Identification and Validation of Tryptophan Metabolism-Related Genes in Diabetic Kidney Disease and Construction of a Clinical Prediction Model.

Abstract

Background: Diabetic kidney disease (DKD) is a common microvascular complication of diabetes mellitus (DM). Amino acid (AA) homeostasis has an important impact on renal hemodynamics and glomerular hyperfiltration in patients with DKD, and the metabolite level of tryptophan (TRP), an AA, has been associated with various diseases. Methods: In this study, DKD tubule- and glomerulus-related microarray datasets were collected from the GEO database, and DKD-related modular genes were identified by weighted gene coexpression network analysis (WGCNA). TRP metabolism-related genes (TRGs) were downloaded from the MSigDB database, and the key genes were obtained by taking the intersection of DKD differentially expressed genes, TRGs, and modular genes. Validated with the Nephrseq v5 database and performed clinical prediction model construction. The association of pivotal genes with immune cell infiltration was verified using CIBERSORTx software. The protein expression of the key genes was verified by qPCR, Western blot, immunohistochemistry, and immunofluorescence. Results: Four hundred and seventy seven DEGs were identified in the GSE30529 dataset, 392 DEGs were identified in the GSE30528 dataset, and the intersection of the DEGs in the two datasets, the module with the most significant correlation with DKD obtained by WGCNA, and the TRGs were taken, respectively. Five key genes were finally obtained (AOC1, HAAO, STAT1, OGDHL, and TDO2). Compared with control-group mice, the expression of AOC1, HAAO, and OGDHL was significantly downregulated, and the expression of STAT1 and TDO2 was significantly elevated in DKD mice. The diagnostic model was constructed using the key genes AUC = 0.996. Conclusion: Our study suggests that the AOC1, HAAO, and STAT1 genes may be potential diagnostic biomarkers of tubular injury in DKD. OGDHL and TDO2 may be potential diagnostic biomarkers of glomerular injury in DKD. The model constructed using AOC1, HAAO, STAT1, OGDHL, and TDO2 had good disease differentiation.