Adenovirus vectors encoding a mouse AMIGO3 short hairpin RNA (AV/AMIGO3-shRNA) were constructed and authenticated by the Sangon Biotech Company (Shanghai, China). Furthermore, a green fluorescent protein (GFP) tag was encoded in the adenovirus vector sequence. Three adenovirus vectors were constructed for use in the study: one was used to silence the AMIGO3 gene (designated AV, sequence: 5′-GCACCACAACCAGACACTTGA-3′), one caused AMIGO3 gene overexpression (designated OAV), and the third was a negative control adenovirus vector carrying a nonsense gene sequence (designated NC, sequence: 5′-GTTCTCCGAACGTGTCACGTA-3′).

Postnatal day 21 [P21] male immature C57/BL6 mouse were provided by the experimental animal center of Chongqing Medical University (China). One hundred and fifty-four immature mice were sacrificed in total. All the mice were housed in a controlled environment (12/12-h light/dark cycle, humidity 50–60%, room temperature 22–24°C). To analyze the sequential changes of Lingo-1 and AMIGO3 in the cortex after SC, eight mice were sacrificed at four time-points (1, 3, 7, and 20 days after SC and at the same age in the control group). To determine the effect of regulating AMIGO3 expression on myelin sheaths, axon growth and synaptic plasticity, we allocated mice to five groups: (1), control group; (2), SC group; (3), SC group treated with AV; (4), SC group treated with OAV; and (5), SC group treated with NC. Six mice were evaluated at 1, 5, and 20 days after SC. All animal procedures were licensed and approved by the Laboratory for Animal Care of Chongqing Medical University.

Five days before inducing SC, the mice were anesthetized via intraperitoneal (i.p.) injection of pentobarbital sodium (7.5 mg/mL, 0.05 mL/10 g) and fixed in a stereotactic frame (RWD Company, Shenzhen, China). A 5-μL Hamilton syringe was placed 0.4 mm posterior and 1 mm lateral to the bregma at a depth of 2 mm relative to the parietal bone and located in the right lateral cerebral ventricle. Each group was injected with 1 μL (0.5 μL/min) virus at the following titers: AV group, 5.21 × 10¹⁰ PFU/mL; OAV group, 9.95 × 10¹⁰ PFU/mL; and NC group, 1 × 10¹¹PFU/mL. 5 days after the ICV injection, SC was induced in the mice of all SC subgroups by i.p. injection of kainic acid (2 mg/kg, Cayman Chemical Company, Ann Arbor, United States). After 30 min, all mice received a dose of pentobarbital sodium (7.5 mg/mL, 0.05 mL/10 g i.p.) to terminate SC. Mice were excluded from the study due to failure to induce SC failure or death after SC.

Mice were anesthetized with pentobarbital sodium (7.5 mg/mL, 0.05 mL/10 g i.p.), and then perfused with 0.9% sodium chloride and a mixture containing 4% paraformaldehyde and 0.5% glutaraldehyde. Hippocampal tissue blocks (1 mm³) were dissected and fixed in precooled 2.5% dedicated glutaraldehyde for 2 h at 4°C, and then osmicated in 1% osmium tetroxide (OsO₄) for 2 h at 4°C. After dehydration in an ascending acetone series, the blocks were infiltrated and embedded with epoxy resin. A single 50-nm section from each epoxy resin block was stained with uranyl acetate for 10–15 min and lead citrate for 1–2 min. Sections were examined under a transmission electron microscope (Tecnai G2 Spirit 120kV, Thermo Fisher Scientific United States) to evaluate the morphological changes in the myelin sheath and synapse. For each section, eight fields of vision were randomly chosen and photographed at magnifications of 6,800× and 30,000×. The diameters of axons and nerve fibers were calculated by measuring the circumference of each by using ImageJ software (NIH Image; U.S. National Institutes of Health, Bethesda, MD, United States). The g-ratios were defined by dividing the axon diameters by the fiber diameters, to provide a reliable index of myelination independent of axonal diameter (Song et al., 2018). The thicknesses of postsynaptic density (PSD) and length of active zone (AZ) were measured for three times in the field of view (30,000× magnification) for each group using ImageJ software. The mean values of PSD thickness and AZ length are used to perform statistical analysis.

The cortex of each mouse (n = 6 per time-point) was dissected on ice and immediately frozen in liquid nitrogen. Protein samples were extracted using a total protein extraction kit (cat#BB-3101-100T, BestBio, China) and protein concentrations were determined using a protein assay kit (lot# UB276924, Thermo scientific, United States). Samples containing 50 μg of protein and 6 μL Precision Plus Protein Dual Color Standards marker (Cat#1610374, Bio-Rad, California, United States) were separated by 10% polyacrylamide gel electrophoresis (Epizyme, Shanghai, China). Proteins [Lingo-1: 83 kDa; AMIGO3: 55 kDa; myelin basic protein (MBP): 21 kDa; myelin oligodendrocyte glycoprotein (MOG): 28 kDa; neurite outgrowth inhibitor protein A (NogoA): 180 kDa; myelin associated glycoprotein (MAG): 53 kDa; GAPDH 38 kDa; and β-actin: 43 kDa] were transferred to polyvinylidene difluoride (PVDF) membranes (0.2 μm, Bio-Rad, United States) in rapid transfer buffer (Cat. No: WB4600, NCM Biotech, Suzhou, China). The membranes were then blocked for 10 min in blocking buffer (Cat. No: P30500, NCM Biotech, China) before incubation overnight at 4°C with the following primary detection antibodies: rabbit polyclonal anti-Lingo-1 (1:1,000; ab23631, RRID:AB_2135216, Abcam, Cambridge, United States), rat monoclonal anti-AMIGO3 (1 μg/ml; MAB2375, RRID:AB_2226632, Bio-techne, Minneapolis, Minnesota, United States), mouse monoclonal anti-MBP (1:1,000; SMI99, RRID:AB_2314772, Biolegend, California, United States), NogoA (rabbit polyclonal anti-NogoA (1:1,000; ab62024, RRID:AB_956171, Abcam, United States), mouse monoclonal anti-MOG (1:5,000; MAB5680, RRID:AB_1587278, Millipore, Boston, Massachusetts, United States), mouse monoclonal anti-MAG (1:250; ab89780, RRID:AB_2042411, Abcam, United States), mouse monoclonal anti-β-actin (1:1,000; 700068, ZEN BIO, China), or mouse monoclonal anti-GAPDH (1:5,000; 200306-7E4, RRID:AB_2722713, ZEN BIO, Chengdu, China). The membranes were then incubated with the corresponding secondary antibodies for 60 min: goat anti-rabbit IgG (1:5,000; ZB-2301, ZSGB-BIO, Beijing, China), goat anti-rat IgG (1:5,000; ZB-2307, ZSGB-BIO, Beijing, China) or rabbit anti-mouse IgG (1:5,000; 701051, ZEN BIO, Chengdu, China). All proteins are detected on the same blot and sorted out later by targets’ molecular weights. Protein bands were visualized using clarity Western electrochemiluminescence substrate (Cat.1705060, Bio-Rad, United States) and quantified by densitometry using the Syngene imaging system. Protein expression was normalized against β-actin and GAPDH using Image Lab 6.0 software (Bio-Rad, United States).

Commercial ELISA kits (JiangLai Biology, Shanghai, China) were used to determine the concentrations of proteins in the mouse cortex [Lingo-1 (JL49841), AMIGO3 (JL51020), AKT (JL46534), p-AKT (JL20453), RhoA (JL51302), and p-RhoA (JL51309)] according to the manufacturer’s instructions. Briefly, tissue homogenates were added to the ELISA plates and incubated overnight at 4°C. After washing, horseradish peroxidase-conjugated detection antibodies in the commercial kits were added followed by the tetramethylbenzidine substrate solution. Protein concentrations were then quantified by measuring the absorbance of each well at 450 nm using a microplate spectrophotometer (Rayto RT-6100, Shenzhen, China).

At 1, 5, and 20 days after SC, mice were anesthetized by i.p. injection of pentobarbital sodium (7.5 mg/mL, 0.05 mL/10 g) and then perfused with 0.9% sodium chloride (50–100 mL) and 4% paraformaldehyde (50 mL) in sequence. After dissection, the brains were then fixed in 4% paraformaldehyde for 24 h, followed by gradient dehydration in sucrose solution (30% followed by 15%). A rotary microtome was used to prepare successive coronal paraffin-embedded sections (3 μm thickness), which were stored at room temperature before incubation overnight at 4°C with the following primary detection antibodies: rabbit polyclonal anti-Lingo-1 (1:200; ab23631, Abcam, United States), rat monoclonal anti-AMIGO3 (25 μg/mL; MAB2375, Bio-techne, Minneapolis, MN, United States). The slides were then incubated with the Alexa Fluor 488 labeled anti-rabbit or Cy3 labeled anti-rat secondary antibodies (Beyotime biotechnology, Shanghai, China, 1:250) for 2 h at room temperature away from light. Images were captured under a fluorescence microscope with the same exposure time (200× magnification).

Statistical analysis was performed using SPSS 22.0 statistical software. All data were presented as the means ± standard deviation (SD) and were analyzed by one-way or two-way analysis of variance followed by Tukey’s or Sidak’s post hoc test. P < 0.05 was considered to indicate statistical significance.

Western blot, ELISA and immunofluorescence analyses were used to detect the changes of the expression of Lingo-1 and AMIGO3 proteins in the cortex. Western blot analysis showed that the expression of both Lingo-1 and AMIGO3 proteins increased in the cortex after SC compared with the levels detected in the control group at the same age (Figures 1A–C). ELISA analysis revealed that Lingo-1 expression in the cortex was significantly increased at days 1, 3, 7, and 20 after SC (P < 0.05) (Figure 1D), while significant increases in AMIGO3 expression were detected at days 3, 7, and 20 after SC (P < 0.01) (Figure 1E). Furthermore, AMIGO3 expression in the cortex was higher than that of Lingo-1 expression (Figures 1D,E). Paragraph before as shown in Figure 2, immunofluorescence analysis of Lingo-1 (green fluorescence) and AMIGO3 (red fluorescence) revealed similar distributions of the two proteins in the different zones of hippocampus and cortex.

The AMIGO3 levels in the cortex after ICV injection with the recombinant adenoviruses was detected by ELISA (Figure 3). No statistical significances were detected between the control and NC groups at different timepoints (Figure 3). Compared with the control group at the same age, AMIGO3 expression in cortex of the AV group was significantly decreased at days 1, 5, and 7 after ICV injection (P < 0.05) (Figure 3A). In the OAV group, AMIGO3 expression in the cortex was significantly increased at day 1 after ICV injection, and the levels were maintained to day 7 (P < 0.05) (Figure 3B). Based on these observations, SC was induced 5 days after ICV injection with the different recombinant adenoviruses in the subsequent experiments.

The functional structures of the presynaptic membrane and postsynaptic membrane were further observed by TEM (Figures 7A–C). In each group, the number of synapses with clearly identifiable structure in the hippocampus was counted in four randomly selected fields of view (20,000× magnification). Compared with the control group, number of synapses was significantly decreased at days 1 and 5 after SC (P < 0.05). In contrast, the number of synapses in the AV group was increased at days 1 and 5 compared with the numbers in the SC group (P < 0.05) (Figure 7D). The length of the active zone (AZ) of the presynaptic membrane and thickness of postsynaptic density (PSD) in each group was measured and analyzed using ImageJ software (NIH Image; U.S. National Institutes of Health, Bethesda, MD, United States). Compared with the control group, the length of the AZ in presynaptic membrane was increased in the SC group at day 20 (P < 0.001). Compared with the SC group, the length of the AZ in presynaptic membrane was increased in the OAV group at day 1, but decreased in the AV group at days 1 and 20 (P < 0.05) (Figure 7E). The thickness of PSD significantly decreased from days 1 to 20 after SC when compared with the control group (P < 0.0001) (Figure 7F). Compared with the SC group, the thickness of PSD in the AV groups was significantly increased at days 5 and 20 (P < 0.0001), while that in the OAV group significantly decreased at day 20 (P < 0.05) (Figure 7F).

Expression of ROCK/RhoA and PI3K/AKT signaling pathway proteins in the cortex was detected by ELISA (Figure 8). Compared with the normal group, p−AKT and AKT expression levels decreased significantly at days 5 to 20 after SC (P < 0.01) (Figures 8A,B). Compared with the SC group, the expression levels of p−AKT and AKT in the AV group were higher at day 20 (P < 0.05) (Figures 8A,B). The ratio of p−AKT/AKT showed no statistical significance among groups at different timepoints (Figure 8C). The expression of p−RhoA increased significantly from days 1 to 20 after SC (P < 0.05) (Figure 8D). Compared with the SC group, the expression of p−RhoA in the OAV group was higher than at day 5, while the level in the AV group was lower at days 1 and 20 (P < 0.05) (Figure 8D). The expression of RhoA increased significantly at days 5 and 20 after SC (P < 0.05) (Figure 8E). Compared with the SC group, the expression of RhoA in the OAV group was higher at day 1, while that in the AV group was lower at days 5 and 20 after SC (P < 0.05) (Figure 8E). The ratio of p−RhoA/RhoA at day 20 showed statistical significance in the AV group compared with SC group (Figure 8F).