressor elements which confer resist-ance to cisplatin. Oncogene, 14: 185–193, 1997.

16. Righetti, S. C., Della Torre, G., Pilotti, S., Menard, S., Ottone, F.,Colnaghi, M. I., Pierotti, M. A., Lavarino, C., Cornarotti, M., Oriana, S.,Bohm, S., Bresciani, G. L., Spatti, G., and Zunino, F. A comparative studyof p53 gene mutations, protein accumulation and response to cisplatin-based chemotherapy in advanced ovarian carcinoma. Cancer Res., 56:689–693, 1996.

17. Perego, P., Righetti, S. C., Supino, R., Delia, D., Caserini, C., Carenini,N., Bedogne, B., Broome, E., Krajewski, S., Reed, J. C., and Zunino, F.Role of apoptosis and apoptosis-related proteins in the cisplatin-resistantphenotype of human tumor cell lines. Apoptosis, 2: 540–548, 1997.

18. Harris, C. C. Structure and function of the p53 tumor suppressorgene: clues for rational cancer therapeutic strategies. J. Natl. Cancer Inst.,88: 1442–1455, 1996.

19. Kirsch, D. G., and Kastan, M. B. Tumor suppressor p53: implicationsfor tumor development and prognosis. J. Clin. Oncol., 16: 3158–3168,1998.

20. Goldie, J. H., and Coldman, A. J. A mathematic model for relating thedrug sensitivity of tumours to their spontaneous mutation rate. CancerTreat. Rep., 63: 1727–1733, 1979.

21. O’Connor, P. M., Jackman, J., Bae, I., Myers, T. G., Fan, S., Mutoh,M., Scudiero, D. A., Monks, A., Sausville, E. A., Weinstein, J. N., Friend, S.,Fornace, A. J., Jr., and Kohn K. W. Characterization of the p53 tumorsuppressor pathway in cell lines of the National Cancer Institute antican-cer drug screen and correlation with the growth-inhibitory potency of 123anticancer agents. Cancer Res., 57: 4285–4300, 1997.

22. Harris, C. C. Structure and function of the p53 tumor suppressorgene: clues for rational cancer therapeutic strategies. J. Natl. Cancer Inst.,88: 1442–1455, 1996.

23. Perego, P., Romanelli, S., Carenini, N., Magnani, I., Leone, R., Bonetti,A., Paolicchi, A., and Zunino, F. Ovarian cancer cisplatin-resistant celllines: multiple changes including collateral sensitivity to taxol. Ann. Oncol.,9: 423–430, 1998.

24. Alley, M. C., Scudiero, D. A., Monks, A., Hursey, M. L., Czerwinski,M. J., Fine, D. L., Abbott, B. J., Mayo, J. G., Shoemaker, R. H., and Boyd,M. R. Feasibility of drug screening with panels of human tumor cell linesusing a microculture tetrazolium assay. Cancer Res., 48: 589–601, 1988.

25. Sambrook, J., Fritsch, E. F., and Maniatis, T. Molecular cloning: ALaboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Labo-ratory Press, 1989.

26. Donghi, R., Longoni, A., Pilotti, S., Michieli, P., Della Porta, G., andPierotti, M. A. Gene p53 mutations are restricted to poorly differentiatedand undifferentiated carcinomas of the thyroid gland. J. Clin. Invest., 91:1753–1760, 1993.

27. Baker, S. J., Markowitz, S., Fearon, E. R., Willson, J. K. V., andVogelstein, B. Suppression of human colorectal carcinoma cell growth bywild-type p53. Science (Washington DC), 249: 912–915, 1990.

478 Cisplatin Resistance and Mutant p53 Cell Selection