Discovery of KDM5D as a novel biomarker for traumatic brain injury identified through bioinformatics analysis

Abstract

Background and aim: Traumatic brain injury (TBI) poses a significant burden on the global economy due to its poor treatment and prognosis. Current TBI markers do not comprehensively reflect the disease status. Therefore, identifying more meaningful biomarkers is beneficial for improving the prognosis and clinical treatment of TBI patients.

Methods: The gene expression profile of TBI was obtained from the Gene Expression Omnibus (GEO) database. Differentially expressed genes (DEGs) were subjected to enrichment analysis, and key potential genes were identified through the protein-protein interaction network and cytoHubba modules. ROC curves were used to construct diagnostic models for hub genes. Immunofluorescence experiments were conducted to detect the expression of candidate biomarkers in TBI rat models. Finally, we investigated the expression of TBI biomarkers in normal human organs and pan-cancer tumor tissues, and evaluated their correlation with immune infiltration in different tumors.

Results: A total of 44 DEGs were identified across four brain regions of TBI patients. Enrichment analysis revealed that these genes were primarily involved in intracellular and cell signal transduction pathways. Furthermore, three hub genes- RPS4Y1, KDM5D and NLGN4Y-were identified through different module analysis. The ROC curve diagnostic model also confirmed that these genes also have high diagnostic value in serum. Subsequently, the presence of Kdm5d was detected in the brain tissue of TBI rats through immunofluorescence experiments. Compared to normal rats, Kdm5d expression increased in the cortical area of ​​TBI rats, with no significant change in the hippocampus area, aligning with observations in TBI patients. Immune infiltration analysis demonstrated changes in immune cell subsets in HIP and PCx, revealing that plasma cells and CD8 T cells were lowly expressed in TBI (HIP) and while neutrophils was under-expressed in TBI (PCx). Pan-cancer analysis indicated that KDM5D was significantly up-regulated in 23 cancers, down-regulated in 3 cancers, and significantly associated with immune infiltration in 10 cancers.

Conclusion: Based on the results of bioinformatics analysis and animal experiments, KDM5D serves as a potential biomarker for the diagnosis and prognosis of TBI. Additionally, research on KDM5D may develop into new serum markers, providing new indicators for further clinical liquid biopsy and aiding in the prevention of both TBI and tumors to a certain extent.

Keywords: KDM5D; bioinformatics analysis; biomarker; neuroinflammation; traumatic brain injury.

Figure 1

Research design flow chart.

Figure 2

Identification of DEGs across four brain areas in TBI and non-TBI. (a–d) Cluster heat map displaying all DEGs in four brain regions between TBI and non-TBI. (e–h) Volcano map highlighting DEGs between TBI and non-TBI, with upregulated genes marked in light red and downregulated in light blue. The threshold was set to |log2FC (fold change)> 0, and p value < 0.05. (i–l) Bar charts illustrating DEGs in four brain areas of TBI and non-TBI.

Figure 3

Functional enrichment analysis of DEGs. (a, c, e) GO analysis of DEGs in FWM, HIP and TCx, respectively. (b, d, f) KEGG pathway analysis of DEGs in FWM, HIP and TCx, respectively. P < 0.05.

Figure 4

PPI network analysis and identification of hub genes. (a) Venn diagrams showing the distribution and overlap of DEGs across four brain regions. (b) PPI network highlighting communal hub genes. (c–h) Top scoring genes in various PPI network analysis.

Figure 5

Diagnosis model construction of hub genes. ROC curve of diagnostic model based on three hub genes in GSE254880. (a) RPS4Y1 diagnostic model, AUC=0.81. (b) KDM5D diagnostic model, AUC=0.90. (c) NLGN4Y Diagnosis model of, AUC=0.86.

Figure 6

KDM5D expression in TBI patients and TBI rat. (a–d) Expression of KDM5D in four brain regions of TBI and non-TBI (GSE104687). (e, f) Expression of Kdm5d in cortical areas of TBI rats (p=0.0383). (g, h) Expression of Kdm5d in hippocampus areas of TBI rats (p=0.2329). Blue fluorescence indicates the nucleus, red fluorescence indicates Kdm5d and pink indicates the merge diagram. Scale bars: 50μm.

Figure 7

Analysis of immune cell infiltration. (a) Color bar plot illustrating the distribution of 22 immune cell types across various HIP samples. (b) Boxplot showing the expression profiles of two immune cell significant dysregulated in TBI compared to non-TBI. (c) Color bar plot illustrating the distribution of 22 immune cell types across PCx samples. (d) Boxplot showing the expression profile of one immune cell type significant dysregulated in TBI compared to non-TBI. *P<0.05, **P<0.01.

Figure 8

Analysis of KDM5D in pan-cancer. (a) Differences expression of KDM5D between tumors and normal tissues in the Cancer Genome Atlas (TCGA). *P<0.05, ** P<0.01 and *** P<0.001. (b) Immune correlation analysis of KDM5D across various immune infiltrates in human tumors. Positive values ​​indicate positive correlation, and negative values ​​indicate negative correlation.