Epigenetic regulation of the Wnt signaling inhibitor DACT2 in human hepatocellular carcinoma

Abstract

DACT2 (Dapper, Dishevelled-associated antagonist of β-catenin homolog 2) is a member of the DACT family involved in the regulation of embryonic development. Human DACT2 is localized on 6q27, a region of frequent loss of heterozygosity in human cancers. However, the regulation of DACT2 expression and function in hepatocellular carcinoma (HCC) remains unclear. In this study, genetic and epigenetic changes of DACT2 were analyzed in HCC cell lines and primary cancer. We found no single-nucleotide polymorphism (SNP) associated with HCC. Promoter region methylation was correlated with loss or reduction of DACT2 expression, and restoration of DACT2 expression was induced by 5-aza-2’-deoxycytidine (5-AZA) in HCC cell lines. Promoter region methylation was found in 54.84% of primary HCC. Reduction of DACT2 expression was associated with promoter hypermethylation, and expression of DACT2 was inversely related to β-catenin expression in primary HCC. DACT2 suppressed cell proliferation, induced G₂-M arrest in cell lines and inhibited tumor growth in xenograft nude mice. The transcriptional activity of TCF-4 and the expression of Wnt signaling downstream genes were suppressed by DACT2 re-expression and reactivated by depletion of DACT2. In conclusion, DACT2 is frequently methylated in HCC and its expression is regulated by promoter hypermethylation. DACT2 suppresses HCC by inhibiting Wnt signaling in human HCC.

Keywords: :

• DACT2
• hepatocellular carcinoma
• epigenetics
• DNA methylation
• Wnt signaling

Introduction

In men, hepatocellular carcinoma (HCC) is the fifth most frequently diagnosed cancer worldwide and the second cause of global cancer death. In women, it is the seventh most commonly diagnosed cancer and the sixth leading cause of cancer death.¹ Carcinogenesis is a multi-step process with cumulative genetic and epigenetic changes involving oncogenes and tumor suppressor genes.^2-6 An increasing number of publications are focusing on identifying novel genes regulated by DNA methylation, histone modification and miRNAs.^7-12 These studies are mainly focused on elucidating the mechanisms of inactivation of tumor suppressors in different signaling pathways in order to find new therapeutic strategies in human cancer, including HCC. Wnt signaling plays an important role in the physiology, development, cell differentiation, proliferation and growth.¹³ Abnormal activation of Wnt signaling is a major driving force in HCC.^14-16

Dapper, a Dishevelled-associated antagonist of β-catenin (DACT), was isolated by a screen for proteins interacting with Dishevelled, a key factor in the Wnt signaling. Dapper and Dishevelled were co-localized intracellularly and formed a complex with Axin, GSK3 and β-catenin.¹⁷ Human DACT1 and DACT2 were identified by Katoh et al. in 2003.¹⁸ Human and murine DACT3 were both identified by Fisher et al.¹⁹ DACT1 has been reported frequently to be methylated in HCC, and DACT3 has been found to be regulated by histone modifications in colorectal cancer.^12,20 Human DACT2 is localized in chromosome 6q27, a region of frequent loss of heterozygosity in human cancers.^18,21-27 However, the regulation of DACT2 expression and its function in human HCC remains unknown.

In this study, we first analyzed genetic and epigenetic changes of DACT2, and then studied its expression and function in hepatic carcinogenesis in vitro and in vivo.

Results

No single-nucleotide polymorphism (SNP) in the DACT2 gene is associated with HCC

The sequencing of the full length cDNA and genomic DNA of DACT2 in seven hepatic cancer cell lines and one immortalized hepatocyte cell line (LO2) revealed five SNP in exon 4, an important functional region also known as PDZ (post synaptic density-95/discs large/zonula occludens-1) binding domain.²⁸ Although no new mutations were discovered, four of the above SNPs were found both in patients with HCCs and in healthy controls. The respective locations and frequencies of these SNPs in both patients with HCCs and in healthy controls are as follows: 26.25% vs. 23.10% for A/G (rs6925614), 2.50% vs. 1.28% for T/C (rs79931308), 15.00% vs. 15.38% for A/C (rs10945501) and 1.25% vs. 1.28% for G/T (rs73789362). No significant differences were found in SNPs between HCCs patients and healthy individuals (p > 0.05).

DACT2 is silenced by promoter hypermethylation in HCC cell lines

DACT2 was silenced in the HepG2 cell line and reduced in cell lines SNU182, BEL7402, SMMC7721 and SNU449. DACT2 was normally expressed in PLC/PRF/5, 97H and in the immortalized cell line (LO2) (Fig. 1A). To investigate if silencing of DACT2 is associated with promoter region hypermethylation, we first analyzed the CpG island of DACT2 DNA sequence using a CpG Island search program (http://cpgislands.usc.edu). One CpG island was found in the promoter region (Fig. 1B). Then DACT2 promoter region methylation was analyzed by MSP and bisulfite sequencing (BSSQ). Complete methylation was found in the HepG2 cells, and partial methylation was observed in the SNU182, BEL7402, SMMC7721 and SNU449 cell lines. No methylation was detected in LO2, PLC/PRF/5 and 97H cell lines (Fig. 1C). The methylation density within DACT2 promoter region was characterized and validated by BSSQ (Fig. 1D). Bisulfite sequencing of 10 individual clones of PCR products from HepG2 revealed dense methylation of CpGs within the promoter region. The mixed methylation pattern of CpGs observed with BSSQ in the SNU182 cell line may represent both methylated and unmethylated alleles or both methylated and unmethylated clonal subpopulations within cultured cells. No methylation was found by BSSQ in PLC/PRF/5 and LO2. These results indicate that our MSP assays results accurately represent DACT2 promoter region methylation status in these cell lines.

Figure 1.DACT2 is silenced by promoter region hypermethylation in HCC cell lines.(A) Expression of DACT2 was analyzed by semiquantitative RT-PCR in HCC cell lines and one immortalized hepatocyte cell line (LO2). (-) 5-AZA untreated; (+) 5-AZA treated; GAPDH was used as an internal control for RT-PCR. (B) Distribution of CpG sites in the promoter region of DACT2 and the location of the MSP primers as well as BSSQ region are shown. (C) DACT2 methylation was determined by MSP in HCC cell lines and LO2. IVD, in vitro methylated DNA (methylation control); NL, normal blood lymphocyte DNA (unmethylated control); M, methylated band; U, unmethylated band. (D) Promoter region methylation status of DACT2 was analyzed by BSSQ in three HCC cell lines. Open circles denote unmethylated CpG site and filled circles represented methylated CpG site. The region amplified by MSP is indicated by arrows. TSS, transcription start site.

Concomitant loss of DACT2 expression together with promoter region complete methylation was found in HepG2 cells. Normal expression without concomitant methylation was observed in LO2, PLC/PRF/5 and 97H cells. Partial methylation and reduced expression were detected in SNU182, BEL7402, SMMC7721 and SNU449 cell lines. These results indicate that promoter region methylation is correlated with DACT2 silencing. DACT2 expression was restored after 5-AZA treatment in HepG2 cells, and increased expression was observed in the SNU182 and BEL7402 cell lines. All of the above results demonstrated that DACT2 expression was regulated by promoter region hypermethylation.

DACT2 is frequently methylated in primary HCCs

DACT2 promoter region hypermethylation was not limited to cultured HCC cell lines. Frequent methylation was found in primary HCC (Fig. 2A). In 62 HCCs, 34 cases (54.84%) were methylated and 28 cases (45.16%) were unmethylated. No association was found between DACT2 methylation and clinicopathological variables such as age, gender, hepatitis B/C virus infection, cirrhosis, AFP levels, tumor size or tumor stage in HCCs (Table 1).

Figure 2.DACT2 expression is associated with promoter hypermethylation in primary HCCs. (A) Representative MSP results of DACT2 in primary HCCs. (B) DACT2 expression in a representative HCC case. Left image: adjacent non-tumor tissue (100×) and the indicated fields enlarged (400×). Right image: tumor tissue (100×) and the indicated fields enlarged (400×). (C) DACT2 expression scores are shown as box plots, the horizontal lines represent the median score; the bottom and top of the boxes represent the 25th and 75th percentiles, respectively; the vertical bars represent the range of data. Expression of DACT2 was different between non-tumor tissues and tumor tissues in 62 matched primary HCCs. **p < 0.01. (D) The correlation of DACT2 hypermethylation and expression level were analyzed in 62 matched primary HCCs. *p < 0.05. (E) β-catenin expression in the representative methylated case (400×). β-catenin expression is positive mostly in the membranes of hepatic cells (left image) and in the cytoplasm and nucleus in tumor cells (right image).

Table 1. Association of clinical factors with DACT2 methylation and expression in HCC patients

Clinical factors	n	DACT2 Methylation (%)	p (χ^2)	Loss of expression (%)	p (χ^2)
Age (years)			1		0.01
< 50	21	12 (57.14)		17 (80.95)
≥ 50	41	22 (53.65)		18 (43.9)
Gender			1		0.27
Male	54	30 (55.55)		32 (59.26)
Female	8	4 (50.00)		3 (75.00)
Virus infection			0.6		0.27
HBV	49	28 (57.14)		28 (57.14)
HCV	2	1 (50.00)		2 (100.00)
No	11	5 (45.45)		5 (45.45)
AFP (ng/ml)			1		0.04
< 500	44	24 (54.54)		21 (47.73)
≥ 500	18	10 (55.55)		14 (77.78)
Tumor size (cm) ^a			0.79		0.7
< 8.07	36	19 (52.77)		21 (58.33)
≥ 8.07	26	15 (57.69)		14 (53.84)
ALTSG stage			1		1
T1,T2	15	8 (53.33)		8 (53.33)
T3,T4	47	26 (55.32)		27 (57.44)
AJCC stage			0.17		0.78
T1,T2	36	24 (66.67)		24 (66.67)
T3,T4	21	10 (47.62)		11 (52.38)
Differentiation			0.96		0.67
Poor	10	6 (60.00)		6 (60.00)
Poor to moderate	8	4 (50.00)		3 (37.50)
Moderate	36	20 (55.56)		20 (55.56)
Moderate to well	3	2 (66.67)		2 (66.67)
Well	5	2 (40.00)		4 (80.00)
Cirrhosis			0.77		0.57
No	16	8 (50.00)		8 (50.00)
Yes	46	26 (56.52)		27 (58.70)
Capsule invasion ^b		0.315			0.1
No	39	20 (51.28)		25 (64.10)
Yes	20	13 (65.00)		8 (40.00)
Encapsulated ^b			1		0.30
No	12	6 (50.00)		5 (41.67)
Yes	31	16 (51.61)		20 (64.52)
Vascular invasion			0.54		0.53
No	49	28 (57.14)		29 (59.18)
Yes	13	6 (46.15)		6 (46.15)

^a The average of tumor size of all 62 samples is 8.07 cm. ^bInformation on capsule invasion and encapsulation is only available for 59 and 43 samples, respectively.

DACT2 expression is associated with promoter hypermethylation in primary HCCs

DACT2 expression was analyzed by IHC in tumor tissues and adjacent non-tumor tissues. DACT2 was expressed in the cytoplasm of hepatic cells and its expression was reduced significantly in tumor tissues compared with adjacent tissues (p < 0.01) (Fig. 2B and C). Reduced DACT2 expression was related to younger age (p < 0.01) and high serum levels of AFP (p < 0.05) patients (Table 1). These results suggest that DACT2 may be a potential prognosis marker.

The association of DACT2 expression and methylation status was analyzed in 62 HCC patients. Reduced expression of DACT2 was found in 42 cases of tumor tissue, of which 28 cases were methylated (66.67%). Twenty cases were normally expressed, of which 6 cases were methylated (30%). Reduced DACT2 expression was significantly related to promoter methylation (p < 0.05, Figure 2D). This suggests that DACT2 expression may be regulated by promoter region methylation in primary HCC.

To evaluate whether activation of Wnt signaling was related to DACT2 expression, β-catenin expression was analyzed by IHC in DACT2 reduction cases. In cancer tissues, positive cytoplasm and nucleus β-catenin staining was found in 25 cases (59.52%). In adjacent tissues, β-catenin expression was mainly located on the membrane (31/42, 73%) (Fig. 2E). The expression of DACT2 was associated with β-catenin location in HCC (p < 0.05).

HCC cell growth is suppressed by DACT2 re-expression

Frequent loss of DACT2 expression in HCCs suggests that it may be a potential tumor suppressor. To evaluate the effect of DACT2 on cancer cell proliferation, colony formation assay was employed. DACT2 expression resulted in significant suppression of long-term cell growth in colony formation assays, as much as 60% (Fig. 3A). Concomitantly, the inhibitory effect of DACT2 was evaluated by BrdU incorporation assay. HepG₂ growth was inhibited about 30% by DACT2 (Fig. 3B).

Figure 3. The effect of DACT2 on cell proliferation, cell cycle and apoptosis in vitro. (A) The effect of DACT2 on HCC cell proliferation was evaluated by colony formation assay. Left panel: Colony formation results of transfected with pCMV6 (empty vector) or DACT2 expression vector in HepG2 cells. Right panel: Quantitative analysis of colony numbers in different group. **p < 0.01. (B) BrdU incorporation was analyzed in pCMV6 or DACT2 expression vector transfected HepG2 cells. Left panel: cell nuclei were counterstained by BrdU and DAPI. Right panel: quantitative analysis of BrdU positive cells in different group. *p < 0.05. (C) Representative results of cell cycle distribution for pCMV6 (left panel) or DACT2 expression vector (right panel) transfected in HepG2 cells. Cycle distributions were measured by propidium iodide (PI) staining followed by flow cytometry after transfection for 48 h. (D) Quantitative analysis of cell cycle distribution in DACT2 expression and control group. **p < 0.01. (E) No effect was found on apoptosis after expression of DACT2 in HepG2 cell line. HepG2 cells were transfected with pCMV6 (left panel) or DACT2 expression vector (Right panel). 48 h serum starvation was performed 24 h after transfection. Then Annexin V-FITC/PI double staining was performed. Annexin V-positive and PI-negative staining cells are indicated as apoptotic cells. (F) Quantitative analysis of cell apoptosis.

G₂-M arrest is induced by re-expression of DACT2

The effect of DACT2 on the cell cycle was evaluated by flow cytometry. The ratio of G₂ phase was increased after re-expression of DACT2 in HepG2 cells (G₁: 56.60 ± 1.72 vs 51.70 ± 4.27; S: 25.95 ± 1.07 vs 16.64 ± 2.50; G2:17.45 ± 2.81 vs 31.0 ± 1.27) (p < 0.01) (Fig. 3C and D). These results indicate that DACT2 induces G₂-M phase arrest in HCC.

DACT2 has no effects on cell apoptosis

To investigate the effects of DACT2 on apoptosis, an annexin V-FITC assay was employed in HepG2 cells. No difference was found between the DACT2 expression group and control group (11.98% ± 2.19 vs. 13.60% ± 1.67, p > 0.05) (Fig. 3E and F). These data suggest that DACT2 is not involved in apoptosis in HCC cells.

Tumor growth is retarded by DACT2 in vivo

Mice burdened with subcutaneous tumors were shown in Figure 4A. Figure 4B shows that the volume is significantly different in DACT2 expressed and unexpressed HepG2 groups (251.09 ± 174.24 mm³ vs. 533.20 ± 370.86 mm³, p < 0.05). As shown in Figure 4C and D, the tumor weight is reduced in DACT2-expressing nude mice compared with an empty vector group (0.18 ± 0.11 g vs. 0.42 ± 0.25 g; p < 0.05). The results indicate that DACT2 inhibits tumor growth in vivo.

Figure 4. Tumor growth was inhibited by DACT2 in vivo. (A) Nude mice inoculated with HepG2 cells expressing or not expressing DACT2. Left rear flank: empty vector group. Right rear flank: DACT2-expressing group (n = 10/group). (B) Tumors in the upper panel are from the empty vector group and tumors in the lower panel are from the DACT2-expressing group. (C) Histogram represents tumor weights of each mouse in the two groups. (D) Tumor weights are significantly different between DACT2-expressing and empty vector groups. *p < 0.05.

DACT2 is an inhibitor of Wnt signaling

DACT2 was reported to be involved in Wnt signaling during zebra fish and mice development. To explore the effects of DACT2 on Wnt signaling, promoter-luciferase activity assays were employed in this study. The transcriptional activity of TCF-4 was inhibited in both wild type and mutant β-catenin vectors groups by the re-expression of DACT2 (Fig. 5A). Western blotting shows that the expression of TCF-4/β-catenin downstream targets, c-Myc and cyclin D1, was decreased after re-expression of DACT2 in HepG2 cells (Fig. 5B). These results strongly suggest that DACT2 is a Wnt signaling inhibitor in HCC.

Figure 5. DACT2 is a Wnt signaling inhibitor. (A) The transcriptional activity of TCF-4 was inhibited by DACT2. Left panel: DACT2 was co-transfected with wt β-catenin and OT into HepG2 cells. Right panel: DACT2 was co-transfected with mut β-catenin and OT into HepG2 cells. OF acted as a negative control reporter and its transcriptional activity was defined as 1. *p < 0.05. (B) Expression of c-Myc and cyclin D1 was analyzed by western blot in DACT2-expressing or -not expressing HepG2 cells. Actin was used as an internal control. (C) The transcriptional activity of TCF-4 was increased after two individual siRNAs targeting DACT2 vectors were transfected into PLC/PRF/5 cells. *p < 0.05. (D) Knockdown of DACT2 increases the expression of c-Myc and cyclin D1 in PLC/PRF/5 cells. Actin was used as internal reference.

To further analyze the effect of DACT2 on Wnt signaling, DACT2 was knocked down by RNA interference (RNAi) in PLC/PRF/5 cells. As shown in Figure 5C and D, the transcriptional activity of TCF-4 was increased in DACT2-depleted cells, compared with control group, and the expression of Wnt signaling downstream genes was increased when depletion of DACT2 in PLC/PRF/5 cells. These results suggest that DACT2 is a Wnt signaling inhibitor in HCC.

Discussion

Abnormal activation of Wnt signaling is a major driving force in cancer, which can be initiated by genetic or epigenetic changes.^4,29 Frequent methylation of Wnt signaling antagonists suggest an important role for the activation of this pathway during carcinogenesis.^5,30

DACT2 is located on human chromosome 6q27, a region frequently associated with loss of heterozygosity in human cancers.^18,21-27

DACT genes encode a small family of vertebrate intracellular proteins that can regulate intercellular signaling pathways by a conserved leucine zipper motif near the N-terminus and a binding motif for PDZ domain at the C-terminus.^17,18,31DACT1 and DACT2 were discovered by two independent groups who were screening partners of the Dvl scaffold protein, which is central to the developmentally and clinically important Wnt signaling.^31-36 The initial functional analyses of DACT2 were based on its overexpression and morpholino-based knockdown technologies in Xenopus laevis and zebrafish.^37,38

In this study, we describe for the first time that DACT2 is frequently absent or downregulated in HCC cell lines, and is also significantly reduced in primary HCC samples. We first sequenced DACT2 in primary HCC and cell lines and found no mutation related to HCC. This suggests that genetic changes may not play important roles in HCC carcinogenesis. It was previously reported that DACT2 is frequently inactivated by DNA methylation in colorectal cancer cell lines (RKO and HT29).¹² Our previous study in lung cancer also reported that the DACT2 gene was silenced by promoter region methylation.³⁹ To determine whether hypermethylation was responsible for the silencing of DACT2 in HCC, promoter region methylation was analyzed in HCC cell lines and primary cancer. Our data suggest that DACT2 is regulated by promoter region hypermethylation in HCC. Further analysis indicates that reduction of DACT2 expression was related to younger age and high level of AFP in serum. These results suggest that methylation of DACT2 may serve as a diagnostic tool and prognostic marker in HCC.

It has been reported that DACT2 binds to the TGF-β receptors ALK5 and ALK4, accelerating lysosomal degradation of these receptors in zebrafish.³⁸ Although recent studies showed that murine DACT proteins formed a weak complex with Alk5 in HEK293T cells, no such complex can be detected between Alk4 or Alk5 and DACT2 protein.³⁶ DACT2 strongly co-immunoprecipitated with β-catenin or δ-catenin in HEK293T cells, as well as formed even stronger complexes with CDK1δ/ε, Dvl or Vangl family members.³⁶ This report suggests that DACT2 is a Wnt signaling inhibitor. Our study found that HCC proliferation was suppressed by DACT2 both in vitro and in vivo. In addition, Wnt signaling activity was inhibited by DACT2 in HCC cells. These results suggest that DACT2 inhibits human HCC by inactivating Wnt signaling.

In conclusion, DACT2 is silenced by promoter region hypermethylation in human hepatic cancer and its methylation may serve as a detection marker for HCC. DACT2 suppressed cell proliferation, induced G2-M arrest in cell cycle and inhibited tumor growth in nude mice. DACT2 also inhibited TCF-4 transcriptional activity and its downstream targets (c-Myc and cyclin D1) in HCC. Based on these findings and its important human chromosomal localization, DACT2 may be a tumor suppressor in HCC.

Materials and Methods

Ethics statement

For the use of clinical materials for research purposes, prior patients’ consent and approval were obtained from Institutional Review Board of the Chinese PLA General Hospital.

Cell lines and tissue samples

Eight HCC cell lines (LO2, SNU182, BEL7402, HepG2, PLC/PRF/5, SMMC7721, SNU449 and 97H) were used in this study. Cell lines were cultured at 37°C in an atmosphere containing 5% CO₂ in 90% RPMI 1640 (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum. Cells were passaged 1:3 once 80% confluence (approximately 1 × 10⁶ cells) was reached on a 75 cm² culture flask (NEST Biotechnology).

Sixty-two paired tumor tissue samples of primary HCC and their adjacent non-tumor tissues were obtained from surgical resected HCC patients at the Chinese PLA General Hospital, Beijing. For this study, all of the tumor and non-tumor tissue were re-examined and confirmed by pathologists.

RNA isolation and semi-quantitative RT-PCR

RNA Isolation and semi-quantitative RT-PCR. Total RNA was isolated using the RNeasy mini kit (Qiagen). First strand cDNA (cDNA) was synthesized using the Superscript II-reverse transcriptase kit (Invitrogen). Expression of DACT2 mRNA (mRNA) was determined by RT-PCR using the LightCycler system (Roche Diagnostics,). PCR amplification of DACT2 was performed using primers: 5′-GGCTGAGACAACAGGACATCG-3′ (forward) and 5′-GACCGTCGCTCATCTCGTAAAA-3′ (reverse). The primer set for DACT2 was designed to span intronic sequences between exons in order to control genomic DNA contamination. A total of 35 cycles of amplification were performed for each of the RT-PCR experiments. As an internal control, GAPDH was amplified with 25 cycles to ensure cDNA quality and quantity for each RT-PCR. Amplified products were analyzed on 1.5% agarose gel.

Methylation-specific PCR (MSP) and bisulfite sequencing

Genomic DNA from cell lines and tissue samples was prepared using the proteinase-K method. After chloroform/phenol extraction, DNA was precipitated in ethanol and dissolved in low TE buffer and stored at -20°C. Genomic DNA from HCC tissues and cell lines was bisulfite modified as previously described.⁴⁰ MSP primers were designed according to genomic sequences flanking the presumed transcription start site (TSS). Primer sequences were oligo-synthesized (Invitrogen) to allow MSP to detect bisulfite-induced changes affecting unmethylated (U) and methylated (M) alleles. MSP of DACT2 was performed using primers: 5′-GATTTTAGTTTATTTTGGCGATTTGC-3′ (M-forward); 5′-CACATCTCCCGAACAAAATCCCG-3′ (M-reverse); 5′-TAGATTTTAGTTTATTTTGGTGATTTGT-3′ (U-forward) and 5′- TCCACATCTCCCAAACAAAATCCCA -3′ (U-reverse). Each MSP reaction incorporated approximately 100 ng of bisulfite-treated DNA, 25 pmoles of each primer, 100 pmoles dNTPs, 2.5 μl 10 × PCR buffer, and 1 unit of Taq Polymerase (Invitrogen) in a final reaction volume of 25 μl. Cycle conditions were: 95°C × 5 min, 1 cycle; 35 cycles × (95°C × 30 sec, 60°C × 30 sec, 72°C × 30 sec); 72°C × 5 min, 1 cycle. Each PCR assay included a methylated control (in vitro methylated DNA, IVD), an unmethylated control (normal blood lymphocyte DNA, NL) and a negative control (water). MSP products were analyzed using 2% agarose gel electrophoresis.

Bisulfite-treated DNA was subjected to PCR using primers flanking the targeted MSP regions above. Sequencing primers were as follows: 5′-TGGTTATAGATTTTAGTTTATTTTGG-3′ (forward) and antisense 5′-CAACCCCTACAACTCCTACAAC-3′ (reverse). PCR cycle conditions were as follows: 95°C × 5 min, 1 cycle; 35 cycles × (95°C × 30sec, 58°C × 30sec, 72°C × 40 sec); 72°C × 5 min, 1 cycle. PCR products were gel purified and cloned into pCR2.1 vector according to the manufacturer’s instructions (Invitrogen). Sequencing was performed as previously reported.^{5, 41}

5-aza-2’-deoxycytidine (5-AZA) treatment

Cell lines (HepG2, SNU182, BEL7402, PLC/PRF/5 and LO2) were split to a low density (30% confluence) 12 h before treatment. Cells were treated with 5-AZA (Sigma) at a concentration of 2 μM. Growth medium, conditioned with 5-AZA at 2 μM, was exchanged every 24 h for 96 h. At the end of the treatment, RNA was extracted from the cells as described above.

Immunohistochemistry staining

Rabbit anti-DACT2 antibody (Prosci) and mouse anti-β-catenin antibody (ZSGB Biotech.) were employed. Immunohistochemistry (IHC) was performed on 4μm thick serial sections derived from formaldehyde fixed paraffin wax embedded tumor tissue blocks. After de-paraffinization and rehydration, endogenous peroxidase activity was blocked for 30 min in methanol containing 0.3% hydrogen peroxide. The slides were then incubated with anti-DACT2 antibodies (1:1600 dilution) or with mouse anti-β-catenin antibody (1:400) overnight at 4°C in a humidified chamber. IHC was performed on the tissue sections according to the Polink-2 plus^® Polymer HRP Detection System (ZSGB Biotech). Hematoxylin was used for counterstaining. The expression of DACT2 and β-catenin was evaluated according to previous report.⁴²

Construction of DACT2 expression vector and isolation of DACT2 expressed cells

The expression construct for DACT2 was generated by cloning a PCR-amplified full-length human DACT2 cDNA fragment (GenBank accession number NM_214462) into pCMV6-AC-GFP or pCMV6-Entry-myc (Origene Technologies, Inc.). The vector was verified by DNA sequencing and western Blotting. Transient transfection was performed using FuGENE 6 (Roche Applied Science) according to the manufacturer’s instructions. GFP positive cells were isolated by flow cytometry (Becton Dickinson) 48 h after transfection.

Western blotting

Cell lysates were collected and western blotting was performed. The antibodies used included rabbit anti-DACT2 antibodies (Prosci) and monoclonal antibodies against cyclin D1 (Santa Cruz Biotechnology) as well as c-Myc and actin (Bioworld Technology)

Colony formation assay

HepG2 cells were transfected with either an empty or a DACT2 expression vector using FuGENE HD (Roche Applied Science) according to manufacturer’s instructions. 60% GFP positive cells were isolated by flow cytometry 48 h after transfection. Cells were collected and reseeded in triplicate at 1,500 cells per well in 6-well plates. Growth medium, conditioned with G418 (Invitrogen) at 500 μg/mL, was exchanged every 24 h. After 10–14 d, cells were fixed with 75% ethanol for 30 min, stained with 0.2% crystal violet for visualization, and counted.

BrdU incorporation analysis

HepG2 cells, transfected with either an empty or a DACT2 expression vector, were seeded in 24-well plates. Ten milimolar BrdU was added in growth medium and incubated at 37°C for 2 h. Cells were stained with mouse anti-BrdU antibody and DAPI for double labeling. The BrdU (red) /DAPI (blue) positively staining cells are indicated as positive BrdU incorporation.

Cell cycle analysis

DACT or empty vector was transfected into HepG2 cells using FuGENE. Forty-eight hours after transfection, cells were harvested, washed with phosphate-buffered saline (PBS) and fixed with ice-cold 70% ethanol at -20°C overnight. Samples were then washed with PBS and stained with propidium iodide (Sigma) containing RNase A (Sigma) for 30 min at 37°C. Cell cycle distribution in different phases was determined using flow cytometry.

In vivo tumor genesis assay

Six-week-old female nu/nu mice were bred under specified pathogen-free conditions. HepG2 cells were transiently transfected with either an empty or a DACT2 expression vector using FuGENE HD and GFP positive cells were isolated by flow cytometry 48 h after transfection. Cells (3 × 10⁶) were diluted in 100 μl PBS and injected subcutaneously into the left or right rear flank of the mice. When tumors reached approximately 1.0 cm in size, the mice were sacrificed and the volume as well as the weight of the tumors were measured. Tumor volume (mm³) was estimated by measuring the longest and shortest diameter of the tumor and calculated with the formula: tumor volume = (length) × (width)²/2. All experimental procedures were approved by the Animal Ethics Committee of the Chinese PLA General Hospital, Beijing.

Dual-luciferase reporter assay

To explore the effect of DACT2 on Wnt signaling, DACT2 was co-transfected with TCF-4 reporter, pTOPFlash (OT), and β–catenin wide type (wt) or β-catenin mutant (mut) vector into HepG2 cells, pFOPFlash (OF) acted as a negative control reporter.⁴³ Cell lysates were collected and luciferase enzymatic luminescent activity was measured 48 h after transfection according to the manufacturer’s instructions (GLOMAX luminometer, Dual Luciferase Reporter Assay system, Promega).

RNA interference

Two selected siRNAs targeting DACT2 and RNAi Negative Control Duplex were used in this study. The sequences are as follows: siRNA-1 (sense sequence: 5′-CCAGCUGUCCUGAGUCUAATT-3′ and antisense sequence: 5′-UUAGACUCAGGACAGCUGGTT-3′); siRNA-2 (sense sequence: 5′-GUCGGUUGAUGAGACUACUTT-3′ and antisense sequence: 5′-AGUAGUCUCAUCAACCGACTT-3′). RNAi Negative Control Duplex (sense sequence: 5′-UUCUCCGAACGUGUCACGUTT-3′; and antisense sequence: 5′-ACGUGACACGUUCGGAGAATT-3). Fifty to sixty percent confluent PLC/PRF/5 cells were transfected with 50 nM of siRNAs using Lipofectamine 2000 (Invitrogen) following the manufacturer’s direction.

Statistical analysis

We evaluated the relationship between methylation status in human hepatocellular carcinomas and clinicopathologic characteristics using the Pearson’s chi-square test or the Fisher’s exact test for independence for dichotomous variables as appropriate. Continuous variables were analyzed with Student’s t-test. Results were judged to be statistically significant at p < 0.05 and all p values were two-sided. All analyses were done using SPSS PASW Statistics 18.0.

Abbreviations:
HCC	=	hepatocellular carcinoma
DACT2	=	Dapper, a Dishevelled-associated antagonist of β-catenin, homolog 2
MSP	=	methylation specific polymerase chain reaction
BSSQ	=	Bisulfite sequencing
GAPDH	=	glyceraldehyde-3-phosphate dehydrogenase
TSS	=	transcription start site
M	=	methylated
U	=	unmethylated
5-AZA	=	5-aza-2’-deoxycytidine
NL	=	normal blood lymphocyte DNA
IVD	=	in vitro methylated DNA
PDZ	=	Post synaptic density-95/Discs large/Zonula occludens-1
PBS	=	phosphate-buffered saline
AJCC	=	American Joint Committee on Cancer
AFP	=	alpha-fetoprotein
ALTSG	=	American Liver Tumor Study Group
BrdU	=	bromodeoxyuridine
TCF	=	T-cell factor
LEF	=	lymphoid enhancer factor
NT	=	non-tumor tissue
T	=	tumor tissue
SNP	=	single-nucleotide polymorphism

Acknowledgments

This work was supported by grants from the National Basic Research Program of China (973 Program No. 2012CB934002, 2010CB912802); National High-tech R&D Program of China (863 Program No. SS2012AA020314, SS2012AA020821, SS2012AA020303); National Key Scientific instrument Special Programme of China (Grant No. 2011YQ03013405); National Science Foundation of China (Grant No. 81121004, 81071953 and 81161120432).

Disclosure of Potential Conflicts of Interest

J.G.H. is a consultant to MDxHealth. The other authors declare no conflict of interest.