This article is mentioned in:

Introduction

Diabetic foot ulcer (DFU) is a common, highly morbid consequence of longstanding and poorly-managed diabetes, exhibiting a complex underlying pathophysiology. Among the estimated 537 million individuals with diabetes worldwide (1), 19-34% will develop a DFU in their lifetime and 17% of patients with complications associated with DFU undergo lower-extremity, minor (below the ankle), major (above the ankle) or minor and major amputation (2). Additionally, the severity of DFU causes death in 10% of these patients within 1 year of their first diagnosis (3). Therefore, researchers have been investigating clinical treatments and underlying mechanisms that can facilitate DFU wound healing and enhance patient survival rates.

Normal skin wound healing has four consecutive and overlapping stages: i) The 'hemostasis' stage; ii) the 'inflammation' stage; iii) the 'proliferation' stage; and iv) the 'reshaping' stage (4). In the diabetes wound, tissue ischemia, hypoxia and the high glucose microenvironment interfere with the progress of these programmed healing stages, resulting in delayed or absent healing of the wound and some clinical complications (5). Re-epithelialization and dermal repair play a crucial role in the DFU healing process. Research has shown that during the re-epithelialization stage of wounds, keratinocytes can migrate to the wound site and proliferate and differentiate into different structures to restore the integrity of the epidermal structure and function (6). However, the high glucose environment and the chronic inflammatory state of the diabetes wound damage the normal function of keratinocytes, resulting in delayed re-epithelialization of the wound (7). Therefore, exploring therapeutic targets to improve the function of diabetes skin keratinocytes to promote chronic wound healing has become one of the research hotspots in the DFU field in recent years.

RNA-binding proteins (RBPs) play a key role in inflammation and immune regulation by affecting the metabolism and stability of mRNA. The fused in sarcoma (FUS) and interleukin enhancer-binding factor 2 (ILF2) proteins are RBPs that interact with RNA molecules, exerting pivotal biological functions and regulating processes such as metabolism, transport, stability and translation (8). FUS is an RBP containing 526 amino acids in the FET-binding protein family namely FUS, EWS RNA binding protein 1 and TATA-box binding protein associated factor 15. FUS is a widely expressed protein that can shuttle between the nucleus and the cytoplasm (9). A study has shown that the expression of RBP genes, such as FUS, in retinal microvascular endothelial cells of high glucose-induced diabetic mice is markedly different from that in the normal control group (10). Another study has shown that FUS can bind to paired box 3 mRNA and positively regulate its expression, increasing cardiac fibroblast fibrosis (11). In addition, a study has shown that circular (circ)FNDC3B participates in the regulation of angiogenesis through interaction with FUS (12). Although the research on FUS has made progress in some diseases, its role in diabetic skin wound healing is rarely reported. ILF2, also known as nuclear factor 45 (NF45), is essential for cell growth and inflammatory responses, participating in DNA damage repair and cell division as well as affecting cyclin expression (13). Numerous studies have shown that ILF2 plays a key role in promoting the proliferation of cancer cells (14,15). In addition, Jin et al (16) found that ILF2 inhibited NLR family pyrin domain containing 3 inflammasome activation in macrophages. However, the role of ILF2 in diabetic skin keratinocytes should be investigated further.

Negative pressure wound therapy (NPWT) is a widely adopted strategy in contemporary wound care, particularly recommended for managing complex wounds, such as foot wounds in individuals with diabetes mellitus (DM). Evidence shows that NPWT reduces rehospitalizations, associated surgical procedures, dressing changes, personnel commitments, hospitalization, treatment time and costs (17,18). A number of findings suggest that treating DFU with NPWT reduces ulcer size, enhances granulation tissue formation, shortens hospital stay and confers complete wound healing (19-21). This therapy also demonstrates promising improvements in healing rates without a notable increase in wound complications (21). However, the specific mechanism through which NPWT promotes DFU wound healing remains unclear and deserves further investigation.

RNA sequencing (RNA-seq) is a powerful genomic technique used to study the transcriptome, which involves capturing the sequence information of RNA molecules using high-throughput sequencing (HTS) technologies. Hence, it enables the identification of upregulated (i.e., activated) or downregulated (i.e., repressed) genes under specific biological conditions, allowing crucial insights into the mechanisms of various diseases and in biomarker discovery. As such, the aim of the present study was to identify key genes that promote wound healing in patients with DFU by analyzing transcriptome sequencing data before and after NPWT, and to explore the clinical significance of these genes in promoting wound healing.

Materials and methods

Study participants and grouping

The present study included 27 patients with DFU who were hospitalized in the Department of Endocrinology at The First Affiliated Hospital of Anhui Medical University (Hefei, China) from October, 2022 to March, 2023, and received NPWT for the first time. The inclusion criteria for all participants were as follows: i) Type 2 DM (T2DM) diagnosis; ii) age range, 18-80 years old; iii) ulcer duration ≥4 weeks; iv) ulcer area range, 2-20 cm2 with a Wagner grade 2-3 (22); and v) ankle-brachial index (ABI) range, 0.7-1.3. The exclusion criteria were as follows: i) Severe kidney, liver and cardiac dysfunctions; ii) a history of malignant tumors; iii) autoimmune diseases; iv) recent use of glucocorticoids, immunosuppressants or exogenous cytokines within the last 6 months; and v) severe sepsis. All participants received a standard systemic treatment, including lipid regulation, nerve-nutrition, hypoproteinemia improvement, enhancement of blood supply of a lower limb wound and anti-infection and antihypertensive treatments. In addition, the patients received an appropriate glycemic control treatment. The participants underwent wound debridement to remove blackened and necrotic soft and bone tissues, followed by NPWT. For NPWT, a vacuum assisted closure negative pressure-assisted healing therapy system (Kinetic Concepts, Inc.; 3M) was used according to published protocols (23), at a dose of −125 mmHg (1 mmHg=0.133 kPa) for 1 week. Full-thickness skin tissue was collected within 0.5 cm of the wound edge before and after 1 week of NPWT, only after removing the negative pressure device. The collected granulation tissue was stored at −80°C for further examination.

The experimental grouping was as follows: i) RNA-seq was performed on the granulation tissue of 3 patients with DFU before and after NPWT; and ii) 24 patients with DFU receiving NPWT were defined as the pre-NPWT group and the post-NPWT group based on their pre-treatment and post-treatment status. Fig. 1 shows the flowchart of the study procedure.

Figure 1 Flow chart of the study design. DFU, diabetic foot ulcer; NPWT, negative pressure wound therapy; PCA, principal component analysis; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; RT-qPCR, reverse transcription-quantitative PCR; FUS, fused in sarcoma; ILF2, interleukin enhancer binding factor 2; RBP, RNA-binding protein; RNA-seq, RNA sequencing.

All procedures involving human participants in the present study complied with the 1964 Helsinki Declaration and subsequent amendments or similar ethical standards. The present study was approved by the Medical Ethics Committee of The First Affiliated Hospital of Anhui Medical University (approval no. CDEC000004982), and written informed consent was obtained from the subjects.

RNA extraction, library construction and RNA-seq

Total RNA was extracted from the granulation tissue of full-thickness wounds collected from 3 patients before and after NPWT using TRIzol Reagent (cat. no. 15596026CN; Thermo Fisher Scientific, Inc.). DNA digestion was carried out after RNA extraction by DNaseI. RNA quality was determined by examining the A260/A280 with a Nanodrop™ OneCspectrophotometer (Thermo Fisher Scientific, Inc.). RNA Integrity was confirmed by 1.5% agarose gel electrophoresis. Next, 2 μg total RNA was used for stranded RNA sequencing library preparation using a KC™ Stranded mRNA Library Prep kit for Illumina® (cat. no. DR08402; Wuhan Seqhealth Co., Ltd.; Wuhan Kangee Technology) following the manufacturer's instruction. After the library construction was completed, RNA quantification was performed using the Qubit3.0 with Qubit™ RNA Broad Range Assay kit (cat. no. Q10210; Thermo Fisher Scientific, Inc.) and diluted to 4 nM. The Illumina NovaSeq 6000 sequencer was for sequencing according to the instructions of the sequencing kit (cat. no. 12310ES96; Shanghai Yeason Biotechnology Co., Ltd.) and generated a paired-end reading of 150 bp. The RNA-seq data for every sample are shown in Table SI.

Bioinformatics analysis

The sequenced fragments were transformed into sequence data with CASAVA (version1.6; https://gaow.github.io/genetic-analysis-software/c/casava/) base recognition using image data measured by HTS machines. Differential expression analysis was conducted to identify differentially expressed genes (DEGs) between samples using DESeq2 (version 1.19.40; https://github.com/thelovelab/DESeq2), and the significance was inferred by P-value and false discovery rate (FDR). In the present study, the screening criteria for DEGs were FDR<0.05 and the absolute value of log2FoldChange (FC) >1.

Principal component analysis (PCA) was utilized to examine the grouping of samples based on gene expression data from the granulation tissue of full-thickness wounds collected from 3 patients before and after NPWT. The PCA, conducted using the prcomp function in the stats package of R software (version 4.4.1; https://cran.r-project.org), revealed distinct clusters of gene expression patterns.

Gene Ontology (GO) enrichment analysis was performed on the DEGs using GOSeq (version 1.0; https://bioconductor.org/packages/release/bioc/html/goseq.html) and GO terms in DEGs with a corrected P<0.05 were deemed significantly enriched. A Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis (http://www.genome.jp/kegg/) was conducted to assess the pathways enriched in the DEGs. Gene Set Enrichment Analysis (GSEA) was conducted using the clusterProfiler package (version 4.12.0; https://github.com/YuLab-SMU/clusterProfiler) under the R software with P<0.05. Gene Set Variation Analysis (GSVA) analysis was used to explore the differences in biological pathways between different pattern clusters based on the enrichment scores. The GSVA (version 1.52.3; https://github.com/rcastelo/GSVA) R software package was used to perform functional enrichment analysis on DFU wound granulation tissue samples before and after NPWT to obtain enrichment pathways. A corrected P<0.05 was considered to indicate a statistically significant difference.

Reverse transcription-qPCR (RT-qPCR)

Total RNA was extracted from the wound granulation tissue of patients with DFU, HaCaT cells and the skin tissue of mice using TRIzol reagent (cat. no. 15596026CN; Thermo Fisher Scientific, Inc.), which was reverse transcribed into cDNA using a PrimeScript RT kit (Takara Biotechnology Co., Ltd.) as per the provided protocol. qPCR was performed with 2X Q3 SYBR qPCR Master Mix (Universal) (cat. no. 22204; Tolo Biotech Co., Ltd.) and the PCR procedure was as follows: Initial denaturation at 94°C for 30 sec, followed by 40 cycles of 94°C denaturation for 5 sec, 56°C annealing for 30 sec and 72°C extensions for 10 sec. The 2−ΔΔCq method (24) was used to calculate the relative expression level of mRNA, with β-actin or 18S ribosomal RNA as the internal reference. All primers used in the present study are listed in Table SII.

Collection of general information and laboratory parameters

The laboratory parameter data were collected from medical records. The demographic information including sex and age were obtained from the patients. The area of the wound ulcers was quantified using digital photography and ImageJ (version 1.8.0; National Institutes of Health). The ABI was determined using a doppler blood flow detector (DPL-03; Shanghai Hanfei Medical Equipment Co., Ltd.). The transcutaneous oxygen pressure (TcPO2) near the ulcer was measured using a TCM 400 monitoring device (Radiometer).

After fasting for ≥8 h, venous blood samples from the elbow vein were collected between 6 and 7 am. The collected venous blood samples were used to determine various laboratory indicators, including C-reactive protein (CRP), white blood cell (WBC) count and neutrophil percentage (NEUT). Other parameters were tested using the following methods: Fasting plasma glucose (FPG) was measured using the glucose oxidase method; glycosylated hemoglobin A1c (HbA1c) was measured using high-pressure liquid chromatography; triglyceride (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C) and low-density lipoprotein cholesterol (LDL-C) were measured using oxidase-linked colorimetry; and renal function was measured using the estimated glomerular filtration rate (eGFR).

Detection of related biomarkers in granulation tissue

Granulation tissue frozen at −80°C was reduced to a fine powder in a nitrogen-chilled mortar and transferred to a liquid nitrogen pre-cooled centrifuge tube. The tube was centrifuged at 9,659.5 × g for 10 min, 4°C, followed by aspirating the supernatants. The reactive oxygen species (ROS) levels were assessed using a dichloro-dihydro-fuorescein diacetate assay (cat. no. KGA7501-1000; Jiangsu Kaiji Biotechnology Co., Ltd.) and the malondialdehyde (MDA) levels were assessed using a thiobarbituric acid assay (cat. no. KGA7101-100; Jiangsu Kaiji Biotechnology Co., Ltd.), following the manufacturer's instructions. The tumor necrosis factor-α (TNF-α; cat. no. E-EL-H0109; Elabscience Biotechnology Co., Ltd.; Wuhan Eliorite Biotech Co., Ltd.) and IL-4 (cat. no. E-EL-H0101; Elabscience Biotechnology Co., Ltd.; Wuhan Eliorite Biotech Co., Ltd.) levels were determined using ELISA kits following the manufacturer's instructions. The activities of the matrix metalloproteinase (MMP) 2 and MMP9 enzymes were determined with a gelatin zymogram kit (Cosmo Bio Co., Ltd.) according to kit's instructions.

Western blotting

Total protein was extracted from cells and the granulation tissue from patients using radioimmunoprecipitation assay buffer (cat. no. P0013B; Beyotime Institute of Biotechnology), and protein concentrations were determined by bicinchoninic acid protein assay (cat. no. P0010; Beyotime Institute of Biotechnology). Equal amounts (20 μg) of protein were loaded per lane and separated by 10% sodium dodecyl sulfate polyacrylamide gel electrophoresis, and transferred to polyvinylidene fluoride membranes, which were blocked with 5% skimmed milk at room temperature for 2 h. The membranes were probed overnight at 4°C with the following primary antibodies: Rabbit anti-FUS (1:5,000; cat. no. 11570-1-AP; Proteintech Group, Inc.), rabbit anti-ILF2 (1:1,000; cat. no. 14714-1-AP; Proteintech Group, Inc.) and mouse β-actin (1:1,000; cat. no. TA-09; ZSGB BIO; OriGene Technologies, Inc.). Subsequently, the membranes were incubated with horseradish peroxidase (HRP)-conjugated goat anti-rabbit secondary antibody (1:2,000; cat. no. ab205718; Abcam) or anti-mouse secondary antibody (1:5,000; cat. no. 550010; Chengdu Zen-Bioscience Co., Ltd.) at room temperature for 2 h. Enhanced chemiluminescence substrates (cat. no. SB-WB004; ShareBio; Shanghai Shenger Biotechnology Co., Ltd.) were used for signal detection, and ImageJ was used to identify the grayscale value of the target protein compared with the internal reference grayscale ratio.

Immunohistochemistry

The granulation tissue obtained from patients was fixed in 4% paraformaldehyde at room temperature for 24-36 h. The tissue was embedded in paraffin, cut into 5-μm-thick slices and placed on a glass slide. The slices were immersed in xylene for 5 min 3 times for dewaxing, then in anhydrous alcohol, 95% alcohol and 80% alcohol for 5 min each. After removal, the slices were washed several times with tap water. Subsequently, the slices were completely immersed in 40 ml EDTA repair solution (pH 8.0; cat. no. ZLI-90667; ZSGB-BIO; Beijing Zhongshan Jinqiao Biotechnology Co., Ltd.), and the boiling state of the pressure cooker was maintained for 2.5 min. Then, the pressure cooker was depressurized and washed with cold water, followed by further drip cooling. The slices were removed, cooled and placed in a reaction chamber. Then, the slices were washed with PBS for 3 min 3 times, and the endogenous peroxidase activity was blocked with 3% H2O2 solution at room temperature for 10 min, before again washing with PBS buffer for 3 min 3 times. The FUS (1:200; cat. no. 11570-1-AP; Proteintech Group, Inc.) and ILF2 (1:100; cat. no. 14714-1-AP; Proteintech Group, Inc.) primary antibodies were added to the slices dropwise and incubated at 37°C for 60 min, before washing with PBS for 3 min 3 times. The HRP-conjugated secondary antibody (1:2,000; cat. no. ab205718; Abcam) was added dropwise and incubated at room temperature for 10 min, before washing with PBS for 3 min 3 time. Next, the slices were incubated with DAB color reagent kit at room temperature for 5-8 min, then washed with tap water to terminate staining after the color development completed. The slices were counterstained with hematoxylin for 1 min, washed with warm water and differentiated with hydrochloric acid and alcohol. The slices were finally dehydrated, treated with xylene for transparency and sealed with neutral resin. Images were obtained using a light microscope and further analyzed using ImageJ.

Cell culture and transfection

Human epidermal keratinocyte (HaCaT) cells were purchased from Wuhan Pricella Biotechnology Co., Ltd. (cat. no. CP-H113), which detected the quality of the cell line by PCK immunofluorescence, with a purity of >90%. The cell line used in the present study was also confirmed to be the HaCaT cell line by the STR analysis method. The HaCaT cells were cultured in low-glucose Minimum Essential Medium (MEM) containing NEAA (cat. no. MP150410; Procell Life Science & Technology Co., Ltd.) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin and cultured at 37°C in a cell culture incubator maintained at 5% CO2. When the cell growth reached >90% confluency, trypsin digestion and passage culture were carried out.

After cell passage, the HaCaT cells were cultured in a 6-well plate and randomly divided into three groups: i) The normal glucose (NG) group (5.5 mM D-glucose); and ii) the high glucose (HG) groups (25 mM or 50 mM D-glucose). Subsequently, FUS short-interfering RNA (siRNA) and ILF2 siRNA (Guangzhou RiboBio Co., Ltd.) were transfected into HaCaT cells in the NG (5.5 mM) and HG (50 mM) groups. Briefly, 0.1 nM siRNA/NC was prepared in in 200 μl Opti-MEM (cat. no. 31985070; Thermo Fisher Scientific, Inc.), then incubated with Hieff Trans (12 μl) in vitro siRNA transfection reagent (cat. no. 40806ES03; Shanghai Yeason Biotechnology Co., Ltd.) at room temperature for 10 min. Then, the siRNA mixture was added dropwise to each well and incubate for 48 h before subsequent experiments. After transfection, the successful knockdown using siRNA in HaCaT cells was confirmed using RT-qPCR (Fig. S1A). The siRNA sequences used in the present study are shown in Table SIII. In addition, to confirm whether the results were affected by osmotic pressure, a supplementary experiment comparing the high osmotic pressure (HO) group (5.5 mM D-glucose + 44.5 mM mannitol) with the HG group (50 mM) was conducted.

Cell viability assay

Cell viability was measured using the Cell Counting Kit-8 (CCK-8) assay (Biosharp Life Sciences). The normal digested resuspension of each group of cells was added to the wells of a 96-well plate at a density of ~1×105 cells/100 μl/well. The edge wells were filled with sterile phosphate-buffered saline (PBS) and the inoculated cell culture plate was placed in the incubator (37°C, 5% CO2). After 48 h of culture, the CCK-8 reagent was added to each well for 30-60 min according to the manufacturer's instructions. The absorbance values of each well were measured at a wavelength of 450 nm using an enzyme-linked immunosorbent assay plate reader.

Wound healing assay

Logarithmic cells were inoculated at a density of 1×106 cells/well into a 6-well plate, with three independent wells set up for each group. When cell monomers adhered to the wall and proliferated to a cell density of 90-100%, a 200-μl pipette tip was used to vertically scratch the plate. The suspended cells were washed three times with PBS to remove the scraped cells, then serum-free MEM was added, and the plate was cultured in an incubator (37°C, 5% CO2). Images were obtained using an optical microscope at 0 and 48 h and the experiment was repeated three times. ImageJ (version 1.8.0; National Institutes of Health) was used to calculate the wound healing rate. The wound healing rate was calculated as a percentage using the following formula: Wound healing rate (%)=[(0 h scratch area-48 h scratch area)/0 h scratch area] ×100%.

Cell apoptosis assay

Cell apoptosis was evaluated using the Annexin V-FITC/PI apoptosis detection kit (Beijing Solarbio Science & Technology Co., Ltd.). HaCaT cells from each independent sample were resuspended in 100 μl binding buffer and the membrane was stained with annexin V-FITC (5 μl) and PI (5 μl) in the dark for 15 min. Flow cytometry (CytoFLEX LX; Beckman Coulter Inc.) was performed and cell apoptosis images were analyzed using CytExpert (version 2.4.0.28; Beckman Coulter Inc.).

Animal experiments

All animal experiments were approved and performed in accordance with the guidelines of the Animal Care and Use Committee of Anhui Medical University (Hefei, China; approval no. LLSC20201040). Male C57BL/6 mice (6 weeks old, 18-20 g) were purchased from GemPharmatech Co., Ltd., raised in controlled habitats and provided with water and food. A total of 20 mice (n=5 per group) were placed in a stainless steel cage covered with wood shavings, at a temperature of ~22°C, with a light/dark cycle of 12:12 h and a humidity of ~ 50%. The health of the mice was monitored by observing the temperature, humidity, noise and lighting conditions in the animal room, and monitoring the weight and general status of mice, such as mental state, activity, hair loss, excretion, food intake and water intake, to evaluate the health and behavioral status of mice. After the mice became familiar with their habitat for 1 week, 15 mice were randomly selected and were intraperitoneally injected with streptozotocin (STZ) to create a diabetes model (50 mg/kg for 5 consecutive days). Starting from day 12 after the first injection of STZ, the fasting blood-glucose was measured using the tail cut blood sampling method twice a week for 2 weeks. When the blood glucose was ≥250 mg/dl (16.7 mmol/l) twice in a row, diabetes was diagnosed. All mice met the diagnostic criteria for diabetes.

The adeno-associated virus (AAV) short-hairpin (sh)FUS or shILF2 vectors (Shanghai GenePharma Co., Ltd.) were generated by replacing the FUS/ILF2 transcriptional sequence with the shRNA target of FUS/ILF2 (shFUS, 5′-CTTCAAGCAGATTGGAATTAT-3′; shILF2, 5′-CCGGCAGGTAGGATCATATAA-3′). AAV-shRNA NC was used as the negative control (shRNA NC, 5′-ACTACCGTTGTTATAGGTG-3′). The back hair of the mice was shaved off and AAV-shFUS, AAV-shILF2 or AAV-shRNA NC vector was subcutaneously injected into the back of the mice (1.0×1011 vector genomes/mouse, 100 μl), and the knockout efficiency in mouse skin tissue was confirmed through RT-qPCR after 4 weeks.

Mice were anesthetized by intraperitoneal injection of 1% sodium pentobarbital (50 mg/kg) (25), then wiped with 5% iodophor disinfectant. After drying, the skin was disinfected with 75% alcohol by rotating outward extension, which was repeated twice. Under anesthesia, a 6-mm biopsy punch was used to symmetrically form two full-thickness skin excision wounds near the midline of the back. The mice were divided into the normal control (Ctrl), the DM-Ctrl, the DM FUS knockdown (DM-shFUS) and the DM ILF2 knockdown (DM-shILF2) groups. Images were collected and the size of the wound area was measured every other day. A ruler was placed around the wound as a reference to correct the distance between the camera and the animal. After the study was completed, mice were euthanized by intraperitoneal injection of 3% pentobarbital sodium (100 mg/kg). The breathing, heartbeat and consciousness of the mice were observed for >3 min until their death was confirmed (26). If the mice exhibited persistent pain behavior, severe dehydration, inability to eat, extreme fatigue or even severe infection during the research process, they were euthanized; however, no animals in the present study reached these humane endpoints. ImageJ (version 1.8.0; National Institutes of Health) was used to calculate the wound area in pixels. The wound closure rate was calculated as a percentage using the following formula: Wound healing (%)=[(initial wound area-final wound area)/initial wound area] ×100%.

Statistical analysis

Data analysis was conducted using SPSS (version 22.0; IBM Corp.). The analysis software used to intersect DEGs with the RBP2GO dataset was the Venny tool (version 2.1; https://bioinfogp.cnb.csic.es/tools/venny/index.html) (27). Measured data with a normal distribution are presented as the mean ± standard deviation, and those with a non-normal distribution are presented as the median [interquartile range: P25, P75]. Comparisons before and after NPWT were performed using paired t-test or the Wilcoxon rank-sum test. Ranked data were analyzed using the Fisher's exact test or Kruskal Wallis test followed by the Dunn's post hoc test. One-way analysis of variance were used to determine the statistical significance between multiple groups followed by the Bonferroni post hoc test. Changes before and after NPWT were calculated based on the values of various indicators measured before and after the therapy (Δ value). The associations between the expression changes of target proteins and other clinical variables were evaluated with Pearson's correlation coefficient. P<0.05 was considered to indicate a statistically significant difference.

Results

General information on the 3 subjects

A total of 3 patients with DFU were subjected to RNA-seq (Table I; patients 1, 2 and 3). Of the 3 patients, 1 was female and 2 were male, all with T2DM and aged 47-64. The patients had a long disease course, poor blood-glucose control, a long DFU time and normal blood perfusion. Generally, their wounds had similar infection statuses and blood supplies before NPWT. In addition, after 1 week of NPWT and a systemic standard treatment, laboratory results demonstrated that the WBC, NEUT and CRP levels in the blood decreased, indicating that 1 week of NPWT improved the wound infection of these 3 patients with DFU (Table I).

Table I Clinical characteristics of 3 participants with a diabetic foot ulcer subjected to RNA sequencing.

Table I

Clinical characteristics of 3 participants with a diabetic foot ulcer subjected to RNA sequencing.

Clinical feature	Patient 1	Patient 2	Patient 3
Sex	Male	Male	Female
Age, years	58	64	47
BMI, kg/m2	22.89	24.52	21.94
Ulcer duration, weeks	5	9	7
Ulcer area, cm2	8.7	12.5	10.9
Diabetes duration, years	11.0	19.0	9.0
Diabetes type	2	2	2
Wagner grade	3	3	3
ABI	1.03	0.91	0.98
TcPO2, mmHg
Right	70.2	73.6	71.4
Left	68.9	72.5	70.7
FPG, mmol/l	7.9	9.0	8.4
HbA1c, %	8.0	8.8	8.6
WBC, ×109
Before NPWT	11.8	13.2	12.8
After NPWT	8.5	9.0	8.7
NEUT
Before NPWT	80.3	88.9	85.5
After NPWT	69.7	73.5	71.0
CRP, mg/dl
Before NPWT	19.3	22.1	20.4
After NPWT	10.9	13.2	11.3

[i] BMI, body mass index; ABI, ankle brachial index; TcPO2, transcutaneous oxygen pressure; FPG, fasting plasma glucose; HbA1c, glycated hemoglobin A1c; WBC, white blood cell; NPWT, negative pressure wound therapy; NEUT, neutrophil percentage; CRP, C-reactive protein.

Transcriptome sequencing analysis of granulation tissue

To explore the biological significance of the DFU wound tissue samples before and after NPWT, GSEA was performed to assess the association between gene sets from the GO and KEGG databases and the transcriptome. The enrichment analysis of the top 10 GO gene sets was mainly related to cell tissue formation and lipid transport (P<0.05; Fig. 2A). The enrichment analysis of the top 10 KEGG gene sets was mainly related to the 'IL-17 signaling pathway' and 'Insulin resistance' (P<0.05; Fig. 2B). In addition, GSVA was also conducted and it was found that before NPWT, the genes were mainly enriched in amino acid metabolism-related signaling pathways and after NPWT, the genes were mainly enriched in signaling pathways related to cell apoptosis (Fig. 2C).

Figure 2 GSEA. (A) Top 10 GO gene sets showing significant enrichment in the GSEA, P<0.05. (B) Top 10 KEGG gene sets showing significant enrichment in the GSEA, P<0.05. (C) Heat map of the Gene Set Variation Analysis showing the significantly different signaling pathways. GO, Gene Ontology; GSEA, Gene Set Enrichment Analysis; KEGG, Kyoto Encyclopedia of Genes and Genomes.

Clusters of distinct groups were evident in the principal component analysis (Fig. 3A). Using |log2FC|≥1 and adjusted P<0.05 as the screening threshold, 199 DEGs were identified, of which 101 were upregulated and 98 were downregulated genes (Fig. 3B and C). To elucidate the biological characteristics of the DEGs, GO and KEGG pathway analyses were used to demonstrated that that DEGs have notable roles in cell proliferation and extracellular matrix (ECM) synthesis, mainly involving pathways in inflammation and immune cell signaling, amino acid metabolism and lipid metabolism (Fig. 3D and E; Tables SIV and SV).

Figure 3 DEGs identified in the RNA-sequenced granulation tissue of 3 patients with diabetic foot ulcer before and after NPWT, using |log2FoldChange|≥1 and P<0.05 as the selection criteria. (A) PCA. (B) DEGs before and after NPWT represented as a volcano plot. (C) Cluster diagram to show the DEGs before and after NPWT. (D) GO enrichment analysis of the DEGs. (E) KEGG enrichment analysis of the DEGs. Gene Ratio was calculated and shown on horizontal ordinates. The size of the dots indicates gene numbers and distinct colors of the dots indicate the adjusted P-values. PCA, principal component analysis; DEGs, differentially expressed genes; NPWT, negative pressure wound therapy; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes.

Validation of FUS and ILF2 expression in 24 patients with DFU

Previous studies have shown that some RNA-encoded proteins are RBPs (28,29). The DEGs before and after NPWT described in the aforementioned results were intersected with the human RBP dataset (https://rbp2go.dkfz.de/), and it was shown that a total of 43 DEGs encoded RBPs. Among them, the top two with the highest RBP2GO score were FUS and ILF2 (Fig. 4A). Subsequently, the expression levels of FUS and ILF2 in granulation tissue of skin wound margin collected from 24 patients with DFU before and after NPWT were detected by RT-qPCR, western blotting and immunohistochemistry As anticipated, the mRNA and protein [tissue (T)-FUS; T-ILF2] expression levels of FUS and ILF2 in granulation tissue were markedly upregulated after 1 week of NPWT (Fig. 4B and C). Moreover, immunohistochemical staining showed that the expression of FUS and ILF2 in the granulation tissue of the wound margin increased significantly after 1 week of NPWT (Fig. 4D), confirming the effectiveness of RNA-seq and the accuracy of the results.

Figure 4 Expression of FUS and ILF2 before and after NPWT. (A) Venn diagram showing the intersection of the RBP dataset and the DEGs from the transcriptome sequencing. (B) The relative expression of FUS and ILF2 mRNA in granulation tissue of wounds before and after NPWT. (C) Changes in protein expression of FUS and ILF2 in the pre-NPWT group (designated as minus symbol '-') and post-NPWT group (designated as plus symbol '+'). (D) Immunohistochemical staining was used to observe the expression of FUS and ILF2 in the granulation tissue of wounds before and after NPWT. Brown is the targeted gene-specific staining, and blue is other nuclei; magnification, ×400. RBP, RNA-binding protein; NPWT, negative pressure wound therapy; FUS, fused in sarcoma; ILF2, interleukin enhancer binding factor 2; DEGs, differentially expressed genes.

Comparison of primary parameters among 24 patients before and 1 week after NPWT

After 1 week of NPWT and 1 week systemic standard treatment, the FPG (P<0.05) and the inflammatory markers, WBC, NEUT and CRP, of patients significantly decreased (P<0.05) (Table II). In addition, the levels of ROS, lipid peroxides (MDA) and the pro-inflammatory factor, TNF-α, significantly decreased, and those of the anti-inflammatory factor, IL-4, significantly increased in the wound granulation tissue after NPWT (P<0.05). Notably, the MMP2 and MMP9 levels also significantly decreased (P<0.05). Except for the aforementioned indicators, no significant change was observed in other indicators before and after NPWT (P>0.05).

Table II Comparison of the primary parameters among the 24 participants before and 1 week after NPWT.

Table II

Comparison of the primary parameters among the 24 participants before and 1 week after NPWT.

Variables	Pre-NPWT	Post-NPWT	P-value
Sex, male/female	18/6	18/6	-
Age, years	49.2±11.9	49.2±11.9	-
BMI, kg/m2	24.22±2.13	24.09±2.26	0.086
Ulcer duration, weeks	6.7±0.7	7.75±1.29	0.010
Ulcer area, cm2	9.5±1.3	8.00±1.12	<0.001
ABI	0.89±0.04	0.89±0.03	0.663
FPG, mmol/l	9.50±3.37	6.14±0.89	<0.001
HbA1c, %	8.11±1.19	8.08±1.27	0.565
TG, mmol/l	1.70±0.41	1.70±0.40	0.435
TC, mmol/l	5.52 (4.32, 6.01)	5.51 (4.36, 6.05)	0.156
LDL-C, mmol/l	3.05 (2.63, 3.21)	3.06 (2.56, 3.25)	0.435
HDL-C, mmol/l	1.26±0.17	1.26±0.17	0.784
eGFR, ml/min/1.73 m2	94.88±13.48	93.75±11.40	0.200
WBC, ×109	15.78 (11.40, 25.53)	8.32 (6.92, 12.40)	<0.001
NEUT	78.00 (71.2, 83.8)	64.66 (59.25, 71.25)	<0.001
CRP, mg/dl	35.85 (17.53, 47.28)	9.06 (7.74, 15.07)	<0.001
TNF-α, pg/ml	97.70±3.02	65.85±3.27	<0.001
IL-4, ng/l	45.75±3.28	75.72±3.57	<0.001
ROS, U/ml	8.58±0.48	4.16±0.47	<0.001
MDA, nmol/mg	3.71±0.33	1.62±0.36	<0.001
MMP2, ng/ml	204.79±5.58	152.09±5.70	<0.001
MMP9, ng/ml	145.99±3.49	106.80±4.96	<0.001
T-FUS	0.44 (0.38, 0.46)	0.84 (0.80, 0.94)	<0.001
T-ILF2	0.45±0.08	0.74±0.07	<0.001

[i] Data are presented the mean ± standard deviations or the median and interquartile range. Differences between two groups were analyzed using paired t-test or the Mann Whitney U test. NPWT, negative pressure wound therapy; BMI, body mass index; ABI, ankle brachial index; FPG, fasting plasma glucose; HbA1c, glycated hemoglobin A1c; TG, triacylglycerol; TC, total cholesterol; LDL-C, low density lipoprotein cholesterol; HDL-C, high density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate; WBC, white blood cell; NEUT, neutrophil percentage; CRP, C-reactive protein; TNF-α, tumor necrosis factor-α; IL-4, interleukin-4; ROS, reactive oxygen species; MDA, malondialdehyde; MMP2/9, matrix metalloproteinase 2/9; T-FUS, the protein expression levels of fused in sarcoma in the granulation tissue; T-ILF2, the protein expression levels of interleukin enhancer binding factor 2 in the granulation tissue.

Correlation between ΔT-FUS and ΔT-ILF2 and the Δvalues of clinical parameters before and after NPWT

The changes in the T-FUS and T-ILF2 (ΔT-FUS and ΔT-ILF2) before and after NPWT were negatively correlated with ΔFPG, ΔWBC, ΔNEUT, ΔCRP, ΔTNF-α, ΔROS, ΔMDA, ΔMMP2 and ΔMMP9 (P<0.05), but positively correlated with ΔIL-4 (P<0.05) (Fig. 5). In addition, no significant correlation with other clinical indicators was noted (P>0.05).

Figure 5 Pearson's correlation coefficient between changes in T-FUS and T-ILF2 and changes in other clinical parameters (r). T-FUS, the protein expression levels of FUS in the granulation tissue; T-ILF2, the protein expression levels of ILF2 in the granulation tissue; NPWT, negative pressure wound therapy; ABI, ankle brachial index; FPG, fasting plasma glucose; HbA1c, glycated hemoglobin A1c; TG, triacylglycerol; TC, total cholesterol; LDL-C, low density lipoprotein cholesterol; HDL-C, high density lipoprotein cholesterol; eGFR, estimated glomerular filtration rate; WBC, white blood cell; NEUT, neutrophil percentage; CRP, C-reactive protein; TNF-α, tumor necrosis factor-α; IL-4, interleukin-4; ROS, reactive oxygen species; MDA, malondialdehyde; MMP2, matrix metalloproteinase 2; MMP9, matrix metalloproteinase 9; FUS, fused in sarcoma; ILF2, interleukin enhancer binding factor 2.

Correlation between ΔT-FUS and ΔT-ILF2, and the wound outcomes

The median values of ΔT-FUS and ΔT-ILF2 were used as the cut-off values to investigate the relationship between ΔT-FUS and ΔT-ILF2 and the wound healing rate after 4 weeks of NPWT. Patients with values below the median were classified as the low change group (LCG), while those with values equal to or greater than the median were designated the high change group (HCG). It was observed that ΔT-FUS and ΔT-ILF2 were positively correlated with the wound healing rate 4 weeks after NPWT (ΔT-FUS, P=0.043; ΔT-ILF2, P=0.011; Table III). In addition, the wound healing rate of patients with DFU after stopping NPWT for 4 weeks was 41.67% (n=10/24).

Table III Association between ΔT-FUS and ΔT-ILF2 and wound outcomes.

Table III

Association between ΔT-FUS and ΔT-ILF2 and wound outcomes.

	ΔT-FUS			ΔT-ILF2
	HCG, n=12	LCG, n=12	P-value	HCG, n=11	LCG, n=13	P-value
Mean age ± SD, years	49.7±13.0	48.7±11.1	0.842	49.0±13.1	49.3±11.3	0.951
Sex, n (%)			1.000			0.166
Male	9 (75.0)	9 (75.0)		10 (90.9)	8 (61.5)
Female	3 (25.0)	3 (25.0)		1 (9.1)	5 (38.5)
Ulcer duration, n (%)			0.742			0.860
≤6 weeks	4 (33.3)	5 (41.7)		4 (36.4)	5 (38.5)
6-10 weeks	7 (58.3)	5 (41.7)		5 (45.5)	7 (53.8)
>10 weeks	1 (8.4)	2 (16.6)		2 (18.1)	1 (7.7)
Ulcer area, n (%)			0.688			0.440
≤10 cm2	6 (50.0)	5 (41.7)		6 (54.5)	5 (38.5)
>10 cm2	6 (50.0)	7 (58.3)		5 (45.5)	8 (61.5)
Wagner, n (%)			0.680			0.918
Ⅱ	4 (33.3)	5 (41.7)		4 (36.4)	5 (38.5)
Ⅲ	8 (66.7)	7 (58.3)		7 (63.6)	8 (61.5)
Ulcer healing rate after			0.043			0.011
4 weeks, n (%)
Healing	8 (66.7)	2 (16.7)		8 (72.7)	2 (15.4)
Non-healing	4 (33.3)	10 (83.3)		3 (27.3)	11 (84.6)

[i] Data are presented as the mean ± standard deviation or numbers (%). The differences between two groups were analyzed using an unpaired t-test, Fisher's exact test or Kruskal Wallis test followed by the Dunn's post hoc test. T-FUS, the protein expression levels of fused in sarcoma in the granulation tissue; T-ILF2, the protein expression levels of interleukin enhancer binding factor 2 in the granulation tissue; HCG, high change group; LCG, low change group.

Effect of FUS and ILF2 on HaCaT cell function

To investigate the effect of FUS and ILF2 on HaCaT cell function, the following experiment was conducted: HaCaT cells were cultured in 5.5, 25 and 50 mM glucose for 72 h, and it was found that the mRNA and protein levels of FUS and ILF2 in the HG group (50 mM) were significantly reduced compared with the NG group (Fig. 6A). In addition, it was confirmed that the effect of high sugar on target genes was not related to the effect of osmotic pressure (Fig. S2). Therefore, a high glucose concentration of 50 mM was selected as the concentration for the HG group for subsequent experiments. After transfecting HaCaT cells of the NG and HG groups with FUS and ILF2 siRNA (Fig. 6B), the expression of FUS and ILF2 was significantly reduced (P<0.05), indicating successful transfection.

Figure 6 Effects of FUS and ILF2 on the proliferation, migration and apoptosis of HaCaT cells. (A) mRNA and protein expression levels of FUS and ILF2 in HaCaT cells under different glucose concentration were measured by RT-qPCR and western blotting, respectively. (B) mRNA expression levels of FUS and ILF2 in HaCaT cells under different conditions were measured by RT-qPCR. (C) Cell viability was monitored by the Cell Counting Kit-8 assay. (D) Cell migratory capability was measured by wound healing assay. (E) Cell apoptosis was measured by flow cytometry (Annexin V-PI). Data are presented as the mean ± SEM. One-way analysis of variance was used to determine the statistical significance between groups followed by the Bonferroni post hoc test. vs. NG, **P<0.01; vs. NG + si(FUS/ILF2)-NC, #P<0.05; vs. HG + si(FUS/ILF2)-NC, &P<0.05. FUS, fused in sarcoma; ILF2, interleukin enhancer binding factor 2; RT-qPCR, reverse transcription-quantitative PCR; NG, normal glucose; HG, high glucose; siRNA, short-interfering RNA; NC, negative control.

The CCK-8 assay results indicated that the cell viability of the HG-siFUS/ILF2-NC group was significantly reduced compared with the NG-siFUS/ILF2-NC group (P<0.05). In addition, under different glucose concentrations, knockdown of the FUS and ILF2 genes significantly reduced the cell viability (P<0.05; Fig. 6C).

The results of the wound healing experiment showed that, compared with the NG-siFUS/ILF2-NC group, the HG-siFUS/ILF2-NC group exhibited a significant decrease in cell migration ability (P<0.05). In addition, HaCaT cells under different glucose concentration conditions that were transfected with FUS or ILF2 siRNA showed a significant decrease in the cell migration rate (P<0.05; Fig. 6D).

Flow cytometry analysis was used to evaluate the apoptosis of HaCaT cells, and the results revealed a significant increase in cell apoptosis rate under HG-siFUS/ILF2-NC conditions compared with NG-siFUS/ILF2-NC conditions. Moreover, under both NG and HG conditions, the apoptosis rate of HaCaT cells transfected with FUS or ILF2 siRNA was significantly increased (P<0.05; Fig. 6E). In summary, HG culture conditions caused a degree of damage to the function of HaCaT cells, while FUS and ILF2 enhances the proliferation and migration ability of HaCaT cells, while reducing cell apoptosis.

Effects of FUS and ILF2 on wound healing in diabetic mice

After subcutaneously injecting AAV-shFUS, AAV-ILF2 or AAV-shCtrl vector into the backs of diabetic mice for 4 weeks, the knockout efficiency was verified by RT-qPCR (Fig. S1B). As shown in Fig. 7A, compared with DM-Ctrl, the expression level of the target genes in the DM-shFUS and DM-shILF2 groups significantly decreased (P<0.05). Additionally, compared with healthy mice, wound healing in diabetic mice was significantly delayed. Furthermore, compared with the DM-Ctrl mice, wound healing in diabetic mice was slower after the administration of AAV-shFUS (P<0.05) or AAV-shILF2 (P<0.05) (Fig. 7B and C).

Figure 7 Effects of FUS and ILF2 on wound healing in diabetic mice. (A) Relative FUS and ILF2 expression at the edge of skin wounds in each mouse model group. Total RNA was extracted and the levels of FUS and ILF2 were quantified by reverse transcription-quantitative PCR with 18S as an internal reference. (B) Representative images and (C) quantification showing changes in the area of excisional wounds on days 0, 2, 4 and 8 in diabetic mice injected with either AAV-shCtrl or AAV-shFUS/ILF2. Data are presented as the mean ± SEM; n=4. One-way analysis of variance was used to determine the statistical significance between groups followed by the Bonferroni post hoc test. vs. Ctrl, #P<0.05; vs. DM-Ctrl, &P<0.05. FUS, fused in sarcoma; ILF2, interleukin enhancer binding factor 2; sh, short hairpin; Ctrl, control; DM, diabetes mellitus.

Discussion

DFUs affect ~18.6 million individuals worldwide annually, and these ulcers are associated with impaired physical function and reduced quality of life (30). If not treated promptly, foot ulcers can progress to soft tissue infections, gangrene and limb loss (31). Therefore, identifying therapeutic targets for the healing of DFU wounds is particularly important.

There are numerous clinical adjuvant treatments for DFU, including wound dressings such as hydrogels and alginates, local treatments, placenta-derived treatments, hyperbaric oxygen therapy and NPWT (32). NPWT has been widely used as an adjuvant treatment technique for complex wounds in DFU, mainly for wounds with soft tissue infection, bone and tendon exposure, osteomyelitis, skin graft or flap graft after surgery and amputation/toe surgery (33-35). A retrospective study of 75 patients with chronic diabetic ulcers treated with surgical debridement, mesh skin grafts and NPWT for biofilm-related infections showed that all 75 wounds healed successfully, with an average complete wound healing time of 3.5±1.8 weeks (36). A 16-week-long, 18-center randomized clinical trial conducted in the United States included 162 diabetic foot amputees, and the results showed that more patients in the NPWT group healed compared with the control group, and the wound healing speed was faster than that in the control group (37). As has been previously reported, NPWT can markedly increase the number of endothelial progenitor cells in the peripheral blood of patients with DFU, thereby promoting wound healing (23). In addition, NPWT can promote wound healing in DFUs by affecting the expression of microRNAs (miRs) (38). Although previous studies have shown that NPWT has high clinical application value in promoting wound healing in DFU (39,40), its potential mechanism of action is still unclear. The present study used rigorous RNA-seq technology to accurately and systematically describe the transcriptome differences in wound granulation tissue before and after 1 week of NPWT in patients with DFU. In addition, the mechanisms of action of the top two genes encoding the RBPs, FUS and ILF2, were explored in wound healing in DFU through in vitro and in vivo experiments.

Previous studies have shown that the incidence rate of DFU is higher among male than female patients with diabetes (41). Seghieri et al (42) conducted a study based on the diabetes population in Italy and showed that of the 4,589 patients with diabetic foot (DF), 3,119 were male and 1,470 were female. In addition, a retrospective study by Iacopi et al (43) showed that 615 male patients and 227 female patients were hospitalized for DF from January, 2011 to December, 2015. Sex differences may arise from potential risk factors, access to care, differences in screening and treatment compliance (44). Additionally, the prevalence of peripheral neuropathy and peripheral vascular disease is higher in male patients with diabetes (), and the occurrence and development of DFU are typically related to poor blood-glucose control, blood circulation disorders, peripheral neuropathy and secondary infection wounds in patients with diabetes (45). This explains the sex difference in the number of male and female patients with DFU included in the present study and highlights the importance of comprehensive treatment for patients with DFU, with optimizing blood-glucose levels being the first priority. High blood-glucose can have a notable impact on endothelial cell function, and reduced endothelial cell numbers and dysfunction can lead to the development of macrovascular and microvascular complications in diabetes (46). Multiple observational studies have found a positive correlation between blood-glucose control and wound healing (47-49). In addition, a retrospective study found that intensive control of blood-glucose reduced the risk of lower limb amputation in patients by 35% (50). Similarly, the in vitro experiments of the present study showed that high glucose significantly inhibited the proliferation and migration of keratinocytes, while promoting their apoptosis. Further animal experiments showed that high glucose significantly inhibited skin wound healing.

Dyslipidemia can lead to the formation of atherosclerotic plaques and arteriosclerosis, ultimately resulting in vascular stenosis and occlusion. Peripheral blood vessels are some of the main blood vessels that are often present in atherosclerosis. In addition, vascular inflammation is one of the mechanisms of vascular disease formation (51). Therefore, effective lipid regulation is essential for the treatment of DFU. It is noteworthy that strict blood-glucose control and lipid-lowering therapy can also markedly improve diabetic peripheral neuropathy (52). More notably, antibiotic treatment is needed in cases of DFU complicated by infection. The Infectious Diseases Society of America recommends a 1-2-week course of antibiotics for mild infections and a 2-3-week course of antibiotics for moderate-to-severe infections, but antibiotics can typically be stopped once the clinical symptoms and signs of infection have resolved (53). As demonstrated in the present study, the hematological inflammatory indicators (WBC, CRP and NEUT) in patients with DFU after NPWT were significantly decreased compared with those before treatment, which may be due to the rational use of antibiotics during the entire treatment period, although it cannot be ruled out that NPWT use may have also contributed. Previous studies have shown that NPWT can inhibit bacterial growth in tissues, reduce inflammation and alleviate oxidative stress (54). This was also reflected in the results of the present study, where the markers of oxidative stress (ROS and MDA) were significantly decreased after NPWT.

RBPs are associated with numerous diseases, particularly metabolic disorders such as hyperuricemia, hyperlipidemia, hypertension, non-alcoholic fatty liver and diabetes (7), but there are relatively fewer studies on their role in wound healing. There has been a study reporting that long non-coding RNA (lncRNA) TINCR binds directly to the RBP, staphylococcal nuclease and tudor domain containing 1, and mediates TGF-β1 expression to promote excessive proliferation and inflammation in burn-induced skin fibroblasts (55). Meder et al (56) reported that overexpression of the RBP, lin-28 homolog B (LIN28B), upregulates VEGFA mRNA and miR-21 expression, thereby enhancing angiogenesis and accelerating wound healing. Guo et al (57) reported that the RBP, LIN28A, promotes proliferation and ECM synthesis in human skin fibroblasts after thermal injury. ECM is a complex three-dimensional network of fibronectin and matrix, which supports and connects tissue structures, regulates tissue growth and cell physiological activities and plays an important role in organism development, tissue dynamic equilibrium and wound healing (58). MMPs, as a family of Zn2+-dependent metalloproteinases, degrade ECM components and participate in wound healing (59). A study has found that MMP2 and MMP9 are expressed at higher levels in diabetic tissue and high blood-glucose levels can increase the activity of MMP2 in vascular cells, stimulating the degradation of ECM and causing imbalance in diabetes (60). Furthermore, certain studies have shown that excessive MMP9 is a predictor of poor wound healing in diabetic skin lesions (61,62). Cui et al (63) showed that the RBP, HuR, binds to MMP9 mRNA, enhancing its stability and promoting wound healing. The results of the present study indicated that expression of MMP2 and MMP9 in the DFU wound tissue was significantly reduced after NPWT. Conversely, expression of FUS and ILF2 was significantly increased, and the changes in these proteins were negatively correlated with the changes in MMPs. Therefore, the promotion of DFU wound healing by NPWT may be related to a significant downregulation in MMPs, and this downregulation may be related to the changes in FUS or ILF2, which warrants further investigation in the future.

FUS is an RBP with multiple functions and domains. Previous studies have shown that lncRNA GAS6-AS1 promotes tripartite motif containing 14-mediated cell proliferation, migration and invasion of colorectal cancer through competing endogenous RNA networks and FUS-dependent pathways (64). Another study has shown that FUS-mediated circRHOBTB3, a tumor activator, promotes proliferation of pancreatic ductal adenocarcinoma cells by regulating autophagy regulated by the miR-600/nucleus accumbens associated 1/Akt/mTOR axis (65). In addition, Wang et al (66) found that FUS can inhibit the proliferation and migration of human umbilical vein endothelial cells and reduce inflammation in atherosclerosis. It can therefore be inferred that FUS plays an important role in promoting cell proliferation and reducing inflammation in certain diseases. This is consistent with the findings of the present study, as it was shown that the upregulation of FUS in the granulation tissue of DFU wounds after NPWT was positively correlated with the anti-inflammatory cytokine, IL-4, and wound outcomes. By contrast, FUS was negatively correlated with inflammation-related (WBC, NEUT and CRP) and oxidative stress (ROS and MDA) markers. The in vitro experiments in the present study further demonstrated that knockdown of FUS inhibited the proliferation and migration of HaCaT cells in different glucose environments, while it promoted cell apoptosis. The in vivo experiments in the present study confirmed that knockdown of FUS delayed the skin wound healing in diabetic mice. On the contrary, a study has shown that FUS can reduce the expression of proliferation factors such as cyclin D1, thereby preventing the growth of prostate cancer cells (67). Therefore, the functions of FUS are diverse and worth exploring in depth.

In addition, ILF2 is also crucial for cell growth and the inflammatory response (13). As a previous study has shown, circ406961 inhibits the activation of the STAT3/c-Jun N-terminal kinase pathway by interacting with ILF2 protein, thereby inhibiting the PM2.5-induced inflammatory response (68). In addition, lncRNA LINC00470 can promote cell proliferation by binding to NF45/NF90 complexes (39). Previous studies have also confirmed that ILF2 promotes the proliferation of tumor cells (70,71). However, other studies have demonstrated that ILF2 promotes keratinocyte proliferation and inflammatory response in a long non-coding RNA KLHDC7B-DT-dependent manner (72). In the present study, it was shown through the analysis of clinical samples that the expression of ILF2 was significantly upregulated after NPWT. This change had the opposite trend to the proinflammatory factors and oxidative stress markers (ROS and MDA) and was positively correlated with wound outcomes. In addition, it was observed that knocking down ILF2 inhibited the proliferation and migration of HaCaT cells and caused a significant delay in the healing of skin wounds in diabetic mice.

According to the KEGG pathway analysis, the DEGs of DFU before and after NPWT were associated with the TNF signaling pathway. Previous studies have shown that ginsenoside combined with bone marrow mesenchymal stem cells can reduce the inflammatory reaction and promote the healing of diabetic skin ulcers by downregulating the expression of TNF-α, a key component of the TNF signaling pathway (73). In addition, curcumin treatment of wounds in diabetic mice can inhibit the expression of TNF-α and MMP-9 and accelerate the healing of diabetic skin wounds (74). In the present study, a significant decrease in TNF-α expression in the clinical samples after NPWT compared with before treatment was observed, and this change was significantly negatively correlated with the upregulation of FUS and ILF2. Although a series of studies have explored the mechanism by which FUS and ILF2 may promote the proliferation and migration of keratinocytes by regulating the TNF signaling pathway, reducing inflammatory reaction and oxidative stress and thus promoting the healing of diabetic skin wounds from different aspects, additional studies are warranted to further clarify the mechanism in the future.

When reviewing the present study, certain limitations were identified. First, although the present study revealed the crucial roles of the RBPs, FUS and ILF2, in promoting DFU wound healing, the downstream targets of their effects were not thoroughly identified. To comprehensively understand the mechanisms of FUS and ILF2 in promoting DFU wound healing, high-throughput studies such as RNA immunoprecipitation and transcriptome sequencing will be conducted in the future to identify and validate key genes that interact with FUS or ILF2. Second, the present study mainly focused on the effects of FUS and ILF2 on skin keratinocytes, but it is not yet clear how they affect other key cells in skin wound healing, such as dermal microvascular endothelial cells and skin fibroblasts, which requires further exploration. Third, in terms of signaling pathway research, although the present study focused on the importance of the TNF signaling pathway in promoting DFU wound healing by NPWT, there is a lack of direct experimental evidence to intervene in this pathway clinically. Finally, the present study is a single-center study with a limited clinical sample size, which may lead to slight selection bias. To more reliably evaluate the potential role of these DEGs in promoting DFU wound healing, data collection from multicenter clinical samples will be carried out in the future. Additionally, in vitro and in vivo experiments will be carried out to explore the molecular mechanisms involved, including inflammation and immune related and amino acid metabolism related-mechanisms, providing a more scientific basis for future clinical applications.

In summary, to the best of our knowledge, the present study revealed for the first time the genomic changes in the granulation tissue of DFU wounds before and after NPWT. The present study further demonstrated that the RBPs, FUS and ILF2, accelerate DFU wound healing by promoting proliferation and migration of skin keratinocytes, inhibiting inflammation and oxidative stress. These findings provide new insights for the early diagnosis, treatment monitoring and prognostic evaluation of DFUs.

Supplementary Data

Availability of data and materials

The RNA sequencing data generated in the present study may be found in the Gene Expression Omnibus database under accession no. GSE272918 or at the following URL: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE272918. All other data generated in the present study may be requested from the corresponding author.

Authors' contributions

YT contributed to the conception and design of the study, carried out the high-throughput sequencing experiments, performed the bioinformatics analysis and was a major contributor in writing the manuscript. HJ performed the collection and detection of granulation tissue. YY and DH completed the cell culture and cell function experiments. MuX and MiX conducted the data integration and analysis and participated in manuscript writing. XZ and MC analyzed and interpreted the data and reviewed the manuscript. XZ and MC confirm the authenticity of all the raw data. All authors have read and approved the final version of the manuscript.

Ethics approval and consent to participate

The patient study was approved by the Medical Ethics Committee of The First Affiliated Hospital of Anhui Medical University (approval no. CDEC000004982). Informed consent was obtained from all individual participants included in the study. All animal experiments were approved and performed in accordance with the guidelines of the Animal Care and Use Committee of Anhui Medical University (Hefei, China; approval no. LLSC20201040).

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Acknowledgements

Not applicable.

Funding

This study was supported by the Natural Science Foundation of Anhui Province in China (grant no. 2108085MH269), the Anhui Provincial Health Research Project (grant no. AHWJ2023BAc10012) and the Postgraduate Innovation Research and Practice Program of Anhui Medical University (grant no. YJS20230124).

References