Cancer Cell Biology
EBAG9 is a tumor-promoting and prognostic factor for bladder cancer
Version of Record online: 9 SEP 2008
Copyright © 2008 Wiley-Liss, Inc.
International Journal of Cancer
Volume 124, Issue 4, pages 799–805, 15 February 2009
How to Cite
Kumagai, J., Urano, T., Ogushi, T., Takahashi, S., Horie-Inoue, K., Fujimura, T., Azuma, K., Muramatsu, M., Ouchi, Y., Kitamura, T. and Inoue, S. (2009), EBAG9 is a tumor-promoting and prognostic factor for bladder cancer. Int. J. Cancer, 124: 799–805. doi: 10.1002/ijc.23982
- Issue online: 11 DEC 2008
- Version of Record online: 9 SEP 2008
- Accepted manuscript online: 9 SEP 2008 12:00AM EST
- Manuscript Accepted: 11 AUG 2008
- Manuscript Received: 10 JUN 2008
- Genome Network Project and the DECODE from the Ministry of Education, Culture, Sports, Science and Technology
- Japan Society for the Promotion of Science
- Ministry of Health, Labor and Welfare
- nude mouse;
- bladder cancer;
Upregulation of EBAG9 expression has been observed in several malignant tumors such as advanced breast and prostate cancers, indicating that EBAG9 may contribute to tumor proliferation. In the present study, we assess the role of EBAG9 in bladder cancer. We generated human bladder cancer EJ cells stably expressing FLAG-tagged EBAG9 (EJ-EBAG9) or empty vector (EJ-vector), and investigated whether EBAG9 overexpression modulates cell growth and migration in vitro as well as the in vivo tumor formation of EJ transfectants in xenograft models of BALB/c nude mice. EBAG9 overexpression promoted EJ cell migration, while the effect of EBAG9 to cultured cell growth was rather minimal. Tumorigenic experiments in nude mice showed that the size of EJ-EBAG9-derived tumors was significantly larger than EJ-vector-derived tumors. Loss-of-function study for EBAG9 using small interfering RNA (siRNA) in xenografts with parental EJ cells showed that the intra-tumoral injection of EBAG9 siRNA markedly reduced the EJ tumor formation compared with control siRNA. Furthermore, immunohistochemical study for EBAG9 expression was performed in 60 pathological bladder cancer specimens. Intense and diffuse cytoplasmic immunostaining was observed in 45% of the bladder cancer cases. Positive EBAG9 immunoreactivity was closely correlated with poor prognosis of the patients (p = 0.0001) and it was an independent prognostic predictor for disease-specific survival in multivariate analysis (p = 0.003). Our results indicate that EBAG9 would be a crucial regulator of tumor progression and a potential prognostic marker for bladder cancer. © 2008 Wiley-Liss, Inc.
Bladder cancer is the fourth most common type of cancer in men and the ninth most common in women.1 A radical cystectomy remains the most common treatment for patients with muscle-invasive bladder cancer. Despite advances in surgical techniques, the 5-year disease survival rate after radical cystectomy remains at 50–60%.2 Moreover, current clinical and pathological variables have a limited ability to predict tumor recurrence, progression, or patient survival; thus, prognostic information regarding transitional cell carcinoma of the urinary bladder is needed. Biomarkers may be helpful for selecting patients who are best suited for adjuvant therapy.
EBAG9 (Estrogen receptor-binding fragment-associated antigen 9) is a primary estrogen-responsive gene that has been originally cloned by our group from human breast cancer MCF-7 cells using the CpG-genomic binding site cloning method.3 EBAG9 protein is predominantly expressed in estrogen-target organs as well as several other tissues, such as brain, liver, and kidney.4 The expression of EBAG9 protein is inducible by estrogen in the uterus, as previously shown in ovariectomized mice treated with 17β-estradiol.4 The physiologic function of EBAG9 has not been well defined; nevertheless, the molecule has been implicated in cancer pathophysiology, with several lines of evidence showing protein expression in malignancies like breast,5 ovarian,6 prostate,7 and hepatocellular carcinomas8 as well as renal cell carcinoma.9 In the previous report, positive EBAG9 immunoreactivity was closely correlated with a poor prognosis in renal cell carcinoma patients (p = 0.0007), and murine renal cell carcinoma Renca cells harboring EBAG9 developed larger tumors than Renca cells transfected with a vector alone, suggesting that an abundance of EBAG9 might influence the progression of malignant tumors.9
In the present study, we investigated whether EBAG9 expression is critical to the tumor development of bladder cancer. Using EJ cells, a model for human bladder transitional cell carcinoma, we investigated whether the overexpression of EBAG9 affected EJ cell and tumor growth. We also explored the effect of EBAG9 on EJ cell migration. Moreover, the expression of EBAG9 in human bladder tissues was studied using immunohistochemistry.
Material and methods
Rabbit anti-EBAG9 polyclonal antibody was generated against a fusion protein of glutathione S-transferase and EBAG9.10 Anti-FLAG M2 antibody and anti-β-actin antibody were obtained from Sigma (St. Louis, MO). Amino-terminal FLAG-tagged human EBAG9 cDNA (FLAG-EBAG9) was cloned into a mammalian expression vector pcDNA3 (Invitrogen, Carlsbad, CA).
A human bladder cancer cell line, EJ cells, was maintained in RPMI 1640 media supplemented with 2 mM glutamine, 1% nonessential amino acids, 100 U/mL streptomycin/penicillin, 10% fetal calf serum (FCS) and antibiotics.
Generation of EJ cells stably expressing FLAG-EBAG9
EJ cells were transfected with an expression vector, pcDNA3, containing human FLAG-EBAG9 cDNA or the vector alone using Lipofectamine 2000 (Invitrogen). G418-resistant cells were selected, and several independent clones were isolated. To validate the expression of exogenous human FLAG-EBAG9, Western blotting was done using an anti-FLAG antibody and an anti-EBAG9 antibody.
Western blot analysis
Cells were lysed in 200 mL of Nonidet P-40 lysis buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 10 mM NaF, 5 mM EDTA, 5 mM EGTA, 2 mM sodium vanadate, 0.5% sodium deoxycholate, 1 mM dithiothreitol [DTT], 1 mM phenylmethylsulfonylfluoride [PMSF], 2 mg/mL aprotinin and 0.1% Nonidet P-40). Proteins were resolved by sodium dodecyl sulfate (SDS)-12.5% polyacrylamide gels and electrophoretically transferred onto polyvinylidene difluoride (PVDF) membranes (Immobilon, Bedford, MA). Membranes were probed with rabbit anti-EBAG9 antibody, anti-FLAG antibody or anti-β-actin monoclonal antibody.
Cell proliferation assay
Cells were seeded on day 0 at a density of 1 to 6 × 104 cells per dish into 10-cm dishes, and the cell numbers were counted using the Coulter counter Z1 (Coulter Japan, Tokyo, Japan) on days 1, 2 and 3.
BALB/c nu/nu mice (Nisseizai, Tokyo, Japan) were kept under specific pathogen-free conditions and fed dry food and water. All mice used for the experiments were 5-week-old males. All animal experiments were performed with the approval of the Animal Study Committee of the University of Tokyo, and conformed to relevant guidelines and laws.
In vivo tumor challenge
For subcutaneous implantation, transfected EJ cells (3 × 105 cells per mouse) suspended in 0.1 mL of complete medium were injected with 0.1 mL of Matrigel® (Becton Dickinson Labware, Bedford, MA) into the flanks of BALB/c nude mice. Tumor size was calculated with a micrometer in 2 dimensions, and the tumor volume was estimated according to the following formula: tumor volume (mm3) = 1/2 × (long diameter) × (short diameter).2
The cell migration assay was performed using a Cell Culture Insert with an 8.0-μm pore size PET filter (Becton Dickinson). Before the assay, the lower surface of the filter was immersed for 30 min in 10 μg/mL of fibronectin (Sigma) diluted with PBS. Next, 700 μL of RPMI 1640 medium with 10% FCS was added to the lower chamber. Then, 5 × 104 cells were suspended in 300 μL of RPMI 1640 medium with 10% FCS and added to the upper chamber. After incubation for 4 hrs at 37°C in a humid 5% CO2 atmosphere, the cells on the upper surface of the filter were completely removed by wiping with cotton swabs. The cells on the lower surface of the filter were fixed in methanol for 30 min, washed with PBS, and then stained with Giemsa's stain solution (Muto Pure Chemicals, Tokyo, Japan) for 30 sec. After washing 3 times with PBS, the filters were mounted on a glass slide. The cells on the lower surface were counted in at least 5 fields at a magnification of ×200 under the microscope. A Student t-test was used to analyze the data from these experiments.
Tumor regression by siRNA targeting EBAG9
siRNA duplex targeting EBAG9 was generated by Proligo LLC (Boulder, CO). The target sequence of the EBAG9 siRNA was 5′-GCACAACGGCTAATGAAGAAG-3′. A nontargeting control siRNA that was not homologous with any known gene targets in mammalian cells was also made. The target sequence of the control siRNA was 5′-GTACCGCACGTCATTCGTATC-3′. Transfection of the constructs into EJ cells was performed using Lipofectamine 2000 according to the manufacturer's instructions. To investigate the in vivo silencing effect of EBAG9 siRNA in EJ tumors, the intra-tumoral injection of siRNA duplexes was performed twice a week. Briefly, EJ cells (5 × 106 cells) were implanted in the flank of BALB/c nude mice. When the volumes of the tumors reached 200 mm3, siRNA duplexes (10 μg) were injected directly into the tumors twice a week, along with 2 μL of Lipofectamine 2000 dissolved in 0.1 mL of RPMI 1640 medium. The mice were sacrificed after 4 weeks of treatment. Tumor size was measured weekly.
Patients and tissue preparation
Sixty patients with bladder cancer who underwent a radical cystectomy at Tokyo University Hospital between 1982 and 2000 were included in this study. All patients were consented for use of the tissues at the time of cystectomy. The mean patient age was 61 years (range, 39–80 years), and the patients comprised 53 men and 7 women. The mean follow-up time was 40 months (range, 1–136 months). The staging and grading of the tumors were done according to the World Health Organization classification11 and the TNM classification 2002 International Union Against Cancer/UICC, respectively.
Immunohistochemical studies were done using the streptavidin-biotin amplification method with horseradish peroxidase detection. Paraffin sections of the tumors were blocked in 0.3% H2O2 (30 min) and in 10% FCS (30 min), then incubated overnight with purified rabbit anti-EBAG9 antibody for human bladder cancer (1:100 dilution). Sections were incubated with biotinylated rabbit immunoglobulin G or anti-rabbit EnVison+ reagent (DakoCytomation Japan, Kyoto, Japan), and developed using diaminobenzidine (Sigma). Negative controls were done for each slide, using nonimmune immunoglobulin G. The immunoreactivity scores for EBAG9 expression were determined by two investigators (JK and TF) according to the percentage of positive cells. Human breast cancer sections (DakoCytomation) were used as a positive control. The positivity was 0–4% for an immunoreactivity score of 0 (negative), 5–24% for a score of 1+, 25–49% for a score of 2+, and 50–100% for a score of 3+. Immunoreactivity scores of 0 or 1+ were defined as negative, and scores of 2+ or 3+ were defined as positive. If the immunoreactivity scores assigned by the two independent investigators differed, a third investigator (ST) evaluated the samples and the most frequent immunoreactivity score was adopted.
Comparisons between the different groups of EJ cells were analyzed using a nonparametrical Mann-Whitney U test. The associations between EBAG9 immunoreactivity and clinicopathologic characteristics were evaluated using a Student t-test or a Fisher exact probability test. Disease-specific survival was computed using the Kaplan-Meier method and compared using a log-rank test. A multivariate analysis of prognostic factors was performed using the Cox proportional hazard regression model. Computations were done using StatView 5.0J software (SAS Institute, Cary, NC). All p values are two sided and were regarded as significant if p < 0.05.
Generation of EJ cells stably expressing EBAG9
We selected two EJ cell clones stably expressing FLAG-EBAG9 (EJ-EBAG9 no. 19 and EJ-EBAG9 no. 28), as confirmed by Western blotting using an anti-FLAG antibody and an anti-EBAG9 antibody (Fig. 1a, top and middle). The amounts of EBAG9 proteins in EJ-EBAG9 (EJ-EBAG9 no. 19 and EJ-EBAG9 no. 28) were clearly higher than those in the EJ cells transfected with the vector alone (EJ-vector no. 1 and EJ-vector no. 5). The cell growth rate of EJ-EBAG9 was lower than that of EJ-vector on Day 2 (p < 0.05) and on Day 3 (p < 0.001) (Fig. 1b). These results suggest that the stable expression of EBAG9 had a negative or rather minimal effect on the proliferation of cultured EJ cells.
EBAG9 promotes in vivo tumor growth of EJ cells
Contrary to the in vitro results for cell growth, EJ-EBAG9 cells implanted into BALB/c mice developed into > 4-fold larger tumors, compared with EJ-vector cells implanted into BALB/c mice, at 5 weeks after inoculation (Figs. 1c and 1d). The mean tumor volumes at 5 weeks were 624.8 ± 101.1 mm3 and 456.6 ± 94.7 mm3 for EJ-EBAG9 no. 19 and EJ-EBAG9 no. 28, respectively. However, those for EJ-vector no. 1 and EJ-vector no. 5 were 113.2 ± 43.8 mm3 and 153.8 ± 85.7 mm3, respectively (p = 0.006; Fig. 1d).
Over-expression of EBAG9 promotes cell migration
Cell migration was significantly enhanced in EJ-EBAG9 cells, compared with EJ-vector cells (p < 0.0001, Figs. 2a and 2b). The number of EJ-EBAG9 no. 19 and EJ-EBAG9 no. 28 cells that migrated was about 5.5 times higher than that for EJ-vector no. 1 and EJ-vector no. 5.
Gene silencing of EBAG9 suppressed in vivo tumor growth of EJ cells
si-EBAG9 reduced the protein level of endogenous EBAG9, compared with the levels of EBAG9 in parental EJ cells or the control siRNA group (Fig. 3a). We investigated the effect of si-EBAG9 utilizing in vivo mouse xenograft model. When treated with control siRNA, the implanted EJ cells developed prominent tumors; however, the injection of EBAG9 siRNA suppressed the tumor growth of EJ cells (Figs. 3b and 3c). After 4-weeks of treatment, the tumor volume in the EBAG9 siRNA treatment group was significantly smaller than that in the control siRNA group (1017.4 ± 369.3 mm3vs. 1939.0 ± 350.6 mm3, p < 0.05).
Expression of EBAG9 protein in human bladder cancer
In noncarcinomatous lesions, a weak and scattered immunostaining of EBAG9 was observed in the cytoplasm of the bladder epithelial cells (Fig. 4a). The levels of EBAG9 immunostaining in normal bladder tissues corresponded to an immunoreactivity score of 0. In bladder cancer, 33 of 60 cases (55%) had negative immunoreactivity for EBAG9 (Fig. 4b), whereas 27 cases (45%) showed EBAG9 positivity (Fig. 4c). With regard to the EBAG9-positive bladder cancers, the cancer cells generally retained an intense and diffuse staining pattern in the cytoplasm or on the membrane. Eighteen bladder cancers (30%) had an immunoreactivity score of 2+, while 9 cancers (15%) had an immunoreactivity score of 3+. Lymph node-metastatic cancers exhibited the strongest EBAG9 staining, predominantly in the cytoplasm (Fig. 4d; immunoreactivity score, 3+). A significant association between EBAG9 immunoreactivity and clinicopathologic variables was observed among the bladder cancer patients (Table I). EBAG9 positivity (immunoreactivity score, 2+ or 3+) was significantly correlated with a positive lymph node metastasis status and lymphatic vessel infiltration (p = 0.0008, p = 0.0017, respectively). In a Kaplan-Meier analysis of the bladder cancer patients, those with positive EBAG9 immunoreactivity showed a shorter disease-specific survival period (Fig. 4e), compared with those with negative EBAG9 immunoreactivity (immunoreactivity score, 0 or 1+) (p = 0.0001, log rank test). The 5-year disease-specific survival rate was 40.5% for patients with positive EBAG9 immunoreactivity and 79.8% with those with negative EBAG9 immunoreactivity. In a univariate Cox proportional hazards analysis of cancer-specific survival, established prognostic factors including pathologic stage, lymph node metastasis and immunoreactivity were significant univariate variables of survival (Table II; p = 0.018, and 0.005, respectively). Negative EBAG9 immunoreactivity was also a significant survival predictor in the univariate analysis (p < 0.001). In a multivariate Cox proportional hazards analysis, positive EBAG9 immunoreactivity (immunoreactivity score, 2+ or 3+) was the only factor associated with disease-specific death (p = 0.003; relative risk, 4.11) (Table II). These results indicate that positive EBAG9 immunoreactivity is a potential predictor of a poor prognosis in bladder cancer patients.
|No. patients||Immunoreactivity of EBAG91||p|
|Negative (n = 33)||Positive (n = 27)|
|Age||60||60.72 ± 10.0||62.3 ± 9.9||0.54|
|pT Ta, T1, T2||30||20||10||0.069|
|Lymph node status|
|Hazard ratio||95% CI||p||Hazard ratio||95% CI||p|
|Pathological tumor grade (G3 vs. G2)||1.35||0.58–3.11||0.483||1.28||0.54–3.16||0.57|
|Pathological tumor stage (pT3-4 vs. pTa-2)||2.82||1.20–6.62||0.018||1.13||0.41–3.16||0.81|
|Lymph node metastasis (positive vs. negative)||3.59||1.48–8.73||0.005||2.69||0.94–7.67||0.06|
|EBAG9 (negative vs. positive1)||5.09||2.03–12.7||<0.001||4.11||1.69–10.5||0.003|
The present study suggested that EBAG9 could play a tumor-promoting role in bladder cancer. We showed that the tumor volumes derived from EJ-EBAG9 cells were larger than those derived from EJ-vector cells implanted in BALB/c nude mice. We explored the mechanism by which EBAG9 promoted tumor growth and found that the migration ability of EJ-EBAG9 cells was augmented. The intra-tumoral injection of siRNA against EBAG9 clearly prevented tumor growth following subcutaneous inoculation with EJ cells. EBAG9 immunoreactivity was also detected in almost half of the human bladder cancer samples, and a positive EBAG9 immunoreactive status was associated with a poor patient outcome.
EBAG9 gene has a functional estrogen-responsive-element in the 5′-promoter region (−86 to −36).10 The mRNA levels of EBAG9 in MCF-7 cells are significantly increased within 6 hrs of estrogen treatment.3 Estrogen may not be a critical factor for bladder cancer. It is likely that other mechanisms independent of estrogen signaling pathways could be involved in the regulation of EBAG9 expression, as the EBAG9 is expressed in many tissues and tumors.4, 10
It is notable that the 32-kDa EBAG9 protein predominantly expressed in the cytosol is distinct from the 78-kDa tumor-associated cell surface antigen recognized by 22-1-1 monoclonal IgM antibody,12 which was raised by immunization with human uterine cervical cancer cells.13 Nakashima et al. isolated a cDNA that was assumed as a responsible gene for the 22-1-1 antigen through expression cloning and designated the gene as RCAS1.14 Since RCAS1 cDNA was eventually identical to EBAG9 cDNA, EBAG9 protein is identical to the protein encoded by RCAS1 cDNA, but distinct from the cell surface 22-1-1 antigen. It has been reported that 22-1-1 antigen is tumor-associated O-glycan, and that EBAG9 increases O-glycan expression.12 Up to date, there are a number of reports in regard to the immunohistochemical studies using the 22-1-1 antibody, yet careful attention needs to be paid to distinguish the functions of the protein detected by the 22-1-1 antibody from those of the protein encoded by EBAG9/RCAS1 cDNA.
In the previous study of renal cell carcinoma Renca cells, we showed that inoculated Renca cells harboring EBAG9 (Renca-EBAG9) in BALB/c mice developed into larger tumors than Renca cells transfected with the vector alone (Renca-vector), whereas both Renca-EBAG9 and Renca-vector tumors had similar volumes in BALB/c nude mice.9 The mechanism responsible for this difference in tumor growth was assumed to arise from a smaller number of infiltrating CD8+ T lymphocytes in Renca-EBAG9 subcapsular tumors, leading to an impaired antitumor immunity. In the present study using a human bladder transitional cell carcinoma model, EJ cells, we noticed that EJ-EBAG9 tumors developed into significantly larger tumors than EJ-vector tumors in nude mice with abnormal T cell activity.
Several mechanisms explaining the promotion of EJ-EBAG9 tumor growth in nude mice can be considered. Firstly, local immune responses other than T cells, such as the upregulation of inflammatory cytokines, natural killer cells, neutrophils, or macrophages may be suppressed in nude mice with EJ-EBAG9 tumors. These immunological responses are important for cancer growth and cancer maintenance.15 Nude mice specimens were stained using hematoxylin-eosin, to observe immune cell invasion or apoptotic cells, as other reports did.16, 17 However, we could not find an overt difference in the numbers of invaded immune cells and apoptotic cells in the EJ-EBAG9 tumors, compared with those in the EJ-vector tumors (data not shown). Secondly, EBAG9 may stimulate angiogenesis by up-regulating growth factors or cytokines. In bladder cancer, angiogenesis has been reported to be associated with the prognosis of patients with invasive bladder cancer.18 Again, we could not find an overt increase in vascularity in EJ-EBAG9 tumors compared with EJ-vector tumors when nude mice specimens were immunostained with anti-VEGF receptor antibody (data not shown). Thirdly, EBAG9 may potentiate cell migration. The present observations suggest that EBAG9 may potentiate tumor growth through a mechanism that involves altering the ability of the tumor cells to migrate. Several candidate genes are associated with both migration and tumor formation. Genes associated with integrin traffic could promote migration19 and might be involved in the effect of EBAG9 overexpression. Accordingly, we examined the change in tyrosine-phosphorylated proteins between EJ-EBAG9 cells and EJ-vector cells using Western blotting with a 4G10 antibody, although no significant difference was observed. Julien et al. showed that sialyl-Tn expression and concomitant changes in the overall O-glycan profiles induced a decrease in cell adhesion and an increase in cell migration in MDA-MB-231 breast cancer cells.20 Moreover, sialyl-Tn-positive clones exhibited increased tumor growth in severe combined immunodeficiency mice.20 It is tempting to speculate that EBAG9 may enhance tumor growth and cell migration by modifying glycosylation, such as sialyl-Tn. Actually, EBAG9 has been suggested to be involved in the glycosylation process.12, 21
Despite advances in our knowledge of the biological behaviors of bladder cancer, physicians and patients have few tools to help decide whether adjuvant therapy after cystectomy is indicated. Shariat et al. demonstrated that the immunohistochemical expression of a panel of established cell cycle regulators (p53, pRB, p21, p27, and cyclin E1) could help to identify patients with pTa-3 N0 M0 bladder cancer, who have an increased risk of disease progression.22 Since EBAG9 immunoreactivity was associated with a poor outcome, its use as a biomarker may help physicians to decide whether postoperative chemotherapy or radiotherapy should be used after a total cystectomy.
In summary, we showed that EBAG9 is a tumor-promoting factor in human bladder cancer. We propose that EBAG9 immunoreactivity could be a potential biomarker for determining the prognosis of patients with bladder cancer; furthermore, treatment modalities targeting EBAG9 would provide a novel therapeutic option for patients with advanced bladder cancer.
The authors thank Dr. T. Takeuchi (The University of Tokyo, Tokyo, Japan) for his technical assistance and critical discussion.