Gynecologic Oncology 138 (2015) 647–655

Contents lists available at ScienceDirect

Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno

RUNX3 contributes to carboplatin resistance in epithelial ovarian cancer cells Samir H. Barghout b,1, Nubia Zepeda a,1, Krista Vincent a, Abul K. Azad a, Zhihua Xu a,b, Christine Yang a, Helen Steed a,b, Lynne-Marie Postovit a,b, YangXin Fu a,b,⁎ a b

Department of Oncology, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, AB, Canada Department of Obstetrics and Gynecology, Faculty of Medicine and Dentistry, University of Alberta, Edmonton, AB, Canada

H I G H L I G H T S • RUNX3 expression is elevated in platinum-resistant ovarian cancer cells and tissues • RUNX3 overexpression renders ovarian cancer cells more resistant to carboplatin • RUNX3 inhibition sensitizes platinum-resistant ovarian cancer cells to carboplatin

a r t i c l e

i n f o

Article history: Received 16 May 2015 Received in revised form 2 July 2015 Accepted 8 July 2015 Available online 15 July 2015 Keywords: Epithelial ovarian cancer Chemoresistance RUNX3 Carboplatin cIAP2

a b s t r a c t Objective. Resistance to platinum-based therapeutic agents represents a major hurdle in the treatment of epithelial ovarian cancer (EOC). There is an urgent need to better understand the underlying mechanisms. Here, we investigated the role of RUNX3 in carboplatin resistance in EOC cells. Methods. Expression of RUNX3 was determined in human EOC cell line A2780s (cisplatin-sensitive) and A2780cp (cisplatin-resistant), human ovarian surface epithelium (OSE) and primary EOC cells. The effects of RUNX3 expression on sensitivity to carboplatin were determined in A2780s and A2780cp cells using neutral red uptake and clonogenic assays. Carboplatin-induced apoptosis was determined by measuring cleaved PARP using Western blotting. The expression of cellular inhibitor of apoptosis protein-2 (cIAP2) and its regulation by RUNX3 were assessed by quantitative RT-PCR and Western blotting. Results. The expression of RUNX3 was elevated in A2780cp cells compared to A2780s cells and in EOC tissues from chemoresistant patients compared to those from chemosensitive patients. Overexpression of RUNX3 rendered A2780s cells more resistant to carboplatin, whereas inhibition of RUNX3 increased sensitivity to carboplatin in A2780cp cells. Inhibition of RUNX3 potentiated carboplatin-induced apoptosis in A2780cp cells as demonstrated by more pronounced PARP cleavage. Interestingly, the expression of cIAP2 was elevated in A2780cp cells compared to A2780s cells. Overexpression of RUNX3 increased cIAP2 expression in A2780s cells, whereas inhibition of RUNX3 decreased cIAP2 expression and potentiated carboplatin-induced decrease of cIAP2 in A2780cp cells. Conclusions. RUNX3 contributes to carboplatin resistance in EOC cells and may hold promise as a therapeutic target to treat EOC and/or a biomarker to predict chemoresistance. © 2015 Elsevier Inc. All rights reserved.

1. Introduction Ovarian cancer is the leading cause of death due to gynecologic malignancies in the United States [1]. Epithelial ovarian cancer (EOC), which comprises approximately 90% of ovarian cancer, is believed to arise from the ovarian surface epithelium (OSE) [2] or fallopian tube ⁎ Corresponding author at: Department of Oncology, University of Alberta, 5124M, Katz Building, 114th St & 87th Ave, Edmonton, AB, T6G 2E1, Canada. E-mail address: [email protected] (Y. Fu). 1 These authors equally contributed to the paper.

http://dx.doi.org/10.1016/j.ygyno.2015.07.009 0090-8258/© 2015 Elsevier Inc. All rights reserved.

fimbria [2,3]. Approximately 75% of ovarian cancers are diagnosed at advanced stages and require a combination of cytoreductive surgery and chemotherapy. Platinum-based compounds have been used to treat EOC, with carboplatin currently being used as a first-line therapeutic agent, in combination with paclitaxel. Despite the initial positive response to the first-line treatment, relapse occurs in most cases and chemoresistance eventually develops [4]. Current chemotherapy regimens are ineffective against platinum-resistant EOC, resulting in a 5year survival rate of only 30% [4]. There is an urgent need to better understand the mechanisms of chemoresistance in order to develop novel therapeutic strategies to treat EOC.

648

S.H. Barghout et al. / Gynecologic Oncology 138 (2015) 647–655

Platinum-based chemotherapy has been the mainstay of EOC management for several decades [5]. Although a number of platinumbased compounds have been developed, cisplatin and carboplatin remain to be the major agents in this therapeutic class. These two agents share similar modes of action and a broad range of cross-resistance [6]. They exert their cytotoxic effects primarily through the formation of DNA adducts that are detected by DNA damage recognition machinery. As a result, multiple signaling pathways are activated, such as cell cycle checkpoints, p53 and MAP kinases, leading ultimately to cell death [7,8]. A plethora of pharmacokinetic and pharmacodynamic mechanisms have been implicated in platinum resistance that include decreased platinum uptake, conjugation and inactivation by glutathione or metallothioneins, enhanced DNA damage repair, and activation of survival signaling pathways, resulting in tolerance to DNA damage and reduced apoptosis [6–8]. Additionally, the intra- and inter-tumor heterogeneity further complicates the management of platinum-resistant EOC [6]. The RUNX family of transcription factors, RUNX1-3, are essential regulators of cell fate during development and are implicated in various human diseases [9]. The three RUNX proteins share an evolutionarilyconserved 128 amino acid runt domain which is responsible for DNA binding and dimerization with a common cofactor named CBFβ (corebinding factor β) or PEBP2β (polyomavirus enhancer binding protein 2β). CBFβ does not interact with DNA itself, but enhances DNA binding and stability of RUNX proteins [10]. All three RUNX proteins are implicated in cancer [9]. RUNX3, which regulates gene expression in a tissue-specific manner, plays either a tumor suppressive or oncogenic role in a cancer-specific manner [9]. For example, RUNX3 has been shown to be a tumor suppressor in gastric cancer [11] and in several other malignancies [12]. However, in basal cell carcinoma (BCC) [13], head and neck squamous cell carcinoma (HNSCC) [14], and EOC [15, 16], RUNX3 is shown to play an oncogenic function. Specifically, Nevadunsky et al. and Lee et al. reported that the expression of RUNX3 is increased in EOC samples and overexpression of RUNX3 promotes proliferation of ovarian cancer cells [15,16]. The other two RUNX proteins, RUNX1 and RUNX2, have also been shown to act in an oncogenic manner in EOC [17,18] and knockdown of CBFβ leads to growth inhibition of EOC cells [19]. Taken together, these studies provide evidence that all three RUNX proteins play an oncogenic role in EOC. RUNX3 is implicated in chemoresistance in several types of cancer, including hepatocellular carcinoma, gastric cancer, and lung adenocarcinoma [20–22]; down-regulation of RUNX3 expression in these cancer cells is associated with increased resistance to chemotherapy [20–22]. It appears that the loss of RUNX3 contributes to chemoresistance of cancers where RUNX3 acts as a tumor suppressor. Given the oncogenic role played by RUNX3 in EOC, we speculate that the elevated expression of RUNX3 may contribute to chemoresistance of EOC. In this study, we demonstrate that expression of RUNX3 is elevated in carboplatin resistant EOC cells and tissues, and that the elevated expression of RUNX3 contributes to carboplatin resistance of chemoresistant EOC cell line A2780cp cells. 2. Materials and methods 2.1. Reagents Puromycin, neutral red dye, and β-actin antibody were purchased from Sigma-Aldrich. The protein inhibitor cocktail was purchased from Roche. Antibodies against RUNX3 and tubulin were purchased from Abcam. Antibodies against RUNX1, RUNX2, PARP, cleaved PARP, and cIAP2 were purchased from Cell Signaling Technology. 2.2. Culture of cell lines Human ovarian cancer cell lines A2780s and A2780cp cells were cultured in DMEM/F12 medium. HEK 293T and Phoenix-Ampho cells were cultured in DMEM medium. All culture media were supplemented

with 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 μg/ml streptomycin. The cisplatin-resistant A2780cp cells were derived from cisplatin-sensitive A2780s cells by exposing A2780s cells to stepwiseincreasing concentrations of cisplatin [23]. The paired A2780s and A2780cp cells were provided by Dr. Benjamin Tsang (Ottawa Hospital Research Institute). 2.3. Isolation and culture of human ovarian surface epithelial cells and primary EOC cells Human ovarian surface epithelial (OSE) cells were obtained by scraping the surface of human ovaries surgically removed due to benign gynecologic diseases and primary EOC cells were isolated from ascites of EOC patients by following the protocol described by Shepherd et al. with minor modifications [24]. The isolation of OSE and primary EOC cells was performed in a tissue culture laminar flow hood and the isolated cells were cultured in M199/MCDB105 medium supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 μg/ml streptomycin. Institutional approval for research with human materials was received prior to the initiation of these studies (Health Research Ethics Board of Alberta-Cancer Committee, #25132), and samples were obtained after receiving an informed consent. For OSE cell isolation, we placed the ovarian tissue samples with the OSE-side down into 2.5 ml pre-warmed 0.25% trypsin/EDTA in a well of a 6-well tissue culture dish and incubated the tissues in an incubator for 30 min. The dish was swirled every 10 min. After the incubation, we transferred the ovarian tissues to another well of the 6-well dish containing pre-warmed M199/MCDB105 medium and gently scraped the surface of the ovary to collect OSE cells into the medium with a sterile scalpel blade. We then removed the tissue and cultured the scraped OSE cells for 3 days prior to the first medium change. When cells became confluent, we split them into two 6 cm dishes. After the cells in the 6 cm dishes became confluent, one dish was used to prepare cell lysates and the other dish for RNA. If we noticed fibroblast contamination in the culture, we discarded the cells and excluded the sample from the study. For EOC isolation, we added 50 ml ascites of EOC patients and 50 ml M199/MCDB105 medium to each T-125 flask with 0.2 μm vented cap. After gentle mix, we cultured the cells in an incubator undisturbed for 3 days prior to the first medium change. We then changed medium every 3 days until cells became confluent. We then froze numerous vials of cells as passage 0 stocks in liquid N2. A portion of cells was reseeded as passage 1 cells for experiments. In this study, all EOC cells were used at passage 1. 2.4. Generating overexpression and knockdown cells Phoenix-Ampho cells were transfected with an empty retroviral vector (MSCVpac) or a retroviral vector containing the human RUNX3 cDNA (MSCVpac-hRUNX3) by the calcium phosphate method as previously described [25]. A medium collected from the transfected PhoenixAmpho cells was used to infect A2780s cells. Positively infected cells were selected with 1 μg/ml of puromycin to generate A2780s/Vector and A2780s/RUNX3 cells. 293T cells were transfected with a lentivirus vector (pLentiLox) containing an shRNA targeted against a random sequence (shRandom: 5′-GTT GCT TGC CAC GTC CTA GAT-3′) or an shRNA targeted against the RUNX3 gene (shRUNX3A: 5′-GGA CCC TAA CAA CCT TCA AGA-3′ or shRUNX3B: 5′-GCC GTC TCA TCC CAT ACT TCT-3′) and packaging plasmids by the calcium phosphate method as previously described [25]. A2780cp cells were infected with lentivirus containing shRandom, shRUNX3A or shRUNX3B. Positively infected cells were purified by fluorescence-activated cell sorting (FACS) for green fluorescent protein (GFP) positive cells; the lentiviral vector contains a cytomegalovirus (CMV)-driven GFP, to generate A2780cp/ shRandom, A2780cp/shRUNX3Aand A2780cp/shRUNX3B cells.

S.H. Barghout et al. / Gynecologic Oncology 138 (2015) 647–655

649

cells seeded in these wells × PE) × 100. Cell viability was expressed as a percentage relative to the respective untreated controls.

A2780cp cells were stably transfected with pcDNA3.1 vector or pcDNA-FLAG-RUNX3 (1–187) (kindly provided by Dr. Yoshiaki Ito, Cancer Science Institute of Singapore) using the Novagen GeneJuice® Transfection Reagent and selected with 500 μg/ml of G418 (Invitrogen). pcDNA-FLAG-RUNX3 (1–187) expresses a truncated form of RUNX3 that contains the runt domain but lacks the transactivation domain at the carboxyl terminus. RUNX3 (1–187) functions as a dominant negative form of RUNX3 [26,27] and was therefore referred as to dominant negative RUNX3 (dnRUNX3) in this study.

RNA isolation, reverse transcription (RT) and quantitative RT-PCR were performed as described previously [31]. PCR primer sequences are available upon request.

2.5. Cytotoxicity assays

2.7. Preparation of whole cell lysates and cytosolic and nuclear fractions, and Western blotting

Cytotoxicity of carboplatin was determined by the neutral red uptake assay as we previously described [28] and the clonogenic assay adapted from protocols described by Munshi et al. and Franken et al. [29,30]. For the neutral red uptake assay, cells were treated with increasing concentrations of carboplatin for 72 h and cell viability was determined as previously described [28]. For clonogenic assay, cells were seeded in 6-well plates at a plating density starting from 50 cells per well for control wells (untreated) up to 6400 cells per well for wells treated with the highest concentration of carboplatin. According to the protocol [29], cells should be treated once they attach and before they start replication to avoid cell number change. Since A2780s and A2780cp attach and grow rapidly, we treated the cells 6 h postseeding with carboplatin for 24 h. The medium with carboplatin was then replaced with fresh medium and cells were allowed to grow for 9–11 days. The colonies formed were then gently washed with phosphate-buffered saline (PBS), fixed with methanol/acetic acid (3:1) solution and stained with crystal violet (0.5% in methanol). Colonies of ≥ 50 cells were counted and the viability was calculated using these equations: plating efficiency (PE) = count of colonies formed in control wells / number of cells seeded in control wells; percent viability = (count of colonies formed in treated wells / number of

2.6. RNA isolation and quantitative reverse transcription-PCR (qRT-PCR)

Whole cell lysates were prepared using modified radioimmunoprecipitation assay (RIPA) buffer as described previously [32]. Cytosolic and nuclear fractions were prepared as previously described [31]. Protein concentration was quantified using the DC protein assay (BioRad) and equal amount of proteins were loaded into each lane of an SDS polyacrylamide gel and transferred to nitrocellulose membrane. Immunoblotting was performed using an anti-RUNX1 (1:1000), antiRUNX2 (1:1000), anti-RUNX3 (1:1000), anti-β-actin (1:1000), antitubulin (1:1000), anti-PARP (1:1000), anti-cleaved PARP (1:1000), and anti-cIAP2 antibodies. Membranes were scanned and analyzed using an Odyssey® IR scanner and Odyssey® imaging software 3.0. 2.8. Statistical analysis Data are shown as mean ± SE of three to five independent experiments. Statistical analysis and IC50 calculation was performed using GraphPad Prism 5. Statistical significance between the two groups treated with the same concentration of carboplatin and between the two groups in the published microarray data from GEO database was determined by unpaired t-test and defined as P b 0.05.

Fig. 1. RUNX3 expression is elevated in human primary EOC cells and in cisplatin resistant EOC cells and tissues. (A) Expression of RUNX3 in primary EOC and normal OSE cells was examined by Western blotting. β-actin was used as a loading control. (B) Subcellular localization of RUNX3 in human primary EOC cells was measured by subcellular fractionation and Western blotting. Poly(ADP-ribose) polymerase (PARP) was used as the nuclear loading control. Tubulin was used as the cytosolic loading control. (C) Expression of RUNX1, RUNX2, RUNX3, and CBFβ in A2780s and A2780cp cells was examined by qRT-PCR. Data are shown as mean ± SEM of three independent experiments. *Significantly different (P b 0.05). (D) Expression of RUNX1, RUNX2, RUNX3, and CBFβ in A2780s and A2780cp cells was examined by Western blotting. β-actin was used as a loading control. (E) and (F) Published microarray data in GEO database was analyzed for RUNX3 expression and cisplatin resistance. RUNX3 expression is higher in cisplatin resistant cell line A2780cp70 compared to cisplatin sensitive counterpart A2780 cells (E) and in EOC tissues collected from 15 chemoresistant patients (recurrence within 6 months) compared to EOC tissues collected from 10 chemosensitive patients (recurrence after 30 months) (F). The boxes represent 25th and 75th percentiles with the median shown by the line bisecting the box. Points represent individual parameters and triangles denote outliers. *Significantly different (P b 0.05).

650

S.H. Barghout et al. / Gynecologic Oncology 138 (2015) 647–655

3. Results

3.3. RUNX3 overexpression in EOC cells confers resistance to carboplatin

3.1. RUNX3 is expressed in primary EOC cultures, but not in normal OSE cultures

Because carboplatin is currently more often used than cisplatin as the first-line therapeutic agent in the treatment of ovarian cancer due to its low toxicity profile when compared to cisplatin [33], we decided to use carboplatin instead of cisplatin for this study. Because A2780cp cells were generated to be cisplatin resistant, we first determined whether A2780cp cells are also resistant to carboplatin compared to A2780s cells. As expected, A2780cp cells were indeed more resistant to carboplatin (Fig. 2). The IC50 for carboplatin is 10.1 fold higher in A2780cp cells than that in A2780s cells (150.8 μM vs. 13.6 μM) as determined by the neutral red uptake assay and 8.6 fold higher in A2780cp cells than that in A2780s cells (35.5 μM vs. 3.7 μM) as determined by the clonogenic assay. Both assays validate the use of these paired cell lines to investigate carboplatin resistance. To determine whether elevated expression of RUNX3 is associated with carboplatin resistance, we stably overexpressed RUNX3 in A2780s cells to generate A2780s/Vector and A2780s/RUNX3 cells (Fig. 3A). These cells were treated with increasing concentrations of carboplatin for 72 h. Neutral red uptake assay showed that A2780s/ RUNX3 cells were more resistant to carboplatin compared to A2780s/ Vector; RUNX3 expression increased the IC50 for carboplatin from 10.9 μM to 19.9 μM (Fig. 3B). The clonogenic assay also showed that A2780s/RUNX3 cells were significantly more resistant to carboplatininduced cytotoxicity than A2780s/Vector cells (Fig. 3C); RUNX3 expression increased the IC50 for carboplatin from 3.6 μM to 7.9 μM. These results indicate that RUNX3 overexpression confers carboplatin resistance to A2780s cells.

To confirm the elevated expression of RUNX3 in EOC reported in the two previous studies [15,16], we examined the protein expression of RUNX3 in five primary EOC cultures and seven normal OSE cell culture. As shown in Fig. 1A, RUNX3 was expressed in all five primary EOC cultures, but in none of the normal OSE cells. We detected two RUNX3 bands in EOC cells, which is consistent with previous studies in EOC [16], human basal cell carcinomas [13] and human endothelial cells [25]. These two bands could represent two isoforms of RUNX3 or are generated by phosphorylation modification or proteolytic cleavage [13]. We also examined the subcellular localization of RUNX3 to ensure that it is localized to the nucleus in primary EOC cells, where it can functionally act as a transcription factor. Consistent with the results reported by Lee et al. [16], we confirmed that RUNX3 was localized in the nucleus of primary EOC cells (Fig. 1B).

3.2. RUNX3 expression is elevated in A2780cp cells and chemoresistant EOC tissues To determine whether RUNX proteins are involved in chemoresistance of EOC, we examined their expression in cisplatin-sensitive A2780s cells and the cisplatin-resistant counterpart A2780cp cells. qRT-PCR showed that the mRNA expression of RUNX1, RUNX3, and CBFβ was higher in A2780cp cells by 7.3, 20.2 and 2.2 fold, respectively, than that in A2780s (Fig. 1C). Western blotting confirmed the marked increase of RUNX3 and the slight increase of RUNX1 expression in A2780cp cells compared to A2780s cells (Fig. 1D). However, increased protein level of CBFβ in A2780cp cells was not confirmed by Western blotting (Fig. 1D). RUNX2 expression was similar between A2780s and A2780cp cells at the mRNA level (Fig. 1C), but was undetectable at the protein level by Western blotting (data not shown). The analysis of the microarray data from the Gene Expression Omnibus (GEO) database for RUNX3 and cisplatin resistance showed that RUNX3 expression was significantly higher in cisplatin-resistant A2780cp70 cells compared to the cisplatin-sensitive A2780 cells [GSE28648 of RUNX3 (204198_s_at)] (Fig. 1E). Importantly the analysis showed that RUNX3 expression was significantly higher in 15 serous EOC tissues from chemoresistant patients (recurrence time b 6 months) compared to 10 serous EOC tissues form chemosensitive patients (recurrence time N 30 months) [GSE28739 of RUNX3 (A_23_P51231)] (Fig. 1F). Together, these data indicate that RUNX3 expression is elevated in the chemoresistant EOC cells and tissues.

3.4. Knockdown of RUNX3 modestly increases the sensitivity of A2780cp cells to carboplatin Next we determined whether inactivation of RUNX3 could sensitize A2780cp cells to carboplatin. We stably knocked down RUNX3 expression in A2780cp cells using two lentivirus-delivered shRNA constructs (shRUNX3A and shRUNX3B) targeting two distinct sequences of human RUNX3 gene (Fig. 4A) [25]. RUNX3 knockdown and control A2780cp cells were treated with increasing concentrations of carboplatin for 72 h. Neutral red uptake assay showed that knockdown of RUNX3 increased the sensitivity of A2780cp cells at 200 μM carboplatin, but not at the lower doses of carboplatin (Fig. 4B). The IC50 values for carboplatin in A2780cp/shRandom, A2780cp/shRUNX3A and A2780cp/shRUNX3B were 123.4, 92.8 and 103.9 μM, respectively. Therefore, RUNX3 knockdown moderately decreased IC50 values by 25% and 16% in shRUNX3A and shRUNX3B cells, respectively. Reduced apoptosis is one mechanism for cancer cells to be resistant to cisplatin

Fig. 2. A2780cp cells are more resistant to carboplatin than A2780s cells. A2780s and A2780cp cells were treated with increasing concentrations of carboplatin and cell viability was measured by the neutral red uptake assay (A) and the clonogenic assay (B). Cell viability was expressed as a percentage relative to the respective untreated controls (0 μM carboplatin). Data are shown as mean ± SEM of three independent experiments. *Significantly different from the A2780s cells treated with the same concentration of carboplatin (P b 0.05).

S.H. Barghout et al. / Gynecologic Oncology 138 (2015) 647–655

651

Fig. 3. Overexpression of RUNX3 renders EOC cells more resistant to carboplatin. (A) Overexpression of RUNX3 in A2780s/RUNX3 cells was confirmed by Western blotting. β-actin was used as the loading control. (B) and (C) A2780s/Vector and A2780s/RUNX3 cells were treated with increasing concentrations of carboplatin and cell viability was determined by the neutral red uptake assay (B) and the clonogenic assay (C) and expressed as a percentage relative to the respective untreated controls. Data are shown as mean ± SEM of five independent experiments for the neutral red assay and three independent experiments for the clonogenic assay. *Significantly different (P b 0.05).

[7,8]. To determine whether RUNX3 knockdown renders A2780cp cells more sensitive to carboplatin-induced apoptosis, we treated A2780cp/ shRandom, A2780cp/shRUNX3A and A2780cp/shRUNX3B cells with 200 μM carboplatin for 72 h and detected carboplatin-induced cleavage

of poly(ADP-ribose) polymerase (PARP) as a marker of apoptosis by Western blotting. Carboplatin-induced cleavage of PARP was more pronounced in RUNX3 knockdown cells compared to the control cells, suggesting that RUNX3 knockdown potentiates carboplatin-induced

Fig. 4. Knockdown of RUNX3 moderately sensitizes A2780cp cells to carboplatin. (A) Stable knockdown of RUNX3 by two different shRNA constructs (shRUNX3A and shRUNX3B) in A2780cp cells was confirmed by Western blotting. Tubulin was used as the loading control. (B) These cells were treated with increasing concentrations of carboplatin. Cell viability was determined by the neutral red uptake assay and expressed as a percentage relative to the respective untreated controls. Data are shown as mean ± SEM of four independent experiments. (C) A2780cp/shRandom, A2780cp/shRUNX3A and A2780cp/shRUNX3B cells were left untreated or treated with 200 μM carboplatin for 72 h. Carboplatin-induced PARP cleavage was measured by Western blotting using antibody against the cleaved PARP. Tubulin was used as the loading control. (D) The Western blotting results of carboplatin-induced PARP cleavage were quantified using Odyssey imaging software. The density of the cleaved PARP bands was normalized to that of tubulin. The density of the bands in the carboplatin-treated A2780cp/shRandom cells was designated as 1. The relative level (fold change) of cleaved PARP was shown as mean ± SEM of four independent experiments. *Significantly different (P b 0.05).

652

S.H. Barghout et al. / Gynecologic Oncology 138 (2015) 647–655

apoptosis in A2780cp cells (Fig. 4C). Quantification of four independent experiments showed that shRUNX3A and shRUNX3B increased carboplatin-induced production of cleaved PARP by 2.2 and 1.9 fold, respectively, as compared to shRandom A2780cp cells (Fig. 4D). 3.5. Overexpression of dnRUNX3 increases the sensitivity of A2780cp cells to carboplatin A truncated form of RUNX3 (RUNX3 1–187, dnRUNX3) containing the intact Runt domain but lacking the transactivation domain at the carboxyl-terminus functions in a dominant negative manner and has been shown to inhibit the functions of other RUNX proteins [26,27]. To determine whether dnRUNX3 is more potent in sensitizing A2780cp cells to carbopaltin than RUNX3 knockdown, we stably overexpressed the empty vector pcDNA3.1 or the dominant-negative vector pcDNA-FLAG-RUNX3 (1–187) in A2780cp cells. The resultant cells were referred as to A2780cp/Vector and A2780cp/dnRUNX3 cells, respectively. Expression of dnRUNX3 in A2780cp/dnRUNX3 was confirmed by Western blotting (Fig. 5A). Overexpression of dnRUNX3 did not affect the expression of the endogenous RUNX1, RUNX3 and CBFβ in A2780cp cells. Neutral red uptake assay and clonogenic assay showed that A2780cp/dnRUNX3 cells were more sensitive to carboplatin than A2780cp/Vector cells (Fig. 5B and 5C). Overexpression of dnRUNX3 decreased IC50 for carboplatin in A2780cp cells by 27.5% (from 110 μM to 79.8 μM) as determined by the neutral red uptake assay and by 52.3% (from 36.9 μM to 17.6 μM) as determined by the clonogenic assay. Western blotting showed that carboplatin-induced cleavage of PARP was more pronounced in A2780cp/dnRUNX3 cells than that in A2780cp/

vector cells (Fig. 5D and 5E), suggesting that dnRUNX3 potentiates carboplatin-induced apoptosis in A2780cp cells.

3.6. dnRUNX3 decreases the expression of cellular inhibitor of apoptosis protein-2 (cIAP2) in A2780cp cells The inhibitor of apoptosis proteins (IAPs) inhibit apoptosis by interfering with caspase activation [34]. Gene expression profiling of A2780s and A2780cp by a microarray analysis (Fu et al., unpublished data) showed that expression of the IAP family member cIAP2, but not that of other family members (cIAP1, XIAP and survivin), was elevated in A2780cp cells compared to A2780s cells. qRT-PCR showed that cIAP2 mRNA level was 37.9 fold higher in A2780cp cells compared to A2780s cells (Fig. 6A), which was confirmed at the protein level by Western blotting (Fig. 6B). RUNX3 overexpression increased the mRNA level of cIAP2 in A2780s cells by 3.6 fold (Fig. 6C). dnRUNX3 decreased the mRNA level of cIAP2 by 36% in A2780cp cells (Fig. 6D), which was confirmed by Western blotting (Fig. 6E). Carboplatin treatment (200 μM carboplatin for 48 h) significantly reduced the expression of cIAP2, which was potentiated by dnRUNX3 (Fig. 6F). Quantification of three independent experiments showed that overexpression of dnRUNX3 and carboplatin treatment decreased cIAP2 protein by 43% and 29%, respectively. However, the combination of dnRUNX3 and carboplatin treatment decreased cIAP2 protein by 60% (Fig. 6G). Taken together, these results suggest that RUNX3 regulates the expression of cIAP2 in A2780s and A2780cp cells, and dnRUNX3 potentiates carboplatin-induced decrease of cIAP2 in A2780cp cells.

Fig. 5. dnRUNX3 increases the sensitivity of A2780cp cells to carboplatin. (A) Overexpression of dnRUNX3 was confirmed by Western blotting using anti-FLAG antibody. Expression of RUNX1 and CBFβ was also measured by Western blotting. β-actin was used as the loading control. (B) and (C) A2780s/Vector and A2780s/dnRUNX3 cells were treated with increasing concentrations of carboplatin. Cell viability was determined by the neutral red uptake assay (B) and the clonogenic assay (C) and expressed as a percentage relative to the respective untreated controls. Data are shown as mean ± SEM of five independent experiments for the neutral red assay and three independent experiments for the clonogenic assay. *Significantly different (P b 0.05). (D) A2780cp/Vector and A2780cp/dnRUNX3 cells were left untreated or treated with 200 μM carboplatin for 48 or 72 h. Carboplatin-induced PARP cleavage was measured by Western blotting using antibodies against total PARP and the cleaved PARP. β-actin was used as the loading control. (E) The Western blotting results of carboplatin-induced PARP cleavage were quantified using Odyssey imaging software. The density of the cleaved PARP bands was normalized to that of β-actin. The density of the bands in the carboplatin-treated vector cells at 48 h was designated as 1. The relative level (fold change) of cleaved PARP was shown as mean ± SEM of three independent experiments. *Significantly different (P b 0.05).

S.H. Barghout et al. / Gynecologic Oncology 138 (2015) 647–655

653

Fig. 6. RUNX3 regulates cIAP2 expression in A2780s and A2780cp cells. (A) mRNA level of cIAP2 in A2780s and A2780cp was determined by qRT-PCR. Data are shown as mean ± SEM of three independent experiments. (B) Protein level of cIAP2 in A2780s and A2780cp was determined by Western blotting. β-actin was used as the loading control (C) mRNA level of cIAP2 was determined by qRT-PCR in A2780s/Vector and A2780s/RUNX3 cells. Data are shown as mean ± SEM of three independent experiments. (D) mRNA level of cIAP2 in A2780cp/Vector and A2780cp/dnRUNX3 was determined by qRT-PCR. Data are shown as mean ± SEM of three independent experiments. (E) Expression of cIAP2 in A2780cp/Vector and A2780cp/ dnRUNX3 was determined by Western blotting. β-actin was used as the loading control. (F) A2780cp/Vector and A2780cp/dnRUNX3 cells were left untreated or treated with 200 μM carboplatin for 48 h. cIAP2 protein level was measured by Western blotting. β-actin was used as the loading control. (G) The Western blotting results of cIAP2 were quantified using Odyssey imaging software. The density of cIAP2 bands was normalized to that of β-actin. The density of the bands in the untreated A2780cp/Vector cells was designated as 1. The relative level (fold change) of cIAP2 was shown as mean ± SEM of three independent experiments. *Significantly different (P b 0.05).

4. Discussion A better understanding of the molecular mechanism underlying the acquired drug resistance is necessary to improve the management of EOC. Two studies have demonstrated that RUNX3 plays an oncogenic role in EOC [15,16]. However, the role for RUNX3 in chemoresistance of EOC has not been reported. In this study we confirmed that RUNX3 is expressed in primary EOC cells isolated from ascites of EOC patients, but not in normal OSE cells. Conflicting results have been reported regarding the cellular localization of RUNX3 in EOC cells. Nevadunsky et al. reported that RUNX3 is localized to the cytoplasm in EOC cell lines [15], whereas Lee et al. observed nuclear localization of RUNX3 in EOC cells [16]. We confirmed that RUNX3 is localized to the nucleus in primary EOC cells, which is consistent with the localization reported by Lee et al. and supports the functional role of RUNX3 as a transcription factor. Cisplatin resistant EOC cell line A2780cp was established by exposure of cisplatin sensitive A2780s cells to stepwise-increasing concentrations of cisplatin [23] and the paired A2780s and A2780cp cell lines have been widely used to study the acquired chemoresistance of EOC [35,36]. Our results show that the IC50 for carboplatin in A2780cp cells is more than 10 fold higher compared to A2780s cells, validating their use to study carboplatin resistance. Our finding that RUNX3 expression

is markedly elevated in A2780cp cells compared to A2780s cells is supported by the data from the GEO database; RUNX3 expression is higher in cisplatin resistant A2780cp70 cells and chemoresistant EOC tissues compared to their respective controls. Importantly, overexpression of RUNX3 renders A2780s cells more resistant to carboplatin. To ascertain the role of RUNX3 in carboplatin resistance of A2780cp cells, we performed loss-of-function experiments. We found that overexpression of dnRUNX3, and to a lesser extent knockdown of RUNX3, are able to partially sensitize A2780cp cells to carboplatin. Three RUNX proteins (RUNX1-3) have been shown to have overlapping and distinct biological functions depending on cell context [9]. All three RUNX proteins play a pro-tumorigenic role in EOC by increasing proliferation, migration and invasion [15–18] and knockdown of CBFβ leads to growth inhibition of SKOV3 cells due to cell cycle arrest or increase in apoptosis [19]. Our results showed that RUNX1 is also expressed in A2780cp cells at a slightly higher level compared to A2780s cells. Thus, the less pronounced effect of RUNX3 knockdown on carboplatin resistance could be due to either the incomplete knockdown of RUNX3 or the compensation for decreased RUNX3 expression by RUNX1 or both. On the other hand, overexpression of dnRUNX3 that inhibits the function of other RUNX proteins [26,27] likely inhibits both RUNX3 and RUNX1 functions and thus renders A2780cp cells more sensitive to carboplatin.

654

S.H. Barghout et al. / Gynecologic Oncology 138 (2015) 647–655

IAPs that include cIAP1 (cellular inhibitor of apoptosis protein-1), cIAP2, XIAP (X-linked inhibitor of apoptosis protein) and survivin inhibit apoptosis by interfering with caspase activation [34]. Our unpublished microarray data show that cIAP2 expression is elevated in A2780cp cells compared to A2780s cells, which is confirmed by qRT-PCR and Western blotting. Upregulation of cIAP2 has been associated with cisplatin resistance in prostate cancer and lung cancer [37,38] and inhibition of IAPs including cIAP2 results in increased apoptosis in EOC cells [39]. Our results showed that RUNX3 regulates cIAP2 expression; overexpression of RUNX3 increases the expression of cIAP2 in A2780s cells and dnRUNX3 decreases the expression of cIAP2 in A2780cp cells. Interestingly, we found that carboplatin treatment decreases the expression of cIAP2, suggesting that downregulation of cIAP2 can be one mechanism for carboplatin to induce apoptosis of A2780cp cells. Our finding that dnRUNX3 decreases the expression of cIAP2 and potentiates carboplatin-induced downregulation of cIAP2 in A2780cp cells suggests that dnRUNX3-induced sensitization of A2780cp cells to carboplatin could be attributed to the downregulation of cIAP2. Further studies are required to determine the role of cIAP2 in carboplatin resistance in A2780cp cells. Cancer cells develop a variety of mechanisms to resist platinum cytotoxicity, including decreasing intracellular platinum accumulation and reducing platinum-induced apoptosis [6–8]. In this study, we demonstrate that RUNX3 inhibits carboplatin-induced apoptosis in A2780cp cells, suggesting that RUNX3 is one of the factors that contribute to carboplatin resistance in EOC. Because the interaction between cancer cells and tumor microenvironment plays an important role in chemoresistance, it is important that the role of RUNX proteins in carboplatin resistance be tested in pre-clinical in vivo models. In this regard, it is our future interest to test whether dnRUNX3 can sensitize A2780cp xenografts to carboplatin in mouse models. Also, it is important to examine the role for RUNX3 in carboplatin resistance in additional EOC cell lines. Given the multifactorial nature of resistance to platinum-based compounds [6–8], it becomes crucial to identifying all potential mechanisms contributing to the resistance. Identifying the major mechanisms governing the resistant phenotype in a given patient will help determine the best-suited combination therapy to achieve an optimal therapeutic outcome. This copes with the emerging trend of precision oncology that aims at effectively managing cancer heterogeneity and drug resistance [40]. In summary, we confirm that RUNX3 expression is expressed in EOC cells, but not in normal OSE cells and demonstrate that expression of RUNX3 is elevated in chemoresistant cells and tissues compared to their respective chemosensitive controls. Overexpression of RUNX3 renders A2780s cells more resistant to carboplatin, whereas inhibition of RUNX3 increases the sensitivity of A2780cp cells to carboplatin. These results provide insight into the molecular mechanism underlying the chemoresistance of EOC. Further study is required to determine whether targeting RUNX3 in combination with other therapeutics could be an effective strategy to tackle the chemoresistance of EOC. The elevated expression of RUNX3 in chemoresistant cells and tissues suggests that RUNX3 can be a potential biomarker to predict chemoresistance of EOC. Conflicts of interest statement The authors have no conflict of interest to declare.

Role of the funding source The study sponsors played no role in any aspect of this study. Acknowledgments This study was generously supported by a start-up fund from the Women and Children's Health Research Institute (WCHRI) with funding donated by the Royal Alexandra Hospital Foundation (RAHF) to Dr. Fu.

We thank the Gynecologic Oncology group at the University of Alberta for assistance in collection of normal OSE and primary EOC cells; Dr. Benjamin Tsang (Ottawa Hospital Research Institute) for providing A2780s and A2780cp cells; and Dr. Yoshiaki Ito, Cancer Science Institute of Singapore for providing pcDNA-FLAG-RUNX3 (1–187). Nubia Zepeda was supported by a WCHRI graduate studentship and the Canadian Institutes of Health Research (CIHR) Master's Award: Frederick Banting and Charles Best Canada Graduate Scholarships. Samir Barghout was supported by the MatCH program at the University of Alberta, Faculty of Medicine and Dentistry/Alberta Health Services graduate studentship, and the Medical Science Graduate Program Scholarship. We thank the Flow Cytometry Facility of the Oncology Department at the University of Alberta at the Cross Cancer Institute for cell sorting.

References [1] R.L. Siegel, K.D. Miller, A. Jemal, Cancer statistics, 2015, CA Cancer J. Clin. 65 (2015) 5–29. [2] N. Auersperg, A.S. Wong, K.C. Choi, S.K. Kang, P.C. Leung, Ovarian surface epithelium: biology, endocrinology, and pathology, Endocr. Rev. 22 (2001) 255–288. [3] K. Levanon, C. Crum, R. Drapkin, New insights into the pathogenesis of serous ovarian cancer and its clinical impact, J. Clin. Oncol. 26 (2008) 5284–5293. [4] T.A. Yap, C.P. Carden, S.B. Kaye, Beyond chemotherapy: targeted therapies in ovarian cancer, Nat. Rev. Cancer 9 (2009) 167–181. [5] S. Vaughan, J.I. Coward, R.C. Bast Jr., A. Berchuck, J.S. Berek, J.D. Brenton, G. Coukos, C.C. Crum, R. Drapkin, D. Etemadmoghadam, M. Friedlander, H. Gabra, S.B. Kaye, C.J. Lord, E. Lengyel, D.A. Levine, I.A. McNeish, U. Menon, G.B. Mills, K.P. Nephew, A.M. Oza, A.K. Sood, E.A. Stronach, H. Walczak, D.D. Bowtell, F.R. Balkwill, Rethinking ovarian cancer: recommendations for improving outcomes, Nat. Rev. Cancer 11 (2011) 719–725. [6] L. Kelland, The resurgence of platinum-based cancer chemotherapy, Nat. Rev. Cancer 7 (2007) 573–584. [7] Z.H. Siddik, Cisplatin: mode of cytotoxic action and molecular basis of resistance, Oncogene 22 (2003) 7265–7279. [8] L. Galluzzi, L. Senovilla, I. Vitale, J. Michels, I. Martins, O. Kepp, M. Castedo, G. Kroemer, Molecular mechanisms of cisplatin resistance, Oncogene 31 (2012) 1869–1883. [9] Y. Ito, S.C. Bae, L.S. Chuang, The RUNX family: developmental regulators in cancer, Nat. Rev. Cancer 15 (2015) 81–95. [10] G. Huang, K. Shigesada, K. Ito, H.J. Wee, T. Yokomizo, Y. Ito, Dimerization with PEBP2beta protects RUNX1/AML1 from ubiquitin–proteasome-mediated degradation, EMBO J. 20 (2001) 723–733. [11] Q.L. Li, K. Ito, C. Sakakura, H. Fukamachi, K. Inoue, X.Z. Chi, K.Y. Lee, S. Nomura, C.W. Lee, S.B. Han, H.M. Kim, W.J. Kim, H. Yamamoto, N. Yamashita, T. Yano, T. Ikeda, S. Itohara, J. Inazawa, T. Abe, A. Hagiwara, H. Yamagishi, A. Ooe, A. Kaneda, T. Sugimura, T. Ushijima, S.C. Bae, Y. Ito, Causal relationship between the loss of RUNX3 expression and gastric cancer, Cell 109 (2002) 113–124. [12] M.M. Subramaniam, J.Y. Chan, K.G. Yeoh, T. Quek, K. Ito, M. Salto-Tellez, Molecular pathology of RUNX3 in human carcinogenesis, Biochim. Biophys. Acta 1796 (2009) 315–331. [13] M. Salto-Tellez, B.K. Peh, K. Ito, S.H. Tan, P.Y. Chong, H.C. Han, K. Tada, W.Y. Ong, R. Soong, D.C. Voon, Y. Ito, RUNX3 protein is overexpressed in human basal cell carcinomas, Oncogene 25 (2006) 7646–7649. [14] T. Tsunematsu, Y. Kudo, S. Iizuka, I. Ogawa, T. Fujita, H. Kurihara, Y. Abiko, T. Takata, RUNX3 has an oncogenic role in head and neck cancer, PLoS One 4 (2009) e5892. [15] N.S. Nevadunsky, J.S. Barbieri, J. Kwong, M.A. Merritt, W.R. Welch, R.S. Berkowitz, S.C. Mok, RUNX3 protein is overexpressed in human epithelial ovarian cancer, Gynecol. Oncol. 112 (2009) 325–330. [16] C.W. Lee, L.S. Chuang, S. Kimura, S.K. Lai, C.W. Ong, B. Yan, M. Salto-Tellez, M. Choolani, Y. Ito, RUNX3 functions as an oncogene in ovarian cancer, Gynecol. Oncol. 122 (2011) 410–417. [17] M. Keita, M. Bachvarova, C. Morin, M. Plante, J. Gregoire, M.C. Renaud, A. Sebastianelli, X.B. Trinh, D. Bachvarov, The RUNX1 transcription factor is expressed in serous epithelial ovarian carcinoma and contributes to cell proliferation, migration and invasion, Cell Cycle 12 (2013) 972–986. [18] Z.Q. Wang, M. Keita, M. Bachvarova, S. Gobeil, C. Morin, M. Plante, J. Gregoire, M.C. Renaud, A. Sebastianelli, X.B. Trinh, D. Bachvarov, Inhibition of RUNX2 transcriptional activity blocks the proliferation, migration and invasion of epithelial ovarian carcinoma cells, PLoS One 8 (2013) e74384. [19] A.H. Greer, T. Yong, K. Fennell, Y.W. Moustafa, M. Fowler, F. Galiano, S.W. Ng, R.S. Berkowitz, J. Cardelli, S. Meyers, J.N. Davis, Knockdown of core binding factorbeta alters sphingolipid metabolism, J. Cell. Physiol. 228 (2013) 2350–2364. [20] N. Xu, C. Shen, Y. Luo, L. Xia, F. Xue, Q. Xia, J. Zhang, Upregulated miR-130a increases drug resistance by regulating RUNX3 and Wnt signaling in cisplatin-treated HCC cell, Biochem. Biophys. Res. Commun. 425 (2012) 468–472. [21] Y. Zhang, Q. Lu, X. Cai, MicroRNA-106a induces multidrug resistance in gastric cancer by targeting RUNX3, FEBS Lett. 587 (2013) 3069–3075. [22] Y. Zheng, R. Wang, H.Z. Song, B.Z. Pan, Y.W. Zhang, L.B. Chen, Epigenetic downregulation of RUNX3 by DNA methylation induces docetaxel chemoresistance in human

S.H. Barghout et al. / Gynecologic Oncology 138 (2015) 647–655

[23]

[24]

[25]

[26]

[27]

[28]

[29] [30] [31]

lung adenocarcinoma cells by activation of the AKT pathway, Int. J. Biochem. Cell Biol. 45 (2013) 2369–2378. B.C. Behrens, T.C. Hamilton, H. Masuda, K.R. Grotzinger, J. Whang-Peng, K.G. Louie, T. Knutsen, W.M. McKoy, R.C. Young, R.F. Ozols, Characterization of a cisdiamminedichloroplatinum(II)-resistant human ovarian cancer cell line and its use in evaluation of platinum analogues, Cancer Res. 47 (1987) 414–418. T.G. Shepherd, B.L. Theriault, E.J. Campbell, M.W. Nachtigal, Primary culture of ovarian surface epithelial cells and ascites-derived ovarian cancer cells from patients, Nat. Protoc. 1 (2006) 2643–2649. Y. Fu, A.C. Chang, M. Fournier, L. Chang, K. Niessen, A. Karsan, RUNX3 maintains the mesenchymal phenotype after termination of the Notch signal, J. Biol. Chem. 286 (2011) 11803–11813. K. Ito, Q. Liu, M. Salto-Tellez, T. Yano, K. Tada, H. Ida, C. Huang, N. Shah, M. Inoue, A. Rajnakova, K.C. Hiong, B.K. Peh, H.C. Han, T. Ito, M. Teh, K.G. Yeoh, Y. Ito, RUNX3, a novel tumor suppressor, is frequently inactivated in gastric cancer by protein mislocalization, Cancer Res. 65 (2005) 7743–7750. T. Yano, K. Ito, H. Fukamachi, X.Z. Chi, H.J. Wee, K. Inoue, H. Ida, P. Bouillet, A. Strasser, S.C. Bae, Y. Ito, The RUNX3 tumor suppressor upregulates Bim in gastric epithelial cells undergoing transforming growth factor beta-induced apoptosis, Mol. Cell. Biol. 26 (2006) 4474–4488. N. Gupta, Z. Xu, A. El-Sehemy, H. Steed, Y. Fu, Notch3 induces epithelial–mesenchymal transition and attenuates carboplatin-induced apoptosis in ovarian cancer cells, Gynecol. Oncol. 130 (2013) 200–206. N.A. Franken, H.M. Rodermond, J. Stap, J. Haveman, C. van Bree, Clonogenic assay of cells in vitro, Nat. Protoc. 1 (2006) 2315–2319. A. Munshi, M. Hobbs, R.E. Meyn, Clonogenic cell survival assay, Methods Mol. Med. 110 (2005) 21–28. Y. Fu, A. Chang, L. Chang, K. Niessen, S. Eapen, A. Setiadi, A. Karsan, Differential regulation of transforming growth factor beta signaling pathways by Notch in human endothelial cells, J. Biol. Chem. 284 (2009) 19452–19462.

655

[32] Y. Fu, H. Sies, X.G. Lei, Opposite roles of selenium-dependent glutathione peroxidase-1 in superoxide generator diquat- and peroxynitrite-induced apoptosis and signaling, J. Biol. Chem. 276 (2001) 43004–43009. [33] R.F. Ozols, B.N. Bundy, B.E. Greer, J.M. Fowler, D. Clarke-Pearson, R.A. Burger, R.S. Mannel, K. DeGeest, E.M. Hartenbach, R. Baergen, Phase III trial of carboplatin and paclitaxel compared with cisplatin and paclitaxel in patients with optimally resected stage III ovarian cancer: a Gynecologic Oncology Group study, J. Clin. Oncol. 21 (2003) 3194–3200. [34] A.D. Schimmer, Inhibitor of apoptosis proteins: translating basic knowledge into clinical practice, Cancer Res. 64 (2004) 7183–7190. [35] L. Farrand, S. Byun, J.Y. Kim, A. Im-Aram, J. Lee, S. Lim, K.W. Lee, J.Y. Suh, H.J. Lee, B.K. Tsang, Piceatannol enhances cisplatin sensitivity in ovarian cancer via modulation of p53, X-linked inhibitor of apoptosis protein (XIAP), and mitochondrial fission, J. Biol. Chem. 288 (2013) 23740–23750. [36] L. Farrand, J.Y. Kim, S. Byun, A. Im-Aram, J. Lee, J.Y. Suh, K.W. Lee, H.J. Lee, B.K. Tsang, The diarylheptanoid hirsutenone sensitizes chemoresistant ovarian cancer cells to cisplatin via modulation of apoptosis-inducing factor and X-linked inhibitor of apoptosis, J. Biol. Chem. 289 (2014) 1723–1731. [37] T. Nomura, M. Yamasaki, Y. Nomura, H. Mimata, Expression of the inhibitors of apoptosis proteins in cisplatin-resistant prostate cancer cells, Oncol. Rep. 14 (2005) 993–997. [38] H.H. Wu, J.Y. Wu, Y.W. Cheng, C.Y. Chen, M.C. Lee, Y.G. Goan, H. Lee, cIAP2 upregulated by E6 oncoprotein via epidermal growth factor receptor/phosphatidylinositol 3-kinase/AKT pathway confers resistance to cisplatin in human papillomavirus 16/18-infected lung cancer, Clin. Cancer Res. 16 (2010) 5200–5210. [39] C. Zeng, S. Vangveravong, J.E. McDunn, W.G. Hawkins, R.H. Mach, Sigma-2 receptor ligand as a novel method for delivering a SMAC mimetic drug for treating ovarian cancer, Br. J. Cancer 109 (2013) 2368–2377. [40] F.S. Collins, H. Varmus, A new initiative on precision medicine, N. Engl. J. Med. 372 (2015) 793–795.