Tideglusib is a non-competitive GSK-3β inhibitor which contain 1,2,4-thiadiazolidine-3,5-dione moiety, and now mainly used for progressive supranuclear palsy due to the lack of some primary cognitive endpoints and secondary endpoints in a phase IIb trail for Alzheimer's disease. Additionally, insufficient evidence exists to support that there are obvious covalent bonds between Tideglusib and GSK-3β. Targeted covalent inhibition strategy could improve the binding efficiency, selectivity and duration of kinase inhibitors. Based on the above premise, two series of targeted compounds with acryloyl warheads were designed and synthesized. The kinase inhibitory activity of the selected compound 10a with better neuroprotective effect improved 2.7 fold than that of Tideglusib. After the preliminary screening of GSK-3β inhibition and neuroprotective activity, the mechanism action of the selected compound 10a was investigated in vitro and in vivo. The results confirmed that 10a with excellent selectivity among the whole tested kinases could significantly reduce the expressions of APP and p-Tau via increasing the level of p-GSK-3β. The pharmacodynamic assay in vivo showed that 10a could markedly improve the learning and memory functions in AD mice induced by AlCl3 combined with d-galactose. At the same time, the damage of hippocampal neurons in AD mice was obviously reduced. Accordingly, the introduction of acryloyl warheads could increase the GSK-3β inhibitory activity of 1,2,4-thiadiazolidine-3,5-dione derivatives, and the selected compound 10a deserves further research as an effective GSK-3β inhibitor for the potential treatment of AD.