Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Alterations of the Wnt signaling pathway during the neoplastic progression of Barrett's esophagus


Aberrant activation of the Wnt signaling pathway has been reported during neoplastic progression in Barrett's esophagus (BE). However, mutations in APC and CTNNB1 genes were rarely observed. In this study, expression pattern of Wnt ligands, Frizzled receptors and APC, as well as the methylation status of the APC, SFRP1 and SFRP2 promoter genes were investigated in normal esophageal mucosa and in preneoplastic and neoplastic lesions of BE patients. Promoter methylation of APC was found in all BE samples and in 95% of esophageal adenocarcinomas (EAC). Full methylation of APC correlated with lack of expression. In EAC, nuclear translocation of β-catenin was observed regardless of the expression of APC. WNT2 expression was higher in dysplasia and EAC than in BE, with 20/26 (77%) of the EAC showing high expression of WNT2. SFRP1 methylation occurred in all BE samples and in 96% of EAC, while SFRP2 was methylated in 73% of the normal squamous esophageal mucosa samples. In conclusion, (1) alterations of key regulators of the Wnt signaling are frequent in the pathogenesis of BE; (2) the APC and SFRP1 genes are inactivated by promoter methylation in BE; (3) the WNT2 gene is upregulated along the progression from low-grade dysplasia to EAC.


Barrett's esophagus (BE) is an acquired condition in which the normal stratified squamous epithelium in the distal esophagus is replaced by a metaplastic columnar epithelium, in response to chronic gastro-esophageal reflux. The importance of this condition is its association with a predisposition to esophageal adenocarcinoma (EAC), which arises through progression along a sequence of increasing degrees of dysplasia (Flejou, 2005). BE-associated EAC is the most rapidly increasing solid tumor in the Western World: its incidence has increased by over 70% during the last two decades (Blot and McLaughlin, 1999). Despite advances in treatment of EAC, its prognosis is still poor (Falk, 2002). The molecular basis of the development of EAC, although extensively studied (Jenkins et al., 2002; Brabender et al., 2004; McManus et al., 2004; Metzger et al., 2004), is not completely understood. Better understanding of the molecular alterations during its development might improve prevention and tumor control and ultimately lead to better disease management.

Wnt glycoproteins comprise a family of extracellular signaling ligands that play essential roles in the regulation of cell growth, motility and differentiation during embryonic development. Wnts are also required for adult tissue maintenance and perturbations in Wnt signaling can promote both human degenerative diseases and cancer (Logan and Nusse, 2004). A key component of this pathway is β-catenin, which is phosphorylated by glycogen synthase kinase β (GSK3β) in a multiprotein complex including Adenomatous Polyposis Coli (APC), Axin and casein kinase 1 (CK1). Phosphorylation of β-catenin induces its degradation via the ubiquitin–proteasome pathway. Binding of Wnt proteins to their frizzled transmembrane receptors activates the Wnt signal transduction pathway by triggering the phosphorylation of a cytoplasmic effector, Dishevelled (Dsh), which then inhibits the activity of GSK3β in phosphorylation of the APC–Axin complex. Unphosphorylated and therefore stable β-catenin can accumulate in the cytoplasm and becomes translocated to the nucleus, where it binds to member of the T-cell factor/lymphocyte enhancer factor (TCF/LEF) family of DNA binding proteins (Logan and Nusse, 2004). This leads to the transcriptional activation of growth promoting genes such as myc, cyclooxygenase 2, matrilysin/matrix metalloproteinase 7, cyclin D1 (see for a detailed list). Modulation of Wnt signaling occurs through two classes of Wnt antagonists. The first includes secreted Frizzled-related protein (SFRP) and Wnt inhibitory factor 1, which bind directly to Wnt ligands. The second class includes the Dickkopf (Dkk) family and binds to low-density lipoprotein receptor-related protein 5/6, which functions as coreceptor in Wnt signaling (Kawano and Kypta, 2003).

The role of Wnt signaling in cancer became evident with the discovery that ectopic expression of mouse Wnt1 causes mammary tumors in mice (Nusse and Varmus, 1982). A number of Wnt genes, including WNT2, WNT5A, WNT7B, WNT10B, WNT13, has now been associated with the development of various human cancers, including colon, gastric, head and neck, brain, endometrial, ovarian cancer and melanoma (Bui et al., 1997; Holcombe et al., 2002; Howng et al., 2002; Rhee et al., 2002; Ricken et al., 2002; Saitoh et al., 2002; Pham et al., 2003). The role of Wnt antagonists in oncogenesis, as well as their downregulation by promoter hypermethylation, has been extensively studied (Kobayashi et al., 2002; Suzuki et al., 2004; He et al., 2005). Mutations of Wnt signaling molecules are carcinogenic through activation of β-catenin-TCF signaling. APC or CTNNB1 (β-catenin) genes are frequently mutated in human colorectal cancer (Morin et al., 1997; Bienz and Clevers, 2000; Polakis, 2000). Mutant APC and Axin are unable to assist GSK3β in phosphorylating β-catenin; similarly, mutations in the phosphorylated residues of β-catenin stabilize the protein. This causes constitutive signaling independent of the upstream Wnt signal.

Previous studies have reported that nuclear accumulation of β-catenin is an indicator of activation of Wnt/β-catenin signaling, which has been found during progression of BE towards EAC (Bian et al., 2000; Osterheld et al., 2002). Mutations in CTNNB1 and Axin have not, and in APC rarely been detected in Barrett's-associated EAC (Gonzalez et al., 1997; Bian et al., 2000; Koppert et al., 2004), and therefore the mechanism of Wnt pathway activation remains unclear in this cancer. In this study, we investigated whether activation of Wnt and Frizzled proteins and/or silencing of APC, and/or downregulation of Wnt antagonists through promoter methylation, are responsible for the activation of the Wnt pathway in the development of EAC in BE.


The APC promoter 1A is methylated during neoplastic progression of BE mucosa

APC promoter 1A methylation status was analysed by MS-SSCA and MS-DBA after microdissection of formalin-fixed paraffin-embedded tissue sections. Samples were considered as methylated when at least 25% of the amplified DNA was methylated. Results are listed in Tables 1 and 2. Representative examples of the results obtained by MS-SSCA and MS-DBA are illustrated in Figure 1a and b, respectively. Different patterns of methylation were found in the BE and tumor cells: no methylation, full methylation, or a mixture of unmethylated and fully methylated alleles in varying ratios.

Table 1 APC, SFRP1 and SFRP2 promoter methylation during neoplastic progression of BE
Table 2 Immunostaining of β-catenin and APC, WNT2 expression level and epigenetic alterations of APC, SFRP1 and SFRP2 in 30 EAC
Figure 1

(a) APC promoter methylation by MS-SSCA in four normal squamous epitheliums (N), in four BE mucosa (BE) surrounding EAC and in six EAC (1, 2, 6, 9, 14, 20). Numbers correspond to the cases listed in Table 2. Square brackets indicate the migration of either the unmethylated DNA (unmethyl. CpGs) or the methylated DNA (methyl. CpGs). The percentage of methylated alleles is shown under each sample. (b) APC promoter methylation by MS-DBA in the cases illustrated in (a). Dots hybridized with an oligoprobe specific for either the unmethylated DNA (TG probe) or the methylated DNA (CG probe) are shown. The percentage of methylated alleles is indicated under each sample. (c) The presence or absence of APC expression by immunohistochemistry analysis in three EAC (1, 2, 9) showing different pattern of APC promoter methylation (50 or 100%) using a mouse monoclonal antibody to the COOH terminus of human APC.

APC expression was studied by immunohistochemistry in 30 EAC on consecutive sections (Table 2). Of the seven cases with full APC promoter methylation, six (86%) did not show APC staining (Figure 1c). Loss of APC expression was observed in four of 13 (31%) tumors with methylation of 50% of the amplified DNA, while in the other nine (69%) diffuse positive staining was found. APC did not stain in the unmethylated case.

Activation of the Wnt signaling pathway is not APC dependent

APC and β-catenin staining were compared in 29 EAC. Nuclear staining (Table 2, N1 to N3 staining) of β-catenin was detected in nine of 14 (64%) cases that had lost expression of APC. However, nine of 15 (60%) EAC that expressed APC also expressed nuclear β-catenin, indicating that APC independent mechanisms activate the Wnt signaling pathway.

To further assess whether loss of APC is compatible with Wnt signaling, we used siRNA to silence APC expression in TE-7, OE19 and OE33 cell lines. The APC gene contains two promoters (1A and 1B) that initiate transcription from distinct sites. TE-7 and OE33 expressed the APC mRNA from promoter 1A. APC transcripts from promoter 1A were not found in OE19, since this cell line showed methylation of both alleles of the promoter 1A. APC transcripts from promoter 1B were found in all three cell lines (Figure 2a). As shown in Figure 2b, a high level of Wnt signaling was only observed in OE19, the cell line with methylation of the APC promoter 1A. RNA oligonucleotides against APC reduced its mRNA level for about 90% in TE7, and for 50% in OE19 (transcripts from promoter 1B only) and OE33. After treatment with the siRNA against APC, TOPFLASH activity remained at a basal level in TE-7 and did not change in OE19. In contrast, a three- to seven-fold increase of the TOPFLASH activity was detected in OE33 cells after transfection of the siRNA against APC (Figure 2b). Treatment with siRNA against APC affected proliferation of OE33 cells, but not of TE-7 and OE19, as indicated by AlamarBlue analysis (Figure 2c). Therefore, the three- to seven-fold increase in TOPFLASH activity might be an underestimation.

Figure 2

APC expression and RNA interference of the APC gene in three EAC cell lines (TE-7, OE19 and OE33). RT–PCR analyses of promoter 1A specific transcripts (APC/ex1A-2; 191-bp) and combined transcripts from promoter 1A and 1B (APC/ex3-5; 352-bp). A 274-bp fragment of the ACTB gene was used as an internal control. C+ and C− corresponded to HeLa RNA and H2O, respectively, used as control during the RT–PCR. The primers are described in Table 4 (a). RNA interference of the APC gene and of an unrelated gene (siRNA control). Firefly luciferase activity of the TOPFLASH reporter is represented relative to the percentage of the firefly luciferase activity of the pGL3-control vector. Each sample was normalized with the Renilla luciferase activity (b). AlamarBlue analysis after 40 h of siRNA treatment in TE-7, OE19 and OE33 (c).

Overexpression of WNT2 occurs in dysplastic lesions

The mRNA expression levels of WNT1, WNT2, WNT5A, WNT10B, WNT13 and FZD7 were analysed by RT–PCR in 20 formalin-fixed paraffin-embedded EAC tissues and their surrounding normal squamous epithelium. WNT1 was not found in normal and tumor tissues. The expression of WNT5A, WNT10B and WNT13 was variable – in 80, 80 and 100% of cases in the squamous epithelium and in 80, 85 and 75% of the tumor tissues, respectively (Figure 3). Expression of the WNT2 gene was higher in EAC tissues than in normal tissue. FZD7 expression in normal squamous mucosa and EAC tissues did not differ, both tissues strongly expressed FZD7. In five samples, analysis through cell-specific microdissection of the normal squamous mucosa showed that FZD7 is expressed in the normal squamous epithelium as well as in smooth muscle cells.

Figure 3

Expression of WNT ligands and Frizzled receptor by RT–PCR analysis in four normal squamous epitheliums (N) and the corresponding EAC (T) (cases 5, 6, 14 and 16). Cases correspond to those listed in Table 2. The fragments amplified were: WNT1, 175-bp; WNT2, 129-bp; WNT5A, 120-bp; WNT10B, 132-bp; WNT13, 161-bp; FZD7, 133-bp. A 274-bp fragment of the ACTB gene or a 180-bp fragment (ACTB*) was used as an internal control to ensure RNA quality. C+ and C− corresponded to normal placenta RNA and H2O, respectively, used as control during the RT–PCR.

The mRNA expression level of WNT2 was quantified in BE, dysplasia and in EAC (Table 3 and Figure 4). For the analysis, microdissection of metaplastic and dysplastic cells was performed in eight BE and nine dysplastia (low and high grade), respectively. None of the normal squamous and nondysplastic BE samples expressed a detectable level of WNT2 transcript. A moderate to high level of WNT2 mRNA was detected in 4/9 (44%) samples with dysplasia (Table 3). None to weak WNT2 mRNA expression level was reported in 6/26 (23%) of the EAC samples, while 20/26 (77%) EAC expressed moderate to high level of WNT2 transcripts. The results showed that during neoplastic progression of BE, WNT2 overexpression was associated with the occurrence of low-grade dysplasia.

Table 3 mRNA expression level of WNT2 during the neoplastic progression of BE
Figure 4

Dot blot readout of the quantitative RT–PCR of the WNT2 gene in six normal squamous epitheliums (N), BE mucosa (BE), low-grade (LGD) or high-grade (HGD) dysplastic mucosa and EAC (cases 1, 4, 5, 6, 8 and 11). Cases correspond to those listed in Table 2. WNT2 mRNA expression level is shown under each sample. Samples were previously normalized against ACTB.

The WNT antagonist SFRP1 is regulated by promoter methylation

As the promoters of SFRP1, SFRP2 and WNT2 are located in CpG islands, we hypothesized that these genes are regulated by DNA methylation during neoplastic progression of BE. We performed methylation analysis by MS-SSCA or MS-DBA after microdissection of squamous, BE and EAC samples. We observed in the methylation analysis of SFRP1 significantly more often promoter methylation in BE (6/6; 100%) and EAC (23/24; 96%) than in normal esophageal mucosa (1/12; 8%) (P<0.0005, Fisher exact test) (Table 1). Methylation of SFRP2 was not different between normal and BE and EAC (11/15; 73%, 6/6; 100% and 19/25; 76%). Representative examples of the methylation analysis results for SFRP1 and SFRP2, obtained by MS-DBA, are illustrated in Figure 5a. Although WNT2 expression is higher in dysplasia and EAC, hypermethylation of the WNT2 promoter was not observed in the non-WNT2 expressing samples (data not shown), suggesting that WNT2 silencing is not a result of promoter methylation.

Figure 5

SFPR1 and SFRP2 promoter methylation analysis by MS-DBA in five normal squamous epithelium, in BE mucosa surrounding EAC and in EAC (cases 1, 5, 6, 7, 9) (a). Cases correspond to those listed in Table 2. Dots hybridized with an oligoprobe specific for either the unmethylated DNA (TG probe) or the methylated DNA (CG probe) are shown. The percentage of methylated alleles is indicated under each sample. RT–PCR analysis of SFRP1 and SFRP2 gene in two EAC cell lines (TE-7 and OE33) after or without treatment with the demethylating agent 5-azadC (b). The fragments amplified were: SFRP1, 167-bp and SFRP2, 160-bp. A 274-bp fragment of the ACTB gene was used as an internal control to ensure RNA quality. C+ corresponded to normal placenta RNA used as control during the RT–PCR.

To explore the role of DNA methylation in the regulation of SFRP1 and SFRP2 genes, we subjected the TE-7 and OE33 cell lines to the demethylating agent 5-aza-2′-deoxycitidine. In control cells, SFRP1 showed complete methylation of both alleles, whereas SFRP2 was methylated in only one allele (monoallelic methylation). As expected, the SFRP1 gene was not transcribed in untreated cells, while demethylation permitted its expression. Although only one of its alleles was methylated, the SFRP2 gene was neither expressed in untreated nor in treated cells (Figure 5b).


The implication of the Wnt signaling pathway in the pathogenesis of a broad range of human cancers is becoming more clearly understood. Key mediator of this pathway is β-catenin, which, unphosphorylated, can migrate into the nucleus and in interaction with members of the TCF/LEF transcription factor family activates tumor growth promoting genes (Logan and Nusse, 2004). Neoplastic progression of BE involves a multistep process from intestinal metaplasia to low- and high-grade dysplasia and finally to adenocarcinoma (Flejou, 2005). Nuclear accumulation of β-catenin occurs early in the transition from BE to EAC: it is associated with the transition from metaplasia to low-grade dysplasia (Seery et al., 1999; Bian et al., 2000). In contrast to colorectal cancer, where mutations in either APC or CTNNB1 account for more than 90% of the cancers, many other cancers harbor no or only few mutations in APC and CTNNB1 (Morin, 1999; Polakis, 1999). This holds true for carcinogenesis in BE, where very few mutations of APC or CTNNB1 have been detected (Gonzalez et al., 1997; Bian et al., 2000; Koppert et al., 2004). The mechanism of Wnt pathway activation in EAC in BE remains unclear. In this study, we looked for genetic alterations of the main Wnt pathway components that might explain its activation during the development of EAC.

We detected methylation of APC promoter 1A in all BE and in 95% of the EAC samples. Full methylation of the APC promoter correlated with lack of expression. In the 13 EAC showing an equidense pattern of methylated and unmethylated alleles, nine expressed the APC protein, while the expression was silenced in four cases. This indicates that promoter methylation is not the only mechanism of inactivation, since transcription remains possible from the unmethylated allele. APC promoter methylation was found already in BE without dysplasia, whereas nuclear accumulation of β-catenin was detected only in the presence of dysplasia. Furthermore, in EAC, nuclear translocation of β-catenin was observed regardless of the expression of APC. All these data indicate that loss of APC expression is neither necessary nor sufficient for activation of the Wnt pathway in EAC.

By treating three EAC cell lines with siRNA against APC, we observed that silencing of APC led to nuclear translocation of β-catenin in one of the analysed cell lines. Not all the esophageal tumor cell lines showed increased activity of the TOPFLASH reporter assay, confirming that loss of APC is not sufficient for activation of the Wnt pathway. This result is consistent with other published data, where reduction of the APC level resulted in slightly increased nuclear staining of β-catenin (Verma et al., 2003). Worm et al. (2004) explored in a recent study the role of APC alterations in malignant melanoma and provided evidence that hypermethylation of the APC promoter may have a tumorigenic effect that is independent from the Wnt signaling pathway. In EAC, the role of APC inactivation in activating the Wnt pathway does not seem to be dominant but to fully understand the mechanism involved more EAC cell lines have to be investigated. The promoter 1B is probably activated in several EAC cell lines, as well as in primary tumors, as is the case in gastric cancer and melanoma (Tsuchiya et al., 2000; Worm et al., 2004). As the C-terminal anti-APC antibody may not recognize the protein translated from promoter 1B, the potential role of transcripts from 1B in altered Wnt signaling should be clarified.

We wondered whether overexpression of the Wnt ligands might play a role in the activation of the Wnt signaling pathway in EAC tissues, as it has been reported for WNT2, WNT5A, WNT7B, WNT10B, WNT13 in various human cancers (Bui et al., 1997; Holcombe et al., 2002; Howng et al., 2002; Rhee et al., 2002; Ricken et al., 2002; Saitoh et al., 2002; Pham et al., 2003). The results revealed a clear disparity in WNT2 expression level between the normal squamous (esophageal mucosa) and EAC samples. By quantitative RT–PCR, we observed high WNT2 expression in 77% of EAC. WNT2 expression is higher in dysplasia than in normal mucosa and BE without dysplasia, which goes along with nuclear accumulation of β-catenin (Bian et al., 2000). Of the 20 EAC samples with high WNT2 mRNA, 14 (70%) showed nuclear staining of β-catenin (see Table 2). Five cases without WNT2 expression also showed nuclear staining for β-catenin, indicating that other mechanisms not involving WNT2 expression must be implicated in the activation of Wnt signaling. Surprisingly, none of the three EAC cell lines expressed a detectable level of WNT2 mRNA. This might be an indication of a role for tumor cell microenvironment in modulating WNT2 expression in EAC. However, more EAC cell lines should be analysed to confirm that loss of WNT2 transcripts occurs in vitro and is not a tissue culture artifact.

WNT2 is frequently upregulated in primary gastric cancer and colorectal cancer and less frequently upregulated in primary breast cancer (Katoh 2001a, 2001b, 2003). It has been shown that Wnt-2 shares several biological and biochemical characteristics with Wnt-1, notably its capacity for cell transformation and tumorigenesis (Blasband et al., 1992). In a recent study, Le Floch et al. (2005) demonstrated that Wnt-2 can act as a proinvasive agent through noncanonical Wnt transcription using GSK3β and the AP-1 oncogene in colonic and kidney cells. A role for Wnt-2 in EAC, eventually through noncanonical Wnt signaling, needs to be further explored.

The activation of Wnt signaling through downregulation of SFRP expression is becoming well documented, notably their epigenetic inactivation in human cancers (Caldwell et al., 2004; Lee et al., 2004; Suzuki et al., 2004; He et al., 2005). We observed SFRP1 promoter methylation in 96% of the EAC samples. Moreover, SFRP1 promoter methylation was found in all BE samples, indicating that epigenetic alteration of SFRPs genes is an early event in the carcinogenesis of BE. We confirmed this through exposure of the cells to the agent 5-aza-2′-deoxycitidine, which upregulated SFRP1 expression. Interestingly, we found in 53% of the normal esophageal mucosa samples an equidense pattern of unmethylated and methylated alleles for SFRP2. We recently reported monoallelic methylation pattern of the APC gene in normal gastric mucosa, which is either decreased or increased in morphologically normal mucosa adjacent to BE-associated neoplasia (Clement et al., 2004). This suggests that SFRP2 shows monoallelic methylation in normal esophageal mucosa, which is altered in the normal epithelium adjacent to EAC. Indeed, 27% of the esophageal mucosa samples were found unmethylated, while 20% showed methylation of both alleles. However, additional studies are necessary to confirm this hypothesis. Our data suggest that SFRP2 promoter methylation is not involved in the neoplastic progression of BE, since it is methylated in the normal mucosa and a trend to hypermethylation is not observed along with the transition from normal mucosa to BE and to EAC. However, in spite of monoallelic methylation of SFRP2 in the two cell lines transcripts were not found, indicating that promoter methylation is not the only regulator.

In conclusion, we found (1) APC silencing by promoter hypermethylation to occur frequently in BE and EAC, (2) the WNT2 gene to be upregulated when low-grade dysplastic BE progresses to EAC, (3) SFRP1 to be regulated by promoter methylation, which occurs already in BE without dysplasia, and (4) the SFRP2 promoter to be methylated in the normal mucosa without showing a trend towards hypermethylation during the transition from BE to EAC. Alteration of key regulators of the Wnt signaling pathway is therefore frequent in EAC and could play an important role in the neoplastic progression of BE. A better understanding of the critical factors of the Wnt pathway may contribute to the development of targeted therapies and novel drugs for EAC patients.

Materials and methods

Tissue samples and cell lines

Samples were taken from histologically characterized lesions, including normal squamous esophageal epithelium (20 samples), intestinal metaplasia (BE) (13 samples), dysplasia (nine samples) and adenocarcinoma (30 samples). All the samples were from 30 esophagectomy specimens with an adenocarcinoma in BE occurring between 1983 and 2001. None of the patients had received chemotherapy or radiation therapy before surgery. All samples had been formalin-fixed and paraffin-embedded. Normal human placenta was used as a positive control for the RT–PCR analyses and human normal colon mucosa was used as positive control for the methylation analysis (Clement and Benhattar, 2005). The use of human tissues in this study was authorized by the local ethics committee.

The human esophageal carcinoma cell lines OE19 and OE33 (Rockett et al., 1997) were obtained from the European Collection of Cell Culture (Salisbury, UK). The human esophageal carcinoma cell line TE-7 was a kind gift from Professor T Kudo (Tohoku University, Japan) (Nishihira et al., 1993; Kitadai et al., 1998). The cells were routinely cultured in RPMI-1640 medium supplemented with 10% heat-inactivated fetal bovine serum and an antibiotics mixture (20 U/ml penicillin, 100 μg/ml streptomycin and 0.5 μg/ml fungizone) (all products from Invitrogen, Basel, Switzerland).

Microdissection, DNA extraction and sodium bisulfite conversion

After deparaffinization and staining in 0.1% toluidine blue, histologically selected areas in tissue sections were manually microdissected (Baisse et al., 2000). Only the tissues of interest were retained and final histological control before collection of the tumor or Barrett's epithelium cells confirmed that contamination with other cells, including stromal cell, was negligible. The selected tissue was suspended in 50 μl of Tris-EDTA buffer and digested with proteinase K. Genomic DNA was extracted according to the standard phenol/chloroform method (Baisse et al., 2000). The total amount of DNA obtained from microdissected tissues was modified in 40 μl of water with sodium bisulfite as described previously (Benhattar and Clement, 2004).

Methylation-sensitive single-strand conformation analysis of the APC, SFRP2 and WNT2 promoters

PCR amplifications were performed with 2 μl of modified DNA in a total volume of 20 μl for 40 cycles. A 159-bp fragment of the APC gene promoter was amplified by nested PCR using the following primers: FW 5′-IndexTermGGGGTTAGGGTTAGGTAGG-3′ and RV 5′-IndexTermAACTACACCAATACAACCACATA-3′ for the outer PCR and FW 5′-IndexTermGGGTTAGGGTTAGGTAGGTTGT-3′ and RV 5′-IndexTermCCCACACCCAACCAATC-3′ for the inner PCR amplification. In total, 20 cycles were performed for the inner PCR. Our amplified nested product include 13 CpG sites between –210 and –52 relative to the APC major transcription start site (Horii et al., 1993). A 188-bp fragment of the SFRP2 gene promoter containing 20 CpG sites located between –395 and –208 relative to the SFRP2 transcription start site was amplified with the following primers: FW 5′-IndexTermGATTGGGGTAAAATAAGTT-3′ and RV 5′-IndexTermCCCTAAAATTTCTTTAAACA-3′. The primers FW 5′-IndexTermTTTGGGGAATTGGGTTGTG-3′ and RV 5′-IndexTermCCCTTACCTCTACATCAAATC-3′ were used to amplify a 231-bp fragment of the WNT2 gene containing 29 CpG sites located between +62 and +292 relative to the transcription start site. PCRs for the APC and SFRP2 gene promoter amplification were carried out in the presence of 5% dimethyl sulfoxide. Amplification was confirmed by analysis on a 2% agarose gel. Single-strand conformation analysis was performed with 5 μl of PCR products as described previously (Bian et al., 2001).

Methylation-sensitive dot blot assay of the APC, SFRP1 and SFRP2 promoters

An MS-DBA was performed for the APC, SFRP1 and SFRP2 gene promoters using two different commercially synthesized oligonucleotide probes specific for the amplified sequences, as previously described (Clement and Benhattar, 2005). The 146-bp fragment of the SFRP1 promoter was amplified with the following primers: FW 5′-IndexTermGGAGGTTTTTGGAAGTTTG-3′ and RV 5′-IndexTermCCAACAACACTCCCAAAACTAC-3′. The amplified product includes 19 CpG sites located between +214 and +359 relative to SFRP1 transcription start site. The probes used for the MS-DBA were labeled with the digoxigenin oligonucleotide 3′-End Labeling Kit (Roche, Rotkreuz, Switzerland) and were designed to hybridize to either methylated (CG-probe) or unmethylated DNA (TG-probe). Probe sequences were as follows: APC: CG-probe: 5′-IndexTermGATGCGIndexTermGATTAGGG CG IndexTermTTTT-3′ and TG-probe: 5′-IndexTermGGATG TG IndexTermGATTAGGG TG IndexTermTTTT-3′; SFRP1: CG-probe: 5′-IndexTermAATAGGG CG IndexTermTAGAGT CG IndexTermGTA-3′ and TG-probe: 5′-IndexTermGAATAGGG TG IndexTermTAGAGTTGIndexTermGTA-3′; SFPR2: CG-probe: 5′-IndexTermGTTAGG CG CG IndexTermTTTT CG IndexTermTTAG-3′ and TG-probe: 5′-IndexTermGGGTTAGG TG TG IndexTermTTTT TG IndexTermTTAG-3′. The bold letters represent the two CpG sites investigated in each gene promoter. The hybridization was performed with 5–10 pmol of the respective probes for 2–3 h at 50°C.

RT–PCR for WNT and frizzled genes

Using an H&E section as a reference, the tissue lesion of interest (normal or tumoral) was macrodissected from the paraffin block. After deparaffinization and Proteinase K digestion, total RNA was extracted using Trizol-LS (Invitrogen, Basel, Switzerland) as previously described (Guillou et al., 2001). Metaplastic and dysplastic tissues were manually microdissected as described above before RNA extraction in order to dispose of histologically homogeneous starting material. All the RNAs were treated with DNaseI and the efficiency of the treatment was checked by PCR amplification. Both cDNA synthesis and PCR were performed in a single tube using the SUPERSCRIPT one-step RT–PCR with Platinum Taq Kit (Invitrogen, Basel, Switzerland). The reverse transcriptase step was performed at 50°C during 30 min followed by 2 min at 94°C to release the Taq polymerase. Subsequently, 40 cycles of PCR were performed, including denaturation at 94°C for 30 s, annealing at Tm for 45 s and extension at 72°C for 75 s. Primers sequences and Tm are listed in Table 4. The housekeeping gene ACTB was coamplified as internal control. RNA extracted from frozen human placenta was used as positive control. Amplifications were analysed on a 2% agarose gel.

Table 4 Primer sequences and Tm used for the RT–PCR analysis

Quantitative RT–PCR and dot blot analysis for WNT2

DNaseI treated RNA from normal, metaplastic, dysplastic and tumor tissues was coamplified for the WNT2 gene and the housekeeping gene ACTB for 22, 25 and 28 cycles. A dilution scale of RNA (200–0 ng) was amplified simultaneously with all the samples. The following primers were used for the amplification of the ACTB gene: FW 5′-IndexTermAGGCCAACCGCGAGAAGATGA-3′ and RV 5′-IndexTermCCTCGTAGATGGGCACAGTGTGG-3′. PCR products were denatured in 100 mM (final concentration) NaOH at 50°C for 10 min. Denatured PCR products were deposed onto two nylon membranes and fixed with UV. DIG-labeled probes were synthesized by RT–PCR from placenta extracted RNA using the following primers: WNT2 probe: FW 5′-IndexTermCGCATTTGTGGATGCA AAGGA-3′ and RV 5′-IndexTermTCGCCCGTTTTCCTGAAGTCG-3′ and ACTB probe: FW 5′-IndexTermAGGCCAACCGCGAGAAGATGA-3′ and RV 5′-IndexTermGCCGTGGTGAAGCTGTAG-3′. Hybridization was performed for 2 h at 42°C with 30 ng of probe per ml of hybridization buffer. After incubation with anti-Digoxigenin-AP Fab fragments (Roche, Rotkreuz, Switzerland) and addition of CDP-Star ready-to-use solution (Roche, Rotkreuz, Switzerland), the membranes were exposed to a X-OMAT film (Eastman Kodak Company, NY, USA) for different times (30 s, 2, 4, 32 and 1 h 20 min). Samples were first normalized to a uniform intensity of ACTB by choosing the adequate blot on each film. Blots were then quantified by comparing the intensities to those of the reference dilution scale.

Immunohistochemistry of β-catenin and APC

Paraffin sections (4 μm) were cut and deparaffinized using standard methods. Antigen retrieval was performed in a pressure cooker for 1 min in 10 mmol/l sodium citrate buffer. The slides were incubated for 30 min in 10% normal goat immunoglobulin G (Dako, Glostrup, Denmark). Sections were then incubated for 2 h at room temperature with the monoclonal anti-β-catenin (Transduction Laboratories) or with the monoclonal anti-APC antibody (ab120, Abcam, Cambridge, UK) diluted 1:500 in PBS containing 0.2% BSA. The ENVISION+ System/horseradish peroxidase kit (Dako) was used according to the manufacturer's instructions. Results were visualized with diaminobenzidine and counterstained with hematoxylin.

Cell line transfection

Cells were seeded at a concentration of 2.5 × 105 cells/well in a 12-well plate 24 h before transfection. Lipofectin Reagent (2 μl) (Invitrogen, Basel, Switzerland) were used to transfect 1 μg of pGL3-OT vector (TOPFLASH, kind gift of Dr Vogelstein, The Johns Hopkins Oncology Center, Baltimore, MD, USA; Shih et al., 2000), a promoterless reporter comprising three copies of the wild-type TCF-4 binding site cloned into the pGL3 basic vector containing the firefly luciferase gene. The pGL3 control vector (1 μg) containing the firefly luciferase gene under the control of SV40 promoter (Promega, Madison, WI, USA) was used as control. The Renilla luciferase reporter vector (0.1 μg) (Promega, Madison, WI, USA) was cotransfected as control for transfection efficiency. The cells were harvested 48 h after transfection. Luciferase assays were performed using the Dual-Luciferase Reporter Assay System (Promega).

SiRNA transfection assays

Double-stranded annealed siRNA oligonucleotides with two thymidine residues (dTdT) at the 3′-end of the sequence were designed to APC (sense, 5′-IndexTermGCAACAGAAGCAGAGAGGU-3′ and antisense, 5′-IndexTermACCUCUCUGCUUCUGUUGC-3′) (Verma et al., 2003). Unrelated siRNA (Qiagen, Hilden, Germany) was used as control. SiRNAs were resuspended at a concentration of 20 μ M in buffer (Qiagen), according to the manufacturer's instructions. Transfection of siRNA oligonucleotides was performed with jetSI™-ENDO (Polyplus-Transfection, Illkirch, France).

To determine the effects of APC silencing on β-catenin expression, the TE-7 and OE19 cells were transfected 24 h after APC siRNA transfection with the TOPFLASH reporter and the Renilla luciferase vector. OE33 cells were transfected simultaneously with RNA oligonucleotides and plasmids according to the jetSITM-ENDO protocol. Efficiency of RNA silencing was checked by quantitative RT–PCR with the following primers: FW 5′-IndexTermCTGCCAGGATATGGAAAAAC-3′ and RV 5′-IndexTermTTCTCCCACTCCTTGACCT-3′. The primers used for the APC DIG-labeled probe were the following: FW 5′-IndexTermTATGGGTTCATTTCCAAGAAG-3′ and RV 5′-IndexTermCGGT TTCATGCTTGTTCTGAGA-3′.

Proliferative activity after siRNA treatment was assessed using AlamarBlue (Serotec Ltd, Oxford, UK). At 40 h after transfection, the cells were washed and AlamarBlue was added for an additional incubation of 8 h. Fluorescence was read at 530 and 580 nm.

5-azadC treatment

Cells were treated 48 h after seeding and at 48 h intervals with 1.5 μ M of 5-azadC in fresh RPMI-1640 medium. Treatment was maintained for two passages to allow efficient demethylation. Cells were harvested at the end of the treatment and stored at −80°C. The level of demethylation was checked by MS-SSCA. Expression of SFRP1 and SFRP2 were analysed by RT–PCR. Primer sequences are listed in Table 4.


  1. Baisse B, Bian YS, Benhattar J . (2000). Biotechniques 28: 856–858, 860, 862.

  2. Benhattar J, Clement G . (2004). Methods Mol Biol 287: 181–193.

  3. Bian YS, Osterheld MC, Bosman FT, Fontolliet C, Benhattar J . (2000). Am J Clin Pathol 114: 583–590.

  4. Bian YS, Yan P, Osterheld MC, Fontolliet C, Benhattar J . (2001). Biotechniques 30: 66–72.

  5. Bienz M, Clevers H . (2000). Cell 103: 311–320.

  6. Blasband A, Schryver B, Papkoff J . (1992). Oncogene 7: 153–161.

  7. Blot WJ, McLaughlin JK . (1999). Semin Oncol 26: 2–8.

  8. Brabender J, Marjoram P, Salonga D, Metzger R, Schneider PM, Park JM et al. (2004). Oncogene 23: 4780–4788.

  9. Bui TD, Zhang L, Rees MC, Bicknell R, Harris AL . (1997). Br J Cancer 75: 1131–1136.

  10. Caldwell GM, Jones C, Gensberg K, Jan S, Hardy RG, Byrd P et al. (2004). Cancer Res 64: 883–888.

  11. Clement G, Benhattar J . (2005). J Clin Pathol 58: 155–158.

  12. Clement G, Bosman FT, Fontolliet C, Benhattar J . (2004). Cancer Res 64: 6867–6873.

  13. Falk GW . (2002). Gastroenterology 122: 1569–1591.

  14. Flejou JF . (2005). Gut 54: i6–i12.

  15. Gonzalez MV, Artimez ML, Rodrigo L, Lopez-Larrea C, Menendez MJ, Alvarez V et al. (1997). J Clin Pathol 50: 212–217.

  16. Guillou L, Coindre J, Gallagher G, Terrier P, Gebhard S, de Saint Aubain SN et al. (2001). Hum Pathol 32: 105–112.

  17. He B, Lee AY, Dadfarmay S, You L, Xu Z, Reguart N et al. (2005). Cancer Res 65: 743–748.

  18. Holcombe RF, Marsh JL, Waterman ML, Lin F, Milovanovic T, Truong T . (2002). Mol Pathol 55: 220–226.

  19. Horii A, Nakatsuru S, Ichii S, Nagase H, Nakamura Y . (1993). Hum Mol Genet 2: 283–287.

  20. Howng SL, Wu CH, Cheng TS, Sy WD, Lin PC, Wang C et al. (2002). Cancer Lett 183: 95–101.

  21. Jenkins GJ, Doak SH, Parry JM, D'Souza FR, Griffiths AP, Baxter JN . (2002). Br J Surg 89: 824–837.

  22. Katoh M . (2001a). Int J Mol Med 8: 657–660.

  23. Katoh M . (2001b). Int J Oncol 19: 1003–1007.

  24. Katoh M . (2003). Int J Mol Med 12: 811–816.

  25. Kawano Y, Kypta R . (2003). J Cell Sci 116: 2627–2634.

  26. Kitadai Y, Haruma K, Tokutomi T, Tanaka S, Sumii K, Carvalho M et al. (1998). Clin Cancer Res 4: 2195–2200.

  27. Kobayashi K, Ouchida M, Tsuji T, Hanafusa H, Miyazaki M, Namba M et al. (2002). Gene 282: 151–158.

  28. Koppert LB, van der Velden AW, van de Wetering M, Abbou M, van den Ouweland AM, Tilanus HW et al. (2004). Br J Cancer 90: 892–899.

  29. Le Floch N, Rivat C, De WO, Bruyneel E, Mareel M, Dale T et al. (2005). FASEB J 19: 144–146.

  30. Lee AY, He B, You L, Dadfarmay S, Xu Z, Mazieres J et al. (2004). Oncogene 23: 6672–6676.

  31. Logan CY, Nusse R . (2004). Annu Rev Cell Dev Biol 20: 781–810.

  32. McManus DT, Olaru A, Meltzer SJ . (2004). Cancer Res 64: 1561–1569.

  33. Metzger R, Schneider PM, Warnecke-Eberz U, Brabender J, Holscher AH . (2004). Onkologie 27: 200–206.

  34. Morin PJ . (1999). Bioessays 21: 1021–1030.

  35. Morin PJ, Sparks AB, Korinek V, Barker N, Clevers H, Vogelstein B et al. (1997). Science 275: 1787–1790.

  36. Nishihira T, Hashimoto Y, Katayama M, Mori S, Kuroki T . (1993). J Cancer Res Clin Oncol 119: 441–449.

  37. Nusse R, Varmus HE . (1982). Cell 31: 99–109.

  38. Osterheld MC, Bian YS, Bosman FT, Benhattar J, Fontolliet C . (2002). Am J Clin Pathol 117: 451–456.

  39. Pham K, Milovanovic T, Barr RJ, Truong T, Holcombe RF . (2003). Mol Pathol 56: 280–285.

  40. Polakis P . (1999). Curr Opin Genet Dev 9: 15–21.

  41. Polakis P . (2000). Genes Dev 14: 1837–1851.

  42. Rhee CS, Sen M, Lu D, Wu C, Leoni L, Rubin J et al. (2002). Oncogene 21: 6598–6605.

  43. Ricken A, Lochhead P, Kontogiannea M, Farookhi R . (2002). Endocrinology 143: 2741–2749.

  44. Rockett JC, Larkin K, Darnton SJ, Morris AG, Matthews HR . (1997). Br J Cancer 75: 258–263.

  45. Saitoh T, Mine T, Katoh M . (2002). Int J Mol Med 9: 515–519.

  46. Seery JP, Syrigos KN, Karayiannakis AJ, Valizadeh A, Pignatelli M . (1999). Acta Oncol 38: 945–948.

  47. Shih IM, Yu J, He TC, Vogelstein B, Kinzler KW . (2000). Cancer Res 60: 1671–1676.

  48. Suzuki H, Watkins DN, Jair KW, Schuebel KE, Markowitz SD, Chen WD et al. (2004). Nat Genet 36: 417–422.

  49. Tsuchiya T, Tamura G, Sato K, Endoh Y, Sakata K, Jin Z et al. (2000). Oncogene 19: 3642–3646.

  50. Verma UN, Surabhi RM, Schmaltieg A, Becerra C, Gaynor RB . (2003). Clin Cancer Res 9: 1291–1300.

  51. Worm J, Christensen C, Gronbaek K, Tulchinsky E, Guldberg P . (2004). Oncogene 23: 5215–5226.

Download references


We thank Patricia Martin, Maude Muriset, Adeline Cottier and Tian Lu for skillful laboratory work. This study was supported by a grant from the Ligue Suisse contre le Cancer (Grant KLS-01327-02-2003).

Author information



Corresponding author

Correspondence to J Benhattar.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Clément, G., Braunschweig, R., Pasquier, N. et al. Alterations of the Wnt signaling pathway during the neoplastic progression of Barrett's esophagus. Oncogene 25, 3084–3092 (2006).

Download citation


  • Barrett's esophagus
  • Wnt pathway
  • DNA methylation
  • APC
  • SFRP

Further reading


Quick links