Cardiac tissue samples were obtained from normal and pathological adult human hearts. Samples of atrial appendages from normal hearts (n = 11, 6 males and 5 females, mean age 35 ± 12 years) were collected from heart waste fragments of donors who died for reasons other than cardiovascular diseases. Pathological samples were taken from atrial appendages of explanted hearts of patients with end-stage heart failure associated with ischemic cardiomyopathy and undergoing heart transplantation (n = 20, 14 males and 6 females, mean age 56 ± 5.5 years, mean ejection fraction 18.75 ± 3.6%) (Supplementary Table S1). All patients provided written informed consent for use of heart tissue for experimental studies and specimens were collected, without patient identifiers, following protocols approved by the Monaldi Hospital and in conformity with the principles outlined in the Declaration of Helsinki. Ethical approval for this study was gained from the University of Naples Federico II ethics committee (reference number 79/17).

Cardiac tissue samples were dissected, minced, and enzymatically disaggregated by incubation in 0.25% trypsin and 0.1% (w/v) collagenase II (both from Sigma-Aldrich, St. Louis, MO, United States), for 30 min at 37°C. The digestion was stopped by adding a double volume of Hank’s balanced salt solution (HBSS) supplemented with 10% fetal bovine serum (FBS) (Sigma-Aldrich). This preparation was further disaggregated by pipetting and tissue debris and cardiomyocytes were removed by sequential centrifugation at 100 g for 2 min, passage through 40 μm cell strainer and centrifugation at 400 g for 5 min. Cell population was seeded on culture dishes in Dulbecco’s Modified Eagle Medium/Nutrient Mixture F-12 (DMEM/F-12) supplemented with 10% FBS, 5% horse serum, 0.2 mM glutathione, 5U/L erythropoietin, 50 μg/ml porcine gelatine, 50 IU/ml Penicillin G-Streptomycin (all from Sigma-Aldrich) and 10 ng/ml basic fibroblast growth factor (bFGF) (Peprotech, Rocky Hill, NJ, United States) and cultured in a humidified incubator at 37°C and 5% CO₂ in air. The medium was changed every 3 days to remove cell debris and maintain a physiological pH. Once the adherent cells were more than 75% confluent, they were detached with 0.25% trypsin-EDTA (Sigma-Aldrich) and cell suspension was used to isolate CPCs by immunomagnetic cell sorting (Di Meglio et al., 2007; Castaldo et al., 2008; Di Meglio et al., 2010a; Di Meglio et al., 2010b; Castaldo et al., 2013; Nurzynska et al., 2013).

Both CPC from normal (CPC-N) and pathological hearts (CPC-P) were pooled to reduce the biological variability among patients (Metzger et al., 2020) and plated at a 2 × 10⁴ /cm² density and cultured with Nutrient Mixture F-12 medium, supplemented with 10% FBS, 5% horse serum, 0.2 mM glutathione, 5 U/L erythropoietin, 50 μg/ml porcine gelatin, 50 IU/ml penicillin G-streptomycin (all from Sigma-Aldrich) and 10 ng/ml bFGF (Peprotech). Culture dishes were checked daily at a phase contrast microscope (Olympus, Tokyo, Japan) and the medium was replaced every 3 days to remove cell debris and maintain a physiological pH.

Fibroblasts from the hearts of patients with end-stage heart failure associated with ischemic cardiomyopathy (CF-P) were isolated by outgrowth as previously described (Pagano et al., 2017; Sacco et al., 2019; Belviso et al., 2020a). Briefly, small fragments of myocardium were placed in culture plates under sterile coverglasses and incubated with Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% FBS and 0.5% penicillin and streptomycin (all from Sigma-Aldrich) in a humidified incubator at 37°C and 5% CO₂ in air. After a time ranging between 7- and 21-days cells outgrew from fragments. Coverglasses were removed and fibroblasts were pooled and passaged. Culture dishes were checked daily at a phase contrast microscope (Olympus) and the medium was replaced every 3 days to remove cell debris and maintain a physiological pH.

Once CPC-N and CPC-P were 80–90% confluent, complete medium was replaced with a serum-free medium, which was collected after 48 h and processed to isolate EVs. The isolation was carried out by ExoQuick-TCTM (System Biosciences, Mountain View, CA, United States), following the protocol supplied with the reagent without introducing any modification. Briefly, culture medium from each cell population was pooled and centrifuged at 3000 g for 15 min at 4°C to eliminate cells and cellular debris. The supernatant was transferred to sterile tubes adding 1 ml of ExoQuick-TCTM Exosome Isolation Reagent to each 5 ml of cell culture medium. Tubes were mildly agitated until the separation between the two phases was no longer visible, then incubated at 4°C overnight. The next day tubes with ExoQuick-TC/culture medium mixture were centrifuged at 1500 g for 30 min at 4°C to allow the precipitation of the white/beige pellet. Then the supernatant was discarded, and the residual ExoQuick-TC solution was spun down by centrifugation at 1500 g for 5 min at 4°C. The EVs were quantified using Bradford Assay (Gupta et al., 2021), then pooled and stored at −80°C until their use for specific assays (Belviso et al., 2016).

Pellet obtained from both EV-CPC-P and EV-CPC-N was lysed using RIPA buffer containing Tris-HCl 50 mM pH 7.4, NaCl 150 mM, Nonidet P-40 1%, sodium deoxycholate 0.25%, Na₃VO₄ 1 mM and NaF 1 mM (all from Sigma-Aldrich) supplemented with protease inhibitor cocktails complete ULTRA Tablets, Mini, EASYpack (Roche Diagnostics Corporation, Mannheim, Germany). Protein concentration was determined by Bio-Rad Protein Assay (Bio-Rad Laboratories, Hercules, CA, United States) using albumin from bovine serum (Sigma-Aldrich) as a standard. Protein array to simultaneously detect 41 targets between growth factors and growth factor receptors was performed on both EV-CPC-P and EV-CPC-N lysates, using Human Growth Factor Antibody Array C1 kit (Raybiotech, Norcross, GA, United States). Briefly, the array membranes binding primary antibodies were blocked for 30 min and then incubated for 2.5 h with 100 µg of protein for each sample. Both steps were performed at room temperature. After washing, membranes were incubated with a cocktail of biotin-conjugated antibodies overnight at 4°C, and then with horseradish peroxidase (HRP)-conjugated streptavidin for 2 h at room temperature. Membranes were then developed, and signal was detected by chemiluminescence and autoradiography. Numerical comparison of the signal densities of growth factors was performed strictly according to the guidelines for data extraction supplied with the array protocol. Briefly, spot signal densities from the scanned images of arrays were obtained using ImageJ densitometry software (https://imagej.nih.gov/ij/download.html). The background was then subtracted from the densitometry data, and the obtained values were normalized to the positive control signals. Data were expressed as the mean ± SEM (Supplementary Table S2).

Total RNA was extracted from EVs isolated from both CPC-N (n = 11) and CPC-P (n = 20), combining Trizol Reagent (Invitrogen, Thermo Fisher Scientific, Carlsbad, CA, United States) and RNeasy Micro kit (Qiagen, Hilden, Germany), as previously described (Di Meglio et al., 2010c). RNA obtained was resuspended in RNase-free water and the final concentration was determined by spectrophotometric analysis with Nanodrop 2000 (Thermo Scientific). RNA extracted was then retrotranscribed with QuantiTect Reverse Transcription kit (Qiagen) following the protocol provided by the supplier as previously described (Putame et al., 2020). In order to assess the presence and the quantity of specific transcripts, a Real-Time PCR was performed using PrecisionPLUS^TM MasterMix kit (Primerdesign, Southampton, United Kingdom) according to manufacturer’s protocol. The detection was carried out by measuring the binding of the fluorescent dye SYBR Green I to double-stranded DNA. DNA was then amplified using Mastercycler ep realplex^4S (Eppendorf, Hamburg, Germany). All samples were tested in triplicate and comparative quantification of target genes expression was based on cycle threshold (Ct) (Livak and Schmittgen, 2001). All the primers used, listed in Table 1, were designed with Primer3 software (http://frodo.wi.mit.edu) starting from the coding sequence of mature mRNA available on GeneBank. Melt curve analysis was used to assess amplification of non-specific products, uniformity of product and formation of primer dimers. Data were expressed as the mean ± SEM.

EV-CPC-N were used to prepare a conditioned medium (EV-CM), to test effects produced by EVs on both CPC-P and CF-P. EV-CPC-N were resuspended in the medium specific for each cell population and added to CPC-P and CF-P culture dishes at the final concentration of 0.1 mg/ml (Barile et al., 2014) and cultured for 48 h. During the culture, the morphology of cells was evaluated by phase contrast microscope observation with Nikon Eclipse Ti-E DS-Qi2 Microscope (Nikon Corporation, Tokyo, Japan).

Total RNA extracted from CPC-P and CF-P treated or not with EV-CPC-N was retrotranscribed and Real-Time PCR was performed as previously described. All the primers used are listed in Table 2.

To evaluate the effects of EV-CPC-N, CPC-P and CF-P cultured with EV-CM or with regular medium (control groups) were fixed in 4% paraformaldehyde for 20 min at room temperature after 3 days of culture. After blocking with 10% donkey serum, cells were incubated with primary antibody against human α-sarcomeric actin (Di Meglio et al., 2010b) or fibronectin (both from Sigma-Aldrich), for 1 h at 37°C, then washed and incubated with secondary antibody conjugated with fluorescein (Jackson ImmunoResearch Europe, Newmarket, United Kingdom) at 37°C, for 1 h. Nuclei were counterstained with 4′-6-diamidino-2-phenylindole (DAPI) (Merck Millipore) and stained area of culture dishes was mounted in Vectashield (Vector Labs). Microscopic analysis was performed with Nikon Eclipse Ti-E DS-Qi2 Microscope (Nikon) and marker expression was quantified by three independent observers by Nikon Imaging Analytical Software (NIS Elements Analysis 4.50) (Nikon) and expressed as mean fluorescence intensity ± SEM.

To evaluate proliferation and apoptosis, both CPC-P receiving EV-CM and control group were plated at a density of 2.5 × 10⁴ /cm² and stained with Cell-Clock Cell Cycle Assay and Cell-ApoPercentage Apoptosis Assay (both from Biocolor Ltd., Carrickfergus, United Kingdom), respectively, strictly following manufacturer’s instructions. In the first assay, cells were incubated at 37°C for 1 h with the redox dye supplied by the kit. Such dye is taken up by live cells and its uptake induces a distinct change in cell color, from light green to blue, specifically associated with G1- S- G2- or M-phase of the cell cycle. For ApoPercentage Apoptosis Assay, CPC-P receiving EV-CM or from control group, were first incubated with 3% hydrogen peroxide in the complete medium for 12 h, then stained with the ApoPercentage dye that is selectively imported by cells undergoing apoptosis. Microscopic analysis and quantification were performed by three independent observers with a Leica DM2000 LED microscope (Leica Microsystems) equipped with a digital camera Leica ICC50 HD (Leica Microsystems). Data were expressed as mean percentage of cycling cells over total cells ± SEM, for Cell-Clock Cell Cycle Assay, and as mean percentage of stained cells ± SEM over total cells for Cell-ApoPercentage Apoptosis Assay.

To evaluate the speed of migration of CPC-P cultured with EV-CM and of CPC-P of control group, scratch wound assay was performed as previously described (Castaldo et al., 2013). Briefly, cells were grown to confluence and a thin scratch was produced in straight line on culture plates with a 10 µl sterile pipette tip, leaving a cell-free zone. Cell culture dishes were first washed with medium to remove debris and then fresh medium was pipetted in the dishes. Plates were placed under Nikon Eclipse Ti-E DS-Qi2 Microscope (Nikon) equipped with stage incubator (Okolab, Pozzuoli, Italy) and the migration was documented acquiring pictures every 10 min for 8 h by digital camera (Nikon). Data were analyzed by NIS Elements software (Nikon) and expressed as mean speed of migration ± SEM.

All experiments were performed at least in triplicate, each sample was tested three to five times, and all numerical data were expressed as mean ± SEM. Statistical differences between groups were evaluated with Student’s two-tailed unpaired t-test. A value of p ≤ 0.05 was considered statistically significant.

Protein array performed to analyze growth factor cargo of EV-CPC revealed differences in the content of growth factors (GFs) between EV-CPC-N and EV-CPC-P. The comparison clearly showed that EV-CPC-N carried factors mostly involved in biological processes like cell migration, proliferation, and differentiation as EGF (Epidermal Growth Factor), FGF-6 and FGF-7 (Fibroblasts Growth Factors-6 and 7). Further, they showed a considerably higher amount of IGF-1 (Insulin-like Growth Factor) and SCF (Stem Cell Factor). On the contrary, EV-CPC-N did not carry factors involved in myofibroblast activation and angiogenesis processes, as TGF-β (Transforming Growth Factor), HGF (Hepatocyte Growth Factor) and VEGF (Vascular Endothelial Growth Factor) that were clearly present in EV-CPC-P. Additionally, a higher expression of receptors VEGFR2, VEGFR3, and SCFR in EV-CPC-P emerged. The assay did not reveal, in EV-CPC-N, the presence of receptors like SCFR, EGFR, PDGFRα, and PDGFRβ (Figure 2). Using the array map as a reference (Figure 2A), the representative images of the protein array membranes clearly showed that EV-CPC-N (Figure 2B) carried higher amounts of EGF, FGF, IGF-1, and SCF (Figure 2D), while EV-CPC-P (Figure 2C) were enriched with HGF, SCF-R, TGF, VEGF and its receptors (Figure 2D).

Real-time PCR analysis revealed also that EV-CPC-N and EV-CPC-P transported specific transcripts for cardiac cell progenitors or precursors of cardiomyocyte, as ACTC1, MEF2C, and NKX2.5, and smooth muscle cells (SMCs), as ACTA2 and GATA6. They also carried THY1, VCAM1 and SOX9 (Figure 3), markers of mesenchymal cells. Gene expression analysis revealed differences between EV-CPC-P and EV-CPC-N cargo. Precisely, statistically significant differences (p < 0.05) emerged for MEF2C and NKX2.5 transcripts, mainly carried by EV-CPC-N, and for ACTA2 contained in a higher amount in EV-CPC-P (Figure 3).

Real-time PCR analysis showed an up-regulation of ACTC1 gene, encoding for α-sarcomeric actin, in CPC-P treated with EV-CM with the respect to the control group (Figure 4A).

Observation by phase contrast microscope showed a similar morphology for CPC-P cultured with a regular medium (control group) or with EV-CM (Figures 4C,D). Immunofluorescence supported the Real-time PCR results, revealing that both populations expressed α-sarcomeric actin, even though CPC-P treated with EV-CM showed a significantly (p < 0.05) stronger immunopositivity (3442.15 ± 164.2 mean fluorescence intensity in CPC in EV-CM versus 2899.41 ± 96.73 mean fluorescence intensity in control CPC-P) (Figure 4B) and a better organized α-actin pattern and distribution (Figures 4E,F).

The presence of EV-CPC-N had also effects on proliferation and apoptosis of cells. CPC-P treated with EV-CM showed higher proliferation rate when compared to control group, as indicated by the ratio of cells in S to M phases (Figures 5A,B). Indeed, in the presence of EV-CM the proportion of proliferating cells in M phase reached a percentage of 20.52 ± 4.43%, while the control group showed a significantly lower number of proliferating cells in the same phase, equal to 1.068 ± 0.79%. The percentage of cells in G2-S phase was 14.19 ± 0.63% in CPC-P treated with EV-CM, against a 27.48 ± 2.13% observed in CPC-P in culture with regular medium. No statistically significant differences were observed for cells in G1 phase that showed a percentage of 65.28 ± 5.04 and 71.45 ± 2.42% for CPC-P treated with EV-CM and control group, respectively (Figure 5C). From microscopic analysis a protective effect on oxidative stress-induced apoptosis exerted by EV-CM culture of CPCs also emerged, as demonstrated by the significantly lower percentage of nuclei stained with the dye selective for apoptotic cells (Figures 5D,E). In fact, oxidative stress induced apoptosis in only 2.25 ± 0.31% in case of CPC-P that received EV-CM in vitro, while CPC-P cultured in regular medium resulted more susceptible to apoptosis showing a dramatically higher percentage of apoptotic nuclei (26.74 ± 8.08%) (Figure 5F).

After the scratch wound assay, cell migration was monitored in real time by time-lapse imaging which also yielded valuable cell morphology and localization. Images from scratched areas of cell monolayers were recorded every 10 min for 8 h. Results obtained indicated that migration was not affected by administration of EV-CM, as the speed of migration did not differ significantly between CPC-P treated with EVs (11.28 ± 1.34 μm/h) and control group (11.43 ± 1.31 μm/h), and complete wound healing occurred after 8 h in both CPC populations (Figure 6).

Real-time PCR analysis revealed that CF-P cultured with EV-CM showed a significant up-regulation of genes encoding for specific ECM proteins, like collagen type IV and fibronectin, while no statistically significant differences emerged for laminin, tenascin, collagen type I and type III with respect to the control group (Figure 7). At the microscopic observation, CF-P cultured with EV-CM exhibited less elongated morphology with respect to the control group (Figures 8A,B). Fibronectin expression was clearly extracellular, and its synthesis and deposition increased after EV-CM administration to CF-P (1118.96 ± 118.5 versus 2349.31 ± 261.0 mean fluorescence intensity in control CF-P and in CF-P cultured with EV-CM, respectively; p = 0.005). Furthermore, fibronectin was secreted and assembled into distinctive fibrillary form that resulted homogeneously distributed in CF-P cultured in EV-CM with respect to the control group where the protein was fragmented and disarranged (Figures 8C,D).

The findings of this study have to be seen in light of a relevant limitation.

Indeed, in vivo studies, aimed at investigating the feasibility of cardiac regenerative therapies based on the use of EV-CPC and to explore the potential of EV-CPC as cost-effective off-the-shelf alternative therapeutic modality, are needed. Although based on in vitro experiments, evidence emerging from our study offers an intriguing perspective on the underlying mechanisms responsible for effects of CPC-based therapy of myocardial infarction (Ostovaneh et al., 2021).