Screening of key genes associated with m6A methylation in diabetic nephropathy patients by CIBERSORT and weighted gene coexpression network analysis

Abstract

Diabetic nephropathy (DN) is a common complication of diabetes. Due to its complex pathogenesis, there is no effective treatment. M6A is a newly discovered epigenetic mechanism that may be involved in the development of diabetic nephropathy. In this study, we analyzed differentially expressed genes (DEG) in the GEO database (GSE96804) and paid attention to genes with m6A methylation. 623 DEGs in glomerular tissue were identified by comparing diabetic nephropathy with normal. Correlation analysis with 21 genes involved in m6A modification showed that 492 genes were associated with m6A methylation. According to the CIBERSORT algorithm, the infiltration of M1 macrophages in DN patients was significantly higher than that in normal samples. Weighted gene coexpression network analysis (WGCNA) was used to screen for the modules most correlated with the clinical features of M1 macrophages. The genes in the selected modules and 492 m6A-related DEGs were intersected by a Venn diagram, and 43 key genes were obtained. GO and KEGG analyses showed that these genes were mainly related to the positive regulation of protein aggregation and the transforming growth factor β receptor signaling pathway. According to a literature review, among the top 10 genes, HSPA1A, HSPA1B, CHI3L1, TYRO3 and PTH1R are markers in diabetic nephropathy, and their abnormal expression is associated with renal hypertrophy, proteinuria and glomerulosclerosis. These findings may provide evidence for the diagnosis and treatment of diabetic nephropathy.

Keywords: CIBERSORT; Diabetes; diabetic nephropathy; differential analysis; enrichment analysis; m6A; weighted gene co-expression network.

Figure 1

Screening of differentially expressed genes in diabetic nephropathy patients and healthy controls in the GSE96804 dataset; DEG analysis Volcano map of differentially expressed genes (red dots represent up-regulated genes, green dots represent down-regulated genes).

Figure 2

Screening of DEGs associated with m6A. The orange dots represent the m6A-related genes, the green dots represent the DEGs, and the lines represent the correlations between the dots.

Figure 3

Boxplot of the proportion of immune cells in diabetic nephropathy patients and normal samples. The abscissa is the type of immune cells, and the ordinate is the proportion of immune cells. Green represents the proportion of immune cells in diabetic nephropathy samples, red represents the proportion of immune cells in normal samples, 0.01

Figure 4

Construction of the co-expression network. A: Sample clustering diagram (delete 2 outlier samples by setting the height to 94); B: Determination of the optimal soft threshold (in the process of module selection, the adjacency matrix is converted into a topology matrix, and the optimal soft threshold β=3 is determined); C: Cluster tree of coexpressed gene modules (similar genes are grouped into the same module through dynamic splicing and cluster analysis).

Figure 5

Immune cell module and gene screening in diabetic nephropathy patients. A: Correlation between module genes and immune cells (the redder the color, the higher the correlation; Pearson correlation coefficient between module characteristic genes and sample characteristic vectors, and the number in brackets represents the corresponding p value); B: Scatter plot of GS and MM of red module genes of M1 macrophages.

Figure 6

Venn diagram. A: The intersection of m6A-related DEGs with the genes of the red modules; B: Functional enrichment analysis of intersecting genes in M1 macrophages.