Bradykinin-pretreated Human cardiac-specific c-kit+ Cells Enhance Exosomal miR-3059-5p and Promote Angiogenesis Against Hindlimb Ischemia in mice

Protection of cardiac function following myocardial infarction was largely enhanced by bradykinin-pretreated cardiac-specific c-kit+ (BK-c-kit+) cells, even without significant engraftment, indicating that paracrine actions of BK-c-kit+ cells play a pivotal role in angiogenesis. Nevertheless, the active components of the paracrine actions of BK-c-kit+ cells and the underlying mechanisms remain unknown. This study aimed to define the active components of exosomes from BK-c-kit+ cells and elucidate their underlying protective mechanisms. Matrigel tube formation assay, cell cycle, and mobility in human umbilical vein endothelial cells (HUVECs) and hindlimb ischemia (HLI) in mice were applied to determine the angiogenic effect of condition medium (CM) and exosomes. Proteome profiler, microRNA sponge, Due-luciferase assay, microRNA-sequencing, qRT-PCR, and Western blot were used to determine the underlying mechanism of the angiogenic effect of exosomes from BK-c-kit+. As a result, BK-c-kit+ CM and exosomes promoted tube formation in HUVECs and the repair of HLI in mice. Angiogenesis-related proteomic profiling and microRNA sequencing revealed highly enriched miR-3059-5p as a key angiogenic component of BK-c-kit+ exosomes. Meanwhile, loss- and gain-of-function experiments revealed that the promotion of angiogenesis by miR-3059-5p was mainly through suppression of TNFSF15-inhibited effects on vascular tube formation, cell proliferation and cell migration. Moreover, enhanced angiogenesis of miR-3059-5p-inhibited TNFSF15 has been associated with Akt/Erk1/2/Smad2/3-modulated signaling pathway. Our results demonstrated a novel finding that BK-c-kit+ cells enrich exosomal miR-3059-5p to suppress TNFSF15 and promote angiogenesis against hindlimb ischemia in mice.


Introduction
Numerous studies have suggested that stem/progenitor cell transplantation is an innovative strategy that can restore cardiac function in heart failure [1]. Among the various types of progenitor/stem cells studied, autologous cardiac specific c-kit + cells (c-kit + ), a cluster of multipotent progenitor cells found in the adult heart, is one of the most promising cell types presently used for cardiac regeneration [2]. Although there is strong evidence of the therapeutic potential of c-kit + cells in heart failure at a preclinical level, there is mounting evidence that transplanted c-kit + cells do not engraft in the diseased heart, not differentiate into mature cardiac myocytes, and do not regenerate dead myocardium. This suggests that their major mechanism may involve paracrine modulation rather than cell-directed differentiation and functional integration of damaged tissue [3,4]. A better understanding of the paracrine signaling of c-kit + cells involved in the repair of damaged tissue is essential in the treatment of cardiovascular disease. Exosomes are small extracellular vesicles (50-150 nm) released from cells by a lipid bilayer. Functionally, they mediate intercellular communication under physiological and pathological conditions by stably transferring active proteins, lipids, mRNAs, and microRNAs [5][6][7]. Studies have shown that exosomes can be considered essential elements in homeostasis regulation, cardiac repair, and cardioprotection by the paracrine effect of c-kit + cells [8][9][10]. Despite the importance of exosomes in paracrine signaling, their mediators and therapeutic mechanisms remain unidentified.
We previously reported that bradykinin (BK), a vasoactive nonapeptide messenger, can regulate the proliferation and migration of c-kit + cells by regulating the PLC/pAkt/ pERK/Cyclin D1 pathway via the B2 receptor [11]. We also demonstrated that Ca 2+ channels in c-kit + cells, such as the IP3R/SOCE channel, play a vital role in BK-regulated cell proliferation and migration [12]. More recently, we revealed that BK could attenuate H 2 O 2 -induced c-kit + cells apoptosis and promote c-kit + cells survival in a rat model of myocardial ischemia (MI) [13]. Despite the poor long-term survival of transplanted c-kit + cells, enhanced angiogenic effects of BK-c-kit + cells in the rat model of MI were observed over 6 weeks, indicating that BK-c-kit + may exert their beneficial effects through paracrine signaling. Our previous studies have demonstrated that stem cell-secreted trophic factors play critical roles to protect hosted tissues against cell death [14]. Nonetheless the key mediators and mechanisms that underlie the paracrine actions of BK-c-kit + in angiogenesis have not been elucidated. Recent progress in stem cell research suggested that stem cell-secreted extracellular vesicles including exosomes have emerged as a novel therapeutic tool for tissue regeneration [15,16] and hindlimb ischemia [17,18]. We recently also found that BKpretreated heart tissue-derived c-kit + cells (BK-c-kit + ) produced abundant exosomes that might benefit for hindlimb ischemia in a pilot study. To this end, we conducted this study to explore the protective effects of exosomes from BK-c-kit + cells by focusing on their active components and identifying the underlying mechanism in a hindlimb ischemia model.

Cell Culture
Human tissue collection and animal experiments were performed in accordance with the Declaration of Helsinki guidelines, and were approved by the ethics committee of Xiamen Cardiovascular Hospital of Xiamen University (Ethical approvals: No. XMH201710). Patient consent was obtained at the start of the study. Human heart atrial tissue was obtained from 0-2-year-old patients with congenital heart disease who underwent cardiovascular surgery at the Xiamen Cardiovascular Hospital of Xiamen University. c-kit + cells were isolated from atrial tissue as previously described [19,20]. The isolated c-kit + cells were cultured in Iscove's Modified Dulbecco's Medium (IMDM, Gibco, USA) supplemented with 10% (v/v) fetal bovine serum (FBS, Gibco, USA), 1% penicillin/streptomycin (Gibco, USA), and 50 µM 2-mercaptoethanol (Gibco, USA).

Conditioned Medium and Exosome Isolation from c-kit + Cells
Conditioned medium (c-kit + -CM) was obtained from the c-kit + cells culture supernatant containing a complex of c-kit + cells secreted products, and treated with FBS-free IMDM supplemented with phosphate-buffered saline (Gibco, USA, vehicle control, PBS-CM) or BK (Sigma-Aldrich, USA, 10 nM, BK-CM) for 72 h. After treatment, PBS-CM or BK-CM was obtained by centrifugation at 2,000 g and 4 °C for 10 min to eliminate dead cells and at 10,000 g for 30 min to eliminate cell debris. For exosome isolation, exosomes from c-kit + cells (PBS-Exos) or BK-c-kit + cells (BK-Exos) were isolated via ultracentrifugation at 100,000 g for 2 h ( Fig. 2A), then resuspended at a concentration of 1 mg/mL in PBS. The shape of exosomes was detected by transmission electron microscope (TEM), and their surface marker proteins were verified by Western blot. Exosomes were quantified by protein concentration in the suspension with PBS using BCA kits (Beyotime Biotech Inc, China). To identify whether exosomes were uptake by HUVECs, we labelled exosomes with 40 µg/ml Dil (Beyotime Biotech Inc, China) for 30 min and then incubated them with HUVECs at 37 °C for 3 h. After washing with PBS, fixed by 2% paraformaldehyde and incubated with DAPI for 20 min, cells were imaged by a Leica SP5 II confocal fluorescent microscope.

Mouse Model of Hindlimb Ischemia and Treatment
All animal studies were conducted following the Declaration of Helsinki guidelines and were approved by the Ethics Committee of Xiamen University (Ethical approval No. XMULAC20190120). All procedures conformed with the guidelines of Directive 2010/63/EU of the European Parliament on the protection of animals used for scientific purposes. C57BL/6 mice (male, 6-8 weeks old) were purchased from Shanghai SLAC Laboratory Animal Co., Ltd (Shanghai, China). Animals were housed in a controlled environment with a 12 h light/dark cycle and fed ad libitum. A murine model of hindlimb ischemia was established as previously described [21]. Briefly, after anesthetizing mice with 2% vaporized inhaled isoflurane, the femoral artery was ligated in the ischemia group, and the femoral artery was exposed but not ligated in the sham group. The femoral artery and areas around the arterial branches and small segments were removed. Mice were then randomly assigned to receive intramuscular injections of PBS (n = 10), PBS-CM (n = 10), BK-CM (n = 10), PBS-Exos (n = 7), BK-Exos (n = 7), Scr-PBS-Exos (n = 10), Scr-BK-Exos (n = 10), miR3059 −/− -PBS-Exos (n = 10), or miR3059 −/− -BK-Exos (n = 10). CM (125 µL) or exosomes (10 µg in 125 µL PBS), both from 1.25 × 10 5 c-kit + cells, were directly injected into the ischemic hindlimb at 5 locations with 25 µL per site. Mice in the control group received PBS in the same routes as that in the treatment groups. In the present animal studies, there are not any mice who died or were excluded from the analysis.

Laser Speckle Blood Flow Monitoring System for Limb Functional Recovery
The PeriCam PSI NR laser speckle blood flow monitoring system was purchased from PeriMed (Sweden). To measure blood flow and the perfusion ratio of hindlimb ischemia, mice were anesthetized with isoflurane (2%), and measurements were obtained 0 and 28 days after the establishment (Abcam, UK). After further incubation of membranes with enhanced chemiluminescence select reagent (GE Healthcare Life Sciences, Sweden), they were imaged using the Bio-Rad Chemi Doc™ chemiluminescence imaging system. Gray density was calculated using Image J software. Exosomes used in both BK-and PBS-treated c-kit + cells were mixed with five c-kit + cell lines from five donors.

MicroRNA (miRNA) Sequencing
To determine the different profiles of miRNAs in exosomes from BK-treated or untreated c-kit + cells CM, we used a miRNA sequence to observe the differentially expressed miRNAs in the two groups. Exosomes isolated from c-kit + cells of at least three donors were used for the experiments (n = 3). Total RNA was isolated from the exosomes of PBStreated c-kit + cells and BK-treated c-kit + cells using TRIzol (Life Technology, USA). The NEBNext® Multiplex Small RNA Library Preparation Set kit (Illumina, USA) was used for library preparation using the Agilent 2100 Bioanalyzer (USA). Sequencing was performed using an Illumina Next-Seq 500 Bioanalyzer sequencing system (Aksmics Biotech, China).

miRNA Expression Analysis
To confirm the expression of miRNAs, we determined the differentially expressed miRNAs from BK-or PBS-treated c-kit + exosomes by qRT-PCR. Total RNA was extracted from exosomes using TRIzol, and subsequently reverse transcribed into cDNA using All-in-One™ miRNA qRT-PCR Reagent Kits and validated primers (Genecopoeia, USA) according to the manufacturer's instructions. The expression of candidate genes was determined via qRT-PCR using Bio-Rad CFX96 (USA). The 2 −ΔΔCt method was used to calculate changes in mRNA expression level. The expression value of miRNAs was normalized against that of endogenous control U6.

Luciferase Reporter Assay
The total length 3'UTR of human TNFSF15 (VEGI) was cloned into a luciferase reporter vector, and mutagenesis of the putative miR-3059-5p binding sites to create a mutant of TNFSF15. HUVECs in 96-well plates were transfected with TNFSF15 wild type (100 nM) or TNFSF15 mutant (100 nM) and co-transfected with miR-3059-5p mimic (miR3059, 50 nM) or negative control (miNC, 50 nM) using Lipofectamine 2000 transfection reagent (Invitrogen, USA). A Dual-Glo Luciferase Assay System (Promega, USA) was used to quantify miR-3059-5p binding 48 h posttransfection. The intensity of the signal was measured using of hindlimb ischemia. Statistical analysis of defects of the hind limbs for ischemic mice was performed as described previously [22].

Immunohistology for Neovascularization
After 28 days of ischemia, mice were euthanized by intraperitoneal administration of ketarom (10 mg/kg xylazine mixed with 100 mg/kg ketamine, 0.2 mL/ 20 g) and blood samples were obtained. Subsequently, ischemic limb muscle tissue was harvested and embedded in OCT (Sakura, Japan). Cross-Sect. (6 μm) of muscle were subjected to CD31 histological immunostaining. Briefly, sections were blocked with 10% donkey serum at 37 °C for 1 h. Primary antibodies against CD31 (Abcam, UK), an endothelial cell marker, were applied at 4 °C overnight, and subsequently, Alexa 488 fluor-conjugated secondary antibodies (Abcam, UK) were used. Cell nuclei were counterstained with DAPI (Vector Laboratories, USA). CD31-positive cells were observed and quantified under a Leica SP5 II confocal fluorescent microscope. For quantification of capillary density, 5 mice in each group were used for analysis, 5 tissue sections from one mouse were picked up, and 5 areas per tissue section were randomly captured. The total number of CD31 + cells was counted. The quantification of capillary density was presented as a format of cell counts/mm 2 by using Image J software (NIH, USA).

Exosomes from miR3059-5p-sponge Infected c-kit + Cells
c-kit + cells were infected with HBLV-hus-miR3059-5psponge-ZsGreen-PURO or HBLV-ZsGreen-PURO negative control lentivirus (Hanbio Biotechnology, China) for 48 h. The cells were then washed and treated with BK or PBS for an additional 72 h. Exosomes were isolated using a previously described protocol via ultracentrifugation and then used in the in vitro Matrigel tube formation assay or model of mouse hindlimb ischemia.

Proteome Profiler Human Angiogenesis Array
Angiogenesis-related protein profiling of exosomes from BK-treated or untreated c-kit + -CM was performed using a proteome profiler human angiogenesis array kit (R&D systems, USA) that contained 55 different capture antibodies printed in duplicate, according to the manufacturer's protocol. Briefly, exosomes from the CM of BK-treated or untreated c-kit + cells were incubated with membranes for 1 h. The reconstituted detection antibody cocktail was incubated with membranes at 2-8 °C overnight. After washing, the membranes were incubated with streptavidin-HRP chambers were washed with PBS three times and fixed with 4% paraformaldehyde (Sigma-Aldrich, USA). Crystal violet (Sigma-Aldrich, USA) was used to stain cells for 15 min. After rinsing thoroughly, non-migrated cells on the upper membranes were wiped off with cotton swabs. The migrated cells of the membrane were photographed by microscopy and counted with Image J software (NIH, USA).

Statistical Analysis
GraphPad Prism 9.0 was used to analyze data and prepare graphs. Comparisons between two groups were evaluated using Student's t-test, and between three or more groups using ANOVA. P < 0.05 was considered statistically significant. Data are expressed as mean ± SEM.

Conditioned Medium (CM) from BK-pretreated c-kit + Cells Promoted Angiogenic Activity in vitro and in vivo
Our previous study demonstrated that BK-pretreated c-kit + cells improved heart function in a rat model of MI. The underlying mechanism may be related to the paracrine effect a microplate reader (Tecan, Swiss), and the intensity normalized to firefly luciferase signal intensity.

Bulk RNA Sequencing
Bulk RNA sequencing was conducted as described previously to investigate the angiogenic target of miR3059 in HUVECs [23] (for miR3059-or scramble miRNA-treated HUVECs infected with HBLV-hus-miR3059-5p-sponge-ZsGreen-PURO or HBLV-ZsGreen-PURO negative control (Hanbio Biotechnology, China) for 48 h. Briefly, total RNA was isolated from miR3059-or scramble miRNA-treated HUVECs using TRIzol. The mRNA Library Preparation Set kit (Illumina, USA) was used for library preparation using the Agilent 2100 Bioanalyzer. Illumina NextSeq 500 Bioanalyzer was used to sequence the different expressed mRNAs. Differential gene expression analysis was performed by DESeq2 [24]. Principal component analysis was performed using Log 2 FPKM values and R ggplot2 package. Heatmaps were generated using the R plots package. GO term and KEGG enrichment were analyzed using the R clusterProfiler package [25] and bar plots were generated using GraphPad Prism software. Gene Set Enrichment Analysis was performed using GSEA software (version 4.0.3).

Cell Cycle
Cell cycle distribution was detected using flow cytometry in HUVECs stained with propidium iodide (PI, Invitrogen, USA), as described previously [11,12,26]. After transfection with miRNA or siRNA particles for 48 h, cells were digested with 0.25% trypsin. Cells were washed and fixed in ice-cold 70% ethanol overnight at 4 °C. Cells were then washed and incubated with a staining solution (20 µg/mL PI and 10 µg/mL RNase A in 0.1% Triton X-100 in PBS) for 30 min. The distribution of different stages of cell cycling was analyzed by fluorescence-activated cell sorting using a flow cytometry analyzer (LSR Fortessa, BD Biosciences, USA). The percentage of cells at G0/G1, S, and G2/M phases was calculated using MODFIT software (BD Biosciences, USA).

Cell Mobility
Transwell assay was employed to determine cell migration in HUVECs [11,12]. Briefly, after being transfected with miRNA or siRNA particles, cells in 200 µL 1% FBS medium were placed for 30 min in a modified Boyden chamber with an 8 μm-pore polycarbonate membrane (Corning, USA) pre-coated with serum-free medium. The lower chamber contained 600 µL 1% FBS medium. Plates were incubated at 37℃ in 5% CO 2 for 12 h. Then the compared with 20% and 80% respectively PBS-CM-treated mice. Most importantly, in BK-CM-treated mice, only 20% of mice showed foot necrosis with limb salvage in 80%. These results indicate that CM from BK-treated c-kit + cells were rich in active components that provided protective effects against hindlimb ischemia.

Exosomes from BK-treated c-kit + Cells Significantly Promoted Angiogenesis in vitro and in vitro
Exosomes are key components enriched with paracrine secretomes. To further investigate whether BK-treated c-kit + cells exerted their pro-angiogenic effects via exosomes, we separated the exosomes from CM by ultracentrifugation. The procedure for exosome separation is shown in Fig. 2A. The morphology of separated exosomes was examined under a transmission electron microscope (Fig. 2B). Western blot assay showed that exosomes from c-kit + of BK-c-kit + cells [13]. To determine whether BK regulates paracrine secretion in c-kit + cells, we first examined the vasculogenic activity of BK-treated or untreated c-kit + -CM in HUVECs (Fig. 1A and B) using a Matrigel tube formation assay. We established that control CM (PBS-CM) induced angiogenesis whereas BK-CM significantly enhanced the angiogenic effect in HUVECs by promoting total master segment length, total mesh area, total length, and total segment length (Fig. 1A and B).
Angiogenesis is a vital factor in the repair of ischemic limbs. We further validated the effects of BK-CM in mice with hindlimb ischemia (HLI). As shown in Fig. 1C and D, PBS-treated mice exhibited significant hindlimb muscle atrophy induced by ischemia. Interestingly, hindlimb ischemia was improved in PBS-CM-treated mice, although the most significant improvement was observed in BK-CM-treated mice ( Fig. 1C and D). Notably, in PBS-treated mice, 80% showed limb loss and 20% showed foot necrosis We first explored the different profiles of angiogenesisrelated proteins in BK-or PBS-treated c-kit + cells exosomes using a proteome profiler human angiogenesis array kit. The results showed that BK-Exos had no significant effect on angiogenesis-related proteins compared with PBS-Exos ( Fig. 3A and B). We then performed miRNA sequencing and qRT-PCR to compare the differential expression of miR-NAs in exosomes isolated from PBS-and BK-treated c-kit + cells. miRNA sequencing revealed that BK-Exos exhibited a range of differentially expressed miRNAs compared with PBS-Exos including eight upregulated miRNAs and eight downregulated miRNAs (Fig. 3C and D). We then selected and validated six upregulated miRNAs, miR-3059-5p, miR-200a-5p, miR636, miR-141-3p, miR-200a-3p, and miR-34c-5p, to determine those that most significantly associated in the angiogenic activity of exosomes from BK-mediated c-kit + cells. Among them, the expression of miR-3059-5p in BK-Exos was the most prominently increased, approximately 20-fold that of PBS-Exos, and verified by qRT-PCR (Fig. 3E). These results suggest that miR-3059-5p may be responsible for the angiogenic function of BK-regulated c-kit + cells exosomes.

Knockdown of miR-3059-5p in c-kit + Cells Abolished the pro-angiogenic Effects of Exosomes from BKtreated c-kit + Cells in vitro and in vivo
To further validate whether miR-3059-5p was the main contributor to the pro-angiogenic effect of BK-treated c-kit + cells exosomes, we isolated exosomes infected with miR-3059-5p sponge lentivirus from c-kit + cells. To determine the angiogenic activity of exosomes from BK-treated c-kit + cells with miR-3059-5p knockdown, we treated HUVECs with PBS-Scr-Exos, BK-Scr-Exos, miR3059 −/− -PBS-Exos, or miR3059 −/− -BK-Exos, and applied a Matrigel tube formation assay. Interestingly, the pro-angiogenic effect of BK-Exos on tube formation was abolished in the presence of miR-3059-5p-sponge-infection ( Fig. 4A and B) in HUVECs. These results illustrate that miR-3059-5p is an essential mediator of BK-treated c-kit + cells exosomemediated angiogenic function in vitro.
To determine whether miR-3059-5p was responsible for the pro-angiogenic function of BK-treated c-kit + cells exosomes in the repair of ischemic limbs, we injected the exosomes of PBS-or BK-treated c-kit + cells transfected with miR-3059-5p-sponge into HLI mice. Similar to the in vitro results, the recovery of hindlimb ischemia was dynamically decreased in miR3059 −/− -PBS-Exo and miR3059 −/− -BK-Exo-treated mice after 42 days (Fig. 4C). Notably, the decreased rate of limb loss and foot necrosis by BK-Exos was abrogated by miR-3059-5p-sponge infection ( Fig. 4D and E). These results imply that BK-treated c-kit + cells displayed the exosome surface marker proteins CD63 and CD81, and did not express the endoplasmic reticulum proteins Grp94 and Calnexin (Fig. 2C). All corresponding uncropped full-length gels and blot are shown in the supplemental Fig. 2. To determine whether exosomes can affect HUVECs, we labeled exosomes with Dil and treated HUVECs with them. The results showed that the exosomes separated from c-kit + cells could be transferred into HUVECs (Supplemental Fig. 1). We then examined whether exosomes participated in the pro-angiogenic effect of BKtreated c-kit + cells. Exosomes and exosome-depleted CM were used to detect angiogenesis in HUVECs. The results showed that control exosomes (PBS-Exos) exhibited lower pro-angiogenic activity, and BK-exosomes (BK-Exos) significantly promoted this effect, similar to the effect of CM from both control and BK-treated c-kit + cells. Intriguingly, when exosomes were depleted from CM, both control CMdepleted exosomes (PBS-Exos-depl) and BK-CM-depleted exosomes (BK-Exos-depl) demonstrated no pro-angiogenic effect on tube formation in HUVECs, implying that effective promotors in HUVECs of angiogenesis mainly existed in exosomes ( Fig. 2D and E).
To investigate whether the angiogenic effect of exosomes from BK-treated c-kit + cells could promote the repair of HLI, we examined blood flow in the ischemic limb of HLI mice using a laser speckle blood flow monitor system. As shown in Fig. 2F H, PBS-injected ischemic hindlimbs showed no restoration of blood flow, while control exosomes from c-kit + cells (PBS-Exos) slightly improved blood flow and BK exosomes (BK-Exos) showed a substantially improved blood flow as the effect of BK-CM. Importantly, improved blood flow following BK-Exos treatment was in agreement with increased capillary density in ischemic tissues. As shown in Fig. 2I J, compared with PBS control or PBS-Exos-treated mice, BK-Exos-treated mice significantly enhanced capillary density detected by CD31 positive immunostaining. These results demonstrated that the angiogenic effect of BK-promoted c-kit + cells were mainly regulated by exosomes.

Propeomic Profiling and miRNA Sequencing Revealed that MiR-3059-5p was Highly Enriched in BK-treated c-kit + cells-exosomes
Exosomes function as mediators of cell-cell communication by selectively carrying biologically active molecules in the form of proteins and miRNAs [27]. To identify the involvement of particles in the angiogenic effect of BKtreated c-kit + cells exosomes, we examined the expression of angiogenesis-related proteins and microRNAs in exosomes from PBS-or BK-treated c-kit + cells.
angiogenic effect (Fig. 4H) were also significantly promoted by miR-3059-5p mimic. In addition, phosphorylation of Akt, Erk1/2, and Smad2/3 was determined by Western blot and revealed that miR-3059-5p significantly promoted the phosphorylation level of cell survival and proliferation kinase Akt and Erk1/2, and Smad2/3 pathway. This suggests that miR-3059-5p may exert its angiogenic effects by regulating these two signaling pathways (Fig. 4I and J). cells exosomes promoted angiogenesis and repaired blood vessels mainly through the expression of miR-3059-5p.
To further explore the mechanism of exosomal miR-3059-5p in angiogenesis, we examined the cell cycle, cell mobility, and vascular tube formation in miR-3059 mimic-or control mimic-transfected HUVECs. The cell cycle of HUVECs was promoted in S-phase and decreased in the G0/G1 phase after transfection with miR-3059-5p mimic (miR3059) compared to the control mimic (miNC) (Fig. 4F). Meanwhile, migratory function (Fig. 4G) and significant change in the mutant TNFSF15 3'UTR group after co-transfection with miR-3059-5p (Fig. 5F). Additionally, the protein level of VEGI was decreased by miR-3059-5p mimic (Fig. 5G and H). Collectively, these results confirmed that TNFSF15 was the target of miR-3059-5p.

Deficiency of TNFSF15 Promotes Vascular tube Formation, cell Cycling, and cell Mobility via the Akt/ Erk/Smad2/3 Pathway
To determine whether miR-3059-5p-targeted TNFSF15 is the critical mechanism for the regulation of angiogenesis, we used a VEGI (TNFSF15)-siRNA to silence TNFSF15 expression. Consistent with the effect of overexpression of miR-3059-5p, compared with the control group, deficiency of TNFSF15 expression in HUVECs led to more vascular tube formation ( Fig. 6A and B), enhanced cell cycle progression ( Fig. 6C and D) and cell migration ( Fig. 6E and F). Phosphorylation of Akt, Erk1/2, and Smad2/3 was also promoted in VEGI-siRNA transfected cells compared with control siRNA transfected cells ( Fig. 6G and H). These results demonstrated that the Akt/Erk1/2/Smad2/3 pathway was involved in the regulation of VEGI-modulated angiogenic functions.

Discussion
Cardiovascular disease results in a decline in heart function and ultimately irreversible heart failure, the primary cause of morbidity and mortality [28]. Despite decades of research, long-term outcomes of cardiac transplantation remain unsatisfactory, and no significant survival improvement has been observed over the past decade [29]. Novel therapeutics are required to improve the survival of patients with heart disease and to decrease morbidity due to heart failure. Stem cell therapy has emerged as a promising therapeutic option to improve the outcomes for patients with heart failure. Among stem cells/progenitor cells, c-kit + cells show great potential as an ideal cell type for cell transplantation to treat heart failure. At least 50 studies from 26 independent laboratories have reported the beneficial effects of c-kit + cells in mice, rats, pigs, and cats [3]. Although c-kit + cells have potential benefits in heart failure, their therapeutic benefit cannot be attributed to stem cell survival and differentiation since they are not retained by the host myocardium, suggesting that cardiac repair occurs primarily through paracrine signaling mechanisms. Exosomes are considered a key factor in intercellular communication based on the transport and transfer of peptides, lipids, and nucleic acids that have the potential to modulate signaling pathways and participate in different functions, including cardiovascular protection

miR-3059-5p Exerts Angiogenic Effects by Targeting TNFSF15(VEGI)
It has been well established that miRNAs exert their effects by binding directly to their target mRNA and triggering translational repression and degradation. We conducted mRNA sequencing to search for the negatively regulated target gene of miR-3059-5p by transfecting the control mimic and miR-3059-5p mimic. The results revealed 267 up-regulated and 251 down-regulated genes in miR-3059-5p overexpressed HUVECs ( Fig. 5A and B). Among them, we found TNFSF15, which encoded vascular endothelial growth inhibitor (VEGI), was significantly downregulated by miR-3059 ( Fig. 5C and D). Next, we searched for potential targets of miR-3059-5p by employing bioinformatic tools miRDB (http://mirdb.org), an online database for miRNA target prediction and functional annotations. Bioinformatic analysis identified 3'UTR of TNFSF15 as the potential binding site of miR-3059-5p (Fig. 5E). To validate the relationship between TNFSF15 and miR-3059-5p, we transfected HUVECs with miR-3059-5p and analyzed the expression of TNFSF15. To further verify the association of miR-3059-5p with TNFSF15, we constructed the wild-type or mutant 3'UTR of human TNFSF15 (VEGI) into a dual-luciferase reporter system to determine the miR-3059-5p binding site. The results of dual-luciferase reporter assay showed that the expression of TNFSF15 was significantly decreased on co-transfection with miR-3059-5p and wild-type 3'UTR TNFSF15. Nonetheless, there was no Fig. 2 Exosomes isolated from BK-treated c-kit + -CM significantly promoted tube formation in HUVECs and recovered blood perfusion in mice with hindlimb ischemia. A, Flowchart of exosome isolation from c-kit + CM. B, Representative images of exosomes captured using an electron microscope. C, Western blot detection of the expression of CD63, CD81, Grp94 and Calnexin in c-kit + cells and exosomes. Full-length blots are presented in supplementary file 2: Figure. S2A. D, Representative tube formation images of HUVECs treated with 5% FBS ECM (PC), FBS-free ECM (NC), PBS-Exos, BK-Exos, and CM depleted exosome treated with BK (BK-Exos-depl) or PBS (PBS-Exos-depl). E, Analysis of tube formation by ImageJ software and presentation of total master segment length, total mesh area, total length, and total segment length (n = 6). F, Representative blood perfusion images of sham mice (Sham), hindlimb ischemic mice treated with PBS (PBS), PBS-Exos, and BK-Exos detected using the laser speckle blood flow monitoring system at 0 and 28 days after ischemia. G, Data analysis of blood perfusion in sham mice (Sham), hindlimb ischemic mice treated with PBS (PBS), PBS-Exos, and BK-Exos 28 days after ischemia (n = 7 in each group). H, Data analysis of limb salvage, foot necrosis, and limb loss in sham mice (Sham), hindlimb ischemic mice treated with PBS (PBS), PBS-Exos, BK-Exos 28 days after ischemia (n = 7 in each group). I, Representative fluorescence immunostaining images by using anti-CD31. In each group, the hindlimb ischemic mice were treated with PBS, PBS-Exos, and BK-Exos for 28 days after ischemia and the left hindlimb was served as sham group, (scale bar = 125 μm). J, Quantification of capillary density (CD31 + counts/mm 2 ) by CD31 immunostaining in sham, PBS, PBS-Exos, and BK-Exos group at day 28 after hindlimb ischemia. (n = 5 in each group) effects on various stem cells in vitro; BK induces the expression of alpha-smooth muscle actin in human mesenchymal stem cells [35], promotes neuro-generative division of neural progenitor cells [36], and inhibits oxidative stressinduced senescence [37]. BK has also been employed as a protective factor for the pre-treatment of human endothelial progenitor cells to improve MI [38,39]. Nonetheless, the effect of BK on c-kit + cells has not been investigated, and the roles and mechanisms of BK-c-kit + -derived exosomes in angiogenic activity have not been reported previously.
The present study aimed to explore the potential proangiogenic role of paracrine c-kit + cells and may provide [30]. Several reports have suggested that myocardial tissue secretes specific exosomes that are involved in cell-to-cell communication in the adult heart and that common drugs used in cardiac patients may enhance the release of exosomes [21,[31][32][33][34].
Our previous study showed that despite the poor longterm survival of transplanted c-kit + cells, BK-pretreated c-kit + cells (BK-c-kit + ) could significantly prolong the therapeutic benefits in MI rats, suggesting that BK-c-kit + cells may exert their beneficial effects through paracrine functions. BK is a linear nonpeptide messenger that belongs to the kinin group of proteins and exhibits different positive There is emerging evidence that exosome-based therapies effectively treat various diseases, including ischemic diseases such as MI [29,40,41] and hindlimb ischemia [17,18]. Next, we investigated whether the primary regulators are present in exosomes. We first isolated exosomes from the CM of BK-treated or untreated c-kit + cells and compared their angiogenic effects with those of exosome-depleted CM in HUVECs and the ischemic hindlimb of mice. Interestingly, the results showed that after BK treatment, exosomes instead of exosome-depleted CM significantly promoted tube formation in HUVECs and rescued hindlimb ischemia, indicating that BK promotes angiogenesis by stimulating exosome secretion from c-kit + cells.
an essential approach to the treatment of heart failure by improving the microvascular environment of the heart. In this study, the vascular tube-forming effect of c-kit + -CM in HUVECs was significantly higher than that of the negative control. Moreover, the pro-angiogenic effect of c-kit + -CM was enhanced by BK treatment (BK-CM). The pro-angiogenic effect of PBS-CM and BK-CM repaired hindlimb ischemia in mice in vivo. These results indicate that paracrine signaling may play a critical role in the therapeutic effects of BK-treated c-kit + cells. Nonetheless, the primary regulator of the pro-angiogenic action of CM in c-kit + cells requires further exploration. . F, Representative cell cycle images and data analysis of HUVECs transfected with control mimic (miNC) and miR-3059-5p mimic (miR3059) (n = 5). G, Representative transwell images and data analysis of HUVECs transfected with control mimic (miNC) and miR-3059-5p mimic (miR3059) (n = 3). H, Representative tube formation images and data analysis of HUVECs transfected with control mimic (miNC) and miR-3059-5p mimic (miR3059) (n = 5). I, Representative western blot images of p-Akt, Akt, p-Erk1/2, Erk1/2, p-Smad2, Smad2, p-Smad3, and Smad3 in HUVECs transfected with control mimic (miNC) and miR-3059-5p mimic (miR3059) (n = 5). Full-length blots are presented in supplementary file 2: Figure. S2B. J, Representative data analysis of relative expression of p-Akt/Akt, p-Erk1/2/Erk1/2, p-Smad2/Smad2, and p-Smad3/Smad3 in HUVECs transfected with control mimic (miNC) and miR-3059-5p mimic (miR3059) (n = 3-5) It is well known that miRNAs exert their effects by directly regulating their mRNA targets [43]. Further experiments were performed to investigate the target of miR-3059-5p and the mechanism involved in angiogenesis. The results showed that the expression of TNFSF15, encoding the protein vascular endothelial growth inhibitor (VEGI), a predicted target of miR-3059-5p, was significantly decreased after overexpression of miR-3059-5p in HUVECs. This was also confirmed via a dual-luciferase reporter assay by cotransfecting the 3'UTR of TNFSF15 and miR-3059-5p. VEGI is a new member of the tumor necrosis factor family and is predominantly an endothelial cell-specific gene. Recombinant VEGI is a potent inhibitor of endothelial cell proliferation, angiogenesis, and tumor growth [44]. Diverse studies have reported that VEGI is an endogenous regulator of angiogenesis in cancer-related studies, such as in lymphatic endothelial cells [45], colon carcinomas [46], breast cancer [47], human osteosarcoma, and vascular endothelial cells [48]. Nonetheless, few studies have reported the effect of VEGI on cardiovascular-related disorders. In the present study, TNFSF15 was identified as a critical target of miR-3059-5p-enriched exosomes secreted from BK-treated c-kit + cells. We also confirmed the anti-angiogenic role of TNFSF15 by knockdown of TNFSF15 and detected cell cycle, cell viability, and tube formation. Finally, we verified the further mechanism of TNFSF15 downregulation on angiogenesis by regulation of the phosphorylation of Akt, Erk1/2, and Smad2/3. As a shuttle vehicle between cells, exosomes play an indispensable role in transporting functional molecules such as proteins, mRNAs, miRNAs, or other non-coding RNAs to regulate gene expression in recipient cells [42]. Therefore, the objective of this study was to determine the crucial molecules that regulate the pro-angiogenic effect of exosomes in BK-treated c-kit + cells. We first detected angiogenesisrelated proteins using a proteome profiler human angiogenesis array kit, and found that no angiogenic-related protein was elevated in the exosomes of BK-treated c-kit + cells. In addition to proteins and mRNAs, miRNAs are one of the most active molecules in exosomes. They can be functionally delivered to target cells, resulting in the direct modulation of mRNA targets [27]. We used miRNA sequencing to explore miRNA expression in the exosomes of BK-treated or untreated c-kit + cells. The results of miRNA sequencing and consequent detection by qPCR showed that the expression of miR-3059-5p in the exosomes of BK-c-kit + cells increased more than 20-fold, indicating that miR-3059-5p might be a vital factor in the exosomes. miR-3059-5p is a new microRNA that has not been reported to have angiogenic effects or other physiological or pathological functions. We detected the potential role of miR-3059-5p in angiogenesis in HUVECs and ischemic hindlimb mice. After miR-3059-5p sponge transfection in c-kit + cells, the pro-angiogenic effects and the reparative function of BK-Exos in the ischemic limb were abolished, indicating that exosomal miR-3059-5p from BK-c-kit + cells is the key player in mediating angiogenesis within ischemic tissue. issues will help us to understand the molecular basis of the therapy of ischemic diseases.

Conclusions
In summary, our present study has demonstrated that exosomal miR-3059-5p from bradykinin-pretreated human cardiac-specific c-kit + cells is a key mediator in promoting angiogenesis and facilitating the repair of hindlimb ischemia in mice. Mechanistically we also revealed that miR-3059-5p in BK-Exos directly bind to the 3'UTR region of TNFSF15 and downregulates the expression of TNFSF15 through Akt/ There are several limitations to this study. First, in light of the fact that angiogenesis of ischemic tissue is a complex process regulated by producing a precise balance of growth and inhibitory factors, although we found much higher expression of miR-3059-5p in BK-exos, there are miRNAs that are also upregulated, it remains to be determined whether and how miR-3059-5p communicates and coordinates with other regulators to control angiogenesis of ischemic diseases. Second, our current study cannot address the angiogenesis regulation and functions of suppressed Tnfsf15 because of the lack of genetically modified rodent models with Tnfsf15 gene ablation. Further studies on these Fig. 6 The deficiency of VEGI exerted a pro-angiogenic effect in HUVECs. A, Representative tube formation images in HUVECs transfected with control siRNA (siNC) and VEGI siRNA (siVEGI). B, Data analysis of tube formation in HUVECs transfected with control siRNA (siNC) and VEGI siRNA (siVEGI) (n = 6). C, Representative cell cycle images of HUVECs transfected with control siRNA (siNC) and VEGI siRNA (siVEGI). D, Data analysis of cell cycle in HUVECs transfected with control siRNA (siNC) and VEGI siRNA (siVEGI) (n = 6). E, Representative transwell images of HUVECs transfected with control siRNA (siNC) and VEGI siRNA (siVEGI). F, Data analysis of transwell in HUVECs transfected with control siRNA (siNC) and VEGI siRNA (siVEGI) (n = 3). G, Representative western blot images of p-Akt, Akt, p-Erk1/2, Erk1/2, p-Smad2, Smad2, p-Smad3, and Smad3 in HUVECs transfected with control siRNA (siNC) and VEGI siRNA (siVEGI). Full-length blots are presented in supplementary file 2: Figure. S2D. H, Representative data analysis of relative expression of p-Akt/ Akt, p-Erk1/2/Erk1/2, p-Smad2/ Smad2, and p-Smad3/Smad3 in HUVECs transfected with control siRNA (siNC) and VEGI siRNA (siVEGI) (n = 3-6)