Oncogenic Function of DACT1 in Colon Cancer through the Regulation of β-catenin

The Wnt/β-catenin signaling pathway plays important roles in the progression of colon cancer. DACT1 has been identified as a modulator of Wnt signaling through its interaction with Dishevelled (Dvl), a central mediator of both the canonical and noncanonical Wnt pathways. However, the functions of DACT1 in the WNT/β-catenin signaling pathway remain unclear. Here, we present evidence that DACT1 is an important positive regulator in colon cancer through regulating the stability and sublocation of β-catenin. We have shown that DACT1 promotes cancer cell proliferation in vitro and tumor growth in vivo and enhances the migratory and invasive potential of colon cancer cells. Furthermore, the higher expression of DACT1 not only increases the nuclear and cytoplasmic fractions of β-catenin, but also increases its membrane-associated fraction. The overexpression of DACT1 leads to the increased accumulation of nonphosphorylated β-catenin in the cytoplasm and particularly in the nuclei. We have demonstrated that DACT1 interacts with GSK-3β and β-catenin. DACT1 stabilizes β-catenin via DACT1-induced effects on GSK-3β and directly interacts with β-catenin proteins. The level of phosphorylated GSK-3β at Ser9 is significantly increased following the elevated expression of DACT1. DACT1 mediates the subcellular localization of β-catenin via increasing the level of phosphorylated GSK-3β at Ser9 to inhibit the activity of GSK-3β. Taken together, our study identifies DACT1 as an important positive regulator in colon cancer and suggests a potential strategy for the therapeutic control of the β-catenin-dependent pathway.


Introduction
Mutations and the dysregulated expression of components of the ancient Wnt signaling pathway are linked to oncogenesis in multiple systems, and have been particularly implicated in the initiation of colon cancer [1,2,3,4,5,6,7,8]. Over 90% of colorectal cancers originate from active mutations in the Wnt pathway [1]. Mutations have been described in the adenomatous polyposis coli (APC) gene in cases of familial adenomatous polyposis (FAP), and this mutation also occurs in a high proportion of sporadic colorectal cancers [9,10]. APC mutations represent an early event in colorectal tumorigenesis [11].
b-catenin (official symbol CTNB1) is considered to be a central player in the Wnt signaling pathway. Although Wnt activation can occur through mutations that affect phosphorylation sites within exon 3 of b-catenin in a minority of colorectal tumors [12,13], many other components of the Wnt signaling pathway contribute to colorectal cancer via dysregulating the activity or localization of b-catenin [14,15].
The DACT1 (Dapper1/Dpr1) gene, located at chromosome 14q22.3, encodes a 836 amino acid protein with a putative leucine zipper (LZ) domain in the amino-terminal end and a consensus PDZ binding (PDZ-B) motif in the carboxy-terminal end that allows the DACT1 protein to interact with the Dishevelled (Dvl) PDZ domain [16]. Bioinformatic analyses have revealed that DACT1 mRNA is expressed in the amnion, fetal brain, eye, heart, adult brain medulla, gastric cancer (signet ring cell features), RER+ colon tumor, acute lymphoblastic leukemia, germ cell tumor, chondrosarcoma, and parathyroid tumors [17]. Furthermore, based on the evolutionary and functional conservation of Wnt signaling molecules, as well as the human chromosomal localization of DACT1, the DACT1 gene is also predicted to be a potent cancer-associated gene [17]. DACT1 has been reported to be downregulated in hepatocellular carcinoma [18]. A recent report identified a correlation between DACT1 expression in lung cancer and poor histological grade, large tumor size, extent of tumor invasion, and lymph node metastasis [19].
Although some studies have shown associations between DACT1 expression and cancer, the function of DACT1 in the WNT/bcatenin signaling pathway remains unclear. One possible mechanism is that Dpr stabilizes GSK-3b and axin in the APC complex, as shown by co-immunoprecipitation studies [16]. Another possibility is that Dpr competes with Fz for binding to the PDZ domain of Dvl, thereby blocking the signal transduction via Dvl, and hence inhibits the Dvl-mediated stabilization of bcatenin [20]. Yau et al. reported that human Dpr1 was downregulated in hepatocellular carcinoma, and this downregulation was correlated with the cytoplasmic accumulation of bcatenin [18]. However, in this report, we have found that DACT1 is overexpressed in colon cancer, and it acts to enhance cellular proliferation, migration and invasion in colon cancer cell lines. We have shown that DACT1 interacts with GSK-3b and b-catenin. We have further demonstrated that DACT1 stabilizes b-catenin via DACT1-induced effects on GSK-3b and directly interacts with b-catenin. We have also demonstrated DACT1 inhibits the activity of GSK-3b via increasing the level of phosphorylated GSK-3b at Ser9, which alters the subcellular location of b-catenin. It particularly promotes b-catenin levels at the plasma membrane and in the nucleus.

DACT1 is overexpressed in human colon carcinoma
To identify the potential roles of DACT1 in the development and progression of colonic carcinoma, we used quantitative realtime PCR (qRT-PCR) to assess the level of gene expression. We compared the expression in six cancer tissues to those in six paired samples of normal control colonic mucosa. The results showed that the level of DACT1 mRNA was significantly elevated in all six samples of colon cancer ( Figure 1A). The gene expression data were further confirmed by Western blotting, which were performed with the use of soluble cell lysates prepared from surgical colectomy specimens. The results confirmed that high levels of DACT1 protein are present in colon cancer ( Figure 1B).

DACT1 enhances cellular proliferation in vitro and colon tumorigenesis in vivo
To understand the function of DACT1 in colon cancer, we explored the effect of ectopic DACT1 expression on cellular growth in vitro in cell lines with different levels of endogenous DACT1 expression. Following transfection with a DACT1 cDNA expression construct, SW480 colon cancer cells, which express very low endogenous levels of DACT1, demonstrated a significant increase in cellular proliferation ( Figure 2A). Conversely, cellular prolifer-ation decreased when endogenous DACT1 expression was silenced by stable transfection with siRNA vectors that targeted DACT1 in the HCT116, LoVo and HT29 colon cancer cell lines, which have much higher endogenous levels of DACT1 ( Figure 2B-D). To confirm these results, endogenous DACT1 expression was silenced by stable transfection with another siRNA vectors (DACT1 siRNA1) that targeted distinct regions of DACT1 in the HCT116 cells and the same results were observed ( Figure S1A-B).
To examine the effects of the silencing of DACT1 on colon tumor growth in vivo, HCT116 cells that stably expressed a control siRNA or DACT1 siRNA were used. Cells (5610 6 ) were injected into the subcutaneous tissue of the nude mouse (n = 8 animals per group). Tumors were measured weekly with a vernier caliper, and their volumes were calculated according to the following formula: p/ 66length6width 2 [21]. All of the animals injected with HCT116 cells that expressed control siRNA developed tumors by 14 days after injection, while only 75% (6/8) of the DACT1 siRNA animals developed tumors. Fifty-six days after injection, the tumors of the control siRNA animals were significantly larger ( Figure 2F). The mean tumor volume was 2,068.786561.57 mm 3 in mice injected with HCT116 cells that expressed control siRNA, compared to the mean tumor volume in mice injected with HCT116 cells that expressed DACT1 siRNA (503.466178.90 mm 3 ) ( Figure 2E). These data demonstrate the important role of DACT1 in promoting the growth of colon cancer cells.

DACT1 enhances the migration and lack of anchorage of colon cancer cells
Anchorage-independent proliferation is a hallmark of oncogenic transformation and is regarded to be conducive to the proliferation of cancer cells away from their original site. With this knowledge, we examined the capacity of DACT1 to drive the anchorageindependent growth of colon cancer cells by soft agar colony formation assays. The overexpression of DACT1 enhanced the anchorage-independent proliferation of SW480 cells ( Figure 3A). Downregulation of DACT1 reduced the anchorage-independent potential of HCT116, LoVo and HT29 cells ( Figure 3B-D and S1C). Furthermore, we observed that DACT1 overexpression enhanced the migratory potential of SW480 cells by transwell assays (Figure 3E and I). Conversely, the migratory potential of cells was decreased when endogenous DACT1 expression was silenced by stable transfection with siRNA vectors that targeted DACT1 in HCT116, LoVo and HT29 cells ( Figure 3F-I).
Following these results, we conclude that DACT1 enhances anchorage independence in colon cancer cells and their migratory potential.

DACT1 enhances invasion of colon cancer cells in vitro or in vivo
The most dangerous feature of malignancy is the invasive and metastatic potential of tumor cells. To test whether cell invasion was affected by DACT1, the effects of DACT1 on cell invasion through Matrigel was investigated. Overexpression of DACT1 enhanced the invasion of SW480 cells ( Figure 4A and E). Downregulation of DACT1 reduced the invasive potential of HCT116, LoVo and HT29 cells through the Matrigel ( Figure 4B-E). To examine the effects of DACT1 silencing on the invasion of colon cancer cells in vivo, we infected HCT116 cells that expressed a control siRNA or DACT1 siRNA. Then, 5610 6 cells were injected into the subcutaneous tissue of each nude mouse (n = 8 animals per group). Fifty-six days later, the tumors of six out of eight control siRNA-transfected nude mice had invaded into the musculature, while only two of the tumors in DACT1 siRNAtransfected animals had invaded into the musculature. These data further show that DACT1 did affect the invasive potential of colon cancer cells in vitro and in vivo.

DACT1 regulates the cell cycle though increasing bcatenin levels
To elucidate the mechanism which caused the above results further, we tested the protein levels of the b-catenin and the Wnt/ b-catenin target genes, cyclin D1 [8] and c-Myc [22]. Our experiment showed that overexpression of DACT1 resulted in a significant increase in the total levels of b-catenin, cyclin D1 and c-Myc in SW480 cells ( Figure 5A). Conversely, DACT1 siRNA, which mediated the silencing of endogenous DACT1 expression, decreased the total levels of b-catenin, cyclin D1 and c-Myc levels in HCT116, LoVo and HT29 cells ( Figure 5B).
Considering that cyclin D1 and c-Myc are related to cell cycle regulation, we examined the effect of DACT1 overexpression and silencing on cell cycle regulation in colon cancer cells. In these experiments, all of the cells were synchronized at the G0/G1 phases by serum starvation. Cell cycle profiling by FACS indicated that the overexpression of DACT1 in SW480 cells was associated with an increase in the proportion of cells in the G0/G1 phases ( Figure 5C). Consistently, silencing of DACT1 expression with siRNA resulted in an increased number of cells in the S+G2/M phases in the HCT116, LoVo and HT29 cell lines ( Figure 5D-F). These results suggest that the S+G2/M accumulation of cells after DACT1 silencing is specifically associated with the loss of DACT1 protein. Collectively, these results indicate that DACT1 promotes the proliferation of colon cancer cells by facilitating the transition from the G1 to the S phase of the cell cycle.
Besides cyclin D1, CDK4 is another key factor that facilitates the G1/S transition. We noted that the overexpression of DACT1 in SW480 cells increased the expression of CDK4, while CDK4 levels were significantly reduced in HCT116, LoVo and HT29 cells that expressed DACT1 siRNA ( Figure S2). Together, these results clearly show that DACT1 stimulates cell proliferation and tumor growth by promoting the entry of cells into the cell cycle through increasing levels of b-catenin.

DACT 1 enhances migratory and invasive potential of colon cancer cells through changing the subcellular location of b-catenin
To reveal the means by which DACT1 increases b-catenin/ TCF-regulated gene expression, the precise location of b-catenin was observed under a laser scanning confocal microscope (LSCM) by immunofluorescence. The results showed that the higher expression of DACT1 not only increased the nuclear and cytoplasmic fractions of b-catenin, but also increased its membrane-associated fraction. The change in the level of the b-catenin membrane-associated fraction was the most significant finding ( Figure 6A-D and S1D).
To confirm these interesting results, extracts from control/ DACT1-overexpressing SW480 cells and control/DACT1-silenced HCT116 cells were separated into membrane, cytoplasmic, and nuclear fractions. Then, the relative abundance of b-catenin in these fractions was analyzed by Western blotting ( Figure 6E and F). Previous reports have demonstrated an important role for bcatenin in cell adhesion as a part of a protein complex that includes E-cadherin [23]. The E-cadherin expression pattern in colon cancer cells was unaltered by DACT1 silencing or overexpression ( Figure S3). These data suggest that increased levels of DACT1 affect b-catenin levels and b-catenin/TCFdependent transcription by the activation of the Wnt signaling pathway. They also suggest the possibility that DACT1 enhances the migratory and invasive potential of colon cancer cells via bcatenin-mediated stabilization of the adherens junction.
Correlation between DACT1 and membrane-associated b-catenin expression in vitro and in vivo From the results of the above experiments, we found that DACT1 was highly expressed in human colon cancers and that it promoted the cell growth, migration and invasion potential of colon cancer cells through its effects on b-catenin signaling. Based on these findings, we hypothesized that DACT1 and the expression levels of membrane-associated b-catenin would be positively correlated in colon cancer cell lines and primary colon cancers. As shown in Figure 7A and B, the levels of DACT1 and b-catenin in colon cancer cell lines were correlated. The highest levels of bcatenin at the cell membrane were observed in HCT116 cells with  high levels of endogenous DACT1. Intermediate levels of b-catenin at the cell membrane and DACT1 proteins were found in HT29 cells. Both membrane-associated b-catenin and DACT1 were minimally expressed in SW480 cells.
To study the relationship between DACT1 and membraneassociated b-catenin expression in colon cancer further, immuno-histochemical and immunofluorescence analyses were performed of b-catenin in human colon normal and adenocarcinoma tissues (the six above-mentioned cases). We observed focal staining of membrane-associated b-catenin in all six tumors (100%; Figure 7C-H). These results correlate well with the previous results with regards to b-catenin expression in colon cancer cell Cells were seeded in a transwell chamber and allowed to migrate across the chamber toward cell-specific conditioned medium for 24 h. Photomicrographs of stained migrating cells were taken under brightfield illumination (206). (I) Quantification of migration assay. Results were obtained from three separate experiments each performed in triplicate. Migration was determined by counting cells in six random microscopic fields per well (mean 6 SEM; *p,0.05 versus control group cells). doi:10.1371/journal.pone.0034004.g003  We were interested in clarifying the mechanism by which elevated levels of DACT1 resulted in increased b-catenin expression. In the absence of Wnt ligands, cytoplasmic levels of b-catenin are regulated by the destruction complex that contains axin, APC, and glycogen synthase kinase-3b (GSK-3b), where bcatenin is phosphorylated by GSK-3b at multiple serine and threonine residues in its N terminus. Phosphorylated b-catenin is then ubiquitinated, leading to its rapid proteasomal degradation [4,24,25,26,27,28]. To determine whether DACT1 increases bcatenin levels by influencing b-catenin stability, HCT116 cells (with or without DACT1 silencing) were incubated with 10 mg/ml cycloheximide (CHX) to prevent new b-catenin synthesis. Then, b-catenin levels were measured by Western blotting to reflect the rate of b-catenin protein degradation. Silencing of DACT1 expression significantly increased the b-catenin degradation rate, resulting in an evident reduction in the levels of remaining bcatenin ( Figure 8A). This effect of DACT1 on b-catenin stability was confirmed by an immunofluorescence assay. We found that bcatenin was degraded more rapidly in HCT116 cells that expressed DACT1 siRNA than in HCT116 cells expressing control DACT1 siRNA ( Figure 8B).
To ascertain the mechanism by which DACT1 influences bcatenin stability, we examined whether or not DACT1 interacts with the components of the multiprotein complex that regulates bcatenin stability. Our co-immunoprecipitation analysis revealed colocalization of DACT1 and GSK-3b, DACT1 and axin in SW480 cells stably transfected with a GFP-tagged DACT1 construct ( Figure 8C). Moreover, DACT1 did interact with b-catenin in  SW480 cells that overexpressed DACT1 ( Figure 8C). We then examined the colocalization of DACT1 with GSK-3b, and DACT1 with b-catenin in wild type HT29 cells (without APC or b-catenin mutations) ( Figure 8D-E). These consistent results indicated that DACT1 interferes with the GSK-3b-dependent phosphorylation of b-catenin by the destruction complex, and DACT1 affects bcatenin stability by interacting with b-catenin. DACT1 affects the subcellular localized b-catenin through inhibiting GSK-3b activity b-catenin is known to accumulate in the nucleus [29] and its levels are speculated to increase at the membrane lamellipodia, membrane adherens junctions and membrane ruffles [30,31] in response to Wnt signaling or the drug-mediated inhibition of GSK-3b activity. Moreover, based on our findings, we speculated that DACT1 inhibits GSK-3b. Some factors that inhibit GSK-3b via the phosphorylation of Ser9 have been reported [32,33]. We observed the protein expression levels of GSK-3b with phosphorylated Ser9 in HCT116 cells (with or without DACT1 silencing). The results showed that the levels of phosphorylated GSK-3b at Ser9 were significantly increased in HCT116 cells that expressed control siRNA ( Figure 9A and B). To confirm that DACT1 regulated the localization pattern of b-catenin through inhibiting GSK-3b, we tested the effect of GSK-3b inhibition on HCT116 cells that expressed either the control siRNA or the DACT1 siRNA. We observed a clear and significant enhancement in cells that displayed membrane-associated b-catenin staining after they were treated with 40 mM LiC1 for 1 h (Figure 9C and D). We further tested the expression of activated b-catenin in SW480 cells. The results revealed the increased accumulation of nonphosphorylated bcatenin in cytoplasm and particularly in the nuclei of SW480 cells that overexpressed DACT1 (Figure 9E and F). All of these results support the conclusion that DACT1 affects the subcellular localization of b-catenin through the inhibition of GSK-3b activity.   Our studies have shown that DACT1 is overexpressed in human colonic adenocarcinomas. This result is consistent with previous observation that DACT1 was upregulated in invasive breast tumors [34]. Our studies have also revealed that b-catenin and Dvl-2 are upregulated along with DACT1, and demonstrate that there is a trend toward poor survival in patients with tumors that have higher scores for DACT1 expression in colon adenocarcinomas [35]. These results suggest that the higher expression of DACT1 may have oncogenic functions in colon cancer. However, DACT1 has also been reported to be downregulated in hepatocellular carcinoma [18] and lung cancer [19], which suggests that the function of DACT1 may be dependent on cellular context. Here, our observations show that the expression of DACT1 enhances cellular proliferation and promotes tumor growth and metastasis in the SW480, HCT116, LoVo and HT29 cell lines with or without APC or b-catenin mutations.

DACT1 upregulates b-catenin levels in colon cancer cells
Xenopus Dpr, the ortholog of DACT1, was found to be a negative regulator in both Dvl-mediated canonical and noncanonical Wnt pathway by Cheyette et al. in 2002 [16]. Concurrently, Gloy et al. demonstrated that Fordo, another ortholog of DACT1 promotes Wnt signaling in axis formation and in eye and neural development in Xenopus embryos [36]. To date, the function of DACT proteins in the regulation of signaling pathways remains ambiguous, but it is very clear that DACT1 not only regulates embryonic development but may have an important role in tumorigenesis. Indeed, one study demonstrated that DACT1 was downregulated in hepatocarcinoma, and this downregulation was associated with the methylation of its promoter-associated CpG island and significantly correlated with the cytoplasmic accumulation of b-catenin [18]. However, Jiang et al. reported that the expression of DACT1 did not differ significantly between human colorectal tumors and control tissues, and the DACT1 promoter was partially methylated [37]. DACT1 expression in lung cancer was downregulated [19].
We found that DACT1 upregulated b-catenin in colon cancer cell lines. The higher expression of DACT1 not only increased the nuclear and cytoplasmic fractions of b-catenin, but also increased its membrane-associated fraction. In mammals, b-catenin is originally identified as a component associated with cadherins, which are Ca2+-dependent cell-cell adhesion molecules [38]. Besides its function in the cadherin complex, b-catenin has an essential role in the Wnt signaling pathway [39,40,41]. Our results showed that the E-cadherin expression pattern in colon cancer cells was unaltered by DACT1 siRNA or DACT1 overexpression. These results suggest that DACT1 affects the level of b-catenin through the Wnt signaling pathway. Cyclin D1 and c-Myc are Wnt target genes. They can regulate the cell cycle and promote cell proliferation. We found that increased level of DACT1 not only

The mechanisms of DACT1-induced upregulation of bcatenin in colon cancer cells
The stability of b-catenin is known to be properly regulated by the multiprotein complex containing axin, GSK-3b, and APC. In the absence of an activating Wnt signal, mediated via Wnt binding to the Frizzled/LRP5/6 coreceptor complex, cytoplasmic bcatenin is destabilized by the multiprotein complex. Axin acts as the scaffold of this complex and interacts with the other components, b-catenin, APC, and GSK-3b. The interaction of GSK-3b with axin in the complex facilitates efficient phosphorylation of b-catenin by GSK-3b. Phosphorylated b-catenin is then ubiquitinated, leading to its rapid proteasomal degradation. We found that DACT1 binds to axin and GSK-3b in colon cancer cells. The results support the fact that DACT1 interacts with the destruction complex to prevent the phosphorylation and subsequent ubiquitination of b-catenin. We also found that DACT1 interacted with b-catenin in the plasma membrane, which indicated that DACT1 may directly regulate the degradation of b-catenin, leading to the accumulation of cytoplasmic b-catenin. However, the mechanisms by which DACT1 regulates this multiprotein complex and b-catenin are unclear.
DACT1 has been identified as an interacting protein for Dvl and induces Dvl degradation via a lysosome inhibitor-sensitive and proteasome inhibitor-insensitive mechanism [42]. Dvl is a central mediator of Wnt signaling. It is activated by the binding of Wnt ligands to the Frizzled/low-density lipoprotein receptor-related protein (LRP) coreceptor at the cell surface. Activated Dvl binds to the axin/GSK-3b complex and inhibits the GSK-3b-induced degradation of b-catenin, therefore activating the expression of target genes via the b-catenin-T-cell factor-lymphoid enhancer factor complex [4,15]. Our data showed that DACT1 could downregulate the expression of Dvl-2 ( Figure S4), whereas the expression of Dvl-1 and Dvl-3 was not been affected by DACT1 ( Figure S5). The results suggest that DACT1 upregulates b-catenin level in colon cancer cells through direct regulation of the multiprotein complex and b-catenin. It is not clear whether DACT1 regulates the phosphorylation of Dvl-2.
DACT1 mediated the subcellular localization of b-catenin b-catenin can be found in three cell compartments: the plasma membrane, the cytoplasm and the nucleus. The soluble cytoplas-and DACT1 siRNA were treated for 1 h with GSK-3b inhibitory drugs (40 mM LiCl). Cells were stained and analyzed by microscopy to detect the expression of b-catenin. LiCl treatment of HCT116 cells increases the levels of b-catenin at the plasma membrane. (D) Representative Western blots of LiCl-treated HCT116 cells, which express either control siRNA or DACT1 siRNA. LiCl treatment of HCT116 cells increases levels of b-catenin at the plasma membrane. KDEL was used as a loading control. 1: HCT116 cells expressing control siRNA that had been treated with 40 mM LiCl for 1 h. 2: HCT116 cells expressing control siRNA that had not been treated with LiCl. 3: HCT116 cells expressing DACT1 siRNA that were treated by 40 mM LiCl for 1 h; 4: HCT116 cells expressing DACT1 siRNA that had not been treated with LiCl. (E) Photomicrographs of control siRNA and DACT1 siRNA expressing HCT116 cells immunostained with an anti-phospho-GSK-3b(Ser9) antibody (red). Silencing of DACT1 in HCT116 cells decreases levels of phospho-GSK-3b(Ser9). (F) Representative Western blots of HCT116 cells expressing control siRNA and DACT1 siRNA. Silencing of DACT1 in HCT116 cells decreases levels of phospho-GSK-3b(Ser9). GAPDH was used as a loading control. doi:10.1371/journal.pone.0034004.g009 mic pool of b-catenin is highly unstable; when it is stabilized by the Wnt signal, this results in nuclear localization. Nuclear b-catenin is a cofactor for a wide variety of transcription factors that regulate the expression of many genes involved in cell growth. At the membrane, b-catenin plays a pivotal role in cell-cell adhesion, linking the cytoplasmic domain of E-cadherin to a-catenin. The deregulation of b-catenin or E-cadherin leads to the loss of cell-cell adhesion [43]. We found that increased levels of DACT1 led to higher levels of membrane-associated b-catenin, but the Ecadherin expression pattern in colon cancer cells was unaltered by DACT1 siRNA or the overexpression of DACT1. We also found DACT1 enhanced anchorage independence, as well as the migratory and invasive potential of colon cancer cell lines. These results are consistent with the above theories. In addition, our data showed that the DACT1 expression level and membraneassociated b-catenin expression level were positively correlated in colon cancer cell lines and primary colon cancers. These results are consistent with our previous report that showed b-catenin was upregulated and is associated with DACT1 levels. This demonstrates that there is a trend toward poor survival in patients with tumors that have higher scores for DACT1 expression in colon adenocarcinomas [35].

Mechanisms by which DACT1 mediates the subcellular localization of b-catenin
Some reports have shown that b-catenin can shuttle in a bidirectional manner between the nucleus and cytoplasm, or between the nucleus and the cell membrane [44,45,46]. This is necessary to enable the cell to respond quickly to fluctuations in extracellular stimuli that require it to switch between nuclear transactivation and membrane cell adhesion or cell migration activities [30]. Previous studies have shown that the inhibition of GSK-3b promoted the accumulation of b-catenin in the nucleus [29] and at the plasma membrane [30,31], which our data confirmed. In our studies, increased level of phosphorylated GSK-3b at Ser9 were present in HCT116 cells that expressed control siRNA, which led to the higher expression of b-catenin in the nucleus, cytoplasm and at the plasma membrane. We further found that DACT1 interacted with GSK-3b. This result supports the finding that DACT1 regulates the subcellular location of bcatenin through inhibiting the activity of GSK-3b. Reports have shown that Dvl can inhibit the activity of GSK-3b [47,48]. Our studies showed that DACT1 could downregulate the expression of Dvl-2, whereas the expression of Dvl-1 and Dvl-3 was not been affected by DACT1 (Figure S4 and S5). These data further show that b-catenin accumulates in the nucleus and at the plasma membrane as a result of the inhibition of GSK-3b by DACT1.
The model for the mechanism by which DACT1 promotes the oncogenesis of colon cancer cells The data in the present study support the following model for the mechanism by which DACT1 promotes the oncogenesis of colon cancer cells ( Figure 10). In unstimulated, normal colon cells that lack DACT1, GSK-3b is present in the cytoplasm and forms the b-catenin destruction complex with axin and APC. This allows the destruction complex to phosphorylate b-catenin and target it for ubiquitin-mediated degradation. In colon cancer cells that express high levels of DACT1, we propose that DACT1 interacts with b-catenin, after which DACT1 binds to and inhibits GSK-3b, which results in the release of b-catenin from the destruction complex, increased intracellular b-catenin levels, and its subsequent accumulation in the nucleus. This would lead to the activation of downstream b-catenin/TCF-regulated target genes, and at the membrane, regulation of the adhesion or migration properties of the cell. These studies define a functional role of DACT1 in human tumorigenesis and, besides highlighting DACT1 as a potential therapeutic target in colon cancer, define a mechanism for the regulation of b-catenin in cancer.

Tissue specimens and cell lines
Colon carcinoma tissues were obtained from the Peking University Third Hospital, China. Samples were frozen in liquid nitrogen immediately after surgical removal and maintained at 80uC until use. All human tissue was collected using protocols that were approved by the Ethics Committee of the Peking University Health Science Center. Patients diagnosed with colon adenocarcinoma were included if they had no family history of CRC or secondary malignancy, and had not received radiotherapy or chemotherapy before surgery. The human colon cancer cell lines were obtained from the cell resources centers of the Shanghai Institutes for Biological Sciences (Shanghai, China).

Plasmid construction
The cDNA of human DACT1 extracted from 293T cells was subcloned into the pEGFP-N1 expression vector (Invitrogen, Carlsbad, CA, USA). The sequences were verified by DNA sequencing. All of the control siRNAs and DACT1 siRNAs constructs were kindly provided by Dr. Yeguang Chen (Tsinghua University, China).

Quantitative real-time PCR and semi-reverse transcription-PCR
Total RNA from the human tissues and cell lines were isolated using TRIzol reagent (Invitrogen) according to the manufacturer's protocol. The cDNA of cell lines were synthesized using a High-Capacity cDNA Archive Kit (Invitrogen). Primer sequences for DACT1 were as follows: sense, 59-CACAAGCGAACTGAC-TACCG-39; antisense, 59-GTAATTGCTCTGCTCGTCCT-39. The primers for GAPDH were: sense, 59 -A CAG T C CATGCCATCACTGCC-39; antisense, 59-GCCTGCTTCAC-CACCTTCTTG-39. Twenty-five cycles were used. The conditions for hot-start PCR were as follows: 95uC for 10 min, followed by amplification at 95uC for 1 min, 58uC for 30 s and 62uC for 30 s for DACT1 and GAPDH, respectively, and 72uC for 30 s, with a final extension at 72uC for 10 min.
The cDNA of human tissues served as a template in quantitative real-time PCR utilizing TaqMan Gene Expression Master Mix and TaqMan Gene Expression assay probes for DACT1 (Hs00420410_m1), or GAPDH (Hs03929097_g1; Applied Biosystems, Foster City, CA, USA) and an ABI 7500 Fast Sequence Detection System. All reactions were performed in triplicate. The DACT1 mRNA expression of different group specimens was normalized to endogenous GAPDH. The relative DACT1 mRNA levels were presented as unit values of 2 2DCt = 2 2(Ct (GAPDH)2Ct (DACT1)) , where Ct is the threshold cycle value defined as the fractional cycle number at which the target fluorescent signal passes a fixed threshold above the baseline.

Immunoblot analysis
After the frozen tissues were homogenized in liquid nitrogen and the cultured cells were digested by Trypsin/EDTA Solution (pH 7.4; 0.025% trypsin, 0.5 mM EDTA, 1 mM sodium pyruvate and 10 mM HEPES), whole-cell lysates were prepared by incubating cells in ice-cold lysis buffer (50 mM Tris-HCl [pH 8.0], 150 mM NaCl, 1% Triton X-100 and 100 mg/ml of phenylmethylsulfonyl fluoride) for 30 min. The tissue samples and cells were sonicated for 8 s and then placed on ice for 5 min. The lysates were then centrifuged at 12,000 g for 20 min at 4uC. The supernatant was collected, aliquotted and frozen at 280uC before use. Protein concentrations were determined with a protein assay kit

Cell culture
All of the cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum (Invitrogen), nonessential amino acids, L-glutamine, and penicillin/streptomycin in an atmosphere that contained 5% CO 2 at 37uC.

Creation of stable cell lines
SW480 cells were stably transfected with pEGFP-N1-DACT1 and pEGFP-N1 empty vectors so that they would stably overexpress DACT1. The selection for G418 resistance was initiated 72 h after transfection by the addition of 300 mg/ml of G418 (Invitrogen) to the culture medium. The selection media was changed every 3-4 days for several weeks, and clones of resistant G418 cells were isolated and expanded for further characterization. DACT1 expression in stably transfected cells was verified by Western blotting, as described above.
To silence DACT1 expression, HCT116 and LoVo and HT29 cells were stably transfected with DACT1 siRNA and control siRNA expression vectors. Selection for puromycin resistance was initiated 72 h after transfection by the addition of 5 mg/ml puromycin (Invitrogen) to the culture medium. The selection media was changed every 3-4 days for several weeks, and clones of puromycinresistant cells were isolated and expanded for further characterization. The silencing of DACT1 expression in stably transfected cells was verified by Western blotting, as described above.

Proliferation assay
Cells were seeded in 24-well plates at 1610 4 cells per well in a volume of 1 ml of supplemented medium. The supplemented medium was changed every day. Cell numbers were counted every 24 h with the use of a NucleoCounter cell counter (Chemomete, Denmark).

Soft agar colony assay
Cell suspensions (1,000 cells) were prepared using 0.6% Noble agarose (Becton Dickinson) and overlaid onto a 24-well plate that contained a solidified lower layer of 0.6% agarose in medium. Once the top layer had solidified, 1 ml of medium was placed on the top of the cell layer. After treatment, plates were incubated for 3 weeks and colonies were counted with the assistance of a stereomicroscope.

Transwell migration assay
For transwell migration assays, 5610 4 cells were plated into the top chamber on a noncoated membrane (24-well insert; pore size, 8 mm; Corning Costar, NY, USA); cells were then allowed to migrate toward the medium that contained serum in the lower chamber. Cells were fixed with methanol after 24 h of incubation and stained with 0.1% crystal violet (2 mg/mL; Sigma-Aldrich). The number of cells that invaded through the membrane was counted under a light microscope (206, three random fields per well).

Transwell invasion assay
For the invasion assay, 5610 4 cells were plated in the top chamber onto the Matrigel-coated membrane (24-well insert; pore size, 8 mm; Millipore). Cells were plated in medium without serum, and medium supplemented with serum was used as a chemoattractant in the lower chamber. The cells were incubated for 24 h and cells that did invade through the pores were removed with a cotton swab. Cells on the lower surface of the membrane were fixed with methanol and stained with crystal violet. The number of cells invading through the membrane was counted under light microscopy (206, three random fields per well).

Flow cytometry analysis
After all of the cells were synchronized to the G0/G1 phases by serum starvation, the cells were cultured for 72 h. Then, the cells were washed twice with ice-cold phosphate buffered saline (PBS), harvested, fixed with ice-cold PBS in 70% ethanol, and stored at 220uC for 30 min. After fixation, the cells were incubated with RNase A (0.1 mg/ml, Sigma-Aldrich) at 37uC for 30 min, stained with propidium iodide (50 mg/ml, Sigma-Aldrich) for 30 min on ice in the dark and analyzed for DNA content using a flow cytometer (FACSCalibur; BD Biosciences). Data were collected in list mode for at least on 10,000 events and analyzed using Mod Fit 2.0 software.

Immunohistochemical analysis
The paraffin-embedded colon tissue sections (4-mm thick) were cut, deparaffinized, and subjected to a heat-induced epitope retrieval step. Endogenous peroxidase activity was blocked with 1% (v/v) hydrogen peroxide in distilled water. To block nonspecific binding, the sections were incubated with 2% (g/v) BSA for 60 min at room temperature. Subsequently, samples were incubated with a rabbit polyclonal b-catenin antibody (Cell Signaling Technology), diluted to 1:200 in PBS for spending the night at 4uC. For negative controls, sections were incubated with PBS without the primary antibody. For detection, specimens were sequentially incubated with biotinylated mouse anti-rabbit immunoglobulin G (IgG) and streptavidin-horseradish peroxidase. bcatenin expression levels in the stained sections were evaluated by an experienced pathologist.

Nuclear and membrane fractionation
Nuclear and cytoplasmic proteins were extracted using a NE-PERN Nuclear and Cytoplasmic Extraction Reagents Kit (Pierce, Rockford, IL, USA). Membrane proteins were extracted using a Mem-PER Eukaryotic Membrane Protein Extraction Kit (Pierce), according to the instructions of the manufacturer.

In vivo tumorigenicity studies
Six-to 8-week-old female nude mice were obtained from the Animal Center of Peking University Health Science Center (Beijing, China). Eight animals per group were used in each experiment. Briefly, 5610 6 HCT116 cells (expressing control siRNA or DACT1 siRNA) in 200 ml PBS were injected into the subcutaneous tissue of the nude mouse. Tumors were measured weekly using a vernier caliper, and the tumor volume was calculated according to the following formula: p/66length6width 2 [21]. Tumors were collected at the time of sacrifice 8 weeks after infection. All studies were approved by the Animal Care Committee of Peking University Health Science Center.

Statistical analysis
Data are represented as mean 6 standard error of the mean (SEM) from at least three independent experiments. The significance of differences between groups was evaluated by Student's t-test or ANOVA. P-values,0.05 were considered to be significant.