miR-615-3p promotes the epithelial-mesenchymal transition and metastasis of breast cancer by targeting PICK1/TGFBRI axis

The human breast cancer cell lines MCF-7, BT-549, T47D, MDA-MB-468, MDA-MB-231, and HEK293T were purchased from American type culture collection (ATCC, USA). Cell lines were authenticated on the basis of viability, recovery, growth, and morphology. The expression status of ER was further confirmed by western blotting before they were used in the experiments. All cells were cultured in Dulbecco’s modified Eagle’s media medium containing 10% fetal bovine serum (Hyclone) at 37 °C with 5% CO₂ in cell culture incubators.

Cultured cells were transfected with miRCURY LNA Inhibitor Control (Negative control A catalog#199004–00, 5 nmol, EXIQON), miRCURY LNA Inhibitor (hsa-miR-615-3p catalog#411356–00, Batch#224134, 5 nmol, EXIQON), mirVana microRNA mimic negative control #1(catalog#4464058, Lot#ASO0VDIQ, 5 nmol, Ambion), mirVana microRNA mimic hsa-miR-615 (catalog#4464066, 5 nmol, Ambion). Using Ambion by Life Technologies Silencer siRNA Labeling Kit-Cy3 (cat # AM1632) according to manufacturer’s instructions. Expression plasmids of miR-615-3p were created by PCR amplification using human genomic DNA as a template. The primers are as follows: miR-615-3p, forward 5′ CCC AAG CTT GGG CAT AAT TGG ATC ATA GGA AC 3′ and reverse 5′ CCG GAA TTC CGG GTG AAT AGC TTG CAG CGT TC 3′. The reactions were incubated in tubes with 30 cycles of 98 °C for 10 s and 64 °C for 30 s. The PCR product (518 bp) was digested with HindIII and EcoRI, followed by insertion into a HindIII- and EcoRI-open pcDNA 3.1(+) vector (Invitrogen) and confirmed by DNA sequencing.

The region of the wild-type PICK1 mRNA 3′UTR with a putative miR-615-3p binding site or mutant PICK1 mRNA 3′UTR were cloned into pMIR-REPORT Luciferase vector (cat # AM5795, Applied Biosystems) using Spe I and Hind III sites. The sequences of the putative binding site and the regions targeted by mutagenesis and cloned into the reporter gene. All plasmids were verified by sequencing. These constructs were transfected into indicated cells using Lipofectamine LTX with Plus Reagent (cat #18324–012, Life Technologies). Cells were plated at a density of 3600/cm² {(1 × 10⁴) per well, into a 96-well plate and attached overnight. They were co-transfected with 100 ng of wild-type or mutant reporter vector, 10 ng of internal control pRL-TK-Renilla-luciferase plasmid (cat# E2241, Promega) and negative control #1 or mirvana microRNA miR-615-3p mimic, both from Life Technologies final concentration, 80 nM. Twenty- four hours post-transfection, luciferase activities were measured using the Dual-Glo Luciferase Assay System (cat # E2920, Promega) according to the manufacturer’s instructions. Firefly luciferase values were normalized by dividing by the Renilla luciferase values.

Total RNA was isolated with Trizol reagent (Invitrogen, USA), according to the manufacturer’s instructions. Total RNA from each sample was reverse transcribed with oligo (dT)₂₀ using SuperScript III Reverse Transcriptase (Invitrogen, USA) followed by real-time PCR. Real-time PCR was performed with SYBR Green PCR Master Mix reagents using an ABI Prism 7700 Sequence Detection System (Applied Biosystems, USA). Data were analyzed according to the comparative Ct method. U6 was used as an internal reference for miRNAs and β-actin as used as an internal reference for mRNAs. The primers are as follows: miR-615-3p, forward: 5′-ACA CTC CAG CTG GGT CCG AGC CTG GGT CTC-3′, reverse: 5′-TGG TGT CGT GGA GTC G-3′; PICK1 mRNA, forward 5′-TAC TAA CAG CGA GCT TCC GC-3′ and reverse 5′-GGT TCC GAG AGT TGG AGT GG-3′; β-actin mRNA, forward 5′-AGA GAT GGC CAC GGC TGC TT-3′ and reverse 5′-ATT TGC GGT GGA CGA TGG AG-3′; U6, forward 5′-CTC GCT TCG GCA GCA CA-3′ and reverse 5′-AAC GCT TCA CGA ATT TGC GT-3′.

Co-immunoprecipitation assays were carried out by using the Pierce Co-Immunoprecipitation Kit (#26149, Thermo Fisher, USA) according to the manufacturer’s protocol. Western blotting was performed according to the previously described procedures [24]. The cells were lysed in lysis buffer. Protein was separated by SDS-PAGE (10% gels) and transferred onto a 0.22 μm polyvinylidene fluoride (PVDF) membrane. The proteins were probed with specific antibodies overnight. After incubation, the blots were incubated with corresponding anti-rabbit IgG H&L (HRP) or anti-mouse IgG H&L (HRP) for 1 h at room temperature. The proteins were detected using ECL western blot detection system. Anti-PICK1(#ab3420,rabbit) antibody, anti-E-cadherin(#ab15148, rabbit) antibody and anti-vimentin (#ab16700, rabbit) antibody were obtained from Abcam. Anti-smad2(#5339 rabbit) antibody, anti-p-smad2(#3108 rabbit ser465/467) antibody, anti-p-smad3(#9520 rabbit ser423/425) antibody, anti-smad3(#9523 rabbit) antibody, and anti-Dicer (#5362 rabbit) antibody were purchased from Cell Signaling Technologies (CST). Anti-TGFβ RI (#sc-101,574, mouse) antibody and anti-TGFβ RII (#sc-17,791, mouse) antibody were obtained from Santa Cruz Biotechnology. Anti-β-actin (#WL01372, mouse) was obtained from Wanleibio (China).

Breast tumor and adjacent noncancerous tissues were obtained from the Affiliated Hospital of Harbin Medical University, with the informed consent of patients and with approval for experiments from the Affiliated Hospital of University.

Indicated cells (2 × 10⁶) that stably expressed miR-615-3p or control plasmids were injected into the tail vein of 6-week-old female nude mice. All mice were sacrificed 8 weeks after the injection. All the experimental procedures involving animals were conducted in accordance with Institutional Animal Care guidelines and approved ethically by the Administration Committee of Experimental Animals.

The migration and in vitro invasion were performed in 24-well cell culture inserts (Corning Life Sciences). The inserts for invasion assays were coated with 30 μL of Matrigel matrix at 37 °C for 1 h. Cells (1 × 10⁵ cells per Transwell) were added into the upper chamber and allowed 24 h for cell migration. For in vitro invasion, 2 × 10⁵ cells were added into each upper chamber and allowed 24 h for invasion. After the period of migration or invasion, cells on the undersurface of the upper units were stained and counted under a phase-contrast microscope.

Cells were cultured on coverslips overnight and fixed with 4% paraformaldehyde, followed by the treatment of 1% Triton X-100 (Thermo Fisher, MA) for permeabilization. To visualize the E-cadherin and vimentin, we incubated the coverslips with the respective antibodies for 1 h and then rhodamine-conjugated secondary antibody for another hour. The fluorescence staining was observed with the aid of a fluorescence microscope (Axiovert 200 M; Carl Zeiss, USA). 4′,6-diamidino-2-phenylindole (DAPI) was included during staining to visualize nuclei of cells.

Data were presented as the means ±S.E.M. from three or more independent experiments unless indicated otherwise. Statistical analysis was performed using two-tailed t-tests. P < 0.05 was considered a significant difference.(*P < 0.05; **P < 0.01; ***P < 0.001).

To investigate the role of miR-615-3p in breast cancer pathogenesis, we determined the expression of miR-615-3p in breast cancer tissues and adjacent normal tissues by using qRT-PCR. We observed that miR-615-3p expression was significantly increased in breast cancer tissues compared with adjacent normal breast tissues (Fig. 1a). In situ hybridization assays indicated that miR-615-3p was highly expressed in tumor tissues, whereas little signal was observed in the normal breast tissues (Fig. 1b). We further determined the correlation between the miR-615-3p level and the metastatic status of patients with breast cancer. We found that the expression of miR-615-3p was positively associated with lymph node metastasis (Fig. 1c), and tumors expressed increasing levels of miR-615-3p with the progression of the clinical stage (Fig. 1d) in breast cancer patients. We also used qRT-PCR to measure the expression levels of miR-615-3p in breast cancer cell lines. BT-549, MDA-MB-468, MCF-7, MDA-MB-231 and a non-malignant breast epithelial cell MCF-10A were used. Compared to low metastatic MCF-7 cells, the expression of miR-615-3p was obviously increased in highly metastatic cells, while the expression of miR-615-3p was significantly decreased in non-malignant breast epithelial cell MCF-10A (Fig. 1e). These results suggest that the miR-615-3p level is upregulated in breast cancer tissues and breast cancer cell lines and that it is positively correlated with the metastatic ability of breast cancer cells.

To evaluate the effect of miR-615-3p on MCF-7 cells, which have low endogenous miR-615-3p expression, we ectopically expressed miR-615-3p in MCF-7 cells. Ectopic miR-615-3p did not promote MCF-7 cell growth (data not shown). However, the Transwell migration assay and Matrigel invasion assay showed that ectopic expression of miR-615-3p increased migration and invasion of MCF-7 cells significantly (Fig. 2a). To investigate whether silencing of miR-615-3p expression in metastatic cells impedes their ability to metastasize, we knocked down endogenous miR-615-3p levels in MDA-MB-231 cells. MDA-MB-231 cells were infected with lentiviral vectors to establish two stable lines that either expressed antisense RNA against miR-615-3p (Anti-miR-615-3p) or contained a control vector (Anti-NC). Transwell migration assay and Matrigel invasion assay showed that knockdown of miR-615-3p decreased migration and invasion of MDA-MB-231 cells significantly (Fig. 2b). The transfection efficiency of MCF-7(miR-615-p overexpression) and MDA-MB-231(anti-miR-615-p) was shown in Fig. 2f.

To investigate whether miR-615-3p promotes metastasis in vivo, we performed intravenous injection through the tail vein with the same cells. In the group injected with MCF-7-miR-615-3p cells, all 10 mice developed lung metastasis, with a large number of metastatic nodules covering the entire lung. In contrast, only 1 out of 10 mice injected with MCF-7-NC cells developed lung metastasis, with a small number of metastatic nodules (Fig. 2c). Therefore, miR-615-3p expression promoted the colonization of circulating MCF-7 cells in the lung. Similarly, equal numbers of MDA-MB-231-Anti-NC and MDA-MB-231-Anti-miR-615-3p were intravenously injected through the tail vein. In the control group (MDA-MB-231-Anti-NC), all 10 mice (10/10) developed lung metastasis, which is consistent with the high metastatic potential of MDA-MB-231. miR-615-3p knockdown significantly reduced lung metastasis of MDA-MB-231 cells. In the MDA-MB-231-Anti-miR-615-3p group, only 2 out of 10 (2/10) mice developed lung metastasis (Fig. 2c). Thus, miR-615-3p knockdown suppressed metastasis of MDA-MB-231 cells in vivo.

As an EMT-like process has been associated with breast cancer metastasis [25], we examined whether miR-615-3p promotes EMT. We measured the levels of the epithelial marker E-cadherin and the mesenchymal marker Vimentin in MCF-7-miR-615-3p and MCF-7-NC cells using both western blot analysis and immunofluorescence. Compared with the control of NC cells, which had high E-cadherin expression and low Vimentin expression, MCF-7-miR-615-3p cells had significantly downregulated E-cadherin expression and upregulated Vimentin expression (Fig. 2d and e, left panel). Consistent with the EMT markers, SW480-miR-615-3p cells adopted a spindle-shaped, mesenchymal-like morphology in contrast to the epithelial-like morphology of SW480-NC cells (data not shown). Hence ectopic miR-615-3p expression causes MCF-7 cells to undergo EMT. We then examined whether the silencing of miR-615-3p could impede the ability of cells to undergo EMT. miR-615-3p knockdown reduced Vimentin expression and upregulated E-cadherin expression in MDA-MB-231 cells, as shown by immunofluorescence and western blot (Fig. 2d and e, right panel). Therefore, miR-615-3p knockdown impedes the ability of MDA-MB-231 cells to undergo an EMT-like process.

Collectively, the in vitro migration and invasion assays and in vivo metastasis assays, with ectopic expression of miR-615-3p in MCF-7 cells and knockdown of endogenous miR-615-3p in MDA-MB-231 cells, indicate that miR-615-3p promotes breast cancer metastasis.

To determine the molecular mechanism of miR-615-3p in the regulation of breast cancer cell migration, we analyzed the effect of miR-615-3p on the TGF-β1 signal pathway which is a critical EMT inducer in a number of cancer cells including breast cancer cells [26, 27].TGF-β1 treatment induced phosphorylation of Smad2 and Smad3 in a time-dependent manner [28, 29]. However, knockdown of miR-615-3p suppressed this phosphorylation (Fig. 3a). In contrast, overexpression of miR-615-3p resulted in a significant increase in TGF-β1-induced phosphorylation of Smad2 and Smad3 (Fig. 3b), suggesting that miR-615-3p could function as a positive regulator of Smad2 and Smad3 activation in response to TGF-β1.

Although TGF-β1 treatment significantly increased the migratory and invasive potentials of MDA-MB-231 cells, however, knockdown of miR-615-3p inhibited TGF-β1-induced cell migration (Fig. 3c). In contrast, overexpression of miR-615-3p resulted in a significant increase in TGF-β1-induced cell migration and invasion (Fig. 3c), suggesting that miR-615-3p could be involved in the regulation of TGF-β1-mediated responses.

We next investigated whether SB431542, a specific TβRI inhibitor, modulates miR-615-3p-induced cell migration and invasion. Treatment of MCF-7 cells with SB431542 significantly suppressed both basal and miR-615-3p-induced cell migration and invasion of MCF-7 cells (Fig. 3d), suggesting that miR-615-3p could be involved in the regulation of TβRI-mediated responses.

The TGF-β receptor internalization effectively decreasing TGF-β1 signaling [30, 31], we next investigated whether miR-615-3p regulates plasma-membrane expression of TβRI. Western blot analysis revealed that miR-615-3p knockdown in MDA-MB-231 by siRNA decreased the expression level of TβRI, but not that of TβRII (Fig. 3e). In contrast, ectopic expression of miR-615-3p in MCF-7 cells increased the expression level of TβRI, but not that of TβRII (Fig. 3e). However, the TβRI mRNA level remained unchanged (data not shown), suggesting that miR-615-3p may modulate TβRI stability.

To understand the mechanisms by which miR-615-3p promotes breast cancer cell migration, several computational methods were used to help identify miR-615-3p targets in humans. As miR-615-3p increased Smad2 and Smad3 phosphorylation, we hypothesized that its potential targets should decrease modulate TβRI stability. Among those potential targets, protein interacting with C-kinase-1(PICK1) is of particular interest because previously PICK1 was shown to antagonize transforming growth factor-beta (TGF-β) signaling by targeting TGF-β type I receptor (TβRI) for degradation and acted as an important negative regulator of TGF-β signaling [21, 32]. Therefore, we decided to carry out a series of experiments to determine whether PICK1 is a direct target of miR-615-3p. First, we detected the effect of miR-615-3p on PICK1 expression. Although miR-615-3p overexpression did not change PICK1 mRNA levels (Fig. 4a), there was a clear reduction in PICK1 protein levels in MCF-7 and T47D cells (Fig. 4b, upper panel). However, miR-615-3p silencing caused an increase in PICK1 protein levels in MDA-MB-468 and MDA-MB-231 cells (Fig. 4b, lower panel). These findings suggest that miR-615-3p targets PICK1 through translational inhibition. Next, we generated wild-type and mutant PICK1–3′-UTR expression plasmids that were fused to a luciferase reporter according to the matched sequence between PICK1–3′-UTR and miR-615-3p (Fig. 4c). We found miR-615-3p overexpression significantly inhibited luciferase activity of wild-type but not mutant reporter genes in MCF-7 cells (Fig. 4d, left). However, miR-615-3p silencing specifically enhanced the luciferase activity of wild-type but not mutant PICK1–3′-UTR (Fig. 4d, right).

We further found that knockdown of PICK1 produced similar changes in invasion and migration assay to that of miR-615-3p overexpression (Fig. 5a). To determine whether these effects depend specifically on PICK1 suppression, we used an expression construct that encodes the entire PICK1 coding sequence but lacks the 3′UTR. Unsurprisingly, the re-expression of PICK1 completely inhibited miR-615-3p–mediated invasion and migration (Fig. 5a). This suggests that, after miR-615-3p overexpression, a decrease in PICK1 is required for cells to show increased invasiveness and migration.

Then we ask whether PICK1 connects miR-615-3p and EMT. The loss of E-cadherin is the hallmark of EMT. Interestingly, MCF-7 cells changed from round to a spindle-like mesenchymal phenotype after knockdown of PICK1 (Fig. 5b), which mimic the effect of miR-615-3p on EMT. Furthermore, knockdown of PICK1 also significantly reduced the E-cadherin but increased the Vimentin levels (Fig. 5c).

The TGF-β signaling network plays an essential role in the EMT of breast cancer cells. To determine whether PICK1 regulation of EMT is functionally associated with the TGFβ signaling pathway which is important for miR-615-3p-mediated EMT, we analyzed the expression levels of TβRI and TβRII in MCF7 and T47D cells with PICK1 knockdown. Among the TβRs, the amount of TβRI significantly increased in MCF7 and T47D cells on PICK1 knockdown, though little change was detected in TβRII (Fig. 5c). These results clearly show that PICK1 regulates TβRI expression in breast cancer cells.

Taken together, these data indicated that the ability of miR-615-3p to promote metastasis is attributable, in significant part, to its capacity to inhibit PICK1.

We next examined the clinical relevance of altered miR-615-3p and PICK1 expression in human breast cancers. We found that the levels of the PICK1 mRNA were lower in breast cancer tissues than in non-cancerous tissues (Fig. 6a) and correlated with lymphatic metastasis and the histological grade of the tumor (Fig. 6b and c). We also investigated the relationship between miR-615-3p expression and the mRNA and protein levels of PICK1 in breast cancer patients. The level of miR-615-3p was inversely correlated with those of the PICK1 mRNA (Fig. 6d). We measured PICK1 protein levels in human breast cancer and matched non-cancerous tissues using immunohistochemical staining (Fig. 6e). The level of miR-615-3p was inversely correlated with those of the PICK1 protein levels (Fig. 6f). suggesting that miR-615-3p promotes the malignant phenotypes of breast cancer by targeting PICK1.

Given that SMAD effectors of TGF-β1 signaling can interact with DICER1 to promote primary miR-21 processing into precursor miR-21 [33], We hypothesize that overexpression of PICK1 disrupts the association of p-SMAD2 and p-SMAD3 with DICER1 and thus decreases the processing of pre-miR-615-3p to mature miR-615-3p in breast cancer cells. To assess this hypothesis, we transfected MDA-MB-231 cells with PICK1 overexpression or control vectors, and then performed co-immunoprecipitation–RT-qPCR studies using an antibody that targets SMAD2/3. We found that DICER1 co-precipitated along with SMAD2/3 in the control group, but this binding was significantly decreased in MDA-MB-231 cells with PICK1 overexpression (Fig. 7a). After RNA bound to the protein complexes was isolated, a significant decrease of pre-miR-615-3p in MDA-MB-231 cells with PICK1 overexpression was observed (Fig. 7b). When the anti-SMAD2/3 antibody was substituted by a non-specific IgG in control assays, pre-miR-615-3p was absent of the immunoprecipitated complexes (data not shown). RT-qPCR studies showed a 5.1 fold decrease of mature miR-615-3p in MDA-MB-231 cells with PICK1 overexpression compared to control (Fig. 7c). Together, our data suggest that PICK1 inhibits the binding of DICER1 to Smad2/3 and the processing of pre-miR-615-3p to mature miR-615-3p in breast cancer cells, thus exerting a negative feedback loop (Fig. 7d).