Abstract
A recent study reported the cardioprotective effects mediated by cardiac fibroblasts (CFs) during acute phase of ischaemia–reperfusion injury (IRI). Little is known about whether exosomes/microvesicles mediate this beneficial effect and whether ischaemia post-conditioning (Postcon) can regulate this process. Here, we aimed to investigate the cardioprotective effect of CFs-exosomes/microvesicles and whether Postcon can regulate this effect.
By using transwells co-culture system, we found that hypoxia–reoxygenation (H/R) significantly increased the exosomes/microvesicles secretion of CFs and CFs protected H9C2 cells against H/R injury and Postcon could amplify these effects. Inhibition of CFs exosomes/microvesicles secretion led to a significant abrogation on the amplified protective effect of H/R-Postcon. We further demonstrated that Postcon enhanced the cardioprotective effect of CFs-exosomes/microvesicles both in vitro and in vivo. To detect the underlying mechanism, exosomes/microvesicles microRNAs were analysed by RNA sequencing and quantitative polymerase chain reaction, our results revealed that miR-423-3p expression was selectively enhanced by Postcon in CFs exosomes/microvesicles. By co-culture H9C2 cells with CFs-exosomes/microvesicles enriching with miR-423-3p, we demonstrated that H/R-Postcon exerted cardioprotective effects by upregulation of miR-423-3p in CFs-exosomes/microvesicles. RNA-fluorescence in situ hybridization and qPCR demonstrated that the decreasing of miR-423-3p is closely related to IRI, by inhibited miR-423-3p expression with its antagomir in vivo, we demonstrated that miR-423-3p plays an essential mediate role in I/R-Postcon-induced cardioprotection against I/R in vivo, Postcon may exert cardioprotective effect by upregulation of miR-423-3p in CFs exosomes/microvesicles. Gain- and loss-of-function approaches suggested that rescuing the down-regulated miR-423-3p might be a potential strategy to protect the cardiomyocytes against H/R. Using computational predictions tools and luciferase reporter assay, we demonstrated that miR-423-3p regulates the expression of Ras-related protein Rap-2c (RAP2C) in H9C2 cells, and knockdown of RAP2C by siRNA obviously increased cell viability and reduced apoptosis in H9C2 cells under H/R.
In conclusion, we demonstrated, for the first time, that CFs participate in cardioprotective effects via an exosomes/microvesicles pathway during the acute phase of IRI and Postcon can enhance this effect by upregulating the expression of CFs exosomes/microvesicles miR-423-3p, which targets the downstream effector RAP2C.
This article is part of the Spotlight Issue on Cardioprotection Beyond the Cardiomyocyte.
1. Introduction
Early reperfusion with primary percutaneous coronary intervention reduces myocardial infarct size but is linked with increased risk of inducing ischaemia–reperfusion injury (IRI).1 IRI accounts for up to 50% of the final infarct size.2 Experimental studies showed that ischaemic post-conditioning (Postcon) could reduce IRI and attenuate cardiomyocytes (CMs) apoptosis,3,4 but recent clinical trials have encountered challenges in translating the beneficial effects of Postcon from bench to bedside.5–8 Most Postcon studies assumed that the CMs were the target of Postcon.9 However, CMs are found to contribute only 56% of cell numbers of the heart, other cell numbers are fibroblasts (27%), endothelial cells (7%), and vascular smooth muscle cells (10%).10 Thus, it is reasonable to argue that the target cells of Postcon might include other cell types beyond CMs.
Cardiac fibroblasts (CFs), the largest cell population within the heart,11 which play important roles in providing the three-dimensional structure and milieu in which CMs and other myocardial cell types reside, and some of the functions are known to be mediated in paracrine fashion, through releasing various endogenous factors that contribute to the conditioning stimulus, such as gaseotransmitters (NO, CO), acid and basic fibroblast growth factors (aFGF/FGF-1 and bFGF/FGF-2),12,13 might be a promising target of Postcon. In fact, Cooper and Ytrehus14 have demonstrated that CFs mediated the protective effects of ischaemic preconditioning against IRI. However, there is no report about the effect of Postcon on CFs in the process of IRI.
Exosomes/microvesicles, which are able to carry a cargo including proteins, messenger RNA (mRNA) and microRNA (miRNA), and can transfer these cargos to recipient cells, thus serving as major mediators of cell-to-cell communication.15 Recent studies supplied some evidences regarding the important role of exosome encased miRNA in cell-to-cell communication within the cardiovascular system, e.g. from stem cells to cardiovascular cells and from the heart to the bone marrow stem cells.16 CFs were known to secrete exosomes/microvesicles enriched with miR-21* as a crucial paracrine signalling mediator of cardiac hypertrophy.17
Chen et al. demonstrated that ischaemic preconditioning potentiates the protective effect of stem cells through secretion of exosomes/microvesicles by targeting Mecp2 via miR-22.18 Giricz et al.19 reported that exosomes/microvesicles released from the rat heart after ischaemic preconditioning are critical for the cardioprotective effect of remote conditioning. These studies thus suggest that various ischaemia conditioning can affect the production and miRNAs contents of exosomes/microvesicles.
Another recent study demonstrated that CFs participate in cardioprotective effects during the acute phase of ischaemia–reperfusion via a paracrine pathway involving tissue inhibitors of metalloproteinase-1 (TIMP-1).20 In the present study, using both in vitro and in vivo experimental preparations, we tested the hypothesis that CFs might be a target of Postcon and showed that CFs reduce IRI through secreting exosomes/microvesicles, and Postcon enhanced the cardioprotective effect of CFs via increasing exosomes/microvesicles secretion and upregulating the exosomes/microvesicles miR-423-3p expression and the downstream signalling was RAP2C.
2. Methods
Full methods are available in the Supplementary material online and data. All the experiments were performed in accordance with the recommendations of the European Ethical Committee (EEC) (2010/63/EU) and the US National Institutes of Health (NIH Publication No. 85-23, revised 1996) for the care and use of laboratory animals. Protocols were approved by the Institutional Animal Care and Use Committee of Central South University.
2.1 Isolation and enrichment of neonatal rat cardiac fibroblasts
Primary cultures of neonatal rat CFs were prepared from hearts of 2 day-old rat (Sprague-Dawley, Department of Laboratory Animals, Central South University) as described previously.21
2.2 Hypoxia–reoxygenation (H/R) and H/R-Postcon
H9C2, a rat CM cell line, was obtained from the Cell Bank of China Science Academy (Shanghai, China) and, were cultured in complete medium composed of high glucose (4.5 g/L) dulbecco's modified eagle medium (DMEM), 10% fetal bovine serum (FBS), and 1% penicillin/streptomycin. Cells were plated at an appropriate density according to each experimental design. Culture media was changed every 2 days, and cells were synchronized using 1% FBS-high glucose (4.5 g/L) DMEM once reaching 80% confluence.
For H/R, cells were transferred to a hypoxic incubator in a humidified atmosphere for 9 h (95% N2, 5% CO2, PO2 range in the incubator was maintained at <1%) in a culture medium deprived of glucose and serum, followed by reoxygenation under a normoxic condition in complete medium composed by high glucose (4.5 g/L) DMEM, 1% pre-exosome-depleted FBS (FBS is depleted of contaminating bovine exosomes/microvesicles by ultracentrifugation for at least 6 h at 100 000 g) and 1% penicillin/streptomycin for 15 h. The control group was maintained in complete medium composed by high glucose (4.5 g/L) DMEM, 1% pre-exosome-depleted FBS, and 1% penicillin/streptomycin in normoxic conditions for 24 h.
For H/R-Postcon, cells were transferred between hypoxic and normoxic incubators as described previously.22 Briefly, after the completion of 9 h of hypoxia, the cells were transferred to a normoxic incubator for 5 min then returned to the hypoxic incubator for an additional 5 min without a change in culture medium. The Postcon cycle was repeated three times and was followed by 15 h of continuous normoxia with a change to complete culture medium composed of high glucose (4.5 g/L) DMEM, 1% pre-exosome-depleted FBS and 1% penicillin/streptomycin.
2.3 Exosomes/microvesicles purification
Exosomes/microvesicles were purified from CFs conditioned medium after treatment with H/R, H/R-Postcon, and normoxic incubation. Briefly, 30 mL of conditioned solution of various groups was centrifuged at 3000 rpm, 4°C for 15 min to remove cells, followed by filtration through a 0.22 μm filter to remove cell debris. Exosomes/microvesicles in the medium were precipitated with ExoQuick-TC (System Biosciences) according to the manufacturer’s instruction, the pellets of exosomes/microvesicles were re-suspended in 50 μL phosphate buffer saline (PBS) and stored at −80°C. The amount of exosomes/microvesicles was quantified by measuring protein concentrations, total RNA concentrations, and protein levels of exosome markers: CD9 and CD63. The protein concentration was determined by the bicinchoninic acid (BCA) assay and, the protein expression of CD9 and CD63 was measured by western blotting.
2.4 Co-culture experiments and exosomes/microvesicles inhibition
The objective was to determine whether Postcon affects the paracrine interaction between CFs and H9C2 cells during H/R and whether exosomes/microvesicles would play an important role in this interaction. We set up a co-culture model using transwells (0.4 μm pore size, Corning) as described previously.20 CFs were positioned above H9C2 cells in culture dishes, thereby allowing a paracrine interaction with H9C2 cells during all sequences of H/R.
For exosomes/microvesicles inhibition, we treated CFs with a chemical exosome inhibitor (GW4869, Sigma, USA) for 48 h. Thereafter, medium was changed to high glucose (4.5 g/L) DMEM supplemented with 1% pre-exosome-depleted FBS + 10 µM GW4869. As vehicle control, CFs were treated with Dimethyl sulfoxid (DMSO). Furthermore, control cells or CFs treated with GW4869 were fixed in 4% Paraformaldehyde (PFA) for CD63 immunofluorescence staining.
For CFs exosomes/microvesicles treatment of H9C2 cells, H9C2 cells underwent hypoxia while maintained in the culture medium deprived of glucose, serum and were supplemented with H/R, H/R-Postcon, and normoxia induced CFs-exosomes/microvesicles (25 µg/mL), or vehicle (PBS), followed by reoxygenation under a normoxic condition in complete medium composed of high glucose (4.5 g/L) DMEM, 1% pre-exosome-depleted FBS, and 1% penicillin/streptomycin supplemented with H/R, H/R-Postcon, and normoxia induced CFs-exosomes/microvesicles (25 µg/mL), or vehicle (PBS) for 15 h.
2.5 Western blotting
Total protein was extracted with RIPA lysis buffer, and 20 µg protein was used for western blotting as described previously.23 Specific primary antibodies used for western blotting are presented in Supplementary material online, Table S1.
2.6 miRNA library construction and sequencing
Total RNA was extracted from exosomes/microvesicles using the miRNeasy Mini Kit (QIAGEN Sample and Assay Technologies, Germany), according to the manufacturer’s instructions. A total amount of 3 μg total RNA per sample was used as input material for the small RNA library. Sequencing libraries were generated using NEBNext® Multiplex Small RNA Library Prep Set for Illumina® (NEB, USA) following manufacturer’s recommendations and index codes were added to attribute sequences to each sample. The clustering of the index-coded samples was performed on a cBot Cluster Generation System using TruSeq SR Cluster Kit v3-cBot-HS (Illumia) according to the manufacturer’s instructions. After cluster generation, the library preparations were sequenced on an Illumina Hiseq 2500/2000 platform and 50 bp. Mapped small RNA tags were used to look for known miRNA. miRBase20.0 was used as reference. Modified software mirdeep2 and srna-tools-cli were used to obtain the potential miRNA and draw the secondary structures. Custom scripts were used to obtain the miRNA counts as well as base bias on the first position of identified miRNA with certain length and on each position of all identified miRNA, respectively. Differential expression analysis of two conditions/groups was performed using the DESeq R package (1.8.3). The P-values were adjusted using the Benjamini–Hochberg procedure. Corrected P-value of 0.05 was set as the threshold for significantly differential expression by default.
2.7 qPCR analysis
Total RNA was extracted from exosomes, cells, and tissue using the miRNeasy Mini Kit (QIAGEN Sample and Assay Technologies, Germany), according to the manufacturer’s instructions. mRNAs reverse transcription and Real-time Quantitative Polymerase Chain Reaction (qPCR) was performed as described previously.23 The mRNA and miRNAs primers are shown in Supplementary material online, Tables S2 and S3. miRNA PCR was performed using the All-in-One™ miRNA qPCR Kit (GeneCopoeia, USA) in the ABI Vila7 Real-Time PCR System (Applied Biosystems, USA). Each sample was analysed in triplicate. β-Actin was used as the internal control for mRNA normalization and U6 as the internal control for miRNA normalization.
2.8 Data analysis and statistics
All data are presented as mean ± SD. An unpaired, two-tailed Student’s t-test was used for statistical analysis of two groups; Analysis of Variance (ANOVA) followed by Bonferroni’s post-test was used for comparison of three or more groups. Data analysis was performed using SPSS 17.0 (SPSS Inc., Chicago, IL, USA), and figures were prepared using the GraphPad Prism software.
3. Result
3.1 Cardiac fibroblasts protect H9C2 cells against H/R in an exosomes/microvesicles paracrine manner and H/R-Postcon amplifies this protective effect in vitro
As envisaged, cells viability of H9C2 cells was significantly reduced post-H/R compared to normoxia condition. When H9C2 cells were co-cultured with CFs, viability was significantly improved from 32.0% to 59.9% (Figure 1A). To detect the apoptosis of H9C2 cells, the H9C2 cells were harvested for protein collection and western blotting for cleaved-caspase3 and pre-caspase3 after H/R. H/R significantly increased H9C2 cells apoptosis compared to normoxia. When co-cultured with CFs, H9C2 cells apoptosis was significantly reduced compared to H/R H9C2 alone (Figure 1B). These results demonstrated that CFs protected H9C2 cells against H/R insult H9C2. Moreover, H/R-Postcon further significantly amplifies the protective effects of CFs (Figure 1A, B).
Cardiac fibroblasts protect H9C2 cells against H/R in an exosomes/microvesicles paracrine manner and Post-conditioning amplifies this protection. (A) 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2-H-tetrazolium bromide, Thiazolyl Blue Tetrazolium Bromide (MTT) assay measuring H9C2 cells viability was performed after H/R and H/R-Postcon. Results are expressed in % of control (n = 6, **P < 0.01 vs. mono-culture, ##P < 0.01 vs. co-culture during H/R, H/R, Hypoxia/Reoxygenation; H/R-Postcon, Hypoxia/Reoxygenation-Post-conditioning). (B) Western blotting analysis of protein expression of caspase-3 in H9C2 cells after H/R and H/R-Postcon (n = 6, **P < 0.01 vs. mono-culture, ##P < 0.01 vs. co-culture during H/R, H/R, Hypoxia/Reoxygenation; H/R-Postcon, Hypoxia/Reoxygenation-Post-conditioning). (C) Immunofluorescence staining of rat cardiac fibroblasts (CFs) treated with GW-4869, a chemical exosomes/microvesicles inhibitor. Treatment with 10 μM GW-4869 for 48 h resulted in accumulation of CD63-positive microvesicles (green) in CFs. Nuclei are stained with DAPI (blue). Scale bar: 50 μm. exosomes/microvesicles (D) MTT assay measuring H9C2 cells viability while co-cultured with GW4869 or control treated CFs was performed after H/R. Results are expressed in % of control (n = 6, **P < 0.01 vs. control, ##P < 0.01 vs. H/R group, &P < 0.05 vs. H/R+CF group, H/R, Hypoxia/Reoxygenation). (E) Western blotting analysis of the protein expression of caspase-3 in H9C2 cells viability while co-cultured with GW4869 or control treated CFs was performed after H/R (n = 6, **P < 0.01 vs. control, ##P < 0.01 vs. H/R group, H/R, Hypoxia/Reoxygenation). (F) MTT assay measuring H9C2 cells viability while H9C2 cells were co-cultured with GW4869 treated CFs to undergo H/R-Postcon (n = 6, **P < 0.01 vs. H/R group, ##P < 0.01 vs. H/R+CFs group, &&P < 0.05 vs. H/R-Postcon+CFs group, H/R, Hypoxia/Reoxygenation; H/R-Postcon, Hypoxia/Reoxygenation-Post-conditioning). (G) Western blotting analysis of the protein expression of caspase-3 in H9C2 cells viability while H9C2 cells were co-cultured with GW4869 treated CFs to undergo H/R-Postcon (n = 6, **P < 0.01 vs. H/R group, ##P < 0.01 vs. H/R+CFs group, &&P < 0.05 vs. H/R-Postcon+CFs group, H/R, Hypoxia/Reoxygenation; H/R-Postcon, Hypoxia/Reoxygenation-Post-conditioning).
To test if exosomes/microvesicles mediate the paracrine cardioprotective effect of CFs, we treated CFs with GW4869, a chemical inhibitor of neutral sphingomyelinase 2 (NSMASE2), which has been proven to inhibit exosome secretion. Treatment with GW4869 at a concentration of 10 mM for 48 h resulted in a significant stronger accumulation of CD63 (an exosomes/microvesicles maker)-positive vesicles within CFs than control and CFs treated with DMSO, suggesting that GW4869 can effectively inhibit the exosome secretion of CFs (Figure 1C). We also measured the total amount of proteins (concentration) in total exosomes/microvesicles released from the same number of CFs after the treatment protocol as noted above. Treatment with GW4869 reduced protein quantity (Supplementary material online, Figure S1). Inhibition of CFs exosome secretion led to a significant abrogation on the protective effect of CFs to hypoxic H9C2 cells, evidenced by both cell viability and western blotting results for cleaved-caspase-3 and pre-caspase-3 (Figure 1D, E). These results indicate that exosomes/microvesicles are essential for paracrine cardioprotective effects of CFs. While H9C2 cells were co-cultured with GW4869 treated CFs to undergo H/R-Postcon, the amplified cardioprotective effect of CFs was significantly abrogated, evidenced by both cell viability and western blotting results for cleaved-caspase-3 and pre-caspase-3 (Figure 1F, G). These results indicated that exosomes/microvesicles also mediate the amplified paracrine cardioprotective effect of H/R-Postcon.
3.2 H/R and H/R-Postcon stimulate release of exosomes/microvesicles from cardiac fibroblasts and CFs-exosomes/microvesicles are taken up by H9C2 cells
To investigate the role of exosomes/microvesicles in the paracrine cardioprotective effects of CFs, we isolated exosomes/microvesicles from conditioned medium of CFs. Using transmission electron microscopy (TEM), we observed the typical ‘cup-shaped’ vesicles of exosomes/microvesicles that were <100 nm in diameter (Figure 2A). To further investigate the size distribution profile of the CFs-derived exosomes/microvesicles, we performed size analysis using the nanoparticle tracking system, revealing the average size of exosomes/microvesicles purified from normoxia-treated CFs was 90.25 nm (Figure 2B), and the size of exosomes/microvesicles purified from H/R- and H/R-Postcon-treated CFs was similar as control (Supplementary material online, Figure S2). Western blotting confirmed the presence of exosome-associated protein markers such as CD9, CD63, and tumour susceptibility gene (TSG101) in CFs-exosomes/microvesicles, while only TSG101 presented in CFs (Figure 2C). To determine whether the exosomes/microvesicles release is altered by H/R and H/R-Postcon in CFs, we measured the total amount of proteins (concentrations) and expression levels of exosomes/microvesicles specific marker CD9 and CD63 in total exosomes/microvesicles released from the same number of CFs of various groups. As expected, H/R and H/R-Postcon both enhanced protein quantity as well as CD9 and CD63 protein expression in the CFs-exosomes/microvesicles (Figure 2D–G). Interestingly, protein quantity and the protein expression of exosome markers CD9 and CD63 in H/R-Postcon CFs-exosomes/microvesicles were significantly higher than in H/R CFs-exosomes/microvesicles, indicating that Postcon can further enhance the release of exosomes/microvesicles from CFs compared to H/R (Figure 2D–G). To determine whether CFs-exosomes/microvesicles can be efficiently taken up by H9C2 cells in vitro, we labelled CFs-exosomes/microvesicles with PKH26, a red fluorescent marker that is incorporated into the cell membrane. The labelled CFs-exosomes/microvesicles were then incubated with H9C2 cells, which were examined for fluorescence at 0 min/30 min/120 min/12 h. Thirty minutes after PKH26-labelled CFs-exosomes/microvesicles were added to the culture medium, red fluorescence signals were detected in nearly every H9C2 cells. By 12 h, every H9C2 cell exhibited intense red fluorescence. These results suggested an increasing uptake of CFs-exosomes/microvesicles by H9C2 cells over time (Figure 2H, I).
H/R and post-conditioning stimulate release of exosomes/microvesicles from cardiac fibroblasts and CFs-exosomes/microvesicles were uptake by H9C2 cells. (A) Electron microscopy image of rat cardiac fibroblast–derived exosomes/microvesicles, showing a size of approximately 50 to 100 nm in diameter. Scale bar: 100 nm exosomes/microvesicles. (B) Nanoparticle trafficking analysed the diameters distribution of CFs-exosome sizes. (C) Western blotting analysis of several exosome biomarkers (CD9, CD63, and TSG101)in CFs-exosomes/microvesicles. (D) The amount of proteins in total released exosomes/microvesicles from the same number of CFs that were stimulated with vehicle, H/R and H/R-Postcon (n = 6, **P < 0.01 vs. control-CFs-exo, ##P < 0.01 vs. H/R-CFs-exo, H/R, Hypoxia/Reoxygenation; H/R-Postcon, Hypoxia/Reoxygenation-Post-conditioning; exo, exosomes/microvesicles). (E) Western blotting analysis of CD9 and CD63 in total released exosomes/microvesicles from the same number of CFs that were stimulated with vehicle, H/R, and H/R-Postcon (H/R, Hypoxia/Reoxygenation; H/R-Postcon, Hypoxia/Reoxygenation-Post-conditioning). (F) Quantification analysis of CD9 and CD63 (G) expression are showed beneath (n = 6, **P < 0.01 vs. control-CFs-exo, ##P < 0.01 vs. H/R-CFs-exo, H/R, Hypoxia/Reoxygenation; exo, exosomes). (H) H9C2 cells were incubated with PKH26-labelled (red) exosomes/microvesicles from CFs and fixed for confocal imaging. H9C2 cells were incubated with PKH26-labelled exosomes/microvesicles for 30 min, 2 h, and 12 h. H9C2 cells were stained with carboxyfluorescein succinimidyl amino ester (CFSE) (green) and nuclei with DAPI (blue), Scale bar: 25 μm. (I) Quantification analysis of PKH26-labelled (red) exosomes/microvesicles in H9C2 cells (n = 6, **P < 0.01 vs. 0 h group, ##P < 0.01 vs. 0.5 h group, &&*P < 0.01 vs. 2 h group).
3.3 H/R-Postcon enhances the cardioprotective effects of CFs-exosomes/microvesicles both in vitro and in vivo
In vitro, in order to determine the cardioprotective effects of CFs-exosomes/microvesicles on H9C2 cells during H/R, the H9C2 cells were pre-treated with PBS, H/R-CFs-exosomes/microvesicles, or H/R-Postcon-CFs-exosomes/microvesicles, and then subjected to H/R. Although H/R-CFs-exosomes/microvesicles and H/R-Postcon-CFs-exosomes/microvesicles both improved H9C2 cells viability, this effect was more significant for H/R-Postcon-CFs-exosomes/microvesicles (Figure 3A and Supplementary material online, Figure S3). Similarly, reduction of H9C2 cells apoptosis was more significant post H/R-Postcon-CFs-exosomes/microvesicles as compared to H/R-CFs-exosomes/microvesicles (Figure 3B). These results indicate that Postcon amplified the cardioprotective effects of CFs-exosomes/microvesicles in vitro.
H/R-Postcon enhanced the cardioprotective effect of CFs-exosomes/microvesicles both in vitro and in vivo. (A) CCK-8 assay measuring viability of H9C2 cells pre-incubated with H/R, H/R-Postcon induced CFs-exosomes/microvesicles and vehicle (PBS), followed by H/R. Results are expressed as % of control (n = 6, **P < 0.01 vs. control, ##P < 0.01 vs. H/R, &&P < 0.01 vs. H/R + H/R-exo, H/R, Hypoxia/Reoxygenation; H/R-Postcon, Hypoxia/Reoxygenation-Post-conditioning; exo, exosomes/microvesicles). (B) Western blotting analysis of the protein levels of caspase-3 in H9c2 cells pre-incubated with H/R, H/R-Postcon-CF-exosomes/microvesicles, and vehicle (PBS), followed by H/R (n = 6, **P < 0.01 vs. H/R, ##P < 0.01 vs. H/R + H/R-exo, H/R, Hypoxia/Reoxygenation; H/R-Postcon, Hypoxia/Reoxygenation-Post-conditioning; exo, exosomes/microvesicles). (C) Representative images of Evans Blue and TTC-stained hearts intra-myocardially injected with H/R, H/R-Postcon-CFs-exosomes/microvesicles and vehicle (PBS). (D) Quantitative analysis of infarct zone (white line) (n = 6, *P < 0.05 vs. I/R, **P < 0.01 vs. I/R, ##P < 0.01 vs. I/R + H/R-exo, I/R, ischaemia/reperfusion; H/R, Hypoxia/Reoxygenation; H/R, Hypoxia/Reoxygenation; H/R-Postcon, Hypoxia/Reoxygenation-Post-conditioning; exo, exosomes/microvesicles). (E) Quantification of TUNEL-positive area (green) and (F) Quantification of TUNEL positivity nuclei (green), (n = 6, *P < 0.05 vs. I/R, **P < 0.01 vs. I/R, ##P < 0.01 vs. I/R + H/R-exo). (G) Representative images of TUNEL (green)-stained heart tissue from the infarct zones of H/R, H/R-Postcon-CFs-exosomes/microvesicles and vehicle (PBS) treated hearts. Hearts were stained with α-actin (red) and DAPI (blue), Scale bar: upper panel 100 μm, lower panel 2000 μm.
In vivo, we injected PKH26-labelled exosomes/microvesicles in the infarct border zone. After 24 h, the heart was excised and stained with α-actin and 2-(4-Amidinophenyl)-6-indolecarbamidine dihydrochloride (DAPI) and found that these exosomes/microvesicles could be up taken by CMs (Supplementary material online, Figure S4). We then intra-myocardially injected PBS, H/R-CFs-exosomes/microvesicles or H/R-Postcon-CFs-exosomes/microvesicles in the infarct border zone immediately after reperfusion, and area at risk (AAR) was similar among animals (Figure 3C and Supplementary material online, Figure S5). Both H/R-CFs-exosomes/microvesicles and H/R-Postcon-CFs-exosomes/microvesicles treated animals exhibited significantly reduced infarct size compared to PBS group. Again, infarct size reduction was more significant in H/R-Postcon-CFs-exosomes/microvesicles group than in H/R-CFs-exosomes/microvesicles group (Figure 3C, D). We also measured collagen deposition by picro sirius red staining 24 h after reperfusion. Significant collagen deposition in heart tissue was seen in IR group, and delivery of either H/R-CFs-exosomes/microvesicles or H/R-Postcon-CFs-exosomes/microvesicles to ischaemic myocardium before reperfusion significantly reduced collagen deposition compared to animals treated with PBS, and this beneficial effect was more significant in H/R-Postcon-CFs-exosomes/microvesicles group than in H/R-CFs-exosomes/microvesicles group (Supplementary material online, Figure S6).We then assayed TUNEL (+)/α-actin (+) apoptotic CMs 24 h after reperfusion. Delivery of either H/R-CFs-exosomes/microvesicles or H/R-Postcon-CFs-exosomes/microvesicles to ischaemic myocardium before reperfusion significantly reduced CM apoptosis compared to animals treated with PBS, and this beneficial effect was more significant in H/R-Postcon-CFs-exosomes/microvesicles group than in H/R-CFs-exosomes/microvesicles group (Figure 3E–G). These results suggest Postcon enhances the cardioprotective effects of CFs-exosomes/microvesicles in vivo.
3.4 miRNA expression profile of H/R-CFs-exosomes/microvesicles and H/R-Postcon-CFs-exosomes/microvesicles
Exosomal-miRNAs have an important role in exosomes/microvesicles-mediated cellular communication, and the miRNAs expression profiles vary by cell type and conditions.24 Since H/R-Postcon significantly enhanced cardioprotective effects of CFs-exosomes/microvesicles compared to H/R, we sought to investigate the exosomal miRNAs expression profiles of CFs-exosomes/microvesicles between these two groups by using Illumina HiSeq 2500 high-throughput sequencing as described in Methods section. We compared the expression level of miRNAs in H/R-CFs-exosomes/microvesicles and H/R-Postcon-CFs-exosomes/microvesicles. Using a two-fold change and P < 0.01 as the threshold cut-off, we found that 11 miRNAs were differentially expressed. Among the differentially expressed miRNAs, three miRNAs (miR-122-5p, miR-423-3p, and miR-3591) were upregulated in H/R-Postcon-CFs-exosomes/microvesicles (Figure 4A and Supplementary material online, Figure S7). qPCR analysis of miR-122-5p, miR-423-3p, miR-3591, and U6 revealed that miR-423-3p was selectively high expressed in the rat heart, CFs, CFs-exosomes/microvesicles, and H9C2 cells compared to miR-122-5p and miR-3591 (Figure 4B and Supplementary material online, Figure S8), and to identify fibroblast-enriched miRNAs, the expression of those three miRNAs was measured in CFs and H9C2 cells, only miR-423-3p was enriched in CFs in comparison with H9C2 cells (Supplementary material online, Figure S9), so we selected miR-423-3p for further studies. To validate the miRNA-profiles results, levels of miR-423-3p were measured by qPCR in CFs-exosomes/microvesicles. Compared to H/R-CFs-exosomes/microvesicles, H/R-Postcon-CFs-exosomes/microvesicles had significantly increased miR-423-3p level (Figure 4C). H/R-Postcon also significantly increased miR-423-3p level in CFs compared to H/R (Supplementary material online, Figure S10). In vivo, I/R-Postcon significantly reduced the infarct size and reduced CM apoptosis (Supplementary material online, Figure S11, left and middle lane). To explore the relationship between miR-423-3p and the cardioprotective effect of Postcon, we measured the expression of miR-423-3p in the infarct zone, border zone, and remote zone in I/R, I/R-postcon, and sham rat heart. In the infarct zone, MiR-423-3p expression was significantly reduced in I/R rat heart, and I/R-Postcon reversed this effect (Figure 4D). I/R-Postcon also significantly increased the miR-423-3p expression in the infarct border zone (Figure 4E), but this effect was absent in remote zone (Supplementary material online, Figure S12). Parallel to qPCR results, RNA-fluorescence in situ hybridization (RNA-FISH) demonstrated that miR-423-3p expression was significantly downregulated in the infarct zone, while Postcon significantly upregulated the expression of miR-423-3p in the infarct zone (Supplementary material online, Figure S11, right lane). Eighteen hours after H/R-CFs-exosomes/microvesicles and H/R-Postcon-CFs-exosomes/microvesicles were injected to the infarct border zone at three different sites, H/R-Postcon-CFs-exosomes/microvesicles significantly upregulated the expression of miR-423-3p compare to H/R-CFs-exosomes/microvesicles in the infarct zone and infarct border zone (Figure 4F). Similarly, expression of miR-423-3p was selectively increased in H9c2 cells which was co-cultured with H/R-Postcon-CFs-exosomes/microvesicles for 24 h (Figure 4G).To validate a role for miR-423-3p in CFs-exosomal paracrine and cardioprotective effects, CFs were transfected with a miR-423-3p mimics or negative control as described in Methods section. The efficiencies of miR-423-3p overexpression in both CFs and CFs-exosomes/microvesicles were quantified (Supplementary material online, Figure S13 and Figure 4H). Significantly higher cell viability was observed in H9C2 cells incubated with miR-423-3p-mimic-CFs-exosomes/microvesicles than H9C2 cells incubated with mimic-NC-CFs-exosomes/microvesicles and control-CFs-exosomes/microvesicles (Figure 4I). Western blotting confirmed that significantly lower cells apoptosis was observed in H9C2 cells incubated with miR-423-3p-mimic-CFs-exosomes/microvesicles than H9C2 cells incubated with mimic-NC-CFs-exosomes/microvesicles and control-CFs-exosomes/microvesicles (Figure 4J). These results suggested that the downregulating of miR-423-3p is closely related to IRI, and Postcon may exert cardioprotective effects by upregulation of miR-423-3p in CFs-exosomes/microvesicles.
miRNA expression profile of H/R-CFs-exosomes/microvesicles and H/R-Post-conditioning-CFs-exosomes/microvesicles. (A) View of exosomal miRNA profile of H/R-CFs-exosomes/microvesicles and H/R-Postcon-CFs-exosomes/microvesicles by RNA sequencing, a partial heat map of the up-regulated (labelled by redline) and down-regulated miRNAs in H/R-Postcon-CFs-exosomes/microvesicles compared to H/R-CFs-exosomes/microvesicles (H/R, Hypoxia/Reoxygenation; H/R-Postcon, Hypoxia/Reoxygenation-Post-conditioning) exosomes/microvesicles exosomes/microvesicles. (B) Differential expression of miR-423-3p, miR-122-5p, and miR-3591 in rat heart, CFs, CFs-exosomes/microvesicles, H9C2 cells, and rat organs mixture validated by qPCR (n = 6, *P < 0.05 vs. mir-3591, **P < 0.01 vs. mir-3591, ##P < 0.01 vs. mir-122-5p). (C) qPCR validation of miR-423-3p expression in H/R-CFs-exosomes/microvesicles and H/R-Postcon-CFs-exosomes/microvesicles and control-CFs-exosomes/microvesicles (normalized to U6) (n = 6, **P < 0.01 vs. control-CF-exosomes/microvesicles, ##P < 0.01 vs. H/R-CF-exosomes/microvesicles). (D) Expression of miR-423-3p in the infarct zone of I/R, I/R+Postcon, and sham rat heart (n = 8, **P < 0.01 vs. sham, ##P < 0.01 vs. I/R group). (E) Expression of miR-423-3p in the infarct border zone of I/R, I/R+Postcon, and sham rat heart (n = 8, **P < 0.01 vs. sham, ##P < 0.01 vs. I/R group). (F) Expression of miR-423-3p in the infarct zone and infarct border zone of rat heart 18 h after H/R-CFs-exosomes/microvesicles, H/R-Postcon-CFs-exosomes/microvesicles injection, and control (PBS) (n = 8, **P < 0.01 vs. I/R+PBS, ##P < 0.01 vs. I/R + H/R-CFs-exosomes/microvesicles. H/R-exo, H/R-CFs-exosomes/microvesicles; H/R-p-exo, H/R-Postcon-CFs-exosomes/microvesicles). (G) Expression of mir-423-3p in H9C2 cells which was co-cultured with H/R and H/R-Postcon-CFs-exosomes/microvesicles for 24 h (normalized to U6) (n = 6, **P < 0.01 vs. H9C2 control, ##P < 0.05 vs. H9C2+H/R-exo; H/R-exo, H/R-CFs-exosomes/microvesicles; H/R-p-exo, H/R-Postcon-CFs-exosomes/microvesicles). (H) The efficiencies of miR-423-3p overexpression in CFs-exosomes/microvesicles was quantified (n = 6, **P < 0.01 vs. mimic-NC-CFs-exo). (I) CCK-8 assay measuring viability of H9C2 cells pre-incubated with control-CFs-exosomes/microvesicles, mimic-NC-CFs-exosomes/microvesicles and miR-423-3p-mimic-CF-exosomes/microvesicles, followed by H/R. Results are expressed as % of control n = 6, **P < 0.01 vs. mimic-NC-CFs-exo, exo, exosomes/microvesicles). (J) Western blotting analysis of the protein levels of caspase-3 in H9c2 cells pre-incubated with control-CFs-exosomes/microvesicles, mimic-NC-CFs-exosomes/microvesicles, and miR-423-3p-mimic-CF-exosomes/microvesicles, followed by H/R. Results are expressed as % of control n = 6, **P < 0.01 vs. mimic-NC-CFs-exo; exo, exosomes/microvesicles).
3.5 Knockdown of miR-423-3p expression attenuates I/R-Postcon-induced cardioprotection against I/R in vivo
To test the essential role of miR-423-3p in I/R-Postcon-induced cardioprotection against I/R in vivo, miR-423-3p expression was inhibited by its antagomiR. Rats were intra-myocardially injected with 80 μg miR-423-3p antagomiR or scramble control prior to I/R, and 24 h after injection, subjected to I/R or I/R-Postcon. The efficiency of knockdown of mir-423-3p in vivo was determined by qPCR (Supplementary material online, Figure S14).
The myocardial AAR and infarct size were determined by TTC and Evans blue staining. Representative TTC and Evans blue-stained heart slices from the I/R, I/R-Postcon, I/R-Postcon+scramble, and I/R-Postcon+antagomiR-423-3p groups are shown (Figure 5A). There was no difference in AAR among the I/R, I/R-Postcon, I/R-Postcon+scramble, and I/R-Postcon+antagomiR-423-3p groups (Supplementary material online, Figure S15). Compared with the I/R group, the infarct size was significantly attenuated in the I/R-Postcon group, compared with the I/R-Postcon+scramble group, the I/R-Postcon+antagomiR-423-3p group presented a significantly exacerbated myocardial infarct size, and the I/R-Postcon+scramble group presented no effect on IPC-induced cardioprotection (Figure 5B).
Knockdown of miR-423-3p expression attenuates I/R-Postcon-induced cardioprotection against I/R in vivo. (A) Representative TTC and Evans blue-stained heart slices from the I/R, I/R-Postcon, I/R-Postcon+scramble, and I/R-Postcon+antagomiR-423-3p groups. (B) The quantitative analysis of I/R-Postcon and miR-499 inhibition on myocardial infarct size (n = 8, **P < 0.01 vs. I/R-Postcon group, ##P < 0.01 vs. I/R-Postcon+scramble). (C) Representative images of CM apoptosis in heart sections from the I/R, I/R-Postcon, I/R-Postcon+scramble, and I/R-Postcon+antagomiR-423-3p groups. (D) The quantitative analysis of I/R-Postcon and miR-499 inhibition on CM apoptosis in vivo (n = 8, **P < 0.01 vs. I/R-Postcon group, ##P < 0.01 vs. I/R-Postcon+scramble).
To further confirm the mechanism of miR-423-3p in I/R-Postcon-induced cardioprotection against I/R in vivo, apoptosis was determined in rat heart sections by TUNEL staining. Representative images of CM apoptosis in heart sections from the I/R, I/R-Postcon, I/R-Postcon+scramble, and I/R-Postcon+antagomiR-423-3p groups are shown (Figure 5C). CM apoptosis markedly decreased in the I/R-Postcon group as compared with that observed in the I/R group, the I/R-Postcon+antagomiR-423-3p group showed a significant increase in CM apoptosis as compared with that observed in the I/R-Postcon+ scramble group, and the I/R-Postcon+scramble group presented no effect on IPC-induced cardioprotection. These results suggested that miR-423-3p plays an essential mediate role in I/R-Postcon-induced cardioprotection against I/R in vivo.
3.6 MiR-423-3p protected H9C2 cells from hypoxia–reoxygenation by targeting Rap2c
The expression of miR-423-3p was examined in H9C2 cells after H/R. The results showed that miR-423-3p was significantly downregulated in H9C2 cells under H/R (Figure A), suggesting that a possible connection exists between the reduction of miR-423-3p and the apoptosis of H9C2 cells under H/R. To identify a functional cardioprotective role of miR-423-3p, gain and loss of function approaches were used in the H9C2 cells. MiR-423-3p mimics or inhibitors obviously increased or decreased miR-423-3p expression in H9C2 cells (Supplementary material online, Figure S16). CCK-8 result showed that the miR-423-3p mimics obviously increased cell viability, whereas the inhibitors decreased cell viability in H9C2 cell under H/R (Figure 6B). To detect the effects of miR-423-3p on the H/R-induced apoptosis of H9C2 cells, the levels of pre-caspase-3 and cleaved-caspase-3 were detected by western blotting. The results showed that miR-423-3p mimics obviously reduced H9C2 cells apoptosis, whereas the inhibitors increased H9C2 cells apoptosis under H/R (Figure 6C). These results confirmed the cardioprotective function of miR-423-3p and suggested that rescue of the down-regulated miR-423-3p in the CMs might be a potential strategy to protect CMs against IRI.
MiR-423-3p protected H9C2 cells from hypoxia–reoxygenation by targeting Rap2c. (A) Expression of miR-423-3p in H/R treated H9C2 and control (n = 6, **P < 0.01 vs. control). (B) CCK-8 assay measuring H9C2 cells viability after H/R. Cells were transfected with miR-423-3p mimics, inhibitors and both negative controls. Results are expressed in % of control (n = 6, *P < 0.05 vs. H/R group, **P < 0.01 vs. H/R group). (C) Western blotting analysis of the protein levels of caspase-3 in H9C2 cells after H/R. Cells were transfected with miR-423-3p mimics, inhibitors, and both negative controls (n = 6, *P < 0.05 vs. control, **P < 0.01 vs. control). (D) Western blotting analysis of RAP2C protein levels in H9C2 cells transfected with miR-423-3p mimics, inhibitors, and both negative controls (n = 6, **P < 0.01 vs. control). (E) Luciferase activity assay of HEK293T cells transfected with pmiR-RB-REPORT™ plasmid containing WT-3′UTR and Mut-3′UTR of RAP2C (n = 6, **P < 0.01 vs. rno-RAP2C-WT+NC). (F) mRNA and (G) protein expression of RAP2C in H/R treated H9C2 and control (n = 6, **P < 0.01 vs. control). (H) CCK-8 assay measuring H9C2 cells viability after H/R. Cells were transfected with RAP2C siRNA and negative controls. Results are expressed in % of control (n = 6, **P < 0.01 vs. control, ##P < 0.01 vs. H/R+siRNA-NC group). (I) Western blotting analysis of the protein levels of caspase-3 in H9c2 cells after H/R. Cells were transfected with RAP2C siRNA and negative controls (n = 6, **P < 0.01 vs. control, ##P < 0.01 vs. siRNA-NC group).
Schematic of our present hypothesis. CFs participate in cardioprotection via exosomes/microvesicles pathway during the acute phase of ischaemia–reperfusion and Postcon can enhance this cardioprotective effect by elevating the expression of CFs exosomal miR-423-3p, which targeting downstream effecter RAP2C.
Target genes of miR-423-3p were predicted using two computational predictions tools (TargetScan & miRDB), and only target genes with conserved sites were predicted. After matching the putative targets from both TargetScan and miRDB, we found that only Ras-related protein Rap-2c (RAP2C) and BCL6 corepressor-like 1 (Bcorl) were predicted by both methods (Supplementary material online, Figure S17A). miRwalk 2.0 is a prediction tool that automatically compares target genes from 11 different prediction tools, we have used miRwalk 2.0 to predict the target genes of miR-423-3p, we found that RAP2C is also predicted by miRwalk 2.0 (Supplementary material online, Figure S18). In a previous study, Ras inhibitor S-farnesylthiosalicylic acid (FTS) was shown to be effective in reducing the damage caused to the heart in the I/R model.25 We therefore chose RAP2C as the alternative target genes of miR-423-3p. The complementary base pairing is located at 26 to 33 bp of rat RAP2C 3′ UTR (Supplementary material online, Figure S17B). Interestingly, the seed sequence of RAP2C 3′UTR targeting to miR-423-3p is highly conserved among the species of mammals (Supplementary material online, Figure S17C), suggesting a critical role in their physiology. To confirm whether RAP2C is a target of miR-423-3p in H9C2 cells, gain and loss of function assays showed that miR-423-3p inhibitors increased, whereas miR-423-3p mimics decreased RAP2C protein levels in H9C2 cells (Figure 6D). These results indicated that miR-423-3p possibly attenuated cell apoptosis by inhibiting the expression of RAP2C. To further address whether miR-423-3p directly binds the 3′ UTR region of RAP2C, we generated some chimeric constructs which harbour luciferase wild-type 3′ UTR sequence (WT-3′ UTR) or mutant 3′ UTR sequence (Mut-3′ UTR) (Supplementary material online, Figure S17D, E). As expected, the miR-423-3p mimics exclusively inhibited the luciferase activity of WT-3′ UTR, suggesting that the putative binding site is important for miR-423-3p suppression of RAP2C expression (Figure 6E).
To further determine the relationship between RAP2C and apoptosis in H9C2 cells, the expression of RAP2C in H9C2 cells after H/R was measured by qPCR and western blotting. These results showed that RAP2C was significantly upregulated in H9C2 cells after H/R treatment (Figure 6F, G). Then we used siRNA to knockdown RAP2C expression in H9C2 cells. The knockdown efficiency of RAP2C-siRNA was also detected by qPCR and western blotting (Supplementary material online, Figure S19). We found that knockdown of RAP2C obviously increased cell viability (Figure 6H) and reduced apoptosis in H9C2 cells (Figure 6I). Taken together, above results demonstrated that miR-423-3p protected H9C2 cells from H/R by targeting RAP2C.
4. Discussion
In patients with STEMI, PCI is the recommended therapy. Although restoration of myocardial blood flow is crucial, PCI therapy itself may jeopardize the myocardium. This phenomenon, known as reperfusion injury, may account for as much as 50% of the final myocardial infarct size, a major determinant of the prognosis in patients with STEMI.2 Ischaemic Postcon is a safe, easy to perform and feasible protocol. In several animal models, Postcon resulted in a reduction in infarct size.4 In contrast to those marked effects in animal models, studies of the effect of Postcon in humans have shown conflicting results. The DANAMI-3-iPOST trial is, to our knowledge, the largest clinic study to assess the effect of ischaemic Postcon; however, the result is neutral.8 Therefore, it is important to investigate the underlying mechanism why ischaemic Postcon is effective in animal models.
Most research about Postcon only studied CMs alone. However, the heart is a whole composed of a variety of cells including fibroblasts, endothelium, macrophages, and pericytes. To our knowledge, there is no report about the effect of Postcon on other cells. CFs are one of the single largest population of cells within the heart and possess an important paracrine function that may play a potential role in inducing a condition state to surrounding myocardial cells. A recent study demonstrated that CFs participate in cardioprotection during the acute phase of ischaemia–reperfusion via a paracrine pathway involving TIMP-1.20 In our study, CFs protected H9C2 cells from H/R insult in a paracrine manner, which is consistent with the previous study. However, our result indicated that Postcon can amplify this protection. Previous studies have shown that CFs can release a range of growth factors that contribute to cardioprotection.9,26 Recent studies paid more attention to the importance of exosome in cell-to-cell communication within the cardiovascular system. CFs were shown to secrete exosomes enriched with miR-21* that mediated cardiac hypertrophy.17 In our study, after inhibition of exosome production by GW4869, the protective effect of CFs was largely abolished, and inhibition of CFs-exosomes/microvesicles secretion led to a significant abrogation on the amplified protective effect of H/R-Postcon. These results confirm that secreted exosomes/microvesicles are essential for the cardioprotective effect of CFs.
To further confirm the role of CF-derived exosome-mediated paracrine cardioprotection, we isolated exosomes/microvesicles from CFs and identified them by TEM, NAT, and WB. Li et al. found that hypoxia induced the release of exosomes from oral squamous cell carcinoma cells.27 King et al. demonstrated that hypoxia promoted the release of exosomes from breast cancer cells.28 As for CFs, Lyu et al. showed that Ang II stimulates release of exosomes from CFs.29 Currently, there is no report about the effect of H/R on exosome secretion of CFs. In our study, H/R significantly stimulated release of exosomes/microvesicles from CFs. A recent study reported that exosomes/microvesicles are essential for the cardioprotective effect of remote conditioning19 and that remote conditioning can enhance the exosomes secretion in plasma.30 In the present study, our results revealed that compared to H/R, Postcon further enhances the exosome secretion by CFs. Our results further demonstrated that although H/R seemed to enhance the cardioprotection of CFs-exosomes/microvesicles, Postcon can further enhance the cardioprotection of CFs-exosomes/microvesicles compared to H/R alone.
Exosomes/microvesicles may influence recipient cells via three mechanisms16: firstly, exosomes/microvesicles could be internalized into an endocytic compartment or MVB, releasing their cargo into the cytoplasm of the recipient cell31; secondly, the exosome membrane may fuse with the plasma membrane, releasing the cargo directly into the cytoplasm of the recipient cell32; and thirdly, exosomes/microvesicles act as a ligand to mediate a ligand–receptor interactions.30 In our study, CFs-exosomes/microvesicles were uptaken by H9C2 cells within 12 h, an indication that CFs-exosomes/microvesicles might influence H9C2 cells via the first mechanism mentioned above. In vivo, 24 h after intra-myocardially injection of PKH26-labelled exosomes/microvesicles, these exosomes/microvesicles were uptaken by CMs, further confirmed this result. A widely investigated group of exosome cargo is miRNA, which has gained a lot of attention since the first time it was discovered in exosomes/microvesicles. Exosomes/microvesicles and exosomal miRNAs have been found to play an important role in cardiac repair after myocardial infarction.33–35 Logically, because the CMs in the infarct area are unlikely to be exposed to the normoxic CFs-exosomes/microvesicles and Postcon can significantly enhance the cardioprotection of CFs-exosomes/microvesicles compared to H/R, we compared miRNAs profiles between H/R-CFs-exosomes/microvesicles and H/R-Postcon-CFs-exosomes/microvesicles by RNA sequencing. Only three miRNAs (mir-122-5p, mir-423-3p, and mir-3591) were upregulated in H/R-Postcon-CFs-exosomes/microvesicles. Among these three miRNAs, we found that miR-423-3p had selectively high expression in rat heart, CFs, CFs-exosomes/microvesicles, and H9C2 cells. Fibroblast-derived miRNAs can be transported to CMs via exosomes/microvesicles.17 In the present study, when the H9C2 cells were pre-treated with CFs-exosomes/microvesicles, after the exosomes/microvesicles were taken up by H9C2 cells. Expression of miR-423-3p was selectively increased in H9C2 cells that were co-cultured with H/R-Postcon-CFs-exosomes/microvesicles, indicating that the miR-423-3p was transferred into the H9C2 cells by exosomes/microvesicles.
We choose neonatal rat CFs based on previous studies, as for adult CFs, we cultured mouse adult CFs and treated it with H/R and H/R-Postcon. We found that compared to H/R-mouse-CFs-exosomes, H/R-Postcon-mouse-CFs-exosomes had significantly increased miR-423-3p level and H/R-Postcon also significantly increased miR-423-3p level in mouse-CFs compared to H/R (Supplementary material online, Figure S20).
By co-culture H9C2 cells with CFs-exosomes/microvesicles enriching with miR-423-3p, we demonstrated that H/R-Postcon exerted cardioprotective effects by upregulation of miR-423-3p in CFs-exosomes/microvesicles. qPCR and FISH results demonstrated that miR-423-3p is closely related to IRI. By inhibited miR-423-3p expression with its antagomir in vivo, we demonstrated that miR-423-3p plays an essential mediate role in I/R-Postcon-induced cardioprotection against I/R in vivo, we demonstrated that miR-423-3p is closely related to IRI, and Postcon may exert cardioprotection by upregulation of miR-423-3p in CFs-exosomes/microvesicles.
We also found that miR-30b-5p, miR-30c-5p, and miR-374a-5p were downregulated in H/R-Postcon-CFs-exosomes/microvesicles. To the best of our knowledge, miR-30b-5p and miR-30c-5p are pro-angiogenic, Overexpression of miR-30 family in endothelial cells led to increased vessel number and length in an in vitro model of sprouting angiogenesis.36 miR-374a-5p is anti-angiogenic.37 We think it might be that these miRNAs offset each other’s role, but it deserves further study.
There is no report about the relationship between miR-423-3p and cardiovascular disease. Previous studies have shown that miR-423-3p promotes cell growth in hepatocellular carcinoma and colorectal cancer.38,39 Interestingly, we found miR-423-3p was downregulated in H9C2 cells under H/R, indicating that miR-423-3p may be a key regulator of CMs function under H/R. In the gain and loss of function experiments, we found that up-regulated miR-423-3p levels effectively protected H9C2 cells against H/R and down-regulated miR-423-3p levels increased the injury, demonstrating that miR-423-3p played an important role in cardioprotection against H/R. miRNAs mediate cell function by inhibiting the transcription of downstream target genes; one miRNA can target hundreds of mRNA simultaneously. In our study, we used TargetScan and miRDB to predict target genes candidates of miR-423-3p. Only RAP2C and Bcorl-1 were predicted by both tools. As inhibition of Ras could reduce the IRI in an animal model,25 we picked RAP2C as the candidate target gene. We demonstrated that miR-423-3p regulates the expression of RAP2C in H9C2 cells, and we identified RAP2C as a direct target of miR-423-3p using luciferase reporter assay.
The Ras family regulates a wide variety of cellular functions including cell growth, differentiation, and apoptosis.40 The Rap proteins define a family of monomeric low molecular weight guanosinetriphosphate (GTP)-binding proteins that share 50–60% sequence homology with the product of the ras proto-oncogene. Five different members of this family have been identified: RAP1A, RAP1B, RAP2A, RAP2B, and RAP2C.41 The Rap2 proteins have been much less investigated for a long time, probably because they did not share the Ras-antagonizing effects as with Rap1 proteins. It has been recently evidenced that Rap2 proteins may regulate integrin function and cell adhesion similar to Rap1 protein recently.42
RAP2C is a new member of the Rap2 subfamily. RAP2C shows peculiarity in the nucleotide binding properties compared to the highly homologous RAP2B.41 In particular, it is characterized by a reduced binding of GTP and a reduced rate of Guanosine diphosphate (GDP) release.43 RAP2C represents the predominant Rap2 isoform expressed in circulating mononuclear cells. The role of RAP2C in cardiovascular disease is currently poorly understood. In the present study, RAP2C was upregulated in H9C2 cells after H/R. Knockdown of RAP2C with siRNA increased cell viability and reduced apoptosis in H9C2 cells after H/R. Our study demonstrated for the first time that inhibition of RAP2C may represent a novel therapeutic target of IRI.
In conclusion (Figure 6), we have described that CFs can reduce IRI through paracrine secreting exosomes/microvesicles, and Postcon can enhance cardioprotection of CF-derived exosomes/microvesicles. The underlying mechanism might involve elevated expression and cellular delivery of exosomal miR-423-3p by targeting downstream effecter RAP2C. Our findings reveal that fibroblast-derived miR-423-3p acts as a crucial paracrine mediator of CFs cardioprotection and illustrate a novel role RAP2C as a therapeutic target of IRI.
To the best of our knowledge, most of current studies focused on mesenchymal stem cells (MSC)-exosomes and cardiac regeneration, we have not found any study about neonatal fibroblast exosomes and cardiac regeneration. We believe that neonatal fibroblast exosomes may have a positive effect on cardiac regeneration, and we will conduct further research on this aspect.
Conflict of interest: none declared.
Funding
This work was supported by the National Natural Science Foundation of China (81570050).
Footnotes
Time for primary review: 27 days
