Abstract: Objective: To investigate the effect and mechanism of treadmill exercise on cognitive deficits induced by 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)/Probenecid in chronic Parkinson’s diseases （PD） mice. Methods: SPF male C57BL/6 mice were randomly divided into 6 groups, i.e., Control, PD, PD+Exe4, PD+Exe8, PD+Exe8+DMXAA, and PD+3MA+Exe8 group. The PD model was constructed by intraperitoneal injection of MPTP and Probenecid in mice except Control group. The rota-rod test, pole test, and suspension test were employed to assess the motor abilities of mice, and the reliability of the PD model was comprehensively evaluated by using the content of striatal dopamine and its metabolite, as well as the expression levels of nigral and striatal tyrosine hydroxylase (TH). The mice in PD+Exe4 and PD+Exe8 groups were conducted 4-week or 8-week exercise interventions, respectively; the water maze experiment was used to test the cognitive function of mice at the end of week 4 and 8. Immunofluorescence was used to evaluate the hippocampal co-localization of Caspase1 and LC3II with the microglia marker protein Iba-1. The expression of α-syn in the hippocampus was detected by immunohistochemical staining. Nissl staining was employed to ascertain the number and extent of neuronal loss in the hippocampus. Golgi staining was utilized to observe the development of dendritic spines in hippocampal neurons. ELISA was used to measure the levels of serum IL-1β and IL-18. Western blot was employed to detect the expression of p-α-syn/α-syn, synaptic plasticity-related proteins (PSD93, PSD95, and SYN), neurotrophic factor-related proteins (p-CREB, CREB, and BDNF), Iba-1, pyroptosis-related proteins (NLRP3, cleaved Caspase1/Caspase1, GSDMD-N, ASC, IL-1β, and IL-18), STING signal proteins (p-STING/STING and p-TBK1/TBK1), and autophagy proteins (Beclin1, p62, and LC3II/I) in the hippocampus tissues. Results: 1) Compared with control group, MPTP/p-induced chronic PD mice showed a significant reduction in motor coordination, and a decrease in the number of TH-positive cells in the striatum and substantia nigra, accompanied by the accumulation of pathological protein α-syn. 2) Treadmill exercise significantly suppressed p-α-syn/α-syn expression levels and neuronal loss of hippocampus in PD mice, but the inhibitory effect of treadmill exercise on α-syn was reversed by treatment with STING agonist DMXAA and autophagy inhibitor 3MA. 3) Treadmill exercise increased axon length and number of dendritic junctions in hippocampal neurons of PD mice, up-regulated the expression levels of synaptic plasticity-related proteins of PSD93, PSD95, SYN and the neurotrophic factors of p-CREB/CREB and BDNF, and improved the synaptic plasticity of the hippocampus. DMXAA and 3MA treatment blunted the up-regulation of synaptic plasticity-associated proteins by treadmill exercise. 4) Treadmill exercise up-regulated the proteins level of Beclin1, LC3 Ⅱ/Ⅰ, LAPM2, and down-regulated p62 to promote the fusion of autophagosomes and lysosomes and autophagy levels in hippocampus of chronic PD mice. 3MA treatment reversed the beneficial effect of exercise intervention. 5) Treadmill exercise inhibited the pyroptosis levels in hippocampal MGs of PD mice significantly, and the treatment with DMXAA and 3MA blunted the anti-pyroptosis effect of exercise. Treadmill exercise exerted its anti-pyroptosis effect by inhibiting STING signaling through autophagy. 6) DMXAA and 3MA treatment blunted the ameliorative effects of 8 weeks of treadmill exercise on cognitive deficits in PD mice. Conclusions: Both the 4-week and 8-week treadmill exercise intervention could alleviated cognitive deficits in chronic PD mice induced by MPTP/p, and the effects of 8-week intervention is better. The protective effect of exercise intervention may be attributed to the activation of autophagy, which downregulates STING signaling, thereby reducing α-syn accumulation, and attenuating MG activation and pyroptosis, then regulating hippocampal synaptic structural plasticity.