Tibial transverse transport promotes diabetic foot ulcer healing by regulating miR-182-5p Authors: Tongliang Xie, Haojie Li, Zuyan Sun, Rui Li, Wenliang Huang Published: 2026 ABSTRACT Diabetic foot ulcers (DFUs) represent a major complication of diabetes, characterized by impaired wound healing. Tibial transverse transport (TTT), a mechanical intervention technique, has shown clinical efficacy in promoting DFU repair; however, its molecular mechanisms remain largely unclear. This study aimed to explore the role of microRNAs (miRNAs) in the TTT-mediated healing of DFUs and to elucidate their regulatory effects on fibroblast function under hyperglycemic conditions. A diabetic rabbit model with full-thickness foot ulcers was established and treated with TTT. Granulation tissue was harvested for transcriptome sequencing to identify differentially expressed miRNAs. Key candidates (miR-182-5p, miR-150-5p, miR-205-5p) were validated by RT-qPCR. In vitro experiments were conducted on fibroblasts cultured in high-glucose conditions, including gain-of-function assays to assess the impact of miR-182-5p on cell proliferation, migration, and expression of HIF-1α and VEGF. TTT significantly altered the miRNA expression profile in DFU granulation tissue, with miR-182-5p being the most upregulated. Functional studies revealed that miR-182-5p overexpression restored proliferation and migration of high-glucose-exposed fibroblasts, while enhancing HIF-1α/VEGF pathway activity. miR-150-5p and miR-205-5p were also elevated by TTT and may contribute to wound repair via PI3K/Akt activation and VEGFA modulation, respectively. These effects align with findings from recent high-impact studies. TTT promotes DFU healing not only through biomechanical stimulation but also via reprogramming of miRNA-mediated molecular pathways. miR-182-5p, in particular, plays a pivotal role in regulating fibroblast activity and angiogenic signaling, representing a promising target for adjunctive molecular therapies. This study provides novel mechanistic insight into TTT and supports its integration with miRNA-based strategies for enhanced DFU treatment.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-51382-5. Introduction Diabetic foot is a condition characterized by foot infections, ulcers, and deep tissue damage caused by nerve and vascular lesions in the lower limbs of diabetic patients1. Diabetic foot ulcers (DFU) represent a severe complication of diabetes, with high incidence, disability, and mortality rates. Studies indicate that approximately 19%-34% of diabetic patients will experience DFU at some point in their lives2. The underlying pathogenesis of DFU is highly complex, involving immune dysregulation, elevated oxidative stress, neuropathy, microvascular complications, and chronic inflammation3. As the global burden of diabetes increases, the clinical management of DFU faces growing challenges. In recent years, tibial transverse transport (TTT) has gained increasing clinical use in the treatment of DFU, owing to its unique ability to improve lower limb ischemia and promote wound healing. Although clinical evidence shows that TTT effectively enhances local blood supply, reduces inflammation, and accelerates healing4, the molecular mechanisms underlying these effects remain poorly understood, and there is a lack of sufficient research data. This gap represents a critical barrier to the optimization and broader implementation of TTT. The core mechanisms of wound healing involve fibroblast activation, angiogenesis, and inflammation regulation. As negative regulators of gene expression, miRNAs—due to their stability and tissue specificity—have been shown to play a crucial role in these processes5. Previous studies by our team have demonstrated that TTT can promote DFU healing. However, the role of miRNAs in TTT-mediated DFU healing is still largely unexplored. In this study, we aim to utilize a rabbit DFU model combined with transcriptome sequencing, bioinformatics analysis, and functional cellular assays to: (1) systematically identify differentially expressed miRNAs during TTT; (2) analyze their potential target genes and signaling pathway networks; (3) validate the role of miRNA regulation in TTT-mediated DFU healing through qPCR and biological experiments. The goal is to provide molecular-level evidence for the mechanisms of TTT and establish a theoretical foundation for the precise treatment and biomarker development in DFU. Methods and materials Source of experimental animals and ethical statement The tissue samples used in this study were obtained from a previously approved rabbit model of type 2 diabetes mellitus–associated diabetic foot ulcer (DFU) and tibial transverse transport (TTT) intervention. All animal procedures were reviewed and approved by the Institutional Animal Care and Use Committee of Zunyi Medical University (Approval No. Lun Shen [2025] 2-740) and were conducted in accordance with relevant institutional guidelines. All efforts were made to minimize animal pain and stress. Animal care, housing, stocking density, and environmental enrichment Male New Zealand White rabbits aged 3 months (approximately 2.8–3.6 kg) were used in the preliminary animal experiments and were obtained from a qualified laboratory animal supplier. Animals were acclimatized for at least 7 days before any procedures. Rabbits were housed individually (1 rabbit/cage) under controlled conditions (22–24 °C; 50%–60% relative humidity) with a 12-h light/dark cycle. An automatic watering system was available, and cages were cleaned and waste was removed regularly. Environmental enrichment (e.g., hay/chew sticks and a shelter) was provided to reduce stress. Dedicated personnel monitored animals daily for general condition, food and water intake, body weight, wound status, and signs of pain or distress. Any abnormalities (e.g., infection, lameness, or weight loss) were assessed and managed by the facility veterinarian.No unexpected adverse events or unplanned animal losses were observed during the experiment. Experimental interventions, anesthesia, and analgesia Type 2 diabetes mellitus model: Rabbits were fed a homemade high-sugar/high-fat diet (approximately 250 g/day; e.g., containing 10% sucrose and 10% lard) for 4 weeks. Diabetes was then induced by intravenous alloxan (150 mg/kg) administered via the marginal ear vein. Fasting blood glucose was monitored, and a value ≥ 16.7 mmol/L was considered indicative of successful modeling. To prevent ketosis or hypoglycemia, insulin was administered subcutaneously when necessary and adjusted according to serial blood glucose measurements. Diabetic foot ulcer model: General anesthesia was induced with intramuscular xylazine hydrochloride (0.2 mL/kg). After adequate anesthesia, the surgical field was shaved and prepared aseptically. A skin incision was made approximately 3 cm to the right of the abdominal midline; the right femoral artery was isolated and ligated to simulate hindlimb ischemia. A full-thickness skin defect (approximately 3 cm × 2 cm) was then created on the dorsum of the right foot to establish an ulcer wound. Tibial transverse transport (TTT): A curved incision (~ 3–4 cm) was made along the medial tibia. After drilling, a rectangular corticotomy window (~ 20 mm × 10 mm) was created using a circular saw. The bone segment and the lateral pin tract were secured with 1.2-mm and 1.5-mm threaded Kirschner wires, followed by installation of a modified rabbit transverse transport fixator. The inter-frame distance was adjusted to ~ 5 mm before closure. Transport began 1 week postoperatively: outward transport for 1 week followed by inward transport for 1 week, at three turns per day (approximately 0.025 cm per turn; two turns on the final day). Perioperative analgesia and infection prophylaxis: Postoperative analgesia routinely included an NSAID (e.g., meloxicam 0.2 mg/kg, subcutaneously, every 24 h for 2–3 days). Additional opioid analgesia was provided as needed based on pain assessment (e.g., butorphanol 0.2 mg/kg, intramuscularly, every 12 h). Antibiotics (e.g., penicillin) were administered for infection prophylaxis when indicated (route and duration according to the institutional protocol used in the experiment). Outcome measures and sampling Primary outcomes in the preliminary animal experiments included ulcer area/healing rate (serial photography and quantification using ImageJ or equivalent software), hindlimb perfusion/microcirculation (e.g., CT angiography of hindlimb vessels or laser Doppler assessment), histopathology (hematoxylin and eosin staining), and angiogenesis-related markers (e.g., CD31 immunofluorescence and vascular endothelial growth factor A). Sampling: At predefined time points, marginal ear vein blood was collected under anesthesia into PAXgene Blood RNA tubes, and wound tissues were harvested. Wound-edge/granulation tissue was excised, gently rinsed with cold PBS to remove residual blood, blotted dry, snap-frozen in liquid nitrogen, and stored at − 80 °C. For omics and validation analyses in the present study, wound-edge/granulation tissues collected on postoperative day 10 from the TTT and diabetic foot ulcer (DFU) groups were used for re-analysis of miRNA/mRNA sequencing data and RT-qPCR validation. Euthanasia, humane endpoints, and end-of-study disposition Humane endpoints: Animals were euthanized before the planned endpoint if any of the following occurred: persistent pain or distress unrelieved by analgesia; refusal of food and water for 24–48 h or > 15% body weight loss; severe infection/necrosis or uncontrolled bleeding; severe lameness preventing access to food and water; or other irreversible deterioration as determined by the attending veterinarian. Euthanasia: Under deep anesthesia (per institutional standard operating procedures, using either isoflurane inhalation anesthesia or a xylazine/ketamine regimen), sodium pentobarbital was administered as an overdose (≥ 100 mg/kg) via the marginal ear vein until cessation of respiration and cardiac activity, with loss of corneal reflex as confirmation. When required by institutional procedures, a secondary confirmatory method (e.g., cervical dislocation or exsanguination) was performed only under deep anesthesia. At the end of the study, any surviving animals not used for tissue harvesting were euthanized at the predefined endpoint. Carcasses were disposed of by the animal facility as biomedical waste in accordance with institutional regulations. Source and processing of tissue samples The wound edge tissues from the TTT and DFU groups were collected from animals on postoperative day 10 in a prior experiment conducted by our research team. Tissue collection, anesthesia, and postoperative care were all carried out in strict accordance with the established standard procedures. Immediately after collection, the tissues were frozen in liquid nitrogen and stored at − 80 °C for future use. In this study, only the stored tissues were used for RNA extraction and qPCR validation, and no further animal procedures were performed.The present work represents a secondary molecular and cellular analysis based on preserved tissues; no new animals were generated specifically for this study, and all available samples meeting RNA quality requirements were included without additional inclusion/exclusion criteria. RNA extraction and transcriptome sequencing Total RNA was extracted from the wound edge tissues of experimental rabbits using the Trizol Total RNA Quick Extraction Kit. After removing genomic DNA, the RNA was subjected to quality control and quantification. Subsequently, PCR amplification, library construction, and miRNA-seq and mRNA-seq sequencing were performed using the Illumina HiSeq platform at Shanghai Kangcheng Biotechnology Co., Ltd. Transcriptome sequencing analysis (RNA-seq) Sample size, replication, and sequencing depth A total of six tissue samples were included in the miRNA-seq analysis, comprising two groups: the DFU control group and the TTTRT3 group (TTT intervention), with three biological replicates per group. Each sample was independently subjected to library preparation and sequenced once on an Illumina platform, and no technical replicates were performed. Sequencing yielded 5.54–7.69 × 10^6 raw reads per sample, with Q30 values of 90.85%–93.24%. Following adapter removal and length filtering, 81.16%–86.66% of reads were aligned to the reference genome, corresponding to 4.70–6.33 × 10^6 mapped reads per sample (mean: 5.57 × 10^6; median: 5.65 × 10^6). Statistical power analysis The statistical power of the study design to detect differentially expressed miRNAs at the observed sequencing depth and sample size was estimated using the negative binomial model implemented in RNASeqPower for a two-group comparison (n = 3 vs. n = 3). Using the median sequencing depth of 5.65 × 10^6 mapped reads per sample, the mean expression level of a representative miRNA in the control group was assumed to be approximately 3.5 CPM (≈ 20 reads per sample, satisfying the CPM ≥ 1 filtering criterion). The biological coefficient of variation was set to 0.4, the two-sided significance level to α = 0.05, and the minimum effect size of interest to a 2.5-fold change (|log2FC| = 1.32). Under these conditions, the estimated statistical power calculated with RNASeqPower was 0.84. In a previous project, miRNA-seq was performed on wound-edge tissues from the TTT and DFU groups (DFU control vs. TTTRT3; n = 3 per group), with matched mRNA-seq data generated concurrently. Building on these datasets, the present study conducted integrative bioinformatic analyses. miRNA read counts were normalized to counts per million (CPM), and differential expression was assessed using edgeR. Differentially expressed miRNAs were identified using thresholds of FDR (Benjamini–Hochberg correction) < 0.05 and |log2FC| ≥ 0.5. Significant miRNAs were subsequently subjected to target gene prediction (TargetScan), protein–protein interaction network construction (STRING and Cytoscape), and Gene Ontology and KEGG6 pathway enrichment analyses using clusterProfiler. RT-qPCR validation of key miRNAs (tissue level) To ensure that tissue-level RT-qPCR validation of key miRNAs was performed and reported in accordance with the MIQE (Minimum Information for Publication of Quantitative Real-Time PCR Experiments) guidelines, we standardized and documented RNA quality control, reverse transcription conditions, amplification parameters, control settings, and data processing. Primer sequences are provided in the Supplementary Materials for verification. Total RNA was extracted from wound-edge tissues of the TTT and DFU groups using TRIzol (as described above), followed by genomic DNA removal, quality assessment, and quantification. RNA purity was evaluated by the A260/A280 ratio, and RNA integrity as well as potential DNA contamination were assessed by agarose gel electrophoresis. Reverse transcription was performed using TRUEscript H Minus M-MuLV Reverse Transcriptase (Aidlab, PC1703). Briefly, RNA and primers were incubated at 70 °C for 5 min and immediately chilled on ice. A 20-µL reverse transcription reaction was then prepared by adding 5× Reaction Buffer, dNTP Mix, RNase inhibitor, and reverse transcriptase, followed by incubation at 42 °C for 40 min and enzyme inactivation at 85 °C for 5 min. cDNA was stored at − 20 °C until use. Quantitative real-time PCR was conducted on an Applied Biosystems 7500 Real-Time PCR System using 2× SYBR Green qPCR Mix (Aidlab, PC3302). Each 20-µL reaction contained 10 µL of 2× SYBR Green qPCR Mix, 1 µL each of forward and reverse primers (10 µM), 1 µL of cDNA, and 7 µL of nuclease-free water. The cycling conditions were 95 °C for 10 min, followed by 40 cycles of 95 °C for 20 s, 55 °C for 20 s, and 72 °C for 20 s. Melt-curve analysis was performed after amplification to confirm reaction specificity. U6 snRNA served as the internal reference, and relative expression levels were calculated using the 2^−ΔΔCt method to compare candidate miRNA expression between the TTT and DFU groups. Each run included a no-template control (NTC) and a no–reverse transcriptase control (− RT) to monitor contamination and residual genomic DNA. For each miRNA, each sample was analyzed in at least three technical replicates. If the standard deviation of Ct values among replicates exceeded 0.30 Ct, the assay was repeated. Wells were excluded based on predefined criteria if the SD remained > 0.30 Ct after repeat testing or if an obvious outlier was detected (e.g., a Ct value differing by > 0.50 Ct from the other replicates); alternatively, the assay for that sample was repeated. Rabbit fibroblast functional experiments Cell culture and experimental grouping Rabbit skin fibroblasts (purchased from RUIAN, RA-C101) were cultured in Dulbecco’s Modified Eagle Medium (DMEM, Biosharp, BL301A) containing 10% fetal bovine serum (FBS, Biosharp, BL1047A) and 1% penicillin-streptomycin (Biosharp, BL505A) under 37 °C and 5% CO₂ conditions. All experiments used cells in the logarithmic growth phase. Cells were seeded at 2–3 × 10⁵ cells/well in 6-well plates for RT-qPCR and Western Blot experiments, at 0.8 × 10⁴ cells/well in 96-well plates for CCK-8 proliferation assays, and resuspended at 3 × 10⁵ cells/mL for Transwell migration assays. The high-glucose injury model was established by treating the cells with 200 mmol/L glucose in DMEM for 48 h. The cells were randomly assigned to the following six groups: Control group: normal glucose culture; Model group: high-glucose treatment; Model + miR-182-5p mimics NC; Model + miR-182-5p mimics; Model + miR-182-5p inhibitors NC; Model + miR-182-5p inhibitors. Cells were transfected with Lipofectamine™ 3000 Transfection Reagent (Thermo Fisher Scientific, L3000008) following the manufacturer’s protocol; miR-182-5p mimics, inhibitors, and their respective negative controls were used at a final concentration of 50 nM, and after 48 h of incubation, various assays were performed. For in vitro experiments, cells were randomly allocated to treatment groups prior to downstream assays. Real-time quantitative PCR (RT-qPCR) To ensure that cell-based RT-qPCR experiments were performed and reported in accordance with the MIQE (Minimum Information for Publication of Quantitative Real-Time PCR Experiments) guidelines, we standardized the workflow for RNA extraction and quality control, reverse transcription, qPCR reaction setup and cycling conditions, specificity verification, controls, and data processing. Primer sequences are provided in the Supplementary Materials for verification. Total RNA from each cell group was extracted using a universal RNA extraction kit (CTCC, M004). For adherent cells, culture medium was removed and ~ 1 mL lysis buffer was added per ~ 1 × 10^6 cells. Cells were thoroughly pipetted to ensure complete lysis and homogenization, followed by incubation at room temperature for 5 min. Chloroform was added (0.2 mL per 1 mL lysis buffer), vortexed vigorously for 15 s, incubated for 3 min at room temperature, and centrifuged at 12,000 rpm for 15 min at 4 °C. The upper aqueous phase was transferred, mixed with ethanol, and purified on a spin column. After sequential washing steps, RNA was eluted with RNase-free water. RNA concentration and purity were measured using a microvolume spectrophotometer (BIO-RAD Smartspecplus), and purity was evaluated by the A260/A280 ratio. Reverse transcription was performed using ~ 1 µg total RNA as input. RNA and primers were incubated at 70 °C for 5 min and immediately chilled on ice. A 20-µL reverse transcription reaction was prepared by adding 5× Reaction Buffer, dNTP Mix, RNase inhibitor (Aidlab, RN3501), and TRUEscript H Minus M-MuLV Reverse Transcriptase (Aidlab, PC1703), followed by incubation at 42 °C for 40 min and enzyme inactivation at 85 °C for 5 min. cDNA was stored at − 20 °C until use. qPCR was conducted on an Applied Biosystems 7500 Real-Time PCR System using 2× SYBR Green qPCR Mix (Aidlab, PC3302). Each 20-µL reaction contained 10 µL of 2× SYBR Green qPCR Mix, 1 µL each of forward and reverse primers (10 µM), 1 µL of cDNA, and 7 µL of nuclease-free water. The cycling conditions were 95 °C for 10 min, followed by 40 cycles of 95 °C for 20 s, 55 °C for 20 s, and 72 °C for 20 s. Melt-curve analysis was performed after amplification to confirm reaction specificity. Primers were designed using Primer Premier 5.0 and synthesized by Sangon Biotech (Shanghai, China) (see the primer table). U6 snRNA served as the internal reference. Ct values were calculated automatically by the instrument, and the threshold and baseline were adjusted with reference to negative controls to determine Ct values for each sample. Relative expression was calculated using the 2^−ΔΔCt method, where ΔCt = Ct(target) − Ct(U6) and ΔΔCt = ΔCt(experimental group) − mean ΔCt(control group). Each run included a no-template control (NTC) and a no–reverse transcriptase control (− RT) to monitor contamination and residual genomic DNA. For each gene, each sample was tested in at least three technical replicates. If the standard deviation of Ct values among replicates exceeded 0.30 Ct, the assay was repeated. Wells were excluded based on predefined criteria if the SD remained > 0.30 Ct after repeat testing or if an obvious outlier was detected (e.g., a Ct value differing by > 0.50 Ct from the other replicates); alternatively, the assay for that sample was repeated. Western blot analysis Cells were lysed using RIPA lysis buffer (Beyotime, P0013B) containing PMSF and phosphatase inhibitors. After centrifugation at 12,000 rpm at 4 °C for 10 min, the supernatants were collected. Protein concentration was measured using the BCA method. After adding SDS sample buffer, proteins were denatured at 100 °C for 5 min. Equal amounts of protein (30 µg/lanes) were separated on a 10–12% SDS-PAGE gel and transferred to a PVDF membrane (Millipore, IPVH00010). The membrane was blocked with 5% BSA at room temperature for 2 h, and the following primary antibodies were incubated overnight at 4 °C: HIF-1α (LMAI Bio, LM-0737R), VEGF (Bioss, bs-1313R), FGF2 (Bioss, bs-0217R), GAPDH (Abcam, ab125247). After washing with TBST, HRP-conjugated secondary antibody (ZSGB-BIO, ZB2301) was applied, and the membrane was incubated at 37 °C for 1 h. Protein bands were visualized using ECL chemiluminescent reagent and imaged with a ChemiScope 5300 Pro imaging system. CCK-8 cell proliferation assay Cells were seeded at 0.8 × 10⁴ cells/well in a 96-well plate. After overnight attachment, high-glucose treatment was applied at 200 mmol/L glucose in DMEM for 48 h, and miRNA transfection was performed at a final concentration of 50 nM as described in “Cell culture and experimental grouping” section. Subsequently, 20 µL of CCK-8 reagent (CTCC, M007) was added to each well, and the cells were incubated at 37 °C for 2 h. Absorbance was measured at 450 nm using a microplate reader. Each group had 5 replicates. Transwell migration assay Cell migration ability was assessed using Transwell chambers (FALCON, 353097). Following the respective treatment protocols (high-glucose pre-treatment at 200 mmol/L for 48 h and/or miRNA transfection at 50 nM as described in “Cell culture and experimental grouping” section, cells were digested with 0.25% trypsin, washed twice with PBS to remove residual serum, and resuspended in serum-free DMEM at 3 × 10⁵ cells/mL (200 µL per upper chamber insert, i.e. 6 × 10⁴ cells per insert). The lower chamber contained 800 µL of complete medium with 10% FBS, and 200 µL of cell suspension was added to the upper chamber. After incubation at 37 °C and 5% CO₂ for 48 h, non-migrated cells in the upper chamber were removed. Migrated cells in the lower chamber were fixed with 70% ethanol and stained with 0.5% crystal violet. Cells were photographed under an inverted microscope (Olympus IX71), and 5 random fields were counted. Statistical analysis All experiments were repeated at least three times. Data are presented as mean ± standard deviation (mean ± SD). Statistical analysis was performed using one-way ANOVA followed by Tukey’s post-hoc test with GraphPad Prism 8.0 software. P < 0.05 was considered statistically significant.Normality and homogeneity of variance were assessed prior to parametric testing; all P values were two-sided. Results Screening and identification of differentially expressed miRNAs after TTT intervention MiRNA-seq analysis was performed on the wound edge tissues from the TTT and DFU groups, with p < 0.05 and |log₂FC| ≥ 1.2 as the thresholds. Three significantly differentially expressed miRNAs were identified: ocu-miR-150-5p, ocu-miR-182-5p, and ocu-miR-205-5p, all of which were significantly upregulated in the TTT group. The volcano plot of differentially expressed miRNAs provides a clear visualization of the expression distribution across all miRNAs (Fig. 1A), while the heatmap of the top 30 different miRNAs further highlights the expression pattern differences between the groups (Fig. 1B). Fig. 1Identification of differentially expressed miRNAs; (A) shows the volcano plot of differentially expressed miRNAs; (B) presents the heatmap of the top 30 differentially expressed genes. Identification of differentially expressed miRNAs; (A) shows the volcano plot of differentially expressed miRNAs; (B) presents the heatmap of the top 30 differentially expressed genes. Target gene prediction and protein interaction network analysis of differential miRNAs Target gene prediction for the three differential miRNAs was performed using TargetScan, and a miRNA-target gene protein interaction network was constructed using Cytoscape (Fig. 2). Among them, the gene network regulated by miR-182-5p was the most complex, containing 171 nodes and 279 edges, with CTNNB1 identified as the hub gene. The hub genes for miR-150-5p and miR-205-5p were PRKACA and YAP1, respectively. In the network diagram, the more red the node, the higher the connectivity of the protein within the network. Fig. 2Gene-protein interaction network prediction; (A) displays miR-150-5p; (B) displays miR-182-5p; (C) displays miR-205-5p Gene-protein interaction network prediction; (A) displays miR-150-5p; (B) displays miR-182-5p; (C) displays miR-205-5p GO and KEGG pathway enrichment analysis The top 10 genes with the highest enrichment from the target genes predicted for the three differential miRNAs were selected for GO functional and KEGG pathway enrichment analysis. As shown in Fig. 3, at the biological process level, the target genes were significantly enriched in processes such as the Wnt/β-catenin signaling pathway, cell cycle regulation, and TOR signaling regulation. At the cellular component level, the enrichment was primarily in synaptic structures, membrane rafts, and focal adhesions. At the molecular function level, the target genes were mainly enriched in transcription factor binding, protein kinase activity, and β-catenin binding. KEGG analysis revealed significant enrichment of the target genes in pathways associated with various cancers (such as prostate cancer and colorectal cancer), Wnt signaling, estrogen signaling, and platelet activation (Fig. 3). All enrichment results were corrected using the Benjamini–Hochberg method, with an FDR < 0.05 considered significant. Fig. 3Bubble plot of GO (BP/CC/MF) and KEGG pathway enrichment. The x-axis represents the GeneRatio (= number of genes enriched in a specific entry, k, divided by the total number of input genes, K). The size of the dots reflects the number of enriched genes (Count), and the color indicates the p.adjust (q-value after Benjamini-Hochberg correction). The darker the color, the stronger the significance. Each panel displays the top N entries sorted by p.adjust (default N = 20, unless specified otherwise). The significance threshold is set at FDR < 0.05. Bubble plot of GO (BP/CC/MF) and KEGG pathway enrichment. The x-axis represents the GeneRatio (= number of genes enriched in a specific entry, k, divided by the total number of input genes, K). The size of the dots reflects the number of enriched genes (Count), and the color indicates the p.adjust (q-value after Benjamini-Hochberg correction). The darker the color, the stronger the significance. Each panel displays the top N entries sorted by p.adjust (default N = 20, unless specified otherwise). The significance threshold is set at FDR < 0.05. qPCR validation of differential miRNAs in tissue samples RT-qPCR was performed to validate the sequencing results at the tissue level. The relative expression of miRNAs was calculated using ΔCt (= Ct_miRNA - Ct_U6), where a lower ΔCt value indicates higher expression. The results (Fig. 4) show that, compared to the DFU group, the ΔCt values of the three miRNAs in the TTT group were significantly reduced (p < 0.05), indicating that their expression was significantly upregulated after TTT intervention, consistent with the sequencing results. The amplification curves for each miRNA and the internal control U6 (Fig. 5A-D) showed a typical S-shape, with similar slopes in the exponential phase. The melt curves displayed a single sharp peak, with no primer dimer interference, indicating that the qPCR results were reliable and met the conditions required for ΔΔCt analysis. Fig. 4miRNA ΔCt (ΔCt = Ct_miRNA − Ct_U6) in DFU and TTTRT3. The bars represent the mean (n = 3), with error bars showing standard deviation (SD). Welch t-test, p < 0.05. For bars with negative values (TTTRT3), error bars are plotted from the bottom of the bar. The miRNAs are: ① ocu-miR-205-5p; ② ocu-miR-150-5p; ③ ocu-miR-182-5p. miRNA ΔCt (ΔCt = Ct_miRNA − Ct_U6) in DFU and TTTRT3. The bars represent the mean (n = 3), with error bars showing standard deviation (SD). Welch t-test, p < 0.05. For bars with negative values (TTTRT3), error bars are plotted from the bottom of the bar. The miRNAs are: ① ocu-miR-205-5p; ② ocu-miR-150-5p; ③ ocu-miR-182-5p. Fig. 5Amplification Plot (A–D): Amplification curves for ocu-miR-150-5p, U6, and ocu-miR-205-5p/ocu-miR-182-5p (threshold ΔRn = 0.1). The curves show a typical S-shape, with consistent replication and similar slopes. The Ct order is as follows: miR-150-5p < U6 < miR-205-5p ≈ miR-182-5p. Melt Curve Plot (A–D): Melt curves for U6, ocu-miR-205-5p, ocu-miR-150-5p (and miR-182-5p), all exhibiting a single peak (around 82–85 °C), with no obvious secondary peaks or dimer signals, confirming that the amplification products are specific and reliable. Amplification Plot (A–D): Amplification curves for ocu-miR-150-5p, U6, and ocu-miR-205-5p/ocu-miR-182-5p (threshold ΔRn = 0.1). The curves show a typical S-shape, with consistent replication and similar slopes. The Ct order is as follows: miR-150-5p < U6 < miR-205-5p ≈ miR-182-5p. Melt Curve Plot (A–D): Melt curves for U6, ocu-miR-205-5p, ocu-miR-150-5p (and miR-182-5p), all exhibiting a single peak (around 82–85 °C), with no obvious secondary peaks or dimer signals, confirming that the amplification products are specific and reliable. Functional validation of miR-182-5p in the high glucose injury model Effect of miR-182-5p expression intervention In rabbit skin fibroblasts treated with high glucose, miR-182-5p expression was assessed by RT-qPCR (Fig. 6). Compared to the control group, the expression of miR-182-5p was significantly downregulated in the high glucose model group. Transfection with miR-182-5p mimics significantly restored its expression, even surpassing normal levels, while transfection with the inhibitor further suppressed its expression. No significant difference was found between the negative control group and the model group, indicating that the intervention effect was specific. Fig. 6RT-qPCR detection of miR-182-5p relative expression. High glucose significantly downregulated its expression. miR-182-5p mimics restored its expression, while inhibitors further decreased its expression. Data are presented as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.005, ****P < 0.0001, compared to the control group. RT-qPCR detection of miR-182-5p relative expression. High glucose significantly downregulated its expression. miR-182-5p mimics restored its expression, while inhibitors further decreased its expression. Data are presented as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.005, ****P < 0.0001, compared to the control group. Changes in the expression levels of key proteins Western Blot results (Fig. 7) showed that high glucose treatment significantly reduced the protein expression of HIF-1α, VEGF, and FGF2 in fibroblasts. Transfection with miR-182-5p mimics significantly reversed this effect, upregulating the expression of these three proteins. In contrast, miR-182-5p inhibitor further aggravated the suppression of protein expression induced by high glucose. Fig. 7Western Blot bands for HIF-1α, VEGF, and FGF2 and protein quantification by grayscale analysis (normalized to GAPDH). Data are presented as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.005, ****P < 0.0001, compared to the control group. Western Blot bands for HIF-1α, VEGF, and FGF2 and protein quantification by grayscale analysis (normalized to GAPDH). Data are presented as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.005, ****P < 0.0001, compared to the control group. Cell proliferation and migration ability detection CCK-8 assay results (Fig. 8) show that high glucose significantly inhibited the proliferation ability of fibroblasts. Overexpression of miR-182-5p significantly alleviated the inhibitory effect of high glucose on proliferation, while inhibition of miR-182-5p further decreased cell proliferation. Transwell migration assay results (Fig. 9) show that cell migration ability was impaired in high glucose conditions. Transfection with miR-182-5p mimics significantly enhanced cell migration, while miR-182-5p inhibitors further suppressed migration. No significant difference was found between the negative control group and the model group in the above functional experiments, confirming that miR-182-5p plays a key regulatory role in high glucose-induced cellular functional impairment. Fig. 8CCK-8 assay showing differences in cell proliferation among the groups. Data are expressed as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. CCK-8 assay showing differences in cell proliferation among the groups. Data are expressed as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Fig. 9Transwell migration assay showing changes in cell migration ability. miR-182-5p mimics promote, while inhibitors inhibit cell proliferation and migration. Data are presented as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Transwell migration assay showing changes in cell migration ability. miR-182-5p mimics promote, while inhibitors inhibit cell proliferation and migration. Data are presented as mean ± SD; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Discussion This study systematically reveals the mechanisms by which tibial transverse transport (TTT) promotes diabetic foot ulcer (DFU) healing through a combination of omics sequencing and cellular validation. Through transcriptome sequencing and qPCR validation, we identified significant upregulation of three key miRNAs (miR-150-5p, miR-182-5p, and miR-205-5p) in the wound tissue of the TTT treatment group. The PPI networks constructed for miR-182-5p, miR-150-5p, and miR-205-5p each exhibited a high degree of interconnectivity, with EP300, CTNNB1, and MAPK3 identified as their respective core hub genes. These findings suggest that TTT may promote DFU repair through coordinated regulation of multiple pathways, including transcriptional programs, signal transduction cascades, and cellular functional states. In the miR-182-5p–associated PPI network, EP300 was identified as the central hub gene. EP300 is a canonical transcriptional coactivator and histone acetyltransferase that stabilizes HIF-1α through acetylation, thereby enhancing HIF-1 transcriptional activity and downstream angiogenic gene expression such as VEGF7. In the present study, miR-182-5p overexpression in high-glucose fibroblasts restored HIF-1α, VEGF, and FGF2 protein levels (Fig. 7), a pattern consistent with activation of EP300-mediated HIF-1α transcriptional activity. These findings suggest that TTT-induced upregulation of miR-182-5p may promote DFU healing by engaging the EP300/HIF-1α/VEGF axis, thereby creating a pro-angiogenic microenvironment conducive to tissue reconstruction. However, the precise molecular mechanism through which miR-182-5p regulates EP300 — including whether this involves direct targeting of the EP300 3’UTR, modulation of upstream EP300 regulators, or changes in EP300 post-translational activity — was not experimentally addressed in the present study and represents an important direction for future mechanistic investigation. The miR-150-5p–associated PPI network was centered on CTNNB1, a key effector of Wnt/β-catenin signaling. CTNNB1 occupies a central position in controlling cell proliferation and migration, as well as stem/progenitor cell fate decisions. Wnt/β-catenin signaling has been widely implicated in skin regeneration, angiogenesis, and fibroblast activation. Under diabetic conditions, this pathway is frequently attenuated by sustained hyperglycemia and chronic inflammation8 Our findings suggest that TTT-induced upregulation of miR-150-5p may modulate CTNNB1-centered interaction networks and thereby reactivate Wnt/β-catenin signaling, potentially enhancing local cellular proliferative capacity and tissue regenerative potential in DFU. In parallel, MAPK3 (ERK1) was identified as the key hub gene in the miR-205-5p–associated PPI network. MAPK/ERK signaling is a core pathway in cellular stress responses, inflammatory regulation, and tissue repair, with established roles in keratinocyte migration, fibroblast proliferation, and endothelial homeostasis9. Given the hub position of MAPK3 within this network, miR-205-5p may facilitate DFU wound repair by shaping MAPK/ERK signaling cascades and, consequently, cellular adaptation to mechanical cues and the ischemic microenvironment induced by TTT. DFU often results in poor healing due to factors such as local ischemia, chronic inflammation, and neuropathy. TTT, as an innovative orthopedic technique, has demonstrated clinical efficacy in improving DFU prognosis. This is based on the “tensile stress theory,” which states that slow traction of cortical bone stimulates new blood vessel growth and improves local blood flow. Clinical reports have shown that TTT treatment for severe DFU significantly increases the healing rate and reduces the likelihood of amputation10 compared the effects of TTT and platelet-rich plasma (PRP) therapy on severe DFU: after 18 months of follow-up, the wound healing rate in the TTT group reached 96.7%, significantly higher than the 80% in the PRP group; the average healing time was reduced from approximately 6 months to less than 3 months, and the amputation and recurrence rates were also significantly reduced11. These clinical data strongly support the superiority of TTT in refractory DFU. Based on these findings, this study further explores the biological reasons behind the effectiveness of TTT. The upstream mechanisms by which TTT-generated mechanical forces are transduced into miRNA expression changes in the wound microenvironment remain to be fully elucidated. Emerging evidence in mechanobiology supports a multi-step transduction cascade: cortical bone displacement generates cyclic tensile and compressive stresses that activate periosteal mechanosensors, including stretch-activated ion channels (PIEZO1/2) and integrin–focal adhesion complexes12. These signals converge on mechanosensitive transcriptional regulators, notably YAP/TAZ and NF-κB, which modulate primary miRNA (pri-miRNA) transcription and the processing machinery including Drosha and DGCR813. Future studies employing single-cell sequencing or mechano-sensitive reporter systems will help dissect the precise signals driving the coordinated upregulation of miR-182-5p, miR-150-5p, and miR-205-5p observed in TTT-treated wounds. Our molecular results show that TTT is not only a physical therapy but also triggers a broad biological response. The upregulation of miR-182-5p, miR-150-5p, and miR-205-5p by TTT are considered important regulators in the wound healing process. For example, miR-182-5p promotes angiogenesis and epithelial cell migration during normal wound healing, but is often downregulated in the diabetic environment, leading to delayed healing. Li et al. (2023) demonstrated that exosomes rich in miR-182-5p, when applied to diabetic rat wounds, significantly accelerated healing by inhibiting PPARγ and enhancing cell proliferation and matrix deposition14. TTT upregulates miR-182-5p, which may compensate for the deficiency of this miRNA in the diabetic state, restoring normal healing function. Similarly, miR-150-5p promotes healing in normal wounds, but its expression is abnormal in chronic inflammatory environments15. This study found that TTT significantly increased miR-150-5p levels, which may help reverse the adverse effects of long-term inflammation in DFU. Previous studies have shown that miR-150-5p accelerates skin wound closure by activating the AKT pathway and modulating inflammatory cytokines16. Therefore, TTT’s effect is not limited to mechanical stimulation but also “empowers” the body’s own repair potential by regulating key miRNAs and reprogramming local cell behavior to promote healing. It is important to note that the molecular changes triggered by TTT align with the known key pathways in diabetic wound healing. Diabetic wounds often suffer from impaired HIF-1α/VEGF axis due to chronic hypoxia, leading to insufficient angiogenesis17. TTT, through bone trauma stimulation, induces the release of hematopoietic and angiogenic factors from the bone marrow, increasing circulating levels of HIF-1α and VEGF18. Qin et al. (2024) reported significant increases in HIF-1α, VEGF, SDF-1, and other factors in the serum of DFU patients after TTT, indicating that TTT promotes both systemic and local angiogenic responses19. Our study found that the upregulation of miR-205-5p may prevent fibrosis by moderately inhibiting the YAP pathway, which aligns with the observation of orderly collagen deposition and smaller scars in TTT-treated wounds. On the other hand, literature reports that in chronic non-healing DFU, certain miRNAs (such as miR-205-5p) are abnormally elevated, inhibiting angiogenic factors and contributing to impaired healing20. Our results suggest that TTT may shift the function of these “harmful” miRNAs to promote healing. For example, in typical DFU, miR-205-5p primarily inhibits VEGF, leading to reduced angiogenesis, while after TTT improves blood supply, miR-205-5p is more involved in inhibiting excessive YAP signaling, resulting in a net positive effect on vascular remodeling. In summary, TTT corrects the pathological environment of diabetic wounds at both the macro (circulation, blood flow) and micro (gene/miRNA expression) levels: macroscopically, it alleviates ischemia and hypoxia, and microscopically, it reactivates the molecular networks required for healing. This dual mechanism of action is consistent with recent review perspectives, which emphasize that TTT not only improves blood circulation and promotes angiogenesis but also potentially affects inflammatory mediators and cellular signaling pathways, resulting in improved healing of difficult-to-treat wounds21. The clinical significance of TTT in the treatment of diabetic foot is further highlighted by the findings of this study. On one hand, we provide molecular evidence supporting the reliability of TTT’s therapeutic effects. Clinicians often focus on understanding the mechanism of treatment to assess its feasibility and sustainability. This study proves that TTT can regulate specific miRNAs, thereby influencing key healing pathways, thus providing an answer to why TTT is effective. This means that TTT is not just a temporary solution, but it may have a profound impact on the pathophysiology of DFU, including improving local microcirculation, reducing chronic inflammation, and promoting tissue regeneration. For patients, the benefits of TTT are tangible: faster wound closure, lower infection and amputation risks22. Therefore, applying TTT in suitable patients is expected to reduce the disability rate and medical burden of diabetic foot, which is especially important for patients with ineffective conventional treatments who face amputation risk. On the other hand, revealing the role of miRNAs in TTT’s efficacy opens up new possibilities for optimizing diabetic foot treatment. miRNAs themselves have the potential to become therapeutic targets or diagnostic markers. Based on the miRNA changes we observed, future developments could include minimally invasive diagnostic methods, such as detecting miR-182-5p levels in wound exudates or patient blood to assess wound healing tendency or TTT treatment efficacy. This “liquid biopsy” style of monitoring would enable more precise clinical decision-making. Additionally, miRNA-mediated mechanisms provide a blueprint for biological therapy: if drugs can be developed to upregulate miR-150-5p/miR-182-5p or inhibit miRNAs that delay healing, they could mimic the molecular effects of TTT and be used for patients who cannot undergo surgery. For example, Li et al. (2023) envisioned using exosomes enriched with specific miRNAs to promote wound healing23. In fact, similar attempts are already underway: the team of Peng Li has used engineered exosomes to deliver miR-182-5p to diabetic ulcers, achieving accelerated healing14. This suggests that future miRNA-targeted therapies may serve as a powerful complement or alternative to TTT. For example, administering specific miRNA mimics to patients before or after surgery could further promote angiogenesis and tissue regeneration; or for those who are not suitable for TTT, direct local miRNA administration combined with standard wound care could improve healing rates. Of course, these concepts still require extensive research and validation, but this study provides a clear direction: the miRNAs induced by TTT that are beneficial for healing are highly promising intervention targets. Although this study provides a comprehensive perspective, there are limitations and challenges. First, our miRNA omics data are from rabbit models at specific time points, and caution is needed when extrapolating to humans. There may be species differences in miRNA sequences and regulation, and the function of human counterparts to rabbit miR-182-5p needs to be confirmed. Therefore, it is necessary to repeat sequencing analysis in clinical samples to verify whether TTT also regulates the miRNA profile in human DFU. Second, this study primarily used bioinformatics methods to predict miRNA target genes and signaling pathways. Although we cited literature and existing experimental support for these predictions, direct functional experiments are lacking. In the future, it will be necessary to perform gain- and loss-of-function experiments for each key miRNA. Additionally, the absence of an osmotic pressure control (e.g., equimolar L-mannitol) in the high-glucose cell culture model means that a contribution from non-specific hyperosmotic stress to the observed phenotypes cannot be formally excluded, and future studies should address this limitation.For example, knocking down miR-182-5p in animal models to observe whether it weakens TTT’s healing-promoting effects, or administering miR-182-5p mimics to see if it can accelerate healing without TTT. If the results are consistent, this would solidify the causal relationship. Third, we observed that TTT affects multiple pathways, but it is still unclear whether these pathways interact or cooperate. Wound healing is a multi-factorial process, and we speculate that Wnt, Hippo, cAMP, and other signaling pathways may not be isolated in the TTT mechanism, but are jointly involved in regulation. For example, YAP/TAZ and β-catenin have been reported to have interdependencies in wounds. Therefore, further proteomics and signaling pathway analysis (e.g., phospho-protein arrays, single-cell sequencing) will help reveal the full signaling network triggered by TTT. Through these in-depth studies, we can identify which pathway is dominant, which are secondary, and their temporal order, thus pinpointing the most critical intervention nodes. Looking ahead, we believe there are several directions worth pursuing: (1) Clinical validation: Conduct prospective clinical studies to collect wound biopsies/liquid samples from TTT patients and verify the changes in miRNA and target genes through high-throughput sequencing or specific PCR, assessing their value as efficacy indicators or predictive factors; (2) Functional studies: Use transgenic animals or gene-editing technologies to knock out or overexpress key miRNAs or their target genes, and directly observe the effects on wound healing, establishing causal links between mechanical stimulation, gene expression, and functional outcomes; (3) Combined intervention: Based on the miRNA network identified in this study, design animal experiments combining TTT with local delivery of miR-150/182/205 mimics, or TTT with drugs that regulate Wnt/YAP pathways, to assess potential synergistic effects, providing evidence for clinical combination therapies; (4) Long-term follow-up: Conduct long-term follow-up studies of TTT-treated patients, including healing quality (scar conditions), recurrence rates, and changes in lower limb vascular function, and correlate postoperative miRNA levels with these long-term outcomes, to determine the long-term benefits and potential areas for improvement in TTT. Through these efforts, we hope to not only comprehensively reveal the biological effects of TTT but also apply this knowledge to improve the overall treatment strategy for DFU. In conclusion, TTT, as an interdisciplinary innovative therapy, has broad prospects, and understanding its mechanisms of action is key to unlocking these prospects. This study takes an important step forward by integrating multi-level evidence, and future research and exploration will continue to build on this foundation to write a new chapter in the treatment of diabetic foot. Conclusion This study demonstrates that tibial transverse transport (TTT) significantly reshapes the miRNA expression profile in diabetic foot ulcer (DFU) wounds, particularly by upregulating miR-182-5p, miR-150-5p, and miR-205-5p. Among these, miR-182-5p plays a key role in promoting fibroblast proliferation and migration by activating the HIF-1α/VEGF pathway under high glucose conditions. Our findings suggest that the reparative effects of TTT are not only due to biomechanical regulation but also involve miRNA-mediated molecular regulation. These results provide new insights into the molecular therapeutic mechanisms of TTT and support the combination of mechanical interventions with targeted miRNA strategies to enhance the potential for diabetic wound healing. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2