Identification of Hub Genes in Diabetic Nephropathy by an Integrated Bioinformatic Analysis


 BackgroundDiabetic nephropathy (DN) is a progressive kidney disease caused by damage to the capillaries in the kidneys' glomeruli. The dysregulation of genes plays a significant role in the progression of DN. MethodsIn the present study, gene expression profiles were analyzed to identify the key genes and pathways involved in DN. Gene expression profiles of 9 patients with DN and 11 normal subjects were downloaded from the Gene Expression Omnibus (GEO) database. Subsequently, differentially expressed genes (DEGs) were identified and subjected to functional enrichment analysis. In addition, protein-protein interaction (PPI) networks were established and hub genes were identified. ResultsAs a result, a total of 424 DEGs were identified (113 were upregulated and 311 were downregulated). In the protein-protein interaction network, four hub modules and 30 hub genes were identified. To explore potential associations between gene and DN clinical features and to identify hub genes, the top 25% of genes with the greatest variance in the gene expression profiles were extracted for weighted correlation network analysis (WGCNA). There were ten genes (RNASE6, CD1C, SASH3, COL1A2, MS4A6A, CD163, CLEC10A, MOXD1, IQGAP2, GHR) identified as significant DN‐associated genes. Furthermore, the expression level of these hub genes was confirmed in the GSE96804 dataset. ConclusionsThese findings provide new insight into DN pathogenesis, which may enhance our fundamental knowledge of the molecular mechanisms underlying this disease.


31
The prevalence of diabetes and its complications poses a major threat to global health, 32 has contributed tremendously to the burden of mortality and disability (1). Acute 33 metabolic complications of diabetes associated with mortality include hyperglycemia 34 and coma due to hypoglycemia(2). While the most devastating consequence of diabetes 35 is its long-term vascular complications(3, 4). These complications are wide-ranging and 36 result, at least in part, from vascular damage caused by chronically elevated blood 37 glucose levels(5). Diabetic microvascular complications (nephropathy, retinopathy, and 38 neuropathy), which are long-term complications that affect small blood vessels, usually 39 affect those with a chronic or uncontrollable disease, but they can also be observed in 40 those who have been diagnosed or have not yet made a diagnosis of diabetes(6). 41 Diabetic nephropathy (DN) is one of the most fatal long-term complications of diabetes 42 and a leading cause of chronic kidney disease (7). The main signatures of DN usually 43 include glomerular scarring, proteinuria, a progressive decline in renal function, and 44 3 even end-stage renal disease (ESRD), which are attributed to tubular interstitial fibrosis, 45 hypertrophy, and dilatation of the glomerular mesentery, thickening of the glomerular 46 basement membrane, loss of foot cell peduncles, and inflammation due to monocyte 47 and macrophage infiltration (7, 8). The pathophysiology of DN is complex, involving 48 interactions between genetic factors, epigenetic factors, and the environment. 49 Diabetogenic stimuli, such as high blood glucose levels; advanced glycation end 50 products (AGEs); growth factors including transforming growth factor β1 (TGF-β1) 51 and platelet-derived growth factor and inflammatory cytokines, which have been 52 implicated in the pathogenesis of DN due to their detrimental effects on multiple renal 53 cell types(9-11). Although the pathophysiology of DN is continually being elucidated, 54 the underlying molecular mechanisms of DN progression are not fully understood.

55
Advances in histological techniques and the integration of high-dimensional data 56 through systems medicine approaches can provide the molecular mechanism of action 57 for drugs and disease progression pathways (12). Genomic data related to various 58 diseases are stored in public repositories, which can be easily accessed to obtain 59 meaningful information and make novel discoveries (13). Transcriptomic analysis 60 during the development of DN may be of great potential value for timely diagnosis and 61 timely treatment to prevent progression to end-stage renal disease.

65
These DEGs were adopted into the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses. The protein-protein interaction (PPI) network 67 was further constructed to understand cellular mechanisms and interactions between 68 cell's molecular constituents of selected genes. In addition, the top 25% of genes with 69 the greatest variance in the dataset were extracted to perform weighted gene co-

106
The R package Weighted correlation network analysis "(WGCNA)" is a comprehensive 107 collection of R functions for performing various aspects of weighted correlation 108 network analysis. In this study, the top 25% of genes with the greatest variance in the 109 dataset GSE30528 were extracted as the input data for subsequent WGCNA. The

119
Significant differences were measured by one-way ANOVA for two groups of data 120 followed by a Tukey's posthoc comparison. P value<0.05 was considered statistically 121 significant.

124
The workflow for identification, validation, function enrichment, and pathway analysis 125 of DEGs was shown in Figure 1A. The DEGs screening criteria were set in the limma 126 package, and a total of 113 up-regulated and 311 down-regulated significant DEGs were 127 identified from the genes of glomeruli samples with DN patients and normal subjects.

128
The volcano plot displayed the distribution of DEGs between DN and normal glomeruli 129 samples ( Figure 1B). Compared to normal subjects, the gene C1QA was the most 130 significant up-regulated gene (P value=2.66E-08, adjusted P value= 7.11E-06), 131 followed by SERPINE2 (P value=3.10E-07, adjusted P value=3.53E-05) in DN samples.   Figure 5A) and with a relatively high-average connectivity ( Figure 5B). Then, we 177 constructed the gene network and identified modules using a one-step network 178 construction function. Finally, we eventually identified 11 gene co-expression modules 179 ( Figure 6A).

180
Analysis of gene co-expression modules 181 The relationship between identified modules was mapped ( Figure 7A). Subsequently,   (Table 1). Then, gene expression profiles from the GSE96804 dataset 202 were used to validate the expression of these key genes in diabetic nephropathy. As 203 shown in Figure 9A