Pregnant C57BL/6 mice at gestational day 7 (G7) were purchased from Ziyuan Experimental Animal Technology Co. (Hangzhou, China). The mice were housed in a humidity-and temperature-controlled animal room with free access to food and water. The room was maintained on a standard 12:12-h light/dark cycle, with lights on at 07:00 and off at 19:00. The animals were treated in accordance with protocols approved by the Animal Ethics and Welfare Committee of Zhejiang University School of Medicine. All experimental procedures were carried out in accordance with the National Institutes of Health (NIH) Guide for Care and Use of Laboratory Animals (Publication no. 86–23).

Pregnant mice were randomized into a PPD group, a PPD + EGCG group and a control group. The gestational stress-induced PDD model was established as previously described (Kiryanova et al., 2016), with some modifications. From G7 to G16, mice were subjected to a regimen of chronic unpredictable mild stress (CUMS; Table 1; Figure 1B). Stressors included restraint stress (mice were restrained in conical tubes with holes for airflow), white noise, cage tilting (home cage tilted by 45°), fasting and water deprivation, forced swim (5 min), overnight lighting, tail pinching (2 min), housing in pairs (mice were paired with a pregnant dam in their home cage or in the pregnant dam’s cage), soiled cage (with 100 ml fresh water spilled on the bedding), and foreign object in the cage (an unusual novel plastic object). Anxiety and depression-like behaviors and maternal care were assessed about 1 week after delivery (Figure 1A). For EGCG treatment, freshly dissolved EGCG in saline was injected intraperitoneally daily for 5 days at a dose of 50 mg/kg (the first dose of EGCG was given on postpartum day 3). The EGCG was purchased from Sigma-Aldrich (St. Louis, MO, United States).

The number of pups at birth was recorded. The survival rate of the offspring was calculated by dividing the number of pups on postnatal day 8 by the number of pups that were born.

Maternal behavior was evaluated based on pup retrieval (Gerecsei et al., 2018). Five days after parturition, pups were separated from the dam and placed at the opposite end of the cage. The female mouse was allowed 10 min to retrieve the pups and the latency to retrieval of the first and all other pups was recorded. The experiment was terminated when all pups were retrieved or after 10 min.

A white opaque square box (45 × 45 × 45 cm) in a quiet room with dim lighting was used for the open field test (OFT). Each mouse was placed in the corner of the box and allowed 10 min to explore. The distance covered, amount of time spent in the center square, and number of entries into the center square were recorded.

The forced swimming test (FST) was carried out as previously described (Yang et al., 2019). Briefly, each mouse was placed in a transparent cylindrical container (diameter, 25 cm and height, 40 cm) filled with water to a height of 25 cm at a temperature of 25 ± 2°C, and was forced to swim in the container for 6 min. Immobility time during the last 4 min was recorded.

The tail suspension test (TST) was performed as previously described (Lu et al., 2021). Briefly, mice were suspended approximately 30 cm above the floor by tape placed about 1 cm from the tail tip. Each mouse was suspended for 6 min and immobility time during the test was recorded.

C-Fos expression in response to the FST was evaluated as a marker of neuronal activation (Rotllant et al., 2002; Ons et al., 2004). 90 min after the FST, mice were deeply anesthetized and intracardially perfused with phosphate-buffered saline followed by 4% paraformaldehyde solution. The whole brain was removed and postfixed overnight at 4°C in 4% paraformaldehyde before dehydration in 30% sucrose solution for at least 48 h. The brain was then serially sectioned at a thickness of 30 μm, and the sections were blocked for 2.5 h and incubated overnight at 4°C with c-Fos antibody (1:1000; Synaptic Systems, Göttingen, Germany) followed by incubation for 1 h at room temperature with secondary antibody (1:1000; Abcam, Cambridge, MA, United States). Images were acquired with a fluorescence microscope (Olympus, Tokyo, Japan; Model VS120) and analyzed with ImageJ software (NIH, Bethesda, MD, United States).

The hippocampus was dissected and frozen in liquid nitrogen and stored at −80°C. Total RNA was extracted with an RNA Isolation Kit (Vazyme, Shanghai, China) and reverse-transcribed to cDNA using reverse transcriptase (Vazyme, Shanghai, China). mRNA levels were quantified using SYBR qPCR Master Mix (Vazyme, Shanghai, China) using the primers shown in Table 2.

Mice anesthetized with 2% pentobarbital sodium solution were sacrificed and the hippocampus was removed and homogenized in radioimmunoprecipitation assay buffer containing protease and phosphatase inhibitor. The homogenate was centrifuged at 13,200 rpm for 30 min at 4°C. The supernatant was collected and protein concentration was measured with a bicinchoninic acid protein assay kit (Thermo Fisher Scientific, Waltham, MA, United States). Protein samples were electrophoretically separated on 4–12% polyacrylamide gels and transferred to a polyvinylidene difluoride membrane (Bio-Rad, Hercules, CA, United States) that was blocked in 5% nonfat milk for 2 h at room temperature and incubated overnight at 4°C with primary antibodies against Sema3A (1:1000; Abcam, Cambridge, MA, United States) and phosphorylated (p-)GSK3β, GSK3β, p-CRMP2, CRMP2, and β-actin (all 1:1000 and from Cell Signaling Technology, Danvers, MA, United States). The membrane was then incubated with horseradish peroxidase-conjugated secondary antibody (1:1000; Cell Signaling Technology, Danvers, MA, United States) for 1 h at room temperature and immunoreactivity was analyzed using ImageJ.

After behavioral tests, all mice were anesthetized with 2% pentobarbital sodium solution and blood was collected from the eyeballs. The whole blood was stored overnight at 4°C and then centrifuged at 5,000 rpm for 10 min at 4°C. The supernatant was stored at −80°C until use. Sema3A level was quantified with an enzyme-linked immunosorbent assay (ELISA) kit (CUSABIO, Wuhan, China) according to the manufacturer’s instructions.

Statistical analysis was performed using Prism v8.0 software (GraphPad, La Jolla, CA, United States). Data are presented as mean ± SEM. Comparisons were performed across 2 groups with the unpaired t test and across multiple groups by one-way analysis of variance followed by the Tukey post hoc test. Statistical differences were accepted as significant at p < 0.05.

Anxiety and depression-like behaviors in PPD model mice were evaluated with the OFT, FST, and TST (Figure 2A). Representative movement trajectories in the OFT for mice in each group were shown in Figure 2B. There was no significant difference in total distance traveled between PPD and control mice, suggesting that gestational stress had no effect on locomotion (p = 0.4010; Figure 2C). However, compared with control group, PPD mice spent less time in the center area (p < 0.05; Figure 2D) and had decreased entries into the center zone (p < 0.01; Figure 2E), suggesting that gestational stress caused postpartum anxiety-like behavior in mice. Gestational stress also increased immobility time in the FST (p < 0.01; Figure 2F) and TST (p < 0.0001; Figure 2G), indicating that chronic unpredictable mild stress during gestation induced postpartum depression-like behavior in mice.

Depressed mothers exhibit significantly more negative behaviors and disengagement toward their children compared with those without depression (Lovejoy et al., 2000). To determine whether PPD had a detrimental effect on maternal care and pup development, we recorded the litter size at birth and pup survival rate on postpartum day 8, as well as the maternal behaviors of the dam. However, there were no significant differences in offspring litter size (p = 0.8639; Supplementary Figure 1A), survival rate (p = 0.3243; Supplementary Figure 1B), or pup retrieval including latency to retrieve the first pup (p = 0.9408; Supplementary Figure 1C) and all pups (p = 0.1078; Supplementary Figure 1D) between the PPD and control groups.

To determine the involvement of the hippocampus in gestational stress-induced anxiety and depression-like behaviors, we performed immunofluorescence labeling of c-Fos after the FST (Figure 3A). The number of c-Fos–positive cells in the hippocampus was significantly decreased in the PPD group compared with the control group (p < 0.05; Figure 3B), suggesting that decreased activation of the hippocampus was related to gestational stress-induced anxiety and depression-like behaviors. Specifically, gestational stress caused a marked reduction in c-Fos levels in the CA1 (p < 0.05; Figure 3C) and CA3 (p < 0.05; Figure 3D) areas while having no effect on the c-Fos level in the dentate gyrus (p = 0.5882; Figure 3E).

An elevated level of Sema3A in the central nervous system (CNS) including the hippocampus and cerebellum has been observed in various neurologic and psychiatric diseases (Eastwood et al., 2003; Good et al., 2004; Van Battum et al., 2015), but its role in depression is not fully understood. We examined Sema3A expression in the hippocampus and peripheral blood and found that the mRNA (p < 0.05; Figure 4B) and protein (p < 0.001; Figures 4A,C) levels in the hippocampus were significantly higher in the PPD group than the control group. However, there was no difference in serum Sema3A level between groups (p = 0.5537, Figure 4D).

As GSK3β and CRMP2 were linked to major depression, we examined the mRNA and protein levels of GSK3β and CRMP2 in the hippocampus. Compared with the control group, there were no significant changes in GSK3β (p = 0.9569; Figure 4E) and CRMP2 (p = 0.3859; Figure 4F) transcript levels in PPD mice; however, gestational stress decreased the ratio of p-GSK3β to GSK3β (p < 0.05 vs. control; Figures 4A,G) in the hippocampus, indicating an increase in the expression of active GSK3β. There was no difference in phosphorylated CRMP2 level in the hippocampus between groups (p = 0.1551; Figures 4A,H).

As EGCG had anti-inflammatory, antioxidative, and neuroprotective effects in mammals (Chakrawarti et al., 2016), we investigated whether it could alleviate the anxiety and depression-like behaviors associated with PPD. The experimental timelines of drug administration and behavioral testing were shown in Figure 5A. Representative movement trajectories of mice from each group in the OFT were shown in Figure 5B. There was no difference across groups in total distance covered [F_(2,28) = 0.5797, p = 0.5666; Figure 5C], but there were significant differences in the time spent in the center area [F_(2,27) = 9.336, p < 0.001; Figure 5D] and entries into the center zone [F_(2,28) = 9.697, p < 0.001; Figure 5E]. Specifically, mice exposed to gestational stress spent less time in the center area (p < 0.01; Figure 5D) and exhibited fewer entries into the center zone (p < 0.01; Figure 5E) than controls; EGCG treatment abrogated these decreases (time in center area: p < 0.01, Figure 5D; center zone entries: p < 0.01, Figure 5E). There was also a significant difference in immobility time in the FST [F_(2,30) = 8.045, p < 0.01; Figure 5F] and TST [F_(2,29) = 9.735, p < 0.001; Figure 5G] across groups. Multiple comparisons revealed that mice in the PPD group had longer immobility time in both tests (FST: p < 0.01, Figure 5F; TST: p < 0.05, Figure 5G) than control mice. EGCG decreased immobility time in the FST (p < 0.05; Figure 5F) and TST (p < 0.001; Figure 5G) in mice exposed to gestational stress.

To clarify the molecular basis for the effects of EGCG in mice with PPD, we examined the expression of Sema3A, p-GSK3β, and GSK3β in the hippocampus. There were significant differences among groups in Sema3A mRNA [F_(2,12) = 12.25, p < 0.01; Figure 6A] and protein [F_(2,13) = 5.942, p < 0.05; Figure 6B] levels as well as p-GSK3β/GSK3β protein ratio [F_(2,9) = 12.91, p < 0.01; Figure 6C] in the hippocampus. Sema3A mRNA (PPD vs. PPD + EGCG, p < 0.01; Figure 6A) and protein (PPD vs. PPD + EGCG, p < 0.05; Figure 6B) expression was decreased by EGCG treatment, whereas the ratio of p-GSK3β/GSK3β was increased (PPD vs. PPD + EGCG, p < 0.01; Figure 6C), indicating a reduction of active GSK3β in the hippocampus by EGCG treatment. These results suggest that EGCG mitigates anxiety and depression-like symptoms in PPD mice by inhibiting Sema3A and GSK3β signaling.