- Email: [email protected]
E-cadherin gene polymorphism and risk of urothelial cancer
Cancer Letters 195 (2003) 53–58 www.elsevier.com/locate/canlet
E-cadherin gene polymorphism and risk of urothelial cancer Hiromasa Tsukinoa,b, Yoshiki Kurodaa, Hiroyuki Nakaoa, Hirohisa Imaia, Hisato Inatomic, Kiyotaka Kohshid, Yukio Osadab, Takahiko Katoha,* a
Department of Public Health, School of Medicine, Miyazaki Medical College, 5200 Kihara, Miyazaki, 889-1692, Japan b Department of Urology, School of Medicine, Miyazaki Medical College, 5200 Kihara, Miyazaki, 889-1692, Japan c Munakata Suikoukai General Hospital, Munakata, 811-3298, Japan d Department of Hyperbaric Medicine and Neurosurgery, School of Medicine, University of Occupational and Environmental Health (UOEH), 1-1 Iseigaoka, Yahatanishi-ku, Kitakyushu 807-8555, Japan Received 16 January 2003; accepted 20 January 2003
Abstract The E-cadherin is important in cell– cell adhesion and in the development and maintenance of the epithelial phenotype. A 2 160 C ! A polymorphism in the promoter region of E-cadherin has been shown to decrease gene transcription. This allelic variation may be a potential genetic marker that can help identify those individuals at higher risk for invasive/metastatic disease. We studied the effect of E-cadherin gene polymorphism on urothelial cancer susceptibility in a case control study of 314 urothelial cancer patients and 314 age– sex matched controls, to determine whether this polymorphism is a biomarker for the risk and how aggressive the disease is. The frequency with which the subjects carried E-cadherin A/A genotype was significantly higher in the urothelial cancer patients than in the healthy control subjects (OR ¼ 2.32, 95% CI 1.03– 5.22). Subdividing urothelial cancer according to tumor differentiation and stage, we found no association between E-cadherin polymorphism and poorly-differentiation and invasiveness of urothelial cancer. Furthermore, no significant association between E-cadherin polymorphism and recurrence rate of urothelial cancer patients was found. The present study demonstrates for the first time that E-cadherin A/A genotype may be associated with susceptibility to urothelial cancer, but not with the progression of disease. q 2003 Elsevier Science Ireland Ltd. All rights reserved. Keywords: E-cadherin; Polymorphism; Urothelial cancer; Promoter; SNP
1. Introduction Cadherins are transmembrane Ca2þ dependent Abbreviations: PCR, polymerase chain reaction; RFLP, restriction fragment length polymorphism; OR, odds ratio; 95% CI, 95% confidence interval. * Corresponding author. Tel.: þ81-985-85-0874; fax: þ 81-98585-6258. E-mail address: [email protected] (T. Katoh).
homophilic adhesion receptors that have important roles in cell recognition and cell sorting during development [1]. Cadherin are localized in specialized cell – cell adhesion sites termed adherence junctions. At these sites cadherins establish linkages with the actin containing cytoskeleton. The classic cadherins include E, N and P-cadherin. E-cadherin is expressed in all epithelial tissue and is found on the plasma membrane of squamous and transitional cells,
0304-3835/03/$ - see front matter q 2003 Elsevier Science Ireland Ltd. All rights reserved. doi:10.1016/S0304-3835(03)00130-7
54
H. Tsukino et al. / Cancer Letters 195 (2003) 53–58
in contrast to the concentrated expression at the intermediate junction (zonula adherens and belt desmosome) of polarized cells, such as intestinal or prostatic epithelium. Perturbation of E-cadherin mediated cell adhesion is involved in tumor progression and metastasis. In vitro loss of E-cadherin expression is associated with loss of cellular differentiation and increased cellular invasiveness and infiltration, while transfection of E-cadherin negative cell lines with E-cadherin complimentary DNA is able to suppress the invasive behavior [2]. Investigation of the expression of E-cadherin histopathological material of human transitional cell carcinoma demonstrates that aberrant expression of E-cadherin correlates with lack of differentiation, muscle invasion and distant metastasis [3]. Loss of normal E-cadherin expression has also been shown to correlate with decreased recurrence-free and overall survival, although multivariate analysis suggests that it has no independent prognostic value over the grade and stage of the tumor [4]. A number of possible mechanisms have been proposed to explain the documented reduction in E-cadherin function in bladder cells undergoing malignant transformation. They include suppression or mutation of E-cadherin gene [5], transformation disorder [6] or increased proteasemediated degradation [7]. Recently, a C/A single nucleotide polymorphism (SNP) at 2 160 from the transcriptional start site of the E-cadherin gene promoter has been identified by Li et al. [8]. In vitro studies showed that the A allele of this polymorphism decreased the transcriptional efficiency by 68% compared with the C allele. They suggested that 2 160 C/A polymorphism has a direct effect on E-cadherin gene transcriptional regulation, and this allelic variation may be a potential genetic marker that can help identify those individuals at higher risk for invasive/metastatic disease. However, to our knowledge, there are no reports investigating the association between E-cadherin gene polymorphism and urothelial cancer risk. In this study, we hypothesized that E-cadherin polymorphism is responsible for interindividual variation in the production of E-cadherin and in turn leads to individual susceptibility to invasiveness, differentiation and recurrence. We studied E-cadherin genotype in 314 urothelial cancer patients and 314 controls. Additionally, we assessed E-cadherin geno-
type in relation to clinical findings associated with the outcome. These included differentiation, stage and recurrence. We have demonstrated a significant association between E-cadherin genotype and urothelial cancer susceptibility.
2. Materials and methods 2.1. Subjects The case groups comprised 314 patients with urothelial cancer (bladder n ¼ 240, renal pelvis and ureter n ¼ 37, overlap n ¼ 37) (244 men, 70 women; mean age 69.6 years) in Kitakyushu city and Miyazaki city, Japan. This study was designed to recruit all eligible urothelial cancer cases who were newly diagnosed with urothelial cancer during 1992 –2002 in Kitakyushu city and 2000 –2002 in Miyazaki city. The patients were treated at the University of Occupational and Environmental Health Hospital and Miyazaki Medical College Hospital, and had been histologically diagnosed for urothelial transitional cell carcinoma. Histological grading and staging was done according to the Japanese classification system [9]. Thirty-two patients (10.2%) had a well-differentiated disease (G1), 142 (45.2%) had a moderately differentiated (G2) and 140 (44.6%) had a poorly differentiated disease (G3). We used the term ‘superficial tumors’ to refer to those which were limited to the mucosa (pTis, pTa) or the lamina propria (pT1); and ‘invasive tumors’ to refer to those which had invaded the muscle layer (pT2, pT3) or deeper (pT4). One hundred and eighty-two patients (58.0%) had superficial tumors, 124 (39.5%) had invasive tumors and eight (2.5%) had unknown stage tumors. A total of 314 controls, frequency-matched with cases for age (^ 5 years) and gender, were selected from the people who visited a medical institution located in Kitakyushu city for a general health check-up (244 men, 70 women, mean age 67.2 years). All study subjects completed a questionnaire administered by a trained interviewer covering medical, residential, and occupational exposures as well as smoking history. Smoking history was summarized as the total amount of cigarettes consumed during their lifetime up until the time of the interview. The amount of tobacco smoke exposure
H. Tsukino et al. / Cancer Letters 195 (2003) 53–58
was calculated as pack –years [one pack (20 cigarettes)/day £ years of smoking]. All participants were given an explanation of the nature of the study, and informed consent was obtained. This study was approved by the ethics committees of the University of Occupational and Environmental Health and Miyazaki Medical College. 2.2. Genotyping Genomic DNA was isolated from peripheral leukocytes by proteinase K digestion and phenol/chloroform extraction. The genetic polymorphism of E-cadherin was determined by polymerase chain reaction (PCR) amplification followed by digestion with Afl III and Hph I, using the method described previously [10]. Briefly, the two primers used in the PCR were 50 -TCCCAGGTCTTAGTGAGCCA-30 and 50 -GGCCACAGCCAATCAGCA-30 , yielding a single 190-base pair (bp) product. One hundred nanograms of DNA were amplified in a total volume of 30 ml containing 10 pmole of each primer, 2 U of Taq polymerase, 1.5 mM MgCl2, and PCR buffer. Amplification was performed by denaturing at 968C for 30 s, annealing at 608C for 30 s and extending at 728C for 30 s for 35 cycles using a Perkin – Elmer 9700. After confirmation of the amplified fragment of the expected size on 3% agarose gel, the PCR products were digested with 5 U of restriction enzyme HphI or AflIII at 378C for 16 h, and electrophoresed on a 3% agarose gel. The C allele created an HphI site, and the A allele created an AflIII site. All study subjects were analyzed using both restriction enzymes. 2.3. Statistical analysis Crude odds ratios (ORs) and 95% confidence intervals (95% CIs) were calculated for E-cadherin genotypes. ORs were adjusted for age, gender and smoking status (ever- and never- smokers), using multiple logistic regression analysis by SPSS Medical Pack for Windows. To examine the interaction between environmental and genetic factors, stratification analysis of urothelial cancer risk associated with E-cadherin genotypes was carried out for smoking status. All statistical tests were based on two-tailed probability. The x 2 test was used for categorical
55
comparisons of the data and the probability of the Hardy – Weinberg equilibrium.
3. Results Table 1presents the frequencies of E-cadherin alleles and genotypes by case-control status and the association of E-cadherin genotypes with urothelial cancer risk. A significant difference in allele frequencies between urothelial cancer patients and controls was not found ðP ¼ 0:16Þ. Compared to controls with the A/A genotype, the frequency of urothelial cancer patients with A/A genotype was higher ðP ¼ 0:08Þ, and adjusted OR for A/A genotype was statistically significant (OR ¼ 2.32, 95% CI 1.03– 5.22). Genotype distributions in controls were in Hardy –Weinberg equilibrium ðP ¼ 0:829Þ. In order to check the effect of the gene in combination with smoking, we calculated the OR for data that were classified by smoking status and cumulative cigarette dose (pack – years) and by gene genotypes. No significant increased risk of the any genotype in urothelial cancer patients in relation to tobacco smoke exposure was found (data not shown). To evaluate the potential modifying effect of Ecadherin genotypes with tumor differentiation and stage, stratified analyses were performed between urothelial cancer subgroups (Table 2). There were no statistically significant differences in genotype distribution among any of the subgroups when the combination of E-cadherin C/C genotype and G1 þ G2 or superficial patients was used as a referent group to calculate the OR. We analyzed the associations between E-cadherin polymorphism and recurrence rate of superficial tumor patients ðn ¼ 127Þ who had been treated with bladder reserving operation for the first time and been followed for more than 1 year after operations. However, no significant association between Ecadherin polymorphism and recurrence rate of superficial tumor patients was found (data not shown).
4. Discussion Diminished E-cadherin expression is associated with aggressive, poorly differentiated urothelial
56
H. Tsukino et al. / Cancer Letters 195 (2003) 53–58
Table 1 E-cadherin allele and genotype frequencies and adjusted ORs for urothelial cancer Cases (n ¼ 314)
Controls (n ¼ 314)
Statics
Allele frequencies (% )a C allele A allele
78.3 21.7
81.7 18.3
P ¼ 0.16
Genotype frequencies (% )b C/C C/A A/A
63.4 29.9 6.7
66.2c 30.9 2.9
P ¼ 0.08
Adjusted ORs and 95% CId C/C C/A A/A a b c d e
199 94 21
208 97 9
1.00 (reference) 1.01 (0.72–1.43) 2.32 (1.03–5.22)e
Number of alleles/number of chromosomes. Number of participants with genotype/total number of participants. The distribution of genotypes among control individuals was in Hardy–Weinberg equilbrium. ORs were adjusted for age, gender and smoking status. P ¼ 0.042.
cancer [4]. The loss of E-cadherin function is caused by mutation, loss of heterozygosity, trans-acting pathway, chromatin rearrangement and hypermethylation of promoter CpG islands of E-cadherin gene [11]. Recently, Li et al. [8] reported a C/A single nucleotide polymorphism in position 2 160 from the transcriptional start site of the E-cadherin gene promoter, with the A allele associated with a decreased transcriptional activity by 68% compared to the C allele. We found the E-cadherin A/A
genotype had a 2.3 fold increased risk of urothelial cancer compared to C/C genotype. However, stratifying by tumor differentiation and stage, we found no association between E-cadherin polymorphism and poorly-differentiation and invasiveness of urothelial cancer. Furthermore, no significant association between E-cadherin polymorphism and recurrence rate of urothelial cancer patients was found. These finding suggest that A/A genotype is a marker of genetic susceptibility rather than a prognostic marker
Table 2 Associations of E-cadherin genotypes with urothelial cancer, according to tumor differentiation and stage Risk (A/A þ C/A vs. C/C) C/C
C/A
A/A
Differentiation G1 þ G2 (n ¼ 174) G3 (n ¼ 140)
62.6%(109) 64.3%(90)
31.0%(54) 28.6%(40)
6.3%(11) 7.1%(10)
1b
Stage Superficial (n ¼ 182) Invasive (n ¼ 124)
63.7%(116) 62.1%(77)
29.7%(54) 31.5%(39)
6.6%(12) 6.5%(8)
1c
a
OR (95% CI)a
P value
0.94 (0.59–1.50)
P ¼ 0.808
1.11 (0.69–1.79)
P ¼ 0.671
ORs were adjusted for age, gender and smoking status. Combination of E-cadherinC/C genotype and G1 þ G2 patients was used as a referent group to calculate the OR and 95% CI of urothelial cancer patients subdivided according to tumor differentiation. c Combination of E-cadherinC/C genotype and superficial patients was used as a referent group to calculate the OR and 95% CI of urothelial cancer patients subdivided according to tumor stage. b
H. Tsukino et al. / Cancer Letters 195 (2003) 53–58
of urothelial cancer. E-cadherin has been observed to act as a tumor suppressor gene [12,13], and our results support this finding. There are several reports regarding E-cadherin polymorphism in cancer. Two case-control studies related to gastric cancer were reported [10,14]. One reported that stratification analyses of the gastric cancer cases according to their location, histology, tumor stage and lymph node metastasis failed to reveal any heterogeneity with respect to E-cadherin genotype. However, the frequency of variant A/A genotype in gastric cancer case was significantly lower than that of controls ðP , 0:005Þ, conferring a five-fold decrease in the risk of gastric cancer (OR ¼ 0.20, 95% CI, 0.06 – 0.56) compared with the C/C genotype [10]. This result was the opposite of that assuming that the A allele was associated with reduced transitional activity and in turn leads to susceptibility to cancer. The other reported that the genotype frequencies did not differ between 433 gastric cancer patients and 466 controls [14]. They found no evidence for difference in risk for the intestinal- and diffuse-type histopathologic subgroups. Furthermore, one individual study on colorectal or breast cancer showed a lack of association between E-cadherin polymorphism and risk of these diseases [15,16]. Verhage et al. [17] reported that C/A and A/A genotypes had an increased risk of prostatic cancer (OR ¼ 3.8, 95% CI 2.1 –6.8 and OR ¼ 1.7, 95% CI 0.4 – 6.6, respectively). Their data showed that the C/A SNP is a novel risk factor of substantial magnitude for prostatic cancer. These results of above studies were inconsistent. One of the reasons for these discrepancies may be insufficient study power. To detect a small relative risk of some disease need a large number of subjects. In this study, we had 80% power (two-sided test of significance, a ¼ 0:05) to detect an OR of 1.6 (C/A), 2.9 (A/A) and 1.6 (any A allele) relative to the C/C genotype, respectively. If subjects are subdivided into some groups, power will be reduced further, and larger studies will be needed. Another reason for these discrepancies may be different frequencies of alleles and genotypes by ethnicity. Previous studies reported that the frequencies of A allele were 24.7 – 31.0% in European populations [14,16,17], 30.1% in Canadian populations [14] and 33.7% in Taiwanese populations [10]. And the frequencies of A/A genotype were 4.8–
57
11.9% in European populations [14,16,17], 6.5% in Canadian populations [14] and 9.7% in Taiwanese populations [10]. In our healthy control groups, the frequencies of A allele and A/A genotype were found in 18.3 and 2.9%, respectively, and those were notably lower than the frequencies reported previously [10,14, 16,17]. The relatively high frequency of A allele and A/A genotype in some populations argues that the Ecadherin polymorphism does not greatly increase cancer risk. In conclusion, this study is the first to test the association between E-cadherin polymorphism and urothelial cancer. We observed that E-cadherin polymorphism was associated with the occurrence of urothelial cancer in Japanese populations. However, the present findings suggested no association of this polymorphism with disease progression.
Acknowledgements This work was supported in part by a Grant-in-Aid for Research on Environmental Health from the Ministry of Health and Welfare of Japan, and a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan.
References [1] M. Takeichi, Cadherin cell adhesion receptors as a morphogenetic regulator, Science 251 (1991) 1451–1455. [2] U.H. Frixen, J. Behrens, M. Sachs, G. Eberle, B. Voss, A. Warda, D. Lochner, W. Birchmeier, E-cadherin-mediated cell –cell adhesion prevents invasiveness of human carcinoma cells, J. Cell Biol. 113 (1991) 173 –185. [3] M. Fujisawa, J. Miyazaki, Y. Takechi, S. Arakawa, S. Kamidono, The significance of E-cadherin in transitionalcell carcinoma of the human urinary bladder, World J. Urol. 14 (1996) 12s–15s. [4] T. Okegawa, Y. Li, R.C. Pong, J.T. Hsieh, Cell adhesion proteins as tumor suppressors, J. Urol. 167 (2002) 1836–1843. [5] I. Taddei, M. Piazzini, R. Bartolett, M. Dal Canto, I. Sardi, Molecular alterations of E-cadherin gene: possible role in human bladder carcinogenesis, Int. J. Mol. Med. 6 (2000) 201 –208. [6] U.H. Frixen, J. Behrens, M. Sachs, G. Eberle, B. Voss, A. Warda, D. Lochner, W. Birchmeier, E-cadherin-mediated cell –cell adhesion prevents invasiveness of human carcinoma cells, J. Cell Biol. 113 (1991) 173 –185.
58
H. Tsukino et al. / Cancer Letters 195 (2003) 53–58
[7] M. Katayama, S. Hirai, K. Kamihagi, K. Nakagawa, M. Yasumoto, I. Kato, Soluble E-cadherin fragments increased in circulation of cancer patients, Br. J. Cancer 69 (1994) 580–585. [8] L.C. Li, R.M. Chui, M. Sasaki, K. Nakajima, G. Perinchery, H.C. Au, D. Nojima, P. Carroll, R. Dahiya, A single nucleotide polymorphism in the E-cadherin gene promoter alters transcriptional activities, Cancer Res. 60 (2000) 873–876. [9] Japanese Urological Association, The Japanese Society of Pathology, General Rules for Clinical and Pathological Studies on Bladder Cancer, 3rd ed., Kanehara Press, Tokyo, 2001. [10] M.S. Wu, S.P. Huang, Y.T. Chang, M.T. Lin, C.T. Shun, M.C. Chang, H.P. Wang, C.J. Chen, J.T. Lin, Association of the 2160 C – . a promoter polymorphism of E-cadherin gene with gastric carcinoma risk, Cancer 94 (2002) 1443–1448. [11] L.A. Ribeiro-Filho, J. Franks, M. Sasaki, H. Shiina, L.C. Li, D. Nojima, S. Arap, P. Carroll, H. Enokida, M. Nakagawa, S. Yonezawa, R. Dahiya, CpG hypermethylation of promoter region and inactivation of E-cadherin gene in human bladder cancer, Mol. Carcinogen. 34 (2002) 187–198. [12] J.C. Machado, C. Oliveira, R. Carvalho, P. Soares, G. Berx, C. Caldas, R. Seruca, F. Carneiro, M. Sobrinho-Simoes, Ecadherin gene (CDH1) promoter methylation as the second hit in sporadic diffuse gastric carcinoma, Oncogene 20 (2001) 1525–1528.
[13] G. Keller, H. Vogelsang, I. Becker, J. Hutter, K. Ott, S. Candidus, T. Grundei, K.F. Becker, J. Mueller, J.R. Siewert, H. Hofler, Diffuse type gastric and lobular breast carcinoma in a familial gastric cancer patient with an E-cadherin germline mutation, Am. J. Pathol. 155 (1999) 337–342. [14] P.D. Pharoah, C. Oliveira, J.C. Machado, G. Keller, H. Vogelsang, H. Laux, K.F. Becker, H. Hahn, S.M. Paproski, L.A. Brown, C. Caldas, D. Huntsman, CDH1 c-160a promotor polymorphism is not associated with risk of stomach cancer, Int. J. Cancer 101 (2002) 196 –197. [15] T.R. Porter, F.M. Richards, R.S. Houlston, D.G. Evans, J.A. Jankowski, F. Macdonald, G. Norbury, S.J. Payne, S.A. Fisher, I. Tomlinson, E.R. Maher, Contribution of cyclin d1 (CCND1) and E-cadherin (CDH1) polymorphisms to familial and sporadic colorectal cancer, Oncogene 21 (2002) 1928–1933. [16] H. Lei, S. Sjoberg-Margolin, S. Salahshor, B. Werelius, E. Jandakova, K. Hemminki, A. Lindblom, I. Vorechovsky, CDH1 mutations are present in both ductal and lobular breast cancer, but promoter allelic variants show no detectable breast cancer risk, Int. J. Cancer 98 (2002) 199–204. [17] B.A. Verhage, K. van Houwelingen, T.E. Ruijter, L.A. Kiemeney, J.A. Schalken, Single-nucleotide polymorphism in the E-cadherin gene promoter modifies the risk of prostate cancer, Int. J. Cancer 100 (2002) 683– 685.
E-cadherin gene polymorphism and risk of urothelial cancer Hiromasa Tsukinoa,b, Yoshiki Kurodaa, Hiroyuki Nakaoa, Hirohisa Imaia, Hisato Inatomic, Kiyotaka Kohshid, Yukio Osadab, Takahiko Katoha,* a
Department of Public Health, School of Medicine, Miyazaki Medical College, 5200 Kihara, Miyazaki, 889-1692, Japan b Department of Urology, School of Medicine, Miyazaki Medical College, 5200 Kihara, Miyazaki, 889-1692, Japan c Munakata Suikoukai General Hospital, Munakata, 811-3298, Japan d Department of Hyperbaric Medicine and Neurosurgery, School of Medicine, University of Occupational and Environmental Health (UOEH), 1-1 Iseigaoka, Yahatanishi-ku, Kitakyushu 807-8555, Japan Received 16 January 2003; accepted 20 January 2003
Abstract The E-cadherin is important in cell– cell adhesion and in the development and maintenance of the epithelial phenotype. A 2 160 C ! A polymorphism in the promoter region of E-cadherin has been shown to decrease gene transcription. This allelic variation may be a potential genetic marker that can help identify those individuals at higher risk for invasive/metastatic disease. We studied the effect of E-cadherin gene polymorphism on urothelial cancer susceptibility in a case control study of 314 urothelial cancer patients and 314 age– sex matched controls, to determine whether this polymorphism is a biomarker for the risk and how aggressive the disease is. The frequency with which the subjects carried E-cadherin A/A genotype was significantly higher in the urothelial cancer patients than in the healthy control subjects (OR ¼ 2.32, 95% CI 1.03– 5.22). Subdividing urothelial cancer according to tumor differentiation and stage, we found no association between E-cadherin polymorphism and poorly-differentiation and invasiveness of urothelial cancer. Furthermore, no significant association between E-cadherin polymorphism and recurrence rate of urothelial cancer patients was found. The present study demonstrates for the first time that E-cadherin A/A genotype may be associated with susceptibility to urothelial cancer, but not with the progression of disease. q 2003 Elsevier Science Ireland Ltd. All rights reserved. Keywords: E-cadherin; Polymorphism; Urothelial cancer; Promoter; SNP
1. Introduction Cadherins are transmembrane Ca2þ dependent Abbreviations: PCR, polymerase chain reaction; RFLP, restriction fragment length polymorphism; OR, odds ratio; 95% CI, 95% confidence interval. * Corresponding author. Tel.: þ81-985-85-0874; fax: þ 81-98585-6258. E-mail address: [email protected] (T. Katoh).
homophilic adhesion receptors that have important roles in cell recognition and cell sorting during development [1]. Cadherin are localized in specialized cell – cell adhesion sites termed adherence junctions. At these sites cadherins establish linkages with the actin containing cytoskeleton. The classic cadherins include E, N and P-cadherin. E-cadherin is expressed in all epithelial tissue and is found on the plasma membrane of squamous and transitional cells,
0304-3835/03/$ - see front matter q 2003 Elsevier Science Ireland Ltd. All rights reserved. doi:10.1016/S0304-3835(03)00130-7
54
H. Tsukino et al. / Cancer Letters 195 (2003) 53–58
in contrast to the concentrated expression at the intermediate junction (zonula adherens and belt desmosome) of polarized cells, such as intestinal or prostatic epithelium. Perturbation of E-cadherin mediated cell adhesion is involved in tumor progression and metastasis. In vitro loss of E-cadherin expression is associated with loss of cellular differentiation and increased cellular invasiveness and infiltration, while transfection of E-cadherin negative cell lines with E-cadherin complimentary DNA is able to suppress the invasive behavior [2]. Investigation of the expression of E-cadherin histopathological material of human transitional cell carcinoma demonstrates that aberrant expression of E-cadherin correlates with lack of differentiation, muscle invasion and distant metastasis [3]. Loss of normal E-cadherin expression has also been shown to correlate with decreased recurrence-free and overall survival, although multivariate analysis suggests that it has no independent prognostic value over the grade and stage of the tumor [4]. A number of possible mechanisms have been proposed to explain the documented reduction in E-cadherin function in bladder cells undergoing malignant transformation. They include suppression or mutation of E-cadherin gene [5], transformation disorder [6] or increased proteasemediated degradation [7]. Recently, a C/A single nucleotide polymorphism (SNP) at 2 160 from the transcriptional start site of the E-cadherin gene promoter has been identified by Li et al. [8]. In vitro studies showed that the A allele of this polymorphism decreased the transcriptional efficiency by 68% compared with the C allele. They suggested that 2 160 C/A polymorphism has a direct effect on E-cadherin gene transcriptional regulation, and this allelic variation may be a potential genetic marker that can help identify those individuals at higher risk for invasive/metastatic disease. However, to our knowledge, there are no reports investigating the association between E-cadherin gene polymorphism and urothelial cancer risk. In this study, we hypothesized that E-cadherin polymorphism is responsible for interindividual variation in the production of E-cadherin and in turn leads to individual susceptibility to invasiveness, differentiation and recurrence. We studied E-cadherin genotype in 314 urothelial cancer patients and 314 controls. Additionally, we assessed E-cadherin geno-
type in relation to clinical findings associated with the outcome. These included differentiation, stage and recurrence. We have demonstrated a significant association between E-cadherin genotype and urothelial cancer susceptibility.
2. Materials and methods 2.1. Subjects The case groups comprised 314 patients with urothelial cancer (bladder n ¼ 240, renal pelvis and ureter n ¼ 37, overlap n ¼ 37) (244 men, 70 women; mean age 69.6 years) in Kitakyushu city and Miyazaki city, Japan. This study was designed to recruit all eligible urothelial cancer cases who were newly diagnosed with urothelial cancer during 1992 –2002 in Kitakyushu city and 2000 –2002 in Miyazaki city. The patients were treated at the University of Occupational and Environmental Health Hospital and Miyazaki Medical College Hospital, and had been histologically diagnosed for urothelial transitional cell carcinoma. Histological grading and staging was done according to the Japanese classification system [9]. Thirty-two patients (10.2%) had a well-differentiated disease (G1), 142 (45.2%) had a moderately differentiated (G2) and 140 (44.6%) had a poorly differentiated disease (G3). We used the term ‘superficial tumors’ to refer to those which were limited to the mucosa (pTis, pTa) or the lamina propria (pT1); and ‘invasive tumors’ to refer to those which had invaded the muscle layer (pT2, pT3) or deeper (pT4). One hundred and eighty-two patients (58.0%) had superficial tumors, 124 (39.5%) had invasive tumors and eight (2.5%) had unknown stage tumors. A total of 314 controls, frequency-matched with cases for age (^ 5 years) and gender, were selected from the people who visited a medical institution located in Kitakyushu city for a general health check-up (244 men, 70 women, mean age 67.2 years). All study subjects completed a questionnaire administered by a trained interviewer covering medical, residential, and occupational exposures as well as smoking history. Smoking history was summarized as the total amount of cigarettes consumed during their lifetime up until the time of the interview. The amount of tobacco smoke exposure
H. Tsukino et al. / Cancer Letters 195 (2003) 53–58
was calculated as pack –years [one pack (20 cigarettes)/day £ years of smoking]. All participants were given an explanation of the nature of the study, and informed consent was obtained. This study was approved by the ethics committees of the University of Occupational and Environmental Health and Miyazaki Medical College. 2.2. Genotyping Genomic DNA was isolated from peripheral leukocytes by proteinase K digestion and phenol/chloroform extraction. The genetic polymorphism of E-cadherin was determined by polymerase chain reaction (PCR) amplification followed by digestion with Afl III and Hph I, using the method described previously [10]. Briefly, the two primers used in the PCR were 50 -TCCCAGGTCTTAGTGAGCCA-30 and 50 -GGCCACAGCCAATCAGCA-30 , yielding a single 190-base pair (bp) product. One hundred nanograms of DNA were amplified in a total volume of 30 ml containing 10 pmole of each primer, 2 U of Taq polymerase, 1.5 mM MgCl2, and PCR buffer. Amplification was performed by denaturing at 968C for 30 s, annealing at 608C for 30 s and extending at 728C for 30 s for 35 cycles using a Perkin – Elmer 9700. After confirmation of the amplified fragment of the expected size on 3% agarose gel, the PCR products were digested with 5 U of restriction enzyme HphI or AflIII at 378C for 16 h, and electrophoresed on a 3% agarose gel. The C allele created an HphI site, and the A allele created an AflIII site. All study subjects were analyzed using both restriction enzymes. 2.3. Statistical analysis Crude odds ratios (ORs) and 95% confidence intervals (95% CIs) were calculated for E-cadherin genotypes. ORs were adjusted for age, gender and smoking status (ever- and never- smokers), using multiple logistic regression analysis by SPSS Medical Pack for Windows. To examine the interaction between environmental and genetic factors, stratification analysis of urothelial cancer risk associated with E-cadherin genotypes was carried out for smoking status. All statistical tests were based on two-tailed probability. The x 2 test was used for categorical
55
comparisons of the data and the probability of the Hardy – Weinberg equilibrium.
3. Results Table 1presents the frequencies of E-cadherin alleles and genotypes by case-control status and the association of E-cadherin genotypes with urothelial cancer risk. A significant difference in allele frequencies between urothelial cancer patients and controls was not found ðP ¼ 0:16Þ. Compared to controls with the A/A genotype, the frequency of urothelial cancer patients with A/A genotype was higher ðP ¼ 0:08Þ, and adjusted OR for A/A genotype was statistically significant (OR ¼ 2.32, 95% CI 1.03– 5.22). Genotype distributions in controls were in Hardy –Weinberg equilibrium ðP ¼ 0:829Þ. In order to check the effect of the gene in combination with smoking, we calculated the OR for data that were classified by smoking status and cumulative cigarette dose (pack – years) and by gene genotypes. No significant increased risk of the any genotype in urothelial cancer patients in relation to tobacco smoke exposure was found (data not shown). To evaluate the potential modifying effect of Ecadherin genotypes with tumor differentiation and stage, stratified analyses were performed between urothelial cancer subgroups (Table 2). There were no statistically significant differences in genotype distribution among any of the subgroups when the combination of E-cadherin C/C genotype and G1 þ G2 or superficial patients was used as a referent group to calculate the OR. We analyzed the associations between E-cadherin polymorphism and recurrence rate of superficial tumor patients ðn ¼ 127Þ who had been treated with bladder reserving operation for the first time and been followed for more than 1 year after operations. However, no significant association between Ecadherin polymorphism and recurrence rate of superficial tumor patients was found (data not shown).
4. Discussion Diminished E-cadherin expression is associated with aggressive, poorly differentiated urothelial
56
H. Tsukino et al. / Cancer Letters 195 (2003) 53–58
Table 1 E-cadherin allele and genotype frequencies and adjusted ORs for urothelial cancer Cases (n ¼ 314)
Controls (n ¼ 314)
Statics
Allele frequencies (% )a C allele A allele
78.3 21.7
81.7 18.3
P ¼ 0.16
Genotype frequencies (% )b C/C C/A A/A
63.4 29.9 6.7
66.2c 30.9 2.9
P ¼ 0.08
Adjusted ORs and 95% CId C/C C/A A/A a b c d e
199 94 21
208 97 9
1.00 (reference) 1.01 (0.72–1.43) 2.32 (1.03–5.22)e
Number of alleles/number of chromosomes. Number of participants with genotype/total number of participants. The distribution of genotypes among control individuals was in Hardy–Weinberg equilbrium. ORs were adjusted for age, gender and smoking status. P ¼ 0.042.
cancer [4]. The loss of E-cadherin function is caused by mutation, loss of heterozygosity, trans-acting pathway, chromatin rearrangement and hypermethylation of promoter CpG islands of E-cadherin gene [11]. Recently, Li et al. [8] reported a C/A single nucleotide polymorphism in position 2 160 from the transcriptional start site of the E-cadherin gene promoter, with the A allele associated with a decreased transcriptional activity by 68% compared to the C allele. We found the E-cadherin A/A
genotype had a 2.3 fold increased risk of urothelial cancer compared to C/C genotype. However, stratifying by tumor differentiation and stage, we found no association between E-cadherin polymorphism and poorly-differentiation and invasiveness of urothelial cancer. Furthermore, no significant association between E-cadherin polymorphism and recurrence rate of urothelial cancer patients was found. These finding suggest that A/A genotype is a marker of genetic susceptibility rather than a prognostic marker
Table 2 Associations of E-cadherin genotypes with urothelial cancer, according to tumor differentiation and stage Risk (A/A þ C/A vs. C/C) C/C
C/A
A/A
Differentiation G1 þ G2 (n ¼ 174) G3 (n ¼ 140)
62.6%(109) 64.3%(90)
31.0%(54) 28.6%(40)
6.3%(11) 7.1%(10)
1b
Stage Superficial (n ¼ 182) Invasive (n ¼ 124)
63.7%(116) 62.1%(77)
29.7%(54) 31.5%(39)
6.6%(12) 6.5%(8)
1c
a
OR (95% CI)a
P value
0.94 (0.59–1.50)
P ¼ 0.808
1.11 (0.69–1.79)
P ¼ 0.671
ORs were adjusted for age, gender and smoking status. Combination of E-cadherinC/C genotype and G1 þ G2 patients was used as a referent group to calculate the OR and 95% CI of urothelial cancer patients subdivided according to tumor differentiation. c Combination of E-cadherinC/C genotype and superficial patients was used as a referent group to calculate the OR and 95% CI of urothelial cancer patients subdivided according to tumor stage. b
H. Tsukino et al. / Cancer Letters 195 (2003) 53–58
of urothelial cancer. E-cadherin has been observed to act as a tumor suppressor gene [12,13], and our results support this finding. There are several reports regarding E-cadherin polymorphism in cancer. Two case-control studies related to gastric cancer were reported [10,14]. One reported that stratification analyses of the gastric cancer cases according to their location, histology, tumor stage and lymph node metastasis failed to reveal any heterogeneity with respect to E-cadherin genotype. However, the frequency of variant A/A genotype in gastric cancer case was significantly lower than that of controls ðP , 0:005Þ, conferring a five-fold decrease in the risk of gastric cancer (OR ¼ 0.20, 95% CI, 0.06 – 0.56) compared with the C/C genotype [10]. This result was the opposite of that assuming that the A allele was associated with reduced transitional activity and in turn leads to susceptibility to cancer. The other reported that the genotype frequencies did not differ between 433 gastric cancer patients and 466 controls [14]. They found no evidence for difference in risk for the intestinal- and diffuse-type histopathologic subgroups. Furthermore, one individual study on colorectal or breast cancer showed a lack of association between E-cadherin polymorphism and risk of these diseases [15,16]. Verhage et al. [17] reported that C/A and A/A genotypes had an increased risk of prostatic cancer (OR ¼ 3.8, 95% CI 2.1 –6.8 and OR ¼ 1.7, 95% CI 0.4 – 6.6, respectively). Their data showed that the C/A SNP is a novel risk factor of substantial magnitude for prostatic cancer. These results of above studies were inconsistent. One of the reasons for these discrepancies may be insufficient study power. To detect a small relative risk of some disease need a large number of subjects. In this study, we had 80% power (two-sided test of significance, a ¼ 0:05) to detect an OR of 1.6 (C/A), 2.9 (A/A) and 1.6 (any A allele) relative to the C/C genotype, respectively. If subjects are subdivided into some groups, power will be reduced further, and larger studies will be needed. Another reason for these discrepancies may be different frequencies of alleles and genotypes by ethnicity. Previous studies reported that the frequencies of A allele were 24.7 – 31.0% in European populations [14,16,17], 30.1% in Canadian populations [14] and 33.7% in Taiwanese populations [10]. And the frequencies of A/A genotype were 4.8–
57
11.9% in European populations [14,16,17], 6.5% in Canadian populations [14] and 9.7% in Taiwanese populations [10]. In our healthy control groups, the frequencies of A allele and A/A genotype were found in 18.3 and 2.9%, respectively, and those were notably lower than the frequencies reported previously [10,14, 16,17]. The relatively high frequency of A allele and A/A genotype in some populations argues that the Ecadherin polymorphism does not greatly increase cancer risk. In conclusion, this study is the first to test the association between E-cadherin polymorphism and urothelial cancer. We observed that E-cadherin polymorphism was associated with the occurrence of urothelial cancer in Japanese populations. However, the present findings suggested no association of this polymorphism with disease progression.
Acknowledgements This work was supported in part by a Grant-in-Aid for Research on Environmental Health from the Ministry of Health and Welfare of Japan, and a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan.
References [1] M. Takeichi, Cadherin cell adhesion receptors as a morphogenetic regulator, Science 251 (1991) 1451–1455. [2] U.H. Frixen, J. Behrens, M. Sachs, G. Eberle, B. Voss, A. Warda, D. Lochner, W. Birchmeier, E-cadherin-mediated cell –cell adhesion prevents invasiveness of human carcinoma cells, J. Cell Biol. 113 (1991) 173 –185. [3] M. Fujisawa, J. Miyazaki, Y. Takechi, S. Arakawa, S. Kamidono, The significance of E-cadherin in transitionalcell carcinoma of the human urinary bladder, World J. Urol. 14 (1996) 12s–15s. [4] T. Okegawa, Y. Li, R.C. Pong, J.T. Hsieh, Cell adhesion proteins as tumor suppressors, J. Urol. 167 (2002) 1836–1843. [5] I. Taddei, M. Piazzini, R. Bartolett, M. Dal Canto, I. Sardi, Molecular alterations of E-cadherin gene: possible role in human bladder carcinogenesis, Int. J. Mol. Med. 6 (2000) 201 –208. [6] U.H. Frixen, J. Behrens, M. Sachs, G. Eberle, B. Voss, A. Warda, D. Lochner, W. Birchmeier, E-cadherin-mediated cell –cell adhesion prevents invasiveness of human carcinoma cells, J. Cell Biol. 113 (1991) 173 –185.
58
H. Tsukino et al. / Cancer Letters 195 (2003) 53–58
[7] M. Katayama, S. Hirai, K. Kamihagi, K. Nakagawa, M. Yasumoto, I. Kato, Soluble E-cadherin fragments increased in circulation of cancer patients, Br. J. Cancer 69 (1994) 580–585. [8] L.C. Li, R.M. Chui, M. Sasaki, K. Nakajima, G. Perinchery, H.C. Au, D. Nojima, P. Carroll, R. Dahiya, A single nucleotide polymorphism in the E-cadherin gene promoter alters transcriptional activities, Cancer Res. 60 (2000) 873–876. [9] Japanese Urological Association, The Japanese Society of Pathology, General Rules for Clinical and Pathological Studies on Bladder Cancer, 3rd ed., Kanehara Press, Tokyo, 2001. [10] M.S. Wu, S.P. Huang, Y.T. Chang, M.T. Lin, C.T. Shun, M.C. Chang, H.P. Wang, C.J. Chen, J.T. Lin, Association of the 2160 C – . a promoter polymorphism of E-cadherin gene with gastric carcinoma risk, Cancer 94 (2002) 1443–1448. [11] L.A. Ribeiro-Filho, J. Franks, M. Sasaki, H. Shiina, L.C. Li, D. Nojima, S. Arap, P. Carroll, H. Enokida, M. Nakagawa, S. Yonezawa, R. Dahiya, CpG hypermethylation of promoter region and inactivation of E-cadherin gene in human bladder cancer, Mol. Carcinogen. 34 (2002) 187–198. [12] J.C. Machado, C. Oliveira, R. Carvalho, P. Soares, G. Berx, C. Caldas, R. Seruca, F. Carneiro, M. Sobrinho-Simoes, Ecadherin gene (CDH1) promoter methylation as the second hit in sporadic diffuse gastric carcinoma, Oncogene 20 (2001) 1525–1528.
[13] G. Keller, H. Vogelsang, I. Becker, J. Hutter, K. Ott, S. Candidus, T. Grundei, K.F. Becker, J. Mueller, J.R. Siewert, H. Hofler, Diffuse type gastric and lobular breast carcinoma in a familial gastric cancer patient with an E-cadherin germline mutation, Am. J. Pathol. 155 (1999) 337–342. [14] P.D. Pharoah, C. Oliveira, J.C. Machado, G. Keller, H. Vogelsang, H. Laux, K.F. Becker, H. Hahn, S.M. Paproski, L.A. Brown, C. Caldas, D. Huntsman, CDH1 c-160a promotor polymorphism is not associated with risk of stomach cancer, Int. J. Cancer 101 (2002) 196 –197. [15] T.R. Porter, F.M. Richards, R.S. Houlston, D.G. Evans, J.A. Jankowski, F. Macdonald, G. Norbury, S.J. Payne, S.A. Fisher, I. Tomlinson, E.R. Maher, Contribution of cyclin d1 (CCND1) and E-cadherin (CDH1) polymorphisms to familial and sporadic colorectal cancer, Oncogene 21 (2002) 1928–1933. [16] H. Lei, S. Sjoberg-Margolin, S. Salahshor, B. Werelius, E. Jandakova, K. Hemminki, A. Lindblom, I. Vorechovsky, CDH1 mutations are present in both ductal and lobular breast cancer, but promoter allelic variants show no detectable breast cancer risk, Int. J. Cancer 98 (2002) 199–204. [17] B.A. Verhage, K. van Houwelingen, T.E. Ruijter, L.A. Kiemeney, J.A. Schalken, Single-nucleotide polymorphism in the E-cadherin gene promoter modifies the risk of prostate cancer, Int. J. Cancer 100 (2002) 683– 685.