Human Mesenchymal Stem Cell Derived Exosomes Endowed with miR-13474 as a Therapeutic Delivery Vehicle for Diabetic Wound Healing by Targeting the CPEB2/TWIST1 Axis

Abstract

A delayed healing process in diabetic wounds is intractable. In this study, a high-glucose condition was found to be responsible for skin structure destruction, inflammatory infiltration, and vital cell dysfunction. Extracellular vesicles, particularly exosomes secreted by hucMSCs, contribute to improved diabetic wound healing, largely by promoting tissue repair and re-establishing normal function in affected cells. Small RNA-sequencing revealed that hucMSC-derived exosomes (hucMSC-Ex) were highly enriched in NC_000019.10_13474 (miR-13474), which was predicted to be an miRNA with an undiscovered function. miR-13474 showed a reduced expression level in high-glucose-treated skin cells as well as diabetic foot ulcer (DFU) rats. Moreover, there is also a significant expression difference between the wound area and the wound edge in DFU patients, indicating the potential clinical value of miR-13474. Blocking miR-13474 in hucMSC-Ex obviously diminished the therapeutic effects. Furthermore, exosomal miR-13474 was found to target the CPEB2/TWIST1 axis to improve the impaired function of skin cells. On this basis, hucMSC-Ex were used as a vehicle for the delivery of therapeutic miR-13474 to optimize the repairing effect. The study has revealed the role of hucMSC-derived exosomes and the underlying molecular mechanism in diabetic wound healing and proposes a cell-free-based modification strategy for refractory wound management.

Introduction

Diabetes has become an increasingly significant public health concern worldwide and is frequently associated with a range of complications. Its chronic and persistent nature renders both the disease and its related complications a continuing challenge that requires ongoing attention. , Among diabetes-related complications, diabetic foot ulcer (DFU) is considered one of the most devastating, with reports suggesting that nearly 84% of lower-limb amputations not caused by trauma are associated with a history of DFU. Multiple mechanisms drive DFU formation, notably neuropathy, vascular pathology, and immune disturbances. Hyperglycemia acts as a central factor, promoting microvascular injury and metabolic impairment in muscle, which synergistically worsens tissue damage. , Tissue repair relies on an integrated sequence of biological activities, cell movement and expansion, extracellular matrix generation, angiogenesis, and remodeling, working together to restore the structural and functional integrity. In chronic wounds, these overlapping phases are often impaired, with particularly pronounced deficiencies observed during the proliferative and matrix remodeling stages. −

Various types of cells such as immune cells and endothelial cells are involved in the healing process, while dermal fibroblasts (DFs) are considered as the main target and effector cells in the above-mentioned two phases by secreting collagen-rich ECM in the wound bed and facilitating the migration of keratinocytes to close wounds. , However, microenvironmental glycation in diabetic patients is likely to inhibit cell proliferation, migration, and angiogenic sprout formation. − Compared with fibroblasts from nondiseased controls, those obtained from leptin receptor-deficient diabetic mice or patients suffering from chronic venous insufficiency show significant defects in migration. , Furthermore, production of NO by DFs, which is also critical in fibroblast proliferation and function, is reduced in high-glucose conditions. These cellular dysfunctions hinder granulation tissue formation and extracellular matrix (ECM) deposition, ultimately resulting in nonhealing wounds. Thus, using proper ways to recover the functions of DFs in skin tissue is probably half the work, with double results in diabetic wound healing.

Cellular therapies, with a particular emphasis on mesenchymal stem cell (MSC) applications, are regarded as powerful tools for driving regenerative processes. , Growing evidence indicates that MSCs primarily restore tissue function through paracrine mechanisms, with exosomes emerging as key mediators and receiving considerable research attention. ,,, Almost all cell types release exosomes, nanometer-scale vesicles that deliver lipids, metabolites, nucleic acids, and proteins. Through this molecular cargo, they establish a communication system that operates both proximally and distally, underpinning their diverse biological effects. MSC-derived exosomes (MSC-Ex) have beneficial functions that are similar to and even better than those of MSCs. Exosomes show distinct advantages such as the lipid bilayer membrane protection of cargos from fast degradation, ,, a low chance of triggering immune response, a smaller size, high stability, and easy storage. , Furthermore, when compared with synthetic nanoparticles (NPs) such as metal NPs, polymer NPs, and carbon-based NPs, which still show some uncertainty in intravital toxicity, exosomes are more natural, are safer, and show certain targeting properties as well as the capacity to cross the biological barriers like the blood–brain barrier. Immediately after nanoparticles are introduced into the vascular system, proteins present in the serum bind to their surfaces, leading to the establishment of a protein corona; , exosomes show weak nonspecific interactions with circulating proteins. , In this particular case, we previously found that human umbilical cord MSC-derived exosomes (hucMSC-Ex) could alleviate type 2 diabetes mellitus (T2DM) and achieve regulation of blood glucose through the restoration of insulin sensitivity in peripheral tissues and attenuation of β-cell injury, which could also be an advantage of hucMSC-Ex when applied in DFUs.

Intriguingly, exosomes as natural bioparticles can also be designed and engineered with selected therapeutic payloads, including drugs, immune modulators functional proteins, and RNAs. , Exosomes combine high stability with biocompatibility, making them suitable carriers for functional payloads, and their endogenous constituents may additionally promote superior biodistribution and safety. All these characteristics make exosomes an ideal “cell-free”-based therapeutic strategy for DFU.

Current findings suggest that microRNAs are crucial mediators in the overall process of diabetic wound repair. We are thus motivated to screen out key miRNA(s) in the hucMSC-Ex-mediated healing process. A miRNA with the provisional ID NC_000019.10_13474 selected by small RNA sequencing and bioinformatics analysis is studied. We interrogate whether the selected miRNA participates in exosome-mediated functional recovery of recipient cells and the underlying functioning mechanism. On this basis, combining the accurate, effective, and protective roles of exosomes with specific miRNA delivery by electroporation, which is easy to achieve, may be a promising strategy for potential DFU management and clinical application in the near future.

Materials and Methods

Ethics

All of the experiments in this study were approved by the Medical Ethics Committee of Jiangsu University (approval protocol number: 2020161) and the Ethics Committee of the Affiliated Hospital of Jiangsu University (approval protocol number: 201701).

Cell Culture

HucMSCs were collected from the Affiliated Hospital of Jiangsu University with maternal informed consent and processed within 6 h of acquisition. We isolated HucMSCs following established protocols and cultured them in low-glucose DMEM (Gibco) supplemented with 10% fetal bovine serum (FBS; Bioind). DFs were obtained from neonatal rat skin as described previously. Briefly, the skin tissues were minced into 1–2 mm pieces, transferred to 10 cm culture dishes, and maintained in α-MEM (Invitrogen) containing 10% FBS (Bioind) and 1% penicillin–streptomycin (Bioind). Cells migrating from the explants were collected after 7–9 days, and passages 3–7 were employed for subsequent experiments. DFs were visualized under an Olympus microscope and confirmed by Western blot analysis using fibroblast activation protein (FAP) and vimentin antibodies.

Isolation and Characterization of Exosomes

Exosomes derived from hucMSCs and HFL1 cells were isolated from the conditioned medium according to previously reported protocols. ,, Following exosome depletion, the supernatant was sterilized through a 0.22 μm filter and preserved at −80 °C for later use. For subsequent studies, hucMSC-Ex were applied at 400 μg/mL in vitro, while animals received an intraperitoneal dose of 10 mg/kg for in vivo experiments. The exosomal markers were verified through Western blot analysis. Exosome morphology and particle size were examined by TEM (transmission electron microscopy) and assessed by NTA (nanoparticle tracking analysis).

Exosome Labeling and Internalization

To investigate exosome uptake by cells, 0.25 mL of exosomes was first incubated with 5 μL of the red membrane dye Dil (Invitrogen) at 37 °C for 30 min for labeling. Following labeling, the exosome suspension was centrifuged at 1000g for 30 min and subsequently filtered through a 100 kDa molecular weight cutoff hollow fiber membrane to remove free dye. A total of 1 × 10⁴ cells were then plated onto cover glasses in 12-well plates and coincubated with the prepared Dil-labeled exosomes at 37 °C for 24 h to allow sufficient interaction between exosomes and cells. After incubation, cells were fixed with 4% paraformaldehyde and subjected to immunofluorescence staining using specific cellular markers (CD31 for HUVECs and vimentin for dermal fibroblasts) along with Hoechst 33342 (Sigma-Aldrich), enabling the simultaneous visualization of cell nuclei and exosome uptake. Finally, confocal microscopy was employed to capture fluorescence images and analyze the distribution and internalization of the exosomes.

Cell Migration Assays

Both a Transwell migration test and a scratch wound healing experiment were conducted to analyze the cellular motility. To allow cells to adequately respond, they were cultured in six-well plates with high glucose and hucMSC-Ex for 48 h prior to subsequent testing. A total of 2 × 10⁴ cells from each group were seeded into the upper compartment of a 24-well Transwell insert (8 μm pore, Corning) in 200 μL serum-free medium. To promote migration, 600 μL of complete medium was added to the bottom chamber. The exosome-depleted supernatant was filtered with a 0.22 μm membrane and maintained at −80 °C until it was required. Migrated cells attached to the lower surface were subsequently fixed in 4% paraformaldehyde and subjected to crystal violet staining. Using a Nikon microscope, stained cells were photographed and enumerated with a minimum of three randomly selected microscopic fields examined per group for accuracy. For scratch assays, cells were inoculated at a density of 2 × 10⁵ per well in six-well plates and grown to confluence. To create the wound, a 200 μL pipet tip was used to draw a straight scratch across the cell layer. Detached cells were gently removed by PBS washing, and photographs were collected at 0 and 24 h to track cell movement. Cell migration was quantified using the following formula: migration area (%) = (A ₀ – A _t )/A ₀ × 100%. Here, A ₀ denotes the wound area at 0 h, and A _t corresponds to the residual wound area at the indicated time.

Cell Proliferation Assay

The CCK-8 assay was performed in strict accordance with the manufacturer’s protocol to evaluate the proliferation of cells subjected to high-glucose injury and those treated with HucMSC-Ex. Cells were seeded at a density of 1500 per well in 96-well plates and incubated with 10 μL of the CCK-8 reagent (Vazyme) every 24 h. Absorbance at 450 nm was measured by using a microplate reader (BioTek) to quantify cell viability. Cells (2 × 10³) were plated in 35 mm culture dishes. Following a 9 day culture period, the cell colonies were fixed using 4% paraformaldehyde and subsequently stained with crystal violet. Colony number and size were subsequently recorded and photographed for analysis.

Tube Formation Assay

HUVECs were cultured in six-well plates with high glucose and hucMSC-Ex for 48 h as a pretreatment step before the assay. Briefly, prechilled Matrigel was added to the wells of a culture plate and allowed to polymerize at 37 °C to form HUVECs that were then seeded onto the polymerized Matrigel and incubated at 37 °C in a humidified atmosphere containing 5% CO₂ for 6–12 h, allowing the cells to organize into capillary-like tubular networks. Using a phase-contrast microscope, the emergence of tubular structures was monitored, and representative fields were photographed. Image analysis software was employed to evaluate angiogenesis by determining metrics, such as tube length, junction frequency, and complete loop formation.

Diabetic Wound Healing Evaluation

Eight week-old male Sprague–Dawley rats (180–220 g) were sourced from Jiangsu University’s Laboratory Animal Center and maintained on a 5 week high-fat feeding protocol with fat contributing 45% of total caloric intake. High-fat-diet-fed rats underwent a 12 h fasting period without access to food or water, after which streptozotocin (35 mg/kg, dissolved in 0.1 M citrate-buffered saline, pH 4.5) was administered via the tail vein to induce diabetes mellitus. Three days later, glucose levels were measured with test strips (Roche), and rats with levels exceeding 16.7 mM were considered diabetic. Rats were kept on a normal diet for a further 2 months to allow stabilization of the diabetic state, and blood glucose levels were verified again immediately before the wounding procedure. Following anesthesia with 2% pentobarbital sodium administered intraperitoneally, the dorsal skin was shaved, and 2 cm circular full-thickness wounds were generated. Two groups were randomly formed, with one receiving hucMSC-Ex (2 mg in 200 μL of PBS) and the other 200 μL of PBS, each injected subcutaneously at four sites around the wound. All animals were individually caged and had unlimited access to standard food and water for the duration of the study. Wound healing was monitored by photographing the wounds using a digital camera at designated time points. At the end of the 21 day postwounding period, rats were euthanized, and skin specimens were harvested. Histological staining (HE, Sirius Red), immunohistochemical analyses (IHC, IHF), and Western blotting were performed. Wound areas at different intervals were analyzed using the ImageJ software, and the percentage of closure was computed as wound closure (%) = (A ₀ – A_t )/A ₀ × 100%, providing a detailed quantitative assessment of tissue repair.

Histologic, IHC, and IHF Analysis

Following sacrifice on day 15, skin specimens were obtained and immediately immersed in 4% paraformaldehyde for fixation. Samples underwent graded ethanol dehydration and paraffin embedding and were sliced at a thickness of 4 μm. Standard protocols were followed to stain the sections with hematoxylin–eosin and Sirius Red. Sections were incubated with primary antibodies, including PCNA and CD31, at the specified dilutions, and detection was performed according to standard procedures. Stained sections were visualized and scanned by using a digital slide scanner (Pannoramic MIDI, 3D Histech). IHF was performed on skin sections following the established protocols. Sections were exposed overnight to a β-catenin antibody (CST, 1:100) at 4 °C and subsequently incubated with an FITC-conjugated secondary antibody (SAB, 1:200) for 60 min. Nuclear counterstaining was done using Hoechst 33342 (1:300), and fluorescent images were recorded with a Nikon microscope.

Dual Luciferase Reporter Assay

(1) An appropriate number of 293T cells were placed in each well of a 24-well plate. After incubation overnight, roughly half of the cells had fused. (2) Dual luciferase reporter plasmid and miR-13474 mimics were cotransfected. Four groupswild-type plasmid (pmirGLO-CPEB2-WT) + mimic NC, wild-type plasmid + miR-13474 mimics, mutant plasmid (pMIRGLO-CPEB2-MUT) + mimic NC, and mutant plasmid + miR-13474 mimicswere formed, and each group had six repeated wells. (3) After being incubated for 24 h, luciferase activity was measured. The previous medium was sucked out from each well, the solution was rinsed twice with PBS, and 150 μL 1× lysate buffer PLB was then added to each 24-well plate to cover the cells and spun for 30 min to make the cells lyse fully. After cell lysis, 10 μL of the cell lysate was added into an EP tube preloaded with 50 μL of LAR II (luciferase substrate), a biochemiluminescence analyzer, to measure. After reading for the first time, 50 μL of reaction stopping solution was added, the solution was blown and mixed two to three times, and the analyzer was used to read again. Each sample will have three values: the first reading, firefly luciferase activity; the second reading, Renilla activity; and the ratio between them. The ratio was recorded for statistical analysis.

Electroporation Treatment

The aim of electroporation is to increase the permeability of the exosome membrane under pulse high-voltage electrical currents, forming small pores through which miR-13474 mimics enter hucMSC-Ex so that hucMSC-Ex can overexpress miR-13474. The steps of electroporation experiment are briefly described as follows: (1) The FAM-miR-13474 mimic powder was dissolved in PBS to 1 μg/μL, and the concentration of hucMSC-Ex was diluted to 20 μg/μL. (2) The 5 μL dissolved FAM-MiR-13474 mimics and 45 μL hucMSC-Ex were thoroughly mixed, added into the electrode cup, and then placed in the electrode cup groove. The parameters of the electroporation instrument were as follows: perforation voltage (PPV), 110 V; perforation voltage duration (Pp on), 3 ms; resting time after perforation (Pp off), 10 ms; driving voltage (Pd V), 25 V; driving voltage duration (Pd on), 50 ms; resting time after driving (Pd off), 50 ms; and driving cycles (Pd Cycle N), 10. (3) Electroporation was performed repeatedly until sufficient hucMSC-Ex were obtained for experiments in vivo and in vitro. (4) The hucMSC-Ex were repeatedly concentrated and washed with a 100 kDa ultrafiltration centrifuge tube to remove the free fragments that did not enter the exosome. After filtration, these were sterilized and stored at −80 °C. Mimic NC was imported into hucMSC-Ex as the control. (5) The average size of hucMSC-Ex is 100 nm, which cannot be observed directly under an ordinary microscope. However, Dil-hucMSC-Ex with FAM-miR-13474 mimics could be coincubated with target cells for 12 h through a cell uptake experiment. Ultrahigh-resolution fluorescence microscopy could test the success of the electroporation experiment. The final preparations used in cell and animal experiments were 50-fold diluted, resulting in a working concentration of miR-13474-enriched hucMSC-Ex of 400 μg/mL.

Western Blotting

A RIPA buffer containing protease inhibitors was used to lyse cells, exosomes, and skin tissues for the extraction of total proteins.The protein lysates were incubated on ice for half an hour, with occasional mixing, and subsequently spun at 12,000g for 15 min. A NanoDrop One instrument (Thermo Fisher Scientific) was used to measure protein absorbance at 280 nm, allowing for normalization of protein amounts across samples. Using 12% SDS-PAGE, 150 μg of protein per sample was electrophoresed and moved onto PVDF membranes with a semidry transfer setup. Membranes were incubated in 5% nonfat milk in TBS/T for 2 h at room temperature to block nonspecific sites. Afterward, they were exposed overnight at 4 °C to primary antibodies for FAP, vimentin, CD9, CD63, ALIX, calnexin, PCNA, Bcl-2, and Bax, with β-actin as the internal standard. Upon completion of primary antibody incubation, membranes were washed in TBS/T thrice and then subjected to a 1 h treatment with HRP-tagged goat antirabbit or antimouse IgG antibodies (Invitrogen, 1:2000) at room temperature. Subsequent to washing, protein bands were revealed using an ECL kit (General Electric) and recorded digitally. Quantification of band density was performed in ImageJ, normalizing the values to β-actin.

QRT-PCR

The total RNA from the cultured cells was obtained with a commercial RNA extraction kit according to the manufacturer’s guidelines, ensuring high-quality and intact RNA. Before synthesizing cDNA, RNA samples were evaluated for quality and concentration with a NanoDrop spectrophotometer. cDNA was generated using a reverse transcription kit (Invitrogen) according to the supplied protocol, and quantitative PCR with SYBR Green detection was employed to quantify target gene expression. β-Actin served as a control gene to correct for intersample differences and standardize target gene measurements. Gene-specific primers were obtained from Invitrogen (Shanghai, China), with detailed sequences provided in Table S1. All amplification reactions were conducted in triplicate, and expression ratios were calculated using the comparative 2^–ΔΔCt method.

Statistical Analysis

Data are expressed as the mean ± SD. Statistical evaluations were conducted with GraphPad Prism 5.0 (GraphPad, San Diego, USA). Unpaired t tests with two tails assessed differences between two groups, while one-way ANOVA accompanied by Bonferroni post hoc testing was used for comparisons across three or more groups. In cases where measurements were repeated over time, repeated-measures ANOVA was applied to account for intrasubject variability. A threshold of p < 0.05 was used to determine statistical significance. Significance annotations in figures are indicated by *, **, and ***, corresponding to p < 0.05, p < 0.001, and p < 0.0001.

Results

High Glucose Exacerbates Skin Destruction and DF Dysfunction

Even though diabetes disrupts multiple processes involved in wound repair, elevated glucose levels seem to significantly influence healing outcomes, as decreasing hyperglycemia promotes better tissue recovery. SD rats underwent T2DM modeling through the combination of a high-fat diet and streptozotocin treatment, and this condition was preserved for a duration of 2 months (Figure A). T2DM rats showed a persistent high blood glucose level over 16.7 mmol/L (Figure B) and a distinct loss of weight (Figure C), along with typical diabetic symptoms such as polydipsia, hyperphagia, and polyuria. Long-term hyperglycemia caused a series of pathologic changes in cutaneous tissue structures. Subcutaneous fat is much thinner and even disappeared in T2DM rats, which would be closely related to the weight reduction. Research suggests that hyperglycemia can independently induce both the production of proinflammatory cytokines and the activation of inflammatory signaling. An inflammatory infiltration was also noticed in the epidermis and dermis of T2DM rats. The thickness of the epidermis layer was reduced significantly in diabetic skin, with obscure multilayer epithelium features, while the stratum corneum almost disappeared (Figure D). These abnormal changes collectively inflict a vulnerable status on the diabetic skin. High glucose treatment in in vitro studies also showed impaired viability, proliferation, and migration capacity in DFs, which are the major cellular agents in the proliferative phase of wound healing (Figure E–G). Additionally, HUVECs for vessel regeneration were functionally impaired (Figure S1). Together, the above findings confirmed that hyperglycemia exacerbated the destruction of skin tissues and dysfunction of important repair-related cells.

1.

High glucose exacerbates skin destruction and DF dysfunction. (A) Schematic diagram of the T2DM rat model establishment. (B) Blood glucose was monitored in T2DM rats (n = 6, ***p < 0.001). (C) Body weight was monitored in T2DM rats (n = 6, ***p < 0.001). (D) HE staining to observe the structure of skin in normal rats and skin in diabetic rats. (E) CCK8 assay was used to detect the cell viability of DFs after 96 h of treatments with different concentrations of high glucose (n = 6, *p < 0.05, ***p < 0.001). (F) A Transwell assay was used to detect the migration ability of DFs after 48 h of treatments with different concentrations of high glucose (scale bar = 200 μm). (G) Colony formation assays were conducted to determine how exposure to distinct levels of high glucose affected DF proliferation.

HucMSC-Ex Accelerate Wound Healing in Diabetic Skins

From the supernatants of hucMSCs and HFL1 cultures, exosomes were collected and subsequently confirmed by analyzing their morphologies, particle sizes, and marker proteins. Both hucMSC-Ex and HFL1-ex were observed to have vesicles around 120 nm with a spherical or cup-like morphology, expressing typical markers CD9, CD81, and Alix, while no signal was detected for calnexin and albumin (Figure S2). The T2DM rats were subjected to a diabetic wound creation and received different treatments after being wounded. Wound closure and other functional indicators were assessed up to 21 days postwounding (Figure A). Wound closure was obviously delayed in hyperglycemic rats. There is no visible improvement in wound size of T2DM rats that received PBS treatment within 7 days. HucMSC-Ex treatment distinctly promoted the wound healing compared to the exosome-free conditioned medium (Ex-free CM), HFL1-Ex, and clinically used alginate dressing (Figure B). It is worth noting that the repair effect of hucMSC-Ex was most striking from days 7 to 10 (Figure C), in accordance with the reported study that showed that fibroblasts surrounding the wound displayed swift activation after injury, and α-SMA reached its greatest expression approximately 7 days later. Evaluation of PCNA, Bcl-2, and Bax levels confirmed that hucMSC-Ex support skin cell proliferation, which in turn enhances wound healing efficiency (Figure D). The results from the professional pathological scanning system HALO show that the repair group exhibited significantly better recovery in terms of epidermal and dermal repair area, number of hair follicles and sweat glands, nerve fiber bundles, and tactile corpuscles compared to the control group (Figure E). Histological evaluation of wounds showed a complete re-epithelialization with barely visible inflammatory cells and red blood cells in hucMSC-Ex-treated DFU rats (Figure F). Furthermore, hucMSC-Ex treatment led to an obviously active proliferation in epidermal basal layer cells (Figure G), and large amounts of wavy-shaped collagen fibers were also seen to be deposited and well organized (Figure H). Neovascularization with more vessel numbers and complete lumen structure, bringing nutrition and oxygen to the wound, was also enhanced after hucMSC-Ex treatment (Figure I). These results collectively confirmed the promoting role of hucMSC-Ex in diabetic skins, which was even better than that of Ex-free CM, exosomes from other source, and certain clinical materials.

2.

HucMSC-Ex accelerate wound healing in diabetic skins. (A) Schematic diagram of the DFU rat model establishment. (B) General view of the wounds in DFU rats. (C) Line chart for percentages of wound closure (n = 6, *p < 0.05, **p < 0.01, ***p < 0.001). (D) Western blot assays were applied to evaluate the levels of proteins involved in proliferation and apoptosis within the DFU model. (E) A professional pathological scanning system, HALO, was used to reflect the wound healing status, including the epidermal and dermal repair area, numbers of hair follicles and sweat glands, nerve fiber bundles, and tactile corpuscles. (F) HE staining of skin sections (scale bar = 100 μm). (G) PCNA immunohistochemical staining of skin sections (scale bar = 100 μm). (H) Sirius red staining was used to detect the distribution of collagen (scale bar = 100 μm). (I) CD31 immunohistochemical staining of skin sections (scale bar = 100 μm).

HucMSC-Ex Protect Dermal Fibroblasts from High-Glucose-Induced Functional Damage

We first coincubated red fluorescence-labeled HucMSC-Ex with DFL cells for 12 h. Confocal fluorescence microscopy clearly showed that DFs cells were able to internalize a large number of exosomes (Figure A). In the in vitro cell models, we established six groups: 5.5, 30, and 45 mmol/L glucose groups, along with the corresponding hucMSC-Ex treatment groups under each glucose concentration. We found that hucMSC-Ex significantly promoted DF proliferation (Figure B,C,D), enhanced DF cell viability (Figure .8C), and improved DF migration ability (Figure .8D) regardless of the glucose concentration. Interestingly, we found that, under the same treatment conditions (the same concentration of exosomes and the same time duration), DFs were likely to take in more exosomes than HUVECs, which were also a vital cell type in wound healing (Figure A, Figure S3A). Although less quantity was taken up, hucMSC-Ex still protect HUVECs from high glucose damage and promote their tube forming capacity (Figure S3B–D). HucMSC-Ex exerted immunoregulatory actions by inhibiting proinflammatory mediators (IL-1, IL-6, and IL-8) while promoting the anti-inflammatory cytokine IL-10. (Figure E).

3.

HucMSC-Ex protect dermal fibroblasts from high-glucose-induced functional damage. (A) Confocal fluorescence microscopy was used to observe the uptake of hucMSC-Ex by DFs (red fluorescent membrane dye Dil was used to label hucMSC-Ex, green fluorescence was for DFs characteristic protein α-SMA, and Hoechst 33342 was used to label nuclei). (B) Colony formation assay was performed to assess DF proliferation in response to high glucose and HucMSC-Ex. (C) The viability of DFs subjected to high glucose and hucMSC-Ex for 96 h was measured by a CCK-8 assay (n = 6, **p < 0.01, *p < 0.001). (D) DF migration was quantified by a Transwell assay following 48 h of treatment with high glucose and hucMSC-Ex (scale bar = 200 μm). (E) Expression levels of proinflammatory factors (IL-1, IL-6, and IL-8) and anti-inflammatory cytokine (IL-10) were detected by qRT-PCR (n = 3, *p < 0.05, **p < 0.01, ***p < 0.001).

Increased miR-13474 in hucMSC-Ex Contributes to Diabetic Wound Healing

Research has confirmed that miRNAs possess a remarkable capacity to govern complicated regulatory circuits. We thus performed small RNA sequencing to screen out the differentially expressed miRNAs between hucMSC-Ex and HFL1-Ex. Among the selected upregulated miRNA candidates, NC_000019.10_13474 (miR-13474), the predicted miRNA showed a remarkable increase in hucMSC-Ex (Figure A–C). Subsequent identification using qRT-PCR to detect the content of exosomal miR-13474 was consistent with the results of small RNA sequencing (Figure D). We also detected the content of miR-13474 in different cell origins, while hucMSC showed the highest, which indicated that the elevating exosomal miR-13474 may be attributed to its mother cell (Figure E). Adding hucMSC-Ex to the DF culture medium resulted in an obvious increase in miR-13474 expression (Figure F). These findings suggested that exosomes could act as a vehicle to enrich and transport miR-13474 to target cells. Further detection of miR-13474 in DFU rats, cell models, and DFU patients showed an impaired expression of miR-13474 in the high-glucose environment, which could be saved by hucMSC-Ex treatment (Figure G). We therefore centered on miR-13474 as the impotant participator in hucMSC-Ex-mediated diabetic wound healing. miR-13474 mimics facilitated the DF cell viability, proliferation, migration, and ECM secretion capacity, while the miR-13474 inhibitor showed the opposite (Figure H–J, Figure S4). Additionally, miR-13474 also promotes the biological function of HUVECs, which might also help wound healing (Figure S5). Collectively, the increased miR-13474 in hucMSC-Ex contributed to diabetic wound healing.

4.

Increased miR-13474 in hucMSC-Ex contributes to diabetic wound healing. (A) Heat map of miRNA sequencing in hucMSC-Ex and HFL1-Ex. (B) Scatter plot with cutoff at twofold within exosomes derived from hucMSC-Ex and HFL1-Ex. (C) Schematic structure of NC_000019.10_13474 (miR-13474). (D) RNA sequencing results of miR-13474 were verified by qRT-PCR in hucMSC-Ex and HFL1-Ex (n = 3, *p < 0.05, **p < 0.01, ***p < 0.001). (E) The expression level of miR-13474 in selected cell lines was detected by qRT-PCR. (F) The expression level of miR-13474 in DFs treated with hucMSC-Ex was detected by qRT-PCR (n = 3, ***p < 0.001). (G) The expression level of miR-13474 in DFU rats, cells, and patients were detected by qRT-PCR (DFU rats, n = 6; cell lines, n = 3, DFU patients, n = 10; **p < 0.01, ***p < 0.001). (H) A CCK8 assay was used to detect the changes in the cell viability of DFs transfected with miR-13474 mimics or inhibitor for 5 consecutive days (n = 6, ***p < 0.001). (I) The proliferative capacity of DFs transfected with miR-13474 mimics or inhibitors was evaluated using a colony formation assay. (J) A Transwell assay was used to detect the changes in the migration ability of DFs transfected with miR-13474 mimics or inhibitor (scale bar = 200 μm).

HucMSC-Ex Promote Diabetic Wound Healing by Delivering miR-13474 to Target CPEB2/TWIST1 in Cutaneous Tissue and Cells

Conjoint consideration of bioinformatic prediction, the high sequence matching, favorable binding stability, and expressive abundance in skin. CPEB2, a highly conserved cytoplasmic polyadenylation-element-binding protein, was selected as the preferential target regulated by miR-13474. Currently, little is known about CPEB2 except for its role in suppressing certain transcription factor expression by binding to the mRNA 3′UTR region. We thus assumed that exosomal miR-13474 potentially targeted CPEB2/TWIST1 to work (Figure A). To verify this hypothesis, we primarily identified the expressive correlation of miR-13474 and its target axis. We found that miR-13474 mimics could inhibit the expression of CPEB2 and upregulate the expression of TWIST1 at both the mRNA and protein level, while the effect of the miR-13474 inhibitor is just the opposite (Figure B,C). HucMSC-Ex treatment played a consistent role as miR-13474 mimics (Figure D). In DFU rats and patients, the downregulated expression of CPEB2 accompanied by the upregulation of TWIST1 was observed in the situation where miR-13474 is relatively abundant, such as hucMSC-Ex treatment in rats and the wound edge area rather than the wound area in DFU patients (Figure E,F, Figure S7). The results of the dual luciferase reporter assay subsequently proved the direct binding and regulation between miR-13474 and CPEB2, which means that CPEB2 was the direct target of miR-13474 (Figure G). TWIST1 knockdown did impair the proliferative, migrative capacity and cell viability of DFs, which could not be rescued by miR-13474 mimics (Figure H–J). It is worth mentioning that the miR-13474/CPEB2/TWIST1 axis consistently came into play in HUVECs (Figure S6). The results together indicated that miR-13474 enhanced skin cell function by targeting CPEB2/TWIST1.

5.

HucMSC-Ex function by delivering miR-13474 to target CPEB2/TWIST1 in cutaneous tissues and cells. (Ai) The predicted functioning axis of miR-13474. (Aii) The KEGG enrichment top 20 for the targeted function of CPEB2. (B, C) The expression level of miR-13474, CPEB2, and TWIST1 in DFs transfected with miR-13474 mimics and inhibitor was detected by qRT-PCR (n = 3, *p < 0.05, **p < 0.01, ***p < 0.001) and Western blot. (D) The mRNA level of CPEB2 and TWIST1 in DFs was detected after hucMSC-Ex treatment (n = 3, *p < 0.05, **p < 0.01). (E) Immunohistochemical staining for detecting the expression levels of CPEB2 and TWIST1 in the DFU rats (scale bar = 100 μm). (F) The expression level of CPEB2 and TWUST1 in DFU rats and patients were detected by qRT-PCR (DFU rats, n = 6; DFU patients, n = 10; *p < 0.05, ***p < 0.001). (G) A dual-luciferase reporter gene assay in 293T cells verified that miR-13474 could directly bind to CPEB2. (H) A CCK8 assay was used to detect the cell viability of DFs cotransfected with TWIST1 siRNA and miR-13474 mimics at 96 h (n = 6, ***p < 0.001). (I) A cell colony formation assay was used to detect the proliferation ability of DFs cotransfected with TWIST1 siRNA and miR-13474 mimics. (J) To examine the effect of cotransfection with TWIST1 siRNA and miR-13474 mimics on DF migration, a Transwell assay was employed (scale bar = 200 μm).

miR-13474-Modified hucMSC-Ex Enhanced the Biofunctions of DFs and HUVEC cells

Given the therapeutic role of miR-13474, we assumed that miR-13474-modified hucMSC-Ex could be more effective for DFU management. miR-13474 mimics or the negative control was thus encapsulated into hucMSC-Ex by electroporation to treat both skin cell models and DFU rats (Figure A). There is no obvious difference between exosomes with or without electroporation in the morphology and size distribution (Figure A). The coincubation of tailored exosomes and DFs showed that miR-13474 could be taken into recipient cells (Figure B) and reached a significantly increasing content (Figure C). Compared with control exosomes, miR-13474-modified hucMSC-Ex showed a better effect on promoting cell viability (Figure D). Moreover, miR-13474-enriched hucMSC-Ex could also enhance the proliferative capacity of DFs (Figure E) as well as the migration capacity in DFs (Figure F). The same prohealing effects were also observed in HUVECs (Figure S8), and the tube forming ability of HUVECs was also enhanced by the tailored exosomes (Figure S8). In DFU rats, miR-13474 modification enhanced the repairing effect of hucMSC-Ex (Figure A,B). More importantly, compared with PBS injection, hucMSC-Ex did reorganize the impaired skin tissue and promoted neovascularization ; however, there was still a thickening of the stratum spinosum layer indicating an incomplete repair process. By contrast, exosomes encapsuled with miR-13474 treatment achieved the desire effect of making the skin return to relatively normal conditions with normal epidermal layers and thickness as well as abundant accessory structures that were absent in the control-exosome-treated skin (Figure C). The detection of proteins extracted from DFU rats also showed that miR-13474-modified exosomes could upregulate the expression level of proliferative related proteins while downregulate that of apoptotic ones (Figure D). The miR-13474-modified exosome-treated DFU rat skin also presented a higher miR-13474 expression level, indicating the possibility that there might be an extraneous reception of miR-13474 transferred by exosomes (Figure E). Consistently, the expressive correlation of miR-13474/CPEB2/TWIST1 was also confirmed in both skin cell models and DFU rats (Figures G and F,G).

6.

MiR-13474-modified hucMSC-Ex enhanced the biofunctions of DFs and HUVEC cells. (A) The diagram illustrates how HucMSC-Ex were engineered by electroporation to carry miR-13474 mimics or a corresponding negative control. (B) HucMSC-Ex were labeled with Dil in advance. FAM-miR-13474 mimics were encapsuled into Dil-labeled hucMSC-Ex by electroporation. Dil-HucMSC-Ex-FAM-miR-13474 mimics were incubated with DFs for 12 h; finally, the transport efficiency was observed under an ultrahigh-resolution fluorescence microscope. (C) The expression level of miR-13474 in DFs treated with Ex-miR-13474 mimics was detected by qRT-PCR (n = 3, ***p < 0.001). (D) DF cell viability was measured by a CCK-8 assay after 96 h of incubation with Ex-miR-13474 mimics (n = 6, *p < 0.001). (E) DF proliferation after Ex-miR-13474 mimic exposure was quantified using a colony formation assay. (F) A Transwell assay was used to detect the migration ability of DFs treated with Ex-miR-13474 mimics (scale bar = 200 μm). (G) Western blot was used to detect the protein levels of CPEB2 and TWIST1 in DFs treated with Ex-miR-13474 mimics.

7.

miR-13474 modification enhanced hucMSC-Ex functions in diabetic wound healing. (A) General view of the wounds. (B) Line chart for percentages of wound closure (n = 6, *p < 0.05, **p < 0.01, ***p < 0.001). (C) HE staining of skin sections (scale bar = 100 μm). PCNA immunohistochemical staining of skin sections (scale bar = 100 μm). Sirius red staining was used to detect the distribution of collagen (scale bar = 100 μm). CD31 immunohistochemical staining of skin sections (scale bar = 100 μm). (D) Western blot was used to detect the expression levels of proliferation-and-apoptosis related proteins in the DFU rats. (E–G) The expression levels of miR-13474, CPEB2, and TWIST1 in the DFU model were detected by qRT-PCR (n = 3, *p < 0.05, ***p < 0.001).

Discussion

Wound closure entails a multifaceted biological process that rebuilds tissue organization and supports its functional capacity. Generally, there are four overlapping phases of wound healing, which usually begins with hemostasis and is rapidly followed by inflammation, proliferation, and remodeling phases. However, the conditions caused by high glucose and its toxic products in diabetic wounds impair each phase of wound healing and delay the whole repair process. Wounds under diabetic conditions are more likely to feature prolonged inflammation, which hinders the development of mature granulation tissue and weakens tensile strength. Dermal fibroblasts are widely recognized as cells that produce collagen-rich ECM scaffolding and are the main force in the proliferation and remodeling phases. The proliferation and migratory capacities of fibroblasts are essential factors influencing tissue regeneration after injury. In this current study, we found that hucMSC-Ex reduced the inflammatory infiltration (such as neutrophil infiltration) in diabetic rats and enhanced the proliferative, migrative, and collagen-secreting capacities of dermal fibroblasts, which provided the basis for the following repair process. HucMSC-Ex also functioned on vascular endothelial cells and promoted neovascularization to bring more nutrients and oxygen to accelerate wound healing. Intriguingly, an active proliferation status of stratum basale cells, where stem cells and transient proliferating cells reside, was observed in hucMSC-Ex-treated DFU rats. The results indicated that the hucMSC-Ex treatment of the chaotic niche might also activate a range of stem cells to replenish and replace the dead cells. The epidermal layer, especially the stratum spinosum, was obviously thickened after hucMSC-Ex treatment; however, the persistent proliferation holds the risk for scar formation and analogous function impairment. In contrast, the hyperplasia had already returned to silence with a favorable differentiation and almost normal structure after miR-13474-encapsuled hucMSC-Ex treatment, which showed a faster and lower-risk propulsion and more complete healing process mediated by miR-13474 modification. However, it further remains unclear how the balance between proliferation and differentiation is achieved.

As dermal fibroblasts, vascular endothelial cells, and even the epidermal stem cells could be the potential targets of hucMSC-Ex, the exosome intake varied in different recipient cells under the same conditions of quantity and incubating time. The difference could be attributed to heterogeneity as well as the preference of exosomes. , The preferential uptake of exosomes by certain cells is thought to be mediated by the recognition of exosomal surface proteins by the corresponding receptors on the plasma membrane. Integrins α6β4 and α6β1 on exosomes are implicated in lung metastasis, while αvβ5 is associated with metastasis to the liver. Exosomal tetraspanins could also regulate cell targeting to promote exosome docking and uptake by selected recipient cells, such as endothelial cells and pancreatic cells. Exosomes presenting CD47, which acts as a “do not eat me” signal, have been reported to strongly suppress phagocytosis by monocytes. It is acknowledged that exosomes did display a preference when taken up by certain tissue or cell types. In our research, DFs showed more intake of exosomes than HUVECs, while hucMSC-Ex played the most prominent role in days 7–10, in accordance with the reported study that, upon wounding, fibroblasts surrounding the wound responded promptly to injury, showing maximal α-SMA expression 1 week after wounding. However, the exact cellular and molecular basis for the specific targeting of recipient cells requires further exploration. It is also worth mentioning that although not so large in quantitative terms, hucMSC-Ex were still demonstrated to be effective on glucose-damaged HUVECs.

Exosomes serve as carriers of lipids, proteins, and nucleic acids, including mRNA, miRNA, and lncRNA, with microRNAs recognized as key functional components responsible for many of the therapeutic effects of exosomes. MiRNA diversity within an organism is usually associated with morphological complexity, while genomic density is not, which indicates that each miRNA has a range of downstream target genes. The capacity to influence numerous targets at once allows them to serve as highly effective regulators of genetic and signaling networks. In our study, we found there were in total 322 differently expressed miRNAs between hucMSC-Ex and HFL1-Ex, among which 102 miRNAs were upregulated, while the other 220 ones were downregulated. miR-13474 was selected and identified as the most promising cargo in exosomes to mediate the healing process for the following reasons: (1) Based on the results of small RNA sequencing, miR-13474 was significantly enriched and had a highlighted expression in hucMSC-Ex with a high reliability score. Subsequent validation of the miR-13474 expression level by qRT-PCR showed a consistent result. The distinct miRNA signatures in exosomes compared to donor cells suggest that miRNA loading into these vesicles is an orchestrated process controlled by specific cellular mechanisms. (2) miR-13474 is a predicted miRNA without any functional research. (3) HucMSC as the mother cell of therapeutic exosomes displayed a higher content of miR-13474 than other cell types, indicating that the enrichment in exosomes might be traced back to the origin cell. (4) HucMSC-Ex treatment did obviously increase the miR-13474 content in target cells, suggesting that miR-13474 might be transferred to recipient cells. Herein, whether the increased expression of miR-13474 was caused by exosome-mediated transportation or alterant gene expression networks of target cells regulated by exosome treatment still requires more evidence. The following functional experiments verified that high-glucose stimulation impaired the expression of miR-13474, and hucMSC-Ex treatment as well as miR-13474 mimics could make injured skin cells and tissues recover. Although the role of miR-13474 was demonstrated, we still prefer that there would be other miRNAs that would function in the healing process. Whether there will be a synergistic effect of other upregulated miRNAs or even the downregulated miRNAs remains unclear.

To boost the therapeutic potential of exosomes, exogenous molecules can be incorporated in addition to their native contents. Methods include modifying the parent MSCs to produce loaded exosomes or introducing therapeutic molecules directly into the exosomes. Studies indicate that direct incorporation into recipient cells is significantly more efficient than relying on exosome-mediated transfer, which we adopted in our research as well. However, defects like the possibility of triggering aggregation of RNA cargo and the varying work conditions between different laboratories when using electroporation should also be taken into account. To enhance molecular loading into exosomes, techniques including pH gradient modification have been investigated; these methods can match the efficiency of electroporation while preserving the integrity of sensitive nucleic acid cargo. In this regard, better ways to load miRNA cargos into exosomes are worth further exploration.

Conclusions

In conclusion, we found that hyperglycemia exacerbated skin inflammatory infiltration, structure destruction, and skin cell dysfunction. HucMSC-Ex could accelerate diabetic wound healing by delivering predicted miR-13474 to target the CPEB2/TWIST1 axis. MiR-13474 was demonstrated to be positively reversible and modified into hucMSC-Ex to acquire optimized therapeutic effects. Our study provided a research basis for further application of hucMSC-Ex in clinical diabetic wound treatment.

The data of the study are available from the corresponding author upon reasonable request.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsabm.5c01352.

• Additional experimental details including photographs of experimental results (PDF)

§.

H.S., X.H., and Y.Lu contributed equally to this manuscript.

H.S.: Conceptualization [lead], Data curation [lead], Formal analysis [lead], Investigation [lead], Methodology [lead], Funding acquisition [lead], Resources [lead], Validation [lead]. X.H.: Data curation [lead], Formal analysis [equal], Investigation [equal], Validation [equal]). Y.Lu: Data curation [equal], Formal analysis [equal], Investigation [equal], Validation [lead]. Y.Li: Data curation [equal], Formal analysis [equal], Investigation [equal], Validation [equal]. H.L.: Conceptualization [equal], Supervision [lead], Validation [equal], Resources [lead]). H.Q.: Conceptualization [equal], Supervision [lead], Resources [lead]. W.X.: Conceptualization [lead], Supervision [lead], Validation [equal], Resources [equal].

This work was supported by the National Natural Science Foundation of China (Grant 82472573), the Natural Science Foundation of Jiangsu Province (the Basic Research Program, Grant BK20241925), and The Technology Development Project of Jiangsu Province (Grant HX20240572).

All of the experiments in this study were approved by the Medical Ethics Committee of Jiangsu University (approval protocol number: 2020161) and the Ethics Committee of Affiliated Hospital of Jiangsu University (approval protocol number: 201701).

All authors have approved the publication of this manuscript.

The authors declare no competing financial interest.