Hypoxic ucMSC-secreted exosomal miR-125b promotes endothelial cell survival and migration during wound healing by targeting TP53INP1

ucMSC-derived exosomes mediate the paracrine effect of ucMSCs on the cell proliferation and migration of endothelial cells under hypoxic conditions

To investigate the role of hypoxia in the interaction of ucMSCs and endothelial cells, we treated human umbilical vein endothelial cells (HUVECs) with the supernatants from ucMSCs cultured under normoxic or hypoxic conditions. The results from the Cell Counting Kit-8 (CCK-8) assay showed that both the normoxic and hypoxic supernatants of ucMSCs increased the viability of HUVECs, and the hypoxic supernatants resulted in higher cell viability than the normoxic supernatants (Figure 1A). The migration of endothelial cells plays an important role in tissue repair. To study the effect of the hypoxic supernatants of ucMSCs on HUVEC migration, we performed wound healing and Transwell migration assays. The wound-healing assay demonstrated that the hypoxic supernatants of ucMSCs increased cell migration, as indicated by the significantly smaller wound area in the hypoxic group than in the normoxic group at 16 h (Figure 1B). Then, an indirect coculture model using a Transwell system was established (Figure 1C). The HUVECs exposed to the hypoxic supernatants of ucMSCs displayed the greatest migratory ability among the three groups of cells (Figure 1C). These data indicated that hypoxia promoted the paracrine effect of ucMSCs on the proliferation and migration of endothelial cells.

Exosomes, as major paracrine components of mesenchymal stem cells, mediate intercellular communication in various physiological or pathological processes.23, 24, 25 To further explore the role of exosomes in the paracrine function of ucMSCs, we pretreated ucMSCs with GW4869 (10 μM, an inhibitor of exosome biogenesis and release) for 48 h. The culture media were replaced by fresh complete media, and the cells were exposed to hypoxic conditions for 12 h. Then, the supernatants were harvested to treat HUVECs. The CCK-8 assay showed that GW4869 pretreatment significantly blocked the promoting effect of ucMSCs on the viabilities of HUVECs (Figure 1D), demonstrating that exosomes are an essential mediator of ucMSC functions. Then, ucMSCs were cultured for 72 h under normoxic and hypoxic conditions, and normoxic and hypoxic exosomes were isolated from the culture media by ultracentrifugation. The particle size and zeta potential of the normoxic and hypoxic exosomes were assessed by a particle detector. In good agreement with the range of standard values of exosomes, the average diameters of the two types of exosomes were within 104.2~118.5 nm (Figure 1E), and the average zeta potential value ranged from −3.9~−6.9 mV (Figure 1F). Under transmission electron microscopy, the two types of exosomes displayed representative cup-shaped morphology (Figure 1G), which was consistent with previous reports.^25,26 Known exosomal markers, including CD9, CD63, CD81, and HSP70, were also expressed in both types of exosomes (Figure 1H). Then, exosomes labeled with 3,3′-dioctadecyloxacarbocyanine perchlorate (DIOC18(3), a green fluorescent lipophilic tracer) were used to treat HUVECs. The images showed that HUVECs indeed took up the exosomes, which were mainly distributed in the perinuclear region of HUVECs (Figure 1I). Then, exosomes were added to the culture media of HUVECs to determine their effect on the function of endothelial cells. As Figures 1J and 1K show, hypoxic exosomes induced higher endothelial cell viabilities and migratory activities than normoxic exosomes. These data indicated that exosomes derived from ucMSCs mediated the paracrine effect of ucMSCs on HUVECs, and the hypoxic microenvironment in local wounds might create a favorable context for the function of ucMSCs.

Hypoxic exosomes lead to a better outcome of regenerated skin by enhancing cell proliferation and inhibiting cell apoptosis compared to normoxic exosomes

A mouse full-thickness cutaneous injury model was established to verify the effect of normoxic or hypoxic exosomes on wound healing. The wounds were photographed, and wound healing curves were generated. As shown in Figure 2A, the wound areas in both normoxic and hypoxic exosome-treated groups were smaller than those in the PBS-treated group at day 3 and day 8, and the wounds of the three groups almost healed at day 12. Statistical analysis showed the faster healing rates of the normoxic and hypoxic exosome-treated groups than of the PBS group, while the healing rate of the hypoxic exosome-treated group was very close to that of the normoxic exosome-treated group (Figure 2B). With an in vivo tracking assay, we observed a remarkable colocalization of exosomes (DIOC18(3), green) and endothelial cells (von Willebrand factor, red) at the areas adjacent to the injury sites, which confirmed the uptake of exosomes by endothelial cells (Figure 2C). Hematoxylin and eosin (H&E) staining also showed that there were no statistically significant differences in the thickness of the epidermis among normoxic exosome-, hypoxic exosome-, and PBS-treated groups; however, the epidermis in both the normoxic (26.4 ± 2.9 μm) and hypoxic (27.3 ± 9.8 μm) exosome-treated groups was slightly thinner than that in the PBS group (30.4 ± 4.5 μm) (Figure 2D). Ki67 and TUNEL staining were performed to evaluate cell proliferation and apoptosis (Figure 2D). The exosome injection groups had significantly more Ki67-positive cells and fewer TUNEL-positive cells than the PBS control group. Interestingly, we observed a significantly increased number of Ki67-positive cells and a decreased number of TUNEL-positive cells in the hypoxic exosome-treated group compared to the normoxic exosome-treated group (Figure 2D), suggesting a better outcome of hypoxic exosome treatment.

ucMSC-derived exosomal miR-125b, transcriptionally induced by hypoxia, modulates the behaviors of endothelial cells

Exosomes always carry and deliver mRNAs, microRNAs, and proteins to target cells, and the levels of exosome contents are strongly correlated with their levels in the cells of origin.^27,28 Hypoxic conditions have been implicated in modulating the expression profile of microRNAs in different types of cells as well as in the exosomes derived from these cells.29, 30, 31 A previous study investigated the exosomal microRNA expression profile in ucMSCs and screened several abundant microRNAs, including miR-21-5p, miR-125b-5p, miR-23a-3p, let-7f-5p, let-7a-5p, and miR-145-5p.¹⁸ We speculated that the abundance of microRNAs in exosomes was important for their functions, so we confirmed the expression of these microRNAs and analyzed their differential expression between hypoxic exosomes and normoxic exosomes from ucMSCs by quantitative PCR. The expression of miR-125b was found to be most notably increased in hypoxic exosomes compared with normoxic exosomes (Figure 3A). To explore the correlation of the miR-125b levels in exosomes and their source ucMSCs, we assessed the expression of miR-125b in ucMSCs exposed to hypoxic conditions, including a tri-gas culture for 3 h and 6 h or stimulation with cobalt chloride (CoCl₂) to create a chemical hypoxia model. We observed an upregulation of miR-125b expression in ucMSCs cultured under hypoxic conditions (Figures 3B and 3C). A luciferase assay also confirmed increased reporter activity of the miR-125b promoter under hypoxic conditions (Figure 3D). These data suggested that the accumulation of miR-125b in exosomes was largely due to its upregulation in ucMSCs under hypoxic conditions.

To further confirm whether exosomal miR-125b could be taken up by endothelial cells, we cocultured normoxic exosomes or hypoxic exosomes with HUVECs for 48 h, and the HUVECs were pretreated with 5,6-dichloro-1-β-D-ribofuranosylbenzimidazole (DRB, an RNA polymerase II inhibitor, 20 mM) for the previous 48 h. Then, we detected the expression of mature miR-125b in the HUVECs. DRB treatment markedly decreased the endogenous expression of mature miR-125b in HUVECs; exosome addition obviously improved the level of mature miR-125b, indicating that the increase in miR-125b expression in HUVECs was due to their uptake from exosomes rather than the enhanced transcription of miR-125b in HUVECs. Consistent with the dramatic increase in miR-125b expression in hypoxic exosomes, as shown in Figure 3A, the level of mature miR-125b expression in hypoxic exosome-treated HUVECs was much higher than that in normoxic exosome-treated HUVECs. The above data, together with the cytoplasmic localization of exosomes in endothelial cells (Figure 2C), further demonstrated much more uptake by HUVECs of miR-125b from hypoxic exosomes than of miR-125b from normoxic exosomes (Figure 3E).

To further address whether the function of exosomes depends on miR-125b, we treated HUVECs with hypoxic exosomes with or without miR-125b inhibitors. As CCK-8 (Figure 3F) and migration assays (Figure 3G) showed, inhibition of miR-125b blocked the promoting effect of hypoxic exosomes on cell growth and migration. Consistent with the in vivo study results (Figures 2F and 2G), hypoxic exosome treatment significantly decreased the cell apoptosis induced by high concentrations of CoCl₂, while inhibition of miR-125b abrogated the protective effect of hypoxic exosomes on cell apoptosis (Figure 3H). Taken together, we demonstrated that ucMSC-derived exosomal miR-125b, whose transcription was induced by hypoxic conditions, was necessary for cell migration and survival of endothelial cells.

The interaction of miR-125b and TP53INP1 is crucial for the function of hypoxic exosomes

MicroRNAs participate in diverse cellular signaling pathways by regulating their targets. To examine the functional mechanism of exosomal miR-125b in endothelial cells, we analyzed the targets of miR-125b using online databases, including TargetScan (http://www.targetscan.org/vert_72/), miRTarBase (https://mirtarbase.cuhk.edu.cn/∼miRTarBase/miRTarBase_2019/php/download.php), miRDB (http://mirdb.org/), and PicTar (https://pictar.mdc-berlin.de/), and screened 30 common targets (Figure 4A). DAVID analysis showed that these targets were enriched in the cell apoptotic process (p = 0.0018), and tumor protein p53 inducible nuclear protein 1 (TP53INP1) was an interesting target mRNA of them (Figures 4B and 4C). Luciferase reporter assay confirmed the direct binding of miR-125b to the 3′ untranslated region of wild-type TP53INP1 (Figure 4D). Then the expression of TP53INP1 was detected by quantitative PCR and western blotting. The mRNA and protein levels of TP53INP1 were obviously downregulated in endothelial cells by exposure to hypoxia (Figures 4E and 4F), indicating the involvement of TP53INP1 in hypoxia-induced cell damage. miR-125b overexpression suppressed the level of TP53INP1 under both normoxic and hypoxic conditions, and vice versa, suggesting that miR-125b increased the mRNA degradation of TP53INP1 (Figures 4E and 4F).

To further study the interaction of miR-125b and TP53INP1, we constructed stable TP53INP1-overexpressing or TP53INP1-knockdown HUVECs using a lentivirus delivery system (Figure 4G). Then, we verified the role of the interaction of miR-125b and TP53INP1 in the process of endothelial cell apoptosis by determining the protein levels of TP53INP1 and cell apoptotic markers, including caspase-3, B cell lymphoma 2-associated X protein (Bax), and B cell lymphoma 2 (Bcl2). As Figure 4H demonstrates, overexpression of miR-125b decreased the levels of cleaved caspase-3 and Bax and increased the level of Bcl2 in HUVECs treated with high concentrations of CoCl₂. Cotransfection of TP53INP1 without the 3′ untranslated region reversed the effect of miR-125b overexpression on the above protein levels. Moreover, inhibition of miR-125b upregulated cleaved caspase 3, and Bax expression and downregulated Bcl2 expression, whereas cotransfection of shTP53INP1 impaired the promotion effect of miR-125b inhibitors on cell apoptosis, indicating that the function of miR-125b during hypoxia was due, at least in part, to the downregulation of TP53INP1 expression (Figure 4H).

To identify whether the protective effect of hypoxic exosomes on endothelial cells was dependent on TP53INP1 downregulation, we analyzed the mRNA and protein levels of TP53INP1 following exosome treatment. Hypoxic exosome treatment reduced the level of TP53INP1, while inhibition of miR-125b abrogated this suppressive effect under hypoxic conditions (Figure 4I). Then, cell growth, migration, and survival were assessed under both normoxic and hypoxic conditions. As CCK-8 assays showed (Figure 4J), TP53INP1 overexpression led to lower cell viability than the corresponding control. Hypoxic exosome treatment improved cell viability, whereas the protective effect was reversed by TP53INP1 overexpression (Figure 4J). A similar phenomenon was observed in the migration assay; hypoxic exosomes increased the number of migrated cells, whereas TP53INP1 overexpression inhibited the promoting effect of hypoxic exosomes following hypoxia stimulation (Figure 4K). Apoptosis of TP53INP1-overexpressing endothelial cells exposed to hypoxia was also detected. As expected, we observed a higher cell apoptotic rate in hypoxic exosome-treated TP53INP1-overexpressing cells than in hypoxic exosome-treated control cells (Figure 4L). The data described above further illustrated that exosomal miR-125b targeted TP53INP1 mRNA in endothelial cells and that the interaction of miR-125b and TP53INP1 was crucial for the function of exosomes.

Artificial miR-125b promotes cutaneous wound healing

To further determine the function of miR-125b in skin wound repair in vivo, we injected agomiR-125b or agomiR control oligos into the areas adjacent to skin injury sites by subcutaneous injection. Compared with the agomiR control group, agomirR-125b led to an apparent decrease in wounds at day 3 and the following days (Figure 5A). The wound closure rate of the agomiR-125b group was higher than that of the agomiR control group, as shown in Figure 5B. AgomiR-125b injection led to a slight increase in the thickness of the newly formed epidermis of 49.8 ± 12.3 μm compared with that of the agomiR control group of 40.7 ± 9.9 μm, but the difference was not statistically significant (Figure 5C). The number of Ki67-positive cells increased and the number of TUNEL-positive cells decreased following agomiR-125b treatment (Figure 5C). The expression of TP53INP1 was also tested, and we observed a reduction in TP53INP1 levels in the agomiR-125b group (Figure 5C). Then, the RNA levels of miR-125b and Tp53inp1 in the wound sites of exosome- or oligo-treated animals at the end point of injury repair were detected by quantitative PCR, and the results showed that the expression of miR-125b was increased following exosome injection (Figure 5D, left). Although there was no significant difference in the mRNA levels of Tp53inp1 among the five groups, Tp53inp1 expression was lower in the exosome- or agomiR-125b-injected groups (Figure 5D, right). All these data further defined the interaction of miR-125b and TP53INP1 during exosome-mediated skin tissue repair.

Animals

All experimental procedures involving animals were approved by the Animal Care and Use Committee of Wuhan Union Hospital. All mice were housed in an environment with controlled light (12 h light/12 h dark), temperature (23°C ± 2°C), and humidity. 8-week-old male BALB/C mice were used. For the full-thickness skin wound model, mice were anesthetized by intraperitoneal (i.p.) administration of 50 mg/kg pentobarbital sodium. A skin wound 8 mm in diameter was created on the dorsum. In total, 25 mice were randomly divided into control group (PBS), normoxic or hypoxic exosome groups (200 μg in 100 μL PBS), and agomiR-125b or agomiR-control groups, with 5 mice in each group. The above exosomes or oligos were subcutaneously injected at four points around the wound sites. The wounds were photographed every day and the wound size was analyzed by ImageJ software. 12 days after injection, mice were sacrificed, and skin samples were harvested for further analysis. The wound-healing rate on a specific day (day X) was the ratio of the wound area differences between day 0 and day X to wound area at day 0.

Histology staining and TUNEL staining

Skin samples were fixed in 4% paraformaldehyde, dehydrated, embedded in paraffin, sectioned into 5-μm-thick pieces, and stained with H&E. For each mouse, epidermal thickness of regenerated skin flanked by the normal skin with hair follicles was measured from stratum basal to stratum corneum at six equidistant sites. Immunofluorescence staining for Ki67 (Servicebio, #GB111141, Wuhan, China) and TP53INP1 (ABclonal, #A8274) were performed to evaluate cell proliferation and the level of TP53INP1, respectively. TUNEL (terminal deoxynucleotidyl transferase dUTP nick end labeling) assays were performed to detect cell apoptosis according to the manufacturer’s instructions (Servicebio, #G1507, Wuhan, China).

Exosome isolation and identification

ucMSCs (Cyagen, Guangdong, China) were cultured to a density of 2 × 10⁷ cells/75 mm flask. Culture media were removed, and cells were washed three times with PBS and then cultured in serum-free media for 48 h. The media were collected, and cell debris was discarded by gradient centrifugation at 300 × g for 10 min and 10,000 × g for 30 min at 4°C. Further centrifugation (100,000 × g, 4°C, 70 min) was performed to obtain exosomes. The size distribution and zeta potential of exosomes was measured by Zetasizer Nano ZS90 (Malvern Panalytical, Westborough, MA, USA). The morphologies of exosomes were photographed with transmission electron microscopy (Hitachi HT 7800, Tokyo, Japan). The exosomes were measured for their protein content using the BCA protein assay kit (Pierce Protein Biology, Thermo Fisher Scientific Life Sciences). For labeling, exosomes were incubated with DIOC18(3), a green fluorescent lipophilic tracer (Beyotime), at room temperature for 30 min and centrifuged at 100,000 × g, 4°C, for 70 min, incubated with HUVECs in normoxic and hypoxic conditions for 3 h, and then photographed.

In vivo exosome tracking assay

To assess exosome uptake by skin endothelial cells, 20 μL DIOC18(3)-labeled exosomes (1 μg/μL) were injected into the around areas of wounded skin. Mice were sacrificed at 5 h after injection. The skin was dehydrated and frozen, then sliced into 10-μm-thick cryosections. The sections were incubated with primary antibody against von Willebrand factor (Servicebio, #GB11020, Wuhan, China) overnight at 4°C and then dyed with DAPI. The internalization of exosomes into skin endothelial cells was observed by confocal microscope.

Cell culture

ucMSCs and HUVECs were cultured in specific complete media (Cyagen, Guangdong, China). 293T were cultured in DMEM (Gibco) supplemented with 10% fetal bovine serum (FBS) and 100 U/mL penicillin-streptomycin (Gibco). For hypoxic treatment, cells were exposed to 1% O₂, 95% N₂, and 5% CO₂ for 3–6 h in a tri-gas incubator. When required, cells were treated with a high concentration of CoCl₂ (200 μM) for 24 h to induce cell growth arrest and apoptosis, while low concentration of CoCl₂ with 50 μM was used to induce the production of exosomes The optimal concentration of CoCl₂ (50 μM) was determined by dose-dependent (using 25–200 μM) experiments to produce maximal induction of HIF1α and miR-125b at 12 h, without affecting cell viabilities (data not shown). For coculture of exosomes and cells, 50 μg/mL of exosomes were incubated with HUVECs at 37°C for 24 h.

Cell viability assay

CCK-8 assays were performed to analyze cellular proliferation and activity. HUVECs were seeded at a density of 2 × 10³ cells/well in 96-well plates. At 24, 48, 72, 96, and 120 h, 10 μL of CCK-8 solution (Dojindo Laboratories, Kumamoto, Japan) was added to each well of the plate and incubated for 1 h. Cell viability was then determined using a spectrophotometer set (ELx800, BioTek, Winooski, VT, USA) at a wavelength of 450 nm.

Wound-healing assay

HUVEC migrations were performed in 12-well plates by wound-healing assays in vitro. Cells were seeded at a density of 5 × 10⁵ cells per well and starved for 24 h. Wounds were made by scratching with a 200 μL pipette tip (3 wounds per well). The cell debris was removed with PBS washing. Then cells were incubated with supernatants from ucMSCs treated with or without hypoxic treatment. The scratch of each group was observed using a microscope at 0 h and 24 h. The area of the scratch was measured using ImageJ software.

Transwell migration assay

For Transwell assay, 3 × 10⁴ cells were suspended in serum-free medium and seeded into the upper chamber of Transwell 24-well plates (8 μm pore filters, Corning, Corning, NY, USA). Complete medium with or without exosomes was added to the lower chamber. After 16 h, the cells on the upper surface of the inserts were wiped. Migrated cells of the lower surface were stained with crystal violet and photographed with a microscope.

Flow cytometry analysis

Cells treated with indicated exosomes or oligos were exposed to 200 μM CoCl₂ for 24 h, and cell apoptosis was tested using PI/Annexin V-FITC staining (#556547, BD Biosciences, St. Louis, MO, USA) according to the manufacturer’s protocol for another 10 min. After incubation, the cells were subjected to apoptosis analysis using BD Accuri C6 Plus cytometer (Beckman Coulter, Fullerton, CA, USA), and the results were analyzed with BD Accuri C6 Plus software.

Plasmid and stable transfected cell lines

The mimics, inhibitors, and corresponding negative control oligos for hsa-miR-125b-5p were purchased from Guangzhou RiboBio (Guangzhou, China). Oligo transfection was performed as per manufacturer’s instructions. Human TP53INP1 cDNA was amplified following the primers: 5′-ATGGACAATATGTCTATTAC-3′ (forward) and 5′-TCAGTCTAAAGGTTGTGGG-3′ (reverse). The short hairpin RNA (shRNA) sequences against TP53INP1 were as follows: (1) 5′-CGAGTTGTATCACCTGGAATT-3′, (2) 5′-GTACTTCATACCATGCCGATT-3′, and (3) the scramble sequence, 5′-AATTCTCCGAACGTGTCACGT-3′. The cDNAs were ligated into pHAGE-flag vector, and shRNAs were ligated into pLKO.1 vector. For stable cell line construction, HEK293T cells were transfected with overexpression or knockdown plasmids with packaging plasmids pSPAX2 and pMD2G for 60 h. The supernatants were harvested, filtered with 0.22 μm membrane, and added into pre-seeded HUVECs in a 6-well plate. Puromycin (2 μg/mL) was utilized to screen stable expression clones.

Luciferase reporter assay

293T cells were seeded in 24-well plates at a density of 2 × 10⁵ cells/well and transfected with the indicated plasmids for 24 h. Luciferase assays were performed using a dual-luciferase assay kit (#E1960, Promega, Madison, WI, USA) according to the manufacturer’s instructions.

Quantitative PCR

Total RNA was extracted using TRIzol reagent (#9109, TaKaRa, Japan). cDNA was obtained by using an oligo (dT) primer and reverse transcriptase (Thermo Fisher Scientific, Waltham, MA, USA) following standard protocols. MicroRNA was reverse transcribed by using specific Bulge-Loop RT primers (RiboBio, Guangzhou, China). The relative expression levels of mRNA and microRNA were normalized to GAPDH or U6, respectively.

Western blot

Cells were lysed in RIPA lysis buffer (#P00138, Beyotime Biotechnology, Shanghai, China). Equal amounts of protein were separated by SDS-PAGE gels and transferred to polyvinylidene fluoride (PVDF) membrane. Antibodies against CD9 (#A19027), CD63 (#A5271), CD81 (#A5270), and HSP70 (#A0284) were obtained from ABclonal Technology (ABclonal, Wuhan, China). The antibodies against HIF1α (#36169, Cell Signaling Technology [CST], Danvers, MA, USA), Bax (#5023, CST, Danvers, MA, USA), Bcl-2 (#15071, CST, Danvers, MA, USA), caspase3 (#9665, CST, Danvers, MA, USA), cleved-caspase3 (#9661, CST, Danvers, MA, USA), p53 (#9282, CST, Danvers, MA, USA), and β-tubulin (#2128, CST, Danvers, MA, USA) were purchased from Cell Signaling Technology. A ChemiDoc MP Imaging System (Bio-Rad, Hercules, CA, USA) was used for signal detection.

Statistical analysis

All the data in this study are expressed as means ± SD. Two treatment groups were compared by Student’s t test. Multiple groups were analyzed by one-way ANOVA analysis. GraphPad Prism version 8.0 was used for statistical analyses. Statistical significance was considered when p < 0.05.