Skip to main content

Thank you for visiting nature.com. 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.

Wnt-5a has tumor suppressor activity in thyroid carcinoma

Abstract

Stabilization of β-catenin by inhibition of its phosphorylation is characteristic of an activation of the canonical Wnt/β-catenin signaling pathway and is associated with various human carcinomas. It contrasts to an as yet incompletely characterized action of an alternative noncanonical Wnt signaling pathway on neoplastic transformation. The aim of the present study was to test the effects of a member of the noncanonical Wnt signaling pathway, Wnt-5a, in primary thyroid carcinomas and in thyroid carcinoma cell lines. Compared to normal tissue Wnt-5a mRNA expression was clearly increased in thyroid carcinomas. Immunohistochemically, a bell-shaped response was observed with low to undetectable levels in normal tissue and in anaplastic tumors whereas differentiated thyroid carcinomas showed strong positive immunostaining for Wnt-5a. Transfection of Wnt-5a in a thyroid tumor cell line FTC-133 was able to reduce proliferation, migration, invasiveness and clonogenicity in these cells. These effects of Wnt-5a are associated with membranous β-catenin translocation and c-myc oncogene suppression and are mediated through an increase in intracellular Ca2+ release, which via CaMKII pathways promotes β-catenin phosphorylation. Specific inhibition of β-catenin phosphorylation by W-7, a calmodulin inhibitor, or by KN-93, a CaMKII inhibitor, supports these findings whereas PKC inhibitors were without effect. This interaction occurs downstream of GSK-3β as no Wnt-5a effect was seen on the Ser9 phosphorylation of GSK-3β. Our data are compatible with the hypothesis that Wnt-5a serves as an antagonist to the canonical Wnt-signaling pathway with tumor suppressor activity in differentiated thyroid carcinomas.

Introduction

Wnts are secreted proteins that exert their effects through activation of distinct intracellular signaling pathways. Based on their biological activities in specific assays, vertebrate Wnt have been divided into canonical signaling with transforming activities in mammary epithelial cells and noncanonical pathways. Members of the canonical Wnt signaling pathway act via Fz receptors to induce β-catenin stabilization by prevention of its phosphorylation, which targets β-catenin for proteosomal degradation. Stabilized β-catenin in this canonical Wnt pathway enters the cell nucleus and associates with members of the T-cell factor and lymphoid enhancer factor (TCF/LEF-1) family of transcription factors where it stimulates expression of Wnt/β-catenin target genes including regulators of cell growth and proliferation, modulators of cell death pathways and cell–cell communication (He et al., 1998; Van der Heyden et al., 1998; Mann et al., 1999; Tetsu and McCormick, 1999; Yamada et al., 2000; Zhang et al., 2001). This transforming activity of the Wnt/β-catenin pathway, which was initially delineated in the mammary epithelial cell line, C57MG, appears to be involved in the carcinogenesis of a number of human epithelial carcinomas (Kitaeva et al., 1997; Fukuchi et al., 1998; Miyoshi et al., 1998; Palacios and Gamallo, 1998; Chan et al., 1999; Gamallo et al., 1999; Kobayashi and Sagae, 1999; Koch et al., 1999; Mirabelli-Primdahl et al., 1999; Nhieu et al., 1999; Park et al., 1999; Wright et al., 1999; Jeng et al., 2000; Woo et al., 2001).

Signaling and physiological function of the noncanonical group of Wnt-proteins is currently only incompletely understood. These Wnts do not transform C57MG cells. Indirect data in mammalian cells suggest that these nontransforming subclasses of Wnt proteins, which includes Wnt-4, -5a and -11, may trigger intracellular Ca2+ release to activate Ca2+-sensitive enzymes in a G-protein-dependent manner (Slusarski et al., 1997a, 1997b; Kuhl et al., 2000a, 2000b). Further signal transduction and the potential interaction of canonical and noncanonical Wnt pathways are still elusive. In addition, noncanonical Wnt signaling in mammalian cells can activate a route that is equivalent to the planar polarity pathway in Xenopus and Drosophila, which is known to participate in control of the cytoskeleton via activation of RhoA and JNK (Pandur et al., 2002; Lustig and Behrens, 2003; Veeman et al., 2003).

Here, we studied Wnt-5a expression in normal thyroid tissue and in thyroid carcinomas related to their grades of differentiation. We link Wnt-5a expression in thyroid carcinoma cell lines to Ca2+ release and by a CamKII-dependent mechanism directly to β-catenin phosphorylation and degradation. Finally, we demonstrate that Wnt-5a is able to decrease cell proliferation, migration, invasiveness and clonogenicity in a thyroid carcinoma cell line and propose Wnt-5a as a tumor suppressor.

Results

Expression of Wnt-5a in human thyroid tissues and thyroid carcinoma cell lines

Northern blot analysis showed a characteristic differential expression of Wnt-5a in thyroid carcinomas and normal tissues (Figure 1a). Normal thyroid tissue expressed very low or undetectable steady-state Wnt-5a mRNA levels. In thyroid carcinomas expression of Wnt-5a was significantly increased by an average of 3.2-fold (Figure 1b). This was supported in thyroid carcinoma cell lines where a strong signal for Wnt-5a was observed in one follicular (FTC-238), two papillary (BcPAP and TPC) and two anaplastic carcinoma cell lines (Hth-74 and C-643).

Figure 1
figure1

Expression of Wnt-5a in thyroid carcinomas and thyroid normal tissues by Northern blot analysis. (a) Northern blot showing 5-kb Wnt-5a and 1.2-kb GAPDH specific bands in same tissues. In total, 3 μg per lane of mRNA from each thyroid tissue was blotted onto a nylon membrane. The cDNA probe of Wnt-5a gene was labeled with [α-32P]deoxy-CTP and used to detect Wnt-5a expression. (b) Summary of Wnt-5a mRNA expression in thyroid carcinomas and thyroid normal tissues. The results are corrected for GAPDH and displayed in units of phosphostimulated luminescence (PSL; arbitrary units). Horizontal bars indicate the average score within each group of specimens

Immunohistochemistry using Wnt-5a specific antibodies verified the differential expression of Wnt-5a in the same normal and malignant thyroid tissues on a translational level. None of the normal thyroid samples stained positive for Wnt-5a (Figure 2a and b), whereas 10 of 11 papillary (PTC) and all follicular (FTC) carcinomas exhibit strong positive immunostaining for Wnt-5a (Figure 2c–f). All positive tumors showed a cytoplasmatic Wnt-5a staining with a high rate of Wnt-5a positive cells (84% on average for PTC and 67% for FTC). In contrast, in all undifferentiated thyroid carcinomas, Wnt-5a immunostaining was completely lost (Figure 2g and h).

Figure 2
figure2

Immunohistochemical analysis of Wnt-5a protein expression in normal and carcinoma thyroid tissues. (a) Negative Wnt-5a expression in normal thyroid tissue; (b) details of (a); (c and e) Strong Wnt-5a expression in PTC and FTC, respectively; (d) details of (c) and (f) details of (e); (g) loss of Wnt-5a expression in ATC; (h) details of (g). Magnifications: 10 × 40 in (a, c, e, g); 10 × 63 in (b, d, f, h)

Wnt-5a stimulates an increase in cytosolic free Ca2+ in thyroid carcinoma cell line

The noncanonical Wnt signaling pathway has been linked to PKC and intracellular calcium activity (reviewed in Veeman et al., 2003). In order to explore the possible mechanism of Wnt-5a action, we analysed the ability of Wnt-5a to increase cytosolic free Ca2+ ([Ca2+]i) in FTC-133 cells. The cells were stimulated with increasing concentration of recombinant Wnt-5a (R&D Systems, Wiesbaden, Germany) and [Ca2+]i was measured by using the fluorescent calcium indicator Fura-2/AM. Stimulation with Wnt-5a caused a rapid rise (occurred after 10 –20 s) in [Ca2+]i in a concentration-dependent manner. Incubation with 0.5, 1.0 or 2 μg/ml Wnt-5a increased [Ca2+]i 1.7-, 4.3- or 12-fold, respectively (Figure 3a). The Ca2+ signal was followed by a gradual decay (measured within 6 min) and returned to baseline at the time point of 180 s after the substance had been washed off. Stable transfection with Wnt-5a induces a two-fold increase of [Ca2+]i in FTC-133 cells, whereas transfection with empty vector had no effect on [Ca2+]i levels. In addition, stimulation of these cells with low micromolar concentration (0.01 mM) of adenosine triphosphate (ATP), a Ca2+-phosphatidylinositol signaling pathway activator, caused a increase in [Ca2+]i with a large initial peak in Wnt-5a transfected cells and fast undetectable response in FTC-133/empty vector transfectants. Moreover, in Wnt-5a transfected cells the dose response of intracellular Ca2+ activity was shifted to higher sensitivity (Figure 3b).

Figure 3
figure3

Effect of Wnt-5a on [Ca2+]i in FTC-133 cells. The cells were loaded with 5 μ M fura-2/AM and the fluorescence from a single cell was recorded with a dual excitation spectrofluorometer system. The FTC-133 cells were stimulated with indicated concentrations of recombinant Wnt-5a (a) and the FTC-133/Wnt-5a or FTC-133/pCIN4 cells were stimulated with indicated concentrations of ATP (b). The mean and s.e. of six to eight cells from two independent experiment are shown (*P<0.05; **P<0.02)

A direct signaling interaction between Wnt-5a and β-catenin

Previous studies of our group (Helmbrecht et al., 2001) have shown that degradation resistant β-catenin is active as part of the canonical Wnt signaling pathway and increases TCF/LEF-1 dependent activity indicated by an increased luciferase activity in the reporter gene assay. This was supported by a stimulatory effect of mutated β-catenin on FTC-133 proliferation and clone formation in specific proliferation and clone formation assays (Figure 4).

Figure 4
figure4

Effect of β-catenin on FTC-133 cell growth and clonogenicity. Cell proliferation and clonogenicity were determined in FTC-133 cells stable transfected with an expression vector carrying the cDNA of a mutated β-catenin or with empty vector (pCi neo). (a) Cell proliferation was defined by MTT colorimetric method and data were expressed as a change in absorbance at 570 nm after culture for 2, 4 and 6 day relative to initial values obtained 3 h after plating. (b) For clonogenic assay, the cells were plated at a density of 1000 viable and cultured 7 days. After staining with 1% crystal violet, the colonies containing more than 50 cells were counted by microscopy. The FTC-133 cells displays increase in clonogenicity after β-catenin overexpression. All experiments were performed in triplicate and the results from three separate sets of experiments were averaged. Results are shown as mean±s.e. (*P<0.05)

The increase in proliferation induced by stabilized β-catenin is small nevertheless significant. An endogenously activated Wnt/β-catenin pathway in these tumor cells may explain these findings. However, to test the specificity and physiological importance of β-catenin dependent proliferation control in thyrocytes, we transfected FTC-133 cells with dominant-negative TCF-4 (dnTCF-4) that lack the β-catenin binding domain. Expression of dnTCF-4 in thyroid cells resulted in a significantly reduction of the cell proliferation (Figure 5). This effect was apparent after 48 h of dnTCF-4 transfection. These results confirm that β-catenin/TCF-driven transcription, which is upregulated in response to Wnt/β-catenin signaling, has an essential role in the thyroid tumorigenesis.

Figure 5
figure5

Control of cell proliferation by Wnt signaling. The FTC-133 cells were transient transfected with dnTCF-4 or with empty vector pCi-neo as a negative control. Cell proliferation was determined by MTT colorimetric method after culture for 48 h. Data are expressed as a change in absorbance at 570 nm. Data points represent mean±s.e. of three independent experiments plated in triplicate (*P<0.03)

We then evaluated a potential effect of Wnt-5a on the phosphorylation of β-catenin. The N-terminus of β-catenin contains a series of serine and threonine residues (Ser33, Ser37, Ser45, Thr41), which may be phosphorylated by casein kinase I (CKI) or glycogen synthetase kinase-3β (GSK-3β). These phosphorylation events mark β-catenin for ubiquitination and subsequent rapid proteasomal degradation. Activation of Wnt/β-catenin pathway induces the stabilization of β-catenin by prevention of its phosphorylation. To analyse the possible mechanism of Wnt-5a action in thyroid carcinomas we investigated the ability of Wnt-5a to modulate β-catenin phosphorylation and degradation in FTC-133 cells. Using immunoprecipitation (IP) and Western blotting techniques, we found that β-catenin levels were significantly decreased and phosphorylated β-catenin was detected only in Wnt-5a transfected cells. In contrast, in control cells (FTC-133/pCIN4), immunoreaction for β-catenin Ser37 was negative and β-catenin levels were unaltered (Figure 6a).

Figure 6
figure6

Immunoblot analysis of total and phosphorylated β-catenin (pSer37) and phosphorylated GSK-3β (pSer9 GSK-3β) in FTC-133 transfectants. (a) For detection of total β-catenin total cell lysates from FTC-133/Wnt-5a and FTC-133/pCIN4 were extracted and 10 μg of protein/lane were separated by SDS–PAGE and detected using β-catenin antibody. Lysates were probed with antitubulin to ensure equal loading of samples. For detection of pSer37 β-catenin the cells were extracted and immunoprecipitated with anti-β-catenin. The immunoprecipitates were immunoblotted with anti-pSer37 β-catenin. Overexpression of Wnt-5a increases the β-catenin degradation and phosphorylation at Ser37 in FTC-133/Wnt-5a cells. (b) Treatment with calmodulin inhibitor W-7 (75 μ M) or CaMKII inhibitor KN-93 (50 μ M) led to significantly decreased levels of pSer37 β-catenin in FTC-133/Wnt-5a cells. (c) Phosphorylation of GSK-3β was detected by performing immunoprecipitation with the antibody to GSK-3β followed by Western blotting for pSer9 GSK-3β. Treatment with neither calmodulin inhibitor W-7 (75 μ M) or CaMKII inhibitor KN-93 (50 μ M) did alter the phosphorylation pattern of GSK-3β whereas the positive control following treatment with lithium chloride (20 mM) significantly increased Ser9 phosphorylation of GSK-3β. The W-5 (75 μ M) and KN-92 (50 μ M) were used as negative controls for W-7 and KN-93, respectively. The amount of protein in cell lysates used in immunoprecipitation was analysed on a separate gel to probe for α-tubulin on Western blots

Given that Wnt-5a is able to increase cytosolic free Ca2+ in thyroid cells, we investigated the possibility that Wnt-5a might promote β-catenin degradation and, thus, suppress the canonical Wnt pathway activity through a Ca2+/calmodulin-dependent mechanism. Treatment with the calmodulin inhibitor, W-7, but not with its less potent dechlorinated analog, W-5, block Wnt-5a induced β-catenin phosphorylation in FTC-133/Wnt-5a cells. This effect is dependent on CaMKII activity. Treatment FTC-133/Wnt-5a cells with the CaMKII inhibitor, KN-93, again significantly decreased phosphorylated β-catenin in contrast to its ineffective analog, KN-92, which showed no such effect (Figure 6b). Finally, we investigated whether these effects depend on GSK-3β phosphorylation, which is well known to inhibit β-catenin phosphorylation. Treatment with neither calmodulin nor CaMKII inhibitors did alter the phosphorylation pattern of GSK-3β, whereas the positive control following treatment with lithium chloride (LiCl) significantly increased Ser9 phosphorylation of GSK-3β (Figure 6c).

As recent studies in other cell systems demonstrate stimulating result of Wnt-5a on PKC activity (Sheldahl et al., 1999; Kuhl et al., 2000a; Topol et al., 2003), we examined a potential effect of PKC in Wnt-5a dependent signaling in thyroid. Using the PKC inhibitor peptides 19–27 (100 μm) and 19–36 (0.5 μm) (Calbiochem, Schwalbach, Germany), no alteration in β-catenin stability was found that argues against an important impact of PKC in mediating Wnt-5a dependent effects.

Wnt-5a decrease thyroid carcinoma cell proliferation, migration, invasiveness and clonogenicity

Given that Wnt-5a expression is sufficient to modulate β-catenin phosphorylation and degradation, we next examined the effect of Wnt-5a on cell proliferation, migration, invasiveness and clonogenicity in a thyroid carcinoma cell line. Wnt-5a overexpression (by stable transfection of plasmid pCIN4 containing full-length Wnt-5a as compared to empty vector) or treatment of FTC-133 cells with recombinant Wnt-5a protein (0.5 μg/ml) significantly inhibited proliferation of these cells (Figure 7a). Furthermore, FTC-133/Wnt-5a transfectants were associated with a reduced number of clonogenic cells compared with the FTC-133/empty vector transfectants in a clonogenic assay (Figure 7b). The effect of Wnt-5a on motility and invasive capacity was studied by scratch wound assay and by a modified Boyden chamber assay, respectively. In these conditions, the invasiveness of FTC-133/Wnt-5a cells was reduced to 58% and cell spreading alonbg the edges of the wound was significantly decreased as compared to the FTC-133/pCIN4 cells (Figure 7c and d).

Figure 7
figure7

Effect of Wnt-5a on proliferation, clonogenicity, invasiveness and migration of FTC-133 cells. (a) FTC-133 cells were stable transfected with expression vector encoding Wnt-5a or with empty vector as control or were treated with recombinant Wnt-5a protein as described in Materials and methods. Cell proliferation was determined by MTT colorimetric method and data were expressed as a change in absorbance at 570 nm after culture for 2, 4 and 6 day relative to initial values obtained 3 h after plating. Wnt-5a inhibited proliferation of FTC-133 cell lines. (b) For clonogenic assay, the cells were plated at a density of 1000 viable and cultured 7 days. After staining with 1% crystal violet, the colonies containing more than 50 cells were counted by microscopy. The FTC-133 cells displays decrease in clonogenicity after Wnt-5a overexpression (*P<0.05). (c) Invasive profiles of FTC-133 cells were measured using Matrigel in modified Boyden chambers. The invasiveness of FTC-133/pCIN4 cells was taken as 100% (control) and that of FTC-133/Wnt-5a cells was expressed as a percentage of control. Compared with the control, Wnt-5a reduced invasion to 58%). (d) The FTC-133/pCIN4 and FTC-133/Wnt-5a cell motility was determined by wound migration assay. Confluent cells were carefully wounded using sterile pipette tips and then (after 12 and 24 h) were photographed under a phase contrast microscope. FTC-133/Wnt-5a cell spreading along the edges of the wound was significantly decreased as compared to the FTC-133/pCIN4 cells. All experiments were performed in triplicate and the results from three separate sets of experiments were averaged. Results are shown as mean±s.e. (**P<0.01, *P<0.05)

Immunohistochemical analysis of β-catenin and E-cadherin expression in FTTC-133/pCIN4 and FTC-133/Wnt-5a cells was performed to test whether these effects of Wnt-5a on proliferation, migration, invasiveness and clonogenicity are associated with membranous β-catenin translocation and with E-cadherin reexpression. In wild-type FTC-133 cells, E-cadherin is not expressed on mRNA nor protein level (data not shown). Similarly, Wnt-5a transfected cells are negative for E-cadherin, indicating that the effects are not mediated via E-type cadherins. However, using a pan-cadherin antibody, which recognizes all classical cadherins, we could show a clear stabilization of membranous cadherins along with a translocation of β-catenin to the membrane in FTC-133/Wnt-5a cells (Figure 8a and b).

Figure 8
figure8

Effect of Wnt-5a on β-catenin localization, pan-cadherin and c-myc expression in FTC-133 cells. (a) Induction of pan-cadherin by Wnt-5a expression in FTC-133 cell line. Equal amounts (10 μg) of membrane extracts from FTC-133/pCIN4 and FTC-133/Wnt-5a cells were analysed by Western blot with anti-pan-cadherin antibodies using standard procedures. Tubulin detection confirmed equal loading. (b) Immunohistochemical staining of β-catenin in FTC-133/pCIN4 and FTC-133/Wnt-5a transfectants. Sections from formalin-fixed, paraffin-embedded samples of FTC-133/PCIN4 or FTC-133/Wnt-5a cells were made on poly-L-lysine coated slides. Slides were dewaxed, rehydrated and pretreated with microwave irradiation, incubated with the primary monoclonal β-catenin antibody and subjected to tyramine amplification technique. Negative controls were performed by omitting the primary antibody in all series. Wnt-5a expression results in redistribution of the cytoplasmic β-catenin to the cell membrane. Nuclear accumulation is only seen in FTC-133/pCIN4 cells. (c) Localization of β-catenin in FTC-133 cells after treatment with recombinant Wnt-5a protein (0.5 μg/ml, 24 h) by immunofluorescence confocal microscopy. After treatment, FTC-133 cells were incubated with monoclonal β-catenin antibodies and with FITC-conjugated secondary antibodies. Confocal microscopy was performed on a Leica Inverted TCS with × 63 oil immersion objective. Wnt-5a stabilized the β-catenin within cell membrane and decreased β-catenin nuclear accumulation. (d) Downregulation of c-myc expression by Wnt-5a overexpression. The FTC-133/Wnt-5a and FTC-133/empty vector trancfectants were analysed for c-myc expression by Western blot and confocal immunofluorescence microscopy methods. For detection of c-myc, total cell lysates from FTC-133/Wnt-5a and FTC-133/pCIN4 were extracted and 10 μg of protein/lane were separated by SDS–PAGE and detected using c-myc antibody. Tubulin detection confirmed equal loading. The fluorescence stacks were as described in Materials and methods

To directly link these effects to Wnt-5a we treated FTC-133 cells with recombinant Wnt-5a protein (0.5 μg/ml). In comparison to untreated cells, Wnt-5a clearly stabilized the β-catenin within cell membrane and decreased β-catenin nuclear accumulation (Figure 8c). Western blot analysis and further immunofluorescence confocal microscopy confirmed that c-myc, a well-known target gene of Wnt/β-catenin signaling, was suppressed in Wnt-5a transfected cells when compared to FTC-133/empty vector transfectants (Figure 8d).

Discussion

The present study demonstrates an increased expression of steady-state Wnt-5a mRNA in thyroid tumors, whereas in normal thyroid tissue Wnt-5a expression was low or undetectable. This pattern fit to observations in other epithelial type tumors where Wnt-5a mRNA is overexpressed namely in carcinomas of the lung, prostate, metastatic melanoma and in squamous head and neck carcinomas (Iozzo et al., 1995; Rhee et al., 2002). However, not all epithelial type tumors exhibit high Wnt-5a expression as reports on endometrial carcinomas show low Wnt-5a mRNA levels (Bui et al., 1997).

Immunohistochemically only thyrocytes but not stroma cells stained positive for Wnt-5a, suggesting that Wnt-5a upregulation is restricted to the thyrocyte. Comparable to studies in mammary carcinomas the expression pattern did not parallel mRNA studies. Using immunostaining a bell-shaped expression pattern was observed with low to absent staining in normal thyroid, strong positive immunostaining for Wnt-5a in differentiated thyroid carcinomas but a complete loss in all anaplastic tumors, indicating a post-translational dysregulation in anaplastic thyroid carcinomas. This discrepancy between transcriptional and translational control fits to observations in other tumor types such as aggressive invasive breast ductal tumors, where in the dedifferentiated state a selective loss of protein expression was found in parallel to an aggressive state of the tumor (Jonsson et al., 2002).

Our findings of an up- and downregulation of Wnt-5a expression are compatible with the idea that Wnt-5a has a tumor suppressor activity. If that is true it may be speculated that in differentiated thyroid carcinomas, Wnt-5a plays a role in the control of differentiation, proliferation and invasiveness of cells. This cellular defence mechanism may subsequently be lost in dedifferentiated thyroid carcinomas. To test this hypothesis we examined the effect of Wnt-5a on cell proliferation, migration, invasiveness and clonogenicity in a thyroid carcinoma cell line. The transfection of FTC-133 cells with full-length Wnt-5a significantly decreased cell proliferation, migration, invasiveness and clonogenicity of the cells. Moreover, Wnt-5a was able decrease nuclear expression of c-myc, a well-known target gene of Wnt/β-catenin cascade involved in the control of proliferation and differentiation. These results support a tumor suppressor activity of Wnt-5a, indirectly suggested from data of other cellular background. In a human renal cell carcinoma cell line, RCC233, ectopic expression of Wnt-5a results in suppression of cell growth (Olson et al., 1998). Similarly, by stable transfection of Wnt-5a into a human uroepithelial cell line, MC-T116, loss of tumor formation was induced in athymic nude mice and anchorage-independent cell growth was lost in soft agar (Olson et al., 1997). In mouse mammary cells, Wnt-5a was associated with decreased proliferation (Olson and Parkoff, 1994).

We next examined the potential mode of Wnt-5a action. Proliferation in FTC-133 cells was increased by transfection of a degradation resistant form of β-catenin supporting an important role of stabilized β-catenin in thyroid proliferation control. This effect mimics an activation of the canonical Wnt pathway, which interferes with β-catenin inactivation induced by GSK-3β dependent phosphorylation. When the canonical Wnt signaling pathway is activated, β-catenin phosphorylation is inhibited by an inactivation of GSK-3β through its phosphorylation at Ser9. This effect is specific and appears to be pathophysiologically important as transfection with dominant-negative TCF-4 clearly suppressed proliferation of FTC-133 cells. These data suggest that under basal conditions β-catenin dependent pathways stimulating proliferation are endogenously activated.

We thus analysed the possibility of Wnt-5a to interact with β-catenin and its phosphorylation status. In Wnt-5a transfected cells, β-catenin levels were significantly decreased as compared to controls and only in Wnt-5a transfected cells phosphorylated β-catenin was detectable. This suggests that Wnt-5a has direct effects on β-catenin degradation and indicates a role of Wnt-5a in the counterregulation of the canonical Wnt–β–catenin pathway. Our results are consistent with recent findings of Topol et al. (2003) describing a significant decrease in total β-catenin protein levels in human embryonic kidney cells by overexpression of Wnt-5a. In this work phosphorylated-β-catenin was not studied but the Wnt-5a induced downregulation of total β-catenin appeared to be independent of GSK-3β in these cells. To further evaluate the mechanism of Wnt-5a dependent β-catenin phosphorylation, we investigated the levels of Ser9 GSK-3β in Wnt-5a transfected vs shame transfected cells. The lack of any difference supports the idea that Wnt-5a is acting downstream of GSK-3β on the phosphorylation of β-catenin (data not shown).

To better evaluate the impact of Wnt-5a on migration and invasiveness, we investigated the distribution of β-catenin and E-cadherin as well as pan-cadherin. Wnt-5a overexpression clearly redistributed β-catenin to the membrane. The effects on migration appear however not to be mediated via E-cadherin but rather through the action of other cadherins. In contrast to recent publications in keratinocytes where E-cadherin is upregulated (Taki et al., 2003), FTC-133 cells negative for E-cadherin are not capable to upregulate E-type cadherin.

To better characterize the mechanism of Wnt-5a induced β-catenin phosphorylation, we evaluated a potential activation of the Wnt/Ca2+ pathway. Recent investigations have shown that Wnt-5a may stimulate [Ca2+]i signaling. In support of this hypothesis, experiments in zebrafish embryos and mouse teratocarcinoma cells indicate that Wnt-5a may signal through G-protein-linked phosphatidylinositol signaling and release of intracellular calcium (Slusarski et al., 1997a, 1997b; Ahumada et al., 2002). Elevated [Ca2+]i can activate calcineurin/NF-AT and/or CaMKII/TAK1-NLK cascades that lead to inhibition of Wnt/β-catenin signaling at the different level (Saneyoshi et al., 2002; Ishitani et al., 2003). We first analysed the ability of Wnt-5a to modulate cytosolic free Ca2+ in a thyroid carcinoma cell line. Stimulation with recombinant Wnt-5a protein or transfection of FTC-133 cells with full-length Wnt-5a revealed significantly increased levels of cytosolic free Ca2+ relative to controls indicating that Wnt-5a/Ca2+ signaling is active in thyroid carcinoma cells. This was supported by dynamic testing of the pathway with ATP, a Ca2+-phosphatidylinositol signaling pathway activator. In contrast to the almost undetectable response in control transfected FTC-133 cells, [Ca2+]i was increased with a large initial peak in Wnt-5a transfected cells and shifting of the dose–response curve to higher sensitivity. We further hypothesized that the effects on β-catenin phosphorylation may be mediated via Ca2+/calmodulin pathways. Treatment of Wnt-5a transfected cells with the calmodulin inhibitor W-7 significantly reduced levels of phosphorylated β-catenin, suggesting that Wnt-5a induced β-catenin phosphorylation and degradation is at least part of the noncanonical Ca2+/calmodulin pathway action. It is interesting to note that treatment with the CaMKII inhibitor KN-93 again induced a significant decrease of phosphorylated β-catenin. These findings contrast to the work of Topol et al. (2003) suggesting that Wnt-5a-induced β-catenin degradation does not require activation of CaMKII or NF-AT in embryonic kidney cells. They however corroborate a model proposed by Ishitani et al. (2003), in which Wnt-5a requires CaMKII to activate not β-catenin phosphorylation but to act on the TAK1-NLK pathway. Active NLK (NEMO-like kinase), a target of TAK1, phosphorylates TCF and prevents the β-catenin/TCF-LEF complex from binding DNA, thereby inhibiting the ability of β-catenin/TCF-LEF to activate transcription. In contrast to the present work, these data suggest an interference of Wnt-5a and CaMKII with the canonical Wnt/β-catenin pathway downstream of β-catenin. However, as phospho-β-catenin was not directly monitored, this discrepancy not necessarily represents exclusive pathways. On the other hand, treatment with PKC inhibitors are without effects indicating that PKC dependent signaling is of minor importance in mediating Wnt-5a effects.

In summary, Wnt-5a is overexpressed in differentiated thyroid carcinomas but lost in the anaplastic cell type. Wnt-5a transfection of a thyroid carcinoma cell line clearly demonstrates a decreased proliferation, migration, invasiveness and clonogenic activity. This suggests that Wnt-5a acting via an increase in intracellular calcium activity and leading to an increased phosphorylation of β-catenin may decrease the level of free (stabilized) β-catenin. This effect of Wnt-5a is associated with suppression of c-myc, a Wnt/β-catenin pathway target gene. This antagonistic action of Wnt-5a to the canonical Wnt signaling and the bell-shaped expression pattern of Wnt-5a in primary thyroid carcinoma suggest a tumor suppressor activity and may represent a rescue mechanism for the control of β-catenin dependent pathways in differentiated thyroid carcinomas.

Materials and methods

Thyroid tissues

Thyroid tissues were obtained from patients undergoing surgery for thyroid carcinomas (papillary (PTC), n=12; follicular (FTC), n=8; anaplastic (ATC), n=5). Normal thyroid tissues (n=11) were obtained by surgery for benign diffuse goiter or during surgery for a parathyroid adenoma. Tissue fragments were snap-frozen in liquid nitrogen and stored at −80°C. Histopathological diagnosis and classification was performed at the Department of Pathology of the Medical School of Hannover. All patients involved in this study gave informed consent.

Cell culture and transfection

The human papillary (BcPAP, NPA and TPC-1), follicular (WRO, FTC-133 and FTC-238) and anaplastic (HTH-74 and C-643) thyroid carcinoma cell lines were studied. The WRO, BcPAP, NPA, TPC-1 and HTH-74 cell lines were grown in RPMI-1640 medium (with HEPES and L-Glutamine, PAA, Cölbe, Germany). The C-643, FTC-133 and FTC-238 cell lines were cultured in DMEM/HAM'S F-12 medium (with L-Glutamine). Growth mediums were supplemented with 10% fetal bovine serum, 100 μg/ml streptomycin and 100 U/ml penicillin. All cells were grown at 37°C in 5% CO2 and 95% air.

The human mutant β-catenin cDNA (S33Y) inserted into the XhoI–XbaI sites of expression vector pCI-neo (provided by H Clevers, Utrecht, the Netherlands) or human Wnt-5a cDNA inserted into the EcoRI sites of expression vector pCIN4 (provided by A Kispert, Hannover, Germany) or empty vectors (pCI-neo or pCIN4) were transfected into FTC-133 cells using a FuGENE transfection reagent (Roche, Mannheim, Germany) following the manufacturer's instruction. After 24 h of transfection, the cells were fed with fresh medium in the presence of G418 (neomycin, 1 mg/ml). After 2 weeks, the cells were picked up and plated in a 60-mm culture dish at a density of 1–2 cells per dish. The cells were subsequently passaged five times separately in selective medium and Wnt-5a or β-catenin positive clones were isolated and used for experiments.

For transient transfection, FTC-133 cells were cultured overnight on six-well dishes and then transfected with dominant-negative TCF-4 (provided by H Clevers, Utrecht University, Utrecht, the Netherlands) using FuGENE transfection reagent (Roche, Mannheim, Germany) following the manufacturer's instruction.

Treatment with pharmacological inhibitors

The calmodulin or CaM kinase II activity in FTC-133/Wnt-5a or FTC-133/pCIN4 cells was blocked with 75 μ M W-7 (N-(6-Aminohexyl)-5-chloro-1-naphthalenesulfonamide) or 50 μ M KN-93 (2-[N-(2-hydroxy-ethyl)]-N-(4-methoxybenzenesulfonyl)]amino-N-(4-chlorocinnamyl)-N-methylbenzyl-amine), respectively, for 24 h. A measure of 75 μ M W-5 (N-(6-Aminohexyl)-1-naphthalenesulfonamide) and 50 μ M KN-92 2-[N-(4-Methoxybenzenesulfonyl)]amino-N-(4-chlorocinnamyl)-N-methylbenzylamine) were used as a negative controls for W-7 and KN-93, respectively, at same time. Treatment with 20 mM LiCl was performed for 24 h. After incubation with indicated agents, the cells were implicated for IP and Western analysis. All chemicals were obtained from Calbiochem (Schwalbach, Germany).

RNA extraction and Northern blot

Total RNA was prepared from tissue samples and cultured cells by using Tri-Reagent (Sigma, Taufkirchen, Germany). Messenger RNA was extracted from total RNA using Oligotex mRNA Mini Kit (Qiagen, Hilden, Germany). For Northern blotting, 3 μg of mRNA/lane were separated on a 1.2% denaturing formaldehyde–agarose gel and capillary transferred onto a nylon membrane (Biodyne, PALL, Dreieich, Germany) and mRNA was fixed by UV crosslinking. The cDNA probe of Wnt-5a gene was labeled with [α-32P]deoxy-CTP (3000 Ci/mmol; Amersham, Freiburg, Germany) using random priming technique (Invitrogen, Karlsruhe, Germany) and hybridized with filters at 68°C for 1 h in the ExpressHyb solution (Clontech, Heidelberg, Germany). Filters were subsequently washed, visualized by autoradiography and then exposed to the storage phosphor screens and quantitated by Phosphorimager (Fujifilm Bas-1000, Berthold, Bad Wildbach, Germany). Obtained images were analysed using AIDA software (RZPD, Germany). The intensity of signals was normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a housekeeping gene.

Western blot analysis of total β-catenin, pan-cadherin and c-myc

Cells were lysed in 100 μl lysis buffer (50 mM HEPES (pH 7.4), 150 mM NaCl, 10% glycerol, 1 mM EDTA, 1 mM sodium orthovandate, 10 mM sodium pyrophosphate, 100 mM NaF, 1% Triton X-100, 10 μg/ml leupeptin, 10 U/ml aprotinin, 1 mM phenylmethylsulfonyl fluoride) or membrane protein extraction was performed by using Mem-PER Eukariotic Membrane Protein Extraction Reagent Kit (Pierce, Bonn, Germany). For Western blot analysis, proteins (10 μg/lane) were resolved in 7.5% SDS–PAGE and electrotransferred onto a nitrocellulose membrane (Millipore, Schwalbach, Germany) using standard procedures. After blocking with 4% nonfat dry milk in TBS-T buffer (20 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.1% Tween-20), the blots were probed with primary antibodies (anti-β-catenin mAb, BD Transduction Laboratories, Heidelberg, anti-pan-cadherin mAb, Sigma, Munchen, Germany or anti-c-myc mAb, Santa Cruz Biotechnology, Santa Cruz, CA, USA). The blots were then reacted with a secondary alkaline phosphate-conjugated anti-mouse antibodies (Dianova, Hamburg, Germany), followed by detection of the proteins with CSPD® chemiluminescent reagent (Roche, Mannheim, Germany). Finally, the blots were reprobed using antibodies against α-tubulin (1 : 1000; Santa Cruz Biotechnology, Santa Cruz, CA, USA) to ensure equal loading and transfer of proteins.

Detection of phosphorylated β-catenin and GSK-3β

Phosphorylation of β-catenin or GSK-3β was detected by performing IP with the antibody to β-catenin or GSK-3β followed by Western blotting for β-catenin-phosphoserine37 or phosphoserine9 GSK-3β. The cells were washed in PBS and lysed for 30 min in 1 ml lysis buffer at 4°C. The lysates were centrifuged (10 000 r.p.m. 10 min) and the supernatants were transferred to a new tube. For IP, lysates (800 μg protein) were incubated with 2 μg anti-β-catenin or anti-GSK-3β antibody (BD Transduction Laboratories, Heidelberg, Germany) for 6 h at 4°C. Specific antibody–antigen complexes were collected with 20 μl of protein-G Sepharose overnight at 4°C with gentle over head rotation. The samples were centrifuged, washed and suspended in SDS–PAGE sample buffer. After SDS–PAGE and transfer of proteins, membranes were incubated with phospho-β-catenin (Ser37) or phospho-GSK-3β (Ser9) antibody. For detection of phospho-β-catenin or phospho-GSK-3β, the membranes were incubated with alkaline phosphate-conjugated anti-mouse second antibodies (Dianova, Hamburg, Germany) at 1 : 5000 for 1 h, and washed again. Expressed proteins were visualized with CSPD® chemiluminescent reagent (Roche, Mannheim, Germany). The amount of protein in cell lysates used in IP was determined on a separate gel and Western blot probed for α-tubulin by using anti-α-tubulin antibody.

Measurement of cytosolic free Ca2+ ([Ca2+])i

For Ca2+ measurements the FTC-133, FTC-133/Wnt-5a or FTC-133/pCIN4 cells were loaded with 5 μ M fura-2/AM for 30 min at 37°C in medium containing 130 mM NaCl, 4.7 mM KCl, 1.2 mM KH2PO4, 1.2 mM MgSO4, 1.5 mM CaCl2, 10 mM glucose, 20 mM HEPES, 2% BSA and 0.1% pluronic acid, pH 7.4. After loading, the coverslips were washed, mounted in a temperature-controlled superfusion chamber (37°C) and placed on the stage of a Zeiss Axiovert IM 135 equipped with a × 40 Achrostigmat oil immersion objective (Zeiss, Oberkochen, Germany). The chamber was superfused at a flow rate of 0.9 ml/min using a similar medium as above but with 0.1 BSA and without pluronic acid. Fura-2/AM fluorescence from a single cell was recorded with a dual excitation spectrofluorometer system (Deltascan 4000, Photon Technology International, Wedel, Germany). Cytosolic free Ca2+ concentrations were calculated according to the following formula:

where Kd=225 nM, Rmax, Rmin and B are constants, which were determined in the superfusion chamber from solutions containing fura-2 free acid (1 μ M) and various concentrations of free Ca2+.

Proliferation and clonogenic assays

For growth experiment, the FTC-133 cells treated with recombinant Wnt-5a protein (0.5 μg/ml) or stable transfected with Wnt-5a or with mutated β-catenin or transient transfected with dnTCF-4 were seeded into a six-well plate at 2 × 104 cells/well and maintained in conditioned medium. Cell proliferation was determined by dimethyl-thiazol-diphenyltetrazolium bromide (MTT, Sigma, Taufkirchen, Germany) colorimetric method after culture for 2, 4 and 6 day by the following procedure: MTT (400 μg/ml in culture medium) was added to each well and incubated for 3 h at 37°C. Subsequently, it was solubilized with 0.04 N HCl/iso-propanol/3% sodium dodecyl sulfate and incubated for 20 min. Color development was determined using spectrophotometer (Ultrospec K, Biochrom, Berli, Germany) at a wavelength 570 nm. The experiment was repeated two times in triplicates (three parallel samples were used at each time point).

For clonogenic assay, the FTC-133 cells stable transfected with Wnt-5a or with mutant β-catenin or empty vector (pCIN4 or pCI-neo) were plated at a density of 1000 viable cells/well in six-well plates and cultured for 7 days at 37°C incubator. Then the cells were washed with PBS, fixed with a mixture of methanol and acetic acid (1 : 1) and stained with 1% crystal violet. The colonies containing more than 50 cells were counted by microscopy. Assays were performed in triplicate.

Matrigel invasion and migration assays

The invasive potential of FTC-133, FTC-133/pCIN4 and FTC-133/Wnt-5a cells was analysed using modified Boyden Matrigel Invasion Chambers (Costar Corporation, Cambridge, MA, USA) as described in the manufacture's protocol. Briefly, 700 μl of full media with fibronectin as a chemoattactant (10 μg/chamber) was added to the lower compartment of the chamber. In total, 105 cells suspended in 100 μl of the serum-free medium were placed in the upper compartment coated with 25 μg of basement membrane Matrigel (Becton Dickinson, Bedford, MA, USA) for 24 h at 37°C in a humidified 95% air/5% CO2 atmosphere. At the end of incubation, the cells on the upper surface of the filter and the invaded cells that migrated to the lower surface of the filter were collected separately. An MTT assay was then performed on harvested cells. The relative invasive rate was calculated as percent OD of the cells from the top of the membrane to the overall OD from the total cells. The assay was performed in triplicate.

Cell motility was assessed using a scratch wound assay as described (Meng et al., 2000). FTC-133/pCIN4 (control) and FTC-133/Wnt-5a cells were cultured in six-well dishes in the cultured medium. The confluent plates were carefully wounded using sterile pipette tips and then recultured in the same medium. Fibronectin (10 μg/ml) was added to the culture. After incubation for 12 and 24 h, the cells were photographed under a phase contrast microscope. The pattern of migration was observed. The experiments were repeated in triplicate.

Immunofluorescence and confocal microscopy

Cells were seeded on 12-mm diameter glass coverslips and used at a confluency of 70–90%. The cells were fixed in 4% paraformaldehyde for 20 min and permeabilized with 0.1% Triton X-100 (Fluka Sigma-Aldrich Chemie, Deisenhofen, Germany) in PBS for 5 min at room temperature. Incubations with primary antibodies (anti-β-catenin mAb, BD Transduction Laboratories, Heidelberg or anti-c-myc mAb, Santa Cruz Biotechnology, Santa Cruz, CA, USA) were followed by FITC-conjugated secondary antibodies (Sigma, Munchen, Germany) for 1 h at room temperature. Finally the cells were embedded in Vectashield mounting medium and viewed by confocal laser scanning microscopy (Leica Inverted TCS, Leica Microsystems, Heidelberg, Germany). Optical sections were recorded using × 63 oil immersion objective with a numerical aperture of 1.4 at a resolution of 1024 × 1024 pixels.

Multitissue block preparation and immunohistochemistry

Multitissue blocks were assembled according to tissue array technology developed in the Department of Pathology, Medical School of Hannover. In brief, all diagnostic H&E slides from each case were reviewed and transferred to the cutting side of the corresponding paraffin block. Using a sharpened needle as punching device, tumor tissue cores of 35 cases were melted together into one new homogenous multiblock. Sections were stained using the Shandon coverplate system in a Tecan Genesis Autostainer (Tecan, Crailsheim, Germany). Tissue peroxidase activity was blocked by incubation with 3% hydrogen peroxide for 8 min. The primary Wnt-5a antibody (Jonsson et al., 2002) was applied in the 1 : 200 dilution. Biotinylated secondary antibodies were used for the catalysed signal amplification technique (CSA, Dako, Hamburg, Germany). The final color reaction was carried out using new fuchsin as chromogen, hemalaun as light counterstaining and Kaiser's glycerine (VWR, Darmstadt, Germany) to mount the coverslips. Evaluation of the immunostainings was carried out by two independent observers (RvW and NK). We defined the positive cases those with more than 10% of neoplastic cells exhibiting immunoreactivity. The percentage of tumor cells showing Wnt-5a positivity was evaluated by counting at least four areas of 100 cells per slide, and an overall percentage (average) was calculated. The staining intensity was scored as weak, moderate or strong.

For β-catenin immunohistochemical detection 2 μ M thick sections from formalin-fixed, paraffin-embedded samples of FTC-133/PCIN4 or FTC-133/Wnt-5a cells were made on poly-L-lysine coated slides. Slide were dewaxed using xylene, rehydrated and pretreated with microwave irradiation for antigen retrieval as described above, incubated with the primary monoclonal β-catenin antibody (Transduction Laboratories, Inc., Lexington, KY, USA) at 4°C overnight (1 : 500 dilution) and subjected to tyramine amplification technique. Negative controls were performed by omitting the primary antibody in all series.

Statistical analysis

The results are expressed as a mean±s.e. Statistical analysis was carried out with Student's t-test. Differences were considered significant at P<0.05.

References

  1. Ahumada A, Slusarski DC, Liu X, Moon RT, Malbon CC and Wang HY . (2002). Science, 6, 2006–2010.

  2. Bui TD, Tortora G, Ciardiello F and Harris AL . (1997). Biochem. Biophys. Res. Commun., 239, 911–917.

  3. Chan EF, Gat U, McNiff JM and Fuchs E . (1999). Nat. Genet., 21, 410–413.

  4. Fukuchi T, Sakamoto M, Tsuda H, Maruyama K, Nozawa S and Hirohashi S . (1998). Cancer Res., 58, 3526–3528.

  5. Gamallo C, Palacios J, Moreno G, Calvo de Mora J, Suarez A and Armas A . (1999). Am. J. Pathol., 155, 527–536.

  6. He TC, Sparks AB, Rago C, Hermeking H, Zawel L, da Costa LT, Morin PJ, Vogelstein B and Kinzler KW . (1998). Science, 281, 1509–1512.

  7. Helmbrecht K, Kispert A, von Wasielewski R and Brabant G . (2001). Endocrinology, 142, 5261–5266.

  8. Ishitani T, Kishida S, Hyodo-Miura J, Ueno N, Yasuda J, Waterman M, Shibuya H, Moon RT, Ninomiya-Tsuji J and Matsumoto K . (2003). Mol. Cell. Biol., 23, 131–139.

  9. Jeng YM, Wu MZ, Mao TL and Chang MH . (2000). Cancer Lett., 152, 45–51.

  10. Jonsson M, Dejmek J, Bendahl PO and Andersson T . (2002). Cancer Res., 15, 409–416.

  11. Kitaeva MN, Grogan L, Williams JP, Dimond E, Nakahara K, Hausner P, DeNobile JW, Soballe PW and Kirsch IR . (1997). Cancer Res., 57, 4478–4481.

  12. Kobayashi K and Sagae S . (1999). Jpn. J. Cancer Res., 90, 55–59.

  13. Koch A, Denkhaus D, Albrecht S, Leuschner I, von Schweinitz D and Pietsch T . (1999). Cancer Res., 59, 269–273.

  14. Kuhl M, Sheldahl LC, Malbon CC and Moon RT . (2000a). J. Biol. Chem., 28, 12701–12711.

  15. Kuhl M, Sheldahl LC, Park M, Miller JR and Moon RT . (2000b). Trends Genet., 16, 279–283.

  16. Iozzo RV, Eichstetter I and Danielson KG . (1995). Cancer Res., 55, 3495–3499.

  17. Lustig B and Behrens J . (2003). J. Cancer Res. Clin. Oncol., 129, 199–221.

  18. Mann B, Gelos M, Siedow A, Hanski ML, Gratchev A, Ilyas M, Bodmer WF, Moyer MP, Riecken EO, Buhr HJ and Hanski C . (1999). Proc. Natl. Acad. Sci. USA, 96, 1603–1608.

  19. Meng Q, Goldberg ID, Rosen EM and Fan S . (2000). Breast Cancer Res. Treat., 63, 147–152.

  20. Mirabelli-Primdahl L, Gryfe R, Kim H, Millar A, Luceri C, Dale D, Holowaty E, Bapat B, Gallinger S and Redston M . (1999). Cancer Res., 59, 3346–3351.

  21. Miyoshi Y, Iwao K, Nagasawa Y, Aihara T, Sasaki Y, Imaoka S, Murata M, Shimano T and Nakamura Y . (1998). Cancer Res., 58, 2524–2527.

  22. Nhieu JT, Renard CA, Wei Y, Cherqui D, Zafrani ES and Buendia MA . (1999). Am. J. Pathol., 155, 703–710.

  23. Olson DJ, Gibo DM, Saggers G, Debinski W and Kumar R . (1997). Cell Growth Differ., 8, 417–423.

  24. Olson DJ and Parkoff J . (1994). Cell Growth Differ., 5, 197–206.

  25. Olson DJ, Oshimura M, Otte AP and Kumar R . (1998). Tumour Biol., 19, 244–252.

  26. Palacios J and Gamallo C . (1998). Cancer Res., 58, 1344–1347.

  27. Pandur P, Maurus D and Kuhl M . (2002). Bioessays, 10, 881–884.

  28. Park WS, Oh RR, Park JY, Lee SH, Shin MS, Kim YS, Kim SY, Lee HK, Kim PJ, Oh ST, Yoo NJ and Lee JY . (1999). Cancer Res., 59, 4257–4260.

  29. Rhee CS, Sen M, Lu D, Wu C, Leoni L, Rubin J, Corr M and Carson DA . (2002). Oncogene, 21, 6598–6605.

  30. Saneyoshi T, Kume S, Amasaki Y and Mikoshiba K . (2002). Nature, 417, 295–299.

  31. Sheldahl LC, Park M, Malbon CC and Moon RT . (1999). Curr. Biol., 1, 695–698.

  32. Slusarski DC, Corces VG and Moon RT . (1997a). Nature, 27, 410–413.

  33. Slusarski DC, Yang-Snyder J, Busa WB and Moon RT . (1997b). Dev. Biol., 1, 114–120.

  34. Taki M, Kamata N, Yokoyama K, Fujimoto R, Tsutsumi S and Nagayama M . (2003). Cancer Sci., 94, 593–597.

  35. Tetsu O and McCormick F . (1999). Nature, 398, 422–426.

  36. Topol L, Jiang X, Choi H, Garrett-Beal L, Carolan PJ and Yang Y . (2003). J. Cell Biol., 162, 899–908.

  37. van der Heyden MA, Rook MB, Hermans MM, Rijksen G, Boonstra J, Defize LH and Destree OH . (1998). J. Cell Sci., 111 (Part 12), 1742–1749.

  38. Veeman MT, Axelrod JD and Moon RT . (2003). Dev. Cell, 5, 367–377.

  39. Woo DK, Kim HS, Lee HS, Kang YH, Yang HK and Kim WH . (2001). Int. J. Cancer, 95, 108–113.

  40. Wright K, Wilson P, Morland S, Campbell I, Walsh M, Hurst T, Ward B, Cummings M and Chenevix-Trench G . (1999). Int. J. Cancer, 82, 625–629.

  41. Yamada T, Takaoka AS, Naishiro Y, Hayashi R, Maruyama K, Maesawa C, Ochiai A and Hirohashi S . (2000). Cancer Res., 60, 4761–4766.

  42. Zhang T, Otevrel T, Gao Z, Ehrlich SM, Fields JZ and Boman BM . (2001). Cancer Res., 61, 8664–8667.

Download references

Author information

Affiliations

Authors

Corresponding author

Correspondence to G Brabant.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Kremenevskaja, N., von Wasielewski, R., Rao, A. et al. Wnt-5a has tumor suppressor activity in thyroid carcinoma. Oncogene 24, 2144–2154 (2005). https://doi.org/10.1038/sj.onc.1208370

Download citation

Keywords

  • Wnt-5a
  • β-catenin
  • thyroid cancer
  • tumor suppression

Further reading

Search

Quick links