SPF male Sprague–Dawley rats (180–210 g) were purchased from the Experiment Animal Center of the Academy of Military Medical Sciences (Beijing, China) and bred and maintained at the SPF Animal Research Building of China Astronaut Research and Training Center. The temperature was controlled at 23°C and the animals had free access to food and water. After 7 days of acclimatization in conditions of 12-h light and 12-h dark cycles, the rats were randomly divided into the following four groups (eight animals per group): control (Ctrl), SLSE, SLSE plus DS (DS), and SLSE plus piracetam (Pir) groups.

The rats in the Ctrl group were kept five per cage with a natural light-dark cycle (12-h light and 12-h dark). Other rats were given stimulated SLSE. The rats in DS groups received intragastric administration with DS (50 mg⋅kg^-1; Pegasus Pharmaceuticals Group Inc, Richmond, BC, Canada) once a day. Those in Pir groups were given Pir (216 mg⋅kg^-1; Huazhong Pharmaceutical Co. Ltd., Xiangyang, China) by gavage once a day. The Ctrl and SLSE groups were given an equal volume of distilled water, which was the vehicle used for DS and Pir. Behavioral tests were performed from 9:00 to 12:00 after 6 weeks of drug administration. The drug and SLSE administration lasted until the day when behavioral tests finished (Figure 1C).

This study was carried out in accordance with the recommendations of guidelines for the use and care of live animals, which was approved by the Animal Care and Use Committee of China Astronaut Research and Training Center.

Simulated long-duration spaceflight environment was established as the method in our previous study (Liu et al., 2015). Four factors were chosen to simulate the spaceflight environment: hind-limb unloading, isolation rearing, steady-state noise, and altered light-dark cycle (Figure 1B). Rats were reared individually in cages with frosted glass to simulate isolation and confinement in manned space explorations. The hind-limb-unloading procedure was achieved by tail suspension, as described previously (Morey-Holton and Globus, 2002; Wang et al., 2013), to simulate microgravity in space. Briefly, the rats were individually caged and suspended by the tail using a strip of adhesive surgical tape attached to a chain hanging from a pulley. The rats were suspended at a 30° angle to the floor with only the forelimbs touching the floor, which allowed them to move and access food and water freely. At the same time, 90 min cycles with 45-min light and 45-min dark were generated using an automatic light controller to simulate the light-dark cycle in low-earth-orbit flight. A continuous steady-state noise of 65 ± 5 dB sound pressure level was generated with a white noise generator (UZ-3; Great Wall Electronic Equipment Co., Beijing, China) and passed to a loudspeaker (D40, SAST, Shenzhen, China) by a balancer (DEQ2496; Behringer, Zhongshan, China) and a power amplifier (MP3-50; TAMO, Shenzhen, China). This was to simulate the noise in the spaceflight environment, which is closed to the noise intensity in a manned aircraft (Si et al., 2016). The noise was measured using sound level meters (Type 2270; B&K, Naerum, Denmark) and was applied consistently to rats 12 h a day.

Learning and memory were assessed using MWM test and SDPAT. MWM tests were used to evaluate spatial learning and memory (D’Hooge and De Deyn, 2001) with slight modifications from the method reported previously (Shetty et al., 2013). A circular pool (120 cm in diameter, 50 cm in depth) filled with water (30 cm in depth, made opaque by the addition of black ink and maintained at 23 ± 1°C) was divided into four equally spaced quadrants (I, II, III, and IV) with numerous extra maze cues on its walls. A platform (10 cm in diameter) was placed in the center of quadrant III, with its surface 1 cm below the water level. The water maze was placed in a quiet room with dimmed lights. The rats were placed individually in water in one of the three quadrants not containing the platform. They were placed facing the wall around the rim of the pool. The rats that failed to find and stand on the platform within 60 s were guided to the platform and allowed to stay there for 10 s. The place navigation trials were carried out for 4 days (three times a day). On the 5th day, the spatial probe trial was conducted in the absence of the platform. The rats were placed in the water from quadrant I. Their activities were videotaped and analyzed with a custom-built MWM automatic display and analysis processing system, and then the escape latency (times to find the platform) in place navigation trials and the percentage of time spent in the platform quadrant in the spatial probe trial were calculated.

Step-down passive avoidance test was used to measure memory retention and the ability for non-spatial learning and memory with some modifications from the method reported previously (Zarrindast et al., 2006; Bortolatto et al., 2013). The apparatus was a box (30 cm × 30 cm × 40 cm high) with parallel stainless steel bars (0.3 cm diameter spaced 1 cm apart) that was placed on the floor. A cylindrical platform (10 cm in diameter and height) was placed on the center of the grid floor. On the 1st day of the training session, each rat was placed in the box to acclimatize for 3 min, and the bars were then electrified (3 mA, 65 V) for 5 min. Memory retention tests were carried out 24 h later after the training session. Without acclimatization, each rat was placed in the box again with the bars electrified (3 mA, 65 V) for 5 min. The activities were videotaped and analyzed using an SDPAT automatic display and analysis processing system, and then the latency time (time to step onto the platform from the electrified bars), the duration on the platform (over a 4-min period), and error times (the number of times stepped down from the platform during 4 min) were calculated.

1 day after the behavioral tests, rats were deeply anesthetized with chloral hydrate (500 mg⋅kg^-1) and decapitated to collect brain tissue. Four brains in each group were fixed in 4% paraformaldehyde for histopathological evaluation. The other four brains in each group were dissected into hippocampus and cortices and then were stored at -80°C until analysis. The latter were ground in liquid nitrogen for liquid chromatography-tandem mass spectrometry (LC–MS/MS) analysis, microarray analysis, and western blotting analysis.

Liquid chromatography-tandem mass spectrometry method was used to measure seven major neurotransmitters simultaneously in rat brain, including 5-HT, DA, NE, E, Glu, ACh and GABA. The analytical method was carried out as the previous study with some modifications (Huang et al., 2014). Briefly, 20 mg of hippocampus and cortices were homogenized with 9 times the amount of ice-cold normal saline. The homogenate and 3 times the ice-cold acetonitrile was vortex-mixed and placed for 30 min at 4°C, and then centrifuged at 12,000 r⋅min^-1 at 4°C. The supernatant was collected. 200 μL of the supernatant was mixed with 20 μL internal standard solution (0.3 mg⋅mL^-1 dihydroxy-benzoic acid), then the mixture (50 μL) was injected into LC–MS/MS system for assay.

Liquid chromatography-tandem mass spectrometry method was run on a Q-Trap 5500 hybrid triple quadrupole-linear ion trap mass spectrometer (SCIEX, Vaughan, ON, Canada) equipped with an electrospray ionization ion source and an Agilent 1290 HPLC system with auto-sampler (Agilent Technologies, Santa Clara, CA, United States). The chromatographic separation was carried out using a TSKgel Amide-80 HPLC column (2 mm × 15 cm, 3 μm, Tosoh, Tokyo, Japan) at 35°C. The mobile phase system was a mixture of acetonitrile-ammonium formate solution (15mmol⋅L^-1, pH = 5.5, 40:60, v/v) with flow rate of 0.4 mL⋅min^-1. The injection volume was 5 μL. The instrument was operated in positive electrospray ionization multiple reaction monitoring mode with the following parameter settings: ion spray voltage, 5500 V; gas temperature, 500°C; gas 1, 50 Psi; gas 2, 70 Psi. The multiple reaction monitoring parameters used for analysis were summarized in Table 1.

After post-fixing in 4% paraformaldehyde for 48 h, the whole brains were dehydrated and embedded in paraffin. Brain sections (5 μm) were then prepared and subjected to Nissl’s staining. The populations of intact neurons in different sub-regions were counted by a blinded observer.

Each group consisted of six samples (including three hippocampi and three cerebral cortices). The KAM-850 antibody microarray (Kinexus, Vancouver, Canada) was used to perform an unbiased characterization of the signaling proteins that were regulated in the different groups, as described previously (Kupcova Skalnikova et al., 2013), in accordance with the manufacturer’s instructions. Briefly, 50 μg of protein lysate from each sample was covalently labeled with a proprietary fluorescent dye. Free dye molecules were then removed by gel filtration. After blocking non-specific binding sites on the array, an incubation chamber was mounted onto the microarray to permit the loading of three samples side by side on the same chip and prevent sample mixing. Following sample incubation, unbound proteins were washed away. The arrays were then scanned using a Scan Array Reader laser array scanner (Perkin-Elmer, Waltham, MA, United States) with a resolution of 10 μm, and the acquired images were quantified using ImaGene 9.0 (BioDiscovery, El Segundo, CA, United States) with predetermined settings for spot segmentation and background correction. Z scores were calculated by subtracting the overall mean intensity of all spots within a sample from the raw intensity for each spot and then dividing it by the SD of all of the measured intensities within each sample. Z ratios were further calculated by taking the difference between the means of the observed protein Z scores and dividing by the SD of all of the differences for that particular comparison. A Z ratio above 1.5 was inferred as significant. By converting the corresponding genes to proteins, KEGG enrichment analysis was then performed to detect altered signaling pathways in different groups, as described previously (Kanehisa and Goto, 2000).

The rat protein interaction datasets were downloaded from STRING database v10¹, and only the high confidence interactions with experimental evidence were selected (Szklarczyk et al., 2015). The network graphs were produced using Cytoscape software (Shannon et al., 2003). Their topological parameters, such as degree centrality, betweenness centrality, clustering coefficients, shortest path lengths, and closeness, were calculated using Cytoscape plug-in Network Analyzer (Assenov et al., 2008). The proteins were then ranked by protein degree.

Hippocampal and cortical tissues were extracted using Tissue Protein Extraction Reagent (Thermo, Rockford, IL, United States) plus Protease Inhibitor Cocktail (Roche, South San Francisco, CA, United States) and Phosphatase Inhibitor Mixture (Applygen, Beijing, China) on ice for 30 min. Protein fractions were then collected by centrifugation at 15,000 g at 4°C for 10 min, subjected to SDS-PAGE, and transferred to polyvinylidene difluoride membranes. The membranes were blocked with 5% BSA and incubated with specific antibodies overnight. After washing and incubating with horseradish peroxidase-labeled secondary antibodies, the specific bands were visualized using Immobilon Western Chemiluminescent HRP Substrate (Millipore, Billerica, MA, United States). The following antibodies were used: PI3K (rabbit, 1:1000; CST, 4249), p-PI3K (rabbit, 1:1000; CST, 2971), AKT (rabbit, 1:1000; CST, 9272), p-Akt (rabbit, 1:1000; CST, 9271), mTOR (rabbit, 1:1000; CST, 2983), p-mTOR (rabbit, 1:1000; CST, 2971), CREB (rabbit, 1:1000; CST, 9197), p-CREB (rabbit, 1:1000; CST, 9198), BDNF (sheep, 1:1000; Abcam, ab75040), JNK (mouse, 1:1000; CST, 3708), p-JNK (mouse, 1:1000; CST, 4668), p38 (rabbit, 1:1000; CST, 8690), p-p38 (rabbit, 1:1000; CST, 4511), and GAPDH (mouse, 1:1000; Santa Cruz, sc-47724). The ratio of the protein band intensities relative to that of GAPDH was calculated for each sample using Image J software (US National Institutes of Health, Bethesda, MD, United States).

All data are presented as means ± SEM and were analyzed using GraphPad Prism 5.0 software (GraphPad Software, San Diego, CA, United States). Multiple comparisons were performed using one-way ANOVA followed by Bonferroni’s post hoc test. P values less than 0.05 were considered statistically significant.

The common behavioral methods MWM tests and SDPAT were used to assess the cognitive effects of SLSE exposure in rats (Figure 2). In the MWM test, representative searching swimming paths on the 4th day of place navigation trials displayed clear differences among groups (Figure 2A). Escape latencies to the hidden platform from quadrant I decreased in all groups over 4 days of the place navigation test (Figure 2B). SLSE-treated rats had higher latencies than the control, and DS-and piracetam-treated rats had lower latencies than the rats in the SLSE group on the 4th day. Figure 2C shows that SLSE-treated rats had a tendency to swim everywhere in the MWM, whereas rats in other groups preferred to move in the platform quadrant in the spatial probe test performed on the 5th day. The percentage of time spent in the platform quadrant was decreased in SLSE-treated rats compared with that in the control, and was increased in the DS and Pir groups compared with that in the SLSE group (Figure 2D). These results suggest that SLSE can lead to memory loss and cognitive decline, and that DS and piracetam can improve spatial learning and memory.

In SDPAT, SLSE significantly increased the latency time and error times compared with those in the control group, and DS and piracetam decreased the latency time and error times compared with those in the SLSE group (Figures 2E,F). Meanwhile the total duration spent on the platform was decreased in the SLSE group compared with that in the control, and was increased in the DS group compared with that in the SLSE group; however, there was no significant difference between the SLSE and Pir groups (Figure 2G). These results indicate that SLSE can decrease the efficiency of passive avoidance, and that the cognition-enhancing effects of DS are similar to that of piracetam. Above all, these findings suggest that SLSE can induce neurocognitive functional impairment in rats, and that DS has cognition-enhancing effects in SLSE-exposed rats.

Neurotransmitters are closely linked to cognitive function. A rapid and sensitive LC–MS/MS method was used to measure seven major neurotransmitters simultaneously in the rat brain (Figure 3), including four monoamine (Figures 3A–D), two amino acid (Figures 3E,F), and ACh (Figure 3G). In the cerebral cortices, compared with those in the control group, the contents of 5-HT, DA, NE, E, Glu, and ACh (Figures 3A–E,G) were decreased in the SLSE group, whereas the GABA content (Figure 3F) was increased. The GABA content (Figure 3F) in the cerebral cortices was decreased in the DS treatment group compared to that in the SLSE group, while the levels of all other markers were increased. In the hippocampus, the changes in the 5-HT, NE, and ACh contents were the same as those in the cerebral cortices (Figures 3A,C,G); The other neurotransmitters had no significant changes (Figures 3B,D–F). These results show that DS had significant effects on neurotransmitter levels in the SLSE-exposed rat brain.

To assess the effects of SLSE and DS on neurons, Nissl’s staining and microphotography were used to study the changes in the hippocampal and cortical tissues in the different groups (Figure 4). Compared with those in the control and DS groups, neurons in the hippocampal subfields 1, 2 and 3 (Figures 4A–D) and cerebral cortices (Figure 4E) of SLSE-exposed rats were larger in the intercellular space, loosely and irregularly arranged, and weakly stained, indicating that they were injured (Figures 4A–E). There were significantly fewer neurons in the hippocampal subfields 1, 3 and cerebral cortices in SLSE-exposed rats than in the control and DS groups (Figure 4F). According to the hippocampal regionalism, subfields 1, 2 and 3 are in the region of CA3. DS significantly increased the number of neurons in the hippocampus CA3 and cortex and prevented neuronal injury in SLSE rats.

Protein antibody microarrays are the most efficient and economical way to analyze multiple samples or proteins at the same time (Wingren and Borrebaeck, 2004). To investigate further the underlying molecular mechanism of action of DS in SLSE rats, Kinex^TM antibody microarrays were used to profile protein expression in rat brains. A total of 854 molecules were detected in the microarray.

In the hippocampus, KEGG pathway enrichment analysis showed that 8 pathways were significantly altered in the SLSE group compared with those in the control (Figure 5A) and 14 pathways were significantly altered in the DS group compared with those in the SLSE group (Figure 5B). In the cerebral cortices, the numbers of pathways altered in the respective groups were 11 (Figure 6A) and 16 (Figure 6B). The MAPK and neurotrophin signaling pathways were significantly altered in most groups; their “–log₁₀P” values were larger than those of most other pathways, which suggested that they could be most strongly related to the changes in nervous function.

There were 51 differentially expressed proteins whose trends of change in the Ctrl and DS groups compared with the SLSE group were consistent. A hippocampus protein interaction network was constructed with 165 interactions among the 51 differentially expressed proteins based on STRING database (Figure 5C). In this network, there were more proteins connecting with p38a, CDK1, JNK, and S6K than with the others; therefore, these may be potential target proteins of DS. Similarly, a cortex protein interaction network composed of 193 interactions among 65 differentially expressed proteins was constructed (Figure 6C). The network revealed that CDK1, GSK3b, S6Kb1, Bcl-2, and ATF2 proteins were more likely than the others to play important roles in the cognition-enhancing effects of DS.

To validate further which pathways are influenced by DS in the cerebral cortex and hippocampus of SLSE rats, the main proteins in the MAPK and neurotrophin signaling pathways and some other key proteins in the protein interaction networks were analyzed using western blotting (Figures 5D,E, 6D,E). As shown in Figures 5D, 6D, both in the hippocampus and in the cortices, the expression of p-PI3K, p-AKT, and p-mTOR was significantly decreased in the SLSE group compared with that in the control group, and increased in the DS group compared with that in the SLSE group. In the two tissues, the expression of p-CREB and BDNF was significantly decreased in the SLSE group compared with that in the control, and increased in the DS group compared with that in the SLSE group. But only in the hippocampus, the expression of p-JNK and p-p38 was increased in the SLSE group compared with that in the control group, and decreased in the DS group compared with that in the SLSE group. There were no significant changes in the expression of p-JNK and p-p38 in the cortices (Supplementary Figure 1).