Mice

Cre-inducible miR-214 expression construct was generated by cloning a miR-214 expression cassette downstream of a CAGGS promoter and a LoxP-flanked transcription STOP element. This construct was targeted into the mouse Collagen A1 locus using a Flippase (FLP) recombinase-mediated genomic integration. mouse Embryonic Stem cells (mESCs) carrying a single copy of the miR-214STOP construct were identified by resistance to the antibiotic marker hygromycin and Southern blotting. Selected clones were injected into blastocysts to generate pups. To obtain total body overexpressing miR-214over, miR-214STOP mice were bred to a Balancer-Cre transgenic strain [16], kindly provided by E. Hirsch. To generate PyMT miR-214over transgenic mice, miR-214over mice were crossed with Mouse Mammary Tumor Virus Polyoma Middle T antigen MMTV-PyMT transgenic mice [17], kindly provided by F. Cavallo’s laboratory, University of Torino, Italy. miR-214ko mice [18] were kindly provided by Eric Olson’s laboratory, UT Southwestern Medical Center, Dallas, USA. The sources of primers used for genotyping are available upon request. All experiments performed with live animals complied with ethical animal care and were approved by the MBC Animal Care Committee and the Italian Ministry of Health (13/2014-PR to DT; 847/2020-PR to DT).

Cell culture

B16-F10 murine melanoma cells, EO771 mouse mammary tumor cells, NIH3T3 fibroblasts and human bone marrow stromal HS5 cell lines were obtained from The American Type Culture Collection. B16-F10, NIH3T3 and HS5 were maintained in Dulbecco's Modified Eagle's Medium containing 10 mM Glutamax and 4.5 g/L glucose (DMEM Glutamax™, GIBCO Invitrogen Life Technologies, Carlsbad, CA), supplemented with 10% heat-inactivated FCS (Seromed, GmbH), 1 mM sodium pyruvate, 25 mM HEPES pH 7.4 and 100 μg/mL gentamicin (all from GIBCO Invitrogen Life Technologies, Carlsbad, CA). EO771 cells were cultured in RPMI (Roswell Park Memorial Institute) medium containing 10 mM Glutamax and 4.5 g/L glucose (DMEM Glutamax™, GIBCO Invitrogen Life Technologies, Carlsbad, CA), supplemented with 10% heat-inactivated FCS (Seromed, GmbH), 25 mM HEPES pH 7.4 and 100 μg/mL gentamicin (all from GIBCO Invitrogen Life Technologies, Carlsbad, CA). 4175-TGL breast cancer cells were kindly provided by J. Massaguè and maintained as in [19]. MA-2 melanoma cells were a kind gift from Lei Xu and cultured as indicated in [19, 20]. Mesenchymal Stem Cells (MSCs) were derived, characterized and maintained as described in [21]. Cancer Associated Fibroblasts (CAFs) were isolated form PyMT-miR-214wt, miR-214over and miR-214ko PyMT tumors. Briefly, mammary tumors (around 0.5 mm3) were excised and minced into 2–4 mm fragments, which were then incubated for 3 h with Collagenase A 1 mg/ml. Digested fragments were filtered (70 μm cell strainer) and fibroblast cell populations were enriched through pre-plating and subsequent differential trypsinization. CAFs were maintained in Dulbecco's Modified Eagle's Medium containing 10 mM Glutamax and 4.5 g/L glucose (DMEM Glutamax™, GIBCO Invitrogen Life Technologies, Carlsbad, CA), supplemented with 10% heat-inactivated FCS (Seromed, GmbH), 1 mM sodium pyruvate, 25 mM HEPES pH 7.4 and 100 μg/mL gentamicin (all from GIBCO Invitrogen Life Technologies, Carlsbad, CA). The purity of the isolated population was assessed by Western Blot analysis of the main CAF markers. Tumor Associated Macrophages (TAMs) were derived and maintained as described in [22]. Murine Embryo Fibroblasts (MEFs) were derived and maintained as described in [23]. MEFs Stat3ko and Stat3wt were derived and maintained as described in [24]. For IL-6 experiments, MEFs Stat3ko and Stat3wt were treated with recombinant IL-6 (500 ng/ml) plus soluble receptors (250 ng/ml) as described in [25] for 6 h before RNA extraction. Co-culture experiments were performed as follows. Cultures containing either stroma (MEFs, hMSCs) or melanoma (GFP+ B16-F10, GFP+ MA-2) cells or both were prepared. A cell preparation with a 1:1 (stroma:tumor) ratio was used. Cells were allowed to attach for 24 h, then media were replaced with fresh DMEM. After 24 h, cells were detached and sorted based on GFP expression as described below.

Reagents and antibodies

TaqMan® MicroRNA assays for miRNA detection: Hsa-miR-214 ID 002306, Hsa-miR-148b ID000471, Hsa-miR-223 ID 002295 U6 snRNA ID001973, (all from Applied Biosystems, Foster City, USA). TaqMan® Gene expression assays for S100b: ID Mm00485897_m1. Primary antibodies: anti-ITGA5 pAb RM10 kindly provided by G. Tarone laboratory (Molecular Biotechnology Center, University of Torino), anti-CD166/ALCAM mAb MOG/07 (Novocastra Laboratories), anti-E-cadherin mAb #610,182 (BD Transduction Laboratories, Franklin Lakes, USA), anti-N-cadherin pAb ab18203 and anti-αSMA pAb ab15734 (Abcam, Cambridge, United Kingdom), anti-TFAP2C mAb 6E4/4 mAb, H-77 pAb, anti-GAPDH pAb V-18, anti-ACTIN I-19 pAb, anti-Hsp90 mAb F-8 (all from Santa Cruz Biotechnology). Secondary antibodies: HRP-conjugated goat anti-mouse IgG, goat anti-rabbit IgG (all from Santa Cruz Biotechnology, Santa Cruz, CA), goat anti-rabbit IgG Alexa-Fluor-488 (Molecular Probes, Invitrogen Life Technologies, Carlsbad, USA). All antibodies were used at the producer’s suggested concentration.

Vectors, generation of stable cell lines

Stable miR-214 down-modulation in NIH3T3 and HS5 cells were obtained following transduction of pLenti-CMV-GFP-Puro-miR-214sponge (miR-214sponge) or pLenti-CMV-GFP-Puro (control) expression vectors [19]. GFP+ melanoma cells (B16-F10 or MA-2) were obtained following transduction of pLenti-CMV-GFP-Puro lentiviral expression vectors. Lentiviruses were produced according to Trono’s lab protocol (http://tronolab.epfl.ch). Supernatants were harvested 48 h post-transfection, filtered with 0.45 μm filters, diluted and used to infect 3.5 × 105 cells in 6-well plates, in presence of 8 μg/mL Polybrene (Sigma-Aldrich, St Louis, MO). Infected cells underwent puromycin selection to obtain a pure population.

Fluorescence-activated cell sorting

GFP+ B16-F10 subcutaneous tumors were harvested at the end point of the experiment, dissociated with Collagenase A for 1 h and cells sorted based on GFP expression. For co-culture experiments, melanoma (GFP+ B16-F10 or GFP+ MA-2) and stromal cells (MEFs, hMSCs) were detached and sorted based on GFP expression using a BD FACS Aria III (Becton Dickinson) cell sorter: GFP+ tumor fraction; GFP− stroma fraction. Cell pellets were washed and snap-frozen before RNA isolation.

RNA isolation and qRT-PCR

Total RNA was isolated using TRIzol® Reagent (Invitrogen Life Technologies, Carlsbad, CA). qRT-PCRs for miR detection or gene expression analysis were performed with the indicated TaqMan® MicroRNA or Gene Expression Assays (Applied Biosystems, Foster City, CA) on 10 ng total RNA according to the manufacturer's instructions. qRT-PCRs were carried out using gene-specific primers, using a 7900HT Fast Real Time PCR System (Applied Biosystems, Foster City, CA). Quantitative normalization was performed on RNU6 or RNU44 small nucleolar RNAs expression or 18S expression. The relative expression levels between samples were calculated based on the comparative delta CT (threshold cycle number) method (2-ΔΔCT) using the sample median as reference point as described in [26]. For experiments with Conditioned Medium (CM) or Extracellular Vescicles (EVs), RNA was extracted from tumor or stroma cells following 24-48 h treatments in serum-free medium. When RNA was extracted from tumors or dissected metastases, samples were disrupted with an Ultra TURRAX Homogenizer (IKA®-Werke GmbH) prior Trizol extraction.

In situ hybridization

Control and miR-214over mouse embryos were collected at 12.5 days post-coitum (E12.5), fixed in 4% PFA/0.1 M Phosphate Buffer (PB, pH 7.4) for 12–16 h, washed in PBS, dehydratated in methanol, processed for paraffin embedding and sectioned at 6 μm. Hybridization was carried out with Digoxigenin (DIG)-labelled Locked Nucleic Acid (LNA) probes, specific for the detection of the mature murine miR-214 (Exiqon, Vedbaek, Denmark) according with manufacturer’s instruction. The sections were hybridized with the probe for 16 h, washed, incubated with an anti-DIG-Alkaline Phosphates (AP) antibody (Roche, GmbH) and developed with NBT-BCIP (Sigma-Aldrich, St Louis, MO). To control the efficiency of the procedure and RNA preservation, the adjacent sections were hybridized with a specific probe for the U6 small nucleolar RNA.

Northern blot analysis

Total RNA was isolated from miR-214wt and miR-214over embryos collected at 12.5 days post-coitum (E12.5) using TRIzol® Reagent (Invitrogen Life Technologies, Carlsbad, CA) according to manufacturer’s instruction. 25 μg of total RNA were resolved on 12.5% (w/v) TBE–Urea–polyacrylamide gel electrophoresis, transferred to a Hybond N + membrane (GE Healthcare Life Sciences, Piscataway, NJ, USA) and UV crosslinked to membrane. The filter was hybridized overnight at 45 °C with a specific miR-214 digoxigenin-labeled LNA Detection probe (Exiqon, Vedbaek, Denmark), washed and visualized with a specific DIG antibody (1: 10,000) using the DIG Nucleic Acid Detection kit, according to manufacturer’s instructions (all from Roche, GmbH). The filter was then stripped and re-probed overnight at 45 °C using a specific U6 digoxigenin-labeled LNA Detection probe (Exiqon, Vedbaek, Denmark).

Protein preparation and immunoblotting

Total protein extracts were obtained using a boiling buffer containing 0.125 M Tris/HCl, pH 6.8 and 2.5% Sodium Dodecyl Sulphate (SDS) (Sigma-Aldrich, St Louis, MO). 20–30 µg of proteins were separated by SDS polyacrylamide gel electrophoresis (PAGE) and electroblotted onto nitrocellulose membranes (BioRad). Membranes were blocked in 5% non-fat milk PBS-Tween 0.1% buffer for 1 h at 37 °C, then incubated with appropriate primary and secondary antibodies in PBS-Tween 0.1% buffer, respectively, overnight at 4 °C or for 1 h at room temperature and developed using Chemidoc Touch Imaging System (Bio Rad). For experiments in the presence of Extracellualr Vescicles (EVs), proteins were extracted from tumor cells previously treated with stroma EVs (5000 EVs/cell) for 24-48 h in serum-free medium.

Conditioned medium (CM) from stroma and tumor cells

miR-214wt and miR-214over MEFs or CAFs and HS5 or NIH3T3 control and miR-214sponge or B16-F10, MA-2 cells were grown to sub-confluence and treated for 48 h with serum-free medium to obtain the corresponding CM to use on recipient cells which were then kept with the CM for 24-48 h, before RNA/protein extractions or biological experiments. To obtain Extracellular Vescicle-depleted CM (EVs-depleted), CM was harvested and centrifuged for 30 min at 3,000 g to remove cell debris and apoptotic bodies. After that, the supernatant was centrifuged for 2 h at 100,000 g, 4 °C using the Beckman Coulter Optima L‐100 K Ultracentrifuge with the rotor type 45 Ti 45,000 rpm. The supernatant was then collected and centrifuged again for 2 h at the same conditions to remove remaining EVs. For anti-IL-6R and anti-IL-6 blocking antibody experiments, CAFs were treated with EV-depleted B16-F10-derived CM plus 50 μg/ml of anti-mouse IL-6R (rat MAb 15A7 clone) or 10 μg/ml of anti-mouse IL-6 (rat Mab clone MP5-20F3, BioXCell), respectively, or control IgG (Thermo Fisher Scientific) for 6 h, before RNA extraction.

Extracellular Vesicle (EV) isolation and characterization

Isolation of EVs was performed as described in [27]. Briefly, sub-confluent miR-214wt and miR-214over MEFs or CAFs or miR-214sponge and control NIH3T3 or HS5 cells were cultured in serum‐free DMEM for 18 h. The medium was then centrifuged for 30 min at 3,000 g to remove cell debris and apoptotic bodies. After that, the supernatant was centrifuged for 2 h at 100,000 g, 4 °C using the Beckman Coulter Optima L‐100 K Ultracentrifuge with the rotor type 45 Ti 45,000 rpm. The pellet of EVs obtained was resuspended in DMEM supplemented with 1% DMSO. Suspension of isolated EVs was then stored at − 80 °C until further use. Alternatively, EV pellets were resuspended in Trizol Reagent for RNA extraction. EVs were analyzed using the NanoSight NS300 system (Malvern Instruments, Ltd). For isolation of EVs from mouse, blood was collected, plasma was derived and Exoquick™ reagent (System Bioscience, Palo Alto, CA) was used according to manufacturer’s instructions. Immediately after EV isolation, RNA was extracted using the miRNeasy Serum/Plasma kit (Qiagen, Stanford CA) following manufacturer’s standard protocol.

FACS Characterization of EVs

EVs were characterized by cytofluorimetric analysis using the following fluorescein isothiocyanate (FITC), allophycocyanin (APC) or phycoerythrin (PE) conjugated antibodies: CD63 (Cat.n. 130–100-160, Miltenyi Biotec, Germany), PDGFRβ (Cat.n. 130–105-280, Miltenyi Biotec, Germany), CD73 (Cat. n. 130–095-182, Miltenyi Biotec, Germany), αSMA (Cat. n. C6198, Invitrogen), CD44 (Cat. n. 130–095-195, Miltenyi Biotec, Germany), PDGFRα (Cat. n. LS-C107240, LSBio), FAP (ab207178, Abcam, Cambridge, United Kingdom) and fluorescent secondary Rabbit IgG antibody (A-11012, Thermo Fisher Scientific, Waltham, MA, USA). Conjugated mouse non-immune isotypic immunoglobulin G (IgG) (Miltenyi Biotec, Germany) was used as control. Briefly, 10 µl of EVs were labeled for 15 min at 4 °C with antibodies and immediately diluted 1:3 with saline solution and acquired [28]. Cytofluorimetric analysis were performed using the CytoFLEX flow cytometer (Beckman Coulter) with CytExpert software. Each analysis includes 3 biological replicates.

Transmission electron microscopy analysis of EVs

EVs were analyzed using transmission electron microscopy analysis. For this, EV samples were placed on 200 mesh nickel formvar carbon-coated grids (Electron Microscopy Science, Hatfield, Pennsylvania, USA) and left to adhere for 20 min. Next, grids were incubated with 2.5% glutaraldehyde containing 2% sucrose. After washing in distilled water, samples were negatively stained with Nano-W and Nano-Van (Nanoprobes, Yaphank, New York, USA) and analyzed using a Jeol JEM 1010 electron microscope (Jeol, Tokyo, Japan) as described in [29].

Proliferation assay

5 × 103 cells/well were plated in 96 well plates in complete medium and starved for 24 h. Complete medium was then added and cells were allowed to grow for 1, 2, 3 and 5 days, fixed with 2.5% glutaraldehyde and stained with 0.1% crystal violet. The dye was solubilized using 10% acetic acid and optical density measured directly in plates using Promega GloMax®-Multi Detection System (Promega, Madison, WI) at 600 nm wavelength. For experiments with CM, tumor cells were pretreated for 24 h with CM from the different stroma cells, then complete medium was added and the assay performed as described above.

Transwell migration, Transendothelial migration and Wound healing assays

To measure migration 3 × 105 B16-F10 or 1 × 105 EO771 cells or 1 × 105 MEFs or CAFs were seeded in serum-free media in the upper chambers of cell culture inserts (Transwells) with 8.0 μm pore size membrane (24‐well format, Becton Dickinson, NJ). The lower chambers were filled with complete growth media. After 18–20 h, the migrated cells present on the lower side of the membrane were fixed in 2.5% glutaraldehyde, stained with 0.1% crystal violet and photographed using an Olympus IX70 microscope [19]. For transendothelial migration, 105 HUVECs labelled with CellTracker™ Green CMFDA (Molecular Probes, Invitrogen Life Technologies) according to the manufacturer's instructions were seeded in complete medium in the upper part of transwell inserts with 5.0 μm pore size membrane (24-well format, Costar, Corning Incorporated, NY) coated by gelatin, and grown for 72 h, till confluency. Then, 3 × 105 B16-F10 or 1 × 105 EO771 cells were labelled with CellTracker™ Orange CMRA (Molecular Probes, Invitrogen Life Technologies), according to the manufacturer's instructions and seeded in HUVEC's complete medium onto the HUVEC–CMFDA monolayer on the upper side of the transwell. After 20 h, HUVECs and non-transmigrated cells were removed and the red-fluorescent (CMRA) cells that migrated on the lower side of the membrane were fixed in 4% paraformaldheyde and photographed using Zeiss AxioObserver microscope with ApoTome Module. Migration, invasion and transendothelial migration were evaluated by measuring the area occupied by migrated cells using the ImageJ software (http://rsbweb.nih.gov/ij/). For CM or EV experiments, cells were pretreated for 24 h with CM or EVs (5000 EVs/cell) before seeding. The wound healing motility assay was used to measure two dimensional movements. Cells were grown to confluency in six-well plates, serum starved or treated with CM or EVs (5000 EVs/cell) for 24 h, then a cross wound was made on the monolayer using a sterile 200 μl pipette tips. Cells were rinsed three times with Phosphate Buffered Saline (PBS) and placed in either serum-free DMEM or 10% FBS-DMEM. Two-dimensional cell movements were quantitated by measuring the distance covered by the migrating cells. For each experiment the four arms near the cross were photographed. Photos were taken at t = 0 h and at t = 6 h for B16-F10 cell or at t = 24 h for NIH3T3, HS5, EO771, MA-2 and 4175-TGL using a Zeiss AxioObserver microscope (Zeiss). Images were analyzed with ImageJ Software (http://rsbweb.nih.gov/ij/). The two-dimensional movement of the cells was quantitated by measuring the distance between the two edges of the wound and the formula described in [30] used to estimate cell speed.

In vivo tumor and metastasis assays

For experimental metastasis assays, 5 × 105 syngeneic B16-F10 (in 200 µL of PBS) were injected into the tail vein of 8–10 weeks old wild type (miR-214wt) or miR-214over or miR-214ko mice. Mice were dissected 8 days later and lung surface metastases counted in fresh total lungs using a Nikon SMZ1000 stereomicroscope (Nikon, Japan), then lungs were formalin-fixed, cut in small pieces, paraffin-embedded, sectioned and haematoxylin & eosin (H&E)-stained. Micrometastases were evaluated on specimens, with an Olympus BH2 microscope (Olympus, Japan). Spontaneous dissemination was evaluated in 8–10 weeks old wild type (miR-214wt) or miR-214over or miR-214ko mice subcutaneously injected with 5 × 105 syngeneic B16-F10 or 5 × 105 EO771 cells (in 200 µL of PBS). Mice were sacrified 45 days (B16-F10) or 30 days (EO771) after injections and tumors were harvested and weighed. For B16-F10 cells, subcutaneous tumors were surgically removed 15 days after injection. 30 days later, animals that were free of any local recurrence were further analyzed for the presence of Circulating Tumor Cells (CTCs), derived as described in [31]. Briefly, blood was collected by heart puncture with a 25G needle syringe in the presence of heparin. Blood was plated in tissue culture medium, and 3 days later tumor cells were washed, and then colonies or total number of cells were counted one week later. For some experimental groups, lungs were formalin-fixed, cut in small pieces, paraffin-embedded, sectioned and haematoxylin & eosin (H&E)-stained. Micrometastases were evaluated on specimens, with an Olympus BH2 microscope (Olympus, Japan). For Extracellular Vesicle (EV) treatments, 5 µg of EVs were administered (tail vein) to miR-214ko mice 24 h before the tail vein injection of B16-F10 cells. For MA-2 experiments, subcutaneous tumors and experimental lung metastases were obtained in immunosuppressed mice as described in [11], primary tumors and metastases were dissected and RNA isolated.

In vivo extravasation assay

Extravasation of B16-F10 cells was evaluated as described in Orso et al. [19]. Briefly, 1 × 105 B16-F10 cells, previously labeled with CellTracker Orange CMRA (Molecular Probes, Invitrogen Life Technologies), were injected into the tail vein of 4- to 6-week-old miR-214wt, miR-214over or miR-214ko mice. 2 h or 48 h later, mice were sacrificed, lungs were dissected and photographed in toto using a Leica MZ16F fluorescence stereomicroscope and red fluorescence was quantified 48 h post- injection using the ImageJ software (http://rsbweb.nih.gov/ij/). Lungs were embedded in OCT (Killik, BioOptica), frozen, cryostat-cut in 6-μm-thick sections. Localization of tumor cells, inside/outside the vessels, was evaluated on sections at a Leica TCS SP8 confocal system (Leica Microsystems), following blood vessels staining with an anti-CD31 primary antibody in immunofluorescence. For Extracellular Vesicle (EV) treatment experiments, B16-F10 cells were pretreated for 24 h with 5,000 EVs/cell of the indicated sources before tail vein injections. Alternatively, 5ug of EVs were administered (tail vein) to miR-214ko mice and subsequently tail vein injected with B16-F10 cells 24 h later. Relative extravasated cells were measured as fold change of extravasated cells in miR-214over mice versus the values obtained in miR-214wt or miR-214ko animals. In detail, when miR-214wt versus miR-214over were compared, we performed the following calculation: we counted the number of extravasated cells in each miR-214wt mouse and evaluated the average, we divided each figure by the average, we plotted the single values and showed the average as a line (equal to 1). Similarly, we counted the number of extravasated cells in each miR-214over mouse and calculated the average, we divided each value by the average of extravasated cells in miR-214wt animals and plotted the obtained results plus the average. A similar approach was used when miR-214over versus miR-214ko were analyzed.

Stroma and IL-6/STAT3 signature correlation analysis in human tumors

TCGA data have been downloaded from the gdc portal (https://portal.gdc.cancer.gov/). Tumor purity data have been obtained from: http://genboree.org/theCommons/documents/569 (EDec), https://bioinformatics.mdanderson.org/estimate/disease.html (Stromal scores), and the TCGA biolinks R package (ABSOLUTE, IHC). TCGA samples corresponding to tumors (“Tumor metastatic” for SKCM and “Primary tumor” for BRCA) have been selected for the analyses and matched wih the corresponding stromal scores. The correlation between miR-214 expression and IL-6/STAT3 activity was inferred from the Pearson’s correlation between miR expression and a set of IL-6/STAT3 activity signatures’ expression. More in detail, for each TCGA sample having both mRNA and miR expression profiles, the sum of log transformed expression values of each IL-6/STAT3 signature’s genes was computed, and correlated with the corresponding log transformed miR expression value. AZARE_sig [32], DAUER_sig [33], IL-6_sig (from MSigDB, [34]), Alvarez_sig [35], TH_sig [36], stat3_sig [37], Jak/STAT (from MSigDB [34]). Analyses have been performed in R (version 3.5.1). Plots have been generated with pheatmap and ggplot2 R packages.

Statistical analysis

The results are shown as mean ± Standard Deviation (SD) or ± Standard Error of Mean (SEM), as indicated. Each data group was first evaluated with Shapiro–Wilk normality test and, based on results, values were examined with parametric or non-parametric tools. t-test was used for parametric analyses between two groups. Instead, Mann–Whitney test was applied for non-parametric evaluations. When comparisons for more than two groups were performed, 1-way or 2-way ANOVA tests were chosen for parametric analyses, instead Kruskal–Wallis tests were applied for non-parametric evaluations. * = P ≤ 0.05; ** = P ≤ 0.01; *** = P ≤ 0.001 were considered to be statistically significant.