MicroRNA-138 Inhibits Cell Growth, Invasion, and EMT of Non-Small Cell Lung Cancer via SOX4/p53 Feedback Loop.

Many studies have shown that downregulation of miR-138 occurs in a variety of cancers including non-small cell lung cancer (NSCLC). However, the precise mechanisms of miR-138 in NSCLC have not been well clarified. In this study, we investigated the biological functions and molecular mechanisms of miR-138 in NSCLC cell lines, discussing whether it could turn out to be a therapeutic biomarker of NSCLC in the future. In our study, we found that miR-138 is downregulated in NSCLC tissues and cell lines. Moreover, the low level of miR-138 was associated with increased expression of SOX4 in NSCLC tissues and cell lines. Upregulation of miR-138 significantly inhibited proliferation of NSCLC cells. In addition, invasion and EMT of NSCLC cells were suppressed by overexpression of miR-138. However, downregulation of miR-138 promoted cell growth and metastasis of NSCLC cells. Bioinformatics analysis predicted that SOX4 was a potential target gene of miR-138. Next, luciferase reporter assay confirmed that miR-138 could directly target SOX4. Consistent with the effect of miR-138, downregulation of SOX4 by siRNA inhibited proliferation, invasion, and EMT of NSCLC cells. Overexpression of SOX4 in NSCLC cells partially reversed the effect of miR-138 mimic. In addition, decreased SOX4 expression could increase the level of miR-138 via upregulation of p53. Introduction of miR-138 dramatically inhibited growth, invasion, and EMT of NSCLC cells through a SOX4/p53 feedback loop.

INTRODUCTION

One of the most common and lethal malignant tumors worldwide is non-small cell lung cancer (NSCLC), which accounts for about 80% of all lung cancer cases1–3. Although clinical diagnosis and therapeutic strategies have improved, the 5-year survival rate for patients with NSCLC is still less than 20%1–6. Currently, surgery, radiotherapy, chemotherapy, and photodynamic therapy are available treatment strategies for NSCLC, but these therapeutic modalities remain generally unsuccessful7. To provide new insight into the development of new diagnosis and therapeutic strategies, it is very important to understand the precise molecular mechanisms that contribute to the progression and metastasis of NSCLC cells.

SOX4, a 47-kDa protein, belongs to a member of the sex-determining region Y (SRY)-related high-mobility group (HMG)-box (SOX) transcription factor family and is highly conserved in vertebrates, and its clinical importance has attracted more and more attention recently, with many studies indicating that SOX4 may lead to progression of multiple cancers8,9. Upregulation of SOX4 is found in colon, prostate, and bladder cancers as well as in NSCLC10–13. Moreover, overexpression of SOX4 correlated with increased cell proliferation, migration, and metastasis in NSCLC14,15. Its high expression has also been closely related to poor prognosis of patients with NSCLC, which makes it a marker to predict the outcome of patients with NSCLC16.

MicroRNAs (miRNAs) are small (about 22 nucleotides in length) noncoding RNAs17, which can degrade or suppress their translation and regulate a series of cell functions such as proliferation, apoptosis, invasion, and differentiation by binding to complementary sequences in the 3′-untranslated regions (3′-UTRs) of targeted mRNAs18,19. An increasing number of studies have demonstrated that miRNAs are involved in a variety of cancers20. Many miRNAs have been identified to act as tumor suppressors in NSCLC such as miR-59021, miR-187-5p22, miR-18623, and miR-13424. These findings provide a strong basis for the importance of miRNAs in the pathogenesis of NSCLC and emphasize the implications of miRNAs in the diagnosis, therapy, and prognosis of NSCLC.

Currently, miR-138 has attracted much attention because several studies have reported that miR-138 is frequently decreased and functions as a tumor suppressor in colorectal, esophageal, and bladder cancers as well as in NSCLC25–28. It has been shown that miR-138-5p is a tumor suppressor in colorectal cancer, and its effects are exerted at least partially through programmed death ligand 1 (PD-L1) downregulation25. Besides, miR-138 inhibits tumor growth through repression of enhancer of zeste homolog 2 (EZH2) in NSCLC28. Moreover, miR-138 may play a suppressive role in the growth and metastasis of NSCLC cells partly by targeting yes-associated protein 1 (YAP1)29. However, until now, the precise mechanism of miR-138 in NSCLC has remained unclear. In this study, we also demonstrated that the level of miR-138 was frequently downregulated in NSCLC cell lines and tissues, which was consistent with previous studies28,29. Introduction of miR-138 suppressed cell proliferation, invasion, and epithelial–mesenchymal transition (EMT) of NSCLC cells. Furthermore, we found that miR-138 could directly target a novel tumor suppressor gene SOX4 in NSCLC cells. Overexpression of SOX4 reversed the inhibitory effects of the miR-138 mimic on NSCLC cells. In addition, decreased SOX4 expression could increase the level of miR-138 via upregulation of tumor protein 53 (p53). Hence, our results showed the important roles for miR-138 in the pathogenesis of NSCLC and suggested miR-138’s potential application in NSCLC treatment.

MATERIALS AND METHODS

Cell Culture and Human Tissues

One human bronchial epithelial (HBE) cell line and NSCLC cell lines such as H522, H1299, H460, A549, PAa, H1975, and Calu-3 were purchased from the American Type Culture Collection (ATCC; Manassas, VA, USA). These cells were cultured in Roswell Park Memorial Institute (RPMI)-1640 medium (Gibco, Grand Island, NY, USA) containing 10% fetal bovine serum (FBS; Gibco) and 1% penicillin/streptomycin (Gibco) at 37°C in a humidified atmosphere of 5% CO2 on 0.1% gelatin-coated culture flasks (Thermo Fisher Scientific, Carlsbad, CA, USA). Ten pairs of human lung carcinoma and their corresponding adjacent normal tissues were collected from the Cancer Hospital of the Harbin Medical University (Harbin, P.R. China). The specimens were immediately frozen in liquid nitrogen and then stored at −80°C for analysis. Prior informed consent was obtained, and the study protocol was approved by the Ethics Committee of the Cancer Hospital of the Harbin Medical University.

Bioinformatics Analysis for miR-138

The online database TargetScan 6.2 (www.targetscan.org) was accessed to identify potential binding partners for miR-138, and their expression was verified by quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR).

miRNA Transfection

To upregulate or downregulate the level of miR-138 in A549 and H1975 cells, the cells were transfected with the miR-138 mimic and miR-negative control (miR-NC) or the miR-138 inhibitor and miR-negative control of inhibitor (anti-miR-NC) that were purchased from GenePharma (Shanghai, P.R. China). One day before transfection, cells were changed to antibiotic-free medium. After 24 h, cells were transfected with 50 nM miR-138 mimic, miR-NC, miR-138 inhibitor, and anti-miR-NC using Lipofectamine 3000 reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s instructions. The pcDNA3.1 vector, pcDNA-SOX4, small interfering RNA (siRNA) for SOX4 (si-SOX4), siRNA for p53 (si-p53), and siRNA-negative control (si-NC) were synthesized and purified by GenePharma.

qRT-PCR

Total RNA of A549 and H1975 cells was extracted using TRIzol reagent (Invitrogen) for analyzing miRNA and mRNA levels according to the manufacturer’s protocols. For quantification of miR-138, the TaqMan MicroRNA Reverse Transcription Kit and TaqMan miRNA assay (Applied Biosystems, Carlsbad, CA, USA) were used to perform reverse transcription and PCR according to the manufacturer’s instructions. U6 small nuclear RNA (snRNA) was used as the internal control. The gene expressions of SOX4, epithelial cadherin (E-cadherin), neural cadherin (N-cadherin), vimentin, matrix metalloproteinase-2 (MMP-2), and MMP-9 were detected using the SYBR Green PCR Kits (Qiagen, Germany). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) served as an internal control. The following primers were used: SOX4, 5′-GTGAGCGAGATGATCTCGGG-3′ (forward) and 5′-CAGGTTGGAGATGCTGGACTC-3′ (reverse); E-cadherin, 5′-TACACTGCCCAGGAGCCAGA-3′ (forward) and 5′-TGGCACCAGTGTCCGGATTA-3′ (reverse); N-cadherin, 5′-CGAATGGATGAAAGACCCATCC-3′ (forward) and 5′-GGAGCCACTGCCTTCATAGTCAA-3′ (reverse); vimentin, 5′-GCTGAATGACCGCTTCGCCAACT-3′ (forward) and 5′-GCTCCCGCATCTCCTCCTCGTA-3′ (reverse); MMP-2, 5′-CTGCGGTTTTCTCGAATCCA-3′ (forward) and 5′-GGGTATCCATCGCCATGCT-3′ (reverse); MMP-9, 5′-CCCTGGAGACCTGAGAACCA-3′ (forward) and 5′-CCACCCGAGTGTAACCATAGC-3′ (reverse); U6, 5′-CTCGCTTCGGCAGCACA-3′ (forward) and 5′-AACGCTTCACGAATTTGCGT-3′ (reverse); and GAPDH, 5′-GAGTCAACGGATTTGGTCGTATTG-3′ (forward) and 5′-CCTGGAAGATGGTGATGGGATT-3′ (reverse). U6 snRNA and GAPDH mRNA were used to normalize. Each sample was assessed in triplicate.

Enzyme-linked Immunosorbent Assay-Bromodeoxyuridine (ELISA-BrdU) Assay

To investigate the effect of the miR-138 mimic or inhibitor on the cell proliferation of A549 and H1975 cells, ELISA-BrdU assay was used to detect the cell proliferation using the Cell Proliferation ELISA-BrdU Kit (Roche Applied Science, Mannheim, Germany) following the manufacturer’s instruction. Briefly, 5 × 103 cells were seeded in a 96-well plate (Corning, Corning, NY, USA) and allowed to grow overnight in complete RPMI-1640 medium. The medium was then removed, and the cells were transfected with the miR-138 mimic or inhibitor for 24 h at 37°C. After 24 h of incubation, the cells were treated with BrdU labeling solution for the remaining 16 h. The culture medium was then removed, and the cells were fixed and the DNA denatured. The cells were incubated with anti-BrdU-peroxidase (POD) solution for 90 min, followed by washing three times to remove the antibody conjugates. After incubation with a 3,3′,5,5′-tetramethylbenzidine (TMB) substrate for 15 min, absorbance at 405 and 490 nm was measured to determine immune complexes.

Transwell Invasion Assay

Transwell chambers (8-mm pore size; Millipore, Billerica, MA, USA) precoated with Matrigel (BD Biosciences, Franklin Lakes, NJ, USA) were used to determine cell invasion. In brief, 1 × 105 cells in 100 μl of RPMI-1640 containing 1% FBS were seeded in the upper chamber, and 600 μl of RPMI-1640 containing 1% FBS was added to the lower chamber. After 24 h of incubation at 37°C in a 5% CO2 atmosphere, the cells that remained in the upper chamber were removed by cotton swabs, and penetrating cells were fixed in methanol and then stained with 0.1% crystal violet. The membranes were rinsed with 30% glacial acetic acid. The washing solution was read at 540 nm to count the number of cells.

Measurement of MMP-2 and MMP-9 Levels

The levels of MMP-2 and MMP-9 in the culture supernatants were assessed using the human MMP-2 and MMP-9 ELISA Kits (R&D Systems, Minneapolis, MN, USA). Samples were diluted to obtain the appropriate concentrations, and an ELISA was performed following the manufacturer’s protocols. The standard curves were made by supplied standards. The standards and samples were added to the wells. Unbound protein was removed by washing, and conjugate was added. After the addition of substrate to generate a color reaction, the optical density was recorded using an automated ELISA reader at a wavelength of 450 nm. The absorbance at 450 nm was converted to ng/ml for MMP-2 and MMP-9. The minimal detection limit was 0.05 ng/ml for MMP-2 and MMP-9.

Western Blot Analysis

A549 and H1975 cells were washed twice in cold phosphate-buffered saline (PBS) and then lysed in radioimmunoprecipitation assay (RIPA) lysis buffer (Beyotime Institute of Biotechnology, Jiangsu, P.R. China) containing protease inhibitor cocktail (Millipore). The protein concentration of cell lysates was quantified by a bicinchoninic acid (BCA) kit (Beyotime Institute of Biotechnology), and 50 μg of each protein was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 10% gels and then transferred to a polyvinylidene fluoride (PVDF) membrane (Millipore). The membranes were blocked in 5% nonfat milk diluted with Tris-buffered saline + Tween 20 (TBST) at room temperature for 1 h and incubated overnight at 4°C with the primary antibodies anti-SOX4 (1:500; Abcam, Cambridge, UK), anti-E-cadherin, anti-N-cadherin, anti-vimentin, anti-proliferating cell nuclear antigen (PCNA), and anti-p53 (1:1,000; Cell Signaling Technology Inc., Danvers, MA, USA). The membranes were then incubated with a goat anti-rabbit or anti-mouse IgG conjugated to horseradish peroxidase secondary antibody (1:1,000; Cell Signaling Technology Inc.) for 2 h. The proteins were visualized using Enhanced Chemiluminescence (ECL) Plus reagents (Amersham Biosciences Corp., Piscataway, NJ, USA). The density of the bands was measured using the ImageJ software (NIH, Bethesda, MD, USA), and values were normalized to the densitometric values of α-tubulin (1:1,000; Sigma-Aldrich, St. Louis, MO, USA) or β-actin (1:1,000; Cell Signaling Technology Inc.) in each sample.

Luciferase Reporter Assay

A549 and H1975 cells were seeded in 24-well plates and incubated for 24 h before transfection. The pMIR-SOX4-3′UTR wild-type (WT) or mutant plasmid was cotransfected with the miR-138 mimic or miR-NC, and pRL-TK plasmid (Promega, Madison, WI, USA) into A549 and H1975 cells. After transfection for 24 h, luciferase reporter gene assay was implemented using the Dual-Luciferase Reporter Assay system (Promega) following the manufacturer’s protocols. All experiments were performed at least three times.

Statistical Analysis

All statistical analyses were performed using GraphPad Prism 5.0 (GraphPad Software Inc., La Jolla, CA, USA). Data from each group were expressed as mean ± standard error of the mean (SEM) and statistically analyzed by Student’s t-test. Differences were considered statistically significant at a value of p < 0.05.

RESULTS

The Level of miR-138 Is Downregulated in NSCLC Cell Lines and Tissues

It has been reported that miR-138 was downregulated in multiple cancers25–28, including NSCLC. In this study, the level of miR-138 was detected by qRT-PCR in a human bronchial epithelial cell (HBE) and seven NSCLC cell lines including H522, H1299, H460, A549, PAa, H1975, and Calu-3. Our results showed that the level of miR-138 was significantly downregulated in these seven NSCLC cell lines compared to that in the human bronchial epithelial cell line HBE (p < 0.05) (Fig. 1A). Furthermore, we found that the level of miR-138 in human NSCLC tissues was significantly lower in comparison to the adjacent tissues (p < 0.001) (Fig. 1B). Subsequently, the online database TargetScan 6.2 showed that SOX4 was predicted to be a direct target of miR-138. So we detected the mRNA level of SOX4 in seven NSCLC cell lines and tissues, respectively. Our findings suggested that the expression of SOX4 was significantly increased in all NSCLC cell lines compared to that in HBE at the mRNA level (p < 0.01) (Fig. 1C). The expression of SOX4 in NSCLC tissues was also significantly upregulated compared to adjacent normal tissues (p < 0.001) (Fig. 1D). From the above data, we predicted that SOX4 might be negatively regulated by miR-138.

Figure 1

The expression of microRNA-138 (miR-138) in non-small cell lung cancer (NSCLC) tissues and cell lines. (A) Relative miR-138 level analyzed by quantitative real-time reverse transcriptase polymerase chain reaction (qRT-PCR) in seven lung cancer cell lines (H522, H1299, H460, A549, PAa, H1975, and Calu-3) and a human bronchial epithelial cell line (HBE) were normalized with U6 small nuclear RNA (snRNA). (B) Relative miR-138 expression levels in NSCLC tissues (n = 10) and their corresponding adjacent normal tissues (n = 10). (C) Relative sex-determining region Y (SRY)-related high-mobility group (HMG)-box 4 (SOX4) level analyzed by qRT-PCR in seven lung cancer cell lines (H522, H1299, H460, A549, PAa, H1975, and Calu-3) and a human bronchial epithelial cell line (HBE) were normalized with glyceraldehyde 3-phosphate dehydrogenase (GAPDH). (D) Relative SOX4 expression levels in NSCLC tissues (n = 10) and their corresponding adjacent normal tissues (n = 10). All data are presented as mean ± standard error of the mean (SEM), n = 6. *p < 0.05, **p < 0.01, ***p < 0.001 versus HBE or normal tissues.

miR-138 Inhibited Cell Proliferation of Both A549 and H1975 Cells

Because the level of miR-138 was dramatically reduced in multiple cancers, we believed that miR-138 could act as a suppressor of cell growth. After transfection with the miR-138 mimic or inhibitor, the qRT-PCR analysis showed that the level of miR-138 was significantly upregulated or downregulated in the miR-138 mimic or inhibitor group compared to the miR-NC or anti-miR-NC group in A549 and H1975 cells (p < 0.01) (Fig. 2A). To determine the role of miR-138 in the proliferation of NSCLC cells, the ELISA-BrdU assay demonstrated that overexpression of miR-138 dramatically inhibited the proliferation of A549 and H1975 cells, whereas knockdown of miR-138 promoted NSCLC cell proliferation in a time-dependent manner (p < 0.05) (Fig. 2B). To further confirm this result, we detected the expression of the PCNA protein. We found that the miR-138 mimic could significantly reduce the expression of PCNA, and the miR-138 inhibitor had the reverse effect on PCNA expression (p < 0.05) (Fig. 2C).

Figure 2

Effects of miR-138 on cell proliferation in A549 and H1975 cells. A549 and H1975 cells were transfected with miR-138 mimic or inhibitor. (A) The mRNA levels of miR-138 in A549 and H1975 cells were determined by RT-PCR. (B) Cell proliferation of A549 and H1975 cells was assessed by enzyme-linked immunosorbent assay-bromodeoxyuridine (ELISA-BrdU) assay. (C) The protein expression of proliferating cell nuclear antigen (PCNA) by A549 and H1975 cells was determined by Western blot. α-Tubulin was detected as a loading control. All data are presented as mean ± SEM, n = 6. #p < 0.05, ##p < 0.01 versus miR-negative control (NC) or anti-miR-NC.

Effect of miR-138 on the Invasion of NSCLC Cells

To study the role of miR-138 in the invasion of NSCLC cells, we evaluated the invasive capacities of A549 and H1975 cells transfected with the miR-138 mimic or inhibitor by Transwell invasion assays. The results from the Transwell assays showed that the invasive capability of A549 and H1975 cells was remarkably suppressed in the miR-138 mimic group compared to the miR-NC group, but was significantly promoted in the miR-138 inhibitor group (p < 0.05) (Fig. 3). These results confirmed that miR-138 might play an important role in inhibiting invasion of A549 and H1975 cells.

Figure 3

The effects of miR-138 on invasion in A549 and H1975 cells. The invasion of A549 and H1975 cells transfected with miR-138 mimic or miR-NC (left) and miR-138 inhibitor or anti-miR-NC (right) was assessed by Transwell assay. All data are presented as mean ± SEM, n = 6. #p < 0.05, ##p < 0.01 versus miR-NC or anti-miR-NC.

Effects of miR-138 on Secretions and Expressions of MMP-2 and MMP-9 in NSCLC Cells

Because MMPs are closely correlated with cell invasion, we determined the protein expressions of MMP-2 and MMP-9. Our ELISA assay indicated that secretions of MMP-2 and MMP-9 in the culture supernatants were significantly decreased by the overexpression of miR-138 in A549 and H1975 cells and significantly increased by knockdown of miR-138 (p < 0.01) (Fig. 4A). In addition, we detected the mRNA expression of MMP-2 and MMP-9 by qRT-PCR. After transfection with the miR-138 mimic, the mRNA levels of MMP-2 and MMP-9 were distinctly reduced, and the miR-138 inhibitor enhanced the mRNA levels of both (p < 0.01) (Fig. 4B). Our results suggested that decreased expressions of MMP-2 and MMP-9 might be a possible mechanism that contributed to the inhibitory effect of the miR-138 mimic on the invasion of A549 and H1975 cells.

Figure 4

Overexpression of miR-138 suppressed expressions and secretions of matrix metalloproteinase 2 (MMP-2) and MMP-9. A549 and H1975 cells were transfected with miR-138 mimic or inhibitor. (A) Levels of MMP-2 and MMP-9 were detected in the culture supernatants of cultured A549 and H1299 cells by ELISA assay. (B) The mRNA levels of MMP-2 and MMP-9 in A549 and H1299 cells were examined by RT-PCR. All data are presented as mean ± SEM, n = 6. ##p < 0.01, ###p < 0.001 versus miR-NC or anti-miR-NC.

Effects of miR-138 on EMT in NSCLC Cells

To determine the effect of miR-138 on EMT in NSCLC cells, we examined the effect of the miR-138 mimic or inhibitor on the expressions of EMT markers in A549 and H1975 cells using qRT-PCR and Western blotting. The miR-138 mimic could upregulate the expression of E-cadherin, an epithelial marker, and downregulate the expressions of N-cadherin and vimentin, two mesenchymal markers, in A549 and H1975 cells at both the mRNA and protein levels (p < 0.01) (Fig. 5A), but the miR-138 inhibitor had the opposing effects on the expressions of these EMT markers (p < 0.01) (Fig. 5B). Taken together, our results suggested that overexpression of miR-138 could effectively inhibit the EMT of NSCLC cells.

Figure 5

The effects of miR-138 on the expression of epithelial–mesenchymal transition (EMT)-related molecules in A549 and H1975 cells. (A) The mRNA and protein levels of epithelial (E)-cadherin, neural (N)-cadherin, and vimentin were determined by qRT-PCR and Western blotting in A549 and H1975 cells transfected with miR-138 mimic or miR-NC, respectively. (B) The mRNA and protein levels of E-cadherin, N-cadherin, and vimentin were determined by qRT-PCR and Western blotting in A549 and H1975 cells transfected with miR-138 inhibitor or anti-miR-NC, respectively. α-Tubulin was detected as a loading control. All data are presented as mean ± SEM, n = 6. #p < 0.05, ##p < 0.01, ###p < 0.001 versus miR-NC or anti-miR-NC.

SOX4 Is a Direct Target of miR-138 in NSCLC Cells

The online database TargetScan 6.2 predicted that SOX4 was a binding target of miR-138. Next, we performed qRT-PCR and Western blotting to determine the expression of SOX4 at the mRNA and protein levels in A549 and H1975 cells transfected with the miR-138 mimic or inhibitor. We found that the mRNA and protein levels of SOX4 were prominently reduced after transfection with the miR-138 mimic and were enhanced after transfection with the miR-138 inhibitor (p < 0.01) (Fig. 6A and B). Furthermore, luciferase reporter assay had demonstrated that miR-138 directly targeted SOX4. SOX4 3′-UTR was cloned into a luciferase reporter vector, and the putative miR-138 binding site in the SOX4 3′-UTR was mutated (Fig. 6C). The results showed that introduction of miR-138 dramatically inhibited the luciferase activity of pMIR-SOX4 3′-UTR WT (p < 0.001) (Fig. 6D). Mutation of the miR-138 binding site in the SOX4 3′-UTR abolished the effect of miR-138, which suggested that SOX4 was directly and negatively regulated by miR-138.

Figure 6

SOX4 was a direct target of miR-138. (A) The mRNA level of SOX4 was determined by qRT-PCR in A549 and H1975 cells transfected with miR-138 mimic or inhibitor, respectively. (B) The protein expression of SOX4 was determined by Western blot in A549 and H1975 cells transfected with miR-138 mimic or inhibitor, respectively. (C) Schematic representation of SOX4 3′-untranslated regions (3′-UTRs) showing putative miRNA target site. (D) The analysis of the relative luciferase activities of SOX4 wild type (WT) and SOX4 mutation (MUT) in NSCLC cells. All data are presented as mean ± SEM, n = 6. ##p < 0.01, ###p < 0.001 versus miR-NC or anti-miR-NC.

Knockdown of SOX4 Had Similar Effects to Overexpression of miR-138

To investigate the function of SOX4 in NSCLC cells, A549 and H1975 cells were transfected with si-NC or si-SOX4. The qRT-PCR and Western blot analysis showed that expression of SOX4 was significantly decreased at the mRNA and protein levels after 24 h in both A549 and H1975 cells transfected with si-SOX4 (p < 0.001) (Fig. 7A). The results from the ELISA-BrdU assay indicated that knockdown of SOX4 could significantly inhibit proliferation of NSCLC cells (p < 0.01) (Fig. 7B). Moreover, the Transwell assay revealed that downregulation of SOX4 suppressed invasion of NSCLC cells (p < 0.01) (Fig. 7C). Silencing of SOX4 contributed to the upregulation of E-cadherin and the downregulation of N-cadherin and vimentin (p < 0.01) (Fig. 7D). Therefore, knockdown of SOX4 induced a very similar phenotype to miR-138 overexpression in NSCLC cells. All the above results confirmed that miR-138 downregulated the expression of SOX4 to inhibit NSCLC cell growth, invasion, and EMT.

Figure 7

SOX4 silencing also could inhibit cell proliferation, invasion, and EMT in NSCLC cells. A549 and H1975 cells were transfected with small interfering RNA SOX4 (si-SOX4) or si-NC. (A) The mRNA and protein levels of SOX4 were determined by qRT-PCR and Western blot, respectively. (B) Cell proliferation was assessed by ELISA-BrdU assay. (C) The invasion of A549 and H1975 cells was assessed by Transwell assay. (D) The expressions of E-cadherin, N-cadherin, and vimentin were determined by qRT-PCR in A549 and H1975 cells, respectively. All data are presented as mean ± SEM, n = 6. #p < 0.05, ##p < 0.01, ###p < 0.001 versus si-NC.

Overexpression of SOX4 Partially Reversed the Effects of the miR-138 Mimic on the Proliferation, Invasion, and EMT of NSCLC Cells

To confirm whether miR-138 inhibited the proliferation, invasion, and EMT of NSCLC cells through a SOX4-dependent mechanism, we cotransfected A549 and H1975 cells with miR-NC, miR-138 mimic and pcDNA, and miR-138 mimic and pcDNA3.1-SOX4 vector. Our data showed that the miR-138 mimic could dramatically decrease SOX4 expression, whereas the expression of SOX4 was significantly increased after transfection with miR-138 and pcDNA-SOX4 compared with the miR-138 mimic and pcDNA vector group in both A549 and H1975 cells (p < 0.001) (Fig. 8A). Data from the ELISA-BrdU assay showed that introduction of SOX4 promoted the proliferation of NSCLC cells transfected with the miR-138 mimic (p < 0.01) (Fig. 8B). The Transwell assay indicated that upregulation of SOX4 could reverse the inhibitory effect of miR-138 overexpression on the invasion of NSCLC cells (p < 0.01) (Fig. 8C). Moreover, introduction of SOX4 decreased the expression of E-cadherin and increased the expressions of N-cadherin and vimentin in A549 and H1975 cells after transfection with the miR-138 mimic (p < 0.01) (Fig. 8D). Therefore, our results clearly demonstrated that knockdown of SOX4 was essential for the miR-138 mimic-induced inhibition of cell proliferation, invasion, and EMT in NSCLC cells.

Figure 8

Overexpression of SOX4 partially rescued miR-138-inhibited cell proliferation, invasion, and EMT in NSCLC cells. A549 and H1975 cells were transfected with miR-NC, miR-138 mimic with or without pcDNA-SOX4 vector. (A) The mRNA and protein levels of SOX4 were determined by qRT-PCR and Western blot, respectively. (B) Cell proliferation was assessed by ELISA-BrdU assay. (C) The invasion of A549 and H1975 cells was assessed by Transwell assay. (D) The expressions of E-cadherin, N-cadherin, and vimentin were determined by qRT-PCR in A549 and H1975 cells, respectively. All data are presented as mean ± SEM, n = 6. ##p < 0.01, ###p < 0.001 versus miR-NC; &p < 0.05, &&p < 0.01 versus miR-138 mimic + pcDNA.

Inhibition of SOX4 Increased the Level of miR-138 in NSCLC Cells via Activation of p53

It has been reported that SOX4 can suppress the effects of p5330. In addition, p53, a downstream target of SOX4, has been shown to control the expression of a number of prominent miRNAs involved in tumorigenesis31. Based on these previous findings, we considered that inhibition of SOX4 by the upregulation of miR-138 induced expression of p53, culminating in greater levels of miR-138. To confirm this hypothesis, we cotransfected A549 and H1975 cells with si-SOX4 or si-p53 to detect the protein expression of p53 or levels of miR-138 by Western blot or qRT-PCR, respectively. Our results showed that overexpression of miR-138 significantly enhanced the expression of p53, but upregulation of SOX4 induced by pcDNA-SOX4 could dramatically reduce the p53 expression (Fig. 9A). In addition, silencing SOX4 induced p53 expression (Fig. 9B). Furthermore, downregulation of SOX4 silenced by si-SOX4 made a significant increase in the level of miR-138, and decreasing p53 expression reduced the level of miR-138, while silencing p53 by si-p53 could reverse the effect of si-SOX4 on the level of miR-138 (Fig. 9C). Finally, downregulation of p53 could significantly reverse the inhibitory effects of the miR-138 mimic and SOX4 knockdown on cell proliferation, invasion, and EMT of NSCLC (Fig. 9D–F). From all the above results, we concluded that increased miR-138 dramatically inhibited growth, invasion, and EMT of NSCLC cells via a SOX4/p53 feedback loop.

Figure 9

A regulatory network linking miR-138, SOX4, and p53. (A) The protein expression of tumor protein 53 (p53) was determined by Western blot. A549 cells were transfected with miR-138 and pcDNA or pcDNA-SOX4 (B) The protein expression of p53 was determined by Western blot. A549 cells were transfected with si-NC or si-SOX4. (C) The level of miR-138 was determined by qRT-PCR in A549 cells. A549 cells were transfected with si-NC or si-SOX4 with or without si-p53. (D) Cell proliferation was assessed by ELISA-BrdU assay. A549 cells were transfected with miR-NC, miR-138 mimic with or without pcDNA-SOX4, si-tumor protein 53 (p53), si-SOX4, and si-SOX4 with si-p53. (E) The invasion of A549 cells was assessed by Transwell assay. (F) The expressions of E-cadherin, N-cadherin, and vimentin were determined by qRT-PCR in A549. All data are presented as mean ± SEM, n = 6. *p < 0.05, **p < 0.01 versus miR-NC; ##p < 0.01, ###p < 0.001 versus si-NC or miR-138 mimic + si-NC; &p < 0.05, &&p < 0.01 versus si-SOX4.

DISCUSSION

Up to now, miRNAs have been considered as critical regulators involved in different biological processes such as cell proliferation, apoptosis, differentiation, invasion, transcriptional regulation, and tumorigenesis32. Globally miRNA dysregulation of cancers has provided major insights into the precise mechanisms of oncogenesis33. As a main tumor suppressor, miR-138 was found to be reduced in many human cancers such as colorectal, esophageal, and bladder cancers as well as NSCLC25–28. The precise mechanism of miR-138 in NSCLC remained unclear. Therefore, in this study, we set out to elucidate the biological function and mechanism of miR-138 in NSCLC. Our results demonstrated that miR-138 was frequently downregulated in NSCLC cell lines and tissues. According to these findings, we speculated that miR-138 might be a potential antioncogene in NSCLC, which was consistent with a previous study34. As expected, introduction of miR-138 inhibited proliferation, invasion, and EMT of A549 and H1975 cells. Our findings suggested that miR-138 played a crucial role in regulating proliferation, invasion, and EMT of NSCLC, and might be a possible diagnostic and predictive biomarker.

Afterward, we explored the precise mechanism of miR-138 on the inhibition of proliferation, invasion, and EMT of NSCLC cells. In our study, qRT-PCR, Western blotting, and luciferase reporter assay confirmed that miR-138 directly targeted SOX4. Importantly, we also demonstrated that upregulating SOX4 expression partly reversed the inhibitory effects of miR-138 overexpression on the proliferation, invasion, and EMT of NSCLC cells. Hence, we concluded that miR-138 played an important role in the inhibition of proliferation, invasion, and EMT of NSCLC cells, partially by decreasing the protein expression of SOX4.

In this study, the ELISA-BrdU assays showed that overexpression of miR-138 could result in remarkable inhibition of proliferation of A549 and H1975 cells. The Western blot analyses also showed that the expression of PCNA, a proliferative marker, was significantly decreased in cells transfected with a miR-138 mimic compared to cells transfected with miR-NC. However, the miR-138 inhibitor could significantly promote proliferation of A549 and H1975 cells compared with the anti-miR-NC group. The Transwell assay showed that overexpression or knockdown of miR-138 dramatically suppressed or promoted the invasion of A549 and H1975 cells compared with the miR-NC or anti-miR-NC group, respectively. Moreover, in order to study the effect of miR-138 on EMT of NSCLC cells, we determined the changes of EMT markers in A549 and H1975 cells transfected with the miR-138 mimic or inhibitor. It turned out that introduction of miR-138 could markedly inhibit the invasive ability of NSCLC cells by dramatically increasing the epithelial marker E-cadherin and decreasing the mesenchymal markers N-cadherin and vimentin, whereas the miR-138 inhibitor had the opposing effect on the expression of EMT markers.

Next, our findings confirmed that SOX4 was a direct target of miR-138 in NSCLC cells. Many studies have indicated malignancies associated with high expression of SOX4 including colorectal cancer35, prostate cancer36, cervical cancer37, osteosarcoma38, and NSCLC14, indicating that SOX4 may be an important oncogene contributing to the progression and metastasis of cancers. For instance, SOX4 promoted the proliferation and invasion of esophageal tumor cells by silencing miR-31 via activation and stabilization of a corepressor complex with EZH2 and histone deacetylase 3 (HDAC3)39. A proapoptosis molecule, p53, may be responsible for induction of apoptosis, as SOX4 interacts with and stabilizes p53 protein by blocking mouse double minute 2 homolog (MDM2)-mediated p53 ubiquitination and degradation40. More recently, overexpression of SOX4 was closely correlated with progression and metastasis of tumors9,10,13. In addition, SOX4 is also a regulator of EMT41. EMT is a pivotal developmental program that is often activated during tumor progression and may enhance resistance to treatment42. Zhang et al. found that upregulation of SOX4 in human mammary epithelial cells contributed to the acquisition of mesenchymal traits and promoted cell migration and invasion12. A previous study has demonstrated that miR-138 can inhibit invasion and metastasis of ovarian cancer cell by downregulation of SOX443. In the present study, SOX4 was also found to be upregulated in NSCLC cells and tissues. Furthermore, we found that knockdown of SOX4 using siRNA oligos inhibited the proliferation, invasion, and EMT of NSCLC cells, which had similar effects to miR-138 overexpression. Restoration of SOX4 reversed the inhibitory effects of the miR-138 mimic, suggesting that SOX4 might play a critical role in NSCLC progression and metastasis.

It has been reported that the depletion of SOX4 mRNA by siRNA can induce apoptosis in cancer cells and can regulate p53 stability40,44, suggesting that SOX4 may inhibit p53-mediated apoptosis. The tumor suppressor p53 plays a central role in tumor suppression. Recent studies have demonstrated that p53 and its network are regulated by miRNAs at multiple levels45. Some miRNAs regulate the level and function of p53 through directly targeting p53. However, some other miRNAs target regulators of p53, such as MDM2 and MDM4, to indirectly regulate the activity and function of p5346,47. On the other hand, p53 also regulates the transcriptional expression and the biogenesis of a group of miRNAs, which contributes to the tumor suppressive function of p5348. For example, a tumor suppressor miR-34 directly and negatively regulates the expression of nuclear receptor subfamily 4, group A, member 2 [NR4A2; or nuclear receptor-related 1 protein (NURR1)], while inhibition of NR4A2 can enhance the activation of p53, leading to upregulation of miR-34, which indicated that the positive feedback loop miR-34/NR4A2/p53 is a novel mechanism for the treatment of tumors49. Ye et al. reported that miR-138 can regulate p53 signaling50. In this study, we found that miR-138 could increase the expression of p53 through downregulation of SOX4. Furthermore, inhibition of SOX4 significantly increased p53 expression, resulting in the upregulation of miR-138 levels. For further study, we found that downregulation of p53 could significantly reverse the inhibitory effects of the miR-138 mimic and SOX4 knockdown on cell proliferation, invasion, and EMT of NSCLC cells. This formed a miR-138/SOX4/p53+ feedback loop in regulating the proliferation, invasion, and EMT of NSCLC cells.

In conclusion, we have demonstrated that relative miR-138 expression was dramatically reduced in NSCLC tissues and cells. Upregulation of miR-138 inhibited proliferation, invasion, and EMT of NSCLC cells by direct regulation of SOX4/p53. This miR-138/SOX4/p53 feedback loop might provide new insights into the molecular mechanisms underlying progression and metastasis of NSCLC, and overexpression of miR-138 might be a possible therapeutic strategy for the therapy of NSCLC in the future.