- Email: [email protected]
The Expression of Claudin-1, Claudin-2, Claudin-3, and Claudin-4 in Gastric Cancer Tissue
Journal of Surgical Research 167, e185–e191 (2011) doi:10.1016/j.jss.2010.02.010
The Expression of Claudin-1, Claudin-2, Claudin-3, and Claudin-4 in Gastric Cancer Tissue Hun Jung, M.D.,* Kyong Hwa Jun, M.D., Ph.D.,†,1 Ji Han Jung, M.D., Ph.D.,‡ Hyung Min Chin, M.D., Ph.D.,† and Woo Bae Park, M.D., Ph.D.† *Department of Surgery, Seoul St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea; †Department of Surgery; and ‡Department of Pathology, St. Vincent’s Hospital, College of Medicine, The Catholic University of Korea, Suwon, Korea Submitted for publication December 31, 2009
Background. The claudins (CLDNs) are a family of functional tight junction proteins, and are involved with the epithelial-to-mesenchymal transition (EMT). The claudin proteins have a significant influence on the biological behavior of tumor progression in several types of cancers. In this study, we aimed to evaluate the expression pattern of claudin-1, claudin-2, claudin-3, and claudin-4 in gastric cancer tissue. Materials and methods. Tissue was obtained from surgically resected specimens of 72 patients who were diagnosed with gastric adenocarcinoma at a single institution. The expressions of claudin-1, claudin-2, claudin-3, and claudin-4 were determined by immunohistochemical staining with the ABC method. Results. Claudin-2 demonstrated the highest expression rate (73.6%) and claudin-4 demonstrated the lowest expression rate (44.4%). The expression of claudin-1 was significantly lower in cases of intestinal type adenocarcinoma based on the Lauren classification. The expressions of claudin-3 and claudin-4 were significantly lower in cases with positive lymphatic invasion. The expression of claudin-3 was significantly lower in cases with an advanced T-stage (T3 and T4). The expression of claudin-3 showed significantly positive correlations with the expression of the other claudin proteins. In survival analysis, the expression of claudin-4 was related to good overall survival rate with significance (P [ 0.046). Conclusion. We suggest that claudin-3 and claudin4 represent useful molecular markers for gastric cancer. Claudin-3 and claudin-4 would be the most important proteins related to the lymphatic invasion 1 To whom correspondence and reprint requests should be addressed at Department of Surgery, St. Vincent’s Hospital, The Catholic University of Korea, Ji-dong, Paldal-gu, Suwon, Gyeonggi-do, 442-723 Republic of Korea. E-mail: dkkwkh@catholic. ac.kr.
process, and claudin-4 would be useful with prognostic marker based on our results. Further investigations with a greater number of subjects are required to identify the action mechanism of claudin in gastric cancer. Ó 2011 Elsevier Inc. All rights reserved. Key Words: gastric cancer; claudin-1; claudin-2; claudin-3; claudin-4.
INTRODUCTION
Gastric cancer is the second most frequent cause of death from malignancy worldwide and is the most common cancer in Asian countries, though the disease is less frequently encountered in Western countries [1, 2]. For tumor progression and metastasis, the epithelialto-mesenchymal transition (EMT), which includes loss or redistribution of tight junction proteins such as claudins or E-cadherin, is considered to be an important pathway and trigger mechanism for the ability to progress through the basement membrane [3, 4]. In humans, claudins (CLDNs) are transmembranous tetraspan proteins consisting of approximately 24 subtypes. These proteins, which are essential for the formation of tight junctions in the epithelium, have been identified in numerous tissues, maintain the polarity of epithelial cells and determine the tightness of strands [5–7]. Among the many tight junction proteins, claudins are key functional proteins and expression in humans varies in different cells and tissue. Furthermore, claudins have a significant influence on the biological behavior of tumor progression [8, 9]. For the claudin protein family, claudin-1, claudin-2, claudin-3, and claudin-4 are frequently expressed in human tissue. Expression of claudin-1 is related to a poor prognosis in colon cancer
e185
0022-4804/$36.00 Ó 2011 Elsevier Inc. All rights reserved.
e186
JOURNAL OF SURGICAL RESEARCH: VOL. 167, NO. 2, MAY 15, 2011
and increased re-expression during apoptosis of breast cancer cells [10, 11]. Claudin-2 has a potential barrier function within gastrointestinal organs and expression is related to gastric adenocarcinoma [12, 13]. Claudin-3 and claudin-4, previously classified as intestinal claudin phenotypes, have an important role in the metastatic pathway [14]. However, previous studies have demonstrated the heterogeneities of claudin immunoactivities within various cancer tissues, and the biological function of the claudins has not been clarified. In this study, we aimed to evaluate the expression pattern of claudin-1, claudin-2, claudin-3, and claudin-4 proteins in gastric cancer tissue. MATERIALS AND METHODS Patients and Materials Tissues were obtained from surgically resected specimens of 72 patients, who were diagnosed as having gastric tubular adenocarcinoma, between 2005 and 2007 at the Department of Surgery, St. Vincent’s Hospital, The Catholic University of Korea. The Institutional Review Board of St. Vincent’s Hospital approved this study. All subjects signed a written informed consent form approved by the Institutional Review Board. All patients had undergone a curative radical gastrectomy with standard lymph node dissection except stage IV patients. None of the patients had received any therapy prior to surgery. The tumors were staged according to the sixth International Union Against Cancer (AJCC/UICC) TNM staging system. After surgery, clinical follow-up data for all patients was collected at the outpatient clinic. For survival analysis, stage IV patients were excluded from analysis because those cases had not undergone a curative intended resection. A total 56 patients were finally included in survival analysis. The follow-up period ranged from 2.7 to 48.8 mo, and median time was 36.8 mo.
Immunohistochemistry Claudin expression was determined by the use of immunohistochemical staining with the avidin-biotin complex (ABC) immunoperoxidase method. The tissue, obtained from primary cancer tissue (slides 4 mm), were deparaffinized in xylene and rehydrated in a graded series of ethanol, followed by microwave antigen retrieval. Endogenous peroxidase activity was blocked with the use of 0.3% hydrogen peroxide. The sections were incubated for 1 h or overnight at 4 C using the following primary antibodies: anti-claudin-1, anti-claudin-2, anti-claudin-3, and anti-claudin-4 (1:200 dilution; Lab Vision, Fremont, CA). Immunohistochemical staining was conducted using the rabbit or mouse DAKO ChemMate EnVision system and Peroxidase/DAB kit (DAKO, Glostrup, Denmark). The sections were then counterstained with Meyer’s hematoxylin and were dehydrated, cleared, and mounted. As positive controls, colon cancer samples and normal gastric mucosa samples were used. One pathologist, who was blinded to the specific diagnosis for each individual case, interpreted the results of the immunohistochemical staining. Membranous staining for the four claudin proteins was considered in the evaluation, and the results were scored with the use of a semiquantitative scale from 0 to 4þ as follows. The scores were 0, no staining; 1þ, <25% cells positive and incomplete membranous staining; 2þ, 25%–50% cells positive and incomplete membranous staining; 3þ, 50%–75% cells positive and complete or incomplete membranous staining; 4þ, >75% cells positive and complete membranous staining. For statistical analysis, the expression of claudin-1, claudin-2, claudin-3, and claudin-4 were grouped into negative
(0, 1þ) and positive (2þ, 3þ, 4þ) levels. Clinical data was reviewed from medical records and immunohistochemical staining findings were obtained from pathology reports. We analyzed the association between claudin expression and pathologic results.
Statistics The results from the analysis of the continuous variables are expressed as means 6 standard deviation (SD). Univariate analysis was performed using the c2 or Fisher’s exact test for categorical data. Survival analysis was performed using Kaplan-Meier methods with univariate analysis. All statistical analyses were performed using SPSS software (version 10.0; SPSS, Chicago, IL) and P < 0.05 was considered to indicate a statistically significant difference with a 95% confidence interval (95% CI).
RESULTS
Of the patient, 43 patients (59.7%) were men and 29 (40.3%) patients were women, with a mean age of 60.46 6 11.26 y (age range, 32–80 y). Twenty (27.8%) patients had stage I, 19 (26.4%) patients had stage II, 17 (23.6%) patients had stage III, and 16 patients (22.2%) had stage IV (Table 1). Claudin-1, claudin-2, claudin-3, and claudin-4 expression are shown in Table 2 and Fig. 1. Claudin-2 was determined to have the highest expression rate (73.6%) and claudin-4 was determined to have the lowest expression rate (44.4%) (Table 2, Fig. 1). The expression of claudin-1 was significantly lower in cases of intestinal type adenocarcinoma based on the Lauren classification (P < 0.001). The expressions of claudin-3 and laudin-4 were significantly lower in cases with positive lymphatic invasion (P ¼ 0.01 for claudin-3 expression; P < 0.001 for claudin-4 expression). The expression of claudin-1 was significantly weaker in the M1 stage compared with the M0 stage (P ¼ 0.046) and expression of claudin-3 was significantly lower for an advanced T-stage (T3 and T4) compared with an early stage (T1 and T2) (P ¼ 0.04) (Tables 3 and 4). Based on a correlation analysis of each claudins, a significant positive correlation was found between claudin3 and claudin-1 expression (P ¼ 0.02), claudin-3 and claudin-2 expression (P ¼ 0.01), and claudin-3 and laudin-4 expression (P < 0.001) (Table 5), respectively. Regarding survival results, there were only significant differences of 4-year overall survival result; according to the expression pattern of claudin-4, patients with positive expression in claudin-4 significantly better compared with those with negative ones (95.8% versus 72.2%, P ¼ 0.046) (Table 6, Fig. 2). DISCUSSION
The epithelial-to-mesenchymal transition (EMT) has been considered an important mechanism for cancer progression and metastasis. For the EMT pathway,
e187
JUNG ET AL.: EXPRESSION OF CLAUDINS IN GASTRIC CANCER TISSUE
TABLE 1
TABLE 2
The Baseline Clinical Characteristics
The Results of Claudin Expression in Immunohistochemical Staining
Basic characteristics
Values
Age Gender Male Female Pathophysiologic features Differentiation Well Moderate Poor Lauren’s classification Intestinal Diffuse Lymphatic invasion Positive Negative Venous invasion Positive Negative Neural invasion Positive Negative Growth pattern Expanded Expanded and infiltrative Infiltrative T stage T1 T2 T3 T4 N stage N0 N1 N2 N3 M stage M0 M1 UICC/TNM stage Stage I Stage II Stage III Stage IV Total cases
60.46 6 11.26
Positive expression* (%) 43 (59.7) 29 (40.3)
3 (4.2) 29 (40.3) 40 (55.6) 37 (51.4) 35 (48.6) 46 (63.9) 26 (36.1) 14 (19.4) 58 (80.6) 33 (45.8) 39 (54.2) 16 (22.2) 27 (37.5) 29 (40.3) 6 (8.3) 43 (59.7) 22 (30.6) 1 (1.4) 24 (33.3) 24 (33.3) 16 (22.2) 8 (11.1) 61 (84.7) 11 (15.3) 20 (27.8) 19 (26.4) 17 (23.6) 16 (22.2) 72
tight junction proteins such as claudins, E/M-cadherin, and vimentin are important proteins that are required to preserve the integrity of the cell layer and to control cell proliferation. The functions of tight junction proteins in tumor progression are complex in action and are associated with multiple interactions with other proteins. However, the loss or down-regulation of tight junction proteins in cancer cells has been reported in previous studies [5, 16]. The expression rate and clinicopathologic correlation of E-cadherin expression has been frequently reported for gastric cancer and other malignancies. However, expression of claudin family proteins has been rarely reported, and outcomes have been limited to evaluation
Claudin 1 Claudin 2 Claudin 3 Claudin 4
41 (56.9) 53 (73.6) 37 (51.4) 32 (44.4)
* Grouped into negative (0, 1þ) and positive (2þ, 3þ, 4þ) level expression.
of one or two claudin proteins. Furthermore, the expression of claudin has been reported for numerous types of malignancies such as for the breast, pancreas, liver, and esophagus in previous studies [10–15]. Claudins, however, could be clinically considered as feasible molecular markers for targeting of progression or metastasis in gastric cancer. Matsuda et al. have proposed a novel three-phenotype classification according to the type of claudin expression in gastric cancer [14,16]. In our study, immunohistochemical staining was performed for four subtypes of the claudin family (claudin-1, claudin-2, claudin-3, and claudin-4); expression patterns were evaluated and the probability of interaction was predicted. Though the sample size was small compared with other single claudin studies, variable expression of claudin family members was demonstrated in gastric cancer specimens. The lowest frequency of expression was seen for claudin-4 (44.4%) and the most prominent expression was seen for claudin 2 (73.6%) in the gastric cancer specimens. The expression rate of claudin-1 was 56.9 % (41/72) in our study and significantly decreased for the intestinal type of adenocarcinoma compared with the diffuse type based on the Lauren classification (52.9% versus 74.2%), but there was no significant difference for other pathologic features. However, the expression of claudin-1 also decreased for M1 stage gastric cancer compared with the M0 stage (27.3% versus 62.3%). Disruption of claudin-1 may be associated with carcinogenesis and progression of intestinal type gastric cancer. Paradoxically, it was reported that expression of claudin-1 was significantly decreased for the diffuse type of adenocarcinoma and claudin-1 expression has been demonstrated to be related to the determination of the diffuse phenotype in gastric cancer [9]. It has also been reported that up-regulation of claudin-1 is related to transformation in invasive and metastatic gastric cancer [17]. But, based on the tumor biology, it appears that down-regulation or destruction of tight junction proteins result in tumor progression. This paradoxical issue requires further investigation.
e188
JOURNAL OF SURGICAL RESEARCH: VOL. 167, NO. 2, MAY 15, 2011
FIG. 1. Histological expression of (A) claudin-1, (B) claudin-2, (C) claudin-3 and (D) claudin-4 in gastric cancer samples is shown. Membranous staining for the four claudin proteins was considered in the evaluation. Scoring consisted of 0, no staining; 1þ, < 25% cells positive and incomplete membranous staining; 2þ, 25 -50% cells positive and incomplete membranous staining; 3þ, 50 -75% cells positive and complete or incomplete membranous staining; 4þ, > 75% cells positive and complete membranous staining.
TABLE 3 The Correlation Between Pathologic Features and Claudin Family Expression Pattern Claudin 1 (þ) Differentiation Well 2 Moderate 14 Poor 25 Lauren’s classification Intestinal 18 Diffuse 23 Lymphatic invasion (þ) 23 (–) 18 Venous invasion (þ) 10 (–) 31 Neural invasion (þ) 20 (–) 21 Growth pattern Expanded 6 Exp and inf 13 Infiltrative 12
Claudin 2
Claudin 3
Claudin 4
(–)
P value
(þ)
(–)
P value
(þ)
(–)
P value
(þ)
(–)
P value
1 15 15
0.54
3 19 31
– 10 9
0.40
2 13 22
1 16 18
0.60
2 12 18
1 17 22
0.71
16 8
<0.001
24 25
10 6
0.38
14 20
20 11
0.17
13 16
21 15
0.54
23 8
0.11
33 20
13 6
0.63
18 19
28 7
0.01
13 19
33 7
<0.001
4 27
0.22
10 43
4 15
1.00
4 33
10 25
0.06
4 28
10 30
0.18
13 18
0.56
27 26
6 13
0.15
17 20
16 19
0.98
12 20
21 19
0.20
10 14 17
0.77
4 7 8
12 20 21
0.98
7 14 14
9 13 15
0.88
6 18 16
10 9 13
0.18
e189
JUNG ET AL.: EXPRESSION OF CLAUDINS IN GASTRIC CANCER TISSUE
TABLE 4 The Correlation Between Clinical Stage and Claudin Family Expression Pattern Claudin 1 (þ) T stage T1 4 T2 22 T3 15 T4 – N stage N0 16 N1 12 N2 7 N3 6 M stage M0 38 M1 3 UICC/TNM stage Stage I 13 Stage II 12 Stage III 10 Stage IV 6
Claudin 2
Claudin 3
(–)
P value
(þ)
(–)
P value
(þ)
(–)
P value
(þ)
(–)
P value
2 21 7 1
0.34
4 34 15 –
2 9 7 1
0.26
6 22 9 –
– 21 13 1
0.04
5 17 10 –
1 26 12 1
0.17
8 12 9 2
0.36
17 20 10 6
7 4 6 2
0.52
15 13 6 3
9 11 10 5
0.39
13 13 4 2
11 11 2 6
0.16
23 8
0.046
45 8
16 3
1.00
33 4
28 7
0.28
28 4
33 7
0.75
7 7 7 10
0.35
14 15 12 12
6 4 5 4
0.94
14 11 6 6
6 8 11 10
0.11
12 9 6 5
8 10 11 11
0.29
Prominent expression of claudin-2 was demonstrated in our study (73.6%, 53/72). It was reported that expression of claudin-2 was organ-specific for 2.1% of gastric TABLE 5 The Correlation Analysis Between Claudin Family P value
Claudin 1
Claudin 2
Claudin 3
Claudin 4
Claudin 2 Claudin 3 Claudin 4
0.33 0.02 0.18
– 0.01 0.44
–
– – –
<0.001
Claudin 3–1 Claudin 1
Claudin 3
Positive Negative
Positive
Negative
P value
26 (63.4) 15 (36.6)
11 (35.5) 20 (64.5)
0.02
cancers and 25.3% of colorectal cancers, but there was no obvious correlation with clinicopathologic parameters [18]. There has been no previous study demonstrating expression of claudin-2 in gastric cancer. Claudin-3 and claudin-4 are considered clinically important among the claudins, particularly in the possible application of Clostridium perfringens enterotoxin (CPE) as a new chemotherapeutic compound. CPE is a ligand material for claudin-3 and claudin-4 and binding of the toxin to these claudin proteins leads to a rapid cytotoxic effect in vivo [19, 20]. Several recent studies have shown that CPE may be clinically effective against expression of claudin-3 and claudin-4 in pancreatic and breast cancer [21]. Satake et al. have reported that claudin-3 and claudin-4 have been pathologically classified as intestinal claudins, with expression not only in the intestinal metaplasia of gastric
Claudin 3–2
TABLE 6 The Results of Survival Analysis According to Claudin Family Expression Pattern
Claudin 2 Positive
Negative
P value
32 (60.4) 21 (39.6)
5 (26.3) 14 (73.7)
0.01
Claudin 1 Claudin 3
Positive Negative
(þ)
Claudin 3–4 Claudin 4
Claudin 3
Positive Negative
Claudin 4
Positive
Negative
P value
26 (81.3) 6 (18.8)
11 (27.5) 29 (72.5)
<0.001
(–)
Claudin 2 (þ)
(–)
Disease free survival (%) 2y 81.8 81.0 77.5 92.9 4y 59.3 66.3 59.6 70.7 P value 0.94 0.46 Overall survival (%) 2y 90.8 94.4 89.2 100.0 4y 82.3 83.3 76.0 100.0 P value 0.95 0.06
Claudin 3
Claudin 4
(þ)
(þ)
(–)
(–)
80.0 83.3 68.7 56.5 0.54
80.0 82.8 70.8 55.2 0.39
93.2 95.5 85.8 80.5 0.84
95.8 89.0 95.8 72.2 0.046
e190
JOURNAL OF SURGICAL RESEARCH: VOL. 167, NO. 2, MAY 15, 2011
This study has some limitations. Although the analysis included evaluation of four claudins involved in gastric cancer, a relatively small number of cases were obtained, and transcriptional analysis was not performed. However, this study could be considered as a preliminary study for further clinical transcriptional research of claudin function in gastric cancer. In conclusion, we suggest that claudin-3 and claudin4 represent adequate molecular markers for prediction of cancer progression and prognosis in gastric cancer. Claudin-3 and claudin-4 would be the most important proteins related to the lymphatic invasion process, and claudin-4 would be useful as a prognostic marker, based on our results. Further investigations with a greater number of subjects are required to identify the claudin action pathway and to perform clinically long-term survival analysis. FIG. 2. The results of 4-y overall survival analysis according to expression of claudin-4. There was significant differences according to the expression pattern of claudin-4 (P ¼ 0.046).
mucosa but also in gastric cancer, with regulation of the proteins by Cdx2 [13]. In our study, the expression rates of claudin-3 (51.4%) and claudin-4 (44.4%) were lower compared with the two other claudins. However, expression of claudin-3 and claudin-4 was significantly weaker for cases with positive lymphatic invasion. Especially, lower expression of claudin-3 was determined for cases with an advanced T-stage. It is suggested that claudin-3 and claudin-4 may be involved as important proteins during the lymphatic invasion process in gastric cancer. Author thinks that the administration of CPE with chemotherapeutic agents may be able to apply clinically in cases of claudin-3 and claudin-4 over-expression gastric cancer patients. Based on the findings of the inter-relationship between claudin expressions, we consider that claudin-3 is an important protein for the mechanism of action of the claudin proteins. The expression of claudin-3 showed significantly positive correlation compared with the expression of the other claudin proteins; especially claudin-3 and claudin-4 were revealed strong correlation in expression. We think that claudin-3 may have a role as a universally functional mediator protein, but this point should be studied further for more for definitive evidence. There has been rarely reported survival outcomes associated with claudin expression until now. It was reported that patients with tumors expressing claudin-3 show a better survival than negative ones [9]. In our result, tumors expressing claudin-4 were only related to good prognosis in 4-y overall survival results. But, we could not show 5-y survival results due to short follow-up periods.
ACKNOWLEDGEMENT The authors acknowledge support for this study by a research grant from St. Vincent’s Hospital.
REFERENCES 1. Shibuya K, Mathers CD, Boschi-Pinto C, et al. Global and regional estimates of cancer mortality and incidence by site: II. Results for the global burden of disease 2000. BMC Cancer 2002;26:37. 2. Central Cancer Registry in Korea. 2002 Annual report of the Central Cancer Registry in Korea. Seoul: Ministry of Health and Welfare, Republic of Korea, 2007. 3. Jechlinger M, Grunert S, Tamir IH, et al. Expression profiling of epithelial plasticity in tumor progression. Oncogene 2003; 22:7155. 4. Reichmann E, Schwarz H, Deiner EM, et al. Activation of an inducible c-FosER fusion protein causes loss of epithelial polarity and triggers epithelial-fibroblastoid cell conversion. Cell 1992; 71:1103. 5. Tsukita S, Furuse M. Pores in the wall: Claudins constitute tight junction strands containing aqueous pores. J Cell Biol 2000; 149:13. 6. Furuse M, Sasaki H, Fujimoto K, et al. A single gene product, claudin-1 or -2, reconstitutes tight junction strands and recruits of claudin in fibroblasts. J Cell Biol 1998;143:391. 7. Hewitt KJ, Agarwal R, Morin PJ. The Claudin gene family: Expression in normal and neoplastic tissues. BMC Cancer 2006;12:186. 8. Turksen K, Troy T-C. Barriers built on claudins. J Cell Sci 2004; 117:2435. 9. Soini Y, Tommola S, Helin H, et al. Claudins 1, 3, 4, and 5 in gastric carcinoma, loss of claudin expression associates with the diffuse subtype. Virchows Arch 2006;448:52. 10. Resnick MB, Konkin T, Routhier J, et al. Claudin-1 is a strong prognostic indicator in stage II colonic cancer: A tissue microarray study. Mod Pathol 2005;18:511. 11. Hoevel T, Macek R, Swisshelm K, et al. Re-expression of the TJ protein CLDN1 induces apoptosis in breast tumor spheroids. Int J Cancer 2004;108:374. 12. Xin S, Huixin C, Benchang S, et al. Expression of Cdx2 and claudin-2 in the multistage tissue of gastric carcinogenesis. Oncology 2007;73:357.
JUNG ET AL.: EXPRESSION OF CLAUDINS IN GASTRIC CANCER TISSUE 13. Satake S, Semba S, Matsuda Y, et al. Cdx2 transcription factor regulates claudin-3 and claudin-4 expression during intestinal differentiation of gastric carcinoma. Pathol Int 2008;58:156. 14. Matsuda Y, Semba S, Ueda J, et al. Gastric and intestinal claudin expression at the invasive front of gastric carcinoma. Cancer Sci 2007;98:1014. 15. Mitic LL, Anderson JM. Molecular architecture of tight junctions. Annu Rev Physiol 1998;60:121. 16. Jechlinger M, Grunert S, Tamir IH, et al. Expression profiling of epithelial plasticity in tumor progression. Oncogene 2003; 22:7155. 17. Wu YL, Zhang S, Wang GR, et al. Expression transformation of claudin-1 in the process of gastric adenocarcinoma invasion. World J Gastroenterol 2008;14:4943.
e191
18. Aung PP, Mitani Y, Sanada Y, et al. Differential expression of claudin-2 in normal human tissues and gastrointestinal carcinomas. Virchows Arch 2006;448:428. 19. Katahira J, Sugiyama H, Inoue N, et al. Clostridium perfringens enterotoxin utilizes two structurally related membrane proteins as functional receptors in vivo. J Biol Chem 1997; 272:26652. 20. Michl P, Buchholz M, Rolke M, et al. Claudin-4: A new target for pancreatic cancer treatment using Clostridium perfringens enterotoxin. Gastroenterology 2001;121:678. 21. Kominsky SL, Vali M, Korz D, et al. Clostridium perfringens enterotoxin elicits rapid and specific cytolysis of breast carcinoma cells mediated through tight junction proteins claudin 3 and 4. Am J Pathol 2004;164:1627.
The Expression of Claudin-1, Claudin-2, Claudin-3, and Claudin-4 in Gastric Cancer Tissue Hun Jung, M.D.,* Kyong Hwa Jun, M.D., Ph.D.,†,1 Ji Han Jung, M.D., Ph.D.,‡ Hyung Min Chin, M.D., Ph.D.,† and Woo Bae Park, M.D., Ph.D.† *Department of Surgery, Seoul St. Mary’s Hospital, College of Medicine, The Catholic University of Korea, Seoul, Korea; †Department of Surgery; and ‡Department of Pathology, St. Vincent’s Hospital, College of Medicine, The Catholic University of Korea, Suwon, Korea Submitted for publication December 31, 2009
Background. The claudins (CLDNs) are a family of functional tight junction proteins, and are involved with the epithelial-to-mesenchymal transition (EMT). The claudin proteins have a significant influence on the biological behavior of tumor progression in several types of cancers. In this study, we aimed to evaluate the expression pattern of claudin-1, claudin-2, claudin-3, and claudin-4 in gastric cancer tissue. Materials and methods. Tissue was obtained from surgically resected specimens of 72 patients who were diagnosed with gastric adenocarcinoma at a single institution. The expressions of claudin-1, claudin-2, claudin-3, and claudin-4 were determined by immunohistochemical staining with the ABC method. Results. Claudin-2 demonstrated the highest expression rate (73.6%) and claudin-4 demonstrated the lowest expression rate (44.4%). The expression of claudin-1 was significantly lower in cases of intestinal type adenocarcinoma based on the Lauren classification. The expressions of claudin-3 and claudin-4 were significantly lower in cases with positive lymphatic invasion. The expression of claudin-3 was significantly lower in cases with an advanced T-stage (T3 and T4). The expression of claudin-3 showed significantly positive correlations with the expression of the other claudin proteins. In survival analysis, the expression of claudin-4 was related to good overall survival rate with significance (P [ 0.046). Conclusion. We suggest that claudin-3 and claudin4 represent useful molecular markers for gastric cancer. Claudin-3 and claudin-4 would be the most important proteins related to the lymphatic invasion 1 To whom correspondence and reprint requests should be addressed at Department of Surgery, St. Vincent’s Hospital, The Catholic University of Korea, Ji-dong, Paldal-gu, Suwon, Gyeonggi-do, 442-723 Republic of Korea. E-mail: dkkwkh@catholic. ac.kr.
process, and claudin-4 would be useful with prognostic marker based on our results. Further investigations with a greater number of subjects are required to identify the action mechanism of claudin in gastric cancer. Ó 2011 Elsevier Inc. All rights reserved. Key Words: gastric cancer; claudin-1; claudin-2; claudin-3; claudin-4.
INTRODUCTION
Gastric cancer is the second most frequent cause of death from malignancy worldwide and is the most common cancer in Asian countries, though the disease is less frequently encountered in Western countries [1, 2]. For tumor progression and metastasis, the epithelialto-mesenchymal transition (EMT), which includes loss or redistribution of tight junction proteins such as claudins or E-cadherin, is considered to be an important pathway and trigger mechanism for the ability to progress through the basement membrane [3, 4]. In humans, claudins (CLDNs) are transmembranous tetraspan proteins consisting of approximately 24 subtypes. These proteins, which are essential for the formation of tight junctions in the epithelium, have been identified in numerous tissues, maintain the polarity of epithelial cells and determine the tightness of strands [5–7]. Among the many tight junction proteins, claudins are key functional proteins and expression in humans varies in different cells and tissue. Furthermore, claudins have a significant influence on the biological behavior of tumor progression [8, 9]. For the claudin protein family, claudin-1, claudin-2, claudin-3, and claudin-4 are frequently expressed in human tissue. Expression of claudin-1 is related to a poor prognosis in colon cancer
e185
0022-4804/$36.00 Ó 2011 Elsevier Inc. All rights reserved.
e186
JOURNAL OF SURGICAL RESEARCH: VOL. 167, NO. 2, MAY 15, 2011
and increased re-expression during apoptosis of breast cancer cells [10, 11]. Claudin-2 has a potential barrier function within gastrointestinal organs and expression is related to gastric adenocarcinoma [12, 13]. Claudin-3 and claudin-4, previously classified as intestinal claudin phenotypes, have an important role in the metastatic pathway [14]. However, previous studies have demonstrated the heterogeneities of claudin immunoactivities within various cancer tissues, and the biological function of the claudins has not been clarified. In this study, we aimed to evaluate the expression pattern of claudin-1, claudin-2, claudin-3, and claudin-4 proteins in gastric cancer tissue. MATERIALS AND METHODS Patients and Materials Tissues were obtained from surgically resected specimens of 72 patients, who were diagnosed as having gastric tubular adenocarcinoma, between 2005 and 2007 at the Department of Surgery, St. Vincent’s Hospital, The Catholic University of Korea. The Institutional Review Board of St. Vincent’s Hospital approved this study. All subjects signed a written informed consent form approved by the Institutional Review Board. All patients had undergone a curative radical gastrectomy with standard lymph node dissection except stage IV patients. None of the patients had received any therapy prior to surgery. The tumors were staged according to the sixth International Union Against Cancer (AJCC/UICC) TNM staging system. After surgery, clinical follow-up data for all patients was collected at the outpatient clinic. For survival analysis, stage IV patients were excluded from analysis because those cases had not undergone a curative intended resection. A total 56 patients were finally included in survival analysis. The follow-up period ranged from 2.7 to 48.8 mo, and median time was 36.8 mo.
Immunohistochemistry Claudin expression was determined by the use of immunohistochemical staining with the avidin-biotin complex (ABC) immunoperoxidase method. The tissue, obtained from primary cancer tissue (slides 4 mm), were deparaffinized in xylene and rehydrated in a graded series of ethanol, followed by microwave antigen retrieval. Endogenous peroxidase activity was blocked with the use of 0.3% hydrogen peroxide. The sections were incubated for 1 h or overnight at 4 C using the following primary antibodies: anti-claudin-1, anti-claudin-2, anti-claudin-3, and anti-claudin-4 (1:200 dilution; Lab Vision, Fremont, CA). Immunohistochemical staining was conducted using the rabbit or mouse DAKO ChemMate EnVision system and Peroxidase/DAB kit (DAKO, Glostrup, Denmark). The sections were then counterstained with Meyer’s hematoxylin and were dehydrated, cleared, and mounted. As positive controls, colon cancer samples and normal gastric mucosa samples were used. One pathologist, who was blinded to the specific diagnosis for each individual case, interpreted the results of the immunohistochemical staining. Membranous staining for the four claudin proteins was considered in the evaluation, and the results were scored with the use of a semiquantitative scale from 0 to 4þ as follows. The scores were 0, no staining; 1þ, <25% cells positive and incomplete membranous staining; 2þ, 25%–50% cells positive and incomplete membranous staining; 3þ, 50%–75% cells positive and complete or incomplete membranous staining; 4þ, >75% cells positive and complete membranous staining. For statistical analysis, the expression of claudin-1, claudin-2, claudin-3, and claudin-4 were grouped into negative
(0, 1þ) and positive (2þ, 3þ, 4þ) levels. Clinical data was reviewed from medical records and immunohistochemical staining findings were obtained from pathology reports. We analyzed the association between claudin expression and pathologic results.
Statistics The results from the analysis of the continuous variables are expressed as means 6 standard deviation (SD). Univariate analysis was performed using the c2 or Fisher’s exact test for categorical data. Survival analysis was performed using Kaplan-Meier methods with univariate analysis. All statistical analyses were performed using SPSS software (version 10.0; SPSS, Chicago, IL) and P < 0.05 was considered to indicate a statistically significant difference with a 95% confidence interval (95% CI).
RESULTS
Of the patient, 43 patients (59.7%) were men and 29 (40.3%) patients were women, with a mean age of 60.46 6 11.26 y (age range, 32–80 y). Twenty (27.8%) patients had stage I, 19 (26.4%) patients had stage II, 17 (23.6%) patients had stage III, and 16 patients (22.2%) had stage IV (Table 1). Claudin-1, claudin-2, claudin-3, and claudin-4 expression are shown in Table 2 and Fig. 1. Claudin-2 was determined to have the highest expression rate (73.6%) and claudin-4 was determined to have the lowest expression rate (44.4%) (Table 2, Fig. 1). The expression of claudin-1 was significantly lower in cases of intestinal type adenocarcinoma based on the Lauren classification (P < 0.001). The expressions of claudin-3 and laudin-4 were significantly lower in cases with positive lymphatic invasion (P ¼ 0.01 for claudin-3 expression; P < 0.001 for claudin-4 expression). The expression of claudin-1 was significantly weaker in the M1 stage compared with the M0 stage (P ¼ 0.046) and expression of claudin-3 was significantly lower for an advanced T-stage (T3 and T4) compared with an early stage (T1 and T2) (P ¼ 0.04) (Tables 3 and 4). Based on a correlation analysis of each claudins, a significant positive correlation was found between claudin3 and claudin-1 expression (P ¼ 0.02), claudin-3 and claudin-2 expression (P ¼ 0.01), and claudin-3 and laudin-4 expression (P < 0.001) (Table 5), respectively. Regarding survival results, there were only significant differences of 4-year overall survival result; according to the expression pattern of claudin-4, patients with positive expression in claudin-4 significantly better compared with those with negative ones (95.8% versus 72.2%, P ¼ 0.046) (Table 6, Fig. 2). DISCUSSION
The epithelial-to-mesenchymal transition (EMT) has been considered an important mechanism for cancer progression and metastasis. For the EMT pathway,
e187
JUNG ET AL.: EXPRESSION OF CLAUDINS IN GASTRIC CANCER TISSUE
TABLE 1
TABLE 2
The Baseline Clinical Characteristics
The Results of Claudin Expression in Immunohistochemical Staining
Basic characteristics
Values
Age Gender Male Female Pathophysiologic features Differentiation Well Moderate Poor Lauren’s classification Intestinal Diffuse Lymphatic invasion Positive Negative Venous invasion Positive Negative Neural invasion Positive Negative Growth pattern Expanded Expanded and infiltrative Infiltrative T stage T1 T2 T3 T4 N stage N0 N1 N2 N3 M stage M0 M1 UICC/TNM stage Stage I Stage II Stage III Stage IV Total cases
60.46 6 11.26
Positive expression* (%) 43 (59.7) 29 (40.3)
3 (4.2) 29 (40.3) 40 (55.6) 37 (51.4) 35 (48.6) 46 (63.9) 26 (36.1) 14 (19.4) 58 (80.6) 33 (45.8) 39 (54.2) 16 (22.2) 27 (37.5) 29 (40.3) 6 (8.3) 43 (59.7) 22 (30.6) 1 (1.4) 24 (33.3) 24 (33.3) 16 (22.2) 8 (11.1) 61 (84.7) 11 (15.3) 20 (27.8) 19 (26.4) 17 (23.6) 16 (22.2) 72
tight junction proteins such as claudins, E/M-cadherin, and vimentin are important proteins that are required to preserve the integrity of the cell layer and to control cell proliferation. The functions of tight junction proteins in tumor progression are complex in action and are associated with multiple interactions with other proteins. However, the loss or down-regulation of tight junction proteins in cancer cells has been reported in previous studies [5, 16]. The expression rate and clinicopathologic correlation of E-cadherin expression has been frequently reported for gastric cancer and other malignancies. However, expression of claudin family proteins has been rarely reported, and outcomes have been limited to evaluation
Claudin 1 Claudin 2 Claudin 3 Claudin 4
41 (56.9) 53 (73.6) 37 (51.4) 32 (44.4)
* Grouped into negative (0, 1þ) and positive (2þ, 3þ, 4þ) level expression.
of one or two claudin proteins. Furthermore, the expression of claudin has been reported for numerous types of malignancies such as for the breast, pancreas, liver, and esophagus in previous studies [10–15]. Claudins, however, could be clinically considered as feasible molecular markers for targeting of progression or metastasis in gastric cancer. Matsuda et al. have proposed a novel three-phenotype classification according to the type of claudin expression in gastric cancer [14,16]. In our study, immunohistochemical staining was performed for four subtypes of the claudin family (claudin-1, claudin-2, claudin-3, and claudin-4); expression patterns were evaluated and the probability of interaction was predicted. Though the sample size was small compared with other single claudin studies, variable expression of claudin family members was demonstrated in gastric cancer specimens. The lowest frequency of expression was seen for claudin-4 (44.4%) and the most prominent expression was seen for claudin 2 (73.6%) in the gastric cancer specimens. The expression rate of claudin-1 was 56.9 % (41/72) in our study and significantly decreased for the intestinal type of adenocarcinoma compared with the diffuse type based on the Lauren classification (52.9% versus 74.2%), but there was no significant difference for other pathologic features. However, the expression of claudin-1 also decreased for M1 stage gastric cancer compared with the M0 stage (27.3% versus 62.3%). Disruption of claudin-1 may be associated with carcinogenesis and progression of intestinal type gastric cancer. Paradoxically, it was reported that expression of claudin-1 was significantly decreased for the diffuse type of adenocarcinoma and claudin-1 expression has been demonstrated to be related to the determination of the diffuse phenotype in gastric cancer [9]. It has also been reported that up-regulation of claudin-1 is related to transformation in invasive and metastatic gastric cancer [17]. But, based on the tumor biology, it appears that down-regulation or destruction of tight junction proteins result in tumor progression. This paradoxical issue requires further investigation.
e188
JOURNAL OF SURGICAL RESEARCH: VOL. 167, NO. 2, MAY 15, 2011
FIG. 1. Histological expression of (A) claudin-1, (B) claudin-2, (C) claudin-3 and (D) claudin-4 in gastric cancer samples is shown. Membranous staining for the four claudin proteins was considered in the evaluation. Scoring consisted of 0, no staining; 1þ, < 25% cells positive and incomplete membranous staining; 2þ, 25 -50% cells positive and incomplete membranous staining; 3þ, 50 -75% cells positive and complete or incomplete membranous staining; 4þ, > 75% cells positive and complete membranous staining.
TABLE 3 The Correlation Between Pathologic Features and Claudin Family Expression Pattern Claudin 1 (þ) Differentiation Well 2 Moderate 14 Poor 25 Lauren’s classification Intestinal 18 Diffuse 23 Lymphatic invasion (þ) 23 (–) 18 Venous invasion (þ) 10 (–) 31 Neural invasion (þ) 20 (–) 21 Growth pattern Expanded 6 Exp and inf 13 Infiltrative 12
Claudin 2
Claudin 3
Claudin 4
(–)
P value
(þ)
(–)
P value
(þ)
(–)
P value
(þ)
(–)
P value
1 15 15
0.54
3 19 31
– 10 9
0.40
2 13 22
1 16 18
0.60
2 12 18
1 17 22
0.71
16 8
<0.001
24 25
10 6
0.38
14 20
20 11
0.17
13 16
21 15
0.54
23 8
0.11
33 20
13 6
0.63
18 19
28 7
0.01
13 19
33 7
<0.001
4 27
0.22
10 43
4 15
1.00
4 33
10 25
0.06
4 28
10 30
0.18
13 18
0.56
27 26
6 13
0.15
17 20
16 19
0.98
12 20
21 19
0.20
10 14 17
0.77
4 7 8
12 20 21
0.98
7 14 14
9 13 15
0.88
6 18 16
10 9 13
0.18
e189
JUNG ET AL.: EXPRESSION OF CLAUDINS IN GASTRIC CANCER TISSUE
TABLE 4 The Correlation Between Clinical Stage and Claudin Family Expression Pattern Claudin 1 (þ) T stage T1 4 T2 22 T3 15 T4 – N stage N0 16 N1 12 N2 7 N3 6 M stage M0 38 M1 3 UICC/TNM stage Stage I 13 Stage II 12 Stage III 10 Stage IV 6
Claudin 2
Claudin 3
(–)
P value
(þ)
(–)
P value
(þ)
(–)
P value
(þ)
(–)
P value
2 21 7 1
0.34
4 34 15 –
2 9 7 1
0.26
6 22 9 –
– 21 13 1
0.04
5 17 10 –
1 26 12 1
0.17
8 12 9 2
0.36
17 20 10 6
7 4 6 2
0.52
15 13 6 3
9 11 10 5
0.39
13 13 4 2
11 11 2 6
0.16
23 8
0.046
45 8
16 3
1.00
33 4
28 7
0.28
28 4
33 7
0.75
7 7 7 10
0.35
14 15 12 12
6 4 5 4
0.94
14 11 6 6
6 8 11 10
0.11
12 9 6 5
8 10 11 11
0.29
Prominent expression of claudin-2 was demonstrated in our study (73.6%, 53/72). It was reported that expression of claudin-2 was organ-specific for 2.1% of gastric TABLE 5 The Correlation Analysis Between Claudin Family P value
Claudin 1
Claudin 2
Claudin 3
Claudin 4
Claudin 2 Claudin 3 Claudin 4
0.33 0.02 0.18
– 0.01 0.44
–
– – –
<0.001
Claudin 3–1 Claudin 1
Claudin 3
Positive Negative
Positive
Negative
P value
26 (63.4) 15 (36.6)
11 (35.5) 20 (64.5)
0.02
cancers and 25.3% of colorectal cancers, but there was no obvious correlation with clinicopathologic parameters [18]. There has been no previous study demonstrating expression of claudin-2 in gastric cancer. Claudin-3 and claudin-4 are considered clinically important among the claudins, particularly in the possible application of Clostridium perfringens enterotoxin (CPE) as a new chemotherapeutic compound. CPE is a ligand material for claudin-3 and claudin-4 and binding of the toxin to these claudin proteins leads to a rapid cytotoxic effect in vivo [19, 20]. Several recent studies have shown that CPE may be clinically effective against expression of claudin-3 and claudin-4 in pancreatic and breast cancer [21]. Satake et al. have reported that claudin-3 and claudin-4 have been pathologically classified as intestinal claudins, with expression not only in the intestinal metaplasia of gastric
Claudin 3–2
TABLE 6 The Results of Survival Analysis According to Claudin Family Expression Pattern
Claudin 2 Positive
Negative
P value
32 (60.4) 21 (39.6)
5 (26.3) 14 (73.7)
0.01
Claudin 1 Claudin 3
Positive Negative
(þ)
Claudin 3–4 Claudin 4
Claudin 3
Positive Negative
Claudin 4
Positive
Negative
P value
26 (81.3) 6 (18.8)
11 (27.5) 29 (72.5)
<0.001
(–)
Claudin 2 (þ)
(–)
Disease free survival (%) 2y 81.8 81.0 77.5 92.9 4y 59.3 66.3 59.6 70.7 P value 0.94 0.46 Overall survival (%) 2y 90.8 94.4 89.2 100.0 4y 82.3 83.3 76.0 100.0 P value 0.95 0.06
Claudin 3
Claudin 4
(þ)
(þ)
(–)
(–)
80.0 83.3 68.7 56.5 0.54
80.0 82.8 70.8 55.2 0.39
93.2 95.5 85.8 80.5 0.84
95.8 89.0 95.8 72.2 0.046
e190
JOURNAL OF SURGICAL RESEARCH: VOL. 167, NO. 2, MAY 15, 2011
This study has some limitations. Although the analysis included evaluation of four claudins involved in gastric cancer, a relatively small number of cases were obtained, and transcriptional analysis was not performed. However, this study could be considered as a preliminary study for further clinical transcriptional research of claudin function in gastric cancer. In conclusion, we suggest that claudin-3 and claudin4 represent adequate molecular markers for prediction of cancer progression and prognosis in gastric cancer. Claudin-3 and claudin-4 would be the most important proteins related to the lymphatic invasion process, and claudin-4 would be useful as a prognostic marker, based on our results. Further investigations with a greater number of subjects are required to identify the claudin action pathway and to perform clinically long-term survival analysis. FIG. 2. The results of 4-y overall survival analysis according to expression of claudin-4. There was significant differences according to the expression pattern of claudin-4 (P ¼ 0.046).
mucosa but also in gastric cancer, with regulation of the proteins by Cdx2 [13]. In our study, the expression rates of claudin-3 (51.4%) and claudin-4 (44.4%) were lower compared with the two other claudins. However, expression of claudin-3 and claudin-4 was significantly weaker for cases with positive lymphatic invasion. Especially, lower expression of claudin-3 was determined for cases with an advanced T-stage. It is suggested that claudin-3 and claudin-4 may be involved as important proteins during the lymphatic invasion process in gastric cancer. Author thinks that the administration of CPE with chemotherapeutic agents may be able to apply clinically in cases of claudin-3 and claudin-4 over-expression gastric cancer patients. Based on the findings of the inter-relationship between claudin expressions, we consider that claudin-3 is an important protein for the mechanism of action of the claudin proteins. The expression of claudin-3 showed significantly positive correlation compared with the expression of the other claudin proteins; especially claudin-3 and claudin-4 were revealed strong correlation in expression. We think that claudin-3 may have a role as a universally functional mediator protein, but this point should be studied further for more for definitive evidence. There has been rarely reported survival outcomes associated with claudin expression until now. It was reported that patients with tumors expressing claudin-3 show a better survival than negative ones [9]. In our result, tumors expressing claudin-4 were only related to good prognosis in 4-y overall survival results. But, we could not show 5-y survival results due to short follow-up periods.
ACKNOWLEDGEMENT The authors acknowledge support for this study by a research grant from St. Vincent’s Hospital.
REFERENCES 1. Shibuya K, Mathers CD, Boschi-Pinto C, et al. Global and regional estimates of cancer mortality and incidence by site: II. Results for the global burden of disease 2000. BMC Cancer 2002;26:37. 2. Central Cancer Registry in Korea. 2002 Annual report of the Central Cancer Registry in Korea. Seoul: Ministry of Health and Welfare, Republic of Korea, 2007. 3. Jechlinger M, Grunert S, Tamir IH, et al. Expression profiling of epithelial plasticity in tumor progression. Oncogene 2003; 22:7155. 4. Reichmann E, Schwarz H, Deiner EM, et al. Activation of an inducible c-FosER fusion protein causes loss of epithelial polarity and triggers epithelial-fibroblastoid cell conversion. Cell 1992; 71:1103. 5. Tsukita S, Furuse M. Pores in the wall: Claudins constitute tight junction strands containing aqueous pores. J Cell Biol 2000; 149:13. 6. Furuse M, Sasaki H, Fujimoto K, et al. A single gene product, claudin-1 or -2, reconstitutes tight junction strands and recruits of claudin in fibroblasts. J Cell Biol 1998;143:391. 7. Hewitt KJ, Agarwal R, Morin PJ. The Claudin gene family: Expression in normal and neoplastic tissues. BMC Cancer 2006;12:186. 8. Turksen K, Troy T-C. Barriers built on claudins. J Cell Sci 2004; 117:2435. 9. Soini Y, Tommola S, Helin H, et al. Claudins 1, 3, 4, and 5 in gastric carcinoma, loss of claudin expression associates with the diffuse subtype. Virchows Arch 2006;448:52. 10. Resnick MB, Konkin T, Routhier J, et al. Claudin-1 is a strong prognostic indicator in stage II colonic cancer: A tissue microarray study. Mod Pathol 2005;18:511. 11. Hoevel T, Macek R, Swisshelm K, et al. Re-expression of the TJ protein CLDN1 induces apoptosis in breast tumor spheroids. Int J Cancer 2004;108:374. 12. Xin S, Huixin C, Benchang S, et al. Expression of Cdx2 and claudin-2 in the multistage tissue of gastric carcinogenesis. Oncology 2007;73:357.
JUNG ET AL.: EXPRESSION OF CLAUDINS IN GASTRIC CANCER TISSUE 13. Satake S, Semba S, Matsuda Y, et al. Cdx2 transcription factor regulates claudin-3 and claudin-4 expression during intestinal differentiation of gastric carcinoma. Pathol Int 2008;58:156. 14. Matsuda Y, Semba S, Ueda J, et al. Gastric and intestinal claudin expression at the invasive front of gastric carcinoma. Cancer Sci 2007;98:1014. 15. Mitic LL, Anderson JM. Molecular architecture of tight junctions. Annu Rev Physiol 1998;60:121. 16. Jechlinger M, Grunert S, Tamir IH, et al. Expression profiling of epithelial plasticity in tumor progression. Oncogene 2003; 22:7155. 17. Wu YL, Zhang S, Wang GR, et al. Expression transformation of claudin-1 in the process of gastric adenocarcinoma invasion. World J Gastroenterol 2008;14:4943.
e191
18. Aung PP, Mitani Y, Sanada Y, et al. Differential expression of claudin-2 in normal human tissues and gastrointestinal carcinomas. Virchows Arch 2006;448:428. 19. Katahira J, Sugiyama H, Inoue N, et al. Clostridium perfringens enterotoxin utilizes two structurally related membrane proteins as functional receptors in vivo. J Biol Chem 1997; 272:26652. 20. Michl P, Buchholz M, Rolke M, et al. Claudin-4: A new target for pancreatic cancer treatment using Clostridium perfringens enterotoxin. Gastroenterology 2001;121:678. 21. Kominsky SL, Vali M, Korz D, et al. Clostridium perfringens enterotoxin elicits rapid and specific cytolysis of breast carcinoma cells mediated through tight junction proteins claudin 3 and 4. Am J Pathol 2004;164:1627.