Overexpression of RNF146 in Non-Small Cell Lung Cancer Enhances Proliferation and Invasion of Tumors through the Wnt/β-catenin Signaling Pathway

Studies have suggested a possible correlation between the newly identified E3 ubiquitin ligase ring finger protein 146 (RNF146) and tumor development. However, until now, studies on RNF146 have been restricted to poly(ADP-ribosyl)ation and ubiquitin ligation, whereas the role of RNF146 in tumor biology has rarely been reported. In the present study, the role of RNF146 in non-small cell lung cancer (NSCLC) was investigated. The results showed that the expression of RNF146 was increased in clinical lung cancer samples and cell lines. RNF146 expression correlated with tumor size, differentiation level, lymphatic metastasis, pTNM staging, and prognosis of patients in stage I. RNF146 expression was negatively correlated with Axin expression but positively correlated with the nuclear expression of β-catenin in NSCLC tissues. RNF146 downregulated the expression of Axin in lung cancer cell lines and induced the expression and nuclear distribution of β-catenin. Overexpression of RNF146 in NSCLC cell lines increased the levels of cyclinD1, cyclinE, and CDK4, promoted cell cycle G0/G1-S transitions, and regulated cell proliferation. Overexpression of RNF146 led to upregulated levels of matrix metalloproteinases 2 and 7 and enhanced lung cancer cell invasiveness, events that were mediated by the classical Wnt/β-catenin signaling pathway. In summary, the data in the present study indicate that RNF146 regulated the development and progression of NSCLC by enhancing cell growth, invasion, and survival, suggesting that RNF146 may be a potential treatment target in NSCLC.


Introduction
E3 ubiquitin ligases play important roles in regulating cell functions including proliferation, cell cycle arrest, and apoptosis. They may also have additional functions that depend on the identity of their substrates. For example, if an E3 ubiquitin ligase targets an oncogene for degradation, it may be considered a tumor suppressor. Similarly, if an E3 uniquitin ligase degrades a tumor suppressor protein, it may be considered an oncogene. Many proteins containing RING-finger domains possess ubiquitin ligase activity, some of which participate in tumorigenesis and tumor metastasis. The newly identified E3 ubiquitin ligase RING finger protein 146 (RNF146) interacts with poly(ADP-ribose) (PAR) through a PAR-binding motif in the Trp-Trp-Glu (WWE) domain. The RNF146 gene is located on human chromosome 6q22, 33 [1]. RNF146 has neuroprotective activity due to its inhibition of Parthanatos via binding with PAR [2]. RNF146 may facilitate DNA repair against cell death induced by DNAdamaging agents or c-irradiation [3]. In response to cellular damage, RNF146 translocates to the nucleus and enhances the ubiquitination and degradation of various nucleoproteins that participate in DNA damage repair. In addition, as a poly(ADPribosyl)ation (PARsylation)-directed E3 ubiquitin ligase, RNF146 regulates the Tankyrase-dependent degradation of Axin and positively regulates the Wnt signaling pathway [1].
The Wnt signaling pathway is highly active in lung cancer cells, leading to metastasis and proliferation of these cells [4]. Wnt signaling can be aberrantly activated by various mechanisms, and a main function is to inhibit the proteolysis of b-catenin, which is controlled by phosphorylation [5]. Free b-catenin can enter the nucleus and activate the target genes of Wnt. Steady-state levels of Axin are very important, as this scaffolding protein initiates formation of the b-catenin degradation complex. Researchers have demonstrated that the transfer of PAR to residues of Axin catalyzed by Tankyrase leads to the PARsylation of Axin [1,6]. RNF146 participates in the degradation of PARsylated Axin through its PAR-binding motif. This interaction leads to destruction of the b-catenin degradation complex, aggregation of intracellular b-catenin, and increased signaling through the Wnt pathway [1]. Despite many studies on RNF146, its exact role in tumorigenesis remains unclear. In the present study, the roles of RNF146 in lung cancer were investigated.

NSCLC Tissue Samples
Primary NSCLC samples and control tissues were collected from 133 patients. Normal lung samples were taken from lung tissue more than 5 cm from the cancer resection site. Procedures took place at the Fourth Affiliated Hospital of China Medical University. The patients did not receive any radiation or chemotherapy before the operation. NSCLC staging was based on the TNM Classification of Malignant Tumors, Seventh Edition [7]. The survival time was calculated from the operation day to death via the evaluation of recurrence and metastasis or until the last follow-up date. Fresh specimens were frozen in liquid nitrogen and stored at 280uC. For experiments involving human tissues, approval was obtained from the institutional review committee of China Medical University. Written informed consent was provided according to the Declaration of Helsinki.

Cell Lines and Cell Culture
The normal bronchial epithelial cell line HBE, the human lung adenocarcinoma cell lines A549 and H1299 were purchased from the American Type Culture Collection (Manassas, VA, USA
Two pathology physicians assessed the staining results on a semiquantitative scale. Five high magnification fields were randomly selected from each section, and 100 tumor cells in each field were counted to evaluate the intensity and range of RNF146 staining. A negative result was given a score of 0 (no staining), faint yellow staining was recorded as a 1, yellow was scored as a 2, and deep brown staining was recorded as a 3. Percentage scores were assigned as follows: ,5%, 0; 5-25%, 1; 26-50%, 2; 51-75%, 3; $75%, 4. Two independent scores were multiplied together to obtain the patient's coloring coefficient, which was categorized as ''low expression'' for coloring coefficient ,4 and ''high expression'' for coloring coefficient $4. Axin and b-catenin were scored as previously described [8][9].

Cell Transfection
Cells were plated on 24-well plates at approximately 50% confluence and transfected with plasmid or siRNA 24 hours later. Transfection of lung cancer cell lines A549 and LTE was performed using Lipofectamine 2000 (Invitrogen) according to the manufacturer's directions. For co-transfection experiments, 1 to 1.5 mg of co-transfected or empty vector and 45 pmol siRNA were added to the transfection mixture. After 5 h at 37uC and 5% CO 2 , complete medium was added to the transfection mixture. Transfection efficiency was determined by Western blot.

Western Blot
Cells were lysed in buffer containing 150 mM NaCl, 1% NP-40, 0.1% SDS, 2 mg/ml aprotinin, and 1 mM PMSF. Total protein was harvested, and 60 mg were separated by SDS-PAGE and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, USA). Proteins on PVDF membranes were blocked with 5% bovine serum albumin (BSA) in phosphate buffered saline with tween (PBST), incubated with primary antibody at 4uC overnight, and then incubated with secondary antibodies at room temperature for 2 hours. Membranes were developed with enhanced chemical liquid (ECL) (Thermo Fisher Scientific), and the results were recorded using a bioimaging system (UVP Inc., Upland, CA, USA). Protein expression levels were evaluated using b-actin as a loading control.

MTT Assay
Transfected cells and control cells were plated on 96-well plates at 10 3 cells per well and cultured for 24 hours. Cell proliferation was detected with the cell counting Kit-8 solution (Dojindo, Gaithersburg, MD) according to the manufacturer's instructions. The 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was repeated on 4 consecutive days. Optical density values were measured using a microplate reader (Spectra Thermo, Mä nnedorf, Switzerland) at an absorbance of 550 nm.

Wound-healing Assay
Each well of a 6-well plate was seeded with 5610 5 cells, cultured for 24 hours, and scratched with a spearhead. Cells were washed with PBS and cultured in serum-free medium at 37uC and 5% CO 2 . Samples were harvested and photographs were taken at 0, 6, 12 and 24 hours. Each experiment was repeated at least three times and average values were calculated.

Transwell
Cell suspensions of 100 ml containing 2.5610 6 cells/ml were added to the upper chamber of a transwell insert, and 600 ml medium containing 10% fetal bovine serum was placed in the lower chamber. A polycarbonate microporous membrane with a diameter of 8 mm separated the upper and the lower chambers. Cells were cultured at 37uC and 5% CO 2 for 6 hours. After washing with PBS, cells were fixed with methanol at room temperature for 15 minutes. Finally, the treated cells were stained with hematoxylin, photographs were taken, and stained cells were counted. To measure the invasive capacity of cells, precooled medium without serum and Matrigel (BD Biosciences, USA) were mixed at a volume ratio of 1:7 and were added into the upper chamber. The volume of the mixture liquid was 100 ml. The cells were treated for 3 hours at room temperature, and the results were obtained after the cells were cultured for 24 hours.

Cell Cycle Assay
Cells were seeded at 5610 6 cells per 6 cm culture dish, cultured for 12 hours, and transfected. Cells were synchronized by serum starvation for 20 hours. Cells were maintained in medium containing 10% serum for 24 hours. Cells were harvested and fixed in cold 70% ethanol. After washing with PBS, cells were incubated with 0.1 mg/ml RNaseA (Roche Indianapolis, IN) at 37uC for 30 minutes and treated with 5 mg/ml propidium iodide for 30 minutes. Finally, the treated cells were analyzed by flow cytometry.

Statistical Analysis
Databases were set up using Excel. Data analysis was conducted using SPSS17.0. Relationships between the expression of RNF146, Axin, b-catenin, and clinical and pathological factors were evaluated by the x 2 test and Fisher test. Kaplan-Meier survival curves and Log-rank were used to detect and evaluate differences. Other data were compared by t-test. Data from three independent experiments were shown as x6s. A P,0.05 was considered statistically significant.

RNF146 Overexpression in NSCLC
To investigate the expression of RNF146 in NSCLC, 20 paired NSCLC and non-cancerous tissue samples were collected, and RT-PCR and Western blot were used to analyze the RNF146 mRNA and protein expression levels, respectively. Expression of RNF146 mRNA in the NSCLC samples was higher than that of the non-cancerous lung tissues (P = 0.030) ( Figure 1A). The results of the Western blot and immunohistochemical staining confirmed that RNF146 protein expression was increased in NSCLC (P = 0.014, P = 0.000) ( Figure 1B), consistent with the RT-PCR results. The expression of RNF146 was also analyzed in the HBE and NSCLC cell lines. Different expression levels of RNF146 were observed among the NSCLC cell lines ( Figure 1C). The expression of RNF146 mRNA and protein in both the HBE and the LK2 (squamous) cell lines were low. The expression of RNF146 protein in the LTE (adenocarcinoma) and LK2 (squamous) cell lines were moderate, and the expression of RNF146 protein in the A549 and H1299 (adenocarcinoma) cell lines were high. The data indicated that expression of RNF146 was increased in the clinical NSCLC samples and lung cancer cell lines.
To explore the role of RNF146 in NSCLC further, the relationships between protein expression and clinical pathological characteristics were analyzed. There were correlations between the expression of RNF146 and tumor size (P = 0.044), histological type (P = 0.005), poor differentiation (P = 0.002), lymphatic metastasis (P = 0.009), and pTNM stage (P = 0.015). However, RNF146 expression was not correlated with the age or gender of NSCLC patients (Table 1, Figure 1D). Furthermore, Kaplan-Meier survival analysis demonstrated that overexpression of RNF146 was correlated with the survival of lung cancer patients in stage I (P = 0.016). There was no correlation between RNF146 expression and survival time in all patients (P = 0.616) or those patients with stage II (P = 0.437), or stage III NSCLC (P = 0.520) ( Figure 1E).

Expression of RNF146, Axin, and Nuclear Distribution of b-catenin
As an E3 ubiquitin ligase, RNF146 regulates the Tankyrasedependent degradation of Axin and positively regulates the Wnt signaling pathway [1]. To investigate the relationship between RNF146, Axin, and b-catenin, the expressions of Axin and bcatenin in 133 NSCLC samples were detected by immunohistochemical methods and correlations of the proteins were calculated. RNF146 was negatively correlated with Axin expression (P = 0.003), but it was not correlated with the decreased expression of b-catenin in membranes or the increased expression of bcatenin in the cytoplasm (P = 0.524). High RNF146 expression was positively correlated with nuclear b-catenin staining (P = 0.047) (Table 2, Figure 2A). LTE cell lines were transiently transfected with the pEX4-RNF146 overexpression plasmid or the control pEX4 empty plasmid. Experiments confirmed that RNF146 was upregulated. A549 and H1299 cell lines were transiently transfected with RNF146-shRNA1, RNF146-shRNA2 or control shNC, and experiments confirmed that RNF146 expression was reduced ( Figure 2C and Figure S1A). Furthermore, expression levels of Axin and b-catenin were analyzed by RT-PCR, Western blot, and immunofluorescence. Overexpression or knock-down of RNF146 did not affect the mRNA expression levels of Axin and b-catenin ( Figure. 2B). However, Western blot results indicated that the overexpression of RNF146 in LTE lung cancer cells inhibited the protein expression of Axin and increased the expression of b-

Expression of RNF146 and the Proliferation of Lung Cancer Cells
To explore the roles of RNF146 in cell proliferation, LTE and A549 cells were transfected with plasmids, as described above, and cell proliferation was detected by the MTT assay. Compared with the control group and the group transfected with the empty plasmid, overexpression of RNF146 significantly enhanced cell proliferation in LTE cells, whereas knockdown of RNF146 inhibited cell proliferation in A549 and H1299 cells ( Figure 3A and Figure S1B). The cell cycle was analyzed by flow cytometry, and the results demonstrated that overexpression of RNF146 decreased the percentage of G 0 /G 1 phase LTE cells and increased the percentage of S phase cells. In contrast, knockdown of RNF146 in A549 and H1299 cells increased the percentage of cells in the G 0 /G 1 phase and decreased the percentage of cells in S stage ( Figure 3B and Figure S1C).
Western blot results also revealed correlations between the levels of RNF146 in lung cancer cells and the expression of cell cyclerelated regulatory proteins including cyclinD1, cyclinE and CDK4 ( Figure 3C). The data indicated that RNF146 regulated the expression of cyclinD1 and cyclinE and regulated cell cycle progression by inducing the G 0 /G 1 -S transition.

RNF146 Expression and Invasion of Lung Cancer Cells
Overexpression and knockdown of RNF146 affected the cell migration and invasion abilities of NSCLC cells. Wound-healing and transwell experiments demonstrated that overexpression of RNF146 enhanced cell migration and invasion in LTE cells, and that knockdown of RNF146 decreased cell migration and invasion in A549 cells ( Figure 4A, B, C). Similar results were obtained in H1299 cells in which RNF146 expression was depleted ( Figure S2, S3). Proteins related to migration and invasion were also detected by Western blot. The expression levels of MMP2 and MMP7 were increased in LTE cells that overexpressed RNF146, whereas the levels of RhoA, RhoB, RhoC, Rock1, and TIMP-1 were unchanged. Furthermore, levels of MMP2 and MMP7 were decreased significantly when RNF146 expression was knocked down ( Figure 4D). alveolar epithelium. D2, RNF146 was expressed in the cytoplasm of lung cancer cells, and weak RNF146 expression was detected in the adjacent bronchial epithelium. D3 and D5 showed weak RNF146 staining in the well-differentiated squamous cell carcinoma and adenocarcinoma. D4 and D6 showed strong RNF146 staining in the poorly differentiated squamous cell carcinoma and adenocarcinoma. (E) Relationship between RNF146 expression and the prognosis of NSCLC patients. E1 showed correlation between RNF146 expression and prognosis of stage I patients (P,0.05). E2 and E3, no correlation between RNF146 expression and prognosis of stage II and stage III patients (P.0.05

RNF146-regulated Expression of cyclinD1 and MMP7
We proposed that RNF146 regulated cell migration and invasion by indirectly regulating target-gene transcription through Axin/b-catenin signaling. To test this hypothesis, we knocked down RNF146 in cells treated with siRNA targeting Axin. We found that the levels of cyclinD1, cyclinE, and MMP7 recovered as compared to cells treated with shRNF146 alone ( Figure 5A). Overexpression of RNF146 in combination with siRNA knockdown of b-catenin did not cause levels of cyclinD1, cyclinE, or MMP7 to increase, whereas the MMP2 levels were still increased ( Figure 5B). Knockdown of endogenous TCF-4 in LTE cells combined with overexpression RNF146 did not cause levels of cyclinD1 and MMP7 to increase ( Figure 5C). However, when RNF146 and TCF-4 were knocked down in A549 cells simultaneously, expression of cyclinD1 and MMP7 decreased more dramatically ( Figure 5D). Taken together, the data indicated that RNF146 probably regulated expression of cyclinD1 and MMP7 by the Wnt/b-catenin signaling pathway.

Discussion
In the present study, we found that RNF146 was overexpressed in NSCLC tissue compared to adjacent normal lung tissues. The RNF146 expression level showed correlations with several clinical pathological factors, including tumor size, differentiation level, lymphatic metastasis, pTNM staging, and the prognosis of patients in stage I, which are important factors that represent the potential for lung cancer malignancy. These data suggest that overexpression of RNF146 in NSCLC may enhance metastasis in lung cancer, and that RNF146 might be an indicator of poor prognosis in stage I NSCLC.
Our data are consistent with the findings of two recent reports [10][11] that RNF146 might be involved in the development of tumors. Gold et al. discovered a breast cancer risk gene in the Jewish Ashkenazi population located on 6q 22.33 in a genomewide association study. This location contained two genes, RNF146 and enoyl CoA hydratase domain containing 1 [11]. Another important discovery was that RNF146 interacted with PARsylated Axin through its PAR-binding motif, leading to Axin degradation and positive regulation of the Wnt signaling pathway. These results provided new evidence for the potential role of RNF146 in tumor development [1,6]. It has been indicated that the newly identified RNF146 substrates basic leucine zipper nuclear factor 1 (BLZF1) and cancer susceptibility candidate 3 (CASC3) are involved in tumor development and cell proliferation [12][13]. These reports suggest that RNF146 might play important roles in tumors by altering gene expression or degrading target proteins.
The Wnt signaling pathway affects gene expression and cell migration. As an important molecule in the Wnt signaling pathway, Axin participates in formation of the Axin/glycogen synthase kinase 3b (GSK-3b)/adenomatous polyposis coli (APC)/ Casein kinase I (CKI) degradation complex, which enhances the degradation of b-catenin and inhibits tumor proliferation, invasion, and metastasis [14][15]. Previous studies have established that Axin plays important roles in NSCLC [14]. Axin expression has been shown to inhibit the proliferation and invasion of lung cancer cells [9]. Cytoplasmic stabilization and accumulation of nuclear b-catenin, an important signal transducer, are important processes in Wnt signal transduction.
We found that RNF146 was negatively correlated with the expression of Axin and positively correlated with nuclear expression of b-catenin. Immunofluorescent staining demonstrated that RNF146 regulated the distribution and nuclear accumulation of b-catenin. Previous studies have demonstrated that nuclear bcatenin expression leads to the loss of cell adhesion. Moreover, this protein increased the transcription of target genes, including cmyc, cyclinD1, VEGF, and MMP7, by interacting with the T cell factor/lymphoid enhancer factor, thereby leading to the proliferation and metastasis of tumor cells [16][17]. In the present study, RNF146 regulated Axin and b-catenin of the Wnt signaling pathway, indicating that RNF146 affected the proliferation and invasion of tumor cells.
Our MTT and cell cycle experiments showed that RNF146 not only regulated cell proliferation and cell cycle progression but also affected the expression levels of proteins related to the cell cycle. Cyclin D1 mainly regulates G 1 phase progression [18], and cyclin E enhances G 1 /S transition [19][20]. Cyclin D1 could interact with CDK4/6 and enhance the transcription of downstream genes [18]. Thus, it can be concluded that RNF146 controls tumor proliferation by regulating the cell cycle.
Our data also showed that RNF146 affected the migration and invasion of lung cancer cells, suggesting that RNF146 has multiple functions. The scratch test and transwell assay results showed that RNF146 enhanced the migration and invasion of lung cancer cells. MMPs degrade various proteins in the extracellular matrix, disrupt tissue barriers to tumor cell invasion, and play key roles in the invasion and metastasis of tumors [21]. The data in the present study revealed that MMP2 and MMP7 were regulated by RNF146. Degradation of the extracellular matrix and bone matrix by MMP2 and MMP7 has been shown to be closely related to NSCLC invasion and metastasis [22][23]. The data suggest that RNF146 might regulate the migration and invasion of lung cancers by regulating MMP2 and MMP7.
Prior to our study, the roles of RNF146 in the Wnt signaling pathway were confined to the degradation of Axin and the RNF146 in adenocarcinoma (A1). Low Axin expression and cytoplasmic and nuclear-expression of b-catenin was found in squamous cancers with high expression of RNF146 (A2 and A3). Expression and distribution of RNF146, Axin and b-catenin detected by (B) RT-PCR, (C) Western blot, and (D) immunofluorescence. Cells transfected with empty vectors served as controls. doi:10.1371/journal.pone.0085377.g002   by Axin and b-catenin, indicating that MMP2 may not be a direct target of the Wnt signaling pathway, and RNF146 may regulate MMP2 by other mechanisms. The data demonstrated that RNF146 probably regulated cyclinD1 and MMP7 by the Wnt/ b-catenin signaling pathway. However some caveats need to be considered in the potential mechanism of RNF146 in Wnt signaling pathway. First, Wnt is a very complex process and lots of proteins are involved in Wnt regulation. Wnt signaling is regulated by altered expression of various regulatory proteins in lung cancer and RNF146 may be the only one of those regulators. So, further studies are required to verify the potential mechanism of RNF146 in Wnt signaling pathway. Second, RNF146 was confirmed to be the novel E3 ubiquitin ligase [1], however whether the biologic function of RNF146 in tumors is dependent or independent on ubiquitin ligase activity remains unknown. This potential link needs to be fully investigated.
In summary, we observed high expression levels of RNF146 in clinical NSCLC samples, consistent with the finding that RNF146 overexpression enhanced the proliferation and invasion of lung cancer cells. In addition to regulating the growth and survival of tumor cells, RNF146 played tumor-enhancing roles and affected the migration and invasion of tumor cells in NSCLC. These effects probably depended on key factors, including Axin and b-catenin in the Wnt signaling pathway. The results reveal new roles for RNF146 in the development and progression of NSCLC and identify RNF146 as a potential target for lung cancer treatment.