Tumor Biol. DOI 10.1007/s13277-015-3350-1

RESEARCH ARTICLE

Decreased expression of claudin-3 is associated with a poor prognosis and EMT in completely resected squamous cell lung carcinoma Juanjuan Che 1 & Yifan Yang 1 & Jing Xiao 1 & Pengfei Zhao 1 & Bo Yan 2 & Shuo Dong 3 & Bangwei Cao 1

Received: 27 November 2014 / Accepted: 16 March 2015 # International Society of Oncology and BioMarkers (ISOBM) 2015

Abstract The deregulation of claudin-3 has been reported to correlate with the invasion and metastasis of various cancers, but little is known about its expression level and the prognostic value in squamous cell lung carcinoma (SqCC). The purpose of this study is to determine the expression levels and the prognostic value of claudin-3 in completely resected SqCC tissues, and the potential underlying mechanism. The protein expression of claudin-3, E-cadherin, β-catenin, and vimentin in the tumor tissues from 103 patients with surgically resected SqCC was examined using immunohistochemistry, western blots, as well as semi-quantitative estimation. The claudin-3 protein level was significantly associated with E-cadherin, βcatenin, and vimentin protein expression. A decreased claudin-3 protein level was significantly correlated with TNM stage, lymph node metastasis, and disease recurrence. Similarly, downregulation of E-cadherin was significantly correlated with lymph node metastasis and disease recurrence. Decreased β-catenin expression also had a significant correlation with disease recurrence. Univariate analyses indicated that the T stage, lymph node metastasis, the TNM stage, and the expression of claudin-3, β-catenin, and vimentin were significant predictors for overall survival (OS). Moreover, multivariate analyses demonstrated that the TNM stage and

* Bangwei Cao [email protected] 1

Department of Oncology, Beijing Friendship Hospital, Capital Medical University, #95 Yong An Road, Xicheng District, Beijing 100050, People’s Republic of China

2

Department of Microbiology and Immunology, University of Texas Health Science Center at San Antonio, San Antonio, TX 75390, USA

3

Department of Medicine, Baylor College of Medicine, Houston, TX, USA

protein levels of claudin-3, β-catenin, and vimentin were independent predictors for OS of SqCC patients. Claudin-3 plays an important role in the epithelial–mesenchymal transition of SqCC and might be used as a potential prognostic factor for SqCC.

Keywords Squamous cell lung carcinoma . Claudin-3 . Epithelial–mesenchymal transition . Prognosis

Introduction Lung cancer is the main cause of cancer-related mortality worldwide [1]. Despite great advances in the management of lung cancer patients during the past decades, the 5-year survival rate remains less than 15 %, with most patients dying from recurrence after resection [2]. Histologically, non-small cell lung carcinoma (NSCLC) accounts for approximately 85 % of lung cancer cases, of which 70 % of cases are adenocarcinoma (AC) and squamous cell lung carcinoma (SqCC) [3]. Recent molecularly targeted therapies, particularly epidermal growth factor receptor-tyrosine kinase inhibitors, have dramatically improved the prognosis of AC patients, especially those with advanced AC [4, 5]. In contrast, very few molecularly targeted therapeutic agents are available for the treatment of recurrent SqCC. Thus, it is urgent to elucidate the molecular biology of SqCC recurrence following surgery and identify potential targets for the development of molecularly targeted therapies for SqCC. Tight junctions (TJ) are important structural components at the apicobasal region of the epithelium and serve as a barrier to maintain homeostasis by controlling the passage of ions, water, and different macromolecules. Tight junctions play critical roles in lung defense by preventing pathogen penetration and other exogenous molecules across the interstitial tissues [1]. Recently, deregulation

Tumor Biol.

of claudins, essential constituents of tight junctions, has been found in various human cancers, such as colon, pancreatic, ovarian, and breast cancers [6–9]. In most patients, altered expression of claudins correlates to metastasis or a poor prognosis. For example, a reduction in claudin-1 expression was found to be correlated to a higher tumor grade and a poor survival in patients with colon cancer [7], and downregulation of claudin-1 in AC patients was linked to a shorter survival [10]. In addition, ectopic overexpression of claudin-1 attenuated the invasion and aggressive growth of cultured AC cells [10]. Moreover, low claudin-3 expression was associated with a poor prognosis in ovarian cancers [11]. Claudin-3 expression levels varied more in different histological types of lung cancers than in normal bronchial cells or lung tissues [12–14]. Claudin-3 mRNA was found to be 16-fold higher in small cell lung carcinoma than that in normal lung tissue [13]. However, compared to normal bronchial cells, AC, small cell carcinoma, bronchioloalveolar carcinomas, and carcinoid tumors, a decreased expression of claudin-3 was found in SqCC [12, 14–16]. These findings suggest that there might be an association between the deregulation of claudin-3 and the invasion and metastasis of SqCC, and that claudin-3 could possibly be a prognostic factor for SqCC patients. Epithelial–mesenchymal transition (EMT) involves a dynamic change of epithelial cells, including loss of epithelial cell–cell contacts, increase of mobility, and acquisition of mesenchymal phenotype alterations [17]. EMT plays a crucial role in the invasion, migration, and metastasis of epithelial cancer cells [18]. Accumulating evidence has demonstrated that EMT-related markers, such as E-cadherin, vimentin, and β-catenin, are associated with the overall survival (OS) of NSCLC patients [19–22]. Multiple transcription factors including Slug, Snail, Twist, and ZEB1 regulate EMT [23, 24]. Interestingly, it has been recently found that Twist and ZEB1 regulate claudin expression in various cancers [25, 26]. These findings suggest that claudins might be involved in EMT. However, the relationship between claudin-3 expression and EMT-related markers as well as the clinicopathological characteristics of SqCC patients is unknown. Claudin proteins are the main components of tight junctions, which functions as a Bfence^ continuously surrounding the apical part of the cells. Tight junctions play an important role in the infiltration, metastasis, transformation, and transport of tumor cells. The process of EMT is initiated by the disappearance of the basic structure of epithelial tissue, followed by epithelial cell degeneration, migration capacity enhancement, loss of cell polarity, and tight connection. Therefore, we speculate that claudin proteins play an important role in the EMT process [18]. Immunohistochemical studies have found the expression rate of claudin-3 is distinct in different pathological types of NSCLC. For example, the expression of claudin-3 in SqCC is higher than that in adenocarcinoma [27]. However, it is not clear whether claudin-3 has prognosis significance and if there

is a direct functional interaction between claudin-3 and EMT in SqCC. In addition, claudin-3 is a receptor of Clostridium perfringens toxin (CPE). Binding of claudin-3 with CPE leads to rapid cell lysis. Furthermore, previous studies have shown that CPE could inhibit the malignancy that expresses claudin3 [28, 29]. These facts suggest that claudin-3 is a potential therapeutic target. To evaluate the prognostic value of claudin-3 in SqCC and the relationship between claudin-3 and EMT-related markers, we analyzed the correlation between claudin-3, E-cadherin, β-catenin, and vimentin expression, and investigated the association of these markers with clinicopathological parameters of SqCC patients.

Materials and methods The inclusion criteria of patients and data extraction The medical records of SqCC patients who received surgical resection at Beijing Friendship Hospital from 2010 to 2014 were reviewed. One hundred and three cases were included in this study based on the following inclusion criteria: (1) SqCC diagnosis was confirmed by pathological tissue examination, and patients received complete pulmonary resection and systematic node dissection of the hilar and mediastinal lymph nodes; (2) tumor tissues were available. The exclusion criteria were as follows: (1) patients underwent preoperative chemotherapy or radiotherapy; (2) data regarding the tumor stage was not available; (3) surgical resection guidelines were unavailable; (4) follow-up data were incomplete or patients died within 1 month after surgery; (5) a secondary malignant neoplasm was found within 5 years after surgery. All the patients included in this study were Chinese. Our study was approved by the Research Ethics Committee of Beijing Friendship Hospital. Written consent for tissue specimen collection was obtained from each patient. The following data were extracted from the medical records: age, gender, history of smoking, surgical procedure, T or N stage, TNM stage, and other treatments (surgical resection, chemotherapy, or radiotherapy). Patients were restaged based on the 7th edition of the American Joint Committee on Cancer TNM classification of lung cancer (Table 1). All patients were followed up until December 30, 2014. Immunohistochemistry The procedures were performed according to standard protocols. Lung tissues were first deparaffinized with xylene, followed by a descending series of ethanol concentrations, and immersed in 3 % hydrogen peroxide for 10 min to block endogenous peroxidase activity. Following antigen retrieval in 0.01 M sodium citrate buffer (pH=6.0), the samples were incubated with the following primary antibodies: rabbit anti-

Tumor Biol. Table 1 Demographic and clinical characteristics of 103 squamous cell lung carcinoma patients

Characteristic Gender Male Female Age (years) <65 ≥65 Smoking history Smoker Never T stage T1 T2 T3 N stage N0 N1 N2 TNM Stage I Stage IIIA

No.

Percent (%)

85 18

82.5 17.5

59

57.3

44

42.7

86 17

83.5 16.5

30 62 11

29.1 60.2 10.7

68 3 32

66.0 2.9 31.1

68 35

66.0 34.0

Western blot We selected 16 tissue samples of SqCC patients to detect the expression of claudin-3, E-cadherin, β-catenin, and vimentin. Total proteins were isolated from tissues using T-PER Mammalian Protein Extraction Reagent (Thermo) supplemented with protease and phosphatase inhibitors and pre-cleared using centrifugation, followed by measuring protein concentrations using the bicinchoninic acid (BCA) Protein Assay kit (Pierce). Equal amounts of protein extracts (40 μg) were separated by 10 % sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto a PVDF membrane. Primary antibodies used were rabbit anti-human claudin-3 (1:500 dilution; Assay Biotech, Sunnyvale, CA, USA), mouse anti-E-cadherin (1:200; Invitrogen, Carlsbad, CA, USA), mouse anti-β-catenin (1:200 dilution; Millipore, Billerica, MA, USA), and mouse anti-vimentin (1:500 dilution; Zymed Laboratories, San Francisco, CA, USA). Statistical analysis

claudin-3 (1:100 dilution; Assay Biotech, Sunnyvale, CA, USA), mouse anti-E-cadherin (1:200; Invitrogen, Carlsbad, CA, USA), mouse anti-β-catenin (1:200 dilution; Millipore, Billerica, MA, USA), and mouse anti-vimentin (1:400 dilution; Zymed Laboratories, San Francisco, CA, USA). After washing three times in phosphate-buffered saline, the sections were incubated with the corresponding secondary antibody for 30 min at 37 °C, followed by incubation for 1 min with 3,3′-diaminobenzidine. The cover slips were counterstained with hematoxylin and dehydrated in 85–100 % ethanol. Following treatment with xylol, the cover slips were mounted permanently with resinous mounting medium. In addition, positive specimens of claudin-3, E-cadherin, vimentin, and β-catenin expression were used as positive controls, respectively.

Continuous variables were presented as mean±standard deviation, and categorical variables were expressed as numbers (%). The χ2 test was used to analyze the correlation between the expression of these markers and clinic-pathologic parameters. The χ2 test for linear trends was applied to evaluate the correlations between the expression of claudin-3, E-cadherin, β-catenin, and vimentin. Cox proportional hazards regression for univariate and multivariate analyses were used to identify potential predictors of OS. Only variables with statistical significance according to univariate Cox regression analysis were analyzed by multivariate analysis, and the results were expressed as hazard ratios (HRs) with 95 % confidence intervals (CIs). Kaplan–Meier analysis was used to analyze the statistical difference in OS rates. A p value <0.05 was considered significant, using the two-tailed test. All statistical analyses were performed by using SPSS 18.0 software (SPSS Inc., Chicago, IL, USA).

Results Assessment of immunohistochemical staining Patient characteristics The immunohistochemical staining results were examined independently by two pathologists. Membranous staining of claudin-3, E-cadherin, vimentin, and β-catenin was evaluated by a semi-quantitative scale from 0 to 3+ as follows: 0, no color; 1+, ≤25 % of cells were positively or partially stained; 2+, 25–75 % of cells were positively or partially stained; 3+, >75 % of cells were positively or partially stained. The intensity of staining of claudin-3 and β-catenin was classified as negative (0, 1+) and positive (2+, 3+) expression.

The baseline characteristics of the 103 SqCC patients included in this study are shown in Table 1. The study cohort included 85 men and 18 women. Eighty-six patients had a history of smoking (83.5 %), and 42.7 % of the patients were older than 65 years old. The numbers of patients at pathological stages T1, T2, and T3 were 30 (29.1 %), 62 (60.2 %), and 11 (10.7 %), respectively, while those at pathological stages N0, N1, and N2 were 68 (66.0 %), 3 (2.9 %), and 32 (31.1 %),

Tumor Biol.

respectively. There were 68 (66.0 %) patients at pathological stage I and 35 (34.0 %) at IIIA. Claudin-3, E-cadherin, β-catenin, and vimentin expression Sixty-seven (65.0 %) patient samples were positively stained for claudin-3 (Fig. 1a). E-cadherin and β-catenin protein expression was observed in the intercellular junctions (Fig. 1c, e). Downregulation of E-cadherin and β-catenin protein expression was observed at the membranes of some of the tumor cells (Fig. 1d, f). Negative or low E-cadherin and β-catenin protein levels were found in 53.3 and 60.1 % of all cases, respectively. Vimentin was localized to the stromal structures of normal or neoplastic tissues as well as the cytoplasm of the neoplastic tissue (Fig. 1g, h). Negative or low vimentin protein expression was found in 78.6 % of patients. To analyze the correlation between claudin-3 and EMT markers, we selected 16 samples of SqCC patients to detect the expression of claudin-3, E-cadherin, β-catenin, and vimentin by western blots. The western results were consistent with the immunohistochemical results (Fig. 3a, b). Correlation between claudin-3, E-cadherin, β-catenin, and vimentin expression and clinicopathological parameters As shown in Table 2, there were significant correlations between claudin-3 protein expression and lymph node metastasis (p=0.001), histological TNM stage (p=0.001), and recurrence (p<0.001). Negative or low claudin-3 expression was found in 69.8 % (30/43) of the tumors from patients with an eventual recurrent tumor, whereas 10 % (6/60) of the tumors from patients without a history of tumor recurrence displayed negative or low claudin-3 protein expression (p<0.001). Moreover, negative or low claudin-3 expression was observed in 57.1 % (20/ 35) of the tumors from patients with eventual lymph node Fig. 1 The expression of claudin3, E-cadherin, β-catenin, and vimentin of SqCC patients by immunohistochemistry. Claudin3 (a, b), E-cadherin (c, d), βcatenin (e, f), and vimentin (g, h) expression were examined by immunohistochemistry. In squamous cell lung carcinoma patients, the negative or low expression of claudin-3 (b), Ecadherin (d), or β-catenin (f) and increased expression of vimentin (g) were observed in tumor tissues (×400)

metastasis, whereas 23.5 % (16/68) of tumors from patients who did not have a history of lymph node metastasis displayed negative or low claudin-3 protein expression (p=0.001). Decreased E-cadherin protein expression was significantly correlated with lymph node metastasis (p=0.006) and recurrence (p<0.001). Negative or low E-cadherin expression was found in 81.3 % (35/43) of the patients with an eventual recurrent tumor, whereas 36.7 % (22/60) of the patients who did not have a history of recurrence showed weak E-cadherin protein expression (p<0.001). Weak E-cadherin protein expression was found in 74.2 % (26/35) of the patients with eventual lymph node metastasis, whereas 45.5 % (31/68) of patients without a history of lymph node metastasis showed weak E-cadherin protein expression. Moreover, decreased βcatenin protein expression was correlated with age (p=0.025) and disease recurrence (p=0.037). However, no significant association between vimentin expression and clinicopathological parameters was observed (p>0.05). Correlation between claudin-3, E-cadherin, β-catenin, and vimentin expression As shown in Table 3, the claudin-3 protein level was significantly correlated with E-cadherin, β-catenin, and vimentin expression (p=0.003, p=0.024, and p=0.007, respectively), while E-cadherin protein expression was closely correlated with βcatenin expression (p<0.001). However, no significant correlation was observed between vimentin and E-cadherin expression (p=0.377). Similarly, no significant correlation was found between vimentin and β-catenin expression (p=0.710). Survival analysis Forty-one patients were alive and 62 patients had died by the end of the follow-up period. According to univariate analysis (Table 4), the OS was significantly correlated with the T stage (p=0.001), lymph node metastasis (p<0.001), the TNM stage

Tumor Biol. Table 2 Correlation between claudin-3, E-cadherin, β-catenin, and vimentin expression and clinicopathological parameters

Characteristic

Claudin-3

β-Catenin

E-Cadherin

Vimentin

−

+

p

−

+

p

−

+

p

−

+

p

27 9

58 9

0.140

47 10

38 8

0.984

50 12

35 6

0.537

66 15

19 3

0.593

18 18

41 26

0.273

30 27

29 17

0.288

30 32

29 12

0.025

47 34

12 10

0.770

30 6

56 11

0.974

49 8

37 9

0.452

53 9

33 8

0.504

69 12

17 5

0.375

4 29 3

2 56 9

0.202

4 46 7

2 39 5

0.816

4 51 7

2 34 5

0.940

4 67 10

2 18 2

0.715

11 21 4

19 41 7

0.961

19 30 8

11 32 3

0.189

22 32 8

8 30 3

0.091

24 49 8

6 13 3

0.875

16 20

52 15

0.001

31 26

37 9

0.006

40 22

28 13

0.692

56 25

12 10

0.217

16 20

52 15

0.001

37 20

31 15

0.792

40 22

28 13

0.692

56 25

12 10

0.286

6 30

54 13

<0.001

22 35

38 8

<0.001

31 31

29 12

0.037

49 32

11 11

0.376

Gender Male Female Age (years) <65 ≥65 Smoking history Smoker Never Grade of SqCC Well Moderate Poor T T1 T2 T3 LN metastasis N0 N1–2 TNM Stage I Stage IIIA Recurrence No Yes

p values <0.05 were considered to be significant and are indicated by bold font N0 no lymph node metastasis, N1–2 lymph node metastasis

(p<0.001, Fig. 2a), and claudin-3 (p=0.008, Fig. 2b), βcatenin (p=0.003, Fig. 2c), and vimentin (p=0.001, Fig. 2d) expression, respectively. Patients with a high expression of claudin-3 as well as E-cadherin and β-catenin had a favorable prognosis, compared with those with negative or low expression of claudin-3, E-cadherin, or β-catenin. On the contrary, patients with a negative or low vimentin protein level had a favorable prognosis, compared with those with a high vimentin protein level. Multivariate analyses demonstrated that the TNM stage (HR=2.430; 95 % CI, 1.439–4.104; p= 0.001) as well as the claudin-3 (HR=0.593; 95 % CI, 0.356– 0.987; p=0.044), β-catenin (HR=0.506; 95 % CI, 0.292– 0.876; p = 0.015), and vimentin (HR = 1.918; 95 % CI, 1.107–3.324; p=0.020) expression levels were independent predictors for OS (Table 5).

Discussion In the current study, we semi-quantitatively examined the expression of claudin-3 as well as EMT-related markers in

postoperative patients with SqCC and further analyzed the relationships between the expressions of these proteins with Table 3 Correlation between claudin-3, E-cadherin, β-catenin, and vimentin expression Variable

β-Catenin

Vimentin −

Claudin-3 − 23 + 58 E-cadherin − 43 + 38 β-Catenin − 48 + 33

E-cadherin

+

p

−

+

p

−

+

p

13 9

0.007

27 35

9 32

0.024

27 30

9 37

0.003

14

0.377

50

7

<0.001

12

34

8 14 8

0.710

p values <0.05 were considered to be significant and are indicated by bold font

Tumor Biol. Table 4

Univariate analysis of OS for 103 SqCC patients

Characteristic Gender Age (years) Smoking history Tumor

Lymph node TMN stage Chemotherapy (IIIA) Radiotherapy (IIIA) Claudin-3 E-cadherin β-Catenin Vimentin

No.

5-year OS (%)

p value

Male Female <65 ≥65 Smoker Never T1 T2 T3 N0 N1–2

85 18 59 44 86 17 30 62 11 68 35

45.8 48.9 50.0 41.1 45.5 48.6 55.3 46.4 0.00 60.8 18.6

0.509

Stage I Stage IIIA Yes No Yes No + − + − + − + −

68 35 23 12 11 24 67 36 46 57 41 62 22 81

60.8 18.6 19.3 16.7 22.7 16.7 55.4 29.6 54.9 39.5 60.2 37.0 27.3 51.8

<0.001

0.358 0.493 0.001

<0.001

0.776 0.821 0.008 0.107 0.003 0.001

p values <0.05 were considered to be significant and are indicated by bold font N0 no lymph node metastasis, N1–2 lymph node metastasis, OS overall survival, SqCC squamous cell lung carcinoma

the clinicopathological variables. Furthermore, we investigated the relationship between claudin-3 and EMT markers (Fig. 3). We found that decreased claudin-3 protein expression was significantly correlated with lymph node metastasis, the TNM stage, recurrence, and a shorter OS. The claudin-3 protein level was significantly associated with the protein levels of E-cadherin, β-catenin, and vimentin. Additionally, the expression of claudin-3 as well as β-catenin and vimentin were independent predictors for the OS. The diagnostic gold standards of SqCC in the current clinical application, such as CK5, CK6, and P63, had no prognostic value. They were just used to indentify SqCC, AC, and other pathological types of lung cancer. Our study suggested that the different expression levels of claudin-3 were associated with the prognosis of SqCC. Furthermore, this study suggested that claudin-3 played a critical role in the EMT of SqCC and might be used as a prognostic factor for SqCC. Tight junctions play a key role in maintaining the normal structure and function of epithelial cells. The claudin protein family is one of the main components of tight junctions. The

abnormal expression of claudins on the cell membrane correlates with the progression and prognosis in various tumors. For example, a low expression of claudin-3 predicts a poor prognosis of breast cancer, esophageal cancer, colon cancer, and ovarian cancer [8, 9, 30]. Moreover, claudin-3 expression was observed to be markedly different in normal lung tissues compared to lung tissues with different types of histopathological lung cancers, including SqCC, pulmonary AC, small cell lung cancer, and carcinoid lung tumors [13, 14, 31]. However, little is known regarding the correlation of claudin-3 expression with the metastasis and prognosis of NSCLC. A previous study has indicated that claudin-3 expression is irrelevant to the differentiation of SqCC, the T stage, lymph node metastasis, and the TNM stage [13]. In sharp contrast, our findings showed that the positive rate of claudin-3 protein expression in SqCC was 65 % and was significantly correlated with lymph node metastasis, the TNM stage, and recurrence of SqCC. Our results were different from the results obtained by Judit et al. [13]. It could be partially explained by the following reasons: different antibodies, different experimental methods, and the patients from different ethnic have exhibited a tumor heterogeneous cell population that displayed variable behavior. Furthermore, compared with patients with high expression levels of claudin-3, patients with negative or low expression of claudin-3 showed a poor survival and a high relapse rate. Additionally, the main point of our research was the OS, not the relapse-free survival (RIS). Therefore, we compared the OS using the Kaplan–Meier curve. Univariate and multivariate survival analyses indicated that claudin-3 was an independent predictor of SqCC. Our results suggest that claudin-3 plays a key role in the progression of SqCC and that claudin-3 is a potential prognostic factor for SqCC. EMT is a key regulator in the invasion and metastasis of cancer cells, including NSCLC. During EMT, E-cadherin and mucin 1 are downregulated, and vimentin, fibronectin, and S100 calcium-binding protein A4 are upregulated. Ecadherin binds to β-catenin to improve cell–cell adhesion, while changes of E-cadherin and β-catenin levels are correlated with tumor invasion and metastasis. Among different types of tumors, the negative or low E-cadherin and βcatenin expression rates have been determined to be 38.0– 85.1 % [19, 32, 33] and 31.1–82.6 % [19, 21, 34, 35], respectively. Consistent with these studies, the current study demonstrated that the negative E-cadherin and β-catenin expression rates in SqCC were 55.3 and 60.2 %, respectively. The findings from available studies on changes of E-cadherin and βcatenin expression levels in NSCLC are not consistent. Some studies have suggested that negative or low expression of Ecadherin or β-catenin correlates with the T stage, lymph node metastasis, the histopathological type, and the degree of differentiation of lung cancer [19, 20, 32, 35]. For example, Kase et al. have reported that the low expression rate of E-cadherin correlates with lymph node metastasis of SqCC; however, the

Tumor Biol. Fig. 2 Survival curves of 103 patients with squamous cell lung carcinoma. a Cum survival curves of patients with I or IIIA (log-rank test, p=0.001). b Cum survival curves of patients with tumors expression positive or negative levels of claudin-3 (log-rank test, p=0.044). c Cum survival curves of patients with tumors expression positive or negative levels of β-catenin (log-rank test, p=0.044). d Cum survival curves of patients with tumors expression positive or negative levels of vimentin (log-rank test, p=0.020)

5-year survival rate was not statistically significant and Ecadherin expression was not an independent predictor in NSCLC [19]. Meanwhile, Zhang et al. have found that negative or low E-cadherin expression correlates with the postoperative recurrence of SqCC and a decrease in the OS or disease-free survival (DFS) and that E-cadherin is an independent predictor in SqCC [37]. Our study indicated that negative E-cadherin expression was significantly correlated with lymph node metastasis and postoperative recurrence of SqCC, but it was not significantly correlated with the OS. Thus, Ecadherin expression did not appear to be an independent prognostic factor in SqCC. We found that in 57 SqCC tissues with negative expression of E-cadherin, 50 cases (87.7 %) showed negative expression of β-catenin, demonstrating a positive correlation between Ecadherin and β-catenin expression. Our observation is Table 5

consistent with a study by Kase et al. on NSCLC [19], further supporting the conclusion that the EMT pathway involving Ecadherin is related to β-catenin expression. High β-catenin expression has been suggested to be a favorable predictor in NSCLC [21, 34]. However, the correlation between β-catenin expression and the prognosis of SqCC remains elusive. Choi et al. found that low β-catenin expression was irrelevant to DFS of stage I SqCC [20]. Similarly, Zhang et al. determined

Multivariate analysis of OS for 103 SqCC patients

Variables

HR

95 % CI

p value

TNM (I/IIIA) Claudin-3 β-Catenin Vimentin

2.430 0.593 0.506 1.918

1.439–4.104 0.356–0.987 0.292–0.876 1.107–3.324

0.001 0.044 0.015 0.020

P values<0.05 were considered to be significant and are indicated by bold font OS overall survival, SqCC squamous cell lung carcinoma, HR hazard ratio, CI confidence interval

Fig. 3 Claudin-3, E-cadherin, β-catenin, and vimentin expression in the tissue samples of SqCC patients by western blotting. a The positive expression of claudin-3 was correlated with a positive expression of Ecadherin and β-catenin, while a negative expression of vimentin. b The negative expression of claudin-3 was correlated with a negative expression of E-cadherin and β-catenin, while a positive expression of vimentin

Tumor Biol.

that β-catenin expression was unrelated to postoperative recurrence, OS, or DFS of SqCC [37]. In contrast, Pirinen et al. showed that a lower β-catenin expression correlated with poor differentiation of SqCC [36]. We found that β-catenin protein expression obviously correlated with SqCC postoperative recurrence and that patients with negative β-catenin expression had a poor prognosis. Thus, our results demonstrated that βcatenin was related to OS in SqCC and an independent prognostic factor for SqCC. This is the first report on the correlation between the negative expression of β-catenin and a poor prognosis for SqCC patients. The downregulation of E-cadherin with the upregulation of vimentin are important signs of EMT. Seok et al. have reported that in 59 patients with SqCC, the high vimentin expression rate was 10.2 % [38]. Meanwhile, Zhang et al. found the high vimentin expression rate to be 54.4 % in 204 patients with SqCC [37], while Prudkin et al. showed a rate of 42.4 % [32]. We found that the positive vimentin expression rate was 21.4 %. The apparently different findings from these studies suggest that vimentin expression in SqCC needs to be determined in a large cohort of patients. Moreover, the value of vimentin in NSCLC prognosis is also currently controversial. Some studies have indicated that increased expression of vimentin is associated with poor differentiation and lymph node metastasis of NSCLC, while other studies have suggested that vimentin is not an independent predictor of SqCC. Our study found that a positive vimentin expression was not obviously correlated to lymph node metastasis or the TNM stage of SqCC, which was consistent with the report by Seok et al. [37, 38]. Nevertheless, our univariate and multivariate survival analyses indicated that vimentin was an independent predictor in SqCC. However, Seok et al. found that vimentin was an independent predictor in pulmonary AC but not in SqCC. They also found that a high vimentin level correlated with Twist 1 expression in SqCC [38]. Therefore, we propose that vimentin is a predictor for the prognosis of SqCC. Dysregulation of cell–cell adhesion is a key factor in the invasion and metastasis of lung cancer. Tight junctions, adherens junctions, and gap junctions should work together in maintaining the normal epithelial structure. They can interact to connect cells as well as generate and transmit signals inside and outside cells. Many transcription factors, including Snail, Slug, Twist, and ZEB1, regulate EMT [23, 24, 39]. For example, Angelow et al. found that Snail in rat mammary epithelial cell lines induced EMT and downregulated the expression of claudin-3 and E-cadherin [40]. Similarly, a negative correlation between claudins and Twist or ZEB1 was also found by other studies [25, 41]. These results suggested that claudin-3 expression might be regulated by the relevant EMT transcription factors and that claudins may participate in EMT. Indeed, depletion of claudin-3 resulted in increased lung metastasis of ovarian cancer, accompanied with the downregulation of E-cadherin but activation of the β-catenin signaling

pathway [9]. Thus, claudin-3 can regulate the interactions between cells to inhibit tumorigenesis and metastasis through the E-cadherin and β-catenin signaling pathway. Our study showed that claudin-3 protein expression is significantly correlated with E-cadherin, vimentin, and β-catenin expression, suggesting that claudin-3 is involved in EMT of SqCC and that EMT-associated transcription factors can regulate claudin-3 expression in SqCC. Shang found that knockdown of claudin-3 expression in the ovarian carcinoma 2008 cells was accompanied by a decrease in E-cadherin expression and a increase of β-catenin in the cytoplasm and nuclei [9]. However, our data showed that negativity of claudin-3 was associated with reduced expression of both E-cadherin and β-catenin. The discrepancy may be due to the fact that we detected the expression of β-catenin in the membrane, instead of the cytoplasm and nuclei. This is the first report to demonstrate the correlation between claudin-3 expression and EMT in SqCC. Tight junction proteins including claudin-3 and claudin-4 have been becoming novel targets for the development of molecularly targeted therapy for tumors. These claudins are cellular surface receptors for the cytolytic pore-forming CPE. Claudin-3 and claudin-4 over-expressing cells are efficiently lysed by CPE in a cell cycle-independent manner. Moreover, the epithelial tissues from normal small bowel, colon, pancreas, lung, and kidney expressed claudin-3 and claudin-4. Therefore, CPE may be toxic to these tissues. Victor Romanov et al. [28] employed a sequential dual targeting strategy to limit the cytotoxicity of CPE and found that CPE could only be activated by PSA-producing targeted prostate cancer cells simultaneously expressing claudin-3 and claudin-4. Thus, the targeted cytotoxicity of CPE toward PSA-negative, claudin-3, and claudin-4 expressed cells could be greatly decreased. Therefore, claudin-3- and claudin 4-targeted CPE may be a novel therapy strategy for SqCC. In conclusion, our study revealed claudin-3 was an independent prognostic factor for the postoperative patients of SqCC. The decreased expression of claudin-3 was correlated with a downregulation of E-cadherin and βcatenin while an increased expression of vimentin. Our findings suggest that claudin-3 is a potential target for the development of molecularly targeted therapy for SqCC. Acknowledgments This work was supported by Grants No. 81272615 from the Natural Science Foundation of China (to Bangwei Cao), No. 2011-3-007 from the Beijing Municipal Health System High-level Health Person Foundation Project (to Bangwei Cao), and No. yyqdkt2014-11 from the Beijing Friendship Hospital Scientific Research Foundation Project (to Juanjuan Che). Author contributions Juanjuan Che performed the experiments and wrote the manuscript. Yifan Yang, Jing Xiao, and Pengfei Zhao designed the experiment and collected the tumor samples. Bo Yan and Shuo Dong analyzed the experimental data. Bangwei Cao directed the work and edited the manuscript.

Tumor Biol. Conflicts of interest None Informed consent Obtained. Ethics approval The study was approved by the Research Ethics Committee of Beijing Friendship Hospital, Capital Medical University, Beijing, China. Data sharing statement The authors declare that they have no conflicts of interest. This work is original and not under consideration for publication elsewhere.

References 1. 2.

3. 4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

Soini Y. Tight junctions in lung cancer and lung metastasis. A review. Int J Clin Exp Pathol. 2012;5:126–36. PubMed: 22400072. Moore W, Talati R, Bhattacharji P, Bilfinger T. Five-year survival after cryoablation of stage I non-small cell lung cancer in medically inoperable patients. J Vasc Interv Radiol. 2015;26:312–9. PubMed: 25735518. Dempke WC, Suto T, Reck M. Targeted therapies for non-small cell lung cancer. Lung Cancer. 2010;67:257–74. PubMed: 19914732. Tokumo M, Toyooka S, Kiura K, Shigematsu H, Tomii K, Aoe M, et al. The relationship between epidermal growth factor receptor mutations and clinicopathologic features in non–small cell lung cancers. Clin Cancer Res. 2005;11:1167–73. PubMed: 15709185. Tsao MS, Sakurada A, Cutz JC, Zhu CQ, Kamel-Reid S, Squire J, et al. Erlotinib in lung cancer—molecular and clinical predictors of outcome. N Engl J Med. 2005;353:133–44. PubMed: 16014883. Nichols LS, Ashfaq R, Iacobuzio-Donahue CA. Claudin 4 protein expression in primary and metastatic pancreatic cancer support for use as a therapeutic target. Am J Clin Pathol. 2004;121:226–30. PubMed: 14983936. Resnick MB, Konkin T, Routhier J, Sabo E, Pricolo VE. Claudin-1 is a strong prognostic indicator in stage II colonic cancer: a tissue microarray study. Mod Pathol. 2005;18:511–8. PubMed: 15475928. Szasz AM, Nemeth Z, Gyorffy B, Micsinai M, Krenacs T, Baranyai Z, et al. Identification of a claudin-4 and E-cadherin score to predict prognosis in breast cancer. Cancer Sci. 2011;102:2248–54. PubMed: 21883696. Shang X, Lin X, Alvarez E, Manorek G, Howell SB. Tight junction proteins claudin-3 and claudin-4 control tumor growth and metastases. Neoplasia. 2012;14:974–85. PubMed: 23097631. Chao YC, Pan SH, Yang SC, Yu SL, Che TF, Lin CW, et al. Claudin-1 is a metastasis suppressor and correlates with clinical outcome in lung adenocarcinoma. Am J Respir Crit Care Med. 2009;179:123–33. PubMed: 18787218. Heinzelmann-Schwarz VA, Gardiner-Garden M, Henshall SM, Scurry J, Scolyer RA, Davies MJ, et al. Overexpression of the cell adhesion molecules DDR1, claudin 3, and Ep-CAM in metaplastic ovarian epithelium and ovarian cancer. Clin Cancer Res. 2004;10: 4427–36. PubMed: 15240533. Merikallio H, Kaarteenaho R, Pääkkö P, Lehtonen S, Hirvikoski P, Mäkitaro R, et al. Impact of smoking on the expression of claudins in lung carcinoma. Eur J Cancer. 2011;47:620–30. PubMed: 21106366. Moldvay J, Jäckel M, Páska C, Soltész I, Schaff Z, Kiss A. Distinct claudin expression profile in histologic subtypes of lung cancer. Lung Cancer. 2007;57:159–67. PubMed: 17418912. Gao F, Duan X, Lu X, Liu Y, Zheng L, Ding Z, et al. Novel binding between pre-membrane protein and claudin-1 is required for

efficient dengue virus entry. Biochem Biophys Res Commun. 2010;391:952–7. PubMed: 19962368. 15. Van Itallie CM, Betts L, Smedley JG, McClane BA, Anderson JM. Structure of the claudin-binding domain of Clostridium perfringens enterotoxin. J Biol Chem. 2008;283:268–74. PubMed: 17977833. 16. Saeki R, Kondoh M, Kakutani H, Matsuhisa K, Takahashi A, Suzuki H, et al. A claudin-targeting molecule as an inhibitor of tumor metastasis. J Pharmacol Exp Ther. 2010;334:576–82. PubMed: 20442222. 17. Kupferman ME, Jiffar T, El-Naggar A, Yilmaz T, Zhou G, Xie T, et al. TrkB induces EMT and has a key role in invasion of head and neck squamous cell carcinoma. Oncogene. 2010;29:2047–59. PubMed: 20101235. 18. Iwatsuki M, Mimori K, Yokobori T, Ishi H, Beppu T, Nakamori S, et al. Epithelial–mesenchymal transition in cancer development and its clinical significance. Cancer Sci. 2010;101:293–9. PubMed: 19961486. 19. Kase S, Sugio K, Yamazaki K, Okamoto T, Yano T, Sugimachi K. Expression of E-cadherin and beta-catenin in human non-small cell lung cancer and the clinical significance. Clin Cancer Res. 2000;6: 4789–96. PubMed: 11156236. 20. Choi YS, Shim YM, Kim S-H, Son DS, Lee H-S, Kim GY, et al. Prognostic significance of E-cadherin and beta-catenin in resected stage I non-small cell lung cancer. Eur J Cardiothorac Surg. 2003;24:441–9. PubMed: 12965318. 21. Chiu CG, Chan SK, Fang ZA, Masoudi H, Wood-Baker R, Jones SJ, et al. Beta-catenin expression is prognostic of improved nonsmall cell lung cancer survival. Am J Surg. 2012;203:654–9. PubMed: 22402266. 22. Richardson F, Young GD, Sennello R, Wolf J, Argast GM, Mercado P, et al. The evaluation of E-cadherin and vimentin as biomarkers of clinical outcomes among patients with non-small cell lung cancer treated with erlotinib as second- or third-line therapy. Anticancer Res. 2012;32:537–52. PubMed: 22287743. 23. Schmalhofer O, Brabletz S, Brabletz T. E-cadherin, beta-catenin, and ZEB1 in malignant progression of cancer. Cancer Metastasis Rev. 2009;28:151–66. PubMed: 19153669. 24. De Wever O, Pauwels P, De Craene B, Sabbah M, Emami S, Redeuilh G, et al. Molecular and pathological signatures of epithelial–mesenchymal transitions at the cancer invasion front. Histochem Cell Biol. 2008;130:481–94. PubMed: 18648847. 25. Merikallio H, Kaarteenaho R, Pääkkö P, Lehtonen S, Hirvikoski P, Mäkitaro R, et al. Zeb1 and twist are more commonly expressed in metastatic than primary lung tumours and show inverse associations with claudins. J Clin Pathol. 2011;64:136–40. PubMed: 21131312. 26. Lee KW, Lee NK, Kim JH, Kang MS, Yoo HY, Kim HH, et al. Twist1 causes the transcriptional repression of claudin-4 with prognostic significance in esophageal cancer. Biochem Biophys Res Commun. 2012;423:454–60. PubMed: 22668877. 27. Jung JH, Jung CK, Choi HJ. Diagnostic utility of expression of claudins in non-small cell lung cancer: different expression profiles in squamous cell carcinomas and adenocarcinomas. Pathol Res Pract. 2009;205:409–16. PubMed: 19231096. 28. Romanov V, Whyard TC, Waltzer WC, Gabig TG. A claudin 3 and claudin 4-targeted Clostridium perfringens protoxin is selectively cytotoxic to PSA-producing prostate cancer cells. Cancer Lett. 2014;351:260–4. PubMed: 24952257. 29. Yao Q, Zheng QM, Wen JF, Lü T, Wei MQ, Dai SZ. Target-specific cytotoxic activity of recombinant fusion toxin C-CPE-ETA’ against CLDN-3,4-overexpressing ovarian cancer cells. Zhonghua Zhong Liu Za Zhi. 2010;32:897–902. PubMed: 21223796. 30. Ersoz S, Mungan S, Cobanoglu U, Turgutalp H, Ozoran Y. Prognostic importance of claudin-1 and claudin-4 expression in colon carcinomas. Pathol Res Pract. 2011;207:285–9. PubMed: 21493012.

Tumor Biol. 31.

Merikallio H, Pääkkö P, Harju T, Soini Y. Claudins 10 and 18 are predominantly expressed in lung adenocarcinomas and in tumors of nonsmokers. Int J Clin Exp Pathol. 2011;4:667–73. PubMed: 22076167. 32. Prudkin L, Liu DD, Ozburn NC, Sun M, Behrens C, Tang X, et al. Epithelial-to-mesenchymal transition in the development and progression of adenocarcinoma and squamous cell carcinoma of the lung. Mod Pathol. 2009;22:668–78. PubMed: 19270647. 33. Hirata T, Fukuse T, Naiki H, Wada H. Expression of E-cadherin and lymph node metastasis in resected non-small-cell lung cancer. Clin Lung Cancer. 2001;3:134–40. PubMed: 14659029. 34. Hommura F, Furuuchi K, Yamazaki K, Ogura S, Kinoshita I, Shimizu M, et al. Increased expression of beta-catenin predicts better prognosis in nonsmall cell lung carcinomas. Cancer. 2002;94:752–8. PubMed: 11857309. 35. Zhang Y, Han Y, Zheng R, Yu J-H, Miao Y, Wang L, et al. Expression of Frat1 correlates with expression of β-catenin and is associated with a poor clinical outcome in human SCC and AC. Tumour Biol. 2012;33:1437–44. PubMed: 22528942. 36. Pirinen RT, Hirvikoski P, Johansson RT, Hollmén S, Kosma VM. Reduced expression of alpha-catenin, beta-catenin, and gamma-

37.

38.

39.

40. 41.

catenin is associated with high cell proliferative activity and poor differentiation in non-small cell lung cancer. J Clin Pathol. 2001;54: 391–5. PubMed: 11328840. Zhang H, Liu J, Yue D, Gao L, Wang D, Zhang H, et al. Clinical significance of E-cadherin, β-catenin, vimentin and S100A4 expression in completely resected squamous cell lung carcinoma. J Clin Pathol. 2013;66:937–45. PubMed: 23853314. Kim SH, Kim JM, Shin M, Kim CW, Huang SM, Kang DW, et al. Correlation of epithelial-mesenchymal transition markers with clinicopathologic parameters in adenocarcinomas and squamous cell carcinoma of the lung. Histol Histopathol. 2012;27:581–91. PubMed: 22419022. Peinado H, Olmeda D, Cano A. Snail, Zeb and bHLH factors in tumour progression: an alliance against the epithelial phenotype? Nat Rev Cancer. 2007;7:415–28. PubMed: 17508028. Angelow S, Ahlstrom R, Yu AS. Biology of claudins. Am J Physiol Renal Physiol. 2008;295:F867–76. PubMed: 18480174. Soini Y, Tuhkanen H, Sironen R, Virtanen I, Kataja V, Auvinen P, et al. Transcription factors zeb1, twist and snai1 in breast carcinoma. BMC Cancer. 2011;11:73. PubMed: 21324165.