关键词: 慢性肾病, 肾间质纤维化, 肾功能, 间充质转化, 铁死亡, 自噬, 自噬依赖性铁死亡, GEO数据库, 生物信息学

Abstract: BACKGROUND: Autophagy and ferroptosis play important roles in the development of chronic kidney disease, but the molecular mechanisms and gene targets related to autophagy and ferroptosis in renal tissue of chronic kidney disease are still unclear. 
OBJECTIVE: To screen differentially expressed genes in chronic kidney disease-related datasets based on bioinformatics, and to explore potential key biomarkers suitable for screening renal function progression in patients with chronic kidney disease. 
METHODS: (1) The GSE137570 dataset was obtained from GEO database to screen the differentially expressed genes by Networkanalyst database analysis. Ferroptosis and autophagy related targets were obtained by OMIM, GENECARD, FerrDb and HAMdb databases. The respective data were intersected to obtain autophagy-ferroptosis related differentially expressed genes in chronic kidney disease for parallel enrichment analysis. The STRING website was used to construct the protein-protein interaction network of differentially expressed genes, which was imported into Cytoscape software and analyzed by MCODE and Cytohubba plug-in to screen potential core targets. Enrichment analysis was performed to obtain the functions of these potential core targets. (2) In the in vitro experiment, mouse renal tubular epithelial cells were divided into two groups: the control group received no intervention, while the model group was stimulated with 5 ng/mL transforming growth factor β1 for 24 hours to induce mesenchymal transformation of renal tubular epithelial cells. Flow cytometry was used to measure the levels of reactive oxygen species and changes in mitochondrial membrane potential in the cells. RT-PCR was employed to assess ferroptosis, autophagy-related markers, and the mRNA expression of potential core targets in the cells. 
RESULTS AND CONCLUSION: After screening the GSE137570 dataset, a total of 480 differentially expressed genes were obtained, including 104 upregulated genes and 376 downregulated genes (log2| (FC) | > 1, P < 0.05). There were 562 ferroptosis-related targets and 1 266 autophagy-related targets obtained from the OMIM, GENECARD, FerrDb, and HAMdb databases. Intersection of differentially expressed genes with ferroptosis- and autophagy-related targets yielded 15 ferroptosis-related targets and 18 autophagy-related targets, respectively. The enrichment analysis results indicate that ferroptosis-related differentially expressed genes are primarily involved in biological processes such as sulfur amino acid metabolism, neutrophil degranulation, and ferroptosis signaling pathways. Autophagy-related differentially expressed genes are mainly enriched in biological processes such as platelet degranulation, extracellular matrix degradation, and receptor tyrosine kinase signaling. After screened by MCODE and CytoHubba, key genes were identified in the protein-protein interaction network, including CD44, ALB, TIMP1, PLG, CCL2, and DPP4. Immune infiltration analysis results indicate that immune cells such as B cells, CD4+ T cells, NK cells, and monocytes show significant differential expression in renal tissue after chronic kidney disease, and the core targets are also significantly correlated with these immune cells (P < 0.05). The results of receiver operator characteristic curve analysis further demonstrate that the pathological progression of chronic kidney disease can be effectively diagnosed by CD44, ALB, TIMP1, PLG, CCL2, and DPP4. Single-cell sequencing results show that, except for PLG, the expression of target genes in the renal tissue of mice in each model group is generally consistent with the results of this experiment. RT-PCR results demonstrate that, for the validation of autophagy and ferroptosis phenotypes, compared with the control group, the model group shows a significant decrease in mRNA expression of LC3B, Nrf2, and SLC7A11 (P < 0.05), and a significant increase in P62 mRNA expression (P < 0.05). Regarding the validation of potential core targets, compared with the control group, the model group exhibits a significant decrease in mRNA expression of ALB and PLG (P < 0.05), and a significant increase in TIMP1 and CCL2 mRNA expression (P < 0.05). Overall, these findings indicate that, through bioinformatics analysis and experimental validation, CD44, ALB, TIMP1, PLG, and CCL2 are abnormally expressed in the renal tissue of patients with chronic kidney disease, closely correlated with estimated glomerular filtration rate and tubulointerstitial fibrosis, and maybe play a predictive role in the progression of chronic kidney disease. 

Key words: chronic kidney disease, tubulointerstitial fibrosis, renal function, interstitial transformation, ferroptosis, autophagy, autophagy-dependent ferroptosis, GEO database, bioinformatics