Gene Expression Profile in Response to Doxorubicin−Rapamycin Combined Treatment of HER-2−Overexpressing Human Mammary Epithelial Cell Lines

Published OnlineFirst November 14, 2011; DOI: 10.1158/1535-7163.MCT-11-0033

Molecular Cancer Preclinical Development Therapeutics

Gene Expression Proﬁle in Response to Doxorubicin– Rapamycin Combined Treatment of HER-2–Overexpressing Human Mammary Epithelial Cell Lines

Adriana Priscila Trape1, Maria Lucia Hirata Katayama1, Rosimeire Aparecida Roela1, Helena Brentani2, Graziela Rosa Ravacci3, Leandro de Araujo Lima4, and Maria Mitzi Brentani1

Abstract HER-2–positive breast cancers frequently sustain elevated AKT/mTOR signaling, which has been associated with resistance to doxorubicin treatment. Here, we investigated whether rapamycin, an mTOR inhibitor, increased the sensitivity to doxorubicin therapy in two HER-2–overexpressing cell lines: C5.2, which was derived from the parental HB4a by transfection with HER-2 and SKBR3, which exhibits HER-2 ampliﬁcation. The epithelial mammary cell line HB4a was also analyzed. The combined treatment using 20 nmol/L of rapamycin

and 30 nmol/L of doxorubicin arrested HB4a and C5.2 cells in S to G2–M, whereas SKBR3 cells showed an

increase in the G0–G1 phase. Rapamycin increased the sensitivity to doxorubicin in HER-2–overexpressing cells by approximately 2-fold, suggesting that the combination displayed a more effective antiproliferative action. Gene expression profiling showed that these results might reflect alterations in genes involved in canonical pathways related to purine metabolism, oxidative phosphorylation, protein ubiquitination, and mitochondrial dysfunction. A set of 122 genes modulated by the combined treatment and specifically related to HER-2 overexpression was determined by finding genes commonly regulated in both C5.2 and SKBR3 that were not affected in HB4a cells. Network analysis of this particular set showed a smaller subgroup of genes in which coexpression pattern in HB4a cells was disrupted in C5.2 and SKBR3. Altogether, our data showed a subset of genes that might be more robust than individual markers in predicting the response of HER-2–overexpressing breast cancers to doxorubicin and rapamycin combination. Mol Cancer Ther; 11(2); 464–74. 2011 AACR.

Introduction because it has been described as conferring either sensitivity (3–5), relative resistance (6–8), or exerting no influences HER-2/neu, a member of the ErbB receptors, is amplified (9) on anthracycline-based regimes, including doxorubicin. and/or overexpressed in about 25% to 30% of human breast Doxorubicin intercalates in the DNA double strand and cancers and generally correlates with poor prognosis (1). forms a stable complex with the enzyme topoisomerase II Although it has been used in the clinical practice as a (TOP2A), the direct target of anthracyclines. The fact that parameter to indicate those patients who would better HER-2 and TOP2A genes are frequently coamplified in benefit from trastuzumab treatment, almost all HER-2– breast cancers led some researchers to speculate the pre- positive tumors develop resistance to this monotherapy dictive value of these simultaneous alterations (10–12). (2). Controversial data arise out of the predictor role of One mechanism of resistance to doxorubicin-induced this receptor in response to conventional chemotherapy, apoptosis in HER-2–positive tumors seems to be the ability of this receptor to activate several cell survival Authors' Affiliations: 1Departamento de Radiologia e Oncologia, LIM24, pathways, such as the phosphoinositide 3-kinase (PI3K)/ Faculdade de Medicina, Universidade de Sao~ Paulo; 2Departamento de Akt/mTOR signaling, which can induce cell proliferation, Psiquiatria, Faculdade de Medicina, Universidade de Sao~ Paulo; 3Depar- tamento de Gastroenterologia, LIM35, Faculdade de Medicina, Universi- transformation, and increased cell motility (13). More- dade de Sao~ Paulo; and 4Laboratorio de Bioinformatica e Bioestatstica, over, microarray analysis of breast tumor biopsies Centro Internacional de Pesquisa e Ensino (CIPE) - Hospital AC Camargo, showed significant correlation between TOP2A and genes Sao~ Paulo, Brazil involved in mTOR pathway (14). Note: Supplementary material for this article is available at Molecular Considering the role of PI3K/AKT/mTOR signaling in Cancer Therapeutics Online (http://mct.aacrjournals.org/). the acquisition of resistance to the antitumoral effects of Corresponding Author: Maria Mitzi Brentani, Faculdade de Medicina da chemotherapy, it seems reasonable to investigate whether Universidade de Sao~ Paulo, Departamento de Radiologia e Oncologia, Avenida Dr. Arnaldo, 455, 4 andar, sala 4115, CEP: 01246-903, Sao~ Paulo, blockade of this pathway could increase the response of SP, Brazil. Phone: 55-11-3082-6580; Fax: 55-11-3082-6580; E-mail: breast cancer cells to the antitumoral effects of doxorubi- [email protected] cin. Previously published data obtained from different doi: 10.1158/1535-7163.MCT-11-0033 cancer models have shown optimized results after the 2011 American Association for Cancer Research. combination of several chemotherapeutic agents with

464 Mol Cancer Ther; 11(2) February 2012

Genes Affected by Rapamycin–Doxorubicin in Mammary Cells

PI3K/AKT/mTOR inhibitors, such as rapamycin (14–15). Treatments A more recent study showed clinical benefits from met- Cells were seeded in 25 cm2 flasks or 96-well culture astatic breast cancer patients treated with trastuzumab plates at a density of 5 103 cells/cm2 to 15 103 cells/cm2. associated with everolimus, a rapamycin analog (16). Flow cytometry and luminescence assays were con- Rapamycin inhibits specifically the mTOR protein, ducted after 24 and 72 hours of treatment, respectively. leading to dephosphorylation of its main downstream Cells were treated with DMSO (0.04% v/v), 20 nmol/L targets, p70S6K and 4EBP1, which results in the suppres- of rapamycin, and 30 to 1,000 nmol/L of doxorubicin. sion of translation of mRNAs related to cell cycle and The combined treatment was carried out by adding a proliferation (17). Despite preclinical studies have fixed dose of rapamycin (20 nmol/L) to each doxoru- described favorable effects of rapamycin on different bicin dose. Microarray experiments and Western blot- breast cancer models, clinical trials showed only a modest ting were carried out after 24 hours of treatment with antitumoral activity in patients, which may be attributed 30 nmol/L of doxorubicin and 20 nmol/L of rapamycin, to the fact that this drug is essentially cytostatic and not doses clinically achieved in plasma of patients under- cytotoxic as chemotherapeutic agents (18, 19). going these therapies (22, 23). Altogether, these evidences highlight the relevance of combining rapamycin with doxorubicin to improve the Flow cytometry and cell viability assays response of HER-2–positive tumors to both drugs. Nev- Cell-cycle distribution was determined by a FACS- ertheless, the molecular mechanisms involved in this Calibur flow cytometer (Beckton Dickinson) after PI treatment need to be elucidated to provide insights about incorporation. Cell viability was measured with Cell the biological processes modulated by the combination. Titer Glo Reagent. GraphPad (version 5.0) software Therefore, we aimed to investigate whether rapamycin was used for statistical analyses, carried out by ANOVA could increase the sensitivity of HER-2–expressing breast and Tukey post test. Dose–response curves were con- epithelial cell lines to doxorubicin. Gene expression pro- structed by nonlinear regression analyses (P < 0.05). file generated from this approach might lead to a better understanding about the molecular mechanisms associ- Western blotting ated with the response of HER-2-overexpressing tumors Proteins were extracted in lysis buffer (50 mmol/L to the combined treatment. Tris-HCl,pH7.4;150mmol/LNaCl;1%NP-40;0.5% sodium deoxycholate; 0.1% SDS; and protease inhibi- Materials and Methods tors) and the concentration was determined using the Bradford Protein Assay Reagent. Proteins (40 mg) were Reagents boiled for 5 minutes in Laemli buffer, separated by 12% Rapamycin (Calbiochem) and doxorubicin hydrochlo- SDS-PAGE, and transferred to nitrocellulose mem- ride (Sigma-Aldrich) were solubilized in dimethyl sulf- branes, which were blocked for 1 hour in 5% skimmed oxide (DMSO) and stored at 20C. Propidium iodide (PI) milk in TBS-T (0.05% Tween-20) and incubated with was purchased from Sigma-Aldrich and Cell Titer Glo monoclonal antibodies (1:500) against human p-AKT was purchased from Promega. Chemical structures of (Thr308), p-p70S6K (Thr389), p-4EBP1 (Thr37/46) and rapamycin and doxorubicin are presented in Supplemen- total proteins Akt, p70S6K, and 4EBP1 (Cell Signaling tary Fig. S1. Technology) for 16 hours at 4C. Following incubation with antirabbit immunoglobulin (1:3,000), the results Cell culture were detected with ECL Western blotting (GE Health- The HB4a and C5.2 cell lines were kindly provided in care) and quantitative densitometric analysis of the 2006 by Dr. Michael O’Hare from Ludwig Institute for bands was conducted by Scion Image Software. Cancer Research, London, United Kingdom. The SKBR3 cells from the American Type Culture Collection were Microarray procedures kindly provided in 2009 by Dr. A.A. Camargo from Lud- Total RNA was isolated with TRIzol reagent (Invi- wig Institute for Cancer Research, Sao~ Paulo Branch, trogen) and purified by the RNeasy Mini Kit (QIA- Brazil. The C5.2 cells were established by transfection GEN). The RNA quality and purity were analyzed by of parental mammary luminal epithelial HB4a cells with NanoDrop ND-1000 (Thermo Scientific) and Agilent cDNA full-length corresponding to HER-2 gene (20). HB4a 2100 Bioanalyzer (Agilent). Only total RNA samples expresses basal levels of HER-2 receptor, and C5.2 with 260/280 ratios between 1.9 and 2.0 and RNA expresses high levels of this receptor similarly to SKBR3. integrity number more than 7 were further processed. The cells were cultured as described elsewhere (21), tested Total RNA (250 ng) from samples were amplified by the periodically for Mycoplasm contamination, and authenti- GeneChip 30IVT Expression Kit Assay and hybridized cated by quantitative real-time reverse transcriptase (RT)- to Human Genome U133 Plus 2.0 GeneChips. After PCR to evaluate HER-2 overexpression in C5.2 and SKBR3 hybridization, genechips were scanned with a Gene- compared with HB4a cells (data not shown). The 3 cell Chip Scanner 3000 7G and data were recovered using lines express EGFR (21) and are negative for estrogen Expression Console Software (Affymetrix). All experi- receptor and progesterone receptor (data not shown). ments presented the indexes of quality control as

www.aacrjournals.org Mol Cancer Ther; 11(2) February 2012 465

Trape et al.

specified in the Affymetrix guidelines for assessing from 26.8% in control cells to 74.6% after exposure to sample and array quality. a 30 nmol/L dose. Increased doses of doxorubicin showed no significant alterations in this cell-cycle dis- Gene expression analysis tribution up to 250 nmol/L dose (P > 0.05). Neverthe- The data were normalized with RMA Express Release less, higher concentrations (500–1,000 nmol/L) resulted 1.0 software. Statistical analyses were conducted by Multi- in depletion of cells in S to G2–M, which could be Experiment Viewer v.4.5 (Mev) software. Differentially explained by an increase in the percentage of cells in P < expressed genes were identified after comparing each G0–G1 and in sub-G1 ( 0.05; Fig. 1A). The C5.2 cells treatment to the control cells by SAM method using seemed to be more resistant to the effects of doxorubicin criteria based on FDR less than 5%. The genes were on S to G2–M arrest, because these cells displayed 50% assigned to canonical pathways and functions by Inge- increase in S to G2–M only after exposure to 120 nmol/L nuity Pathway Analysis software (P < 0.05). dose (P < 0.01). Although these results were maintained up to 500 nmol/L of doxorubicin, the 1,000 nmol/L dose Network construction reduced the percentage of cells in the S to G2–M phase, P < Network analysis was carried out for the 122 genes with a concomitant increase in G1 phase ( 0.01; related to HER-2–mediated response to the combined Fig. 2A). The ability of doxorubicin to induce S to treatment, based on the interactome database (human G2–M arrest was less prominent in SKBR3 cells com- protein data base, HPRD; ref. 24). For each cell line pared with the others. The maximum increase in the (HB4a, C5.2, and SKBR3), we selected the shortest path- percentage of cells in S to G2–M phase consisted in 30% ways between 2 nodes and the first neighbors. Next, the and occurred after treating cells with a 120 nmol/L dose R software (25) was used to calculate Pearson correla- (P < 0.001; Fig. 3A). We observed no substantial cell tion coefficient (PCC) among normalized expression death as shown by absence of peaks in sub-G1 phase for values of genes in each sample. PCC values of both C5.2 and SKBR3, even at higher doses (P > 0.05). C5.2 and SKBR3 cells were compared with HB4a The effects of the combined treatment on cell-cycle cells and we selected for further investigation only those distribution were similar to those generated by doxo- gene pairs that displayed differences greater than 1.5. rubicin alone in HB4a and C5.2 cells, because the Data analysis and visualization were conducted by increase in S to G2–M also occurred after exposure Cytoscape software (version 2.8.0; www.cytoscape.org). to 30 nmol/L doxorubicin and was only 10% smaller after the combined treatment than doxorubicin alone Quantitative real-time RT-PCR (Figs. 1B and 2B). These effects were attenuated after Validation experiments were carried out for 11 out of increasing the doses of the chemotherapeutic agent. 122 genes related to HER-2–mediated response to the In SKBR3 cells; the response to the combined treat- combined treatment. Samples of C5.2 and SKBR3 cells ment was different from that with doxorubicin alone. were reverse transcribed with 2 mg of total RNA, random Rapamycin seemed to cause an increase of 10% of primers (Promega Corporation), and Superscript III cells in G0–G1 phase with a concomitant decrease in Reverse Transcriptase (Invitrogen Corporation). Primers StoG2–M after its addition to 30 nmol/L of doxoru- were designed with Primer-Blast software (http://www. bicin (P < 0.05). Increased doxorubicin doses resulted ncbi.nlm.nih.gov/tools/primer-blast/) and synthesized in a cell-cycle distribution similar to that of control cells by IDT (Integrated DNA Technologies). Real-time RT PCR (P > 0.05; Fig. 3B). reactions were carried out in duplicate using SYBR Green Although 20 nmol/L of rapamycin alone caused no cell- PCR Power MasterMix in the 7900HT Fast Real Time cycle alterations in HB4a cells compared with the control System (Applied Biosystems) with the following thermal (P > 0.05), C5.2 and SKBR3 cells displayed 12% and 15% cycling: 95 C (10 minutes) and 40 cycles of 95 C (15 sec- increase, respectively, in the percentage of cells in G0–G1 onds) and 60C (1 minute). b-Actin (ACTB) gene expres- (P < 0.01; Figs. 1C, 2C, and 3C). These increases were sion was used as endogenous control, and a pool of C5.2 accompanied by a depletion of cells accumulated in S to and SKBR3 cells was used as calibrator for calculating G2–M. Increased doses of rapamycin caused no changes in DDC relative expression of target genes with 2 t method. cell-cycle distribution for all cell lines studied (data not Statistical analyses were carried out with Student’s t test. shown).

Results Effects of doxorubicin and rapamycin on cell growth Doxorubicin inhibited cell growth of the 3 cell lines in a Effects of doxorubicin and rapamycin on cell-cycle dose-dependent manner (Fig. 4). The results showed that distribution 30 nmol/L doxorubicin caused a 35% inhibition in cell Theeffectsofdoxorubicinandrapamycinoncell- growth of HB4a and C5.2 cells, reaching 90% to 100% after cycle distribution were determined by ﬂow cytometry. exposure to doses ranging from 250 nmol/L to 1,000 The HB4a cells showed a remarkable 50% increase in nmol/L (P < 0.01). According to the ﬂow cytometric P < cells situated in S to G2–M ( 0.001) by doxorubicin, results, SKBR3 cells seemed to be more resistant to the because the percentage of cells in this phase shifted antiproliferative effects of doxorubicin, which consisted

466 Mol Cancer Ther; 11(2) February 2012 Molecular Cancer Therapeutics

Genes Affected by Rapamycin–Doxorubicin in Mammary Cells

Figure 1. Cell-cycle distribution of HB4a cells treated with varying doses of doxorubicin (30–1,000 nmol/L) and/or a ﬁxed dose of rapamycin (20 nmol/L). Representative DNA histograms of cells treated with doxorubicin alone (A), combination of both drugs (B), and rapamycin alone (C). Results are expressed as mean of 3 independent assays. Values of SEM were less than 5% (data not shown).

in 23% of growth inhibition at a 30 nmol/L dose (P < 0.01). Impact of rapamycin, doxorubicin, and the Escalating doses of doxorubicin caused 77% of cell growth combination of these two drugs on the inhibition, reaching up to 90% at 1,000 nmol/L. Rapamy- phosphorylation level of AKT, p70S6K, and 4EBP1 cin caused 35% growth inhibition in HB4a cells, 47% in Determination of the impact of rapamycin, doxoru- C5.2 cells, and 40% in SKBR3 cells (P < 0.001; data not bicin, and their combination on activation of the mTOR shown). pathway was carried out in HB4a, C5.2, and SKBR3 The addition of rapamycin was 20% more effective in cells. We analyzed the effects of each treatment on inhibiting cell growth than doxorubicin alone, parti- phosphorylation levels of AKT protein (upstream reg- cularly after adding rapamycin to 30 to 120 nmol/L ulator), p70S6K, and 4EBP1 (downstream targets of doxorubicin for HB4a (P < 0.05) and C5.2 cells (P < 0.001) mTOR; Supplementary Fig. S3). Constitutive levels of and to 30 nmol/L for SKBR3 cells (P < 0.01). Higher phosphorylated proteins were observed for control concentrations of doxorubicin had similar effects to samples of all cell lines and, compared with the others, those exerted by this agent alone (P > 0.05; Fig. 4). SKBR3 cells showed about 2-fold increased levels for p- Nonlinear regression analyses showed that rapamycin AKT and p-4EBP1 proteins. After 24 hours of treatment, was able to reduce about 2-fold the IC50 of doxorubicin Western blotting analysis detected no changes in the (P < 0.0001), which changed from 54.2, 56.3, and 88.2 levels of p-AKT by each drug individually or the com- nmol/L to 23.4, 22.9, and 47.1 nmol/L for HB4a, C5.2, bined therapy in HB4a and SKBR3 cells. On the con- and SKBR3, respectively (Supplementary Fig. S2). trary, C5.2 cells showed increased levels of p-AKT only Moreover, the higher IC50 values of SKBR3 cells sug- in response to the combined treatment. The levels of gested that these cells were more resistant to doxoru- p-p70S6K were signiﬁcantly decreased by all treatments bicin than the other cell lines. in the 3 cells, mainly after treatment with rapamycin

www.aacrjournals.org Mol Cancer Ther; 11(2) February 2012 467

Trape et al.

Figure 2. Cell-cycle distribution of C5.2 cells treated with varying doses of doxorubicin (30–1,000 nmol/L) and/or a ﬁxed dose of rapamycin (20 nmol/L). Representative DNA histograms of cells treated with doxorubicin alone (A), combination of both drugs (B), and rapamycin alone (C). Results are expressed as mean of 3 independent assays. Values of SEM were less than 5% (data not shown).

and its association with doxorubicin. Only SKBR3 cells cells (n ¼ 518 genes) compared with HB4a (n ¼ 190) and displayed decreased levels of p-4EBP1 by rapamycin C5.2 (n ¼ 188). and combination therapies. Genes deregulated by the combined treatment were predominantly upregulated in HB4a cells (1,439 up vs. Gene expression proﬁling 821 down), downregulated in C5.2 cells (509 up vs. 1,203 Gene expression proﬁles were obtained for HB4a, C5.2, down), and altered in both directions in SKBR3 (926 and SKBR3 cells treated with doxorubicin, rapamycin, up vs. 1,109 down). Altogether, up- and downregulated and the association of both drugs. Figure 5 is a Venn genes showed a high percentage (40%–50%) of genes diagram showing that doxorubicin consistently upregu- exclusively regulated by the combined treatment of lated global gene expression and this response was more HB4a cells (1,129 of 2,260), C5.2 cells (656 of 1,712), and remarkable in HB4a cells (n ¼ 1,453) than C5.2 (n ¼ 468) SKBR3 (1,023 of 2,035). Overlapping genes between each and SKBR3 (n ¼ 174). Repressed genes by doxorubicin drug alone and the combined treatment showed that treatment were also observed predominantly in HB4a the contribution of rapamycin and doxorubicin on gene cells (n ¼ 231) compared with C5.2 (n ¼ 12) and SKBR3 expression consisted mainly in decrease by the former (n ¼ 11). and upregulation by the latter. Unlike doxorubicin, rapamycin substantially downregulated global gene expression and these effects occurred Canonical pathways modulated by the combined predominantly in C5.2 (n ¼ 1,589) and SKBR3 (n ¼ 867) treatment compared with HB4a cells (n ¼ 167). The induction of gene To better understand the molecular mechanisms expression by rapamycin occurred prevalently in SKBR3 involved in the response to the association between the

468 Mol Cancer Ther; 11(2) February 2012 Molecular Cancer Therapeutics

Genes Affected by Rapamycin–Doxorubicin in Mammary Cells

Figure 3. Cell-cycle distribution of SKBR3 cells treated with varying doses of doxorubicin (30–1,000 nmol/L) and/or a ﬁxed dose of rapamycin (20 nmol/L). Representative DNA histograms of cells treated with doxorubicin alone (A), combination of both drugs (B), and rapamycin alone (C). Results are expressed as mean of 3 independent assays. Values of SEM were less than 5% (data not shown).

2 drugs, differentially expressed genes by the combined Genes upregulated fell into the following most signiﬁ- treatment were classiﬁed in canonical pathways. These cant pathways: PI3K signaling in B lymphocytes, p53 annotations included up- and downregulated genes and ATM signaling, G2–M DNA damage checkpoint representing 3 gene sets: (i) overlapped with doxorubicin regulation, antigen presentation, and NF-kB signaling. alone, (ii) overlapped with rapamycin alone, and (iii) Some pathways in C5.2 cells were also regulated by the exclusively regulated by the combined treatment. The combination in the parental HB4a, such as glycolysis. complete list of genes referred to the canonical pathways Nevertheless, we decided to focus on canonical pathways found in each cell line is available in Supplementary commonly regulated in both C5.2 and SKBR3 and not Tables S1A, S1B, S2A, S2B, S3A, and S3B. evidenced in HB4a cells to identify those related to the Ingenuity Pathway Analysis of the transcripts differ- HER-2–mediated response to the combined treatment. entially regulated in C5.2 cells by the combined treatment Among these pathways, we found PI3K signaling in B revealed that downregulated genes were involved in lymphocytes, mitochondrial dysfunction, protein ubiqui- pathways such as aminoacyl-tRNA biosynthesis, glycol- tination, oxidative phosphorylation, cyclins and cell-cycle ysis/gluconeogenesis, pyrimidine and purine metabo- regulation, ubiquinone biosynthesis, and citrate cycle, lism, cell-cycle control of chromosomal replication, mito- which, therefore, might underlie some molecular mechan- chondrial dysfunction, and oxidative phosphorylation. isms of this response (Supplementary Fig. S4A).

www.aacrjournals.org Mol Cancer Ther; 11(2) February 2012 469

Trape et al.

cells, which showed that 27% (139 of 509) of the upregulated and 65% (780 of 1,203) of the downregulated genes by the combined treatment were found only in C52 compared with HB4a cells. Our next step was to identify which of these genes were also regulated in SKBR3, which resulted in a set containing 122 genes (16 up- and 106 downregulated; Supplementary Table S4). Although most of these genes have seemed not to exert known roles in canonical pathways, they were functionally anno- tated according to Gene Ontology database (Supplemen- tary Table S5).

Experimental validation by real-time RT PCR Among the 122 genes corresponding to HER-2–mediated response to the combined treatment, we selected 11 genes to be validated by real-time RT-PCR (4 up- and 7 downregulated; Supplementary Table S6). We observed validation of 6 genes (ALDOC, ELOVL6, PGHDH, TIPIN,

Figure 4. HB4a, C5.2, and SKBR3 cell lines were treated with varying doses of doxorubicin alone (30–1,000 nmol/L) and associated with a ﬁxed dose of rapamycin (20 nmol/L) for 72 hours and growth inhibition was determined by luminescence assays. Results are expressed as percentage of the control. Experiments were conducted in 3 different days. Error bars represent the values of SEM. Dox, doxorubicin; Dox þ Rapa, combined treatment.

Genes exclusively regulated by the combined treatment might play relevant roles in the HER-2–mediated response as they were altered irrespective of the action of each drug alone (Supplementary Fig. S4B). According- ly, a signiﬁcant percentage of genes in the canonical pathways related to HER-2 overexpression belonged to this category, mainly in SKBR3 cells, which displayed 60% to 80% of genes represented in citrate cycle, mitochondrial dysfunction, and oxidative phosphorylation as exclusively regulated by the combination therapy.

Genes commonly regulated in HER2-overexpressing cell lines Considering that our results showed the relevance of HER-2 in the modulation of canonical pathways by the Figure 5. Venn diagrams of the number of differentially expressed genes combined treatment, we decided to determine a group of generated from HB4a, C5.2, and SKBR3 cell lines after exposure to genes related to this response. Firstly, we compared the doxorubicin alone (D), rapamycin alone (R) and the combination of the 2 gene expression proﬁles generated from C52 and HB4a drugs (D þ R).

470 Mol Cancer Ther; 11(2) February 2012 Molecular Cancer Therapeutics

Genes Affected by Rapamycin–Doxorubicin in Mammary Cells

WWC3 ZNF323 , and ) in both C5.2 and SKBR3 cells, rubicin-mediated S to G2–M arrest of HB4a and C5.2 cells, because their expression levels were significantly altered whereas SKBR3 cells displayed prominent effects of rapa- in the same direction of microarray experiments (Supple- mycin over doxorubicin as shown by G0–G1 arrest of these mentary Fig. S5). In addition, 2 genes were individually cells. validated in C5.2 cells (BIK and DET1) and SKBR3 cells According to previous works (29, 15), our cell growth (AMD1 and CORO1A). The gene TXNIP showed a trend to assays showed that inhibition of mTOR caused a de- P ¼ be validated in SKBR3 cells ( 0.0513). Overall, our crease of doxorubicin IC50 for all cell lines. Because HB4a results showed a validation rate of microarray results cells lack HER-2 overexpression, we expected they were from 54% (6 of 11) considering both cell lines to 72% (8 more resistant to the effects of the combined treatment of 11) for each cell line individually. compared with the HER-2–overexpressing cell lines. The ability of rapamycin to increase the chemosensitivity Network construction of cell lines depends on constitutive activation of the It was recently shown that examination of human AKT pathway (30). Indeed, our Western blotting results interactome dynamics can evidence substantial altera- showed that the HB4a cells, like C5.2 and SKBR3, had tions in its organization, besides quantitative gene expres- constitutive levels of phosphorylation of AKT and down- sion changes (26). Using the 122 aforementioned genes, stream targets of mTOR (p70S6K and 4EBP1). We sup- we constructed a gene–gene interaction network based on pose that the high levels of EGFR receptor described the human interactome to explain the molecular mechan- for HB4a cells lead to the formation of EGFR–HER3 isms whereby HER-2 might mediate the response to the heterodimers, which might trigger an alternative signal- combined treatment. After selecting the shortest path- ing cascade that activates AKT/mTOR pathway and ways among these genes, we calculated the PCC based increases the response of HB4a cells to the growth inhi- on normalized expression values in each cell line (HB4a, bitory effects of the combined treatment (21). C5.2, and SKBR3). Subsequently, we compared the PCC Our Western blotting analyses also showed a significant values of both C5.2 and SKBR3 cells with those obtained decrease of p70S6K activation after rapamycin and its for HB4a cells. association with doxorubicin for all cell lines whereas AKT Among the 7,381 gene pairs resulting from this analysis, remained activated irrespective of the treatment. These we found that 38 showed differences larger than 1.5 results might be explained, in part, by the fact that doxo- (Supplementary Table S7). Interestingly, these gene pairs rubicin has been shown to activate AKT pathway (31). allowed us to construct a network in which interactions In addition, mTOR leads to a feedback inhibition of were inverted by HER-2 overexpression, because these PI3K/AKT through activation of p70S6K so that its inhi- genes were coexpressed in response to the combined bition by rapamycin decreases p70S6K activity and, there- treatment in HB4a cells (positive PCC) and became inde- fore, abolishes this regulatory loop, resulting in AKT pendently expressed in both HER-2–overexpressing cell activation (32). Decreased phosphorylation levels of 4EBP1 lines (negative PCC; Fig. 6). The TCF19 (SC1) was a hub were observed only for SKBR3 cells undergoing rapamy- gene displaying a highly correlated pattern of coexpres- cin and combined treatment. These data do not necessarily sion with several genes and the gene C17orf95 was exclu- indicate resistance to treatment because p-4EBP1 levels sively regulated by the combined treatment. We also might not correlate with rapamycin sensitivity (33). observed that 4 genes represented in this network Gene expression profiling was carried out to better (PHGDH, TXNIP, WWC3, and CORO1A) were validated understand the processes involved in HER-2–mediated by real-time RT-PCR. response to the combined treatment. For this purpose, we identified canonical pathways commonly regulated in Discussion C5.2 and SKBR3 and not evidenced in HB4a cells. Down- regulated genes were categorized in purine metabolism, In this study, our purpose was to investigate whether cyclins and cell-cycle regulation, protein ubiquitination, rapamycin could modulate the response of 2 cell lines lysine degradation, oxidative phosphorylation, mito- harboring HER-2 overexpression or amplification to chondrial dysfunction, ubiquinone biosynthesis, and cit- doxorubicin. Immortalized breast epithelial cells obtained rate cycle. Considering that our recent report showed from normal mammary epithelium (HB4a) were also enrichment of oxidative phosphorylation for genes upre- analyzed. gulated by HER-2 overexpression (34), these data suggest Consistent with previous works, we observed that the that the pathways identified by our approach were, in fact, effects of doxorubicin on cell-cycle distribution consisted specifically related to HER-2 overexpression. of increased arrest in S to G2–M phase (27) whereas Increasing evidence has shown that protein-based rapamycin caused a G0–G1 arrest (28). Moreover, the 3 network analysis might provide a more effective means cell lines displayed differences about the response to these to identify molecular biomarkers involved in response drugs. Results of cell-cycle distribution showed that HB4a to treatments (35), because it allows to study gene–gene cells were the most sensitive to doxorubicin, whereas C5.2 interactions related to a set of biological processes instead and SKBR3 cells were more sensitive to rapamycin. The looking at a specific pathway (Fig. 6). Using more strin- combined treatment showed a preponderance of doxo- gent criteria to understand the effects of HER-2 on the

www.aacrjournals.org Mol Cancer Ther; 11(2) February 2012 471

Trape et al.

Figure 6. Network generated from genes related to HER-2–mediated response to the combined treatment and based on the human interactome. PCC calculated among normalized expression values of the genes is represented by the lines connecting genes and colors vary according to these values: the red lines correspond to PCC values close to 1 and the green lines represent PCC values close to 1. Only gene pairs related to signiﬁcant differences (>1.5) of PCC values in HB4a compared with C5.2 and SKBR3 cells were selected for network construction.

response to the combined treatment, we selected only the it presented disruption of interactions with several genes genes exclusively regulated in HER-2–overexpressing cell in response of HER-2–overexpressing cells to the com- lines for network construction. Among the 122 genes bined treatment. These data are consistent with the role analyzed, we found a subgroup of genes in which inter- of this gene as a relevant mediator of cell growth and connections in HB4a were inverted in C5.2 and SKBR3 transcription of targets required for G1–S transition and cells and, therefore, might represent a putative set of entry to S to G2–M (36). Among the genes interacting markers for HER-2–mediated response to the combined with TCF19, we found PHGDH, which encodes for treatment. 3-phosphoglycerate dehydrogenase, an important medi- Our network analyses showed the TCF19 (SC1) gene as ator of production of biosynthetic precursors required a central player (hub) of the interaction network, because for cell proliferation. The diversion of glycolysis into

472 Mol Cancer Ther; 11(2) February 2012 Molecular Cancer Therapeutics

Genes Affected by Rapamycin–Doxorubicin in Mammary Cells

serine biosynthetic pathway by this enzyme leads to sion and chromosome segregation required for cell accumulation of serine and a-ketoglutarate essential for growth (44). the synthesis of proteins, nucleotides, carbohydrates, We also noticed genes encoding for enzymes involved and lipids that occurs through citrate cycle in mitochon- in energy homeostase, such as CKB (creatine kinase) dria (37). and FKBP11 (peptidyl prolylisomerase), with the latter In addition to its role in these biosynthetic processes, described as inhibited by rapamycin (45). mitochondria also supplies 40% to 75% of ATP require- ments in cancer cells, which indicates that therapeutic Conclusions interventions targeting this organelle might contribute to mitochondrial dysfunction and could lead to loss of cell Our results suggest that HER-2–mediated response to viability. Indeed, other genes represented in the network the combined treatment might lead to alterations of path- encode for mitochondrial components, such as mitochon- ways related to mitochondrial metabolism. A protein drial ribosomal protein (MRPS7) and translocases network analysis of a smaller set of genes corresponding (TIMM50, TIMM10, and TOMM40L), which recognize to this response showed a loss of coordinated coexpres- and mediate the import of nucleus-encoded mitochondri- sion of genes related to oxidative phosphorylation, citrate al precursor proteins (38). cycle, mitochondrial proteins, and DNA metabolism, Altogether, these data suggest that the combined which might contribute to the impairment of processes treatment might affect the mitochondria, which is sup- relevant for increased cell survival mediated by HER-2. ported by previous reports unraveling the coordinated We suppose that, compared with strategies based on relationship between mTOR and mitochondrial metab- individual gene markers analyses, our network might be olism. Inhibition of mTOR by rapamycin has been more robust in predicting response of HER-2–overexpres- described to decrease mitochondrial membrane poten- sing breast carcinomas to the combined treatment with tial, oxygen consumption, and ATP synthetic capacity doxorubicin and rapamycin. (39, 40). Other genes represented in the network involved Disclosure of Potential Conflicts of Interest upstream regulators of AKT/mTOR (IMPA2, CORO1A, and TXNIP). The CORO1A (41) and IMPA2 (42) genes have No potential conflicts of interest were disclosed. been described as involved in the phosphatidylinositol Acknowledgments signaling, which triggers a signaling cascade that activates the AKT protein. The gene TXNIP was related to the The authors thank Dr. Tiago Goss dos Santos for carrying out some stimulation of PTEN protein, a negative regulator of this assays, Dr. Walcy Paganelli R. Teodoro for the data interpretation, Dr. AnaMaria A. Camargo for providing SKBR3 cell lines, Mrs Maria Jose pathway, which indicates a relevant role of this gene in the Goncalves¸ Benevides for secretarial help, and Mrs. Cristina Pineiro~ attenuation of AKT/mTOR effects on cell growth (43). Grandal for figure edition. Other genes in which interactions with TCF19 were disrupted play functions related to DNA metabolism Grant Support (ORC1L, GINS2, and DTYMK). The DTYMK gene encodes for thymidylate kinase, a key enzyme for dTTPs supply in This work was supported by FAPESP and CnPq. The costs of publication of this article were defrayed in part by the DNA synthesis of cancer cells. Knockdown of GINS2 was payment of page charges. This article must therefore be hereby marked reported to lead to growth inhibition and polyploidy by advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. the suppression of M-phase progression in human breast cancer cells, showing that besides its roles in DNA rep- Received January 26, 2011; revised October 10, 2011; accepted October lication initiation, this gene might also facilitate cell divi- 26, 2011; published OnlineFirst November 14, 2011.

References 1. Goldhirsch A, Ingle JN, Gelber RD, Coates AS, Thurlimann€ B, Senn HJ 5. Rody A, Karn T, Gatje€ R, Ahr A, Solbach C, Kourtis K, et al. Gene Panel members. Thresholds for therapies: highlights of the St Gallen expression profiling of breast cancer patients treated with docetaxel, International Expert Consensus on the primary therapy of early breast doxorubicin, and cyclophosphamide within the GEPARTRIO trial: cancer 2009. Ann Oncol 2009;20:1319–29. HER-2, but not topoisomerase II alpha and microtubule-associated 2. Slamon DJ, Leyland-Jones B, Shak S, Fuchs H, Paton V, Bajamonde A, protein tau, is highly predictive of tumor response. Breast 2007;16: et al. Use of chemotherapy plus a monoclonal antibody against HER-2 86–93. for metastatic breast cancer that overexpresses HER-2. N Engl J Med 6. Dressler LG, Berry DA, Broadwater G, Cowan D, Cox K, Griffin S, et al. 2001;344:783–92. Comparison of HER-2 status by fluorescence in situ hybridization and 3. Gennari A, Sormani MP, Pronzato P, Puntoni M, Colozza M, Pfeffer U, immunohistochemistry to predict benefit from dose escalation of et al. HER-2 status and efficacy of adjuvant anthracyclines in early adjuvant doxorubicin-based therapy in node-positive breast cancer breast cancer: a pooled analysis of randomized trials. J Natl Cancer patients. J Clin Oncol 2005;23:4287–97. Inst 2008;100:14–20. 7. Vargas-Roig LM, Gago FE, Tello O, Martin de Civetta MT, Ciocca DR. 4. Pritchard KI, Shepherd LE, O'Malley FP, Andrulis IL, Tu D, Bramwell c-ErbB2 (HER-2/neu) protein and drug resistance in breast cancer VH, et al. HER-2 and responsiveness of breast cancer to chemother- patients treated with induction chemotherapy. Int J Cancer 1998; apy. N Engl J Med 2006;54:2103–11. 2:129–34.

www.aacrjournals.org Mol Cancer Ther; 11(2) February 2012 473

Trape et al.

8. DiGiovanna MP, Stern DF, Edgerton S, Broadwater G, Dressler LG, 26. Taylor IW, Linding R, Warde-Farley D, Liu Y, Pesquita C, Faria D, et al. Budman DR, et al. Influence of activation state of ErbB-2 (HER-2) on Dynamic modularity in protein interaction networks predicts breast response to adjuvant cyclophosphamide, doxorubicin, and fluoroura- cancer outcome. Nat Biotechnol 2009;27:199–204. cil for stage II, node-positive breast cancer: study 8541 from the 27. Wang YA, Johnson SK, Brown BL, McCarragher LM, Al-Sakkaf K, Cancer and Leukemia Group B. J Clin Oncol 2008;26:2364–72. Royds JA, et al. Enhanced anti-cancer effect of a phosphatidylinositol- 9. Arpino G, Ciocca DR, Weiss H, Allred DC, Daguerre P, Vargas-Roig L, 3 kinase inhibitor and doxorubicin on human breast epithelial cell lines et al. Predictive value of apoptosis, proliferation, HER-2, and topo- with different p53 and oestrogen receptor status. Int J Cancer isomerase IIalpha for anthracycline chemotherapy in locally advanced 2008;123:1536–44. breast cancer. Breast Cancer Res Treat 2005;1:69–75. 28. Noh WC, Mondesire WH, Peng J, Jian W, Zhang H, Dong J, et al. 10. Jarvinen TAH, Tanner Tanner M, Rantanen V, Barlund M, Borg A, Determinants of rapamycin sensitivity in breast cancer cells. Clin Grenman S, et al. Amplification and deletion of topoisomerase IIa Cancer Res 2004;10:1013–23. associate with ErbB-2 amplification and affect sensitivity to topoisom- 29. Sokolosky ML, Stadelman KM, Chappell WH, Abrams SL, Martelli AM, erase II inhibitor doxorubicin in breast cancer. Am J Pathol 2000;3: Stivala F, et al. Involvement of Akt-1 and mTOR in sensitivity of breast 839–47. cancer to targeted therapy. Oncotarget 2011;2:538–50. 11. Glynn RW, Miller N, Whelan MC, Kerin MJ. Topoisomerase 2 alpha and 30. Wendel HG, Malina A, Zhao Z, Zender L, Kogan SC, Cordon-Cardo the case for individualized breast cancer therapy. Ann Surg Oncol C, et al. Determinants of sensitivity and resistance to rapamycin- 2010;17:1392–7. chemotherapy drug combinations in vivo. Cancer Res 2006;66: 12. Oakman C, Moretti E, Galardi F, Santarpia L, Di Leo A. The role of 7639–46. topoisomerase IIalpha and HER-2 in predicting sensitivity to anthra- 31. Li X, Lu Y, Liang K, Liu B, Fan Z. Differential responses to doxorubicin- cyclines in breast cancer patients. Cancer Treat Rev 2009;35:662–7. induced phosphorylation and activation of Akt in human breast cancer 13. Miller TW, Forbes JT, Shah C, Wyatt SK, Manning HC, Olivares MG, cells. Breast Cancer Res 2005;7:R589–97. et al. Inhibition of mammalian target of rapamycin is required for 32. O'Reilly KE, Rojo F, She QB, Solit D, Mills GB, Smith D. mTOR inhibition optimal antitumor effect of HER-2 inhibitors against HER-2-overex- induces upstream receptor tyrosine kinase signaling and activates Akt. pressing cancer cells. Clin Cancer Res 2009;15:7266–76. Cancer Res 2006;66:1500–8. 14. Mondesire WH, Jian W, Zhang H, Ensor J, Hung MC, Mills GB, et al. 33. Meric-Bernstam F, Gonzalez-AnguloAM.TargetingthemTOR Targeting mammalian target of rapamycin synergistically enhances signaling network for cancer therapy. J Clin Oncol 2009;27: chemotherapy-induced cytotoxicity in breast cancer cells. Clin Cancer 2278–87. Res 2004;10:7031–42. 34. Carraro DM, Ferreira EN, de Campos Molina G, Puga RD, Abrantes EF, 15. Steelman LS, Navolanic PM, Sokolosky ML, Taylor JR, Lehmann BD, Trape AP, et al. Poly (A)þ transcriptome assessment of ERBB2- Chappell WH, et al. Suppression of PTEN function increases breast induced alterations in breast cell lines. PLoS One 2011;6:e21022. cancer chemotherapeutic drug resistance while conferring sensitivity 35. Chuang HY, Lee E, Liu YT, Lee D, Ideker T. Network-based classifi- to mTOR inhibitors. Oncogene 2008;27:4086–95. cation of breast cancer metastasis. Mol Syst Biol 2007;3:140. 16. Morrow PK, Wulf GM, Ensor J, Booser DJ, Moore JA, Flores PR, et al. 36. Ku DH, Chang CD, Koniecki J, Cannizzaro LA, Boghosian-Sell L, Alder Phase I/II study of trastuzumab in combination with everolimus H, et al. A new growth-regulated complementary DNA with the (rad001) in patients with her2-overexpressing metastatic breast cancer sequence of a putative trans-activating factor. Cell Growth Differ who progressed on trastuzumab-based therapy. J Clin Oncol 1991;2:179–86. 2011;10:3126–32. 37. Possemato R, Marks KM, Shaul YD, Pacold ME, Kim D, Birsoy K, et al. 17. Guertin DA, Sabatini DM. Defining the role of mTOR in cancer. Cancer Functional genomics reveal that the serine synthesis pathway is Cell 2007;12:9–22. essential in breast cancer. Nature 2011;476:346–50. 18. Chan S, Scheulen ME, Johnston S, Mross K, Cardoso F, Dittrich C, 38. Schmidt O, Pfanner N, Meisinger C. Mitochondrial protein import: from et al. Phase II study of temsirolimus (CCI-779), a novel inhibitor of proteomics to functional mechanisms. Nat Rev Mol Cell Biol mTOR, in heavily pretreated patients with locally advanced or meta- 2010;11:655–67. static breast cancer. J Clin Oncol 2005;23:5314–22. 39. Schieke SM, Finkel T. Mitochondrial signaling, TOR, and life span. Biol 19. Hernandez-Aya LF, Gonzalez-Angulo AM. Targeting the phosphati- Chem 2006;387:1357–61. dylinositol 3-kinase signaling pathway in breast cancer. Oncologist 40. Schieke SM, Ma M, Cao L, McCoy JP Jr, Liu C, Hensel NF, et al. 2011;16:404–14. Mitochondrial metabolism modulates differentiation and teratoma 20. Harris RA, Eichholtz TJ, Hiles ID, Page MJ, O'Hare MJ. New model of formation capacity in mouse embryonic stem cells. J Biol Chem ErbB-2 over-expression in human mammary luminal epithelial cells. Int 2008;283:28506–12. J Cancer 1999;80:477–84. 41. Mueller P, Massner J, Jayachandran R, Combaluzier B, Albrecht I, 21. Timms JF, White SL, O'Hare MJ, Waterfield MD. Effects of ErbB-2 Gatfield J, et al. Regulation of T cell survival through coronin-1- overexpression on mitogenic signalling and cell cycle progression in mediated generation of inositol-1,4,5-trisphosphate and calcium human breast luminal epithelial cells. Oncogene 2002;21:6573–86. mobilization after T cell receptor triggering. Nat Immunol 2008; 22. Benjamin RS, Riggs CE Jr, Bachur NR. Plasma pharmacokinetics of 9:424–31. adriamycin and its metabolites in humans with normal hepatic and 42. Berridge MJ, Downes CP, Hanley MR. Neural and developmental renal function. Cancer Res 1977;37:1416–20. actions of lithium: a unifying hypothesis. Cell 1989;59:411–9. 23. Raymond E, Alexandre J, Faivre S, Vera K, Materman E, Boni J, et al. 43. Hui ST, Andres AM, Miller AK, Spann NJ, Potter DW, Post NM, et al. Safety and pharmacokinetics of escalated doses of weekly intrave- Txnip balances metabolic and growth signaling via PTEN disulfide nous infusion of CCI-779, a novel mTOR inhibitor, in patients with reduction. Proc Natl Acad Sci U S A 2008;105:3921–6. cancer. J Clin Oncol 2004;15:2336–47. 44. Rantala JK, Edgren H, Lehtinen L, Wolf M, Kleivi K, Vollan HK, et al. 24. Prasad TSK, Goel R, Kandasamy K, Keerthikumar S, Kumar S, Mathi- Integrative functional genomics analysis of sustained polyploidy phe- vanan S, et al. Human Protein Reference Database - 2009 update. notypes in breast cancer cells identifies an oncogenic profile for Nucleic Acids Res 2009;37:D767–D72. GINS2. Neoplasia 2010;12:877–88. 25. R Development Core Team (2009). R: A language and environment for 45. Rulten SL, Kinloch RA, Tateossian H, Robinson C, Gettins L, Kay JE. statistical computing. R Foundation for Statistical Computing, Vienna, The human FK506-binding proteins: characterization of human Austria. ISBN 3-900051-07-0, URL http://www.R-project.org. FKBP19. Mamm Genome 2006;17:322–31.

474 Mol Cancer Ther; 11(2) February 2012 Molecular Cancer Therapeutics

Gene Expression Profile in Response to Doxorubicin−Rapamycin Combined Treatment of HER-2−Overexpressing Human Mammary Epithelial Cell Lines

Adriana Priscila Trapé, Maria Lucia Hirata Katayama, Rosimeire Aparecida Roela, et al.

Mol Cancer Ther 2012;11:464-474. Published OnlineFirst November 14, 2011.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-11-0033

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2011/11/11/1535-7163.MCT-11-0033.DC1

Cited articles This article cites 44 articles, 15 of which you can access for free at: http://mct.aacrjournals.org/content/11/2/464.full#ref-list-1

Citing articles This article has been cited by 1 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/11/2/464.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 Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/11/2/464. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.