Altered Erbb Receptor Signaling and Gene Expression in Cisplatin-Resistant Ovarian Cancer

Research Article

Altered ErbB Receptor Signaling and Gene Expression in Cisplatin-Resistant Ovarian Cancer

Kenneth Macleod, Peter Mullen, Jane Sewell, Genevieve Rabiasz, Sandra Lawrie, Eric Miller, John F. Smyth and Simon P. Langdon

Cancer Research UK Centre, University of Edinburgh, Edinburgh, United Kingdom

Abstract The erbB receptor family consists of the EGFR (erbB-1 and The majority of ovarian cancer patients are treated with HER-1), erbB-2 (HER-2), erbB-3 (HER-3), and erbB-4 (HER-4; refs. platinum-based chemotherapy, but the emergence of resistance 6, 7). Ligands of the EGF family, including transforming growth factor-a (TGFa), activate the EGFR, whereas members of the to such chemotherapy severely limits its overall effectiveness. We have shown that development of resistance to this neuregulin (NRG)/heregulin family activate erbB-3 and erbB-4 (6). treatment can modify cell signaling responses in a model ErbB-2 is activated via interaction with other ligand-stimulated system wherein cisplatin treatment has altered cell respon- family members. Ligand binding to EGFR promotes either siveness to ligands of the erbB receptor family. A cisplatin- dimerization with another EGFR (homodimerization) or binding resistant ovarian carcinoma cell line PE01CDDP was derived to another erbB receptor family member (heterodimerization) in from the parent PE01 line by exposure to increasing concen- which case, erbB2 is the preferred option (7). The erbB receptor trations of cisplatin, eventually obtaining a 20-fold level of family are widely reported to have key roles in regulating a network resistance. Whereas PE01 cells were growth stimulated by the of signaling pathways, including the Ras/Raf/mitogen-activated erbB receptor-activating ligands, such as transforming growth protein kinase kinase (MEK)/extracellular signal-regulated kinase factor-A (TGFA), NRG1A, and NRG1B, the PE01CDDP line was (ERK) pathway (8), the phosphatidylinositol 3-kinase (PI3K)/Akt g growth inhibited by TGFA and NRG1B but unaffected by pathway (9), and the PLC cascades (10). These are implicated in NRG1A. TGFA increased apoptosis in PE01CDDP cells but the translation of ligand-mediated signaling at the cell membrane to the nucleus where induction of gene expression provides the decreased apoptosis in PE01 cells. Differences in extracellular signal-regulated kinase and phosphatidylinositol 3-kinase basis for the cell response. In response to external stimuli, cells may signaling were also found, which may be implicated in the be directed to grow, differentiate, migrate, or apoptose; outcomes altered cell response to ligands. Microarray analysis revealed which will ultimately determine the rate and the manner in which 51 genes whose mRNA increased by at least 2-fold in PE01CDDP the disease will progress and also how it will respond to treatment. cells relative to PE01 (including FRA1, ETV4, MCM2, AXL, MT3, The erbB receptors and their ligands play key roles in the growth TRAP1, and FANCG), whereas 36 genes (including IGFBP3, and progression of several cancers including ovarian cancer (6, 11). TRAM1, and KRT4 and KRT19) decreased by a similar amount. Overexpression of both EGFR and erbB2 separately have been Differential display reverse transcriptase-PCR identified al- linked to poor survival in ovarian cancer (12–15) and EGFR tered mRNA expression for TCP1, SLP1, proliferating cell activators are mitogens in ovarian cancer cell lines in culture (16–18). Several EGFR-targeted inhibitors are currently being nuclear antigen, and ZXDA. Small interfering RNA inhibition of FRA1, TCP1,andMCM2 expression was associated with considered for evaluation in advanced ovarian cancer and the reduced growth and FRA1 inhibition with enhanced cisplatin effect of prior treatment with cisplatin on the signaling pathways sensitivity. Altered expression of these genes by cytotoxic may be important in influencing tumor response. These agents exposure may provide survival advantages to cells including include the tyrosine kinase inhibitors gefitinib (‘‘Iressa’’; ZD 1839; deregulation of signaling pathways, which may be critical in the refs. 19, 20) and erlotinib (‘‘Tarceva’’; OSI-774; ref. 21). To understand further the changes in EGF signaling after the development of drug resistance. (Cancer Res 2005; 65(15): 6789-800) onset of cisplatin resistance, we have developed an in vitro model wherein the parent cell line PE01 has been made resistant to Introduction cisplatin by exposure to increasing levels of drug. Coupled with this Platinum-based chemotherapy is a standard first-line treatment change in drug sensitivity is a modified responsiveness to ligands of for ovarian cancer (1). In many instances, however, resistance to the erbB receptor family. In this report, we have identified a chemotherapy develops, limiting its use. The changes in cellular number of changes that might be significant in explaining the signaling after drug treatment are currently attracting a great deal altered phenotype. of interest and this is of particular relevance when cytotoxic therapy is combined with signaling inhibitors. Among the key regulators of growth in this disease is the group 1 receptor tyrosine Materials and Methods kinase family of erbB receptors. Cisplatin is known to interact with Cell lines and culture conditions. The PE01 cell line was derived from the epidermal growth factor receptor (EGFR) pathway (2) either a poorly differentiated human ovarian adenocarcinoma (22). The PE01CDDP promoting (3, 4) or inhibiting apoptosis and cell death (5). variant was established by in vitro exposure of the PE01 line to cisplatin commencing at 25 nmol/L cisplatin and increasing to 1 Amol/L over a period of 4 months. The resistant cell line thus obtained was then passaged Requests for reprints: Simon P. Langdon, Cancer Research UK Centre, University in the absence of cisplatin for a further 4 months to ensure that the of Edinburgh, Crewe Road South, Edinburgh EH4 2XR, United Kingdom. Phone: 44- 131-777-3537; Fax: 44-131-777-3520; E-mail: [email protected]. resistance remained stable. Both cell lines were maintained in RPMI 1640 I2005 American Association for Cancer Research. (Life Technologies, Paisley, Scotland) containing 10% heat-inactivated FCS, www.aacrjournals.org 6789 Cancer Res 2005; 65: (15). August 1, 2005

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 2005 American Association for Cancer Research. Cancer Research penicillin (100 units/mL), and streptomycin (100 Ag/mL). Cells were left for 24 hours. Medium was changed every 2 to 3 days for 12 to 14 days j maintained routinely at 37 C in a humidified atmosphere of 5% CO2 in air. until colonies (>50 cells) were obtained in the untreated control wells. The Growth assays. Cells in log-phase growth were seeded in 24-well plates surviving fraction in each of the drug-treated samples was then counted (Falcon, Franklin Lakes, NJ) at a density of 2.5 Â 104 cells per well in using an inverted microscope and the colony counts obtained for drug- quadruplicate in RPMI 1640 containing 10% heat-inactivated FCS, penicillin treated wells expressed as a percentage of the counts in untreated control (100 units/mL), and streptomycin (100 Ag/mL). After 24 hours, medium was wells. Dose-response curves were plotted for each drug-cell line combina- replaced by phenol red-free RPMI 1640 containing 5% double charcoal- tion and IC50 values extrapolated from summed data obtained from at least stripped FCS (DCS-FCS), penicillin (100 units/mL), streptomycin (100 Ag/ three experiments in every case. Cytotoxic drugs were obtained from the mL), and glutamine (2 mmol/L). After a further 24 hours, medium was following sources: Carboplatin (CBDCA) and iproplatin (CHIP) were gifts removed and replenished. Ligand additions were made at this time point, from Dr. Mervyn Jones (Institute of Cancer Research, Sutton, United designated day 0, and medium plus additions was replaced on day 2. In Kingdom); cisplatin and JM40 were obtained from Bristol Myers certain experiments, inhibitors were also added alongside ligands. Cells Pharmaceuticals; chlorambucil and melphalan from Sigma; doxorubicin were harvested on days 0, 2, and 5 and counted using a cell counter (Coulter from Farmitalia Carlo Erba Ltd.; and prednimustine from Aktiebolaget Leo Electronics, Ltd., Luton, England). TGFa and EGF were obtained from (Helsingborg, Sweden). Boehringer Mannheim (Indianapolis, IN), NRG1a from Sigma Ltd. (St. Louis, Apoptosis assay. PE01 and PE01CDDP cells were grown to 80% MO), and NRG1h from Neomarkers (Fremont, CA). Gefitinib was a gift from confluence in the presence of RPMI 1640 containing 10% FCS. Following AstraZeneca (Macclesfield, United Kingdom) and LY294002 (40202) and an overnight incubation in phenol red–free RPMI 1640 containing 5% DCS- PD98059 (513000) were obtained from Calbiochem (La Jolla, CA). FCS, the medium was replenished before the addition of TGFa. Apoptosis Clonogenic assay. Cells in logarithmic growth were plated in 35-mm was measured using the TACS Annexin V-FITC kit (R&D Systems, 6-well plates (Costar, Cambridge, MA) at densities that produced 100 to 200 Minneapolis, MN) following the prescribed protocol. colonies in untreated wells (2 Â 103/mL for PE01 and 103/mL for PE01CDDP mRNA extraction and reverse transcriptase-PCR. Total cellular RNA cells). The medium described above was used but with the addition of was extracted from cells in log-phase growth using TRI reagent (Sigma, 1 mmol/L sodium pyruvate to enhance plating at low cell densities. After Poole, United Kingdom), following the standard protocol provided. Samples 48 hours, drug was added at appropriate concentrations to the wells and were treated with 20 units DNase 1 (Roche, Nutley, NJ) to remove genomic

Figure 1. In vitro growth responses of the PE01 and PE01CDDP cell lines to cytotoxic agents and erbB receptor activating ligands. A, clonogenic assay to assess IC50 growth inhibitory concentrations. PE01 and PE01CDDP cells were exposed to a range of concentrations of cytotoxic agents for 24 hours and colonies then allowed to grow for 12 to 14 days. Dose-response curves were plotted for each drug-cell line combination and IC50 values extrapolated from summed data obtained from at least three experiments in every case. B, response to TGFa of cells grown in DCS-FCS medium. PE01 and PE01CDDP cells were treated with 1 nmol/L TGFa for 8 days and cell counts made daily. C, effect of TGFa (10À8 to 10À12 mol/L), EGF, NRG1a and NRG1h (all 10À9 mol/L) on growth response. Cells were grown as described in Materials and Methods and treated for 5 days with ligands. Cells were then harvested and counted by cell counter. Columns, means; bars, FSD. Statistical comparisons were made with the relevant control line. *, P < 0.05, (ANOVA followed by Tukey-Kramer multiple comparisons test).

Cancer Res 2005; 65: (15). August 1, 2005 6790 www.aacrjournals.org

Figure 2. Effects of TGFa (1 nmol/L) on morphology and apoptosis in PE01 and PE01CDDP cell lines. A, cells exposed to TGFa were observed and photographed daily for 10 days. Images shown are of day 5 incubations. B, apoptotic cell number as detected by Annexin V positivity on day 4 of treatment. Cells were treated and collected as described in Materials and Methods and % cells showing Annexin V positivity measured by FACScan. *, P < 0.05 (t test, treated group versus control).

DNA contamination. RNA was then re-extracted using a phenol/chloroform Quantitative reverse transcriptase-PCR. RNA extraction, DNase protocol. Reverse transcription was done with a first-strand cDNA Synthesis treatment, and cDNA synthesis were conducted as described above; kit (Roche) using the oligo dT primer provided. One microgram of RNA 2.5 AL of cDNA was analyzed by real-time PCR using a LightCycler (Idaho yielded 20 AL of cDNA of which 2 AL were used for each subsequent PCR Technology). Reactions included LC Master Mix (Biogene, Kimbolton, reaction with each primer pair. PCR reactions were done in a final volume United Kingdom), SYBR green dye at a final concentration of 1:20,000 (1765; of 20 AL containing the following: 1Â PCR buffer, 1.5 mmol/L MgCl2, Biogene), magnesium chloride at 4 mmol/L, and forward and reverse 0.2 mmol/L deoxynucleotide triphosphate mixture, 2.5 units Taq polymer- primers at 0.5 Amol/L. The primer pairs used are were as follows: Fanconi ase (Cancer Research UK, Clare Hall, South Mimms, United Kingdom), anemia complementation group G (FANCG): CTGTAGCTGCCACGTTTTGA, 400 Amol/L each primer. The amplification reaction was carried out for GGTGGTGGCAGAGATTGTTT; tumor necrosis factor type 1 receptor– 35 cycles using the following variables: denaturation, 94jC for 5 minutes; associated protein (TRAP1): TGCGAGATGTGGTAACGAAG, cycling, 94jC for 30 seconds, 55jC for 30 seconds, 72jC for 45 seconds CGGTGCGTCCGTCTTATAGT; Axl oncogene (AXL): TTTCCTGA- (repeated for 35 cycles); extension, 72jC for 5 minutes. For semiquantitative GTGAAGCGGTCT, CATCTGAGTGGGCAGGTACA; minichromosome reverse transcriptase-PCR (RT-PCR), 2 AL cDNA were amplified using the maintenance 2 (MCM2): ACCAGGACAGAACCAGCATC, CAGG- standard PCR protocol but reactions were terminated at 15, 20, or 25 cycles ATGTCAAAGCGTGAGA; ETV4: GCAGATCCCCACTGTCCTAC, to fall within the linear region of PCR product synthesis. Activator protein GTTCTCTGTGGTTGGGGAAA; metallothionein 3 (MT3): TGTGAGAA- (AP-1) primers (designed using Primer 3 software) were as follows: JUNB: GTGTGCCAAGGA, GTCATTCCTCCAAGGTCAGC; T-complex protein 1 TCTCTCAAGCTCGCCTCTTC, ACGTGGTTCATCTTGTGCAG; JUN: GGAG- (TCP1): CCCAGGTTCTCAGAGCTCA, GGATGACACACAAAGCATCG; pro- TGTCCAGAGAGCCTTG, GAAAGGCTTGCAAAAGTTCG; JUND: TTCT- liferating cell nuclear antigen (PCNA): CGGGGGAATGTTAAGAGGAT, ACTCGGGGAACAAACG, GGCGAACCAAGGATTACAAA; FOSB: CCAGCCACGAAAGTGAAAGT; insulin-like binding protein 3 (IGFBP3): GACTCAAGGGGGTGACAGAA, AAAATGTCACAGCCCCTCAC; FOS: TTTA- CGTCAACGCTAGTGCCGTCAGCCG, GACCATATTCTGTCTCCCGCTTG- TAGTGGGCGGAAGTGG, ACGTCCTGGACAAAGGTCAC; fos-related anti- GACT; KRT19: GGTCAGTGTGGAGGTGGATT, TCAGTAACCTCG- gen 2 (FRA2): GGAGCTGGAGGAGGAGAAGT, GGGCTCCTGTTTCACC- GACCTGCT; translocating chain-associating membrane protein (TRAM): ACTA; FRA1: CCCTGCCGCCCTGTACCTTGTATC, AGACATTGGC- ACAGCTGGCTTACTGGCTTC, CGGGAAATGTGGAAAAGAAA; KRT4: TAGGGTGGCATCTGCA. GGCAGCAGAAGCCTCTACAA, CCACCCTTACCACTGAAGGA; ZXDA: PCR products were visualized after electrophoresis on polyacrylamide GGACTCTTTGGCCATGAAAA, ACTGGGTTTCTCCCTCCTGT; secretory gels and sized using a 100-bp ladder (Life Technologies). PCR reactions were leukoprorease inhibitor 1 (SLP1): GGGAAGTGCCCAGTGACTTA, AAAG- also carried out on RNA which had not been reverse transcribed to check GACCTGGACCACACAG; h-actin: CTACGTCGCCCTGGACTTCGAGC, for genomic DNA contamination but these were routinely negative. GATGGAGCCGCCGATCCACACGG. www.aacrjournals.org 6791 Cancer Res 2005; 65: (15). August 1, 2005

The standard protocol used was as follows: premelt: 95jC for 10 seconds; clone EGFR1 (Imperial Cancer Research Fund, Clare Hall, London, United PCR: 95jC for 0 second, ramp at 20jC/s, 55jC for 0 second, ramp at 20jC/s, Kingdom); erbB2, clone CB11 (Novocastra); and erbB3, clone RTJ1 72jC for 6 seconds, ramp at 5jC/s, XjC for 0 second, ramp at 20jC/s (Novocastra). Cells were harvested by trypsinisation, washed in cold PBS (where X was an empirically determined temperature high enough to melt containing 5% FCS, and aliquots of f106 cells were then incubated for primer dimer but leave PCR product intact) followed by a melt step from 60 minutes with antibody. For erbB2 and erbB3 staining, 1% saponin (BDH, 72jCto95jC at 0.1jC/s. Poole, Dorset, United Kingdom) was added to the cells before antibody Standard curves were obtained by performing reactions with predeter- addition (23). mined amounts of target template DNA for each primer pair. Contamina- Cells were then washed in PBS/FCS and incubated with sheep-antimouse tion of RNA by genomic DNA was excluded by doing reactions on RNA, FITC (1:20) for 60 minutes and washed twice with PBS/FCS. Cells were which had not been reverse transcribed. resuspended in PBS and analyzed on the FACScan flow cytometer. Determination of epidermal growth factor receptor, erbB2, and Western blotting. Western analysis was undertaken as previously erbB3 by immunofluorescence. Expression of EGFR, erbB2, and erbB3 described (24). The following antibodies were used: phosphotyrosine proteins were measured in PE01 and PE01CDDP cells by immunofluores- (PY20; Santa Cruz Biotechnology, Santa Cruz, CA), phospho-ERK (9101; cence using a flow cytometer. The following antibodies were used: EGFR, New England Biolabs), phospho-AKTS473 (9271; New England Biolabs), ERK I/II (9102; New England Biolabs), AKT (9272; New England Biolabs), actin (CP01; Oncogene Research Products), FRA1 (N-17, sc-183; Santa Cruz Biotechnology), PCNA (NA03; Oncogene Research Products), ETV4 (sc-114; Santa Cruz Biotechnology), TRAP1 (616480; Calbiochem), and IGFPB3 (GF60; Oncogene Research Products). Antibodies were used at 1/1,000 dilution apart from PY20, which was used at 1/200. The following inhibitors were used: gefitinib (‘‘Iressa’’; ZD1839; AstraZe- neca; 1 Amol/L) is an EGFR-tyrosine kinase inhibitor and was added to cells 5 minutes before growth factor; LY294002 (440202; Calbiochem; 10 Amol/L) inhibits PI3K and was added to cells 30 minutes before growth factor; PD98059 (513000; Calbiochem; 50 Amol/L) inhibits MEK and was added 60 minutes before growth factor. For the signaling study, 1 nmol/L TGFa was added for 15 minutes. Extracellular signal-regulated kinase activity. ERK activity was measured using a Biotrak mitogen-activated protein kinase activity assay (RPN 84; Amersham Life Science, Amersham, United Kingdom) following the recommended protocol.1 Differential display. RNA extraction, DNase treatment, and re-extraction were as described above. Forty-cycle PCR reactions were undertaken with three single base anchored primers (AAGCTTTTTTTTTTTA, AAGCT- TTTTTTTTTTC, and AAGCTTTTTTTTTTTG) each paired with seven random primers (AAGCTTTCTACCC, AAGCTTTGGCTCC, AAGCTTATA- CAGG, AAGCTTGTCATAG, AAGCTTCAAGTCC, AAGCTTCTGACAC, and AAGCTTCAATCGC) in a mix which contained 32P-dATP. Samples were run on a 6% sequencing gel, bands of interest were revealed by exposure to film, excised, amplified by PCR, sequenced, and BLAST searched to identify genes. Small interfering RNA transfection. Small interfering RNA (siRNA) was used to inhibit the expression and function of target mRNAs. Cells (30-50% confluence) were treated with 100 nmol/L siRNA which had been precomplexed with oligofectamine (Invitrogen, San Diego, CA). Transfection was done in Opti-MEM (Invitrogen) as per manufacturer’s guidelines for 4 hours. After 4 hours, medium was supplemented with RPMI 1640 and DCS-FCS (to a final concentration of 5%), 1 nmol/L TGFa was added as appropriate. RNA was harvested after 24 hours and cell counts obtained after 2 or 5 days. Sense sequences used were as follows: FRA1, GCAUCAACACCAUGAGUGGTT; AXL, GGUACAUUGG- CUUCGGGAUTT; TCP1, GGCCCUCAAGUCUCAUAUATT; MCM2, GGAUGGAGAGGAGCUCAUUTT: FRA1 (second sequence), GUAUCCCA- CAUCCAACUCCTT. Antisense sequences used were the complementary sequences and sense sequences were annealed to antisense sequences before use. Negative control siRNA (Ambion, Austin, TX) was also included and showed nonsignificant effects on growth. For the Western Figure 3. Effects of TGFa on erbB receptor signaling and ERK activation in analysis of FRA1 inhibition, a second negative control siRNA (Ambion) PE01 and PE01CDDPcells. A, expression of EGF receptor, erbB2, and erbB3 in was used. CDDP PE01 and PE01 cells. Expression was measured as described in Materials Clontech array. Atlas human cancer cDNA expression arrays (Human and Methods by flow cytometer. Columns, means; bars, FSD. There were no significant differences in expression between these two cell lines for these three cancer 1.2; Clontech, Palo Alto, CA) were used to analyze differential receptors. B, time course of TGFa-induced phosphorylation of EGF receptor and gene expression between PE01 and PE01CDDP cell lines, with and without erbB2. Western blot analysis was done as described in Materials and Methods. 48 hours treatment with TGFa (1 nmol/L). Isolation of total RNA, cDNA C, down-regulation of EGF receptor expression after 2 and 5 days exposure to 10 nmol/L TGFa. Expression was assessed by flow cytometer. Columns, means; bars, FSD. Statistical comparisons were made with the comparative non–TGFa-treated samples. *, P < 0.05 (ANOVA followed by Tukey-Kramer multiple comparisons test). D, activation of ERK in the PE01 and PE01CDDP cell lines by TGFa. ERK activity was measured by a kinase method as described in 1 http://www.jp.amershambiosciences.com/tech_support/manual/pdf/cellasy/ Materials and Methods. Points, means of quadruplicate samples; bars, FSD. rpn84plaa.pdf

Cancer Res 2005; 65: (15). August 1, 2005 6792 www.aacrjournals.org

Figure 4. TGFa activation of the Akt and ERK pathways: effects of specific inhibitors on signaling, growth responses to TGFa, and subsequent gene expression changes in PE01 and PE01CDDP cell lines. A, Western blots of phospho-Akt and phospho-ERK1/2 in the presence and absence of TGFa (1 nmol/L for 15 minutes) and inhibitors. LY294002 (10 Amol/L), PD 98059 (50 Amol/L), and gefitinib (1 Amol/L) were added before growth factor as described in Materials and Methods. B, effects of inhibitors on cell growth of PE01 and PE01CDDP cell lines. Inhibitors were used at the concentrations described above and cells were treated for 5 days. Columns, means of quadruplicate samples; bars, FSD. Statistical comparisons were made with the comparative no inhibitor control sample. *, P < 0.05 (ANOVA followed by Tukey-Kramer multiple comparisons test). C, quantitative RT-PCR of identified up-regulated (FRA1) and down-regulated (IGFBP3) mRNAs from cell lines exposed to TGFa F inhibitors for 24 hours. Expression was measured as described in Materials and Methods. Columns, mean of quadruplicate samples; bars, FSD.

synthesis, labeling of cDNA, and hybridization of cDNA probes to the PE01CDDP cells showed a reduced dependency on serum-containing Atlas Array filters were done according to the manufacturer’s protocol hormones and growth factors as indicated by similar growth rates (Clontech). To normalize the signal intensity between a pair of arrays, in either complete or charcoal-stripped serum, whereas PE01 cells the global (sum) normalization method was used. showed a dramatically reduced growth rate when cultured in charcoal-stripped conditions (Fig. 1B). Results These two cell lines exhibited markedly different responses to Differential growth responses to ErbB ligands after devel- erbB receptor activating ligands. Whereas the parent PE01 cell line CDDP opment of cisplatin resistance. The PE01CDDP ovarian cancer cell was growth stimulated by TGFa (10 nmol/L), the PE01 cell line line was derived from the parental wild-type PE01 cell line by was growth inhibited (Fig. 1C). EGF produced similar effects exposure to increasing concentrations of cisplatin and eventually (Fig. 1C). Both NRG1a and NRG1h stimulated the growth of the showed a 20-fold increase in resistance to this agent as measured PE01 cell line at 1 nmol/L with the latter producing a more potent by clonogenic assay (Fig. 1A). Cross-resistance to a range of other effect. Against the PE01CDDP line, NRG1h inhibited growth to 20% platinum analogues, such as carboplatin, iproplatin, and JM40, and of control cell number, whereas NRG1a had no effect (Fig. 1C). also certain cytotoxics, including prednimustine and melphalan, The effect of TGFa on growth over an 8-day period is shown in was observed (Fig. 1A). In addition to becoming chemoresistant, Fig. 1B. PE01 cells failed to grow in 5% DCS-containing medium, www.aacrjournals.org 6793 Cancer Res 2005; 65: (15). August 1, 2005

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 2005 American Association for Cancer Research. Cancer Research whereas the PE01CDDP cells achieved log-phase growth under these Differential gene expression in PE01CDDP cells. Differential conditions. Addition of TGFa (0.1 mmol/L) markedly stimulated display RT-PCR and microarray analysis were used to identify the growth of PE01 cells but reduced the cell counts of PE01CDDP differences in gene expression between PE01 and PE01CDDP cells. cells relative to untreated controls after 4 days exposure. Comparisons were made in the presence and absence of TGFa Detachment of TGFa-treated PE01CDDP cells was observed at this (1 nmol/L). Differential display RT-PCR identified a number of and later time points (Fig. 2A) and increased apoptosis was shown mRNAs that were differentially expressed when run on a 6% by an increase in Annexin V positivity (Fig. 2B). sequencing gel. Bands of interest were excised, amplified by PCR, Altered response is not associated with erbB receptor sequenced, and BLAST searched. Putative changes were then expression or activation. Immunofluorescence antibody mea- confirmed by quantitative RT-PCR. Expression of mRNAs for TCP1 surement of the EGFR, erbB2, and erbB3 by FACscan indicated that and PCNA were increased over 4-fold in PE01CDDP cells relative both cell lines have similar levels of each receptor (Fig. 3A). to PE01, whereas expression of ZXDA and SLP1 were decreased by Western blot analysis and mRNA transcript expression levels >4-fold (Fig. 5A). (normalized to g-actin) also indicated no major differences in Differential mRNA expression was also analyzed using receptor expression between these cell lines (data not shown). the Clontech Atlas human Cancer 1.2 array. Of 1,185 genes Treatment with TGFa (10 nmol/L) showed that phosphorylation of analyzed, 51 increased by at least 2-fold in PE01CDDP relative EGFR and erbB2 proteins were not significantly different in to PE01, whereas 36 decreased by a similar amount (Table 1). PE01CDDP cells compared with PE01 cells (Fig. 3B). Both EGFR Confirmation of at least a 2-fold increase in mRNA levels and erbB2 were tyrosine phosphorylated by TGFa at 30 minutes in PE01CDDP cells relative to PE01 cells has been verified by and 1 hour followed by a reduced level of phosphorylation of erbB2 real-time PCR for the following genes: FRA1, FANCG/XRCC9, at 6 hours and a return to control value at 24 hours. The identity of TRAP1, AXL, MCM2, ETV4, and MT3 (Fig. 5A). In contrast, these bands had been confirmed in previous experiments using expression of the following genes was reduced in PE01CDDP antibodies specific for the EGFR and erbB2 (data not shown). A cells by over 2-fold: IGFBP3, TRAM1, and keratins 4 and 19 similar level of EGFR down-regulation was shown in both cell lines (Fig. 5A). after 2 and 5 days exposure to TGFa (Fig. 3C). Expression of these genes was further modified by TGFa Activation of extracellular signal-regulated kinase and (Fig. 5A). TGFa increased expression in PE01CDDP cells of seven phosphatidylinositol 3-kinase pathways in PE01 and of the nine genes already increased relative to PE01. In contrast, PE01CDDP cells by transforming growth factor-A. Because no TGFa tended to decrease expression in PE01 cells in 11 of these differences in the erbB receptor levels were observed between the genes (Fig. 5A). wild-type and resistant lines in either the presence or absence of Selected genes with known associations to cell proliferation were TGFa, modifications in intracellular signaling pathways were next explored further. Western analysis confirmed the increased investigated. ERK activity was measured in both cell lines at expression of FRA1, PCNA, E1AF, and TRAP1 in PE01CDDP cells intervals between 5 and 60 minutes after TGFa addition and in relative to PE01 and reduced expression of IGFBP3 (Fig. 5B). To untreated controls (Fig. 3D). PE01CDDP cells had a higher basal level investigate the roles of signaling cascades downstream of EGFR in of ERK activity compared with that found in PE01 cells. Both cell the regulation of these genes, specific signaling inhibitors were lines were stimulated to a similar extent over basal levels by added to cells with TGFa before RNA extraction. Gefitinib blocked the addition of TGFa with maximal stimulation between 5 and both TGFa-induced FRA1 expression and IGFBP3 inhibition 15 minutes and increased activity sustained over the duration of (Fig. 4C). LY294002 and PD98059 partially blocked the FRA1 the 60-minute period measured. The increased basal activation of modulations, implicating the involvement of both PI3K/Akt and ERK in PE01CDDP cells relative to PE01 cells was also observed on MEK/ERK pathways in regulating these gene responses (Fig. 4C). Western blot analysis and this is consistent with an enhanced Similar data to that obtained for FRA1 was obtained with TCP1 growth rate (Fig. 4A). Whereas the EGFR inhibitor gefitinib could and MCM2 (data not shown). eliminate TGFa-stimulated ERK activation in PE01 cells, it could To explore the functionality of the genes associated positively only partially reduce its activation in PE01CDDP cells suggesting with proliferation, siRNA knockdown of expression was used. either constitutive activation of ERK- or EGFR-independent Specific siRNAs targeted to FRA1, MCM2, AXL, and TCP1 activation in the resistant line (Fig. 4A). expression selectively reduced expression of these mRNAs by Activation of PKB/Akt by 1 nmol/L TGFa for 15 minutes was f50% with minimal effects on the expression of the other targets observed in PE01 cells consistent with enhanced survival/reduced (Fig. 6A). Protein reduction for FRA1 was shown after 48 hours apoptosis but not in PE01CDDP cells (Fig. 4A). This was reversed by relative to either no treatment or with negative control siRNAs either the PI3K inhibitor LY294002 or gefitinib. (Fig. 6B). Growth was significantly reduced after targeting MCM2, Gefitinib reduced basal growth of PE01 and PE01CDDP cells and TCP1, and FRA1, but only minimal effects were obtained by was able to reverse both the TGFa-induced growth of PE01 cells targeting AXL (Fig. 6C). and the TGFa-induced growth inhibition of PE01CDDP cells Effects of transforming growth factor-A on activator (Fig. 4B). Basal growth of PE01 cells was reduced by both protein transcription complex expression. Cisplatin resistance LY294002 and the MEK inhibitor PD98059 suggesting growth has frequently been shown to be associated with changed dependency on both PI3K/Akt and MEK/ERK pathways (Fig. 4B). expression of AP-1 family members, in particular c-fos (25, 26). TGFa-stimulated growth was reduced by both LY294002 and Furthermore, because expression of FRA1 was markedly elevated in gefinitib consistent with enhanced survival. Basal growth of PE01CDDP cells relative to PE01 cells, it was of interest to compare PE01CDDP cells was again decreased by both LY294002 and FRA1 against other members of the AP-1 family. PD98059 (Fig. 4B). The TGFa reduction of growth was reversed Semiquantitative RT-PCR was done on RNA extracted from by PD98059 consistent with the MEK/ERK pathway being the two cell lines at time points between 15 minutes and 24 associated with modified growth rates/survival. hours after TGFa addition. Results showed different patterns of

Cancer Res 2005; 65: (15). August 1, 2005 6794 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 2005 American Association for Cancer Research. ErbB Receptor Signaling in Resistant Ovarian Cancer expression of AP1 complex members in the two cell lines both report, we describe changes that occur after treatment with in the presence and absence of TGFa (Fig. 7A). The most cisplatin, one of the cytotoxic agents most widely used in the dramatic difference between the two cell lines was in the treatment of ovarian cancer. Resistance to platinum-based therapy expression of FRA1 mRNA. Nonstimulated PE01 cells had occurs frequently in the course of disease management and when undetectable expression. FRA1 mRNA was induced by TGFa this occurs, relapsed tumors often respond poorly to other after 60 minutes and continued to increase up to 24 hours. cytotoxic agents as well. As such, there is a need to understand PE01CDDP cells had higher FRA1 expression under nonstimulated the signaling changes that can occur in resistant tumor cells to conditions and mRNA induction was much stronger at time better develop new approaches to their treatment. points between 60 minutes and 24 hours. Differences were also We observed three major sets of differences in this cancer cell observed in other family members; both JUN and JUND were line model after prolonged exposure to cisplatin: the develop- down-regulated by TGFa at 24 hours in PE01 relative to control, ment of platinum resistance, the ability to grow in charcoal- whereas FRA2 seemed sharply elevated at 30 minutes in PE01 stripped serum conditions, and differential growth responses to compared with PE01CDDP cells. Differences in FRA1 expression erbB activation. These differences could not be explained by were quantified by real-time PCR and are shown in Fig. 7B. changed expression of the erbB receptors or by initial upstream Because FRA1 was elevated in the cisplatin-resistant PE01CDDP signaling effects in response to ligands. However, TGFa activated line, we determined to see if FRA1 reduction (by siRNA) could the PI3K/Akt pathway in PE01 cells but not in PE01CDDP cells influence cisplatin sensitivity of this line. PE01CDDP cells were and this could be linked with reduced apoptosis observed in treated with FRA1 siRNA for 48 hours to reduce FRA1 protein these cells. Conversely, ERK activity was increased in PE01CDDP expression (cf. Fig. 6B) and treated with either 12 or 24 Amol/L cells relative to PE01 cells and this would be consistent with an cisplatin for 1 hour. Cells were plated on plastic (at equivalent enhanced rate of proliferation. Whereas the EGFR inhibitor numbers to allow for FRA1 siRNA reduction of growth) and gefitinib could completely block TGFa-stimulated ERK activation colonies counted after 14 days (Fig. 8). Reduction of FRA1 was in PE01 cells, it could only partially reduce ERK activation in associated with enhanced sensitivity to cisplatin suggesting that PE01CDDP cells. This may suggest constitutive activation of ERK- the elevated level of FRA1 expression in PE01CDDP cells was indeed or EGFR-independent activation, which could contribute to the associated with cisplatin resistance. changed functionality. There is accumulating evidence that the duration and intensity of ERK signaling is critical in determining the type of biological response (30, 31). Transient activation of Discussion ERK is generally associated with proliferation, whereas sustained The erbB receptors are major regulators of cell signaling and activation links with growth arrest (30, 31). Sustained stimulation function in ovarian cancer cells. Most published data supports the through high levels of activated Ras or Raf can similarly produce view that EGFR, erbB2, and erbB3 positively regulate cell growth growth inhibition (32, 33). and invasion, whereas erbB4 mediates an inhibitory growth Differential display and microarray analyses identified several function (16–18, 27, 28). Furthermore, the level of erbB receptor gene expression changes that can be linked to platinum resistance expression may also influence the type of response (29). In this or insensitivity (MT3, XRCC9, PCNA) or with cell proliferation and

Figure 5. Differential expression of selected differential display and microarray products in the PE01 and PE01CDDP cell lines. A, quantitative RT-PCR of selected up-regulated and down-regulated mRNAs. Expression was measured as described in Materials and Methods. Columns, means of actin-corrected values from quadruplicate samples; bars, FSD. B, Western blot analysis of selected proteins identified as being up-regulated (FRA1, PCNA, ETV4, and TRAP1) or down-regulated (IGFBP3) in the microarray and differential display experiments. Actin is shown as a loading control. www.aacrjournals.org 6795 Cancer Res 2005; 65: (15). August 1, 2005

Table 1. Genes differentially expressed between PE01 and PE01CDDP cells

c Gene symbol Gene name Accession no.* Fold difference

FRA1 Fos-related antigen; FOS-like antigen 1 (FOSL1) X16707 24.3 ETV4 ets variant gene 4 (E1A enhancer binding protein, E1AF) U18018 5.3 TOP2A Topoisomerase (DNA) II a 170 kDa J04088 5.0 CD74 CD74 antigen (invariant polypeptide of major X00497 4.1 histocompatibility complex, class II antigen associated) CKS1B CDC28 protein kinase regulatory subunit 1B X54941 3.9 DTYMK Thymidylate kinase deoxythymidylate kinase (thymidylate kinase L16991 3.6 EPR1 Effector cell protease receptor U75285 3.5 GEM GTP binding protein overexpressed in skeletal muscle U10550 3.5 MCM2 MCM2 minichromosome maintenance deficient 2, mitotin D21063 3.4 TIE1 Tyrosine kinase with immunoglobulin-like and EGF-like domains 1 X60957; S89716 3.4 HIST1H4B Histone 1, H4b X67081 3.1 IGFBP6 Insulin-like growth factor binding protein 6 M62402 3.0 FANCG Fanconi anemia, complementation group G U70310 3.0 PCDHGC3 Protocadherin g subfamily C, 3 L11373 3.0 MT3 Metallothionein 3 [growth inhibitory factor (neurotrophic)] D13365; M93311 3.0 AXL AXL receptor tyrosine kinase M76125 2.9 TRAP1 TNF receptor-associated protein 1 U12595 2.9 TNFSF7 tumor necrosis factor (ligand) superfamily, member 7 L08096; S69339 2.9 E2F1 E2F transcription factor 1 M96577 2.9 MYBL2 v-myb myeloblastosis viral oncogene homologue (avian)-like 2 X13293 2.8 MDM4 Mdm4, transformed 3T3 cell double minute 4, AF007111 2.8 p53 binding protein RRM2 ribonucleotide reductase M2 polypeptide X59618 2.8 AURKB aurora kinase B AF008552 2.8 PLK1 polo-like kinase 1 U01038 2.7 CDK4 cyclin-dependent kinase 4 M14505 2.6 GART Phosphoribosylglycinamide formyltransferase, X54199 2.6 phosphoribosylglycinamide synthetase, phosphoribosylaminoimidazole synthetase CKS2 CDC28 protein kinase regulatory subunit 2 X54942 2.6 CDC25B cell division cycle 25B M81934; S78187 2.5 TIMP1 tissue inhibitor of metalloproteinase 1 (erythroid potentiating X03124 2.5 activity, collagenase inhibitor) CCNA2 cyclin A2 X51688 2.5 RGS3 regulator of G protein signaling 3 U27655 2.5 MK167 antigen identified by monoclonal antibody Ki-67 X65550 2.5 RAD54L RAD54-like X97795 2.3 HLA-DRA major histocompatibility complex, class II, DR a K01171 2.3 RAD51 RAD51 homologue D13804 2.3 REA repressor of estrogen receptor activity U72511 2.2 LIG1 ligase I, DNA, ATP dependent M36067 2.2 PPM1G protein phosphatase 1G (formerly 2C), magnesium-dependent, Y13936 2.2 g isoform CCNB1 cyclin B1 M25753 2.1 BAX BCL2-associated X protein L22474 2.1 MTHFD1 Methylenetetrahydrofolate dehydrogenase (NADP+ dependent) 1, J04031 2.0 methenyltetrahydrofolate cyclohydrolase, formyltetrahydrofolate synthetase ARHGDIA Rho GDP dissociation inhibitor (GDI) a X69550 2.0 TYRO3 TYRO3 protein tyrosine kinase D17517 2.0 EGFR Epidermal growth factor receptor X00588; K03193; 2.0 X00663; U48722 NEK2 NIMA (never in mitosis gene a)–related kinase 2 U11050 2.0 ARHGDIB Rho GDP dissociation inhibitor (GDI) h L20688 2.0 SIVA CD27-binding (Siva) protein U82938 2.0 PARP1 Poly(ADP-ribose) polymerase family, member 1 M18112; J03473 2.0 GSS glutathione synthetase U34683 2.0

(Continued on the following page)

Cancer Res 2005; 65: (15). August 1, 2005 6796 www.aacrjournals.org

Table 1. Genes differentially expressed between PE01 and PE01CDDP cells (Cont’d)

c Gene symbol Gene name Accession no.* Fold difference

PSMB9 proteasome (prosome, macropain) subunit, h type, 9 Z14977 2.0 (large multifunctional protease 2) AK3 Adenylate kinase 3 X60673 2.0 KRT10 Keratin 10 (epidermolytic hyperkeratosis; keratosis M19156 0.5 palmaris et plantaris) COPS5 COP9 constitutive photomorphogenic homologue subunit 5 U65928 0.5 KRAS2 v-Ki-ras2 Kirsten rat sarcoma 2 viral oncogene homologue M54968 0.5 PPP3CC Protein phosphatase 3 (formerly 2B), catalytic subunit, g isoform S46622 0.5 (calcineurin A g) IRFD1 Interferon-related developmental regulator 1 Y10313 0.5 UGP2 UDP-glucose pyrophosphorylase 2 U27460 0.5 MMP9 matrix metalloproteinase 9 (gelatinase B, 92-kDa gelatinase, J05070; D10051 0.48 92-kDa type IV collagenase) LIF leukemia inhibitory factor (cholinergic differentiation factor) X13967; M63420 0.48 MX1 myxovirus (influenza virus) resistance 1, interferon-inducible M33882 0.48 protein p78 GBP2 guanylate binding protein 2, IFN-inducible M55543 0.45 KLF6 Kruppel-like factor 6 U44975 0.45 COL8A1 collagen, type VIII, a 1 X57527 0.45 FN1 fibronectin 1 X02761; K00799; K02273; 0.45 X00307; X00739 CREBBP CREB binding protein (Rubinstein-Taybi syndrome) U47741 0.43 ARHE ras homologue gene family, member E X95282 0.43 IL15 interleukin 15 U14407 0.43 KRT19 keratin 19 Y00503 0.43 CDH6 cadherin 6, type 2, K-cadherin (fetal kidney) D31784 0.43 LRRC17 leucine rich repeat containing 17 U32907 0.43 DSG2 desmoglein 2 Z26317 S64273 0.42 CASP4 caspase 4, apoptosis-related cysteine protease U28014 U28015 0.40 KRT4 keratin 4 X67683 0.37 SMAD2 SMAD, mothers against DPP homologue 2 U68018 0.37 KRT14 keratin 14 J00124 0.36 GADD45A growth arrest and DNA damage–inducible, a M60974 0.36 CREM cAMP responsive element modulator D14825 0.36 TRAM1 translocation associated membrane protein 1 X63679 0.33 MMP7 Matrix metalloproteinase 7 (matrilysin, uterine) X07819 0.33 IGFBP3 Insulin-like growth factor binding protein 3 M31159; M35878 0.31 KIAA0033 KIAA0033 D26067 0.29 ATF1 Activating transcription factor 1 X55544 0.25 SLK STE20-like kinase D86959 0.25 UGT2B15 UDP glycosyltransferase 2 family, polypeptide B15 U08854; X63359; U06641; 0.25 J05428; Y00317 TNFSF10 tumor necrosis factor (ligand) superfamily, member 10 U57059 0.17 HRH1 Histamine receptor H1 D28481 0.17 XDH Xanthine dehydrogenase U39487 0.17

*GenBank accession number. cFold increase in PE01CDDP relative to PE01.

invasion (ETV4, MCM2, TCP1, IGFBP3, SLP1). Significantly, the platinum (35). The XRCC9/FANCG gene which is linked with gene showing the largest change in expression between the two cell Fancomi anemia was also increased (by 3.6-fold) in PE01CDDP cells lines, FRA1, a member of the AP-1 transcription complex family of relative to PE01 cells. Interestingly, TGFa increased expression 2- proteins, is implicated in the regulation of expression of many fold in PE01CDDP while inducing a 10-fold reduction in PE01 cells. genes and may be critical in determining the cell response to Knockout of this gene results in hypersensitivity to cross-linking external stimuli. agents; hence, increased expression might account for reduced MT3 was elevated 5.5-fold in PE01CDDP cells and this is sensitivity to cisplatin (36). PCNA was increased 5.4-fold in the consistent with previous reports associating increased expression resistant line and this is in line with data for lung cancer wherein of metallothioneins with exposure to (34) and resistance to PCNA expression was higher in cisplatin-resistant tumors (37). www.aacrjournals.org 6797 Cancer Res 2005; 65: (15). August 1, 2005

ETV4 (E1AF or PEA3) is a member of the Ets-related Several expression changes reflect the enhanced growth of transcription factor family and has been associated with both PE01CDDP cells relative to PE01. The MCM2 protein is required for rapid cell growth (38) and invasion (39) in ovarian cancer. DNA replication and cell division and has been proposed as a Platinum treatment has been reported to up-regulate ETV4 (40) proliferation marker in esophageal cancer (44), and in PE01CDDP and this change seems to have been maintained in the cells, it was found to be up-regulated. The TCP1 chaperonin is also PE01CDDP line with a 12-fold increase in expression compared up-regulated in PE01CDDP cells. Expression of this chaperonin is CDDP with PE01. We observed increased ERK activity in the PE01 particularly linked with the S to G2-M phase transition of the cell cells relative to PE01 cells and ERK is reported to regulate ETV4 cycle and again is likely to be associated with proliferation (45). (41). siRNA reduction of both these targets was associated with growth Expression of the tyrosine kinase receptor AXL, which acts as a inhibition in the PE01CDDP cell line suggesting a role for these receptor for GAS6, was increased in PE01CDDP cells. GAS6 is molecules. expressed in both PE01 and PE01CDDP cells (data not shown) and In PE01CDDP cells, IGFBP3 is dramatically down-regulated acting through AXL is reported to induce cell cycle division in (54-fold). Numerous studies have described its growth inhibitory serum-starved cells and activate the ERK pathway (42, 43). This effects either through blockade of IGF-stimulated mitogenesis may be responsible for the ability of PE01CDDP cells to thrive in and cell survival or via antiproliferative activity unrelated to its charcoal-stripped serum conditions and could also contribute to ability to bind to IGFs (43). Recently, it has also been shown in an enhanced basal level of phospho-ERK in PE01CDDP cells; breast epithelial cells that IGFBP3 can potentiate cell prolifer- however, reduction of AXL by RNA interference (RNAi) had little ation stimulated by EGF (46). This ability may assist growth effect on growth. stimulation in PE01 cells but not in PE01CDDP cells. SLP1 (or

Figure 6. Effects of specific siRNAs on target mRNA expression, protein expression and cell growth. Cells were treated as described in Materials and Methods with 100 nmol/L siRNA for 4 hours. A, quantitative RT-PCR of mRNA from cells 24 hours after siRNA treatment. Expression of FRA1, AXL, MCM2, and TCP1 mRNA was measured as described in Materials and Methods. Columns, means of actin-corrected values from quadruplicate samples; bars, FSD. B, Western analysis of FRA1 expression in cells 48 hours after siRNA treatment. Actin is shown as a loading control. C, cell counts 5 days after siRNA exposure. TGFa was added after siRNA treatment and was present throughout the 5 days. Statistical comparisons were made with the comparative control sample. *, P < 0.05 (ANOVA followed by Tukey-Kramer multiple comparisons test).

Cancer Res 2005; 65: (15). August 1, 2005 6798 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 2005 American Association for Cancer Research. ErbB Receptor Signaling in Resistant Ovarian Cancer antileukoproteinase) has been reported to be overexpressed in ovarian cancer relative to normal ovary (47, 48). Expression of this is reduced 12-fold in PE01CDDP cells relative to PE01 cells. FRA1 expression is dependent on ERK activity (49) and can also be induced by Akt (50). In untreated PE01CDDP cells, FRA1 expression was 18-fold higher compared with PE01 cells and exposure to TGFa resulted in a further 2-fold increase. FRA1 expression under both basal conditions and TGFa regulation was associated with both Akt and ERK signaling in PE01 cells but with only ERK signaling in PE01CDDP cells. The functional AP-1 complex is a dimer composed of JUN (JUN, JUNB, and JUND) and FOS (FOS FOSB, FRA1, and FRA2) family members (51). The composition of the AP-1 dimers has been shown to confer promoter specificity (51) suggesting that the choice of dimerization partner will affect both the genes up-regulated and Figure 8. Effect of FRA1 siRNA on the sensitivity of the PE01CDDP cell line to the cell response. Furthermore, it is reported that in control of cisplatin. Cells were plated in 25-cm2 flasks and treated with siRNA as previously. After 48 hours, medium was changed for fresh RPMI supplemented Cyclin D1 (and cell cycle), a system of chromatin trafficking with 10% FCS. Flasks were then treated with a range of cisplatin concentrations exists whereby early expression and recruitment of c-fos is for 1 hour (across the previously determined IC50 value). Two washes with PBS superseded by later and prolonged FRA1 recruitment on ERK-1- were done and cells were trypsinised and counted. Cells were seeded into 6-well trays at a density of 2 Â 103 cells per well. Medium was changed after 7 days and ERK-2-dependent promoter sites (52). The elevated FRA1 in and cells were left for a further 7 days. Medium was removed and cells were the PE01CDDP cell line is consistent with rapid growth of this cell fixed with acetone/methanol (50:50) for 5 minutes before staining with hematoxylin. Colonies of >50 cells were counted microscopically. Statistical line even in reduced serum conditions but the further increase comparisons were made with the comparative no siRNA control sample. we observed after treatment with TGFa may serve to promote *, P < 0.05 (ANOVA followed by Tukey-Kramer multiple comparisons test). AP-1 dimers that favor an apoptotic response. In the C6 glioma cell line, overexpression of FRA1 was shown to both inhibit growth and increase apoptosis post treatment with hexam- ethylene bisacetamide (53). The observed temporal changes in expression seen not only with FRA1 but also FRA2, JUN and JUN D suggest that in these two cell lines the AP-1 dimer composition may be different which could explain their differing responses to the same erbB ligands. FRA1 has been shown to have a role in cell motility and attachment in fibroblastoid L929 cells (54) and in the proliferation, invasiveness, and motility of breast cancer cells (55). PC-12 pheochromocytoma cells transfected with FRA1 showed a significant enhancement of proliferation and were able to proliferate in low-serum–containing medium (56). This is in agreement with our observation of a more rapid growth rate of PE01CDDP cells relative to PE01 and their ability to proliferate in charcoal-stripped conditions. Reduced expression of FRA1 following RNAi inhibited growth consistent with this view. Similarly, reduced expression of FRA1 also enhanced cisplatin sensitivity suggesting a role for FRA1 in cisplatin resistance as has been reported for FOS (25, 26). In conclusion, these results indicate that exposure to cisplatin can lead not only to the onset of cisplatin resistance but also modified signaling with consequent changes in growth regulation. In this report, we have identified a number of potential gene changes, which might be significant in explaining the changed phenotype. In particular, we have highlighted differences in the AP-1 transcription factor family members, which may contribute to differential gene regulation and responses within this model system. Further studies are required to explore the prevalence of these changes in primary ovarian cancers. Figure 7. RT-PCR expression analysis of AP-1 family members in the PE01 and PE01CDDP cell lines after addition of TGFa (1 nmol/L). A, gel RT-PCR of AP-1 family members after addition of 1 nmol/L TGFa. RT-PCR analysis was performed as described in Materials and Methods. Expression levels of FOS Acknowledgments and FOSB were undetectable under these conditions. B, quantitative assessment of FRA1 expression in the PE01 and PE01CDDP cell lines after Received 7/28/2004; revised 4/29/2005; accepted 5/23/2005. addition of TGFa (1 nmol/L). Columns, means of actin-corrected values from The costs of publication of this article were defrayed in part by the payment of page quadruplicate samples; bars, FSD. Statistical comparisons were made with the charges. This article must therefore be hereby marked advertisement in accordance comparative control sample. *, P < 0.05 (ANOVA followed by Tukey-Kramer with 18 U.S.C. Section 1734 solely to indicate this fact. multiple comparisons test). We thank Astrazeneca for supplies of gefitinib. www.aacrjournals.org 6799 Cancer Res 2005; 65: (15). August 1, 2005

References 20. Fujimura M, Hidaka T, Saito S. Selective inhibition of correlation to invasive potential. Gynecol Oncol 2000; the epidermal growth factor receptor by ZD1839 79:256–63. 1. McGuire WP, Markham M. Primary ovarian cancer decreases the growth and invasion of ovarian clear cell 39. Kaya M, Yoshida K, Higashino F, Mitaka T, Ishii S, chemotherapy: current standards of care. Br J Cancer adenocarcinoma cells. Clin Cancer Res 2002;8:2448–54. Fujinaga K. A single ets-related transcription factor, 2003;89:S3–8. 21. Finkler N, Gordon A, Crozier M, et al. Phase 2 E1AF, confers invasive phenotype on human cancer 2. Benhar M, Engelberg D, Levitzki A. Cisplatin-induced evaluation of OSI-774, a potent oral antagonist of the cells. Oncogene 1996;12:221–7. activation of the EGF receptor. Oncogene 2002;21: EGFR-TK in patients with advanced ovarian carcinoma. 40. Funaoka K, Shindoh M, Yoshida K, et al. Activation of 8723–31. Proc Am Soc Clin Oncol 2001;20:208a(abs 831). the p21(Waf1/Cip1) promoter by the ets oncogene 3. Christen RD, Hom DK, Porter DC, et al. Epidermal 22. Wolf CR, Hayward IP, Lawrie SS, et al. Cellular family transcription factor E1AF. Biochem Biophys Res growth factor regulates the in vitro sensitivity of human heterogeneity and drug resistance in two ovarian Commun 1997;236:79–82. ovarian carcinoma cells to cisplatin. J Clin Invest 1990; adenocarcinoma cell lines derived from a single patient. 41. Ito E, Sweterlitsch LA, Tran PB, Rauscher FJ III, 86:1632–40. Int J Cancer 1987;39:695–702. Naryanan R. Inhibition of PC-12 cell differentiation 4. Cenni B, Aebi S, Nehme A, Christen RD. Epidermal 23. Brotherick I, Shenton BK, Angus B, Waite IS, Horne by the immediate early gene fra-1. Oncogene 1990;5: growth factor enhances cisplatin-induced apoptosis by a CHW, Lennard TWJ. A flow cytometric study of c-erbB-3 1755–60. caspase 3 independent pathway. Cancer Chemother expression in breast cancer. Cancer Immunol Immun- 42. Goruppi S, Ruaro E, Schneider C. Gas6, the ligand of Pharmacol 2001;47:397–403. other 1995;41:280–6. Axl tyrosine kinase receptor, has mitogenic and survival 5. Frankel A, Mills GB. Peptide and lipid growth factors 24. Mullen P, McPhillips F, MacLeod K, Monia B, Smyth activities for serum starved NIH3T3 fibroblasts. Onco- decrease cis-diamminedichloroplatinum-induced cell JF, Langdon SP. Antisense oligonucleotide targeting of gene 1996;12:471–80. death in human ovarian cancer cells. Clin Cancer Res Raf-1:importance of Raf-1 mRNA expression levels and 43. Allen MP, Zeng C, Schneider K, et al. Growth arrest- 1996;2:1307–13. Raf-1 dependent signaling in determining growth specific gene 6 (Gas6)/adhesion related kinase (Ark) 6. Yarden Y, Slikowski MX. Untangling the ErbB signal- response in ovarian cancer. Clin Cancer Res 2004;10: signaling promotes gonadotropin-releasing hormone ling network. Nat Rev Mol Cell Biol 2001;2:127–37. 2100–8. neuronal survival via extracellular signal-regulated 7. Olayioye MA, Neve RM, Lane NE, Hynes NE. The ErbB 25. Moorehead RA, Singh G. Influence of the proto- kinase (ERK) and Akt. Mol Endocrinol 1999;13:191–201. signaling network: receptor heterodimerization in de- oncogene c-fos on cisplatin sensitivity. Biochem Phar- 44. Kato H, Miyazaki T, Fukai Y, et al. A new proliferation velopment and cancer. EMBO J 2000;19:3159–67. macol 2000;59:337–45. marker, minichromosome maintenance protein 2, is 8. Egan SE, Weinberg RA. The pathway to signal 26. Scanlon KJ, Jiao L, Funato T, et al. Ribozyme- associated with tumor aggressiveness in esophageal achievement. Nature 1993;365:781–2. mediated cleavage of c-fos mRNA reduces gene squamous cell carcinoma. J Surg Oncol 2003;84:24–30. 9. Verbeek BS, Adriaansen-Slot SS, Vroom TM, Beckers T, expression of DNA synthesis enzymes and metallothio- 45. Dittmar G, Schmidt G, Kopun M, Werner D. Mapping Rijksen G. Overexpression of EGFR and c-erbB2 causes nein. Proc Natl Acad Sci U S A 1991;88:10591–5. of G2/M-phase prevalences of chaperon-encoding tran- enhanced cell migration in human breast cancer cells 27. Gilmour LM, Macleod KG, McCaig A, Gullick WJ, scripts by means of a sensitive differential hybridization and NIH3T3 fibroblasts. FEBS Lett 1998;425:145–50. Smyth JF, Langdon SP. Expression of erbB-4/HER-4 approach. Cell Biol Int 1997;21:383–91. 10. Chen P, Xie H, Sekar MC, Gupta K, Wells A. growth factor receptor isoforms in ovarian cancer. 46. Martin JL, Weenink SM, Baxter RC. Insulin-like Epidermal growth factor receptor-mediated cell motil- Cancer Res 2001;61:2169–76. growth factor-binding protein-3 potentiates epidermal ity: phospholipase C activity is required, but mitogen- 28. Gilmour LM, Macleod KG, McCaig A, et al. Neu- growth factor action in MCF-10A mammary epithelial activated protein kinase activity is not sufficient for regulin expression, function, and signaling in human cells. Involvement of p44/42 and p38 mitogen-activated induced cell movement. J Cell Biol 1994;127:847–57. ovarian cancer cells. Clin Cancer Res 2002;8:3933–42. protein kinases. J Biol Chem 2003;278:2969–76. 11. Yarden Y. The EGFR family and its ligands in human 29. Xu F, Yu Y, Le XF, Boyer C, Mills GB, Bast RC Jr. The 47. Shigemasa K, Tanimoto H, Underwood LJ, et al. cancer. Signalling mechanisms and therapeutic oppor- outcome of heregulin-induced activation of ovarian Expression of the protease inhibitor antileukoprotease tunities. Eur J Cancer 2001;37:S3–8. cancer cells depends on the relative levels of HER-2 and and the serine protease stratum corneum chymotryptic 12. Bartlett JMS, Langdon SP, Simpson BJB, et al. The HER-3 expression. Clin Cancer Res 1999;5:3653–60. enzyme (SCCE) is coordinated in ovarian tumors. Int J prognostic value of epidermal growth factor receptor 30. Marshall CJ. Specificity of receptor tyrosine kinase Gynecol Cancer 2001;11:454–61. mRNA expression in primary ovarian cancer. Br J signaling: transient versus sustained extracellular signal- 48. O’Hagan RC, Tozer RG, Symons M, McCormick F, Cancer 1996;73:301–6. regulated kinase activation. Cell 1995;80:179–85. Hassell JA. The activity of the Ets transcription factor 13. Berchuck A, Kamel A, Whitaker R, et al. Over- 31. Cowley S, Paterson H, Kemp P, Marshall CJ. PEA3 is regulated by two distinct MAPK cascades. expression of HER-2/neu is associated with poor Activation of MAP kinase kinase is necessary and Oncogene 1996;13:1323–33. survival in advanced epithelial ovarian cancer. Cancer sufficient for PC12 differentiation and for transforma- 49. Tiwari G, Sakaue H, Pollack JR, Roth RA. Gene Res 1990;50:4087–91. tion of NIH 3T3 cells. Cell 1994;77:841–52. expression profiling in prostate cancer cells with Akt 14. Berchuck A, Rodriguez GC, Kamel A, et al. Epidermal 32. Sewing A, Wiseman B, Lloyd AC, Land H. High- activation reveals Fra-1 as an Akt inducible gene. Mol growth factor receptor expression in normal ovarian intensity Raf signal causes cell cycle arrest mediated by Cancer Res 2003;6:475–84. epithelium and ovarian cancer. I. Correlation of receptor p21Cip1. Mol Cell Biol 1997;9:5588–97. 50. Vial E, Marshall C. Elevated ERK-MAP kinase activity expression with prognostic factors in patients with 33. Woods D, Parry D, Cherwinski H, Bosch E, Lees E, protects the FOS family member FRA-1 against ovarian cancer. Am J Obstet Gynecol 1991;164:669–74. McMahon M. Raf-induced proliferation or cell cycle proteasomal degradation in colon carcinoma cells. 15. Slamon DJ, Godolphin W, Jones LA, et al. Studies of arrest is determined by the level of Raf activity with J Cell Sci 2003;116:4957–63. the HER-2/neu proto-oncogene in human breast and arrest mediated by p21Cip1. Mol Cell Biol 1997;17: 51. Bakiri L, Matsuo K, Wisniewska M, Wagner EF, Yaniv ovarian cancer. Science 1989;244:707–12. 5598–611. M. Promoter specificity and biological activity of 16. Morishige K, Kurachi H, Amemiya K, et al. Evidence 34. Ferguson PJ. Mechanisms of resistance of human tethered AP-1 dimers. Mol Cell Biol 2002;13:4952–64. for the involvement of transforming growth factor a and tumours to anticancer drugs of the platinum family: a 52. Burch PM, Yuan Z, Loonen A, Heintz NH. An epidermal growth factor receptor autocrine growth review. J Otolaryngol 1995;24:242–52. extracellular signal-related kinase 1- and 2-dependent mechanism in primary human ovarian cancers in vitro. 35. Koo MS, Kwo YG, Park JH, Choi WJ, Billiar TR, Kim program of chromatin trafficking of c-Fos and Fra-1 is Cancer Res 1991;51:5322–8. YM. Signaling and function of caspase and c-jun N- required for cyclin D1 expression during cell cycle 17. Scambia G, Benedetti-Panici P, Battaglia F, terminal kinase in cisplatin-induced apoptosis. Mol reentry. Mol Cell Biol 2004;11:4696–709. Ferrandina G, Gaggini C, Mancuso S. Presence of epi- Cells 2002;13:194–201. 53. Shirsat NV, Shaikh SA. Overexpression of the dermal growth factor (EGF) receptor and proliferative 36. Koomen M, Cheng NC, van de Vrugt HJ, et al. immediate early gene Fra-1 inhibits proliferation, response to EGF in six human ovarian carcinoma cell Reduced fertility and hypersensitivity to mitomycin C induces apoptosis, and reduces tumourigenicity of c6 lines. Int J Gynecol Cancer 1991;1:253–8. characterize Fancg/Xrcc9 null mice. Hum Mol Genet glioma cells. Exp Cell Res 2003;291:91–100. 18. Crew AJ, Langdon SP, Miller EP, Miller WR. Mitogenic 2002;11:273–81. 54. Tkach V, Tulchinsky E, Lukanidin E, Vinson C, effects of epidermal growth factor and transforming 37. Ogawa J, Iwazaki M, Inoue H, Koide S, Shohtsu A. Bock E, Berezin V. Role of the Fos family members, growth factor-a on EGF-receptor positive human Immunohistochemical study of glutathione-related c-Fos, Fra-1 and Fra-2 in the regulation of cell motility. ovarian carcinoma cell lines. Eur J Cancer 1992;28: enzymes and proliferative antigens in lung cancer. Oncogene 2003;22:5045–54. 337–41. Relation to cisplatin sensitivity. Cancer 1993;71: 55. Belguise K, Kersual N, Galtier F, Chalbos D. FRA-1 19. Sewell JM, Macleod KG, Ritchie A, Smyth JF, Langdon 2204–29. expression level regulates proliferation and invasiveness SP. Targeting the EGF receptor in ovarian cancer with 38. Nishikawa A, Iwasaki M, Akutagawa N, et al. of breast cancer cells. Oncogene 2005;24:1434–44. the tyrosine kinase inhibitor ZD 1839 (‘‘Iressa’’). Br J Expression of various matrix proteases and Ets family 56. Shaulian E, Karin M. AP-1 in cell proliferation and Cancer 2002;86:456–62. transcriptional factors in ovarian cancer cell lines: survival. Oncogene 2001;20:2390–2400.

Cancer Res 2005; 65: (15). August 1, 2005 6800 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 2005 American Association for Cancer Research. Altered ErbB Receptor Signaling and Gene Expression in Cisplatin-Resistant Ovarian Cancer

Kenneth Macleod, Peter Mullen, Jane Sewell, et al.

Cancer Res 2005;65:6789-6800.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/65/15/6789

Cited articles This article cites 51 articles, 15 of which you can access for free at: http://cancerres.aacrjournals.org/content/65/15/6789.full#ref-list-1

Citing articles This article has been cited by 14 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/65/15/6789.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://cancerres.aacrjournals.org/content/65/15/6789. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.