Healthy male specific-pathogen-free (SPF) Sprague–Dawley (SD) rats were kept under a controlled environment condition (12 h:12 h light/dark cycle, temperature of 21 ± 1°C and humidity of 60%) with free access to food and water. We collected 12-month-old and 3-month-old female SD rats (n = 12 in each group) with no prior pregnancy experience. The virgin rats were mated individually with a randomly selected 3-month-old male rat. After pregnancy, female rats were removed from their cages for natural delivery. The offspring of 12-month-old rats were randomly selected from each litter and assigned to the AMA group (n = 45 in total). The offspring of 3-month-old female rats were randomly selected (both male and female) as the control group (Ctl; n = 45 in total) according to the quantity and sex of the AMA group (Supplementary Material Table 1). A novel object recognition behavior experiment was performed at P28 to assess cognitive function in the AMA offspring. At three time points (P14, P28, and P60), animals were sacrificed, and then the samples were prepared for immunofluorescence, western blot analysis and transmission electron microscopy. All animal experimental procedures were conducted in accordance with international guidelines for the care and use of laboratory animals and were approved by the Animal Care Committee of Chongqing Medical University, Chongqing, China.

The Novel object recognition experiment was conducted at P28 to assess the cognitive function of the rats. The rats were placed in a 1 × 1 m square black box. Each experimental rat was placed in the tank for 10 min to allow it to adapt to the box environment to avoid interference during the test period. On day 1, two objects of the same size and color were placed in the two corners of the tank, and the rats were placed in the tank to become familiar with the two objects for 5 min. On day two, one object was replaced by another object whose shape and color were completely different from those of the object present on day 1. The rat was put into the box and allowed to move freely for 5 min. Effective contact was defined as directly touching the object or a distance of <2 cm from the object recorded by the nose of the rat. The time and numbers of rats touching the object were recorded by the computerized video tracking system. The discrimination ratio was calculated as the time spent exploring the new objects/time spent exploring the old objects.

At 14, 28, and 60 days after birth, rats (n = 5 rats for each group) were intraperitoneally (i.p.) anesthetized deeply with sodium pentobarbital and then perfused with 0.9% sodium chloride. Brains were removed and fixed in 4% paraformaldehyde. Each brain was divided into 30-μm-thick frozen coronal sections. Complete eyebrow-shaped hippocampal slices were taken from the same septotemporal level for MBP staining. Anti-MBP primary antibody (Cat. No:SMI-99P, mouse monoclonal antibody, Biolegend, San Diego, CA) was used at a concentration of 1:500. The primary antibody was incubated with the tissue overnight. Then, the slides were incubated with the Alexa Fluor 555 labeled donkey anti-mouse secondary antibodies (Cat. No: A0460, Beyotime biotechnology, Shanghai, China) at 1:500 for 2 h at room temperature away from light.A laser scanning confocal microscope was used to capture images with the same exposure time and blindly analyzed the average fluorescence intensity of MBP-positive cells in the hippocampi and surrounding callosum of rats using Nis-element BR3.2 software.

The hippocampi of rats at different time points (P14, P28, and P60) were rapidly dissected and immediately stored in liquid nitrogen. The protein concentration was determined using a Bio–Rad protein assay kit (Bio–Rad Laboratories, USA). Samples containing 30 μg of total protein were loaded onto 12% sodium dodecyl sulfate (SDS)-polyacrylamide gels and electrophoretically separated. The proteins were then transferred to polyvinylidene difluoride membranes (0.22 μm, Millipore Corp, Billerica, MA, USA). The membranes were blocked for 1 h in 5% bovine serum albumin before incubation overnight at 4°C with the following specific primary antibodies: anti-MBP (Cat. No: SMI-99P, mouse monoclonal antibody, Biolegend, San Diego, CA, 1:1,000), anti-extracellular signal-regulated kinases 1 and 2 (Cat. No:4695, anti-ERK1/2, rabbit monoclonal antibody, Cell Signaling Technology, Massachusetts, USA, 1:2,000), anti-phospho-ERK1/2 (Cat. No:4370, anti-p-ERK1/2, rabbit monoclonal antibody, Cell Signaling Technology, Massachusetts, USA, 1:2,000) and anti-β-actin antibody (Cat. No:700068, mouse monoclonal antibody, horseradish peroxidase conjugate, Zen Bioscience, Chengdu, 1:1,000). The membranes were incubated with the corresponding secondary antibody (Cat. No: ZB-2305/ZB-2301, goat anti-mouse/rabbit IgG; ZSGB-BIO, Beijing, China, 1:5,000) for 90 min. Protein bands were visualized using clear western electrochemiluminescence (ECL) substrate (Bio–Rad). The protein bands were analyzed by densitometry using a Bio–Rad imaging system and Image Lab software (n = 6 in each group).

The rats (n = 3 in each group) were transcardially perfused with 0.9% sodium chloride and 2.5% glutaraldehyde in 0.2 M phosphate-buffered saline (PBS). Under an anatomical microscope, 1 mm³ tissue slabs were sampled from the CA1 region of the hippocampus, which was separated from the cerebrum, and fixed in 4% dedicated glutaraldehyde for 2 h at 4°C. Next, the samples were osmicated in osmium acid in 0.1 M PBS for 2 h at 4°C. The tissues were cut into single 60 nm sections using an ultramicrotome before being stained with uranyl acetate for 15 min and lead citrate for 2 min. A transmission electron microscope (Hitachi-7500P, Phillips, Ltd., Holland) was used to observe the ultrastructural morphological changes in hippocampal myelin. For each section, three fields of vision were randomly chosen and photographed at a magnification of 40,000 × .

The g-ratio, defined as the axon diameter (AD) to the nerve fiber diameter (FD), is a reliable index of myelination that is independent of axonal diameter (13, 14). We used g-ratios to evaluate the thickness of myelin sheaths. The diameters of the nerve fibers and axons were measured by hand tracing using ImageJ software, and then the areas from which the diameters were derived were calculated. Nine sections were randomly selected from each rat for analysis.

SPSS 17.0 software was used to analyze the data. All data are presented as the mean ± standard error of the mean (SEM). The results in the Ctl and AMA groups were compared and analyzed using an independent samples t-test at the same time point. For multiple comparisons, one-way analysis of variance (ANOVA) with Dunnett's T3 test was used. Statistical significance was considered if p < 0.05.

We first analyzed the offspring's learning and memory function by novel object recognition behavioral experiments (26). As shown in Figure 1A, compared with the Ctl group, offspring born to AMA rats had worse learning and memory function in the early postnatal period. We observed no significant difference in the number of rats touching the new object (Ctl vs. AMA: 8.33 ± 2.08 vs. 5.33 ± 1.11; t = 1.47, P = 0.17; Figure 1B). However, a significant difference was detected over time (Ctl vs. AMA: 14.43 ± 2.43 vs. 7.33 ± 1.62; t = 2.42, P = 0.036; Figure 1C) and discrimination ratio (Ctl vs. AMA: 1.21 ± 0.22 vs. 0.52 ± 0.10; t = 2.63, P = 0.025; Figure 1D) to explore new objects in Ctl and AMA offspring. Compared with the Ctl group, AMA offspring reflected a worse ability to explore new objects.

MBP is one of the major proteins in compact myelin. To identify whether AMA can affect myelin sheath development in offspring, we labeled cells with MBP in the hippocampus with immunofluorescence, and the relative intensity was quantified to indirectly measure the myelin level. Normally, the myelin sheath matures gradually from childhood to adulthood, as shown in Figure 2A. Expression was initiated in the pleomorphic layer of the hippocampal CA1 region and corpus callosum on P14. Gradually, MBP appeared in the stratum lacunosum-moleculare and fimbria of the hippocampus on P28, and MBP was expressed in the corpus callosum, subcortex, and around the ventricle by P60. The expression site did not differ significantly between the AMA and Ctl groups.

Figure 2B, a representative graph on P28 of Ctl, indicates that MBP is typically expressed as fibrous strips in CA1 area and callosum which radiating to the subventricular white matter area. For the reason, the fluorescence intensity of MBP-positive in CA1 and surrounding callosum was calculated. As shown in Figure 2C, the intensity of MBP-positive cell colonies increased with age in the corpus callosum around the hippocampus, peaking on P60 in both the AMA (1,535.20 ± 10.57) and Ctl (1,458.41 ± 63.11) groups. The results showed that the fluorescence intensity of MBP was a lower trend in the AMA group than in the Ctl group at each time point, but the difference was not statistically significant (t = 1.2, P = 0.31).

To enhance the evidence of AMA on myelin sheath development deficits in offspring, we subsequently conducted western blotting to detect the MBP expression level. As shown in Figure 3A, the hippocampal expression of MBP in the AMA group was lower trend than that in the control group at all timepoints. In addition, as shown in Figure 3B, a significant difference in the MBP/actin expression ratio was observed in the hippocampus on P14 after birth relative to those in the Ctl group (Ctl vs. AMA: 1.68 ± 0.27 vs. 0.43 ± 0.55, t = 4.5, p = 0.005 on P14), indicating that myelin sheath damage was initiated in the immature brains of offspring rats during the early phase after birth and that the severity of this damage increased throughout the developmental period into adulthood.

The hippocampal myelin sheath in rat offspring on P14 and P60 was observed using TEM. As previously mentioned, the critical time point for myelin formation in rats is 14 days after birth, and myelination is largely complete 60 days after birth. The myelin sheath in the Ctl groups contained neural fibers with regular thickness and a dense structure. However, the myelin in the AMA group showed varying degrees of stratification, vacuolation, collapse, disruption, and an irregular shape (Figures 4A–D) at days 14 and 60 after birth.

Myelin sheath thickness in different groups was measured by determining the g-ratios (Figures 4E–G). ANOVA revealed a significant difference in the g-ratio between the groups (F = 4.405, P = 0.006). The g-ratio in the Ctl group on P14 (0.74 ± 0.01) was significantly lower than that in the AMA group (0.81 ± 0.01). However, the g-ratio in the AMA group on P60 (0.79 ± 0.02) was not significantly different from that in the Ctl group (0.78 ± 0.02). This indicated that the myelin in the offspring of mothers with advanced maternal age was thinner during the early postnatal period after birth.

To determine whether the ERK pathway mediates AMA-induced injury to the myelin sheath in offspring, the protein expression levels of ERK1/2 and p-ERK1/2 in the hippocampus were measured using western blot analysis (Figure 5). The ratio of the p-ERK1/2/ERK1/2 protein expression intensities was used to represent the level of ERK phosphorylation.

In the control group, the p-ERK1/2 protein expression level dramatically peaked at 14 days, and then decreased with increasing age, reaching a relatively low level in adulthood. It is interesting to note that the p-ERK1/2 expression in the AMA group significantly lower than that in Ctl group at 14 days (Ctl vs. AMA: 0.54 ± 0.13 vs. 0.11 ± 0.02, t = 3.26, P = 0.021). The p-ERK1/2 expression at 28 days postnatal was sharply increased in the AMA group and was statistically different than the control group (Ctl vs. AMA: 0.17 ± 0.03 vs. 0.80 ± 0.18, t = −3.43, P = 0.017). However, the difference in p-ERK1/2 expression at 60 days was not significantly different between the AMA and control groups (Ctl vs. AMA: 0.36 ± 0.09 vs. 0.36 ± 0.09, t = −1.46, P = 0.16). These data indicate that AMA results in abnormal levels of ERK phosphorylation in the immature brains of rat offspring.