Knockdown of SIPA1L3 Inhibits the Proliferation and Invasion of Non–Small Cell Lung Cancer Cells Through the Hippo Pathway

Background: SIPA1L3 is a member of the signal-induced proliferation-associated (SIPA) protein family, only limited data about the SIPA1L3 are currently available in human carcinoma. Our present study provides the expression pattern and function of SIPA1L3 in non-small cell lung cancer (NSCLC). Methods: We performed immunohistochemistry to detect the distribution of SIPA1L3 in NSCLC specimens and analyzed the relationship between SIPA1L3 expression and patients clinicopathological feature. We used small interfering RNA (siRNA) to specically silence SIPA1L3, then investigated its effect on cell growth and invasion in human lung cancer cell lines. Western blot and immunoprecipitation were performed to show the interaction of SIPA1L3 with the core proteins of Hippo pathway, and the Hippo pathway activity. Results: Immunohistochemical staining showed that SIPA1L3 was cytoplasmic increasing in 147 of 217 cases. High levels of SIPA1L3 expression were associated with poor differentiation and a high tumor node metastasis stage, positive lymph nodal status, and poor prognosis. Downregulation of SIPA1L3 inhibited cell EMT, growth, and invasion. As well as SIPA1L3 inhibited Hippo pathway. SIPA1L3 interacted with AMOT, which inhibited AMOT binding to Pals, then decreased nuclear location of YAP. Conclusions: SIPA1L3 overexpression may be a marker for advanced NSCLC. SIPA1L3 reduction inhibits the proliferative and invasive ability of the lung cancer cells, which involves the Hippo pathway by contacting with AMOT.


Introduction
Of all the malignant tumors, lung cancer currently has the highest incidence and mortality, with 85% to 90% of these cancers being non-small cell lung cancer (NSCLC) [1][2][3]. A variety of factors critical for cell adhesion, cell polarity and cellular structure are mutated in human lung cancer, however not all are currently known.
SIPA1L3 is a member of the signal-induced proliferation-associated (SIPA) protein family, which was also known as the part of the Rap GTPase activating protein (RapGAPs) superfamily. The SIPA protein family comprises four members, SIPA1, SIPA1L1, SIPA1L2, and SIPA1L3 [4][5][6]. All SIPA proteins are expressed in the central nervous system, and basically most of the studies focused on their function within the brain.
Beside it, SIPA1L3 expression has been described in the embryonic lens and mutations of the human SIPA1L3 gene result in congenital cataracts [7][8][9]. These results showed that SIPA1L3 has an important role in epithelial cell morphogenesis, establishing or maintaining cell polarity, cell adhesion, and cytoskeletal organization. They also indicated that SIPA1L3 is a novel molecular factor contributing to epithelial cellular function, and SIPA1L3 is a causative disease gene. However, the functional role of Sipa1l3 in human carcinoma, including NSCLC, remain largely unclear.
All SIPA family members share common domains, namely an N-terminal Rap GAP domain (Rap-GTPase activating domain), a PDZ domain and a C-terminal coiled-coil domain carrying a leucin-zipper motif. SIPA1L1-SIPA1L3 contains an additional domain of unknown function (DUF3401) [6][7][8]. Base on its functional structure, we hypothesized that SIPA1L3 play an important functional role in lung cancer development. In this study, we investigated whether SIPA1L3 has the abnormal expression in NSCLCs, and its in uence on cell proliferation and invasion. As well as the underlying molecular pathways involving in the progression of NSCLC.

Immunohistochemistry Staining
Immunohistochemistry assays were performed as described previously [10]. Brie y, tissue sections were incubated with normal goat serum to block nonspeci c antibody binding and then with mouse monoclonal anti-SIPA1L3. (1:100; Cell Signaling Technology (CST) Inc., Danvers, Massachusetts).
Positive SIPA1L3 expression was determined by the intensity of cytoplasm staining of tumor cells and the proportion of SIPA1L3-positive cells in the tissue sections. The immunostaining assessment scale as follows: negative (no staining), weak (light yellow staining, no clear granular or yellow staining, and clear granular staining area <10%), moderate (yellow staining and clear granular staining area ≥10% or brown staining and clear granular staining area <10%), or strong (brown staining and clear granular staining area ≥10%). Negative to weak were considered to be normal SIPA1L3 expression and moderate to strong were considered to be SIPA1L3 overexpression. Two investigators blinded to the clinical data independently and randomly examined all tumor slides.

Cell lines and transfection
The human bronchial epithelial cells (HBEs), a human adenocarcinoma cell line (A549, LTEP-a-2, H1299), and an SCC cell line (LK-2, SK-MES-1) were purchased from ATCC (Manassas, VA, USA). The SIPA1L3-siRNA and the negative controls were purchased from Genechem Biological Technology Co. Ltd. (Shanghai, China). The cells were transfected with plasmids using Attractene transfection reagent (Qiagen, Hilden, Germany) according to the manufacturer's instructions.

Matrigel Invasion Assay
The cell invasion assay was performed in a 24-well Transwell chamber with an 8-μm pore size (Costar, Cambridge, MA). Inserts were coated with 20 μL Matrigel (1:3 dilution; BD Bioscience). Forty-eight hours after transfection, 3×10 5 cells in 100 μL of serum-free medium were transferred to the upper Matrigel chamber and incubated for 16 hours, whereas 600 μL of culture medium supplemented with 10% fetal bovine serum was placed in the lower chamber. After 18 hours of incubation, the noninvaded cells were removed from the upper membrane surface with a cotton swab, and the cells that passed through the lter were xed with 4% paraformaldehyde and stained with hematoxylin. The number of invaded cells was counted in 10 randomly selected high-power elds under a microscope. This experiment was performed in triplicates.

Colony Formation Assay
Cells were plated into 6-cm cell culture dishes (1000 cells per dish) and incubated for 14 days.

Overexpression of SIPA1L3 correlated with malignant phenotype in NSCLC patients
The database from TCGA indicated that lung cancer patients with increased SIPA1L3 showed higher mRNA level (Supplementary Fig.1) and poorer overall survival ( Supplementary Fig. 2). We performed immunohistochemical analysis of 217 NSCLC tissues to testify them. SIPA1L3 was undetectable (−) or weak positive in normal bronchial epithelia or pneumocytes. In 147 of 217 (67.74%) lung cancer specimens, SIPA1L3 showed increased cytoplasmic expression (Fig. 1). To identify the clinical signi cance of SIPA1L3 in NSCLC tissues, we analyzed the correlation between SIPA1L3 overexpression and clinicopathologic parameters. As summarized in Table 1, tumors with increased SIPA1L3 tended to display more malignant phenotypes, such as the lower differentiation (P = 0.016), the greater clinical tumor size (P = 0.003), the higher tumor node metastasis (TNM) stage (P = 0.010), and positive lymphatic metastasis (P = 0.004). In terms of survival, patients with SIPA1L3 overexpression had a poorer overall survival than patients with normal SIPA1L3 expression (P < 0.001; Fig. 2). Western blot analysis was performed in lung cancer tissues and cell lines, which were consistent with the IHC results (Fig. 3). These results suggest that SIPA1L3 acts as a tumor promoter in NSCLC.

Downregulation of SIPA1L3 by siRNA in lung cancer cells inhibited lung cancer cell growth and invasion
After transfection with SIPA1L3-siRNAs in LTEP-a-2 and LK-2 cell lines, decreased SIPA1L3 inhibited the proliferation of LTEP-a-2 and LK-2 cells (Fig. 4A). As shown in Fig. 4B, SIPA1L3 silencing resulted in a signi cant increase in G1-phase cells and a signi cant reduction in S-phase cells. In addition, in the cells with SIPA1L3 knockdown, a signi cant increase was observed in the population of cells undergoing early and late apoptosis (Fig. 4C). As well as SIPA1L3 signi cantly reduced colony numbers and the size of the lung cancer cells (Fig. 4D). These data indicated that SIPA1L3 has the effect on the regulation of the cell cycle and apoptosis, thus regulated the cell growth of lung cancer cells. Furthermore, we observed that decreased expression of SIPA1L3 reduced the cell invasion signi cantly, comparing with the controls (Fig.   4E).
The cell-biological program termed the epithelial-to-mesenchymal transition (EMT) plays an important role in the process of malignant progression [12][13][14]. In the LTEP-a-2 and LK-2 cell lines with SIPA1L3-siRNA, the expression of E-cadherin and ZO-1 was signi cantly up-regulated, while the expression of Ncadherin, vimentin, snail and a-SMA was signi cantly down-regulated. EMT is inhibited in the SIPA1L3deleption lung cancer cells. Thus, SIPA1L3 may contribute to the carcinoma processes by inducing EMT.
3.4. SIPA1L3 depletion is accompanied with Hippo signal pathway activation.
The Hippo signaling pathway was rst discovered and studied in Drosophila melanogaster, and it has been demonstrated to play crucial roles in lung cancer [15][16][17]. We are using immunoblotting to detect the difference in the expression of Hippo pathway-related proteins between the lung cancer cells and the cells with SIPA1L3-siRNA. The expression of p-MST1 and p-LATS1 was upregulated, whereas Cyclin E, CTGF, and the nuclear YAP was downregulated, compared to those in control LTEP-a-2 and LK-2 cells. These results indicated that advanced carcinoma by SIPA1L3 may be closely related to the inhibition of the Hippo pathway.

Combination of SIPA1L3 and AMOT Inhibits the Hippo Pathway
AMOTs is involved in Hippo signaling pathway through interacting with multiple core proteins on this pathway. It formed AMOT-mediated complex, such as AMOT-Pals-Merlin-Patj complex, activating Hippo pathway. In the C-terminal region of AMOT, it composes of the PDZ-binding domain [18][19][20]. Base on SIPA1L3 possessing PDZ domain, we used SIPA1L3 and AMOT antibody for immunoprecipitation, respectively, and the obtained proteins were detected by western blot. We found that SIPA1L3 and AMOT interacted with each other. Furthermore, AMOT binding to Pals increased after interfering with SIPA1L3, and nuclear YAP decreased. In the cell lines with SIPA1L3-cDNA, immunoprecipitation revealed that the binding of AMOT to SIPA1L3 increased but the binding of AMOT to Pals decreased. YAP is upregulated and translocated into the nucleus.

Discussion
Recently, the SIPA1L3 has also been associated with congenital human cataracts [6][7][8]. As expression data of the SIPA1L3 during the embryogenesis and disease, it provides a good basis for further functional studies. Here, we present the expression and signi cance analysis of SIPA1L3 in lung cancer.
Firstly, we examined the subcellular localization of endogenous SIPA1L3 in lung tissues by immunohistochemical staining. The SIPA1L3 protein was weakly detectable or undetectable in normal bronchial epithelium, and 67.74% of examined NSCLC tissues showed elevated cytoplasm expression. Clinically, a direct correlation was observed between upregulated SIPA1L3 and an advanced tumor stage, lymph nodal status, and poor differentiation. This implies the tumor-promoting function of SIPA1L3 as a cytoplasmic protein in NSCLC.
Following the evaluation of SIPA1L3 expression patterns in NSCLC, we explored the role of activated SIPA1L3 in regulating NSCLC cell proliferation and invasion. Knockdown of SIPA1L3 in LTEP-a-2 and LK-2 cells, resulted in a signi cant decrease in the cell proliferation and the invasion capacity of the lung cancer cells.
The Sipa1l3-/-mutant mice showed a failure in the maintenance of epithelial cell properties in the lens. Failure in the maintenance of epithelial cell properties can result in abnormal EMT, which is characterized by the disassembly of cell adhesion, loss of polarity and the acquisition of migratory capacity [8]. We further testi ed that SIPA1L3 prompted EMT program. The increased motility/invasiveness associated with the mesenchymal cell state has linked the EMT program with metastasis, in which cell separation from the primary tumor mass can be considered as the rst step of the invasion-metastasis cascade. Thus, a variety of studies have demonstrated that induction of an EMT program allows carcinoma cells to lose cell-cell junctions, degrade local basement membrane via elevated expression of various matrixdegrading enzymes, and thus support their migration and invasion [21][22][23][24]. Therefore, SIPA1L3 overexpression promotes malignant lung cell growth, invasion, and EMT, which may explain our ndings that SIPA1L3 overexpression was the major de ning characteristic of NSCLC tumors and correlated with several clinicopathological factors.
We further investigated possible mechanisms underlying the role of SIPA1L3 in cell proliferation and invasion next. The Hippo pathway controls cell proliferation, differentiation, invasion, and survival by regulating the main downstream effector molecules Yes-associated protein (YAP). The major YAP regulators are the kinases LATS1/2 and MST1/2, which phosphorylate and inhibit YAP [25][26][27]. After interference with SIPA1L3, we found that p-MST1 and p-LATS1 was upregulated, while the nuclear YAP was decreased. So, SIPA1L3 inhibited the Hippo signal pathway in lung cancer.
An AMOT-dependent complex, which comprised of AMOT, Pals, Merlin, and Patj, functions upstream of the core Hippo pathway kinases LATS1/2 and MST1/2 [18,19]. Based on the predicted domain structure of SIPA1L3, it possesses an N-terminal PDZ domains, which is predicted to be required for protein-protein interactions. While AMOT has the C-terminal PDZ-binding motifs, we use immunocoprecipitation to identify the interaction between SIPA1L3 and AMOT. In accordance with previous ndings, immunocoprecipitation also indicated an interaction between AMOT and Pals. Furthermore, we found that SIPA1L3 inhibited the binding of AMOT to Pals and the phosphorylation of YAP, then increased the nuclear location of YAP. It has been shown that the PDZ-binding motifs of the Amot binds Pals, thus SIPA1L3 and Pals could competitively bind with AMOT PDZ-binding motifs. Knockdown of SIPA1L3 decreased its combination with AMOT, then enhanced the combination of AMOT PDZ-binding motifs with Pals, which prompted the AMOT-dependent complex (AMOT, Pals, Merlin, and Patj). The phosphorylate kinases LATS1/2 and MST1/2 were increased, then inactives YAP through phosphorylation, resulting in YAP degradation. Reduced YAP nuclear localization ultimately prevents the proliferation and invasion of lung cancer cells.

Conclusion
Clinical tumor analysis in this study strongly supports that SIPA1L3 is a tumor oncogene in NSCLC. Our results from cellular experiments demonstrate that SIPA1L3 reduction inhibits cells' proliferative and invasive ability, which involves the Hippo pathway. Although AMOT-medicated complex is one manner of the regulation Hippo pathway by SIPA1L3, but AMOT play multiple functions in its in uence on Hippo signaling. Further studies will be required to determine the other manners of SIPA1L3 regulating the Hippo pathway, which mediated by AMOT. Taken together, although there is a limited understanding of SIPA1L3's role in cancer tumorigenesis, we identi ed SIPA1L3 as a novel candidate gene. It paves the way for improvements in our understanding of lung cancer malignant phenotype development.

Declarations
Ethics approval and consent to participate The present study was conducted with the approval of the China Medical University Local Institutional Review Board (2017025). Informed consent was obtained from all the patients for analysis of their specimens.

Consent for publication
Not applicable.
Availability of data and material Not applicable.