Human umbilical vein endothelial cells (Cat. no. 8000; ScienCell Research Laboratories, Carlsbad, CA, United States) were cultured in ECM (Cat. no. 1001; ScienCell Research Laboratories) containing 5% fetal bovine serum (FBS; cat. no. 0025; ScienCell Research Laboratories), 1% penicillin/streptomycin (PS; cat. no. 0503; ScienCell Research Laboratories), and 1% endothelial growth factor (EGF; cat. no. 1052; ScienCell Research Laboratories). A293 cells (Fenghbio, Changsha, China) were cultured in Dulbecco’s modified Eagle medium (DMEM; 4.5 g/L D-glucose, HyClone, Logan, UT, United States) containing 10% FBS and 1% PS. All cells were cultured in a humidified incubator at 37°C with 5% CO₂. ECM containing 2% FBS, 1% PS, and 1% EGF along with D-glucose (Sigma-Aldrich, St. Louis, MO, United States) was used to generate hyperglycemic conditions. D-mannitol (Sigma-Aldrich) was also used to induce a similar osmotic pressure. Lipofectamine 2000 (Invitrogen, Carlsbad, CA, United States) was used to transfect cells according to the manufacturer’s protocol. NRF2 inhibitor ML385 (3 μM) and HO-1 inhibitor ZnPP (10 μM) were used to incubate cells.

Exosomes from three diabetic foot ulcer patients’ plasma and three non-diabetes foot trauma patients’ plasma were considered as DFU-Exos and NDF-Exos. This experiment was approved by the Ethics Committee of The Second Hospital of Hebei Medical University. Exosomes from culture supernatant of HUVECs were grown in an exosome-free culture medium. Briefly, the plasma or culture medium was centrifuged at 2,000 × g for 30 min at 4°C. The supernatant was centrifuged at 10,000 × g in 4°C for 70 min and filtered through a 0.22 μm filter. Next, the supernatant was ultracentrifuged at 100,000 × g in 4°C for 70 min and the resulting liquid was discarded. Finally, the precipitate was washed with 5 mL PBS and ultracentrifuged (100,000 × g, 70 min) at 4°C. The precipitate contained the exosomes used to perform further experiments.

Twenty microliters of exosomes were used for TEM. The morphology of the exosomes was observed using a transmission electron microscope (HT-7700, Hitachi, Japan). NTA (N30E, NanoFCM, China) was used to analyze the size and concentration of exosomes.

1,1′-dioctadecyl-3,3,3′,3′-tetramethylindocarbocyanine perchlorate (DiO; Sigma, St. Louis, MO, United States) was used to label exosomes according to the manufacturers’ protocol (10). DiO labeled the exosomes as our previous study described. The exosomes uptaken were imaged by a fluorescence microscope (IX73, Olympus, Japan).

Total RNA was isolated from cells using TRIzol reagent (Invitrogen). Exosomal miRNA was extracted using an miRcute miRNA isolation kit (Tiangen Biotech, Beijing, China). The concentration of RNA was measured using a NanoDrop One (Thermo Fisher Scientific, Wilmington, DE, United States). cDNA was synthesized using a high-capacity cDNA reverse transcription kit (Thermo Fisher Scientific). qRT-PCR was performed via SYBR Green assays (Genecopoeia, Guangzhou, China) using a Bio-Rad system (Bio-Rad, Hercules, CA, United States). U6 or β-actin was used as the internal control to normalize miR-181b-5p or mRNA expression, respectively. The 2^–ΔΔCT formula was used to calculate the fold change of miRNA or mRNA expression. Table 1 showed the sequences of primers.

A293 cells were seeded in a six-well plate and co-transfected with pmiR-GLO-NRF2 3′ UTR wild-type (pmiR-GLO-NRF2 3′ UTR wt) or mutation (pmiR-GLO-NRF2 3′ UTR mut) and miR-181b-5p mimics. After 48 h, the cells were lysed using RIPA (Solarbio, China). A Dual Luciferase Assay System (Promega, Madison, WI, United States) was used to measure the luciferase activity using a Flash and Glow reader.

Human umbilical vein endothelial cells or tissue slices were treated with 4% paraformaldehyde, permeabilized in 0.2% Triton X-100 for 10 min, and blocked with goat serum for 1 h at room temperature. The cells or tissue slices were incubated overnight with primary antibodies of NRF2 (1:50; 66504-1-lg, Proteintech, China) or CD31 (1:50; AF6191, Affinity, China). They were then incubated with secondary antibodies (1:50; Affinity) after washing with PBS. Nuclei were stained with DAPI (Solarbio) and imaged using a fluorescence microscope (IX73, Olympus, Japan).

Immunohistochemical staining was performed according to the manufacturer’s protocol (ZSBio, China). After dewaxing and hydration, paraffin sections were immersed in endogenous peroxidase blocker at room temperature for 10 min and then washed with PBS. The sections were incubated with CD31 (1:50; AF6191, Affinity) at 37°C for 60 min and then washed with PBS. An enhancer (100 μL) was added for 20 min and then the sections were again washed with PBS. Next, the sections were incubated with enzyme-labeled goat anti-rabbit IgG polymer for 20 min and then washed with PBS. The labeling was visualized using DAB (ZSBio) and hematoxylin (Solarbio) and captured using a microscope (IX73).

We seeded 3000 cells per well in a 96-well plate after transfection or incubation with exosomes. After 24 h, 10 μL of CCK-8 (Solarbio) solution was added to the medium. The OD value was measured at 450 nm using a microplate reader (MD2, Molecular Devices, United States).

β-galactosidase staining was performed according to the manufacturer’s protocol (Beyotime, China). The cells were seeded in a six-well plate and incubated with 1 mL of β-galactosidase staining fixation fluid for 15 min at room temperature after washing the cells with PBS. After that, the fixation fluid was removed and the cells were incubated with 1 mL staining working solution which was composed of 10 μL staining solution A, 10 μL staining solution B, 10 μL staining solution C and 50 μL X-Gal solution. The cells were incubated at 37°C overnight and captured the images using a microscope (IX73).

Matrigel was coated in a 24-well plate in a humidified incubator at 37°C for 2 h. After incubation with exosomes or transfection, HUVECs (50,000 cells per well) were seeded into the plate. After 24 h, tube formation was imaged using a microscope (IX73).

C57BL/6J mice (male, 6 weeks old, weighing 20–30 g, SPF-Tsinghua, China) were used to establish a DFU mouse skin wound model. The mice were randomly separated into three groups with 5 mice each group. Mice in the normal group were injected with saline under normal feeding. Mice in the diabetes group were fed a high-glucose and high-fat diet and intraperitoneally injected with streptozotocin (STZ; 45 mg/kg, Sigma-Aldrich, St. Louis, MO, United States). After 48 h, the blood glucose level was measured using a blood glucose monitoring system (Roche, Switzerland). Blood glucose levels ≥16.7 mM continuously for 10 days were considered characteristic of the diabetes model. Two weeks later, a full-thickness excisional skin wound was inflicted on the upper backs of the mice (11, 12). The mice were then randomly divided into three groups with 5 mice each group: normal control group was C57BL/6J mice injected with exosomes from HUVECs transfected with ASO-NC (NC group), diabetes group was diabetic mice injected with exosomes from HUVECs transfected with ASO-NC (DM + Exo-NC), and Exo-miR-181b-5p inhibitor group was diabetic mice injected with exosomes from HUVECs transfected with miR-181b-5p inhibitor (DM + Exo-miR-181b-5p inhibitor). On days 1, 4, 7, 11, and 14 post wounding, the wounds were photographed and measured using a ruler. The diameter was the average of maximum diameter and minimum diameter. All the animal experiments were conducted in accordance with the National Institute of Health guidelines (NIH Publication No. 8023, revised 1978) and were approved by the Research Ethics Committee of The Second Hospital of Hebei Medical University.

The proteins isolated from exosomes or cells were homogenized in RIPA buffer (Solarbio) and the protein concentration was measured using a BCA protein assay kit (Beyotime, China). The proteins were separated by SDS-PAGE and transferred to PVDF membranes (Millipore, Burlington, MA, United States). The membranes were incubated in BSA (Solarbio) and then with the antibody TSG101 (67381-1-lg, Proteintech), CD63 (67605-1-lg, Proteintech), Calnexin (Cat No. 10427-2-AP, Proteintech), PCNA (610664, BD, United States), VEGF (2E2H9, Proteintech), p16^INK4A (10883-1-AP, Proteintech), or NRF2 (66504-1-lg, Proteintech) overnight. Secondary antibodies (5230-0336 and 5230-0341, KPL, IN, United States) were added 1 h after washing the PVDF membranes with TBST (Solarbio). Bands were developed using ECL solution (Solarbio) and imaged using a ChemiDoc MP (Bio-Rad).

All statistical analyses were performed using GraphPad Prism software (Version 5.0; GraphPad Software, Inc., San Diego, CA, United States). The data was expressed as means ± standard deviations. Statistical differences between two groups were determined using Student’s t-test. Statistical differences >2 groups were determined using One-way ANOVA or two-way ANOVA followed by Tukey’s post hoc test. All data were tested for normality and equal variance. Statistical significance was set at p ≤ 0.05.

Plasma exosomes of patients with non-diabetic foot wound (NDF-Exos) and DFU (DFU-Exos) were isolated by ultracentrifugation. NDF-Exos and DFU-Exos were observed using TEM (Figure 1A). NTA showed that the diameters of both DFU-Exos and NDF-Exos groups are in 40–150 nm. The median diameters of NDF-Exos and DFU-Exos were 70.25 and 66.25 nm (Figure 1B). Western blotting showed that exosome-associated markers (CD63 and TSG101) were positively expressed and Calnexin was negatively expressed (Figure 1C). We then assessed the effects of NDF-Exos or DFU-Exos on HUVECs. CCK-8 assay indicated that cell viability decreased when HUVECs were incubated with DFU-Exos (Figure 1D). DFU-Exos promoted cell senescence compared with the NDF-Exos group; this was detected using the senescence-associated β-galactosidase activity in HUVECs (SA-β-Gal) (Figure 1E). Moreover, tube formation reduced in HUVECs treated with DFU-Exos (Figure 1F). Western blotting showed that PCNA and VEGF expression decreased, whereas p16^INK4A expression increased, in DFU-Exos-treated HUVECs compared with that in NDF-Exos-treated cells (Figure 1G).

To determine the molecular mechanism by which DFU-Exos impaired HUVECs functionality, we retrieved a dataset from the Gene Expression Omnibus (GEO) of the National Center for Biotechnology Information and found that exosomal miR-181b-5p expression was elevated in patients with DFU compared with that in patients with NDF or healthy people (13). qRT-PCR verified the higher level of expression of miR-181b-5p in exosomes from the plasma of patients with DFU (Figure 2A). DiO-labeled DFU-Exos or NDF-Exos showed the uptake of exosomes by HUVECs (Figure 2B). Furthermore, exosomes from HUVECs grown with a high concentration of glucose (25 mM) exhibited higher levels of miR-181b-5p than those from cells grown with normal glucose (5.5 mM) levels (Figure 2C). In addition, the level of miR-181b-5p was higher in HUVECs which were incubated with DFU-Exos than those incubated with NDF-Exos (Figure 2D). Because the miRNAs might be attached outside the exosomes, we treated exosomes with proteinase K and found that proteinase K did not affect the miR-181b-5p level in HUVECs which were incubated with DFU-Exos (Figure 2D). In contrast, treatment with the phospholipid membrane disruptor Triton X-100 led to a degradation of miR-181b-5p which indicated that the high level of miR-181b-5p is from the DFU-Exos (Figure 2D). Overexpression of miR-181b-5p by transfecting miR-181b-5p mimics in HUVECs inhibited cell viability and promoted cell senescence (Figures 2E,F). We also found that miR-181b-5p could inhibit tube formation in HUVECs compared with the control group (Figure 2F). In addition, Western blotting showed the PCNA and VEGF expression decreased, whereas p16^INK4A expression increased, when miR-181b-5p was overexpressed in HUVECs (Figure 2G). To further evaluate the roles of miR-181b-5p carried by DFU-Exos on cell senescence and tube formation of HUVECs, we performed a rescue experiment. Inhibition of miR-181b-5p by miR-181b-5p inhibitor could partly rescue the effects of cell viability, senescence and tube formation exerted by DFU-Exos (Figures 2H,I). Nevertheless, the expression of PCNA, VEGF, and p16^INK4A in HUVECs incubated with DFU-Exos also partly rescued by inhibiting miR-181b-5p expression (Figure 2J).

It is well known that miRNAs carry out their functions by binding to target mRNA 3′ UTRs. We used Starbase¹ to predict miR-181b-5p targets and found that NRF2 might be one (Figure 3A). The binding sites are shown in Figure 3A. To verify their relationship, a dual luciferase reporter assay was performed. We cloned a NRF2 3′ UTR fragment (wild-type or mutation) containing binding sites or mutation for miR-181b-5p into pmiR-GLO (pmiR-GLO-NRF2 3′ UTR wt or mut). Cells co-transfected with pmiR-GLO-NRF2 3′ UTR wt and miR-181b-5p mimics inhibited luciferase activity, unlike cells co-transfected with pmiR-GLO-NRF2 3′ UTR wt and NC mimics. No effects on luciferase activity were observed when pmiR-GLO-NRF2 3′ UTR mut was transfected into A293 cells along with miR-181b-5p mimics (Figure 3B). To further identify the effects of miR-181b-5p on NRF2 mRNA and protein expression, we transfected miR-181b-5p or NC mimics into HUVECs. qRT-PCR, Western blotting, and immunofluorescence staining showed that miR-181b-5p overexpression decreased the mRNA and protein levels of NRF2 (Figures 3C–E). In addition, DFU-Exos reduced the mRNA and protein levels of NRF2 (Figures 3F–H). More importantly, overexpression of NRF2 partly rescued the inhibition of NRF2 induced by miR-181b-5p (Figure 3I). These results indicated that exosomal miR-181b-5p inhibited NRF2 expression by targeting its 3′ UTR.

Considering that HO-1 gene is the downstream gene of NRF2, we also verified that overexpression of miR-181b-5p inhibited HO-1 expression in HUVECs (Figure 4A). To further identify whether NRF2/HO-1 pathway is involved in the HUVECs senescence and tube formation, we knocked down NRF2 or HO-1 expression by siRNA. CCK-8, β-galactosidase activity and tube formation were performed, which indicated that knock down of NRF2 or HO-1 expression promoted cell senescence, and inhibited cell activity and tube formation in HUVECs (Figures 4B,C). Western blotting showed that the PCNA and VEGF expression decreased, whereas p16^INK4A expression increased, when NRF2 or HO-1 was knocked down in HUVECs (Figure 4D). In addition, NRF2 and HO-1 inhibitors, ML385 and ZnPP, could also rescue the effects on cell activity, senescence and tube formation exerted by miR-181b-5p inhibitor (Figures 4E,F).

To further confirm the effects of exosomal miR-181b-5p on wound repair in vivo, full-thickness cutaneous wounds were generated on the backs of mice. Equal volumes of exosomes isolating from culture medium of HUVECs transfected with inhibitor NC (Exo-NC) or miR-181b-5p inhibitor (Exo-miR-181b-5p inhibitor) were injected around the wounds. The disappearance of wounds showed that the Exo-miR-181b-5p inhibitor accelerated wound healing in the diabetes mice group compared with that in the DM + Exo-NC group (Figures 5A,B). CD31 immunohistochemical staining indicated that the CD31⁺ area was higher in the DM + Exo-miR-181b-5p inhibitor group than in the DM + Exo-NC group which indicated that Exo-miR-181b-5p inhibitor group showed higher vascular density than DM + Exo-NC group (Figures 5C,D). Immunofluorescence staining showed that the expression of NRF2 in ECs was higher in the DM + Exo-miR-181b-5p inhibitor group than in the DM + Exo-NC group (Figure 5D). These results indicated that exosomal miR-181b-5p could impair wound healing by inhibiting angiogenesis in vivo.