ABSTRACT: Purpose:A long-term "memory" of hyperglycemic stress, even when glycemia is normalized, has been previously reported in endothelial cells. However, the molecular mechanism of "metabolic memory" (MM) remains unknown. In this report, we sought to screen at the whole transcriptome level the genes that participate in MM. Methods:In the present research, RNA sequencing was used to determine the protein-coding mRNA expression profiles of human umbilical vein endothelial cells (HUVECs) under normal-glucose concentration (LG), high-glucose concentration (HG), and MM. A series of bioinformatic analyses was performed. HG-induced MM-involved up-regulated genes (up-HGMMGs) and HG-induced MM-involved down-regulated genes (down-HGMMGs) were identified. Afterward, based on up-HGMMGs and down-HGMMGs, the biological functions and signaling pathways were analyzed using Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG). In addition, several of the identified genes were validated by RT-qPCR. Results:A total of 726 HGMMGs were identified, including 210 down- and 516 up-HGMMGs, which were enriched in the cell cycle (hsa04110), oocyte meiosis (hsa04114), p53 signaling pathway (hsa04115), and oxidative phosphorylation (hsa00190), among others. The protein-protein-interaction (PPI) network consisted of 462 nodes and 2656 connections, and four main modules were identified by MCODE. The cell cycle (hsa04110), oocyte meiosis (hsa04114), p53 signaling pathway (hsa04115), and oxidative phosphorylation (hsa00190), among others, could be potential therapeutic targets of HG-induced MM in endothelial cells. The real-time PCR results validated the RNA-seq data. Conclusion:This study identified crucial mRNAs related to MM-persistent injury in endothelial cells even after switching the cells from high- glucose to normal glucose levels. Further research focusing on these mRNA may unravel new ways to modify MM in diabetes.