Treadmill exercise ameliorates hippocampal synaptic injury and recognition memory deficits by TREM2 in AD rat model

Objective: The impairment of cognitive function has been associated with Alzheimer's disease (AD). Exercise exerts a positive modulatory effect on cognition by reducing synapse injury. However, limited in vivo evidence is available to validate the neuroprotective effect of TREM2 on synaptic function in this phenomenon. Here, we aim to explore whether physical exercise pretreatment alters A beta-induced recognition memory impairment in structural synaptic plasticity within the hippocampus in AD rats. Methods:: In study 1, fifty-two Sprague-Dawley (SD) rats were randomly divided into following four groups: control group (C group, n =13), Alzheimer's disease group (AD group, n =13), 4 weeks of physical exercise and Alzheimer's disease group (Exe+AD group, n =13), 4 weeks of physical exercise and blank group (Exercise group, n =13). Four weeks of treadmill exercise intervention was performed, and AD model were established by intra-cerebroventricular injection (ICV) injection of A beta 1-42 protein. After 3 weeks, we also conducted a novel object test to evaluate recognition memory in the behavior assessment. Golgi staining and transmission electron microscopy were used to evaluate the morphology and synaptic ultrastructure of neurons. Western blotting was used to measure the expression of hippocampal synaptic proteins. Extracellular neurotransmitters in the hippocampus were detected by microdialysis coupled with high-performance liquid chromatography. In study 2, 33 SD rats were randomly divided into three groups: 4 weeks of physical exercise and Alzheimer's disease group (Exe+AD group, n =11), AAV-Control and physical exercise and Alzheimer's disease group (AAV-Control+Exe+AD group, n = 11), AAV-TREM2 and physical exercise and Alzheimer's disease group (AAVTREM2 +Exe+AD group, n = 11). Stereotactic intracerebral injection in the bilateral hippocampus was performed to achieve microglial TREM2 down-expression by using adeno-associated virus (AAV) with CD68 promoter. After 4 weeks treadmill exercise and 3 weeks A beta injection, all rats received behavior test and molecular experiment, which the same with experiment 2. Results: Novel recognition index in novel object recognition test significantly decreased, and western blot demonstrate that hippocampal TREM2 protein is significantly decreased (P < 0.001). But physical exercise reversed this phenomenon(P < 0.001). In addition, compared with Con group, the neuron from Exe+AD group exhibited a more complex branching pattern (P < 0.05). And impaired synaptic ultrastructure was observed in AD group. Hippocampal synaptic-related protein (SYX, SYP, GAP43, PSD95) and neurotransmitter (DA, Glu, GABA) was also significantly decreased (P <0.01) in AD group. But the neuroprotection effect can be found in Exe+AD group, which are associated with the inhibition of synaptic injury by activate hippocampal TREM2 (P < 0.05). However, when blockade of hippocampal TREM2 reduced brain protective effect of exercise in AD rat model, including increased the damage of neuronal dendritic complexity, synaptic ultrastructure, and the decrease of hippocampal synapses-related protein, typical neurotransmitter.