DKK1 promotes migration and invasion of non–small cell lung cancer via β-catenin signaling pathway

Abstract

Disregulation of dickkopf-related protein 1 (DKK1) has been reported in a variety of human cancers. However, how DKK1 functions in Non-small cell lung cancer has not been revealed. In the current study, DKK1 was knocked out by the lentivirus-mediated short hairpin RNA interference approach in H1299 and 95C non-small cell lung cancer cell lines. Subsequently, the migration and invasion ability were assessed by wound-healing and transwell assays. In addition, epithelial-mesenchymal transition markers and β-catenin were examined by Western blot analysis. The signaling pathway downstream of DKK1 was characterized using the Wnt signaling pathway inhibitor, IWP2, and glycogen synthase kinase 3 beta inhibitor, LiCl. Immunofluorescence analysis investigated the subcellular localization of β-catenin. The results suggested that knockdown of DKK1 caused reduced migration and invasion ability of H1299 and 95C cells. DKK1 silencing resulted in the downregulation of epithelial-mesenchymal transition-related proteins, such as Snail and zinc finger E-box binding homeobox 1. Besides, DKK1 silencing inhibited β-catenin and promoted the phosphorylation of β-catenin. Mechanism results indicated that the expression of β-catenin was reduced in H1299 or 95C cells after being treated with Wnt signaling inhibitor, IWP2. In addition, the inhibition of β-catenin phosphorylation by glycogen synthase kinase 3 beta inhibitor, LiCl, significantly enhanced the migration and invasion capacities in DKK1-knockdown cell lines. Furthermore, cell immunofluorescence revealed that nuclear β-catenin was reduced when DKK1 was knocked down. Taken together, these findings suggest that DKK1 induces the occurrence of epithelial-mesenchymal transition and promotes migration and invasion in non-small cell lung cancer cells. Mechanically, β-catenin plays a vital role in DKK1-induced non-small cell lung cancer cell migration and invasion, and DKK1 inhibits the phosphorylation of β-catenin, resulting in the increased nuclear localization of β-catenin.

Introduction

Non–small cell lung cancer (NSCLC) accounts for 80%–85% of all lung cancers, the leading cause of cancer-related deaths worldwide.^1,2 Most patients have advanced or metastatic disease at the time of diagnosis, and poor clinical outcome is common. ³ Patients with NSCLC have a low 5-year survival rate, although several approaches such as radiotherapy, chemotherapy, and surgical intervention can improve the clinical symptoms. ⁴ NSCLC progression and metastasis have been the primary causes for poor clinical outcomes of patients. Therefore, it is essential to explore the underlying molecular mechanisms of NSCLC migration and invasion; also, a diagnostic marker for NSCLC is an indispensable prerequisite.

The dickkopf (DKK) family of proteins include DKK1, DKK2, DKK3, DKK4, and a unique DKK3-related gene, Soggy. ⁵ DKK1 encodes a secreted protein, which was first found in Xenopus, ⁶ and involved in the head formation in embryonic development. In addition, DKK1 is an inhibitor of the canonical Wnt signaling pathway. ⁷ In the absence of Wnt, β-catenin degradation complex consisting of Axis inhibition protein (AXIN), adenomatous polyposis coli (APC), and glycogen synthase kinase 3 beta (GSK3β) can induce the phosphorylation of β-catenin, resulting in the proteolytic degradation. Canonical Wnt signaling pathway is activated by Wnt ligands by binding to Frizzled receptor and low-density lipoprotein receptor–related protein-5/6 (LRP5/6), which leads to nuclear translocation of β-catenin. Nuclear β-catenin activates T cell factor (TCF)/lymphoid enhancing factor (LEF) complex, regulating the expression of downstream genes.^8,9 However, DKK1 competitively binds to LRP5/6 and prevents the formation of Wnt-Frizzled-LRP5/6 complexes, thereby blocking the Wnt signaling pathway. ⁷ An increasing number of studies have indicated that DKK1 is involved in a variety of carcinomas. However, the role of DKK1 is very different in various cancers. Some researchers suggest that DKK1 acts as a tumor suppressor gene that inhibits proliferation and metastasis of cancer cells. For example, DKK1 downregulation was found in renal cell carcinoma ¹⁰ and colorectal cancer,^11,12 and a recent study showed that DKK1 is responsible for the bone metastasis of breast cancer cells. ¹³ By contrast, DKK1 overexpression was found in hepatocellular carcinoma ¹⁴ and myeloma, ¹⁵ acting as an oncogene. As for in lung cancer, the role of DKK1 is also controversial. Xu et al. ¹⁶ found that serum DKK1 is significantly lower in patients with lung cancer, but is rapidly normalized after treatment, indicating that DKK1 functions as a suppressor. However, other researchers indicated that DKK1 was upregulated at both messenger RNA (mRNA) and protein levels,^17–19 supporting its oncogenic role.

Although most studies have designated DKK1 as an oncogene, promoting NSCLC migration and invasion, the precise mechanism underlying NSCLC is yet unclear. Thus, in our study, we aimed to affirm the role of DKK1 in NSCLC cells and explore the specific mechanism. First, we depleted DKK1 with lentivirus-mediated short hairpin RNA (shRNA) in H1299 and 95C NSCLC cell lines and investigated the effect of DKK1 knockdown on cell migration and invasion, as well as epithelial–mesenchymal transition (EMT). Subsequently, the downstream of DKK1 was characterized using the Wnt signaling pathway inhibitor, IWP2, and GSK3β inhibitor, LiCl. Our results show that the knockdown of DKK1 inhibits the occurrence of EMT and decreases the migration and invasion in both NSCLC cell lines. Mechanically, DKK1 promotes cancer cell migration and invasion process via inhibiting the phosphorylation of β-catenin in NSCLC.

Materials and methods

Cell culture

Human lung adenocarcinoma cell lines, H1299 and 95C, were preserved in the Department of Cellular and Molecular Biology of Beijing Tuberculosis and Thoracic Tumor Research Institute. The cells were cultured in RPMI-1640 medium (Invitrogen, Carlsbad, CA, USA) containing 10% fetal bovine serum (FBS; Gibco, Los Angeles, CA, USA), 1000 U/L penicillin, and 1000 U/L streptomycin and maintained at 37°C in a humidified incubator containing 5% CO₂ and 95% humidity.

Transfection

ShRNA lentivirus was used to knockdown the expression of DKK1. A negative control (NC) was constructed using scrambled shRNA without silencing. ShRNA was provided by GeneChem (Shanghai, China).

For transfection, H1299 cells were seeded on 96-well plates at a density of 5000 cells/well and incubated at 37°C in a humidified incubator containing 5% CO₂. On the following day, the cells were transfected with combinant virus carrying DKK1-shRNA or NC-shRNA for 8–12 h and the virus diluent was replaced with RPMI-1640 medium containing 10% FBS. The cells with green fluorescence indicated stable transfection with lentivirus-mediated DKK1-shRNA or NC-shRNA. In total, >85% cells were examined with green fluorescence under a fluorescence microscope, thereby indicating a successful establishment of the cell models classified as H1299-NC and H1299-DKK1-KD (knockdown), respectively. Similarly, 95C-NC and 95C-DKK1-KD cell models were established.

Wound-healing assay

The wound-healing assay was performed to assess the cell migration ability. Cells were seeded on six-well plates in RPMI-1640 medium containing 10% FBS. After 24 h, straight lines were drawn by scraping the confluent cells with a 20 µL pipette tip and ruler. Then, the medium and floating cells were carefully removed, and the adherent cells were rinsed with normal saline thrice and allowed to culture with serum-free RPMI-1640 medium. Following further 24-h incubation, the wound-healing process was monitored under a phase-contrast microscope, and representative images were acquired at 0 and 24 h, respectively. All experiments were performed thrice to obtain a consensus.

Cell migration and invasion ability assays

Transwell assays were conducted to evaluate the cell migration and invasion using the HTS transwell-24 system (Corning, NY, USA), which is an array of 24 individual Boyden chambers with 8-µm pore size transwell membranes. For the assay, the cells were starved in serum-free RPMI-1640 medium for 24 h, following which 100 µL of 5 × 10⁴ cells were seeded onto the upper chamber while the lower chamber was filled with 600 µL RPMI-1640 medium containing 10% FBS as a chemoattractant. The cells in the upper chamber without serum can be stimulated to pass through the membrane by the high nutrient culture medium in the lower compartment, allowing the assessment of the cell migration ability. After incubation at 37°C for 24 h, the non-invading cells remained on the upper surface of the membrane and were removed with cotton swabs, whereas the cells that passed through the membrane were stained with 0.2% crystal violet, fixed with 30% glycerin on the slides, and then counted under a light microscope.

Before the invasion transwell assay, the transwell chambers were prepared with Matrigel (BD Biosciences, Bedford, MA, USA) on the upper surface. A volume of 100 µL of cell suspension with serum-free RPMI-1640 medium (containing 15 × 10⁴ cells) were seeded as described above. After incubation at 37°C for 48 h, the invaded cells were stained, fixed, and counted. Each experiment was performed in triplicate.

Western blot analysis

The adherent cells were harvested and washed three thrice with cold phosphate-buffered saline (PBS) and lysed with lysis buffer for 10 min followed by centrifugation at 14,000g for 10 min at 4°C. The protein concentration was determined using BioTek EPOCH (Winooski, VT, USA). Protein samples were heat denatured in 5× loading buffer at 100°C for 5 min. Equivalent protein lysates were fractionated by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membranes. The membranes were first blocked with 5% (w/v) dried skimmed milk in Tris-buffered saline with Tween 20 (TBST) for 1 h at room temperature and then probed with the indicated primary antibodies overnight at 4°C. Then, the membranes were washed thrice with 1× TBST followed by incubation with the corresponding horseradish peroxidase–conjugated secondary antibodies for 2 h at room temperature. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH; anti-GAPDH from Kangcheng, Inc., Shanghai, China) was used as an internal control. After washing thrice with TBST, the proteins were detected using the enhanced chemiluminescence (ECL; Pierce Biotechnology, Rockford, IL, USA) system and Alpha Innotech Fluorchem SP (San Leandro, CA, USA).

Cell immunofluorescence analysis

Cells were cultured on coverslips in six-well plates for 24 h and rinsed thrice with PBS on the following day. The cells were fixed with 4% paraformaldehyde for 10 min, permeabilized with 0.2% Triton X-100 for 15 min, and blocked with 3% bovine serum albumin for 30 min at room temperature. After incubating with rabbit anti-β-catenin (Abcam, Cambridge, UK) overnight at 4°C, the coverslips were incubated with rhodamine (TRITC)-conjugated AffiniPure goat anti-rabbit IgG secondary antibodies for 20 min and counterstained with Hoechst 33342 (Sigma-aldrich, St. Louis, MO, USA) for nuclear imaging. Finally, the cells were examined using confocal laser scanning microscopy (Leica, Wetzlar, Germany).

Statistical analysis

Each experiment was repeated at least three times and the results were expressed as mean ± standard error of the mean (SEM) from a representative experiment. Unpaired t tests were used to determine the statistical differences between two groups in each analysis. *p < 0.05 and **p < 0.01 were considered to be statistically significant. Western blot analysis was quantitatively analyzed by ImageJ (NIH Image, CA, USA). All statistical analyses were performed using GraphPad Prism 5.01 (GraphPad, La Jolla, CA, USA).

Results

DKK1 is knockdown in H1299 and 95C NSCLC cells

In order to explore the physiological role of DKK1 in migration and invasion capability of NSCLC cells, we knocked down the expression of DKK1 in H1299 and 95C cells by DKK1-shRNA; NCs were constructed using a non-silencing-shRNA. Considering that the green fluorescence protein (GFP) sequence was fused in the recombinant virus, we performed a fluorescence analysis to examine the transfection efficiency as illustrated in Figure 1(a) and (b). The results revealed that >85% cells were successfully transfected with DKK1-shRNA or non-silencing-shRNA. Western blot results demonstrated that the expression of the DKK1 protein was dramatically decreased in H1299-DKK1-KD (Figure 1(c)) and 95C-DKK1-KD cells (Figure 1(d)), respectively.

Figure 1.

DKK1 is knocked down in H1299 and 95C NSCLC cells. (a) H1299 and (b) 95C successfully transfected with lentivirus-mediated DKK1-shRNA or non-silencing shRNA (green) were examined under a fluorescence microscope. All the nuclei were visualized with Hoechst 33342 (blue), and over 85% cells were successfully transfected. (c and d) The expression of DKK1 protein was detected by Western blot analysis. DKK1 was significantly decreased in DKK1-KD group as compared to cells without transfection or NC group. GAPDH was used as an internal sample loading control (scale bar: 25 µm; **p < 0.01).

Cell migration and invasion abilities were decreased in DKK1 knockdown cells

First, we examined the role of DKK1 in cell migration using wound-healing and migration transwell assays. For the wound-healing assay, as illustrated in Figure 2(a) and (b), wound closure was lesser in the DKK1-KD cell group as compared to the control groups. In addition, we performed the migration transwell assay without Matrigel (Figure 2(c) and (d)) and found that fewer cells migrated to the other side of chamber membrane in the DKK1-KD group. Next, we explored the function of DKK1 in cell invasion ability using invasion transwell assay with Matrigel (Figure 2(e) and (f)). In agreement with the cell migration results, the number of cells that passed through Matrigel was lower in the DKK1-KD cell group. Collectively, these data indicated that the knockdown of DKK1 inhibited cell migration and invasion in vitro. Each analysis was performed in three independent experiments, and the data are represented as mean ± SEM (**p < 0.01).

Figure 2.

Cell migration and invasion abilities were decreased in DKK1 knockdown cell groups. The cell migration ability was examined using (a and b) wound-healing assay and the (c and d) migration transwell assay without Matrigel, and the cell invasion ability was examined with (e and f) transwell assay with Matrigel. In wound-healing assay, the migration distance ((scratch area of 0 h − scratch area of 24 h)/scratch area of 0 h.) in DKK1-KD group was significantly decreased as compared to controls. Transwell assay with and without Matrigel indicated that fewer cells invaded and migrated to the other side of chamber membrane in the DKK1-KD group as compared to controls. Related data statistics was shown in the right panel. Each analysis was performed in triplicate (**p < 0.01).

DKK1 silencing downregulated β-catenin and EMT-associated gene expression

Based on the above results from migration and invasion experiments, we investigated whether DKK1 regulates EMT. Western blot analysis (Figure 3(a) and (b)) revealed that the downregulation of DKK1 decreased the expression of zinc finger E-box-binding homeobox 1 (ZEB1) and Snail as compared to the control groups.

Figure 3.

DKK1 silencing downregulated β-catenin and EMT-associated gene expression. Western blot analysis revealed that silencing of DKK1 inhibited EMT-associated protein, ZEB1 and Snail, respectively in H1299 (a) and 95C (b). Besides, DKK1 downregulation decreased the expression of β-catenin and promoted the phosphorylation of β-catenin as compared to controls in H1299 (c) and 95C (d) respectively. GAPDH was served as an internal control. Gray scale value analysis was presented in the right panel. The analysis was performed based on three independent experiments and the data represented as mean±SEM. **p<0.01, *p<0.05.

DKK1 is an inhibitor of common Wnt/β-catenin signaling pathway, in which β-catenin plays a key role in regulating the expression of downstream genes. ⁵ Hence, we examined whether DKK1 knockdown could increase the expression of β-catenin. However, Western blot analysis (Figure 3(c) and (d)) demonstrated that as compared to the NC group, the downregulation of DKK1 decreased the expression of β-catenin, whereas the phosphorylation of β-catenin is significantly increased. Each analysis was performed in three separate experiments, and the data represented as mean ± SEM (**p < 0.01).

DKK1 promoted NSCLC cell EMT, migration, and invasion via β-catenin signaling pathway

To evaluate the underlying role of β-catenin in DKK1-induced EMT, migration, and invasion in NSCLC cells, we employed IWP2, the inhibitor of Wnt signaling pathway. ²⁰ Western blot analysis (Figure 4(a) and (b)) demonstrated that IWP2 inhibited β-catenin, the key molecule of Wnt signaling pathway. The expression of ZEB1 and Snail was decreased significantly in H1299 or 95C cells treated with IWP2 as compared to those without treatment; however, the EMT gene expression did not get altered significantly in the DKK1-KD cell group. With respect to the migration transwell without Matrigel (Figure 4(c) and (d)) and invasion transwell with Matrigel (Figure 4(e) and (f)) results, after treatment with IWP2, fewer cells migrated and invaded the other side of the chamber membrane in H1299 or 95C or NC groups as compared to their corresponding groups; no significant difference was observed in the DKK1-KD group. Together, these results revealed the key role of β-catenin in DKK1-induced EMT, migration, and invasion in NSCLC cells. Representative images are shown in Figure 4. Each result was analyzed based on at least three independent experiments, and the data are represented as mean ± SEM (**p < 0.01).

Figure 4.

DKK1 promoted NSCLC cells EMT, migration, and invasion via β-catenin signaling pathway. Cells were treated with 10 µm IWP2 for 24 h after adherence and then harvested for (a and b) Western blot analysis. Results demonstrated that IWP2, the inhibitor of Wnt signaling pathway, decreased the expression of β-catenin significantly and inhibited EMT-related proteins in H1299/95C and NC groups, except in DKK1-KD groups. Cells treated as in (a and b) were used for (c and d) migration transwell assay and (e and f) invasion transwell assay with Matrigel. The results revealed that IWP2 inhibited the migration and invasion of H1299/95C cells, as well as in NC groups as compared to the corresponding IWP2-free groups, whereas DKK1-KD groups showed no obvious difference. Each analysis was performed in triplicate and the data were represented as mean ± SEM (**p < 0.01).

DKK1 regulated the expression of β-catenin by inhibiting its phosphorylation

LiCl, the inhibitor of GSK3β, inhibits the phosphorylation of β-catenin. ²¹ As shown in Figure 5(a) and (b), LiCl regained the expression of β-catenin, especially in the DKK1-KD cell group. In addition, after treatment with LiCl, the expression of ZEB1 and Snail increased as compared to the corresponding cells without LiCl, especially in the DKK1-KD cell group. Migration (Figure 5(c) and (d)) and invasion (Figure 5(e) and (f)) results demonstrated that after using LiCl, more cells migrated and invaded the other side of chamber membrane, especially in the DKK1-KD group. These findings reconfirmed the key role of DKK1 in regulating the expression of β-catenin by inhibiting its phosphorylation. Each result was analyzed in at least three separate experiments, and data are represented as mean ± SEM (**p < 0.01).

Figure 5.

DKK1 regulated the expression of β-catenin by inhibiting its phosphorylation. Cells were treated with 5 mM LiCl for 24 h after adherence and then harvested for (a and b) Western blot analysis. Results demonstrated that LiCl, the inhibitor of GSK3β, recovered the expression of β-catenin and also rescued the inhibition of sh-DKK1 on EMT in DKK1-KD groups. Cells treated as in (a and b) were used for (c and d) migration and (e and f) invasion transwell assays. The results revealed that LiCl rescued the inhibition of sh-DKK1 on cell migration and invasion in DKK1-KD group cells as compared to the corresponding cells without LiCl treatment. Each analysis was performed in triplicate and the data were represented as mean ± SEM (**p < 0.01).

DKK1 induced the nuclear localization of β-catenin

Nuclear localization of β-catenin may decrease due to phosphorylation. To investigate whether DKK1 exerts its effects on nuclear β-catenin, we performed cell immunofluorescence analysis and found that β-catenin is localized in the cytoplasm, nucleus, and cytomembrane. Furthermore, we also observed that nuclear β-catenin was significantly decreased in cells transfected with DKK1-shRNA than the controls. Representative images are shown in Figure 6(a) and (b). Each result was analyzed in at least three independent experiments, and data are represented as mean ± SEM (**p < 0.01).

Figure 6.

DKK1 induced the nuclear localization of β-catenin. (a and b) Cells with sh-NC or sh-DKK1 were stained with anti-β-catenin antibodies followed by a rhodamine (TRITC)-conjugated AffiniPure goat anti-rabbit IgG (red for β-catenin). The nuclei were visualized by Hoechst 33342 (blue). The results indicated that nuclear β-catenin was significantly decreased in cells transfected with DKK1-shRNA than controls (scale bar: 25 µm). The proposed mechanism of DKK1 in (c) NSCLC cells suggested that DKK1 inhibited the phosphorylation of β-catenin, inducing its nuclear expression and promoting the occurrence of EMT.

Discussion

The abnormal expression of DKK1 is closely related with tumorigenesis in several cancers.^22,23 In the case of NSCLC, although Xu et al. ¹⁶ considered DKK1 as a suppressor, most studies support that DKK1 serves as an oncogene.^18,19 And in our study, we found that the downregulation of DKK1 inhibited cell migration and invasion, affirming the oncogenic role of DKK1 in NSCLC cells. EMT is a complex process known as the transdifferentiation of epithelial cells into mesenchymal cells, which plays an essential role in embryonic development and tumor progression. ²⁴ Cancer cells enhance their migration and invasion through EMT. ²⁵ For example, Zhou et al. ²⁶ reviewed that the process of EMT is involved in tumor progression, metastasis, and treatment resistance in pancreatic cancer. Some researchers have investigated that DKK1 and Wnt signaling pathway are correlated with EMT occurrence. For instance, Yao et al. ²⁷ reported that DKK1 induced EMT in NSCLC, promoting vasculogenic mimicry formation. In our study, we found that knockdown of DKK1 decreased the expression of EMT-related proteins, ZEB1 and Snail, in NSCLC cells. Combined with the result of DKK1-mediated migration and invasion alteration, all these results indicated the significance of DKK1 to EMT occurrence and migration and invasion.

β-catenin is known as the key component of Wnt signaling pathway, and Masszi et al. ²⁸ manifested that β-catenin was related to EMT occurrence. Moreover, Gonzalez and Medici ²⁹ summarized the signaling mechanisms of EMT and mentioned that stabilization of β-catenin could promote its occurrence during embryonic development. In this study, we found that downregulation of DKK1 promotes the phosphorylation of β-catenin, resulting in the degradation of β-catenin. Combined with Western blot analysis results (Figure 3), we predicted that β-catenin might play a vital role in the occurrence of EMT, cell migration, and invasion. IWP2 inhibits the generation of Wnt by acting on Porcupine (Porcn), inducing the degradation of β-catenin. Hence, we employed IWP2 to authenticate the role of β-catenin. The results indicated that IWP2 inhibited the DKK1-mediated EMT, migration, and invasion in H1299 and 95C cells as compared to the corresponding controls without IWP2 treatment, revealing the significance of β-catenin.

Western blot analysis in Figure 3 showed that the knockdown of DKK1 promoted the phosphorylation of β-catenin, indicating that DKK1 may regulate the expression of β-catenin by inhibiting its phosphorylation. GSK3β is known to localize upstream of β-catenin, promoting its phosphorylation. ³⁰ Hence, to confirm the prediction, we utilized LiCl to inhibit GSK3β-mediated phosphorylation of β-catenin and regain its expression. In addition, LiCl rescued the occurrence of EMT, migration, and invasion as compared to the corresponding LiCl-free groups, especially the DKK1-KD group. These findings indicated that DKK1 inhibited the phosphorylation of β-catenin, inducing EMT and promoting migration and invasion of NSCLC cells.

β-catenin is found at multiple subcellular localizations, including cell–cell junctions, cytoplasm, and the nucleus. ³¹ The phosphorylation of β-catenin may influence its nuclear localization and has also been shown to be involved in the transcriptional regulation of downstream target genes. ³¹ Bernaudo et al. ³² reported that decreased nucleus β-catenin could induce the anti-tumor effect in epithelial ovarian cancer cells. Hence, we examined whether DKK1 affected the nuclear β-catenin. In the immunofluorescence analysis, we found that the knockdown of DKK1 decreased the nuclear localization of β-catenin, thereby suggesting that DKK1 influenced the nuclear localization of β-catenin and regulated EMT, cell migration, and invasion. However, the mechanism underlying the effect of nuclear β-catenin on mesenchymal transition, cell migration, and invasion remains to be elucidated.

In this study, we demonstrated that the downregulation of DKK1 inhibited the migration and invasion of NSCLC cells. This phenomenon suggests the potential therapeutic value of DKK1 on advanced NSCLC. Our findings supplement the molecular mechanism of DKK1 (Figure 6(c)). Specifically, DKK1 inhibits the phosphorylation of β-catenin, which reduces the degradation of β-catenin and elevates the nuclear expression. Subsequently, the nuclear β-catenin regulates the EMT-related gene expression, resulting in mesenchymal transition.