Extracellular vesicles fail to trigger the generation of new cardiomyocytes in chronically infarcted hearts

Background: Extracellular vesicles (EV) mediate the therapeutic effects of stem cells but it is unclear whether this involves cardiac regeneration mediated by endogenous cardiomyocyte proliferation. Methods: Bi-transgenic MerCreMer/ZEG (n = 15/group) and Mosaic Analysis With Double Markers (MADM; n = 6/group) mouse models underwent permanent coronary artery ligation and received, 3 weeks later, 10 billion EV (from human iPS-derived cardiovascular progenitor cells [CPC]), or saline, injected percutaneously under echo guidance in the peri-infarcted myocardium. Endogenous cardiomyocyte proliferation was tracked by EdU labeling and biphoton microscopy. Other end points, including cardiac function (echocardiography and MRI), histology and transcriptomics were blindly assessed 4-6 weeks after injections. Results: There was no proliferation of cardiomyocytes in either transgenic mouse strains. Nevertheless, EV improved cardiac function in both models. In MerCreMer/ZEG mice, LVEF increased by 18.3 ± 0.2% between baseline and the end-study time point in EV-treated hearts which contrasted with a decrease by 2.3 ± 0.2% in the PBS group; MADM mice featured a similar pattern as intra-myocardial administration of EV improved LVEF by 13.3 ± 0.16% from baseline whereas it decreased by 14.4 ± 0.16% in the control PBS-injected group. This functional improvement was confirmed by MRI and associated with a reduction in infarct size, the decreased expression of several pro-fibrotic genes and an overexpression of the anti-fibrotic miRNA 133-a1 compared to controls. Experiments with an anti-miR133-a demonstrated that the cardio-reparative effects of EV were partly abrogated. Conclusions: EV-CPC do not trigger cardiomyocyte proliferation but still improve cardiac function by other mechanisms which may include the regulation of fibrosis.


CD63 and HSC70
After thaw, EV-CPC aliquots were concentrated again with ultrafiltration 0. 5 (Merck Millipore,PMNL 30Kda,Ref UFC 5030). The retentate was divided into two parts: one was directly   (BioRad, 1620094) during 20 min at 110 V, following which the membranes were incubated with Ponceau Red to verify the efficiency of the transfer process.
Membranes were blocked with 5% (w/v) milk in TBS supplemented with 0.1% Tween-20. To detect the protein of interest, membranes were then incubated with primary antibodies (Antihuman CD9, Merck Millipore CBL162) overnight at 4 °C, with constant agitation. After three 10-minute washes, membranes were incubated with a secondary anti-mouse antibody coupled with HRP (Amersham, GE Healthcare, 1/3000) for 1 h at room temperature. Immunodetection was performed using Clarity™ Western ECL Substrate, and the chemiluminescent signal was revealed using the ImageQuant™ LAS 4000. and the chemiluminescent signal was revealed using the ImageQuant™ LAS 4000.

Dose of EV-CPC
The doses used are expressed using different metrics according to MISEV guidelines for in vivo experiments. The EV-CPC secreted by an equivalent amount of 1.4x10 6 of CPC correspond to 1x10 10 (+/-3000) particles, as quantified by NTA and 100 µg of proteins, as measured by BCA (ThermoScientific). given for 14 days, followed by a 5-day wash-out before induction of infarction. In the C57BL/6J model, one part of the animals underwent an acute protocol with treatment implemented 3 days after infarction and sacrifice 2 days later for tissue collection and transcriptomic analysis; the remainder of mice went through the chronic protocol with sacrifice 4-6 weeks after treatment.

Generation of the MerCreMer/ZEG mouse model
Regardless of the timing, EV or a control Phosphate-Buffered Saline (PBS) solution were injected trans-cutaneously to avoid a second surgical procedure and its attendant mortality. To this end, mice were anesthetized with 2.5% isoflurane and immobilized on the VisualSonics (Amsterdam, The Netherlands) platform, as previously described [6]. were also co-delivered with EV. An equivalent volume of 45 μL was delivered in all PBSinjected control hearts. The different protocols are summarized in Figure 1.

Echocardiography
Three weeks after MI, all mice underwent a two-dimensional echocardiography (2 VEVO 2100, Visualsonics [Amsterdam, The Netherlands]). Images were acquired in long axis view in Bmode (VEVO Lab) and all parameters were calculated according to a previously described protocol [7]. All animals with a left ventricular ejection fraction (LVEF) > 45% were excluded.
They were then assigned to the control and experimental groups to ensure that baseline values of LVEF and left ventricular enddiastolic (LVEDV) and endsystolic (LVESV) were well balanced between groups, thereby minimizing biases in the subsequent assessment of treatment effects. Four to six weeks after treatment, cardiac echography was performed again under the same conditions by the same operator blinded to the treatment groups.

Magnetic Resonance Imaging (MRI)
To validate echocardiographic results with a more operator-independent technique, 11 mice in the MerCreMer/ZEG series (EV n = 5, PBS n = 6) underwent MRI imaging in a Bruker BioSpec (Bruker, Wissembourg, France) adapted for small animal studies with a 4.7 Tesla magnetic field. Images were acquired for 85 s each using a FLASH-cine IntraGate technology and 1.03 mm-thick slices were oriented along the heart's short axis. Image analysis was carried out using the Cvi42 software (Circle Cardiovascular Imaging Inc. Calgary, Canada) for Windows 7 by two blinded operators.

HL-1 cells and 3T3 cells culture
Murine cardiomyocyte cells (HL-1 cells generously provided by Pr. Claycomb, LSU Health, School of Medicine, New Orleans, LA) or NIH/3T3 fibroblasts (ATCC® CRL-1658™, ATCC, Molsheim, France) were seeded at 250,000 cells per well and cultured until confluence in a 6well plate. When confluence was reached, culture media was replaced with serum-deprived media and cells were treated with either EV-CPC (10x10 9 particles per well) or 0.1 µm filtered PBS for 24 h. Cells were then washed with filtered PBS and subsequently lysed with 800 µl QIAzol reagent (Qiagen, Courtaboeuf, France). Sample lysates were transferred into new tubes for miRNA extraction and quantification of miR 133-a by qRT-PCR.

Myocyte enhancer factor-2 (MEF-2) detection
Nuclear extraction of EV-CPC was performed using the Nuclear Extraction Kit (Abcam) and the extracts were quantified using BCA (ThermoScientific) for protein content. High throughput ELISA assay was used to quantify the MEF2 transcription factor (Abcam, Cambridge, UK) in EV-CPC.

Immunohistochemistry and immunofluorescence microscopy
After sacrifice, explanted hearts were washed in PBS, submerged in Optimal Cutting Temperature compound (OCT, Tissue-Tek; Sakura Finetek, Villeneuve-d'Ascq, France), frozen in liquid nitrogen and stored at -80 °C until sectioning. Hearts were cryo-sectioned into 5-10 µm thin-sections and an average of 12 tissue sections of different ventricular locations per heart were used for staining. To stain slides for Green Fluorescent Protein (GFP), cardiac Troponin T (TnT) and Fibroblast Activated Protein (FAP), fixation was performed with 4% paraformaldehyde for 15 min at room temperature (RT). Membrane permeabilization was achieved with 1% triton X-100 in PBS for 20 min at RT and non-specific epitopes were blocked with 5% bovine serum albumin (BSA) for 30 min at RT.

Triple GFP/TnT/ EDU staining
In the MerCreMer/ZEG mouse model, it was first necessary to check that after administration of Tamoxifen, mice expressed GFP exclusively in cardiomyocytes. A double staining for GFP and cardiac TnT was then performed with the relevant antibodies (FITC-conjugated anti-GFP, unconjugated anti-Troponin T and Texas red-conjugated secondary anti-rabbit antibody, all from Abcam). EdU was revealed according to the manufacturer's instruction by a click chemical reaction which yields a fluorescent dye (Cy5 Azide alternative). Nuclei were stained with 4', 6-diamidino-2-phenylindole (DAPI, Calbiochem, Lyon, France).

Double FAP/ TnT staining
Slides stained for FAP and TnT underwent a similar protocol using unconjugated anti-FAP (Fisher Scientific, Illkirch, France) and anti-TnT (Abcam) antibodies and Alexa 488-conjugated anti-rabbit (Life Technologies) and Alexa 594-conjugated anti-mouse secondary antibodies, respectively (Life Technologies). Nuclei were stained with DAPI.

Lectin Wheat Germ Agglutinin (WGA) and Isolectin staining
Slides stained for WGA and Isolectin underwent a similar protocol except that they were fixed for 10 min in acetone. Isolectin B4 and WGA were coupled to FITC and tetramethylrhodamine, respectively.

Hematoxylin/eosin staining
Thawed tissue sections were fixed in acetone for 10 min, washed 5 min in PBS and immersed for 2 s in hematoxylin and eosin (Sigma, Lyon, France). All slides stained were scanned with a slide scanner (NanoZoomer, Hamamatsu, Massy, France) and analyzed through the NDP view 2.5 software or the Metamorph® software (Molecular Devices, San Jose, CA) for fibrotic area calculation. Infarct size was calculated as the perimeter of the stained infarct area divided by the total perimeter of the section and expressed as a percentage.

Sirius Red staining
The Sirius Red staining was used to visualize and quantify fibrosis (area or interstitial).
Picrosirius Red stains collagen fibers type I and III in red. All other cellular and extra-cellular elements acquire a variable yellowish to pinkish unspecific coloration. Tissues were fixed in 3.7% formaldehyde for 10 min and then immersed in Picrosirius Red for 16 min. Dehydrated was achieved through three subsequent baths each of 100% ethanol and 100% xylene. Slides were then mounted with Eukitt.

Masson Trichrome staining
The Masson Trichrome staining was used to visualize and quantify the scar area, Tissues were fixed in 3.7% formaldehyde for 10 min, washed twice (5 min), then immersed for 10 min in hematoxylin, washed three times (3 min), immersed in Culvert Fuchsin for 5 min, and washed again twice for 3 min, immersed this time in 1% of phosphomolybdic acid for 7 min, colored in light green for 5 min, and immersed in 1% of acetic acid for 3 s. Dehydrated was achieved through two subsequent baths each of 80% and 100% ethanol and three subsequent baths each of xylene. Slides were then mounted with Eukitt.

Image analysis
Images were analyzed by Qcapture (QImaging, Surrey, Canada), observed using a Leica DM 2000 optical microscope and quantified manually using the Metamorph® or Image J softwares.
Ribonucleic Acid (RNA) expression on mice hearts or murine cultivated cells

RNA extraction
RNAs were extracted from heart tissue cryo-sections (200-400 µm) or cultured cardiomyocytes and fibroblasts following the manufacturer's instructions and the concentration and purity of samples were evaluated with the NanoDrop® Spectrophotometer (ThermoFisher Scientific, Illkirch, France).

Affymetrix microarray analysis
RNA quantification and quality control were performed using the HT Standard RNA LabChip Kit and the Caliper LabChip Microfluidics System (Perkin Elmer, Villebon sur Yvette, France).
One-hundred nanograms of total RNA were amplified, labeled, and fragmented using GeneChip WT. Each sample was hybridized onto Mouse Clariom D, washed, and stained with the Affymetrix® Fluidics Station 450. Array scanning was performed with the Affymetrix® GeneChip Scanner 3000 7G using the Command Console software (ThermoFisher Scientific, Illkirch, France) and then analyzed using the Affymetrix® rmasketch routine. Subsequently, differentially expressed genes (DEGs) were selected using oneway analysis of variance (ANOVA) using the Affymetrix Transcriptome Analysis Console (TAC) software v.1.0. The fold change (FC) (2 ≥ -2) of every gene, together with their corresponding p value (0.05) were used for selection of DEGs.

Real-time quantitative PCR (qRT-PCR)
qRT-PCR was used to assess the expression of genes involved in fibrosis, apoptosis, autophagy, proliferation and miR-133-a1 expression. The list of primers is indicated below. The primers    diastolic volume expressed at baseline (pre-transplantation, 3 weeks following MI) and at the end of study or B. As percent changes from baseline. Each group comprised 6 mice.