Tissue inhibitor of metalloproteinase 2 inhibits activation of the β-catenin signaling in melanoma cells

Abstract

The tissue inhibitor of metalloproteinase (TIMP) family, including TIMP-2, regulates the activity of multifunctional metalloproteinases in pathogenesis of melanoma. The Wnt/β-catenin pathway is constitutively activated and plays a critical role in melanoma progression. However, the relationship between TIMP-2 expression and β-catenin activity is still unclear. We hypothesize that TIMP-2 over expression inhibits the activation of the Wnt/β-catenin pathway in melanoma cells. Protein expression, distribution, and transcriptional activity of β-catenin were assayed in established stable melanoma cell lines: parental A2058 expressing, A2058 T2-1 over-expressing (T2-1), and A2058 T2R-7 under-expressing (T2R-7) TIMP-2. Compared to T2-1 cells at the basal level, T2R-7 showed significantly lower amount protein and weaker immunofluorescence staining of β-catenin. This regulation is through posttranslational level via ubiquitination. Functionally, proliferation and cell growth were lower in T2R-7 compared to A2058 and T2-1. Lithium treatment was used to mimics activation of the Wnt/β-catenin pathway. In T2R-7 cells under-expressing TIMP2, lithium significantly increased total β-catenin, nuclear β-catenin, and its downstream protein phosphor-c-Myc (S62). Nuclear β-catenin staining was enhanced in T2R-7. Beta-catenin transcriptional activity and cell proliferation were also increased significantly. Axins inhibit β-catenin pathway via GSK-3 β. We further found the ratio of p-GSK-3 β (S9) to β-catenin and protein levels of Axins were significantly lower, whereas downstream Wnt 11 was high in T2R-7 treated with lithium. Collectively, the high level of TIMP-2 protein inhibits the activation of the Wnt/β-catenin pathway, thus suppressing proliferation. Insights in the molecular mechanisms of TIMP-2 may provide promising opportunities for anti-proliferative therapeutic intervention.

Keywords:

• axin1
• axin2
• β-catenin
• lithium
• melanoma
• TIMP-2
• Wnt

Abbreviations

APC	=	Adenomatous polyposis coli
GSK-3b	=	Glycogen synthase kinase 3b
MMPs	=	Metalloproteinases
TCF/LEF	=	T-cell factor/L-lymphoid enhancer factor
TIMP	=	Tissue inhibitor of metalloproteinases

Introduction

Tissue inhibitor of matrix metalloproteinase (TIMP) family, including TIMP-1, 2, 3, and 4, naturally inhibits the activity of matrix metalloproteinases (MMPs).^1,2 MMP-2 expression is associated with the progression of melanoma.³ TIMP-2 can inhibit active MMP-2 as well as promote the activation of pro-MMP-2 by MT1-MMP.^4,5 TIMP-2 can act as an endogenous inhibitor of both angiogenesis and tumor growth, via effects dependent or independent of MMP inhibition. TIMP-2 over-expression reduces invasion and angiogenesis and protects B16F10 melanoma cells from apoptosis.⁶ TIMP-2 modulates critical signaling pathways, such as the NF-κB pathway,^7,8 vascular endothelial growth factor pathway,⁹ p38 MAPKs pathway,¹⁰ insulin-like growth factor,^11,12 etc. A recent study reports that TIMP-2 modulates cancer cell transcriptional profile and enhances E-cadherin/β-catenin complex expression in A549 lung cancer cells.¹³

Beta-catenin is a multi-functional protein.¹⁴ As an adhesion protein, β-catenin is associated with α-catenin and E-cadherin which plays a role in tumor suppressor gene.¹⁵ As a transcriptional factor, β-catenin acts as a key transcriptional co-activator and transmits extracellular signals for the activation of target genes in the canonical Wnt pathway,^16,17 which plays an important role in embryonic development and neoplasia.¹⁸ Under normal conditions, β-catenin associates with Axin and adenomatous polyposis coli (APC), a negative regulator of β -catenin. Beta-catenin is constitutively phosphorylated by glycogen synthase kinase 3β (GSK-3β), ubiquitinated, and degraded by the proteasome, thus maintaining appropriate cytoplasmic levels.¹⁹ Without the constitutive β-catenin degradation, cytoplasmic levels rise and allow β-catenin to accumulate in the nucleus. Once in the nucleus, β-catenin binds to the transcription factor T-cell factor/L-lymphoid enhancer factor (TCF/LEF) and activates multiple target genes, such as c-Myc, which promotes cell proliferation, and cancer progression.

Melanoma is the fifth leading cancer in males and the seventh in females in the United States.²⁰ Constitutive activation of the Wnt/β-catenin signaling pathway is often observed in melanoma.^21,22 TIMP-2 expression modulates human melanoma cell adhesion and motility.²³ However, it is not clear how TIMP-2 directly contributes the activity of Wnt/β-catenin pathway. In this study, we employed stable melanoma cell lines: parental A2058 expressing, A2058 T2-1 over-expressing (T2-1), and A2058 T2R-7 (T2R-7) under-expressing TIMP-2^24,7 to investigate how TIMP-2 regulates Wnt/β-catenin pathway. We examined the signaling pathway of β-catenin at different levels, including total and phosphorylated of β-catenin, ubiquitinated β-catenin, cellular distribution, transcriptional activity, target genes, and cell growth. To mimic the enhanced Wnt/β-catenin micro-environment in melanoma, we treated these cells with lithium. We found that under activated states induced by lithium, TIMP-2 under- expression promoted β-catenin protein levels by activation of Wnt 11 and inhibition of GSK-3β, thus leading to enhance proliferation. On the contrary, TIMP-2 overexpression inhibited Wnt 11 and p-c-Myc levels to decrease tumor cells proliferation. Insights in the molecular mechanisms of TIMP-2 in cancer cells will provide promising opportunities for therapeutic intervention.

Results

TIMP-2 under-expression decreases total β-catenin

We first determined the expression level of β-catenin in the A2058, T2R-7 and T2-1 cell lines at the basal level. The western blot results showed that TIMP-2 over-expression increased the level of β-catenin, compared to that in T2R-7 cells. The densitometry of β-catenin v.s. β-actin was significantly higher in T2-1 cells than that in T2R-7 cells (Fig. 1A). Beta-catenin is associated with the adhension protein α-catenin. We detected the α-catenin protein level in these cells. In contrast, α-catenin was not altered by TIMP-2 level. Interestingly, real-time PCR did not showed any difference of β-catenin at the mRNA level among 3 cell lines (data not shown).

Figure 1. TIMP-2 affects the total, phosphorylated, and ubiquitinated β-catenin in A2058 cells. (A) β-catenin protein expression in human melanoma cell lines: parental A2058 expressing, A2058T2R-7 (T2R-7) under-expressing and A2058T2-1 (T2-1) over-expressing TIMP-2.Densitometry of β-catenin. (B) Immunoblot of phospho-β-catenin in A2058, T2R-7 and T2-1 cells. Densitometry of p-β-catenin (ser 552). (C) Immunoblot of total and phosphor-AKT in A2058, T2R-7 and T2-1 cells.Data are expressed as mean ± SD. #P< 0.05. n= 3 separate experiments. (D) Representative ubiquitination of β-catenin in cells treated with 20μM MG262 for 4 hours. Using anti-β-catenin antibody pulled down, detected with anti-total ubiquitin antibody. β-catenin and actin showed input.

TIMP-2 overexpression elevates the post-translational modification of β-catenin

Beta-catenin activity is regulated by posttranslational modification. The increased level of phosphorylated β-catenin (p-β-catenin) indicates the high activity of β-catenin pathway. We further determined the p-β-catenin expression by Western blot. Interestingly, p-β-catenin (Ser552) increased in T2-1 cells compared with those in the A2058 and T2R-7 cells (Fig. 1B). But there were no significant differences of phosphylated and total AKT (Fig. 1C). Neither did β-catenin upstream regulators GSK-3β, Axin 1 (data not shown) among the 3 A2058 cell lines. To explore the potential mechanism of TIMP-2 increasing β-catenin, we detected ubiquitination of β-catenin. Our data showed that down-expressing TIMP-2 increased the ubiquitination of β-catenin in T2R-7 cells (Fig. 1D). Taken together, our data suggested that TIMP-2 expression elevates β-catenin at the phosphorylation and ubiquitination level in cells at the basal level.

TIMP-2 overexpression affects distribution of β-catenin

We further examined the distribution of β-catenin by immunofluorescence staining. Beta-catenin (red staining) showed the strongest immunofluorescence in circum frontal rings around each cell at sites of cell-cell contact in T2-1 with TIMP-2 over-expression, weaker staining in T2R-7 with TIMP-2 under-expression (Fig. 2A). We can also see small amount of β-catenin existing in nuclear area (green arrow) in T2-1 cells. The immunostaining of β-catenin was consistent with the Western blot result (Fig. 1A).

Figure 2. Distribution of β-catenin and cell growth and proliferation in A2058, T2R-7, and T2-1 cell lines. (A) Distribution of β-catenin by immunostaining. Blue: DAPI; Red: β-catenin. Green arrow shows small amount of β-catenin at nuclear area. (B) Cell growth curve. (C) Cell proliferation measured by the MTT. Data are expressed as mean ± SD. #P< 0.05, n= 3 separate experiments.

TIMP-2 under-expression inhibits cell growth

Cell growth is one of the biological effects regulated by the Wnt/β-catenin pathway. We then determined whether the TIMP-2 expression influences cell growth. As showed here, without any stimulation, TIMP-2 under-expressing T2R-7 cells had significantly lower growth rate compared to the parental A2058 and T2-1 cells (Fig. 2B). Cell proliferation was further measured by MTT assay. The optical density (O.D.) was significant lower in T2R-7 cells (Fig. 2C). We further assessed the transcriptional activity of β-catenin by luciferase reporter assay in the parental A2058, T2-1, and T2R-7 cells. Cells were transfected with a plasmid containing either the wild-type (WT) TCF binding site (pGL3-OT) or a defective TCF binding site (pGL3-OF) upstream of the SV40 promoter and luciferase reporter gene.²⁵ There were no significant signals detected in the cells. Taken together, these data indicate that TIMP-2 under-expression is able to inhibit proliferation in the basal level without any stimulation to activate the β-catenin activity.

E-cadherin expression is enhanced by TIMP-2

Beta-catenin is also one of the adhesion proteins including E-cadherin, α-catenin and P120 catenin, etc, which play an essential role in the tumorigenesis of melanoma. We also assessed the expression level of cadherin in cells with different levels of TIMP-2. Cadherin including epithelial (E), neural (N), placental (P)-cadherin is a family cell surface single-pass transmembrane protein that controls critical events in cell-cell adhesion, recognition and tissue development. E-cadherin plays a role of tumor suppressor gene. N-cadherin is expressed in highly invasive tumor cell lacking E-cadherin. It is known that there is a shift of E-cadherin to N-cadherin in the melanoma cells. We found that the expression level of E-cadherin was significantly increased in the T2-1 over expressing TIMP-2. Meanwhile, another adhesion protein, P120 catenin was increased in T2-1 cells. There is no significant difference in the expression of α-catenin (Fig. 1A) and N-cadherin (Fig. 3A). We then further examined the distribution of E-cadherin by immunofluorescence staining. E-cadherin showed the strongest immunofluorescence in the membrane of T2-1 cells, weaker staining in A2058 and T2R-7 cells (Fig. 3B), which was consistent with the immunoblot result of the E-cadherin.

Figure 3. TIMP-2 and E-cadherin. (A) E-cadherin and N-cadherin protein expression in human melanoma cell lines. Densitometry of E-cadherin v.s. β-actin was measured. Data are expressed as mean ± SD. #P< 0.05. n= 3 separate experiments. (B) Distribution of E-cadherin by immunostaining. Blue: DAPI; Red: E-cadherin.

TIMP-2 under expression enhances β-catenin protein induced by lithium

Constitutive activation of the Wnt/β-catenin signaling pathway is often observed in melanoma,²¹ which may cause immune suppression and resistance in human melanoma cells.²⁶ β-catenin levels control tumor differentiation in Braf-activated Pten-deficient melanoma mouse model.²⁷ Lithium is capable of increasing β-catenin levels possibly through the inhibition GSK-3β, therefore mimicking the effects of Wnt signaling activation. Lithium treatment stabilizes cytosolic β-catenin, which translocates to the nucleus, binds to T-cell factor/L-lymphoid enhancer factor (TCF/LEF), and stimulates target genes such as c-Myc.^28-30 To mimic the activation of Wnt signaling pathway in melanoma, we treated cells with 20 mM lithium chloride for 30 h. We then detected protein levels of β-catenin and found β-catenin was significantly increased in T2R-7 with under-expressed TIMP-2. C-Myc is downstream of β-catenin pathway. We also found that p-c-Myc was increased in T2R-7 compared to A2058 and T2-1 cells (Fig. 4A). However, we did not find difference of c-Myc or clyclin-D1 among 3 cell lines.

Figure 4. TIMP-2 inhibits β-catenin induced by lithium chloride. (A) Immunoblot of β-catenin and p-c-Myc (Ser62) in A2058 (A), A2058T2R-7 (R), and A2058T2-1 (T) cell lines treated with 20 mM lithium chloride (LiCl) 30h. Densitometry of β-catenin and p-c-Myc in cells treated with LiCl. (B) Western blot analysis of β-catenin levels in cytosolic (Cy) and nuclear (Nu) extracts isolated from cells. SP1 serves as a nuclear protein loading control. Densitometry of nuclear β-catenin. Data are reported as mean ± SD of 3 independent experiments. (C) Distribution of β-catenin by immunostaining in A, R and T cells treated with 20mM LiCl for 30h. Blue: DAPI; Red: β-catenin. (D) TIMP-2 inhibited Wnt/T cell factor (TCF) responsive reporter in A, R, T cell lines treated with 20mMLiCl 30h.Luciferase activity was normalized to the internal control, set A2058 as 1. (E) Cell Proliferation measured by the MTT.

TIMP-2 decreases translocation of β-catenin from cytoplasm to nuclear induced by lithium

To confirm the distribution of β-catenin, we performed fractionation analysis. Our data showed that nuclear β-catenin protein was significantly lower in T2-1 cells than that in the parental A2058 and T2R-7 cells. The nuclear β-catenin level in A2058 showed a decreasing trend compared to T2R-7 (Fig. 4B). We further examined the distribution of β-catenin by immunofluorescence staining. Our data shows that the nuclear fluorescence of β-catenin (red staining) in T2R-7 was enhanced compared to those in T2-1 with TIMP-2 overexpression (Fig. 4C). These data indicates that TIMP-2 inhibited lithium induced translocation of β-catenin from cytoplasm to nucleus in cells with activated β-catenin signaling.

Transcriptional activity of β-catenin induced by lithium is inhibited by TIMP-2

Once in the nucleus, β-catenin binds to the transcriptional factors of TCF family to form specific DNA binding complexes. We then sought to ascertain whether TIMP-2 affected the transcriptional activity of β-catenin when cells were stimulated by lithium. Our luciferase reporter data showed that the transcriptional activity of β-catenin was significantly (8 folds) increased in T2R-7 compared to the T2-1 cells. In contract, in the parental A2058, T2-1, and T2R-7 cells without lithium treatment, there were no significant signals detected in the cells (Fig. 4D). These data clearly showed that TIMP-2 overexpression significantly decreased lithium induced transcriptional activity of β-catenin.

TIMP-2 inhibits cell proliferation induced by lithium

To explore the role of TIMP2 on the biological effects of Wnt/β-catenin pathway, we then examined cell proliferation by MTT assay in A2058, T2R-7 and T2-1 cells with lithium treatment. Our data show that absorbance was highest in the TIMP-2 under-expressed T2R-7 cells, whereas, in TIMP-2 overexpressing cells it was lowest (Fig. 4E). These data indicate that TIMP-2 over-expression inhibited lithium-induced proliferation.

TIMP-2 inhibits protein and mRNA levels of Wnt 11 induced by lithium

Lithium has been shown to be a direct, reversible inhibitor of GSK-3.³¹ GSK-3 phosphorylates the amino-terminal region of β-catenin, targeting it for ubiquitination and degradation by proteasome.³² We detected total and phosphorylated protein level of GSK-3β through Western blot. There were no significant differences of P-GSK-3β (Y216) and total GSK-3β among A2058, T2R-7 and T2-1 cells. P-GSK-3β (Y216) correlates with an increase of GSK-3β kinase activity.³³ In contrast, p-GSK-3β (S9) correlates with the inhibition of its kinase activity.³⁴ We found that the ratio of p-GSK-3β (S9) to β-catenin was significantly lower in T2R-7 with under-expressing TIMP-2 (Fig. 5A). Axins inhibit β-catenin pathway and are regulated by GSK-3 β.³⁵ We found that the protein levels of Axin 1 and Axin 2 increased in T2-1 compared to T2R-7 cells (Fig.  5A). The data indicate TIMP-2 under-expressing enhanced lithium-associated accumulation of β-catenin through GSK-3β.

Figure 5. TIMP-2 inhibits protein and mRNA level of Wnt11 induced by LiCl. (A) Immunoblot of phospho-GSK3β, Axins in the cells. Densitometry of p-GSK3β (Ser9), Axin1 and Axin2. (B) Immunoblot of Wnt 11 in the LiCl treated A2058 (A), A2058 T2R-7 (R), and A2058 T2-1 (T) cell lines. Densitometry of Wnt 11 in cells treated with LiCl. (C) TIMP-2 inhibits mRNA levels of Wnt 11 and Wnt 6 in A, R, T treated with LiCl. Data are expressed as mean ± SD. n= 3 separate experiments. # compared to A2058, P< 0.05; # compared to T2R-7, P<0.05. (D) The working model of TIMP-2 regulation of β-catenin activity in melanoma cells with or without lithium stimulation.

Wnt 11 is a direct target of the canonical β-catenin pathway.³⁶ Wnt11 transcripts were robustly induced upon treatment with lithium chloride in mouse myoblast C2C12 cells.³⁶ We measured mRNA levels of Wnts. The data show lithium increases mRNA levels of Wnt 11 in A2058 and T2R-7 cells. There was no significant change of Wnt11 mRNA levels in T2-1 with TIMP-2 overexpression cells after lithium treatment. The response of mRNA of Wnt 6 was similar to that of Wnt 11 (Fig. 5B). Furthermore, our protein gel blots show that TIMP-2 inhibits the increase of Wnt 11 protein level in T2-1 cells treated with lithium (Fig. 5C).

Discussion

In the current study, we have used stable melanoma cell lines, parental A2058, T2R-7 under-expressing and T2-1 over-expressing TIMP-2, to determine the TIMP-2 regulation of β-catenin activity. Without stimulation, TIMP-2 under-expression led to a decrease of β-catenin at the basal level in melanoma cells. With lithium treatment, which mimics the enhanced Wnt/β-catenin microenvironment in melanoma, TIMP-2 over-expression inhibits the increased protein expression and translocation of β-catenin, thus inhibiting the transcriptional activity of β-catenin and its consequences on increasing target genes and cell proliferation. Our data indicate that TIMP-2 plays an essential role in the homeostasis of β-catenin, thus inhibiting cancer cell proliferation when cells have an active β-catenin state (Fig. 5D).

TIMP-2 affects extracellular matrix accumulation, which plays a crucial role in the regulation of cell proliferation through various pathways. Overexpression of TIMP-2 increases the proliferation of human choriocarcinoma cells through the mitogen-activated protein kinase (MAPK) signaling pathway.³⁷ TIMP-2 reduced tumor growth in diffuse-type gastric cancer cell lines through the transforming growth factor β (TGF-β) pathway.³⁸ Our data suggest that TIMP-2 over-expression is able to enhance proliferation in human melanoma A2058 cells at the basal level when β-catenin activity is relative low. It is consistent with previous studies which show that TIMP-2 overexpression protects melanoma cells from apoptosis.^1,6,7 TIMP-2 was shown to stimulate proliferation in human cells, including osteosarcoma cells, fibroblasts, and A549 lung adenocarcinoma cells.³⁹ We would like to put emphasis on the stage of cells with high β-catenin activity. Our data clearly demonstrate that TIMP-2 under-expression was not able to suppress the activation of β-catenin signaling, whereas over-expressing TIMP-2 inhibited β-catenin activity.

Our results of cadherin are consistent with recent studies which show that TIMP-2 over expression increased E-cadherin in human lung carcinoma A549 cells.^13,40 The C-terminus of E-cadherin binds directly to β-catenin to stabilize cell-cell adhesion. The distribution of β-catenin and E-cadherin are in cytoplasm and membrane of the cells. Meanwhile, p120 catenin, another adhesion junction protein, was increased under TIMP-2 overexpression. The regulation of assembly and disassembly of the cadherin-catenin complex underlie the dynamics of the adhesive interactions between cells during tissue development and cancer metastasis.^41,42 Cancer metastasis may include morphological changes characterized by epithelial-to-mesenchymal transition with down-regulation of E-cadherin and induction of N-cadherin, release of β-catenin from junctional complexes.⁴³ The loss of E-cadherin may result in a reduction in N-cadherin-mediated adhesiveness and facilitate the migration of tumor cells across the junctions. Rearrangements and defects of E-cadherin adhesions were found in many carcinomas.^40,44 Furthermore, E-cadherin-positive clones displayed a decrease of invasive abilities in bronchial tumor cells.⁴⁵ Silencing of N-cadherin can inhibit cell proliferation in human melanoma cell.⁴⁶

In the current study, we showed that TIMP-2 overexpression inhibited protein level of β-catenin and its translocation from cytoplasm to nuclear, and cell proliferation induced by lithium. Depends on the sites of phosphorylation, β-catenin can be activated or degraded. Phosphorylation mediated by AKT at a c-terminal serine residue (S552) can promote β-catenin transcriptional activity.⁴⁷ Beta-catenin phosphorylated by GSK3 at T41, S37 and S33, and recognized by the E3-ligase component, β-TrCP, for ultimate ubiquitination and destruction by proteasome.^48-50 Without lithium, TIMP2 increased p-β-catenin at S552, but had no effects on S33, S37 and other sites. Meanwhile, TIMP-2 didn't affect the protein amount of total and phosphorylated AKT. Fraction assay showed no nuclear β-catenin changed in the cell lines. We speculate that TIMP-2 may regulate cell proliferation through other pathways, in addition to the β-catenin signaling. For future study, we will illustrate the mechanism through investigation of other pathways, such as MAPK, TGF-β, etc.

TIMP-2 has been widely used as a therapeutic transgene in various animal models. TIMP-2 can significantly prolong survival time and decrease tumor size in ovarian cancer model.⁵¹ Several other animal studies also showed that TIMP-2 can suppress tumor growth, migration and metastasis in vitro²³ and in various cancer models, such as, prostate cancer,⁵² pancreatic cancer,⁵³ breast adenocarcinoma,⁵⁴ Colon Carcinoma,⁵⁵ melanoma,⁵⁶ etc. The roles of TIMP-2 on MMP-2, MMP-9 and E-cadherin may explain its inhibition tumor migration and metastasis. Our current findings illustrate the mechanism of TIMP-2 reduction of tumor growth in melanoma cells.

Altogether, we demonstrated that TIMP-2 inhibits the Wnt/β-catenin pathway, thus inhibiting proliferation of melanoma cells with high activity of-β-catenin. Insights in the molecular mechanisms of TIMP-2 in cancer will provide promising opportunities for therapeutic intervention.

Materials and Methods

Cell culture and treatment

Human melanoma A2058, T2-1, and T2R-7 cells were maintained in DMEM supplemented with 10% FBS and penicillin-streptomycin as previously described.²⁴ Equal number of cells was plated in 6-well plates. After growing for 24 h, cells were harvested for assays. For lithium treatment, cells were treated with 20 mM lithium chloride (Fisher Scientific, Pittsburgh, PA) for 30 h. For growth curve, cells were plated in triplicate in 6-well plates at 5 × 10⁴ cells/well. The cells were trypsinized and counted at various time points after plating by trypan blue staining using a hemocytometer. For MTT assays, cells were plated in triplicate in 96-well plates at 1 × 10⁴ cells/well with 200 μl medium. At various time points, each well was incubated with 50 μl MTT solution (1 mg/ml in PBS) at 37°C for 4 h. To resolve formazan crystals, 200 μl of DMSO was added to each well. The absorbance was measured at 570 nm on a microtiter plate ELISA reader while DMSO was used as blank.

Cell fractionation

Cells were planted in 6 cm dishes for 24h, then washed with HBSS 3 times and lysed. The cytoplasmic and nuclear proteins were extracted using NE-PER nuclear and cytoplasmic extraction reagents (Thermo Scientific, Rockford, IL, USA).

Immunoblotting

Cells were rinsed twice in ice-cold HBSS, lysed in protein loading buffer (50 mMTris, pH 6.8, 100 mM dithiothreitol, 2% SDS, 0.1% bromophenol blue, 10% glycerol), and sonicated. Equal amount of proteins or equal volumes of total cultured cell lysates were separated by SDS-polyacrylamide gel electrophoresis, transferred to nitrocellulose, and immunoblotted with primary antibodies: anti-β-catenin, GSK-3β, E-cadherin (BD Transduction, San Jose, CA, USA), p-β-catenin (Ser33/37/Thr41), p-β-catenin(Thr41/Ser45), p-β-catenin (Ser552), TIMP-2, Axin1 (Cell Signaling Technology, Billerica, MA, USA), SP-1, P120 catenin, Cyclin D1 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), α-catenin, N-cadherin (Life Technologies, Grand Island, NY), Wnt11, p-c-Myc (Abcam, Cambridge, MA) or β-actin (Sigma-Aldrich) antibodies and visualized by ECL (Thermo Scientific, Rockford, IL, USA) as previously described.^57,58

Immunoprecipitation

Cells were rinsed twice in ice-cold HBSS then lysed in cold immunoprecipitation buffer (1% Triton X-100, 150 mM NaCl, 10 mM Tris·HCl, pH 7.4, 1 mM EDTA, 1 mM EGTA, pH 8.0, 0.2 mM sodium orthovanadate) containing protease inhibitor cocktail (Boehringer Mannheim, Roche, Indianapolis, IN, USA). Samples were pre-cleared with protein A-agarose (Invitrogen). Pre-cleared lysates were then incubated with 2 μg of anti- β-catenin antibody overnight at 4°C. A 50% slurry of Protein A-agarose was added to the lysate and incubated for 1h with agitation at 4°C and then washed with cold immunoprecipitation buffer. The pellet was resuspended in protein loading buffer and centrifuged, boiled for 5 min, separated by SDS-PAGE, and transferred onto a nitrocellulose membrane. Membrane blots were probed with anti-ubiquitin antibody (UG9510, Enzo Life).

Immunofluorescence

Cells were grown in the Lab-Tek chambered coverglass system (Thermo Scientific), with/without 20 mM lithium chloride treatment 30 h. Cells were fixed with 4% paraformaldehyde and permeabilized with 0.2% Tritonx-100, incubated with anti-E-cadherin, anti-β-catenin antibody at 4°C overnight after blocking with 5% bovine serum albumin. Samples were then incubated with goat anti-mouse AlexaFluor 594 and DAPI (Life technologies) for 1 hour at room temperature. Slides were mounted with SlowFade (Life technologies), followed by a coverslip, and the edges were sealed to prevent drying. Specimens were examined with a Zeiss laser scanning microscope 710 (Carl Zeiss Oberkochen, Germany).

Quantitative real-time PCR analysis

Total RNA was extracted from epithelial cell monolayers using TRIzol reagent (Life technologies). The RNA integrity was verified by electrophoresis gel. RNA reverse transcription was done using the iScript cDNA synthesis kit (Bio-Rad, Hercules, CA, USA) according to the manufacturer's directions. The RT cDNA reaction products were subjected to quantitative real-time PCR using the CFX96 Real-Time system C1000 Touch thermal cycler (Bio-Rad) and iTaq Universal SYBR green supermix(Bio-Rad) according to the manufacturer's directions with primers (Table 1). All real-time PCR reactions were performed in triplicate. Percent expression was calculated as the ratio of the normalized value of each sample to that of the corresponding control cells.

Table 1. Real-Time PCR Primers.

Gene name	Primer
Wnt 6-F	AACAGGACATTCGGGAGACG
Wnt 6-R	GCAGCTCGCCCATAGAACA
Wnt 11-F	GGCTTGTGCTTTGCCTTCA
Wnt 11-R	TTTGATGTCCTGCCCTCCTT
Actin-F	AGAGCAAGAGAGGCATCCTC
Actin-R	CTCAAACATGATCTGGGTCA

Luciferase reporter assay

TOPFlash luciferase reporter (Millipore, Billerica, MA) containing the TCF/LEF consensus sequence was used to assay β-catenin dependent promoter activity, with control FOPFlash reporter (Millipore) containing mutated TCF binding sites. Cells were transfected with TOPFlash and FOPFlash using Lipofectamine (Life technologies) in 24-well plates in triplicates. To normalize transfection efficiency, cells were co-transfected with internal control reporter Renilla luciferase plasmid (pRL-TK; Promega, Madison, WI, USA). 24h after transfection, cells were incubated with 20 mM lithium chloride 30 h, luciferase activity was determined using the dual luciferase reporter assay system (Promega) with a Sirius single tube luminometer (Berthold, Pforzheim, Germany). The luciferase activity was normalized with Renilla luciferase activity, and the activity was expressed as relative units as previously described.⁵⁷

Statistical analysis

Data were expressed as mean ± SD. Differences between 2 groups were analyzed by Student's t-test. More than 2 groups were analyzed by analysis of variance (ANOVA) followed by the Tukey-Kramer test, P values less than or equal to 0.05 were considered significant.

Disclosure of Potential Conflicts of Interest

The authors declare no conflict of interest.

Author Contributions

Conceived and designed the experiments: SW. Performed the experiments: YX, SW. Analyzed the data: SW. Wrote the paper: YX, SW. All authors read and approved the final manuscript.

Acknowledgments

We want to thank Dr. Jun Sun for providing her resources, cell lines, and extremely insightful advice on the manuscript. We want to thank Lester Liu for his excellent technical support and Dapeng Jin for critical editing of this manuscript.