Esophageal squamous cell carcinoma (ESCC) is increasing in incidence, but the knowledge of the genetic underpinnings of this disease remains limited. In this study, we identified the tetraspanin cell surface receptor uroplakin 1A (UPK1A) as a candidate tumor suppressor gene (TSG), and we investigated its function and mechanism in ESCC cells. UPK1A downregulation occurred in 68% of primary ESCCs examined, where it was correlated significantly with promoter hypermethylation (P < 0.05). Ectopic expression of UPK1A in ESCC cells inhibited cell proliferation, clonogenicity, cell motility, and tumor formation in nude mice. Mechanistic investigations suggested that these effects may be mediated by inhibiting nuclear translocation of β-catenin and inactivation of its downstream targets, including cyclin-D1, c-jun, c-myc, and matrix metalloproteinase 7 (MMP7). Cell cycle arrest elicited by UPK1A at the G1-S checkpoint was associated with downregulation of cyclin D1 and cyclin-dependent kinase 4, whereas metastasis suppression was associated with reduction of MMP7. These findings were consistent with evidence derived from clinical samples, where UPK1A downregulation was correlated with lymph node metastasis (P = 0.009), stage (P = 0.015), and overall survival (P < 0.0001). Indeed, multivariate cyclooxygenase regression analysis showed that UPK1A was an independent prognostic factor for overall survival. Taken together, our findings define a function for UPK1A as an important TSG in ESCC development. Cancer Res; 70(21); 8832–41. ©2010 AACR.

Esophageal cancer is one of the most common malignancies and has been ranked as the sixth leading cause of cancer death over the world (1). Throughout the world, the major type of esophageal cancer is esophageal squamous cell carcinoma (ESCC). The incidence ratio of ESCC in China varies widely compared with other common cancers, wherein the high- and low-risk regions can be as high as 500:1 (2). More than 50% of ESCC cases occurred in Asia, wherein Linzhou and the nearby countries in Henan province of Northern China have the highest incidence in the world (3, 4). Like other types of cancers, the development of ESCC is also believed as a multiple-step process caused by the accumulation of activation of oncogenes and inactivation of tumor suppressor genes (TSG). Studies also showed that there is a strong tendency toward familial aggregation in the Northern China, which suggests that genetic susceptibility may be involved in the etiology of ESCC (5). Hence, to have a better understanding of ESCC, it is important to identify these key genes, elucidate their roles, and discover new biomarkers for improving clinical management.

Downregulation is a common feature of a TSG, which can be caused by gene mutation, allele deletion, promoter hypermethylation, and posttrancriptional silencing by microRNA. Therefore, comparison of expressing profiles by cDNA microarray between tumor and nontumorous tissues and characterization of downregulated genes in tumor specimen is a useful strategy to identify TSGs. Recently, our group performed an Affemetrix cDNA microarray to compare differentially expressed genes between 10 pairs of ESCC tumors and their adjacent nontumor tissues. About 220 downregulated genes were identified including uroplakin 1A (UPK1A). Uroplakin family includes four transmembrane proteins, UPK1A (27 kDa), uroplakin 1B (28 kDa), uroplakin II (15 kDa), and uroplakin III (47 kDa), which are highly conserved in many organisms, such as human, pig, dog, mouse, and rabbit. Each of these integral membrane proteins has 16-nm particles forming two crystals hexagonally, which they constitute as the urothelial plates of the asymmetrical unit membrane in urothelium (68). UPK1A gene comprises eight exons and encodes a protein of 256 amino acids, which belongs to the transmembrane 4 superfamily (TM4SF), also known as the tetraspanin family (9, 10). At first, the expression of UPK1A was found to be highly specific to urothelium normal tissues. However, EST databases revealed that UPK1A could also be found in normal genitourinary tract, uterus, and prostate (11). In addition, during carcinogenesis, the expression of uroplakins decreased markedly and sometimes even totally disappeared in invasive carcinomas (12, 13). In the present study, we studied the UPK1A expression status and its promoter methylation in primary ESCCs and ESCC cell lines. Functional assays with a UPK1A reexpressing ESCC cell line were performed to characterize the biological effects of UPK1A in esophageal tumorigenicity, both in vitro and in vivo. Tumor-suppressive mechanism of UPK1A and its potential to be a new biomarker in ESCC were also addressed.

ESCC samples and cell lines

Primary ESCC tumor tissues were collected immediately after surgical resection at Linzhou Cancer Hospital. All patients did not receive preoperative treatment. Samples used in this study were approved by the committees for ethical review of research involving human subjects at Zhengzhou University and University of Hong Kong. The Chinese ESCC cell line HKESC1 was kindly provided by Professor Srivastava (Department of Pathology, University of Hong Kong), and the other two Chinese ESCC cell lines (EC18 and EC109) were kindly provided by Professor Tsao (Department of Anatomy, University of Hong Kong). Six Japanese ESCC cell lines (KYSE30, KYSE140, KYSE180, KYSE410, KYSE510, and KYSE520) were obtained from DSMZ, the German Resource Centre for Biological Material (14). All ESCC cell lines used in this study were regularly authenticated by checking the morphology and were tested for the absence of Mycoplasma contamination (MycoAlert, Lonza). Cells were cultivated as previously described (15).

Semiquantitative reverse transcription-PCR

Total RNA was extracted from frozen ESCC tissues and cell lines by the TRIzol reagent (Invitrogen), and 2 μg of total RNA were reverse-transcribed using Advantage reverse transcription-PCR (RT-PCR) kit (Clontech Laboratories, Inc.) for first-strand complementary DNA synthesis. RT-PCR was carried out with primers for UPK1A, cyclin D1, matrix metalloproteinase 7 (MMP7), c-jun, E-cadherin, and c-myc (Supplementary Table S1).

Tissue microarrays and immunohistochemistry

A total of 300 formalin-fixed and paraffin-embedded ESCC tumor specimens were kindly provided by Linzhou Cancer Hospital. According to the method described previously (16), tissue microarrays (TMA) containing 300 pairs of primary ESCC tumor samples and their corresponding nontumorous tissues (duplicate 0.6-mm tissue cores for each ESCC) were constructed.

Immunohistochemical staining was carried out following standard streptavidin-biotin-peroxidase complex method (16). Briefly, TMA sections were deparaffinized, and nonspecific bindings were blocked with 10% normal goat serum for 10 minutes. The TMA section was then incubated with anti-UPK1A polyclonal antibody (Abcam, 1:200 dilution) or anti-MMP7 monoclonal antibody (Santa Cruz Biotechnology, 1:50 dilution) at 4°C overnight. Slides were then incubated with biotinylated goat anti-rabbit or anti-mouse immunoglobulin (Santa Cruz Biotechnology, 1:100 dilution) at room temperature for 30 minutes.

5-Aza-2′-deoxygcitidine treatment

To study whether demethylation could restore UPK1A expression in KYSE510 cells, 2 × 105 cells were treated with 50 μmol/L 5-aza-dC (Sigma-Aldrich Corporation) for 3 days, changing the 5-aza-dC and the medium every 24 hours. Total RNA was then extracted, and UPK1A expression was detected by RT-PCR.

Bisulfite modification and promoter methylation analysis

Genomic DNA was treated by bisulfite and then studied by methylation-specific PCR (MSP) as previously described (15). MSP primers are listed in Supplementary Table S1.

Tumor-suppressive function of UPK1A

To test the tumor-suppressive function, UPK1A was cloned into pcDNA3.1 vector (Invitrogen) and transfected into the ESCC cell lines KYSE30 and KYSE510. Stable UPK1A-expressing clones (UPK1A-30 or UPK1A-510) were selected for further study. Empty vector–transfected cells (Vec-30 or Vec-510) were used as control. Cell proliferation assay, foci formation assay, and colony formation in soft agar were carried out as described previously (15). For in vivo tumorigenicity assay, 2 × 106UPK1A-expressing cells (UPK1A-30 or UPK1A-510) and control cells (Vec-30 or Vec-510) cells were injected s.c. into the right and left dorsal flank, respectively, of 4- to 5-week-old nude mice (10 mice per group). Tumor formation in nude mice was monitored by measuring the tumor volume, which was calculated by the formula, V = 0.5 × L × W2, over a 4-week period. To visualize the tumor structure, sections (5 μmol/L) of a paraffin-embedded tumor were stained with H&E. All animal experiments were approved by the University of Hong Kong Committee on the Use of Live Animals in Teaching and Research (CULATR).

Wound healing and invasion assays

For wound healing assay, the cell layer was wounded using a sterile tip. The spread of wound closure was observed after 24 and 48 hours and photographed under a microscope. Invasion assay was performed with a chamber containing a polycarbonate membrane (8-μm pore size) and coated with a layer of extracellular matrix (Chemicon International) according to the manufacturer's instructions. The number of cells that invaded through the Matrigel was counted in 10 fields under 20× objective lens. The experiments were repeated three times.

In vivo metastasis assay

All animal procedures were performed with the approval of CULATR. Briefly, 2 × 105 cells (four groups, including Vec-30, UPK1A-30, Vec-510, and UPK1A-510) were injected i.v. through the tail vein into 4- to 5-week-old severe combined immunodeficient (SCID)-Beige mice (six mice per group). After 12 weeks, mice were euthanized. The number of tumor nodules formed on the liver surfaces was then counted. The livers and lungs were excised and embedded in paraffin for further study.

Flow cytometry

UPK1A-510 and Vec-510 cells (2 × 105) were fixed in 70% ethanol and stained with propidium iodide, and DNA content was analyzed by Cytomics FC (Beckman Coulter). The results were analyzed by ModFit LT2.0 software.

Immunofluoroscence

UPK1A-510 and Vec-510 cells were grown on gelatin-coated coverslips, fixed with 4% paraformaldehyde, permeabilized in PBS, which contain 0.1% Triton-X 100, and blocked with 1% bovine serum albumin. The cells were then treated with antibodies targeting β-catenin (Cell Signaling Technology) and UPK1A (Abcam) at 4°C overnight. After washing with PBS, cells were incubated with FITC-conjugated antimouse secondary antibody and Texas red–conjugated antirabbit secondary antibody at room temperature for 1 hour. Antifade 4′,6-diamidino-2-phenylindole (DAPI) solution was added, and images were captured.

Western blot analysis

Western blotting was done according to the standard protocol with antibodies UPK1A (Abcam), UPK1A (Santa Cruz Biotechnology), β-catenin (Cell Signaling Technology), E-cadherin (Santa Cruz Biotechnology), and cyclin-D1 (Cell Signaling Technology). β-Actin (Santa Cruz Biotechnology) was used as loading control.

Statistical analysis

Statistical analysis was carried out using Statistical Package for Social Sciences 14.0 for Windows (SPSS, Inc.). χ2 test or Fisher's exact test was used to analyze the association of UPK1A gene expression and clinicopathologic parameters. Student's t test was used to analyze the results expressed as mean ± SD. The overall survival curve was plotted by using Kaplan-Meier analysis. The association of UPK1A gene expression, stage, and lymph node status was examined by univariable and multivariable Cox proportional hazard regression model (17). Differences were considered significant when P value was <0.05.

UPK1A is frequently downregulated in ESCC

Semiquantitative RT-PCR was used to study the expression of UPK1A in nine ESCC cell lines and 100 pairs of primary ESCC tumors and their corresponding nontumorous tissues. The results showed that absent expression of UPK1A was detected in 68 of 100 primary ESCC tissues (Fig. 1A) and five of nine ESCC cell lines (Fig. 1B). Sequencing analysis was performed in five primary ESCCs, and no mutation was detected (data not shown). UPK1A expression in protein level was also studied by immunohistochemical staining using a TMA containing 300 pairs of primary ESCC tumor samples and their corresponding nontumorous tissues. Informative immunohistochemical results were obtained from 186 (62%) pairs of ESCCs. Noninformative samples included lost samples, unrepresentative samples, samples with too few tumor cells, and samples with inappropriate staining; such were not used in data complication. The immunohistochemical analysis showed that expression of UPK1A was detected in all nontumorous tissues. Absent expression of UPK1A was observed in 104 of 186 (56%) ESCC specimens (Fig. 1C).

Figure 1.

Downregulation of UPK1A in ESCC. A, downregulation of UPK1A was frequently detected in primary ESCCs by RT-PCR. N, nontumorous tissue; T, tumor tissue. B, absent expression of UPK1A in primary ESCC was detected in five ESCC cell lines (KYSE30, KYSE140, KYSE180, KYSE-510, and HKESC1). N, pool of normal tissues. C, representatives of immunostaining with anti-UPK1A antibody in an ESCC case. Positive staining (brown) was only detected in normal esophageal tissue. The slide was counterstained with hematoxylin. D, Kaplan-Meier overall survival plot. UPK1A(+), patients with UPK1A expression; UPK1A(-), patients without UPK1A expression.

Figure 1.

Downregulation of UPK1A in ESCC. A, downregulation of UPK1A was frequently detected in primary ESCCs by RT-PCR. N, nontumorous tissue; T, tumor tissue. B, absent expression of UPK1A in primary ESCC was detected in five ESCC cell lines (KYSE30, KYSE140, KYSE180, KYSE-510, and HKESC1). N, pool of normal tissues. C, representatives of immunostaining with anti-UPK1A antibody in an ESCC case. Positive staining (brown) was only detected in normal esophageal tissue. The slide was counterstained with hematoxylin. D, Kaplan-Meier overall survival plot. UPK1A(+), patients with UPK1A expression; UPK1A(-), patients without UPK1A expression.

Close modal

UPK1A downregulation is associated with ESCC metastasis and poor prognosis

Clinical association study found that the downregulation of UPK1A was significantly associated with advanced clinical stage (P = 0.015, Fisher's exact test) and lymph node metastasis (P = 0.009, Fisher's exact test; Table 1). No significant association was observed between downregulation of UPK1A and other clinicopathologic parameters (Table 1). Kaplan-Meier analysis found that the downregulation of UPK1A was significantly associated with poor overall survival (P < 0.0001; Fig. 1D). By univariable analysis, downregulation of UPK1A (P < 0.0001), presence of lymph node metastasis (P < 0.0001), and advanced stage (P < 0.0001) were significant negative prognostic factors for overall survival of ESCC patients. Nevertheless, multivariable Cox proportional hazard regression analysis showed that the downregulation of UPK1A (P = 0.004) was the only independent factor for the prediction of overall survival (Supplementary Table S2).

Table 1.

Association of UPK1A downregulation with clinicopathologic features of 186 ESCC patients

Clinicopathologic featuresTotal case (%)UPK1A expression
Negative (%)Positive (%)P
Sex 
    Male 44.62 48.19 51.81 0.159 
    Female 55.38 60.19 39.81  
Age 
    ≦57 45.70 55.29 44.71 0.443 
    >57 54.30 57.43 42.57  
Clinical stage 
    Early stage (I–II) 71.50 51.13 48.87 0.015* 
    Advanced stage (III–IV) 28.50 69.81 30.19  
Lymph node status 
    N0 60.21 49.11 50.89 0.009* 
    N+ 39.79 67.57 32.43  
Clinicopathologic featuresTotal case (%)UPK1A expression
Negative (%)Positive (%)P
Sex 
    Male 44.62 48.19 51.81 0.159 
    Female 55.38 60.19 39.81  
Age 
    ≦57 45.70 55.29 44.71 0.443 
    >57 54.30 57.43 42.57  
Clinical stage 
    Early stage (I–II) 71.50 51.13 48.87 0.015* 
    Advanced stage (III–IV) 28.50 69.81 30.19  
Lymph node status 
    N0 60.21 49.11 50.89 0.009* 
    N+ 39.79 67.57 32.43  

*Significant difference.

UPK1A promoter region is frequently hypermethylated in ESCC

A 524-bp CpG island containing 28 CpG sites was found on the 5′ upstream of the UPK1A gene, which is located within a 2147-bp region reported to have the promoter activity (18). To explore the effect of promoter methylation on UPK1A downregulation in ESCC, MSP using methylation- or unmethylation-specific primers was carried out to analyze the methylation status of UPK1A. The result showed that both methylated and unmethylated alleles were detected in six of nine cell lines (Fig. 2A). In the remaining three cell lines, either methylated allele (KYSE510 and HKESC1) or unmethylated allele (KYSE410) was detected (Fig. 2A). MSP was also used to investigate the methylation frequency of UPK1A CGI in 50 primary ESCC tumors. The result showed that methylated allele was detected in 31 of 50 (62%) primary ESCCs (Fig. 2A). The frequency of methylation in ESCC with UPK1A downregulation (27 of 41, 65.9%) was obviously higher that that in ESCC with normal UPK1A expression (four of nine, 44.4%). To further evaluate whether the methylation of UPK1A directly mediates its repression, KYSE510 cells were treated with demethylating agent 5-aza-dC. After the treatment, transcription of UPK1A was restored (Fig. 2B).

Figure 2.

Tumor-suppressive function of UPK1A in ESCC cells. A, methylation-specific PCR (M, methylated allele; U, unmethylated allele) was performed to investigate the methylation status in the promoter region of UPK1A in ESCC cell lines (top) and primary ESCCs (bottom). N, nontumorous tissue; T, tumor tissue. B, restoration of UPK1A expression in KYSE510 cells after 5-aza-dC (Aza) treatment. +, Aza treated; -, untreated. C, UPK1A expression in stably UPK1A-transfected KYSE30 (UPK1A-30) and KYSE510 (UPK1A-510) cells detected by RT-PCR. Empty vector–transfected cells (Vec-30 and Vec-510) were used as control. D, the efficiency of foci formation was significantly inhibited by UPK1A (left). The results are expressed as mean ± SD of three independent experiments. **, P < 0.001. E, the frequency of colony formation in soft agar was significantly higher in Vec-510 cells compared with UPK1A-510 cells. The results are expressed as mean ± SD of three independent experiments. **, P < 0.001. F, XTT assay showed that cell growth rates in 30-UPK1A (left) and 510-UPK1A (right) cells were inhibited by UPK1A. The results are expressed as mean ± SD of three independent experiments. *, P < 0.05.

Figure 2.

Tumor-suppressive function of UPK1A in ESCC cells. A, methylation-specific PCR (M, methylated allele; U, unmethylated allele) was performed to investigate the methylation status in the promoter region of UPK1A in ESCC cell lines (top) and primary ESCCs (bottom). N, nontumorous tissue; T, tumor tissue. B, restoration of UPK1A expression in KYSE510 cells after 5-aza-dC (Aza) treatment. +, Aza treated; -, untreated. C, UPK1A expression in stably UPK1A-transfected KYSE30 (UPK1A-30) and KYSE510 (UPK1A-510) cells detected by RT-PCR. Empty vector–transfected cells (Vec-30 and Vec-510) were used as control. D, the efficiency of foci formation was significantly inhibited by UPK1A (left). The results are expressed as mean ± SD of three independent experiments. **, P < 0.001. E, the frequency of colony formation in soft agar was significantly higher in Vec-510 cells compared with UPK1A-510 cells. The results are expressed as mean ± SD of three independent experiments. **, P < 0.001. F, XTT assay showed that cell growth rates in 30-UPK1A (left) and 510-UPK1A (right) cells were inhibited by UPK1A. The results are expressed as mean ± SD of three independent experiments. *, P < 0.05.

Close modal

UPK1A has tumor-suppressive ability

To investigate whether UPK1A has tumor-suppressive ability, UPK1A was cloned into an expressing vector and stably transfected into two ESCC cell lines, KYSE30 and KYSE510. Expression of UPK1A in UPK1A-30 and UPK1A-510 cells was confirmed by RT-PCR (Fig. 2C). Protein expression of UPK1A was also tried to be confirmed by using antibodies from two companies. However, they failed to give a specific band. To confirm the protein expression of UPK1A in the clone, immunofluoroscence has been done (Supplementary Fig. S1). With UPK1A-expressing clones, we showed tumor-suppressive ability by cell proliferation assay, foci formation assay, and soft agar assay. The results showed that the efficiencies of foci formation (Fig. 2D) and colony formation in soft agar (Fig. 2E) were significantly inhibited in UPK1A-510 cells compared with Vec-510 cells (P < 0.001). With XTT [2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide] cell growth assay, cell growth rates of UPK1A-30 and UPK1A-510 clones were also significantly suppressed compared with Vec-30 and Vec-510 cells (P < 0.05; Fig. 2F).

To confirm the in vivo tumor-suppressive ability of UPK1A, tumor formation in nude mice was performed by injecting UPK1A-30 cells (n = 10) or UPK1A-510 cells (n = 10), and Vec-30 or Vec-510 cells were used as controls. The results showed that tumor formation in nude mice was significantly inhibited in UPK1A-expressing cells (P < 0.01; Fig. 3A and B). With immunohistochemical staining using anti-UPK1A antibody, we confirmed that UPK1A was expressed in UPK1A-30– or UPK1A-510–derived tumors (Fig. 3C). In addition, a clear boundary between the tumor and its adjacent nontumor tissue was observed in UPK1A-30– and UPK1A-510–derived tumors (Fig. 3D). However, venous infiltration and irregular tumor invasion was observed in the Vec-30– and Vec-510–generated tumors, respectively (Fig. 3D). These data showed that UPK1A had strong tumor-suppressive ability.

Figure 3.

UPK1A inhibits tumor formation in nude mice. A, representatives of tumors formed in nude mice induced by vector-transfected cells (left dorsal flank, indicated by arrows) and UPK1A-transfected cells (right dorsal flank). B, UPK1A effectively suppressed tumorigenicity in nude mice. Points, mean of 10 mice. Bars, SD. **, P < 0.001. C, immunohistochemical staining was performed to confirm the expression of UPK1A in the tumor induced by UPK1A-transfected cells (right) but not in the tumor induced by vector-transfected cells. (magnification, 200×). D, representative of H&E staining shows venous infiltration and irregular tumor invasion in Vec-30–generated (left, indicated by an arrow) and Vec-510–generated tumor, respectively, as well as a clear boundary between the tumor and its adjacent nontumor tissue in UPK1A-30– and UPK1A-510–derived tumor (right).

Figure 3.

UPK1A inhibits tumor formation in nude mice. A, representatives of tumors formed in nude mice induced by vector-transfected cells (left dorsal flank, indicated by arrows) and UPK1A-transfected cells (right dorsal flank). B, UPK1A effectively suppressed tumorigenicity in nude mice. Points, mean of 10 mice. Bars, SD. **, P < 0.001. C, immunohistochemical staining was performed to confirm the expression of UPK1A in the tumor induced by UPK1A-transfected cells (right) but not in the tumor induced by vector-transfected cells. (magnification, 200×). D, representative of H&E staining shows venous infiltration and irregular tumor invasion in Vec-30–generated (left, indicated by an arrow) and Vec-510–generated tumor, respectively, as well as a clear boundary between the tumor and its adjacent nontumor tissue in UPK1A-30– and UPK1A-510–derived tumor (right).

Close modal

UPK1A arrests cell cycle at G1-S checkpoint by downregulating cyclin D1

To understand how UPK1A suppresses tumor cell growth, flow cytometry was carried out to compare the DNA content between Vec-510 and UPK1A-510 cells. The result showed that the proportion of S-phase cells was significantly lower in UPK1A-510 than that in Vec-510 cells (P < 0.05), suggesting that UPK1A is able to arrest cells in at G1-S checkpoint (Fig. 4A and B). Further study found that G1-S checkpoint promoting factors cyclin D1 and cyclin-dependent kinase 4 (CDK4) were downregulated in UPK1A-510 cells compared with Vec-510 cells (Fig. 4C).

Figure 4.

A, flow cytometry was used to compare the DNA content between Vec-510 and UPK1A-510 cells. B, summary of cell proportions in different phases of cell cycle. The results are expressed as mean ± SD of three independent experiments. *, P < 0.05. C, downregulations of cyclin D1, CDK4, and E-cadherin were detected by Western blot analysis in UPK1A-expressing cells. β-Actin was used as loading control. D, representative images of immunofluorescence with anti-β-catenin antibody. β-Catenin (green color) was mainly located in nucleus in Vec-510 cells, but mainly in cell membrane in UPK1A-510 cells. DNA was counterstained by DAPI (blue color). E, expressions of downstream targets of β-catenin were compared by RT-PCR between Vec-510 and UPK1A-510 cells. 18S rRNA was used as internal control. F, expression of MMP7 was compared between tumors formed in nude mice generated from Vec-510 and UPK1A-510 cells by immunohistochemistry using an anti-MMP7 antibody. Positive staining (brown) was observed in tumor derived from Vec-510 cells.

Figure 4.

A, flow cytometry was used to compare the DNA content between Vec-510 and UPK1A-510 cells. B, summary of cell proportions in different phases of cell cycle. The results are expressed as mean ± SD of three independent experiments. *, P < 0.05. C, downregulations of cyclin D1, CDK4, and E-cadherin were detected by Western blot analysis in UPK1A-expressing cells. β-Actin was used as loading control. D, representative images of immunofluorescence with anti-β-catenin antibody. β-Catenin (green color) was mainly located in nucleus in Vec-510 cells, but mainly in cell membrane in UPK1A-510 cells. DNA was counterstained by DAPI (blue color). E, expressions of downstream targets of β-catenin were compared by RT-PCR between Vec-510 and UPK1A-510 cells. 18S rRNA was used as internal control. F, expression of MMP7 was compared between tumors formed in nude mice generated from Vec-510 and UPK1A-510 cells by immunohistochemistry using an anti-MMP7 antibody. Positive staining (brown) was observed in tumor derived from Vec-510 cells.

Close modal

UPK1A inhibits membrane-to-nucleus translocation of β-catenin

We next studied the involvement of E-cadherin/β-catenin signaling pathway in the upregulation of cyclin D1 and CDK4. Western blotting showed that E-cadherin was upregulated in UPK1A-expressing cells compared with empty vector–transfected cells (Fig. 4C). However, no difference was detected for β-catenin. Because E-cadherin is an important component to hold β-catenin in the membrane, we examined the location of β-catenin by immunofluoresence. The result showed that β-catenin was mainly localized in nucleus of Vec-510 cells whereas mainly localized in cell membrane of UPK1A-510 cells, suggesting that UPK1A could inhibit the translocation of β-catenin from membrane to nucleus (Fig. 4D).

We next examined some downstream target genes of β-catenin, including MMP7, c-jun, and c-myc, by RT-PCR. The results showed that all tested genes were downregulated in UPK1A-510 cells compared with Vec-510 cells. However, the mRNA level of E-cadherin has no change (Fig. 4E). In addition, expression of MMP7 between tumors formed in nude mice generated by injecting Vec-510 or UPK1A-510 was compared by immunohistochemical staining using anti-MMP7 antibody. The result showed that the expression of MMP7 in UPK1A-510–generated tumors was lower compared with Vec-510–generated tumors (Fig. 4F).

UPK1A suppresses tumor invasion and metastasis

To validate whether UPK1A can suppress metastasis, wound healing and cell invasion assays were performed to study cell motility. Wound healing assay showed that cell migration rate was dramatically reduced in UPK1A-expressing cells compared with empty vector–transfected cells (Fig. 5A). Cell invasion assay found that the number of invasive cells was significantly decreased in UPK1A-expressing cells compared with empty vector–transfected cells (P < 0.01; Fig. 5B).

Figure 5.

UPK1A inhibits cancer cell invasion and metastasis. A, wound healing assay showed that cell motility was inhibited by UPK1A. Representative of images were photographed at time 0, 24, and 48 h after scratching. B, Matrigel invasion assay was used to compare invisibility between vector- and UPK1A-transfected cells. The cells that invaded through the Matrigel were fixed and stained with crystal violet (magnification, 200×). The results are expressed as mean ± SD of three independent experiments. **, P < 0.001. C, representatives of livers derived from SCID mice after tail vein injection of Vec-510 or UPK1A-510 cells. The metastatic nodules at liver surface are indicated by arrows. The summary of metastatic nodules at liver surface is mean of six SCID mice for each group. Bars, SD. **, P < 0.001. D, representatives of H&E staining show normal liver tissue (left) and liver cancer (right) observed in SCID mice injected with UPK1A-510 and Vec-510 cells, respectively. E, representatives of H&E staining show normal lung tissue (left) and lung cancer (right) observed in SCID mice injected with UPK1A-510 and Vec-510 cells, respectively.

Figure 5.

UPK1A inhibits cancer cell invasion and metastasis. A, wound healing assay showed that cell motility was inhibited by UPK1A. Representative of images were photographed at time 0, 24, and 48 h after scratching. B, Matrigel invasion assay was used to compare invisibility between vector- and UPK1A-transfected cells. The cells that invaded through the Matrigel were fixed and stained with crystal violet (magnification, 200×). The results are expressed as mean ± SD of three independent experiments. **, P < 0.001. C, representatives of livers derived from SCID mice after tail vein injection of Vec-510 or UPK1A-510 cells. The metastatic nodules at liver surface are indicated by arrows. The summary of metastatic nodules at liver surface is mean of six SCID mice for each group. Bars, SD. **, P < 0.001. D, representatives of H&E staining show normal liver tissue (left) and liver cancer (right) observed in SCID mice injected with UPK1A-510 and Vec-510 cells, respectively. E, representatives of H&E staining show normal lung tissue (left) and lung cancer (right) observed in SCID mice injected with UPK1A-510 and Vec-510 cells, respectively.

Close modal

To study the metastatic effect of UPK1A in vivo, experimental metastasis assay was performed by comparing the metastatic nodules formed in the liver and lung of SCID mice. After 12 weeks, the mice were sacrificed, and metastatic nodules were counted in the surface of lung and liver. The number of nodules formed in liver was significantly higher in mice injected with Vec-510 cells (24.2 ± 4.4) than mice injected with UPK1A-510 (3.3 ± 1.2, P < 0.001, independent Student's t test; Fig. 5C). No visible metastatic nodules were observed in the surface of lungs. H&E staining was performed on serial sections of liver and lung, and the results showed that metastatic nodules were not only detected in liver (Fig. 5D) but also in lung (Fig. 5E).

UPK1A is an integral protein, which belongs to TM4SF family, and is thought to be specific to normal urothelium (7, 11). Although several TM4SF molecules have been identified and implicated in the regulation of cell development, differentiation, proliferation, motility, and tumor cell invasion (11), the biological function of UPK1A is largely unrevealed. In the present study, downregulation of UPK1A was frequently detected in ESCCs in both mRNA and protein levels. Further study found that hypermethylation played a crucial role in the inactivation of UPK1A, which was shown by restoring UPK1A expression with dementhylation treatment.

The tumor-suppressive function of UPK1A was shown by both in vitro and in vivo assays. Ectopic expression of UPK1A in ESCC cell lines KYSE30 and KYSE510 could effectively suppress cell growth rate, colony formation in soft agar, foci formation, and tumor formation in nude mice. Our studies also showed the ability of UPK1A to suppress cell motility and invasion. Ectopic expression of UPK1A in ESCC cell lines was able to inhibit cell migration and invasion. In vivo animal model showed that UPK1A could reduce tumor nodule formation in lung and liver, which was consistent with our TMA results, showing that the downregulation of UPK1A was significantly associated with lymph node metastasis in ESCC (P = 0.009).

Molecular study found that the tumor-suppressive effect of UPK1A was closely associated with its role in cell cycle arrest at G1-S checkpoint. Ectopic expression of UPK1A could downregulate cyclin D1 and CDK4 expression. Activation of cyclin D-CDK4/6 complex, resulting in the cyclin D-CDK4/6–medicated Rb phosphorylation, can destruct Rb-E2F binding. The releasing E2F activates the transcription of genes necessary for S-phase entry and cell cycle progression (19, 20). Further study found that UPK1A downregulated E-cadherin and subsequently held β-catenin in the membrane. β-Catenin maintained at low level in quiescent cells by interacting with protein kinases, adenomatous polyposis coli and axin, casein kinase 1, and glycogen synthase kinase 3 (GSK3; ref. 21). However, when Wnt is present, it will inhibit the GSK3, resulting in the accumulation of β-catenin, which is then translocated to nucleus and causes carcinogenesis (22). E-cadherin, which contains the β-catenin binding site, is another important molecule to hold β-catenin in membrane. The loss of E-cadherin will cause the nuclear translocation of β-catenin (23). In the ESCC cell line KYSE510, β-catenin was mainly expressed in the nucleus. However, when UPK1A was transfected into the KYSE510 cells, E-cadherin expression was upregulated and β-catenin was mainly expressed in cell membrane. These data suggest that UPK1A is able to hold β-catenin in the membrane through the upregulation of E-cadherin. In addition, because the protein level of E-cadherin is upregulated in UPK1A-expressing cells, but not in mRNA level, it is possible that UPK1A may reduce E-cadherin ubiquitination. Nevertheless, more lines of evidence are required to confirm this conclusion. The influence of β-catenin translocation was further confirmed by examining the expression of its downstream targets, including cyclin-D1, c-jun, c-myc, and MMP7. In UPK1A-expressing cells, expressions of cyclin-D1, c-jun, c-myc, and MMP7 were all downregulated.

Another target of β-catenin is MMP7, which plays an important role in cancer invasion and metastasis (24, 25). In the in vivo model, we found that UPK1A can inhibit the metastasis in lung and liver. In nude mice model, we also found that tumors induced by empty vector–transfected cells showed a higher level of MMP7 expression compared with tumors induced by UPK1A-transfected cells. In addition, a clear boundary between the tumor and its adjacent nontumor tissue was often observed in the tumors induced by UPK1A-30 and UPK1A-510 cells; however, irregular tumor invasion was frequently observed in tumors induced by Vec-30 and Vec-510 cells (Fig. 3C). The results further implied that the inhibiting effect of UPK1A on cancer metastasis might be through the downregulation of MMP7.

Although different treatment methods have been developed, the survival rate of ESCC patients remains unsatisfactory (26). Therefore, it is necessary to find novel risk markers to guide disease management of ESCC patients for the improvement of their survival. By Kaplan-Meier analysis, downregulation of UPK1A in ESCC patients was significantly associated with poor survival (P < 0.0001), suggesting that UPK1A may be a novel marker to predict overall survival. To further validate which of factors (UPK1A expression, lymph node metastasis, or stage) is an independent factor for predicting the overall survival, univariable and multivariable Cox proportional hazard regression analysis was performed. The result showed that UPK1A expression is an independent factor and the most influential factor in predicting overall survival. In conclusion, our findings indicated that UPK1A is a potent TSG and plays an important role in inhibiting cell proliferation and metastasis.

No potential conflicts of interest were disclosed.

Grant Support: Research Council grant HKU 7393/04M, Hong Kong Research Grant Council Central Allocation grant HKU 1/06C, Sun Yat-Sen University “Hundred Talents Program” grant 85000-3171311, and Major State Basic Research Program of China grant 2006CB910104.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Pisani
P
,
Parkin
DM
,
Ferlay
J
. 
Estimates of the worldwide mortality from eighteen major cancers in 1985: implications for prevention and projections of future burden
.
Int J Cancer
1993
;
55
:
891
903
.
2
Li
JY
. 
Epidemiology of esophageal cancer in China
.
Natl Cancer Inst Monogr
1982
;
162
:
113
20
.
3
Parkin
DM
,
Bray
FI
,
Devesa
SS
. 
Cancer burden in the year 2000. The global picture
.
Eur J Cancer
2001
;
37
:
S4
66.2
.
4
Ke
L
. 
Mortality and incidence trends from esophagus cancer in selected geographic areas of China circa 1970-90
.
Int J Cancer
2002
;
102
:
271
4
.
5
Zhang
W
,
Bailey-Wilson
JE
,
Li
W
, et al
. 
Segregation analysis of esophageal cancer in a moderately high-incidence area of northern China
.
Am J Hum Genet
2000
;
67
:
110
9
.
6
Lin
JH
,
Wu
XR
,
Kreibich
G
,
Sun
TT
. 
Precursor sequence, processing, and urothelium-specific expression of a major 15-kDa protein subunit of asymmetric unit membrane
.
J Biol Chem
1994
;
269
:
1775
84
.
7
Wu
XR
,
Manabe
M
,
Yu
J
,
Sun
TT
. 
Large scale purification and immunolocalization of bovine uroplakins I, II, and III. Molecular markers of urothelial differentiation
.
J Biol Chem
1990
;
265
:
19170
9
.
8
Wu
XR
,
Sun
TT
. 
Molecular cloning of a 47 kDa tissue-specific and differentiation-dependent urothelial cell surface glycoprotein
.
J Cell Sci
1993
;
106
:
31
43
.
9
Wu
XR
,
Medina
JJ
,
Sun
TT
. 
Selective interactions of UPIa and UPIb, two members of the transmembrane 4 superfamily, with distinct single transmembrane-domained proteins in differentiated urothelial cells
.
J Biol Chem
1995
;
270
:
29752
9
.
10
Liang
FX
,
Riedel
I
,
Deng
FM
, et al
. 
Organization of uroplakin subunits: transmembrane topology, pair formation and plaque composition
.
Biochem J
2001
;
355
:
13
8
.
11
Maecker
HT
,
Todd
SC
,
Levy
S
. 
The tetraspanin superfamily: molecular facilitators
.
FASEB J
1997
;
11
:
428
42
.
12
Olsburgh
J
,
Harnden
P
,
Weeks
R
, et al
. 
Uroplakin gene expression in normal human tissues and locally advanced bladder cancer
.
J Pathol
2003
;
199
:
41
9
.
13
Ogawa
K
,
St John
Margaret
,
Oliveira
M
, et al
. 
Comparison of uroplakin expression during urothelial carcinogenesis induced by N-dButyl-B-(4_hydroxybutyl)Nitrosamine in rats and mice
.
Toxicol Pathol
1999
;
27
:
645
51
.
14
Shimada
Y
,
Imamura
M
,
Wagata
T
,
Yamaguchi
N
,
Tobe
T
. 
Characterization of 21 newly established esophageal cancer cell lines
.
Cancer
1992
;
69
:
277
84
.
15
Fu
L
,
Qin
YR
,
Xie
D
, et al
. 
Characterization of a novel tumor-suppressor gene PLCδ1 at 3p22 in esophageal squamous cell carcinoma
.
Cancer Res
2007
;
67
:
10720
6
.
16
Xie
D
,
Sham
JS
,
Zeng
WF
, et al
. 
Heterogeneous expression and association of β-catenin, p16 and c-myc in multistage colorectal tumorigenesis and progression detected by tissue microarray
.
Int J Cancer
2003
;
107
:
896
902
.
17
Cox
DR
. 
Regression models and life tables
.
J R Stat Soc
1972
;
34
:
187
220
.
18
Hall
GD
,
Weeks
RJ
,
Olsburgh
J
, et al
. 
Transcriptional control of the human urothelial-specific gene, uroplakin Ia
.
Biochim Biophys Acta
2005
;
1729
:
126
34
.
19
Lu
Z
,
Ghosh
S
,
Wang
Z
,
Hunter
T
. 
Downregulation of caveolin-1 function by EGF leads to the loss of E-cadherin, increased transcriptional activity of β-catenin, and enhanced tumor cell invasion
.
Cancer Cell
2003
;
4
:
499
515
.
20
Sherr
CJ
,
Roberts
JM
. 
CDK inhibitors: positive and negative regulators of G1-phase progression
.
Genes Dev
1999
;
13
:
1501
12
.
21
Noort
M
,
Meeldijk
J
,
van der Zee
R
,
Destree
O
,
Clevers
H
. 
Wnt signaling controls the phosphorylation status of β-catenin
.
J Biol Chem
2002
;
277
:
17901
5
.
22
Taurin
S
,
Sandbo
N
,
Qin
Y
,
Browning
D
,
Dulin
NO
. 
Phosphorylation of β-catenin by cyclic AMP-dependent protein kinase
.
J Biol Chem
2006
;
281
:
9971
6
.
23
Thiery
JP
. 
Epithelial-mesenchymal transitions in tumour progression
.
Nat Rev Cancer
2002
;
2
:
442
54
.
24
Fang
YJ
,
Lu
ZH
,
Wang
GQ
, et al
. 
Elevated expressions of MMP7, TROP2, and survivin are associated with survival, disease recurrence, and liver metastasis of colon cancer
.
Int J Colorectal Dis
2009
;
24
:
875
84
.
25
Yue
W
,
Sun
Q
,
Landreneau
R
, et al
. 
Fibulin-5 suppresses lung cancer invasion by inhibiting matrix metalloproteinase-7 expression
.
Cancer Res
2009
;
69
:
6339
46
.
26
Rubinfeld
R
,
Albert
I
,
Porfiri
F
, et al
. 
Binding of GSK3β to the APC-β-catenin complex and regulation of complex assembly
.
Science
1996
;
272
:
2023
6
.

Supplementary data