The human normal lung cell lines CCD-19Lu and MRC-9, and human lung cancer cell lines A549 and HCC827 were obtained from the American Type Culture Collection (ATCC). Cells were maintained in complete medium with high-glucose RPMI-DMEM (Hyclone; GE Healthcare) and 10% fetal bovine serum (FBS; Gibco), at 37°C in a humidified atmosphere containing 5% CO2. Antibodies against FOXQ1 (ab194564), TGF-β (ab9758), total Smad2/3 (ab63672), phospho-Smad2/3 (ab63399), E-cadherin (ab76055), N-cadherin (ab18203), vimentin (ab92547), and GAPDH (ab181602) were all purchased from Abcam. Hsa-miR-133 mirVana miRNA mimic (MC10029) and mirVana miRNA negative control (4464061) were both purchased from Life Technologies. Transfection into cell lines was performed following manufacturer's instructions. After transfection, cells were incubated for another 48h before being subjected to other experimental assays.

Cells were seeded in 6-well plates at a density of 1×106 cells/well with 2ml complete DMEM medium. Twenty-four hours after transfection of either hsa-miR-133 or negative control, total protein was extracted using RIPA buffer (50mM Tris–HCl, pH 7.4, 150mM NaCl, 1mM EDTA, 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, and protease inhibitors). The protein concentration was determined using the BCA assay. Equal quantities of proteins were separated by SDS-PAGE and transferred to nitrocellulose membrane using a semi-dry transfer unit (Bio-Rad Laboratories). The membrane was then immersed in blocking buffer (PBS, 0.1% Tween-20) with 5% nonfat milk for 20min and then incubated with appropriate primary antibodies overnight at 4°C. After washing with blocking buffer, blots were incubated with HRP-conjugated secondary antibody (Life Technologies) at room temperature for 1h, washed again with blocking buffer, and visualized using Luminata Forte Western HRP Substrate (EMD Millipore).

A total of 1×105 cells transfected with miRNA negative control or hsa-miR-133 were seeded into a 6-well plate after transfection. A linear wound was carefully made with a 100μl sterile pipette tip across the confluent cell monolayer, and the cell debris was removed by washing with PBS. The migrated distance of the growing edge on wounded monolayers was measured at 24h after being wounded, and migration rate was expressed as a percentage relative to respective controls.

Cell invasion assay was performed in a 24-well plate with 8μm pore size chamber inserts (BD Biosciences). A total of 1×105 cells transfected with miRNA negative control or hsa-miR-133 were placed into the upper chamber per well with the Matrigel-coated membrane, diluted with serum-free culture medium. The lower compartment was filled with 500μl of medium containing 10% FBS to attract cells. The cells were incubated at 37°C in a 5% CO2 humidified incubator for 24h, followed by exposure to 20μM 5-ethynyl-2′-deoxyuridine (EdU) for an additional 4h at 37°C. Membrane inserts were removed from the plate and stained using the ENU kit (Invitrogen). The cells were counted under six random microscopic fields for each well, and invasion rate was expressed as a percentage relative to respective controls.

The cells were seeded in a 24-well plate (1×105 cells/well). Briefly, for attachment assay, after 1h incubation, non-attached cells were washed twice with PBS, and the attached cells were counted after trypsinization. The attachment rate was expressed as percentage of cell numbers of the attached cells to total cells, and normalized to respective control. For the detachment assay, the cells were detached with 0.05% trypsin for 3min and counted after the seeded cells were incubated 24h. The remaining attached cells were further trypsinized with 0.25% trypsin and counted. Cells detachment rate was expressed as a percentage of the detached cells to total cells, and normalized to respective control.

All values were presented as mean±standard error mean (SEM). A 2-tailed Student's t-test was used to establish significant differences between groups. Data were determined to be statistically different when P was <.05.

We found that the miR-133 sequence has a putative binding site on the 3′-UTR of FOXQ1 mRNA, using the microRNA.org resource. Moreover using RT-PCR, upon examination of its expression in two human lung cancer cell lines A549 and HCC827, we found miR-133 levels were significantly lower than in two normal lung cell lines CCD-19Lu and MRC-9. MiR-133 levels in CCD-19Lu and MRC-9 were similar, whereas in A549 it was down-regulated to approximately 30% of the normal cell lines, and in HCC827 to about 40% of the normal cell lines. These results suggested that, compared with normal human lung cells, miR-133 was significantly down-regulated in lung cancer cell lines.

We next validated the putative binding site of miR-133 on the 3′-UTR of FOXQ1, using a luciferase reporter assay. The sequences from wild type (WT-Luc) 3′-UTR of FOXQ1 mRNA and its mutated version (mut-Luc) were fused to the downstream of luciferase ORF, respectively. The two luciferase constructs were then separately transfected into both A549 and HCC827 cells, together with either miRNA negative control or miR-133 mimic. The luciferase reporter activity was then assessed. As a result, miR-133 dramatically reduced WT-Luc activity to approximately 30% of the control transfected experiments, consistently in both A549 and HCC827 cell lines. On the other hand, activity of mut-Luc was largely unaffected in either miR-133 or control miRNA transfected cells. The complementary sequences on FOXQ1 3′-UTR were then a direct target of miR-133.

We continued to determine whether FOXQ1 was indeed an in vivo target of miR-133. MiR-133 and a negative miRNA control were transfected into both A549 and HCC827 cells, which were then subjected to RT-PCR to examine FOXQ1 levels in mRNA. Levels of miR-133 in both A549 and HCC827 cells were indeed greatly increased upon transfection. Compared with control miRNA transfections, miR-133 significantly reduced FOXQ1 messenger levels in both lung cancer cell lines examined (Fig. 1A), clearly indicating that FOXQ1 mRNA is a bona fide target of miR-133. Using anti-FOXQ1antibody, we confirmed that its protein level was also markedly reduced in miR-133, but not in control transfected cells (Fig. 1C).

Fig. 1.

FOXQ1 is down-regulated by transfection of miR-133 into lung cancer cell lines.

(A and B) Relative FOXQ1 (A) and TGF-β (B) mRNA levels in A549 and HCC827 cell lines, after transfection with either miRNA negative control or miR-133. (C and D) FOXQ1 (C) and TGF-β (D) protein levels in A549 and HCC827 cell lines, after transfection with either miRNA negative control or miR-133. GAPDH was used as a loading control. Quantification of FOXQ1 and TGF-β protein expressions normalized to GAPDH was also shown in the lower panels. Values were mean±standard error mean (SEM) of three independent experiments. **P<.01 to respective control.

(0.24MB).

Other authors have reported that FOXQ1 could promote the expression of TGF-β,16 and was involved in the crosstalk between TGF-β and Wnt signaling pathways in the progression of colorectal cancer.7 We were curious to investigate whether a similar correlation also existed between these two genes in lung cancer. Interestingly, down-regulation of FOXQ1 by miR-133 transfection also greatly reduced TGF-β expressions in both lung cancer cell lines (Fig. 1B and D), suggesting that FOXQ1 and TGF-β are both likely to be involved in tumorigenesis of lung cancer. More importantly, we found that in both lung cancer cell lines A549 and HCC827, miR-133 expression resulted in a significant inhibition of TGF-β pathway, as indicated by reduced phosphorylation of Smad2/3 (Fig. 2A), both of which are downstream effectors of the TGF-β pathway.17 Of note, we further confirmed that the regulation of TGF-β pathway by miR-133 was indeed mediated through FOXQ1, since in a co-transfection experiment using both miR-133 and a plasmid expressing FOXQ1 without its 3′-UTR, TGF-β level was also upregulated upon FOXQ1 expression, even in the presence of miR-133 (Fig. 2B–D).

Fig. 2.

MiR-133 inhibits TGF-β pathway through down-regulation of FOXQ1 in lung cancer cell lines.

(A) Total and phosphorylated Smad2/3 protein levels in A549 and HCC827 cell lines, after transfection with either miRNA negative control or miR-133. GAPDH was used as loading control. Quantification of p-Smad2/3 expression normalized to GAPDH was also shown in the lower panels. (B–D) Relative levels of miR-133 (B), FOXQ1 mRNA (C) and TGF-β mRNA (D) in A549 and HCC827 cell lines, after transfection with miR-133 alone, or co-transfection with miR-133 and plasmid expressing FOXQ1 in the absence of its 3′-UTR. Values were mean±SEM of three independent experiments. ns not significant to respective control; **P<.01 to respective control.

(0.25MB).

FOXQ1 is a marker for metastasis and poor prognosis of lung cancer,8,9 so introducing miR-133 to down-regulate FOXQ1 should inhibit the migration and invasion of lung cancer cells. To test this hypothesis, we performed wound-healing and cell invasion assays in both miR-133 and control miRNA-transfected A549 and HCC827 cells. As expected, both cell lines exhibited significantly reduced migration rate after transfection of miR-133, compared with control experiments (Fig. 3A and B). Moreover, in a cell invasion assay, miR-133 also consistently reduced the invasion rates of both A549 and HCC827 cells, compared with control transfected experiments (Fig. 3C and D). Taken together, the above results strongly demonstrated the role of miR-133 in inhibiting lung cancer tumorigenesis.

Fig. 3.

MiR-133 expression inhibits migration and invasion of lung cancer cells.

(A) Wound-healing assay for A549 and HCC827 cell lines, after transfection with either miRNA negative control or miR-133. (B) The migration distances relative to control in wound-healing assay were measured at 24h after being wounded. (C) Cell invasion assay for A549 and HCC827 cell lines, after transfection with either miRNA negative control or miR-133. (D) The invaded cell numbers relative to control were quantified 24h after cells were seeded. Values were mean±SEM of three independent experiments. *P<.05 to respective control.

(0.39MB).

During the wound-healing and cell invasion assays, we often saw a morphological change in cells after transfection with miR-133. Both A549 and HCC827 cells initially displayed elongated and fibroblastoid morphology, whereas upon miR133 transfection both cell lines assumed rounded cell shapes (Fig. 4A), suggesting a loss of EMT phenotype. During transition, epithelial cells at the primary tumor site acquire the mesenchymal phenotype, resulting in enhanced invasion and metastasis, spreading through the blood stream to reach secondary tissues. To confirm this observation, we performed cell attachment and detachment assays in the two lung cell lines, after transfection with either miRNA negative control or miR-133. As shown in Fig. 4B and C, miR-133 significantly reduced attachment and detachment rates of both A549 and HCC827 cells, compared with control transfected cells. These results confirmed the loss of EMT features at the cell morphological level.

Fig. 4.

MiR-133 expression inhibits epithelial–mesenchymal transition (EMT) of lung cancer cells.

(A) Cell morphology was observed for A549 and HCC827 cell lines, after transfection with either miRNA negative control or miR-133. (B and C) Cell attachment (B) and detachment (C) assays were assessed in A549 and HCC827 cell lines, after transfection with either miRNA negative control or miR-133. (D) Expression of EMT phenotypic protein markers were analyzed in A549 and HCC827 cell lines, after transfection with either miRNA negative control or miR-133. Quantification of protein expressions normalized to GAPDH was also shown in the right panels. Values were mean±SEM of three independent experiments. *P<.05, **P<.01 to respective control.

(0.33MB).

At the molecular level, EMT is characterized by reduced epithelial marker E-cadherin and increased mesenchymal marker vimentin. Accordingly, we went on to assess if the same shift in EMT phenomena could also be observed at the molecular level. As expected, miR-133 increased E-cadherin expression, but reduced vimentin and N-cadherin expressions, compared with control transfected cells (Fig. 4D), which clearly demonstrated a reverse of EMT phenotype at the protein level.