Abstract: Objective To explore the hub genes and the potential targets in the treatment of diabetes with TREM2 mutation-related cognitive dysfunction with bioinformatics analysis. Methods The cases of differential genes (DEGs) of diabetes mellitus and TREM2 mutation-related cognitive dysfunction were obtained respectively by microarray data analysis, the common differential genes were obtained by intersection between the two diseases. GO analysis, KEGG and Reactome pathway analysis were performed on the selected differential genes. The protein-protein interaction(PPI) network was constructed using online database. Finally, the effects of diabetes and TREM2 on spatial learning and memory of mice were detected by water maze, and the expression of hub gene SNAP25 was detected by Western blot. Results In both datasets, 19 genes showed similar changes, mainly enriched in biological processes and pathways related to neurons and metabolism. According to PPI analysis, DNER, GFAP, GRM5 and SNAP25 were identified as hub genes. Trem2 gene knockout aggravated spatial learning and memory impairment in diabetic mice. The expression of SNAP25 in hippocampus of diabetes mice was significantly increased, and then decreased after Trem2 knockout. Conclusions This study identified 19 TREM2-related genes in diabetes with cognitive dysfunction, among which 4 hub genes were found. These results provide a new experimental basis for the treatment of diabetes patients with cognitive impairment.

Key words: diabetes mellitus, cognitive dysfunction, triggering receptor expressed on myeloid cells-2(TREM2), bioinformatics