SEARCH

SEARCH BY CITATION

Keywords:

  • FRAT1;
  • ESCC;
  • β-catenin;
  • c-Myc;
  • overexpression

Abstract

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Esophageal squamous cell carcinoma (ESCC) is an aggressive tumor with a poor prognosis. Although aberrant activation of β-catenin/T-cell factor (TCF) pathway has been observed in ESCC, mechanisms underlying this phenomenon remain unknown. Frequently rearranged in advanced T-cell lymphomas-1 (FRAT1), overexpressed in some ESCC lines, is a positive regulator of β-catenin/TCF pathway. However, little is known about the molecular relationship between FRAT1 and β-catenin/TCF in ESCC. In this study, we analyzed freshly resected ESCC specimens and demonstrated that FRAT1 was overexpressed in approximately 74% of tumor samples compared with matched normal tissue. Overexpression of FRAT1 significantly promoted esophageal cancer cells growth, whereas suppression of FRAT1 level by RNAi markedly inhibited their growth. In addition, FRAT1 overexpression induced the nuclear accumulation of β-catenin and promoted the transcriptional activity of β-catenin/TCF. These effects were reversed by coexpression of GSK 3β or ΔN TCF4. Furthermore, accumulation of β-catenin was correlated with FRAT1 overexpression in ESCC and the basal layer of normal esophageal epithelium. Finally, continued expression of c-Myc is necessary and sufficient for maintenance of the growth state in cells expressing FRAT1. Taken together, these results support the novel hypothesis that aberrant activation of β-catenin/TCF pathway in esophageal cancer appears to be due to upstream events such as FRAT1 overexpression, and c-Myc may be an important element in oncogenesis of human ESCC induced by FRAT1. © 2008 Wiley-Liss, Inc.

Human esophageal cancer, an aggressive tumor with a poor prognosis, occurs at a high frequency rate in Asia (especially certain areas of China) and South America.1 More than 90% of esophageal cancers worldwide are esophageal squamous cell carcinomas (ESCC).2 Despite recent advances in therapy and management, esophageal cancer remains a killer. Even in the developed world more than 85% of patients die within 2 years of diagnosis.3 In China, the esophageal cancer mortality rate is ranked fourth among cancer-related deaths.4

Improvement in the efficacy of esophageal cancer treatment is a major public health goal. New therapies based on a better understanding of the biology of esophageal cancer are necessary. The molecular pathogenesis of esophageal cancer includes alterations of expression and function of multiple genes including dominant oncogenes, recessive tumor-suppressor genes and abnormalities in growth-regulatory signaling pathways.2, 5, 6

Recently, accumulation of nuclear and cytoplasmic β-catenin has been observed in ESCC.7–11 It is involved in β-catenin/TCF pathway that regulates cellular differentiation and proliferation during embryonic development and leads to tumor formation when aberrantly activated.12–14 In the absence of Wnt signal, cytoplasmic β-catenin is low because the protein is phosphorylated and targeted for degradation by a multiprotein complex, which includes the adenomatous polyposis coli (APC) protein, glycogen synthase kinase 3β (GSK 3β), casein kinase I and Axin. Mutations in the APC or CTNNB1 (β-catenin) genes that interfere with β-catenin degradation cause accumulation of β-catenin protein. The increased concentration of this protein in the cytoplasm favors its translocation to the nucleus as a coactivator for the T-cell factor (TCF)/ lymphocyte enhancer binding factor (LEF) family and activates the transcription of Wnt/β-catenin target genes such as c-myc,15cyclin D116 and cyclooxygenase-2.17 These target genes play important roles in the development and formation of some neoplasias.

Although compelling evidence has indicated a crucial role for signaling by β-catenin in the tumorigenesis of ESCC,7–11, 18 genetic mutations of APC, CTNNB1 or AXIN are rarely found in ESCC.8, 11, 19, 20 Therefore, additional mechanisms may exist to upregulate β-catenin levels in ESCC. Previous studies have shown that frequently rearranged in advanced T-cell lymphomas-1 (FRAT1) is strikingly overexpressed in some human esophageal cancer cell lines and some other cancer cells.21 Furthermore, FRAT1 is a positive regulator of the Wnt/β-catenin pathway, which can inhibit GSK3 activity toward β-catenin, at least in part, by preventing Axin binding to GSK3.13, 22, 23 However, whereas FRAT1 is a candidate for the regulation of cytoplamsic β-catenin, little is known with regard to the molecular relationship between FRAT1 and β-catenin in ESCC. There have been no studies to date that explore how FRAT1 regulates β-catenin in esophageal cancer.

To investigate whether and how FRAT1 regulates β-catenin in ESCC, the FRAT1 cDNA was cloned and transfected into human ESCC cell lines, and the effects of FRAT1 overexpression on cellular growth and transcriptional activity of β-catenin/TCF were analyzed. RNA interference (RNAi) was also used to determine the functions of FRAT1. In situ hybridization and immunohistochemical analysis were performed to identify the level of FRAT1 mRNA and the localization of β-catenin in surgical specimens of ESCC and normal esophageal tissues. We demonstrate that FRAT1 is overexpressed in ESCC and plays a role in cellular growth. Furthermore, we show that aberrant activation of Wnt/β-catenin pathway appears to be due to upstream events such as FRAT1 overexpression in ESCC. Also, we find that continued expression of c-Myc is necessary for maintenance of the growth state in cells expressing FRAT1, implying that c-Myc may be a critical element in oncogenesis induced by FRAT1.

Material and methods

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Tissue samples

We analyzed tissues from excised ESCC of patients treated at Cancer Hospital, Chinese Academy of Medical Sciences, Beijing, China. Two tissue samples were taken from each patient: 1 from the tumor and 1 from the resection margin. The protocol was approved by the Institutional Review Board of the hospital. Tissue samples were immediately frozen at −80°C until analysis, while the remaining tissues were used for histological classification by the local pathology department. All tumors were verified as squamous cell carcinoma with various degrees of differentiation. The tissues from the resection margins likewise were examined histologically and all were found to be free of tumor cells.

RNA extraction and RT-PCR

RT-PCR was performed as described previously.24 All PCR reactions were performed using standard PCR conditions: 95°C for 5 min, 95°C for 1 min, annealing at 56°C for 1 min and extension at 72°C for 1 min for 30 cycles and a final extension at 72°C for 10 min. The PCR products were visualized by electrophoresis in 2% agarose gels, followed by staining with ethidium bromide, and quantified using a Gel EDAS 290 analysis system (Cold Spring USA) and Gel-Pro Analyzer 3.1 software (Media Cybernetics). The primer pair used for amplification of the human FRAT1 was forward primer, 5′-GCCCTGTCTAAAGTGTATTTTCAG-3′, and reverse primer, 5′-CGCTTGAGTAGGACTGCAGAG-3′. As an internal standard, a fragment of human β-actin was coamplified by PCR using the following primers: forward primer, 5′-GGCGGCACCACCATGTACCCT-3′, and reverse primer, 5′-AGGGGCCGGACTCGTCATACT-3′.

Plasmid constructions, cell culture and transfections

To generate the FRAT1 and GSK 3β expression vector, the open reading frame of the human FRAT1 or GSK 3β cDNA was cloned into the eukaryotic expression vector pcDNA3 (Invitrogen) and pCEFL (kindly provided by Dr. J.S. Gutkind, National Institute of Dental and Craniofacial Research, NIH) or pcDNA3.1/Myc (Invitrogen) and was fused to an amino-terminal hemaglutinin (HA) tag or Myc tag, respectively. The resulting plasmid was designated FRAT1/pcDNA3 and FRAT1/pCEFL or GSK3β/pcDNA. The FRAT1 knockdown vector (pSilencer-FRAT1), the c-Myc knockdown vector (pSilencer-c-Myc) and the control RNAi vector (pSilencer-control) were constructed as described previously.25 The FRAT1 RNAi is targeted against the human FRAT1 sequence: 5′-GGCTTCATTCGCGACGGCT-3′, the c-Myc RNAi is targeted against the human c-myc sequence: 5′-CAGAAATGTCCTGAG CAAT-3′; while the control RNAi is targeted against the sequence: 5′-GACTCTTCCAGTGGTTTAA-3′, which is with limited homology to any known sequences in the human genome. The construct expressing a dominant-negative mutant of TCF4, commonly referred to as ΔN TCF4, was a kind gift from Prof. E.R. Fearon (University of Michigan School of Medicine, MI). c-Myc reporter was made by cut E-box and PTAL from pMyc-SEAP (Clontech) and inserted into pGL3-Basic (Promega). TOPFLASH and FOPFLASH were purchased from Upstate, and pRL-thymidine kinase (TK) was purchased from Promega.

The human ESCC cell line, EC9706 (kindly provided by Prof. M. Wang, Peking Union Medical College, China) was maintained in medium 199 (Invitrogen). Other ESCC cell lines (KYSE150, KYSE180, KYSE410 and KYSE510 cells, kindly provided by Dr. Y. Shimada, Kyoto University, Japan) were grown in RPMI 1640 medium (Invitrogen). All media were supplemented with 10% fetal bovine serum (Gibco BRL), penicillin (50 U/ml) and streptomycin (50 U/ml). These cells were maintained in a humidified 37°C under 5% CO2, fed every 3 days with complete medium and were subcultured when confluence was reached.

Transfections were performed as described previously.24 For the establishment of stable transfectants of FRAT1/pCEFL or pCEFL, after transfection, cells were selected with 600 μg/ml G418, and resistant clones were pooled and cultured. HA-FRAT1 expression was confirmed by Western blot. The resulting resistant clones populations were designated 150/FRAT1 and 150/control.

Protein preparation and western blot analysis

Standard procedures were used for protein preparation and western blotting.24 For preparation of cytoplasmic and nuclear proteins, NE-PER® Nuclear and Cytoplasmic Extraction Reagents (Pierce) were used in accordance with the manufacturer's protocol. The antibodies and dilutions used included anti-HA (Y11; 1:1000; Santa Cruz), anti-FRAT1 (C-17; 1:1000; Santa Cruz), anti-c-Myc (9E10; 1:1000; Santa Cruz), anti-β-catenin (C19220; 1:1000; Transduction Lab), anti-β-actin (AC-15; 1:5000; Sigma) and anti-Lamin B (M-20; 1:1000; Santa Cruz). Membranes were reprobed for β-actin to normalize for loading and to allow for accurate quantification. Protein expression was quantified using a Gel EDAS 293 analysis system (Cold Spring USA) and Gel-Pro Analyzer 3.1 software (Media Cybernetics).

Cell growth assay

Cells transfected with indicated plasmids were harvested and replated at a density of 50 cells/mm2 in triplicate. The total cell number was determined each day by using a hematocytometer and an Olympus inverted microscope. Cell viability was assessed by using trypan blue exclusion.

Immunofluorescence, in situ hybridization and immunohistochemistry

For immunofluorescence, 150/control and 150/FRAT1 cells were grown on cover slips. The next day, cells were fixed with ice-cool methanol-acetone and permeabilized with 0.2% Triton X-100. After blocking in 10% normal blocking serum at room temperature for 1 hr, slides were incubated with antibody to β-catenin (C19220; Transduction Lab; 1:100) at room temperature for 1 hr and then washed with PBS 3 times. Slides were then incubated with tetramethyl rhodamine isothiocyanate-conjugated anti-mouse secondary antibody and followed by staining with DAPI to identify cell nuclei.

In situ hybridization and immunohistochemistry were performed as described previously.24, 26 The digoxigenin-labeled FRAT1 probe for in situ hybridization was generated via in vitro transcription with FRAT1/pcDNA3 linearized with Hind III as templates using a digoxigenin RNA labeling kit (Roche).26 For immunohistochemistry, slides were treated with an anti-β-catenin antibody (C19220; Transduction Laboratories; 1:50) or an anti-c-Myc antibody (9E10; Santa Cruz; 1:50) at 4°C in a humidified chamber for 12 hr. All hybridization or immunostaining experiments were assessed by an experienced pathologist who was and the samples were blinded. FRAT1 mRNA expression within the cancer tissue was evaluated and categorized according to the percentage and intensity of cancer cells stained. Tumors were then further grouped into low and high expression of FRAT1. β-catenin protein expression profiles in cancer tissues were grouped into cytoplasm or nuclear expression and membrane expression only.10, 11, 24, 26

Reporter assay

Approximately 60% confluent cells were transfected in triplicate in 24-well dishes with LipofectAMINE™ 2000 Reagent (Invitrogen). Reporter assays were performed with the Dual-luciferase reporter assay system (Promega) at 36–72 hr after transfection, as described previously.24 pRL-TK (Promega) was used as an internal control for transfection efficiency. All results are expressed as means ± SD for independent triplicate cultures.

Statistical analysis

SPSS for Windows (SPSS Inc.) was used for statistical analysis. Correlation between the FRAT1 expression levels and β-catenin or c-Myc expression profiles on a per case basis was further analyzed by Spearman's rho correlation coefficient test. Values of p < 0.05 were considered statistically significant.

Results

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Overexpression of FRAT1 in human ESCC

Expression of FRAT1 mRNA in human esophageal cancer was determined by RT-PCR analysis. Specific FRAT1 primers yielded a 325-bp FRAT1 cDNA fragment and RT-PCR with detection primers for β-actin as an internal control yielded a 200-bp β-actin cDNA fragment (Fig. 1). FRAT1 mRNA was detected in all 27 cases of ESCC that were analyzed. Expression levels of FRAT1 mRNA were elevated in 20 cases of ESCC compared with corresponding surrounding noncancerous esophageal tissues (Fig. 1a and Supplementary Table S1). In addition, multiple ESCC cell lines were examined for FRAT1 mRNA expression by RT-PCR analysis. The highest level of FRAT1 expression was detected in KYSE180 and KYSE510 cells (Fig. 1b). These results are consistent with an earlier report.21 The finding that FRAT1 is overexpressed in human ESCC tissue samples and some ESCC cell lines suggests that FRAT1 may play a role in the development of this malignancy.

thumbnail image

Figure 1. Overexpression of FRAT1 in human ESCC tissues and cell lines. (a) Expression of FRAT1 mRNA in human ESCC tissues. Five representative cases are shown. T: ESCC tissue; N: corresponding normal adjacent mucosa; Marker: 100-bp DNA Marker. The densitometry data presented below the bands are fold change compared with corresponding normal ones. (b) Variable FRAT1 mRNA expression in human ESCC cell lines. The highest level of FRAT1 expression was noted in KYSE180 and KYSE510 cells. Marker: 100-bp DNA Marker; neg: negative control.

Download figure to PowerPoint

Role of FRAT1 in the growth of esophageal cancer cells

Next, we evaluated the effects of FRAT1 on the growth of the ESCC cell line KYSE150 and KYSE510. Human FRAT1 gene is a human homologue of mouse proto-oncogene Frat1.27, 28Frat1 was identified as a proto-oncogene that conveyed selective advantage to cells at later stages of murine T cell lymphomagenesis.27, 29 It remains unclear, however, whether human FRAT1 has potential roles in regulating cellular growth. To address this question, KYSE150 cells were transfected with pCEFL/HA-FRAT1 or pCEFL and selected with G418 to obtain stable transfectants (Fig. 2a). The resulting pool of stable transfectants was designated 150/FRAT1 and 150/control, and cell growth assays were performed. 150/FRAT1 and 150/control cells were plated at equal densities and cells were counted each day. Overexpression of FRAT1 dramatically promoted the growth of these cells (Fig. 2c). Meanwhile, downregulation of FRAT1 expression by RNAi (Fig. 2a and 2b) markedly inhibited the growth rates of 150/FRAT1 and KYSE510 cells, with a 60–80% decrease in 150/FRAT1 cells and a 45–50% decrease in KYSE510 cells at different time points, consistently in 3 separate experiments (Fig. 2d and 2f). We also observed that expression of GSK 3β or ΔN TCF4 in 150/FRAT1 cells decreased the growth rate of these cells (Fig. 2e). Similar effects of FRAT1 in cellular growth were observed in HEK293 cells (Supplementary Fig. S1). Also, we observed that expression of FRAT1 in HEK293 cells resulted in a significant increase in colony formation (Supplementary Fig. S2a and S2b). Meanwhile, suppression of FRAT1 expression by RNAi in 293/FRAT1 cells markedly inhibited colony formation in soft agar of these cells (Supplementary Fig. S2c and S2d). These results suggest that FRAT1 may play a role in mediating the growth advantage seen in esophageal cancer and that this effect might be exerted via affecting the function of the GSK 3β, thereby activating the β-catenin/TCF pathway.

thumbnail image

Figure 2. Role of FRAT1 in esophageal cancer cell growth. (a–b) Western blot demonstrating stable expression of HA-FRAT1 in KYSE150 (a) and suppression of FRAT1 expression by RNAi in KYSE150 (a) or KYSE510 (b) cells. Expression levels were normalized for loading by probing for β-actin. The densitometry data presented below the bands are fold change compared with control. (c–f) Effects of FRAT1 on esophageal cancer cell growth. 150/control and 150/FRAT1 (c) or 150/FRAT1 (d–e) and KYSE510 (f) cells transfected with plasmids indicated were harvested and replated at a density of 50 cells/mm2 in triplicate. Cells were trypsinized and counted each day. Cell viability was assessed by using trypan blue. error bars, SD.

Download figure to PowerPoint

FRAT1 activates β-catenin-dependent gene expression

To further test the model that FRAT1 regulates the β-catenin/TCF pathway in esophageal cancer, we investigated whether FRAT1 can affect the subcellular localization of β-catenin and its transcriptional activity in ESCC. To address these issues, immunofluorencence and reporter assays were performed. We found increased nuclear staining of β-catenin in 150/FRAT1 cells compared with 150/control cells (Fig. 3a3d). We also prepared cytoplasmic and nuclear proteins in KYSE510 cells following FRAT1 RNAi treatment. As shown in Figure 3e, knockdown of FRAT1 expression by RNAi in KYSE510 cells reduced the expression of β-catenin in both the cell cytoplasm and nucleus, especially in cell nucleus. When β-catenin is present in the nucleus, it is able to bind in conjunction with TCF/LEF to elements found in the promoters of a variety of cellular genes.13, 30 Hence, we studied the effect of FRAT1 on the transcriptional activity of β-catenin/TCF. The luciferase reporters TOPFLASH and FOPFLASH, which have a minimal TK promoter and either wild type (TOP) or mutated (FOP) binding sites for the β-catenin/TCF complex, have been widely used to characterize β-catenin/TCF-dependent gene expression.31 These reporter constructs were transfected into KYSE150 and KYSE510 cells, and luciferase activity was determined.

thumbnail image

Figure 3. FRAT1 activates β-catenin-dependent gene expression. (a–d) Immunofluorencence analysis of β-catenin localization in esophageal cancer cells without or with FRAT1 overexpression. (e) Cytosol and nucleus were isolated from the KYSE510 cells after FRAT1 RNAi transfection at different time points indicated and subjected to western blot with indicated antibodies. The densitometry data presented below the bands are fold change compared with control. (f–g) Effects of FRAT1 on the transcriptional activity β-catenin/TCF. KYSE150 (e) cells were cotransfected with HA-FRAT1/pCEFL (1 μg), TOPFLASH (100 ng) and pSilencer-FRAT1 (1 μg), GSK 3β expression vector (1 μg) or ΔN TCF4 (1 μg) for 36 hr, followed by luciferase assay. KYSE510 (g) cells were transfected with TOPFLASH (100 ng) and pSilencer-FRAT1 (1 μg), ΔN TCF4 (1 μg) or control plasmid for 72 hr, followed by luciferase assay. The pRL-TK Renilla luciferase reporter construct was cotransfected in each sample to normalize transfection efficiency. The activity of the reporter luciferase is expressed relative to the activity in control-vector-transfected cells, which is defined as 1.0. All experiments were performed in triplicate and are expressed as means and SD.

Download figure to PowerPoint

In KYSE150 cells, overexpression of FRAT1 can increase TOPFLASH activity (Fig. 3f). Cotransfection with pSilencer-FRAT1, GSK3β or ΔN TCF4 can partially or completely abolish this effect (Fig. 3f). There was little effect of FRAT1 on the FOPFLASH reporter (data not shown). Similar results were observed in HEK293 cells (Supplementary Fig. S3). This result was consistent with the increased nuclear levels of β-catenin seen in 150/FRAT1 when analyzed by immunofluorencence (Fig. 3a3d). In addition, in KYSE510 cells, suppression of FRAT1 by RNAi or expression of ΔN TCF4 markedly reduced β-catenin/TCF-dependent transcriptional activity (Fig. 3g), consistent with the decreased nuclear levels of β-catenin demonstrated in Figure 3e.

To determine whether FRAT1 expression leads to increased expression of β-catenin/TCF regulated genes in esophageal cells, we investigated the effect of FRAT1 on the expression of 1 important cellular gene known to be regulated by β-catenin/TCF. The endogenous expression levels of c-Myc, but not β-actin, were elevated more than 4-fold in 150/FRAT1 cells compared with 150/control cells (Fig. 4a). Suppression of FRAT1 expression by RNAi in 150/FRAT1 cells or expression of GSK3β or ΔN TCF4 in 150/FRAT1 cells reduced the protein level of c-Myc (Fig. 4b and c). FRAT1 knockdown also reduced the expression of c-Myc in KYSE510 cells (Supplementary Fig. S4).

thumbnail image

Figure 4. Effects of FRAT1 on the expression and transcriptional activity of c-Myc and RNAi directed against c-myc interferes with FRAT1-mediated growth promotion effects. (a–c) FRAT1 regulates c-Myc expression and reduction in c-Myc protein level by RNAi. Whole cell lysates of 150/control and 150/FRAT1 cells or 150/FRAT1 cells transfected with indicated plasmids were separated by SDS-PAGE and immunoblotted with an antibody against c-Myc. Expression levels were all normalized for loading by probing for β-actin. The densitometry data presented below the bands are fold change compared with control. (d–e) FRAT1 regulates c-Myc-dependent transcriptional activity. KYSE150 cells (d) were transfected with c-Myc reporter (100 ng) and HA-FRAT1/pCEFL (1 μg) or pCEFL (1 μg) for 36 hr, followed by luciferase assay. Meanwhile, KYSE510 cells (e) were transfected with c-Myc reporter (100 ng) and pSilencer-FRAT1 (1 μg) or control plasmid for 48 hr, followed by luciferase assay. The pRL-TK Renilla luciferase reporter construct was cotransfected in each sample to normalize transfection efficiency. The activity of the reporter luciferase is expressed relative to the activity in control-vector-transfected cells, which is defined as 1.0. The reporter assays were performed in triplicate and are expressed as means and SD. (f) 150/FRAT1 cells transfected with plasmids indicated were harvested and replated at a density of 50 cells/mm2 in triplicate. Cells were trypsinized and counted each day. Cell viability was assessed by using trypan blue.

Download figure to PowerPoint

These results indicate that FRAT1 overexpression causes the nuclear accumulation of β-catenin and results in the activation the β-catenin/TCF-dependent gene expression in esophageal cancer.

Correlation between subcellular localization of β-cateninand expression of FRAT1 in human ESCC tissues and normal esophageal epithelium

Our data have shown that FRAT1 is capable to promote the growth of esophageal cancer cells and transformation by activating β-catenin/TCF pathway. To further examine whether FRAT1 expression is correlated with cytoplasmic and nuclear localization of β-catenin in primary esophageal cancer tissues, we compared the expression of FRAT1 and the subcellular localization of β-catenin in 67 primary human ESCC tissues by in situ hybridization and immunohistochemical staining respectively. As we reported previously, β-catenin has an aberrant expression pattern in esophageal cancers.10, 11 Approximately 67% (45/67) of esophageal cancer samples showed loss of β-catenin localization at membranes accompanied by cytoplasmic/nuclear accumulation. More importantly, in most cases, β-catenin accumulation was correlated with high levels of FRAT1 expression in tumor specimens, whereas tumor tissues showing membrane localization of β-catenin generally contained relatively low levels of FRAT1 (Fig. 5c5f and Table I). In 67 ESCC tumor tissues examined, there was a significant correlation between FRAT1 and β-catenin expression, as determined by the Spearman rank correlation test (Table I, p = 0.002). In addition, FRAT1 expression was detected in all cases of normal squamous cell epithelium of the esophagus in the basal layer (Fig. 5a). Meanwhile, cytoplasmic/nuclear accumulation of β-catenin was exclusively observed in the basal layer of normal esophageal epithelium (Fig. 5b). The FRAT1 and β-catenin staining pattern of the normal squamous cell epithelium were consistent in different specimens. These results further confirm that FRAT1 may modulate the β-catenin/TCF pathway in esophageal cancer and maybe also in normal esophagus.

thumbnail image

Figure 5. Correlation between FRAT1 overexpression and β-catenin localization in human normal epithelium of the esophagus and ESCC tissues. Human tissues were stained with in situ hybridization probes against FRAT1 or anti-β-catenin antibodies and visualized by NBT/BCIP or DAB staining, respectively. Left panel, representative expression patterns of FRAT1 (a) and β-catenin (b) in normal esophageal epithelium. Figures are ×40, and insets are ×200; Middle panel, a representative high-FRAT1 staining (c) specimen with accumulation of β-catenin in the cytoplasm/nucleus (d); Right panel, a representative specimen with low-FRAT1 staining (e) with membrane staining of β-catenin (f). Figures are ×100, and insets are ×400.

Download figure to PowerPoint

Table I. Comparison of FRAT1 and β-catenin expression in ESCC
 FRAT1 staining 
β-Catenin stainingLow level(n = 34)High level (n = 33)Total
  1. The level of FRAT1 expression and localization of β-catenin were determined in 67 surgical specimens of ESCC, as shown in Figure 5. The correlation was analyzed using a Spearman rank correlation test P = 0.002.

Cytoplasm/nucleus17 (25.4%)28 (41.7%)45 (67.1%)
Membrane only17 (25.4%)5 (7.5%)22 (32.9%)
Total34 (50.8%)33 (49.2%)67 (100.0%)

RNAi directed against c-myc interferes with FRAT1-mediated growth promotion effects

As shown above by western blot, FRAT1 regulates the expression of c-Myc by β-catenin/TCF pathway (Fig. 4a-4c). The significance of this observations is further illustrated by our findings that upregulation of c-Myc in a panel of primary ESCC specimens is significantly correlated with elevated FRAT1 expression (Supplementary Fig. S5 and Supplementary Table S2). In addition, we find that FRAT1 can modulate the transcriptional activity of c-Myc (Fig. 4d and 4e). The c-Myc protein forms a heterodimer complex with the c-Max protein, which binds to the E-box DNA binding element and initiates transcription of genes responsible for its function.32, 33 The c-Myc reporter plasmid contains the luciferase reporter gene and 6 tandem copies of the E-box consensus sequence fused to a TATA-like promoter (PTAL) region from the Herpes simplex virus thymidine kinase promoter. After c-Myc proteins bind E-box, transcription is induced and the reporter gene is activated. In KYSE150 cells, overexpression of FRAT1 can increase the activity of c-Myc reporter (Fig. 4d), while in KYSE510 cells, suppression of FRAT1 expression by RNAi reduced c-Myc-dependent transcriptional activity (Fig. 4e). c-Myc has been reported to be an important oncogene, which plays essential roles in the pathogenesis of many human neoplastic diseases including human ESCC.6, 33, 34 To evaluate the possibility that production of c-Myc is required to maintain the growth state of cells expressing ectopic FRAT1, we attempted to block the expression of the c-myc gene in those cells by RNAi.

To this end, we constructed a plasmid encoding short hairpin RNA (shRNA) based on the human c-myc gene sequence25 (pSilencer-c-Myc) and a control plasmid (pSilencer-control) encodes a sequence, which is with limited homology to any known sequences in the human genome. 150/FRAT1 cells were transfected with pSilencer-c-Myc or pSilencer-control. We initially showed that pSilencer-c-Myc could reduce the amounts of c-Myc protein in 150/FRAT1 cells (Fig. 4b). When we attempted to determine the growth rate of these cells, we found that the cells in which c-Myc production was inhibited (150/FRAT1 cells transfected with pSilencer-c-Myc) grew more slowly, whereas cells expressing control shRNA grew normally (Fig. 4f). We also found that RNAi-mediated reduction of c-Myc expression in FRAT1 overexpression stable cell line 293/FRAT1 also inhibited their growth and anchorage-independent growth ability significantly (Supplementary Fig. S1c and Supplementary Fig. S2c and S2d). We conclude that continued production of c-Myc is necessary for maintenance of the growth state in cells expressing FRAT1, including human esophageal cancer cells.

Discussion

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

ESCC is 1 of the least studied and deadliest cancers worldwide.35 Oncogene and tumor suppressor gene alterations in ESCC are observed frequently.5, 6 Among these alterations are mutations of p53 and p16, and activation of oncogenes (e.g., c-myc, cyclin D1 and EGFR). Although accumulation of cytoplasmic β-catenin has been observed in ESCC, APC,19CTNNB18, 11 and AXIN,20 mutations have not been found in ESCC. The mechanism of accumulation of β-catenin in ESCC might be independent of genetic alterations of the APC, the CTNNB1 or the AXIN1 gene as shown in other tumors.13 Further mechanisms to stabilize β-catenin protein may be hidden in esophageal carcinogenesis. It has been demonstrated that the subcellular localization of β-catenin is affected by some molecules in ESCC, including frizzled E3 (FzE3),36 Wnt-137 and end-binding protein 1 (EB1).24 In this study, we report the activation of β-catenin/TCF pathway by FRAT1 in ESCC cells. We demonstrate the accumulation of β-catenin in the nucleus by transfection of FRAT1 in ESCC cell lines. In addition, FRAT1 can activate the TCF promoter luciferase reporter gene, and we find a significant correlation between FRAT1 and β-catenin expression in esophageal tumor samples and normal esophageal epithelium. To our knowledge, this is the first report of functional β-catenin/TCF activation by FRAT1 in ESCC cells.

FRAT1 gene, mapping to human chromosome 10q24.1, is a human homologue of mouse proto-oncogene Frat1.27, 38Frat1 was identified as a proto-oncogene that conveyed selective advantage to cells at later stages of murine T cell lymphomagenesis.27, 29 Its biological function, however, remained elusive until its Xenopus homolog GSK3-binding protein (GBP) was isolated as a GBP.28 Depletion of the endogenous GBP pool in oocytes prevented formation of a normal body axis in developing embryos, and GBP was thus shown to be a core component of the canonical Wnt pathway in Xenopus.28, 39, 40 Like GBP, Frat is able to induce secondary axis formation upon ectopic expression in Xenopus embryos by stabilizing β-catenin levels41, 42 and overexpression of Frat is sufficient to induce β-catenin/TCF-dependent reporter gene activity.43 However, recent results from van Amerongen and colleagues demonstrated that murine Frat(s) are not essential components of the canonical Wnt pathway in mice.44, 45 The paradoxical phenomenon is not surprising, especially when knockout mice are used to investigate the gene function. For instance, it is well known that the mammalian pituitary tumor transforming gene (PTTG) acts as an oncogene and plays an important role in regulating sister chromatid separation during mitosis. However, mice deficient in PTTG are unexpected viable only showing defects in glucose homeostasis leading to diabetes during late adulthood.46 Similar phenomenon was also observed in c-Src knockout mice.47, 48 c-Src is the first discovered oncogene. However, when c-Src gene is totally disrupted, the homozygous null mice appear completely normal and are only deficient in bone remodeling. Taken together, since that the regulatory mechanisms about Wnt pathway are complex and still obscure, there may be the presence of parallel pathways that are functionally redundant and may compensate for the loss of Frat. Moreover, there are some supportive reports showing that Frat plays a role in tumor progression.27, 29 Therefore, we speculate that in abnormal conditions, such as in the FRAT-overexpressed cancer cells, FRAT can be an important molecular trigger that might lead to the aberrant activation of β-catenin/TCF pathway. It is indeed the case in our previous observation that FRAT1 overexpression may lead to the aberrant activation of β-catenin/TCF pathway in ovarian serous adenocarcinomas.26

To determine whether FRAT1 expression might also be relevant in ESCC, in our current study, we first detect that FRAT1 is overexpressed in freshly resected ESCC and established ESCC cell lines. Then, we find that overexpression of FRAT1 in esophageal cancer cells KYSE150 promotes the growth of these cells and that this effect can be reversed by coexpression of GSK 3β or ΔN TCF4, demonstrating that the growth promoting effect of FRAT1 overexpression is mediated by the GSK 3β-β-catenin/TCF pathway. Conversely, RNAi mediated depletion of FRAT1 inhibits the growth rate of 150/FRAT1 and KYSE510 cells. Furthermore, overexpression of FRAT1 in KYSE150 cells also induces the nuclear accumulation of β-catenin and causes increased expression of β-catenin/TCF target genes, such as c-myc. FRAT1 can promote the transcriptional activity of β-catenin/TCF in KYSE150 and HEK293 cells, and coexpression of GSK 3β or ΔN TCF4 can partially or completely abolish these effects. Also, downregulation of FRAT1 expression by RNAi significantly reduces the nuclear levels of β-catenin and the transcriptional activity of β-catenin/TCF. Because of the lack of commercial available antibodies against FRAT1, which are suitable for immunohistochemical detection, we developed in situ hybridization probes to detect the expression of FRAT1 in paraffin-embedded human normal esophagus and esophageal cancer tissues as described previously.26 The significance of those observations in cell lines is thus further strengthened by our findings that upregulation of FRAT1 is significantly correlated with aberrant β-catenin expression and localization among a panel of primary ESCC specimens. In addition, we also find that in normal squamous cell epithelium of the esophagus, FRAT1 is expressed predominantly in the basal layer, in which the rapidly dividing cells reside. This observation is consistent with previous reports.44, 45 Meanwhile, cytoplasmic or nuclear accumulation of β-catenin is also exclusively detected in the basal layer of normal esophageal epithelium, indicating further that FRAT1 may also regulate β-catenin/TCF pathway in normal epithelium of the esophagus. Therefore, our data strongly suggest that FRAT1 plays some important roles in carcinogenesis of human ESCC.

Upon dysregulated activation of Wnt signaling, some human cancers ensue. The contribution of constitutive β-catenin/TCF activity, the most important event of the activation of Wnt signaling, to the colorectal transformation process has been extensively illustrated during the last decade. It is notable that van de Wetering and colleagues showed recently that c-Myc and p21CIP1/WAF1 activity, controlling by the β-catenin/TCF complex, is a major contributor to inhibit differentiation and imposes a crypt progenitornetics, most like the phenotype during the colorectal cells transformation process in vivo.49 The similar result is observed in our present study. Upon overexpression of FRAT1, the expression and the transcriptional activity of c-Myc is upregulated via constitutive β-catenin/TCF activity in ESCC cell lines and HEK293 cells. Moreover, knockdown of c-Myc expression by RNAi demonstrated that continued expression of c-Myc is necessary and sufficient to maintain the growth state in cells expressing FRAT1. Furthermore, upregulation of c-Myc is significantly correlated with elevated FRAT1 expression in a panel of primary ESCC specimens. All of our data imply that c-Myc may be a critical element during oncogenesis of human ESCC induced by FRAT1. When our manuscript is under preparation, the idea that c-Myc plays an extremely important role during mammalian oncogenesis due to the activation of β-catenin/TCF is further proved by Sansom's group.50 By means of double deleted both Apc and Myc in the adult murine small intestine simultaneously, they showed that Myc is absolutely required for the cellular and molecular changes that occur following Apc loss in the murine small intestine.

Our study suggests that FRAT1 may interact with the GSK 3β protein and affect the formation of the destruction complex for β-catenin formed by APC, GSK 3β, axin/conductin, resulting in subsequent aberrant nuclear accumulation of β-catenin and elevated transcriptional activity of β-catenin/TCF. The transcription of several downstream targets of β-catenin/TCF pathway, such as c-myc, are activated and contributed to the enhanced cellular growth. In conclusion, our data support the novel hypothesis that FRAT1 overexpression is important to β-catenin/TCF signaling activation in ESCC, and c-Myc may be a critical element in oncogenesis induced by FRAT1. Since key genetic, epigenetic and environmental factors associated with ESCC remain incompletely defined, our findings could shed some light on the unresolved molecular mechanism of ESCC and may lead to a better understanding of the role of dysregulated activation Wnt signaling contributing to oncogenesis of human ESCC, and further may indicate some components of Wnt signaling, such as FRAT1 and/or c-Myc, as possible therapeutic targets for the development of novel molecular treatments for human ESCC.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We thank Dr J.S. Gutkind for pCEFL and for his reading and suggestions in preparation of the manuscript. The authors also thank Prof. E.R. Fearon for ΔN TCF4, Dr Y. Shimada for KYSE cells, Prof. Mingrong Wang for EC9706 cells and Mr. Xiao Liang for the help in immunofluorescence microscopic photographing.

References

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Material and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

This article contains supplementary material available via the Internet at http://www.interscience.wiley.com/jpages/0020-7136/suppmat

FilenameFormatSizeDescription
ijc23600-Supplementfigure1.tif1879KSupporting Information file ijc23600-Supplementfigure1.tif
ijc23600-Supplementfigure2.tif2587KSupporting Information file ijc23600-Supplementfigure2.tif
ijc23600-Supplementfigure3.tif416KSupporting Information file ijc23600-Supplementfigure3.tif
ijc23600-Supplementfigure4.tif232KSupporting Information file ijc23600-Supplementfigure4.tif
ijc23600-Supplementfigure5.tif2440KSupporting Information file ijc23600-Supplementfigure5.tif
ijc23600-supplement.doc42KSupporting Information file ijc23600-supplement.doc

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.