Introduction

Foot ulcer is one of the serious complications of diabetes, and the treatment is difficult and expensive, which brings huge economic burden to the family and society. Tibial cortex transverse distraction (TCTD) technique, designed based on the tension-pressure rule, is a new treatment method. TCTD can stimulate the affected limb with diabetic foot ulcers, improve the circulation of microvessels, and thus provide blood supply and nutrients required for the healing of diabetic foot wounds1,2,3. However, the molecular mechanism of TCTD in the treatment of diabetic foot ulcers is still unclear.

Gene expression profile chip is to detect the expression level of genes at the mRNA level, which can usually be used to analyze the changes in the gene expression level of the studied biological subjects in specific biological processes (such as disease, apoptosis, development, differentiation, etc.). Detection of differentially expressed genes and analysis of corresponding signaling pathways in biological processes can provide a theoretical basis for the mechanistic study of biological processes.

In this study, we collected marginal tissue samples from patients with diabetic foot ulcers. By comparing the gene expression in the marginal tissues before and after TCTD, the differentially expressed genes in the tissues after TCTD were detected, and the associated signaling pathways were analyzed through bioinformatics analysis. This study helps us to better understand the molecular regulatory mechanism of TCTD in the healing process of diabetic foot ulcer.

Materials and methods

Patients

This study was approved by the IRB of the authors’ affiliated institutions. Patients and their families were informed and agreed to participate in the study, and signed informed consent. All methods were performed in accordance with the relevant guidelines and regulations. The participants in this study were patients with diabetic foot ulcers admitted to the authors’ affiliated institutions who underwent TCTD surgery combined with debridement and vacuum closed drainage (VSD).

The inclusion criteria we set include:

  1. 1)

    Patients diagnosed with diabetic foot ulcer and undergoing TCTD surgery;

  2. 2)

    No trauma or tumor around the ulcer;

  3. 3)

    Patients having a detailed medical history;

  4. 4)

    Long-term follow-up results;

  5. 5)

    Wound edge specimens were obtained when the patient underwent TCTD surgery combined with debridement, and wound edge specimens were also obtained at the last debridement suture.

Our exclusion criteria include:

(1) The patient had no detailed medical history;

(2) The patient did not cooperate with routine follow-up or medical treatment;

(3) Diabetic foot ulcer combined with other serious diseases that affect treatment;

(4) The patient failed to obtain wound edge specimens after TCTD.

Specimen acquisition

When patients underwent TCTD surgery combined with debridement and vacuum closed drainage (VSD), tissue specimens were collected from the wound edges and preserved according to standard procedures for tissue storage. In the later stage of TCTD surgery, the wounds of some patients needed debridement and suture surgery, and we also collected the tissue specimens trimmed from the wound edges of patients for storage.

Experimental process

The tissue samples underwent a series of continuous processes such as homogenization, phase separation, water separation, RNA precipitation washing, drying re-dissolution, concentration/purity determination, electrophoresis detection, and then further extraction of total RNA. The integrity of total RNA was detected by electrophoresis using RNA specific agarose.

After constructing the cDNA library, we amplified and enriched the library fragments by PCR. Prior to Next Generation Sequencing (NGS), the RNA needs to be extracted, purified, and constructed into a library, and then the second-generation gene sequencing is further completed in the Illumina gene sequencing analysis platform.

Transcriptome analysis process

After raw data was obtained, high-quality data needed to be selected, the species to be referenced as reference, and the data was compared to the reference genome for analysis and processing.

Data collation and filtering

The processed samples were loaded and sequenced, and the corresponding images could be obtained. Through the corresponding software transformation of sequencing, we can obtain the data off the machine. The obtained disembarkation data still need further statistical analysis.

Base mass and content distribution

The factors that usually affect the sequencing error rate include: the sample itself, the sequencer itself and the sequencing reagent.

The detection results of base content distribution were used to judge the separation phenomena of AT and GC. In RNA-Seq detection, according to the principle of base complementarity.

Comparative analysis

With HISAT2 software, we compared the reference genome after processing and filtering Reads values. When comparing Reads with HISAT2, the specified library needs to be selected if it is a chain-specific library, while the default parameters can usually be used directly if it is a non-chain-specific library.

Expression analysis

The Read Count value of each gene is the data obtained by statistical analysis and comparison using HTSeq software (Version: HISAT2 2.2.1, daehwankimlab.github.io, https://htseq.readthedocs.io/en/latest/), and the original expression of all genes in the experiment is also measured by Read Count.

FPKM can standardize gene expression in the analysis of experimental process data. After FPKM standardization, Researchers can compare the expression levels of different genes and the genes in different samples.

Statistical analysis

According to previous literature reports, we set the differential gene screening conditions, in which DESeq was used to analyze the differential expression of all genes. In this study, the screening conditions for differential genes were set as follows: (1) Expression multiple |log2FoldChange|> 1; (2) P < 0.05.

Cluster analysis

We can perform cluster analysis to determine the different expression patterns under various experimental conditions. The R language Pheatmap software package was used to perform cluster analysis on the differential expression genes and samples of each group. The algorithms we need to use in this study include Euclidean method and Complete Linkage method.

GO enrichment analysis

We performed this through topGO during the GO enrichment analysis. This study calculated the gene number and gene list corresponding to each term through topGO, and further annotated and analyzed the differential genes and GO term. The results of GO enrichment analysis of DEGs were presented in terms of cell component CC, biological process BP and molecular function MF.

GO directed acyclic graph

The results of analysis of cell components CC, biological processes BP and molecular functions MF are also presented with corresponding directed acyclic graphs. In directed acyclic graphs, the term near the root node is the general term, while the farther the branch is the more detailed term.

KEGG enrichment analysis

KEGG Enrichment Analysis (kegg.jp/kegg/kegg1.html) is often used to study enrichment signaling pathways of differential gene (including all upregulated and downregulated differential genes) sets and to screen key genes4,5,6. Generally, after obtaining the target gene set, it is necessary to analyze which pathways these genes participate in, what functions they have, and which pathways may play a leading role in the change of phenotype. The pathway enrichment analysis also involved foreground genes and background genes.

qRT-PCR

After RNA was extracted from tissue cells, reverse transcription studies were performed using the PrimeScript™ RT reagent Kit (Japan). In this study, four differentially expressed genes in postoperative tissue samples of TCTD were selected for qRT-PCR verification, so as to confirm their consistency with transcription sequencing results and further prove the reliability of NGS results.

Results

In this study, wound edge specimens were obtained during lateral tibial bone removal combined with debridement and wound edge specimens were also obtained during the last debridement suture of 3 patients. There were a total of six preoperative and postoperative specimens for TCTD. The patients were 2 men and 1 woman aged 52, 61 and 54 years. All three patients had been diagnosed with type 2 diabetes mellitus for more than a decade and had developed ulcers on their feet that had failed to heal well after up to three months or more of dressing changes and other treatments. After TCTD, these 3 patients had good healing of postoperative foot ulcers, which were typical examples of such cases.

RNA quantitative and quality results are shown in Table 1 (NanoDrop ND-1000). The results of this study showed that all samples were qualified.

Table 1 Quantification and quality of RNA identified by NanoDrop ND-1000.

The situation of RNA integrity and gDNA contamination detected by variable electrophoresis in this experiment is qualified.

After the hybridized chip is scanned by the scanner, the scan map is obtained (Fig. 1A). T (Treatment) was the treatment group after transverse tibial bone transfer surgery, and C (Control) was the control group before transverse tibial bone transfer surgery. The fluorescence signal is standardized scanned by the gene chip scanning instrument for detection and analysis, and the image of gene chip hybridization in this experiment can be obtained. Meanwhile, the fluorescence signal intensity of the probe under study can also be quantitatively reflected and processed with specific values (Fig. 1B).

Fig. 1
figure 1

Gene chip quality testing status. (A) The hybridized chip was scanned by the scanner and the scan map is obtained. T is the treatment group after TCTD, and C is the control group before TCTD surgery. (B) A boxplot of the standardized strength values for each of the two sets of samples.

The results of volcano map analysis (Fig. 2) showed that 2441 differential genes were significantly up-regulated and 3904 differential genes were significantly down-regulated in the lower extremity experimental group after TCTD surgery compared with the control group without TCTD surgery before surgery.

Fig. 2
figure 2

Differential gene and cluster analysis. (A) Volcanic map of differential genes. Red dots indicate up-regulated genes, green dots indicate down-regulated genes, and black dots indicate non-significant differentially expressed genes. (B) Cluster analysis results.

The top 20 upregulated genes with the most significant differences were HLA-DRB5, CXCL5, PTX3, HLA-DRB1, THRSP, ADH1A, SPP1, CXCL6, OSM, OR52E8, CD163, EBAG9, OPN1SW, CYP1B1, HLA-DRB5, CXCL5, PTX3, HLA-DRB1, THRSP, ADH1A, SPP1, CXCL6, OSM, OR52E8. GPR61, OR14J1, CASTOR3, ST8SIA5, SIX6, IL21. The specific information of the 20 genes with the most significant differences is shown in Table 2.

Table 2 The top 20 up-regulated genes with the most significant differences.

The top 20 down-regulated genes with the most significant differences were CATG00000038258.1, PAK6, IGFL1, PPP1R12A, LMF1, CASP14, CATG00000036350.1, ARHGAP30, UMOD, HAPLN2, KRT9, ZFP57, CNEP1R1, METTL14, MTRNR2L5, DUSP12, OR9K2, APLN, MKRN3, OR2M7. The specific information of the 20 genes with the most significant differences is shown in Table 3.

Table 3 The top 20 down-regulated genes with the most significant differences.

GO enrichment analysis of biological processes showed the concentration of regulation of cellular metabolic process was the most significant. The top 20 terms with the most significant enrichment are shown in Table 4. The pie chart, column chart and bubble chart analysis results of the top 10 terms with the most significant enrichment are shown in Fig. 3. The results of GO directed acyclic graph analysis for biological processes are shown in Fig. 4. It can be seen that the most significant and farthest Term is the cellular protein modification process.

Table 4 GO enrichment analysis results show the top 20 terms that are most significant in terms of biological processes.
Fig. 3
figure 3

GO enrichment analysis results show the top 10 Term analysis charts that are most significant in terms of biological processes. (A) The pie chart shows the biological process classifications. (B) The column chart of most significant in terms of biological processes. (C) The bubble chart analysis of most significant in terms of biological processes.

Fig. 4
figure 4

GO directed acyclic graph analysis results for biological processes.

The results of GO enrichment analysis in cell components showed that differential genes were most significantly enriched in cytoplasm, intracellular and cytosol terms. The top 20 terms with the most significant enrichment are shown in Table 5. The pie chart, column chart and bubble chart analysis results of the top 10 terms with the most significant enrichment are shown in Fig. 5. The analysis results of GO directed acyclic graph in biological processes are shown in Fig. 6. It can be seen that the most significant and farthest Term is nucleoplasm.

Table 5 GO enrichment analysis results show the top 20 terms that are most significant in terms of cell components.
Fig. 5
figure 5

GO enrichment analysis results show the top 10 Term analysis charts that are most significant in terms of cell components. (A) The pie chart shows the cell component classifications. (B) The column chart of most significant in terms of cell components. (C) The bubble chart analysis of most significant in terms of cell components.

Fig. 6
figure 6

GO directed acyclic graph analysis results in terms of cell components.

The results of GO enrichment analysis showed that in terms of molecular function, differential genes were most significantly enriched in protein binding, ion binding, and protein domain specific binding terms. The top 20 terms with the most significant enrichment are shown in Table 6. The pie chart, bar chart and bubble chart analysis results of the top 10 terms with the most significant enrichment are shown in Fig. 7. The analysis results of GO directed acyclic graph in terms of molecular function are shown in Fig. 8. It can be seen that the most significant Term at the farthest end is metal ion binding (GO:0046872).

Table 6 GO enrichment analysis results show the top 20 terms that are most significant in terms of molecular function.
Fig. 7
figure 7

GO enrichment analysis results show the top 10 Term analysis charts that are most significant in terms of molecular function. (A) The pie chart shows the cell molecular function. (B) The column chart of most significant in terms of molecular function. (C) The bubble chart analysis of most significant in terms of molecular function.

Fig. 8
figure 8

GO directed acyclic graph analysis results in terms of molecular function.

KEGG analysis showed that differential genes were most significantly enriched in Adherens junction, Inflammatory mediator regulation of TRP channels and Wnt signaling pathway. The histogram of the top 10 most significantly enriched signal pathway graphs and the results of the bubble analysis group are shown in Fig. 9. Table 7 provides detailed information about the top 20 most significantly enriched signal pathway maps.

Fig. 9
figure 9

Histogram and bubble analysis results of the top 10 signal pathway graphs with the most significant KEGG enrichment. (A) The column chart of the top 10 signal pathways. (B) The bubble chart analysis of most significant in terms of the topsignal pathways.

Table 7 Specific information of the top 20 signaling pathways with the most significant KEGG enrichment.

In the qRT-PCR experiment, we selected the significantly upregulated HLA-DRB1, HLA-DRB5, CXCL5 and significantly upregulated IGFL1 in the transcriptional sequencing results, with GAPDH as the internal reference. The qRT-PCR verification results showed that the expressions of HLA-DRB1, HLA-DRB5 and CXCL5 in the postoperative treatment group were significantly up-regulated, while the expression of IGFL1 was significantly down-regulated. These results are consistent with the results of transcriptome sequencing. The comparison results of the expression of verified genes in the two groups of samples are shown in Fig. 10.

Fig. 10
figure 10

The qRT-PCR results. The expression of HLA-DRB1, HLA-DRB5 and CXCL5 was significantly up-regulated in the postoperative treatment group compared with the preoperative control group, while the expression of IGFL1 was significantly down-regulated in the postoperative treatment group compared with the preoperative control group.

Discussion

TCTD technology is a new method for the treatment of diabetic foot, the molecular mechanism is not yet clear. The results of this study showed that 2441 differential genes were significantly up-regulated and 3904 differential genes were significantly down-regulated in the treatment group compared with the control group. Through RT-PCR verification of some differentially expressed genes, it was found that the results of gene expression profile chip detection were consistent, which further proved the reliability of gene expression profile chip data.

The top 20 up-regulated genes were mainly related to the immune system level and the nervous system. Among them, CXCL5 and CXCL6 belong to the CXC subfamily and are pleiotropic cytokines. In the context of angiogenesis, CXC chemokines are a unique family of cytokines known for their expression in different ways in the regulation of angiogenesis7,8. Members containing the “ELR” motifs are powerful promoters of angiogenesis and mediate their angiogenic activity by binding to and activating CXCR2 on endothelial cells. In contrast, in general, those members that can be induced by interferon and lack the ELR motif (ELR-) are potent inhibitors of angiogenesis and bind to CXCR3 on endothelial cells. In this study, the expression of CXCL5 and CXCL6 was significantly up-regulated in the patients after treatment, and clinical examination of CTA also observed that the vascular network of the lower limbs of the patients after surgical treatment had regeneration and remodeling, which was consistent with the up-regulated expression trend of CXCL5 and CXCL69,10.

The insulin growth factor-like family member (IGFL1), whose expression was verified to be downregulated by PCR, belongs to the IGF-like (IGFL) family11. IGF L1 is a multifunctional cellular regulator, most of which is synthesized in the liver and is mainly found in the blood circulation. IGFL1 mediates growth hormone stimulation, regulates tissue growth and development, and plays an important role in the regulation of muscle size and strength, maintenance of body composition, and nutrient metabolism11,12. When its secretion increases, it promotes the onset and development of inflammatory metabolism in the body. In the present study, it was found that its expression was significantly reduced after surgery, which may be one of the reasons for the reduced inflammatory response in the lower limbs of the patients and the tendency of the wounds to heal. However, the specific regulatory mechanism needs to be further verified.

In the results of GO analysis, differential genes were mainly involved in regulation of cellular process, cellular macromolecule metabolic process, The concentration of regulation of cellular metabolic process was the most significant. In terms of cell components, differential genes are most significantly enriched in cytoplasm, intracellular and cytosol terms. The results of GO directed acyclic graph analysis of biological processes show that nucleoplasm is the most abundant Term at the farthest end. In terms of molecular function, differential genes are most significantly enriched in terms of protein binding, ion binding, and protein domain specific binding. The results of GO directed acyclic graph analysis on molecular function showed that the most significant enrichment of the farthest Term was metal ion binding. Through GO enrichment analysis, we can roughly understand the biological functions, pathways or cell localization of differential gene enrichment.

KEGG analysis of differential genes showed that Differential genes were most significantly enriched in Adherens junction, Inflammatory mediator regulation of TRP channels, Wnt signaling pathway and other signaling pathways. In type 2 diabetes mellitus (T2DM), the sensory nervous system is involved in the progression of T2DM by maintaining low-grade inflammation through TRPV1. In studies of Trpv1-/- mice injected with small molecule Trpv1 inhibitors, oral glucose tolerance and glucose-stimulated insulin secretion were both improved. Therefore, TRPV1 has certain potential in the treatment of type 2 diabetes. TRP channels respond to various chemical and physical stimuli produced by harmful substances and lead to an increase in the concentration of cations within the cell13,14,15. The roles of immune cells mediated by TRP channels are varied, ranging from regulation of cell migration and phagocytosis to the production and release of inflammatory mediators. Wnt signaling pathway is a complex regulatory network, which mainly includes three branches: Wnt/Ca2 + pathway, Wnt/PCP pathway and Wnt/β-cantenin signaling pathway]. Generally speaking, the Wnt signaling pathway mainly refers to the classical Wnt signaling pathway mediated by β-catenin16,17. In previous studies, it was believed that Wnt signaling pathway was closely related to the differentiation of bone progenitor cells into osteoblasts and the division of osteoblasts into bone tissue. In this study, differential genes were significantly enriched in the Wnt signaling pathway, which may be due to the lateral bone transfer technique of the tibia, which artificially caused the tibia fracture18,19,20. During the recovery process, the activity of the Wnt signaling pathway was up-regulated to promote fracture healing, so the result was consistent with clinical observation.

VEGF signaling pathway plays an important role in the formation and growth of blood vessels21,22. In the process of neovascularization, VEGF induces gene expression, regulates vascular permeability and promotes cell migration, proliferation and survival23,24,25. Binding of VEGF ligands to their homologous membrane binding receptors induces VEGF signaling, thereby activating multiple downstream pathways. The VEGF signaling cascade includes PLC-γ pathway, which controls vascular permeability; PI3K/AKT pathway regulates cell survival; The FAK/paxillin pathway is involved in cytoskeletal rearrangement; Ras/MAPK pathway regulates cell proliferation and gene expression26,27. VEGF signaling is essential for normal vascular development and homeostasis. However, in pathological conditions, high levels of VEGF may induce pathological blood vessel formation through angiogenesis. VEGF has also been linked to neurodegeneration and may play a neuroprotective role.

Conclusions

To sum up, this study collected marginal samples of patients with diabetic foot ulcer whose wound surface was close to normal tissue. By comparing the gene expression in marginal tissues before and after TCTD surgery, differentially expressed genes in tissues after TCTD surgery were detected, and it was found that there were more differentially expressed genes in postoperative lower limb tissues of patients. Among them, CXCL5 and CXCL6 may participate in the process of postoperative neovascularization, and HLA-DRB1 may participate in the improvement of the degree of postoperative diabetic peripheral neurodegeneration. Through bioinformatics analysis, the corresponding signaling pathways of differential gene enrichment were identified, among which Inflammatory mediator regulation of TRP channels may participate in the improvement process of postoperative peripheral neurodegeneration. VEGF signaling pathway may be involved in postoperative neovascularization, and Wnt signaling pathway may be involved in postoperative fracture healing. The differentially expressed genes and related signaling pathways explored and analyzed in this study can provide an objective basis for further targeted intervention and further in-depth study of mechanisms.