Circulating exosomes repair endothelial cell damage by delivering miR‐193a‐5p

Abstract Circulating exosomes delivering microRNAs are involved in the occurrence and development of cardiovascular diseases. How are the circulating exosomes involved in the repair of endothelial injury in acute myocardial infarction (AMI) convalescence (3‐7 days) was still not clear. In this study, circulating exosomes from AMI patients (AMI‐Exo) and healthy controls (Normal‐Exo) were extracted. In vitro and in vivo, our study showed that circulating exosomes protected endothelial cells (HUVECs) from oxidative stress damage; meanwhile, Normal‐Exo showed better protective effects. Through the application of related inhibitors, we found that circulating exosomes shuttled between HUVECs via dynamin. Microarry analysis and qRT‐PCR of circulating exosomes showed higher expression of miR‐193a‐5p in Normal‐Exo. Our study showed that miR‐193a‐5p was the key factor on protecting endothelial cells in vitro and in vivo. Bioinformatics analyses found that activin A receptor type I (ACVR1) was the potential downstream target of miR‐193a‐5p, which was confirmed by ACVR1 expression and dual‐luciferase report. Inhibitor of ACVR1 showed similar protective effects as miR‐193a‐5p. While overexpression of ACVR1 could attenuate protective effects of miR‐193a‐5p. To sum up, these findings suggest that circulating exosomes could shuttle between cells through dynamin and deliver miR‐193a‐5p to protect endothelial cells from oxidative stress damage via ACVR1.

and its repair mechanism have always been an academic hotspot.
However, there is still a lack of effective treatment methods to reverse endothelial function damage and promote angiogenesis.
Therefore, it is of great significance to explore the mechanism of occurrence, development and repair of endothelial cell injury, and to find effective new therapeutic targets.
Exosomes are a group of vesicle-like bodies between 30 and 150 nm in diameter, which are actively secreted by cells. They have a lipid bilayer subcellular structure on the outside and contain proteins similar to their cellular origin, as well as bioactive substances such as lipids, coding or non-coding RNAs. 2 Exosomes play an important role in intercellular communication and other processes. Under different pathophysiological conditions, exosomes secreted by cells may undergo diverse changes. 3 Exosomes can directly stimulate recipient cells by interacting with receptors.
Also, it can regulate the function of recipient cells through transporting plasma membrane receptors, functional proteins, mRNA, microRNAs and organelles. 4,5 During the transport process, exosomes can protect the bioactive substances from degradation and dilution in the extracellular environment. 6 Circulating exosomes are involved in the occurrence and development of cardiovascular diseases, suggesting that they can be used as an important basis for the diagnosis and possible treatment of cardiovascular disease. 7 Previous studies reported that heart failure patients showed an increasing level of miR-192, miR-194, miR-34a and other microRNAs related to P53 pathway in circulating exosomes, indicating that microRNAs could be used as predictors of heart failure after myocardial infarction. 8 Jakob et al. reported that miR-126 in exosomes derived from endothelial cells could participate in angiogenesis via regulating VEFG signal response and played a role in evaluating the prognosis of patients with cardiovascular disease and the effect of clinical interventions. 9 In AMI patients, HIF-1α was involved in the regulation of hypoxia-induced autophagy of myocardial cells, and it could increase the expression of miR-30a in serum exosomes. 10 Ge et al. reported that exosomes from patients with myocardial ischaemia could promote angiogenesis through miR-939-5p by targeting iNOS-NO pathway. 11 However, it is still not clear that how circulating exosomes involve in the repair of endothelial cell injury in AMI convalescence.
Therefore, the present study investigated how circulating exosomes protect endothelial cells from H 2 O 2 -induced oxidative stress damage in vitro and balloon injury in vivo. We verified the interaction pattern between exosomes and cells in vitro. miR-193a-5p with its downstream target ACVR1 was found to be a key factor in the protective effects of circulating exosomes.

| Patients
AMI patients in this study were recruited from the first affiliated hospital of Zhengzhou University (China) and Nanyang central hospital (China). The convalescence period of AMI was defined as 3-7 days after the occurrence of AMI. AMI was defined according to the 2017 ESC Guidelines. 12 5 mL venous whole blood samples were collected for exosomes extraction using ExoQuick kit (System Biosciences, USA). Blood samples were centrifuged at 3000 g for 20 minutes at 4°C. ExoQuick reagent was added to the supernatant in a ratio of 1:4 for 30 minutes. After centrifugation at 1500 g for 30 minutes, sediment was suspended in PBS of equal serum volume, and the purified circulating exosomes solution was stored at −80°C.

| Identification of circulating exosomes
The morphological characters of circulating exosomes were observed by transmission electron microscopy (TEM) with uranyl acetate staining. The diameter distribution and concentration of exosomes were analysed using nanoparticle tracking analyzer (NanoSight NS300). BCA kits (Solarbio, China) were used to determine the protein concentration of circulating exosomes.
Expression of exosome surface markers were tested by Western blot.
CCK-8 (Dojindo, Japan) reagent 10 μL was added into each well for another 1 hour. Absorbance was measured at 450 nm using a 96well microplate reader, and the optical density values represented the survival of HUVECs.
Tube formation was evaluated by inverted fluorescence microscope.
The total branching length of HUVECs was measured with ImageJ (version 1.52, NIH).

| Scratch wound assay
For the scratch wound assay, HUVECs (3 × 10 5 cells per well) were seeded in six-well plates After cell adhesion, HUVECs were scraped with 200 μL pipette tip. Then, medium was replaced by serum-free ECM medium containing H 2 O 2 (600 μmol/L) with PBS, Normal-Exo (100 μg/mL) or AMI-Exo (100 μg/mL). Cell migration was evaluated by inverted fluorescence microscope after 12 hours. The images were processed by ImageJ to mark the scratch areas.

| Transwell assay
Human umbilical vein endothelial cells were suspended in serum-free ECM medium containing H 2 O 2 (600 μmol/L) with PBS, Normal-Exo (100 μg/mL) or AMI-Exo (100 μg/mL). Then mixtures were seeded into the upper chamber of transwell 24-well plates (Corning, USA) with 8-μm pore filters for 24 hours. And the lower chamber was filled with complete ECM medium. Cells attached on the lower surface were fixed by 95% methanol for 20 minutes and stained with 0.05% crystal violet for 20 minutes at room temperature. Cell migration was evaluated by inverted fluorescence microscope.

| Fluorescent labelling of exosomes
Circulating exosomes were labelled with PKH26 kit (PKH26 Red Fluorescent Cell Linker Kit, Sigma, USA). 100 μL Solution C was added to 50 μL circulating exosomes suspended in PBS. 0.5 μL PKH26 was dissolved in another 100 μL solution C. The above solutions were mixed and incubated for 5 minutes, and 200 μL 5% BSA was added to stop staining. Circulating exosomes labelled with PKH26 were isolated using ExoQuick-TC kit.

| Uptake capacity of circulating exosomes in vitro
Circulating exosomes (100 μg/mL) labelled with PKH26 were added to ECM. After 24 hours, HUVECs incubated with the above ECM were washed with PBS for three times. Fluorescence signal of PKH26 was observed by inverted fluorescence microscope.
(Shanghai, China) according to standard methods.

| Transfection of miRNA mimics and plasmid into HUVECs
HUVECs were digested by trypsin and then cultured into a six-well culture plate with a density of 50-70%. After cell adhesion, complete ECM was replaced with serum-free ECM medium. For RNA, miR-193a-5p mimics (Sense: 5' UGGGUCUUUGCGGGCGAGAUGA 3', Antisense:

| Transfection of miRNA mimics into circulating exosomes
Circulating exosomes were diluted to 0.5 μg/μL using gene pulser electroporation buffer (Bio-Rad, USA), and 400 μL exosomes were added to the electroporation cup (0.4 mm, Bio-Rad, USA). 10 μL dilution of microRNAs mimics or its negative control (NC) was added to the circulating exosomes in electroporation cup. The Gene Pulser Xcell electroporation system (Bio-Rad, USA) was used with the parameters of 400 mF, 125 μf and 10-15 ms for three times. The mixtures were incubated on ice for 5 minutes after each electroporation. 13 After the electroporation was completed, the mixtures were added to 2 mL 1% BSA PBS solution. Circulating exosomes transfected with mimics or mimic-NC were isolated using ExoQuick-TC kit.

| Animals selection and experimental ethics
Male Sprague Dawley (SD) rats (6-8 weeks old) purchased from Beijing Vital River Laboratory Animal Technology Co. Ltd. (Beijing, China) were raised to weight 320-350 g for experiment. The rats were raised in the standard household animal experiment centre in Zhengzhou University (Zhengzhou, China). The drinking water and standard feed for rats were strictly sterilized, and at least 7 days prior to the start of the experiment were used to recover the rats from transport stress. The surgery protocol was designed to minimize animal pain. Experiments were approved by ethics committee of the first affiliated hospital of Zhengzhou University.

| Establishment of carotid artery balloon injury model in rats
10% chloral hydrate (100 g/mL) was used to anaesthetize rats by intraperitoneal injection. The skin was cut off along the median line, and the salivary glands were dissociated from the median line after the subcutaneous tissue was separated. The common carotid artery and its branches internal and external carotid artery passed between torn masseter and sternocleidomastoid. The arteries were completely separated for the next step. The guide wire was rapidly inserted into the incision on external carotid artery to guide 1.5 mm × 20 mm balloon catheter (MicroPort, China) into the common carotid artery. Balloon was inflated to pressure 6atm. After the balloon was pulled, it was evacuated in vacuum. The above steps were repeated three times.
Circulating exosomes (400 μg per rat) were dropped on the surface of the arteries. Muscle and skin were sutured with surgical sutures of 5-0 and 4-0, respectively. Rats were placed on heating pad until fully awake and then returned to cages.

| Histological analysis
Following anaesthesia, the artery was excised and immediately placed in 4% paraformaldehyde at room temperature for 24 hours.
The arterial specimens were embedded in paraffin and cut into 4 μm sections. Serial artery sections were stained with haematoxylin and eosin (HE, Solarbio, China). Images were analysed using ImageJ.
The membranes were blocked with 5% milk in Tris-buffered saline

| Dual-luciferase report assay
The Firefly luciferase and Renilla luciferase reporter gene detection system was used to verify the direct effect of miR-193a-5p on target genes. The 3'-UTR of the human ACVR1 sequence containing the predicted miR-193a-5p-binding sites, or its mutant was loaded into plasmid vectors (pmirGLO-ACVR1 WT or pmirGLO-ACVR1 MUT) (GenePharma, China). 293T cells were co-transfected with above plasmid vectors, respectively, and human miR-193a-5p mimics or its NC.
After 48 hours, a dual-luciferase report assay (Promega, USA) following manual was used to measure the firefly luciferase and renilla luciferase activities. Luciferase results shown as relative light units (firefly luciferase activity divided by renilla luciferase activity) were detected with multi-mode microplate readers (Molecular Devices, USA).

| Statistical analysis
All the experiments were repeated at least three times. Average data were shown as mean ± standard errors. Comparisons between two groups were conducted by two-tailed t test. One-way ANOVA tests were used for comparison among three or more groups with Bonferroni post hoc correction (version 7.0, GraphPad Software). A P < .05 was regarded as statistically different.

| Biological identification of circulating exosomes
We harvested exosomes from 12 AMI patients and 9 healthy controls, and the baseline characteristics of these patients were shown in Table 1. Western blot was used to detect the specific protein markers of circulating exosomes, including HSP70, Alix ( Figure 1A). There was no statistically significant differences in the expression of markers of circulating exosomes between the two groups. Transmission electron microscopy (TEM) showed that circulating exosomes were in bubblelike balloon shape with diameter distribution of 50-150 nm ( Figure 1B), which was consistent with the results of Nanoparticle tracking analysis (NTA) ( Figure 1C). The two groups of circulating exosomes were similar in physical characters such as morphology and diameter distribution.

| Circulating exosomes protected endothelial cells from oxidative stress damage
Circulating exosomes play an important role in the intercellular materials and information exchange. Previous researches have reported that serum exosomes can promote angiogenesis of endothelial cell, which is a crucial factor in the development of cardiovascular disease. 14,15 In order to verify whether circulating exosomes can reduce the oxidative stress damage induced by H 2 O 2 , HVUECs were treated with H 2 O 2 (600 μmol/L) and circulating exosomes derived from the two groups. In CCK8 assay, Normal-Exo (100 μg/mL) and AMI-Exo (100 μg/mL) protected HUVECs from H 2 O 2 after 24 hours. Besides, Normal-Exo showed better protective effect (Figure 2A). In tube formation assay, total branching length of HUVECs treated with AMI-Exo (100 μg/mL) was lower than that treated with Normal-Exo (100 μg/mL), which meant that AMI-Exo had a weaker protective effect than Normal-Exo, in terms of vascular formation ability ( Figure 2B). In addition, the protective effects of exosomes on cell migration also varied between the two groups.
Both scratch wound assay and transwell assay showed that cir-

| Circulating exosomes shuttled between cells through dynamin
Circulating exosomes interact with cells in a variety of ways, among which the most convincing hypothesis is that exosomes release their contents after being ingested by cells. 16 In order to verify whether the uptake of circulating exosomes via dynamin, HUVECs were cocultured with dynamin inhibitor, dynasore. And circulating exosomes (100 μg/mL) were stained by PKH-26. After 24 hours, fluorescence microscopy showed that compared with the control group, dyansore had a significant, dose-dependent inhibitory effect on the uptake of circulating exosomes, suggesting that circulating exosomes were ingested by HUVECs through dynamin ( Figure 3A). Transwell system was used to verify this function. And the results showed that HUVECs in lower chambers could ingest stained exosomes without direct contact with HUVECs in upper chambers ( Figure 3B-C), suggesting that circulating exosomes can shuttle among cells.

| miR-193a-5p might be the key factor in protective effects of circulating exosomes
All experiments mentioned above revealed that compared with AMI-Exo, Normal-Exo had a more obvious vascular protection effect. And the contents of circulating exosomes, such as proteins, microRNAs and lnc-RNA, might be responsible for their protective effects. To further investigate the underlying mechanism, we explored the microRNAs of the two groups exosomes. After a microarray analysis, the differently expressed microRNAs in the two groups were shown in the heat map ( Figure 4A). qRT-PCR confirmed differential expression of miR-193a-5p in two groups ( Figure 4B).
Mimics of miR-193a-5p were transfected into HUVECs. Then, transfected HUVECs were treated with H 2 O 2 (600 μmol/L) for 24 hours, and CCK-8 assay confirmed that miR-193a-5p showed obvious protective effect on HUVECs when compared with negative control ( Figure 4B). miR-193a-5p is an important target in various diseases, and it plays different roles in different models. 17 To confirm whether miR-193a-5p could protect vascular formation ability of HUVECs, miR-193a-5p mimics and mimics negative control were transfected into HUVECs for function analyses. Tube formation assay showed the same results as previous CCK-8 assay ( Figure 4C), suggesting that miR-193a-5p directly protected HUVECs from H 2 O 2 -induced oxidative stress damage. In rat carotid artery balloon injury models, miR-193a-5p mimics and negative control were transfected into AMI-Exo through electroporation to confirm that miR-193a-5p did play a role through circulating exosomes in vivo. The quantitative analysis of intima/media ratios showed that miR-193a-5p was essential to the protective effects of exosomes (400 μg/rat). ( Figure 4D).
To sum up, miR-193a-5p, delivered by circulating exosomes, played an important role in promoting the repair of endothelial damage in vitro and in vivo.

| AVCR1 as a target gene of miR-193a-5p were involved in protecting endothelial cells
Bioinformatics analyses including GO and KEGG were applied to screen potential downstream targets of miR-193a-5p. The results showed that plasma membrane adhesion molecules managing homophilic cell adhesion were the most correlated targets ( Figure 5A Figure 5D). Dual-luciferase report confirmed that ACVR1 was the target gene of miR-193a-5p ( Figure 5E). be blocked by its inhibitor dynasore. 25 Furthermore, exosomes propagate the toxic materials between cells, spreading pathology in the Alzheimer brain. 26,27 Our study showed that circulating exosomes could be uptaken by cells and undertake the cell-cell communications, while maintaining membrane structure and internal substances, indicating the possible mechanism of the protective effects of circulating exosomes on oxidative stress damage (Figure 3).
In recent years, microRNAs in exosomes had gradually become a research hotspot. Specific microRNAs have been proven to play an important role in disease development and recovery, which meant that microRNAs could be potential targets for the diagnosis, prevention and treatment of disease. 28,29 To find out the func- reduced PC xenograft growth in vivo. 34 Zhou et al. found that overexpression of miR-193a-5p increased the capacity of migration and cisplatin resistance in urothelial cells. 35 In study of cell derived exosomes, it is a mature and simple technique to change the content of exosomes. Therefore, knockdown or overexpress the expression of certain microRNAs in exosomes is a significant step to confirm its function. As the contents of circulating exosomes could not be interfered from their source, we chose electroporation to directly transfect microRNAs into circulating exosomes with reference to relevant studies. 13 We verified the success of this transfection method by fluorescent labelling microRNAs ( Figure S1). In order to clarify the role of miR-193a-5p, we directly transfected miR-193a-5p mimics into AMI-Exo. And animal experiment showed that AMI-Exo transfected with miR-193a-5p mimics had a better protective effect ( Figure 4E). To explore the downstream of miR-193a-5p, bioinformatics analyses including GO and KEGG by Metascape were applied. 36 Through TargetScan and miRDB databases of microRNAs, we obtained 144 and 249 predicted target genes, respectively.  The coincident predicted target genes in the above results, a total of 78 genes, were incorporated into GO and KEGG analyses to identify the possible downstream target of miR-193a-5p ( Figure   S2). The results showed that plasma membrane adhesion molecules managing homophilic cell adhesion were the most related targets ( Figure 5A-B). Among these predicted target genes, activin A receptor type I (ACVR1) and its downstream pathways had been shown to be involved in a variety of physiological and pathological processes. 18 We found that the expression of ACVR1 was inhibited by miR-193a-5p transfection, and the dual-luciferase report confirmed that ACVR1 was the target gene of miR-193a-5p ( Figure 5C-E Overexpression of ACVR1 could reverse the vascular protective effects of miR-193a-5p, which indicated that miR-193a-5p acted as a protector by inhibiting ACVR1 (Figure 6D-F).
In our study, we demonstrated that circulating exosomes could deliver miR-193a-5p to protect endothelial cells from oxidative stress damage via targeting ACVR1 (Figure 7).

| LI M ITATI O N S
However, our study still has some limitations. Therefore, hospitalized patients with less than 50% coronary artery stenosis would be a better control group; (e) Compared with HUVECs, human coronary artery endothelial cells (HCAECs) were more specific and suitable to our study. HUVECs were widely used in cardiovascular research and had accumulated a wealth of evidence. Because of the limited conditions, we used HUVECs for in vitro experiments.

| CON CLUS ION
In conclusion, we demonstrated that circulating exosomes could shuttle between cells through dynamin and deliver miR-193a-5p to protect endothelial cells from oxidative stress damage via targeting ACVR1. Youth Science Foundation of Henan Province.

CO N FLI C T O F I NTE R E S T
The authors confirmed that there was no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets generated and/or analysed during the current study are available from the corresponding author upon reasonable request.