Cancer Chemosensitivity Profiles Identify Polymorphisms in ...cancerres.aacrjournals.org/content/canres/70/1/172.full.pdf · Chemosensitivity Profiles Identify Polymorphisms in the

172

Published OnlineFirst December 8, 2009; DOI: 10.1158/0008-5472.CAN-09-2218

Molecular and Cellular Pathobiology
Cancer
Research
Chemosensitivity Profiles Identify Polymorphisms inthe p53 Network Genes 14-3-3τ and CD44 ThatAffect Sarcoma Incidence and Survival
Alexei Vazquez1,2,3, Lukasz F. Grochola4, Elisabeth E. Bond4, Arnold J. Levine1,2,3, Helge Taubert5,Thomas H. Müller6, Peter Würl7, and Gareth L. Bond4

Abstract

Authors' AJersey; 2Thand DentisBrunswickUniversity oMaxillofaciaHalle, GerNSTOB, SpFranziskus

Note: SupResearch O

A. Vazquez

CorresponResearch, URoad CamPhone: 44bond@ndm

doi: 10.115

©2010 Am

Cancer R

Dow

The p53 regulatory network responds to cellular stresses by initiating processes such as cell cycle arrest andapoptosis. These responses inhibit cellular transformation and mediate the response to many forms of cancertherapies. Functional variants in the genes comprising this network could help identify individuals at greaterrisk for cancer and patients with poorer responses to therapies, but few such variants have been identified asyet. We use the NCI60 human tumor cell line anticancer drug screen in a scan of single nucleotide polymorph-isms (SNP) in 142 p53 stress response genes and identify 7 SNPs that exhibit allelic differences in cellularresponses to a large panel of cytotoxic chemotherapeutic agents. The greatest differences are observed forSNPs in 14-3-3τ (YWHAQ; rs6734469, P = 5.6 × 10−47) and CD44 (rs187115, P = 8.1 × 10−24). In soft-tissue sar-coma patients, we find that the alleles of these SNPs that associate with weaker growth responses to che-motherapeutics associate with poorer overall survival (up to 2.89 relative risk, P = 0.011) and an earlier ageof diagnosis (up to 10.7 years earlier, P = 0.002). Our findings define genetic markers in 14-3-3τ and CD44 thatmight improve the treatment and prognosis of soft-tissue sarcomas. Cancer Res; 70(1); 172–80. ©2010 AACR.

Introduction

The p53 protein plays a central role in eliciting cellular re-sponses to a variety of stress signals and is crucial to an in-dividual's ability to ward off cellular transformation and torespond to many forms of DNA damage–inducing cancertherapies (1, 2). A body of evidence is emerging in the liter-ature that suggests that the inherited genetics of the p53pathway could be used to further define patient populationsin their abilities to respond to stress, suppress tumor forma-tion, and induce p53 activity in response to DNA-damagingtherapies (3, 4). The common inherited genetic variations inthe p53 pathway most frequently studied are single nucleo-

ffiliations: 1The Institute for Advanced Study, Princeton, Newe Cancer Institute of New Jersey and 3University of Medicinetry, New Jersey-Robert Wood Johnson Medical School, New, New Jersey; 4Ludwig Institute for Cancer Research,f Oxford, Oxford, United Kingdom; 5Department of Oral andl Plastic Surgery, Martin-Luther-University Halle-Wittenberg,many; 6German Red Cross Blood Transfusion Serviceringe, Germany; and 7Department of Surgery, Malteser St.Hospital, Flensburg, Germany

plementary data for this article are available at Cancernline (http://cancerres.aacrjournals.org/).

and L.F. Grochola contributed equally to this work.

ding Author: Gareth L. Bond, The Ludwig Institute for Cancerniversity of Oxford, Old Road Campus Research Building, Old

pus, Off Roosevelt Drive, Oxford OX3 7DQ, United Kingdom.-1865-617497; Fax: 44-1865-617515; E-mail: gareth..ox.ac.uk.

8/0008-5472.CAN-09-2218

erican Association for Cancer Research.

es; 70(1) January 1, 2010

Researcon Februarycancerres.aacrjournals.org nloaded from

tide polymorphisms (SNP) in the p53 (p53 codon72; refs. 5, 6)and MDM2 (MDM2 SNP309; ref. 7) genes. The differentalleles of p53 codon72 encode either a proline or arginine res-idue residing in a proline-rich region of p53 that has beenshown to be important in mediating the apoptotic response(8). The MDM2 SNP309 locus results in either a thymine (T)or a guanine (G) in the intronic promoter/enhancer region ofthe MDM2 oncogene, which encodes a key negative regulatorof p53 (7). In tumor-derived cell lines, the elevated levels ofMDM2 associated with the G allele have been shown in mul-tiple studies to result in the attenuation of the p53 apoptoticresponse after exposure to multiple types of chemothera-peutics compared with cells containing the T allele (7, 9,10). These observations for both p53 codon72 and MDM2SNP309 correlate well with studies of many types of cancersfor which allelic differences have been noted for these loci inthe onset of and risk for cancer, as well as in the response totherapy and survival (3, 4, 11).The above-mentioned studies suggest that this important

stress response pathway could harbor more functional inher-ited genetic variants, the study of which could help furtherdefine patient populations in their abilities to respond tostress, suppress tumor formation, and respond to DNA-damaging therapies. Recently, we proposed a methodologyto uncover candidate functional SNPs using data derivedfrom the NCI60 human tumor cell line anticancer drugscreen (4). NCI60 is a panel of 60 cancer cell lines from dif-ferent tissues of origin (12). Much is known of their genetics:the mutational status of more than 20 important cancer-related genes and the genotypes of 100,000 SNPs (13, 14). Agreat deal is also known of their response to therapies: both

h. 28, 2019. © 2010 American Association for Cancer

http://cancerres.aacrjournals.org/

SNPs in YWHAQ and CD44, Chemosensitivity, and STS


the growth response of these cell lines to more than 100 stan-dard chemotherapeutic agents as well as to more than 40,000publicly available potential anticancer agents (12). Thegrowth response was measured 48 hours after treatment ata five-log concentration range for each compound. The neg-ative log10 (GI50) of the concentration required to inhibit thegrowth of the given cell line by 50% relative to the vehicle-treated cells is calculated and reported (12). A previous studyshowed that the cells with mutations in the p53 gene are lesssensitive to many standard chemotherapeutic agents that in-duce DNA damage, suggesting that the cellular response tothese agents is p53 dependent (15). We developed an analyticframework that we implemented to mine the publicly avail-able genotype, mutational, and cellular drug response data toidentify SNPs in the p53 stress response that merited futurestudy (4). In this report, we build upon, expand, and use thismethodology and identify genetic variants in the CD44 andYWHAQ (14-3-3τ) genes that significantly affect both thecellular responses to many chemotherapeutics, as well ashuman cancer incidence and survival.

Materials and Methods

NCI60 human tumor cell line anticancer drug screen.The mutational status of p53 and 20 important cancer-related genes, the genotypes of more than 100,000 SNPs(Affymetrix 125K chip), and the GI50 data for the NCI60 cellpanel were obtained from the National Cancer Institute(NCI)/NIH Developmental Therapeutics Program.8 The ge-nomic DNA from the NCI60 panel of cell lines was a generouscontribution from the NCI-Division of Cancer Treatment andDiagnosis Repository Molecular Characterization Program.Soft-tissue sarcoma patients. One hundred twenty-nine

patients (73 females and 56 males; ages 14–87 y; mean 55.9 y)diagnosed with soft-tissue sarcomas (STS) in the years 1991to 2001 at the Surgical Clinic 1, University of Leipzig, Germany,

8 http://dtp.nci.nih.gov/

www.aacrjournals.org

Researcon Februarycancerres.aacrjournals.org Downloaded from

and at the Institute of Pathology of the Martin-Luther-University Halle, Germany, were included in the study (Sup-plementary Table S4). The mean observation time was40.2 mo (range 2–198 mo). Sixty-three patients died fromtumor-related causes within the observation time, whereas66 patients were still alive at the time of follow-up. Allpatients underwent surgical treatment, and the subsequentmean survival time was 25.1 mo (range 2–119 mo). Eighty-three patients received postoperative radiotherapy and/orchemotherapy; 76 were treated with fractionated radiation(cumulative dose 60.4 Gy) and 26 received a combination treat-ment of doxorubicin and ifosfamide. The DNA of 66 patientswas extracted from whole blood samples (n = 22) or from nor-mal, pathologically confirmed tumor-free tissue, adjacent tothe resection specimen (n = 44). The DNA of the remaining63 patients was obtained from tissue within the confines ofthe tumor. A control cohort consisting of 498 blood donors(Germans of central European origin; 194 females and 304males, ages 19–68 y; mean 44.0 y) from whom samples wereobtained at the German Red Cross Blood Transfusion ServiceNSTOB (Springe, Germany) was included in the study. Eachperson gave written and informed consent. Approval fromthe local ethics committee was obtained.Sequence analysis. Genomic DNA was extracted using

the Innuprep Blood DNA mini-kit and the Innuprep DNAmini-kit (AJ Innuscreen GmbH). The CD44 SNP (rs187115)and the YWHAQ SNP (rs6734469) genotypes were deter-mined by PCR amplification and subsequent allelic dis-crimination using the C_779820_10 and the C_29724290_10genotyping assays. In brief, the allelic discrimination was per-formed as a multiplexed reaction in a 96-well format, usingtwo primer/probe pairs in each reaction with a unique pairof fluorescent dye detectors to allow genotyping of the twopossible variants at the SNP site in a target template se-quence. Each well contained 20 ng DNA, 12.5 μL of TaqManUniversal PCR Master Mix (Applied Biosystems), 0.625 μL ofthe genotyping assay, and 9.875 μL distilled water. FollowingPCR, fluorescence was measured with the AB 7500 Real-TimePCR System (Applied Biosystems) and the genotype clusters

Table 1. Allelic differences for candidate SNP loci in growth responses to chemotherapeutics of NCI60cells either wild-type or mutant for p53

Gene
SNP GI50 p53 mutant
Univariate

h. 28, 2019. © 2010 Amer

Multivariate

P
No. ofsignificant drugs
Fold ratio
Significantgenotypes
Cancer Res; 7

ican Associatio

P

YWHAQ
rs6734469 G<A 1.0E+00 3 1.8 AG, AA 1.3E−02, 2.4E−02 PPP2R2B rs319227 A<C 8.9E−05 32 3.1 AC, CC 2.1E−02, 2.4E−02 PPP2R2B rs319217 T<C 8.7E−04 33 3 CT, CC 2.1E−02, 2.4E−02 CCNG1 rs2069347 C<T 2.2E−03 28 2.5 CC, TT 1.0E−02, 1.1E−02 CD44 rs187115 T<C 3.9E−19 62 3.5 CC, TT 1.3E−03, 3.3E−03 PIAS1 rs1027154 G<C 1.0E+00 5 4.3 GG, CG 2.6E−03, 5.0E−03 KDR rs2168945 G<T 4.8E−04 30 2.6 GG, TT 2.4E−03, 6.1E−03 CSE1L rs2426127 T<C 1.0E+00 6 2 — —
0(1) January 1, 2010 173

n for Cancer


Gene SNP GI50 p53 wild-type

Univariate Multivariate

P No. ofsignificant drugs

Fold ratio Significantgenotypes

P

YWHAQ rs6734469 G<A 5.6E−47 108 5.4 GG, AA 4.0E−05, 1.3E−03PPP2R2B rs319227 A<C 1.3E−29 91 4.1 AA, CC 1.8E−02, 2.1E−02PPP2R2B rs319217 T<C 1.3E−29 91 4.1 TT, CC 1.8E−02, 2.1E−02CCNG1 rs2069347 C<T 3.4E−26 87 5.7 TT, CT 2.0E−03, 2.4E−02CD44 rs187115 T<C 2.1E−10 63 3.7 TT, CC 1.4E−03, 7.2E−03PIAS1 rs1027154 G<C 1.7E−08 59 7.7 CG, GG 0.0E+00, 4.0E−05KDR rs2168945 G<T 1.2E−07 57 3.4 GG, GT 7.0E−04, 1.0E−02CSE1L rs2426127 T<C 1.0E+00 6 1.9 — —

Table 1. Allelic differences for candidate SNP loci in growth responses to chemotherapeutics of NCI60cells either wild-type or mutant for p53 (Cont'd)

Vazquez et al.

174


were scored using Sequence Detection Software (version 1.3;Applied Biosystems). Both SNPs were genotyped successfullyin all controls, but CD44 genotypes for one STS patient couldnot be determined due to irrecoverable DNA loss. Tumor tis-sue was available for 92 patients included in this study andthe somatic p53 mutational status of these tumor tissues wasdetermined by direct sequencing of exons 4 to 10 of the p53gene as previously described (16).Statistical analysis. The survival analysis was performed

using the Cox multivariate proportional hazards regressionmodel with the SPSS 16.0 software (SPSS, Inc.). A permu-tation test was performed to determine the statistical sig-nificance of the noted increase of the mean age of tumordiagnosis, whereby A/A < A/B < B/B. P values <0.05 were con-sidered significant.

Results

Allelic differences in cellular drug responses. The geno-types for more than 100,000 SNPs are publicly available forthe NCI60 cell line panel and were determined using theAffymetrix 125K genotyping platform.9 Two hundred sixty-four genotyped SNPs were found to reside in the 142 genesknown to be important in the p53 stress response (Supple-mentary Table S1). We set out to explore possible allelic dif-ferences of SNPs residing in these genes in their growthresponses to 132 standard chemotherapeutic agents. To dothis, three statistical tests were performed (SupplementaryFig. S1A). First, a univariate test was undertaken for eachdrug and SNP pair and allelic differences were sought. Spe-cifically, the average log GI50 [X = −log10(GI50)] for cells foreach of the three genotypes of a given locus (AA, Aa, andaa) were calculated for cells either wild-type or mutant forp53. Subsequently, the probability (P value) was computedthat just by chance the difference for the following groupings

9 http://dtp.nci.nih.gov/

Cancer Res; 70(1) January 1, 2010


either was equal to or larger than the actual measurement:(a) Xa-XAA or (b) Xaa-XA, or (c) XAA-Xa, or (d) XA-Xaa, or (e)[Xaa-XaA and XaA-XAA], and (f) [XAA-XaA and XaA-Xaa]. Theseprobabilities were estimated using a permutation test (106

permutations) that preserved the allele or genotype group sizesbut permuted the samples among the groups. Results P < 0.05were considered significant. Four different combinatorial hitswere defined for each SNP-drug pair: 1, significant in all cells(SA); 2, significant in cells with a wild-type p53 gene (SWT); 3,significant in cells with mutant p53 (SMT); and 4, significant incells with wild-type p53 but not significant or significant in theopposite allele orientation in cells with a mutant p53 (SWT-NRMT; Supplementary Fig. S1B).Second, a multiple hypothesis test was performed for the

noted observations. This test took advantage of the fact that132 well-characterized compounds were tested against theNCI60 cell panel, which provided a set of independent mea-surements. The test incorporated all 264 p53 stress responseSNPs and 132 agents and used a Fisher's exact test to com-pute the statistical significance of observing h univariate hitsfor a SNP (assuming one of the possibilities 1–4 listed above)on a total of D drugs, given that overall H significant hits areobserved after testing S reference SNPs on the D drugs. Wechose all 109,687 genotyped SNPs as a reference set.The third statistical test served to help eliminate the pos-

sibility that a given allelic difference in the growth responseto an agent could be due to the tumor type of the cells, othersomatic mutations or other SNPs, or a combination of thesevariables. To do this, a two-step multivariate test was per-formed. Using a log-linear regression model, the contributionof each attribute [tissue of origin, somatic mutational statusof 19 cancer-related genes (APC, BRAF, BRCA1, BRCA2, CDH1,CDKN2A, CTNNB1, EGFR, ERBB2, FLT3, HRAS, KRAS, NRAS,MADH4, PDGFRA, PIK3CA, PTEN, RB1, SKT11, VHL), and theSNP genotypes for the 264 p53 stress response SNPs] to thelog GI50 over the set of 132 drugs was computed. In the sec-ond step, identical computations were performed on a sec-ond set of 264 SNPs that were chosen randomly from the

Cancer Research





109,687 genotyped SNPs. These calculations served as a ref-erence set that allowed for the estimation of the statisticalsignificance of the effects observed for a given SNP withinthe p53 stress response, specifically by defining the probabil-ity (P value) that a SNP in the reference set has an equal orlarger contribution to the observed effect. The SNPs withinp53 stress response with at least two genotypes demonstrat-ing effects with P values lower than 5% in cells with wild-typep53 were deemed significant.Only 8 of the 264 SNPs showed significant allelic differ-

ences in cellular growth responses to the standard chemo-therapeutic agents according to the SWT-NRMT univariateanalysis, the associated multiple hypothesis test, and themultivariate test in the cells wild-type for p53 (Supplement-ary Table S2). To complete and validate the genotyping of theeight candidate SNPs, we obtained the genomic DNAs fromall 59 cell lines and regenotyped each SNP using accurateallelic discrimination assays (Applied Biosystems; Materialsand Methods). The publicly available genotypes derived fromthe Affymetrix 125K genotyping platform had, on average,nine missing genotypes and, compared with the regenotyp-ing, six missed calls (Supplementary Table S3). Therefore,we repeated the above-described genotype-response analysisusing the new genotypes. All but one (rs2426127 in CSEIL) ofthe eight SNPs showed significant allelic differences in cellu-lar growth responses to the standard chemotherapeuticagents in the cells that were wild-type for p53 (Table 1). Spe-cifically, on average, the different genotypes of the seven sig-nificant SNPs associate with a different growth response for79 (range 57–108; Table 1) of the 132 agents tested, wherebycells with the homozygote genotypes differed on average 4.9-fold (range 3.4- to 5.7-fold; Table 2) in their respective drugsensitivities in cells wild-type for p53.The strongest effects in p53 wild-type cells were seen for

the SNP in the YWHAQ (14-3-3τ) gene (rs6734469; Table 1).Specifically, the results of the univariate analysis suggest thatcells with the different genotypes of the YWHAQ SNP sig-nificantly differ in their growth response to 108 of the132 agents tested in wild-type p53 cells (P = 5.6 × 10−47;Table 1), but only 3 of the 132 agents in mutant p53 cells.



Significant allelic differences were observed with 35 alkylat-ing agents, 20 and 15 topoisomerase I and II inhibitors, 14and 15 RNA/DNA and DNA antimetabolites, and, to a lesserextent, with 10 antimitotic agents (Fig. 1A and B). Wild-typep53 cells with the homozygote genotypes for YWHAQ SNPdiffered on average 5.4-fold in their respective drug sensitiv-ities (A/A-GI50 > G/G-GI50). For example, one of the topoi-somerase II inhibitors for which the different genotypes ofthe YWHAQ SNP significantly differed was doxorubicin. Cellswild-type for p53 and A/A in genotype required 7-fold moredoxorubicin to arrest growth compared with p53 wild-typecells G/G in genotype. Specifically, as seen in Fig. 1C, the17 cell lines wild-type for p53 required, on average, 0.13μmol/L of doxorubicin to arrest by 50%, and cells rangedfrom 0.01 to 0.64 μmol/L in their respective GI50 values.Those five cell lines that required the highest concentration(average 0.33 μmol/L) were significantly enriched in cellscontaining the A allele compared with those five cell linesneeding the least amount of agent (average 0.03 μmol/L,P = 0.0002). By contrast, no significant differences in the A-allele frequencies were seen in the extreme groups of the p53mutant cell lines (Fig. 1C). Importantly, the allelic differencesremain significant after performing a multivariate analysis,including the tissue of origin and the mutational status ofthe other cancer-related genes for those cells with a wild-typep53 gene (PGG = 0.00004, PAA = 0.0013; Table 1).Interestingly, not all of the regenotyped SNPs retained

the strong dependence of the wild-type p53 gene for thesignificant allelic differences in cellular growth responsesto the standard chemotherapeutic agents (Table 1). For ex-ample, the allelic differences for the SNP in CD44, and to alesser extent for the SNP in KDR, were seen with a similarnumber of agents and with similar strengths in both p53wild-type and p53 mutant cells (Table 1). These data sug-gest that these allelic differences are not dependent on themutational status of the p53 gene and that these two popu-lations of cells can be grouped together in further analysesof these SNPs.The allelic differences in cellular growth responses to the

standard chemotherapeutic agents according to the SA

Table 2. Allelic differences for candidate SNP loci in growth responses to chemotherapeutics of NCI60cell panel

Gene
SNP GI50 Univariate
h. 28, 2019. © 201

Multivariate

P
No. ofsignificant drugs
Fold ratio
Significantgenotypes
Canc

0 American A

P

CD44
rs187115 T<C 8.1E−24 96 2.8 CC, TT 0.0E+00, 0.0E+00 PPP2R2B rs319217 T<C 1.6E−05 63 2.4 CC, CT 5.4E−03, 2.2E−02 PPP2R2B rs319227 A<C 1.6E−05 63 2.4 CC, AC 5.4E−03, 2.2E−02 YWHAQ rs6734469 G<A 1.4E−04 60 2.1 AA, AG 3.4E−04, 2.2E−02 KDR rs2168945 G<T 2.8E−04 59 2.2 GG, TT, GT 0.0E+00, 2.1E−03, 7.6E−03 CCNG1 rs2069347 C<T 1.4E−02 52 2.7 TT, CC 2.1E−03, 1.5E−02 PIAS1 rs1027154 C<G 9.9E−01 29 3 GG, CG 4.5E−04, 2.2E−02 CSE1L rs2426127 C<T l.0E+00 27 2.3 CC 1.50E−002
er Res; 70(1) January 1, 2010 175

ssociation for Cancer


Vazquez et al.

176


univariate analysis, the associated multiple hypothesis test,and the multivariate test in all 59 cells regardless of the mu-tational status of the p53 gene are depicted in Table 2. In-deed, the strongest effects were seen for the SNP in theCD44 gene (rs187115). Specifically, the results of the univar-iate analysis suggest that cells with the different genotypes ofthe CD44 SNP significantly differ in their growth response to96 of the 132 agents tested (P = 8.1 × 10−24; Table 2 andFig. 2A). Significant allelic differences were observed with21 alkylating agents, 24 and 14 topoisomerase I and II inhi-



bitors, 17 and 8 RNA/DNA and DNA antimetabolites, andwith 20 antimitotic agents (Fig. 2A and B). Cells with thehomozygote genotypes for CD44 SNP differed on average2.8-fold in their respective drug sensitivities (C/C-GI50 >T/T-GI50). For example, one of the topoisomerase II inhibi-tors for which the different genotypes of the CD44 SNPsignificantly differed in their growth response was also doxo-rubicin. Cells C/C in genotype required 10-fold more of thedoxorubicin to arrest growth compared with cells T/T in ge-notype. Specifically, as seen in Fig. 2C, those 10 cell lines that

Figure 1. The YWHAQ SNP (rs6734469) associates with significant allelic differences in cellular growth responses to chemotherapeutic agents in p53wild-type cells. A, a heat map summarizing the genotype-drug response associations for a panel of 132 chemotherapeutics in the NCI60 cell lines withwild-type p53. The color intensity is proportional to the difference between the mean log10 GI50 over cells with G/G and A/A homozygous genotypes,with red corresponding to lower GI50 (higher −log10 GI50) in G/G than A/A containing cells and otherwise green. The genotypes of the YWHAQ SNPsignificantly differ in their growth response to 108 of the 132 agents tested. B, a pie chart depicting the different classes of agents that the 108 drugs belongto based on their known mechanisms of action. C, a graphical representation of the GI50 distributions of doxorubicin (a topoisomerase II inhibitor)across the NCI60 cell lines with those wild-type p53 (left) and those with mutant p53 (right). The five cell lines that required the highest concentration(average 0.33 μmol/L) were significantly enriched in cells containing the A allele compared with the five cell lines needing the least amount of agent (average0.03 μmol/L, P = 0.0002). In contrast, no significant differences in the A-allele frequencies were seen in the extreme groups of the p53 mutant cell lines.

Cancer Research





required the highest concentration (average 1.9 μmol/L)were significantly enriched in cells containing the C allelecompared with those 10 cell lines needing the least amountof agent (average 0.3 μmol/L, P = 0.0134). Importantly, for all59 cell lines, the allelic differences remain significant afterperforming a multivariate analysis, including the tissue of or-igin and the mutational status of the other cancer-relatedgenes, and the 264 p53 stress response SNP genotypes(PCC = 0.0000 and PTT = 0.0003; Table 2).Allelic differences in STS survival. The above-described

allelic differences in cellular growth responses to chemother-apeutic agents for these SNPs suggest that individuals carry-



ing the alleles that were associated with weaker growthresponses to chemotherapeutics in cell culture could associ-ate with poorer outcomes upon cancer onset due to a poorerresponse to therapies. To begin to test this, we analyzedthese loci in 129 patients who were diagnosed with and trea-ted for STS (Supplementary Table S4), a well-described, p53-surveilled tumor (17, 18).Tumor tissue was available for 92 patients included in this

study. The somatic p53 mutational status of these tumor tis-sues was determined by direct sequencing of exons 4 to 10 ofthe p53 gene, as previously described (16). Eighteen p53mutations were identified from the 92 tumor DNAs, resulting

Figure 2. The CD44 SNP (rs187115) associates with significant allelic differences in cellular growth responses to chemotherapeutic agents in cells. A, a heatmap summarizing the genotype-drug response associations for a panel of 132 chemotherapeutics with known mechanism of action in the NCI60 celllines. The color intensity is proportional to the difference between the mean log10 GI50 over cells with T/T and C/C homozygous genotypes, with redcorresponding to lower GI50 (higher −log10 GI50) in T/T than C/C containing cells and otherwise green. The genotypes of the CD44 SNP significantly differin their growth response to 96 of the 132 agents tested. B, a pie chart depicting the different classes of agents that the nine drugs belong to based ontheir known mechanisms of action. C, a graphical representation of the GI50 distributions of doxorubicin across the NCI60 cell lines. The 10 cell linesthat required the highest concentration (average 1.9 μmol/L) were significantly enriched in cells containing the C allele compared with the 10 cell linesneeding the least amount of agent (average 0.3 μmol/L, P = 0.0134).

Cancer Res; 70(1) January 1, 2010 177



Vazquez et al.

178


in a p53 mutation frequency of 19.6% in this cohort, whichis extremely similar to previously reported frequencies inSTS (19.7%; International Agency for Research on Cancerdatabase;10 ref. 19). The genotype frequencies in the 129STS patients of the CD44 SNP were 11.7% for the C/C geno-type, 45.3% for the T/C genotype, and 43.0% for patients T/Tin genotype (Supplementary Table S5). The genotype distri-butions of the YWHAQ SNP were 27.1%, 47.3%, and 25.6% forthe A/A, A/G, and G/G genotypes, respectively (Supplemen-tary Table S5). The allelic frequencies of both SNPs did notdiffer significantly from the frequencies of the controls of thesame ethnic origin nor did the frequencies differ between thepatient's DNAs derived from nontumor or tumor tissues(Supplementary Table S5).To assess the impact of both SNPs on patient outcomes,

Cox multivariate regression analysis was performed, adjustingfor known prognostic factors of STS: tumor stage, resectiontype (R-status), and gender (20, 21). Interestingly, the sarcomapatients harboring the alleles of the YWHAQ and CD44 SNPs,which in the NCI60 analysis were associated with weakergrowth responses to chemotherapeutics, associated withpoorer overall survival. Those patients homozygous for theC allele of the CD44 SNP were associated with a 2.16-fold in-creased risk for tumor-related death compared with indivi-duals C/T and T/T in genotype (P = 0.041; SupplementaryTable S7). Furthermore, patients carrying either one or twocopies of the A allele of the YWHAQ SNP associated witha significantly worse prognosis compared with patientshomozygous for the G allele (P = 0.043; relative risk, 1.92;Supplementary Table S7). As predicted by the results of theNCI60 analysis, the allelic differences in overall survivalbecame notably stronger when the survival analysis was re-stricted to those 83 patients who received radiotherapy/che-motherapy. Patients C/C for the CD44 SNP associated with a2.89-fold increased risk for tumor-related death comparedwith patients C/T and T/T for CD44 who also receivedDNA-damaging therapies (P = 0.011; Supplementary TableS7 and Fig. 3A). Patients harboring either one or two copiesof the A allele of the YWHAQ SNP treated with DNA-damagingtherapies associated with a 2.77-fold increased risk of tumor-related death compared with patients who are similarly treatedbut with a G/G genotype (P = 0.011; Supplementary Table S7and Fig. 3B).Allelic differences in STS incidence. The above-described

results support a model that these SNPs affect cellular stressresponses, thereby resulting in altered survival rates. As cel-lular stress responses like those mediated by the p53 path-way are also important in tumor suppression, we exploredthe effects of these loci on the age-dependent incidence ofthese tumors. Interestingly, the alleles of both SNPs, whichin the NCI60 analysis associated with weaker growth re-sponses to chemotherapeutics and, in the survival analysis,associated with decreased survival rates, also associated witha significantly earlier age of tumor onset. Patients with a C/C

10 http://www-p53.iarc.fr



genotype for the CD44 SNP were diagnosed on average at49.1 years of age (range 16–83 years), patients C/T in geno-type at 54.5 years (range 14–84 years), and patients T/T ingenotype at 59.8 years of age (range 24–87 years; Table 3;P = 0.002). Thus, the C/C homozygotes showed on averagea 10.7-year earlier tumor onset than individuals homozygousfor the T allele (Table 3). Similar associations were observedfor the YWHAQ SNP, whereby patients homozygous for the Aallele were diagnosed with a mean age of 52.8 years (range17–85 years), patients A/G in genotype at 55.5 years (range14–85 years), and patients G/G in genotype at 59.9 years(range 22–87 years; Table 3; P = 0.006). Thus, patients withthe A/A genotype showed on average a 7.1-year earlier ageof onset than patients G/G in genotype (Table 3).

Figure 3. The alleles of the CD44 SNP and the YWHAQ SNP, whichassociate with a weaker apoptotic response to chemotherapeutictreatment in the NCI60 cell lines, also correlate significantly with pooreroverall survival in STS patients in a Cox multivariate regression analysisadjusting for the known prognostic factors of STS. The graphs displaythe survival curves of radiotherapeutically or chemotherapeuticallytreated patients for the CD44 SNP genotypes (A) and for the YWHAQSNP (B) plotted against the survival time in months. Patients C/C for theCD44 SNP associated with a 2.89-fold increased risk for tumor-relateddeath compared with patients C/T and T/T (P = 0.011). Patients harboringeither one or two copies of the A allele of the YWHAQ SNP treated withDNA-damaging therapies associated with a 2.77-fold increased risk oftumor-related death compared with patients with a G/G genotype (P =0.011). All P values were calculated with the Cox multivariate regressionanalysis adjusted to tumor stage, resection type (R-status), and gender.

Cancer Research



11 http://www.hapmap.org/



Discussion

In this report, we build upon, expand, and use a method-ology to uncover functional variants in p53 stress responsegenes with data from the NCI60 human tumor cell line an-ticancer drug screen. We scanned 142 genes known to af-fect the p53 stress response and identified seven SNPs in sixgenes that possess allelic differences in cellular growth re-sponses to standard chemotherapeutic agents. The greatestdifferences in the p53 wild-type cells were observed for theYWHAQ SNP (rs6734469, P = 5.6 × 10−47) and the strongesteffects in all 59 cell lines, regardless of p53 mutational sta-tus, were seen for the CD44 SNP (rs187115, P = 8.1 × 10−24).The CD44 gene encodes a transmembrane glycoprotein in-volved in a vast range of cellular processes, such as regula-tion of growth and survival, differentiation, and motility (22,23). CD44 harbors tumor-promoting activities that includestimulating anchorage-independent cell growth and pro-mote metastasis (24–26). Furthermore, its expression ischaracteristic for breast and prostate cancer stem cells(23, 27–29). Aberrated CD44 expression has also been re-ported to associate with tumor initiation and progressionin various malignancies, such as colorectal, mammary,and prostate carcinomas, as well as neuroblastomas andsarcomas (25, 26, 30–32). Certainly, the precise effects of al-tered CD44 expression on the initiation of human cancerand patients' prognosis need to be further elucidated, butthe reported studies and our data suggest a functionallyand clinically important role of CD44 and its polymorphicvariants in human malignancies.The YWHAQ gene encodes 14-3-3τ, a member of the 14-3-

3-protein family. These proteins interact with many signalingpathways that are critical for apoptosis and cell proliferation,mainly through binding phosphoserine/threonine motifs(33). 14-3-3 binding induces conformational changes in thetarget protein that can alter its stability, catalytic activity, cel-lular localization, or susceptibility to intracellular proteases,kinases, and phosphatases (33, 34). The 14-3-3 proteins arehighly conserved and consist of seven family members inmammals: the β, γ, ε, σ, ζ, τ, and η isoforms (34). Interest-ingly, only YWHAQ (14-3-3τ) has been shown to promote ap-optosis directly upon genotoxic stress (35–37). Furthermore,



14-3-3τ has also been shown to bind to wild-type p53 afterionizing radiation of cells and to stimulate transcriptional ac-tivity of p53 (35, 37). This observation is consistent with theobserved dependence of wild-type p53 for the allelic differ-ences in the cellular growth responses to chemotherapeuticsfor the YWHAQ SNP in our study. Importantly, the A allele,which associates with a significantly weaker drug responsein the NCI60 panel, also associates with an early onset ofdisease and an increased risk of tumor-related death inSTS patients, specifically in patients who received radioche-motherapeutic treatment (Fig. 3).The observations reported here remain to be validated in

other patient cohorts and, importantly, the regulatorychanges associated with YWHAQ and CD44 SNPs remainto be determined. The identified YWHAQ SNP (rs6734469)is located in the second intron of the YWHAQ gene andthe CD44 SNP (rs187115) in the first intron of CD44. OtherSNPs in phase II of the HapMap project are closely linked(r2 > 0.8)11 to these two SNPs in the Caucasian populationof northwestern European ancestry (CEU; SupplementaryTable S6). Specifically, for the YWHAQ SNP, 10 SNPs in in-trons 2 and 3 are closely linked and, for the CD44 SNP,3 SNPs in intron 1 are linked (Supplementary Table S6). Toour knowledge, there have been no published attempts tocharacterize the regulatory regions of YWHAQ and no regu-latory role of intron 1 of CD44 has been proposed. Hence, adiscussion of whether one of these SNPs resides in a poten-tial regulatory region would be premature. In addition, moreexhaustive searches for other genetic variants closely linkedto these SNPs, but not included in the HapMap project,will be necessary to develop a comprehensive list of candi-date functional SNPs that warrant further experimental in-vestigation into the molecular and cellular mechanismsunderlying the significant allelic differences in cellular drugresponses, sarcoma incidence, and survival reported in thiswork. However, these data strongly support the model thatboth CD44 and 14-3-3τ play a significant regulatory role incellular stress responses, thereby affecting sarcoma incidenceand survival.

Table 3. Age of diagnosis analysis of CD44 and YWHAQ SNPs

SNP
Genotype n Age of diagnosis*
h. 28

P†

, 2019. ©

Average difference between homozygotes (y)

CD44 SNP
C/C 15 49.1 (53) 0.002 10.7 C/T 58 54.5 (59) T/T 55 59.8 (62)
YWHAQ (14-3-3τ) SNP
G/G 33 59.9 (59) 0.006 7.1 A/G 61 55.5 (61) A/A 35 52.8 (54)
*Average (median) in years.†Permutation test for the differences between the average age of diagnosis for each genotype.

Cancer Res; 70(1) January 1, 2010 179

2010 American Association for Cancer


Vazquez et al.

180


Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Suzanne Christen, Claire Beveridge, and MarkShipman for their help in preparation of the manuscript.



Grant Support

Ludwig Institute for Cancer Research, Simons Foundation,Helen and Martin Choolijian, and Leon Levy Foundation.The costs of publication of this article were defrayed in

part by the payment of page charges. This article musttherefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.Received 6/17/09; revised 9/21/09; accepted 10/7/09; published

OnlineFirst 12/8/09.

References
1. Levine AJ. p53, the cellular gatekeeper for growth and division. Cell
1997;88:323–31.2. Riley T, Sontag E, Chen P, Levine A. Transcriptional control of human

p53-regulated genes. Nat Rev Mol Cell Biol 2008;9:402–12.3. Bond GL, Levine AJ. A single nucleotide polymorphism in the p53

pathway interacts with gender, environmental stresses and tumorgenetics to influence cancer in humans. Oncogene 2007;26:1317–23.

4. Vazquez A, Bond EE, Levine AJ, Bond GL. The genetics of the p53pathway, apoptosis and cancer therapy. Nat Rev Drug Discov 2008;7:979–87.

5. Harris N, Brill E, Shohat O, et al. Molecular basis for heterogeneity ofthe human p53 protein. Mol Cell Biol 1986;6:4650–6.

6. Matlashewski GJ, Tuck S, Pim D, Lamb P, Schneider J, Crawford LV.Primary structure polymorphism at amino acid residue 72 of humanp53. Mol Cell Biol 1987;7:961–3.

7. Bond GL, Hu W, Bond EE, et al. A single nucleotide polymorphism inthe MDM2 promoter attenuates the p53 tumor suppressor pathwayand accelerates tumor formation in humans. Cell 2004;119:591–602.

8. Sakamuro D, Sabbatini P, White E, Prendergast GC. The polyprolineregion of p53 is required to activate apoptosis but not growth arrest.Oncogene 1997;15:887–98.

9. Arva NC, Gopen TR, Talbott KE, et al. A chromatin-associated andtranscriptionally inactive p53-2 complex occurs in mdm2 SNP309homozygous cells. J Biol Chem 2005;280:26776–87.

10. Nayak MS, Yang JM, Hait WN. Effect of a single nucleotide polymor-phism in the murine double minute 2 promoter (SNP309) on the sensi-tivity to topoisomerase II-targeting drugs.CancerRes 2007;67:5831–9.

11. Whibley C, Pharoah PD, Hollstein M. p53 polymorphisms: cancerimplications. Nat Rev Cancer 2009;9:95–107.

12. Shoemaker RH. The NCI60 human tumour cell line anticancer drugscreen. Nat Rev Cancer 2006;6:813–23.

13. Garraway LA, Widlund HR, Rubin MA, et al. Integrative genomic anal-yses identify MITF as a lineage survival oncogene amplified in malig-nant melanoma. Nature 2005;436:117–22.

14. Ikediobi ON, Davies H, Bignell G, et al. Mutation analysis of 24 knowncancer genes in the NCI-60 cell line set. Mol Cancer Ther 2006;5:2606–12.

15. O'Connor PM, Jackman J, Bae I, et al. Characterization of the p53tumor suppressor pathway in cell lines of the National Cancer Instituteanticancer drug screen and correlations with the growth-inhibitory po-tency of 123 anticancer agents. Cancer Res 1997;57:4285–300.

16. Kern A, Taubert H, Scheele J, et al. Association of p53 mutations,microvessel density and neoangiogenesis in pairs of colorectal can-cers and corresponding liver metastases. Int J Oncol 2002;21:243–9.

17. Donehower LA, Harvey M, Slagle BL, et al. Mice deficient for p53 aredevelopmentally normal but susceptible to spontaneous tumours.Nature 1992;356:215–21.

18. Li FP, Strong LC, Fraumeni JF, Jr., et al. Germ line p53 mutations in afamilial syndrome of breast cancer, sarcomas, and other neoplasms.Sciences (New York) 1990;250:1233–8.

19. Petitjean A, Mathe E, Kato S, et al. Impact of mutant p53 functionalproperties on TP53 mutation patterns and tumor phenotype: lessonsfrom recent developments in the IARC TP53 database. Hum Mutat2007;28:622–9.

20. Lahat G, Tuvin D, Wei C, et al. New perspectives for staging andprognosis in soft tissue sarcoma. Ann Surg Oncol 2008;15:2739–48.

21. Molife R, Lorigan P, MacNeil S. Gender and survival in malignant tu-mours. Cancer Treat Rev 2001;27:201–9.

22. Naor D, Wallach-Dayan SB, Zahalka MA, Sionov RV. Involvement ofCD44, a molecule with a thousand faces, in cancer dissemination.Semin Cancer Biol 2008;18:260–7.

23. Ponta H, Sherman L, Herrlich PA. CD44: from adhesion molecules tosignalling regulators. Nat Rev Mol Cell Biol 2003;4:33–45.

24. Barbour AP, Reeder JA, Walsh MD, Fawcett J, Antalis TM, GotleyDC. Expression of the CD44v2-10 isoform confers a metastatic phe-notype: importance of the heparan sulfate attachment site CD44v3.Cancer Res 2003;63:887–92.

25. Weber GF, Bronson RT, Ilagan J, Cantor H, Schmits R, Mak TW. Ab-sence of the CD44 gene prevents sarcoma metastasis. Cancer Res2002;62:2281–6.

26. Yu Q, Toole BP, Stamenkovic I. Induction of apoptosis of metastaticmammary carcinoma cells in vivo by disruption of tumor cell surfaceCD44 function. J Exp Med 1997;186:1985–96.

27. Dou J, Pan M, Wen P, et al. Isolation and identification of cancerstem-like cells from murine melanoma cell lines. Cell Mol Immunol2007;4:467–72.

28. Hurt EM, Kawasaki BT, Klarmann GJ, Thomas SB, Farrar WL. CD44+CD24(−) prostate cells are early cancer progenitor/stem cells thatprovide a model for patients with poor prognosis. Br J Cancer2008;98:756–65.

29. Wright MH, Calcagno AM, Salcido CD, Carlson MD, Ambudkar SV,Varticovski L. Brca1 breast tumors contain distinct CD44+/CD24−and CD133+ cells with cancer stem cell characteristics. BreastCancer Res 2008;10:R10.

30. Wielenga VJ, Heider KH, Offerhaus GJ, et al. Expression of CD44variant proteins in human colorectal cancer is related to tumor pro-gression. Cancer Res 1993;53:4754–6.

31. De Marzo AM, Bradshaw C, Sauvageot J, Epstein JI, Miller GJ.CD44 and CD44v6 downregulation in clinical prostatic carcinoma:relation to Gleason grade and cytoarchitecture. Prostate 1998;34:162–8.

32. Shtivelman E, Bishop JM. Expression of CD44 is repressed in neu-roblastoma cells. Mol Cell Biol 1991;11:5446–53.

33. Morrison DK. The 14-3-3 proteins: integrators of diverse signalingcues that impact cell fate and cancer development. Trends Cell Biol2009;19:16–23.

34. Tzivion G, Gupta VS, Kaplun L, Balan V. 14-3-3 proteins as potentialoncogenes. Semin Cancer Biol 2006;16:203–13.

35. Stavridi ES, Chehab NH, Malikzay A, Halazonetis TD. Substitutionsthat compromise the ionizing radiation-induced association of p53with 14-3-3 proteins also compromise the ability of p53 to induce cellcycle arrest. Cancer Res 2001;61:7030–3.

36. Wang B, Liu K, Lin FT, Lin WC. A role for 14-3-3τ in E2F1 stabilizationand DNA damage-induced apoptosis. J Biol Chem 2004;279:54140–52.

37. Waterman MJ, Stavridi ES, Waterman JL, Halazonetis TD. ATM-dependent activation of p53 involves dephosphorylation and associ-ation with 14-3-3 proteins. Nat Genet 1998;19:175–8.

Cancer Research



2010;70:172-180. Published OnlineFirst December 8, 2009.Cancer Res Alexei Vazquez, Lukasz F. Grochola, Elisabeth E. Bond, et al. Incidence and Survival

and CD44 That Affect SarcomaτNetwork Genes 14-3-3Chemosensitivity Profiles Identify Polymorphisms in the p53

Updated version

10.1158/0008-5472.CAN-09-2218doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2009/12/07/0008-5472.CAN-09-2218.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/70/1/172.full#ref-list-1

This article cites 37 articles, 13 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/70/1/172.full#related-urls

This article has been cited by 2 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] 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/70/1/172To request permission to re-use all or part of this article, use this link

Research. on February 28, 2019. © 2010 American Association for Cancercancerres.aacrjournals.org Downloaded from


http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-09-2218

http://cancerres.aacrjournals.org/content/suppl/2009/12/07/0008-5472.CAN-09-2218.DC1

http://cancerres.aacrjournals.org/content/70/1/172.full#ref-list-1

http://cancerres.aacrjournals.org/content/70/1/172.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/70/1/172


Documents

Cancer Chemosensitivity Profiles Identify Polymorphisms in ...cancerres.aacrjournals.org/content/canres/70/1/172.full.pdf · Chemosensitivity Profiles Identify Polymorphisms in the