The endothelial cell line bEnd.3 and the OB cell line MC3T3-E1, derived from Mus musculus (mouse), were obtained from the cell bank of the Chinese Academy of Sciences. Both cell lines underwent authentication via STR profiling and were confirmed negative for mycoplasma contamination. bEnd.3 cells were maintained in Dulbecco’s Modified Eagle Medium (DMEM; Cat SH30022.01, Hyclone, Salt Lake City, UT, USA) supplemented with 10% fetal bovine serum (FBS, Cat 10437-028; Gibco, Los Angeles, CA, USA). MC3T3-E1 cells were cultured in Minimum Essential Medium $\alpha$ (MEM $\alpha$, Cat SH30265.01, Hyclone, Salt Lake City, UT, USA), supplemented with 10 mM $\beta$-sodium glycerophosphate (Cat C2226462, Aladdin, Shanghai, China) and 50 mg/L L-ascorbic acid (Cat SLBX2310, Sigma, St. Louis, MO, USA) for 21 d to induce osteogenic differentiation. Both cell lines were incubated in a humidified atmosphere at 37 °C with 5% CO${}_2$. Synthetic miR-423-5p mimics, inhibitors, and negative controls (NC) were custom synthesized by Ribobio (Guangzhou, China). Transfection was carried out using RiboFect™ as described by the manufacturer (C10511-05, Ribobio, Guangzhou, China).

Three experimental groups were established: a blank control, MOBs alone, and MOBs treated with GW4869, an inhibitor of exosome release (Cat: HY-19363, MedChemExpress, Lawrenceville, NJ, USA). In all three groups, MC3T3-E1 cells were cultured in six-well plates for 24 h. GW4869 (10 µM) was added to the MOBs+GW4869 group and the cells incubated overnight. Transwell chambers with a pore size of 0.4 µm (Cat 3422, Corning, Corning, NY, USA) were then added to the wells and ECs were seeded into the upper compartment of the chamber. After two days of co-culture, the ECs were harvested by trypsin digestion (Beyotime, Nanjing, China) and RNA was extracted from the cell pellets to evaluate the miR-423-5p content.

After induction for 21 d, the osteogenic medium was replaced by DMEM containing 10% exosome-free FBS and the cells cultured for a further 2 d. MOB-Exos were then isolated using Exoquick reagent (Cat 4478359, Invitrogen, Waltham, MA, USA). The culture medium was initially centrifuged at 3000 $\times$g for 30 min in a centrifuge tube (Corning, USA). Subsequently, the supernatant was transferred to a 100 kDa ultrafiltration tube (Cat UFC9100, Millipore, Burlington, MA, USA) and centrifuged at 4000 $\times$g for 30 min to concentrate the exosomes. The concentrated solution was then mixed with exosome isolation reagent, incubated at 4 °C overnight, and centrifuged at 10,000 $\times$g for 1 h. After discarding the supernatant, the purified exosomes were resuspended in 1$\times$PBS (Cat G4202, Servicebio, Guangzhou, China) and the protein content determined using a BCA Protein Assay Kit (Cat 23225, Thermo, Waltham, MA, USA). An exosome concentration of 30 µg/mL was utilized for subsequent experiments in this study.

The detection of nanoparticles was carried out using a nanoparticle tracking analysis instrument (Particle Metrix, Munich, Germany). A standard solution of 100 nm polystyrene microspheres (Thermo, Waltham, MA, USA) was diluted 250,000-fold with ultrapure water, and 1 mL of this diluted standard solution was then used to automatically calibrate the instrument prior to testing. Exosome samples were diluted 500-fold to 1.0 $\times$ 10${}^8$ particles/mL with a 1$\times$PBS solution, giving an average count of 30 particles per frame. The size and percentage of exosomes was then evaluated.

Exosome samples were adjusted to the appropriate concentration or viscosity. Approximately 15 µL of exosome sample was absorbed onto copper mesh and left for 1 h. They were then dried with filter paper and stained with 15 µL of a 2% uranyl acetate solution (Thermo, Waltham, MA, USA). The stained exosomes were subsequently observed with a transmission electron microscope (JEM-1400, JEOL, Tokyo, Japan).

MOB-Exos (200 µg) were mixed with PKH-67 dye (Cat MKCK8884, Sigma, St. Louis, MO, USA) at room temperature. A 1% BSA solution was used to stop the labeling procedure, and the stained exosomes were then added to EC culture medium. After incubation for 4 h, ECs were fixed with 4% neutral paraformaldehyde, stained with 4’,6-diamidino-2-phenylindole (DAPI), and the images saved and merged by ImageJ for Mac 1.51 (https://imagej.net/ij/).

Cell proliferation was assessed using the Cell Counting Kit-8 (CCK-8; Cat CK04, Dojindo, Sendai-shi, Japan) assay. Each experimental group of bEnd.3 cells was seeded into 96-well plates at 2 $\times$ 10${}^3$ cells per well and in three replicates. The proliferation of ECs at 0, 24, 48 and 72 h was evaluated by light absorption at 450 nm using a Tecan Infinity 200 Pro-multi-well plate reader (Tecan Group Ltd., Männedorf, Switzerland).

bEnd.3 cells were inoculated into 6-well plates at a density of 5 $\times$ 10${}^5$ cells per well. A scratch wound was created in the cell monolayer using a 200 µL pipette tip, and cells were subsequently cultured in serum-free medium. After 0 and 48 h, the wound was photographed and analyzed by ImageJ software. The percentage of migration area = $\frac{\text{scratch area at }0\mathrm{h}-\text{scratch area at }48\mathrm{h}}{\text{scratch area at }0\mathrm{h}}\times 100\%$.

bEnd.3 cells were first diluted to 1 $\times$ 10${}^5$ cells/mL, and 100 µL was then seeded into the upper chamber of a 24-well transwell insert (Corning, Corning, NY, USA). The lower chamber was filled with culture medium. After 24 h incubation, cells that had migrated through the membrane to the lower side were fixed for 15 min and then stained with a 1% crystal violet solution (Cat Co121, Beyotime, Nanjing, China). Stained cells were observed by microscopy (Shinjuku-ku, Tokyo, Japan) and images were captured for analysis using ImageJ software version 1.51 for Mac.

A pipette tip and Matrigel were placed in a 4 °C refrigerator the day before the experiment. The Matrigel was used to coat 96-well plates. bEnd.3 cells were diluted to 4 $\times$ 10${}^5$ cells/mL and 50 µL of this suspension was added vertically to solidified Matrigel while avoiding bubbles. Following incubation for 6 h, the structure of tube segment was observed by microscopy and quantified by ImageJ for Mac 1.51.

After cell lysis with Trizol reagent (Beyotime, Nanjing, China), samples were mixed with chloroform at a 1:5 ratio and then shaken vigorously for 15 sec. Following centrifugation at 12,000 rpm for 10 min, the upper aqueous phase containing RNA was transferred to a new tube and mixed with an equal volume of isopropyl alcohol. After the same centrifugation process, the white RNA precipitate at the bottom of the tube was washed with 75% alcohol. The mRNA was reverse transcribed using TB Green® Premix Ex Taq™ II (Cat RR820L, TakaraBio, Takara Bio Inc., Kyoto, Japan), while miRNA was reverse transcribed using the miRcute miRNA First-strand cDNA Synthesis Kit (KR211, Tiangen, Beijing, China). PCR conditions for mRNA amplification were as follows: initial denaturation at 95 °C for 30 sec, followed by 40 cycles of 95 °C for 45 sec and 60 °C for 34 sec. For miRNA amplification, PCR conditions included an initial denaturation step at 95 °C for 15 min, followed by 40 cycles of 94 °C for 20 sec and 60 °C for 34 sec. Relative expression levels were calculated using the 2${}^{-\mathrm{\Delta }\mathrm{\Delta }\text{Ct}}$ method. Primer sequences are shown in Table 1.

Proteins from cells and exosomes were extracted with protease and phosphatase inhibitors (Servicebio, Guangzhou, China). The protein solution was mixed with loading buffer and heated to 100 °C for 20 min, separated by sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and then transferred onto polyvinylidene difluoride membranes (Millipore, Burlington, MA, USA). After blocking, membranes were incubated overnight at 4 °C with primary antibodies against Glyceraldehyde 3-phosphate dehydrogenase (GAPDH, 1:10,000 dilution; cat. AC002; ABclonal, Waltham, MA, USA), CD63 (1:1000 dilution; cat. no. 10112; CST, Danvers, MA, USA), CD81 (1:1000 dilution; cat. 10037; CST, Danvers, MA, USA), vascular endothelial growth factor (VEGF, 1:1000 dilution; cat. no. A12303; ABclonal, Waltham, MA, USA), or C-X-C motif chemokine ligand 10 (CXCL10, 1:500 dilution; cat. 10937-1-AP; Proteintech, Chicago, IL, USA). Subsequently, the membranes were washed and incubated with secondary antibody (Servicebio, Guangzhou, China) for 1 h. Protein bands were visualized using an electrochemiluminescence (ECL) detection kit (Beyotime, Nanjing, China). Image analysis was performed using ImageJ software version 1.51 for Mac to quantify the gray values of protein blots, with GAPDH serving as the internal reference. GraphPad Prism 9.0 (https://www.graphpad.com/updates/prism-900-release-notes) was used to normalize the relative intensity of target protein bands in the experimental groups against the average intensity observed in the control group.

To identify the predicted direct binding site, a plasmid containing either wild-type CXCL10 3’UTR or a mutant form was co-transfected into HEK293T cells, along with either a miR-423-5p mimic or a scrambled negative control (NC). Luciferase activity in the transfected cells was subsequently quantified using a Dual Luciferase Reporter Assay System (E1910, Promega, Madison, WI, USA). The complete binding site is shown in the Supplementary Material.

Animal experiments were authorized by the Committee for the Ethics of Animal Experiments, East China Normal University (Approval No. m20210910). Experimental animals were obtained from the Experimental Animal Center, Minhang Campus, East China Normal University. MOBs were transfected with miR-423-5p mimics, inhibitors, or negative control (NC), and the exosomes subsequently isolated. Matrigel was then mixed with ECs that had been pretreated with exosomes derived from the different experimental groups. Each experimental group consisted of 5 female nude mice (BALB/c strain, age 7–8 weeks). These were injected with the Matrigel mixture into the dorsal region after anesthetization with 1% pentobarbital sodium (200 mg/kg). After two weeks, mice were euthanized by placing them in a euthanasia box with carbon dioxide until the cessation of vital signs, as observed by the absence of toe reflex and respiratory and cardiac activity. The Matrigel plugs were collected and analyzed by immunofluorescence and immunohistochemistry staining.

RNA-seq analysis was conducted by Lianchuan Biotechnology (Hangzhou, China). Briefly, RNA extracted from bEnd.3 cells underwent quality assessment using a Bioanalyzer 2100 instrument (Agilent Technologies, Santa Clara, CA, USA). Only those samples with a RIN value $>$7.0 were selected for construction of the sequencing library. Original reads obtained from the sequencer included adapters or low-quality bases that could adversely affect subsequent assembly and analysis. They were filtered further by Cutadapt (https://cutadapt.readthedocs.io/en/stable/) to obtain a high-quality, clean read that was subsequently compared to the reference genome using HISAT2 software (https://daehwankimlab.github.io/hisat2/). This removed part of the read according to the quality of information attached to each read, and then mapped it to the reference genome. The analysis of differential gene expression between two groups was performed using DESeq2 software (https://bioconductor.org/packages/release/bioc/html/DESeq2.html), with a significance threshold of p 0.05. Gene Ontology (GO) enrichment analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis were also performed to gain biological insights from the data.

Statistical analyses were conducted using GraphPad Prism 9.0 (https://www.graphpad.com/updates/prism-900-release-notes). Student’s two-tailed t-test was performed to evaluate the difference between two independent samples, while one-way ANOVA was used to evaluate differences between multiple groups. Two-way ANOVA was used to analyze two independent variables. All data are presented as the mean $\pm$ standard deviation (SD). Results were considered to be significantly different at p 0.05.

Exosomes were extracted from the MC3T3-E1 cell line following osteogenic induction for 21 d. The presence of mature MOBs was confirmed by Alizarin Red S staining (Fig. 1A) and ALP staining (Fig. 1B). Transmission electron microscopy revealed the characteristic hollow spherical morphology typical of exosomes (Fig. 1C), consistent with previous reports. Nanoparticle tracking analysis further characterized MOB-Exos and revealed a particle diameter ranging from 88.6 to 195.0 nm, with an average of 139.4 nm and a peak at 120.6 nm (Fig. 1D). Western blotting confirmed the presence of exosome-specific markers CD63 and CD81 (Fig. 1E), thus verifying the successful isolation of exosomes from MOBs.

Fig. 1.

Identification of exosomes derived from MOBs. (A) ARS staining, and (B) ALP staining showed successful osteogenic induction of MC3T3 cells (scale bar: 200 µm) (C) Transmission electron microscopy images showing exosomes (scale bar: 100 nm). (D) Nanoparticle tracking analysis showing exosome size distribution. (E) Western blot analysis showing the expression of exosomal makers CD63 and CD81 in MOB-Exos. MOBs, Mineralized osteoblasts; ARS, Alizarin Red S; ALP, Alkaline Phosphatase; MOB-Exos, exosomes derived from MOBs.

MOB-Exos were labeled with PKH67 and cultured with ECs to study their internalization. As shown in Fig. 2A, the strong green fluorescent signal surrounding the nucleus indicated internalization of PKH67-labeled MOB-Exos. A series of functional assays was performed to evaluate the effect of MOB-Exos on vascularization. Results from the CCK-8 assay demonstrated that MOB-Exos significantly promoted EC proliferation compared to the control group treated with an equal volume of 1$\times$PBS (Fig. 2B). In addition, the scratch wound assay (Fig. 2C) and Transwell migration assay (Fig. 2E) revealed that MOB-Exos enhanced the migration of ECs (Fig. 2F,G). Furthermore, the tube formation assay (Fig. 2D) showed that incubation with MOB-Exos increased capillary-like structure formation by ECs (Fig. 2H). Together, the above results demonstrate that MOB-Exos enhance EC functions that are critical for vascular regeneration and angiogenesis.

Fig. 2.

MOB-Exos promoted endothelial cell (EC) proliferation, migration, and tube formation. (A) Internalization of PKH67-labelled MOB-Exos by ECs (scale bar: 100 µm). (B) The proliferation rate of ECs as determined by the Cell Counting Kit-8 (CCK-8) assay. (C) Scratch wound (scale bar: 500 µm) and (E) Transwell assays (scale bar: 100 µm) were performed to evaluate the migration of ECs. (D) Tube structure formation assay performed on ECs (scale bar: 200 µm). Results are presented as the mean $\pm$ SD of three independent experiments (F,G,H). *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001.

Exosomes are extracellular vesicles containing bioactive substances that participate in the process of cell-cell communication. miR-423-5p was highly expressed in MOB-Exos (p 0.001) (Fig. 3A). A cell co-culture experiment was conducted to investigate whether miR-423-5p was delivered to ECs via the exosomal pathway. This experiment included three groups: control, MOBs, and MOBs+GW4869. The miR-423-5p content of ECs was markedly increased in the MOBs co-culture group (p 0.0001) compared to the control group (Fig. 3B). However, addition of the exosome release inhibitor GW4869 attenuated this increase (p 0.05), indicating that miR-423-5p was transferred to ECs by MOB-Exos.

Fig. 3.

Exosomal miR-423-5p promoted the angiogenic ability of ECs. (A) Quantitative polymerase chain reaction (qPCR) results showing an increased level of miR-423-5p in MOB-Exos. (B) ECs were co-cultured with MOBs and the miR-423-5p content of the ECs was subsequently evaluated in different experimental groups. (C) Successful establishment of the mimic and inhibitor systems. (D) CCK-8, (E) scratch wound (scale bar: 500 µm), (F) tube formation (scale bar: 200 µm), and (G) Transwell migration assays (scale bar: 100 µm) were performed to assess the angiogenic ability of ECs. Results are presented as the mean $\pm$ SD of three independent experiments (H,I,J). ns: no significance, *p 0.05, **p 0.01, ***p 0.001, ****p 0.0001, ${}^{\mathrm{\#}}$p 0.05, ${}^{\mathrm{\#}\mathrm{\#}\mathrm{\#}\mathrm{\#}}$p 0.0001, p 0.001, p 0.0001. NC, negative controls; miR, miR-423-5p.

The specific impact of MOB-Exos-derived miR-423-5p on EC angiogenesis was further investigated to exclude the possible impact of other exosome components. ECs were transfected with miR-423-5p mimic, negative control (NC), inhibitor, or inhibitor NC. Subsequent qPCR confirmed the successful modulation of miR-423-5p levels in each group (Fig. 3C). Results from the CCK-8 assay revealed that overexpression of miR-423-5p significantly increased EC proliferation, whereas inhibition of miR-423-5p attenuated this effect (Fig. 3D). Scratch wound and Transwell migration assays further demonstrated that ECs transfected with miR-423-5p mimic showed increased migration, whereas ECs transfected with miR-423-5p inhibitor showed reduced migratory ability (Fig. 3E,G,H,I). Additionally, ECs that overexpressed miR-423-5p displayed increased total segment lengths of tube-like structures in angiogenesis assays, whereas inhibition of miR-423-5p reversed this effect (Fig. 3F,J). These results establish conclusively that miR-423-5p derived from MOB-Exos plays a pivotal role in promoting the angiogenic potential of ECs.

We next conducted RNA sequencing of the miR-423-5p overexpression group and the control group to investigate changes in the EC transcriptome following transfection with miR-423-5p mimic. The level of miR-423-5p was also assessed to confirm the successful transfection of miRNA in ECs (Fig. 4F). In ECs that overexpressed miR-423-5p, 108 genes were found to be upregulated (FC $>$2, q 0.05) and 197 genes were significantly downregulated (Fig. 4A). These differentially expressed genes are shown in a heatmap (Fig. 4B). GO analysis revealed that miR-423-5p participated mainly in the positive regulation of angiogenesis and cytokine activity (Fig. 4C), while KEGG analysis showed enrichment for cytokine-cytokine receptor interactions (Fig. 4D). Gene set enrichment analysis (GSEA) showed that miR-423-5p was a positive regulator of angiogenesis (Fig. 4E). Among the differentially expressed genes, CXCL10 was significantly downregulated and was highlighted because of its strong anti-angiogenic properties. CXCL10 is an important cytokine that inhibits VEGF-induced EC motility, which is a critical response for angiogenesis. Therefore, CXCL10 is a possible target gene for miR-423-5p and warrants further investigation.

Fig. 4.

RNA sequencing of ECs with miR-423-5p overexpression compared to the mimic NC group (n = 3). (A) Volcano plot and (B) heatmap showing the differentially expressed genes. These were evaluated by (C) GO and (D) KEGG enrichment analyses. (E) GSEA enrichment analyses showed that miR-423-5p was a positive regulator of angiogenesis. (F) qPCR results showing the miR-423-5p level in the three experimental groups. Results are presented as the mean $\pm$ SD of three independent experiments. ns: no significance, ****p 0.0001. NC, negative control; GO, gene ontology; KEGG, kyoto encyclopedia of genes and genomes; GSEA, gene set enrichment analysis.

Western blot and qPCR analyses showed significant reductions in the mRNA and protein levels of CXCL10 following transfection with miR-423-5p mimic, whereas the VEGF protein level increased (Fig. 5A). A luciferase reporter assay was performed to determine whether CXCL10 is a direct target of miR-423-5p. The results showed a marked decrease in firefly luciferase activity when miR-423-5p was co-transfected with wild-type CXCL10 3${}^{\prime }$-UTR plasmid compared to the miR-423-5p NC group (Fig. 5B). These findings provide robust evidence in support of a direct interaction between miR-423-5p and CXCL10.

Fig. 5.

miR-423-5p directly targets CXCL10. (A) qPCR was used to quantify CXCL10 mRNA, and Western blot for the analysis of C-X-C motif chemokine ligand 10 (CXCL10) and vascular endothelial growth factor (VEGF) proteins. (B) One of the predicted binding sequences of miR-423-5p to CXCL10, together with results from a dual luciferase reporter assay showing luciferase activity in the various experimental groups. Results are presented as the mean $\pm$ SD of three independent experiments. *p 0.05, **p 0.01, ****p 0.0001.

Matrigel plug assays were used to investigate the in vivo angiogenic role of MOB-derived exosomal miR-423-5p. MOBs were transfected with miR-423-5p mimics, inhibitors, or NC, and the exosomes subsequently isolated. Exosomes from the different groups were first cultured with ECs. Matrigel was then mixed with the ECs and injected into the dorsal region of nude mice. qPCR was subsequently performed to assess the level of exosomal miR-423-5p in the different experimental groups (Fig. 6C). Additionally, immunofluorescence and immunohistochemical staining were performed to evaluate the expression levels of CD31 and VEGF. The miR-423-5p mimic group showed significantly higher levels of CD31 and VEGF compared to the NC group. Interestingly, these effects were reversed after inhibition of miR-423-5p using a specific inhibitor (Fig. 6A,B,D,E). Taken together, these results demonstrate that exosomal miR-423-5p derived from MOBs plays a crucial role in promoting EC vascularization in vivo.

Fig. 6.

MOB-derived exosomal miR-423-5p promoted angiogenesis in vivo. (A,D) Immunofluorescence of CD31 expression in Matrigel plugs. (B,E) Immunohistochemical detection of VEGF. (C) The level of miR-423-5p in NC-Exos, Mimic-Exos, and inhibitor-Exos was evaluated by qPCR. Scale bar: 100 µm. Results are presented as the mean $\pm$ SD of three independent experiments. *p 0.05, ***p 0.001, ****p 0.0001.