Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Vimentin is a potential prognostic factor for tongue squamous cell carcinoma among five epithelial–mesenchymal transition-related proteins

  • Pei-Feng Liu ,

    Contributed equally to this work with: Pei-Feng Liu, Bor-Hwang Kang

    Affiliations Department of Medical Education and Research, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan, Department of Biotechnology, Fooyin University, Kaohsiung, Taiwan

  • Bor-Hwang Kang ,

    Contributed equally to this work with: Pei-Feng Liu, Bor-Hwang Kang

    Affiliations Department of Otorhinolaryngology-Head and Neck Surgery, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan, Graduate Institute of Aerospace and Undersea Medicine, National Defense Medical Center, Taipei, Taiwan

  • Yi-Min Wu,

    Affiliations Graduate Institute of Oral Health Sciences, College of Dental Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan, Division of Periodontics, Department of Dentistry, Kaohsiung Medical University Hospital, Kaohsiung, Taiwan

  • Ju-Hsin Sun,

    Affiliations Graduate Institute of Oral Health Sciences, College of Dental Medicine, Kaohsiung Medical University, Kaohsiung, Taiwan, Department of Medical Management, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan

  • Liang-Ming Yen,

    Affiliation Department of Pathology and Laboratory Medicine, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan

  • Ting-Ying Fu,

    Affiliation Department of Pathology and Laboratory Medicine, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan

  • Yun-Chung Lin,

    Affiliation Department of Pathology, China Medical University Hospital, Taichung, Taiwan

  • Huei-Han Liou,

    Affiliation Department of Medical Education and Research, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan

  • Yaoh-Shiang Lin,

    Affiliations Department of Otorhinolaryngology-Head and Neck Surgery, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan, Department of Otolaryngology-Head and Neck Surgery, National Defense Medical Center, Taipei, Taiwan

  • Huei-Cin Sie,

    Affiliation Department of Pathology and Laboratory Medicine, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan

  • I-Chien Hsieh,

    Affiliation Department of Pathology and Laboratory Medicine, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan

  • Yu-Kai Tseng,

    Affiliations Department of Orthopedics, National Cheng Kung University Hospital, Tainan, Taiwan, Department of Orthopedics, Show Chwan Memorial Hospital, Changhua, Taiwan

  • Chih-Wen Shu,

    Affiliation Department of Medical Education and Research, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan

  • Yao-Dung Hsieh ,

    lpger@vghks.gov.tw (LPG); mrusmrus@gmail.com (YDH)

    Affiliations Department of Stomatology, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan, Department of Dentistry, Kaohsiung Veterans Gegeral Hospital, Pingtung Branch, Pingtung, Taiwan

  • Luo-Ping Ger

    lpger@vghks.gov.tw (LPG); mrusmrus@gmail.com (YDH)

    Affiliations Department of Medical Education and Research, Kaohsiung Veterans General Hospital, Kaohsiung, Taiwan, Institute of Biomedical Sciences, National Sun Yat-Sen University, Kaohsiung, Taiwan

Vimentin is a potential prognostic factor for tongue squamous cell carcinoma among five epithelial–mesenchymal transition-related proteins

  • Pei-Feng Liu, 
  • Bor-Hwang Kang, 
  • Yi-Min Wu, 
  • Ju-Hsin Sun, 
  • Liang-Ming Yen, 
  • Ting-Ying Fu, 
  • Yun-Chung Lin, 
  • Huei-Han Liou, 
  • Yaoh-Shiang Lin, 
  • Huei-Cin Sie
PLOS
x

Abstract

We aimed to investigate the association of the expression levels of five epithelial–mesenchymal transition (EMT)-related proteins (Snail, Twist, E-cadherin, N-cadherin, and Vimentin) with tumorigenesis, pathologic parameters and prognosis in tongue squamous cell carcinoma (TSCC) patients by immunohistochemistry of tissue microarray. The expression levels of Snail, E-cadherin, N-cadherin and Vimentin were significantly different between the tumor adjacent normal and tumor tissues. In tumor tissues, lower E-cadherin and higher N-cadherin levels were associated with a higher grade of cell differentiation, advanced stage of disease, and lymph node metastasis. However, higher Vimentin expression was associated with poor cell differentiation and lymph node metastasis. Patients with low E-cadherin expression had poor disease-specific survival (DSS). Conversely, positive N-cadherin and higher Vimentin expression levels were associated with poor DSS and disease-free survival. Notably, our multivariate Cox regression model indicated that high Vimentin expression was an adverse prognostic factor for DSS in TSCC patients, even after the adjustment for cell differentiation, pathological stage, and expression levels of Snail, Twist, E-cadherin, and N-cadherin. Snail, E-cadherin, N-cadherin, and Vimentin were associated with tumorigenesis and pathological outcomes. Among the five EMT-related proteins, Vimentin was a potential prognostic factor for TSCC patients.

Introduction

Tongue squamous cell carcinoma (TSCC) is one of the most common cancers of the oral cavity. It is more aggressive than other forms of oral cancer, with a propensity for rapid local invasion and spread [1] along with a high recurrence rate [2]. Because the tongue is rich in lymphatic vessels and neurovascular bundles, the incidence of neck nodal metastasis is higher [3]. Despite overall improvements in surgical and medical management, the clinical outcomes of TSCC patients have remained unchanged, indicating the immediate need for new prognostic biomarkers for TSCC patients.

Epithelial—mesenchymal transition (EMT), recently identified as a key process in carcinogenesis, invasion, and metastasis, is a predictor of TSCC progression [4]. It plays a critical role in promoting metastasis in epithelial carcinoma, accompanied by a decrease in the expression of epithelial markers, including cell-surface E-cadherin, and an increase in the expression of mesenchymal markers, such as transcription factors Snail and Twist, and cell-surface N-cadherin, as well as cytoskeletal Vimentin [5,6]. E-cadherin, a calcium-dependent transmembrane glycoprotein, is expressed in most epithelial cells for cell polarity and tissue structure [7]. Snail and Twist are transcription factors that repress E-cadherin expression in epithelial tumors to promote metastasis by disassembling cellular adhesion junctions [8]. N-cadherin induces a mesenchymal-scattered phenotype associated with reduced E-cadherin levels in squamous cell carcinoma [9]. Vimentin is an intermediate filament protein that is ubiquitously expressed in normal mesenchymal cells to maintain the cellular architecture and tissue integrity, and it also participates in tumorigenesis, EMT, and the metastatic spread of cancer [10].

Many studies have used immunohistochemistry to investigate the expression levels of EMT-related biomarkers in TSCC patients. However, the tumorigenic and prognostic significance of these biomarkers was studied in a relatively small cohort with short follow-up [1114]. Moreover, combinatorial assessment of Snail, Twist, E-cadherin, N-cadherin, and Vimentin expression levels in tumorigenesis and prognosis of TSCC has not been reported previously. In this study, the expression levels of these five EMT-related markers and their correlations with tumorigenesis, pathological outcomes, and survival were extensively investigated in 248 TSCC patients.

Materials and methods

Patients and tissue subjects

Specimens of TSCC tissues (n = 248) and corresponding tumor adjacent normal (TAN, n = 235) tissues were obtained from the Department of Pathology, Kaohsiung Veterans General Hospital between 1993 and 2006. The 2002 American Joint Committee on Cancer (AJCC) system was used to classify pathological TNM stages at the time of initial resection of the tumor. Informed consent forms were provided by all patients involved in this study, and the protocol of this study was approved by the Institutional Review Board at Kaohsiung Veterans General Hospital (IRB number: VGHKS11-CT12-13).

Tissue microarray construction

The detailed procedures of tissue microarray (TMA) construction were described in our previous study [15]. The TMA block consisted of 129 cores, including 43 trios with each trio containing 2 cores from the tumor tissue and 1 core from the non-cancerous epithelium of the same patient. After exclusion of incorrectly identified tissues, cores with too few normal (or tumor) cells (<20 cells), missing cores and those of poor quality, only 485 tumor cores and 212 TAN cores were stained or scored in this study. After construction, 4-μm-thick sections of the TMA blocks were cut and processed for immunohistochemistry.

Immunohistochemistry (IHC)

IHC of all proteins was performed using the Novolink Max Polymer Detection System (Leica, Newcastle Upon Tyne, United Kingdom). The sections were deparaffinized in xylene, rehydrated in graded ethanol, and washed for 5 min with phosphate-buffered saline (PBS). The sections were boiled and immersed in Tris-EDTA (10 mM, pH 9.0) for 10 min at 125°C for antigen retrieval. Endogenous peroxidase activity was blocked for 30 minutes with 3% hydrogen peroxide in methanol. The sections were incubated with primary antibodies [anti-Snail, Abcam (ab53519), 1:100; anti-Twist, Santa Cruz (sc-15393), 1:50; anti-E-cad, BD Biosciences (610181), 1:200; anti-N-cad, Leica (NCL-L-N-Cad), 1:200; anti-Vimentin, SPRING (M3200), 1:400] overnight at 4°C in a humidified chamber after blocking. The sections were incubated with horseradish peroxidase-labeled secondary antibody for 10 minutes at room temperature after washing in PBS and counterstained with hematoxylin [16].

IHC analysis and scoring

The scoring of IHC staining was performed by a senior technician under the supervision of an oral cancer pathologist as described in our previous study [15]. As a first step, digital images of the IHC-stained TMA slides were evaluated by an oral cancer pathologist (Dr. Ting-Ying Fu) along with a senior technician (Mr. Ju-Hsin Sun) until all discrepancies were resolved. Subsequently, the technician independently reviewed all slides, except for cores with incorrect and uncertain content, which need to be scored by the pathologist. Afterward, the pathologist randomly re-evaluated 15% of the core samples. If the agreement of IHC scores between the pathologist and technician was < 95%, the pathologist evaluated slides again with the technician until all discrepancies were resolved. After that, the senior technician re-scored all slides until the scoring agreement was ≥ 95% among random core samples (5–20%) evaluated by the pathologist. Approximately 45% of the cores were scored (for those that were incorrect, special, or uncertain) or re-evaluated (for random core samples) by the pathologist. In the final re-evaluation of the IHC scores, the agreement between the pathologist and technician for Snail, Twist, E-cadherin, N-cadherin, Vimentin expression was 96%, 96%, 97%%, 98%, and 95%, respectively. A semi-quantitative approach was used to grade the immunoreactivity. The intensity score of nuclear staining (for Snail and Twist), cell membranous staining (for E-cadherin and N-cadherin), and cytoplasmic staining (Vimentin) was measured using a numerical scale (-, negative expression; +, weak expression; ++, moderate expression; and +++, strong expression; Fig 1A). The percentage of cells stained at each intensity level was graded as 0 (<5%), 1 (5–25%), 2 (26–50%), 3 (51–75%), and 4 (>75%). The intensity score and percentage of positive cells were multiplied to derive the final scores. For survival analysis, expression levels of all proteins were dichotomized into low expression and high expression with the cutoff value based on a receiver operating characteristic (ROC) curve analysis. The cutoff values were 4, 5, 3, 0, and 2 for Snail, Twist, E-cadherin, N-cadherin, and Vimentin, respectively. Moreover, the cutoff value of 0 was defined as negative expression, and the cutoff value of >0 was defined as positive expression for N-cadherin.

thumbnail
Fig 1. Immunoreactivity of Snail, Twist, E-cadherin, N-cadherin and Vimentin in TSCC.

(A) The representative photomicrographs for negative (–), weak (+), moderate (++) and strong (+++) staining of Snail, Twist, E-cadherin, N-cadherin and Vimentin in TSCC tissues. (B) Comparative representative photomicrographs of Snail, Twist, E-cadherin, N-cadherin and Vimentin between normal and TSCC tissues. (C) The representative photomicrographs of TMA sections for Snail, Twist, E-cadherin, N-cadherin and Vimentin staining.

https://doi.org/10.1371/journal.pone.0178581.g001

Statistical analysis

The Wilcoxon matched pairs signed rank test was used to evaluate the differential expression of Snail, Twist, E-cadherin, N-cadherin, and Vimentin between paired tissues (TSCC vs. paired TAN). The correlation between the expression score of each protein and pathologic parameters was evaluated with Kruskal-Wallis one-way ANOVA, one-way ANOVA, Mann-Whitney U test, or Student’s t-test. Disease-specific survival (DSS) was measured from the time of initial resection of the primary tumor to the date of cancer-specific death or the date of last follow-up (October 2010). Disease-free survival (DFS) was calculated from the date of initial resection of the primary tumor to the date of recurrence or that of last follow-up. The Kaplan-Meier method was used to estimate cumulative survival curves. The log-rank test was used to compare the different survival curves. The independent predictors of survival were determined by multivariate Cox proportional hazards modeling. P values < 0.05 were defined as statistically significant. The raw clinical data along with expression scores of five proteins are provided in the supporting information (S1 Table).

Results

Association of the expression levels of five EMT-related biomarkers with tumorigenesis

The levels of Snail, Twist, E-cadherin, N-cadherin, and Vimentin were analyzed in TSCC tissue cores and were compared with the levels of the respective proteins in paired TAN tissue cores (Fig 1B) on the TMA (Fig 1C). The results showed that higher expression levels of Snail (187 cases, p<0.001) and Vimentin (206 cases, p<0.001) and lower expression levels of E-cadherin (208 cases, p<0.001) and N-cadherin (210 cases, p<0.001) were found in TSCC tissues compared to the paired TAN tissues (Table 1).

thumbnail
Table 1. The pairwise comparisons of Snail, Twist, E-cadherin, N-cadherin, and Vimentin expression between TSCC and corresponding tumor adjacent normal tissues.

https://doi.org/10.1371/journal.pone.0178581.t001

Association of expression levels of the five EMT-related biomarkers with demographic and pathological outcomes of TSCC patients

The association of expression levels of EMT markers with various demographic and pathological outcomes was examined in TSCC tissues (Table 2). Snail expression was significantly increased in advanced stage disease (p = 0.037). Moreover, N-cadherin expression was significantly increased in samples connected with a higher grade of cell differentiation (p<0.001), advanced stage of disease (p = 0.023), and lymph node metastasis (p = 0.007). Similarly, Vimentin expression was significantly increased in samples connected with a higher grade of cell differentiation (p = 0.006) and with lymph node metastasis (p = 0.039). In contrast, decreased expression of E-cadherin was found in samples connected with a higher grade of cell differentiation (p<0.001), advanced stage of disease (p = 0.013), and with lymph node metastasis (p = 0.007).

thumbnail
Table 2. The comparisons of Snail, Twist, E-cadherin, N-cadherin, and Vimentin expression with various demographic and pathologic outcomes in TSCC patients.

https://doi.org/10.1371/journal.pone.0178581.t002

Association of expression levels of the five EMT-related biomarkers with survival in TSCC patients

In univariate analysis, the expression of Snail (Fig 2A and 2F) and Twist (Fig 2B and 2G) was not associated with DSS and DFS (Table 3). However, a low level of E-cadherin expression was correlated with worse DSS (p = 0.003, Fig 2C) but not associated with DFS (p = 0.095, Fig 2H). Moreover, patients with positive N-cadherin expression had poor DSS (p = 0.004, Fig 2D) and DFS (p = 0.009, Fig 2I). Patients with high Vimentin expression levels had poor DSS (p<0.001, Fig 2E) and DFS (p = 0.017, Fig 2J). After adjustment for cell differentiation and pathologic stage in the multivariate Cox regression model, high E-cadherin expression level was significantly correlated with longer DSS (adjusted hazard ratio [AHR] = 0.66; 95% CI = 0.45–0.97; p = 0.033; Table 3) but not with DFS (p = 0.20). Positive N-cadherin expression was still correlated with both shorter DSS (AHR = 1.88; 95% CI = 1.05–3.39; p = 0.034; Table 3) and DFS (AHR = 2.06; 95% CI = 1.11–3.81; p = 0.022; Table 3). Likewise, high Vimentin expression level was correlated with both shorter DSS (AHR = 2.05; 95% CI = 1.38–3.04; p<0.001; Table 3) and shorter DFS (AHR = 1.54; 95% CI = 1.01–2.34; p = 0.045; Table 3). These results confirmed that a high level of E-cadherin expression is an independent favorable prognostic factor for DSS, whereas positive N-cadherin and high Vimentin expression are independent adverse prognostic factors for both DSS and DFS in TSCC patients.

thumbnail
Fig 2.

Kaplan-Meier survival curves for disease-specific survival according to (A) Snail expression status, (B) Twist expression status, (C) E-cadherin expression status, (D) N-cadherin expression status, and (E) Vimentin expression status in TSCC patients. Kaplan-Meier curves for disease-free survival according to (F) Snail expression status, (G) Twist expression status, (H) E-cadherin expression status, and (I) N-cadherin expression status.

https://doi.org/10.1371/journal.pone.0178581.g002

thumbnail
Table 3. The expression levels of Snail, Twist, E-cadherin, N-cadherin, and Vimentin for disease-specific survival and disease-free survival of TSCC patients.

https://doi.org/10.1371/journal.pone.0178581.t003

The impact of E-cadherin on DSS and DFS was further analyzed by stratification based on demographic and pathological outcomes of patients, which has not been studied previously. Younger (≤50 yrs) patients with high level of E-cadherin expression had a longer DSS (AHR = 0.50; 95% CI = 0.30–0.83; p = 0.007, S2 Table). Moreover, a high level of E-cadherin was associated with better DSS in patients with moderate or poor cell differentiation (AHR = 0.66; 95% CI = 0.45–0.98; p = 0.038, S2 Table) and advanced stage of disease (III-IV, AHR = 0.57; 95% CI = 0.33–1.00; p = 0.048, S2 Table) as well as in patients who never received post-operative radiotherapy (AHR = 0.55; 95% CI = 0.34–0.89; p = 0.014, S2 Table). However, the expression level of E-cadherin was not associated with DFS, even after stratification analysis (S3 Table). Moreover, a high level of Vimentin was associated with poor DSS in younger patients (≤50 yrs; AHR = 2.53; 95% CI = 1.51–4.24; p<0.002, S4 Table), in patients with moderate or poor cell differentiation (AHR = 1.99; 95% CI = 1.34–2.97; p = 0.001, S4 Table), and in patients with advanced stage of disease (III-IV, AHR = 1.90; 95% CI = 1.07–3.38; p = 0.03, S4 Table) as well as in those without lymph node metastasis (N0; AHR = 1.97; 95% CI = 1.20–3.23; p = 0.008, S4 Table) but in those with smaller tumor size (T1-T2; AHR = 2.12; 95% CI = 1.34–3.37; p = 0.001, S4 Table). Additionally, high Vimentin expression was associated with poor DFS in patients with early pathological stage disease (I-II; AHR = 1.86; 95% CI = 1.13–3.08; p = 0.016, S5 Table) and in patients without lymph node metastasis (N0; AHR = 1.79; 95% CI = 1.11–2.88; p = 0.017, S5 Table). Nevertheless, the stratification analysis was not performed for N-cadherin expression due to the limited number of patients (n = 20) with positive expression.

Vimentin as a potential prognostic factor among the five EMT-related proteins in TSCC patients

To investigate the interaction of expression of Snail, Twist, E-cadherin, N-cadherin, and Vimentin as well as pathologic outcomes with survival, the data on the expression levels of these proteins and outcome (cell differentiation and pathologic stage) were put into a Cox regression model. The results showed that Vimentin was still a less favorable prognostic factor for DSS in TSCC patients (AHR = 1.90; 95% CI = 1.26–2.87; p = 0.002; Table 4) after adjustment for cell differentiation, pathological stage, and expression of Snail, Twist, E-cadherin, and N-cadherin.

thumbnail
Table 4. Multivariate Cox’s regression analysis of Snail, Twist, E-cadherin, N-cadherin, and Vimentin for disease-specific survival and disease-free survival of TSCC patients.

https://doi.org/10.1371/journal.pone.0178581.t004

Discussion

EMT plays important roles in cancer invasion and metastasis, involving Vimentin as a cytoskeletal regulator of migration of mesenchymal cells and cell-surface E/N-cadherin switch (the expression of E-cadherin is switched off and that of N-cadherin is switched on) mediated by key transcription factors such as Snail and Twist [8]. In this study, we found that Snail, E-cadherin, N-cadherin, and Vimentin were involved in tumorigenesis and the pathologic outcomes of TSCC. We also determined their impact on the survival of TSCC patients with stratification according to various pathologic factors that had previously not been used for such analyses. Most importantly, after evaluating the interactions of E-cadherin, N-cadherin, and pathologic outcomes with survival by Cox regression model, we identified Vimentin as a potential prognostic factor for DSS.

Snail promotes EMT and results in progression and cancer motility in oral SCC [17]. In this study, we found that higher Snail expression was significantly associated with not only tumor development (p<0.001; Table 1) but also advanced stage (p = 0.037, Table 2) of TSCC. In contrast to other two studies on TSCC patients [11,18], the expression of Snail in our study was not correlated with poor cell differentiation [11,18], large tumor size [11], lymph node metastasis [11], or greater invasion depth of the tumor [11,18]. The reason for different clinicopathologic significance may be that the regulation of Snail is complex and modulated by various signals at the transcriptional and post-translational levels in different TSCC populations [19]. One such factor is betel quid chewing, which is an important risk factor for TSCC in Taiwan but not in most Western countries. Moreover, Snail expression was not associated with survival in TSCC (Table 3), thus excluding the possibility of Snail as a potent prognostic factor in TSCC.

Twist is essential for tumor invasion in TSCC cells [13]. However, our data showed that there was no significant difference in Twist expression between TSCC and corresponding TAN tissues (Table 1). Additionally, Twist expression was not correlated with any pathologic outcomes (Table 2) including DSS and DFS in TSCC patients in this study (Table 3). The above results thus suggest that the transcription repression of E-cadherin by Snail but not Twist may promote cell proliferation [20] during tumorigenesis and exclude Twist as a key repressor for tumor metastasis [11,21] in TSCC.

Loss of E-cadherin expression regulated by Snail and Twist promotes the reduction of cell-cell adhesion [22] and results in poor prognosis in OSCC [23]. We found that E-cadherin expression is reduced together with a clearly increase in Snail expression and slight increase in Twist expression in TAN tissues compared to that in tumors (p<0.001; Table 1), indicating the influence of E-cadherin loss in the tumorigenesis of TSCC [24]. In a small cohort of TSCC patients, a previous study indicated that low E-cadherin expression was associated with poor tumor differentiation and higher invasion in TSCC patients [25]. Interestingly, low E-cadherin expression was correlated not only with poor cell differentiation (p<0.001; Table 2) and lymph node metastasis (p = 0.007; Table 2) but also with late pathological stage (p = 0.013; Table 2) and shorter DSS (p = 0.033; Table 3) in our medium-sized cohort of TSCC patients. Additionally, loss of E-cadherin has been shown to be significantly associated with delayed neck metastasis in stage I/II TSCC [21]. Regarding the molecular mechanism of E-cadherin, it has been reported that E-cadherin repression mediated by enhancer of zeste homolog 2 (EZH2) promotes the metastatic phenotype of TSCC cells, and TSCC patients with high expression of EZH2 and low expression of E-cadherin have lower overall survival [26]. Additional analysis is needed to examine the co-expression of EZH2 and E-cadherin in lymph node metastasis in our TSCC patients. In this study, low E-cadherin expression was significantly correlated with shorter DSS in young TSCC patients (≤50 yrs; p = 0.007; S2 Table); this correlation was also observed in gastric cancer patients [27]. The real reason for the impact of E-cadherin on the survival for young patients needs to be further investigated in the near future. Moreover, low E-cadherin expression was significantly correlated with shorter DSS in patients with moderate and poor cell differentiation (p = 0.038; S2 Table), indicating the role of low E-cadherin in poor DSS in TSCC patients with more aggressive phenotypes in terms of cell differentiation. Furthermore, low E-cadherin expression was significantly correlated with shorter DSS in patients without post-operative RT (p = 0.014; S2 Table), which suggests that E-cadherin could be a prognostic marker for certain groups of TSCC patients. However, low E-cadherin expression was not correlated with shorter DFS in patients even after stratification by various clinicopathologic factors (S3 Table).

Usually, high N-cadherin expression is observed in oral tumor tissues [28]. However, our data showed that the level of N-cadherin expression in TAN tissues was higher than that in the tumor (p<0001; Table 1), indicating that N-cadherin may mediate a less stable and more dynamic form of cell-cell adhesion, which may allow both attachment and detachment of individual cells from the primary tumor and selective association with tissues such as the stroma or endothelium in TAN tissues [29]. The other possible reason may be that the higher N-cadherin expression in TAN may influence the gene expression pattern of residual TSCC cells, which may allow them to survive or migrate [30]. In addition, higher N-cadherin expression is characterized by a malignant phenotype associated with worse clinical outcome, increased invasion/metastasis and reduced overall survival in oral cancer patients [12,28]. Co-expression of N-cadherin and β-catenin is essential in TSCC for the down-regulation of the cell cycle inhibitor p21 and up-regulation of MMP-2 and MMP-9, which are considered potential indicators of invasion and metastasis as well as poor survival in TSCC patients [12,31]. Likewise, our results showed that the increased N-cadherin expression was significantly related to poor cell differentiation (p<0.001; Table 2), late pathological stage (p = 0.023; Table 2), and lymph node metastasis (p = 0.007; Table 2), as well as shorter DSS (p = 0.034; Table 3) and DFS (p = 0.022; Table 3) in TSCC.

Vimentin is a predictive biomarker for tumor growth and metastasis, although its significance is limited in TSCC prognosis [32,33]. Our results also showed that Vimentin expression was involved in tumorigenesis (p<0.001; Table 1) and associated with poor cell differentiation (p = 0.006; Table 2) and lymph node metastasis (p = 0.039; Table 2) in TSCC. Surprisingly, Vimentin expression was significantly associated with poor DSS (p<0.001; Table 3) and DFS (p = 0.045; Table 3), and its impact on DSS is appeared to be more important than that of Snail, Twist, E-cadherin and N-cadherin (p = 0.002; Table 4). Three reasons may explain the potential prognostic significance of Vimentin in our study: (1) pathologic outcomes can confound or decrease the impact of Vimentin on DSS; we used the multivariate Cox regression model to adjust the impact of pathologic outcomes on survival; (2) expression levels of E-cadherin, N-cadherin, and Vimentin were quantified simultaneously, and their interaction were also incorporated into the Cox regression model in this study; (3) the sample size of TSCC patients in our study provided adequate statistical power compared to that of studies from other groups, which could have resulted in the identification of prognostic significance of Vimentin in TSCC. Similar to low E-cadherin expression, high Vimentin expression was also significantly correlated with shorter DSS (p = 0.001; S4 Table) in patients with moderate and poor cell differentiation, suggesting that the combination of low E-cadherin expression and high Vimentin expression might be a potential prognostic marker for patients with more differentiated cancer phenotype. However, high Vimentin expression was significantly correlated with shorter DSS in patients with smaller tumors (p = 0.001; S4 Table) and was significantly correlated with shorter DFS in patients with early pathological stage (p = 0.016; S5 Table) and without lymph node metastasis (p = 0.017; S5 Table), implying that Vimentin could be an early prognostic factor in TSCC patients.

Most importantly, Vimentin might be used as a marker to evaluate the prognosis of TSCC patients and help in clinical decision making with respect to the selection of appropriate therapies for individual TSCC patients in the future. More studies are necessary in the future to (a) assess the tumorigenic pathways of Vimentin through its interaction with 14-3-3 and Akt-phosphorylated Beclin 1, which could contribute to control tumorigenesis [34]; (b) understand the regulation of Vimentin through hypoxia-inducible factor-1 and its role in the extracellular space or cell surface during EMT, which could contribute to the invasiveness of cancer cells [35]; (c) understand how Vimentin assembly and expression are regulated, which would help to elucidate the effect of Vimentin expression on metastasis [10]. Furthermore, Vimentin-specific inhibitors as well as antibodies, peptides, aptamers or siRNA directed against Vimentin in combination with other anticancer agents could be used for the treatment of TSCC patients, and more studies are required to further evaluate the therapeutic efficacy of these approached in the clinic.

One possible limitation of our study is that the external validity of our findings is limited to patients in South Asia who are exposed to betel quid chewing, which is not a major risk factor for TSCC among Caucasians. Thus, we further validated these EMT-related proteins in TSCC patients using another independent cohort from The Cancer Genome Atlas (TCGA) database. Consistent with our findings regarding the association of expression levels of the five EMT-related biomarkers with “tumorigenesis”, the expression of Snail (p = 0.009, S6 Table) and Twist (p = 0.001; S6 Table) were significantly higher in tumor tissues than in normal tissues, indicating that the high expression levels of Snail and Twist were reliable biomarkers for tumorigenesis. However, E-cadherin expression (p = 0.621) and Vimentin expression (p = 0.132) were not significantly reduced and elevated, respectively, in tumor tissues compared with the normal tissues. Additionally, N-cadherin expression (p = 0.621) was slightly higher in tumors than in normal tissues; however, no significant difference was found. However, the normal (n = 12) and TSCC tissues (n = 127) were not paired, and few normal tissues are present in the TCGA database; both these factors may help explain the discordant findings between our cohort and the TCGA data. Consistent with our findings regarding the association of expression levels of the five EMT-related biomarkers with “demographic and pathological outcomes”, low E-cadherin expression was only significantly associated with poor cell differentiation (p = 0.001; S7 Table). Moreover, high Vimentin expression was only significantly correlated with lymph node metastasis (p = 0.031; S7 Table). However, the expression levels of Snail and N-cadherin were not associated with any pathological outcomes. For “survival” analysis, the expression levels of all proteins were dichotomized into low and high expression with the same cutoff value as we used in this study. However, the expression levels of these five EMT-related biomarkers were not related to overall survival and disease-free survival (DFS) (S8 Table) since the sample size of patients with DFS was too small (n = 88) compared to the size of our cohort. The possible reasons for the lack of consistency between our study and the results from the TCGA database could be that (a) the five EMT-related proteins in TSCC might be differentially regulated at the mRNA and protein levels[8]; and (b) the regulation of the five EMT-related proteins might be variable due to different environmental exposures and genetic alterations between Asian and Western TSCC patients.

In conclusion, Snail, E-cadherin, N-cadherin, and Vimentin were associated with tumorigenesis and pathologic outcomes in TSCC. Notably, Vimentin may be a potential prognostic factor among the five-EMT related proteins.

Supporting information

S1 Table. The raw clinical data along with expression scores of five EMT-related proteins.

https://doi.org/10.1371/journal.pone.0178581.s001

(XLS)

S2 Table. Impact of E-cadherin expression levels on disease-specific survival by the different clinicopathologic outcomes with TSCC.

https://doi.org/10.1371/journal.pone.0178581.s002

(DOC)

S3 Table. Impact of E-cadherin expression levels on disease-free survival by the different demographic and clinicopathologic factors with TSCC.

https://doi.org/10.1371/journal.pone.0178581.s003

(DOC)

S4 Table. Impact of Vimentin expression levels on disease-specific survival by the different clinicopathologic outcomes with TSCC.

https://doi.org/10.1371/journal.pone.0178581.s004

(DOC)

S5 Table. Impact of Vimentin expression levels on disease-free survival by the different clinicopathologic outcomes with TSCC.

https://doi.org/10.1371/journal.pone.0178581.s005

(DOC)

S6 Table. The comparison of Snail, Twist, E-cadherin, N-cadherin, and Vimentin expression between TSCC and corresponding tumor adjacent normal tissues from TCGA database.

https://doi.org/10.1371/journal.pone.0178581.s006

(DOC)

S7 Table. The correlation of the expression of Snail, Twist, E-cadherin, N-cadherin, and Vimentin to pathological outcomes in TSCC patients from TCGA database.

https://doi.org/10.1371/journal.pone.0178581.s007

(DOC)

S8 Table. The correlation of the expression of Snail, Twist, E-cadherin, N-cadherin, and Vimentin to disease-specific survival and disease-free survival in TSCC patients from TCGA database.

https://doi.org/10.1371/journal.pone.0178581.s008

(DOC)

Acknowledgments

This study was supported by a research grant from Kaohsiung Veterans General Hospital (VGHKS-101-033 and VGHKS-103-011) in Taiwan, ROC.

Author Contributions

  1. Conceptualization: LPG YDH.
  2. Formal analysis: HHL HCS ICH YCL.
  3. Funding acquisition: LPG.
  4. Investigation: LPG YDH.
  5. Project administration: LPG YDH.
  6. Resources: TYF YKT YMW JHS LMY YSL CWS.
  7. Supervision: LPG YDH.
  8. Validation: LPG YDH.
  9. Writing – original draft: PFL BHK.
  10. Writing – review & editing: PFL BHK.

References

  1. 1. Franceschi D, Gupta R, Spiro RH, Shah JP Improved survival in the treatment of squamous carcinoma of the oral tongue. Am J Surg 1993; 4 (166): 360–5.
  2. 2. Fakih AR, Rao RS, Borges AM, Patel AR Elective versus therapeutic neck dissection in early carcinoma of the oral tongue. Am J Surg 1989; 4 (158): 309–13.
  3. 3. Byers RM, Weber RS, Andrews T, McGill D, Kare R, Wolf P Frequency and therapeutic implications of "skip metastases" in the neck from squamous carcinoma of the oral tongue. Head Neck 1997; 1 (19): 14–9.
  4. 4. Liang X, Zheng M, Jiang J, Zhu G, Yang J, Tang Y Hypoxia-inducible factor-1 alpha, in association with TWIST2 and SNIP1, is a critical prognostic factor in patients with tongue squamous cell carcinoma. Oral Oncol 2011; 2 (47): 92–7.
  5. 5. Acloque H, Adams MS, Fishwick K, Bronner-Fraser M, Nieto MA Epithelial-mesenchymal transitions: the importance of changing cell state in development and disease. J Clin Invest 2009; 6 (119): 1438–49.
  6. 6. Scanlon CS, Van Tubergen EA, Inglehart RC, D'Silva NJ Biomarkers of epithelial-mesenchymal transition in squamous cell carcinoma. J Dent Res 2013; 2 (92): 114–21.
  7. 7. Chaw SY, Majeed AA, Dalley AJ, Chan A, Stein S, Farah CS Epithelial to mesenchymal transition (EMT) biomarkers—E-cadherin, beta-catenin, APC and Vimentin—in oral squamous cell carcinogenesis and transformation. Oral Oncol 2012; 10 (48): 997–1006.
  8. 8. Lamouille S, Xu J, Derynck R Molecular mechanisms of epithelial-mesenchymal transition. Nat Rev Mol Cell Biol 2014; 3 (15): 178–96.
  9. 9. Islam S, Carey TE, Wolf GT, Wheelock MJ, Johnson KR Expression of N-cadherin by human squamous carcinoma cells induces a scattered fibroblastic phenotype with disrupted cell-cell adhesion. J Cell Biol 1996; 6 Pt 1 (135): 1643–54.
  10. 10. Kidd ME, Shumaker DK, Ridge KM The role of vimentin intermediate filaments in the progression of lung cancer. Am J Respir Cell Mol Biol 2014; 1 (50): 1–6.
  11. 11. Zheng M, Jiang YP, Chen W, Li KD, Liu X, Gao SY, et al. Snail and Slug collaborate on EMT and tumor metastasis through miR-101-mediated EZH2 axis in oral tongue squamous cell carcinoma. Oncotarget 2015; 9 (6): 6797–810.
  12. 12. Li S, Jiao J, Lu Z, Zhang M An essential role for N-cadherin and beta-catenin for progression in tongue squamous cell carcinoma and their effect on invasion and metastasis of Tca8113 tongue cancer cells. Oncol Rep 2009; 5 (21): 1223–33.
  13. 13. Tang YL, Jiang J, Liu J, Zheng M, He YW, Chen W, et al. Hyperthermia inhibited the migration of tongue squamous cell carcinoma through TWIST2. J Oral Pathol Med 2015; 5 (44): 337–44.
  14. 14. Afrem MC, Margaritescu C, Craitoiu MM, Ciuca M, Sarla CG, Cotoi OS The immunohistochemical investigations of cadherin "switch" during epithelial-mesenchymal transition of tongue squamous cell carcinoma. Rom J Morphol Embryol 2014; 3 Suppl (55): 1049–56.
  15. 15. Chen HC, Yang CM, Cheng JT, Tsai KW, Fu TY, Liou HH, et al. Global DNA hypomethylation is associated with the development and poor prognosis of tongue squamous cell carcinoma. J Oral Pathol Med 2015;
  16. 16. Fu TY, Wu CN, Sie HC, Cheng JT, Lin YS, Liou HH, et al. Subsite-specific association of DEAD box RNA helicase DDX60 with the development and prognosis of oral squamous cell carcinoma. Oncotarget 2016;
  17. 17. Usami Y, Satake S, Nakayama F, Matsumoto M, Ohnuma K, Komori T, et al. Snail-associated epithelial-mesenchymal transition promotes oesophageal squamous cell carcinoma motility and progression. J Pathol 2008; 3 (215): 330–9.
  18. 18. Hayry V, Makinen LK, Atula T, Sariola H, Makitie A, Leivo I, et al. Bmi-1 expression predicts prognosis in squamous cell carcinoma of the tongue. Br J Cancer 2010; 5 (102): 892–7.
  19. 19. Wang Y, Shi J, Chai K, Ying X, Zhou BP The Role of Snail in EMT and Tumorigenesis. Curr Cancer Drug Targets 2013; 9 (13): 963–72.
  20. 20. Barrallo-Gimeno A, Nieto MA The Snail genes as inducers of cell movement and survival: implications in development and cancer. Development 2005; 14 (132): 3151–61.
  21. 21. Sakamoto K, Imanishi Y, Tomita T, Shimoda M, Kameyama K, Shibata K, et al. Overexpression of SIP1 and downregulation of E-cadherin predict delayed neck metastasis in stage I/II oral tongue squamous cell carcinoma after partial glossectomy. Ann Surg Oncol 2012; 2 (19): 612–9.
  22. 22. Tian X, Liu Z, Niu B, Zhang J, Tan TK, Lee SR, et al. E-cadherin/beta-catenin complex and the epithelial barrier. J Biomed Biotechnol 2011; (2011): 567305.
  23. 23. Liu LK, Jiang XY, Zhou XX, Wang DM, Song XL, Jiang HB Upregulation of vimentin and aberrant expression of E-cadherin/beta-catenin complex in oral squamous cell carcinomas: correlation with the clinicopathological features and patient outcome. Mod Pathol 2010; 2 (23): 213–24.
  24. 24. Mohamet L, Hawkins K, Ward CM Loss of function of e-cadherin in embryonic stem cells and the relevance to models of tumorigenesis. J Oncol 2011; (2011): 352616.
  25. 25. Luo SL, Xie YG, Li Z, Ma JH, Xu X E-cadherin expression and prognosis of oral cancer: a meta-analysis. Tumour Biol 2014; 6 (35): 5533–7.
  26. 26. Wang C, Liu X, Chen Z, Huang H, Jin Y, Kolokythas A, et al. Polycomb group protein EZH2-mediated E-cadherin repression promotes metastasis of oral tongue squamous cell carcinoma. Mol Carcinog 2013; 3 (52): 229–36.
  27. 27. Schildberg CW, Abba M, Merkel S, Agaimy A, Dimmler A, Schlabrakowski A, et al. Gastric cancer patients less than 50 years of age exhibit significant downregulation of E-cadherin and CDX2 compared to older reference populations. Adv Med Sci 2014; 1 (59): 142–6.
  28. 28. M DID, Pierantoni GM, Feola A, Esposito F, Laino L, A DER, et al. Prognostic significance of N-Cadherin expression in oral squamous cell carcinoma. Anticancer Res 2011; 12 (31): 4211–8.
  29. 29. Hazan RB, Phillips GR, Qiao RF, Norton L, Aaronson SA Exogenous expression of N-cadherin in breast cancer cells induces cell migration, invasion, and metastasis. J Cell Biol 2000; 4 (148): 779–90.
  30. 30. Shimizu D, Vallbohmer D, Kuramochi H, Uchida K, Schneider S, Chandrasoma PT, et al. Increasing cyclooxygenase-2 (cox-2) gene expression in the progression of Barrett's esophagus to adenocarcinoma correlates with that of Bcl-2. Int J Cancer 2006; 4 (119): 765–70.
  31. 31. Aparna M, Rao L, Kunhikatta V, Radhakrishnan R The role of MMP-2 and MMP-9 as prognostic markers in the early stages of tongue squamous cell carcinoma. J Oral Pathol Med 2015; 5 (44): 345–52.
  32. 32. Wang C, Liu X, Huang H, Ma H, Cai W, Hou J, et al. Deregulation of Snai2 is associated with metastasis and poor prognosis in tongue squamous cell carcinoma. Int J Cancer 2012; 10 (130): 2249–58.
  33. 33. Ding L, Zhang Z, Shang D, Cheng J, Yuan H, Wu Y, et al. alpha-Smooth muscle actin-positive myofibroblasts, in association with epithelial-mesenchymal transition and lymphogenesis, is a critical prognostic parameter in patients with oral tongue squamous cell carcinoma. J Oral Pathol Med 2014; 5 (43): 335–43.
  34. 34. Wang RC, Wei Y, An Z, Zou Z, Xiao G, Bhagat G, et al. Akt-mediated regulation of autophagy and tumorigenesis through Beclin 1 phosphorylation. Science 2012; 6109 (338): 956–9.
  35. 35. Satelli A, Li S Vimentin in cancer and its potential as a molecular target for cancer therapy. Cell Mol Life Sci 2011; 18 (68): 3033–46.