Extracellular vesicle-derived lncRNA VIM-AS1 promotes diabetic wound healing by promoting glycolysis and alleviating cellular senescence

Abstract

Aims

Diabetic wound healing is a significant challenge due to impaired cellular functions, and current therapeutic approaches often prove inadequate. This study aims to explore the role of extracellular vesicles (EVs) derived from human umbilical mesenchymal stem cells (HuMSCs), particularly focusing on their associated long non-coding RNAs (lncRNAs), in promoting diabetic wound repair.

Methods

To investigate this, we employed lncRNA sequencing of EVs, created reprogrammed EVs, and utilized a diabetic rat  model. The impact of HuMSCs-derived EVs on fibroblast glycolysis, proliferation, and migration was assessed, along with the function of lncRNA VIM-AS1 in glucose metabolism via the PPAR-γ pathway.

Results

Our results demonstrate that HuMSCs-derived EVs enhance glycolysis in fibroblasts, which is essential for effective wound healing. We identified lncRNA VIM-AS1 as a pivotal regulator that not only promotes fibroblast proliferation and migration but also significantly enhances endothelial cell function, specifically regarding angiogenesis and tissue vascularization. Furthermore, EVs-derived lncRNA VIM-AS1 was found to reduce reactive oxygen species (ROS) levels, thereby mitigating oxidative stress and cellular senescence in both fibroblasts and endothelial cells. In vivo experiments in rat models confirmed the capacity of EVs-derived lncRNA VIM-AS1 to improve diabetic wound healing.

Conclusions

This study highlights the therapeutic potential of HuMSCs-derived EVs and specifically lncRNA VIM-AS1 as innovative approaches to address the challenges of tissue repair in diabetic conditions, offering promising strategies for enhancing wound healing efficacy.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13287-025-04451-x.

Introduction

Diabetes is an increasingly severe chronic metabolic disease that poses a significant challenge to global public health, with its incidence continuing to rise [1]. Patients with diabetes often experience a range of complications, among which impaired wound healing stands out as a particularly serious issue [2]. The wound healing process in diabetic patients is markedly delayed, often taking two to four times longer than in healthy individuals. This delay not only diminishes patients’ quality of life but can also lead to serious complications such as infections, amputations, and even life-threatening conditions [1, 2]. Alarmingly, despite the availability of various treatment options for chronic wounds, the success rate in treatment remains relatively low, with around 28% of patients eventually needing lower limb amputation [3]. Therefore, it is imperative to explore and create innovative therapeutic approaches aimed at enhancing the healing of diabetic wounds.

Research has shown that several factors critically influence the healing process of diabetic wounds, including hyperglycemia, poor tissue perfusion, immune dysfunction, and metabolic irregularities [4]. Specifically, in the context of hyperglycemia, the functionality of fibroblasts and vascular endothelial cells is significantly impaired, leading to diminished cell proliferation and migration, which ultimately delays the wound healing process [5]. Additionally, cellular senescence emerges as another key factor associated with impaired wound healing in diabetes [6]. Studies have found that the proportion of senescent cells significantly increases in diabetic mouse models, resulting in a population of cells that not only fails to proliferate normally during wound healing but also secretes factors that prolong inflammation and delay recovery [7]. Therefore, strategies aimed at regulating cellular glycolysis to promote the functionality of fibroblasts and vascular endothelial cells, while mitigating cellular senescence, can be crucial for enhancing wound healing in diabetic patients. Among the potential solutions, human umbilical mesenchymal stem cells (HuMSCs) have become a promising alternative owing to their distinctive regenerative capabilities.

The HuMSCs have attracted considerable attention in the realm of regenerative medicine because of their ethical sourcing from discarded umbilical cords, multipotent differentiation potential, and low immunogenicity [8]. Research suggests that HuMSCs enhance wound repair primarily through a paracrine mechanism, in this process, they release extracellular vesicles (EVs), which are essential for cellular communication between cells [9]. These HuMSCs-derived EVs (HuMSCs-EVs) have demonstrated significant efficacy in enhancing the expansion, movement, and differentiation of fibroblasts and vascular endothelial cells, as well as mitigate cellular senescence [10–12]. In the context of diabetic wound healing, EVs have been recognized for their potential to promote healing. These EVs are abundant in bioactive molecules, including proteins, lipids, and RNA, which collectively contribute to tissue regeneration by enhancing cellular responses to injury [13]. Among the various components found within these EVs, long non-coding RNAs (lncRNAs) have recently been recognized as essential managers of cellular metabolism and function [14, 15]. Recent findings suggest that lncRNAs are vital regulators of glycolysis, a critical metabolic pathway that enhances energy availability for cellular processes. This regulation not only provides promising opportunities for advancing diabetic therapies but also alleviates cellular senescence, thereby improving cell viability and functionality during the healing process [16, 17]. However, it remains unclear which specific lncRNAs within EVs play a critical role in promoting diabetic wound healing.

This research seeks to explore the functions and molecular mechanisms of key lncRNAs derived from HuMSCs-EVs in promoting diabetic wound healing through employing lncRNA sequencing of EVs, establishing a diabetic mouse model for wound healing, and constructing reprogrammed EVs. Our research focuses on how these lncRNAs improve the process of wound repair in mice with diabetes by regulating cellular glycolysis and alleviating cellular senescence. This study will provide a theoretical basis for new therapeutic strategies in diabetic wound healing and open new avenues for the application of EVs in regenerative medicine.

Method materials

Cell culture

The umbilical cord mesenchymal stem cells involved in this research received ethical approval from the First Affiliated Hospital of Guangxi Medical University. Fresh human umbilical cords were sourced in a sterile environment from healthy donors who underwent cesarean deliveries at full term at the First Affiliated Hospital of Guangxi Medical University. The HuMSCs were retrieved, grown, and validated following the techniques described in earlier research [39]. The human umbilical vein endothelial cells (HUVECs) (catalog was # CP-H082) and human dermal fibroblasts (HDFs) (catalog was # CP-H103) were purchased from Wuhan Pricella Biotechnology Co., Ltd. (Wuhan, China) (http://www.whprocell.com/Products-36640225.html). HDFs and HUVECs were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) provided by Hyclone. The medium was enriched with 10% fetal bovine serum (FBS) from Gemini, 1% penicillin/streptomycin mixture supplied by Solarbio (China), and 5 mM glucose to simulate normal-glucose conditions. High glucose (HG) treatment was performed using a glucose concentration of 100 mM. These cells were maintained in an incubator set to 37 °C with a humidified atmosphere containing 5% CO₂.

Engineering HuMSCs  with lentivirus carrying LncRNA VIM-AS1

Recombinant shuttle plasmid with distinct short hairpin RNA (shRNA) sequences aimed at lncRNA VIM-AS1 and packaging plasmids were provided by Shanghai Genechem Co., Ltd.. The targeting sequences for lncRNA VIM-AS1 used were as follows: #sh lncRNA VIM-AS1: GTCCCTACTACTTCAACCGAT. Using Lipofectamine 3000, these plasmids were introduced into 293 T cells, which facilitated the production of lentiviral particles harboring lncRNA VIM-AS1 (Lv-lncRNA VIM-AS1). An analogous approach was utilized to generate the negative control lentiviral vectors (Lv-NC). UCMSCs were then transfected with either Lv-LncRNA VIM-AS1 or Lv-NC to generate sh-lncRNA VIM-AS1 engineered HuMSCs and control HuMSCs, respectively. Following transfection, to select stably transfected HuMSCs, puromycin (2 μg/mL) was used. The successfully transfected cells were then observed using a fluorescence microscope.

Collection and characterization of HuMSCs-derived extracellular vesicles (HuMSCs-EVs)

Extracellular vesicles from the various groups were extracted following the methods described in earlier studies [40]. In particular, EVs were extracted from the supernatant of cultured medium through the process of ultracentrifugation. When the cell confluence reached approximately 80%−90%, the culture medium was harvested and subjected to centrifugation at 400 g for a duration of 6 min, followed by a second centrifugation at 2000 g for a duration of 30 min to eliminate cellular debris. The supernatants were subsequently passed through a 0.22 μm filter to remove nano-scale contaminants that are not small extracellular vesicles (SEVs). The resulting solutions of supernatant underwent two rounds of ultracentrifugation at 120,000 g for 90 min at 4 °C. The isolated EVs were subsequently resuspended in PBS for experimental purposes or stored at −80 °C for future use. The nanoparticle tracking analysis (model ZetaView PMX 110), known as NTA, was utilized for characterizing the size and distribution of EVs, and the transmission electron microscope (manufactured by HITACHI), known as TEM, was utilized for characterizing the morphology of EVs. Western blot analysis was performed to detect key biomarkers of EVs, such as TSG101, CD9, and CD63.

KEGG and GO enrichment analysis

To investigate the molecular mechanisms underlying the healing and non-healing of diabetic ulcers, we first identified differentially expressed genes (DEGs) obtained from the GEO database (accession number: GSE143735 ; available at https://www.ncbi.nlm.nih.gov/geo/). To identify DEGs, the criteria required a logarithmic fold change (logFC) exceeding 2 and a p-value below 0.05, comparing diabetic ulcer non-healing cases to healing cases. After obtaining these DEGs, to categorize the biological functions and cellular components related to these genes, Gene Ontology (GO) enrichment analysis was performed, and to identify the specific signaling pathways in which these DEGs are involved, analyses were conducted using the Kyoto Encyclopedia of Genes and Genomes (KEGG).

RNA isolation and quantitative real-time PCR analysis

Following the protocol provided by the manufacturer, total RNA was isolated from the cells utilizing the HiPure Total RNA Mini Kit. To isolate total RNA from extracellular vesicles, the HiPure Exosome RNA Kits was utilized. Complementary DNA (cDNA) synthesis was conducted utilizing the PrimeScript™ RT Reagent Kit that provided by Takara (Beijing, China). Next, quantitative real-time PCR (qRT-PCR) analysis was performed utilizing the LightCycler® 96 SW 1.1 system from Roche, based in Basel, Switzerland, in conjunction with the SYBR Green I system method (Takara in Beijing, China). The results were normalized against β-actin or GAPDH as reference genes. The sequences of the primers employed in this research are presented in Table 1.

Table 1.

Primers for real-time polymerase chain reaction

Genes	Left primer	Right primer
GLUT1	CATGGGCTTCTCGAAACTGG	GGTCCTTGTTGCCCATGATG
LDHA	GGCTACACATCCTGGGCTAT	TCTTCTTCAAACGGGCCTCT
HK1	GGATCCCTCAACCCTGGAAA	TTCTTTGGCATTGTGGAGGC
HK2	GACCAACTTCCGTGTGCTTT	TCCATGAAGTTAGCCAGGCA
PFK1	TGACCATTGGCACTGACTCT	ACCCTCAGCCACAATGATGA
GAPDH	GTCAAGGCTGAGAACGGGAA	AAATGAGCCCCAGCCTTCTC

Western blot

A cell scraper was used to harvest cells from the culture dish, with the entire process conducted on ice. Following a 10-min centrifugation at 4000 g and 4 °C, the resulting supernatant was removed. An appropriate amount of strong RIPA lysis buffer, along with phosphatase inhibitors, was incorporated based on the number of cells. The cells were disrupted on ice for a period of 30 min, after which they were subjected to centrifugation at 4 °C for a period of 10 min at 12,000 g to collect the supernatant. The collected supernatant was then diluted with a buffer at a ratio of 1:5, and then, the samples were subjected to a heating process at 100 °C for a duration of 10 min and subsequently stored at − 20 °C. After that, the SDS-PAGE was carried out, and subsequently, the proteins were transferred onto a polyvinylidene fluoride (PVDF) membrane. The membranes were subjected to incubation by using a blocking solution from Beyotime Biotechnology (China) for a duration of 15 min. The primary antibodies were left to incubate for an extended period at 4 °C overnight to ensure optimal binding. After incubation, the membrane was rinsed using Tris-Buffered Saline with Tween-20 (TBST) and subsequently was treated with horseradish peroxidase-conjugated anti-rabbit IgG (1:5000, ZSGB-BIO, China) and incubated for 1 h at ambient temperature. Bands were detected with an enhanced chemiluminescence reagent kit obtained from a supplier in Shanghai, China. Protein expression levels were represented by measuring grayscale values using ImageJ, with β-actin serving as a loading control for normalization purposes. All antibodies utilized in this study were obtained from the following sources: anti- CD9 (20597-1-AP, Proteintech), anti-CD63 (25682-1-AP, Proteintech), anti-TSG101 (28283-1-AP, Proteintech), anti-GLUT1 (21829-1-AP, Proteintech), anti-LDHA (19987-1-AP, Proteintech), anti-Collagen Type I (67288-1-Ig, Proteintech), anti-PPAR-γ (16643-1-AP, Proteintech), anti-p53 (10442-1-AP, Proteintech), anti-p16 (108831-AP, Proteintech), anti-SIRT1 (R381406, Zenbio).

Investigation of extracellular vesicles uptake by HUVECs and HDFs

Extracellular vesicles were marked using the Exosome Green Fluorescent Dye (PKH67) kit (Umibio, China) in accordance with the manufacturer’s instructions. In brief, EVs were treated with the dye working solution and incubated for a duration of 10 min to allow for effective labeling. After labeling, the sample was subjected to centrifugation at 4 °C for a duration of at 12000 g to obtain the purified labeled EVs. To investigate the internalization of these labeled EVs, HUVECs or HDFs were co-cultured with the labeled EVs for 12 h. After the incubation period, The cells were preserved by treating them with a 4% paraformaldehyde solution to preserve cellular structures and subsequently rinsed with PBS. Finally, the internalization of the labeled EVs was observed using a fluorescence microscope (manufactured by Olympus, Japan).

Cell proliferation and cytotoxicity assay (CCK-8) assay

To investigate cell growth, a total of 5 × 10³ cells were carefully placed into each well of a 96-well culture plate that was supplemented with 100 μL of nutrient-rich growth medium. After allowing the cells to adhere for 6 h, each well was treated with 20 μL of Cell Counting Kit-8 (CCK-8) solution, obtained from Biosharp (Beijing, China). The cells were subsequently placed in an incubator set at 37 °C for a designated period of 1 h. After incubation, the optical density of each well was assessed at a wavelength of 450 nm with the aid of a microplate reader.

Scratch wound assay

Cells were maintained in a six-well culture plate until they attained roughly 90% confluence. To simulate a wound for experimental analysis, a linear scratch in the cell monolayer was created using a 200 μL pipette tip. The dish was subsequently placed in an incubator maintained at 37 °C, and the wound closure process was carefully examined at predetermined time intervals (0, 24, and 48 h) to evaluate the rates of cell migration. A quantitative assessment of the wound area was conducted utilizing ImageJ software for precise measurement and analysis. The calculation is performed using the formula ((D0 − Dt)/D0 × 100%), where D0 represents the original wound size and Dt indicates the residual area at a given time.

Transwell assay

The cells that had been digested were re-dispersed in a medium devoid of serum and transferred to the upper compartment of the transwell insert, setting the stage for subsequent assays to assess cell migration capabilities. Meanwhile, the lower chamber was filled with a volume of 800 μL of DMEM medium to create an optimal environment for cell culture. Following incubation periods of 24 and 48 h, the culture plate was carefully taken out, and the cells that were attached to the surface of the filter membrane were thoroughly rinsed with PBS to eliminate any non-adherent cells. The cells were subsequently stabilized using a 4% paraformaldehyde solution for a duration of 10 min, followed by staining with crystal violet for a duration of 15 min to visualize the cellular structures. Finally, the cells in the upper chamber were observed and quantified using an optical microscope.

Capillary-like structure formation assay

For the tube formation assay, passage 3 (P3) generation HUVECs were used. Two hundred microliters of Matrigel (ABW® Matrigel) were introduced into each well of a pre-chilled 24-well culture plate and permitted to polymerize by incubating at 37 °C for 1 h. Subsequently, 3 × 10⁴ HUVECs from each group ((Normal Glucose (NG), High Glucose (HG), High Glucose with Extracellular Vesicles (HG+EVs ^NC), and High Glucose with sh-VIM-AS1 treated Extracellular Vesicles (HG+EVs ^{sh−VIM−AS1})) were seeded onto the polymerized Matrigel. After an incubation period of 8 h, the formation of tube structures on the Matrigel was examined using an optical microscope for detailed observation. Images of the tube structures were identified and evaluated using ImageJ software to quantify the number of tube structures formed in each group.

ROS measurement

The concentration of reactive oxygen species (ROS) in HDFs and HUVECs was measured using the oxidation-sensitive fluorescent probes 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA) and hydrogen ethylamine (provided by Beyotime, China). The cells were plated in 12-well plates and, on the subsequent day, were exposed to various types of extracellular vesicles or PBS as a reference control treatment. After an incubation period of 24 h, the cells were rinsed with PBS to eliminate unattached EVs. Subsequently, ROS staining was performed using a reactive oxygen species detection kit (CM-H2DCFDA) in accordance with the guidelines provided by the manufacturer. Fluorescent images of HDFs and HUVECs were obtained using a confocal laser scanning microscope (CLSM) from the Leica Application Suite X system, and the fluorescence signal was measured and quantified utilizing ImageJ software to facilitate a detailed analysis of the cellular responses.

Animal experimentation

All animal-related procedures adhered to the ethical guidelines of the Animal Experiment Center of The First Affiliated Hospital of Guangxi Medical University. The study utilized Sprague–Dawley (SD) rats as the experimental subjects, which were obtained from the Animal Center Guangxi Medical University and were kept in a standard laboratory animal environment, strictly following animal ethics guidelines. Four-week-old SD rats were acquired and allowed to acclimate to their new environment for 2 days. Following acclimatization, the rats were fed a diet high in sugar and fat (Boyaohang, China) over an 8-week period to induce metabolic alterations. To evaluate glucose tolerance, a 20% glucose solution (2 g/kg) was delivered through injection in intraperitoneal route. The blood specimens were drawn from the tail vein at intervals of 15, 30, 90, and 120 min following the injection to evaluate blood glucose concentrations. This was accomplished using a glucometer provided by Yuwell (China), allowing for precise monitoring of glycemic changes over time. If no abnormalities in glucose tolerance were observed, the rats (6 in total) continued to the high-sugar, high-fat diet until signs of impaired glucose tolerance were observed. To induce diabetes, a single intraperitoneal injection of 2% streptozotocin (STZ) at a dosage of 40 mg/kg (Solaibio, China) was carried out. Fasting blood glucose levels monitored over the subsequent weeks, and diabetes was confirmed if levels remained above 11.6 mmol/L for two consecutive weeks. Following successful induction of diabetes, a diabetic foot skin defect model (approximately 6 mm) was established, creating wounds approximately 6 mm in diameter on the rats’ feet. The anesthesia protocol for dorsal foot wound modeling in rats used 2% pentobarbital (Sigma, United States) at a dose of 0.2 mL/100 g body weight administered via intraperitoneal injection into the lower left abdominal quadrant at a 45° angle, avoiding vital organs. The solution was injected slowly over 10–15 s using a 1 mL syringe, with anesthesia efficacy confirmed by loss of righting reflex and vital signs monitored for 5 min post-injection. The 24 rats were randomly divided into four groups: NG group (n = 6), HG group (n = 6), HG+EVs ^shNC group (n = 6), and HG+EVs ^{sh−VIM−AS1} group (n = 6). The wound healing process was recorded using a digital camera at predetermined intervals: day 0, day 3, day 7, and day 14 post-injury. Quantitative analysis of wound healing was conducted utilizing ImageJ software to assess the area of the wounds at each specified time point. After the completion of the experiment, the rats were euthanized via intraperitoneal injection of 2% pentobarbital (Sigma, United States) at a dosage of 200 mg/kg. The work has been reported in line with the ARRIVE guidelines 2.0.

Tissue analysis and immunohistochemical analysis

Wound tissues of rats were harvested at 7 and 14 days post-treatment and preserved in a 4% paraformaldehyde solution. After fixation, the tissue samples underwent dehydration, were embedded in paraffin wax, and subsequently sectioned into slices with a thickness of 5 μm for further analysis. Hematoxylin and eosin staining was conducted to evaluate general histological features. While experiment on Masson’s trichrome staining was accomplished to evaluate collagen deposition within the wound tissue. Additionally, immunohistochemical staining was conducted employing antibodies targeted against VEGFR2 (26,415–1-AP, Proteintech) and Ki67 (ab15580, Abcam), which serve as markers for endothelial cells and proliferating cells, respectively.

Statistical analysis

In this study, the data are presented as the average±standard deviation (SD). GraphPad Prism 8.0 software (San Diego, CA, USA) was employed to accomplish statistical evaluations. The unpaired Student’s t-test was utilized for comparing two groups, while a one-way analysis of variance (ANOVA) was employed for comparisons involving three or more groups. A significance threshold of p < 0.05 was determined for all statistical assessments to identify statistically significant differences among the groups.

Result

Isolation and characterization of human umbilical cord mesenchymal stem cells and their derived extracellular vesicles

To investigate the functions of HuMSCs-EVs, we first isolated HuMSCs from human umbilical cord tissue. The extracted HuMSCs exhibited a distinctive radial expansion pattern, indicative of their proliferative capacity (Fig. 1a & Supplementary Figure S1a). To further verify the stem cell characteristics of these cells, we induced osteogenic and adipogenic differentiation, which supported their multipotent nature (Fig. 1b,c). To validate the expression of specific surface markers associated with mesenchymal stem cells, flow cytometry was employed. Our analysis revealed that the surface markers CD73, CD90, and CD105 were expressed at 100% on HuMSCs (Fig. 1d). Additionally, we assessed the expression levels of surface markers related to monocytes, B lymphocytes, hematopoietic stem cells, and other hematopoietic cells. The results indicated minimal expression of CD11b, CD19, CD34, CD45, and HLA-DR, with positive rates of 0.31%, 0.36%, 0.37%, 0.056%, and 0.38%, respectively (Fig. 1e). These findings indicate that HuMSCs were effectively isolated from umbilical cord tissue. Next, EVs were extracted and purified from the HuMSCs using an ultra-high-speed centrifugation method (Supplementary Figure S1b). Observations via transmission electron microscopy exposed that the EVs displayed a distinctive goblet-like structure, with a mean diameter of roughly 100 nm, confirming the expected size range for EVs (Fig. 1f, g). Moreover, Western blot analysis validated the elevated expression levels of key EVs markers TSG101, CD9, and CD63 within the isolated EVs (Fig. 1h). Collectively, these data validate the successful extraction of extracellular vesicles derived from human umbilical mesenchymal stem cells.

Fig. 1.

Extraction and identification of human umbilical cord mesenchymal stem cells and their extracellular vesicles. a Morphology of human umbilical cord MSCs observed under a light microscope. b Osteogenic differentiation assessed by alizarin red staining. c Adipogenic differentiation evaluated using oil red O staining. d Flow cytometry identification of stem cell markers CD73, CD90, and CD105. e Analysis of cells surface markers: CD11b, CD19, CD34, CD45, and HLA-DR by Flow cytometry. f Transmission electron microscopy (TEM) analysis of extracellular vesicles. g Nanoparticle size measurement of extracellular vesicles. h Identification of surface protein markers TSG101, CD9, and CD63 on extracellular vesicles (Full-length blots are presented in Supplementary Figure S5)

Extracellular vesicles enhance glycolysis in fibroblasts

To investigate the role of HuMSCs-EVs in promoting diabetic wound healing, we first analyzed factors associated with diabetic ulcers using the GEO dataset GSE143735. This analysis compared non-healing diabetic ulcers with healing diabetic ulcers and identified that genes with differential expression patterns were predominantly linked to biological pathways concerned with glucose metabolism (Fig. 2a), as well as cellular components (CC) and molecular functions (MF), as shown in Supplementary Figure S2a. These findings indicate a possible connection between HuMSCs-EVs and the processes involved in glucose metabolism. To further explore the effect of glucose concentration on human dermal fibroblasts (HDFs), we conducted a scratch test, which demonstrated that a concentration of 100 mmol/L significantly impacted fibroblast behavior (Supplementary Figure S2c), Consequently, we selected this concentration as the final level for our HG cell model. To examine the impact of HuMSCs-EVs on HG fibroblasts, we carried out a cell uptake experiment, which revealed that human fibroblasts effectively took up the extracellular vesicles (Fig. 2b). Next, to identify the ideal concentration of EVs for fibroblasts, we applied different doses of EVs (0, 25, 50, and 100 µg/mL) to the induced HG fibroblasts. Using a CCK-8 assay, we found that cell proliferation stabilized at an optimal concentration of 50 µg/mL of HuMSCs-EVs (Fig. 2c). Recent studies have indicated that promoting glycolysis can aid in the treatment of fractures in diabetic patients [18]. Subsequently, we assessed the levels of expression for crucial glycolytic genes, including HK2, HK1, GLUT1, PFK1, and LDHA, following treatment with different concentrations of EVs. The qRT-PCR analysis demonstrated that the expression levels of these glycolytic genes were upregulated in response to rising concentrations of EVs. Notably, most glycolytic genes reached a stable expression level at approximately 50 µg/mL (Fig. 2d–g). These findings suggest that HuMSCs-EVs can significantly enhance glycolysis.

Fig. 2.

Extracellular vesicles enhance glycolysis in fibroblasts and determination of the optimal concentration of extracellular vesicles. a Analysis of GO (Biological Processes, BP) for differential genes between diabetic ulcer healing and non-healing. b Endocytosis of extracellular vesicles by human dermal fibroblasts (HDFs). c CCK-8 assay results showing the optimal concentration of extracellular vesicles for HDFs. Expression levels of glycolytic genes: GLUT1d, LDHA e, HK1 f HK2 g and PFK1 h at optimal extracellular vesicle concentrations. *p < 0.05, **p < 0.01, ***p < 0.001

Extracellular vesicle-derived lncRNA VIM-AS1 enhances glucose metabolism in fibroblasts via the PPAR-γ pathway

To identify the components of HuMSCs-EVs that affect glucose metabolism in fibroblasts, we performed sequencing analysis of lncRNAs of HuMSCs-EVs. The sequencing results revealed that the top five ranked lncRNAs present in HuMSCs-derived EVs were VIM-AS1, BCYRN1, SENP3-EIF4A1, MIF-AS1, and LRRC75A-AS1 (Supplementary Figure S3a). Among these, lncRNA VIM-AS1 was found to be the most abundant, accounting for 28% of the total lncRNA content. This suggests that lncRNA VIM-AS1 may play a key role in facilitating glycolysis through HuMSCs-derived EVs. To validate this hypothesis, we first constructed HuMSCs with lncRNA VIM-AS1 knockdown and subsequently isolated their EVs (Supplementary Figure S3b-c & Fig. 1g). We then glucose consumption across four experimental groups: normal glucose (NG) group, high glucose (HG) group, high glucose with EVs from the negative control (HG+EVs ^NC) group, and high glucose with EVs containing lncRNA VIM-AS1 knockdown (HG+EVs ^{sh−VIM−AS1}) group. Our results showed that in the HG+EVs ^NC group, glucose uptake was markedly elevated compared to the HG group. Our results indicate that treatment with normal EVs resulted in a substantial increase in glucose consumption in contrast to the untreated HG group. In contrast, glucose consumption was significantly inhibited in the absence of lncRNA VIM-AS1 in the EVs. This suggests that lncRNA VIM-AS1 within extracellular vesicles is essential for improving the glucose metabolism of human fibroblasts (Fig. 3a). To delve deeper into the function of lncRNA VIM-AS1 in promoting glycolysis via HuMSCs-EVs, we conducted an analysis of glycolysis-related factors. As depicted in Fig. 3b–f, qRT-PCR analysis revealed that the upregulation of genes associated with glycolysis, including GLUT1, LDHA, HK1, HK2, and PFK1, was notably increased following treatment with normal EVs compared to the untreated HG group. However, the normal HuMSCs-EVs-induced upregulation of glycolysis-related genes was significantly disrupted in the absence of lncRNA VIM-AS1 in the HuMSCs-EVs. Western blot analysis supported our findings, demonstrating that the protein expression levels of GLUT1 and LDHA were markedly elevated in the group treated with normal EVs, as compared to the untreated HG group. Conversely, the levels of these proteins were significantly reduced in the HG+EVs ^{sh−VIM−AS1} group (p < 0.05) (Fig. 3g–i). Additionally, type I collagen secretion in human fibroblasts was increased in the HG+EVs ^NC group, whereas it was significantly inhibited in the HG+EVs ^{sh−VIM−AS1} group (Fig. 3g–j). Notably, KEGG enrichment analysis indicated a notable enrichment in the PPAR signaling pathway among differentially expressed genes based on the diabetic ulcers-related dataset GSE143735 (Supplementary Figure S3d). To further investigate this pathway, we examined the expression levels of PPAR-γ through Western blotting and found that stimulation with normal HuMSCs-EVs significantly increased PPAR-γ protein expression. However, this upregulation was markedly abolished when lncRNA VIM-AS1 was knocked down in the HuMSCs-EVs (Fig. 3g, k). Collectively, these findings reinforce the essential significance of lncRNA VIM-AS1 in controlling glucose metabolism and the PPAR pathway in fibroblasts.

Fig. 3.

Extracellular vesicular-derived lncRNA VIM-AS1 promotes glycolysis by affecting the PPAR-γ pathway. a Measurement of intracellular glucose consumption in fibroblasts. The qRT-PCR analysis detecting the expression of glycolytic genes: GLUT1 b, LDHA c, HK1 d, HK2 e, and PFK1 f. GLUT1 b, LDHA c, HK1 d, HK2 e, and PFK1 f. *p < 0.05, **p < 0.01, ***p < 0.001. g Western blot analysis showing the expression of key glycolytic proteins GLUT1 and LDHA, along with the proteins COLLAGEN 1 and PPAR-γ (Full-length blots are presented in Supplementary Figure S5). β-actin was used as a loading control. h–k The protein levels of GLUT1, LDHA, Collagen I, and PPAR-γ in Western blot analysis were quantified. *p < 0.05, **p < 0.01, ***p < 0.001

HuMSCs-EVs-derived LncRNA VIM-AS1 enhances fibroblast function

Given that fibroblast proliferation and migration are undeniably essential for diabetic wound healing [19, 20], we examined the effects of EVs-derived lncRNA VIM-AS1 on human dermal fibroblasts (HDFs) proliferation and migration. The CCK-8 assay demonstrated that under HG conditions, treatment with normal HuMSCs-EVs significantly increased the viability of fibroblasts, while the deletion of lncRNA VIM-AS1 resulted in a limited effect on cell viability (Fig. 4a). To assess migration capacity, we performed a Transwell assay, which showed that high-glucose-induced fibroblasts exhibited significantly reduced migration ability compared to normal fibroblasts at both 24 and 48 h of culture. However, treatment with EVs restored the migration capacity of high-glucose-induced fibroblasts to nearly normal levels. Conversely, the knockout of lncRNA VIM-AS1 impaired the ability of EVs to enhance fibroblast migration (Fig. 4b–d). Additionally, a scratch healing assay revealed that the healing rates in both the EV group and the NG group were markedly better compared to those of the HG group after 24 and 48 h. In contrast, the healing rate in the HG+EVs ^{sh−VIM−AS1} group was markedly reduced compared to that in the HG+EVs ^NC group (p < 0.05) (Fig. 4e–g). Recent research has shown that the accumulation of substantial reactive oxygen species (ROS) in diabetic wounds severely impairs wound healing [21, 22]. Therefore, we further investigated whether HuMSCs-EVs-derived lncRNA VIM-AS1 could mitigate ROS levels. As illustrated in Fig. 4h–i, the fluorescence signals indicated that the HG group had the highest accumulation of ROS. However, treatment with EVs ^NC effectively inhibited the generation of ROS, while the EVs ^{sh−VIM−AS1} treatment group demonstrated limited ability to scavenge ROS. These results indicate that HuMSCs-EVs-derived lncRNA VIM-AS1 can effectively inhibit the production of ROS. In conclusion, our findings support the critical role of HuMSCs-EVs-derived lncRNA VIM-AS1 in enhancing fibroblast function, which may contribute to improved diabetic wound healing.

Fig. 4.

Extracellular vesicular-derived lncRNA VIM-AS1 promotes fibroblast function. a Extracellular vesicular-derived lncRNA VIM-AS1 promotes the proliferation of fibroblast by CCK8 assay. b Transwell assay demonstrating the fibroblast migration characteristics across different treatment groups. The cell migration rates for Transwell assay were quantified at 24 h c and 48 h d; *p < 0.05, **p < 0.01, ***p < 0.001. e Scratch assay illustrating the migration characteristics of cells in various treatment groups. The cell migration rates for Scratch assay were quantified at 24 h f and 48 h g; *p < 0.05, **p < 0.01, ***p < 0.001. The ROS clearance rate of cells in different groups. h–i Measurement of the ROS clearance rate in cells from different groups. *p < 0.05, **p < 0.01, ***p < 0.001

HuMSCs-EVs-derived LncRNA VIM-AS1 enhances endothelial cell function

It is well established that vascular regeneration is crucial for effective wound healing [23, 24]. Thus, we also tested the functional effects of HuMSCs-EVs-derived lncRNA VIM-AS1 on vascular endothelial cells. First, we detected and observed the uptake of EVs by endothelial cells (Fig. 5a). Next, we evaluated the impacts of HuMSCs-EVs-derived lncRNA VIM-AS1 on cell proliferation and migration. The findings from the CCK-8 assay demonstrated that HuMSCs-EVs-derived lncRNA VIM-AS1 significantly increased the cellular activity of human umbilical vascular endothelial cells (HUVECs) (Fig. 5b). Additionally, as illustrated in Fig. 5c, d, the Transwell assay indicated that treatment with EVs ^NC significantly enhanced cell migration in the HG-induced vascular endothelial cells compared to the untreated HG group. However, this EVs ^NC -induced increase in mobility was markedly inhibited following lncRNA VIM-AS1 knockdown (p < 0.05). Furthermore, the scratch healing assay revealed that vascular endothelial cells exhibited a low rate of scratch healing under high glucose conditions. Interestingly, the HG+EVs ^NC group displayed improved rates of scratch healing compared to the HG group. In contrast, in comparison to the the HG+EVs ^NC group, the HG+EVs ^{sh−VIM−AS1} group exhibited a markedly decreased healing rate (Fig. 5e, f). Given that blood vessel formation within the wound is positively correlated with wound healing, we further investigated whether HuMSCs-EVs-derived lncRNA VIM-AS1 could promote angiogenesis. As shown in Fig. 5g, h, Induction with high glucose led to suboptimal cell conditions and a decreased formation of tube-like structures in comparison to the NG group. However, the addition of EVs ^NC restored the number of vessel-forming structures. In contrast, the HG+EVs ^{sh−VIM−AS1} group exhibited a significant reduction in the number of these tube-forming vessels compared to the HG+EVs ^NC group. Moreover, similar to our findings in fibroblasts, treatment with EVs ^NC effectively inhibited ROS production in vascular endothelial cells pre-treated with HG, while the HG+EVs ^{sh−VIM−AS1} group demonstrated diminished ability to scavenge ROS (Fig. 5i, j). Altogether, these results indicate that HuMSCs-EVs-derived lncRNA VIM-AS1 possesses a significant performance in enhancing endothelial cell growth, movement, and angiogenesis, all of which are critical for effective wound healing.

Fig. 5.

Extracellular vesicular-derived lncRNA VIM-AS1 optimizes vascular endothelial cell function. a Endocytosis of extracellular vesicles by vascular endothelial cells. b Extracellular vesicular-derived LncRNA VIM-AS1 promotes the proliferation of vascular endothelial cells by CCK8 assay. c Transwell assay demonstrating the vascular endothelial migration characteristics across different treatment groups. d The cell migration rates for Transwell assay were quantified;. e Scratch assay illustrating the migration characteristics of vascular endothelial cells in various treatment groups. f The cell migration rates for Scratch assay were quantified; *p < 0.05, **p < 0.01, ***p < 0.001. g, h In vitro tube-forming assay showing the ability of vascular endothelial cells to form tubes, with quantification of the number of tubes formed in each group; *p < 0.05, **p < 0.01, ***p < 0.001. i Measurement of the ROS clearance rate in cells from different groups. j ROS clearance rate was quantified; *p < 0.05, **p < 0.01, ***p < 0.001

The role of HuMSCs-EVs-derived lncRNA VIM-AS1 in modulating cellular senescence and SIRT1 expression

Numerous studies have shown that cells from diabetic patients are often in a state of senescence [7, 25, 26]. We investigated whether the EVs-derived lncRNA VIM-AS1 affects the senescence of fibroblasts and vascular endothelial cells. The findings from the β-galactosidase staining revealed that the proportion of β-galactosidase-positive cells in both fibroblasts and vascular endothelial cells following HG stimulation was significantly higher compared to cells without HG induction. Interestingly, treatment with EVs ^NC contributed to the reduction in the proportion of β-galactosidase-positive cells, while treatment with EVs ^{sh−VIM−AS1} exhibited only a weak mitigating effect (Fig. 6a–d). Additionally, we assessed the expression levels of senescence markers p53 and p16 in both fibroblasts and vascular endothelial cells using Western blotting. As shown in Fig. 6e–g and i–k, the expression of these senescence markers was elevated in the HG group in comparison with the NG group. However, these elevated levels could be reversed with treatment using EVs ^NC. Conversely, when lncRNA VIM-AS1 was knocked down, the EVs lost their capacity to reverse cellular aging. Recent research has confirmed that SIRT1 plays a vital role in glucose metabolism, contributing to improved glycemic regulation and increased insulin sensitivity [27, 28]. This led us to investigate whether HuMSCs-EVs-derived lncRNA VIM-AS1 affects SIRT1 levels. Notably, Western blotting results demonstrated that HuMSCs-EVs-derived lncRNA VIM-AS1 promoted the expression of the SIRT1 protein. However, this elevated expression was substantially reduced upon deletion of lncRNA VIM-AS1, as depicted in Fig. 6e, h and i, l. Collectively, our findings suggest that HuMSCs-EVs-derived lncRNA VIM-AS1 plays a critical role in reducing cellular senescence in fibroblasts and vascular endothelial cells by modulating key senescence markers and enhancing SIRT1 expression.

Fig. 6.

Extracellular vesicular-derived lncRNA VIM-AS1 alleviates cellular senescence. a Proportion of senescent human dermal fibroblast (HDF) in different treatment groups. b Quantification of senescent HDF in different treatment groups; *p < 0.05, **p < 0.01, ***p < 0.001. c Proportion of senescent vascular endothelial cells in different treatment groups. d Quantification of senescent vascular endothelial cells in different treatment groups; *p < 0.05, **p < 0.01, ***p < 0.001. e Western blot analysis of senescence-associated proteins p53, p16 and SIRT1 expression in HDF (Full-length blots are presented in Supplementary Figure S5). f–h The protein levels of p53, p16 and SIRT1 in HDF in Western blot analysis were quantified. *p < 0.05, **p < 0.01, ***p < 0.001. i Western blot analysis of senescence-associated proteins p53, p16 and SIRT1 expression in vascular endothelial cells (Full-length blots are presented in Supplementary Figure S5). j–l The protein levels of p53, p16 and SIRT1 in vascular endothelial cells in Western blot analysis were quantified. *p < 0.05, **p < 0.01, ***p < 0.001

HuMSCs-EVs-derived lncRNA VIM-AS1 promotes diabetic wound healing in vivo

To delve deeper into the function of HuMSCs-EVs-derived lncRNA VIM-AS1 in diabetic wound healing in vivo, we first established diabetic rat model and created controlled circular wounds with a diameter of approximately 6 mm. Subsequently, 20 μL of extracellular vesicles were injected subcutaneously into the wounds every two days, a total of 6 subcutaneous injections (20 μL each) were administered to each rat, and macroscopic images of wound healing were recorded on days 0, 3, 7, and 14 (Supplementary Figure S4a-b). As shown in Fig. 7a–c, the fastest wound healing was observed in the NG group, while the slowest healing occurred in the high glucose-treated group. Notably, EVs ^NC treatment significantly promoted diabetic wound healing, however, the healing was significantly delayed in the EVs ^{sh−VIM−AS1} treatment group compared to the HG+EVs ^NC group, indicating the critical importance of lncRNA VIM-AS1 in this process. Histological analysis using hematoxylin and eosin (HE) staining corroborated these findings. On day 7, treatment with EVs ^NC resulted in partial healing of the wounds, but hair follicle proliferation was absent at this stage. By day 14, the wounds in the HG+EVs ^NC group were nearly completely healed, exhibiting a substantial number of hair follicles compared to the NG group. In contrast, the HG+EVs ^{sh−VIM−AS1} group displayed limited dermal thickness and hair follicle regeneration when compared to the HG+EVs ^NC group (Fig. 7d). Masson’s trichrome staining further confirmed that HuMSCs-EVs -derived lncRNA VIM-AS1 promotes collagen deposition, thereby enhancing wound healing (Fig. 7e). To assess the cell proliferative status within the wound, we conducted Ki67 immunohistochemical staining, which validated that HuMSCs-EVs containing lncRNA VIM-AS1 enabled various types of cells to enter an active proliferative state, whereas HuMSCs-EVs lacking lncRNA VIM-AS1 lost this function (Fig. 7f). Additionally, we examined the expression of VEGFR2, a gene associated with angiogenesis. Immunohistochemical staining demonstrated that HuMSCs-EVs-derived lncRNA VIM-AS1 significantly promoted the expression of VEGFR2, suggesting its role in enhancing angiogenesis in diabetic wounds (Fig. 7g). In summary, these findings highlight the significant role of HuMSCs-EVs-derived lncRNA VIM-AS1 in enhancing the in vivo healing of diabetic wounds.

Fig. 7.

Extracellular vesicular-derived lncRNA VIM-AS1 accelerates diabetic wound healing in vivo. a Macroscopic images showing diabetic wound healing in the treatment groups. b Stacked plots illustrating the wound healing progress across different treatment groups. c Quantitative analysis of healing rates at 3, 7, and 14 days for the various treatment groups. d H&E staining evaluations at 7 and 14 days for each group. e Collagen deposition level was analyzed in the wounds for the four groups by Masson staining. f Immunohistochemical staining for the cell proliferation marker Ki67. g Immunohistochemical staining for the angiogenic factor receptor VEGFR2

Discussion

In this research, we successfully isolated human umbilical mesenchymal stem cells and characterized their derived EVs, evaluating their possible application in the healing of diabetic wounds. Our results demonstrate that HuMSCs-derived EVs significantly reprograms glycolysis in fibroblasts through the PPAR-γ pathway, reduces cellular senescence, and facilitates collagen deposition and angiogenesis by enhancing the function of fibroblasts and endothelial cells, which are crucial for effective wound healing, by which their contained lncRNA VIM-AS1 plays a crucial role in recovery process.

One of the most striking findings from our study is the ability of these HuMSCs-EVs to enhance glycolysis in fibroblasts. Glycolysis is a vital metabolic pathway that provides energy and building blocks necessary for cellular proliferation and migration, processes that are often impaired in diabetic wounds [29, 30]. By promoting glycolysis, HuMSCs-derived EVs facilitate the metabolic reprogramming of fibroblasts, thereby enhancing their functional capacity during wound healing. This highlights the crucial role of metabolic regulation in the process of tissue repair and suggests that targeting metabolic pathways could be a novel therapeutic approach to enhance the process of wound recovery in individuals with diabetes. Furthermore, our exploration of the lncRNA VIM-AS1 within HuMSCs-EVs revealed its critical role in modulating glucose metabolism through the PPAR-γ signaling pathway. Existing research has pointed to the activation of this pathway not only enhances glucose uptake but also promotes the expression of other key metabolic enzymes involved in glycolysis [31, 32], which consistent with our results that HuMSCs-EVs-derived lncRNA VIM-AS1 upregulated PPAR expression in fibroblasts. These insights align with emerging research highlighting the significance of non-coding RNAs in cellular metabolism and indicate that lncRNAs such as VIM-AS1 could serve as therapeutic targets to improve metabolic dysregulation in diabetic wounds.

Additionally, we demonstrated that HuMSCs-EVs-derived lncRNA VIM-AS1 substantially enhances fibroblast functions, including proliferation and migration, which are essential for wound closure. Given that impaired angiogenesis is a hallmark of chronic diabetic wounds [33], our findings extend to vascular endothelial cells, where lncRNA VIM-AS1 was demonstrated to contribute to the rapid and extensive spread or expansion of cells, as well as angiogenesis. These processes are integral to the formation of novel vascular structures and restoration of oxygen and nutrient supply to healing tissues [34]. A notable aspect of our findings is the ability of HuMSCs-EVs-derived lncRNA VIM-AS1 to effectively inhibit the production of ROS in both fibroblasts and endothelial cells. Elevated ROS levels are commonly associated with oxidative stress, which can lead to cellular damage and hinder the wound healing process [35]. By reducing ROS production, lncRNA VIM-AS1 helps maintain cellular homeostasis and protects cells from oxidative injury, thereby promoting a more favorable environment for regeneration. This antioxidant effect adds another dimension to the protective roles of lncRNA VIM-AS1 and highlights its potential as a therapeutic agent for mitigating oxidative stress in diabetic wounds.

Our study also highlights the potential of HuMSCs-EVs-derived lncRNA VIM-AS1 in reducing cellular senescence in both fibroblasts and endothelial cells. Cellular senescence contributes to the aging of tissues and is characterized by a permanent state of cell cycle arrest, along with the secretion of pro-inflammatory factors that impede healing [36, 37]. We observed that lncRNA VIM-AS1 can inhibit the expression of proteins related to senescence, and enhance SIRT1 expression, a protein associated with longevity and metabolic regulation, thereby mitigating the effects of cellular senescence [38]. This finding adds another layer of complexity to the role of HuMSCs-EVs-derived lncRNAs in the modulation of cell fate and emphasizes the need for further investigation into how targeting senescence can improve wound healing outcomes.

Conclusions

In conclusion, our findings provide compelling evidence that HuMSCs-derived EVs and their contained lncRNA VIM-AS1 play a pivotal role in enhancing wound healing in diabetic conditions. Mechanically, HuMSCs-EVs-derived lncRNA VIM-AS1 improves glycolysis through the activation of PPAR-γ, reduces cellular senescence by SIRT1 pathways, promotes angiogenesis, and inhibits oxidative stress in fibroblasts and vascular endothelial cells (Fig. 8). These findings suggested that HuMSCs-derived EVs represent a promising therapeutic modality for addressing the challenges posed by diabetes-related impairments in wound healing.

Fig. 8.

Extraction process of HuMSCs-derived extracellular vesicles and the regulatory mechanism of extracellular vesicle-derived lncRNA VIM-AS1 in promoting diabetic wound healing in rats

Author contributions

F.L., N.H., and Y.T. performed all the experimental assays and analyzed the data. B.Z. provided guidance for this study, offered financial support, and provided final approval of the manuscript. J.C. designed and guided this study and provided final approval of the manuscript. L.Z. supported the study and provided final approval of the manuscript. Y.F., B.Q., T.L., and J.Z. conducted the experiments and analyzed the data. All authors read and approved the final manuscript.

Funding

This research was funded by the Guangxi Natural Science Foundation (2023GXNSFAA026002), the Guangxi Scientific Research andTechnological Development Foundation (Grant No. GuikeAB23026049), the National Natural Science Foundation of China (Grant No. 82160418 and 82360426).

Availability of data and materials

The data used to support the findings of this study are available from the corresponding author upon request.

Declarations

Ethics approval and consent to participate

The Medical Ethics Committee of The First Affiliated Hospital of Guangxi Medical University (Nanning, China) authorized all operations involving human samples in this investigation (Medical Ethical approval number: 2021-KY- National Natural Science Foundation of China -228), and the title of the approval project is ‘The effect and mechanism of exosome-derived lncRNA BCYRN1 regulating autophagy activation through hsa-miR-6842-5p/MAPK13 axis on cartilage’ approved on February 26, 2021. The patient provided written informed consent for the use of samples. All the animal operations involved in this study were carried out as prescribed by the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health (Eighth Edition) and were approved by the Animal Ethic Committee of The First Affiliated Hospital of Guangxi Medical University. ((1) Title of the approved project: The effect and mechanism of exosome-derived lncRNA BCYRN1 regulating autophagy activation through hsa-miR-6842-5p/MAPK13 axis on cartilage; (2) Name of the institutional approval committee or unit: Ethic Committee of The First Affiliated Hospital of Guangxi Medical University; (3) Approval number: 2021-KY- National Natural Science Foundation of China -228; (4) Date of approval: February 26, 2021). The human umbilical vein endothelial cells (HUVECs) (catalog was # CP-H082) and human dermal fibroblasts (HDFs) (catalog was # CP-H103) were purchased from Wuhan Pricella Biotechnology Co., Ltd. (Wuhan, China) (http://www.whprocell.com/Products-36640225.html). The original source (Wuhan Pricella Biotechnology Co., Ltd.) has confirmed that there was initial ethical approval for collection of human cells, and that the donors had signed informed consent.

Consent for publication

All authors read and approved the publication of manuscript.

Competing interests

The authors declare no competing interests.