Bioactive Materials 6 (2021) 3314–3327 Contents lists available at ScienceDirect Bioactive Materials journal homepage: www.sciencedirect.com/journal/bioactive-materials Local administration of porcine immunomodulatory, chemotactic and angiogenic extracellular vesicles using engineered cardiac scaffolds for myocardial infarction Marta Monguió-Tortajada a, b, c, d, 1, Cristina Prat-Vidal a, d, e, 1, Miriam Moron-Font b, Marta Clos-Sansalvador b, c, Alexandra Calle f, Paloma Gastelurrutia a, d, e, Adriana Cserkoova a, Anna Morancho g, Miguel Ángel Ramírez f, Anna Rosell g, Antoni Bayes-Genis a, d, h, i, Carolina Gálvez-Montón a, d, 2, ***, Francesc E. Borràs b, c, j, 2, *, Santiago Roura a, d, k, 2, ** a ICREC Research Program, Health Science Research Institute Germans Trias i Pujol (IGTP), Can Ruti Campus, Badalona, Spain b REMAR-IVECAT Group, Health Science Research Institute Germans Trias i Pujol (IGTP), Can Ruti Campus, Badalona, Spain c Department of Cell Biology, Physiology and Immunology, Universitat Autònoma de Barcelona (UAB), Bellaterra, Spain d CIBERCV, Instituto de Salud Carlos III, Madrid, Spain e Institut d’Investigació Biomèdica de Bellvitge-IDIBELL, L’Hospitalet de Llobregat, Spain f Departamento de Reproducción Animal, Instituto Nacional de Investigación y Tecnología Agraria y Alimentaria (INIA), Madrid, Spain g Neurovascular Research Laboratory, Vall d’Hebron Research Institute (VHIR), UAB, Barcelona, Spain h Cardiology Service, Germans Trias i Pujol University Hospital, Badalona, Spain i Department of Medicine, UAB, Barcelona, Spain j Nephrology Service, Germans Trias i Pujol University Hospital, Badalona, Spain k Faculty of Medicine, University of Vic-Central University of Catalonia (UVic-UCC), Vic, Barcelona, 08500, Spain A R T I C L E I N F O A B S T R A C T Keywords: The administration of extracellular vesicles (EV) from mesenchymal stromal cells (MSC) is a promising cell-free Exosomes nanotherapy for tissue repair after myocardial infarction (MI). However, the optimal EV delivery strategy re­ Mesenchymal stem/stromal cells mains undetermined. Here, we designed a novel MSC-EV delivery, using 3D scaffolds engineered from decel­ Migration lularised cardiac tissue as a cell-free product for cardiac repair. EV from porcine cardiac adipose tissue-derived Infiltration MSC (cATMSC) were purified by size exclusion chromatography (SEC), functionally analysed and loaded to Cardiac tissue engineering scaffolds. cATMSC-EV markedly reduced polyclonal proliferation and pro-inflammatory cytokines production (IFNγ, TNFα, IL12p40) of allogeneic PBMC. Moreover, cATMSC-EV recruited outgrowth endothelial cells (OEC) and allogeneic MSC, and promoted angiogenesis. Fluorescently labelled cATMSC-EV were mixed with peptide hydrogel, and were successfully retained in decellularised scaffolds. Then, cATMSC-EV-embedded pericardial scaffolds were administered in vivo over the ischemic myocardium in a pig model of MI. Six days from im­ plantation, the engineered scaffold efficiently integrated into the post-infarcted myocardium. cATMSC-EV were detected within the construct and MI core, and promoted an increase in vascular density and reduction in macrophage and T cell infiltration within the damaged myocardium. The confined administration of multi­ functional MSC-EV within an engineered pericardial scaffold ensures local EV dosage and release, and generates a vascularised bioactive niche for cell recruitment, engraftment and modulation of short-term post-ischemic inflammation. Peer review under responsibility of KeAi Communications Co., Ltd. * Corresponding author. REMAR-IVECAT Group, Health Science Research Institute Germans Trias i Pujol (IGTP), Can Ruti Campus, Badalona, Catalunya, Spain. ** Corresponding author. ICREC Research Program, Health Science Research Institute Germans Trias i Pujol (IGTP), Can Ruti Campus, Badalona, Catalunya, Spain. *** Corresponding author. ICREC Research Program, Health Science Research Institute Germans Trias i Pujol (IGTP), Can Ruti Campus, Badalona, Catalunya, Spain. E-mail addresses: cgalvez@igtp.cat (C. Gálvez-Montón), feborras@igtp.cat (F.E. Borràs), sroura@igtp.cat (S. Roura). 1 These authors contributed equally to this work. 2 These authors contributed equally (senior co-authorship). https://doi.org/10.1016/j.bioactmat.2021.02.026 Received 31 October 2020; Received in revised form 16 February 2021; Accepted 17 February 2021 2452-199X/© 2021 The Authors. Publishing services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 1. Introduction quenched with α-MEM (Sigma Aldrich) supplemented with 10% heat-inactivated foetal bovine serum (FBS), 2 mM L-glutamine, 1% Ischaemic heart disease, including myocardial infarction (MI), is the penicillin-streptomycin (P/S) (all from Gibco Invitrogen Corp.) and 5 main cause of morbidity and mortality for both men and women μg/ml Plasmocin™ (Invivogen) (α-MEM-FBS). Cells were then centri­ worldwide [1]. The myocardial injury caused by MI triggers an in­ fuged at 1200×g for 10 min and cultured by adherence in α-MEM-FBS at flammatory reaction, extending myocardial damage but also a prereq­ 37 ◦ C 5% CO2. The purity of established cell cultures was analysed at uisite to transition toward myocardial healing and scar formation. The passages 4–8 by immunostaining with monoclonal antibodies against extent of the non-contractile scar formed would determine the proclivity CD105 (Abcam), CD44, CD45, CD14 (AbD Serotec), CD29, CD34, CD90 to adverse cardiac remodelling, cardiac malfunction and, ultimately, and CD106 or IgG isotype control (BD Pharmingen) using a FACSCanto overt heart failure [2]. The current best treatment for heart failure is II flow cytometer (BDBiosciencies) and the FACSDiva™ software (BD heart transplantation, but the great demand for organs generated by the Biosciencies). growing epidemic of cardiac-related diseases is presently unmet [3,4]. This urges therapies focused on preserving cardiac function after injury, 2.2. Extracellular vesicle production and isolation such as cell-based therapies and tissue engineering approaches aimed at modulating post-MI inflammation to mitigate myocardial scarring [5]. EV were produced and isolated as previously published [18,19]. All For that purpose, the administration of mesenchymal stromal cells relevant data regarding our experiments have been submitted to the (MSC) has been studied to promote healing of ischaemia-injured tissue EV-TRACK knowledgebase (EV-TRACK ID: EV200061) [20]. given their immunomodulatory and regenerative capabilities [6–9]. In Serum-derived bovine EV were depleted by ultracentrifuging 2 × this context, our laboratory was the first to pose attention to a popula­ α-MEM-FBS at 100,000×g for >16 h (TH641 rotor, adjusted k-Factor = tion of cardiac adipose tissue-derived MSC (cATMSC) which, despite 240.82) in a Sorvall WX Ultra 100 Series ultracentrifuge (Thermo Fisher residing in an adipocytic environment, do not differentiate into adipo­ Scientific). The supernatant was filtered through 0.22-μm (Sarstedt) for cytes but acquire de novo cardiac and endothelial markers in vitro (both sterilization, and diluted to a 1 × working solution with α-MEM human and porcine cells), and contribute to myocardial tissue revas­ medium. cularisation and infarct size reduction in vivo [6,7,10,11]. Moreover, For EV production, the culture media of two T-175 flasks of porcine their combination with decellularized cardiac tissue as natural cATMSC grown to confluence was replaced by 15 mL each of EV- three-dimensional (3D) scaffolds is a promising tissue engineering depleted α-MEM-FBS. After 48 h of culture, the conditioned medium strategy to locally deliver MSC and enhance their therapeutic effect to was harvested and centrifuged at 400×g for 5 min and 2000×g for 10 restore cardiac function after MI [7,10,12,13]. In pre-clinical studies, min to exclude cells and debris, respectively. Cells were counted at time these engineered cardiac grafts have significantly improved ventricular of EV harvesting (~1 × 107), and cell viability was >95% (trypan blue function, showing scaffold integration with the underlying myocardium, staining). The cleared supernatant was then centrifuged at 2000×g for and induced neovascularization and innervation [12,13]. Nevertheless, 35 min through a 100-kDa ultrafiltration device (Amicon Ultra centrif­ scarcely any cell grafted into the host myocardium. ugal filter; Millipore) to obtain concentrated conditioned medium At the same time, the current deficits understanding the mechanistic (CCM). action of cell therapy for cardiac repair and regeneration limits its po­ EV were then purified from the CCM by size-exclusion chromatog­ tential, optimization, and clinical implementation [14]. Numerous evi­ raphy (SEC) following our already published method [18,21,22]. dences suggest that MSC may promote MI healing thanks to the secretion Briefly, the SEC column was prepared packing 12 mL of Sepharose CL-2B of various soluble factors that boost endogenous repair in a paracrine (Sigma Aldrich) into a sterilized Puriflash dry load 12G flash column manner, rather than replicating and differentiating themselves [15,16]. (Interchim-Cromlab). Then, 500 μL CCM were loaded into the column Amongst them, there is growing interest in MSC’s release of extracellular and eluted with sterile 10% sucrose (w/v; Sigma Aldrich) as an elution vesicles (EV), a variety of lipid bilayered nanovesicles having the bio­ buffer, collecting 500-μL fractions immediately after loading. Protein logical characteristics of the originating cells, and containing a plethora elution was monitored by 280 nm absorbance (Nanodrop spectropho­ of functional lipids, RNA, and proteins [17]. EV can transfer these tometer; Thermo Fisher Scientific) or using the Pierce BCA Protein Assay bioactive molecules, and thus are valuable mediators of in intercellular kit (ThermoFisher), interpolating to a Bovine Serum Albumin (BSA; communication. In this context, we corroborated EV as functional Sigma Aldrich) standard curve. Bead-based flow cytometry was players of the paracrine immunosuppressive capability of MSC [18]. routinely performed to pool together the CD44+CD63+ EV-containing However, the mode of action of EV and the optimal administration route fractions (EV pool) for use in further experiments. to enable sufficient EV dose to be locally active remain unknown. In the present study, we explored the anti-inflammatory and healing effects of porcine MSC-EV, unravelling a new role of MSC-EV as re­ 2.3. Bead-based flow cytometry cruiters of endogenous pro-angiogenic cells, and developed novel po­ tency assays for porcine MSC-EV. Also, we designed a new method for Porcine cATMSC-EV were identified by bead-based flow cytometry, the local delivery of multifunctional MSC-EV over the ischemic screening for EV and MSC markers. EV were covalently coupled to 4-μm myocardium in a pig model of MI, by using a 3D-bioengineered scaffold aldehyde/sulphate-latex beads (Invitrogen-ThermoFisher Scientific) showing high EV-embedding capacity, local release and modulation of with a 15 min incubation, and then blocked for 2 h with BCB buffer (PBS, post-ischemic inflammation in vivo. 0.1% BSA, and 0.01% sodium azide (NaN3); Sigma Aldrich). EV-coupled beads were centrifuged at 2000×g for 10 min and re-suspended in BCB 2. Materials and methods buffer. Next, beads were incubated for 30 min at RT with the fluorochrome-conjugated antibodies anti-CD73-PE and anti-CD90-PE- 2.1. Porcine cardiac adipose tissue-derived MSC isolation and culture Cy7 (1:50; both from BD); or the primary Ab anti-CD9 (Clone VJ1/20; 1:10), anti-CD63 (Clone TEA3/18; 1:10), anti-CD81 (Clone 5A6; 1:10), Primary porcine cATMSC were obtained from pigs (Large White × anti-CD29 (1:10; BD), anti-CD44 (1:10; AbD Serotec) or IgG isotype Landrace) undergoing cardiac surgery (n = 39) by cardiac adipose bi­ control (1:10; Abcam) followed by incubation with the FITC-goat F(ab’) opsy (1.63 ± 0.69 g), as previously described [7]. Tissue samples were 2 anti-mouse IgG (1:10; Bionova) or A488-rabbit anti-rat IgG (1:100; washed from debris and red blood cells in phosphate-buffered saline AbD Serotec). EV-coupled beads were washed with BCB buffer and spun (PBS; Oxoid), and digested for 30 min with 0.05% collagenase (Type II; down at 2000×g for 10 min after each step. Data was acquired in a Gibco Invitrogen Corp.) at 37 ◦ C under mild shaking. Digestion was FACSVerse flow cytometer (BD), and analysed with FlowJo® v10 (BD). 3315 M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 2.4. Cryo-electron microscopy & Chemokine 9-Plex Porcine ProcartaPlex™ Panel 1 immunoassay (Invitrogen-ThermoFisher Scientific) in a Luminex 200 Instrument, and EV size and morphology were examined by cryo-transmission elec­ analysed with the xPONENT software (build 3.1.971.0) following tron microscopy (cryo-EM) as previously detailed [18]. manufacturers’ instructions. 2.5. Nanoparticle tracking analysis 2.10. Porcine outgrowing endothelial progenitor cell cultures and Size distribution and particle concentration was analysed in a characterization NanoSight LM10-12 instrument (Malvern Instruments Ltd), with NTA software (version 3.1, build 3.1.46) averaging three videos of 60 s at 30 Outgrowing colonies from endothelial progenitor cell cultures were FPS, with a camera level of 16, and manual control of temperature (21- isolated by processing fresh 20-mL venous blood samples (<2 h) from 25 ◦ C) of samples diluted to 20–120 particles/frame, as previously re­ pigs (Landrace × Large White; 30–35 kg; indistinct sex) by Ficoll ported [18]. gradient (GE Healthcare) as previously described [23]. Remaining erythrocytes were removed with lysis buffer (155 mM NH4Cl, 10 mM 2.6. EV staining NaHCO3, and 0.115 mM EDTA), PBMC were re-suspended in endothelial growth medium-2 (EGM-2; Clonetics®), seeded at 2 × 107 cells/mL in Fluorescently-labelled EV were produced by staining EV-secreting fibronectin-coated plates (10 μg/mL; Sigma-Aldrich), and cultured with cATMSC with the lipophilic dyes NIR815 (excitation/emission 786/ media replacement every 2–3 days. Late outgrowth endothelial cell 815 nm; Thermo Fisher Scientific) or PKH26 (excitation/emission 551/ (OEC) colonies with clonogenic capacity were visible after 14 days in 567 nm; Sigma Aldrich) following manufacturers’ instructions. Fluo­ culture, further expanded and frozen in liquid nitrogen until further use. rescently labelled cells were washed with PBS, cultured in α-MEM-FBS OEC immunophenotyping was performed by fixing cells with 4% and then changed to EV-depleted medium to produce fluorescently- paraformaldehyde (Sigma Aldrich) for 15 min, washing with 0.1% labelled EV, which were isolated from the conditioned medium using Tween-PBS (T-PBS), blocking for 1 h (1% BSA in T-PBS), and then the already described methodology. incubating at 4 ◦ C for 16 h with the primary antibodies anti-VEGFR2 (1:50; Santa Cruz Biotechnology), anti-CD31 (1:50; Abcam), and anti- 2.7. Endotoxin detection vWF (1:100; Sigma Aldrich). After washing with T-PBS, cells were incubated with the secondary antibodies AlexaFluor® 488 and Alexa­ The endotoxin concentration of cATMSC-EV preparations was Fluor® 568 (1:1000; Invitrogen) for 1 h at RT and counterstained with determined using the Pierce Chromogenic Endotoxin Quant Kit (Ther­ Vectashield® mounting medium with DAPI (Vector Laboratories, Inc.) moFisher) following the manufacturer’s instructions. Briefly, the endo­ prior to visualization in an Olympus BX61 microscope (Olympus). toxin concentration was determined for EV at 1 × 106 cATMSC/ml in a Alternatively, OEC were blocked with 25% AB human serum (Sigma- Limulus Amebocyte Lysate (LAL)-based chromogenic reaction using an Aldrich) in 1% BSA and 0.1% NaN3 in PBS for 15 min at RT, and labelled E. coli (011:B4) endotoxin standard curve and analysing different sample with anti-CD34-PE (BD) for 30 min at RT. After washing and filtering dilutions. The absence of LAL inhibition was confirmed in endotoxin through a 30-μm filter (Partec), cells were analysed in a FACSAria flow spike-in reaction. cytometer (BD) to confirm their hematopoietic/progenitor origin. The capacity of OEC to form vessel-like structures was evaluated in 2.8. Porcine peripheral blood mononuclear cells isolation and functional an in vitro Matrigel assay. 200 μL of ice-cold Matrigel™ (growth factor- assays reduced; BD) was placed in 24-well plates and incubated at 37 ◦ C for 30 min. Then, 4 × 104 OEC were seeded in EGM-2. After 24 h, images were Peripheral blood mononuclear cells (PBMC) were obtained by Ficoll taken using the Olympus IX71 microscope (Olympus). density gradient centrifugation (BD) of fresh whole blood samples (10- mL EDTA tubes; BD) from Landrace × Large White pigs (30–35 kg; indistinct sex). After harvesting, PBMC were counted and frozen. To 2.11. Migration assay monitor proliferation, thawed PBMC were labelled with Violet Cell­ Trace™ dye (Thermo Fisher Scientific) following manufacturer’s pro­ Cell migration driven by EV was measured in agarose spots using an tocol. Then, PBMC were seeded in 96-well plates at 1 × 105/well; 100 μl adaptation of a previously published protocol [24,25]. Beforehand, final volume of serum-free X–VIVO 15 media (Lonza) complemented wells dedicated to OEC were coated with 10 μg/mL fibronectin (Bio­ with 2 mM L-glutamine and 1% P/S. chrom) and left dry. In each quadrant drawn in wells of 6-well plates, we PBMCs were stimulated with 0.6 ng/mL PMA (phorbol 12-myristate placed three 5-μL droplets of 0.5% low-melting-point agarose (40 ◦ C, 13-acetate; Sigma Aldrich) and 200 ng/mL ionomycin (Io; Sigma Ecogen) mixed 1:1 (v/v) with either 10% sucrose, vascular endothelial Aldrich), in the presence or not of cATMSC-EV (pool of three cATMSC growth factor (VEGF; 50 ng/mL final concentration; Sigma Aldrich), or donors), at the indicated ratios according to EV-producing cell equiva­ cATMEC-EV (8 × 104 or 4 × 104/spot). The plates were placed at 4 ◦ C for lents (1 × 106/well, 2 × 105/well, 1 × 105/well or 4 × 105/well). After 5 15 min to promote gel solidification and spot adherence to the plate days of culture, supernatants were collected and frozen for downstream surface. Next, either cATMSC in α-MEM-FBS or OEC in EGM-2 complete cytokine production analysis, and the number of FSChighVioletlow medium (10% FBS), were poured into the corresponding wells and proliferating cells was assessed by flow cytometry with an LSRFortessa allowed to seed for 3 h at 37 ◦ C. To limit cell proliferation and favour (BD) and FlowJo® v10 software. chemoattractant detection by cells, culture media was replaced with low Alternatively, 1 × 105 PBMC were cultured together with cardiac serum supplemented medium (0.2% FBS). scaffolds (7.1 mm2) filled with peptide hydrogel and either cATMSC-EV Localizations of the border of the agarose spots were randomly (1 × 106 EV-producing cells/scaffold) or 10% sucrose buffer. After 5 days picked to take pictures (3 × 3 mosaic at 20 × objective) every hour for of culture, supernatants were harvested for cytokine production analysis. 21 h with an Axioskop Z1 microscope (Zeiss), under controlled tem­ perature and CO2 level. We compiled the images to videos, and random 2.9. Cytokine response cells were followed using the Manual tracking plug-in Fiji software (Image J 1.54w, NIH) and the Chemotaxis and Migration tool software The levels of IL1β, IL4, IL6, IL8, IL10, IL12p40, IFNα, IFNγ, and TNFα version 2.0 (Ibidi, Inc.) to quantify distance, velocity, and directionality were measured in the supernatants of cultured PBMC using the Cytokine of cell movement. 3316 M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 2.12. Tube formation assay CLx, detecting the near-infrared fluorescent dye NIR815 and auto­ fluorescence in the 800-nm and 700-nm channels, respectively. To OEC tube formation capacity was measured in 24-well plates pre- confirm cATMSC-EV scaffold loading right before implantation, peri­ coated with 200 μL/well of ice-cold Matrigel (Corning). OEC (2 × 104 cardial scaffolds were scanned in a Pearl Impulse Imager (Li-COR). cells/well) were seeded in complete EGM-2 media without VEGF (Lonza) in the presence of cATMSC-EV (2 × 106/ml, 1 × 106/ml or 4 × 2.16. Scanning electron microscopy 105/ml EV-producing cells), or VEGF (200 ng/mL; R&D systems) as a positive control. Cells were incubated for 6 h at 37 ◦ C 5% CO2 and im­ To confirm cATMSC-EV presence within scaffolds, cardiac scaffolds ages were taken with a phase contrast inverted microscope at 10 × loaded with NIR815-labelled EV were maintained for 1 week at 37 ◦ C, objective magnifications and analysed using the Fiji (ImageJ) Angio­ 5% CO2, after which they were washed with PBS and processed for SEM. genesis Analyser plugin. Briefly, after fixation with 10% formalin (Sigma Aldrich) and washing with distilled water, scaffolds were dehydrated in ethanol solutions of 2.13. Generation of cardiac scaffolds increasing concentrations, and dried using a CO2 critical point dryer (EmiTech K850; Quorum Technologies), as detailed previously [26]. Cardiac scaffolds were generated by decellularisation of either The scaffolds were then sputter-coated in gold using the JFC 1100 ion human pericardial or porcine myocardial tissues. The formers were 19 sputter (Jeol), and scanned under a JSM-6510 scanning electron mi­ healthy pericardial samples from patients (n = 19; 16 males, 3 females; croscope (Jeol) at 15 kV. mean age, 65 ± 12 years; range, 44–85 years) who were undergoing cardiac interventions at our hospital, and who had given their signed 2.17. Myocardial infarction pig model consent. The local ethics committee revised and approved this study, whose protocols conformed to the principles outlined in the Declaration Animal studies were approved by the local Animal Experimentation of Helsinki. Additionally, we used a total of 21 myocardial scaffolds Ethical Committee and Government Authorities (Generalitat de Cata­ obtained from porcine heart left ventricles from a slaughterhouse lunya; Code:10547), and comply with the guidelines defined by the (Landrace × Large White pigs; n = 4). Guide for the Care and Use of Laboratory Animals [27]. Eight crossbreed We proceeded with a detergent-based decellularisation (SDS and Landrace x Large White pigs (31.3 ± 1.5 kg) were subjected to MI and Triton X-100) of either cardiac tissue, followed by DNase I treatment, randomly distributed into control (n = 4; implantation of lyophilisation, and sterilization as previously detailed by our group [7, hydrogel-alone filled scaffold) or treated group (n = 4; implantation of 26]. Scaffold size was standardised to 7.1 mm2 and 50.3 mm2 using cATMSC-EV with hydrogel-filled scaffold). disposable 3-mm and 8-mm biopsy punches respectively, or to 2 cm2 by Animals were sedated with an intramuscular (IM) injection of atro­ using a scalpel. pine (0.04 mg/kg; BBraun) and pre-anesthetized with dexmedetomidine (0.03 mg/kg, IM; Dexdor®, Orion Pharma), midazolam (0.3 mg/kg, IM; 2.14. EV delivery into decellularised scaffolds Laboratorios Normon), and butorphanol (0.3 mg/kg, IM; Butomidor®, Richter Pharma AG). Then, anaesthetic induction was performed with The EV pool, corresponding to a specified number of producing an intravenous (IV) bolus of propofol (2 mg/kg; Propovet®, Zoetis). cATMSC, was concentrated with a sterile 2-mL 100-kDa ultrafiltration Animals were endotracheally intubated, and anaesthesia was main­ device (Merck Millipore) to adjust its final volume according to the size tained by 2% isoflurane (IsoVet®, BBraun) inhalation. Fentanyl (0.075 of the intended scaffold, using 10% sucrose buffer (see Fig. S5). Adjusted mg/kg/45 min, IV; Fentadon®, Dechra) and an IV 1.5 mg/kg atracurium EV volume was then mixed (1:1, v/v) with 0.3% PuraMatrix® self- besylate bolus (Sanofi Aventis S.A.) were administered during inter­ assembling peptide hydrogel (Corning) in 10% sucrose, and this vention. After a left lateral thoracotomy, a double-ligation of the first mixture was laid over the lyophilised, decellularised scaffolds to both marginal branch of the circumflex artery, 1.5 cm distally from the rehydrate and fill the scaffolds with EV. As control, 10% sucrose buffer atrioventricular groove (Optilene 5/0 W-8556 12-S, Ethicon Inc.) was was mixed (1:1, v/v) with the peptide hydrogel. Specifically, a 4.2 μl mix performed to induce MI. After 30 min, a pericardial scaffold, with or (containing hydrogel and cATMSC-EV from 1 × 106 EV-producing cells) without NIR815-labelled cATMSC-EV (from 20 × 106 EV-producing was added to 3-mm ∅ scaffolds and used to analyse its influence on cells), was placed over the infarcted myocardium and secured with allogeneic PBMC cytokine response; a 100 μl mix (containing hydrogel 0.1–0.2 mL of surgical glue (Glubran2®, Cardiolink S.L.). Tulatromicin and cATMSC-EV from either 1 × 105 or 1 × 106 EV-producing cells, (2.5 mg/kg, IM; Draxxin®, Pfizer Animal Health) was administered at previously stained with either NIR815 or PKH26) was added to 8-mm ∅ the end of the surgery as antibiotherapy and a transdermal fentanyl scaffolds to assess EV retention by NIR815-EV scanning, PKH26-EV patch was applied as an analgesic post-operative care (Durogesic®, confocal microscopy analysis or by scanning electron microscopy Janssen-Cilag). Finally, pigs were recovered, and housed until experi­ (SEM); and a 200 μl mix (containing hydrogel and EV from 20 × 106 EV- mental end-point. producing cATMSC previously stained with NIR815) was added to 2 cm2 scaffolds for in vivo implantation over the ischemic area of the MI 2.18. Tissue collection, morphometric and histopathological analyses porcine model. After 15 min of seeding and scaffold rehydration at RT, culture medium was added over scaffolds to promote the salt-triggered Six days after MI induction, animals were euthanized by adminis­ peptide folding and consequent self-assembling hydrogel jellification for tering sodium thiopental overdose (200 mg/kg) and porcine hearts were a minimum of 30 min and up to 2 h at 37 ◦ C before use. excised, then scanned in a Pearl Impulse Imager (Li-COR), sectioned and scanned again for EV tracing within heart sections (Fig. S7). Heart 2.15. Near-infrared scaffold scanning samples, including infarct core, border zones and healthy myocardium were either fixed in 10% formalin for paraffin inclusion or embedded in To check cATMSC-EV retention in scaffolds and tracking in vivo, EV OCT (VWR) and snap-frozen in liquid nitrogen-cooled isopentane. were produced from NIR815-labelled cATMSC. EV-associated labelling Descriptive histopathological analysis and scar fibrosis were ana­ was confirmed by performing a dotblot assay of each SEC elution fraction lysed in 5-μm paraffin sections with Masson’s trichrome and Picrosirius and of the concentrated EV pool, placing 5 μl/spot by superimposing 1 μl- red staining, respectively, as described previously [28]. Total collagen, drops, and scanning for NIR815 fluorescence in an Odyssey CLx (LI-COR and specifically collagen type I and III were quantified in six random Biosciences, Inc.). Then, cATMSC-EV retention was monitored by scan­ fields of each tissue section using Image-Pro Plus software (6.2.1 ning NIR815-labelled EV-loaded 8-mm ∅ cardiac scaffolds in an Odyssey version; Media Cybernetics, Inc.). 3317 M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 2.19. Immunohistochemical analysis 2.20. Statistical analysis To examine cATMSC-EV retention within scaffolds by confocal mi­ Data is shown as mean with SD unless stated otherwise. Statistical croscopy, immunostaining was performed in PKH26-EV loaded scaf­ differences are shown when found significant (p < 0.05) applying the folds. Specifically, cardiac scaffolds with PKH26-labelled EV incubated appropriate statistical tests for each dataset after checking for normality for 2 h were thoroughly washed 3 times with PBS, fixed with 10% of data, using Graphpad Prism (6.0 version) and SPSS 21.0.0.0 (SPSS, formalin for 1 h at RT, rinsed 3 times with PBS, embedded in OCT and Inc, Chicago, IL) softwares. snap-frozen in liquid nitrogen-cooled isopentane. Also, decellularisation was confirmed by performing immunostaining on 10-μm-thick native 3. Results and decellularised tissue cryosections. Succinctly, after blocking for 1 h at RT with TBS (Tris-buffered saline)/0.2% Triton-X-100/10% horse 3.1. Porcine cATMSC-EV purification serum/1% BSA (all from Sigma Aldrich), tissue sections were incubated overnight at 4 ◦ C with phalloidin-Atto 565 (Phal565; 1:500; Sigma) and Primary cultures of porcine cATMSC were established and validated primary antibody against cardiac troponin I (cTnI; 1:100; Abcam) and as positive for CD29 (84%), CD44 (88%), CD90 (99%) and CD105 (92%) type-III collagen (col-III; 1:100; Abcam). Sections were then incubated and negative for CD14, CD34, CD45 and CD106 [7]. EV were then iso­ for 1 h at RT with Cy2-, Cy3-, and Cy5-conjugated secondary antibodies lated from cell- and debris-cleared conditioned media of porcine (1:500; Jackson ImmunoResearch Laboratories), respectively, and cATMSC cultures by SEC. cATMSC-EV were screened for EV markers nuclei counterstained with DAPI (0.1 μg/mL; Sigma-Aldrich). (CD9+ CD63+ CD81low CD5Llow) and MSC markers (CD44+ CD29low Vascular density in the MI pig model was analysed in 10-μm OCT- CD73low CD90low) (Fig. S1), being CD63 and CD44 the most relevantly embedded tissue cryosections immunostained with biotinylated Griffo­ and consistently detected. Thus, they were routinely used to track nia simplicifolia Lectin I isolectinB4 (IsoB4; 1:50; Vector Labs) and porcine cATMSC-EV elution in the SEC fractions (Fig. 1A). Isolation by Streptavidin-Alexa488 (1:500; Invitrogen). Macrophage, B and T cell SEC ensured EV purification from the rest of soluble factors present in infiltration was also analysed by immunostaining tissue cryosections cATMSC conditioned media, as previously published [18]. We pooled using cTnI (Abcam) and CD163 (Novus Biologicals) or CD20 (Biocare the CD63+CD44+ fractions, containing porcine cATMSC-EV, and medical), CD3 and CD25 (both from AbD Serotec) primary antibodies confirmed by cryo-EM the double membrane, round shape, and nano­ (1:100), followed by incubation with the Alexa Fluor 488-, Alexa Fluor vesicle size of cATMSC-EV (Fig. 1B and C). Furthermore, EV mean 647- (Molecular Probes) and Cy3- (Jackson ImmunoResearch Labora­ diameter by NTA was 211.2 (14.66) nm (Fig. 1D and E), and the con­ tories) conjugated secondary antibodies (1:500) and nuclei counter­ centration of particles according to NTA was homogeneous after nor­ staining with DAPI. malisation by the EV-producing cell numbers at EV harvest (Fig. 1E). Immunofluorescence images were acquired in an Axio-Observer Z1 Protein concentration in EV pools was really low (0.554 (0.391) μg total confocal microscope (Zeiss). Quantitative measurements of blood vessel protein; 6.94 × 10− 02 (7.13 × 10− 02) μg/106 cells at EV harvest), and did area (IsoB4 staining) in either the infarct core or scaffold were per­ not correlate well with the EV-producing cell numbers (Fig. S2), thus it formed using Image-Pro Plus software in six different random fields per was not used to normalise EV numbers. Instead, we used the section. Quantification of CD163+, CD20+, CD3+ and CD25+ cells, was EV-producing cell numbers at harvest as a proxy of EV quantitation, as carried out by two independent investigators measuring at least four we and others previously reported [18,29,30]. different optical fields from each section. Fig. 1. EV were isolated from porcine cATMSC by SEC, and phenotypical and morphologically characterised. (A) Representative SEC elution profile of porcine cATMSC-EV, found in the fractions positive for the EV and MSC markers CD63 and CD44 (MFI, median fluorescence intensity by bead-based flow cytometry). The protein eluted later, measured by BCA. (B) Cryo-EM confirmed that porcine cATMSC-EV were double membrane round nanovesicles of 50–300 nm. Scale bars are 500 nm and 200 nm, respectively. (C) Histogram depicting the size distribution of cATMSC-EV according to cryo-EM (n = 280). (D) Representative histogram depicting size and concentration of EV according to NTA analysis. (E) EV diameter by NTA was around 200 nm. Concentration of particles according to NTA, normalised by the EV-producing cell number. 3318 M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 3.2. cATMSC-EV have immune suppressive properties room facilities would further reduce the endotoxin levels to fulfil reg­ ulatory requirements. To test the immunomodulatory properties of cATMSC-EV, we ana­ lysed their potency in terms of proliferation and cytokine secretion of 3.3. Porcine cATMSC-EV recruit allogeneic pro-regenerative cells and allogeneic porcine PBMC modulation by the presence or absence of promote angiogenesis cATMSC-EV. PBMC were activated with PMA plus Io, that simulated the pro-inflammatory environment found in the ischemic injury post-MI, The success of the local delivery of EV for myocardial repair relies on which led to high PBMC proliferation and an inflammatory cytokine the hypothesis that EV will recruit endogenous cells towards the response (Fig. 2). When cATMSC-EV were added to the mix, they dose- infarcted tissue, and among other beneficial effects, promote revascu­ dependently suppressed the proliferation of allogeneic PBMC (Fig. 2A). larisation of the injured myocardium. To start validating this hypothesis Notably, the capacity of cATMSC-EV to inhibit PBMC proliferation was in vitro, we quantitatively analysed the chemotactic action of cATMSC- not dependent on the specific EV pool, but rather on the responder EV towards allogeneic OEC and MSC in agarose spots, using an adap­ PBMC donor. cATMSC-EV also abrogated the secretion of the pro- tation of a previously described protocol [24,25], and their effect on the inflammatory cytokines IFNγ, TNFα, and IL12p40 in stimulated PBMC tube formation capacity of OEC. (Fig. 2B). IL10 was also reduced by cATMSC-EV. However, the lowest EV Swine OEC were isolated from porcine peripheral whole blood dose maintained IL10 production while reducing the pro-inflammatory samples. We confirmed them as genuine hematopoietic vascular pre­ cytokine levels, thus increasing the IL10:IFNγ ratio. IL6 and IL8 secre­ cursors with specific endothelial-lineage marker expression, such as tion was not modified by cATMSC-EV. Other cytokines, including IL1β, VEGFR2, vWF, CD31 (immunocytofluorescence; Figs. S3A–C), and CD34 IL4, and IFNα, were found at negligible levels (<10 ng/ml) and did not (flow cytometry; Fig. S3D), and formed vascular-like networks when change by cATMSC-EV presence (data not shown). Notably, the lack of seeded in Matrigel (Fig. S3E). cytokine response to cATMSC-EV from allogeneic PBMC suggested an Then, we analysed the active migration of MSC and OEC towards absence of alloreactivity. The absence of TNFα induction, and actually agarose spots containing increasing doses of cATMSC-EV. While no cells reduction by cATMSC-EV, suggested that the EV products had low entered the spots containing buffer alone, MSC and OEC invaded the endotoxin levels, as corroborated by endotoxin testing (<2 EU/ml). spots containing cATMSC-EV in a dose-dependent manner (Fig. 3 and Nevertheless, this is an important issue to inspect before clinical trans­ Supplementary Videos 1–6). Of note, OEC also migrated inside the spots lation. EV production under GMP conditions and manufacturing in clean containing VEGF, used as a positive control. OEC and MSC tracking over Fig. 2. Porcine cATMSC-EV abrogate poly­ clonal proliferation and modulate the cyto­ kine response of allogeneic PBMC. (A) Two representative EV pools (#1 and #2) were tested in potency assays (doses of 2 × 105, 1 × 105 and 5 × 104 EV-producing cells) against two different fully allogeneic PBMC responders (1 × 105 cells) with polyclonal stimulation (PMA + Io; Control +). Prolif­ erating PBMC were determined according to dye dilution as the percentage of FSChigh e670low cells. Data are shown as mean ± SD. Statistical differences to the positive control (PMA + Io) are indicated as *p < 0.05; **p < 0.01; ***p < 0.001 by Two-way ANOVA with Tukey’s multiple comparison test. (B) We assessed the cytokine levels in superna­ tants of the two different fully allogeneic PBMC in the presence or absence of two different cATMSC-EV pools (10:1, 2:1, and 1:1 cATMSC-EV:PBMC ratios) at five days after PMA + Io stimulation. Bars indicate mean ± SD. *p < 0.05; **p < 0.01 by one- way ANOVA with Dunn’s multiple compar­ ison test. 3319 ­ M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 24 h indicated a positive directionality in the cell movement towards the 3.4. Biocompatible 3D-Engineered scaffolds maintain the internal EV-containing spots, demonstrated by an increased Directness of ultrastructure and are devoid of cells after decellularisation movement and Euclidean distance (Δxy) of cells, while no changes in the Accumulated distance (ΣΔxy) or Velocity of cells was found (data not Two decellularised cardiac scaffolds were generated from either shown). Also, this was only observed in OEC and MSC in contact with the human pericardial or porcine myocardial tissues, with their intrinsic agarose spot (border versus remotely positioned cells; Fig. 3). structure preserved, including the spatial 3D organization of the native Supplementary video related to this article can be found at https:// matrix fibrils (Figs. S4A–D). Complete decellularisation was confirmed doi.org/10.1016/j.bioactmat.2021.02.026 in both pericardium and myocardium by the absence of actin Secondly, cATMSC-EV were able to promote angiogenesis on allo­ (Figs. S4E–F) and cardiac troponin I (Figs. S4G–H), respectively, and geneic porcine OEC in a dose-dependent manner (Fig. 4). The number of removal of cell nuclei in both acellular scaffolds (Figs. S4F and H). nodes and branch length increased with EV dose, and thickened Additionally, the positive staining for type-III collagen (Figs. S4E–H), vascular-like meshes were observed in the highest EV doses analysed. Of proved the preservation of extracellular matrix (ECM) protein compo­ note, the EV concentration range was equal to that used in the prolif­ nents following decellularisation, as previously described [12]. More­ eration assays. over, as we previously published, a proteomic characterization of decellularised pericardial and myocardial scaffolds revealed the pres­ ence of matrisome/ECM-related proteins, including major cardiac ECM components like ECM glycoproteins and regulators, collagens and pro­ teoglycans, consistent to their corresponding native tissues [12]. Fig. 3. Porcine cATMSC-EV recruit allogeneic swine cATMSC and swine peripheral blood OEC. (A–C) Allogeneic cATMSC were added to plates with agarose spots containing increasing doses of cATMSC-EV (from 4 × 104 and 8 × 104 cells), and cell migration was assessed over 24 h. (D–F) OEC were added to plates with agarose spots containing VEGF (positive control) or cATMSC-EV (from 8 × 104 cells), and cell migration was assessed over 24 h. (A, D) Representative images of agarose spot borders (dotted line) at 0 and 24 h. Migrating cells are observed only inside spots containing EV or VEGF. Scale bars = 100 μm. (B, E) The corresponding rose plots of cell movement during the 24-h time-lapse. Arrow cap indicates direction to the spot. The border and remote positions were determined at 0 h for cells localised or not in the border of the agarose spot, respectively. (C, F) Directness and Euclidean distance (Δxy) of cell movement. Pictures were taken hourly, tracking ten cells in each position of at least three spot positions tracked for each condition. (A–C) N = 2 independent experiments. *p < 0.05, **p < 0.01, and ***p < 0.001 by two-way ANOVA with Tukey’s post-hoc analysis. (D–F) N = 3 independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001 by two-way ANOVA with Tukey’s post-hoc analysis. 3320 M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 Fig. 4. Porcine cATMSC-EV promote angiogenesis to allogeneic swine peripheral blood OEC. OEC were seeded in Matrigel-coated wells, with increasing doses of cATMSC-EV (2 × 106/ml, 1 × 106/ml or 4 × 105/ml producing cells) or VEGF (200 ng/mL). (A) Representative images at 6 h, showing an increased number of junctions and thickening of the vascular-like networks with increasing EV doses. (B) Tube formation was quantified in terms of Number of nodes (top) and Branching length (bottom). Data includes 2 inde­ pendent experiments, each with 3 biological repli­ cates and 2 technical replicates each. **p < 0.01, and ***p < 0.001 and ****p < 0.0001 by One-way ANOVA with Tukey’s post-hoc analysis. 3.5. Porcine cATMSC-EV are successfully retained within 3D-Engineered associated labelling was confirmed by performing a dotblot of each SEC cardiac scaffolds elution fraction, scanning fluorescence at 800 nm (Fig. 5A). The NIR815 fluorescent signal was increased after concentrating the EV pool by 100 To assess the feasibility of EV delivery through confinement in 3D- kDa ultrafiltration (Fig. 5A, right). Then, the concentrated EV pool was engineered cardiac scaffolds, we first evaluated EV retention within mixed with self-assembling peptide hydrogel, and used to fill and the scaffolds. For that, we loaded the decellularised, lyophilised scaf­ rehydrate lyophilised pericardial and myocardial scaffolds. After the folds with cATMSC-EV mixed with self-assembling peptide hydrogel, salt-triggered hydrogel jellification, scaffolds were thoroughly washed necessary to enhance scaffold rehydration and ensure filling with EV with PBS and near-infrared fluorometric scanning revealed cATMSC-EV (Fig. S5). EV were fluorescently stained for tracking with lipophilic dyes, retention in both 3D-bioengineered cardiac scaffolds (Fig. 5B). Note­ but to avoid troublesome EV washing and carry-over of lipophilic dye worthy, pericardial scaffolds held higher amounts of EV according to aggregates from labelled EV, the EV-producing cATMSC were labelled fluorescence intensity quantification and visible in the scanned images, instead, enabling the tracking of fluorescently-labelled EV. NIR815 EV- observing an earlier saturation of myocardial scaffolds (Fig. 5C). Of 3321 M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 Fig. 5. Engineered cardiac scaffolds retained cATMSC-EV, as shown by fluorometric scanning, confocal microscopy and SEM imaging. (A) Dotblot of SEC fractions eluting NIR815-labelled porcine cATMSC-EV, scanning fluorescence at 800 nm. EV-containing fractions (squared) were confirmed NIR815+, pooled together and concentrated by 100 kDa-ultrafiltration, maintaining NIR815 fluorescence (right). (B–C) NIR815-labelled cATMSC-EV mixed with peptide hydrogel and added to pericardial and myocardial scaffolds were detected after thorough washing of scaffolds. (B) Representative near-infrared fluorescent scanning images of pericardial and myocardial scaffolds (autofluorescence at 700 nm, red) embedded with NIR815-cATMSC-EV (NIR815 fluorescence at 800 nm, green). Scale bar = 1 cm. (C) NIR815-cATMSC-EV fluorescence quantification expressed as relative to hydrogel-alone filled scaffold (background). (D) PKH26-labelled cATMSC-EV (red) were detected inside the scaffolds by confocal imaging, after immunostaining for collagen III (col III; grey). Scale bar = 20 μm. (E) SEM imaging confirms retention of porcine cATMSC-EV within scaffolds after washing and week-long culture, appearing as 100–120 nm protuberances embedded in the hydrogel filling the scaffolds. Controls are scaffolds filled with peptide hydrogel alone. Scale bar = 1 μm. note, both scaffolds released the EV with time (Fig. S6). 3.6. Porcine cATMSC-EV combined with cardiac scaffolds abrogate the To corroborate cATMSC-EV retention inside the scaffolds, we loaded cytokine response of allogeneic PBMC them with PKH26-labelled cATMSC-EV, following the same aforemen­ tioned procedure. Indeed, confocal microscope imaging of scaffold mid- To examine the immunomodulatory capacity of cATMSC-EV sections revealed cATMSC-EV presence within the scaffolds (Fig. 5D). embedded in cardiac scaffolds, we studied the cytokine response of Moreover, even after thorough washing and a week-long culture at allogeneic PBMC, stimulated or not with PMA plus Io, and cultured with 37 ◦ C, cATMSC-EV were found inside the scaffolds as confirmed by SEM, cardiac scaffolds filled with hydrogel alone or mixed with cATMSC-EV. distinguishable as 100- to 120-nm protuberances embedded in the The presence of the scaffolds themselves already weakened cytokine hydrogel filling the scaffolds (Fig. 5E). secretion by stimulated PBMC (Fig. S7), which could be attributed to 3322 M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 both biological effects or to technical reasons, as scaffolds would be (0.058) vs 0.93% (0.410) IsoB4 area in Control vs Treated animals; p = greatly hindering PBMC cell contact. There were only a few exceptions: 0.031). Regarding the MI area, although not significant, less fibrotic area IL6 and IL8 were secreted by stimulated PBMC in the presence of was observed in the infarct core (Fig. 7C; 0.46% (0.35) vs 0.13% (0.06) myocardial scaffolds, which was increased with cATMSC-EV, while very Collagen I area in Control vs Treated animals; p = 0.2). Moreover, low levels of IFNγ and TNFα were found. Notably, only IL6 secretion was treated animals had a marked decrease of infiltrating monocytes found in the presence of cATMSC-EV embedded in pericardial scaffolds, (CD163+) in the infarct core (Fig. 7D–E) (26.38% (2.3) vs 1.942% (2.91) whereas none of the other pro-inflammatory cytokines were detected. in Control vs Treated animals; p = 0.002), whereas no differences were Therefore, pericardial scaffold was the preferred construct to be tested in found within the scaffold itself (32.7% (9.26) vs 30.08% (12.60); p = vivo due to its improved EV retention and low immunoreactivity. 0.78). Regarding B cell infiltration, it was rare and no differences were found (Fig. 7F–G). On the other hand, the decrease in macrophages was 3.7. Locally administered porcine cATMSC-EV stay within scaffolds and accompanied with a reduction in T cell infiltration, observed in the infarct core infarcted tissue (Fig. 7F–G) (27.23% (7.86) vs 11.66% (0.97) in Control vs Treated animals; p = 0.008), but not significant in the scaffold area To assess safety and biodistribution of the cATMSC-EV once (22.61% (12.78) vs 16.66% (4.59) in Control vs Treated animals; p = administered in vivo, a study was performed using the acute MI pig 0.53). No significant differences were found either in the amount of model. Pericardial scaffolds containing NIR815-labelled cATMSC-EV CD25+ T cells. mixed in peptide hydrogel were implanted over the ischemic area of treated animals, and were compared to the use of a hydrogel-alone filled 4. Discussion scaffold in control animals. Fluorescent signal of scaffolds loaded with NIR815-cATMSC-EV was confirmed before implantation (Fig. S8). One In the present study, we deepen into the beneficial properties of animal died after ventricular fibrillation during MI induction and one porcine MSC-EV isolated by SEC for reparative medicine, set up poten­ exhibited no evidence of MI and was excluded from the final analysis, tial immunomodulatory, chemotactic and pro-angiogenic potency as­ which included three treated and three control animals. Six days after says and unravel a new role of porcine MSC-EV as recruiters of implantation, porcine hearts and sections, as detailed in (Fig. S9), were endogenous pro-angiogenic cells. At the same time, we describe the scanned to track the presence of NIR815-labelled cATMSC-EV. They development of a new generation of cell-free cardiac grafts comprising were detected in both the graft and infarct core, and had persisted only MSC-derived EV embedded with peptide hydrogels in cardiac decellu­ within the damaged area of treated animals (Fig. 6 and Fig. S10). larised scaffolds for local myocardial delivery. Cardiac tissue engineering, which is based on the scaffolding of reparative cells and/or bioactive factors with supportive materials, is a 3.8. In vivo scaffold neovascularization and reduction of macrophage and potential alternative therapy to limit adverse ventricular remodelling, T cell infiltration and ultimately preserve heart function post-MI [31]. In this context, decellularised cardiac scaffolds are a source of biological ECM derived All the scaffolds integrated to the myocardium, showing neo­ from native cardiac tissues, with conserved micro- and macro-scale ECM innervation and neovascularization regardless of EV loading (Fig. S11) structure and functional cardiac ECM composition [32,33]. Specifically, as previously observed [13]. Nevertheless, cATMSC-EV presence we already have a protocol approved by the Spanish Agency of induced greater vascular density within scaffolds (Fig. 7A–B; 0.15% Fig. 6. Porcine NIR815-cATMSC-EV detection in the scaffold and infarct area in vivo after six days post-implantation. Representative images of one treated animal are shown. (A, B) Anterior side of a treated pig heart six days after MI. Heart sectioning is indicated by yellow dotted line. (A) The scaffold (not visible due to the tissue coverage) was implanted following the coronary branch descending from the MI ligation (*). (B) Fluorescent scanning and bright field photography of the region of the heart indicated in (A). The NIR815-EV fluorescence (green; 800 nm) could be detected through, and despite the covering tissue, where the scaffold was implanted. (C) The scaffold (white dotted line) and MI area (black dotted line) can be observed in the transversal cut of the 1.5 cm-heart section spanning from the MI ligation to the half of the scaffold. (D) Administered NIR815-cATMSC-EV (green; 800 nm) are detected within the scaffold and MI core (red; autofluorescence; 700 nm) by fluorescent tracking. Both fluorescent signals are overlaid on a white light image of the heart section. (E) Close-up image of the scaffold (white dotted line) and MI area (black dotted line). (F) Pseudo-colour intensity image depicting NIR815-EV signal at the 800 nm channel within the scaffold (black dotted line) and infarct core. (G) Quantification of NIR815 signal in the infarct and scaffold area. Bars indicate mean ± SD. p < 0.001 difference in the infarct core and p < 0.01 in the scaffold area between Control (n = 3) and Treated animals (n = 3) by Two-way ANOVA. 3323 M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 Fig. 7. Porcine cATMSC-EV induce more neovascularization within the scaffold and less infiltrating macrophages in the infarct area. (A) Representative images of scaffold areas from a Control and Treated animals stained with Isolectin-B4 (green) showed an increase in vascular density when cATMSC-EV were loaded to the scaffolds. (B) Percentage of vessel area within the scaffold. (C) Percentage of collagen I area within the infarct area, assessed by Sirius Red staining. (D) Immu­ nohistofluorescence against CD163 (red), cTnI (white), and DAPI (blue) of Control and Treated animals of the infarct core show a reduced macrophage infiltration when cATMSC-EV were implanted with scaffolds. Arrows indicate macrophages (CD163+) found within the infarct core. (E) The number of CD163+ cells per field within the infarct core. *p < 0.05; **p < 0.01 by T-Student test. (F) Immunohistofluorescence against CD3 (red), CD25 (white), CD20 (green) and DAPI (blue) of Control and Treated animals of the infarct core show a reduced T cell infiltration when cATMSC-EV were implanted with scaffolds. Arrows indicate activated T cells (CD3+CD25+). (G) The number of infiltrating B cells (CD20+) and T cells (CD3+) per field and % of CD25+ T cells (CD3+) per field within the infarct core. Bars indicate mean ± SD corresponding to three animals per group. **p < 0.01 by Two-way ANOVA. Scale bar = 20 μm. Medicines and Medical Devices (AEMPS; PEI18-140) to produce cell-free cardiac grafts, based on the potential of cATMSC-EV combined clinical-grade scaffolds from healthy pericardium of cadaveric donors, with our tissue engineering approach. Rationally, EV have several ad­ and currently being used in a clinical trial [34]. vantages over cells thanks to their unchanged phenotype, lack of Following our established decellularisation methods, we obtain tumorigenicity, limited immunoreactivity, and better biodistribution acellular cardiac constructs that can be rehydrated with aqueous peptide and stability. In fact, the sustained delivery of distinct EV sources using a hydrogels, and repopulated with cATMSC [7,12,26] or presently, EV. In variety of injectable hydrogels or engineered patches is currently under our previous studies, the implantation of scaffolds with cATMSC scrutiny [41–43]. To this end, we purified porcine cATMSC-EV by SEC, a improved post-MI cardiac function in the preclinical swine model. stringent method that yields a pool of MSC-derived EV already proven to Microscopically, the engineered grafts integrated with the underlying be immunomodulatory when derived from human samples [18]. myocardium, and showed signs of neoinnervation and neo­ In the present work, we show that porcine cATMSC-EV inhibit vascularization [10–13]. However, despite exhibiting promising results, polyclonal activation (proliferation and pro-inflammatory cytokine it remains uncertain how the implanted cells contribute to MI recovery, production) of allogeneic immune cells, comparable to post-MI cir­ since, as many others also report, administered cells display poor sur­ cumstances. Following the ischemic injury, danger-associated signals vival and engraftment, and negligible proliferation and differentiation released by cardiomyocyte death and ECM degradation trigger an in­ within the myocardial tissue. A plausible hypothesis would be that flammatory reaction, including the production of inflammatory cyto­ instead of replacing the injured tissue themselves, cATMSC would kines (IL1β, IFNγ, TNFα and IL6) and chemokines that activate and reduce the ischemic injury-triggered inflammation by modulating the recruit the innate immune system. While this initial response is crucial host’s immune system and foster endogenous regeneration in a para­ for dead cell clearance and tissue repair, its timely repression would crine fashion. Specifically, they induce CD73 expression in M2-like prevent detrimental inflammation and amplification of the myocardial polarised monocytes and reduce the number of activated CD4+CD25+ damage [44,45]. T cells [11,35], along with recruiting peripheral progenitor cells via Here we report that cATMSC-EV abrogated the secretion of TNFα, a trophic factors [15,36–39]. Based on our previous findings and the cytokine extensively associated to cardiac damage and worsening of increasing evidence that EV play important roles in MSC’s paracrine heart failure symptoms [44]. On the other hand, IL6 and IL8 levels were effects [17,18,40], we decided to design and study a new generation of unaltered by porcine cATMSC-EV. While IL6 production has 3324 M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 controversial functions in MI, it has been also related to the initiation of movement and do not respond to the chemokines stimuli [24]. In line, reparative processes [46–49]. As to IL8, it recruits the neutrophils our experiments indicate that cells are indeed sensing chemoattractant needed after MI for clearance of dead cells and degraded ECM, and molecules present in EV, as only the OEC and MSC in contact with assisting the initiation of the inflammatory phase. However, neutrophil cATMSC-EV were attracted towards them. ablation does not correlate to a change in infarct size [48]. On the other side, the administration of cATMSC-EV greatly reduced Regarding IFNγ and IL12, available in vivo data report Th1-polarized the number of infiltrating monocytes and T cells in the MI core six days CD4+ T cells the predominantly present in the ischemic myocardium after MI induction, which would correlate with the anti-inflammatory [50]. Moreover, experiments in RAG1 KO mice reconstituted with CD4+ properties of cATMSC-EV observed in vitro. Within this relatively short T cells from either WT or IFNγ-KO mice beautifully uncovered that time span monocytes are known to be found in great numbers in the IFNγ-mediated Th1 CD4+ T cells contribute to MI injury [51,52]. Thus, it post-ischemic myocardium, and recent studies suggest that infiltrating is encouraging that cATMSC-EV suppress the production of IFNγ and the monocytes are responsible of exacerbating the inflammatory milieu Th1-driving IL12, translating in a decrease on T cell infiltration in vivo. post-MI that further deteriorates myocardial tissue [58–60]. Notably, Finally, the known interferon- and IL12-dependent IL10 production although the percentage of CD25+ T cells was unchanged, the overall might explain its concomitant reduction in our experiments [53,54]. number of T cells within the damaged MI tissue decreased. Thus, this Moreover, there could be other immunoregulatory mediators playing a marked change in infiltrating immune cells induced by MSC-EV could be role besides IL10 that have not been studied. In summary, the decrease indicative of the promotion of a shift from the inflammatory towards a in pro-inflammatory cytokines evidences the powerful immune modu­ regenerative phase, something that merits further investigation in latory effect of cATMSC-EV towards allogeneic polyclonal activation. longer time points. Of note, infiltration within the scaffold was low and The exploratory in vitro experiments incorporating scaffolds to did not change with cATMSC-EV, indicating the lack of immune stimulated PBMC proved difficult to interpret, as the mere presence of reactivity. the unloaded scaffolds already disrupted cytokine production, likely due Finally, with clinical translation in mind, it would be interesting to to minimized cell interaction. Unexpectedly, cATMSC-EV addition to study the bioequivalence of cATMSC-EV and EV derived from other either type of scaffold spiked IL6 production. Although we do not have a sources of human MSC-EV, such as umbilical cord, with better avail­ clear explanation for this phenomenon, and we cannot affirm whether ability, less ethical concerns and less epigenetic/environmental this also happens in vivo, it did not translate into an adverse reaction imprinting than adult sources of MSCs. after in vivo scaffold implantation. Individually, only cATMSC-EV com­ In summary, our data support the potential of multifunctional bined with myocardial scaffolds increased IFNγ, TNFα, and IL8, cATMSC-EV combined with biocompatible cardiac scaffolds generated although still in less amounts than stimulated PBMC alone. On the other by decellularisation of human pericardium for effective local delivery hand, and despite being of human origin, pericardial scaffolds did not over post-infarcted myocardial tissue. Moreover, SEC is a valuable induce such cytokine response, suggesting absence of xenoreactivity. method to obtain highly purified MSC-EV with functional properties Moreover, pericardial scaffolds seemed to retain multifunctional reciprocal to their parental MSC [18,61]. Noticeably, we find MSC-EV to cATMSC-EV in a larger extent in comparison with myocardial scaffolds. be anti-inflammatory, as well as able to recruit pro-regenerative cells Along this line, the bigger caveats of pericardial scaffolds already dis­ and promote angiogenesis in vivo. Hence, MSC-EV delivery into played better cell penetrance and retention. Other differences regarding post-infarcted myocardium presumably generates a highly bioactive EV or immune interaction could be explained by the different niche that triggers endogeneous cardiac repair. Nonetheless, further biochemical composition, variations in signalling ligands between the studies are mandatory to corroborate the effectiveness of this cell-free two scaffolds [12]. These findings support our downstream tissue engineering approach to benefit cardiac healing and ultimately proof-of-concept in vivo experiments combining cATMSC-EV with peri­ improve cardiac function in the long-term. cardial scaffolds. This EV delivery strategy aims at magnifying the local effective dose Funding sources of administered EV, thus downsizing the required EV units and side targets. In fact, fluorescence-aided tracking of EV showed the confine­ This work was supported in part by grants from Fundació la Marató ment and slow release of EV within scaffolds, even once implanted. de TV3 (201516-10, 201502-30), the PERIS (SLT002/16/00234) and Interestingly, fluorescent signal could be also found within the infarct SGR programmes (2017-SGR-301 REMAR Group, 2017-SGR-1427 and core, indicating that either cATMSC-EV were present there, or that cells 2017-SGR-483 ICREC Group) from the Generalitat de Catalunya, the that interacted with EV -and therefore acquired the dye-were later on Spanish Ministry of Economy and Competitiveness-MICINN (SAF2017- migrating towards the MI area. Also, the combination of EV with tissue 84324-C2-1-R; PID2019-110137RB-I00; PID2019-107145RB-I00), engineering, such as injectable ECM hydrogels, allows a prolonged Instituto de Salud Carlos III (PI17/01487, PIC18/00014, ICI19/00039, release of its embedded EV instead of rapid diffusion, and protect EV PI18/00256, PI18/01227, and ICI20/00135), ISCIII-REDinREN (RD16/ from degradation [42,43,55]. 0009 Feder Funds), Red de Terapia Celular-TerCel (RD16/0011/0006), cATMSC-EV showed immune suppressive properties and, for the first CIBER Cardiovascular (CB16/11/00403) projects as a part of the Plan time to our knowledge, we provide evidence of MSC-EV-mediated Nacional de I + D + I, and it was co-funded by ISCIII-Subdirección directed recruitment of allogeneic pro-regenerative cells, including General de Evaluación y el Fondo Europeo de Desarrollo Regional allogeneic MSC and peripheral blood-derived OEC in vitro. This (FEDER), AGAUR (2019PROD00122), Sociedad Española de Car­ concurred with angiogenesis promotion in vitro and translated to an diología, Societat Catalana de Cardiologia, PCMR[C] and EoI Collabo­ enhanced on-site neovasculogenesis in vivo, an essential first step to rative Projects on Regenerative Medicine 2019, and Institut Català de trigger endogenous cardiac repair. The power of EV-mediated recruit­ Salut (ICS). PG is sponsored by the PERIS (SLT002/16/00209) from the ment was shown by De la Fuente and colleagues, using tumour-derived Generalitat de Catalunya. AR is supported by the Miguel Servet program EV as bait to successfully attract and encapsulate circulating metastatic (CPII15/00003) and research grants (PI16/00981) from the Instituto de cells to an artificial 3D scaffold and enable their localised surgical Salud Carlos III, and co-financed by the European Regional Development removal [56]. In addition, EV can bear functional chemokines attached Fund. FEB is a researcher from Fundació Institut de Recerca en Ciències to surface [57], enabling recognition by cells. Ahmed and colleagues de la Salut Germans Trias i Pujol, supported by the Health Department of demonstrated that a concentration gradient is established across and the Catalan Government (Generalitat de Catalunya). within the agarose spot (in their case, purified chemokines), leading to cell recruitment towards the centre of the spot of cells in close contact with the spot, while cells in a remote position have a stochastic 3325 M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 CRediT authorship contribution statement PMA phorbol 12-myristate 13-acetate P/S penicillin-streptomycin Marta Monguió-Tortajada: designed the study, performed experi­ RT room temperature ments, Formal analysis, Data curation, Writing – original draft, Writing – SEC size exclusion chromatography review & editing. Cristina Prat-Vidal: performed experiments, Formal SEM scanning electron microscopy analysis, Data curation, Writing – original draft, Writing – review & 3D three-dimensional editing. Miriam Moron-Font: performed experiments, Writing – review TBS tris-buffered saline & editing. Marta Clos-Sansalvador: performed experiments, Writing – T-PBS Tween-PBS review & editing. Alexandra Calle: performed experiments, Writing – UC ultracentrifugation review & editing. Paloma Gastelurrutia: performed experiments, VEGF vascular endothelial growth factor Writing – review & editing. Adriana Cserkoova: performed experi­ ments, Writing – review & editing. Anna Morancho: performed ex­ Appendix A. Supplementary data periments, Writing – review & editing. Miguel Ángel Ramírez: conceptual advice, Writing – original draft, Writing – review & editing. Supplementary data to this article can be found online at https://doi. Anna Rosell: conceptual advice, Writing – original draft, Writing – re­ org/10.1016/j.bioactmat.2021.02.026. view & editing. Antoni Bayes-Genis: conceptual advice, Writing – original draft, Writing – review & editing. Carolina Gálvez-Montón: References designed the study, performed experiments, Formal analysis, Data curation, Writing – original draft, Writing – review & editing. Francesc [1] World Health Organization, Global Health Estimates 2016: Deaths by Cause, Age, Sex, by Country and by Region, 2000-2016, 2018. E. Borràs: designed the study, Formal analysis, Data curation, Writing – [2] G. Olivetti, J.M. Capasso, L.G. Meggs, et al., Cellular basis of chronic ventricular original draft, Writing – review & editing. Santiago Roura: designed the remodeling after myocardial infarction in rats, Circ. Res. 68 (1991) 856–869, study, performed experiments, Formal analysis, Data curation, Writing – https://doi.org/10.1161/01.RES.68.3.856. [3] M. Metra, P. Ponikowski, K. Dickstein, et al., Advanced chronic heart failure: a original draft, Writing – review & editing. position statement from the study group on advanced heart failure of the heart failure association of the European society of cardiology, Eur. J. Heart Fail. 9 Declaration of competing interest (2007) 684–694. [4] M.C. Alraies, P. Eckman, Adult heart transplant: indications and outcomes, J. Thorac. Dis. 6 (2014) 1120–1128, https://doi.org/10.3978/j.issn.2072- The authors declare that they have no known competing financial 1439.2014.06.44. interests or personal relationships that could have appeared to influence [5] M.W. Curtis, B. Russell, Cardiac tissue engineering, J. Cardiovasc. Nurs. 24 (2009) 87–92, https://doi.org/10.1097/01.JCN.0000343562.06614.49. the work reported in this paper. [6] A. Bayes-Genis, C. Soler-Botija, J. Farré, et al., Human progenitor cells derived from cardiac adipose tissue ameliorate myocardial infarction in rodents, J. Mol. Acknowledgements Cell. Cardiol. 49 (2010), https://doi.org/10.1016/j.yjmcc.2010.08.010. [7] C. Prat-Vidal, C. Gálvez-Montón, V. Puig-Sanvicens, et al., Online monitoring of myocardial bioprosthesis for cardiac repair, Int. J. Cardiol. 174 (2014) 654–661, We express our gratitude to the Flow Cytometry Unit and the Mi­ https://doi.org/10.1016/j.ijcard.2014.04.181. croscopy Unit of IGTP for their valuable technical help and the Center of [8] S. Roura, C. Soler-Botija, J.R. Bagó, et al., Postinfarction functional recovery driven Comparative Medicine and Bioimaging of Catalonia (CMCiB) for their by a three-dimensional engineered fibrin patch composed of human umbilical cord blood-derived mesenchymal stem cells, Stem Cells Transl. Med 4 (2015) 956–966, contribution in the animal model execution. We are grateful to Dr. María https://doi.org/10.5966/sctm.2014-0259. Yáñez-Mó (Dept Biología Molecular, UAM; Centro de Biología Molecular [9] L. Bagno, K.E. Hatzistergos, W. Balkan, J.M. Hare, Mesenchymal stem cell-based Severo Ochoa (CBM-SO), Hospital Universitario de la Princesa (IIS-IP)) therapy for cardiovascular disease: progress and challenges, Mol. Ther. 26 (2018) 1610–1623, https://doi.org/10.1016/J.YMTHE.2018.05.009. and Dr. Francisco Sánchez-Madrid (Servicio de Inmunología, IIS-IP, [10] I. Perea-Gil, C. Prat-Vidal, C. Gálvez-Montón, et al., A cell-enriched engineered UAM; Centro Nacional de Investigaciones Cardiovasculares (CNIC)) myocardial graft limits infarct size and improves cardiac function: pre-clinical for antibody donations, and to Dr. Hernando A. del Portillo (ICREA study in the porcine myocardial infarction model, JACC Basic Transl. Sci. 1 (2016) 360–372, https://doi.org/10.1016/j.jacbts.2016.06.005. Research Professor at ISGLOBAL-IGTP) for access to NTA instrument. [11] C. Gálvez-Montón, R. Bragós, C. Soler-Botija, et al., Noninvasive assessment of an engineered bioactive graft in myocardial infarction: impact on cardiac function and Abbreviations scar healing, Stem Cells Transl. Med. 6 (2017) 647–655, https://doi.org/10.5966/ sctm.2016-0063. [12] I. Perea-Gil, C. Gálvez-Montón, C. Prat-Vidal, et al., Head-to-head comparison of BSA bovine serum albumin two engineered cardiac grafts for myocardial repair : from scaffold characterization cATMSC cardiac adipose tissue-derived MSC to pre-clinical testing, Sci. Rep. 8 (2018) 1–13, https://doi.org/10.1038/s41598- 018-25115-2, 6708. CCM concentrated conditioned medium [13] C. Gálvez-Montón, M.T. Fernandez-Figueras, M. Martí, et al., Neoinnervation and cryo-EM cryo-electron microscopy neovascularization of acellular pericardial-derived scaffolds in myocardial infarcts, DAPI 4′ ,6-diamidino-2-phenylindole Stem Cell Res. Ther. 6 (2015) 108, https://doi.org/10.1186/s13287-015-0101-6. EBM-2 endothelial cell basal medium-2 [14] K.M. Broughton, B.J. Wang, F. Firouzi, et al., Mechanisms of cardiac repair and regeneration, Circ. Res. 122 (2018) 1151–1163, https://doi.org/10.1161/ ECM extracellular matrix CIRCRESAHA.117.312586. EGM-2 endothelial growth medium-2 [15] K. Malliaras, Y. Zhang, J. Seinfeld, et al., Cardiomyocyte proliferation and EV extracellular vesicles progenitor cell recruitment underlie therapeutic regeneration after myocardial infarction in the adult mouse heart, EMBO Mol. Med. 5 (2013) 191–209, https:// FBS foetal bovine serum doi.org/10.1002/emmm.201201737. hEGF human endothelial growth factor [16] A.I. Caplan, Mesenchymal stem cells: time to change the name!, Stem Cells Transl. hFGF-B human basic fibroblast growth factor Med. 6 (2017) 1445–1451, https://doi.org/10.1002/sctm.17-0051. [17] M. Yáñez-Mó, P.R.-M. Siljander, Z. Andreu, et al., Biological properties of IM intramuscular extracellular vesicles and their physiological functions, J. Extracell. Vesicles 4 Io ionomycin (2015) 27066, https://doi.org/10.3402/jev.v4.27066. IV intravenous [18] M. Monguió-Tortajada, S. Roura, C. Gálvez-Montón, et al., Nanosized UCMSC- derived extracellular vesicles but not conditioned medium exclusively inhibit the MFI median fluorescence intensity inflammatory response of stimulated T cells: implications for nanomedicine, MI myocardial infarction Theranostics 7 (2017) 270–284, https://doi.org/10.7150/thno.16154. MSC mesenchymal stromal cell [19] M. Monguió-Tortajada, M. Morón-Font, A. Gámez-Valero, et al., Extracellular- vesicle isolation from different biological fluids by size-exclusion chromatography, OEC outgrowth endothelial cell Curr. Protoc. Stem Cell Biol. e82 (2019) 1–24, https://doi.org/10.1002/cpsc.82. PBS phosphate-buffered saline PBMC peripheral blood mononuclear cells 3326 M. Monguió-Tortajada et al. Bioactive Materials 6 (2021) 3314–3327 [20] J. Van Deun, P. Mestdagh, P. Agostinis, et al., EV-TRACK: transparent reporting hemodynamics, J. Thorac Cardiovasc. Surg. (2019), https://doi.org/10.1016/j. and centralizing knowledge in extracellular vesicle research, Nat. Methods 14 jtcvs.2019.06.017. (2017) 228–232, https://doi.org/10.1038/nmeth.4185. [42] C. Han, J. Zhou, C. Liang, et al., Human umbilical cord mesenchymal stem cell [21] A. Gámez-Valero, M. Monguió-Tortajada, L. Carreras-Planella, et al., Size-Exclusion derived exosomes encapsulated in functional peptide hydrogels promote cardiac Chromatography-based isolation minimally alters Extracellular Vesicles’ repair, Biomater. Sci. 7 (2019) 2920–2933, https://doi.org/10.1039/ characteristics compared to precipitating agents, Sci. Rep. 6 (2016) 33641, https:// C9BM00101H. doi.org/10.1038/srep33641. [43] B. Liu, B.W. Lee, K. Nakanishi, et al., Cardiac recovery via extended cell-free [22] I. Lozano-Ramos, I. Bancu, A. Oliveira-Tercero, et al., Size-exclusion delivery of extracellular vesicles secreted by cardiomyocytes derived from induced chromatography-based enrichment of extracellular vesicles from urine samples, pluripotent stem cells, Nat. Biomed. Eng. 2 (2018) 293–303, https://doi.org/ J. Extracell. Vesicles 4 (2015) 27369, https://doi.org/10.3402/jev.v4.27369. 10.1038/s41551-018-0229-7. [23] M. Navarro-Sobrino, A. Rosell, M. Hernandez-Guillamon, et al., Mobilization, [44] S. Frantz, J. Bauersachs, R.A. Kelly, Innate immunity and the heart, Curr. endothelial differentiation and functional capacity of endothelial progenitor cells Pharmaceut. Des. 11 (2005) 1279–1290. after ischemic stroke, Microvasc. Res. 80 (2010) 317–323, https://doi.org/ [45] N.G. Frangogiannis, The inflammatory response in myocardial injury, repair and 10.1016/j.mvr.2010.05.008. remodelling, Nat. Rev. Cardiol. 11 (2014) 255–265, https://doi.org/10.1038/ [24] M. Ahmed, H.A. Basheer, J.M. Ayuso, et al., Agarose spot as a comparative method nrcardio.2014.28. for in situ analysis of simultaneous chemotactic responses to multiple chemokines, [46] B. Dawn, Y.T. Xuan, Y. Guo, et al., IL-6 plays an obligatory role in late Sci. Rep. 7 (2017) 1–11, https://doi.org/10.1038/s41598-017-00949-4. preconditioning via JAK-STAT signaling and upregulation of iNOS and COX-2, [25] A. Calle, C. Barrajón-Masa, E. Gómez-Fidalgo, et al., Iberian pig mesenchymal Cardiovasc. Res. 64 (2004) 61–71, https://doi.org/10.1016/j. stem/stromal cells from dermal skin, abdominal and subcutaneous adipose tissues, cardiores.2004.05.011. and peripheral blood: in vitro characterization and migratory properties in [47] N. Smart, M.H. Mojet, D.S. Latchman, et al., IL-6 induces PI 3-kinase and nitric inflammation, Stem Cell Res. Ther. 9 (2018) 178, https://doi.org/10.1186/ oxide-dependent protection and preserves mitochondrial function in s13287-018-0933-y. cardiomyocytes, Cardiovasc. Res. 69 (2006) 164–177, https://doi.org/10.1016/j. [26] I. Perea-Gil, J.J. Uriarte, C. Prat-Vidal, et al., In vitro comparative study of two cardiores.2005.08.017. decellularization protocols in search of an optimal myocardial scaffold for [48] N.G. Frangogiannis, The immune system and the remodeling infarcted heart: cell recellularization, Am. J. Transl. Res. 7 (2015) 558–573. biological insights and therapeutic opportunities, J. Cardiovasc. Pharmacol. 63 [27] NIH, Guide for the Care and Use of Laboratory Animals, eighth ed., National (2014) 185–195, https://doi.org/10.1097/FJC.0000000000000003. The. Academies Press, Washington, D.C, 2011. [49] A.E. Mayfield, P. Kanda, A. Nantsios, et al., Interleukin-6 mediates post-infarct [28] C. Gálvez-Montón, C. Prat-Vidal, S. Roura, et al., Transposition of a pericardial- repair by cardiac explant-derived stem cells, Theranostics 7 (2017) 4850–4861, derived vascular adipose flap for myocardial salvage after infarct, Cardiovasc. Res. https://doi.org/10.7150/thno.19435. 91 (2011) 659–667, https://doi.org/10.1093/cvr/cvr136. [50] U. Hofmann, S. Frantz, Role of lymphocytes in myocardial injury, healing, and [29] L. Kordelas, V. Rebmann, A.K. Ludwig, et al., MSC-derived exosomes: a novel tool remodeling after myocardial infarction, Circ. Res. 116 (2015) 354–367, https:// to treat therapy-refractory graft-versus-host disease, Leukemia 28 (2014) 970–973. doi.org/10.1161/CIRCRESAHA.116.304072. [30] M. Tkach, J. Kowal, A.E. Zucchetti, et al., Qualitative differences in T-cell [51] Z. Yang, Y.-J. Day, M.-C. Toufektsian, et al., Myocardial infarct-sparing effect of activation by dendritic cell-derived extracellular vesicle subtypes, EMBO J. 36 adenosine A2A receptor activation is due to its action on CD4+ T lymphocytes, (2017) 3012–3028, https://doi.org/10.15252/embj.201696003. Circulation 114 (2006) 2056–2064, https://doi.org/10.1161/ [31] C. Gálvez-Montón, C. Prat-Vidal, S. Roura, et al., Update: innovation in cardiology CIRCULATIONAHA.106.649244. (IV). Cardiac tissue engineering and the bioartificial heart, Rev. Esp. Cardiol. 66 [52] S.E. Boag, E. Andreano, I. Spyridopoulos, Lymphocyte communication in (2013) 391–399, https://doi.org/10.1016/j.rec.2012.11.012. myocardial ischemia/reperfusion injury, Antioxidants Redox Signal. 26 (2016) [32] P.M. Crapo, T.W. Gilbert, S.F. Badylak, An overview of tissue and whole organ 6940, https://doi.org/10.1089/ars.2016.6940, ars.2016. decellularization processes, Biomaterials 32 (2011) 3233–3243, https://doi.org/ [53] S. Rutz, W. Ouyang, Regulation of interleukin-10 and interleukin-22 expression in 10.1016/j.biomaterials.2011.01.057. T helper cells, Curr. Opin. Immunol. 23 (2011) 605–612, https://doi.org/10.1016/ [33] D.A. Taylor, L.C. Sampaio, Z. Ferdous, et al., Decellularized matrices in J.COI.2011.07.018. regenerative medicine, Acta Biomater. (2018), https://doi.org/10.1016/j. [54] C.A. Stewart, H. Metheny, N. Iida, et al., Interferon-dependent IL-10 production by actbio.2018.04.044. Tregs limits tumor Th17 inflammation, J. Clin. Invest. 123 (2013) 4859–4874, [34] C. Prat-Vidal, L. Rodríguez-Gómez, M. Aylagas, et al., First-in-human PeriCord https://doi.org/10.1172/JCI65180. cardiac bioimplant: scalability and GMP manufacturing of an allogeneic [55] M.J. Hernandez, K.L. Christman, Designing acellular injectable biomaterial engineered tissue graft, EBioMedicine 54 (2020) 102729, https://doi.org/ therapeutics for treating myocardial infarction and peripheral artery disease, JACC 10.1016/j.ebiom.2020.102729. Basic Transl. Sci. 2 (2017) 212–226, https://doi.org/10.1016/J. [35] M. Monguió-Tortajada, S. Roura, C. Gálvez-Montón, et al., Mesenchymal stem cells JACBTS.2016.11.008. induce expression of CD73 in human monocytes in vitro and in a swine model of [56] A. de la Fuente, L. Alonso-Alconada, C. Costa, et al., M-trap: exosome-based myocardial infarction in vivo, Front. Immunol. 8 (2017) 1577, https://doi.org/ capture of tumor cells as a new technology in peritoneal metastasis, J. Natl. Cancer 10.3389/fimmu.2017.01577. Inst. 107 (2015) djv184, https://doi.org/10.1093/jnci/djv184. [36] R.H. Lee, A.A. Pulin, M.J. Seo, et al., Intravenous hMSCs improve myocardial [57] J. Berenguer, T. Lagerweij, X.W. Zhao, et al., Glycosylated extracellular vesicles infarction in mice because cells embolized in lung are activated to secrete the anti- released by glioblastoma cells are decorated by CCL18 allowing for cellular uptake inflammatory protein TSG-6, Cell Stem Cell 5 (2009) 54–63, https://doi.org/ via chemokine receptor CCR8, J. Extracell. Vesicles 7 (2018) 1446660, https://doi. 10.1016/j.stem.2009.05.003. org/10.1080/20013078.2018.1446660. [37] E. Eggenhofer, V. Benseler, A. Kroemer, et al., Mesenchymal stem cells are short- [58] Y. Maekawa, T. Anzai, T. Yoshikawa, et al., Prognostic significance of peripheral lived and do not migrate beyond the lungs after intravenous infusion, Front. monocytosis after reperfused acute myocardial infarction:a possible role for left Immunol. 3 (2012) 297, https://doi.org/10.3389/fimmu.2012.00297. ventricular remodeling, J. Am. Coll. Cardiol. 39 (2002) 241–246, https://doi.org/ [38] K. Le Blanc, D. Mougiakakos, Multipotent mesenchymal stromal cells and the 10.1016/s0735-1097(01)01721-1. innate immune system, Nat. Rev. Immunol. 12 (2012) 383–396, https://doi.org/ [59] G. Zhao, S. Wang, Z. Wang, et al., CXCR6 deficiency ameliorated myocardial 10.1038/nri3209. ischemia/reperfusion injury by inhibiting infiltration of monocytes and IFN- [39] T. Ben-Mordechai, R. Holbova, N. Landa-Rouben, et al., Macrophage γ-dependent autophagy, Int. J. Cardiol. 168 (2013) 853–862, https://doi.org/ subpopulations are essential for infarct repair with and without stem cell therapy, 10.1016/j.ijcard.2012.10.022. J. Am. Coll. Cardiol. 62 (2013) 1890–1901, https://doi.org/10.1016/j. [60] A.M. van der Laan, A. Hirsch, L.F.H.J. Robbers, et al., A proinflammatory monocyte jacc.2013.07.057. response is associated with myocardial injury and impaired functional outcome in [40] L. Timmers, S.K. Lim, F. Arslan, et al., Reduction of myocardial infarct size by patients with ST-segment elevation myocardial infarction, Am. Heart J. 163 (2012) human mesenchymal stem cell conditioned medium, Stem Cell Res. 1 (2008) 57–65, https://doi.org/10.1016/j.ahj.2011.09.002, e2. 129–137, https://doi.org/10.1016/J.SCR.2008.02.002. [61] I. Perea-Gil, M. Monguió-Tortajada, C. Gálvez-Montón, et al., Preclinical evaluation [41] J.J. Chung, J. Han, L.L. Wang, et al., Delayed delivery of endothelial progenitor of the immunomodulatory properties of cardiac adipose tissue progenitor cells cell-derived extracellular vesicles via shear thinning gel improves postinfarct using umbilical cord blood mesenchymal stem cells: a direct comparative study, BioMed Res. Int. (2015) 439808, https://doi.org/10.1155/2015/439808, 2015. 3327