Exosomal let-7f-5p Derived from Mineralized Osteoblasts Promotes the Angiogenesis of Endothelial Cells via DUSP1 / Erk1/2 Signaling Pathway

Background: Angiogenesis is essential for the tissue engineering bone formation, and osteoblasts (OBs) has been proved to play an important role in angiogenesis via various pro-angiogenic factors. However, whether the mineralized osteoblast derived exosomes (MOB-Exos) and containing let-7f-5p can promote the angiogenesis of endothelial cells (ECs) is still unknown. Methods: MOB-Exos, let-7f-5p mimicked MOB-Exos (miR mimic group) and let-7f-5p inhibited MOB-Exos (miR inhibitor group) were respectively harvested from mineralized osteoblasts (MOBs) and then co-cultured with bEnd.3. Besides, the Erk1/2 signaling pathway in ECs in miR mimic group was inhibited. Subsequently, CCK-8 assays, wound healing assays, transwell migration assays and tube formation assays were performed to detect the angiogenic capability of ECs. Dual luciferase reporter assays were conducted to verify the target genes of exosomal let-7f-5p. Results: The results showed that MOB-Exos could signicantly promote the angiogenesis of ECs, which could be enhanced by mimicking exosomal let-7f-5p, and attenuated by inhibiting exosomal let-7f-5p. And the angiogenic capability of ECs was partly impaired after inhibiting the Erk1/2 signaling pathway despite co-cultured with let-7f-5p mimicked MOB-Exos. Moreover, let-7f-5p suppressed the luciferase activity of wide-type DUSP1, while mutation of DUSP1 abrogated the repressive ability of let-7f-5p. Conclusion: Based the results, our study concluded that exosomal let-7f-5p derived from MOBs could promote the angiogenesis of ECs via activating DUSP1/Erk1/2 signaling pathway, which might be a promising target for tissue engineering bone formation. tube representative statistical in ECs in miR mimic let-7f-5p


Introduction
Angiogenesis plays an extremely important role, and even has become the key to the success of tissue engineering bone formation [1]. Studies had shown that seeded cells within 200 µm from capillaries could obtain su cient nutrition through the diffusion of blood and interstitial uid [2]. Our previous study on ectopic osteogenesis of tissue-engineered bone found that the bone formation were extremely matched with the blood vessel growth, but most specimens in 12th week still exhibited incomplete osteogenesis and scaffold degradation in the center areas with the potential reason of lacking su cient blood vessel in-growth in the center of the grafts [3,4]. Thus, angiogenesis determined the speed and range of bone formation, and the tissue-engineered bone formation depended on the "angiogenesis-osteogenesis" coupling mechanism to a greater extent [5]. In view of the importance of angiogenesis, investigating the regulatory effects of osteoblasts (OBs) on endothelial cells (ECs) could not only deepen the understanding of "angiogenesis-osteogenesis" coupling mechanism, but also explore a promising way for promoting tissue-engineered bone formation and its clinical application.
There is an important synergistic effect between OBs and ECs through a variety of cytokines. Platelet derived growth factor BB (PDGF-BB) secreted by MC3T3-E1 could not only promote the osteogenic differentiation of OBs, but also enhance the angiogenic capability of ECs [6]. Besides that, many other factors, such as vascular endothelial growth factor (VEGF), angiopoietin-1 (Angpt-1) and broblast growth factor-2 (FGF-2), could be secreted by OBs and was proved of pro-angiogenic effects [7]. VEGF and insulin growth factor-1 (IGF-1) could promote the osteogenesis and angiogenesis of human carious dental pulp stem cells via AKT signaling pathway [8]. Zinc-nger transcription factor (ZEB1) secreted by ECs could reverse osteoporosis via triggering the synergy between angiogenesis and osteogenesis [9]. Therefore, OBs could synergize with ECs by various factors to ultimately promote the tissue-engineered bone formation through "angiogenesis-osteogenesis" coupling mechanism.
Besides the numerous cytokines, exosomes, microvesicles mostly secreted by eukaryotic cells with sizes ranged from 30-150nm in diameter [10], also participated in the osteogenesis and angiogenesis as important intercellular messengers. Exosomes derived from bone mesenchymal stem cells (BMSCs) could not only promote the osteogenic differentiation of MSCs [11], but also enhance the angiogenic capability of ECs [12], and then accelerated the bone formation in fracture site via simultaneously promoting osteogenesis and angiogenesis [13]. Additionally, miRNAs in exosomes, as the most important exosomal information carriers due to their de nite biological functions [14], were also closely associated with the osteogenesis and angiogenesis in different ways. Exosomal miR-935 could promote the osteogenic differentiation of OBs [15], while miR-23a-5p inhibited the osteogenic differentiation [16].
Exosomal miR-141 promoted the angiogenesis of lung cancer [17], while miR-6785-5p played an opposite role in gastric cancer [18]. However, BMSCs-derived exosomal miR-29a could not only promote the osteogenic differentiation of BMSCs, but also enhance the angiogenic capability of ECs [19]. Thus it can be seen that, exosomes and containing miRNAs could positively and negatively regulate the osteogenesis and angiogenesis, which indicated that some speci c exosomal miRNAs could be the potential factors for promoting the formation of tissue-engineered bone.
Our previous study demonstrated that mineralized osteoblasts (MOBs) derived exosomes (MOB-Exos) could promote the angiogenesis of ECs [20], and the MSCs derived exosome let-7f was inferred of proangiogenic effect by detected the expression after ECs were intervened by exosomes [12]. However, whether the MOBs derived exosomal let-7f-5p can promote the angiogenesis of ECs has not been studied or systematically veri ed. The present study was designed to investigated the regulatory effects of exosomal let-7f-5p derived from MOBs on angiogenic capability of ECs, expecting to provide a potential target for promoting tissue-engineered bone formation.

Materials And Methods
Cell culture, osteogenic induction and identi cation Wound healing assay ECs of each group were respectively resuspended with DMEM medium containing 10% FBS without exosomes and inoculated into wells of the 6-well plates in density of 5×10 5 cells/well. 3 duplicate were set for each group at each time point. ECs were cultured overnight and adherent, and then 200 μl pipette heads were used to draw lines vertically at the well bottom. After wash out the exfoliated cells with 1×PBS, ECs were cultured in exosome-free DMEM medium. The scratches were observed and photographed with inverted phase-contrast microscope at 0h, 24h, 48h and 72h after scribing. The scratch areas and scratch heights were measured in Image J software, and then the average migration widths of cells were calculated by using the formula: average migration width = (0 h scratch areascratch area at measurement time) / scratch area height.

Transwell migration assay
ECs of each group were respectively resuspended with pure DMEM medium to density of 5×10 5 cells/ml. 100 μl cell suspension was added into the upper chamber of Transwell chamber (Corning, USA), and 600 μl exosome free medium was added into the lower chamber. After cultured for 24 h, the medium and the cells in the upper chambers were wiped up. ECs migrating to the lower chambers were washed with ddH 2 O for three times, xed with methanol for 30 minutes, and stained with 0.1% crystal violet solution for 20 minutes after air drying. Inverted-phase contrast microscope was used to observe and take pictures. Five elds were randomly selected to count the number of migrated cells in Image J software.

Tube formation assay
ECs of each group were respectively resuspended with DMEM medium containing 10% FBS without exosomes to density of 5×10 5 cells/ml. Matrigel (Corning, USA), 96-well plates and 200 μl pipette tips were placed in 4 ℃ refrigerator for precooling overnight. 50 μl Matrigel was added to each well and incubated at 37 ℃ for 30 mins. Then 50 μl cell suspension was added to each well and cultured for 6 h.
The tubule like structures were observed under inverted phase-contrast microscope and photographed. The total segment lengths of tubule-like structures were measured in Image J software.
Western blotting analysis Western blotting was performed as previously described [20]. Brie y, cells and exosomes were lysed by RIPA buffer containing protease and phosphatase inhibitors (Sevenbiotech, China). The protein concentration of supernatant was quanti ed with BCA Protein Assay kit (Thermo, USA), and adjusted to the same with RIPA buffer for each group. Approximately 20 μg total protein of each sample were loaded, subjected to electrophoresis, and then transferred to PVDF membranes (Millipore, USA). Subsequently, the membranes were blocked with with 5% non-fat milk in TBST at room temperature for 2 h, and then respectively incubated with related primary antibodies for CD63 (Invitrogen, USA), CD81 (Invitrogen, USA), Erk1/2 (CST, USA), pErk1/2(Santa Cruz, USA), GAPDH(CST, USA), β-actin(CST, USA) at 4 ℃ overnight.
After that, the membranes were washed with TBST for 3 times and incubated with secondary antibodies (CST, USA) at room temperature for 1 h and visualized ECL detection kit (Bio-channel, China). Finally, the photographs of membranes were taken with Fluor Chem E system (Proteinsimple, USA), and analyzed in Image J software.
Dual luciferase reporter assays The dual luciferase reporter assays were performed with some modi cations. Brie y, wild-type and mutant 3′-UTRs of dual speci city phosphatase-1 (DUSP1) and dual speci city phosphatase-4 (DUSP4) were respectively ampli ed and cloned downstream of the luciferase gene within the pmirGLO vectors (Promega, USA), with the Renilla luciferase plasmid (Promega, USA) co-transfected as a control. After transfected with 1 μg vectors and subjected to 60 μM let-7f-5p mimics, the luciferase activities of the transformed cells were measured with Dual Luciferase Reporter assay kits (Promega, USA) at room temperature according to the manufacturer's instructions.

Statistical analysis
The experimental data were analyzed with graphpad prism 8 software. Unpaired t-test was used for two groups of independent measurement samples, one-way ANOVA test was used for multiple groups of independent measurement samples, and chi square test was used for counting samples. P < 0.05 and P < 0.01 were both regarded as statistical difference.

Identi cation of MOBs and MOB-Exos
Before exosome extraction, pre-osteoblasts MC3T3-E1 were osteogenically induced for 21 days and then subjected to ALP and ARS staining to identify the osteogenic differentiation of OBs. ALP staining showed that the intracellular expressed ALP increased gradually with time and signi cantly increased at the 21 st day (Fig. 1A). In addition, the calcium nodules deposited on the surface of OBs exhibited the similar trend in ARS staining (Fig. 1B). The staining demonstrated that pre-osteoblasts MC3T3-E1 were successfully induced to MOBs.
Microvesicles were extracted from MOBs and identi ed via analyzing particle size distribution, observing the shape and detecting the expression of biomarker CD63 and CD81. The diameters of extracted microvesicles detected by Nanosight light scattering technology ranged from 30.25 nm to 149.25nm with mean of 73.80±18.70 nm (Fig. 1C). And the microvesicles exhibited "saucer"-like shape and diameters less than 200nm in transmission electron microscope scanning (Fig. 1D). Besides, high expression of biomarkers CD63 and CD81 was determined by Western blot and NanoFCM (Fig. 1 E, F and G). The results above showed that the extracted microvesicles exhibited the typical characteristics of exomes and so demonstrated that the microvesicles extracted from MOBs were exosomes.

MOB-Exos enhanced the angiogenic capability of ECs
In order to regulate the function of ECs, exosomes needed to be absorbed by ECs. PKH-67 immuno uorescence staining was performed to examine the uptake of MOB-Exos, and showed that the green uorescent labeled MOB-Exos were absorbed by ECs, wrapped the blue uorescent labeled nucleus and evenly distributed in the cytoplasm of ECs ( Fig. 2A). On the basis of verifying that MOB-Exos could be absorbed by ECs, the regulatory effects of MOB-Exos on ECs were investigated by respectively detecting the proliferation, migration and tube formation of ECs after intervened with equal volume of PBS (control group) and MOB-Exos (MOB-Exos group). CCK-8 showed that the proliferation of ECs treated with MOB-Exos signi cantly increased compared to the control group at 48 h and 72h (Fig. 2E). The migration of ECs detected by wound healing assays and transwell migration assays exhibited remarkable increase after intervened with MOB-Exos (Fig. 2B, C, F and G). And also, the enhancement of tubule forming ability of ECs was con rmed by tube formation assays in MOB-Exos group (Fig. 2D, H). The results above demonstrated that MOB-Exos could be absorbed by ECs and then remarkably promoted the proliferation, migration and tube formation of ECs.
let-7f-5p in MOB-Exos was crucial for promoting the angiogenesis of ECs Based on the promoting effects of MOB-Exos on the angiogenesis of ECs, qPCR was performed to detect whether exosomal let-7f-5p participated in enhancing the angiogenic capability of ECs. The let-7f-5p expressing in ECs increased in several folds after co-culture of ECs with MOBs. While, prior addition of the EV secretion inhibitor GW4896 to MOBs blocked the exosome production and delivery of let-7f-5p from MOBs to ECs (Fig. 3F), which revealed that MOBs transported extracellular let-7f-5p into ECs via an exosome-dependent manner. In order to systematically verify the effects of let-7f-5p, a mimic-inhibiting system was established by extracting exosomes from MOBs pretreated with let-7f-5p mimics and inhibitors (miR mimic and inhibitor groups). Cy3 immuno uorescence staining was performed and found that Cy3 labeled let-7f-5p mimics and inhibitors evenly distributed in MOBs (Fig. 3A), revealing the successfully transfecting mimics and inhibitors into MOBs. Afterward, let-7f-5p expression in relevant MOB-Exos was determined by qPCR, which con rmed that let-7f-5p in MOB-Exos in miR mimic and inhibitor groups was remarkably increased and decreased by let-7f-5p mimics and inhibitors respectively (Fig. 3G), revealing the success and effectiveness of the mimic-inhibiting system.
The role of let-7f-5p in MOB-Exos promoting angiogenesis of ECs was systematically veri ed via contrastively investigating the angiogenic ability of ECs regulated by MOB-Exos (MOB-Exos group) and let-7f-5p mimicked and inhibited MOB-Exos (miR mimic and inhibitor groups). CCK-8 showed that the proliferation of ECs was signi cantly increased in miR mimic group compared to that in MOB-Exos group, while which was signi cantly decreased in miR inhibitor group (Fig. 3E). As was shown in Fig. 3B, D, H and I, wound healing assays and transwell migration assays revealed that the migrating ability of ECs was markedly enhanced by mimicking let-7f-5p and weakened by inhibiting let-7f-5p. In addition, the total segment lengths of tubule-like structures formed in miR mimic group were signi cant higher than those in MOB-Exos group, which were conversely lower in miR inhibitor group (Fig. 3C, J). These functional assays con rmed that let-7f-5p played a positive regulatory role in MOB-Exos promoting the angiogenesis of ECs.
MOB-Exo let-7f-5p promoted the angiogenesis of ECs via activating Erk1/2 signaling pathway MOB-Exo let-7f-5p was proved of positive regulatory effect on angiogenesis of ECs, but whether Erk1/2 signaling pathway participated in the process was unde ned. Therefore, western blotting assays were performed to detect the pErk1/2 expression in ECs in MOB-Exos group, miR mimic group and inhibitor group, and found that the pErk1/2 expression could be remarkably increased by mimicking let-7f-5p and conversely decreased by inhibiting let-7f-5p (Fig. 4D, E), revealing the important role of Erk1/2 signaling pathway in MOB-Exo let-7f-5p promoting angiogenesis of ECs. To verify the effects, Erk1/2 signaling pathway in ECs in miR mimic group was inhibited by PD98059 (Erk inhibitor group), and the inhibiting effectiveness was con rmed by that the relevant pErk1/2 signi cantly decreased compared to that in miR mimic group (Fig. 4D, E). The proliferation of ECs in Erk inhibitor group was signi cantly weaker than that in miR mimic group, but greater than that in MOB-Exos group (Fig. 4F). The enhancement of MOB-Exo let-7f-5p on migrating ability of ECs was alleviated by inhibiting the Erk1/2 signaling pathway in ECs in miR mimic group (Fig. 4A, B, G and H). And also the tubule-like structures formed by ECs remarkably decreased in Erk inhibitor group, but still was more than that in MOB-Exos group (Fig. 4C, I) . Thus it could be seen that inhibiting the Erk1/2 signaling pathway in ECs in miR mimic group could partially inhibit MOB-Exo let-7f-5p from promoting the angiogenesis of ECs, which demonstrated that Erk1/2 signaling pathway was the main but not the exclusive channel for the positive regulatory effects of MOB-Exo let-7f-5p.
MOB-Exo let-7f-5p activated Erk1/2 signaling pathway via suppressing DUSP1 It was proved that MOB-Exo let-7f-5p promoted the angiogenesis of ECs via Erk1/2 signaling pathway, but the target genes by which MOB-Exo let-7f-5p activated Erk1/2 signaling pathway were still unde ned. After retrieving TargetScan, PicTar, miRanda and microT databases, 198 shared target genes were obtained via Venn analysis (Fig. 5A). GO database analysis found that 12 target genes were involved in the regulation of mitogen-activated protein kinase (MAPK) signaling pathway, of which 10 target genes except DUSP1 and DUSP4 positively regulated MAPK signaling pathway (Fig. 5B). And DUSP1 and DUSP4 enriched in Mitogen-activated protein kinase phosphatase (MKP) site of Erk1/2 signaling pathway (Fig. 5C). Based on above, DUSP1 and DUSP4 were supposed most likely to be the target genes of let-7f-5p.
To verify the presumption, qPCR and dual luciferase reporter assays were performed. The results showed that mimicking exosomal let-7f-5p signi cantly suppressed the expression of DUSP1 in ECs, which was conversely increased by inhibiting exosomal let-7f-5p (Fig. 5D). let-7f-5p mimics suppressed the luciferase activity of cells transfected with vectors of wide-type DUSP1 by 39%, while mutation of let-7f-5p seeding region within DUSP1 abrogated the repressive ability of let-7f-5p (Fig. 5E), con rming the speci city of target sequence for DUSP1. Besides, DUSP4 was detected and con rmed to be unregulated by let-7f-5p (Fig. 5F, G).

Discussion
As important structures for supplying nutrition and removing metabolic wastes, blood vessels are crucial for maintaining the normal physiological function of tissues. While angiogenesis is an important physiological process to maintain vascular stability and involved in various pathophysiological processes, such as wound healing, in ammatory response, immune response, tumor growth, invasion and metastasis and so on [21][22][23]. There had been a close synergy between angiogenesis and bone formation: angiogenesis could promote the continuous differentiation of BMSCs into OBs, and then participate in the bone formation and homeostasis maintenance [24]; OBs could secrete VEGF, angpt-1, FGF-2 and some other angiogenic factors to enhance the angiogenic ability of ECs [7]. In view of the key role of angiogenesis in the formation of tissue-engineered bone [1], this "angiogenesis-osteogenesis" coupling mechanism might provide a new potential way to promote the formation of tissue-engineered bone.
Exosomes, as an important medium of intercellular information transmission, were proved of an important role in "angiogenesis-osteogenesis" coupling mechanism by which exosomes could promote fracture healing by accelerating angiogenesis and bone formation [13]. However, the functions of exosomes were closely related to the source cell types and pathophysiological states. Exosomes derived from MSCs and ovarian cancer cells had been shown to promote angiogenesis of ECs respectively through containing miRNAs [12] or soluble cadherin on cell membrane [25]. While the effects of exosomes derived from OBs and osteosarcoma cells on T cells were different: former exhibited relatively weaker inhibition on T cell proliferation activity, and had no effect on promoting T cell regulatory phenotype compared to the latter[26]. And exosomes derived from OBs in patients with hip osteoarthritis could signi cantly inhibit the activity and osteogenic differentiation of BMSCs compared to those in patients with osteoporosis [27]. Therefore, although exosomes derived from BMSCs and MSCs were proved of proangiogenic effects on ECs [12,13], whether these effects existed in exosomes derived from MOBs were unde ned. Our current and previous [20] studies demonstrated that exosomes derived from MOBs could promote the proliferation, migration and tube formation of ECs, which revealed that MOB-Exos might be the potential factors for accelerating tissue engineering bone formation by promoting angiogenesis.
miRNAs contained in exosomes has become hot targets for disease diagnosis and treatment for their signi cant sensitivity and speci city in the process of activating signaling pathways to regulate cell functions and phenotypes [28,29]. The effects of miRNAs in OBs derived exosomes on osteogenic differentiation were con rmed to be different: let-7, miRNA-335-5p and miRNA-378b promoted osteogenic differentiation of OBS, while miRNA-30d-5p, miRNA-133b-3p and miRNA-140-3p exhibited the opposite effects [30]. However, there was no relative research on MOB-Exo miRNAs regulating angiogenesis of ECs.
let-7f-5p, as a member of let-7 family, had been proved closely relating to immune response, in ammatory response and cell differentiation, for which it had become a hot target for early diagnosis and treatment of a variety of diseases [31][32][33][34]. Studies had shown that let-7f-5p could effectively reverse the inhibition of angiogenesis induced by smoking [35], and the plasma derived exosomal let-7f-5p had signi cant sensitivity and speci city in this process [36]. Current study demonstrated the exosome-dependent manner of MOBs transporting let-7f-5p into ECs, indicating the positive regulatory effects of MOBs derived exosomal let-7f-5p on enhancing the angiogenesis ability of ECs, which were con rmed by our further experiments.
Erk1/2 signaling pathway, as the most important pathway in MAPK signaling pathway, was closely related to cell proliferation, migration, apoptosis, skeleton formation and maintenance [37], and had been considered as an important target for the treatment of malignant tumors[38] for its close relationship with angiogenesis [39,40]. It had been demonstrated that exosomal miRNAs regulating angiogenesis was also closely associated with Erk1/2 signaling pathway: exosomal miRNA-21 from adipose-derived stem cells promoted angiogenesis by signi cantly up-regulating the expression of pErk1/2 [41], and BMSCs derived exosomal miRNA-21-5p promoted angiogenesis by inhibiting spry2 gene expression and then activating Erk1/2 signaling pathway [42]. Current study found that mimicking let-7f-5p of MOB-Exos could promote the expression of pErk1/2, while inhibiting let-7f-5p obtained opposite results, indicating that Erk1/2 signaling pathway played an important role in MOB exosomal let-7f-5p promoting angiogenesis of ECs, which was further con rmed by comparative studies via inhibiting the Erk1/2 signaling pathway in let-7f-5p mimic group. These results were also supported by the important effects of Erk1/2 signaling pathway on exosomal miRNAs regulating angiogenesis.
miRNAs could not directly activate Erk1/2 signaling pathway, but by reducing the target mRNA stability or promoting its degradation, thus inhibiting the expression of target genes in protein translation level [43]. DUSPs could inactivate MAPK by dephosphorylating threonine and tyrosine at T-X-Y site of MAPK kinase domain, thus antagonizing cell signaling cascades [44]. DUSP1 and DUSP4, as members of DUSPs family, negatively regulated Erk1/2 signaling pathway [45,46], which was closely associated with some miRNAs' function: miRNA-101 in the ventrolateral orbital cortex could improve the depression-like behavior of rats by inhibiting the expression of DUSP1 to activate the downstream Erk1/2 signaling pathway[47]; miRNA-122-5p promoted the expression of DUSP4 and inhibited Erk1/2 signaling pathway in pulmonary microvascular ECs, which was closely related to acute lung injury [48].Our study found that let-7f-5p could regulate DUSP1, but not DUSP4 expression, and con rmed DUSP1 was the target gene of let-7f-5p activating Erk1/2 signaling pathway. However, our study didn't investigate the effects of DUSP1 on pErk1/2 expression or the angiogenic capability of ECs.

Conclusion
Our ndings demonstrated that exosomal let-7f-5p derived from MOBs could signi cantly promote the angiogenesis of ECs, and the underlying mechanism might be let-7f-5p activated Erk1/2 signaling pathway by suppressing the expression of DUSP1. MOBs derived exosomal let-7f-5p might be a promising target for tissue engineering bone formation. Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
47. Zhao Y, Wang S, Chu Z, Dang Y, Zhu J, Su X. MicroRNA-101 in the ventrolateral orbital cortex (VLO) modulates depressive-like behaviors in rats and targets dual-speci city phosphatase 1 (DUSP1  Figure 1 Identi cation of MOBs and MOB-Exos. ALP staining (A) and ARS staining (B) respectively showed the gradual increase of ALP expressed intracellularly and calcium nodules deposited on the surface of OBs, con rming the successful osteogenic induction of OBs to MOBs. The extracted microvesicles exhibited characteristics with the sizes ranged from 30.25 nm to 149.25nm in diameters according to Nanosight light scattering technology (C), "saucer"-like shape and diameters less than 200nm in transmission electron microscope scanning (D), and high expression of biomarkers CD63 and CD81 detected by Western blot (E) and NanoFCM (F, G), which identi ed the microvesicles as exosomes.  let-7f-5p in MOB-Exos was crucial for promoting the angiogenesis of ECs. After co-cultured with MOBs, ECs exhibited high expression of let-7f-5p, which was inhibited after co-cultured with MOBs pretreated with the EV secretion inhibitor GW4896 (F). Cy3 immuno uorescence staining showed that Cy3 labeled let-7f-5p mimics and inhibitors evenly distributed in MOBs (A). qPCR con rmed that let-7f-5p in MOB-Exos in miR mimic and inhibitor groups was remarkably increased and decreased by let-7f-5p mimics and inhibitors respectively (G). As was shown in the representative images  MOB-Exo let-7f-5p promoted the angiogenesis of ECs via activating Erk1/2 signaling pathway. Western blotting assays showed that the pErk1/2 expression could be remarkably increased by mimicking MOB-Exo let-7f-5p and conversely decreased by inhibiting the let-7f-5p, and that pErk1/2 expression signi cantly decreased after Erk1/2 signaling pathway in ECs in miR mimic group was inhibited by PD98059 (D, E). The migration, tube formation and proliferation of ECs were respectively detected and described as representative images (A,B,C) and statistical results (F,G,H,I), which demonstrated that inhibiting the Erk1/2 signaling pathway in ECs in miR mimic group could partially inhibit MOB-Exo let-7f-5p from promoting the angiogenesis of ECs. *P<0.05, **P<0.01, &P<0.05, &&P<0.01, #P<0.05.