Deregulated molecular signaling pathways are responsible for the altered adhesive, migratory, and invasive properties of cancer cells. The different breast cancer subtypes are characterized by the expression of distinct miRNAs, short non-coding RNAs that posttranscriptionally modulate the expression of entire gene networks. Profiling studies have revealed downregulation of miR149 in basal breast cancer. Here, we show that miR149 expression severely impairs cell spreading, migration, and invasion of basal-like breast cancer cells. We identify signaling molecules, including the small GTPases Rap1a and Rap1b, downstream of integrin receptors as miR149 targets, providing an explanation for the defective Src and Rac activation during cell adhesion and spreading upon miR149 expression. Suppression of cell spreading by miR149 could be rescued, at least in part, by expression of constitutively active Rac. Finally, we demonstrate that increased miR149 levels block lung colonization in vivo. On the basis of our findings, we propose that miR149 downregulation in basal breast cancer facilitates the metastatic dissemination of tumor cells by supporting aberrant Rac activation. Cancer Res; 74(18); 5256–65. ©2014 AACR.

Cancer is one of the leading causes of death worldwide, mainly due to the development of metastases at later stages of the disease. Metastasis of cancer cells is a stepwise process starting with local invasion, intravasation, survival in the circulation, followed by extravasation and colonization of the target organ (1). In breast cancer, the basal subtype is characterized by its frequent reoccurrence and high metastatic potential. Basal breast cancers constitute approximately 20% of all breast cancer cases and they typically lack expression of the estrogen and progesterone receptors and ErbB2/HER2 amplification/overexpression (2, 3). The identification of molecular changes and aberrantly activated signaling pathways is, therefore, a prerequisite for the development of anticancer therapies targeting this specific breast cancer subtype.

Small GTPases of the Rho family are important modulators of cellular actin dynamics, the deregulation of which is associated with virtually every step of the metastatic cascade (4). For example, the Rac subfamily comprising Rac1, Rac2, and Rac3 is linked to the formation of lamellipodia and membrane ruffles and required for directed migration (5). The formation and maintenance of cell junctions and polarization are dependent on Rac and its hyperactivation disrupts epithelial architecture (5). Although there are reports on the overexpression of Rac isoforms in different cancers including those of the breast, the most common cause of Rac hyperactivation is the deregulation of upstream molecules (6). Rac is active when bound to GTP and inactive in the GDP-bound state. GTP loading requires the guanine exchange factors (GEF), whereas inactivation is facilitated by the GTPase-activating proteins, which accelerate the intrinsic GTPase activity, thus leading to inactivation. Rac is furthermore negatively regulated by guanine dissociation inhibitors that sequester the protein in the cytoplasm (7). Several Rac-specific GEFs such as Tiam, pREX-1, and Ect2 play important roles in tumorigenesis and metastasis and are upregulated or hyperactive in cancer (6).

Rac activation occurs downstream of integrins, transmembrane receptors that link the extracellular matrix (ECM) with the intracellular actin cytoskeleton at sites termed focal adhesions. Upon ECM engagement, integrins recruit the cytoplasmic tyrosine kinases FAK and Src which together phosphorylate various focal adhesion components and contribute to Rac activation (8). Integrin activation is potentiated by the small GTPase Rap1 which facilitates the membrane recruitment of the cytoskeletal adaptor protein talin. Upon integrin binding, talin induces a conformation change whereby integrins are converted to their high-affinity state (9). By establishing the connection between the cell and the ECM, integrins are essential for epithelial tissue organization and they are also directly involved in the adhesive migration and invasion of tumor cells (10).

miRNAs are a class of short non-coding RNAs of approximately 17 to 24 nucleotides that regulate gene expression at the posttranscriptional level (11). miRNAs act by binding to their target mRNAs usually within the 3′-untranslated region, resulting in translational repression and/or mRNA degradation (12). Because miRNAs act by partial complementary binding of relatively short sequences they target multiple mRNAs, thereby regulating entire gene networks (13). By doing so, miRNAs orchestrate fundamental biologic cellular processes such as development, proliferation, and apoptosis. Furthermore, miRNAs are implicated in various diseases including cancer, making them attractive targets for therapeutic intervention or as diagnostic markers (14, 15). In recent years, miRNA profiling studies undertaken in different tumor types have identified sets of miRNAs with altered expression in tumor versus normal tissue. Although certain upregulated miRNAs, so-called oncoMiRs, have been characterized to possess transforming potential, there are also examples of miRNAs with inherent tumor-suppressive activity that are downregulated in cancer cells (14, 15). Of those miRNAs specifically associated with metastatic progression, some target key regulators of cytoskeletal reorganization. For example, miR31 whose expression levels correlate inversely with disease relapse in patients with breast cancer was found to promote metastasis via the combined suppression of integrin-α5, the adaptor protein radixin and the small GTPase RhoA (16).

The different molecular breast cancer subtypes are not only characterized by their distinct gene expression profiles but also by specific miRNA signatures (17, 18). In a recent study, microarray analyses of primary human breast tumors identified 26 miRNAs that separated almost perfectly the basal-like and luminal A samples (18). Decreased expression of miR149 in particular further correlated with mutant p53 status (18), and miR149 was also observed to be downregulated in prostate cancer (19), but, thus, far functional analyses of miR149 in these tumor types are still lacking. Using a combination of biochemical and cellular assays, we show here that miR149 interferes at multiple levels with signaling downstream of integrin receptors, impairs Rac activation, and efficiently blocks basal-like breast cancer cell migration and invasion both in vitro and in vivo. Suppression of Rac activation by miR149 associated with defective cell spreading and migration was also observed in prostate cancer cells, implicating a conserved tumor-suppressive function for miR149.

Cell culture and transfection

MCF7 and T47D cells were obtained from Cornelius Knabbe (Institute of Clinical Pharmacology, Stuttgart, Germany), BT474 cells from Nancy Hynes (Friedrich Miescher Institute, Basel, Switzerland), and MDA-MB-468, Hs578T, and MDA-MB-436 cells from Bernhard Lüscher (RWTH Aachen University, Germany) in 2005. Cell lines were authenticated by morphology and growth characteristics, tested for Mycoplasma and frozen, and cultured for less than 4 months. MDA-MB-231, SKBR3, and BT549 cells were obtained from CLS in 2010, 2011, and 2014, respectively, and not reauthenticated. MDA-MB-231 and SKBR3 cells were cultured in DMEM (Life Technologies), all other cell lines in RPMI-1640 (Life Technologies) supplemented with 10% FCS (PAA). Cells were transiently transfected with miRNA and siRNAs using RNAiMax (Life Technologies). miR149-5p mimic (miR149) and miRNA-negative control#1 (miR-con) were from Thermo Scientific, Rap1a (#1: s11780; #2: ss11781) and Rap1b (#1: s11782; #2: s224515). Silencer Select siRNAs were from Life Technologies and siControl (ON-TARGETplus non-targeting pool) was from Dharmacon. Plasmid transfection was performed using Lipofectamine LTX with Plus Reagent (Life Technologies). The pEGFP-Rac G12V plasmid was a gift from Francisco Sanchez-Madrid (Centro Nacional de Investigaciones Cardiovasculares, Madrid, Spain).

Cell lysis, SDS–PAGE, and Western blotting

Cells were lysed in RIPA buffer [50 mmol/L Tris (pH 7.5), 150 mmol/L NaCl, 1% Triton-X-100, 0.5% sodium deoxycholate, 0.1% SDS, 1 mmol/L sodium orthovanadate, 10 mmol/L sodium fluoride, and 20 mmol/L β-glycerophosphate plus Complete protease inhibitors (Roche)]. Lysates were clarified by centrifugation at 16,000 × g for 10 minutes. Proteins were separated by SDS–PAGE and transferred to nitrocellulose membrane using the iBlot Gel Transfer Device (Invitrogen). Membranes were blocked with 0.5% blocking reagent (Roche) in PBS containing 0.1% Tween-20 and then incubated with primary antibodies, followed by horseradish peroxidase–conjugated secondary antibodies. Visualization was carried out using the ECL detection system (Thermo Fisher).

qRT-PCR

Total RNA was isolated from 5 × 105 to 1 × 106 cells using the mirVANA miRNA Isolation Kit (Life Technologies). qRT-PCR was performed with a Cfx96 device (Bio-Rad). Reverse transcription and qPCR of miR149 and RNU6B were performed with TaqMan microRNA Assays (Life Technologies). Relative miRNA expression levels were calculated with the |$2^{ - \Delta \Delta C_{\rm t} }$| method (Bio-Rad CFX manager software 3.1.). Vav2, Rap1a, Rap1b, and GAPDH mRNAs (QuantiTect Primer Assay, Qiagen) were amplified using the QuantiTect SYBR Green PCR Kit (Qiagen). Relative expression levels were calculated using the |$2^{ - \Delta \Delta C_{\rm t} }$| method (Bio-Rad CFX manager software 3.1.).

Rac activity assay

Rac activity was assessed with the G-LISA Rac1/2/3 Activation Assay Biochem Kit (Cytoskeleton). Absorbance was measured with the multimode reader Infinite 200 PRO.

Immunofluorescence microscopy

Cells seeded onto glass coverslips coated with 10 μg/mL collagen (Serva) were fixed with 4% paraformaldehyde (PFA), permeabilized with PBS containing 0.1% Triton X-100, and blocked with 5% goat serum (Life Technologies) in PBS containing 0.1% Tween-20. Cells were then incubated with primary antibodies in blocking buffer, washed with PBS containing 0.1% Tween-20, and incubated with secondary antibodies in blocking buffer. Slides were mounted in Fluoromount-G and analyzed on a confocal laser scanning microscope (LSM 700, Zeiss) using 488, 561, and 633nm excitation with the objectives Plan-Apochromat 20x/0.8 M27 and oil objective lenses EC Plan-Neofluar 40/1.30 DIC M27 and Plan-Apochromat 63/1.40 DIC M27. Images were processed with the ZEN software.

Impedance measurement

A total of 1 × 104 cells were plated into collagen-coated (10 μg/mL) 96-well E-plates (Roche) in medium containing 0.5% FCS, except for MDA-MB-436 cells, which were plated in medium containing 10% FCS. The impedance of cells was measured using the xCELLigence device (Roche).

Life cell imaging

Three days after transfection, 1 × 104 cells were seeded onto collagen-coated (10 μg/mL) glass bottom dishes (Greiner Bio-One) in medium containing 10% FCS and were allowed to adhere for 5 hours before imaging with a wide-field fluorescence microscope (Axio Cell Observer Z1, Zeiss) using the objective Plan-Apochromat 20x/0.8 M27. Raw data were exported to ImageJ and quantification was done with the ImageJ plugin MTracjJ.

Proliferation assays

One day after transfection, 2.5 × 103 cells were plated into 96-well plates. The next day, medium was replaced by fresh medium containing 10% FCS. Three and 5 days after transfection, cells were fixed with 4% PFA and stained with crystal violet. Crystal violet was dissolved in methanol and measured at 550 nm with the multimode reader Infinite 200 PRO (Tecan).

Migration and invasion assays

For haptotactic migration, 0.5 × 105 cells were seeded in medium containing 0.5% FCS into Transwells (8-μm pore size; Costar) coated with 10 μg/mL collagen on the underside and allowed to migrate for 3 hours (MDA-MB-231) or 4 hours (MDA-MB-436, BT549, Hs578T). The bottom chamber contained medium with 0.5% FCS. For invasion, Transwells were coated on the upper side with 50 μL growth factor–reduced Matrigel (BD Biosciences) diluted in medium containing 0.5% FCS and cells (0.5 or 1 × 105 in medium containing 0.5% FCS) were allowed to invade overnight. The bottom chamber contained medium with 10% FCS supplemented with 190 ng/mL EGF. In the case of MDA-MB-436 cells, the bottom chamber contained medium with 0.5% FCS supplemented with 190 ng/mL EGF. Cells on the underside of the membrane were fixed, stained with crystal violet, and counted (five independent microscopic fields at a 20-fold magnification or all cells/filter).

Animal experiments

MDA-MB-231 cells (0.5 × 106 in 100 μL HBSS) were injected into the tail veins of 8-week-old female CB17 SCID mice (Harlan Laboratories GmbH). Four weeks later, mice were sacrificed, and the lungs were fixed and stained in Bouin solution. The number of macro-metastases on the lung surface was counted. Animal care and experiments were in accordance with federal guidelines and approved by the University and state authorities.

Reanalysis of microarray-based miRNA expression data of 101 primary breast tumors revealed reduced expression of miR149 in basal compared with luminal A/B, ErbB2/HER2 positive and normal-like cancers (Fig. 1A; ref. 18). This was confirmed in an independent dataset (17), in which miR149 expression was also significantly reduced in basal-like (n = 16) compared with luminal A/B (n = 24) breast tumors (Fig. 1B). Moreover, the expression of miR149 further inversely correlated with higher tumor stage (Fig. 1C). We first sought to verify that this expression pattern is reflected also in established breast cancer cell lines. Indeed, qRT-PCR analysis showed that in five basal-like cell lines (MDA-MB-231, MDA-MB-468, BT549, Hs578T, and MDA-MB-436), miR149 levels were lower compared with those in the luminal MCF7 and T47D and the HER2-positive BT474 and SKBR3 cell lines (Fig. 1D). We chose MDA-MB-231 cells, a well-characterized invasive basal B-cell line, for our initial studies. Overexpression of a mature miR149 mimic did not have any impact on cell proliferation (Fig. 2A) demonstrating that expression as such is not toxic to cells. However, in Transwell assays measuring haptotactic migration toward a collagen gradient, miR149 strongly reduced the number of migrated cells (Fig. 2B) and in the presence of a Matrigel layer, cell invasion toward a serum plus EGF gradient was almost completely blocked (Fig. 2C). Tracking of single cells revealed that general cell motility was not affected by miR149 overexpression, because the distance (track length) and, thus, the velocity were similar for control and miR149-transfected cells; however, the net displacement was reduced by miR149 expression (Fig. 2D and E), indicating a defect in the directionality of cell movement. Suppression of directed cell migration and invasion by miR149 was further confirmed in Hs578T, MDA-MB-436, and BT549 cell lines (Fig. 3).

Figure 1.

Reduced expression of miR149 in basal breast cancer. A, boxplot visualization of miR149 expression in different tumor subtypes based on datasets by Enerly et al. (18). Data, shown as the mean ± SEM, were analyzed by two-way ANOVA followed by Bonferroni post-test; **, P < 0.01. B, boxplot visualization of miR149 expression in luminal A/B (n = 24) and basal-like (n = 16) breast cancers based on datasets by Blenkiron et al. (17); ***, P = 0.00015, one-sided t test. C, tumor stage classification of breast cancers shown in B compared with the expression of miR149 (17). D, RNA was extracted from the indicated breast cancer cell lines, and expression levels of miR149 were determined by qRT-PCR. Data were normalized to levels of RNU6B as an internal control; MCF7 cells were set as 1.A representative experiment performed with triplicate samples is shown.

Figure 1.

Reduced expression of miR149 in basal breast cancer. A, boxplot visualization of miR149 expression in different tumor subtypes based on datasets by Enerly et al. (18). Data, shown as the mean ± SEM, were analyzed by two-way ANOVA followed by Bonferroni post-test; **, P < 0.01. B, boxplot visualization of miR149 expression in luminal A/B (n = 24) and basal-like (n = 16) breast cancers based on datasets by Blenkiron et al. (17); ***, P = 0.00015, one-sided t test. C, tumor stage classification of breast cancers shown in B compared with the expression of miR149 (17). D, RNA was extracted from the indicated breast cancer cell lines, and expression levels of miR149 were determined by qRT-PCR. Data were normalized to levels of RNU6B as an internal control; MCF7 cells were set as 1.A representative experiment performed with triplicate samples is shown.

Close modal
Figure 2.

Expression of miR149 impairs directed cell migration and invasion. MDA-MB-231 cells were transiently transfected with control miRNA (miR-con) or miR149. A, one day after transfection, cells were plated into 96-well plates and stained with crystal violet after 3 and 5 days, respectively. A representative experiment is shown. B and C, three days post-transfection, cells were harvested and subjected to Transwell assays. B, for haptotactic migration, the bottom side of the filters was coated with collagen. C, for invasion, Transwells were coated with Matrigel, and cells were left to migrate overnight toward a serum and EGF gradient. Shown is the mean ± SEM of three independent experiments. Data were analyzed using the Student t test (unpaired, two tailed); **, P < 0.01. D and E, three days after transfection, cells were plated onto collagen-coated glass bottom dishes and subjected to time-lapse microscopy. Images were taken every 10 minutes over a time period of 5 hours. D, representative trajectories of 10 cells for each condition are plotted. Axis origin refers to the starting point of cell tracking. E, quantitative analysis of cell displacement, cell velocity, and track length. Data correspond to a representative experiment with ≥25 cells per condition and are reported as mean ± SEM. Statistical significance was analyzed using the Student t test (unpaired, two tailed). ***, P < 0.001; n.s. (nonsignificant), P > 0.05.

Figure 2.

Expression of miR149 impairs directed cell migration and invasion. MDA-MB-231 cells were transiently transfected with control miRNA (miR-con) or miR149. A, one day after transfection, cells were plated into 96-well plates and stained with crystal violet after 3 and 5 days, respectively. A representative experiment is shown. B and C, three days post-transfection, cells were harvested and subjected to Transwell assays. B, for haptotactic migration, the bottom side of the filters was coated with collagen. C, for invasion, Transwells were coated with Matrigel, and cells were left to migrate overnight toward a serum and EGF gradient. Shown is the mean ± SEM of three independent experiments. Data were analyzed using the Student t test (unpaired, two tailed); **, P < 0.01. D and E, three days after transfection, cells were plated onto collagen-coated glass bottom dishes and subjected to time-lapse microscopy. Images were taken every 10 minutes over a time period of 5 hours. D, representative trajectories of 10 cells for each condition are plotted. Axis origin refers to the starting point of cell tracking. E, quantitative analysis of cell displacement, cell velocity, and track length. Data correspond to a representative experiment with ≥25 cells per condition and are reported as mean ± SEM. Statistical significance was analyzed using the Student t test (unpaired, two tailed). ***, P < 0.001; n.s. (nonsignificant), P > 0.05.

Close modal
Figure 3.

Expression of miR149 impairs invasive migration of Hs578T, MDA-MB-436, and BT549 cells. Cells were transiently transfected with miR-con or miR149. A, for haptotactic migration, the bottom side of the filters was coated with collagen. B, for invasion, Transwells were coated with Matrigel, and cells were left to migrate overnight. Shown is the mean cell number ± SEM of a representative experiment; in the case of BT549 cell invasion, photos of representative filters are shown.

Figure 3.

Expression of miR149 impairs invasive migration of Hs578T, MDA-MB-436, and BT549 cells. Cells were transiently transfected with miR-con or miR149. A, for haptotactic migration, the bottom side of the filters was coated with collagen. B, for invasion, Transwells were coated with Matrigel, and cells were left to migrate overnight. Shown is the mean cell number ± SEM of a representative experiment; in the case of BT549 cell invasion, photos of representative filters are shown.

Close modal

Cell adhesion and spreading are prerequisites for matrix-dependent cell migration. To study this in more detail, we transiently transfected MDA-MB-231 cells with control miRNA and miR149, respectively, and plated the cells onto collagen-coated dishes for different periods of time (Fig. 4A). Although control cells were fully spread 4 hours after plating, miR149-expressing cells had a rounder morphology and after 24 hours, they were still less spread (Fig. 4A). To measure cell adhesion and spreading in real-time, we plated cells onto collagen-coated E-plates and analyzed the impedance (cell index) using the xCELLigence device. Figure 4B shows that the plateau reached by miR149-expressing cells was lower than that of the control cells. The number of adherent cells was unaffected by miR149 expression (data not shown), indicating a defect in cell spreading rather than cell adhesion. Similar results were obtained for Hs578T, MDA-MB-436, and BT549 cell lines (Supplementary Fig. S1). Next, we analyzed cellular architecture at the molecular level at early times of adhesion by phalloidin and vinculin staining. Compared with the control, miR149-expressing MDA-MB-231 cells exhibited a depolarized actin cytoskeleton and failed to establish prominent cell protrusions and lamellipodia (Fig. 4C). Finally, we examined biochemically how miR149 expression affected the activation of key signaling molecules downstream of integrin engagement by analyzing Src and paxillin phosphorylation. In accordance with the impaired spreading phenotype in miR149-expressing cells, Src phosphorylation was strongly reduced especially at early times of adhesion and spreading, whereas paxillin phosphorylation, an indicator of focal adhesion formation and turnover (20), was still low 4 hours after plating onto collagen (Fig. 4D).

Figure 4.

miR149 interferes with cell spreading on collagen. MDA-MB-231 cells were transiently transfected with miR-con or miR149. Three days post-transfection, cells were harvested and plated onto collagen in 0.5% FCS. A, cells were fixed and stained with crystal violet at the indicated times. B, cell adhesion and spreading on collagen-coated E-plates were measured using the xCELLigence device. Values correspond to the mean ± SEM of duplicate samples. C, cells plated on collagen-coated coverslips for the indicated times were stained with phalloidin and an anti-vinculin antibody. The confocal images shown are stacks of three sections taken from the bottom of the cell. D, cells seeded onto collagen-coated dishes were lysed at the indicated times, and lysates were immunoblotted using the indicated antibodies. The blots were cropped, and in each case, the corresponding panels are from the same gel.

Figure 4.

miR149 interferes with cell spreading on collagen. MDA-MB-231 cells were transiently transfected with miR-con or miR149. Three days post-transfection, cells were harvested and plated onto collagen in 0.5% FCS. A, cells were fixed and stained with crystal violet at the indicated times. B, cell adhesion and spreading on collagen-coated E-plates were measured using the xCELLigence device. Values correspond to the mean ± SEM of duplicate samples. C, cells plated on collagen-coated coverslips for the indicated times were stained with phalloidin and an anti-vinculin antibody. The confocal images shown are stacks of three sections taken from the bottom of the cell. D, cells seeded onto collagen-coated dishes were lysed at the indicated times, and lysates were immunoblotted using the indicated antibodies. The blots were cropped, and in each case, the corresponding panels are from the same gel.

Close modal

miR149 downregulation has been reported in other cancer types, including prostate cancer (19). To explore whether the cellular effects induced by miR149 are conserved in a different cell type, we transfected PC3 prostate cancer cells, which also express very low levels of endogenous miR149 (data not shown), with the control miRNA and miR149 mimic. As seen in MDA-MB-231 cells, upon miR149 transfection, cell spreading on collagen and haptotactic cell migration were severely impaired (Supplementary Fig. S2).

The activation of the small GTPase Rac downstream of integrin engagement is critical for cell spreading, lamellipodia formation, and directed migration. We, therefore, investigated Rac activation by GLISA and found that increased miR149 expression resulted in reduced Rac-GTP levels, preventing the transient increase in Rac activation observed after plating onto collagen (Fig. 5A) or upon EGF stimulation of cells (Fig. 5B), whereas total Rac protein levels were unaffected by miR149 expression (Supplementary Fig. S3). The suppression of MDA-MB-231 cell migration by pharmacologic Rac inhibition in Transwell assays confirmed the importance of Rac for the directed migration of these cells (Fig. 5C). Rac activation was also compromised in PC3 cells (Supplementary Fig. S2), demonstrating that miR149 impacts the same signaling pathway in these cells. To investigate the functional contribution of Rac inactivation in the context of miR149 signaling, we transiently transfected cells expressing either control miRNA or miR149 with a vector encoding constitutively active GFP-Rac G12V. Expression of active Rac, but not GFP alone, rescued the cell spreading defect of miR149-expressing cells (Fig. 5D and E).

Figure 5.

miR149 impairs Rac activation, and active Rac restores spreading of miR149 expressing cells. A, B, and D, MDA-MB-231 cells were transiently transfected with miR-con or miR149, respectively. A, three days post-transfection, cells were seeded in 0.5% FCS onto collagen-coated dishes and lysed at the indicated times. B, three days post-transfection, cells were stimulated with 50 ng/mL EGF for indicated times and lysed. Rac-GTP levels were measured by GLISA and normalized to miR-con at time 0. Data, mean ± SEM of three experiments (A) and analyzed using two-way ANOVA with Bonferroni post-tests; *, P < 0.05, n.s. (nonsignificant). In B, the mean ± SEM of two experiments is shown. C, Transwell migration of MDA-MB-231 cells was measured in the absence (control) and presence of 100 μM NSC23766 in the top chamber. D, two days post-transfection, cells were transiently transfected with a vector encoding GFP alone or GFP-Rac G12V. The next day, cells were replated onto collagen-coated coverslips for 1 hour, fixed, and stained with phalloidin and an anti-vinculin antibody. The confocal images shown are stacks of five sections taken from the bottom of the cell. E, for the quantitative analysis of the area of spread cells pictures (1,024 × 1,024 pixel) were taken at a 20-fold magnification. Analysis was performed with ImageJ. Only cells with a mean pixel intensity of more than 10 were considered as GFP positive. Twenty-two to 59 cells per condition were analyzed. Shown is the mean ± SEM of three independent experiments. Data were analyzed using two-way ANOVA with Bonferroni post-tests; **, P < 0.01, n.s. (nonsignificant).

Figure 5.

miR149 impairs Rac activation, and active Rac restores spreading of miR149 expressing cells. A, B, and D, MDA-MB-231 cells were transiently transfected with miR-con or miR149, respectively. A, three days post-transfection, cells were seeded in 0.5% FCS onto collagen-coated dishes and lysed at the indicated times. B, three days post-transfection, cells were stimulated with 50 ng/mL EGF for indicated times and lysed. Rac-GTP levels were measured by GLISA and normalized to miR-con at time 0. Data, mean ± SEM of three experiments (A) and analyzed using two-way ANOVA with Bonferroni post-tests; *, P < 0.05, n.s. (nonsignificant). In B, the mean ± SEM of two experiments is shown. C, Transwell migration of MDA-MB-231 cells was measured in the absence (control) and presence of 100 μM NSC23766 in the top chamber. D, two days post-transfection, cells were transiently transfected with a vector encoding GFP alone or GFP-Rac G12V. The next day, cells were replated onto collagen-coated coverslips for 1 hour, fixed, and stained with phalloidin and an anti-vinculin antibody. The confocal images shown are stacks of five sections taken from the bottom of the cell. E, for the quantitative analysis of the area of spread cells pictures (1,024 × 1,024 pixel) were taken at a 20-fold magnification. Analysis was performed with ImageJ. Only cells with a mean pixel intensity of more than 10 were considered as GFP positive. Twenty-two to 59 cells per condition were analyzed. Shown is the mean ± SEM of three independent experiments. Data were analyzed using two-way ANOVA with Bonferroni post-tests; **, P < 0.01, n.s. (nonsignificant).

Close modal

To obtain insight into the potential signaling molecules affected by miR149, we used the miRecords online tool (http://c1.accurascience.com/miRecords/prediction_query.php) that integrates eleven miRNA prediction programs, resulting in 325 unique targets predicted by at least four different algorithms. These genes were mapped on the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database (21), enabling the annotation of 79 genes. The pie chart in Fig. 6A depicts pathways that contain at least three annotated genes and comprises 49 genes in total (see Supplementary Table S1 for gene information). Notably, apart from “cytokine–cytokine receptor interaction,” “MAPK signaling pathway,” and “regulation of actin cytoskeleton,” “focal adhesion” contained the highest number of predicted miR149 target genes. The latter pathway includes the small GTPases Rap1a and Rap1b and the guanine nucleotide exchange factor Vav2. Importantly, Rac is known to be activated by Vav2 and Rap1 in a cooperative manner (22, 23).

Figure 6.

miR149 targets Rap1a and Rap1b, and Rap1a/b codepletion reduces cell migration and invasion as well as Rac activity. A, the pie chart depicts the mapping of 325 unique miR149 target genes onto KEGG pathways (Version 2.5). Human pathways containing at least three target genes are shown, comprising 49 genes in total. B, MDA-MB-231 cells were transiently transfected with miR-con or miR149, respectively. Two days post-transfection, RNA was extracted, and Vav2, Rap1a, and Rap1b transcript levels were determined by qRT-PCR. Values were normalized to GAPDH. Data, shown as the mean ± SEM of four independent experiments, were analyzed using two-way ANOVA with Bonferroni post-tests; ***, P < 0.001. C and D, MDA-MB-231 cells were transiently transfected with siControl or cotransfected with two independent siRNA sets specific for Rap1a and Rap1b (siRap1a+b #1; siRap1a+b #2). C, Transwell migration and invasion assays were performed as described in Figure 2. The mean ± SEM of three independent experiments is shown. Data were analyzed using one-way ANOVA with Newman–Keuls post-test; *, P < 0.05; ***, P < 0.001. D, three days post-transfection, Rac-GTP levels were measured by GLISA and normalized to siControl. The mean ± SEM of two independent experiments is shown. Cell lysates were analyzed by immunoblotting with the indicated antibodies.

Figure 6.

miR149 targets Rap1a and Rap1b, and Rap1a/b codepletion reduces cell migration and invasion as well as Rac activity. A, the pie chart depicts the mapping of 325 unique miR149 target genes onto KEGG pathways (Version 2.5). Human pathways containing at least three target genes are shown, comprising 49 genes in total. B, MDA-MB-231 cells were transiently transfected with miR-con or miR149, respectively. Two days post-transfection, RNA was extracted, and Vav2, Rap1a, and Rap1b transcript levels were determined by qRT-PCR. Values were normalized to GAPDH. Data, shown as the mean ± SEM of four independent experiments, were analyzed using two-way ANOVA with Bonferroni post-tests; ***, P < 0.001. C and D, MDA-MB-231 cells were transiently transfected with siControl or cotransfected with two independent siRNA sets specific for Rap1a and Rap1b (siRap1a+b #1; siRap1a+b #2). C, Transwell migration and invasion assays were performed as described in Figure 2. The mean ± SEM of three independent experiments is shown. Data were analyzed using one-way ANOVA with Newman–Keuls post-test; *, P < 0.05; ***, P < 0.001. D, three days post-transfection, Rac-GTP levels were measured by GLISA and normalized to siControl. The mean ± SEM of two independent experiments is shown. Cell lysates were analyzed by immunoblotting with the indicated antibodies.

Close modal

To assess whether Rap1a/b and Vav2 were targeted by miR149, we performed qRT-PCR analyses of mRNA samples extracted from MDA-MB-231 cells transfected with control miRNA and miR149, respectively. Indeed, compared with the control, the transcript levels of Rap1a/b and Vav2 were significantly reduced by miR149 expression (Fig. 6B). Of note, we did not observe any changes in the expression levels of PAK3, Src, or paxillin (Fig. 4D and data not shown), which were also predicted as targets and are associated with the focal adhesion pathway. We next analyzed the contribution of the identified miR149 targets to migration and invasion of MDA-MB-231 cells. Although the knockdown of Vav2 was not sufficient to inhibit cell motility in Transwell assays (data not shown), Rap1a/b codepletion significantly reduced both cell migration and invasion (Fig. 6C). Downregulation of Rap1a/b further reduced Rac-GTP levels without affecting total Rac protein levels (Fig. 6D), supporting the conclusion that miR149 controls invasive cell migration, at least in part, by targeting Rap1a/b.

Finally, to test the impact of miR149 expression on cell migration and invasion in vivo, we analyzed in immunocompromised mice the efficiency of lung colonization of MDA-MB-231 cells transfected with miR149 upon tail vein injection. Four weeks after injection, lungs were dissected and the number of macroscopic metastases counted, revealing a significant block in metastatic lung colonization by miR149 expression compared with the control (Fig. 7). Together, our data provide strong support for a tumor-suppressive role of miR149 in basal breast cancer.

Figure 7.

miR149 blocks lung colonization in vivo. A, MDA-MB-231 cells were transiently transfected with miR-con or miR149. Three days post-transfection, cells were harvested and injected into the tail veins of SCID mice (0.5 × 106 cells; 6 mice/group). Four weeks later, the mice were sacrificed, and the lungs were fixed in Bouin solution and photographed. B, quantification of the number of macro-metastases per lung. Data, mean ± SEM and were analyzed using the Student t test (unpaired, two-tailed); ***, P < 0.001.

Figure 7.

miR149 blocks lung colonization in vivo. A, MDA-MB-231 cells were transiently transfected with miR-con or miR149. Three days post-transfection, cells were harvested and injected into the tail veins of SCID mice (0.5 × 106 cells; 6 mice/group). Four weeks later, the mice were sacrificed, and the lungs were fixed in Bouin solution and photographed. B, quantification of the number of macro-metastases per lung. Data, mean ± SEM and were analyzed using the Student t test (unpaired, two-tailed); ***, P < 0.001.

Close modal

In this study, we show that miR149 expression potently suppressed migration and invasion of basal breast cancer cells in vitro and lung colonization of MDA-MB-231 cells in vivo. Because ectopic expression of miR149 did not impact proliferation, miR149 mainly has an antimetastatic function in these cells, which is in line with its downregulation in more advanced stages of breast cancer. In prostate cancer cells, miR149 expression recapitulated the effects seen in MDA-MB-231 cells, indicating the regulation of conserved pathways. Reduced miR149 expression has been reported in other types of cancer, including colon and gastric cancer (24–26). In this context, the transcription factors SP1, FOXM1, and ZBTB2 were identified as direct miR149 targets and their inhibition was proposed to be responsible for the tumor-suppressive function of miR149 (24–26). miR149, thus, seems to act as a general oncosuppressive miRNA by interfering with different proliferative and metastatic signaling pathways, the precise nature of which are dependent on the cell type.

Increased expression of miR149 in basal breast cancer and PC3 cells impaired cell spreading on both collagen (Fig. 4 and Supplementary Figs. S1 and S2) and fibronectin (data not shown) and, therefore, seems to act ECM-independently. Cell spreading is associated with the activation of Rac, which is also critical for persistent migration. Rac is known to drive the mesenchymal type of cancer cell motility characterized by cell polarization and the matrix-dependent establishment of cell protrusions at the leading edge (5). The compromised Rac activity seen in miR149-expressing cells, thus, provides an explanation for both the suppression of cell spreading and the impaired directionality of single cell motility and haptotactic migration in Transwell assays. Of note, miR149 did not inhibit migration of MDA-MB-468 cells (data not shown), which belong to the basal A subtype and retain epithelial characteristics (27), and thus likely depend on other mechanisms for cell migration. Overexpression of constitutively active Rac1 in MDA-MB-231 cells rescued the initial spreading defect induced by miR149 but was not sufficient to restore migration in Transwell assays. This can be explained by the fact that Rac hyperactivation generates multiple lamellipodia, thereby preventing polarization and inducing random motility (5). Because we did not obtain MDA-MB-231 cells stably expressing miR149, an orthotopic mouse model measuring metastasis formation from the primary tumor could not be investigated. Instead, we injected cells into the tail veins of mice, a procedure compatible with transient miRNA expression. Shortly after injection, cells are trapped in the lung capillaries where they extravasate and then colonize the lung (28). On the basis of the in vitro data, it can be assumed that miR149 expression blocks the first steps of this process that rely on cell adhesion and invasion. However, because miR149 overexpression was still detected 9 days post-transfection in vitro (Supplementary Fig. S4), we cannot rule out that miR149 also negatively affects additional factors that facilitate metastatic colonization of the lung.

Thus far, there is little evidence for the direct miRNA-mediated suppression of the individual Rac isoforms. miR142-3p was reported to target Rac1, regulating migration and invasion of hepatocellular carcinoma cell lines (29). Tiam as a positive upstream regulator of Rac was identified to be subject to miRNA regulation by miR10b, which suppressed cell migration and invasion of breast cancer cells (30). However, in colon cancer cells, downregulation of Tiam by miR21 and miR31 was associated with enhanced motility (31), highlighting the importance of the specific cell context. Interestingly, in head and neck squamous cell carcinoma cells, the loss of let-7i induced a morphologic switch to a mesenchymal, Rac-dependent mode of invasion. let-7i was identified to target Nedd9 and DOCK3, both of which are activators of Rac (32).

The latter report is in accordance with the notion that miRNA function is not mediated by a single target but rather corresponds to the net effect of the regulation of an entire set of targets. Here, we show that miR149 expression reduces the transcript levels of Rap1a, Rap1b, and Vav2, most likely by a direct mechanism that involves miR149 binding to the predicted recognition sites within the 3′UTRs. Rap1a/b codepletion suppressed migration and invasion of MDA-MB-231 cells and reduced basal Rac activity, providing support for their role in miR149 signaling. This agrees with the proposed function of Rap1 in chick embryo and mouse metastasis models of pancreatic cancer and melanoma, respectively (33, 34). Although silencing of Vav2 alone was not sufficient to reduce the invasive behavior of MDA-MB-231 cells in vitro, the finding that Rap1 and Vav2 cooperate in activating Rac (22,23) suggests that Vav2 suppression by miR149 may potentiate the biologic effects resulting from Rap1a/b downregulation. However, it is important to note that Rap1 and Vav2 possess additional Rac-independent functions associated with the metastatic process. Apart from its function in integrin signaling, Rap1 utilizes different effectors to control cell polarity and cell–cell adhesion, the deregulation of which contributes to neoplastic transformation (35). Similarly, Vav2 and Vav3 were recently shown to synergize in sustaining tumor growth, neoangiogenesis, and lung-specific metastasis of breast cancer cells, in part by Rho GTPase-independent pathways (36).

miRNAs are now being integrated into clinical trials as biomarkers for prognosis and clinical response, but they also hold promise as anticancer therapeutics. For example, MRX34, an intravenously injected liposome-formulated miR34 mimic has recently entered clinical trials for patients with advanced or metastatic liver cancer (37). The increased understanding of miRNA biology and function together with improved targeted delivery systems will hopefully also support the development of miRNA-based agents for the treatment of basal breast cancer.

No potential conflicts of interest were disclosed.

Conception and design: M.A. Olayioye

Development of methodology: A. Bischoff, B. Keller, S. Schmid

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A. Bischoff, B. Huck, B. Keller, M. Strotbek, S. Schmid

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Bischoff, B. Huck, M. Strotbek, M. Boerries, H. Busch, M.A. Olayioye

Writing, review, and/or revision of the manuscript: A. Bischoff, M. Boerries, H. Busch, D. Müller, M.A. Olayioye

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A. Bischoff, B. Keller, D. Müller

Study supervision: A. Bischoff, M.A. Olayioye

This work was supported by the German Cancer Aid (#110701) and the DFG (#571692). M.A. Olayioye was funded by the Heisenberg program of the DFG. H. Busch and M. Boerries were supported by the German Federal Ministry of Education and Research (CancerSys – LungSys II; 031P6846).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Nguyen
DX
,
Bos
PD
,
Massague
J
. 
Metastasis: from dissemination to organ-specific colonization
.
Nat Rev Cancer
2009
;
9
:
274
84
.
2.
Di Cosimo
S
,
Baselga
J
. 
Management of breast cancer with targeted agents: importance of heterogeneity. [corrected]
.
Nat Rev Clin Oncol
2010
;
7
:
139
47
.
3.
Hurvitz
SA
,
Finn
RS
. 
What's positive about ‘triple-negative' breast cancer?
Future Oncol
2009
;
5
:
1015
25
.
4.
Karlsson
R
,
Pedersen
ED
,
Wang
Z
,
Brakebusch
C
. 
Rho GTPase function in tumorigenesis
.
Biochim Biophys Acta
2009
;
1796
:
91
8
.
5.
Petrie
RJ
,
Doyle
AD
,
Yamada
KM
. 
Random versus directionally persistent cell migration
.
Nat Rev Mol Cell Biol
2009
;
10
:
538
49
.
6.
Wertheimer
E
,
Gutierrez-Uzquiza
A
,
Rosemblit
C
,
Lopez-Haber
C
,
Sosa
MS
,
Kazanietz
MG
. 
Rac signaling in breast cancer: a tale of GEFs and GAPs
.
Cell Signal
2012
;
24
:
353
62
.
7.
Bos
JL
,
Rehmann
H
,
Wittinghofer
A
. 
GEFs and GAPs: critical elements in the control of small G proteins
.
Cell
2007
;
129
:
865
77
.
8.
Huveneers
S
,
Danen
EH
. 
Adhesion signaling - crosstalk between integrins, Src and Rho
.
J Cell Sci
2009
;
122
:
1059
69
.
9.
Anthis
NJ
,
Campbell
ID
. 
The tail of integrin activation
.
Trends Biochem Sci
2011
;
36
:
191
8
.
10.
Desgrosellier
JS
,
Cheresh
DA
. 
Integrins in cancer: biological implications and therapeutic opportunities
.
Nat Rev Cancer
2010
;
10
:
9
22
.
11.
Ambros
V
. 
The functions of animal microRNAs
.
Nature
2004
;
431
:
350
5
.
12.
Bartel
DP
. 
MicroRNAs: target recognition and regulatory functions
.
Cell
2009
;
136
:
215
33
.
13.
Selbach
M
,
Schwanhausser
B
,
Thierfelder
N
,
Fang
Z
,
Khanin
R
,
Rajewsky
N
. 
Widespread changes in protein synthesis induced by microRNAs
.
Nature
2008
;
455
:
58
63
.
14.
Iorio
MV
,
Croce
CM
. 
MicroRNA dysregulation in cancer: diagnostics, monitoring and therapeutics. A comprehensive review
.
EMBO Mol Med
2012
;
4
:
143
59
.
15.
Farazi
TA
,
Hoell
JI
,
Morozov
P
,
Tuschl
T
. 
MicroRNAs in human cancer
.
Adv Exp Med Biol
2013
;
774
:
1
20
.
16.
Valastyan
S
,
Weinberg
RA
. 
miR-31: a crucial overseer of tumor metastasis and other emerging roles
.
Cell Cycle
2010
;
9
:
2124
9
.
17.
Blenkiron
C
,
Goldstein
LD
,
Thorne
NP
,
Spiteri
I
,
Chin
SF
,
Dunning
MJ
, et al
MicroRNA expression profiling of human breast cancer identifies new markers of tumor subtype
.
Genome Biol
2007
;
8
:
R214
.
18.
Enerly
E
,
Steinfeld
I
,
Kleivi
K
,
Leivonen
SK
,
Aure
MR
,
Russnes
HG
, et al
miRNA-mRNA integrated analysis reveals roles for miRNAs in primary breast tumors
.
PloS ONE
2011
;
6
:
e16915
.
19.
Schaefer
A
,
Jung
M
,
Mollenkopf
HJ
,
Wagner
I
,
Stephan
C
,
Jentzmik
F
, et al
Diagnostic and prognostic implications of microRNA profiling in prostate carcinoma
.
Int J Cancer
2010
;
126
:
1166
76
.
20.
Ballestrem
C
,
Erez
N
,
Kirchner
J
,
Kam
Z
,
Bershadsky
A
,
Geiger
B
. 
Molecular mapping of tyrosine-phosphorylated proteins in focal adhesions using fluorescence resonance energy transfer
.
J Cell Sci
2006
;
119
:
866
75
.
21.
Kanehisa
M
,
Goto
S
,
Sato
Y
,
Furumichi
M
,
Tanabe
M
. 
KEGG for integration and interpretation of large-scale molecular data sets
.
Nucleic Acids Res
2012
;
40
:
D109
14
.
22.
Fukuyama
T
,
Ogita
H
,
Kawakatsu
T
,
Inagaki
M
,
Takai
Y
. 
Activation of Rac by cadherin through the c-Src-Rap1-phosphatidylinositol 3-kinase-Vav2 pathway
.
Oncogene
2006
;
25
:
8
19
.
23.
Arthur
WT
,
Quilliam
LA
,
Cooper
JA
. 
Rap1 promotes cell spreading by localizing Rac guanine nucleotide exchange factors
.
J Cell Biol
2004
;
167
:
111
22
.
24.
Ke
Y
,
Zhao
W
,
Xiong
J
,
Cao
R
. 
miR-149 inhibits non-small-cell lung cancer cells EMT by targeting FOXM1
.
Biochem Res Int
2013
;
2013
:
506731
.
25.
Wang
F
,
Ma
YL
,
Zhang
P
,
Shen
TY
,
Shi
CZ
,
Yang
YZ
, et al
SP1 mediates the link between methylation of the tumour suppressor miR-149 and outcome in colorectal cancer
.
J Pathol
2013
;
229
:
12
24
.
26.
Wang
Y
,
Zheng
X
,
Zhang
Z
,
Zhou
J
,
Zhao
G
,
Yang
J
, et al
MicroRNA-149 inhibits proliferation and cell cycle progression through the targeting of ZBTB2 in human gastric cancer
.
PLoS ONE
2012
;
7
:
e41693
.
27.
Neve
RM
,
Parmar
H
,
Amend
C
,
Chen
C
,
Rizzino
A
,
Benz
CC
. 
Identification of an epithelial-specific enhancer regulating ESX expression
.
Gene
2006
;
367
:
118
25
.
28.
Bos
PD
,
Nguyen
DX
,
Massague
J
. 
Modeling metastasis in the mouse
.
Curr Opin Pharmacol
2010
;
10
:
571
7
.
29.
Wu
L
,
Cai
C
,
Wang
X
,
Liu
M
,
Li
X
,
Tang
H
. 
MicroRNA-142–3p, a new regulator of RAC1, suppresses the migration and invasion of hepatocellular carcinoma cells
.
FEBS Lett
2011
;
585
:
1322
30
.
30.
Moriarty
CH
,
Pursell
B
,
Mercurio
AM
. 
miR-10b targets Tiam1: implications for Rac activation and carcinoma migration
.
J Biol Chem
2010
;
285
:
20541
6
.
31.
Cottonham
CL
,
Kaneko
S
,
Xu
L
. 
miR-21 and miR-31 converge on TIAM1 to regulate migration and invasion of colon carcinoma cells
.
J Biol Chem
2010
;
285
:
35293
302
.
32.
Yang
MH
,
Hsu
DS
,
Wang
HW
,
Wang
HJ
,
Lan
HY
,
Yang
WH
, et al
Bmi1 is essential in Twist1-induced epithelial-mesenchymal transition
.
Nat Cell Biol
2010
;
12
:
982
92
.
33.
Freeman
SA
,
McLeod
SJ
,
Dukowski
J
,
Austin
P
,
Lee
CC
,
Millen-Martin
B
, et al
Preventing the activation or cycling of the Rap1 GTPase alters adhesion and cytoskeletal dynamics and blocks metastatic melanoma cell extravasation into the lungs
.
Cancer Res
2010
;
70
:
4590
601
.
34.
Huang
M
,
Anand
S
,
Murphy
EA
,
Desgrosellier
JS
,
Stupack
DG
,
Shattil
SJ
, et al
EGFR-dependent pancreatic carcinoma cell metastasis through Rap1 activation
.
Oncogene
2012
;
31
:
2783
93
.
35.
Frische
EW
,
Zwartkruis
FJ
. 
Rap1, a mercenary among the Ras-like GTPases
.
Dev Biol
2010
;
340
:
1
9
.
36.
Citterio
C
,
Menacho-Marquez
M
,
Garcia-Escudero
R
,
Larive
RM
,
Barreiro
O
,
Sanchez-Madrid
F
, et al
The rho exchange factors vav2 and vav3 control a lung metastasis-specific transcriptional program in breast cancer cells
.
Sci Signal
2012
;
5
:
ra71
.
37.
Bader
AG
. 
miR-34 - a microRNA replacement therapy is headed to the clinic
.
Front Genet
2012
;
3
:
120
.