Background
The incidence of cardiovascular diseases has increased dramatically in the last decades and remains the predominant cause of mortality worldwide [
1,
2]. Although myocardial dysfunction during heart diseases is often associated with impaired cardiomyocyte activity, cardiac fibrosis is the major cause of end-stage heart damage [
3,
4]. Myocardial fibrosis is a complex process that results from abnormal healing of the heart tissue and ultimately leads to the formation of a non-functional scar, which hampers the activity of the entire organ. This involves both extracellular matrix overproduction and the activation of structural non-excitable fibroblasts to differentiate into contractile myofibroblasts, in a process called fibroblast-to-myofibroblast transition (FMT). Myofibroblasts are characterized by an increased expression of α-smooth muscle actin (α-SMA, encoded by the
ACTA2 gene) and the secretion of a number of pro-fibrotic proteins, such as collagens, fibronectin, or tenascin [
5‐
7]. Transforming growth factor β (TGFβ) is the best-known fibrogenic cytokine described in fibrotic diseases, including heart fibrosis [
6,
8,
9]. By enhanced activation of multiple signaling pathways, particularly those involving SMAD2/3 proteins, TGFβ is able to effectively induce the FMT machinery and propagate the profibrotic signaling cascade [
3,
6].
Currently used therapeutic strategies targeting cardiac fibrosis such as β-blockers or cell-based therapies neither prevent the progression of cardiac fibrosis nor promote the functional recovery of the heart [
10]. Therefore, innovative treatment strategies are an important unmet clinical need.
One of the novel classes of therapeutics which has gained considerable interest in recent years constitutes extracellular vesicles (EVs). EVs are nanometric circular structures secreted by virtually all types of cells under physiological and pathological conditions. They contain bioactive components derived from the parental cell, enclosed in a lipid bilayer that protects them against rapid degradation [
11]. Based on their size and origin, EVs can be classified as exosomes, microvesicles (shedding vesicles), and apoptotic bodies. EVs are considered to be important mediators of intercellular communication and, due to their ability to transport bioactive molecules such as proteins, lipids, and various RNA molecules, they can influence the phenotype and properties of other cells [
12]. In particular, mesenchymal stem/stromal cells (MSCs), cardiac progenitor cells (CPCs), cardiospheres, endothelial cells and pluripotent stem cells were shown to produce EVs with reparative capabilities, including anti-fibrotic activity [
10,
13]. Owing to their unmatched functionality, biocompatibility, and efficiency in delivering components to target cells, EVs are regarded as new generation therapeutics in the treatment of a variety of human diseases [
14].
We have previously shown that EVs derived from induced pluripotent stem cells (iPSCs) carry pro-regenerative potential, with respect to heart regeneration [
15,
16]. Not only were they effective in transferring their cargo to primary cardiac stromal cells influencing their fate, but they also improved heart recovery post-myocardial infarction. Recent reports indicate the usefulness of hiPS-EVs in rejuvenation [
17], immunomodulation [
18], and cytoprotection [
18‐
21]. Importantly, hiPS-EVs were also shown to exert anti-fibrotic function in the treatment of liver and pulmonary fibrosis [
22,
23]. These data indicate that the spectrum of potential medical indications that can benefit from iPS-EV-based therapy is increasing. However, none of the available reports show the possibility of significant enhancement of the anti-fibrotic function of hiPS-EVs by modulating the cell culture environment of iPSCs, which we present in this study.
It is known that stem cells reside in hypoxic niches and that oxygen plays an important role in the regulation of cellular metabolism and function [
24,
25]. In addition, reduced oxygen levels have been demonstrated to affect the cargo and activity of EVs derived from both normal and cancer cells [
26,
27]. In particular, hypoxia was shown to increase the pro-angiogenic activity of EVs derived from various types of cells, including hiPSCs [
28], embryonic stem cell-derived CPCs [
29], MSCs [
30,
31], and microglia [
32], among others. However, the impact of low oxygen conditions on the anti-fibrotic function of hiPS-EVs and the underlying molecular mechanism has not been investigated so far.
Based on this knowledge, we selected oxygen conditions resembling physiological levels within the pluripotent stem cell niche, which were 5% O2 and 3% O2, along with 21% O2, as a standard cell culture setting. Conditioned media harvested from hiPSCs were processed to isolate therapeutic EVs, with the ultimate goal of ameliorating heart fibrosis. Importantly, we also elucidated the mechanisms mediated by hiPS-EVs, pointing to their miRNA cargo. Finally, we examined the anti-fibrotic function of hiPS-EVs in an in vivo model of angiotensin II-induced heart fibrosis.
Materials and methods
Cell culture
In this study, three iPSC lines were used: L1 — a previously published cell line [
15]; L2 — episomal hiPSC line purchased from Gibco, #A18945; and L3 — cell line generated in the laboratory of Prof. Toni Cathomen – Medical Center – University of Freiburg, Freiburg, Germany. iPSCs were cultured in Essential8 medium (Gibco/Thermo Fisher Scientific), supplemented with antibiotics (Penicillin/Streptomycin; P/S; Gibco) on cell culture plates coated with human recombinant vitronectin (Gibco), in an atmosphere containing various values of oxygen concentration: 21% (normoxia; N) — in a standard cell incubator, or in the presence of a reduced oxygen concentration (hypoxia; H) — 5% O
2 (H5) or 3% O
2 (H3). Hypoxic conditions were obtained in an InvivO2 chamber (Ruskinn). Cells were passaged every 4 days using 0.5 mM EDTA solution (Gibco) and plated into new culture vessels coated with vitronectin in a ratio of 1: 6–8, with the addition of 10 µM ROCK kinase inhibitor (Y-27632; Millipore) for the first 24 h (h). Cultures were carried out in an atmosphere of 5% CO
2, 70–90% humidity, with medium change every 24 h. The cells were adapted to each oxygen condition for 4 passages before starting the experiments.
Human dermal fibroblasts (hDF; ATCC, # PCS-201–012) were cultured in Dulbecco’s Modified Eagle’s Medium/High Glucose (DMEM/HG; Sigma-Aldrich/Merck), supplemented with 10% fetal bovine serum (FBS; Gibco) and 1% P/S in standard cell culture conditions (37 °C, 95% humidity, 5% CO2).
Human cardiac fibroblasts (hCF, Cell Applications/Merck, #306-05A) were cultured in Cardiac Fibroblast Growth Medium (Cell Applications, #316–500). Cells were grown under standard conditions. Fibrosis experiments were performed in Advanced DMEM/F12 medium (A/DMEM/F12; Gibco) containing 2% FBS, after adapting the cells to modified conditions.
Isolation of EVs
Conditioned medium (CM) was harvested from iPSCs grown to a density of 70–90%, 24 h after the last medium change. EVs were isolated using the ultracentrifugation (UC) or ultrafiltration (UF) method combined with size-exclusion chromatography (SEC). Media were collected and frozen at -80°C until isolation. CM was sequentially centrifuged at 4°C at increasing speeds: 500 g, 8 min (to remove dead cells and cell debris) and 2000 g, 15 min (to remove apoptotic bodies). Then, EVs were obtained using either the UC or UF + SEC methods. Based on the UC method, EVs were collected at 100,000 g, 4°C for an hour, using the Sorvall WX 90 + Ultracentrifuge equipped with the T-865 fixed angle rotor (Thermo Fisher Scientific). The EVs pellet was washed with PBS and centrifuged again under the same conditions. The obtained EVs were suspended in PBS and frozen in aliquots at – 80°C. When EVs were obtained with the UF + SEC method, CM was concentrated in 15 ml filter tubes (Amicon/Merck) with a protein cut-off of 10, 50, and 100 kDa by centrifugation at 1800 g for 40–60 min at 4°C. Next, the supernatants were transferred to low-adhesive 1.5-ml tubes and kept on ice. Concentrated preparations were purified from proteins, lipids, and other small molecules using the EVs isolation column (qEV/70 nm, Izon), according to the manufacturer’s instructions. Briefly, the column was washed with PBS filtered through filters with a pore size of 0.22 µm. In the optimization stage, 0.5 ml fractions were collected from the column to 1.5 ml tubes by gravity flow, including the void volume (the first 3 ml), and were analyzed for the presence of EV particles and proteins. Subsequently, fractions 0.5–1.5 ml (after the void volume) were verified to contain EVs and were collected in a 1.5-ml low-adhesive tube in the rest of the experiments. EV preps were further concentrated using 4 ml concentrating tubes (Amicon) with a protein cut-off of 10 kDa. The resulting EV solution was aliquoted and stored at − 80 °C until use. Five independent isolations were performed to compare the results from the UF + SEC and UC methods. A sample of Essential8 medium was prepared the same way as EVs with the UF-10 kDa + SEC method, using 25 ml of medium.
To harvest EVs from hDF, cells were grown to 70–90% confluency. Next, cells were washed 2 × with PBS and a serum-free medium composed of DMEM/HG supplemented with 0.1% bovine serum albumin (BSA) was added for 48 h. CM was collected and processed with the UF-10 kDa + SEC method as described above.
Measurement of EVs size and concentration
The size and concentration of EV particles were analyzed using the NanoSight apparatus (Malvern). The samples were diluted 1:1000 in PBS, or 1:100 for the analysis of EV fractions collected from the qEV/70 nm (Izon) column, to obtain the optimal particle density for measurement. Data were collected at camera level 11 and detection threshold 5. A 60-s video was recorded for each sample. Two EV batches were analyzed of each type of EVs (n = 6 for each oxygen condition, or n = 3 for EV fractions).
The amount of protein in the EV preparations was measured based on the reaction with bicinchoninic acid (BCA) (Invitrogen/Thermo Fisher Scientific), according to the manufacturer’s instructions.
Transmission electron microscopy
Negative staining for EV was obtained by adsorbing 20 µl of EV suspended in PBS onto nickel-coated grids (Agar Scientific, Stansted, UK) for 30 min, followed by fixation for 5 min in 2.5% glutaraldehyde solution. After removing excess liquid with filter paper, the samples were stained with 2% uranyl acetate for 30 min, washed three times with distilled water for 1 min, and dried. Then, EVs were imaged with a JEOL JEM 2100HT (Jeol Ltd, Tokyo, Japan) transmission electron microscope (TEM) that was used at an accelerating voltage of 80 kV. Images were taken by using 4 k × 4 k camera (TVIPS) equipped with EMMENU software ver. 4.0.9.87.
Cell metabolic activity was measured with the Cell Counting Kit-8 (Sigma-Aldrich), upon stimulation with EVs. Briefly, hCFs were seeded on 96-well plates (2 × 10e3 cells per well) and were treated with EVs (2.5 × 10e4 EV particles/cell) for 24 h, after which the medium was replaced with a fresh medium. At indicated time points the CCK-8 reagent was added to cells for 2 h and the absorbance resulting from the activity of cellular dehydrogenases was measured in an Infinite M200 Microplate Reader (Tecan).
In vitro model of fibrosis
Cells were seeded at a density of 5 × 10e4 cells/cm2 (for RNA and protein analyzes) or 8 × 10e3 cells/cm2 (for immunofluorescence) 24 h before starting the experiments in a complete cell culture medium. Then, the medium was replaced with A/DMEM/F12 containing 2% FBS, P/S, and TGFβ1 (TGFβ; Corning) at a concentration of 1 ng/ml for 6 h to induce fibrosis. Next, EVs from one cell line and one oxygen condition (hiPSC-L3, H5) were added to the cells at increasing concentrations: 1.25, 2.5, 5 × 10e4 particles/cell to optimize the dose. A concentration of 2.5 × 10e4/cell was used in the rest of the experiments. EVs from all three hiPSC lines and various oxygen concentrations (EV-N, EV-H5, EV-H3) were tested. hDF-EVs at the same dose (2.5 × 10e4/cell) were used for comparison. hCFs were exposed to EVs for 24 h after which the medium was removed. Fresh medium supplemented with TGFβ (1 ng/ml) was changed daily for the next 4 days, when the experiments were terminated.
Immunofluorescence
The presence of pluripotency markers in hiPSCs and selected fibrotic proteins in hCFs were visualized by staining with specific antibodies. Cells were grown in glass bottom 24-well plates (Ibidi, # 82,406). After completion of the experimental procedure, cells were washed with PBS and then fixed with 3.7% formaldehyde solution (v/v) for 20 min at room temperature (RT). Next, cells were washed 3 × with PBS and permeabilized with 0.1% Triton X-100 solution in PBS (v/v) for 8 min. Again, cells were washed with PBS and incubated with 1% BSA solution in PBS (w/v) for 45 min at RT. Incubation with a primary antibody solution in 1% BSA was carried out overnight at 4°C. After washing thoroughly, cells were incubated with secondary antibodies for 45 min at RT. The list of antibodies is provided in Table
1. Cell nuclei were stained with DAPI (4′,6-diamidino-2-phenylindole, dihydrochloride) or Hoechst 33258 solution (1 µg/ml; Sigma-Aldrich), in case of hiPSCs or hCFs, respectively. hCFs were additionally stained with AF546 conjugated phalloidin (1:40; Invitrogen, #A22283), where indicated. Cells were washed with distilled water and analyzed using a Leica6000B fluorescence microscope. The image was adjusted each time to a given type of preparation (exposure time, gain, and binning). For analysis of hCFs, 16 images were taken for each experimental condition using the tile scan method. The obtained images were used to determine the efficiency of phenotypic transitions of hCFs into myofibroblasts. Based on the presence of α-SMA-rich stress fibers in hCFs, the efficiency of the FMT process was determined (% of α-SMA-positive cells in the analyzed population). Fluorimetrical analysis was performed using the ImageJ freeware (NIH, Bethesda, MD, USA). The results are presented as the mean fluorescence intensity of the test protein, relative to DNA fluorescence intensity in all 16 images as described previously [
33]. For the analysis of activation of selected transcription factors in the tested cells, the level of fluorescence of the phosphorylated form of SMAD2 (pSMAD2) or SNAI2 measured in the area of the cell nucleus was determined [
34]. The length and area of the focal contact sites (FCs) from collected images were measured in ImageJ [
35]. The obtained fluorimetry values are presented as relative fluorescence units (RFU).
Table 1
List of antibodies used in the study
Primary antibodies |
Anti-OCT4 | Monoclonal mouse, IgG, Invitrogen #MA1-104 | IF (1:200), WB (1:1000) |
Anti-SSEA4 | Mouse, IgG3, Invitrogen #A24866 | IF (1:100) |
Anti-CD133 | Monoclonal mouse, IgG1, Thermo Fisher Scientific #MA1-219 | IF (1:200), WB (1:1000) |
Anti-α-SMA | Monoclonal mouse, Sigma-Aldrich #A2547 | IF (1:400), WB (1:1000) |
Anti-Vinculin | Monoclonal mouse, Sigma-Aldrich #V9264 | IF (1:200), WB (1:1000) |
Anti-pSmad2 | Polyclonal rabbit, Cell Signaling Technology #3108 | IF (1:100) |
Anti-Slug2 | Polyclonal rabbit, Sigma-Aldrich # PRS3959 | IF (1:100) |
Anti-CD9 | Monoclonal mouse, IgG1, Invitrogen #10626D | WB (1:500) |
Anti-CD81 | Monoclonal mouse, IgG1, kappa, Invitrogen #MA5-13548 | WB (1:200) |
Anti-Flotillin1 | Polyclonal goat, IgG, Invitrogen #PA5-1853 | WB (1:1000) |
Anti-Transferrin | Polyclonal rabbit, IgG, Invitrogen #PA3-913 | WB (1:1000) |
Anti-E-Cadherin | Monoclonal rat, IgG1, kappa, eBioscience #14–3249-82 | WB (1:500) |
Anti-Syntenin | Polyclonal goat, IgG, Invitrogen #PA5-18595 | WB (1:250) |
Anti-Calnexin | Polyclonal goat, IgG, Invitrogen #PA5-19169 | WB (1:1500) |
Anti-COL1A1 | Polyclonal rabbit, IgG, ABclonal #A1352 | WB (1:1000) |
Anti-COL3A1 | Monoclonal rabbit, IgG, ABclonal #A0817 | WB (1:1000) |
Anti-β-tubulin | Monoclonal mouse, IgG2a, Invitrogen #MA5-16308 | WB (1:2000) |
Secondary antibodies |
Anti-mouse, AF 647 | Goat IgG, Invitrogen #A-21235 | IF (1:1000) |
Anti-mouse, AF 488 | Goat IgG3, Invitrogen #A24877 | IF (1:1000) |
Anti-mouse, AF 488 | Goat IgG, Invitrogen #A11001 | IF (1:1000) |
Anti-mouse, HRP | Goat IgG, Invitrogen #31430 | WB (1:2000) |
Anti-rabbit, HRP | Goat IgG, Invitrogen #31460 | WB (1:2000–6000) |
Anti-goat, HRP | Rabbit IgG, Invitrogen #R-21459 | WB (1:2000) |
Anti-rat, HRP | Donkey IgG, Invitrogen #A18739 | WB (1:2000) |
Confocal microscopy
EV uptake by hCFs as well as localization of pSMAD2 in the cells was made using scanning laser confocal microscopy (Zeiss LSM 900 with Airyscan 2) as described before [
36]. To analyze EV uptake by hCFs the cells were stained with a DiO lipid probe (3,3′-dioctadecyloxacarbocyanine, perchlorate; Invitrogen). EVs were stained with the Vybrant DiD cell labeling solution (1,1'–dioctadecyl-3,3,3′,3′-tetramethylindodicarbocyanine, 4-chlorobenzenesulfonate salt; Invitrogen). Briefly, EV-H5 were incubated with 5 µM dye in 1 ml PBS for 30 min at 37°C. Next, EVs were purified by SEC using qEV/70 nm columns and concentrated in 2-ml filter tubes with 10 kDa protein cutoff (Amicon). Purified EVs were added to hCFs for 30 min and imaged. 10 cells were analyzed.
To identify the number of pSMAD2-positive cells, confocal tile images were obtained, composed of 5 × 5 scans. 5 randomly selected areas on the sample for each experimental group were analyzed. Next, to determine the localization of pSMAD2 in the cells, high-resolution Airyscan analysis was performed. First, a 2D image of an entire cell was obtained followed by generating a ‘z-stack’ of the nuclear region. a Total of 10 cells for each experimental group were analyzed.
Cell stiffness measurement by atomic force microscopy (AFM)
AFM analysis was conducted using a Bioscope Catalyst (Bruker) coupled with an inverted optical microscope AxioObserver Z1 (Zeiss). During the analysis, cells were maintained in culture medium at 37°C. Morphology images as well as elasticity maps of the cells were obtained using PeakForce Tapping mode with PeakForce Capture turned ON. This enabled the acquisition of a force curve in every pixel of an image. For AFM imaging a relatively soft cantilever was used with a nominal tip radius of 20 nm and with an experimentally determined spring constant of 0.68 N/m (Bruker Probes). Nanomechanical analysis of cells was made in force spectroscopy mode, which consisted of measuring force–displacement curves. Prior to cell measurement, the AFM probe was positioned on top of the cell and aligned in the center of the cell by optical microscopy. Once aligned, force–displacement curves from a grid of 5 × 5 points were collected at a rate of 1 Hz. 30 cells for each experimental group were analyzed. A detailed description of the mechanical analysis used in this work can be found elsewhere [
37]. In the analysis, a soft cantilever was used with a nominal tip radius of 20 nm and with an experimentally determined spring constant of 0.014 N/m (Bruker Probes). Data analysis of the obtained force–displacement curves from both PeakForce Tapping and force spectroscopy was performed using AtomicJ software [
38].
In vivo studies of heart fibrosis
Animal experiments were approved by the I Local Ethical Committee in Krakow (agreement no. 554/2021). Mice (8-week-old male NOD/SCID strain) were housed in an institutional animal unit (Department of Clinical Immunology, Institute of Pediatrics, Jagiellonian University Medical College, Wielicka 265, 30–663 Krakow, Poland) under conditions: 12 h light cycle, standard rodent chow diet, free access to water, in a sterile environment. Fibrosis was induced by angiotensin II (Ang II; Sigma-Aldrich), which was administered subcutaneously in osmotic pumps (Alzet, Cupertino model 1004) at a dose of 1.4 mg/kg/day, according to a previously published protocol [
39]. 14 days after induction of fibrosis, the animals received 4 doses of EVs by retro-orbital injection at intervals of 3–4 days (4 × 10e10 EV particles each time, as measured by NanoSight). Injections were performed on alternate eyes (2 injections per eye) with no sign of ocular injury. EVs from normoxia and hypoxia of 5% O
2 were applied. To adhere to the 3R principle (Replacement, Reduction, and Refinement), 6 animals per group were used. One animal was excluded from the analysis due to the removal of the osmotic pump. The experiment was terminated 14 or 28 days after induction of fibrosis. Mice were euthanized and hearts were collected for molecular and histopathological analyses.
Histopathology
The hearts were harvested from the experimental animals and fixed in 10% buffered formalin, dehydrated by incubations in EtOH solutions with increasing EtOH content: 50, 70, 96, and 100%, and then in xylene and liquid paraffine. Subsequently, the tissue samples were embedded in paraffin, cut into 4 µm thick sections, and stained with hematoxylin and eosin (H&E) for morphological analyses, using a previously published protocol [
40]. Briefly, sections were deparaffinized, washed in xylene, and then rehydrated in EtOH solutions with increasing content of water. After staining in Mayer’s hematoxylin solution (Merck) and subsequent washing in running tap water, they were stained with 0.5% eosin Y solution with phloxine (Merck) (90 s). Next, the sections were again dehydrated in EtOH solutions and washed in xylene. Finally, the slides were sealed with a Histofluid mounting medium (Paul Marienfeld, Lauda-Königshofen, Germany). Microscopic evaluation of H&E samples was carried out by assessing the presence of inflammation according to a scale: 0 = absent, 1 = infiltration of inflammatory cells around the vessels, 2 = in < 50% of the tissue, 3 = in 50–75% of the tissue, 4 = in > 75% tissue.
Collagens were stained with Sirius red dye as previously described [
41,
42]. Rehydrated tissue sections were stained with Sirius red dye in aqueous solution of picric acid (Merck) for 1 h, followed by two consecutive washes in glacial acetic acid (0.5% v/v) in deionized H
2O. Next, the slides were dehydrated in EtOH solutions, followed by washing in xylene. After the staining sections were sealed with Histofluid mounting medium.
miRNA sequencing
The sequencing of miRNAs in EVs was performed using next-generation Ion TorrentTM technology. First, total RNA was isolated using the Total Exosome RNA & Protein Isolation Kit (Invitrogen), and ribosomal RNA was then removed using the Low Input RiboMinus Eukaryote System 2 kit (Invitrogen). Prior to library preparation, the miRNA fraction was assessed using the Small RNA 2100 Bioanalyzer Kit (Agilent). Libraries were prepared with the Ion Total RNA-Seq v2 kit (Thermo Fisher Scientific) according to the manufacturer's instructions for preparing small RNA libraries. The sequencing template was generated with the Ion PI ™ Hi-Q ™ OT2 200 kit on the Ion OneTouch ™ 2 system. Libraries were sequenced on the Ion Proton Sequencer using the Ion PI Hi-Q Sequencing 200 and the Ion PI Chips v3 chemistry (Thermo Fisher Scientific). Data analysis was performed in Torrent Suite v5.10.0 and miRNA readings were counted using the SmallRNA plugin. The percentage of miRNAs in EVs was calculated in R and in the Excell software. Target genes identification for selected miRNAs was performed using the TargetScanHuman 8.0 [
43], the miRDB [
44], and the DIANA mirPath v.3 [
45] databases. Functional enrichment for miRNA target genes and miRNA-gene network analyses were performed using the MicroRNA Enrichment Turned Network (Mienturnet) [
46].
Transfection of hCFs with miRNA-302 mimic/inhibitor
To verify miR-302b-3p activity on target genes, hCFs were transfected with the miR-302b-3p inhibitor (#AM17000; Ambion), miR-302b-3p mimic (mirVana #4464066; Invitrogen) or anti-miR miRNA negative control (#AM17010; Ambion). Briefly, 3 × 10e4 cells/well were seeded on a 24-well plate a day before. Transfection was done using 50 nM of either control, miR-302b-3p inhibitor, or miR-302 mimic using Dharmafect 1 (Dharmacon) as a transfection reagent, according to the manufacturer’s instruction. Approximately 16 h after transfection the medium was replaced with fresh medium supplemented with 2% FBS and TGF-β (1 ng/ml). Cells were treated with EV-H5 (2.5 × 10e4 EV particles/cell), where indicated, for 24 h. Then, EVs were removed and cells were cultured for the next 24 h, before harvesting for RNA analysis. Three independent experiments were done.
Quantitative real-time PCR (qPCR)
Total RNA was extracted from cultured cells using the GeneMATRIX Universal RNA/miRNA Purification Kit (EURx). In the case of EVs, RNA was purified with the Total Exosome RNA and Protein Isolation Kit (Invitrogen). For RNA isolation from mouse heart tissue and from hCFs transfected with miR302b-3p mimics/inhibitor, the Fenozol Plus reagent (A&A Biotechnology) was used. The procedures were done according to instructions provided by the vendors. RNA concentration and purity were determined using the Implen spectrophotometer. Reverse transcription was performed with 1–2 µg RNA and the NG dART RT-PCR Kit (EURx) in a C1000 Touch Thermal Cycler (BioRad). The synthesized cDNA was used to analyze the expression level of selected genes by the real-time qPCR method or stored at – 20°C until use. The reaction mixes contained cDNA, SybrGreen dye (Applied Biosystems/Thermo Fisher Scientific), and specific primer pairs. The reaction was run in the 7500 Fast Real-Time PCR System (Applied Biosystems) under the following conditions: 50°C, 2 min; 95°C, 10 min; and 40 cycles: 95°C, 15 s; 60°C, 1 min. Relative gene expression level was calculated using the ∆∆Ct method. The list of primer sequences is provided in Table
2.
Table 2
Primer sequences used for qPCR
Human genes |
OCT4 (Octamer-binding transcription factor 4) | 5460 | CCTTCGCAAGCCCTCATTTC | TAGCCAGGTCCGAGGATCAA |
SOX2 (SRY-box transcription factor 2) | 6657 | GGGAAAGTAGTTTGCTGCCTC | CAGGCGAAGAATAATTTGGGGG |
NANOG (Nanog homeobox) | 79923 | ACCTCAGCTACAAACAGGTGAAG | TTCTGCGTCACACCATTGCT |
B2M (beta2-microglobulin) | 567 | AATGCGGCATCTTCAAACCT | TGACTTTGTCACAGCCCAAGATA |
ACTA2 (Actin alpha 2, smooth muscle) | 59 | CTGTTCCAGCCATCCTTCAT | CCGTGATCTCCTTCTGCATT |
CCN2 (Cellular communication network factor 2; connective tissue growth factor) | 1490 | CCCCAGACACTGGTTTGAAGA | CCTCCCACTGCTCCTAAAGC |
COL1A1 (Collagen type I alpha 1 chain) | 1277 | CTTTGCATTCATCTCTCAAACTTAGTTTT | CCCCGCATGGGTCTTCA |
COL3A1 (Collagen type III alpha 1 chain) | 1281 | CTGGTGGTAAAGGCGAAATG | CCAGGAGCACCATTAGCAC |
FN1 (Fibronectin 1) | 2335 | TGTGGTTGCCTTGCACGAT | GCTTGTGGGTGTGACCTGAGT |
TAGLN (Taglin; transgelin) | 6876 | CGTGGAGATCCCAACTGGTT | AAGGCCAATGACATGCTTTCC |
TNC (Tenascin C) | 3371 | GGTCCACACCTGGGCATTT | TTGCTGAATCAAACAACAAAACAGA |
COL1A2 (Collagen type I alpha 2 chain) | 1278 | TGCTGCTGGTCAACCTGGTGC | ACTTCCAGCAGGACCGGGGG |
COL4A1 (Collagen type IV alpha 1 chain) | 1282 | CTAATCACAAACTGAATGACTTGACTTCA | AAATGGCCCGAATGTGCTTA |
TGFB1 (Transforming growth factor beta 1) | 7040 | AGGGCTACCATGCCAACTTCT | CCGGGTTATGCTGGTTGTACA |
SNAI1 (Snail family transcriptional repressor 1) | 6615 | GCTGCAGGACTCTAATCCAGA | ATCTCCGGAGGTGGGATG |
SNAI2 (Snail family transcriptional repressor 2) | 6591 | TGGTTGCTTCAAGGACACAT | GTTGCAGTGAGGGCAAGAA |
PFN1 (Profilin 1) | 5216 | ACCGCCTTCTGGTAATCTTGAG | TCCAGCATCCAGCAGACAAG |
PXN (Paxilin) | 5829 | CCCTGACGAAAGAGAAGCCTAAG | AGATGCGTGTCTGCTGTTGG |
TLN1 (Talin 1) | 7094 | CCCTGATGTGCGGCTTCG | TGTCCTGTCAACTGCTGCTTC |
SMAD2 (SMAD family member 2) | 4087 | CGTCCATCTTGCCATTCACG | CTCAAGCTCATCTAATCGTCCTG |
TGFBR2 (Transforming growth factor beta receptor 2) | 7048 | GACATCAATCTGAAGCATGAGAACA | GGCGGTGATCAGCCAGTATT |
BMPR2 (Bone morphogenetic protein receptor 2) | 659 | CTCAGTCCACCTCATTCATTTAACCG | ACAGAGACTGATGCCAAAGCAAT |
ROCK2 (Rho-associated coiled-coil containing protein kinase 2) | 9475 | CAACTGTGAGGCTTGTATGAAG | TGCAAGGTGCTATAATCTCCTC |
PFN2 (Profilin 2) | 5217 | GAGACTCTGGGTTCTAGCTGC | ACACCTTTCCCCACCAACAG |
CFL2 (Cofilin 2) | 1073 | GACTCCTTCGCTGTATCGTCT | TCTCTTTTTGATCTCCTCTTGTGT |
PTEN (Phosphatase and tensin homolog) | 5728 | CTCAGCCGTTACCTGTGTGT | AGGTTTCCTCTGGTCCTGGT |
CD44 (CD44 molecule) | 960 | CAGCTCATACCAGCCATCCAA | GACTGGAGTCCATATCCATCCTT |
CCND1 (Cyclin D1) | 595 | ATGCCAACCTCCTCAACGAC | TCTGTTCCTCGCAGACCTCC |
CCND2 (Cyclin D2) | 894 | GAAGCTGTCTCTGATCCGCA | TGCTCCCACACTTCCAGTTG |
18S rRNA (18S ribosomal RNA) | 106631781 | GTAACCCGTTGAACCCCATT | CCATCCAATCGGTAGTAGCG |
GAPDH (glyceraldehyde-3-phosphate dehydrogenase) | 2597 | GAAGGTCGGAGTCAACGGAT | AGTTGAGGTCAATGAAGGGGTC |
Mouse genes |
Acta2 (Actin alpha 2, smooth muscle) | 11475 | CAGGTGATCACCATTGGAAACGAACG | GACAGGACGTTGTTAGCATAGAGATCC |
Col1a1 (Collagen type I alpha 1 chain) | 12842 | GGAGAGAGCATGACCGATG | AAGTTCCGGTGTGACTCGTG |
Col3a1 (Collagen type III alpha 1 chain) | 12825 | TGACTGTCCCACGTAAGCAC | GAGGGCCATAGCTGAACTGA |
Ctgf (Connective tissue growth factor) | 14219 | AGACCTGTGCCTGCCATTAC | ACGCCATGTCTCCGTACATC |
Il6 (Interleukin 6) | 16193 | AGTCCTTCCTACCCCAATTTCC | TGGTCTTGGTCCTTAGCCAC |
Tnfα (Tumor necrosis factor alpha) | 21926 | CATCTTCTCAAAATTCGAGTGACAA | TGGGAGTAGACAAGGTACAACCC |
Eef2 (Eukaryotic translation elongation factor 2) | 13629 | ACAATCAAATCCACCGCCATC | AGCCATCCTTGCTCTGCTTA |
Western blot analysis
Cells and heart tissue samples were lysed with RIPA lysis buffer (Sigma) supplemented with protease and phosphatase inhibitors (Thermo Fisher Scientific) and sonicated (3 × 10 s). EVs were lysed by adding 1/5 volume of RIPA buffer. The lysates were centrifuged at 4 °C for 10 min at 10,000
g to isolate the protein fraction. The protein concentration was measured using a BCA assay (Thermo Fisher Scientific), according to the manufacturer’s instructions. 20 or 4 μg protein samples from cell lysates or EVs (respectively), or 20 µl out of 500 µl of each EV fraction collected from qEV/70 nm column, were separated by SDS-PAGE and then transferred onto polyvinylidene fluoride membranes (PVDF; BioRad) using a semi-dry transfer at 25V, 1.3A for 7–10 min (depending on the protein size) in Trans-Blot Turbo RTA Mini PVDF Transfer Kit (BioRad). Membranes were incubated in 5% (w/v) skimmed milk in TBST (0.05 (v/v) Tween 20 in Tris-buffered saline) or 3% BSA in TBST. The membranes were then incubated overnight at 4°C with primary antibodies diluted in 3% BSA in TBST. Next, they were washed with TBST three times and incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies in the same antibody buffer for 50 min at RT. After washing the membranes three times with TBST, the signal was detected with the chemiluminescence HRP substrate (Merck) in the ChemiDoc XRS + (BioRad) imager. Densitometric analysis of the resulting protein bands was performed with the Quantity One (BioRad) software. The list of antibodies is shown in Table
1.
Statistical analysis
All results are presented as the mean of the measurements with error bars representing the standard deviation (SD). Statistical analysis was performed based on the analysis of the normality of the distribution using the Shapiro–Wilk test. Homogeneity of variances was assessed using the F-test (two groups) or Brown–Forsythe test (three or more groups). Thereafter, the unpaired or paired standard Student’s t-test was used for equal variance or the Welch t-test for unequal variance. For multiple comparisons, one-way ANOVA was used followed by the Tukey post hoc test, in case if the data showed normal distribution and equal variances or Welch ANOVA with the Dunnett test for unequal variances. If the data were not normally distributed, the non-parametric Kruskal–Wallis test with Dunn’s post hoc test was used to calculate statistical significance. Statistical analysis was performed using GraphPad Prism 8.4.0 software and the differences were considered statistically significant at p < 0.05. Statistical tests used for data analysis are indicated in each figure legend.
Discussion
Efficient targeting of cardiac fibrosis remains a challenge for modern medicine. Although various therapeutic strategies have been tested to restore the physiological balance between ECM production and degradation to prevent heart scarring, none is routinely used in clinical practice [
10]. The anti-fibrotic treatment is hampered by the complexity of the molecular pathways implicated in the development of fibrosis which makes it difficult to design a specific drug. Currently, applied treatment strategies rely on the use of inhibitors of the renin–angiotensin–aldosterone system (RAAS), including renin blockers, inhibitors of the angiotensin-converting enzyme, angiotensin receptor blockers, and aldosterone antagonists [
56]. Other anti-fibrotic molecules were designed to disrupt the TGFβ signaling pathway [
57]. However, paradoxically, the use of pan-TGFβ antibody drugs in cancer patients caused cardiotoxicity [
58,
59]. In addition to biomolecules, various stem cell-based therapies were developed in order to reduce fibrotic scar formation and improve heart function [
60]. Importantly, the reparative role of stem cell transplants was primarily attributed to the paracrine factors released by these cells, which helped to mitigate the detrimental effects of prolonged inflammation and excessive activation of fibroblasts. Successful suppression of overactive cardiac fibroblasts has also been achieved with cytotoxic T cells modified to express a chimeric antigen receptor (CAR) targeting a myofibroblasts marker — the fibroblast activation protein [
61]. Although promising, this treatment strategy is still in its infancy and much research is needed to obtain an effective anti-fibrotic cure.
In our work, we utilized a novel class of therapeutics, based on EVs. Owing to their excellent biophysical properties, EVs can be successfully used as drug delivery tools, with superior characteristics compared to lipid-based nanocarriers [
14]. EV-based nanomedicines have already progressed to a pre-clinical and clinical stage, demonstrating efficacy as immunomodulatory, pro-regenerative, and anti-apoptotic agents, as well as vaccines [
62]. We took advantage of the exceptional characteristics of EVs and contributed to this rapid development of EV-based bio-active drugs by exploring the functional properties of hiPS-EVs derived from different oxygen conditions, to improve heart function.
The cellular environment has long been recognized as a key factor governing cell identity via direct or indirect mechanisms. Changes in the microenvironment may alter the epigenome of cells and drive evolutional forces [
63]. Particularly in the case of pluripotent stem cells, the partial pressure of oxygen has been shown to play a critical role for the maintenance of pluripotency [
47]. Low oxygen not only supports the expression of core transcription factors linked to pluripotency, such as OCT4 and NANOG, but also increases genome stability and multilineage differentiation potential [
24]. In this work we confirmed that physiological hypoxia (5% O
2) enhances the expression of
OCT4 and
NANOG in hiPSCs (Fig.
1). More importantly, we demonstrated that changes in the environment of hiPSCs affect the molecular cargo of the EVs, altering their biological function. Strikingly, hiPS-EVs derived from hypoxic conditions at 5% O
2 exerted stronger anti-fibrotic activity, compared to EVs derived from normoxia (21% O
2) or hypoxia 3% O
2. This observation was consistent for EVs derived from three independent hiPSC lines, which were used in this study. This allows us to conclude that the significant enhancement of the anti-fibrotic function of hiPS-EV in hypoxia is independent of the cell line, despite the differences that naturally occur between the individual hiPSC lines [
64].
Other critical factors that affect the cargo of EVs, are the composition of the media for cell culture [
65], the storage conditions of EVs [
66], and the isolation procedure [
67,
68]. In our study, we used a serum-free and chemically defined cell culture medium to propagate hiPSCs, which is also available in a good manufacturing practice grade. Therefore, this cell culture system may easily be adapted to pharmaceutical production in the future. To isolate biologically active EVs, we selected the UF + SEC method with UF tubes with a protein cut-off at 10 kDa, which generated a higher number of EV particles and increased protein yield, compared to UF tubes with a higher protein cut-off (Additional file
1: Fig. S2). EVs isolated this way reduced the expression of pro-fibrotic markers in activated hCFs more efficiently than EVs obtained with the UC method (Additional file
1: Fig. S2J). The increased potency of EVs purified with UF + SEC is consistent with a previous study [
48] and may partially be attributed to better homogeneity of the obtained EVs, since the UC method was shown to induce the aggregation of EVs [
69]. Moreover, we cannot exclude the possibility of co-purification with EVs of other bioactive molecules, such as ribonucleoproteins, exogenous RNAs, or exomeres [
70]. Such molecules may decorate the surface of EVs affecting their biological functions.
Comparison of EVs derived from different oxygen concentrations (N, H5, and H3) revealed a higher number of EV particles obtained for EV-H5, compared to other conditions (Fig.
2D). This observation is consistent with data from other researchers, indicating that low oxygen availability triggers the release of EVs via the hypoxia-inducible factor (HIF)-dependent pathway (reviewed in [
26]). We also noted that EVs obtained in hypoxia had a smaller size than EV-N (Fig.
2B, C), which corresponded to a higher level of CD81 protein (Fig.
2G, H). Since CD81 is an established marker for exosomes (small EVs) [
49], we conclude, that we have enriched our EV-H5 preparations with a subpopulation of exosomes. However, further research is needed to distinguish individual fractions in the heterogeneous population of EVs, preferably by using single-vesicle technologies [
71], and to determine their biological function.
Interestingly, our analysis of the EV-associated protein markers showed that the most abundant expression was detected for syntenin, regardless of oxygen condition and the type of producer cells (Fig.
2G). This protein was easily detected in both hiPS-EVs and DF-EVs. Syntenin was described to be involved in exosome biogenesis via direct interaction with ALG-2-interacting protein X (ALIX) [
72] and recently was proposed as a universal biomarker of exosomes [
73]. This study appears to support this notion.
The comparison of the antifibrotic properties of hiPS-EVs revealed that all types of EVs reduced fibrotic markers in activated hCFs. However, treatment with EV-H5 led to the most significant reversal of the fibrotic phenotype (Fig.
3). We showed that EV-H5 ameliorated the formation of α-SMA-positive stress fibers and attenuated the length and area of focal adhesion sites, which resulted in reduced cell stiffness (Fig.
4, Additional file
1: Fig. S4). On the molecular level, EV-H5 prevented the translocation of pSMAD2 to the cell nucleus, and downregulated the expression of several profibrotic gene transcripts (Fig.
5, Additional file 1: Fig. S
5). Thus, by disrupting the fibrotic pathway in hCFs, EV-H5 prevented their transformation to pathogenic myofibroblasts.
Mechanistically, we identified that the microRNA (miRNA) cargo of hiPS-EVs was primarily responsible for attenuating myofibroblast differentiation of hCFs and the production of fibrotic markers.
Although recent studies quantifying the content of miRNAs per single EV particle revealed their small amount (approximately one copy per 10–100 EVs) [
74,
75], the role of miRNAs in the biological activity of EVs exerted on recipient cells has been well documented [
12,
76]. In addition, the miRNA regulatory network has been shown to be an important factor in the regulation of cardiovascular homeostasis [
77].
Our miRNA screening revealed that the pluripotency-associated miR-302 cluster was the most abundant in hiPS-EVs, which is consistent with our previous study [
15] and data from other groups [
22,
23]. However, there are some differences in the individual miRNA species detected at the highest level in hiPS-EVs in different studies, which may result from different cell culture systems used. Particularly, we detected a significantly elevated level of miR-302b-3p in EV-H5, compared to EVs released under other oxygen conditions (Fig.
6). Considering the essential role of miR-302 cluster in the maintenance of pluripotency in hiPSCs, we speculate that the higher level of miR-302b-3p in H5-EVs may be related to the fine-tune regulation of pluripotency factors in hiPSCs. EVs in this case may be utilized to egress the excessive amount of unused RNAs, including miRNAs, and other biomolecules. This hypothesis, however, needs further verification and is currently under investigation.
With respect to its function in mitigating fibrosis, our bioinformatics analysis revealed a critical role of miR302b-3p in the regulation of TGFβ pathway and actin cytoskeleton, which are both severely affected in fibrotic disease. To support this statement, we provide direct evidence of a downregulation of transcript levels for several genes from the TGFβ/SMAD pathway, actin cytoskeleton, the regulators of cell motility and cell cycle, after treatment of hCFs with EV-H5 or miR-302b-3p mimic (Fig.
7). This effect was abolished when EV-H5 were used along with the miR302b-3p inhibitor. Our results are in line with and extended data from other reports showing the inhibitory effect of miR-302 family members on TGFβ/SMAD2 signaling [
23,
78]. Based on sequence similarity, these miRNAs may cooperate and target a similar set of genes, boosting the biological effect.
The achieved enhancement of the antifibrotic properties of EVs obtained from physiological hypoxia was further confirmed in in vivo studies. The development of fibrotic scar during the natural healing process is always preceded by extensive inflammation [
10]. We have shown that EV-H5 inhibit infiltration of immune cells in mouse hearts and reduce the level of pro-inflammatory cytokines (Fig.
8). The resulting decreased level of proinflammatory mediators further translated into reduced deposition of collagens and downregulated expression of α-SMA in the heart tissue (Fig.
9). Our data remain in agreement with other reports, confirming the anti-inflammatory properties of hiPS-EVs [
18] and their potential application as anti-fibrotic drugs [
22,
23]. We advanced these observations by demonstrating that hiPS-EVs derived from physiological hypoxia exhibit significantly enhanced anti-inflammatory, anti-fibrotic, and pro-regenerative properties. Data presented in this work open new avenues in the therapeutic application of EVs to treat heart fibrosis. However, translating these findings into the clinical practice requires further studies in order to (i) fully optimize the dose and the treatment regimen in human patients; (ii) select the route of EV administration; (iii) analyze the pharmacokinetics of infused EVs; and (iv) determine the safety profile of EVs. If successfully established and approved, EV-based therapeutics can offer a new solution in the treatment of end-stage heart disfunction. Moreover, given the common molecular background of fibrotic diseases, our findings can be extended beyond the field of cardiology and used for the treatment of fibrotic disease in other organs, such as the liver, lungs, kidney, and pancreas, among others. Using in vitro models and studies in small animals, we are currently testing the antifibrotic properties of H5-EVs in various tissues. Thus, our work may contribute to the development of innovative EV-based treatment strategies applicable for a large group of patients.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.