Vitamin D receptor regulates proliferation and differentiation of thyroid carcinoma via the E-cadherin-β-catenin complex

Thyroid cancer has the fastest rising incidence among cancers, especially for differentiated thyroid carcinoma (DTC). Although the prognosis of DTC is relatively good, if it changes to anaplastic thyroid carcinoma (ATC), the prognosis will be very poor. The prognosis of DTC is largely depending on the degree of cell differentiation and proliferation. However, whether the vitamin D receptor (VDR) plays a role in regulating the proliferation and the differentiation of DTC cells is unclear. In the present study, we found that VDR was upregulated in DTC tissues compared to the adjacent non-cancerous tissue. Knockdown of VDR increased proliferation and decreased differentiation proliferation in DTC cells in vitro as well as DTC cell-derived xenografts in vivo. In contrast, overexpression of VDR had an opposite effect. Knockdown of E-cadherin abolished VDR-induced suppression of proliferation and enhancement of differentiation of the DTC cells. Knockdown of β-catenin partially reversed the effect of the VDR knockdown. VDR increases the levels of E-cadherin in the plasma membrane and decreases the levels of β-catenin in the nucleus. VDR binds to E-cadherin and β-catenin in the plasma membrane of the DTC cell. Taken together, VDR inhibits DTC cell proliferation and promotes differentiation via regulation of the E-cadherin/β-catenin complex, potentially representing novel clues for a therapeutic strategy to attenuate thyroid cancer progression.


22
Thyroid carcinoma is one of the most common endocrine cancers worldwide and its 23 incidence has been rising in the past decades, becoming one of the tumors with the 24 fastest increase (Lim et al., 2017). It can be divided into papillary thyroid carcinoma 25 (PTC), follicular thyroid carcinoma (FTC), anaplastic thyroid carcinoma (ATC), and 26 medullary thyroid carcinoma (MTC). FTC and PTC are also defined as differentiated 27 thyroid carcinoma (DTC), accounting for 90% of thyroid malignancies (Ito et al., 2013). 28 29 Although most of DTC has an excellent prognosis with 80-95% 10-year survival rates 30 but high risk for the long-term harm, the prognosis significantly worsens when the 31 tumor grows, and the degree of tumor differentiation decreases (Fagin and Wells, 2016;32 Filetti et al., 2019). Furthermore, ATC accounts for 1% of all thyroid cancers with a 33 poor survival (Ain, 1998;Fagin and Wells, 2016). For these reasons, it is important to 34 understand the characteristics of the tumor at the early stage and to better prevent the 35 progression of the tumor. It has been shown that VDR is strongly expressed in DTC tissues compared with 48 adjacent normal thyroid tissues, but is lower or absent in human ATC tissues 49 (Clinckspoor et al., 2012;Izkhakov et al., 2016). VDR has been shown to inhibit growth 50 of several types of tumors by directly or indirectly regulating cell cycling and 51 proliferation, differentiation, and apoptosis. However, the significance of the elevated 52 VDR levels in DTC and the role of VDR in regulating DTC proliferation and 53 differentiation remains unknown. 54 55 In addition to the role of VDR in mediating 1,25(OH) 2 D 3 -regulating cell proliferation 56 and differentiation, VDR regulates cell differentiation and proliferation in a ligand-57 independent and non-genomic fashion in some cell types. It has been reported that mice 58 with an overexpression of the ligand binding site of VDR in the epidermis restored 59 normal hair cycling in VDR knockout mice (Skorija et al., 2005), suggesting that the 60 role of VDR in epidermal hair regulation is independent of its ligand. VDR interacts 61 with β-catenin to regulate the transcription of target genes, thus participating in the 62 regulation of keratinocyte stem cell function (Cianferotti et al., 2007). 63 64 E-cadherin is an adhesion molecule and plays an important role in regulating different 67 Previous studies have revealed an association of the reduction of E-cadherin with 68 stronger nuclear staining for β-catenin in advanced malignancies (Tafrihi and Nakhaei 69 Sistani, 2017). Besides, E-cadherin is expressed in normal thyroid and papillary thyroid 70 carcinoma but decreased and even lost in ATC (Brabant et al., 1993). In thyroid cancer 71 cells, the level of β-catenin in the plasma membrane decreases in undifferentiated or 72 aggressive thyroid cancer cells compared to well-differentiated thyroid cancer cells 73 (Garcia-Rostan et al., 2001). However, the role of E-cadherin-β-catenin signaling in 74 DTC is still not well understood. 75 76 The present study sought to determine whether VDR regulates DTC cell proliferation 77 and differentiation via E-cadherin and β-catenin. The GAPDH (Cat#60004-1-Ig), horseradish peroxidase-conjugated affinipuregoat 112 anti-rabbit IgG (SA00001-2), and horseradish peroxidase-conjugated affinipuregoat 113 anti-mouse IgG (SA00001-1) were purchased from Proteintech (Rosemont, USA). BrdU cell proliferation assay 156 The proliferation of K1 cells and WRO cells was determined by BrdU cell proliferation 157 assay using a BrdU cell proliferation assay kit according to the manufacturer's 158 instructions (Millipore, Billerica, MA). Briefly, cells were plated in 96-well plates at a 159 density of 1 x 10 4 cells/well were cultured for 2 days. Cells were incubated in the serum-160 free medium supplemented without FBS for a further 12 hours, the proliferation assay Colony formation assays 169 For colony formation assay, 1x10 3 cells were cultured in six-well plates with the 170 medium at 37 °C in 5% CO 2 and colony formation was assessed after 2 weeks. Cells 171 were fixed with 4% paraformaldehyde for 20 min and stained with 0.1% crystal violet 172 for 20 min at the end of the experiment. The visible colonies were counted and imaged. assay kit following the manufacturer's instructions. Briefly, sections were exposed to 5 177 μg/ml of proteinase K for 25 minutes at 37°C and then treated with 0.2% Triton X-100 178 solution for 20 minutes at room temperature. The sections were washed extensively and 179 exposed to PBS (PH 7.4). According to the number of sections and tissue size, an 180 appropriate amount of reagent 1 (TdT) and reagent 2 (dUTP) of the TUNEL kit were 181 mixed at 1:9 and added to the covered tissue. The sections were incubated at 37°C for 182 2 hours in a humidified chamber. Each slide was then washed with PBS (PH 7.4) for 5 183 min and repeated for 3 times. Then DAPI staining solution was added and incubated 184 for 10 min at room temperature in a dark room. The sections after sealing were observed 185 using a fluorescence microscope and the images were collected.  Table 1. 198 Total cell lysates were obtained by extraction in RIPA buffer (50 mM HEPES,pH 7.4,199 1% Triton X-100, 0.1% SDS, 150 mM NaCl, 1 mM EDTA) containing a protease

261
The expression of VDR in DTC tissue was higher than that of the adjacent non-262 cancerous thyroid tissue. 263 To determine the expression level of VDR in DTC tissue, we examined the immunostaining level of VDR in 20 paired PTC and adjacent non-cancerous tissues 265 using immunohistochemistry. The results showed that VDR expression was increased 266 significantly in the PTC tissue compared with paired adjacent non-cancerous thyroid 267 tissue, and VDR staining was mainly localized in most cytoplasm and plasma 268 membrane of PTC, whereas VDR staining was localized in the nucleus of paired 269 adjacent non-cancerous thyroid tissue (Fig. 1A). Moreover, we examined the increased β-catenin expression in the nucleus in DTC cells. 304 We examined the expression and location of E-cadherin and β-catenin in K1 cells with 305 VDR knockdown. The results showed that knockdown of VDR reduced E-cadherin, 306 but not β-catenin expression in K1 cell, (Fig. 3A). To determine whether localization 307 of E-cadherin and β-catenin is regulated by VDR, we examined E-cadherin and β-catenin localization after VDR knockdown using cellular fractionation and Western 309 blot analysis. Remarkably, the level of E-cadherin expression in the membrane was 310 attenuated and the level of β-catenin expression in the nucleus was increased in K1 cells 311 after VDR knockdown (Fig.3B). As shown by immunocytochemistry and 312 immunofluorescence, knockdown of VDR reduced the E-cadherin expression in the 313 membrane and enhanced β-catenin expression in the nucleus ( Fig. 3C and 3D). This 314 data suggests that VDR reduces E-cadherin expression and localization of β-catenin in 315 the plasma membrane and enhances localization of β-catenin in the nucleus.

328
cadherin and β-catenin dependent manner. 329 To determine the mechanism by which VDR regulates DTC cell proliferation and 330 differentiation, we firstly investigated whether VDR regulates E-cadherin and β-catenin. 331 The results showed that the expression of E-cadherin was decreased in DTC cells with 332 VDR knockdown (Fig.3A). To further delineate the role of E-cadherin in the 333 proliferation and differentiation regulated by VDR, we use E-cadherin siRNA to 334 knockdown the expression of E-cadherin in DTC cells in which VDR was 335 overexpressed. As shown in Fig. 4, knockdown of E-cadherin blocked VDR suppressed 336 DTC cell proliferation (Fig. 4E) and induced TPO and TSHR expression as indicated 337 by immunoblotting ( Fig. 4A and 4B). These data suggest that E-cadherin is required 338 for VDR suppressed DTC cell proliferation and VDR-induced DTC cell differentiation. To confirm the effect of VDR knockdown on DTC tumor growth, we examined effects 369 of overexpression of VDR in a xenograft experiment using SCID mice (n = 6 for each 370 group). The results showed that the expression of VDR in K1 cells was upregulated in 371 about two folds as shown by immunoblotting (Fig. 6A). As shown in Fig. 6B, 372 xenografts in the VDR overexpression group was smaller compared to the control group. 373 The tumor volume and the mean tumor weight were lower in mice xenografted with 374 VDR overexpressed K1 cells compared to those xenografted with cells transfected with 375 the vector control ( Fig. 6C and 6D), indicating that VDR inhibits tumor growth. 376 Consistent with the expression of VDR in K1 cells, the level of VDR was increased in 377 the VDR overexpressed tumors than that in control tumors (Fig. 6F). The expression of 378 BrdU, PCNA, and Ki67 were decreased in the VDR overexpressed tumors (Fig. 6E). 379 However, TUNEL staining was not significantly different between the two groups ( Fig.   380 6G), suggesting that VDR inhibits the DTC tumor growth and cell proliferation but 381 does not affect apoptosis. Furthermore, the expression of TPO and TSHR in VDR 382 overexpressed tumors were increased than these in control tumors (Fig. 6F)

386
In the current study, we demonstrate that VDR was upregulated in DTC tissues than 387 that in adjacent non-cancerous tissues, and the level of VDR positively correlates with 388 differentiation of DTC. Furthermore, the results provide the first evidence that VDR 389 suppresses proliferation and promotes differentiation both in DTC cells in vitro and 390 DTC cell derived xenografts in vivo. Besides, we demonstrate that VDR binds to the 391 E-cadherin/β-catenin complex in the plasma membrane and works in an E-cadherin and 392 β-catenin dependent manner. Taken as a whole, these results suggest that VDR is 393 involved in DTC progression and will help uncover novel clues for the therapeutic 394 strategy of DTC.
Our results confirm that VDR levels are highly expressed in DTC tissue than that in the 397 adjacent non-cancerous tissue, which is consistent with the previous reports (Khadzkou In the present studies, we investigated the molecular mechanism by which VDR 427 inhibited DTC proliferation and promoted DTC differentiation. This study shows that 428 overexpression of VDR increased the E-cadherin expression in DTC cell lines. 429 However, when E-cadherin is reduced, β-catenin complex is separated from the E-       Densitometric quantification analysis results were presented in the right panel. The data were presented as densitometric ratios normalized to GAPDH. (D) VDR mRNA expression comparison analysis between the normal thyroid gland and PTC & FTC based on the Oncomine database. Each assay was performed in triplicate. Image J was used for WB quantification. Each assay was performed in triplicate (n=3). Statistical were analyzed by independent samples T-test or ANOVA analysis (*p < 0.05 vs normal thyroid cells/tissues).
170x125mm (300 x 300 DPI) Figure 2. VDR suppressed the proliferation and promoted the differentiation of DTC cells in vitro. K1 cells were infected with lentivirus containing VDR shRNA (VDR-shRNA) or control shRNA (Control-shRNA). Cell proliferation was determined by BrdU incorporation (A) and colony formation assay (B) after knockdown of VDR in K1 cells. WRO cells were infected with adenovirus containing VDR cDNA (VDR-adv) or control adenovirus (Control-adv). Cell proliferation was determined by BrdU incorporation (C) and colony formation assay (D) after overexpression of VDR in WRO cells. (E-H) qRT-PCR and western blotting analyses were performed to examine the effects of VDR overexpression or knockdown on differentiation markers including TPO, TSHR and NIS mRNA and protein level. The data were presented as densitometric ratios normalized to GAPDH. Image J was used for WB quantification. Each assay was performed in triplicates (n=3). Statistical were analyzed by independent samples T-test or ANOVA analysis (*p < 0.05 vs control).
170x187mm (300 x 300 DPI) Figure 3. VDR increased E-cadherin expression in the membrane and reduced β-catenin expression in the nucleus in DTC cells. Cells were treated with VDR-shRNA and control-shRNA. (A) The E-cadherin and βcatenin total protein expression by Western blot. Densitometric quantification analysis results were presented in the below panel. The data were presented as densitometric ratios normalized to GAPDH. (B) Cell nuclear and membrane fractions were isolated. Western blot was performed to detect VDR, E-cadherin and β-catenin. Na+/K+ATPase and Histone H3 were used as positive control for membrane and nuclear fractions, respectively. Densitometric quantification analysis results were presented in the below panel. The data were presented as densitometric ratios normalized to Na+/K+ATPase or Histone H3. (C) The immunocytochemistry of E-cadherin and β-catenin in cells with or without VDR knockdown. Positive staining of cells (Brown) was indicated by arrows. (D) The immunofluorescence of VDR (green), E-cadherin and βcatenin (Red) in cells with or without VDR knockdown. (E) Co-IP experiments were performed for VDR using K1 cells membrane protein incubated with anti-VDR antibody or nonspecific IgG. After immunoprecipitation with anti-VDR antibody, western blotting was performed with antibodies of VDR, E-cadherin and β-catenin, respectively. Nonspecific IgG was used as a negative control. Image J was used for WB quantification. Each Figure 4. VDR reduced proliferation and increased differentiation of DTC cells in an E-cadherin and β-catenin dependent manner. WRO cells were treated with or without VDR-adv and then treated with control siRNA (si-NC) or E-cadherin siRNA (si-E-cadherin). K1 cells were transfected with or without VDR shRNA and then treated with control siRNA (si-NC) or β-catenin siRNA (si-β-catenin). (A-D) Differentiation markers including TPO, TSHR and VDR, E-cadherin, β-catenin were evaluated using Western blotting. The data were presented as densitometric ratios normalized to GAPDH. (E-F) Cell proliferation was determined by BrdU incorporation in DTC cells. Each assay was performed in triplicates (n=3). Image J was used for WB quantification. Statistical were analyzed by independent samples ANOVA analysis (*p < 0.05 vs control).
170x194mm (300 x 300 DPI) Figure 5. Effects of VDR knockdown on the proliferation and differentiation of DTC xenograft in vivo. The stable knockdown of VDR in K1 cells was established by selection with 2 μg/ml puromycin. K1 cells infected with control shRNA were used as a negative control. Then 5x106 cells with VDR stable knockdown or control cells were implanted subcutaneously into the flanks of eight-week-old female SCID mice (n = 6 for each group). Tumor diameters were measured once a week for five weeks. BrdU (50 mg/kg body weight) was injected intraperitoneally into mice two hours before euthanization. color indicates apoptotic cells. Nuclei were stained with DAPI (blue). Significance between two groups was analyzed by independent samples T-test and Mauchly's Test followed by Sphericity Assumed or modified Greenhouse-Geisser analysis was performed to distinguish within-subjects' effects over time and Between-Subjects Effects (*p < 0.05 vs control).
170x193mm (300 x 300 DPI) Figure 7. The potential mechanism by which VDR suppresses proliferation and promotes differentiation of DTC cells. When the VDR level is low in the cell membrane, the β-catenin dissociates from E-cadherin in the plasma membrane and enters the nucleus, which leads to cell proliferation. When the VDR level is normal or high, it binds to E-cadherin and β-catenin in the plasma membrane and subsequently prevents β-catenin from entering the nucleus, which inhibits cell proliferation and promote cell differentiation.
109x95mm ( (E-H) Western blotting analyses were performed to examine the effects of VDR overexpression or knockdown on differentiation markers including TPO, TSHR protein level. Densitometric quantification analysis results were presented in the right panel. The data were presented as densitometric ratios normalized to GAPDH. Image J was used for WB quantification. Each assay was performed in triplicate (n=3). Statistical were analyzed by independent samples T-test or ANOVA analysis (*p < 0.05 vs control).