MicroRNA-29b attenuates non-small cell lung cancer metastasis by targeting matrix metalloproteinase 2 and PTEN

Information about tissue specimens, NSCLC cell lines and animals is given in the Additional file 1.

Detailed information about isolation of CD133-positive/negative A549 cells and microarray screening of differentially expressed genes is provided in the Additional file 1.

Using “miRNA” as the index word for predicting target genes in the TargetScan (www.​targetscan.​org), PicTar (www.​pictar.​org), and miRanda (www.​microrna.​org) databases, target genes were identified from overlapping results from the three databases. Subsequently, we extracted the overlap of these results with that of the tumor metastasis PCR array.

Quantitative RT-PCR was performed using kits for U6 and mature miR-29b (ABI, Foster City, CA, USA), according to the manufacturer’s instructions. SYBR green real-time RT-PCR was performed to detect MMP2 and PTEN. Detailed information is provided in the Additional file 1.

All miRNA duplexes (Additional file 2: Table S1) were purchased from Genepharma (Shanghai, P.R. China). H460 cells were transfected with inhibitor or inhibitor NC at a final concentration of 100 nmol/L using Lipofectamine RNAiMAX instructions (Invitrogen). A549 subline stably expressing miR-29b (A549-miR-29b) and its control line (A549-NC) were established as described in the Additional file 1.

The cells were lysed with radioimmunoprecipitation assay buffer (Beyotime, Shanghai, China). The antibodies used for western blotting are described in the Additional file 1.

We purchased psiCHECK-2 plasmids from Promega (Madison, WI, USA). Human genome DNA was used as the template for the MMP2 3’ untranslated region (UTR) and PTEN 3’ UTR PCR. XhoI and NotI restriction sites were introduced at the 5’ ends of both the forward and reverse primers (Additional file 2: Table S1). Following double digestion, the linear psiCHECK-2 fragment was connected with the 3’ UTRs using T4 DNA ligase; positive clones were selected for sequencing validation after transformation. The sequencing-validated, target recombinant plasmid was designated psiCHECK-2-Wt-MMP2/PTEN-3’ UTR, and was used as a template for constructing psiCHECK-2-Mut-MMP2/PTEN-3’ UTR using antisense PCR and a site-specific mutagenesis kit (Toyobo, Osaka, Japan).

To measure the effect of miRNA and inhibitor on cellular proliferation rates, cells were incubated in 10 % CCK-8 (DOJINDO) diluted in normal culture media at 37 °C until visual color conversion appears. Proliferation rates were determined at 24, 48, 72, 96, 120 h post-transfection, and quantification was done on a microplate reader set according to the manufacturer’s protocol.

The 24-well Boyden chamber with 8-μm pore size polycarbonate membrane (Corning, NY) was used to analyze the migration and invasion of tumor cells. Details are in the Additional file 1.

H460 subline stably knockdown miR-29b (H460-LV-miR-29b inhibitor) and its control line (H460-LV-CON), were established as described in Additional file 1. Analysis for tumorigenicity was performed as described in Additional file 1.

All data were analyzed using SPSS 13.0 (SPSS Inc, Chicago, IL, USA); A paired t test was used to investigate the difference in the expression level of miR-29b between normal and cancerous tissues. A 2-sample t test was used to analyse the clinicopathologic characteristics of miR-29b expression in the tissues of patients with NSCLC. Quantitative RT-PCR, CCK-8 assay, migration and invasion assay, and luciferase reporter assay were tested using 1-way analysis of variance for factorial design. P value < 0.05 was considered statistically significant.

To explore the miRNAs related to NSCLC metastasis, miRNA PCR array (MAH-3100A detected 376 human disease–related miRNA) were used to evaluate miRNA expression in primary cultured CD133-positive/negative A549 cells. Fourteen miRNAs were found up-regulated and thirty-seven miRNAs were down-regulated in CD133-positive cells (Fig. 1a, Additional file 3: Table S2). The human tumor metastasis PCR array (PAHS-028A detected 84 metastasis-related genes) was used to further determine the metastasis-related genes that could be controlled by CD133-regulated miRNAs. Nineteen metastasis-related genes were found up-regulated in CD133-positive cells (Fig. 1b, Additional file 4: Table S3). Finally, the target genes of significantly different miRNAs were predicted by bioinformatics analysis. The overlap genes were found between bioinformatics predicted analysis and tumor metastasis PCR array. Among the predicted target genes of the seven down-regulated miRNAs in CD133-positive A549 cells, the tumor metastasis PCR array contained four target genes of miR-29b (Fig. 1c). MiR-29b was down-regulated 7.6-fold in CD133-positive cells. however, its putative target genes PTEN, ETV4, COL4A2, and MMP2 were up-regulated 1.11 to 4.2-fold. Based on PTEN and MMP2 were reported closely related to metastasis process, these two genes were further investigated to confirm their regulation by miR-29b in NSCLC.

Quantitative RT-PCR results revealed that the expression levels of miR-29b were significantly higher in the H460 and 95C cell lines compared to 16HBE cell line, while the expression levels were lower in the PGCL3, PAa, H520, A549, H1299 and 95D cell lines (Fig. 2a). Twenty pairs of paraffin-embedded NSCLC tissues and normal tissues (Fig. 2b) and ten pairs of fresh NSCLC tissues and normal adjacent tissues (Fig. 2c) were also chosen to detect the expression levels of miR-29b, the results showed that the expression level of miR-29b in twenty cases of paraffin NSCLC tissues was (−1.893 ± 1.367), significantly lower than that in the adjacent lung tissue (−0.605 ± 0.639; P = 0.001, t = −3.817). The expression level of miR-29b in ten cases of fresh non-small cell lung cancer tissues was (−1.996 ± 0.460), significantly lower than that in the adjacent lung tissue (−0.463 ± 0.257; P < 0.001, t = −9.016).

Data present from Table 1 showed the clinicopathologic characteristics of miR-29b expression in NSCLC patients. There was no significant relationship of miR-29b expression with age (P = 0.578), gender (P = 0.862), histology (P = 0.625) and differentiation (P = 0.891); while miR-29b expression was found had significant relationships with lymphatic metastasis (P = 0.004) and clinic stage (P = 0.031). Spearman rank correlation analysis was applied to analyze the expression levels of miR-29b, tumor stage and lymphatic metastasis in NSCLC tissues. The expression of miR-29b was positively correlated with lymphatic metastasis (r = −0.547, P = 0.043). Based on our findings, the expression of miR-29b is down-regulated in NSCLC compared to normal tissues and significantly associated with metastasis.

Cell proliferation, migration, and invasion are key steps in tumor metastasis required by tumor cells for metastatic progression in target microenvironments. In A549 cells, after miR-29b lentivirus and negative control (NC) infection, a visible significant difference was seen with fewer miR-29b transfected cells counted than NC transfected cells in migration and invasion assays (Fig. 3a). Consistently, compared with negative controls, the proliferation abilities of A549-miR-29b cells were significantly decreased (Fig. 3b, *P < 0.05, **P < 0.01). Nude mice xenograft model was subsequently applied to evaluate the effect of miR-29b on tumorigenicity (Fig. 3c). Statistical analysis of the mean tumor volume (cm³) demonstated that miR-29b lentivirus infection inhibited the tumor growth comparing to the control groups (Fig. 3d, **P < 0.01). These findings might indicate that upregulation of miR-29b had a potential to inhibit of metastasis of NSCLC.

In the present study, we have observed an apparent endogenous expression of miR-29b in H460 cells. Therefore, miR-29b silencing with antisense oligonucleotides was administrated in H460 cells. Figure 4a showed that cell migration and invasion ability was promoted in miR-29b inhibitor group comparing to the control groups. miR-29b inhibitor also increased H460 cells proliferation in a time-dependent manner (Fig. 4b, *P < 0.05, **P < 0.01). Compared to H460 cells or H460-LV-NC cells group, tumor growth rates and tumor volumes of H460-LV-miR-29b-inhibitor cells group were significantly increased (Fig. 4c, d, *P < 0.05). Collectively, these observations suggested that miR-29b suppressed growth and metastasis of NSCLC cell in vitro and in vivo.

High-invasion (A549-H) and low-invasion (A549-L) NSCLC cell sublines from A549 cells were created by using the 10 times repeated transwell assay. Herein, the role of miR-29b in cell migration and invasion was evaluated these two kinds of cells. miR-29b was found down-regulated in A549-H cells and up-regulated in A549-L cells. The images showed that miR-29b overexpression inhibited A549-H cells migration and invasion (Fig. 5a). The result of statistical analysis was showed in Fig. 5b (*P < 0.05). As expected, miR-29b inhibitor promoted cell migration and invasion (Fig. 5c) of A549-L cells. The result of statistical analysis was showed in Fig. 5d (*P < 0.05). These data confirmed that miR-29b was a metastasis suppressor in NSCLC cells.

To further explore the mechanisms of miR-29b which suppresses lung cancer cell invasion and metastasis, we analyzed probable down stream tumor metastasis-related genes. Computational prediction was used to find out the most likely target genes. The taregets of miR-29b were analyzed by the TargetScan, PicTar and MiRanda databases. The analysis revealed that miR-29b bound to the MMP2 3’ UTR with a partially complementary pattern (Fig. 6a). After A549 cells were infected with lentivirus LV-miR-29b, the expression of miR-29b was increased significantly compared with the NC and Blank groups (Fig. 6b, *P < 0.05). As expected, the expression of MMP2 mRNA was down-regulated significantly (Fig. 6c, *P < 0.05). Moreover, western blot results showed that MMP2 protein expression was decreased significantly (Fig. 6d). Similar results were obtained when miR-29b inhibitor was transfected into the H460 cells, the miR-29b expression level decreased and MMP2 mRNA was up-regulated significantly after 48 h transfection compared with the NC group and Blank groups (Fig. 6b, c, *P < 0.05). MMP2 protein expression was increased after 72 h transfection (Fig. 6d). These findings indicated that miR-29b regulated MMP2 expression negatively. To verify whether MMP2 is a direct target of miR-29b, the 3’UTR of MMP2 cDNA was cloned into the downstream region of the luciferase reporter gene (psiCHECK-2-Wt-MMP2-3’UTR) and co-transfected this vector into 293-T cells with miR-29b mimic (Additional file 5: Figure S1A, B). The luciferase reporter gene study further confirmed that miR-29b bound directly to wild-type MMP2 3’ UTR to inhibit the luciferase activity. To confirm the sequence-specific repression of miR-29b, we designed mutated versions of psiCHECK-2-Wt-MMP2-3’UTR carrying 4-bp substitutions in miR-29b target site (psiCHECK-2-Mut-MMP2-3’UTR) (Additional file 5: Figure S1C, D). There was no inhibition effect on the following site-specific mutagenesis of the miR-29b MMP2 3’ UTR binding sites (Fig. 6e, *P < 0.05), indicating that miR-29b directly regulated the target gene MMP2 negatively in NSCLC.

TargetScan indicated that miR-29b had two highly conserved PTEN 3’ UTR binding sites (Fig. 7a). Luciferase reporter activity assay was used to determine whether miR-29b regulates PTEN directly via the software-predicted binding sites. Figure 7b depicted the insertion sites of wild-type and mutant PTEN 3’ UTR into the constructed plasmids. The 1500-bp fragment of the 3’UTR region of PTEN mRNA that included the predicted miR-29b recognition site was subcloned and inserted into a luciferase reporter plasmid (Additional file 6: Figure S2A, B). Two miR-29b binding sites in the 3’UTR region of PTEN were mutated to obtain psiCHECK-2-Mut1-3-PTEN-3’UTR plasmid (Additional file 6: Figure S2C–F). Following detection, the luciferase activity in the co-transfected with wild-type PTEN-luc reporter and miR-29b mimic group was significantly decreased compared with that in the blank or NC groups (*P < 0.05). The luciferase activity in the co-transfected with mutant-type1 or mutant-type2 PTEN-luc reporter and miR-29b mimic group was also significantly decreased compared with that in the blank or NC groups (*P < 0.05). However, there were no significant changes in the co-transfected with mutant-type3 PTEN-luc reporter and miR-29b mimic group and the blank or NC groups (Fig. 7c, *P < 0.05). This result proved that miR-29b bound directly to both PTEN 3’UTR binding sites. Therefore, PTEN was a direct target of miR-29b. The influence of miR-29b on PTEN mRNA and protein expression levels were also evaluated in the A549 and H460 cells, There were no obvious changes on PTEN mRNA and protein expression levels in LV-miR-29b infected A549 cells or in miR-29b inhibitor transfected H460 cells compared with the NC or Blank groups (Fig. 7d, e).