DACT1 Overexpression in type I ovarian cancer inhibits malignant expansion and cis-platinum resistance by modulating canonical Wnt signalling and autophagy

Type I epithelial ovarian cancer (EOC) is primarily resistant to platinum-based chemotherapies and needs novel therapeutics. Given the aberrant Wnt activation in type I EOC and the involvement of Dapper1 Antagonist of Catenin-1 (DACT1) in Wnt signalling, the role of DACT1 in tumourigenesis of type I EOC was evaluated. Firstly, all tested EOC cell lines and primary EOC tissues, especially type I EOC, were observed to have significantly lower DACT1 expression than normal controls. Next, 3AO cells, which arise from a patient with primary mucinous EOC and express low endogenous levels of DACT1, were transfected with a lentivirus carrying full-length DACT1 (3AO-DACT1), grew slower and formed smaller tumours in nude mice compared to 3AO-NC. Furthermore, 3AO-DACT1 had lower levels of key mediators of canonical Wnt signalling, Dvl2 and β-catenin, GSK-3β with phosphorylated Ser9, and the Wnt/β-catenin target genes, with significantly lower nuclear β-catenin levels. Additionally, 3AO-DACT which contained higher levels of lipidated LC3 (LC3-II) and Beclin1, but lower levels of p62/SQSTM1, were more sensitive to cis-platinum. And chloroquine partially rescued its cis-platinum resistance. We identified DACT1 as a negative regulator in type I EOC, protecting against malignant expansion by inhibiting canonical Wnt signalling and cis-platinum resistance by regulating autophagy.


DACT1 expression was depressed in ovarian cancer cell lines and primary ovarian cancer tissues.
To investigate whether DACT1 is implicated in ovarian cancer, we first examined its expression in several ovarian cancer cell lines by quantitative real-time RT-PCR. We found that DACT1 expression was significantly lower in all tested EOC cell lines (SKOV3, ES-2, OVCAR3, 3AO, A2780 and HEY) than in normal ovarian tissues (Fig. 1a). Interestingly, compared with that in type II EOC, the level of DACT1 expression was significantly lower in type I EOC, including 3AO, OVCAR3, ES-2 and A2780 (Fig. 1b), among which 3AO was reported arising from a patient with primary mEOC 25 .
A typical CpG island was found to span the proximal promoter and exon 1 regions of the DACT1 gene 26 . We thus speculated that DACT1 promoter methylation may be responsible for DACT1 expression depression in EOC cell lines. As indicated in Fig. 1c, after incubation with Decitabine in the presence or absence of histone deacetylase inhibitor TSA, the DACT1 expression in all EOC cell lines was dramatically restored, suggesting that methylation of the DACT1 promoter may contribute to DACT1 silencing in EOC (Fig. 1c). Similar results were observed by semi-quantitative RT-PCR (Supplementary Figs S1 and 2).
To investigate the clinical relevance of these findings, we measured DACT1 mRNA levels in type I ovarian cancer tissues and normal ovarian tissues. Type I primary EOC tissue samples were collected from 18 patients during surgery. As expected, DACT1 expression was significantly lower than in normal ovarian tissues (P < 0.0001) (Fig. 1d).
We further analysed the expression of DACT1 protein in 49 cases of type I EOC tissues and 10 cases of normal ovarian tissues using immunohistochemistry (IHC). The intensity of DACT1 staining in EOC tissues significantly is weaker than in normal control tissues (P < 0.0001). Interestingly, we found that benign, borderline and malignant ovarian mucinous tumour tissues were co-existed in five cases of mEOC specimens. Staining of DACT1 was more intense in benign lesions than in cancer tissues (Fig. 1e,f). Furthermore, in the sections of five cases with mEOC, the expression of DACT1 in benign lesion was observed significantly stronger than that in cancer tissues on the same section ( Supplementary Fig. S3). Such stepwise change of DACT1 expression from normal to benign to malignant tissue further supports the role of DACT1 in the etiopathogenesis of type I EOC. Depressed expression of DACT1 in malignant cells further supports the etiopathogenetic role of DACT1 in mEOC.

DACT1 inhibits ovarian cancer cell growth.
To elucidate the function of DACT1 in mEOC, we examined the effect of DACT1 expression on the growth characteristics of mucinous ovarian cancer cells using colony formation and growth curve assays. 3AO was isolated from a patient with primary mucinous adenocarcinoma 25 and chosen for its histologic type ( Supplementary Fig. S4). 3AO cells, which express very low endogenous levels of DACT1, were transfected with a lentivirus carrying full-length DACT1 (3AO-DACT1) or a control vector (3AO-NC). Stable expression of DACT1 was confirmed using western blot and quantitative real-time RT-PCR ( Fig. 2a,b). MTT assay was used to assess proliferation. The proliferation of 3AO-DACT1 was significantly slower than 3AO-NC (P < 0.001) (Fig. 2c). Consistent with this finding, 3AO-DACT1 formed significantly fewer and smaller colonies than 3AO-NC (P < 0.001) (Fig. 2d,e).
Next we assessed the effect of DACT1 on cell cycle in 3AO cells. Cells were synchronized at the G0/G1 phases by serum starvation for 16 h, then cultured in complete medium for 72 h and cell cycle was assessed by flow cytometry. 3AO-DACT1 progressed from G1 to S phase at a slower rate than 3AO-NC, and thus were arrested in the G0/G1 phase (Fig. 2f,g). Considering that apoptosis also affects tumour growth, we assessed the apoptosis rate of 3AO-DACT1 and 3AO-NC, however, there was no significant difference between the two groups (data not shown).

DACT1 inhibits ovarian tumourigenesis in vivo.
To further elucidate the tumour-suppressive function of DACT1 in vivo, 3AO-NC or 3AO-DACT1 cells were transplanted into five nude mice. Tumour size was measured and the growth-curves of tumours were observed (P < 0.05) (Fig. 3a), then tumours were excised for immunohistochemistry. The graft tumours from 3AO-NC injected-animals were significantly larger than those of 3AO-DACT1 (Fig. 3b). The mean tumour weight was significantly lower in the 3AO-DACT1 group (0.384 ± 0.176 g) than the 3AO-NC group (0.675 ± 0.102 g, P < 0.05) (Fig. 3c).
Proliferation of xenograft tumours was also assessed by Ki-67 assay. Consistent with the results obtained in vitro, DACT1-expressing tumours displayed significantly fewer proliferative cells than control tumours (Fig. 3d). The tumour-suppressive function of DACT1 was also confirmed by intraperitoneal injection of 3AO-DACT1 or 3AO-NC cells ( Supplementary Fig. S5a,b). These findings further implicated that DACT1 may suppress the proliferation of 3AO. DACT1 regulates cell cycle by attenuating β-catenin-dependent Wnt signalling. It is reported that canonical Wnt signalling can affect the EOC growth. To elucidate the mechanism by which DACT1 suppresses ovarian cancer cell growth, we assessed the levels of the key mediators of canonical Wnt signalling, such as Dvl2 and β-catenin, in 3AO-DACT1 (Fig. 4a,b). We found that cellular levels of Dvl2 and β-catenin were lower in 3AO-DACT1 than 3AO-NC, and cellular levels of GSK-3β with phosphorylated Ser9, and the Wnt/β-catenin target genes, CyclinD1 and C-myc, were lower in 3AO-DACT1 than 3AO-NC (Fig. 4c,d).
The subcellular location of β-catenin was assessed to investigate whether DACT1 influenced the nuclear translocation of β-catenin (Figs 4e,f and S6). Nuclear β-catenin levels were significantly lower in 3AO-DACT1 than 3AO-NC. These results further support the theory that DACT1 regulates the subcellular localization of β-catenin by stimulating GSK-3β. DACT1 overexpression induces autophagy and sensitizes 3AO to cis-platinum. DACT1 was found to act as a positive regulator to promote the formation of the Beclin1-Vps34-Atg14L complex and enhance autophagy 23 , and autophagy has been reported to be a putative mechanism of resistance to cis-platinum (CDDP) recently 27,28 . These findings prompted us to explore whether DACT1 could influence the cis-platinum sensitivity of type I EOC by activating autophagy.
We first assessed the response of various EOC cell lines to cis-platinum, and measured viability using the CCK-8 assay (Fig. 5a,b). As expected, EOC cell lines with different backgrounds response differently to cis-platinum. We found that 3AO, a typical mEOC cell line, was resistant to cis-platinum, while HEY, a cell line detached from a high grade serous ovarian cancer tissue, was relatively sensitive to cis-platinum. We thus compared the survival ratio between 3AO-DACT1 incubated with cis-platinum ( Fig. 5c,d) and found that DACT1 expression increased its sensitivity to cis-platinum.
We measured the expression of P62, Beclin1 and lipidated LC3 (LC3-II)/LC3-I protein by Western Blot (Fig. 5e,f). 3AO-DACT1 contained higher levels of LC3-II and Beclin1, and lower levels of p62/SQSTM1. We then tested whether chloroquine, an autophagy inhibitor, could affect the cell survival of 3AO-DACT1 and 3AO-NC cells. However, there was no significant difference in cell survival with autophagy blocked ( Supplementary  Fig. S7). To further examine the role of autophagy in cis-platinum resistance, 3AO-DACT1 cells were pre-treated with autophagy inhibitor chloroquine (Chl). As expected, chloroquine partially rescued cis-platinum resistance of 3AO-DACT1 (Fig. 5g,h). Thus these observations implied that DACT1 expression might be pivotal in enhancing sensitivity of mEOC to cis-platinum by induction of autophagy.

Discussion
As a typical example of type I EOC, mucinous ovarian carcinomas, accounts for 10% of EOC and usually appears as unilateral and expansive tumours mainly featured less aggressiveness 29 . More than half of mEOC patients were diagnosed early (FIGO I-II) and could be removed completely at initial surgery. However, there did have some cases with mEOC were found at advanced stages (FIGO III-IV) in clinical practice. Due to initial resistance to platinum based chemotherapy, those with suboptimal cytoreductive surgery carried even poorer survival in contrast with the same stage type II high grade serous cancer patients. And thus new therapeutic approaches are being needed for these patients. DACT1 is a negative regulator of Wnt signalling via promotion of Dvl degradation 16,17 . In this study we sought to investigate whether DACT1 is involved in type I EOC, especially in mEOC.
We found that expression of DACT1 was lower in EOC cell lines than normal ovarian tissues, and further that DACT1 expression was significantly lower in type I EOC cell lines, which include 3AO, a mEOC cell line arising from a patient with primary mucinous ovarian cancer. Aberrant down regulation of DACT1 was previously observed in hepatocellular carcinoma, gastrointestinal stromal tumours and non-small-cell lung cancer (NSCLC) [18][19][20] , whereas in colon cancer and squamous cell carcinoma DACT1 was reported to be overexpressed 21,22 . underwent surgery. The normal ovarian tissues were obtained from patients with benign disease (adenomyosis, adenoma) who chose hysterectomy and oophorectomy (****P < 0.0001). (e) DACT1 protein expression in type I ovarian cancer and normal ovarian tissues was analysed by immunohistochemistry. Type I primary EOC tissue samples were obtained from 49 patients that underwent surgery. Benign and malignant ovarian mucinous tumour tissues were found in 5 patients, and a representative case was provided. Left: normal ovarian tissue, Right: mucinous carcinoma tissue, (which were captured in the same section). Images were digitally captured at a ×400 magnification ratio. (f) H-score was used to evaluate the level of DACT1 expression in type I ovarian cancer tissues and in normal ovarian tissues (****P < 0.0001). All the experiment was repeated at least three times.
SCiEntiFiC RepoRts | 7: 9285 | DOI:10.1038/s41598-017-08249-7 We attempted to further analyse the association between DACT1 expression and prognosis of mEOC patients by querying the NCI-Cancer Genome Atlas (TCGA) and GEO database for mEOC samples. However, we identified no cases of mEOC preventing analysis, probably on account of its fairly low incidence in clinical practice. Thus the clinical implications of our results should be further confirmed by studies of larger numbers of patients. Epigenetic disruption of tumour suppressor genes, including promoter methylation and histone modification, is a major mechanism of cancer gene regulation 30 . In HCC 18 , bladder urothelial carcinoma 31 and gastric cancer 32 , DACT1 was previously reported to be silenced by promoter methylation. The DACT1 promoter was previously reported to contain a typical CpG island 26 , and incubation with Decitabine or together with histone deacetylase inhibitor TSA restored DACT1 expression in EOC cell lines, suggesting that epigenetic modification may directly   α-tubulin was used as the loading control. Grey values collected by Fusion were analysed (*P < 0.05, **P < 0.01). (g) and (h) Viability of 3AO-DACT1 or 3AO-NC cells incubated with the indicated concentrations of cis-platinum in the presence or absence of chloroquine (***P < 0.001, paired t test, P < 0.05, **P < 0.001, ****P < 0.0001). All the experiment was repeated at least three times. or indirectly influence the transcription levels of DACT1. However, the methylation level of CpG sites within the promoter of DACT1 were not investigated by MSP and BGS analysis, and future studies will be necessary to determine whether DACT1 promoter CpG island hypermethylation could be used as a biomarker of EOC.
3AO is a typical mEOC cell line which is arised from a patient with primary mucinous adenocarcinoma, and expresses very low endogenous levels of DACT1 25 . 3AO cells were transfected with a lentivirus carrying full-length DACT1 (3AO-DACT1) or a control vector (3AO-NC). Over-expression of DACT1 suppressed cell proliferation of 3AO, and when transplanted into nude mice, 3AO-DACT formed smaller tumours. Flow cytometry analysis revealed that DACT1 could reduce the proliferation of 3AO cells through inhibiting the transition from G1 to S phase of the cell cycle and thus resulting in an accumulation of G0/G1 phase population.
Wnt signalling has been reported to play key roles in the regulation of tumourigenesis 14,15 . Since aberrant activation of Wnt signalling has been implicated in the pathogenesis of EOC and DACT1 can regulate Wnt signalling, we investigated whether DACT1 could regulate tumour growth by modulating Wnt signalling. We found that 3AO-DACT1 contained significantly lower levels of key mediators of canonical Wnt signalling, such as Dvl2, both total and nuclear β-catenin, GSK-3β with phosphorylated Ser9, and the Wnt/β-catenin target genes, CyclinD1 and C-myc. These data suggested that DACT1 acted in 3AO cells by downregulating Dvl2 expression, and blocking the total β-catenin and β-catenin/TCF target gene, CyclinD1 and C-myc. β-catenin stability is regulated by the APC complex, containing Axin, adenomatous polyposis coli, glycogen synthase kinase 3β and casein kinase 1. Interaction of GSK-3β with Axin in the complex facilitates efficient phosphorylation of β-catenin 33 . Phosphorylated β-catenin is subsequently ubiquitinated, leading to its rapid proteasomal degradation. Our data demonstrated that over-expression of DACT1 enhances the activity of GSK-3β through attenuating the level of phosphorylated GSK-3β at Ser9, which could make lower expression of β-catenin translocated in the nuclear. These results support the finding that DACT1 regulates the subcellular location of β-catenin through enhancing the activity of GSK-3β 34 .
DACT1 was also recently reported to act as a positive regulator, promoting formation of the Beclin1-Vps34-Atg14L complex and enhancing autophagy 23 . Platinum-derived compounds act by disrupting DNA structure, initiating apoptosis where the damage cannot be repaired 35 . Autophagy has been reported to be a putative mechanism of cis-platinum resistance 27,28 , which can either sensitize cells to cancer killing agents 36 or confer resistance to cis-platinum 37,38 . We thus explored whether DACT1 could influence the cis-platinum sensitivity of mEOC by activating autophagy.
We found that 3AO, a kind of mEOC cell line, was more resistant to cis-platinum than HEY, a typical type II EOC cell line, which is consistent with the clinical facts that mEOC is more resistant to cis-platinum 7,8,29 . During induction of autophagy, the nonlipidated form of LC3 (LC3-I, 18 kDa) is conjugated with phosphatidylethanolamine (PE), and converted into the lipidated form LC3 (LC3-II, 16 kDa), which is associated with autophagosome biogenesis, and activity of autophagy is associated with changes of the cellular level of LC3-II [39][40][41] . Additionally the autophagy receptors SQSTM1/P62 42 and Beclin1 43 have also been used as markers of autophagy induction. 3AO-DACT1 contained higher levels of LC3-II, Beclin1 and lower levels of p62/SQSTM1 and autophagy inhibitor chloroquine partially rescued cis-platinum resistance of 3AO-DACT1. However, the involvement of other mechanisms should also be considered as cis-platinum sensitivity could be only partially reversed by chloroquine treatment. Overall, these results highlight the role of autophagy in the treatment of type I EOC, especially the management of ovarian mucinous carcinoma.
In this study, we found in mEOC, DACT1 can inhibit Wnt signalling and active autophagy. A mutually regulative relationship between autophagy and Wnt signalling has been reported 44,45 and it needs further study to testify whether it be involved in the regulatory role of DACT1 in tumour growth and cis-platinum resistance.
In conclusion, our results suggest that DACT1 expression is depressed in EOC probably by methylation, for the depressed expression could be reversed by pharmacological demethylation. Differences in the level of DACT1 expression between type I and II EOC suggest that this protein plays a key role in the specific behaviour of type I EOC. Overexpression of DACT1 suppressed proliferation of a mEOC cell line, likely by enhancing GSK3β and inhibiting Wnt/β-catenin signalling. In addition, overexpression of DACT1 increased its chemosensivity to cis-platinum, likely by enhancing autophagy activity. Our findings suggest that DACT1 functions as a tumour suppressor in type I EOC and highlight a potential new biomarker of cis-platinum response and a novel therapeutic target against type I EOC, especially against mEOC.

Materials and Methods
Tissue specimens. Primary type I EOC tissue samples were obtained from 49 patients that underwent sur-

DACT1 expression in ovarian cancer cell lines.
A lentivirus carrying full-length DACT1 DNA sequence was constructed by GeneChem (Shanghai, China). 1 * 10 5 cells were seeded per well in 6-well plates and next day the medium was replaced with 4 * 10 6 lentivirus (MOI 40) in 1 mL serum-free 1640 with 1 μg/ml polybrane. 24 hours later, the medium was replaced with complete medium. Stable DACT1 expressing cells were selected by incubation with 1 mg/ml puromycin for 72 h.
Quantitative reverse transcription polymerase chain reaction. Total RNA was isolated from cell lines and tissues using TRIZOL (TAKARA, Japan) according to the manufacturer's protocol, and DACT1 expression was assessed by quantitative Real-time RT-PCR using the following primers: sense, 5′-GACGAGCAGAGCAATTACACC-3′; antisense, 5′-ACCGTTTGAATGGGCAGA-3′. DACT1 expression was normalized to β-actin expression using the comparative CT-method 47  Colony-formation assay. Colony-formation assay was performed using a monolayer culture. Cells were re-suspended in RMPI 1640 and 500 cells were seeded per well in 24-well plates. Plates were incubated at 37 °C for 12 days, and colonies were stained with crystal violet (0.005%; Sigma, USA). All experiments were performed in triplicate.
MTT assay. 3AO Immunohistochemistry. Formalin-fixed, paraffin-embedded tissue sections were dewaxed with xylene and dehydrated by graded alcohol, then permeabilized by incubation in TritonX-100 (0.3%) for 10 min. Endogenous peroxidase activity was blocked with incubation in 3% H 2 O 2 for 15 min at 37 °C. After antigen retrieval in citrate buffer, all sections were blocked with normal goat serum for 30 min to eliminate nonspecific binding, samples were incubated with DACT1-directed antibody (dilution 1:200, Abcam, Cambridge, UK) overnight at 4 °C, then biotinylated anti-rabbit secondary antibody for 30 min at 37 °C. Samples were rinsed in PBS then incubated with SABC reagent (Boster, China) according to the manufacturer's instructions. Sections were stained by DAB and then counterstained with haematoxylin.
Images were digitally captured at a ×400 magnification ratio using an Olympus BX51 optic microscope (Olympus, Japan). Staining intensity was assessed using a histochemistry score (H-score), as previously described 48 . In vivo tumour growth. Six-week-old female nude mice (approximately 18 g) were obtained from the Experimental Animal Centre of Chongqing Medical University. All the mice were bred under SPF (Specific pathogen Free) conditions. All the mice were allowed to acclimate for 1 week, and then 5 * 10 6 3AO-DACT1 or 3AO-NC cells were injected into the subcutaneous tissue of each nude mouse (n = 5 animals per group) in the subcutaneous group. Tumour size was measured weekly with a Venire calliper, and tumour volumes were calculated according to the following formula: π/6* length * width 2, 49 . In the intraperitoneal injection group, 1 * 10 7