Receptor tyrosine kinase , RON , promotes tumor progression by regulating EMT and the MAPK signaling pathway in human oral squamous cell carcinoma

The receptor tyrosine kinase, recepteur d'origine nantais (RON), is known to be associated with the progression, metastasis, and prognosis of various types of cancers. Nevertheless, the role of RON in human oral squamous cell carcinoma (OSCC) is unclear. This study evaluated whether RON affects oncogenic behavior, oncogenic signaling pathways, and clinical outcomes, including survival, in human OSCC. Reverse transcription-PCR, quantitative PCR, western blotting and immunohistochemical staining were used to determine mRNA and protein expression levels of RON. Cell invasion, migration and apoptosis assays were used to assess the functional effects of small interfering RNA-mediated knockdown of RON or snail family transcriptional repressor 2 (SLUG). RON knockdown suppressed tumor cell invasion and migration and enhanced apoptosis in human OSCC cells. RON knockdown also decreased the phosphorylation of MAPK signaling proteins, such as ERK1/2, JNK and p38. In addition, RON knockdown suppressed the expression of the epithelial mesenchymal transition (EMT)-related transcription factor, SLUG. SLUG knockdown blocked the enhancement of cell invasion and migration induced by macrophage-stimulation protein (MSP)-mediated RON activation in OSCC cells. The cell morphology was changed to spindle-like shape under MSP-mediated RON activation in OSCC cells. RON was overexpressed in both fresh and paraffin-embedded human OSCC tissues. Taken together, these results indicate that RON contributed to tumor progression by regulating the EMT-related factor, SLUG, and the MAPK pathway in OSCC. This study may provide a theoretical basis for the application of RON-targeting agents, currently being studied in various cancer fields, for the treatment of OSCC.


Introduction
Oral cancer is the 11th most prevalent cancer worldwide and the 3rd most common cancer in developing nations (1,2). Squamous cell carcinoma represents ~90% of oral cancer cases. The standard treatment strategies for patients with resectable advanced-stage oral squamous cell carcinoma (OSCC) are surgery following adjuvant radiotherapy or concurrent chemoradiotherapy (3). Although significant technological advances in the diagnosis and treatment of patients with OSCC have been made in last few decades, prognosis remains poor, with a 5-year overall survival rate of 50-60%, becoming even lower in advanced stages (2). OSCC frequently leads to lymphatic metastasis, which is associated with the depth of invasion of the primary tumor. This represents the most important predictor of survival in patients with OSCC and has led to a major change in the T stage classification described in the 8th edition of the American Joint Committee on Cancer (4,5). The revised staging system now reflects not only tumor size, but also the depth of invasion (4). Therefore, in order to provide more effective treatment for patients with OSCC and to improve their clinical outcomes, an increased understanding of the molecular and biological mechanisms of tumor invasion and metastasis in OSCC is required.
The receptor tyrosine kinase, recepteur d'origine nantais (RON), also known as macrophage-stimulating 1 receptor (MST1R), is a member of the MET proto-oncogene family that includes c-Met (6). RON is activated by the ligand macrophage-stimulation protein (MSP) (7), and it has essential functional roles during embryonic development and organogenesis (8,9). As with other proteins involved in embryonic development, RON is associated with carcinogenesis, cancer progression and metastasis. RON is overexpressed in various cancers and its overexpression is associated with poor prognosis in several types of cancer (10)(11)(12)(13)(14)(15)(16)(17)(18). Thus, RON has received much interest in the fields of cancer biology and cancer therapy targets over the past 20 years. Agents targeting RON and/or c-Met for cancer therapy, such as low molecular weight kinase inhibitors or monoclonal antibodies that neutralize the receptors or ligands, are in various phases of clinical trials and/or preclinical testing (19)(20)(21). Nevertheless, little is known about the role of RON in human OSCC.
The present study assessed whether RON affects aggressive cancer cell behaviors of human OSCC cells, such as invasion, migration and apoptosis inhibition and the oncogenic signaling pathways involved in these changes, including epithelial-mesenchymal transition and mitogen-activated protein kinase (MAPK) signaling pathways. The expression of RON in human OSCC tissues was analyzed by immunohistochemistry and reverse transcription-quantitative PCR (RT-qPCR). The potential association between RON expression and various clinicopathological variables, including survival, was evaluated in a well-defined series of OSCCs with complete long-term follow-up data. This study may provide an important basis for the application of new RON-targeting therapies for the treatment of OSCC.

Materials and methods
Cell culture and transfection. The human OSCC cell line, SCC25, was purchased from the American Type Culture Collection and the human OSCC cell line, PCI50, was provided by Dr Sung (Seoul National University, Seoul, South Korea). SCC25 cells were cultured in DMEM/F12 medium (Gibco; Thermo Fisher Scientific, Inc.) and PCI50 cells were cultured in DMEM medium (Gibco; Thermo Fisher Scientific, Inc.), supplemented with 10% FBS (Gibco; Thermo Fisher Scientific, Inc.) in a humidified atmosphere of 5% CO 2 at 37˚C. Images of the cells were captured using the EVOS FL microscope (EVOS ® FL; Thermo Fisher Scientific, Inc.). Small interfering RNAs (siRNAs) were used to knockdown the endogenous gene expression of RON or snail family transcriptional repressor 2 (SLUG) in OSCC cells. OSCC cells were seeded in 6-well plates at a density of 2x10 5  ) horseradish peroxidase (HRP)-conjugated secondary antibodies were diluted 1:2,000 and incubated with membranes at room temperature for 2 h. Immunoreactive proteins were visualized using an enhanced chemiluminescence detection system for HRP (EMD Millipore), and images were showed using an LAS-4000 luminescent image analyzer (Image Leader LAS 4000IR, Fujifilm). Band densitometry was analyzed using an Multi gauge V3.2 (Fujifilm).
Cell invasion assay. Cell invasion ability was determined by the number of cells that invaded through a Transwell invasion apparatus with 8.0-µm pores (Costar; Corning, Inc.). At 48 h after transfection, cells transfected with RON or SLUG siRNA or a negative control siRNA were seeded at 3x10 5 cells in 120 µl medium containing 0.2% BSA (BioShop, Inc.) suspension in the upper chamber. The lower chamber was filled with 400 µl medium containing 0.2% BSA containing 100 ng/ml MSP (cat. no. 352-MS; R&D Systems, Inc.) into the lower chamber as the chemoattractant. After incubation for 24 h, cells that had moved to the bottom Transwell surface were stained with Diff-Quik solution (Sysmex Corporation) and were counted in five random squares in the microscopic field of view. Results are presented as the mean ± SE of the number of cells/field from three individual experiments.
Cell migration assay (wound healing assay). At 24 h after transfection, cells transfected with a RON-specific, SLUG-specific or negative control siRNA were seeded in each well of Culture-Inserts (Ibidi GmbH) at 1.5x10 5 cells/well. After incubation for 24 h, each insert was detached and the progression of cell migration was ascertained by imaging at 0, 8, 12 and 24 h, using an inverted microscope. Distances between gaps were normalized to 1 cm after capture of three random sites.
Apoptosis assay. Apoptosis was determined using an Annexin V-FITC assay (cat. no. 556547; BD Biosciences). At 48 h after transfection, cells transfected with a RON-specific siRNA or a negative control siRNA were collected using trypsin, washed twice in PBS, and re-suspended in binding buffer (BD Biosciences). Annexin V-FITC and 7-amino-actinomycin D (BD Biosciences) were added to the cells, which were incubated in the dark for 15 min and then re-suspended in 400 ml binding buffer. Cells were analyzed using a FACSCalibur flow cytometer (BD Biosciences; Becton, Dickinson and Company) and BD CellQuest version 3.3 software (Becton, Dickinson and Company). Data analysis was performed using WinMDI version 2.9 (The Scripps Research Institute).
Patients and tumor specimens. To evaluate RON protein expression, paraffin-embedded tissue sections were collected from 99 patients who had undergone a diagnostic biopsy or definitive surgery for OSCC at Chonnam National University Hwasun Hospital (Jeonnam, Korea) between May 2004 and August 2013. None of the collected tissues were obtained after radiotherapy or chemotherapy. Of these, 10 patients were excluded because of follow-up loss or palliative treatment intent. A total of 67 men and 22 women participated in the present study with a mean age of 62.7±12.8 years (range, 26-87 years). Of the 89 patients, 86 were treated with definitive surgery, with or without adjuvant radiotherapy, or with cisplatin-based concurrent chemoradiotherapy (CRT). Three patients that refused surgery were treated with induction chemotherapy, followed by cisplatin-based concurrent CRT, with curative intent. The incidence of recurrence included locoregional recurrence and/or distant metastasis after primary treatment with curative intent. Patients with locoregional recurrence underwent salvage surgery or CRT. Treatment failure was defined as disease with inoperable locoregional progression or distant metastasis, even with salvage treatment. Patients' clinicopathologic characteristics and complete medical history, were reviewed in hospital records. Tumors were staged according to the 7th edition of the American Joint Committee on Cancer staging system (23). Survival duration (months) was measured from the date of starting treatment to the date of death or the date last seen. Disease-free survival (DFS) was measured to the first observed date of any recurrence or death and disease-specific survival (DSS) to the date of OSCC-related death.
To evaluate RON mRNA expression, fresh OSCC tissues and paired normal oral mucosa were collected from 20 patients that underwent diagnostic biopsy or definitive surgery of OSCC at CNUHH. A total of 12 men and 8 women participated in the present study with a mean age of 66.2±15.4 years (range, 23-86 years). RON mRNA expression was analyzed by RT-PCR and RT-qPCR, performed as described above. Due to the small amount of fresh tissue obtained during diagnostic biopsy, western blotting could not be performed to measure RON protein expression.
Patient dataset from The Cancer Genome Atlas (TCGA). The TCGA (portal.gdc.cancer.gov/) head and neck cancer dataset included 530 samples across all anatomical sites of the head and neck. For this study, only OSCC patients with DFS status information were selected, eliminated duplicate samples of the same patient in the dataset, and finally included 226 patients in this analysis. RNA-seq data was analyzed as expected maximization (RSEM) values of MST1R expression, which were log2 transformed (RSEM+1). The mean of MST1R expression levels was used as a cutoff value for classification into low or high RON expression groups in survival analyses.
Immunohistochemistry. Tissue processing and immunohistochemical analysis were performed as previously described (22). Tissue sections were incubated with primary antibodies against RON (cat. no. sc-25781; Santa Cruz Biotechnology). Two independent observers interpreted RON staining with no knowledge of the associated clinical information. Staining intensity was scored as follows: 0, no staining of tumor cells; 1+, weak-to-comparable staining in the cytoplasm and/or nucleus, compared to non-tumor cell staining; 2+, readily appreciable or dark brown staining distinctly marking the tumor cell cytoplasm and/or nucleus. Percentages of stained cells were scored as follows: 0, 0%; 1, 1-25%; 2, 26-50%; 3, 51-75%; and 4, >75%. Final staining scores were the product of the intensity and percentage scores, with ≤4 defined as low RON expression and >4 defined as high RON expression.
Statistical analysis. The significance of experimental differences was assessed using Student's t-test. Association between RON expression and various clinicopathological parameters were compared using the χ 2 test or Fisher's exact test. Survival curves were calculated by the Kaplan-Meier method and were compared using the log-rank test. Analyses were performed using SPSS version 21.0 (IBM, Corp.). P<0.05 was considered to indicate a statistically significant difference.

Knockdown of RON suppresses tumor cell invasion in human OSCC cells.
To examine the role of RON in tumor progression, the present study used siRNA to inhibit the endogenous expression of RON in the human OSCC cells, PCI50 and SCC25. The mRNA and protein levels of RON were lower in the RON-specific siRNA-treated PCI50 and SCC25 cells than in the negative control siRNA-treated cells (Fig. 1A).

Knockdown of RON suppresses tumor cell migration and RON activation by MSP enhances cell migration in human OSCC cells.
With time, the artificial wound gap became significantly narrower in plates of negative control PCI50 and SCC25 cells at 8, 12 and 24 h, compared with RON-knockdown cells (P<0.05; Fig. 2A and C). Wound gaps were not filled, even after 24 h, in plates of negative control PCI50 and SCC25 cells. Therefore,  in order to achieve greater activation of RON signaling, 100 ng/ml of the RON ligand, MSP, was applied to the plates of PCI50 and SCC25 cells. Migration of PCI50 and SCC25 cells increased after incubation with MSP and the artificial wound gap in plates of negative control cells became significantly narrower than the wound gap in RON-knockdown cells at 8, 12, and 24 h (P<0.05; Fig. 2B and C). This result demonstrated that ligand-mediated activation of RON enhanced cell migration.

Knockdown of RON enhances tumor cell apoptosis in human OSCC cells.
To investigate the effect of RON on apoptosis, we used an Annexin V apoptosis assay. Flow cytometric analysis indicated that RON knockdown significantly increased the proportion of apoptotic cells (Fig. 3A). The proportion of early and late apoptotic cells induced by transfection of RON-specific siRNA were greater than in the cells transfected with negative control siRNA (58.1 vs. 7.0% and 14.2 vs. 4.0%, respectively) in PCI50 and SCC25 cells. To confirm this effect of RON knockdown on tumor cell apoptosis, apoptosis regulatory proteins were evaluated after siRNA treatment. Levels of cleaved caspase-3, -7 and -9 and PARP were higher in RON-knockdown PCI50 and SCC25 cells, compared with negative control cells (Fig. 3B). These results indicate that RON knockdown induces tumor cell apoptosis by modulating apoptosis regulatory proteins, such as caspases-3, -7 and -9 and PARP in human OSCC cells.

factor, SLUG, and the phosphorylation of MAPK signaling proteins in human OSCC cells.
To examine the potential mechanisms involved in the effects of RON knockdown on aggressive tumor cell behavior, we investigated the impact of RON knockdown on the EMT-related protein, SLUG and the MAPK signaling proteins. SLUG is known to be involved in cell invasion and metastasis, and the MAPK signaling pathways are essential for cell growth and survival. RON knockdown reduced the expression of SLUG mRNA and protein in PCI50 and SCC25 cells (Fig. 4A). In addition, the phosphorylation levels of ERK1/2 and p38 were decreased significantly by RON knockdown in PCI50 and SCC25 cells. RON knockdown also decreased the levels of phospho-JNK in PCI50 cells, but not in SCC25 cells (Fig. 4B).

Knockdown of SLUG eliminates the effect of RON activation by MSP to aggressive cell behavior in human OSCC cells.
The mRNA and protein levels of SLUG were lower in SLUG-specific siRNA-transfected PCI50 and SCC25 cells compared with those in the negative control siRNA-transfected cells (Fig. 5A).
To verify the relationship between RON and SLUG, cell invasion and migration were compared in negative control OSCC cells and SLUG-knockdown OSCC cells after RON activation by treatment with 200 ng/ml MSP. The number of invading SLUG-knockdown PCI50 cells was 1.2±0.7, compared with 153.3±14.1 invading negative control PCI50 cells (Fig. 5B). The number of invading SLUG-knockdown SCC25 cells was 5.5±1.4, compared with 170.1±14.3 invading negative control SCC25 cells (Fig. 5B). The difference between the two groups was statistically significant (P<0.05). With time, the artificial wound gap became significantly narrower in plates of negative control PCI50 and SCC25 cells at 4, 8 and 12 h, compared with SLUG-knockdown cells after MSP-induced RON activation (P<0.05; Fig. 5C). This result demonstrated that SLUG knockdown blocks the enhancement of cell migration induced by RON activation in OSCC cells.

RON activation by its ligand, MSP, induces morphological changes in human OSCC cells.
To demonstrate the role of RON in EMT, morphological changes in PCI50 and SCC25 cells treated with 500 ng/ml MSP for 48 and 72 h were analyzed. Light microscopy showed a scattering effect, indicating cell-cell dissociation (Fig. 6). The individual cell morphology also was changed to an elongated and spindle-like shape.
These findings suggested that MSP-induced RON activation is involved in the process of EMT at the cellular level.
Expression of RON is high in human OSCC tissues. The patients in this study included 67 men and 22 women. The mean age was 62.7±12.8 years (mean ± SD), with a range of 26-87 years. The mean patient follow-up period was 59.5±38.1 months, with a range of 3.6-156.3 months. RON protein expression was investigated by immunohistochemical staining of formalin-fixed, paraffin-embedded biopsy tissue obtained from 89 patients with OSCC. Immunostaining patterns were heterogenous, with predominantly nuclear and/or cytoplasmic staining in tumor cells, but weak or no staining in the normal adjacent tissues (Fig. 7). Based on the grading criteria, high RON expression was observed in 43.8% of patients (39/89) and low RON expression was observed in 56.2% of patients (50/89).
Patient data and the correlation between RON expression and various OSCC clinicopathological variables are shown in Table I. RON expression in OSCC was not associated with age, sex, tumor location, tumor stage, T stage, N stage or chemoradiotherapy treatment (P>0.05). Patients with high RON expression tended to have a second primary malignancy more often than those with low RON expression [24.0% (12/50) in the low-expression group vs. 41.0% (16/39) in the high-expression group]. This may be weak evidence that RON is involved in the carcinogenesis of human cancers. Patients with high RON expression had marginally higher rates of recurrence and treatment failure rate than those with low RON expression, but this was not a significant difference (P>0.05; Table II).
For the 89 patients with OSCC enrolled in this study, the 3/5-year DSS was 68/67% and the 3/5-year DFS was 64/58%. The 3-year DSS/DFS rates in patients with high RON expression were lower than those with low RON expression (72/70% in the low-expression group vs. 64/56% in the high-expression group); however, analysis of Kaplan-Meier curves for DSS/DFS using the log-rank test, did not demonstrate a significant difference between these two groups of patients (P>0.05; Fig. 8). For the 226 patients with OSCC in the TCGA dataset, the 3-year DFS rates in patients with high RON expression were lower than in those with low RON expression (60% in the low-expression group vs. 51% in the high-expression group). Similar to the previous results of patients from Chonnam National University Hwasun Hospital, the TCGA data showed no statistically significant association was found between RON expression and DFS (P>0.05; Fig. 9).
To provide further support for the immunohistochemical data on RON protein expression, RON mRNA expression was analyzed using using RT-PCR and RT-qPCR in 20 fresh human OSCC tissue samples and paired normal oral mucosa samples. RON mRNA expression was significantly higher in OSCC tissue compared with paired normal oral tissue (P<0.01; Fig. 10).

Discussion
OSCC typically has locoregional progression, and as it grows, it invades adjacent tissues and undergoes regional lymphatic metastasis, but rarely develops distant metastases (5). The depth of tumor invasion in OSCC is an important parameter for predicting lymphatic metastasis and patient survival (5). The depth of tumor invasion affects adjuvant treatment decisions, including the option of elective lymph node dissection. These are characteristics specific for OSCC, which differ from other types of head and neck cancers. Thus, the discovery of molecular targets involved in tumor invasion in OSCC patients is highly important.
RON is an attractive molecular target for OSCC therapy. In this study, RON knockdown reduced tumor cell invasion and migration. Conversely, RON activation by MSP, enhanced cell migration in human OSCC cells. Several previous studies support these findings. Maggiora et al (11) reported that overexpression of RON increased cell invasion and migration, and that MSP caused a 12-fold increase in migration of human breast carcinoma cells. In human colorectal cancer cells and hepatocellular cancer cells, RON knockdown suppresses tumor cell invasion and migration (24,25). These findings suggest that RON-targeting therapy may control the invasiveness of cancer cells in OSCC.
RON has an important role in EMT. EMT has been highlighted as a key regulator of metastasis, by promoting tumor cell invasion. The EMT process is tightly controlled through a complex system of regulation of EMT-related transcription factors, such as Snail1, SLUG, Zeb1, Zeb2 and  Twist1, which are responsible for inducing and sustaining the mesenchymal phenotype (26). Preliminary experiments were performed to determine the association between RON and various EMT-related transcription factors. Unfortunately, EMT-related factors, other than SLUG, showed variable results that were difficult to interpret among different cell line types, cell conditions and cell preparation timings. Additionally, it was difficult to obtain reproducibility in repeated experiments, even under the same conditions. Therefore, the experiments were focused on the association between RON and SLUG. RON activation by MSP induced cell morphological changes to spindle shape and RON knockdown suppressed SLUG expression. In addition, the promotion of cell invasion and migration by MSP-mediated RON activation was suppressed by SLUG knockdown. These findings suggest that increased Table I. Association between RON expression and clinicopathological parameters in patients with oral squamous cell carcinoma.

RON expression --------------------------------Parameters
Total The MAPK pathway, which includes ERK1/2, JNK and p38, plays a key role in various steps of cancer development (27). These proteins are involved in controlling cell proliferation, differentiation, migration and apoptosis in response to a variety of stresses, including cancers. In this study, RON knockdown suppressed the activation of ERK1/2, JNK and p38 in OSCC cells. In addition, RON knockdown was associated with the reversal of aggressive tumor cell behavior, such as apoptosis inhibition, cell invasion and cell migration in OSCC cells. Although additional evaluation may be required to clarify this mechanism, these findings suggested that RON knockdown inhibits aggressive tumor cell behavior by regulating the MAPK pathway in OSCC.
RON is involved in EMT through the MAPK pathway. Several studies have reported that RON activation induces spindle-shaped cell morphology and that RON is essential for the expression of a specialized mesenchymal marker (28)(29)(30)(31). In addition, a MAPK inhibitor restored the original morphology of the epithelial cells or blocked the expression of EMT-related factors (28)(29)(30). These findings support the hypothesis that RON is associated with aggressive tumor cell behavior and is involved in EMT through the MAPK pathway in OSCC.
The abundant expression of a targeted molecule is the first requirement for an effective molecular target therapy. Elevated RON expression has been found to be associated with poor prognosis in multiple types of cancer, such as breast, stomach, colon, lung, bladder, ovary and cholangiocarcinoma (14,15,(32)(33)(34)(35)(36). It was also demonstrated that RON is overexpressed in both fresh OSCC tissue and paraffin-embedded tissue from human OSCC. However, contrary to expectations, there was no significant association between RON expression and prognostic indicators, including recurrence, treatment failure and survival in human OSCC. Although this is a limitation of our study, this is possible because the treatment outcomes of patients with cancer are affected by many factors, such as multiple genetic mutations, various other signaling pathways and specific microenvironments that contribute to cancer progression.
Several different splice variant isoforms of RON have been identified in several cancers (37)(38)(39)(40). They are involved in RON activation in tumor cells and contribute to oncogenic activities (37). More importantly, efforts to block the RON activity, using low molecular weight kinase inhibitors or monoclonal antibodies, can be disrupted by unknown splice variants, because, despite having similar sequences, they exhibit different localizations and have varied functions (38). Thus, the influence of the RON splice variants needs to be considered when developing RON-targeting anticancer therapies.
RON is a promising target in various types of cancer. Many studies and clinical trials have been conducted to overcome cancer by inhibiting RON. The phase I/II clinical trials of the tyrosine kinase inhibitor, foretinib, multi-targeting c-Met, vascular endothelial growth factor 2 and RON, and anti-RON monoclonal antibody, narnatumab, have reported for advanced solid tumors, although the clinical trials were abandoned (21,(41)(42)(43). Several studies recently reported that the anti-RON antibody Zt/g4-drug conjugate (ADC) is effective in targeted inhibition of colorectal cancer, bladder cancer, triple-negative breast cancer and pancreatic cancer (44)(45)(46)(47). In addition, Ruiz-Torres et al (48) demonstrated that the RON signaling pathway is associated with the immunosuppressive microenvironment in breast cancer. Ekiz et al (49) described that inhibition of RON kinase potentiates checkpoint immunotherapy in breast cancer. It is expected that the effectiveness of RON targeting anticancer therapy can be  improved through the ADC technique and/or the combination therapy with immunotherapy.
In summary, RON knockdown suppressed cell invasion, migration and inhibition of apoptosis mediated by the MAPK pathway and the EMT-related factor, SLUG, in human OSCC cells. RON was overexpressed in human OSCC tissue. The results of the present study provide a theoretical basis for the application of RON-targeting agents, under active investigation, for the treatment of OSCC.