Expression of hepaCAM inhibits bladder cancer cell proliferation via a Wnt/β-catenin-dependent pathway in vitro and in vivo

Abstract

We previously established that hepatocyte cell adhesion molecule (hepaCAM), a typical structure of immunoglobulin (Ig)-like adhesion molecules, inhibited the proliferation and the progression of cultured human bladder cancer cells. As increasing evidence reveals that aberrant activation of canonical Wnt pathway is involved in the pathogenesis of bladder cancer, and β-catenin serves as a pivotal molecule of Wnt pathway. Then, we explored whether the anti-proliferation effect of hepaCAM was associated with Wnt/β-catenin pathway in human bladder cancer cells. The negative correlation between hepaCAM and β-catenin in transitional cell carcinoma of bladder (TCCB) was found. Follow by, studied the effect of hepaCAM on the key elements of Wnt pathway. Here, Our researches showed that hepaCAM played a central role in modulating the Wnt/β-catenin signaling pathway by interfering nuclear protein levels of β-catenin, leading to down-regulate transcriptional activity of LEF/TCF and its target genes c-Myc and cyclinD1. Mechanistically, we demonstrated that hepaCAM-activated GSK3β led to elevate the phosphorylation of β-catenin, contributing to the aberrant translocation of β-catenin. In addition, Anti-proliferation and associated molecular mechanisms of hepaCAM were demonstrated by using vivo experiment. In conclusion, our reports uncover that expression of hepaCAM suppresses the proliferation of bladder cancer cells through a Wnt/β-catenin-dependent signaling pathway in vitro and in vivo.

Abbreviations

Ad

adenvirus

HepaCAM

hepatocyte cell adhesion molecule

Introduction

Reports from United States in 2013 have showed that bladder cancer is the most common genitourinary malignant disease with estimated 140, 430 new cases and 29, 790 deaths.¹ The majority of bladder cancer occurs in males and there is a 14-fold variation in incidence internationally.² Although a few improvements have achieved in diagnosis and treatment, prognosis of transitional cell carcinoma of bladder (TCCB) remains extremely poor, implying the need for development of novel therapeutic approaches. As we know, genetic alterations, including 2 major groups of growth regulatory genes: oncogenes and tumor suppressor genes, have been defined as the hallmark of cancers. The major attention has been focused on identifying oncogenes or tumor suppressor genes that inhibit tumor development processes. Meanwhile, the molecular mechanisms underlying the preventive effects of genetic alterations have not been fully elucidated. Therefore, this study was designed to uncover them.

Hepatocyte cell adhesion molecule(hepaCAM) is first discovered in the liver, and it displays a typical structure of immunoglobulin (Ig)-like adhesion molecules containing 2 extracellular Ig-like domains, a transmembrane segment and a cytoplasmic tail.³ Furthermore, studies showed that down-regulation of hepaCAM was found in many tumor tissues and cell lines, especially in hepatocellular carcinoma.^3,4,5 Moreover, Our group previously demonstrated that hepaCAM was hypermethylated at the exon 2 region in bladder cancer tissues and cell lines,⁶ the expression of hepaCAM was elevated in T24 and BIU87 cells after 5-aza-2′-deoxycytidine (5-Aza-Cdr) and interferon- γ treatment, and re-expression of hepaCAM reduced cells proliferation and colony formation in diverse cancer cells.^3-4,7 In addition, hepaCAM could induce senescence-like growth arrest through a p53-/p21-dependent pathway in human breast cancer cells.⁵ These reports indicate that hepaCAM is considered as a tumor suppressor gene. Besides, overexpression of hepaCAM induced G1 phase arrest and promote c-Myc degradation in renal cancer cells.⁸ HepaCAM also regulated the expression levels of c-Myc and cyclinD1 and inhibited proliferation via an AMPK/mTOR-dependent pathway in human bladder cancer cells.⁹ However, Whether there are other the possible mechanisms modulated the expression of c-Myc, cyclinD1, promoted us to further investigate.

The Wnt's are secreted glycoproteins that are powerful regulators of cell proliferation and differentiation via serving as ligands to stimulate receptor-mediated signal transduction pathways, and activation of Wnt signaling pathway is involved in the pathogenesis of various human tumors.^10,11 An important and most studied Wnt pathway is the canonical Wnt signaling, which functions by modulating the translocation of β-catenin to the nucleus, where it controls critical gene expression programs via interaction with TCF/LEF and other families of transcription factors. In normal condition, casein kinase 1β (CK1β) and glycogen synthase kinase-3β (GSK-3β) sequentially phosphorylate β-catenin in the Axin complex, which is composed of the scaffolding protein Axin, the tumor suppressor adenomatous polyposis coli gene product (APC), CK1β, and GSK-3β. Phosphorylated β-catenin is subsequently ubiquitinated, resulting in its proteasomal degradation. This action keeps β-catenin at low level in the cytoplasm and prevents it from translating to the nucleus leading to repress Wnt target genes. However, when a Wnt ligand binds its cell-surface receptor consisting of Frizzled (Fz) receptor and its co-receptor, low-density lipoprotein receptor related protein 6 (LRP6) or LRP5, the Wnt/β-catenin pathway is thus activated, and this action leads to antagonize the action of GSK-3β and suppress β-catenin phosphorylation, At this stage, β-catenin accumulates in the cytoplasm because of it dissociating from the Axin complex and not being degraded. Follow by, β-catenin translocates to the nucleus, where it binds to TCF/LEF, thereby transcription the expression of various Wnt target genes, for instance, c-Myc and cycinD1, which are related with cell growth.^12-15 Our groups identified that re-expression of hepaCAM could decrease β-catenin expression.¹⁶ Thereby, we hypothesize that overexpression of hepaCAM may suppress proliferation through Wnt/β-catenin pathway in human bladder cancer cells.

Thus, in this study, we explored the correlation between hepaCAM and β-catenin in transitional cell carcinoma of bladder (TCCB) and the regulatory roles of hepaCAM in Wnt/β-catenin signaling pathway. Altogether, our researches demonstrate a new a novel link between hepaCAM and β-catenin, and hepaCAM can inhibit cell growth by Wnt/β-catenin pathway suppression in bladder cancer cells.

Results

Low hepaCAM and high β-catenin protein expression in TCCB

To explore the correlation between β-catenin and hepaCAM in TCCB, mRNA levels of hepaCAM and β-catenin, protein levels of β-catenin were determined in 20 pairs of bladder tumor samples and their corresponding normal bladder samples, which were taken as control groups (Fig. 1.A, B). The results of qRT-PCR indicated that the expression of hepaCAM was reduced in 85% (17/20) of TCCB tissues compared with control groups (P < 0.01, Fig. 1A). These data suggested an association between the reduction of hepaCAM expression and the presence of TCCB. However, the mRNA and protein expression levels of β-catenin were increased in 70% (14/20) of TCCB specimens (Fig1. A, B). A negative linear correlation between hepaCAM and β-catenin was found in the same patient by using the spearman linear correlation analysis (r = −0.723, P = 0.0015, Fig. 1C). Nevertheless, hepaCAM and β-catenin were not related with any clinical or pathological variables including sex, histologic stage, grade or recurrence (Table 2) .

Figure 1.

The expression levels of hepaCAM and β-catenin in TCCB patient specimens. (A) Analysis of hepaCAM and β-catenin expression in 20 pairs of TCCBs and their adjacent noncancerous bladder tissues by using real-time PCR. N: noncancerous bladder tissues. C: transitional cell carcinoma of bladder. (B) Western blot analysis was used to determine β-catenin protein levels in 20 pairs of TCCB specimens and corresponding adjacent tissues. GAPDH was used as a loading control to analyze equal amounts of protein in all lanes. (C) Correlation curve of hepaCAM and β-catenin in TCCB(r = −0.723, P < 0.01). (D)Representative immunohistochemical (IHC) staining for hepaCAM, β-catenin in paired bladder tumors and noncancerous bladder tissues. Magnification, 200×. *P < 0.05, ** P < 0.01.

Table 2.

The mRNA expression of hepaCAM and β-catenin in TCCB and clinicopathological parameters

No. Specimens	hepaCAM	β-catenin		P value
		(Mean±SD)	(Mean±SD)	hepaCAM	β-catenin
Tissue:
Cancer	20	0.507 ± 0.407	1.733 ± 0.401	* <0.01	* <0.01
Adjacent	20	1.520 ± 0.231	0.782 ± 0.321
Gender
M	12	0.458 ± 0.217	1.674 ± 0.365	0.391	0.554
F	8	0.520 ± 0.301	1.753 ± 0.221
Histological grade:
Ta–T1	14	0.627 ± 0.345	1.645 ± 0.383	0.621	0.792
T2–T4	6	0.492 ± 0.351	1.882 ± 0.635
Histological stage:
Low	13	0.687 ± 0.447	1.423 ± 0.540	0.819	0.743
High	7	0.420 ± 0.131	1.502 ± 0.321
Occurrence:
Primary	11	0.407 ± 0.197	1.553 ± 0.381	0.258	0.431
Recurrence	9	0.570 ± 0.471	1.621 ± 0.621

^*P < 0.01 considered significant.

To confirm the correlation between β-catenin and hepaCAM in TCCB immunohistochemistry was used to further analysis hepACAM and β-catenin protein expression and position in 20 pairs of bladder tumor, adjacent tissues samples and another 15 cases TCCB with no adjacent tissue. Results showed that the protein levels of hepaCAM almost lost in total 35 tumor specimens. In contrast, hepaCAM mainly expressed at the cytomembrane and few located at the cell nucleus or whole cell in 20 cases of adjacent tissues. Meanwhile, in β-catenin immunostaining, bladder tumor showed 42.8% (15 of 35) positive for β-catenin nuclear accumulation with cytoplasmic staining. However, the position of β-catenin in noncancerous tumor tissues was almost at membrane and cytoplasm. These data further illustrated a correlation between loss of hepaCAM expression and high β-catenin expression in TCCB.

HepaCAM targets β-catenin and its downstream genes

In our previous research, overexpression of hepaCAM could suppress the proliferation of bladder cancer cells in vitro [9]. Colony formation assay was used to further identify anti-proliferation of hepaCAM. In line with our previous findings, reduction of colony numbers was observed in Ad-GFP-hepaCAM-infected T24 cells, compared with control groups (P < 0.05) (Fig. 2A). Then, the mechanisms that took part in hepaCAM-mediated proliferation were explored in T24 cells. According to microarray profiling data in our previous research, expression of hepaCAM reduced the β-catenin and cyclinD1 expression at mRNA levels,16 and since the close correlation between hepaCAM and β-catenin was found. We examined the levels of β-catenin and its representative downstream targets about proliferation, c-Myc and cyclinD1 by using qRT-PCR and western blot. Data showed that expression of hepaCAM down-regulated the β-catenin, c-Myc and cyclinD1 at mRNA and protein levels, and there was no statistical difference between Ad-GFP group and control group (P < 0.01, Fig. 2B, C). To our knowledge, when Wnt/β-catenin signaling pathway is activated, β-catenin accumulates in the cytoplasm and translocates to the nucleus, where it binds to TCF/LEF transcription factor, thereby stimulating the expression of various Wnt target genes. Thus, we detected the nuclear protein levels of β-catenin after cells treatment with Ad-GFP-hepaCAM. As expected, the levels of β-catenin nucleus protein decreased in Ad-GFP-hepaCAM cells compared with control groups. These findings indicated that hepaCAM inhibited endogenous expression of β-catenin, c-Myc, cyclinD1 and the nuclear translocation of β-catenin in bladder cancer cells .

Figure 2.

Effects of ectopic hepaCAM of expression in T24 Cells. (A) Representative colony-formation assay of hepaCAM in T24 (left panel). Quantitative analyses of colony numbers are shown as values of mean ± SD. * P < 0.05 (right panel). (B, C) Protein and mRNA expression of β-catenin, c-myc, cyclinD1 were detected at 48h after the infection with Ad-GFP and Ad-GFP-hepaCAM, respectively. GAPDH was used as a control; Subcellular localization of β-catenin protein was measured by western blotting at 48 h after the infection with Ad-GFP and Ad-GFP-hepaCAM, respectively. Cell lysates were collected and nuclear proteins were extracted. Histone H3 was used for nuclear protein loading control. (D) Immunofluorescence staining was performed to further demonstrate intracellular distribution of β-catenin at 48 h after cells treatment with Ad-GFP and Ad-GFP-hepaCAM in T24 cells. **P < 0.01,* P < 0.05 vs blank control.

Next, we further detected the effect of hepaCAM on cellular localization of β-catenin protein by using immunofluorescence. we found that β-catenin protein appeared as diffuse staining throughout the cytoplasm, nucleus, and cell membrane in Ad-GFP group and control group cells. However, β-Catenin protein mainly localized to the cell cytoplasm in Ad-GFP-hepaCAM group cells, with low levels of staining also appearing in the nucleus (Fig. 2D).

HepaCAM inhibited the TCF promoter activity

β-catenin acts to regulate the transcription of genes through the binding of a complex of β-catenin and T Cell Factor (TCF) family of transcription factors to specific promoter elements. The decrease of nuclear β-catenin by Ad-GFP-hepaCAM treatment implied that β-catenin nuclear signaling might have been attenuated. Thereby, we evaluated the effect of hepaCAM on the transcriptional activities of β-catenin in T24 cells by using the TOPflash/FOPflash reporter system. The TOP-flash luciferase reporter plasmid contains 3 copies of the consensus T-cell factor (TCF) binding sites upstream of the luciferase gene, whereas its negative control version (FOPflash) carries mutations at these binding sites. As shown in Figure. 3A, Ad-GFP-hepaCAM treatment for 24 h reduced luciferase activity (TOPFlash) in T24 cells (P < 0.01) .

Figure 3.

HepaCAM suppressed the transcriptional activity of β-catenin/Tcf. (A) The T24 cells lines were co-transfected with reporter genes harbouring Tcf-4 binding sites (TOPFlash) or a mutant Tcf-binding site (FOPFlash), respectively, and β-galactosidase gene. After transfection, cells were infected with Ad-GFP and Ad-GFP-hepaCAM. After 24 h of infection, Luciferase activity was determined, normalized against values for the corresponding β-galactosidase activity. Data shown represent mean ± SD of 3 independent observations (**p < 0.01, compared to blank control).

HepaCAM induced the degradation of β-catenin

We further examined the mechanism underlying aberrant nuclear transloction of β-catenin caused by hepaCAM. Since nuclear accumulation of β-catenin is inversely correlated with phosphorylation at certain key residues of β-catenin, then, we measured the effect of hepaCAM on the levels of β-catenin phosphorylation. Western blot analysis showed that treatment of cells with Ad-GFP-hepaCAM increased the phosphorylation of β-catenin compared with control groups (Fig. 4A). As the β-catenin phosphorylation occurs in a multiprotein complex including axin and GSK3 α/β. In addition, GSK3β kinase is known to target β-catenin for proteasomal degradation via combined phosphorylation at key residues of β-catenin, and its activity is regulated by site-specific phosphorylation; full activity of GSK3β generally requires phosphorylation at tyrosine 216 (try216), and conversely, phosphorylation at serine 9 (Ser9) inhibits GSK3β activity.¹⁷ Thereby, Western blot analysis was undertaken to examine total GSK3β, p-GSK3β (ser9) and p-GSK3β (tyr216) protein levels after overexpression of hepaCAM in T24 cells. As shown in Figure. 4A, significantly higher levels of p-GSK3β(tyr216) and less p-GSK3β(ser9) were observed in Ad-GFP-hepaCAM group as compared to control groups(P < 0.05).However, Ad-GFP-hepaCAM treatment had no effect on total GSK3β .

Figure 4.

Effects of overexpression of hepaCAM on p-β-catenin, p-GSK3β, p-AKT, total AKT, total GSK3β in T24 cells. (A) Expression levels of p-β-catenin, p-GSK3β (tyr216/ser9), total GSK3β, p-AKT (ser473) and total AKT were detected at 48h after the infection with Ad-GFP and Ad-GFP-hepaCAM l. All the samples were also probed with anti-GAPDH antibody to verify equal loading.

Furthermore, as GSK3β is the shared intermediate for AKT and Wnt signaling, and it is an important substrate of AKT, and activated AKT phosphorylates and inactivates GSK3β, resulting in β-catenin accumulation and promotion of cellular growth and survival. Therefore, T24 cells were treated with Ad-GFP-hepaCAM, and proteins were detected for AKT phosphorylation on Ser473. As indicated in Figure. 4A, overexpression of hepaCAM had no effect on the phosphorylation of AKT. These findings suggested that hepaCAM through GSK3β modulated Wnt/β-catenin signaling and might suppress cells growth via Wnt/β-catenin signaling.

LiCL reversed the inhibition effect of hepaCAM on cell growth

To validate the role of GSK3β and its effect on the growth of bladder cancer cells in response to hepaCAM overexpression. We performed experiments with pharmacological GSK3β inhibitor (LiCL), which inhibits GSK3β and mimics Wnt signaling by stabilizing β-catenin.¹⁸ CCK-8 and colony formation assay showed that treatment of LiCL (10 uM) partly blocked the effect of hepaCAM on the proliferation of T24 cells (P < 0.01, P < 0.05, (Fig. 5A and B). Moreover, combined treatment of LiCL with hepaCAM, activation of GSK3β and elevated phosphorylation of β-catenin by hepaCAM treatment alone were partly hindered, downregulation of β-catenin and its targets, c-Myc and cyclin D1 by hepaCAM treatment alone were reversed (Fig. 6A) .

Figure 5.

Proliferation effects of hepaCAM alone or co-treatment with Wnt inhibitor DKK1 or GSK3β inhibitor LiCL by using Colony-formation assay and CCK-8 assay in T24 cells. (A) Representative colony-formation assay of hepaCAM alone or co-treatment with Wnt inhibitor DKK1 or GSK3β inhibitor LiCL in T24 cells. Quantitative analyses of colony numbers are expressed as values of mean ± SD. (B) CCK-8 assay assessed the effect of hepaCAM alone or co-treatment with Wnt inhibitor DKK1 or GSK3β inhibitor LiCL on cell proliferation in T24 cells. The values are shown as the mean ± SD. **P < 0.01, *P < 0.05 vs Ad-hepaCAM, ^#, P < 0.05 vs Blank control.

Figure 6.

Effects of hepaCAM alone and co-treatment with Wnt inhibitor DKK1 or GSK3β inhibitor LiCL on protein expression. (A) T24 cells cultured for 12 hours were pretreated with Ad-GFP and Ad-GFP-hepaCAM, followed by treatment with 10uM LiCL for 2h, the expression of p-GSK3β, GSK3β, p-β-catenin, β-catenin, c-Myc and cyclin D1 were measured by western blot. (B) T24 cells cultured for 12 hours were pretreated with Ad-GFP and Ad-GFP-hepaCAM, followed by treatment with 200ng/mL DKK1 for 24h. Western blot was used to examine c-Myc and cyclin D1 protein expression.

Combined treatment of Wnt inhibitor DKK1 with hepaCAM exhibited synergistic inhibitory effects on cell proliferation and Wnt/β-catenin signaling in T24 cells

In order to further confirm that hepaCAM modulation on canonical Wnt/β-catenin signaling, we measured whether there was an overlap between the effect of hepaCAM and DKK1, which inhibit Wnt signaling by binding to the LRP5/6 component of the Wnt receptor complex,¹⁹ As shown in Figure. 6B, findings reflected that combined treatment of Ad-GFP-hepaCAM and Wnt signaling inhibitor DKK1 exhibited synergistic inhibitory effects on the expression of c-Myc and CyclinD1. In addition, the results of CCK-8 and colony formation assay indicated that DKK1 also enforced the inhibition effect of hepaCAM on the growth of T24 cell lines (P < 0.05, Fig. 5A, B).

HepaCAM inhibited bladder tumor growth in vivo

Further to evaluate the tumor-suppressive function and involvement mechanisms of hapaCAM in vivo, tumorigenicity of T24 cells overexpression of hepaCAM was tested in nude mice. Thirty days after injection, tumors were excised from tested mice for further analysis. The weight of tumors induced by hepaCAM-expression T24 cells was significantly decreased, compared with control tumors (P < 0.05, Fig. 7A, B). Immunohistochemistry was further performed to analyze the expression of hepaCAM, β-catenin, p-β-catenin and cell proliferation marker Ki-67 in xenograft tumors. Numerous tumor cells with higher nuclear fragmentation were observed in H&E-stained sections from hepaCAM-expressing T24 cells compared with control groups, along with decreased proliferating cells. In addition, immunohistochemistry showed that low expression levels of hepaCAM, β-catenin, Ki67 and high p-β-catenin immunostaining in Ad-GFP-hepaCAM-infected T24 cells-derived tumors tissues (Fig. 7C). These results indicate that hepaCAM acts as a tumor suppressor genes and suppresses cells proliferation involvement Wnt/β-catenin pathway in bladder tumorigenesis .

Figure 7.

HepaCAM inhibited the tumorigenicity of bladder cancer in vivo. (A) Tumors derived from Blank, GFP, and hepaCAM-expression T24 cells in nude mice. (B) Tumor growth curve of hepaCAM-expressing cells in nude mice compared with control tumors. *P < 0.05. (C) Representative photographs of hematoxylin and eosin (H&E) staining and immunohistochemistry (IHC) analysis of hepaCAM, Ki-67, β-catenin and p-β-catenin antigen and. Magnification, 200

Discussion

In this paper, we further reported that the mRNA and protein expression of hepaCAM were downregulated in TCCB tissues in comparison with their normal tissue counterparts, and hepaCAM was primarily expressed at the cytomembrane with a minor portion in the cell nucleus in human adjacent bladder transitional carcinoma. However, hepaCAM was hardly expressed in cancer samples, which were in accordance with our earlier findings of the hepaCAM expression and localization in TCCB.^7,9 Nevertheless, because of limited number of tissue specimens, there was no relationship between hepaCAM expression and various clinicopathological parameters in the same patients. Moreover, the expression and distribution of hepaCAM were confirmed by using quantitative RT-PCR and immunohistochemistry, which make our findings more convincible.

Wnt/β-catenin signal transduction pathway is constitutively activated in several major human cancers, including bladder cancer. Studies also showed that Wnt/β-catenin signaling played a central in the development of urothelium, and there was a different distribution of extracellular components of the Wnt protein among normal bladder mucosa, superficial bladder tumor and invasive bladder tumor,^20-23 In addition, reports indicated that genetic alterations in the components of the canonical Wnt signaling pathway, for instance, β -catenin and/or the APC gene, played a critical role in the activation of the Wnt/β-catenin signaling pathway.^24-26 While, researches previously showed that only small subsets of mutated β-catenin gene existed in bladder tumor. Meanwhile, genetic mutation of APC seemed to be a rare event in human bladder cancer.²⁷ To our knowledge, β-catenin is a key regulator of Wnt pathway, and recent results suggested that compared with bladder mucosa, bladder tumor showed significantly increasing nuclear β-catenin expression, which was involved in canonical Wnt/β-catenin signal activation. More importantly, β-catenin could transcriptionally regulate cyclin D1 and c-Myc of Wnt target genes that they were involved in cell survival, proliferation,^28-30 Therefore, the expression levels and distribution of β-catenin were further examined in tumor specimens and the corresponding normal bladder specimens. Data showed that the localization of β-catenin in bladder tumor was mainly in the nucleus compared with adjacent bladder tissue, with significant difference in β-catenin protein expression between bladder carcinoma and adjacent samples, which were in line with earlier reports. However, the mRNA expression of β-catenin was a little different from the study by Urakami, S et al. In that work, no obviously difference was found in mRNA expression of β-catenin between bladder tumor and adjacent tissues samples,¹⁷ which may be derived from the difference of samples or the methods of used.

The strong inverse correlation between hepaCAM loss and β-catenin up-regulation we observed suggests that the significance of Wnt signaling in human bladder cancer and hepaCAM -mediated cells proliferation might through Wnt signaling.

Notably, researchers identify that the abnormal levels of nuclear β-catenin, which represent a hallmark of active Wnt/β-catenin signaling, can bind to transcription factor TCF/lymphoid enhancer factor (LEF) activating a cluster of genes that ultimately establish the oncogenic phenotype such as c-Myc, cyclinD1. As we know, c-Myc and cyclinD1 are 2 proto-oncogenes which are central modulators of cellular growth and proliferation.^31,32 Previously, our group showed that expression of hepaCAM down-regulated mRNA and protein levels of c-Myc and cyclinD1 and consequently inhibited cells survival.^8,9 In present study, we observed that re-expression of hepaCAM reduced the total and nuclear protein levels of β-catenin, followed by suppressed the transcription activity of TCF/LEF and its downstream genes. In the canonical Wnt signaling pathway, intracellular accumulation of β-catenin could be reversed by glycogen synthetase kinase3β (GSK-β)-dependent phosphorylation of β-catenin and leads to ubiquitination and degradation through the proteosome.³³ Therefore, preventing β-catenin degradation when the canonical Wnt signaling pathway is activated leads to accumulation of free-β-catenin, and then free-β-catenin translocates to the nucleus and acts as a transcription factor in a complex with TCF/LEF-1 to activate genes such as c-Myc, cyclin-D1 et al. However, since we revealed that hepaCAM and β-catenin were closely related and hepaCAM could modulate β-catenin expression and distribution. We need to explore how they correlated with each other. In most case, GSK3α/β is known to be modulated by several signaling pathways, such as PI3K/AKT, MEK/ERK and Wnt signaling pathway.³⁴ Among them the Wnt/GSK3β pathway is most extensively studied. In normal conditions, β-catenin is controlled by the upstream regulators of the Wingless-type (Wnt) signaling pathway. However, In the presence of Wnt signaling, casein kinase I (CKI) and GSK3β become inactivated, leading to cytoplasmic and subsequently nuclear accumulation of β-catenin, this transcriptional factor in the nucleus forms complexes with members of the lymphoid enhancer factor/T-cell factor (LEF/TCF) family of transcription factors to activate transcription. Ectopic hepaCAM of expression obviously activated GSK3β leading to increase phosphorylation β-catenin, but had no effect on the phosphorylation of AKT, which suggested that hepaCAM might modulate β-catenin expression and anti-proliferation through Wnt/GSK3β pathway in T24 cells, which is a little different from the research by Qiaolin Zhang et al. In that report, hepaCAM-induced c-Myc degradation was independent of GSK3β in human renal cell carcinoma, which may explain that different cells have unique signal mechanism to maintain the proliferation of cancer cells. In our study, we observed that re-expression of hepaCAM induced c-Myc down-regulation, which might dependent on Wnt/GSK3β pathway in human bladder cancer cells. DKK1 is a Wnt signaling inhibitor by binding to the LRP5/6 component of the WNT receptor complex.¹⁹ Thus, we observed that the synergistic inhibitory effects of combination treatment of hepaCAM and DKK1 inhibitor on cell growth. DKK1 inhibitor alone could down-regulate c-Myc and cyclinD1 expression. However, when combined with hepaCAM much more inhibitory effect was obtained. This combination therapy may be a potential and novel therapeutic strategy for bladder cancer in the future. Meanwhile, T24 cells were combined treatment of hepaCAM with LiCL, which inhibits GSK3β and mimics Wnt signaling by stabilizing β-catenin.¹⁸ Data showed that the enhancement of p-β-catenin expression caused by hepaCAM was partly eliminated in the presence of the GSK3β inhibitor LiCL, and the inhibition of p-GSK3β (ser9), β-catenin, c-Myc and cyclin D1 were also partly blocked supporting their downstream position in the signaling cascade. Moreover, hepaCAM-induced growth inhibition in bladder cancer was partly reversed by treatment with Wnt activator LiCL.

In addition, vivo experiment was performed to further clarify the molecular mechanism underlying the suppression of bladder cancer growth mediated by re-expression hepaCAM. We investigated the expression of Ki67, which is frequently used as a marker of cell proliferation and associated with carcinogenesis. Our study revealed that the Ki67 protein levels decreased after the hepaCAM gene was expressed indicating that tumor growth was inhibited. Moreover, β-catenin expression was downregulated after re-expression hepaCAM in vivo, in accordance with our finding in vitro. These results indicate that tumor growth was inhibited and depended on Wnt/β-catenin pathway. Therefore, our researches support the ideas that the observed hepaCAM-elicited inhibition of Wnt/β-catenin pathway mediates suppression of the growth of bladder cancer cells. While, the mechanism by which hepaCAM interacts with GSK3β awaits further investigation.

In conclusion, the results of this report suggest that the hepaCAM has the ability to block growth potential of bladder cancer cells and the indicated anti-proliferation effect of this gene is mediated through the inactivation of the Wnt/β-catenin pathway.

Materials and Methods

Patients and specimens

20 pairs of human bladder tumor samples and their matched adjacent bladder samples were obtained from patients who underwent total cystectomy from the First Affiliated Hospital of Chongqing Medical University during 2012 and 2013. In addition, 15 patients were treated with transurethral resection of bladder tumor but no adjacent tissue was also obtained. All patients provided informed consent. All carcinoma and adjacent bladder samples were identified by histological examination, and histological grade and stage were determined according to UICC guidelines, containing stage Ta-T1 in 21, T2 -T4 in 14. In all TCCB samples, 18 cases were confirmed as primary tumors, while the rest were confirmed as recurrent tumors. All patients have the similar socioeconomic status and the study protocol had been approved by the ethical committee. All specimens were stored at −80°C before experiment.

Immunohistochemical procedures

The assay was performed as described previously,⁷ In brief, Paraffinwax embedded tissue sections were dewaxed, rehydrated and microwaved for 30 minutes in sodium citrate buffer to retrieve antigen epitopes. Endogenous peroxidase activity was suppressed by 3% H₂O₂ and blocked by goat serum(5%BSA. Diluted primary polyclonal rabbit antibody against hepaCAM (ProteinTech, Wuhan, People's Republic of China), β-catenin (abcam, USA), Ki67, p-β-catenin ((Immunaway Biotechnology, Company) were added and left at 4°C overnight. As secondary reagents, we used biotin-labeled anti-IgG and avidin-biotin horseradish peroxidase complex, followed by staining with the chromogen diaminobenzidine (Zhongshan, Bei-jing, People's Republic of China) until a brown color developed. Slides were counterstained with Mayer hematoxylin and differentiated in a solution containing 1%hydrochloric acid and 99% ethanol. Cell nuclei were stained blue using lithium carbonate. Sections were dehydrated and a transparent coverslip was added to enable observation by microscopy.

All immunohistochemical photographs were analyzed by using Image Pro Plus (IPP, version 6.0; Media Cybernetics, Silver Spring, MD USA). The mean optical density (OD), as a quantitative measure of stain intensity, was analyzed to determine average protein expression.

Reverse transcription and quantitative RT-PCR (qRT-PCR)

20 pairs of tissue specimens were pulverized by using liquid nitrogen. The total RNA from cells and tissues was extracted by using Trizol reagent (Takara, Tokyo, Japan) according to the manufactures' instructions. CDNA was generated from total RNA by using reverse transcription kit (Takara, Tokyo, Japan) according to the manufacturer protocol. The primer sequences and PCR condition were listed in Table 1. The qRT-PCR was conducted following the instructions on the SYBR Premix ExTaq™ IIkit (Takara, Japan), and qRT-PCR reactions were performed on a 7500 ABI system (Applied Biosystems, USA). All qRT-PCR data was normalized to β-actin expression .

Table 1.

primer sequences and PCR condition

Gene name	Primer sequence	TM(°C)	cycles	product
Length (bp) β-catenin	5′- GCTTGGAATGAGACTGCTGA -3′	57.5	30	114
	5′- CTGGCCATATCCACCAGAGT -3′
c-Myc	5′- CCTACCCTCTCAACGACAGC -3′	56	34	248
	5′- CTCTGACCTTTTGCAAGGAG -3′
cyclinD1	5′- CCTGTCGCTGGAGCCCGTG -3′	56	34	252
	5′- TCCGCCTCTGGCATTTTGG -3′
β-actin	5′- TGACGTGGACATCCGCAAAG -3′	56	34	205
	5′-CTGGAAGGTGGACAGCGAGG -3′

Protein extraction and western blotting

The experiments were performed as described previously.^7,9 Total protein extraction from cells and tissues was performed using RIPA buffer supplemented with protease inhibitor (PMSF) and phosphatase inhibitor (NaF and Na₃VO₄)(Roche, Switzerland) on ice. The BCA protein assay kit (Beyotime Institute of Biotechnology) was used to detect the quantity of each sample according to manufacturer instructions. Loading buffer (5×) was added to all protein solutions, which were then boiled for 5 minutes. 100ug proteins were separated by 12% or 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Invitrogen, Carlsbad, CA USA). Then the proteins were electrotransferred onto polyvinylidene difluoride membranes (0.45um pore size; Millipore Corporation, Billerica, MA) in transfer buffer. Nonspecific binding was prevented by blocking with 0.1%Tween 20 in TBST containing 5% nonfat dry milk for 2 hours at room temperature. Membranes were first incubated overnight at 4°C with the primary antibodies against phospho-GSK3β (ser9,try216), total GSK3β, phosphor-AKT (ser473), total AKT, phosphor-β-catenin, total β-catenin(Cell Signaling Technology, Boston, MA USA), c-Myc, cyclinD1 and GAPDH(Santa Cruz Biotechnology, Santa Cruz, CA USA). Membranes were then incubated with horseradish peroxidase-conjugated secondary antibody at 1:1000 dilution for 1h at 37°C. Finally, membranes were exposed and developed using an enhanced chemilu-minescence kit (Beyotime) in the darkroom.

Cells and culture

Human TCCB cell line T24 was purchased from American Type Culture Collection (ATCC, Manassas, VA USA). Cells were kept in RPMI 1640 medium (Gibco, USA) supplemented with 10% fetal bovine serum (Gibco, USA) in a humidified incubator with 95% O2 and 5% CO2 at 37°C.

Immunofluorescence

T24 cells were plated on cover slips. When they reached 60% confluence, they were infected with adenovirus-GFP or adenovirus-GFP-hepaCAM, respectively, and cultured for 48h. Thereafter, cells were washed with phosphatebuffered saline (PBS) for 3 times, fixed with paraformaldehyde for 20min and followed by 3 washes in PBS. Permeabilisation was achieved by incubating the fixed cells in 0.1% Triton ×100 for 15 min at room temperature, and they were then washed with PBS for another 3 times. Then, cells were blocked with 10% normal goat serum for 1 h at 37°C. Subsequently, cells were incubated with specific primary antibodies against β-catenin (dilution, 1:100, abcam, USA) overnight at 4°C, and followed by washing with PBS. The secondary antibodies Rhodamine (TRITC) conjugated goat anti-rabbit IgG were added and incubated for 1 h at 37°C. Then the coverslips were washed 3 times with PBS. Nuclei were stained with DAPI (40, 60-diamidino-2-phenylindole). Stained cells were viewed under a fluorescence microscope (ECLIPSE80I, Nikon, Tokyo, Japan).

Nuclear protein extraction

Nuclear proteins were extracted using Beyotime Nuclear and Cytoplasmic Extraction Reagents (Beyotime Institute of Biotechnology) according to the manufacturer's instructions.

β-Catenin/TCF transcription reporter assay

T24 cells were plated into 24-well plates. After overnight incubation, cells were transiently transfected with the TOPFlash/FOPFlash luciferase construct and β-galactosidase-expressing vector with the help of lipofectamine2000. After 24 h incubation, cells were treated with adenovirus-GFP or adenovirus-GFP-hepaCAM, respectively. Cells were then cultured for 24 hours and lysed. Subsequently both luciferase and β-galactosidase activities were determined. The luciferase activity was normalized to the β-galactosidase activity.

Cell proliferation assay

The T24 cells proliferation was analyzed by CCK-8 assay. Cells were seeded in 96 well plates at a density of 1000 cells per well and cultured for 12 h, then it was treated with adenovirus-GFP or adenovirus-GFP-hepaCAM alone or in combination with the GSK3β inhibitor LiCL or Wnt pathway inhibitor DKK1. Cells were kept in a humidified atmosphere of 5% CO2 at 37°C. At the indicated time points, each well was added with 10 ul CCK-8 reagent solution (Beyotime Institute of Biotechnology, Haimen, China), and incubated for 2h. Optical density was determined by a microplate reader at the absorbance of 450 nm (A450). All results from each indicated time points were repeated in triplicate.

Colony formation assay

Cells were seeded into 6-well plates at a density of 300 cells per well. After 48h incubation, cells were treated with adenovirus-GFP or adenovirus-GFP-hepaCAM alone or in combination with the GSK3β inhibitor LiCL or Wnt pathway inhibitor DKK1.After 15 days, cells were fixed with methanol and stained with 0.1% crystalviolet. Visible colonies were counted and taken photos under microscope.

Xenograft transplantation and immunohistochemistry

The T24 cells were infected with adenovirus-GFP, adenovirus-GFP-hepaCAM, without adenovirus were taken as control. After 48 h, cells (2 ×10⁶)were respectively harvested and suspended in 100ul PBS and then injected subcutaneously into nude mice at the side of the posterior flank. The weight of nude mice was evaluated every 6 days for 4 weeks. After 4 weeks, the xenograft tumor weight was measured at the terminal time. Then, all mice were sacrificed, transplanted tumors were excised and tumor tissues were used to performhematoxylin eosin (H&E) staining, and immunohistochemistry was performed to examine hepaCAM, p-β-catenin, Ki67, β-catenin expression. Methods were described previously. The protocols for in vivo animal experiment were approved by the Committee on Ethical Use of Animals of Chongqing Medical University.

Statistical Analysis

All statistical analysis was carried out using the SPSS 17.0 statistical software. Correlation between tumor hepaCAM and β-catenin was measured by using Spearman linear correlation analysis. Student's t-test, Chi-square test, the Fisher's exact test and one way-ANOVA were used to evaluate the associations among categorical variables. Data were shown as the Mean±SD and P < 0.05 was served as the criterion for statistical significance.

Funding

This work was supported by National Natural Science Foundation of PR China (Grant NO.81072086).

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.