Inhibition of profibrotic microRNA-21 affects platelets and their releasate

Fibrosis is a major contributor to organ disease for which no specific therapy is available. MicroRNA-21 (miR-21) has been implicated in the fibrogenetic response, and inhibitors of miR-21 are currently undergoing clinical trials. Here, we explore how miR-21 inhibition may attenuate fibrosis using a proteomics approach. Transfection of miR-21 mimic or inhibitor in murine cardiac fibroblasts revealed limited effects on extracellular matrix (ECM) protein secretion. Similarly, miR-21–null mouse hearts showed an unaltered ECM composition. Thus, we searched for additional explanations as to how miR-21 might regulate fibrosis. In plasma samples from the community-based Bruneck Study, we found a marked correlation of miR-21 levels with several platelet-derived profibrotic factors, including TGF-β1. Pharmacological miR-21 inhibition with an antagomiR reduced the platelet release of TGF-β1 in mice. Mechanistically, Wiskott-Aldrich syndrome protein, a negative regulator of platelet TGF-β1 secretion, was identified as a direct target of miR-21. miR-21–null mice had lower platelet and leukocyte counts compared with littermate controls but higher megakaryocyte numbers in the bone marrow. Thus, to our knowledge this study reports a previously unrecognized effect of miR-21 inhibition on platelets. The effect of antagomiR-21 treatment on platelet TGF-β1 release, in particular, may contribute to the antifibrotic effects of miR-21 inhibitors.

250 µl guanidine hydrochloride (GuHCl) buffer (4 M GuHCl, 50 mM sodium acetate, 20 mM EDTA, pH 5.8). Incubation was performed for 48 h at room temperature and vortexed vigorously. GuHCl extracts were then collected and stored until later analysis. Protein concentrations in all three fractions were then determined by Bradford assay (BioRad) according to manufacturer's instructions. Proteins in the NaCl and GuHCl fractions were then prepared for MS analysis as follows: 15 µg of protein were precipitated by adding a ten-fold volume of 100% ethanol to GuHCl samples and 100% acetone to NaCl samples, followed by overnight incubation at -20°C. Proteins were then precipitated with centrifugation at 16000 x g for 40 min at 0°C. Protein precipitates were fully dried using a SpeedVac Concentrator (ThermoFisher Scientific). Protein pellets were then resuspended in 20 µl deglycosylation buffer as described earlier, but without Nglycosidase F. Samples were briefly vortexed and centrifuged, followed by incubation at 25°C for 2 h and 37°C for 24 h. Samples were then centrifuged and dried using the

Gel-LC-MS/MS
Samples were denatured with sample loading buffer and incubation at 96°C for 5 min.
Gel separation was performed using Bis-Tris discontinuous 4-12% polyacrylamide gradient gels (NuPage, Invitrogen) as described previously (2). Gels were fixed and silverstained to visualize proteins. Each gel lane was divided into 12 pieces without leaving any gap in between, followed by de-staining and in-gel tryptic digestion using an Investigator ProGest (Digilab) robotic digestion system. Eluted peptides were lyophilized under vacuum at -55°C for approximately 5 h (Christ Alpha 1-2 LD Freeze Dryer) and resuspended in 40 µl of 2% acetonitrile, 0.05% trifluoroacetic acid (TFA) in H2O. Tryptic peptides were separated on a reversed phase nanoflow HPLC system (Dionex PepMap C18, 25cm x 75µm, Dionex Ultimate3000 RSLCnano) and eluted with a 70-min gradient Carbamidomethylation of cysteine was set as a fixed modification and oxidation of methionine, proline and lysine as variable modifications. Two missed cleavages were allowed. Search results were loaded into Scaffold (Proteome Software Inc., version 4.3.2) to validate the MS/MS-based peptide and subsequent protein identification and to calculate the spectral count (3,4). Peptide identifications were accepted if they could be established at a probability >95% as specified by the Peptide Prophet Algorithm (3). Only tryptic peptides were included in the analysis. Protein identifications were accepted if they could be established at a probability >99% with at least two unique peptides (4). identifications with the following filters; a peptide probability of greater than 95.0% (as specified by the Peptide Prophet algorithm), a protein probability of greater than 99.0%, and at least two independent peptides per protein. The normalized total precursor intensity was used for quantification.

RNA isolation
Total RNA was extracted using the miRNeasy Mini kit (Qiagen) according to the manufacturer's recommendations, with some modifications. For isolation from heart tissue, a 1-2 mm piece was cut from the apex. The tissue piece was directly placed into a tube containing Lysing Matrix D beads (MP Biomedicals) and 700 µl QIAzol reagent.
Lysis was performed in a FastPrep-24 Homogeniser (MP Biomedicals) at 6000 rpm for two rounds of 20 sec. Cells or 50 µl of whole blood were lysed in 700 µl of QIAzol reagent.
For RNA isolation from plasma, samples were centrifuged for 10 min at 4000 x g at 4°C. Concentration of cellular and tissue RNA was determined by spectrophotometry based on absorbance at 260 nm using NanoDrop 2000c (Thermo Scientific).

Reverse transcription and pre-amplification
For relative quantification of miRNA in plasma by qPCR, 100 ng of RNA (or 3 µl for plasma RNA) were used as input in each reverse transcription (RT) reaction. RT reactions (and a pre-amplification step for plasma RNA) were set up according to the company's recommendations. Briefly, miRNAs were reverse-transcribed using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems), combining 100 ng RNA in 3.5 µl, for 10 min to ensure enzyme inactivation. Pre-amplification reaction products were diluted to a final volume of 40 µl. For gene expression levels in CF, bone marrow or cardiac RNA, RT was performed using the SuperScript VILO cDNA Synthesis Kit (Invitrogen). Per sample, 2 µl of VILO RT Master Mix were combined with 8 µl of sample in a 25-100 ng/µl dilution. Thermal cycler stages were set as follows: incubation at 25°C for 10 min and 42°C for 120 min, followed by termination of the reaction at 85°C for 5 min. RT-PCR and pre-amplification products were stored at -80ºC.

Real-time PCR
TaqMan miRNA or gene expression hydrolysis assays (Supplemental Table 1) were used to assess their relative expression levels, as previously described (5 for 15 sec and 60ºC for 1 min. Relative quantification was performed using the 2 -Cq method with proprietary Viia7 software (Applied Biosystems). For cellular and tissue RNA, the reference gene transcript was selected based on analysis of stability using the online RefFinder tool (6). For plasma RNA, the exogenous cel-miR-39-3p spike-in was used as the reference transcript.

Immunoblotting
Samples were mixed with 4X denaturing sample buffer, heated at 95°C for 10 min and separated on Bis-Tris discontinuous 4-12% polyacrylamide gradient gels (NuPAGE, Invitrogen). Proteins were transferred to nitrocellulose membranes. Membranes were blocked with 5% fat-free milk powder in PBS containing 0.1% Tween20 (PBS-T) and probed overnight at 4ºC with primary antibodies (Supplemental Table 2) in 5% bovine serum albumin in PBS-T at 4°C while shaking. The membranes were incubated with light chain-specific secondary horseradish peroxidase (HRP)-conjugated antibodies (Supplemental Table 2) in 5% fat-free milk powder in PBS-T for 1 h. For analysis of phosphorylated Wiskott-Aldrich Syndrome protein (p-WASp), PBS-T was replaced with Tris-Buffered Saline with 0.1% Tween20 (TBS-T), with blocking and primary/secondary antibody incubation being performed using 5% bovine serum albumin in TBS-T. Blots were imaged using enhanced chemiluminescence (ECL, GE Healthcare) on a Xograph processor. Densitometry was performed using ImageJ software (v.1.48v, NIH, USA).

Platelet counting
For flow cytometry-based platelet counting, ACD-anticoagulated blood was diluted 25x with modified Tyrode-HEPES buffer at 37ºC. Samples were incubated with 5 µl of 0.2 mg/ml allophycocyanin (APC)-labelled anti-mouse CD41 antibody in the dark for 15 min at room temperature, followed by further dilution to 1:1000 with modified Tyrode-HEPES buffer. Unstained control samples were prepared by treatment with an equal amount of modified Tyrode-HEPES buffer. Samples were analyzed using an Accuri C6 flow cytometer (BD Biosciences) with slow flow. A set volume of each sample was analyzed to allow for absolute concentration calculation. Events corresponding to platelets were identified according to forward (FSC) and side scatter (SSC) as well as by identification of APC-positive events. Subsequent quantification of events was performed postacquisition using FlowJo v7.4 (Tree Star). For miR-21 null mice and littermates, blood cell counting was performed using a Hemavet 950FS (Drew Scientific) using proprietary software according to manufacturer's instructions.

Platelet function testing
Blood was collected from the inferior vena cava using syringes containing lepirudin (Refludan, 25 μg/ml; Celgene) for aggregometry experiments and ELISA measurements.
Half-area 96-well microtiter plates (Greiner Bio-One) were pre-coated with hydrogenated gelatin (0.75% w/v; Sigma-Alrich) in PBS to block non-specific activation of blood. 4 μl of vehicle or agonist solution was then added to each well: arachidonic acid (AA; 0.03-0.6 mM; Sigma-Aldrich), Horm collagen (0.1-3 μg/ml; Nycomed) and the PAR-4 activating peptide AYPGKF amide (PAR4-AP, 30-100 μM; Bachem). To each well, 35 μl of PRP or whole blood was added and the plate was then placed onto a heated plate shaker (Bioshake IQ, Q Instruments) at 37°C for 5 min mixing at 1200 rpm. Light transmission of each well was determined using a 96-well plate reader (SunriseTM, Tecan) at 595 nm.

Platelet releasate isolation and analysis
Blood was collected into Acid-Citrate-Dextrose (ACD; 85 mM trisodium citrate dehydrate, 66.6 mM citric acid monohydrate, 111 mM anhydrous D-glucose) by cardiac puncture and pooled per four mice, followed by centrifugation at 150 x g for 10 min at room temperature.
PRP was carefully transferred and combined with 10 µM indomethacin (Sigma-Aldrich) and 1 µM prostaglandin E1 (PGE1, Sigma-Aldrich). Platelet pellets were obtained by centrifugation of PRP at 500 x g for 10 min at room temperature. Supernatant PPP was collected, and the platelet pellet was washed twice with modified Tyrode-HEPES buffer (NaCl 134 mM, KCl 2.9 mM, Na2HPO4 0.34 mM, NaHCO3 12 mM, HEPES 20 mM, MgCl2 1 mM, Glucose 5 mM), supplemented with 1 µM PGE1 and 10 µM indomethacin, with centrifugation at 500 x g for 10 min at room temperature in between. The platelet pellet was then resuspended in 200 µl of modified Tyrode-HEPES buffer without PGE1 and indomethacin, and platelets were activated with 1U/ml thrombin from human plasma (Sigma-Aldrich), followed by incubation for 5 min at 37°C under constant shaking. After 5 min, ice-cold protease inhibitor (cOmplete Mini, EDTA-free, Roche) was added. Subsequent centrifugation for 5 min at 1000 x g and 1 h at 20000 x g, both at 4°C, were performed to isolate the platelet releasate. Samples were concentrated using an Amicon 3K filter device (Merck Millipore) according to manufacturer's instructions. Protein concentration was measured using the CBQCA Protein Quantitation Kit (Molecular Probes) according to manufacturer's instructions. Proteins were separated and digested as described above. Injection volumes for each sample were further adjusted based on quantitative densitometry of the silver-stained gel. Mass spectrometry analysis of the peptides was performed as described above, with the following differences: raw files were searched against UniProt/SwissProt mouse database (2015_02); only oxidation of methionine was set as a variable modification; peptide identifications were accepted if they could be established at a probability > 85% and protein identifications were accepted if they could be established at a probability > 99.9% with at least two unique peptides.

Platelet isolation and lysis
Blood was collected into ACD by cardiac puncture, followed by a 1:1 dilution with filtered Tyrode-HEPES buffer. PRP was obtained by centrifugation at 100 x g for 8 min without brake. The top two-third of supernatant was transferred and supplemented with 1 µl/mg PGE1, immediately followed by centrifugation at 800 x g for 10 min without brake.
Supernatant PPP was transferred and the platelet pellet was washed twice by resuspending in modified Tyrode-HEPES buffer supplemented with 1 µl/mg PGE1, each time followed by centrifugation at 800 x g for 10 min without brake. Platelets were then lysed by a freeze-thaw and by subsequent lysis in ice-cold cell lysis buffer (Cell Signaling) supplemented with protease inhibitor (cOmplete Mini, EDTA-free, Roche) and phosphatase inhibitor (PhosSTOP, Roche). Protein concentration was quantified using a BCA Protein Assay kit (Pierce, Thermo Scientific) according to manufacturer's instructions.

Murine bone marrow cell analysis
Bone marrow cells were isolated from femora using a syringe pre-filled with sterile PBS after cutting the epiphysis at either end of the bone. Cells were flushed out of the bone onto a 100 µm cell strainer (Corning), followed by centrifugation for 5 min at 350 x g at 4ºC. The supernatant was discarded and samples were resuspended in sterile PBS, followed by centrifugation at similar conditions. Supernatant was again removed and cells were resuspended in red cell lysis buffer (0.31 M ammonium chloride supplemented with sodium bicarbonate and EDTA) and incubated at 37ºC for 3 min, followed by adding 5x volume of PBS supplemented with 2% FBS and 0.1% sodium azide. Samples were then centrifuged for 5 min at 350 x g at 4ºC, followed by discarding of the supernatant and resuspension and lysis in 700 µl QIAzol reagent for RNA isolation.

Murine bone marrow immunohistochemistry
Femora were cleaned and placed in 4% formaldehyde and stored at 4°C for three days.
Bones were then washed in PBS twice and decalcified in 0.38M EDTA in H2O, pH 7.0 at 4°C for three weeks. Bone tissues were then dehydrated through a series of graded ethanol baths and embedded in paraffin for subsequent microtome sectioning (5 µm).
After placing on glass slides, sections were baked at 60°C for 2 h. Epitope unmasking was achieved using hot sodium citrate buffer incubation for 20 min. Sections were washed three times with PBS-T and incubated with primary antibodies (Supplemental Table 2) or species-matched isotopes overnight at 4°C after blocking with 10% FBS in PBS-T for 1 h. Following three 5-minute washes in PBS-T, sections were incubated for 1 h at room temperature with secondary antibody (Supplemental Table 2) in 10% FBS/PBS-T, according to the source of the primary antibody. Nuclei were stained with DAPI (1:1000 dilution) for 10 min. Sections were then mounted on a Vecta Mount (Vector Laboratories, cat. no. H-5000). Sections were visualized with a 20X objective using an inverted Nikon NI-E microscope equipped with a Yokogawa CSU-X1 spinning disk confocal unit and an Andor iXon 3 EM-CCD camera. Images were acquired using NIS-elements 4.0 software.

Luciferase reporter assays
The 3'-untranslated region of the mouse Was gene was cloned into the XhoI and PmeI linkers of the dual-luciferase reporter vector psiCHECK-2 (Promega) as described previously (9). The following primer set was used: The reporter constructs (200 ng) were transfected together with miR-21 mimic (5 nmol/L) or mimic control in quadruplicate into HEK293T cells, previously plated (post 12 h) in 12well plates using Lipofectamine RNAiMAX (Invitrogen) as above. After 48 h, cells were harvested in 200 µL Glo Lysis Buffer (Promega) and the activities of both Renilla and firefly were measured. Each lysate (20 µl) was analyzed using Dual-Glo Luciferase reagents (Promega). Renilla luciferase activity was normalized to constitutive firefly luciferase activity for each well. Three independent experiments were performed.

Transfections in hPSC forward programming-derived megakaryocytes
In vitro production of megakaryocytes from hPSCs was carried out using forward programming (FoP) as described previously (10). Both hPSC lines were obtained from the Cambridge Biomedical Research Centre iPSC Core Facility. In short, hPSCs were transduced with replication-deficient lentiviral vectors to overexpress the 3 transcription factors FLI1, TAL1 and GATA1. After 2 days in a mesoderm promoting medium (Essential-6, Gibco) containing bone morphogenetic protein 4 (BMP4, 10ng/mL, Bio-Techne) and fibroblast growth factor 2 (FGF2, 20ng/mL, Bio-Techne) the cells were cultured in a megakaryocyte culture medium (CellGro, CellGenix) containing thrombopoietin (TPO, 20ng/mL, Bio-Techne) and stem cell factor (SCF, 25ng/mL, Gibco).   Supplemental Figure 4. Workflow for the analysis of fibroblast conditioned media. Conditioned media were collected from cardiac fibroblasts after transfections and subsequent TGF-β1 or control treatment. After removing cell debris by centrifugation, samples were filtered and concentrated using 3 kDa cutoff filter columns. Proteins were then deglycosylated by sequential incubation with deglycosylating enzymes. Samples were then reduced and denatured, following separation by polyacrylamide gel electrophoresis (SDS-PAGE). For immunoblotting, proteins were transferred onto nitrocellulose membranes. For mass spectrometry analysis, proteins in the gel were visualized by silver staining to support subsequent gel cutting. Each lane was cut into 12 gel bands without any gap in between. Proteins inside the gel pieces were digested using trypsin, followed by injection into an HPLC-coupled mass spectrometer for identification and quantification.      Figure 13. Effect of miR-21 inhibition on megakaryocyte maturity. Megakaryocytes (MK) were produced using forward programming (FoP) of hPSCs as previously described (10). Two independent FoP-MK lines were transfected with a non-targeting LNA oligonucleotide (control) or miR-21 inhibitor (LNA-21), conjugated to a cationic polymer (TurboFect, ThermoFisher) to enhance transfection efficiency.  Supplemental Table 3. ECM proteins identified in unstimulated or TGF-β1-stimulated murine CF secretome following miR-21 mimic or inhibitor (LNA-21) transfection. Secretome was analysed by gel LC-MS/MS analysis using an LTQ-Orbitrap XL. Values shown are mean±SEM of normalized spectral counts, based on four biological replicates for each condition. Differential expression was determined using a hierarchical Bayes estimation of generalized linear mixed effects model. FDR is calculated using an empirical Bayes method, with an FDR<0.05 considered significant. Proteins that yielded an FDR<0.05 for at least one condition are marked in bold.  Table 4. MiR-21 null heart ECM protein analysis. Analysis was performed using the 3-step extraction method, involving sequential incubation in NaCl, SDS and GuHCl to enrich for ECM proteins. Combined analysis of NaCl and GuHCl fractions by LC-MS/MS was annotated using the Matrisome database for murine ECM and ECM-associated proteins. Values indicate normalized total precursor intensity. FDR-adjusted p-values (q-values) were calculated using the 2-stage step-up method of Benjamini, Krieger and Yekutieli. Q for significant discovery was set to 5%.