Abstract
Fibroblasts are the essential cell type of skin, highly involved in the wound regeneration process. In this study, we sought to screen out the novel genes which act important roles in diabetic fibroblasts through bioinformatic methods. A total of 811 and 490 differentially expressed genes (DEGs) between diabetic and normal fibroblasts were screened out in GSE49566 and GSE78891, respectively. Furthermore, the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways involved in type 2 diabetes were retrieved from miRWalk. Consequently, the integrated bioinformatic analyses revealed the shared KEGG pathways between DEG-identified and diabetes-related pathways were functionally enriched in the MAPK signaling pathway, and the MAPKAPK3, HSPA2, TGFBR1, and p53 signaling pathways were involved. Finally, ETV4 and NPE2 were identified as the targeted transcript factors of MAPKAPK3, HSPA2, and TGFBR1. Our findings may throw novel sight in elucidating the molecular mechanisms of fibroblast pathologies in patients with diabetic wounds and targeting new factors to advance diabetic wound treatment in clinic.
1. Introduction
As the global population ages, the incidence of diabetes is rapidly increasing during recent decades [1]. Diabetic foot ulcers (DFUs) are one of the most common and serious complications of diabetes. It was reported that the incidence of DFUs was up to 4% in diabetes [2]. The mechanism of DFUs remains unclear, and many factors contributed to the delayed healing of it, throwing a significant burden on patients with diabetic wound [3]. Early diagnosis and intervention of diabetic wound are important for reversing the poor prognosis of DFUs [4]. Unfortunately, few distinctive diagnostic biomarkers have been reported and demonstrated in diabetic wound. Thus, it is of great necessity to screen out the novel diagnostic biomarkers involved in the development of diabetic wound.
Fibroblasts are the essential cell type of skin, highly involved in the wound regeneration process, and acted in wound healing by interacting with other cells including keratinocytes and endothelial cells [5]. Exosomal miR-20b-5p derived from the high-glucose impaired fibroblast proliferation and differentiation, and delayed diabetic wound healing, suggesting the crucial role of fibroblasts in diabetic wound healing [6]. Furthermore, accumulative evidences have demonstrated the important role of genetic and epigenetic regulation in diabetic wound healing [7, 8].
In this study, we sought to identify the DEG modulation in diabetic fibroblasts by using bioinformatic methods. These findings may provide useful insights into understanding the molecular mechanisms of fibroblast pathologies in patients with DFUs.
2. Materials and Methods
2.1. DEG Identification
Microarray data of datasets comparing diabetes and the healthy controls were screened out from the Gene Expression Omnibus database (GEO, http://www.ncbi.nlm.nih.gov/geo). DEGs were performed by Limma in R, and values < 0.05 were considered as statistically significant. R package pheatmap was used to visualize Log2 mRNA gene expression. Using Circlize and ComplexHeatmap in R, common DEGs from different datasets were identified and visualized. The circular visualization of chromosomal information of common DEGs was achieved with circular visualization in R.
2.2. GO and KEGG Analyses
DAVID, an online bioinformatics tool, was used to perform GO and KEGG analyses. The top ten GO terms in biological process, molecular function, and cellular component and top five KEGG pathways were identified using the enrichment analysis. The result of enrichment analysis of hub genes was visualized with GOplot. DEGs were imported into Search Tool for the Retrieval of Interacting Genes (STRING) to construct the PPI network. Then, the TSV file of PPI network was imported into Cytoscape 3.7.2. The interactions between enriched KEGG pathways were calculated and visualized by Cytoscape 3.7.2.
2.3. Retrieval of KEGG Pathways Involved in Type 2 Diabetes and Calculation of Shared Pathways between Enriched Pathways and Type 2 Diabetes
miRWalk is an online bioinformatics atlas tool. In this study, the KEGG pathways involved in type 2 diabetes were retrieved from miRWalk. Then, the intersection of enriched KEGG pathways () and type 2 diabetes-related KEGG pathways was obtained with Draw Venn Diagram (http://bioinformatics.psb.ugent.be/webtools/Venn/). The top shared KEGG pathway with the smallest value was selected. The enriched DEG-related part of the KEGG pathway was established with the PPT drawing tool.
2.4. Targeted Transcript Factor Prediction
http://amp.pharm.mssm.edu/Enrichr/, the online predicting tool, was used to predict targeted transcript factors of enriched DEGs in the shared KEGG pathway. The prediction result was visualized by Gephi.
3. Results
3.1. DEG Identification
Datasets of GSE49566 and GSE78891 were obtained from GEO, which are the genes from human skin fibroblasts (Figure 1). There were three type 2 diabetes samples and six normal in GSE49566. There were six type 2 diabetes samples and five normal in GSE78891. 446 upregulated and 365 downregulated DEGs were identified in GSE49566. 242 upregulated and 248 downregulated DEGs were identified in GSE78891. Totally, there were 34 common DEGs identified. They were STMN2, HAPLN1, PTN, POSTN, MAPKAPK3, CDH11, TLE1, ZFAND5, C9orf3, EMX2, TIPRL, MEIS1, FZD6, SLC6A8, SLC7A1, TGFBR1, EMP1, HSPA2, PLCB1, KISS1, HOXD4, EYA2, SERP1, UBL3, GTF2H1, MYO1E, LMAN1, BMP2, CTNNAL1, SDC1, GUCA1A, SUB1, ZC3H15, and MBP (Figure 2).
(a)
(b)
(c)
(d)
(e)
(f)
(a)
(b)
3.2. GO and KEGG Pathway Enrichment Analysis
GO analysis results showed that common DEGs were significantly enriched in skeletal system development, cell surface, protein binding, mesenchymal differentiation, pathway-restricted SMAD protein phosphorylation, cardiac epithelial to mesenchymal transition, heart development, in utero embryonic development, regulation of transcription, DNA-templated, and skeletal system morphogenesis (Table 1). KEGG pathway analysis showed that the common DEGs were significantly enriched in a pathway in cancer, signaling pathways regulating pluripotency of stem cells, Hippo signaling pathway, MAPK signaling pathways, and basal cell carcinoma pathway (Table 2). The information and interaction of the GO and KEGG terms are demonstrated in Figure 3.
(a)
(b)
3.3. Retrieval of KEGG Pathways Involved in Type 2 Diabetes and Calculation of Shared Pathways between Enriched Pathways and Type 2 Diabetes
The KEGG pathways linked with type 2 diabetes were obtained from miRWalk. They are listed in Table 3. Totally, there were 44 KEGG pathways involved in the development of type 2 diabetes. The common KEGG pathway between DEGs and type 2 diabetes with the highest value was the MAPK signaling pathway. The part of the MAPK signaling pathway related to the DEGs was established (Figure 4). MAPKAPK3, HSPA2, TGFBR1, and p53 signaling pathways were involved.
3.4. Targeted Transcript Factor Prediction
The targeted transcript factors of MAPKAPK3, HSPA2, and TGFBR1 were obtained from http://amp.pharm.mssm.edu/Enrichr/, which indicated ETV4 and NPE2 were the potential ones. The relationship of transcript factors, DEGs, and other targeting genes is shown in Figure 5.
(a)
(b)
4. Discussion
High risk of wound infection and healing failure was found in diabetes, and the abnormal function of fibroblasts was assumed as a major issue contributing to the delayed wound healing [9–11]. Noticeably, fibroblasts exert an important role in wound inflammatory response by release of various antibacterial regulators, providing a robust defense of skin against infections [12–14]. Diabetes patients are susceptible to infections due to the dysregulated function of the T cells, leading to the overactivated tissue inflammation. In this bioinformatic research, functional enrichment analysis was performed, and the systematic results suggested that the highest value was the MAPK signaling pathway among DEGs in fibroblasts. And the regulatory roles for diabetic wound healing were identified in MAPKAPK3, HSPA2, and TGFBR1.
Phosphorylation of transcription is one of the modifications of MAPK-dependent regulation in cellular responses [15]. Three subfamilies were found in the MAPK signaling pathway, including the extracellular-signal-regulated kinases (ERK MAPK, Ras/Raf1/MEK/ERK), the c-Jun N-terminal or stress-activated protein kinases (JNK, SAPK), and p38 [16–18]. Once the pathway was activated, a number of downstream target kinases including MAPKAPK3 could be activated [19]. Recently, some researchers have fabricated an in situ injectable hydrogel which can markedly accelerate diabetic wound healing through activating the TGF-β/MEK/MAPK signaling pathway [20]. Similarly, Qian et al. demonstrated that protein tyrosine phosphatase 1B was capable to enhance fibroblast proliferation and mitigation via activation of the MAPK/ERK pathway, thereby promoting diabetic wound healing [21]. In the current study, we found a consistent result that the MAPK signaling pathway plays a key role in the regulation of diabetic wound healing, and MAPKAPK3, HSPA2, and TGFBR1 are the potentially critical genes in this regulation process. Moreover, to uncover the potential targeted transcript factors of MAPKAPK3, HSPA2, and TGFBR1 genes, we used the online software (Enrichr, http://amp.pharm.mssm.edu/Enrichr/) and the results suggested that ETV4 and NPE2 were the potential transcript factors for these genes. Thus, it was assumed that ETV4 and NPE2 may exert a critical role in the regulation of diabetic wound healing.
Some limitations also existed in this bioinformatic research. First, the current results were based on a public database and only two datasets were included in our study; the sample size should be enlarged to minimize the possible confounding factors. Furthermore, this is a pure bioinformatic research; more experimental validation is needed to confirm the candidate pathways and their potential transcript factors. Moreover, clinical specimens of different degrees of DFUs should be collected to validate our current findings.
5. Conclusions
Our findings suggested a functionally enriched MAPK signaling pathway, with a focus on the potential role of ETV4 and NPE2 in the regulation of diabetic wound regeneration. The current study may provide novel therapeutic targets in diabetic wound treatment.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that there is no conflict of interest regarding the publication of this paper.