MiR-133a-3p transferred by circulating microvesicles derived from myocardial ischemic preconditioning protects cardiomyocytes against hypoxia/reoxygenation injury


 Background Microvesicles (MVs) are submicron membrane vesicles as mediators of intercellular communication. The aim of our study was to investigate protective mechanism of circulating MVs derived from ischemic preconditioning (IPC-MVs) on myocardial I/R injury. Results Administration of IPC-MVs reduced infarct size and activity of lactate dehydrogenase (LDH) in myocardial I/R injury in vivo. Meanwhile, IPC-MVs could increase cell viability and reduce LDH activity in hypoxia/reoxygenation (H/R) injured H9c2 cells in vitro. Microarray analysis demonstrated that miR-133a-3p expression in IPC-MVs increased apparently compared with Sham-MVs. We found that miR-133a-3p increased cell viability, decreased LDH activity and apoptosis , as well as suppressed H/R-induced endoplasmic reticulum stress (ERS). MVs induced by hypoxic preconditioning enriched with FAM-miR-133a-3p allowed the transfer of miR-133a-3p to target cells. In addition, miR-133a-3p was significantly increased in H/R injured H9c2 cells by treatment with IPC-MVs. Epidermal growth factor receptor (EGFR) is a target gene of miR-133a-3p. AG1478 (EGFR inhibitor) significantly increased cell viability, decreased LDH activity and ERS-induced apoptosis in H9c2 cells under H/R injury. Conclusions The findings of this study showed that IPC-MVs exerted cardioprotective effects by transferring miR-133a-3p into H/R injured cardiomyocytes targeting EGFR, thus attenuating ERS-induced apoptosis. MiR-133a-3p transferred by IPC-MVs may provide a novel therapy for myocardial I/R injury.


Protective effects of miR-133a-3p on H9c2 cells against H/R injury
In order to determine the effect of miR-133a-3p on cardiomyocyte apoptosis induced by H/R injury, miR-133a-3p mimics or inhibitor was transfected into H9c2 cells to increase or decrease miR-133a-3p expression. As expected, after transfecting H9c2 cells with miR-133a-3p mimics, there was a significant increase in cell viability, reduction in LDH activity after H/R, which was the same as treatment with IPC-MVs. However, after transfecting H9c2 cells with miR-133a-3p inhibitor, there was a further decrease in cell viability, further increase in LDH activity after H/R ( Fig. 5a-b). Meanwhile, detection of caspase 3 activity, and Western blot analysis of caspase 3 and cleaved-caspase 3, was adopted to analyze the protective effect of miR-133a-3p upon apoptosis induced by H/R. Not surprisingly, miR-133a-3p mimics substantially reduced the activity of caspase 3, and expression of caspase 3 and cleaved-caspase 3 in H9c2 cells, after H/R. However, miR-133a-3p inhibitor further substantially increased the activity of caspase 3, and expression of caspase 3 and cleaved-caspase 3 in H9c2 cells, after H/R (Fig. 5c-d).
Additionally, activation of ERS markers had been detected after H/R injury in H9c2 cells, including GRP78, CHOP and caspase 12. As envisaged, consistent with treatment of IPC-MVs, the transfection with miR-133a-3p mimics in H9c2 cells significantly reduced all above proteins compared with H/R group, whereas the expression of these proteins increased significantly after transfection with miR-133a-3p inhibitor in H9c2 cells, which indicated that the suppression of ERS-induced apoptosis may be involved in the protective effect of miR-133a-3p on H/R injured H9c2 cells (Fig. 5e).

HPC-MVs allow the transfer of miR-133a-3p to H9c2 cells
Recent studies suggest that MVs transfer cardioprotective miRNAs between cardiomyocytes [13]. To conform whether MVs could transfer miR-133a-3p to H9c2 cells in vitro, we fluorescently labelled miR-133a-3p mimics and transfected them into H9c2 cells.
The cells were labelled with the strong red fluorescence dye DiI after 24 h, and then the secreted MVs were collected after treatment with hypoxic preconditioning (HPC). After coincubating H9c2 cells with fluorescently labeled HPC-MVs for 4 h, cells exhibited red and green fluorescence in the cytoplasm (Fig. 6a), indicating that HPC-MVs derived from H9c2 cells allow the transfer of miR-133a-3p to H9c2 cells. In contrast, the negative control (NC) group could not detect any fluorescence signal. We further determined the expression level of miR-133a-3p in H9c2 cells after treated with H/R, Sham-MVs + H/R, and IPC-MVs + H/R. As envisaged, qRT-PCR results showed that the expression level of miR-133a-3p increased significantly when H/R injured H9c2 cells were co-cultured with IPC-MVs, but not Sham-MVs (Fig. 6b). Taken together, these results indicated that the protective effect of IPC-MVs on cardiomyocytes against H/R injury may be associated with the elevation of miR-133a-3p in vitro mainly transferred by IPC-MVs.

MiR-133a-3p inhibits EGFR expression in H9c2 cells
To better understand how miR-133a-3p leads to cardioprotection, we honed in on the target genes for miR-133a-3p within IPC-MV-treated H/R injury in H9c2 cells. Among the known targets, EGFR, one of receptor tyrosine kinases (RTKs) on the cell surface, is a direct target gene in Hela cells [16] and other cancer cells[23] confirmed by Luciferase assay, qRT-PCR, and Western blot assay. According to display in KEGG database (https://www.kegg.jp/) and published literatures, we focused on the hypoxia-regulated EGFR. qRT-PCR results showed that miR-133a-3p mimics could apparent reduce, while inhibitor increase the expression of EGFR mRNA compared with NC ( Fig. 7a-b). Meanwhile, Western blot results showed that miR-133a-3p mimics apparent reduce, while inhibitor increase the expression of EGFR protein compared with NC ( Fig. 7c-d), which suggested that miR-133a-3p regulates the expression of EGFR in H9c2 cells.

MiR-133a-3p protects H9c2 cells against H/R injury through suppressing ERS
To further investigate the underlying mechanism of the enhanced ability of miR-133a-3p in myocardial protection, AG1478, an inhibitor of EGFR, was used to pre-treat H9c2 cells before H/R. MTT results showed that AG1478 could dose-dependently increase the cell viability, and the trend became significant in 1 μM AG1478 + H/R group (Fig. 8a). We further detected LDH activity and apoptosis in H9c2 cells after treating with AG1478 (1 μM). There was no doubt that the activity of LDH and caspase 3, and protein levels of caspase 3 and cleaved-caspase 3, reduced significantly compared with H/R group ( Fig. 8bd). Furthermore, related proteins of ERS in cells were detected including GRP78, CHOP and caspase 12. Interestingly, these proteins reduced substantially when pre-treated with AG1478 (1 μM) compared with H/R group (Fig. 8e). Therefore, miR-133a-3p could be a promising therapeutic target by suppressing ERS-induced apoptosis for interventions of the protective effect of IPC-MVs upon H/R injury.

Discussion
Given the wonderful endogenous mechanism of myocardial protection, IPC has been proved to protect from myocardial I/R injury through various pathways [24]. However, the underlying molecular mechanism remains poorly understood. MVs are cystic vesicles secreted from cells under stimulation or apoptosis, which can be involved in the pathogenesis of various diseases. Recent studies have demonstrated that circulating MVs are responsible for the transmission of IPC signals that elicit cardioprotection [5,6]. In the present study, IPC-MVs and Sham-MVs in circulation were successfully obtained, and the total amount of IPC-MVs was slightly higher than that of Sham-MVs. However, there were no significant differences in morphology, size distribution or total amount between these two groups by TEM and flow cytometry. Furthermore, our results confirmed that IPC-MVs could protect myocardium against I/R injury through reducing infarct size and LDH activity in vivo, and protect H9c2 cells against H/R injury through improving the cell viability, decreasing the LDH activity in vitro compared with Sham-MVs.
As a kind of paracrine factor, MVs allow the transfer of cytoplasmic components such as proteins, mRNAs and miRNAs from one cell to another [8]. However, circulating MVs vary in their components, which varies with the different parent cell types or different stimuli, thus playing different biological roles in occurrence and development of diseases [9,12].
Recent studies have revealed that functional miRNAs can be transferred by MVs and appear to specially home to the site of injury within infarcted myocardium, and MVs were considered as major vector of circulating miRNAs [10]. Moreover, miRNAs have emerged as key regulators of cardiovascular physiology and diseases, including myocardial ischemic injury caused by IHD [4]. To dissect the underlying mechanisms about cardioprotective effect of IPC-MVs, microarray was performed to determine components and alterations in miRNA profiles in IPC-MVs and Sham-MVs. As expected, miRNAs could be detected in both IPC-MVs and Sham-MVs, and the types and contents of miRNAs are different. Meanwhile, we found that four miRNAs (miR-1-3p, miR-133a-3p, miR-133b-3p and miR-378b) were upregulated, and one miRNA (miR-702-3p) was downregulated, with 2-fold changes in the IPC-MVs compared with Sham-MVs. In addition, the five miRNAs mentioned above have been reported to be involved in different biological processes in cardiovascular diseases.
MiR-1, a muscle-specific miRNA, has been extensively studied and proven to be . In order to verify the predictive accuracy of microarray assay, the expression of these miRNAs in IPC-MVs and Sham-MVs was detected by qRT-PCR. The results of miR-1-3p, miR-133a-3p, miR-133b-3p were consistent with the expression patterns in microarray. However, miR-378b and miR-702-3p were opposite, miR-378b was downregulated in the IPC-MVs, while miR-702-3p was upregulated compared with Sham-MVs. As a high-throughput technology, microarray is used to interpret and analyze the sequence information of samples efficiently by virtue of the hybridization and pairing characteristics of nucleic acid molecules, which is tremendously affected by background signals, and may lead to inconsistent results. Based on results of GO analysis and KEGG assay (data not shown), and multiple studies, the five miRNAs were involved in a variety of cardiovascular pathways, including aldosterone-regulated sodium reabsorption, leukocyte transendothelial migration, vasopressin-regulated water reabsorption, sphingolipid signaling pathway, phosphatidylinositol signaling system and HIF-1 signaling pathway. However, microarray and qRT-PCR assay all demonstrated that miR-133a-3p expression in the IPC-MVs increased most apparently in the detected miRNAs compared with Sham-MVs.
Increasing evidences have suggested that circulating miRNAs can be potential biomarker candidates due to their highly specific elevation in blood upon stress, including I/R injury.
MiR-133a-3p plays key roles in regulating the differentiation, proliferation and maturation of cardiomyocytes. In patients with myocardial infarction (MI), the expression of miR-133a- , and myocardial ischemic injury [33]. However, elevation of miR-133a-3p reduces hypoxia-induced, oxidative stress-induced and ERS-induced cardiomyocyte apoptosis in vitro [22]. In the present study, we further demonstrated that the decreasing of miR-133a-3p in H9c2 cells was closely related to H/R injury. When treated with IPC-MVs, miR-133a-3p expression was significantly upregulated compared with Sham-MVs.
Therefore, we speculated that miR-133a-3p in IPC-MVs may mediate the protective effect of IPC-MVs on H/R-injured H9c2 cells. Consistent with the results of treatment with IPC-MVs, the ameliorative effect of miR-133a-3p on H9c2 cells upon H/R injury was also proved by the cell viability and the LDH activity detection according to the gain-and loss-offunction approaches.
Cardiomyocyte apoptosis plays central role in the progression of many myocardial disorders, including I/R injury. Here H/R-induced apoptosis was used to determine the ameliorative effect of miR-133a-3p on apoptosis of H9c2 cells. Consistent with the results of treatment with IPC-MVs, the ameliorative effect of miR-133a-3p on cardiomyocyte apoptosis was also proved in H/R injury detected by the expression of caspase 3 and cleaved-caspase 3, and caspase 3 activation in H9c2 cells. Our results could demonstrate the beneficial effect of miR-133a-3p on H/R-induced apoptosis.
IHD can ultimately lead to myocardial tissue death, mainly caused by cardiomyocyte apoptosis induced by tissue ischemia and hypoxia, and the trigger of myocardial injury is closely related to ERS [34,35]. Our previous study has found that the suppression of ERSmediated apoptosis could be involved in the protective mechanism of IPC-MVs against myocardial I/R injury, that is, the addition of IPC-MVs could counteract the I/R-induced increase of ERS-specific proteins GRP78, CHOP and caspase 12 levels [6]. Meanwhile, the importance of miR-133a-3p in cardioprotection was further supported by the finding that transfection with the miR-133a-3p mimics could suppress the ERS-induced apoptosis through reducing the expression of GRP78, CHOP and caspase 12. Instead, transfection with the miR-133a-3p inhibitor could aggravate the ERS-induced apoptosis through increasing the expression of GRP78, CHOP and caspase 12. Hence, the suppression of ERSinduced apoptosis may be involved in the protective effect of miR-133a-3p on H/R injured H9c2 cells, consistent with IPC-MVs.
Recent studies have suggested that MVs can transfer functional miRNAs between cells by various routes, including endocytotic uptaking [8]. Additionally, cardiomyocytes can release and take up MVs both in vivo and in vitro [13]. Our study further found that HPC-MVs isolated from H9c2 cells which subjected to HPC could transfer miR-133a-3p into target cells, which was verified by HPC-MVs labeled with DiI and transfected with FAM-miR-133a-3p, incubating with H9c2 cells for 4 h. Not only were H9c2 cells able to take up the HPC-MVs, but the miRNA content of the HPC-MVs was also able to enter the cytoplasm. miRNAs existing in all kinds of cells can regulate different gene expression, mainly through downregulating the expression of target genes by either promoting mRNA degradation or repressing mRNA translation. To elucidate the mechanism underlying the protective role of miR-133a-3p, we searched for the target genes by bioinformatic technology and literature search. Previous studies have shown that miR-133a-3p is mainly involved in the regulation of pro-apoptotic genes in cardiomyocytes, such as TAGLN2, DAPK2, APAF1, BCL2L11 and BMF, and so on, thus affecting the apoptosis pathways.
Among known target genes, it is well known that EGFR, a hypoxia-associated apoptotic protein involving angiogenesis, is a direct target gene in cardiomyocytes [16] and many cancer cells confirmed by Luciferase assay, qRT-PCR, and Western blot. Our study further verified that EGFR is the target gene of miR-133a-3p in H9c2 cells by qRT-PCR and Western blot. Therefore, we speculated that miR-133a-3p exert a protective effect against H/R-induced apoptosis, and may be related to the inhibition of EGFR.
To further explore whether EGFR is involved in the protective mechanism of miR-133a-3p against cardiomyocyte apoptosis, H9c2 cells were pre-treated with AG1478 (an EGFR inhibitor) before H/R. We found that after treatment with AG1478 (1 µM), the cell viability improved significantly, the LDH and caspase 3 activity decreased significantly, the expression of caspase 3 and cleaved-caspase 3 was downregulated significantly, and the downregulated expression of ERS-related proteins GRP78, CHOP and caspase 12.
Therefore, the suppression of ERS-induced apoptosis might be involved in the protective mechanism of IPC-MVs on H/R injury in H9c2 cells, mediated by miR-133a-3p as evidenced by downregulating the expression of EGFR.

Conclusions
Circulating MVs after the process of IPC could convey survival signals to injured cardiomyocytes that protect cells against ERS-induced cardiomyocyte apoptosis during H/R injury in vitro. The mechanism may be related to miR-133a-3p-mediated cytoprotection targeting EGFR. Furthermore, it is also suggesting that miR-133a-3p in the IPC-MVs appears as a possible novel effective strategy for the protective effect of IPC-MVs on IHD. Furthermore, the discovery of circulating MVs has shed new light on the search for more efficient miRNAs transport tools.

Establishment of the I/R and IPC model in vivo
The myocardial I/R model was established by occluding the left anterior descending (LAD) coronary artery in rats for 30 min followed by 120 min reperfusion and IPC was induced by three cycles of 5 min ischemia and 5 min reperfusion of the LAD as previously described [5]. Successful occlusions were verified by observing the development of STsegment elevation on electrocardiogram (ECG). Sham-operated rats were received as controls.

Isolation of circulating MVs
Circulating IPC-MVs and Sham-MVs were isolated from the blood of rats subjected to IPC or Sham treatment as reported previously [6]. Briefly, healthy male Wistar rats were divided into two groups randomly with n=5 each: (1) Sham-MV group: rats were left untreated for 45 min after a silk ligature was placed around the LAD coronary artery, the blood was drawn from abdominal aorta after 45 min. (2) IPC-MV group: rats were subjected to three cycles of 5 min ischemia and 5 min reperfusion of LAD after being untreated for 15 min, then the blood was drawn from abdominal aorta at once. The blood samples in two groups were collected in sodium citrate coated tubes and centrifugated at 2,600 g, 15 min and 10,000 g, 5 min at room temperature to obtain platelet-free plasma (PFP). Ninety μL PFP was collected and stored at -80 ℃ after fixed with paraformaldehyde (PFA) to a final concentration of 1 % for 1 h at room temperature for flow cytometry, the remaining PFP was ultracentrifuged at 100,000 g, 4 ℃ for 150 min, the supernatant was removed to obtain Sham-MVs and IPC-MVs. The pellet of MVs was resuspended in 100 μL 0.9% sodium chloride and stored at -80 °C. Rats were sacrificed by acute arterial hemorrhage.

Transmission electron microscopy (TEM)
TEM of circulating IPC-MVs and Sham-MVs was conducted as described previously [6].
Briefly, 40 μL MV resuspension was dropped on the carbon-coated copper grids, then grids were blotted dry with filter paper. For negative staining, 40 μL 2% phosphotungstic acid (pH 6.5) was used to stain for 2 min. After drying under the incandescent light, morphology of MVs was viewed in a HT7700 TEM (Hitachi, Japan), and the images were obtained using a digital camera (Olympus, Japan).

Flow cytometry analysis
Flow cytometry analysis was performed as described previously [6]. Briefly, 1 μm and 2 μm For the H/R injury model, H9c2 cells were stimulated by H/R as previously described [17].
Briefly, H9c2 cells were subjected to hypoxic buffer (in mM: 0.9 NaH 2 PO 4 , 6.0 NaHCO 3 , 1.0 For the HPC model, H9c2 cells were transferred between hypoxic and normoxic conditions for five cycles as described previously [18]. Briefly, cells were placed into the hypoxic chamber under hypoxic buffer (pH 6.8) incubating with 95 % N 2 -5 % CO 2 for 10 min, then were cultured normally in 37 °C, 5 % CO 2 for 15 min.

Cell transfection
The FAM-miR-133a-3p mimics or miR-133a-3p mimics, inhibitor and negative control (NC) miRNAs (GenePharma, Shanghai, China) were transfected into H9c2 cells using Lipofectamine 2000 (Invitrogen, USA) according to the manufacturer's instructions. The work concentration of mimics and mimics NC was 50 nM, whereas that of inhibitor and inhibitor NC was 100 nM. The transfection medium was refreshed after 6 h, and cells were harvested for total RNAs and proteins extraction after being cultured for 24 h. The efficiency of mimics or inhibitor was confirmed by qRT-PCR. Following that, the level of EGFR mRNA and the expression of EGFR protein were detected by qRT-PCR and Western blot, respectively. Meanwhile, after being transfected for 24 h, H9c2 cells were treated with H/R injury for further study.

Treatment with IPC-MVs on I/R injured rats in vivo
Rats were anesthetized intraperitoneally by 25% ethyl carbamate, then performed a thoracotomy before ligation of LAD coronary artery. The ends of the silk suture were passed through a polyethylene tube, which was used to occlude the LAD by pulling the thread. After achieving hemodynamic stability for 15 min, all rats were divided into four groups randomly with n=5 each. (1) Sham group, rats were left untreated for 145 min after a silk ligature was placed around the LAD. (2) I/R group, rats received I/R process.
(3) Sham-MV + I/R group, Sham-MVs (7 mg/kg) were infused via the femoral vein in I/R injured rats. (4) IPC-MV + I/R group, IPC-MVs (7 mg/kg) were infused via the femoral vein in I/R injured rats. The same volume of 0.9% sodium chloride was given to Sham and I/R groups. All treatments began at 25-min ischemia, with additional 1 min infusion.

Microarray analysis of circulating MVs
Total RNAs were obtained from the six MV samples (IPC-MVs, n=3, Sham-MVs, n=3), then were quantified by the NanoDrop ND-2000 (Thermo Scientific) and the RNA integrity was detected by Agilent Bioanalyzer 2100 (Agilent Technologies). After all samples passed the quality control analysis, the total RNAs were dephosphorylated, denaturated and then labeled with Cyanine-3-CTP. After purification, the labeled RNAs were hybridized onto the microarray. After being washed, the arrays were scanned using the Agilent Scanner G2505C (Agilent Technologies).

Isolation and quantification of miRNAs from circulating MVs
Total RNAs were extracted from MVs using the miRNeasy kit (Qiagen) according to the manufacturer's instruction. The yield of RNAs was determined using NanoDrop 2000 (Thermo Scientific, USA), and integrity evaluated using agarose gel electrophoresis. Then reverse transcription was performed with miScript Reverse Transcriptase Kit (Qiagen, Germany) and quantification of miRNAs was performed with QuantiFast® SYBR® Green PCR Kit (Qiagen, Germany). Amplification was performed at 95℃ for 10 min, followed by 30~45 cycles of 95℃ for 10 s, 60℃ for 20 s and 72℃ for 15 s. Each sample was run in triplicate for analysis. miRNA primers were subscribed from Generay Biotech (Generay, PRC), and the levels of miRNA analyzed by qRT-PCR were normalized to that of U6 snRNA.
Fold induction was calculated using the Ct method: DDCt= (Ct Target m iRNA -Ct U6 ) IPC-MVs -(Ct Target m iRNA -Ct U6 ) Sham-MVs, and the final data were derived from 2 -DDCt .

HPC-MV preparation and uptake by H9c2 cells
For generating HPC-MVs from miR-133a-3p overexpressed H9c2 cells, the FAM-miR-133a-3p mimics were transfected into cells using Lipofectamine 2000. After 24 h, H9c2 cells were labeled with a fluorescent dye DiI (Beyotime, Nanjing, China) by incubating them in the DiI working solution (5 μM) for 10 min at 37°C, followed by washing with D-hank's solution. And then HPC-MVs were collected from the hypoxic buffer of H9c2 cells after treatment with HPC as described above. Briefly, hypoxic buffer was centrifuged at 2 700 g, 4°C for 20 min to remove cell debris, followed by ultracentrifugation at 100,000 g, 4℃ for stored at -80 °C. The protein concentration of MVs was determined by BCA assay (Beyotime, Nanjing, China). Then HPC-MVs (30 μg/mL) were incubated with H9c2 cells for 4 h at 37℃, 4% paraformaldehyde (Solarbio, Beijing) was used for fixing cells, DAPI was used for nucleus staining, and images were collected using a fluorescence microscope (Leica, Germany). HPC-MVs derived from unlabeled H9c2 cells were used as negative control (NC).

Authors' Contributions
All authors conceived and designed the study; JYZ and QZ conducted the experiments, collected and analyzed the data. JYZ wrote the first draft of the manuscript. All authors contributed to and approved the final manuscript.