Objective To construct an in vitro "metabolic memory" cell model of HT-22 mouse hippocampal neurons induced by high glucose, and to investigate the effect of "metabolic memory" on apoptosis and histone acetylation in HT-22 cells. Methods HT-22 cells were cultured in high glucose medium (glucose concentration was 55 mmol/L) and conventional glucose medium (glucose concentration was 25 mmol/L), and cells were divided into the control group (NG 4, 6 and 8 groups, 25 mmol/L glucose was cultured for 4, 6 and 8 days, respectively), the high glucose group (HG 4, 6 and 8 groups, respectively) and the metabolic memory group (HG2NG2, HG2NG4, HG2NG6, HG4NG2 and HG4NG4 groups, high glucose culture for 2 days to 25 mmol/L glucose culture for 2, 4 or 6 days, high glucose culture for 4 days to 25 mmol/L glucose culture for 2 or 4 days). Cell viability was detected by CCK-8 method. The release of lactate dehydrogenase (LDH) in cell culture supernatant was detected, and the optimal time to establish a "metabolic memory" model was selected. Subsequently, cells were divided into the NG4 group, the NG8 group, the HG4 group, the HG4NG4 group and the HG8 group, and the cell morphology of each group was observed by optical microscope. The apoptosis rate was detected by flow cytometry. The activities of deacetylase (HDAC) and histone acetyltransferase (HAT) were detected by enzyme-linked immunosorbent assay (ELISA). Western blot assay was used to detect expression levels of histone deacetylase 4 (HDAC4), B lymphocyte tumor 2 (Bcl-2), Bcl-2 related X protein (Bax) and Caspase-3 protein. Results The HG4NG4 group was the ideal cell model with high glucose metabolic memory. Cells of the NG4 group and the NG8 group were interwoven into a dense network, growing well, with spindle shaped cells and distinct synaptic structures. However, in the HG4 group and the HG8 group, the cell body became round, synaptic structure disappeared and growth was inhibited. In the HG4NG4 group, the number of cells increased but their morphology was damaged. Results of flow cytometry showed that compared with the NG8 group, the apoptosis rates were significantly increased in the HG8 group and the HG4NG4 group (P<0.05). ELISA results showed that compared with the NG8 group, the expression levels of HDAC4, Bax, and Caspase-3 proteins increased in the HG8 group and the HG4NG4 group, while the expression level of Bcl-2 protein significantly decreased (P<0.05). Compared with the HG8 group, there were no significant differences in protein expression levels of HAT and HDAC in the HG4NG4 group. Western blot reslts showed that compared with the NG8 group, the levels of HDAC4, Bax and Caspase-3 protein increased in the HG8 group and the HG4NG4 group (P<0.05). Compared with the HG8 group, there were no significant differences in protein expression levels in the HG4NG4 group. Conclusion HT-22 mouse hippocampal neurons cultured with 55mmol/L high glucose for 4 days, and then cultured with 25 mmol/L glucose for 4 days are the ideal "metabolic memory" cell model. The mechanism may be related to the increased activity of HDAC, HAT and HDAC4 expression in the hyperglycemic model.
