Fig. 1 Timeline in this experiment.

Treadmill exercise was performed according to previous literature^[19]. Briefly, a treadmill exercise apparatus for rat as used for chronic physical exercise. During the test session, the rat were allocated in the test room for 30 min acclimation, followed by normal exercise intervention. The speed of the treadmill exercise was 15 m/min, and the animal performed exercise 30 min daily for 4 weeks (5 days/week).

Repeated restraint stress model was generated by physical restrain as previous described^[20]. The rat was restrained for 6 h daily (15:00–21:00) in a wire mesh restrainer for 21 days. After restrain treatment, the rat were returned to their own cages. All the procedures were performed after physical exercise.

On the test day, rats were transferred to the testing room and acclimated to the room conditions for at least 1 h. After each individual test session, the apparatus was thoroughly cleaned with 70% alcohol to eliminate the previously tested rat’s odor and trace.

Open field test was to evaluated anxiety-like behavior^[21]. Open field experiment box (100 cm × 100 cm ×50 cm) were made out of plexiglass. Each animal was placed in the open field box for freely exploring the arena during a single 5 min session. During the test session, movement paths in open field experiment box were measured automatically by digital camera.

Passive avoidance test is a common measure of learning memory^[22]. The apparatus consisted of a light compartment and a dark compartment. In acquisition trial, rats we placed them in the light compartment and allowed them to freely explore it. As expected, due to their natural aversion to the light and preference for the dark, rats very quickly entered the dark chamber. The time of rat from the light chamber to enter the dark compartment was noted. Upon entry into the dark chamber, the guillotine door was closed. Then, rats received a single mild electric footshock (0.5mA; 1s) for 30s and then were returned to their homecage. 24 h later, we subjected rats to an retention trial. The procedure was similar to the acquisition trial except this time rats did not receive the electrical foot shock, and we recorded the time to enter the dark compartment (up to 300 s).

Animals were sacrificed and perfused with ice-cold PBS, followed by perfusion with 4% paraformaldehyde. The brain were removed rapidly, immersed in a tube containing 4% paraformaldehyde for 12 hours at 4℃ and cryoprotected in PBS containing 20% and 30% sucrose, respectively. They were then frozen on powdered dry ice and stored at - 80℃ until they were processed. Serial coronal sections (40 μm) were cut through the hippocampus on a cryostat. Brain sections were incubated at 37˚C for 30 min and rinsed in PBS for 10 min, incubated in 0.3% Triton X-100 for 10 min and 10% normal donkey serum for 2 hour. Then subsequently incubated overnight at 4℃ with primary antibody, including rabbit monoclonal to NeuN- neuronal marker. Next day, they were incubated in second antibody (Goat Anti-Rabbit IgG) for 2 hour after washed by PBS. After rinsing three times with PBS, the brain slices were attached to coated slide glasses and evaluated for confocal fluorescence microscopy^[23].

For the quantification of hippocampal morphological changes, Golgi staining kit (FD Rapid Golgi Stain Kit PK-401, USA)was used to analyze neuronal spine density following the manufacturer’s protocols. Briefly, the whole brains were extracted and placed in in EP tube with Golgi-Cox solution for 17 days in the dark. Then the brains were subsequently frozen and coronal sections (60 µm) were obtained using a cryostat (Leica, Nussloch, Germany). And the sections were incubated in the staining solution for 15 min, dehydrated in ethanol solution (30%, 50%, 80%, and 100%). Finally, slides were cover-slipped with Permount. Images of hippocampal pyramidal cells were obtained by panoramic multi-layer scanning with digital slice scanner at a magnification. To calculate the number of spine, a length of dendrite (at least ≥10 µm long, 2nd-oder dendrites) was traced (at 1000×).

The hippocampal brain tissue was lysed in RIPA buffer (Beyotime, Shanghai, China) and processed in ultrasonication and was centrifuged for 10 min to collect the supernatant. Hippocampal dopamine were detected after purification using ELISA kits (Nanjingjiancheng, China). Briefly, the wells of a 48-well flat-bottom high-affinity ELISA plates were coated overnight and add prepared sample and standards. Subsequently, the plate was washed five times in each well, and the plate was incubated for 20 min in the dark at room temperature. Finally, the reaction was stopped and measure the OD values within 10 minutes.

All data in this experiment are presented as the mean ± SEM. Data analysis was performed by GraphPad Prism software. Data analysis was performed by GraphPad Prism software. Student’s t-test was used to compare means between two groups. A statistical significant level was defined when P < 0.05.

The open-field test showed that stress rat displayed decreased central exploration time and central distance relative to the control group (Fig. 2 A-B; p < 0.01, p < 0.001). Locomotor activity, measured by total distance traveled, did not differ between the stress and control groups (Fig. 2 C; p > 0.05). Then we tested whether treadmill exercise could ameliorate stress related behavior abnormality. The results demonstrated that significantly elevated central exploration time and central distance in Exe-stress group when compared with stress group (Fig. 2 A-B; p < 0.01, p < 0.001). and there were no significantly different in total distance between the stress and control groups (Fig. 2 C; p > 0.05). These data demonstrate that stress rat develop an anxiety-like phenotype after repeated restraint stress exposure, which can be rescued by treadmill exercise.

Fig. 2. Treadmill exercise alleviated stress-induced anxiety behavior.^###p<0.001 vs. control;^***p<0.001 vs. stress.

Fig. 3 shows that the latency time of passive avoidance test in each group. In the acquisition trail, there are no significant difference for the four groups (Fig. 3, p > 0.05). In the retention trail: when compare to C group, the time in stress group was significantly shorter (Fig. 3, p < 0.001). As compare to stress group, the time in Exe-stress groups were significantly longer (Fig. 3, p < 0.001). The results indicated that stress-induced working memory impairment was ameliorated by treadmill exercise.

In order to examine the effect of exercise on hippocampal neuron in stress rat, staining with NeuN was used to detect hippocampal neuronal viability in each group (Fig. 4A). The number of NeuN cells in the stress group was significantly increased compared with that observed in the control group (Fig. 4B, p <0.01). But a greater NeuN-positive cells was observed in the hippocampus after treadmill exercise intervention (Fig. 4B, p <0.05). These results suggest that stress-induced hippocampal neuron loss can be rescued by treadmill exercise.