Cisplatin resistance in gastric cancer cells is associated with HER2 upregulation-induced epithelial-mesenchymal transition

Cisplatin remains to be primary chemotherapeutic drug for gastric cancer patients, especially for advanced stage ones. However, primary or acquired resistance often occurs with the mechanisms being not well understood, which results in relapse of the cancer and poor survival. Herein, we found that HER2 upregulation was associated with cisplatin resistance. We observed that cisplatin-resistant gastric cancer cells underwent a morphological change similar to epithelial-mesenchymal transition (EMT) which is mediated by HER2 overexpression. When specific monoclonal antibody Herceptin, small molecular targeted drug CP724714, or small interfering RNA against HER2 was applied, the EMT-like phenotypic change was dramatically reversed. More importantly, the IC50 and Resistance Index of resistant gastric cancer cells to cisplatin were also decreased by any of these treatments.We demonstrated that expression and amplification of HER2 positively correlated with expression of EMT-related transcription factor Snail in gastric cancer tissues. Furthermore, for the first time, we found that HER2/Snail double positive gastric cancer patients had poorer survival than single positive or double negative counterparts, which provided experimental evidence for the necessity of HER2/Snail double testing in gastric cancer. In conclusion, this study provides some clues of the association of cisplatin resistance with HER2 upregulation-induced EMT in gastric cancer cells.

Scientific RepoRts | 6:20502 | DOI: 10.1038/srep20502 properties [12][13][14][15] . EMT of Cancer cells also associates with resistance to chemotherapy [16][17][18][19] . Up-regulation of Twist was associated with resistance to paclitaxel in human nasopharyngeal, bladder, ovarian, and prostate cancers 16 . In colorectal cancer, oxaliplatin-resistant cells can acquire the ability to migrate and invade with phenotypic changes resembling EMT 17 . In pancreatic and ovarian cancer, stable cell lines resistant to gemcitabine and paclitaxel established by continuous exposure can undergo EMT with increased expression of Snail and Twist 18,19 . However, there has no study investigating the role of EMT in mediating cisplatin resistance in gastric cancer.
HER2 (also known as ErbB2), a 185-kDa transmembrane tyrosine kinase (TK) receptor, is a member of the epidermal growth factor receptors (EGFRs) family. This family includes HER1 (also known as EGFR, ErbB1), HER2, HER3 (ErbB3), and HER4 (ErbB4). These receptors share similar molecular structure: an extracellular ligand-binding domain, a short transmembrane domain, and an intracellular domain with TK activity (excepting HER3) 20 . Unlike HER1, HER3 and HER4, HER2 has no ligand 20 . Recent studies indicate a role of HER2 in the development of many types of human cancer, especially in breast cancer 21,22 . HER2 overexpression and/or amplification have been detected in 10-34% of invasive breast cancers and correlate with unfavorable patient outcome 22 . Targeted therapeutic drugs, including monoclonal antibody Herceptin and small molecule tyrosine kinase inhibitors targeting HER2, are playing more and more important roles in breast cancer treatment 23,24 . HER2 overexpression and/or amplification have also been observed in colon 25 , bladder 26 , ovarian 27 , endometrial 28 , lung 29 , uterine cervix 30 , head and neck 31 , esophageal 32 , and gastric carcinomas 33 . However, up to date, there is no study investigating the role of HER2 in mediating cisplatin-resistance in gastric cancer.
Based on the above knowledge, we hypothesized that HER2 and EMT may be involved in cisplatin resistance in gastric cancer. Herein, we found that HER2 is overexpressed in cisplatin-resistant gastric cancer cell models MGC803/DDP and AGS/DDP. Meanwhile, MGC803/DDP and AGS/DDP cells exhibited EMT-like morphological changes, compared with parental gastric cancer cell lines MGC803 and AGS. At molecular level, the alteration of expression patterns of EMT-related protein markers also implied this phenomenon. More interestingly, EMT in the cisplatin-resistant gastric cancer cells could be abrogated by monoclonal antibody Herceptin, small molecular targeted drug CP724714, or small interfering RNAs (siRNAs) against HER2. Furthermore, we demonstrated that amplification of HER2 or expression of HER2 positively correlated with the expression of EMT-related transcription factor Snail in 382 cases of gastric cancer, and HER2/Snail double positive patients had poorer patient outcome compared with single positive or double negative group.

Results
Cisplatin-resistant gastric cancer cell models display higher motility. First of all, we tested cell proliferation rate of cisplatin-resistant gastric cancer cells MGC803/DDP and AGS/DDP. As shown in Fig. 1A, the proliferation rates of parental and cisplatin-resistant cells were similar. Then, we tested the IC 50 and the Resistance Index (RI) of MGC803/DDP and AGS/DDP. As shown in Fig. 1B, IC 50 of MGC803/DDP and AGS/DDP cells were 4.9 and 6.3 μ g/ml respectively, and the RI were 14.8 and 9.5 respectively. However, IC 50 of MGC803 and AGS cells were 0.33 and 0.66 μ g/ml respectively. Interestingly, we found that MGC803/DDP and AGS/DDP cells possess an increased motility compared with parental cells in the wound healing assay, as shown in Fig. 1C,D. Consistently, MGC803/DDP and AGS/DDP cells also showed enhanced migration through the micropore membranes into the lower chambers in the transwell chamber assay (Fig. 1E). The above results demonstrated that cisplatin-resistant gastric cancer cells displayed higher migratory capacity and motility compared with parental cells.

Cisplatin-resistant gastric cancer cells exhibit an EMT-like phenotypic change.
EMT is of paramount importance in a plethora of cancer related events, including cancer metastasis and chemotherapy resistance 14,[16][17][18] . But to the best of our knowledge, the association of EMT with cisplatin resistance in gastric cancer has not been reported. Herein, we found that MGC803/DDP and AGS/DDP cells exhibited a spindle-like fibroblastoid phenotypic change in contrast to parental cell lines MGC803 and AGS ( Fig. 2A). FITC-conjugated phalloidin staining of F-actin also revealed that MGC803/DDP and AGS/DDP cells possessed EMT-like morphology compared with their parental cells (Fig. 2B). At molecular level, the expression patterns of EMT protein markers and EMT-related transcription factors were determined by western blot. MGC803/DDP and AGS/DDP cells showed reduced expression of epithelial cell marker ZO1. Meanwhile, the expression of E-cadherin was down-regulated in MGC803/DDP cells. Furthermore, the expression of EMT-related transcription factor Snail was increased in MGC803/DDP and AGS/DDP cells (Fig. 2C).

HER2 is overexpressed and plays a pivotal role in mediating EMT in cisplatin-resistant gastric cancer cells.
In the following study, we sought to explore the potential molecular mechanism which is responsible for the EMT-like phenotypic changes in cisplatin-resistant gastric cancer cells. Previous study revealed that HER2 overexpression and/or amplification is an unfavourable predictive factor 34 . So, we hypothesized that HER2 overexpression and/or amplification may be associated with cisplatin resistance in gastric cancer. As shown in Fig. 3A, HER2 expression was significantly upregulated in cisplatin-resistant gastric cancer cells, especially in MGC803/DDP cells. Next, we found that MGC803/DDP and AGS/DDP cells showed increased HER2 amplification compared with parental MGC803 and AGS cells through qPCR method, indicating that HER2 amplification played an important role in HER2 protein up-regulation (Fig. 3B).
In order to investigate the role of HER2 in mediating EMT in cisplatin-resistant gastric cancer cells, we utilized multiple strategies, including monoclonal antibody Herceptin blocking method, siRNA technique, and small molecular targeted therapeutic drug CP724714 treatment. Firstly, cell morphology analysis showed that the fibroblastic, spindle-shaped morphology of MGC803/DDP cells was reversed to the epithelial cell phenotype similar to that of parental counterparts after Herceptin treatment in a dose-dependent manner (Fig. 3C). Herceptin at the concentration of 100 μ g/ml completely abrogated the EMT phenomenon observed in MGC803/DDP cells (Fig. 3C). Similar changes were observed by FITC-phalloidin staining for F-actin (Fig. 3D). At molecular level, Scientific RepoRts | 6:20502 | DOI: 10.1038/srep20502 ZO1 down-regulation and Snail up-regulation were also counteracted by Herceptin treatment in MGC803/DDP cells (Fig. 3E). Taken together, these results suggested that HER2 mediates EMT in cisplatin-resistant gastric cancer cells.
Next, a potent molecular targeted agent against HER2, CP724714, was utilized. CP724714 at the concentration of 10 μ M was applied to MGC803/DDP cells for 24 h. As shown in Fig. 3F,G, the morphology of MGC803/ DDP cells was also reversed to that of epithelial cells. Meanwhile, ZO1 down-regulation and Snail up-regulation were reversed by CP724714 in MGC803/DDP cells (Fig. 3H). CP724714 significantly lowered the cell migration of MGC803/DDP cells (Fig. 3I,J), suggesting that HER2 activity contributes to increased cell invasiveness in cisplatin-resistant gastric cancer cells.
Thirdly, we employed siRNA strategy. As shown in Fig. 4A, HER2 expressions were knocked down by three pairs of specific siRNAs. Cell morphology analysis and FITC-phalloidin staining assay showed that EMT of MGC803/DDP cells was completely reversed by these siRNAs (Fig. 4B,C). Snail up-regulation and ZO1 down-regulation were reversed by siRNAs in MGC803/DDP cells (Fig. 4D). Meanwhile, cell migration was inhibited by these siRNAs in MGC803/DDP cells (Fig. 4E,F). Taken together, these data provided evidences that HER2 is required for EMT and cell invasiveness in cisplatin-resistant gastric cancer cells.

HER2 inhibition significantly reverses cisplatin resistance in MGC803/DDP cells. Firstly, we
tested whether multiple HER2 inhibition strategies affect the proliferation of MGC803/DDP cells. After transfection with specific siRNA#3, treated with 100 μ g/ml Herceptin, or treated with 10 μ M CP724714, the proliferation rate remained similar (Fig. 5A). Next, we calculated the IC 50 values and Resistance indexes of each panel. As shown in Fig. 5B, the sensitivity of MGC803/DDP cells towards cisplatin was dramatically increased (Fig. 5B). The IC 50 values was decreased to 0.9, 0.6, or 0.5 respectively after siRNA against HER2, Herceptin, or CP724714 pre-treatment. These data suggested that HER2 is required for the maintenance of cisplatin resistance in gastric cancer cells. Meanwhile, it also provided potential therapeutic strategies for gastric cancer patients being unresponsive with cisplatin.

HER2/Snail double positive gastric cancer patients have unfavorable outcome.
To determine the association of HER2 overexpression and/or amplification with EMT in gastric cancer at tissue levels, we examined HER2 protein expression by immunohistochemistry (IHC) and DNA amplification by fluorescence in situ hybridization (FISH) in a cohort of 382 gastric cancer samples with a 5-year follow-up. Meanwhile, Snail protein expression was also tested by IHC in the same cohort. As shown in Fig. 6A, HER2 expression was positively correlated with Snail expression. In these cases, HER2 amplification was determined as a ratio of HER2 to that of chromosome 17 centromere signal. The typical results of the FISH test were presented in Fig. 6B.
In this cohort of gastric cancer patients, 19.9% (76/382) by IHC and 20.4% (78/382) by FISH were positive for HER2 (Table 1). 42.7% (109/382) by IHC were positive for Snail (Table 1). Although HER2 expression and/ or amplification exhibited no link with patient's gender, age, tumor size, or tumor differentiation, significant correlations with depth of invasion, lymph node metastasis, TNM staging, vascular invasion and metastasis were observed (Table 1). Snail expression obviously correlated with tumor size, differentiation, depth of invasion, lymph node metastasis, TNM staging, vascular invasion and metastasis (Table 1). Next, Chi-square statistical analysis was employed to evaluate the correlation of HER2 overexpression and/or amplification with Snail expression. The results showed that expression of Snail positively correlated with HER2 expression or HER2 amplification (Fig. 6C).
To analyze the correlation of HER2 expression/amplification and Snail expression with patients' outcome, we performed patient survival analysis. The Kaplan-Meier plotting showed that gastric cancer patients with increased HER2 amplification had a poorer overall survival (OS) than those with HER2-negative counterparts (Fig. 6D). Gastric cancer patients with Snail-positive expression also had a poorer OS than those with Snail-negative expression (Fig. 6E). These two findings were consistent with previous reports by other labs [34][35][36] . Importantly, we found gastric cancer patients with HER2/Snail double positive amplification/expression had worst OS than those with single positive expression/amplification or both negative groups (Fig. 6F). These data indicated that combined testing of HER2 and Snail expression and/or amplification had a significantly prognostic value for determining the survival of gastric cancer patients.

Discussion
Gastric cancer remains the second leading cause of cancer mortality in the world 37 . Surgical resection is the primary option of treatment and can cure patients with early-stage cancer. However, many gastric cancer patients are diagnosed when the tumor is at an unresectable stage. For these persons, application of chemotherapeutic agents singly or in combination remains mainstream treatment. Cisplatin, as a first-line chemotherapeutic drug for gastric cancer, could trigger apoptosis by inducing DNA damage through crosslinking of the DNA 38 . However, cancer cells often develop multiple mechanisms to overcome cisplatin-induced DNA damage and apoptosis, leading to cisplatin resistance 39,40 . Therefore, investigation of the molecular mechanism conferring chemotherapy resistance is urgently needed.
Although HER2 plays a povital role in gastric cancer progression and prognosis 34 , the molecular mechanisms of HER2 amplification and/or overexpression in mediating gastric cancer-related biological behaviors is not yet  fully understood. To our knowledge, this is the first study investigating the association of HER2 with cisplatin resistance in gastric cancer cells.
Herein, we found that HER2 protein is overexpressed in cisplatin-resistant gastric cancer cells compared to parental cells, likely due to HER2 gene amplification. Moreover, we found that these cells underwent an EMT-like morphological alteration. EMT is associated with cancer aggressiveness, invasive and metastatic potential, and chemotherapeutic resistance [14][15][16][17][18] . In recent decade, the molecular mechanisms and signaling pathways mediating EMT were extensively studied, leading to the discovery of several pathways involved in the loss of epithelial cell polarity and the acquisition of mesenchymal phenotypic traits [41][42][43] . However, there is few study linking HER2 and EMT with cisplatin resistance in gastric cancer. In this study, we found that HER2 upregulation is responsible for EMT phenomenon and these events are associated with cisplatin resistance in gastric cancer cells. It was demonstrated that progression of carcinoma cells to metastatic tumor cells frequently involved EMT-like epithelial plasticity changes towards a migratory, fibroblastoid phenotype, particularly evident at the invasive front of human tumors 44 . Our current study suggests that EMT is not only a cause of cisplatin resistance in gastric cancer, it may also be a consequence of cisplatin resistance in gastric cancer which results in increased cancer metastasis and tumor relapse after cisplatin treatment for the unresponsive patients.
Herceptin is a monoclonal antibody which specifically targets HER2 protein by directly binding the extracellular domain of the receptor. Herceptin enhances survival rates in both primary and metastatic HER2-positive breast cancer patients 45,46 . The efficacy of Herceptin in breast cancer patients has led to its use in other HER2-positive cancer patients, including gastric cancer [47][48][49] . In this study, we found that gastric cancer patients  with HER2 overexpression and/or amplification had a more unfavourable outcome. Kaplan Meier curve analysis demonstrated that Snail is another prognostic factor in gastric cancer progression, which is consistent with previous report 36 . Our study also suggested that combined testing of HER2 and Snail expression in gastric cancer would provide more useful information for patient prognosis, especially for those respond badly to cisplatin, however more efforts are required to validate this assumption. It is reported that HER2 is upregulated after onset of acquired resistance to cisplatin and fluorouracil combination chemotherapy in gastric cancer patients (P = 0.0065) 50 . The microarray data presented in this article implies that HER2 is upregulated in cisplatin-resistant gastric cancer samples. Furthermore, the mechanism of cisplatin resistance-induced HER2 overexpression in gastric cancer and the signaling pathway mediating HER2-upregulation-induced EMT in cisplatin-resistant gastric cancer cells also need to be elucidated in the future.
In conclusion, this study links HER2 overexpression-mediated EMT with cisplatin resistance in gastric cancer for the first time. Our study also suggests that targeting both HER2 and Snail would be a useful therapeutic strategy for the prevention of gastric cancer metastasis. Quantitative PCR (qPCR) analysis of HER2 amplification. qPCR was performed on an ABI7300 thermocycler (Life Technologies Corporation) by a previously described method 54 . Primers of HER2 and IGF1 (synthesized by Shanghai Sangon) were as follows: HER2 forward primer 5′ -GAACTGGTGTATGCAGATTGC-3′ , HER2 reverse primer 5′ -AGCAAGAGTCCCCATCCTA-3′ , IGF1 forward primer 5′ -AGCTCGGCATAGTCTT-3′ , IGF1 reverse primer 5′ -CCAAGTGAGGGGTGTGA-3′ . The standard amplification protocol consists of an initial denaturation step at 95 °C for 10 min, followed by 50 amplification cycles at 95 °C for 10 sec, 55 °C for 5 sec, and 72 °C for 10 sec.

Methods
Cell Mobility Assay. Cell mobility was assessed using a wound healing assay. The parental and cisplatin-resistant gastric cancer cells were seeded into six-well plates and cultured to monolayers, which were then wounded using sterile pipette tips and any cellular debris was removed by washing with PBS. The wounded monolayers were then incubated in medium containing 10% FBS for 24 h. Photos were captured at 0 and 24 h after wounding.
Cell migration assay. Tissue culture-treated 6.5-mm Transwell chamber with 8.0-μ m pore membranes (Corning, Corning, NY, US) was used in the cell migration assay. The assay was performed as described in 55 . Immunofluorescence. Cells were grown on the coverslips and fixed with 4% fresh paraformaldehyde for 10 min at room temperature, followed by blocking with 5% bovine serum albumin (BSA) at room temperature for 1 h. Next, cells were stained with FITC-Phalloidin for 1 h at room temperature in dark. After washing, cells were counterstained with DAPI for 10 min and mounted on 50% glycerol/PBS. A Leica SP2 confocal microscope (Leica Microsystems, Dresden, Germany) was used to observe the distribution of F-actin.
The MTT cell proliferation assay. 1 × 10 4 cells were plated in 96-well plates before experiments. After HER2 siRNA transfection for 24 h, 100 μ g/ml Herceptin pre-treatment for 2 h or 10 μ M CP724714 pre-treatment for 2 h, different concentrations of cisplatin (0.1, 0.5, 2.5, 12.5 and 62.5 μ g/ml ) were applied for 72 h. Next, 10% MTT was added to the wells. After incubated at 37 °C for 4 h, the mixture reacted in culture media was centrifuged at 4000 rpm for 15 min. The precipitates were dissolved in DMSO. Absorbance was measured using a multiplate reader at the wavelength of 490 nm. Cell Proliferation Rate = Absorbance of drug treatment panel/ Absorbance of control panel × 100%. Resistance Index = IC 50 value of resistant cells/IC 50 value of parental cells.

Specimen and immunohistochemistry (IHC). The study with clinical samples was approved by the
Medical Ethic Committee of Zhejiang Provincial People's Hospital. One cohort of formalin fixed and paraffin embedded (FFPE) gastric cancer tissues were obtained from archives of the Department of Pathology, Zhejiang Provincial People's Hospital, which had 382 cases collected during 1998-2004. Written Informed Consents were obtained from all patients prior to operation. Specimens were diagnosed histopathologically and staged according to the TNM-International Union against Cancer classification system. IHC was performed to detect HER2 and Snail. The degree of immunostaining was reviewed and scored independently by two pathologist (Z.Z. and G.R.) based on the intensity of staining: 0 (no staining), 1 (weak staining = light yellow), 2 (moderate staining = yellow brown), and 3 (strong staining = brown). Moderate and strong stainings were defined as positive, while no and weak stainings were defined as negative.
Fluorescence in situ hybridization (FISH). FISH was performed on fixed-paraffin-embedded samples in the same cohort containing 382 gastric cancer cases. FISH experiments were performed according to the protocol given by the supplier (PathVysion kit, Vysis, Downers Grove, IL, US). The centromeric probe of chromosome 17 was included in FISH analyses in 382 cases. In these cases, HER2 amplification was determined as a ratio of HER2 and chromosome 17 centromere signal counts. Ratios ≤ 2 were determined as no amplification, and those > 2 were defined as positive amplification Human rights statement and informed consent. The study with clinical samples was approved by the Medical Ethic Committee of Zhejiang Provincial People's Hospital and was conducted in accordance with the Helsinki Declaration. Informed consent was obtained from all patients for being included in the study.
Statistical analysis. Data were presented as mean ± SD. The differences were analyzed by ANOVA using SPSS 22.0 software and P < 0.05 was considered statistically significant. IC 50 values were calculated using Graphpad Prism software.