A Peripheral Mechanism of Depression: Disturbed Intestinal Epithelial Per2 Gene Expression Causes Depressive Behaviors in Mice with Circadian Rhythm Disruption via Gut Barrier Damage and Microbiota Dysbiosis

To investigate whether CRD promotes the development of depression, we first established a CRD mouse model through subjecting the mice to a weekly 6 h light–dark (LD) cycle of backward shifting for 2 months (Figure1A). The successful establishment of the model was confirmed by electroencephalogram (EEG) detection^[18^] (Figure S1A–C, Supporting Information) and rectal temperature changes (Figure S1D, Supporting Information). The CRD mice exhibited significantly different power spectral density changes and rectal temperatures compared to the control group. At the end of the intervention, classic behavior tests of depression in rodents including sucrose preference test (SPT), tail suspension test (TST), and forced swimming test (FST) were conducted to evaluate the depressive states of the CRD mice. The results showed that the CRD mice exhibited reduced sucrose intake compared to the control mice, indicating anhedonia (Figure 1B). In both TST and FST, CRD mice displayed an increased immobility time ratio compared to the control group (Figure 1C,D), indicating the presence of despairing behaviors in mice. Additionally, analysis of the intestinal epithelial rhythm genes revealed an abnormal expression of the Per2 gene, which showed a complete loss of normal rhythmicity (Figure 1E). Previous studies have documented that CRD increases intestinal epithelial barrier permeability and leads to gut microbiota dysbiosis in mice.^[19^] In our study, comparative analysis of bacterial abundance and structure between the two groups revealed significant alterations (Figure 1F). A decrease in the abundance of Lachnospiraceae, Prevotellaceae_UCG_001, and Bacteroides was observed, along with an increase in Muribaculaceae, Turicibacter, and Allobaculum. 3D principal component analysis (PCA) analysis also illustrated a clear distinction between the two groups (Figure 1G). Furthermore, Spearman's correlation analysis between the abundances of the differentially abundant taxa and the outcomes of three classical depressive behavioral paradigms (SPT, TST, and FST) showed that the vast majority of these microbial changes (at both the phylum and genus levels) were robustly correlated with the severity of depressive‐like behaviors (Figure S2, Supporting Information). Collectively, these findings suggest that disruption of circadian rhythm alters the expression of intestinal epithelial Per2 gene, leads to gut microbiota dysbiosis, and induces depression‐like phenotypes in mice; and CRD‐related gut microbiota dysbiosis may contribute to the development of depression.

The rhythmic disruption of intestinal epithelial Per2 gene indicates a possible involvement of this gene in CRD‐induced disturbances, we thus explored whether intestinal epithelial‐specific Per2 deletion influences the CRD‐induced depressive phenotype. To this end, we generated intestinal epithelial‐specific Per2 knockout mice (Per2^−/−) by crossing Per2^fl/fl mice with Villin1^Cre mice. The specific ablation of Per2 protein exclusively in the intestinal epithelium was confirmed by Western blotting and immunofluorescence (Figure S3, Supporting Information). Subsequently, these Per2^−/− mice and their control littermates were subjected to the CRD procedure (Figure2A,H). Notably, deletion of Per2 in intestinal epithelial cells effectively prevented the CRD‐induced depressive‐like phenotype, as evidenced by the restoration of SPT (Figure 2B) and a decrease in the immobility time ratio (%) measured by the TST and FST (Figure 2C,D). There were no changes in locomotor ability among the four groups of mice, as indicated by the lack of significant differences in total distance and average speeds (Figure S4A,B, Supporting Information). Tight junction proteins are crucial molecules for maintaining the integrity of intestinal barrier, and a lower protein level of the tight junction protein Occludin is associated with increased intestinal leakiness.^[19^] We subsequently examined intestinal barrier proteins and found that the expression of Occludin and Claudin significantly decreased in the colon of CRD mice, and this decrement was not observed in Per2^−/− + CRD mice when comparing to Per2^−/− mice (Figure 2E,F and Figure S5 (Supporting Information)). Concurrently, Per2 deletion largely modified the overall structure of the gut microbiota and mitigated the CRD‐induced alterations. The gut microbiota structure was comparable between Per2^−/− and Per2^−/− + CRD mice (Figure 2G). Collectively, these results suggest that intestinal epithelial deletion of Per2 is sufficient to prevent the CRD‐induced depressive phenotype, as well as the associated barrier impairment and gut microbiota dysbiosis.

When epithelial barriers in the intestinal or respiratory tract are compromised, the integrity of other barriers, such as the blood–brain barrier (BBB), may also be disrupted. Defects in these barriers can lead to the recruitment and activation of immune cells, and subsequent inflammation.^[20^] We thus investigated whether neuroinflammation and disruption of blood–brain barrier occurred in CRD mice. Microglia, the resident immune cells in the brain that regulate neuroinflammation, were examined for their morphology and characteristics. The number of microglia in the dentate gyrus (DG) region of the hippocampus significantly increased in CRD mice (Figure3A–E), with a typical morphology of activation (Figure 3F,G). Detection of blood–brain barrier proteins showed significant decreases in Occludin and Claudin levels in hippocampal tissues of CRD mice, indicating blood–brain barrier damage (Figure 3K). Notably, intestinal epithelial Per2 deletion protected the mice from most of these above disruptions induced by CRD. Neuroinflammation has been shown to disrupt adult neurogenesis.^[21^] To elucidate the mechanism underlying CRD‐induced depression, we identified newborn immature neurons through immunofluorescence staining with an antibody against Doublecortin (DCX) in the DG area. The results revealed that DCX‐labeled cells were significantly reduced in the hippocampal DG of CRD mice, whereas no significant difference in neurogenesis was observed between the Per2^−/− + CRD and Per2^−/− group (Figure 3H–J). These findings suggest that CRD‐induced neuroinflammation and deficits in neurogenesis may underlie the phenotype of depression, with the protective effect of Per2 intestinal epithelial deletion potentially achieved, in part, by preserving normal BBB barrier function and neurogenesis.

Impaired synaptic transmission has been widely recognized as a cellular basis for the pathogenesis of depression.^[22^] We further investigated whether CRD would affect excitatory and inhibitory synaptic transmission in the hippocampal DG region. To this end, we recorded miniature inhibitory postsynaptic currents (mIPSCs) and miniature excitatory postsynaptic currents (mEPSCs) in the DG using the whole‐cell patch‐clamp technique. Our results demonstrated a reduction in both the frequency and amplitude of mEPSCs (Figure 3L,M), while mIPSCs remained unaffected (Figure 3N,O) in the DG region of CRD mice. Notably, this impairment induced by CRD was alleviated in Per2^−/− mice, suggesting that CRD may disrupt the function of excitatory synaptic transmission, which can be rescued by conditional knockout of intestinal epithelial Per2.

To investigate whether changes in the gut microbiota contribute to behavioral alteration in CRD mice, we conducted fecal microbiota transplantation (FMT) experiments, as illustrated in Figure4A. Interestingly similar to CRD mice, the recipients of CRD microbiota exhibited a decreased sucrose preference in the SPT (Figure 4B) and an increased immobility time ratio (%) in the TST and FST (Figure 4C,D) comparing to con microbiota recipient mice. There were no changes in the locomotor ability among the four groups of mice (Figure S4C,D, Supporting Information). We further examined neuroinflammation and neurogenesis. The results indicated that the number of microglia in the DG region was significantly increased in both CRD and CRD microbiota recipient mice (Figure5A–E). Consistent with the changes in CRD mice, hippocampal barrier protein levels of Occludin and Claudin were reduced in mice transplanted with CRD microbiota (Figure 5I). Additionally, DCX‐labeled cells were significantly reduced in hippocampal DG of these groups (Figure 5F–H), and a reduction in both the frequency and amplitude of mEPSCs (Figure 5J,K) but not of mIPSCs (Figure 5L,M) in the DG region was observed. Both Con‐FMT and CRD‐FMT recipient mice were pretreated with antibiotics before the fecal microbiota transplantation. To rule out potential confounding effects of the antibiotics, a separate cohort of mice receiving antibiotics alone was subjected to a comprehensive assessment. The analysis revealed no evidence of compromised intestinal barrier integrity, neuroinflammation, impaired neurogenesis, or neuronal dysfunction (Figures S6 and S7, Supporting Information). Collectively, these results suggest that gut microbiota transplantation from CRD to normal mice induces blood–brain barrier damage and neuroinflammation, impairs hippocampal neurogenesis and neurological function. Gut microbiota dysbiosis induced by CRD plays a key role in mediating BBB damage, neuroinflammation, and subsequent deficits in neurogenesis and synaptic function.

Rifaximin is a nonabsorbable antibiotic, which can regulate the structure of the gut microbiome, protect intestinal barrier, and reduce gut‐derived inflammation.^[19, 23^] To further identify that gut microbiota dysbiosis is upstream of above pathological changes, we treated CRD mice with rifaximin through gavage during the last week of model establishment. As demonstrated in Figure S8A (Supporting Information), rifaximin effectively mitigated the depression‐like phenotype resulting from rhythmic disturbance, as evidenced by SPT, FST, and TST results (Figure S8B–D, Supporting Information), accompanied by no significant change in motor ability among the four groups (Figure S8E,F, Supporting Information). At the same time, rifaximin effectively protected against CRD‐induced damage to both the gut barrier and blood–brain barrier, as well as subsequent central neuroinflammation, impaired neurogenesis, and functional impairment in the hippocampal DG region (Figures S8G,H and S9, Supporting Information).

Gut microbiota exerts potent regulation on physiological function of host through multiple pathways, one of which is metabolites produced or modified by bacteria.^[12^] To investigate how gut microbiota remodeling in CRD mice causes neuroinflammation and impairments in neurogenesis and synaptic function, fecal metabolome profiling was performed at the end of two months of CRD. The metabolome analysis indicated that significant metabolites from the gut microbiota of CRD mice was predominantly enriched in lipid and amino acid metabolism, with tryptophan metabolism being particularly downregulated compared to that of control mice (Figure6A,B), this indicates a downregulation of the tryptophan‐related pathway in CRD‐exposed mice. Meanwhile, to elucidate the functional ramifications of the CRD‐induced dysbiosis, we conducted a predictive metagenomic analysis using Phylogenetic Investigation of Communities by Reconstruction of Unobserved States 2. This approach pinpointed that pathways pertaining to tryptophan metabolism were the most prominent functional signature distinguishing the microbial profile of the CRD group (Figure S10A,B, Supporting Information). Subsequently, we examined the levels of tryptophan and 5‐hydroxytryptophan (5‐HTP) in serum. Consistent with the metabolome analysis results, a significant decrease in tryptophan and 5‐HTP levels in the serum of CRD mice and mice transplanted with CRD microbiota was observed (Figure 6C,D), and Spearman's correlation analysis showed a significant correlation between metabolites (tryptophan (Trp) and 5‐HTP) and the differential microbiota (Figure S10C, Supporting Information). An increase in serum Tumor Necrosis Factor‐alpha (TNF‐α) and interleukin‐6 (IL‐6) levels (Figure 6E,F) was also observed in CRD mice and mice transplanted with CRD microbiota. Detection of tryptophan and 5‐HTP levels in the hippocampus revealed alterations similar to those observed in serum (Figure 6G,H). Notably, intestinal epithelial Per2 deletion mitigated these changes. Taking together, these results strongly suggest that in CRD mice, disturbed Per2 expression in intestinal epithelium causes intestinal barrier damage and gut microbiota dysbiosis, the latter, contributes to the downregulation of tryptophan and 5‐HTP in serum and brain, which may highly correlate to systemic/central inflammation and neuronal dysfunction.

To further confirm the key role of disturbed tryptophan metabolism in mediating the CRD‐induced pathological changes and depressive‐like phenotype, as illustrated in Figure7A, we tested the effects of tryptophan depletion in normal mice and tryptophan supplementation in CRD mice. The results showed that tryptophan supplementation effectively reverses the CRD‐induced depressive phenotype, as evidenced by an increased sucrose preference in the SPT (Figure 7B) and decreased immobility time ratio (%) in both the TST and FST (Figure 7C,D) in the CRD + Trp group compared to the CRD group. There were no significant changes in the locomotor ability among the four groups of mice (Figure S4E,F, Supporting Information). On the contrary, tryptophan depletion severely impacted the survival and health status of the mice, as evidenced by a mortality rate of up to 75% (Figure 7E) and only half of the body weight of surviving mice relative to normal mice at 9 weeks of age (Figure 7F). This may be attributed to the fact that tryptophan is an essential amino acid that cannot be synthesized by the body, further underscoring its significance for host health. We detected the depressive phenotypes in the surviving mice at the end of 9 weeks tryptophan depletion, and found that these mice exhibited significant depressive‐like behaviors compared with controls. These results reinforce the role of tryptophan metabolism defect in mediating depression in CRD mice. We subsequently examined serum levels of tryptophan and 5‐HTP and found that tryptophan supplementation mitigated the decrease in both tryptophan and 5‐HTP levels (Figure 7G,H) as well as the elevation of the inflammatory factor TNF‐α and IL‐6 (Figure 7I,J) induced by CRD.

Next, we explored whether tryptophan intervention could reverse the pathological changes in the central nervous system induced by CRD. Examination of central inflammation revealed that the significant increase in the number of microglia in the DG area, induced by CRD, was reversed in the CRD + Trp group (Figure8A–E). Consistent with these findings, direct inhibition of active microglia by tryptophan, which was manifested by change in cell morphology and decreases in proinflammatory cytokines, was observed in lipopolysaccharide (LPS)‐stimulated microglia (Figure S11, Supporting Information). These data suggest that tryptophan deficiency may play a key role in CRD‐induced neuroinflammation. With the inhibition of neuroinflammation, the impairment of neurogenesis, as indicated by decreased number of DCX⁺ cells in hippocampal DG area (Figure 8F–H), and the impairment of BBB integrity, as indicated by decreased barrier protein Occludin and Claudin, were also rescued by tryptophan supplementation (Figure 8I). At last, mice in CRD + Trp group showed relatively normal frequency/amplitude of mEPSCs and mIPSCs (Figure 8J–M), indicating a rescue of neuronal function. These results collectively suggest that tryptophan supplementation is an effective strategy to ameliorate blood–brain barrier damage and neuroinflammation, thereby restoring hippocampal neurogenesis and neurological function, as well as the depressive phenotype induced by CRD.

C57BL/6J mice (male, 8 weeks, 22 ± 0.5 g) were purchased from the Experimental Animal Center of Tongji Medical College, Huazhong University of Science and Technology, and maintained under standard laboratory conditions (temperature: 22 ± 2 °C, humidity: 55 ± 5%) with unrestricted access to food and water. Mice in control groups had a consistent 12 h light:12 h dark cycle, with lights on at 8:00 AM and off at 8:00 PM daily. Circadian rhythm disruption was induced by implementing a weekly 6 h phase shift in the light and dark cycle over a period of 8 weeks.^[19^] For rifaximin treatment, rifaximin (Sigma‐Aldrich) was dissolved in corn oil (37.5 g L⁻¹) and administrated to the mice through gavage (250 mg kg⁻¹ per day) daily in the last week of the experiment. The dosage and treatment duration were based on our previous study.^[19^] Control mice were given the same volume of corn oil. The tryptophan dietary intervention was performed by giving feeds that removed tryptophan as well as feeds that contained 0.5% tryptophan throughout the whole experiment. All animal experiments were approved by the Animal Care and Use Committee of Huazhong University of Science and Technology (IACUC Number: 4651), and performed in compliance with the NIH Guide for the Care and Use of Laboratory Animals.

EEG signals were amplified using a Microelectrode AC amplifier model 1700 (A‐M Systems, USA) and digitized at a sampling rate of 500 Hz with a PCIe 6323 data acquisition board (National Instruments, USA). The recordings were conducted using Spikehound software (Neurobiological Instrumentation Engineer, USA). The raw EEG signals were analyzed using multitaper methods from the Chronux toolbox (version 2.1.2, Jarvisbio, Wuhan) within MATLAB 2016a (MathWorks, UK). Briefly, raw EEG data underwent band‐pass filtering (1–80 Hz) and band‐block filtering (48–52 Hz) to eliminate line noise. The analysis was conducted using a window size of 4 s (50% overlapping) within the frequency range of 0.5–50 Hz utilizing a 5‐taper fast Fourier transform. The average power spectral density and normalized power spectral density were calculated for the delta (0.5–4 Hz), theta (4–8 Hz), alpha (8–15 Hz), beta (15–25 Hz), and gamma (25–50 Hz) bands. The normalized power spectral density analysis was performed on the data from the control and CRD groups.

In the open field test, the animals were placed in a 50 cm × 50 cm × 50 cm plastic container for 5 min. The floor of the container was divided into 25 sectors, arranged in a 5 × 5 grid. The central 3 × 3 sectors were designated as the “center area,” while the remaining sectors were classified as the “peripheral area.” The container was cleaned with 75% ethanol between each habituation period. The following parameters were recorded: total distance (mm) and average speed (mm s⁻¹). All mice were acclimated in the testing room for at least for 2 h before the commencement of the testing session.

In this experiment, mice were singly housed and acclimated to two bottles, one containing a 1% sucrose solution and the other containing drinking water for a duration of 12 h. The positions of the bottles were exchanged every 12 h to prevent the development of a position preference. Following a 12 h water deprivation period, the mice were exposed to the 1% sucrose solution and drinking water for another 12 h during the dark phase. Sucrose preference was calculated using the following formula: [sucrose consumption/(sucrose consumption + water consumption)] × 100%.

In the tail suspension test, the posterior third of the mouse's tail was secured with tape and suspended from a bracket, with the head positioned ≈15 cm above the table. All animals were suspended for a total of 5 min; during which the immobility time defined as the absence of any body or limb movements, aside from those associated with respiration was recorded in seconds during the final 5 min.

In the forced swim test, mice were individually placed into a transparent plastic container (18 cm in diameter and 25 cm in height) filled with water (25 ± 2 °C) to a depth of 15 cm, where they had no means of escape or contact with the bottom. The duration of immobility, defined as the time during which the animal did not exhibit escape responses, was recorded over a 5 min session. Subsequently, the animals were removed from the container and allowed to dry in a heated enclosure before being returned to their home cages. The cylinder was emptied, cleaned, and refilled with fresh water between each mouse.

Recipient mice were given an antibiotic cocktail^[57^] including vancomycin (0.5 g L⁻¹), ampicillin (1 g L⁻¹), neomycin (0.5 g L⁻¹), and metronidazole (0.5 g L⁻¹), administered in their drinking water for a duration of 7 consecutive days. All antibiotics were obtained from Sigma‐Aldrich. Fresh fecal transplants were pooled from Con and CRD donor mice, respectively. The homogenates were subsequently filtered through a 20 µm pore nylon filter to eliminate large particulate and fibrous matter. Following antibiotic treatment, the mice were orally challenged with 200 µL of fecal transplants (≈2 × 10⁸ viable probiotic bacteria dissolved in sterile phosphate‐buffered saline (PBS)) via gavage over a period of 28 consecutive days. The animals were colonized through multiple rounds of oral gavage with microbiota. Behavioral testing of the mice was conducted after the intervention.

Fresh fecal samples were collected at the end of the experiment for the preparation of a 16S rRNA gene amplicon library and subsequent sequencing to analyze gut microbiota. To assess bacterial diversity analysis, the V3–V4 variable regions of the 16S rRNA genes were amplified using universal primers 343F and 798R. The amplicon pools were prepared for sequencing, and the size and the quantity of the amplicon library were evaluated using a Qubit 3.0 and an Agilent 2100. Sequencing was conducted on Illumina's high‐throughput sequencing platform. Reads were screened for low‐quality bases and short read lengths, then assembled and assigned to operational taxonomic units with a similarity threshold of 97%. Alpha (α) and beta (β) diversity analyses were performed using the QIIME tool (version 1.9). Alpha diversity was assessed using the Chao 1 index to evaluate the complexity of microbial community composition within the samples. Beta diversity was determined using weighted UniFrac phylogenetic distance matrices, which were visualized through principal component analysis and analyzed via Analysis of Molecular Variance. The relative abundance of species at the genus level was presented to illustrate the community structure across different groups.

The nontargeted metabolomics procedure was performed by electrospray ionization‐Q‐time of flight (TOF)/mass spectrometry (MS) (Xevo 121 G2‐S Q‐TOF, Waters) and UPLC‐QTOF/MS (ACQUITY UPLC I‐Class, Waters). Briefly, cecum samples (100 mg) were dissolved in 500 µL of ice‐cold water, mixed using a vortex, and centrifuged for 15 min at 12 000 rpm. Then, the supernatant was obtained, and the remaining precipitate was further extracted with 500 µL of ice‐cold methanol. The two fecal extracts were combined and centrifuged at 12 000 rpm for 15 min and the supernatant was stored at 4 °C; 10 µL of the supernatant was used for analysis. The MS data of cecum samples were first processed by MarkerLynx (version 4.1, Waters, Milford, MA, USA). The procedure included integration, normalization, and peak intensity alignment. In the positive data set, a list of m/z and retention time with corresponding intensities was provided for all metabolites in every sample. Then, the processed data set was then entered into the SIMCA‐P software package (v13.0, Umetric, Umeå, Sweden). The normalized data were then used to perform analysis.

Mice were anesthetized with sodium pentobarbital (45 mg kg⁻¹) via intraperitoneal injection and subsequently intracardially perfused with an ice‐cold cutting solution composed of (in mm, 209.0 sucrose, 3.1 sodium pyruvate, 22.0 glucose, 1.25 NaH₂PO₄, 12.0 sodium l‐ascorbate, 4.9 MgSO₄‐7H₂O, and 26.0 NaHCO₃, which was aerated with 95% O₂ and 5% CO₂, pH 7.2–7.4). Coronal slices (300 µm thick) of the DG of the hippocampus were prepared using a vibratome (Leica VT1200 S, Leica Biosystems) and incubated in artificial cerebrospinal fluid consisting of (in mm, 128.0 NaCl, 3.0 KCl, 24.0 NaHCO₃, 2.0 MgCl₂, 1.25 NaH₂PO₄, 10.0 d‐glucose, and 2.0 CaCl₂, which was oxygenated with 95% O₂ and 5% CO₂, pH 7.2–7.4, 295–305 mOsm) at 28 °C for 1 h. After that, the slices were returned to room temperature for whole‐cell patch‐clamp recordings. Neurons were voltage‐clamped at −70 mV in voltage‐clamp mode to record mEPSCs or mIPSCs. For mEPSC recording, patch electrodes were filled with an internal solution (in mm): 122.5 cesium gluconate, 17.5 CsCl, 0.2 ethylene glycol‐bis(b‐aminoethylether)‐N,N,N9,N9‐etraacetic acid (EGTA), 10.0 N‐(2‐hydroxyethyl)piperazine‐29‐(2‐ethane‐sulfonic acid) (HEPES), 1.0 MgCl₂, 4.0 magnesium ATP, 0.3 sodium GTP, and 5.0 QX314, pH 7.25 and 280–300 mOsm. Tetrodotoxin (TTX, 10 µm) and bicuculline (20 µm) were included during mEPSC recordings. For mIPSC recordings, pipettes (4–6 MΩ) were filled with an internal solution comprising (in mm) 153.3 CsCl, 1.0 MgCl₂‐6H₂O, 5.0 EGTA, 4.0 magnesium ATP, and 10.0 HEPES, adjusted to pH 7.25 with CsOH (280–300 mOsm). mIPSCs were recorded in the presence of TTX (10 µm) and CNQX (10 µm). All data were obtained by pCLAMP 10 (Axon Instruments, Molecular Devices, San Jose, CA) and the MultiClamp 700B amplifier (Molecular Devices, Sunnyvale, CA). Records were low‐pass filtered at 2–20 kHz and digitized at 5–50 kHz (Molecular Devices, Jarvisbio, Wuhan, China).

Brain tissue was homogenized using RIPA buffer (Beyotime Biotechnology, Shanghai, China) containing phenylmethylsulfonyl fluoride (1:100) and a proteinase inhibitor cocktail (1:100). The homogenates were then boiled for 10 min, and centrifuged at 14 000 g for 10 min. The supernatants were collected, and protein concentrations were quantified using a BCA kit (Thermo Fisher Scientific, Waltham, MA, USA). Proteins from the extracts were separated by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and subsequently transferred to a nitrocellulose membrane. The membranes were blocked with 5% nonfat milk for 1 h and incubated overnight at 4 °C with primary antibodies: anti‐Occludin (#27260‐1‐AP, Proteintech), anti‐Claudin (#MA5‐27605, Thermo Scientific), and anti‐GAPDH (#ab64613, Abcam). Then, the membranes were washed 3 times with TBST for 10 min each and incubated with the secondary antibody at room temperature for 1 h in the dark. Blots were visualized using the Odyssey Infrared Imaging System (Li‐Cor Biosciences, Lincoln, NE, USA), and the protein bands were quantitatively analyzed using ImageJ software (Rawak Software, Germany).

Mice were perfused with a 0.9% saline solution followed by 4% paraformaldehyde in PBS. The brains were postfixed for 24 h in 4% paraformaldehyde and subsequently immersed in a gradient sucrose solution (15–30%) for 3 days at 4 °C. Cryosections (30 µm thick) were incubated in 5% bovine serum albumin and 0.3% Triton X‐100 for 1 h at room temperature, followed by overnight incubation with primary antibodies at 4 °C. Primary antibodies were used as follows: anti‐Iba1 (#019–19741, Wako), anti‐DCX (#91954S, Cell Signaling Technology), and anti‐NeuN (#12943, Cell Signaling Technology). Subsequently, the slices were rinsed and incubated with the appropriate Alexa dye‐tagged secondary antibodies: donkey anti‐mouse 488 (#715‐546‐151, Jackson Immuno Research Labs) and donkey anti‐rabbit 594 (#715‐585‐152, Jackson Immuno Research Labs) for 1 h at room temperature. The slices were then mounted and air‐dried in the dark. All images were observed using the LSM800 confocal microscope (Zeiss, Germany).

Data were analyzed using GraphPad Prism version 8.0 and presented mean ± standard error of the mean (SEM). To assess the normality of the data, a Kolmogorov–Smirnov test was conducted. For data exhibiting a normal distribution, differences between groups were evaluated by Student's t‐test or one‐way analysis of variance (ANOVA). Two‐way repeated‐measures ANOVA followed by Bonferroni's post hoc test, was employed to analyze differences in Sholl analysis. The study was not preregistered; no formal randomization methods were applied, and no blinding was performed. A post‐hoc power analysis was conducted using G*Power to validate the sample size.^[58^] The statistical test method for each figure was indicated in the figure legend. Statistical significance was defined as *p < 0.05, **p < 0.01, ***p < 0.001, and****p < 0.0001. Mice in poor condition (not in health condition and eventually deceased) were excluded from the study.

The data that support the findings of this study are available from the corresponding author upon reasonable request.