METH (98%) was offered by the Hubei Public Security Bureau. Krill oil was provided by the Aker BioMarine Antarctic Company (Norway). Antibodies against protein kinase A (PKA), phosphorylated protein kinase A (p-PKA), cAMP-response element-binding protein (CREB), phosphorylated CREB (p-CREB), extracellular regulated protein kinase 1/2 (ERK1/2), and phosphorylated ERK1/2 (p-ERK1/2) were bought from Cell Signaling Technology (Danvers, United States). Antibodies against the dopamine D1 receptor (DRD1), dopamine D2 receptor (DRD2), dopamine transporter (DAT), and cleaved caspase-3 were obtained from Absin Bioscience Co., Ltd. (Shanghai, China). Antibodies against GAPDH, horseradish peroxides (HRP)-conjugated goat anti-rabbit antibody and HRP-conjugated goat anti-mouse antibody, protein extraction buffer, protease inhibitors, and phosphatase inhibitors were obtained from Wuhan Boster Biological Technology Co., Ltd. (Wuhan, China). Antibodies against the brain-derived neurotrophic factor (BDNF) were purchased from Santa Cruz Biotechnology (Santa Cruz, United States). All other reagents used were of analytical grade.

C57BL/6 mice (male, 8 weeks) were provided by Beijing Vital River Laboratory Animal Technology Co., Ltd. Mice were housed five per cage in a 12-h light–dark cycle and a temperature-controlled environment. All animal experiment procedures were approved by the Ethics Committee of Jianghan University. Specifically, mice were divided randomly into the control group, METH group, krill oil-L group and krill oil-H group, and the experimental procedure is detailed in Figure 1A. Before administration, krill oil was dissolved in ethanol and then diluted with saline, and METH was dissolved in saline at a concentration of 10 mg/ml. For the first 2 weeks, mice in the krill oil-L group and krill oil-H group were intragastrically administrated with 10 mg/kg and 100 mg/kg krill oil every day, respectively, and mice in the control group and METH group were intragastrically administrated with vehicle. At the third week, 1 hour after intragastric administration, mice in the METH, krill oil-L group, and krill oil-H group were intraperitoneally injected with 10 mg/kg METH, and mice in the control group were intraperitoneally injected with saline. Behavioral experiments, sample collection, and the Western blot test were performed as follows (Figure 1).

Memory training on the hidden platform of the Morris water maze was used to measure the associative, spatial memory of mice as a previous report. Briefly, water (23 ± 2°C) was added into a circular pool (21 cm in deep and 120 cm in diameter). A circular hidden platform was placed in the center of the target quadrant and submerged 1.5 cm below the water surface. XR-XM101 software was used to automatically measure the animal escape latency, swimming speed, and the amount of time spent in the target quadrant (Shanghai Xinruan Information Technology Co., Ltd.). Mice underwent four trials each day during the training phase, and the starting position of each trial was different. Mice have a maximum of 60 s in each trial to find the platform, and mice were allowed to stay on the platform for 15 s after boarding the platform. If the mice cannot find the platform within 60 s, it was guided to the platform to rest for 15 s. The time interval between trials was 30 min. Mice were dried with a towel, placed in a cage with a heating lamp, and then returned to their home cage after the last trial. The training phase lasted for 5 days. To test the spatial memory ability of mice, a 60-s free swimming test without a platform was performed on the sixth day.

The potential protein targets of krill oil, related genes of METH, and related genes of memory deficits were collected from the GeneCards database (https://www.genecards.org/). Then, the protein targets of krill oil were mapped with related genes of METH and related genes of memory impairment on the Bioinformatics and Evolutionary Genomics website (http://bioinformatics.psb.ugent.be/webtools/Venn/). To further characterize the molecular mechanism of krill oil on METH-induced memory deficits, the compound–target networks were generated using Cytoscape 3.8.0. In these graphical networks, the compounds and proteins were expressed as nodes, whereas the compound–target interactions were expressed as edges.

The gene ontology (GO) analyses and KEGG pathway analyses were conducted using the functional annotation tool of DAVID Bioinformatics Resources 6.7 (http://david.abcc.ncifcrf.gov/). Terms with thresholds of counts ≥ 10 and p values ≤ 0.05 were chosen in functional annotation clustering. Related target proteins of krill oil on METH-induced memory deficits were analyzed by online STRING 11.0 (https://string-db.org/) to construct a protein–protein interaction (PPI) network. The network was visualized with Cytoscape (v3.1.2) and CytoHubba, a plug-in in Cytoscape, to filter the modules from the PPI network and to obtain the most important hub genes based on the degree score.

Hippocampi were isolated, lysed, and centrifuged for 15 min (4°C) at 12,000 g . After the detection of concentration, the supernatants were mixed with a loading buffer and denatured for 5 min. Protein samples were separated using a gel electrophoresis system and transferred to the polyethylene difluoride membranes. After blocking for 1 h in 5% non-fat milk, the membranes were incubated with the primary antibody and then with the HRP secondary antibody. Enhanced chemiluminescence (Thermo Fisher, United States) was used to observe the bands using a chemiluminescence detector (Gene Corporation, Hong Kong). The intensity of each band was determined quantitatively by ImageJ and calibrated by the corresponding internal reference protein, and the results were shown as normalized for the control group.

Data were expressed as the mean ± standard error (SEM). All results were analyzed using SPSS 23.0 software. Results of the swimming speed and escape latency during the training phase of MWM were analyzed by one-way ANOVA with repeated measures. Other data used one-way ANOVA and Tukey’s HSD post hoc test. A p value less than 0.05 was considered statistically significant.

The Morris water maze test was used to determine whether krill oil alleviated METH-induced spatial learning and memory impairment. As shown in Figure 2, the locomotor activity of mice was not influenced by METH or krill oil treatment since the swimming speed did not differ among the groups (Figure 2A). Mice in the METH group had a higher escape latency than those in the control group on day 4 and day 5 of the training phase (Figure 2B), suggesting that repeated METH exposure triggered a decline in the spatial learning ability of mice (p < 0.05). Moreover, compared with the METH group, 10 mg/kg krill oil treatment greatly reduced the increase of learning latency in mice induced by METH on day 5 (p < 0.01), and mice pretreated with 100 mg/kg of krill oil performed significantly better than those that received METH alone on day 4 and day 5 (p < 0.01), suggesting that learning deficits were improved following krill oil treatment. In the testing section, results revealed significant differences among group effects (Figures 2C,D, p < 0.05). Mice in the METH group had worse performance in the parameter of times of crossing the platform and time in the quadrant of the platform than those of the control group (p < 0.05). Treatment with krill oil (10 or 100 mg/kg), however, remarkably increased the times of crossing the platform and the time in the quadrant of the platform (p < 0.05). The swimming trajectory further confirmed that the krill oil–treated mice stayed in the target quadrant longer than the METH group (Figure 2E). Taken together, these findings demonstrated that krill oil improved METH-induced cognitive deficits of spatial learning and memory.

Krill oil from Euphausia superba (Antarctic krill), an Antarctic marine species, is rich in EPA, DHA, and astaxanthin. Among the three main bioactive components of krill oil, 1,846 protein targets were retrieved from the GeneCards database. The detailed information is shown in Supplementary Table S1. After eliminating the overlaps, 1,445 protein targets were obtained for further analyses. 8400 memory deficit–related genes and 422 METH-related genes were collected from the GeneCards database. The detailed information is shown in Supplementary Tables S2, and S3. Then, these protein targets of krill oil were mapped with related genes of METH and memory impairment on the Bioinformatics and Evolutionary Genomics website. As a result, 210 targets of krill oil were associated with METH-induced memory impairment, and the detailed information of the 210 targets is shown in Supplementary Table S4 and Figure 3A. Among the 210 target genes, 116 were target genes for DHA, 174 were target genes for EPA, and 16 were target genes for astaxanthin. The detailed information is shown in Supplementary Table S5 and Supplementary Figure S1.

As krill oil may exhibit multiple pharmacological activities via multiple targets, it was constructive to investigate the underlying mechanisms of krill oil on complex diseases by network analysis. In the current study, the compound–target network of krill oil on METH-induced memory deficits was constructed with Cytoscape 3.8.1 software (Figure 3B). Among these potential protein targets, there were 12 high-degree targets associated with multiple compounds (Supplementary Figure S1), namely, NFKBIA, COX5A, MAPK14, HMOX1, NOS2, MAPK8, JUN, CAT, MMP1, CYP3A4, CXCL8, and SOD1. These high-degree protein targets in the network may account for the essential protective effects of krill oil on METH-induced memory impairment.

To identify the biological characteristics of putative targets of krill oil on METH-induced memory impairment in detail, the GO and KEGG pathway enrichment analyses of involved targets were conducted via the functional annotation tool of DAVID Bioinformatics Resources 6.8. There were 78 biological processes (BP), 46 cellular components (CC), and 23 molecular function (MF) terms in total, which met the requirements of counts ≥ 10 and p values ≤ 0.05. The detailed GO information is shown in Supplementary Table S6. The top 15 significantly enriched terms in BP, CC, and MF categories are shown in Figure 4A, which indicated that krill oil may improve METH-induced memory impairment via regulation of the neuron apoptotic process, response to oxidative stress, response to hypoxia, dopamine binding, NMDA glutamate receptor activity, and oxidoreductase activity. To explore the underlying involved pathways of krill oil on METH-induced memory impairment, a KEGG pathway analysis of involved targets was conducted. The detailed pathway information of krill oil on METH-induced memory impairment is shown in Supplementary Table S7. There were 94 pathways that met the requirements of counts ≥ 10 and p values ≤ 0.05. The top 15 significantly enriched pathways are shown in Figure 4B. The pathways in neuroactive ligand–receptor interaction exhibited the largest number of involved targets (38 counts).

The PPI network of possible target genes involved in the protective effect of krill oil on METH-induced memory impairment was constructed using STRING (Figure 5A).

There were 205 nodes and 3,388 edges, and the average node degree was 33.1. The average local clustering coefficient was 0.565, and the p value of PPI enrichment was less than 1.0e-16. To obtain the hub genes in the PPI network, these node pairs were entered into Cytoscape software. The scores of nodes were calculated by CytoHubba, and the top 20 hub genes are shown in Figure 5B, and the hub gene symbols, full names, and functions are shown in Table 1. The pathway enrichment analyses of hub genes were conducted via DAVID Bioinformatics Resources 6.8. There were 54 pathways that met the requirements of count ≥ 5 and p values ≤ 0.05. The detailed pathway information of hub genes is shown in Supplementary Table S8. The top 25 significantly enriched pathways are shown in Figure 6A and Table 2. The compound–hub gene–pathway network of krill oil on METH-induced memory deficits was constructed with Cytoscape 3.8.1 software (Figure 6B).

The pathways in the MAPK signaling pathway exhibited the largest number of involved targets (13 counts). Among the 20 hub genes, 17 were target genes for DHA, 17 were target genes for EPA, and 4 were target genes for astaxanthin. The detailed information is shown in Supplementary Table S9 and Supplementary Figure S2.

Network pharmacology analysis predicted that the molecular targets highly associated with the common signaling pathways including the MAPK signaling pathway, dopaminergic synapse, and cAMP signaling pathway may be related to the neuroprotective effect of krill oil on METH-induced memory impairment in regulating neuron functions and apoptosis. We further validated the expressions of the potential hub genes identified via network pharmacology. As shown in Figures 7A and B, compared with the control group, the expression of DRD2 in the hippocampus of mice in the METH group was greatly increased and the expression of DAT was decreased, and 100 mg/kg krill oil treatment significantly decreased the expression of DRD2 and increased the expression of DAT, while 10 mg/kg krill oil treatment significantly decreased the expression of DRD2. There was no significant difference of the expression of DRD1 among different groups. Likewise, METH treatment led to apparent repression of p-PKA, p-ERK1/2, p-CREB, and BDNF (Figures 7A,C–F), and 10 and 100 mg/kg krill oil treatment significantly increased the expression of p-PKA, p-ERK1/2, p-CREB, and BDNF. There were no significant differences of the expression of PKA, ERK1/2, and CREB among different groups. Moreover, METH treatment led to apparent enrichment of cleaved caspase-3 (Figures 7A,F), and 10 and 100 mg/kg krill oil treatment significantly reduced the expression of cleaved caspase-3. These results validated that krill oil may regulate the neuron functions and apoptosis mainly through the MAPK signaling pathway, cAMP signaling pathway, and dopaminergic synapse pathway.