Genetic Alterations Associated with Acquired Temozolomide Resistance in SNB-19, a Human Glioma Cell Line

2182

Genetic alterations associated with acquired temozolomide resistance in SNB-19, a human glioma cell line

Nathalie Auger,1,3 Joe¨lle Thillet,4 parental cells and both temozolomide-resistant variants. Krystell Wanherdrick,4 Ahmed Idbaih,2 None are implicated in known resistance mechanisms, Marie-Emmanuelle Legrier,1 Bernard Dutrillaux,5 such as DNA repair, whereas interestingly, several genes Marc Sanson,4 and Marie-France Poupon1 involved in differentiation were down-regulated. The data suggest that the acquisition of resistance to temozolomide 1INSERM U612 and 2INSERM U509, Research Section, in this model resulted from the selection of less differen- Institut Curie, Paris, France; 3Department of Clinical Biology, tiated preexistent resistant cells in the parental tumor. 4 Institut Gustave Roussy, Villejuif, France; INSERM U711, Hoˆpital [Mol Cancer Ther 2006;5(9):2182–92] de la Salpeˆtrie`re, Paris, France; and 5UMR 5202, Biodiversity, Origin, Structure, and Evolution, Museum d’Histoire Naturelle, Paris, France Introduction Abstract Gliomas are the most common brain tumor. They are usually treated by surgical resection combined with Gliomas are highly lethal neoplasms that cannot be cured irradiation and alkylating agent–based chemotherapy. by currently available therapies. Temozolomide is a However, their prognosis remains poor. The therapeutic recently introduced alkylating agent that has yielded a benefits obtained are transient and recurrence is the rule, significant benefit in the treatment of high-grade gliomas. together with resistance to chemotherapy. The limited However, either de novo or acquired chemoresistance efficacy of chemotherapy is generally attributed to two occurs frequently and has been attributed to increased factors: intrinsic or acquired chemoresistance and the levels of O6-methylguanine-DNA methyltransferase or to blood-brain barrier impeding the delivery of cytotoxic the loss of mismatch repair capacity. However, very few agents (1). gliomas overexpress O6-methylguanine-DNA methyltrans- Temozolomide is a recently introduced, oral, and ferase or are mismatch repair–deficient, suggesting that generally well-tolerated second-generation alkylating other mechanisms may be involved in the resistance to agent. Temozolomide has yielded a significant clinical temozolomide. The purpose of the present study was to benefit in high-grade gliomas (2). Temozolomide-induced generate temozolomide-resistant variants from a human O6-G-methylation is reversed by O6-methylguanine-DNA glioma cell line (SNB-19) and to use large-scale genomic methyltransferase (MGMT) in a reaction leading to irre- and transcriptional analyses to study the molecular basis versible inactivation of the protein. Increased levels of of acquired temozolomide resistance. Two independently MGMT or loss of the mismatch repair capacity confer obtained temozolomide-resistant variants exhibited no resistance to temozolomide (3). However, some tumors cross-resistance to other alkylating agents [1,3-bis(2- display resistance to DNA-methylating drugs indepen- chloroethyl)-1-nitrosourea and carboplatin] and shared dently of either their MGMT level or their mismatch repair genetic alterations, such as loss of a 2p region and loss status, suggesting that other major mechanisms are of amplification of chromosome 4 and 16q regions. The involved in the resistance to alkylating agents, including karyotypic alterations were compatible with clonal selec- the loss of a functional p53 which is able to induce cycle tion of preexistent resistant cells in the parental SNB-19 arrest and apoptosis (4), the p53 independent Chk1- cell line. Microarray analysis showed that 78 out of mediated G -M arrest which protects the cell from mitotic 17,000 genes were differentially expressed between 2 catastrophe (5), or the Akt pathway, frequently activated in glioblastoma, which bypass the temozolomide G2-M arrest on the one hand, and on the other hand, protects cells by suppressing senescence and mitotic catastrophe (6). Received 10/17/05; revised 5/30/06; accepted 6/29/06. Comparative genomic hybridization (CGH) analysis of Grant support: Ligue National contre le Cancer d’Ille et Vilaine. cell lines or primary tumors has revealed regions of The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked chromosomal imbalances associated with acquired chemo- advertisement in accordance with 18 U.S.C. Section 1734 solely to resistance (7–9). None of these studies, however, concerned indicate this fact. glioma cell lines and resistance to temozolomide. Requests for reprints: Marie-France Poupon, INSERM U612, Research In the present study, we used the temozolomide-sensitive Section, Institut Curie, 26 rue d’Ulm, 75248 Paris Cedex 05, France. Phone: 33-1-4234-6629. Fax: 33-1-4234-6619. SNB-19 cell line to generate temozolomide-resistant var- E-mail: [email protected] iants by culturing cells in the presence of incremental Copyright C 2006 American Association for Cancer Research. concentrations of temozolomide. The molecular alterations doi:10.1158/1535-7163.MCT-05-0428 associated with acquired temozolomide resistance were

Mol Cancer Ther 2006;5(9). September 2006

identified using cytogenetic molecular techniques, includ- the surviving adherent cells were fixed with methanol ing conventional cytogenetics, fluorescent in situ hybridi- and stained with methylene blue (1% in PBS buffer), and zation (FISH), CGH, and CGH arrays. Chromosomal cell-fixed dye was eluted with 0.1 N of HCl (12). Absorb- regions having gained or lost DNA during the acquisition ance (A) was measured in an automatic scanning photom- of resistance were disclosed. The additional use of micro- eter at a wavelength of 550 nm. Each experimental point array analysis allowed us to detect quantitative alterations was done in triplicate. The percentage of live cells was of mRNA transcripts associated or not with genetic calculated as follows: P =(A in treated cells / A in control changes. We also evaluated the ‘‘kinetics’’ of these changes cells) Â 100. Each assay was done at least thrice. Results during the acquisition of temozolomide resistance. are expressed as medians with the range. Statistical significance was determined using the Mann-Whitney U test (P V 0.05). Materials and Methods Karyotype Cell Culture,Treatment, and Isolation of Temozolo- Metaphases were harvested after a 2.5-hour colchicine mide-ResistantVariants block. Chromosome spreads were obtained according to The glioma-derived cell line, SNB-19 (10), was cultivated previously described techniques (13). Karyotypes were in DMEM (Sigma, St. Quentin Fallavier, France) supple- established after R-banding. mented with 10% FCS (Dutscher, Brumath, France), DNA Extraction penicillin G (102 IU/mL), streptomycin (50 Ag/mL; Sigma), DNA was extracted from frozen cells: samples were and L-glutamine (2 mmol/L; Sigma). Cells were maintained digested in extraction buffer (10 mmol/L Tris-Cl, 2 mmol/L j at 37 C in a humidified 5% CO2/95% air incubator, and EDTA, and 400 mmol/L NaCl) and 1.25% SDS and harvested for passage with 0.5 mg/mL trypsin and EDTA proteinase K (67 Ag/mL) at 37jC overnight. DNA was (0.2 mg/mL) when they had reached confluence. For subsequently treated with RNase (10 mg/mL) for 1 hour resistant cell selection, temozolomide (Interchim, Montluc¸on, at 37jC. NaCl (5 mol/L) was added before precipitation France) was dissolved in DMSO as a 17 mmol/L stock with ethanol. The concentration and molecular size of DNA solution, and diluted extemporaneously in complete were estimated using a spectrophotometer (Uvikon 923; medium at the desired concentration. At the first step of Fisher Bioblock Scientific, Illkirch, France) and ethidium selection with 3 Amol/L of temozolomide, 12 clones were bromide-stained agarose gels. isolated from among >50, and were cultured separately or CGH and Digital Image Analysis frozen. Two clones were treated with incremental concen- CGH experiments were done according to published trations of temozolomide (3, 5, 10, 20, 30, 60, and 150 Amol/L). protocols (14). DNA was labeled with fluorescein-12-dUTP At each step of selection, cells were exposed to a higher and normal reference DNA obtained from normal lym- temozolomide concentration as soon as regrowth was phocytes, with Texas red-5-dUTP (Vysis, Rungis, France) apparent. Briefly, cells were plated in 25 cm2 flasks in their using a commercially available nick-translation kit (Vysis). usual medium and allowed to attach overnight for drug Four hundred nanograms of labeled tumor and reference selection, then the medium was replaced by 0.5 mL of DNA were coprecipitated with Cot-I DNA, and then dena- temozolomide-containing fresh medium. After the first tured. Metaphase cell preparations obtained from normal 2 hours, temozolomide-free fresh medium was added. 1,3- lymphocytes were denatured and incubated with the Bis(2-chloroethyl)-1-nitrosourea (carmustine, Bristol-Myers probes in a humid chamber at 37jC for 3 days. Slides Squibb) was dissolved in ethanol at 17 mmol/L and were washed and mounted in antifade solution. Digital carboplatin (Merck, Lyon, France) was diluted in water at images of the three fluorochromes were acquired using 10 mg/mL. a computer-driven cooled CCD (Sensys, Photometrics, Cell Proliferation Assay Evry, France) camera mounted on a fluorescence micro- Cell proliferation was determined by using the 3-(4,5- scope (DMRB, Leica, Germany) equipped with selective dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide single bandpass filters and interfaced to the CGH QUIPS (MTT) test (Sigma) and the methyl blue test. The amount software package (Vysis). Chromosomes stained with 4¶,6- of tetrazolium dye or methyl blue was proportional to the diamidino-2-phenylindole were identified by computer- number of viable cells. Briefly, 5 Â 103 cells were plated generated reverse DAPI banding. The green to red per well in 24-well polystyrene plates (ATGC, Noisy- fluorescence ratio along each chromosome was calculated le-Grand, France) in culture medium and allowed to adhere by the appropriate software. CGH analysis of each tumor overnight. The medium was then replaced with fresh was done on 15 metaphases. The threshold values for losses medium containing (or not) increasing concentrations and gains were set at 0.9 and 1.1, respectively. Amplifica- of drug (0.03–150 Amol/L for the parental cells and tion corresponded to a ratio exceeding 1.5. 10–250 Amol/L for the variants) or the same concentration CGH Arrays of the excipient (DMSO or ethanol) as that used for the A genome-wide resource of 3,342 FISH-mapped, sequenced highest drug dose used for 96 hours. For the MTT assay, BAC and PAC clones verified for gene and marker contents the medium was replaced by MTT (2 mg/mL) diluted in were represented as immobilized DNA targets on glass medium (80%) for >3 hours and then replaced by DMSO, as slides for array-based CGH analysis allowing a mean previously described (11). For the methyl blue test assay, resolution of 1 Mb all along the genome. Each clone was

Mol Cancer Ther 2006;5(9). September 2006

Table 1. List of genes verified by real-time PCR

Gene RefSeq Sequence of primers A4 R/S C1 R/S A4 QPCR C1 QPCR symbol Acc

BMPR1B NM_001203 TCAAGAAGTTACGCCCCTCAG- 0.39 0.38 0.13 0.01 TCAGCCTTGATGCAGGATT CTGF NM_001901 AGCCTCAATTTCTGAACACCAC- 2.84 2.13 4.95 3.64 TCCCCTTTGCAAACAATCT CYP4F2 NM_001082 GCTTTGACCCAGAGAACATCAG- 2.54 2.39 Not determined 4.02 CCAGGACCACCTTCATCT EDNRB NM_000115 CCCTTTCCTTCTCCATGTCAAAA- 0.35 0.37 0.32 0.26 AGCCACTGAATGCAATTTT ELL NM_032245 ACCCCAGGTTTAAACGGAACTG- 2.17 2.32 1.72 1.30 TACTCGGCATTGAAGTCG EPHA3 NM_005233 TTCACGGGTGTGGAGTACAGAC- 0.41 0.19 0.70 0.12 TGGGCCATTCTTTGATTG FOS NM_005252 AGAATCCGAAGGGAAAGGAAC- 2.06 4.66 1.54 5.18 TTCTCCTTCAGCAGGTTGG GJA1 NM_000165 TCTTTTGGAGTGACCAGCAAAA- 0.42 0.39 0.17 0.07 GGCATTTGGAGAAACTGG GPM6A NM_005277 GGACCTTCGTCAGTTTGGAACTC- 0.44 0.39 0.37 0.16 CAGCAAGTGCCACAATA HTRA3 NM_053044 GACTTCCCAGAGGTCAGCAGCG- 2.17 2.49 1.59 1.56 TTGACCTTGACGATGATG IGFBP2 NM_000597 CCTCTACTCCCTGCACATCCGTT- 0.45 0.26 0.62 0.18 GGGGTTCACACACCAG IGFBP7 NM_001553 AGTGGTTGATGCCTTACATGACC- 0.39 0.14 0.58 0.07 ATGACTACTTTTAACCATGCAG LEF1 NM_016269 CTTTATCCAGGCTGGTCTGCTCGT- 0.42 0.41 0.60 0.34 TTTCCACCATGTTTCA RAI NM_006663 GGGTGAAGCCTCAAAGGAGTTG- 2.35 2.45 6.90 5.08 CAGATAAAGGCAGCAAAA RGL NM_015149 TGGTCTTTCCAGGAGATTGGGCA- 0.45 0.39 0.55 0.38 CAGAAGCACAAATCGAA SEC3L1 NM_018261 GCCAAACAAAAATACACAGATC- 0.44 0.43 0.58 0.29 ACGAGCTTCAACACCTTCAAA SRP72 NM_006947 AGCAACTGCAGGAGCTTCATTGG- 0.35 0.31 0.51 0.34 TGTCTTGGGGGTATGAT SSB1 NM_025106 GCACAGGGTTGCATTTCTTTCTCG- 2.48 2.05 3.35 3.96 CCCTCCCTTCTTAGTT ZNF436 NM_030634 TGCACAGAGGGACCTTTACCTGCT- 0.47 0.30 0.49 0.10 TGGGATTTACCTCGTT

NOTE: Results are expressed as the resistant/sensitive ratios of expression. The first two columns (A4 R/S and C1 R/S) are the results of microarray experiments for SNB-19-A4 and SNB-19-C1, respectively. The last two columns (A4 QPCR and C1 QPCR) show corresponding real-time quantitative PCR results.

spotted in triplicate on a slide with an Aminosilanne resuspended in hybridization buffer (50% formamide). coating (Corning UltraGAPS, NH3+) with the Microgrid Competitive cohybridization was done on CGH array TAS BioRobotics spotter. These slides were a generous gift slides preblocked by succinic anhydride/N-methyl-2-pyr- from the INSERM Unit U520 (France). rolidinone/borate buffer (Sigma-Aldrich). After extraction, 1.5 Ag of each test and control DNA After a 24-hour hybridization, slides were washed with sample was digested with DpnII enzyme (Ozyme, Saint SDS and saline citrate, dried, and scanned using a 4000B scan Quentin en Yvelines, France) and purified with a QIAquick (Axon Instruments, Union City, CA). Image analysis was PCR purification kit (Qiagen, Courtaboeuf, France). They done with Genepix5.1 software (AxonInstruments) and were then labeled by random priming using a Bioprime processed using software developed at the Curie Institute. DNA labeling kit (Invitrogen, Cergy Pontoise, France) with Any BAC with less than two replicates flagged for not the appropriate cyanine dye (Cy3 or Cy5; Perkin-Elmer, fulfilling qualitative spot criteria was excluded. A ratio of Wellesley, MA). The control and test DNAs were copreci- <0.8 was considered as a loss, a ratio of >1.2 as a gain, and a pitated with Cot-1 DNA (Invitrogen), denatured and ratio of >1.5 as amplification.

Mol Cancer Ther 2006;5(9). September 2006

Table 2. Comparative cytotoxicity of temozolomide, 1,3-bis(2-chloroethyl)-1-nitrosourea, and carboplatin in SNB-19 and its temozolomide-resistant sublines, SNB-19A4 and SNB-19C1

Drugs Cells, IC50 median (range)

SNB-19 SNB-19A4* Pc SNB-19C1* Pc

Temozolomide (Amol/L) 1.03 (0.042–1.7) 101 (90–113) 0.02 55 (40–59) 0.008 1,3-Bis(2-chloroethyl)-1-nitrosourea (Amol/L) 9.5 (8–13) 21 (13–29) 0.038 16 (13–19) 0.037 Carboplatin (Ag/L) 3.1 (0.58–5.8) 13 (5.5–15) 0.12b 3.1 (1.8–6) 0.65b

*Selected at 150 Amol/L of temozolomide. cStatistical significance was determined by the Mann-Whitney test. bNonsignificant.

Chromosome FISH Analysis of Gene Expression FISH was done as previously described (15). Several cDNA was prepared from each cell line using a chromosome-specific biotinylated painting probes (Oncor combination of random primers (Promega, Charbonnie`res, Appligene, Illkirch, France) were hybridized to metaphase France) and Superscript II (Invitrogen) for validation of spreads. The biotinylated probes were detected by goat array results by quantitative PCR. Oligos were designed anti-biotin antibody (Vector Laboratories, Burlingame, CA) around intron-exon boundaries for each gene using and fluoresceinated anti-goat rabbit antibody (Biosys, primer3 software.7 Each PCR was carried out in triplicate Compie`gne, France). Chromosomes were counterstained in a 20 AL volume using SybrGreen Mastermix(Applied with propidium iodide. Fluorescent signals were detected Biosystems, Courtaboeuf, France) in the ABI Prism system. using an epifluorescence microscope (DMRB, Leica). Each primer set was first tested to determine optimal Twenty metaphase spreads were analyzed to evaluate concentrations, to assess the specificity of the PCR product, signal distribution. and to test PCR efficacy. Values for each gene were RNA Extraction normalized to the expression levels of three housekeeping Total RNA was isolated using an RNA extraction kit genes: HMBS, BM, and SDHA, and then a ratio comparing (Macherey-Nagel, Hoerdt, France) according to the manu- expression in resistant versus sensitive cell lines was facturer’s instructions. RNA quality was assessed using an calculated. The sequences of the primers used for real-time Agilent 2100 Bioanalyser (Massy, France). PCR are listed in Table 1. Global Analysis of GeneTranscription In addition to these genes, MGMT expression was also For probe preparation, 20 Ag of RNA from sensitive or tested at the RNA level using the following primers: resistant cell lines were reverse-transcribed and labeled 5¶-CCTGGCTGAATGCCTATTTC-3¶ (forward) and 5¶-GAT- using the CyScribe Post-Labeling Kit procedure (Amer- GAGGATGGGGACAGGATT-3¶ (reverse). The methylation sham Biosciences, Orsay, France), without modifications. status of the MGMT promoter was also evaluated, as After purification with the CyScribe GFX Purification Kit previously described (16). (Amersham Biosciences), the cDNAs were combined and applied to the arrays (Agilent human 1A oligonucleotide microarray), exactly as recommended by the manufacturer. Slides were scanned in both Cy3 and Cy5 channels with the Results GenePix4000 scanner (AxonInstruments) and analyzed by Acquisition of Resistance to Temozolomide by the GenePixPro Software (AxonInstruments). For dye SNB-19 Glioma Cell Line normalization, the LOWESS (locally weighted linear SNB-19 parental cells were very sensitive to temozolo- regression) method was applied, using the VARAN mide (Table 2). Two clones of resistant variants (SNB-19A4 software.6 A log ratio of the red and green channel signals and SNB-19C1) were isolated at an early stage of drug (local background subtracted) was obtained from the treatment (3 Amol/L of temozolomide) and then succes- processed data. Dye-swap replicates were done for each sively exposed to incremental doses of temozolomide experiment and spots were selected on the basis of a 2 SD (up to 150 Amol/L). After initial killing of a large majority cut/of the mean log ratios. Ratios of 2 and 0.5 were set as of cells, surviving cells displayed a normal rate of proli- thresholds to identify differentially expressed genes. They feration. They were then submitted to a new selection A were confirmed as being significantly underexpressed or step up to 150 mol/L. The IC50 of temozolomide was overexpressed in resistant versus sensitive cells, as deter- evaluated by both MTT and methyl blue test techniques, mined by significance analysis of microarrays, using a false and similar results were obtained. The SNB-19A4 and discovery rate of <10%. SNB-19C1 variants showed a 100-fold (P = 0.02) and 55-fold

6 http://www.bionet.espci.fr/varan/varan_info.htm. 7 http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi.

Mol Cancer Ther 2006;5(9). September 2006

(P = 0.008) stronger resistance to temozolomide than the bands, loss of partial amplification of the 4p14-4q21 parental SNB-19 cell line, respectively. After 20 passages in region and loss of amplification of the 16q12-16q22 temozolomide-free medium, the temozolomide resistance region (Fig. 1B and C); these last two regions appeared factor remained constant. with a normal ratio during hybridization versus control The resistance of both variants to different alkylating DNA. During a second step, the same cohybridizations agents was compared with that of the parental cells. The were done with DNA extracted at a lower level of A IC50 of 1,3-bis(2-chloroethyl)-1-nitrosourea was increased temozolomide resistance (3, 30, and 60 mol/L). These 2.2-fold (P = 0.038) and 1.6-fold (P = 0.037) for SNB- alterations were present as early as the 30 Amol/L dose A 19A4 and SNB-19C1, respectively. The IC50 of carboplatin of temozolomide but were not detected at the 3 mol/L was increased 4.1-fold (P = 0.12) for SNB-19A4 and dose (data not shown). The IC50 of temozolomide for remained stable at 3.1 Amol/L for SNB-19C1 (Table 2). the SNB-19C1 variant selected at 30 Amol/L of temozo- Thus, the resistance to temozolomide acquired by these two lomide was the same as that for the variant selected at A A variants was stable and could be considered specific to 150 mol/L (IC50 =45 mol/L). temozolomide. CGH array analyses were done with the same cell lines, Genetic Alterations of Temozolomide-Resistant Cell first cohybridized with normal cell DNA (Fig. 2A), then Lines again cohybridized between the DNA of each resistant CGH was used to analyze the alterations in SNB-19A4 variant (DNA extracted at 150 Amol/L temozolomide) and and SNB-19C1 variants. SNB-19 showed numerous the DNA of the parental SNB-19 cells (Fig. 2B and C). These rearrangements (Fig. 1A), as previously published (10). results confirmed those of the conventional metaphase- DNA was extracted from resistant cells at the highest based CGH analysis. However, the altered regions were temozolomide concentration (150 Amol/L) and hybrid- more accurately defined (2p16.1-2p25.3, 4p14.4-4q21.22, ized with normal DNA (data not shown). Most of the and 16q12.1-16q22.1) and further loss of amplification of numerous imbalances observed were shared by the the 1p13.2-1q21.1 region shared by SNB-19A4 and SNB- parental cells and the temozolomide-resistant cells lines. 19C1 was detected. To eliminate these shared imbalances, the DNA from Cytogenetic Alterations of Temozolomide-Resistant each resistant variant was cohybridized with the DNA Variants as Studied by FISH and Karyotype Analysis from the parental SNB-19 cells. We were therefore able to SNB-19 exhibited a hyperdiploid karyotype with 59 to 61 identify alterations in the temozolomide-resistant clones chromosomes. SNB-19A4 was also hyperdiploid with 54 to that were absent in the parental cell line. Such alterations 55 chromosomes and SNB-19C1 was hypopentaploid with concerned the loss of a 2p region involving the 2p16-2p25 102 to 107 chromosomes (data not shown).

Figure 1. CGH profiles of SNB-19 cells and of SNB-19A4 and SNB-19C1 temozolomide-resistant variants. SNB-19 DNA versus normal control DNA (A), SNB-19A4 versus SNB-19 (B), and SNB-19C1 versus SNB-19 (C). Chromosome ideogram shows a ratio of 1 (central line), a ratio threshold of 0.9 for underrepresentation (left line), and 1.1 for overrepresentation (right line) of DNA/chromosome content. A, SNB-19 cells exhibit many chromosomal imbalances including amplification in chromosomes 4 and 16. Many of these rearrangements are frequently described in gliomas: e.g., gain of the PDGFRA region (on chromosome 4), gain of 7, loss of 10p, 10q and 13. B and C, in both temozolomide-resistant variants, a 2p region and the amplifications present on chromosomes 4 and 16 are lost.

Mol Cancer Ther 2006;5(9). September 2006

Figure 2. CGH array profiles of SNB-19 cells versus a normal control (A) and of temozolomide-resistant variants SNB-19A4 (B) and SNB-19C1 (C) both versus SNB-19. Each point corresponds to a BAC on the slide. They are colored according to the ratio of fluorescence. When the ratio was <0.8 (corresponding to a loss), the point is in green, and in red when the ratio was >1.2 (gain or amplification). Between these two ratios, the points are in yellow. As also shown by CGH, SNB-19 exhibited numerous imbalances, whereas some additional ones were not detected by CGH. Among them, gain of 1q32 is also frequent. Loss of 2p, 4, and 16q are also shown by CGH array profiles.

FISH was done with chromosomal painting probes for and 284 genes between SNB-19C1 and SNB-19. By cross- chromosomes 2, 4, and 16 on metaphases because CGH checking the two lists of genes, 78 genes were found to showed shared rearrangements involving these three exhibit similar alterations in the two temozolomide- chromosomes. FISH revealed a derivative chromosome in resistant variants. Differential expression was confirmed the parental SNB-19 cell line, that involved chromosomes by quantitative PCR analysis for 18 out of these 78 genes 2, 4, and 16 with at least five breakpoints (Fig. 3). This (Table 2). derivative chromosome was lost in both SNB-19A4- Quantitative alterations of gene expression and genomic resistant and SNB-19C1-resistant variants. However, it alterations were confronted. The location of the gene was was difficult to assign the lost chromosomal material found known in 72 out of the 78 differentially expressed genes by CGH to the loss of this marker alone because several (Table 3; Fig. 4). Among them, 11 genes (15%) were located chromosomes were hybridized to these probes. in the rearranged region (Table 3). Gene Expression Alterations inTemozolomide-Resistant MGMT Expression Variants as Studied by MicroarrayAnalysis In order to test whether MGMT expression might be We did a comparative transcriptome analysis between responsible for part of the resistance mechanism in A4 and parental SNB-19 cells and the two temozolomide-resistant C1 variants, we tested MGMT expression at the RNA level, variants. RNA isolated from SNB-19A4 and SNB-19C1 was but it was not detectable. Analysis of the methylation cohybridized with RNA isolated from SNB-19 parental status of the MGMT promoter showed that it was fully cells in a dye-swap experiment using Agilent arrays. The methylated in all cell lines. Consequently, the MGMT level search for differential expression evidenced quantitative did not allow us to discriminate SNB-19 from the resistant alterations of 245 genes between SNB-19A4 and SNB-19, variants.

Mol Cancer Ther 2006;5(9). September 2006

Discussion chromosomal material (13). Among the imbalances found Malignant glial tumors are well known to be refractory to in SNB-19, some are frequent in gliomas: gain of 1q32 and chemotherapy. Significant clinical benefits have been gain of the chromosome 4 region bearing the PDGFRA gene obtained with temozolomide, a recently introduced alky- (15, 27). lating agent (2). The main purpose of the present study was SNB-19 and the two resistant variants were aneuploid to generate temozolomide-resistant variants from SNB-19, a and their karyotypes were very complex, but strikingly, malignant glioma cell line, in an attempt to identify the there were fewer rearrangements in the resistant variants. genetic alterations associated with acquired resistance to The two temozolomide-resistant variants harbored ge- temozolomide. Several temozolomide-resistant variants nomic differences with the parental SNB-19 cell line: loss were obtained by stepwise selection with incremental of a 2p region and loss of amplification of the chromosome doses of temozolomide in the culture medium. Genomic 4 and 16q regions. This observation does not fit with the alterations appeared as early as the 30 Amol/L concentra- commonly accepted hypothesis assuming that the acquisition, which roughly corresponds with the mean peak tion of drug resistance, evolving through accidental concentration obtained after conventional treatment (i.e., chromosome recombinations, harbors more chromosomal 150 mg/m2; ref. 17). rearrangements (28). The resistance of these two variants can be considered A likely explanation is that temozolomide-resistant specific to temozolomide given the difference in the P value variants with less rearranged genomes preexisted in the between temozolomide and 1,3-bis(2-chloroethyl)-1-nitro- parental cell line and were selected through exposure to sourea and no cross-resistance was noted with carboplatin, temozolomide. This could be related to cell passaging, but a third alkylating agent also used to treat glioblastoma- these clones were cultured totally independently after the A bearing patients. This was surprising, because resistance to first 3 mol/L step. Moreover, two additional variants a given alkylating agent often extends to other alkylating displaying the same alterations were further isolated (data agents and even to other drugs (18, 19). Ma et al. (18) found not shown). The less complexrearrangement pattern of the that tumor cell–acquired resistance to temozolomide was temozolomide-resistant variants indicates that they may associated with increased MGMT activity and alterations have more in common with an ancestral genotype than the of apoptosis-controlling gene, i.e., increased Bcl2 and BclXL temozolomide-sensitive parental cells, which is compatible and decreased Bad, Bax, and BclXs. However, none of these with the idea that chemoresistance resulted from the genes were identified in our microarray analysis. enrichment of preexistent cells. Hypermethylation of the MGMT CpG island can cause The expression of 17,000 genes was quantified by transcriptional silencing in cell lines unable to repair O6- microarray analysis. Seventy-eight genes seemed to be methyl (20). We found no changes on MGMT expression differentially expressed in both variants: 15% were located which stay undetectable. The promoter methylation was in regions that had sustained losses (11). Among the 78 also unchanged, with both parental cell lines and resistant genes, only two, CLK1 and RAI, were previously associated variants giving 100% methylation (data not shown). A with a response to cisplatin treatment (29, 30). However, mismatch repair deficiency has not been addressed none were associated with resistance to temozolomide. specifically but such an abnormality is infrequent in Thirty (19 underexpressed and 11 overexpressed) of these gliomas. A differential expression was not found on MGMT 78 genes seemed to be of specific interest because of their genes, mismatch repair genes, or other genes involved in function. drug resistance mechanisms including base excision repair Among the underexpressed genes, some have been (21), drug-metabolism, and detoxification (mainly through described as tumor suppressor genes in solid tumor, higher levels of glutathione or glutathione-S-transferase; ref. 22). Resistance to temozolomide also involves p53 status, Chk1- and Chk2-mediated G2 checkpoint pathways, and Akt activation. In response to temozolomide, p53 induces cell cycle arrest and apoptosis (4). On the other hand, MGMT inactivation may facilitate AT-GC mutations on p53 (23, 24), explaining the correlation between MGMT inactivation and p53 dysfunction (25, 26). In response to temozolomide, Chk1 induces G2 arrest preventing from cell death independently of p53 status (5). Akt pathway activation, which is a frequent feature in glioblastoma, bypasses these mechanisms, preventing both G2 arrest and cell death (6). Figure 3. Gray scale inverted DAPIbanding ( A) and FISH analysis of Again, none of the genes differentially expressed in our metaphase chromosomes of SNB-19 parental tumors. Chromosomes were microarray analysis involved these pathways. counterstained with propidium iodide (red). FISH analysis of metaphases Gliomas are known to display very rearranged karyo- of the parental cell line with human chromosome painting of chromosomes 2(B), 4 (C), and 16 (D) show that the submetacentric marker types with numerous chromosomal imbalances, due to chromosome is a complex derivative chromosome comprising regions of sequential chromosomal endoreduplications and losses of chromosomes 2, 4, 16 and other unidentified chromosomes.

Mol Cancer Ther 2006;5(9). September 2006

Table 3. Differentially expressed genes in SNB-19A4 and SNB-19C1 versus SNB-19 ranging according to level of expression, physical mapping, and involvement in differentiation or stem cell apoptosis (A), mismatch repair or drug resistance (B), or oncogenesis or tumor progression (C)

Gene symbol Name SNB-19A4* SNB-19C1* Physical Function mapping

CNN1 Calponin 1, basic, smooth muscle 3.2 2.27 19 AC KIAA1189 KIAA1189 protein 3.07 2.04 2 CTGF Connective tissue growth factor 2.84 2.13 6 GNA14 Guanine nucleotide binding protein (G protein), a 14 2.8 2.08 9 PF4 Platelet factor 4 2.68 2.1 4 PTGIR Prostaglandin I2 (prostacyclin) receptor (IP) 2.65 2.16 19 OR4K11P OR4K11P 2.64 2.45 21 CYP4F2 CYP4F2 2.54 2.39 19 A SSB1 SPRY domain-containing SOCS boxprotein SSB-1 2.48 2.05 1 PLAC8 Placenta-specific 8 2.47 2.51 12 A ZPBP Zona pellucida binding protein 2.47 2.1 7 KRT14 Keratin 14 (epidermolysis bullosa simplex, Dowling-Meara, Koebner) 2.46 2.53 17 OR6T1 OR6T1 2.44 2.02 11 FLJ44682 FLJ44682 2.39 2.31 RAI RelA-associated inhibitor 2.35 2.45 19 B PRODH Proline dehydrogenase (oxidase) 1 2.33 2.5 22 B CST9L Cystatin 9-like (mouse) 2.24 2.02 20 FZD2 Frizzled homologue 2 (Drosophila) 2.22 2.32 17 A SLC22A1 Solute carrier family 22 (organic cation transporter), member 1 2.19 2.26 6 CEBPB CCAAT/enhancer binding protein (C/EBP), h 2.19 3.12 20 AC ELL Elongation factor RNA polymerase II 2.17 2.32 19 A HTRA3 Serine protease HTRA3 2.17 2.49 4 IAPP Islet amyloid polypeptide 2.16 2.1 12 TAS1R3 TAS1R3 2.13 2.23 1 LIF Leukemia inhibitory factor (cholinergic differentiation factor) 2.09 2.53 22 A FOS v-fos FBJ murine osteosarcoma viral oncogene homologue 2.06 4.66 14 C PTPRCAP 2.01 2.35 11 FLJ13852 Hypothetical protein FLJ13852 2.01 2.02 8 C40 Hypothetical protein C40 0.5 0.45 2 NEDD9 Neural precursor cell expressed, developmentally down-regulated 9 0.5 0.3 6 A TGFB3 Transforming growth factor, h3 0.49 0.49 14 A TPD52L1 Tumor protein D52-like 1 0.48 0.36 6 NET-7 Transmembrane 4 superfamily member tetraspan NET-7 0.48 0.47 10 HOMER1 Homer homologue 1 (Drosophila) 0.48 0.38 5 ENSA Endosulfine a 0.48 0.5 1 JDP1 J domain containing protein 1 0.47 0.31 10 ZNF436 Zinc finger protein 436 0.47 0.3 1 GALNT5 UDP-N-acetyl-a-D-galactosamine:polypeptide 0.46 0.44 2 N-acetylgalactosaminyltransferase 5 (GalNAc-T5) PHKB Phosphorylase kinase, h 0.45 0.39 16 BCMP11 Breast cancer membrane protein 11 0.45 0.43 7 A HADHA Hydroxyacyl-CoA dehydrogenase/a subunit 0.45 0.39 2 IGFBP2 Insulin-like growth factor binding protein 2, 36 kDa 0.45 0.26 2 C CTSL2 Cathepsin L2 0.45 0.36 9 C RGL RalGDS-like gene 0.45 0.39 1 ATP1A3 ATPase 0.45 0.4 19 KLHL13 Kelch-like 13 (Drosophila) 0.45 0.37 X GPM6A Glycoprotein M6A 0.44 0.39 4 A CTNND2 Catenin (cadherin-associated protein), y2 (neural 0.44 0.37 5 plakophilin-related arm-repeat protein) KIAA1212 KIAA1212 0.44 0.41 2 SEC3L1 SEC3-like 1 (S. cerevisiae) 0.44 0.43 4 A FEZ1 Fasciculation and elongation protein ~1 (zygin I) 0.44 0.33 11 AC SLITRK6 SLIT and NTRK-like family, member 6 0.44 0.36 13 C NMU Neuromedin U 0.43 0.38 4

(Continued on the following page)

Mol Cancer Ther 2006;5(9). September 2006

Gene symbol Name SNB-19A4* SNB-19C1* Physical Function mapping

GPR48 G protein-coupled receptor 48 0.43 0.29 11 CLK1 CDC-like kinase 1 0.43 0.37 2 B PRPF4B PRP4 pre-mRNA processing factor 4 homologue B (yeast) 0.42 0.44 6 GJA1 Gap junction protein, A1, 43 kDa (connexin 43) 0.42 0.39 6 ABC AFP a-Fetoprotein 0.42 0.36 4 A LEF1 Lymphoid enhancer-binding factor 1 0.42 0.41 4 A EPHA3 EphA3 0.41 0.19 3 A FBLN5 Fibulin 5 0.4 0.33 14 C TRIM22 Tripartite motif-containing 22 0.4 0.34 11 B IGFBP7 Insulin-like growth factor binding protein 7 0.39 0.14 4 C G3BP2 Ras-GTPase activating protein SH3 domain-binding protein 2 0.39 0.35 4 HELLS Helicase, lymphoid-specific 0.39 0.48 10 BC BMPR1B Bone morphogenetic protein receptor, type IB 0.39 0.38 4 A SCNN1A Sodium channel, non–voltage-gated 1a 0.38 0.47 12 GRIK2 Glutamate receptor, ionotropic, kainate 2 0.38 0.39 6 C C9orf58 Chromosome 9 open reading frame 58 0.38 0.41 9 KIAA1764 KIAA1764 protein 0.38 0.31 8 TKTL1 Transketolase-like 1 0.37 0.45 X BICD1 Bicaudal D homologue 1 (Drosophila) 0.37 0.43 12 SRP72 Signal recognition particle, 72 kDa 0.35 0.31 4 C EDNRB Endothelin receptor type B 0.35 0.37 13 A CAPS Calcyphosine 0.33 0.33 19 A LOC83690 CocoaCrisp 0.32 0.23 8 PLEKHB1 Pleckstrin homology domain containing, 0.3 0.27 11 AC family B (evectins) member 1 ARHGEF4 Rho guanine nucleotide exchange factor (GEF) 4 0.28 0.3 2

NOTE: Genes exhibiting a change in copy number and in expression level are in boldface. *Ratio of expression in the resistant variant versus SNB-19.

IGFBP7 (31), FEZ1/LZTS1 (32), and GJA1 (33), or as GPM6A (36), FBLN-5 (37), NEDD9 (38), BMPR1B (39), inductors of apoptosis, such as CTGF (34). Many are AFP (40), EPHA3 (41), CYP4F2 (42), SEC3L1 (43), involved in differentiation, notably in the nervous HELLS (44, 45), EDNRB (46), and PLEKHB1 (47). PLEKHB1 system, or in embryonic pathways such as GJA1 (35), has recently been found to be down-regulated in grade 4

Figure 4. Chromosomal location of the expression of modified genes. Most of them are located on chromosomes 2, 4, and 16, chromosomes involved in the rearrangement variants had in common, as well as on chromosomes 1, 6, 11, and mainly 19. Most of them were underexpressed.

Mol Cancer Ther 2006;5(9). September 2006

(glioblastomas) versus grade 1 gliomas (pilocytic astrocy- 7. Wasenius VM, Jekunen A, Monni O, et al. Comparative genomic hybridization analysis of chromosomal changes occurring during develop- toma). LEF1 is a transcription factor inhibiting the Wnt ment of acquired resistance to cisplatin in human ovarian carcinoma cells. pathway, a key regulator of cell development and of the Genes Chromosomes Cancer 1997;18:286 – 91. stem cell pool (48). 8. Rao PH, Houldsworth J, Palanisamy N, et al. Chromosomal amplification is associated with cisplatin resistance of human male germ cell Several overexpressed genes in our temozolomide- tumors. Cancer Res 1998;58:4260 – 3. resistant variants are involved in developmental pathways 9. Yasui K, Mihara S, Zhao C, et al. Alteration in copy numbers of genes as a such as FZD2 (implicated in the Wnt pathway; ref. 48), mechanism for acquired drug resistance. Cancer Res 2004;64:1403 – 10. or in stem cell self-renewal and proliferation such as LIF 10. Gross JL, Behrens DL, Mullins DE, Kornblith PL, Dexter DL. (49, 50), ELL (51), and CNN1 (52). Plasminogen activator and inhibitor activity in human glioma cells and modulation by sodium butyrate. Cancer Res 1998;48:291 – 6. Thus, a group of differentially expressed genes in the 11. Dimanche-Boitrel MT, Pelletier H, Genne P, et al. Confluence- SNB-19 temozolomide-resistant variants points to a shift dependent resistance in human colon cancer cells: role of reduced drug towards a cancer stem cell–like genotype. The concept of accumulation and low intrinsic chemosensitivity of resting cells. Int J a cancer stem cell was recently explored in human brain Cancer 1992;50:677 – 82. 12. Raynaud S, Nemati F, Miccoli L, et al. Antitumoral effects of tumors and cell lines (53, 54). Tumor stem cells retain the squamocin on parental and multidrug resistant MCF7 (human breast ability to self-renew and might explain why many cancers adenocarcinoma) cell lines. Life Sci 1999;65:525 – 33. relapse under conditions in which most tumor cells seem 13. Pruchon E, Chauveinc L, Sabatier L, et al. A cytogenetic study of 19 to be killed following therapy (54). An emerging model recurrent gliomas. Cancer Genet Cytogenet 1994;76:85 – 92. proposes that tumor stem cells are naturally resistant to 14. Kallioniemi A, Kallioniemi OP, Sudar D, et al. Comparative genomic hybridization for molecular cytogenetic analysis of solid tumors. Science chemotherapy through their quiescence, their capacity for 1992;258:818 – 21. DNA repair and increased ABC transporter expression 15. Muleris M, Almeida A, Dutrillaux AM, et al. Oncogene amplification in (55). In conclusion, our cytogenetic analyses suggest that human gliomas: a molecular cytogenetic analysis. Oncogene 1994;9: acquired temozolomide resistance in this malignant 2717 – 22. glioma model involves the selection of preexistent 16. Hegi ME, Diserens AC, Gorlia T, et al. MGMT gene silencing and benefit from temozolomide in glioblastoma. N Engl J Med 2005;352: resistant cell variants. The gene expression profile of 997 – 1003. temozolomide-resistant variants displayed a complex 17. Patel M, McCully C, Godwin K, Balis FM. Plasma and cerebrospinal picture with decreased expression of genes involved in fluid pharmacokinetics of intravenous temozolomide in non-human differentiation and increased expression of genes known primates. J Neurooncol 2003;61:203 – 7. 18. Ma J, Murphy M, O’Dwyer PJ, Berman E, Reed K, Gallo JM. to be expressed in stem cells. Altogether, these data Biochemical changes associated with a multidrug-resistant phenotype of a suggest that acquired temozolomide resistance could have human glioma cell line with temozolomide-acquired resistance. Biochem resulted from the selection of less differentiated preexis- Pharmacol 2002;63:1219 – 28. tent resistant cells with stem cell–like characteristics in the 19. Tews DS, Nissen A, Kulgen C, Gaumann AK. Drug resistance- associated factors in primary and secondary glioblastomas and their parental tumor. precursor tumors. J Neurooncol 2000;50:227 – 37. 20. Paz MF, Yaya-Tur R, Rojas-Marcos I, et al. CpG island hypermethylation of the DNA repair enzyme methyltransferase predicts Acknowledgments response to temozolomide in primary gliomas. Clin Cancer Res 2004;10: We thank Olivier Delattre for his generous gift of the CGH array slides and 4933 – 8. for granting access to his analysis platform; Dr. Jean-Gabriel Judde for 21. Trivedi RN, Almeida KH, Fornsaglio JL, Schamus S, Sobol RW. The critical reading of the manuscript; Dr. Franc¸ois Apiou, Anne-Marie- role of base excision repair in the sensitivity and resistance to Dutrillaux, and Miche`le Gerbeault for their technical assistance in temozolomide-mediated cell death. Cancer Res 2005;65:6394 – 400. karyotype, FISH, and CGH analyses; and Lorna Saint Ange (Institut Gustave-Roussy) for editing. 22. Calatozzolo C, Gelati M, Ciusani E, et al. Expression of drug resistance proteins Pgp, MRP1, MRP3, MRP5 and GST-pi in human glioma. J Neurooncol 2005;74:113 – 21. References 23. Esteller M, Herman JG. Generating mutations but providing chemosensitivity: the role of O6-methylguanine DNA methyltransferase in human 1. Fortin D. The blood-brain barrier should not be underestimated in neuro- cancer. Oncogene 2004;23:1 – 8. oncology. Rev Neurol (Paris) 2004;160:523 – 32. 24. Watanabe T, Katayama Y, Komine C, et al. O6-Methylguanine-DNA 2. Stupp R, van den Bent MJ, Hegi ME. Optimal role of temozolomide in methyltransferase methylation and TP53 mutation in malignant astrocy- the treatment of malignant gliomas. Curr Neurol Neurosci Rep 2005;5: tomas and their relationships with clinical course. Int J Cancer 2005;113: 198 – 206. 581 – 7. 3. Esteller M, Hamilton SR, Burger PC, Baylin SB, Herman JG. Inactivation 25. Rolhion C, Penault-Llorca F, Kemeny JL, et al. O(6)-methylguanine- of the DNA repair gene O6-methylguanine-DNA methyltransferase by DNA methyltransferase gene (MGMT) expression in human glioblastomas promoter hypermethylation is a common event in primary human in relation to patient characteristics and p53 accumulation. Int J Cancer neoplasia. Cancer Res 1999;59:793 – 7. 1999;84:416 – 20. 4. Bocangel DB, Finkelstein S, Schold SC, Bhakat KK, Mitra S, Kokkinakis 26. Hermisson M, and Weller M. NF-nB-independent actions of sulfasa- DM. Multifaceted resistance of gliom as to temozolomide. Clin Cancer Res lazine dissociate the CD95L- and Apo2L/TRAIL-dependent death signaling 2002;8:2725 – 34. pathways in human malignant glioma cells. Cell Death Differ 2003;10: 5. Hirose Y, Berger MS, Pieper RO. Abrogation of the Chk1-mediated G(2) 1078 – 89. checkpoint pathway potentiates temozolomide-induced toxicity in a p53- 27. Koschny R, Koschny T, Froster UG, Krupp W, Zuber MA. Comparative independent manner in human glioblastoma cells. Cancer Res 2001;61: genomic hybridization in glioma: a meta-analysis of 509 cases. Cancer 5843 – 9. Genet Cytogenet 2002;135:147 – 59. 6. Hirose Y, Katayama M, Mirzoeva OK, Berger MS, Pieper RO. Akt 28. Duesberg P, Stindl R, Hehlmann R. Origin of multidrug resistance in activation suppresses Chk2-mediated, methylating agent-induced G2 cells with and without multidrug resistance genes: chromosome reassort- arrest and protects from temozolomide-induced mitotic catastrophe and ments catalyzed by aneuploidy. Proc Natl Acad Sci U S A 2001;98: cellular senescence. Cancer Res 2005;65:4861 – 9. 11283 – 8.

Mol Cancer Ther 2006;5(9). September 2006

29. Umehara H, Nishii Y, Morishima M, et al. Effect of cisplatin treatment 42. Nagata T, Takahashi Y, Ishii Y, et al. Profiling of genes on speckled distribution of a serine/arginine-rich nuclear protein CROP/ differentially expressed between fetal liver and postnatal liver using Luc7A. Biochem Biophys Res Commun 2003;301:324 – 9. high-density oligonucleotide DNA array. Int J Mol Med 2003;11: 713 – 21. 30. Bergamaschi D, Samuels Y, O’Neil NJ, et al. iASPP oncoprotein is a key inhibitor of p53 conserved from worm to human. Nat Genet 2003;33: 43. Zhang QH, Ye M, Wu XY, et al. Cloning and functional analysis of 162 – 7. cDNAs with open reading frames for 300 previously undefined genes expressed in CD34+ hematopoietic stem/progenitor cells. Genome Res 31. Drivdahl R, Haugk KH, Sprenger CC, Nelson PS, Tennant MK, Plymate 2000;10:1546 – 60. SR. Suppression of growth and tumorigenicity in the prostate tumor cell line M12 by overexpression of the transcription factor SOX9. Oncogene 44. Geiman TM, Durum SK, Muegge K. Characterization of gene 2004;23:4584 – 93. expression, genomic structure, and chromosomal localization of Hells (Lsh). Genomics 1998;54:477 – 83. 32. Ishii H, Vecchione A, Murakumo Y, et al. FEZ1/LZTS1 gene at 8p22 suppresses cancer cell growth and regulates mitosis. Proc Natl Acad Sci 45. Yano M, Ouchida M, Shigematsu H, et al. Tumor-specific exon U S A 2001;98:10374 – 9. creation of the HELLS/SMARCA6 gene in non-small cell lung cancer. Int J Cancer 2004;112:8 – 13. 33. Huang R, Lin Y, Wang CC, et al. Connexin 43 suppresses human glioblastoma cell growth by down-regulation of monocyte chemotactic 46. Stanchina L, Baral V, Robert F, et al. Interactions between Sox10, protein 1, as discovered using protein array technology. Cancer Res 2002; Edn3 and Ednrb during enteric nervous system and melanocyte develop- 62:2806 – 12. ment. Dev Biol 2006;295:232 – 49. 34. Szeto CC, Chow KM, Lai KB, Szeto CY, Kwan BC, Li PK. Connective 47. Colin C, Baeza N, Bartoli C, et al. Identification of genes differentially tissue growth factor is responsible for transforming growth factor-h- expressed in glioblastoma versus pilocytic astrocytoma using Suppression induced peritoneal mesothelial cell apoptosis. Nephron Exp Nephrol 2006; Subtractive Hybridization. Oncogene 2006;25:2818 – 26. 103:166 – 74. 48. Galceran J, Sustmann C, Hsu SC, Folberth S, Grosschedl R. LEF1- D 35. Bhattacharya B, Miura T, Brandenberger R, et al. Gene expression in mediated regulation of -like1 links Wnt and Notch signaling in somito- human embryonic stem cell lines: unique molecular signature. Blood 2004; genesis. Genes Dev 2004;18:2718 – 23. 103:2956 – 64. 49. Galli R, Binda E, Orfanelli U, et al. Isolation and characterization of tumorigenic, stem-like neural precursors from human glioblastoma. Cancer 36. Mukobata S, Hibino T, Sugiyama A, et al. M6a acts as a nerve growth Res 2004;64:7011 – 21. factor-gated Ca(2+) channel in neuronal differentiation. Biochem Biophys Res Commun 2002;297:722 – 8. 50. Pitman M, Emery B, Binder M, Wang S, Butzkueven H, Kilpatrick TJ. LIF receptor signaling modulates neural stem cell renewal. Mol Cell 37. Albig AR, Schiemann WP. Fibulin-5 antagonizes vascular endothelial Neurosci 2004;27:255 – 66. growth factor (VEGF) signaling and angiogenic sprouting by endothelial cells. DNA Cell Biol 2004;23:367 – 79. 51. Kanda Y, Mitani K, Kurokawa M, Yamagata T, Yazaki Y, Hirai H. Overexpression of the MEN/ELL protein, an RNA polymerase II elongation 38. Merrill RA, See AW, Wertheim ML, Clagett-Dame M. Crk-associated factor, results in transformation of Rat1 cells with dependence on the substrate (Cas) family member, NEDD9, is regulated in human neuroblas- lysine-rich region. J Biol Chem 1998;273:5248 – 52. toma cells and in the embryonic hindbrain by all-trans retinoic acid. Dev Dyn 2004;231:564 – 75. 52. Xie D, Yin D, Wang HJ, et al. Levels of expression of CYR61 and CTGF are prognostic for tumor progression and survival of individuals with 39. Wine-Lee L, Ahn KJ, Richardson RD, Mishina Y, Lyons KM, Crenshaw gliomas. Clin Cancer Res 2004;10:2072 – 81. EB 3rd. Signaling through BMP type 1 receptors is required for development of interneuron cell types in the dorsal spinal cord. 53. Singh SK, Hawkins C, Clarke ID, et al. Identification of human brain Development 2004;131:5393 – 403. tumour initiating cells. Nature 2004;432:396 – 401. 40. Filmus J, Capurro M. Glypican-3 and a-fetoprotein as diagnostic tests 54. Kondo T, Setoguchi T, Taga T. Persistence of a small subpopulation for hepatocellular carcinoma. Mol Diagn 2004;8:207 – 12. of cancer stem-like cells in the C6 glioma cell line. Proc Natl Acad Sci U S A 41. Kudo C, Ajioka I, Hirata Y, Nakajima K. Expression profiles of EphA3 2004;101:781 – 6. at both the RNA and protein level in the developing mammalian forebrain. 55. Dean RA, Talbot NJ, Ebbole DJ, et al. The genome sequence of the J Comp Neurol 2005;487:255 – 69. rice blast fungus Magnaporthe grisea. Nature 2005;434:980 – 6.

Mol Cancer Ther 2006;5(9). September 2006

Downloaded from mct.aacrjournals.org on October 2, 2021. © 2006 American Association for Cancer Research. Genetic alterations associated with acquired temozolomide resistance in SNB-19, a human glioma cell line

Nathalie Auger, Joëlle Thillet, Krystell Wanherdrick, et al.

Mol Cancer Ther 2006;5:2182-2192.

Updated version Access the most recent version of this article at: http://mct.aacrjournals.org/content/5/9/2182

Cited articles This article cites 55 articles, 20 of which you can access for free at: http://mct.aacrjournals.org/content/5/9/2182.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/5/9/2182.full#related-urls

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/5/9/2182. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.