Experiences Shape Hippocampal Neuron Morphology and the Balance of CRHR1 and OTR

doi:10.21203/rs.3.rs-1207767/v1

Download PDF

Research Article

Experiences Shape Hippocampal Neuron Morphology and the Balance of CRHR1 and OTR

https://doi.org/10.21203/rs.3.rs-1207767/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

The dorsal hippocampus (dHIP) is involved in avoidance behaviors regulation. It is unclear how experiences affect avoidance behaviors and the neuron morphology modulated by corticotrophin-releasing hormone (CRH) signaling and oxytocin receptor (OTR) system in the dHIP. We studied the effects of neonatal maternal deprivation (MD) and post-weaning environmental enrichment (EE) on the avoidance behaviors and neuron morphology of the dHIP in male BALB/c mice. We found that MD impaired avoidance memory, whereas EE improved that. MD increased Nissl bodies and neuronal complexity in the CA1 but decreased synaptic connections in the CA2–3 and DG. EE restored the Nissl bodies and neuron complexity alterations induced by MD in the CA1 and the synaptic connections alterations induced by MD in the CA3. MD increased CRHR1 levels in the CA1 and CA3 but decreased CRHR1 levels in the CA2. The effects of EE on the CRHR1 levels depend on whether MD or not. For MD mice, the EE treatment decreased CRH levels of the CA1–3 but only reduced the CRHR1 level of the CA3 region. MD probably increased OTR levels in the CA1, and EE could restore. Our study illustrates how MD and EE affect avoidance behaviors, hippocampal neuron morphology, and the OTR-CRHR1 balance. The results show that the alterations in dHIP’s regions induced by early life stress are partially restored by the late-life environmental improvement.

Maternal deprivation

environmental enrichment

hippocampus

oxytocin receptor

corticotrophin-releasing hormone

Avoidance behavior is the goal-directed actions of an individual to avoid danger when the sensory organs report a possible punishment (Trusel et al. 2019). This behavior is a primary ability for survival that renders individuals adapt to the complex environment with threats (Trusel et al. 2019). Evidence accumulated shows that high levels of childhood adversity impairs passive avoidance behavior in adolescence (Yazgan et al. 2021) and leads to a hippocampal volume reduction (Teicher et al. 2012). When the foot shock intensity that the conditioning parameters are relatively weak in the lab, the dorsal hippocampus (dHIP) will involve rodent avoidance behaviors (Quinn et al. 2008), functional inactivation of the dHIP will impair rats’ active place avoidance-related spatial abilities (Cimadevilla et al. 2000). However, the integrity of the dHIP is not required for the expression context-dependence of avoidance behavior (Oleksiak et al. 2021). Because of the selective vulnerability of the responsive neurons in the HIP to the various stressors or high flow of stress-induced corticosterone, the HIP's sub-regions show different responses to the various stressors such as immobilization, swim, and heat (Meena et al. 2012). It is unclear which region of the dHIP plays a critical role in the avoidance behavior induced by foot shock.

The hypothalamic-pituitary-adrenocortical (HPA) axis mediates experiences and the brain, contributing to avoidance behavior to the risk in humans and animals. Corticotrophin-releasing hormone (CRH) neurons release CRH during HPA activation. CRH binds the target organ’s receptors (CRHRs) and decreases the hippocampal glutamate release evoked by electrical stimulation (Bagosi et al. 2015). CRHRs consist of two types: CRHR1 and CRHR2. CRHR1 is rich in pyramidal neurons in the HIP and primarily serves to mediate the synaptic actions of CRH in hippocampal pyramidal neurons (Maras and Baram 2012). CRHR2 is limited in the HIP (Maras and Baram 2012). Chronic stress increases CRH levels which promote loss of the dendritic spines (Chen et al. 2012). The loss of dendritic spines and synapses begins with actin cytoskeleton collapse within the excitatory synapses on spine heads where CRHR1 receptors reside (Chen et al. 2012). Oppositely, lack of CRHR1 increases total dendritic length and promotes the dendritic branching of principal hippocampal cells (Chen et al. 2004).

Oxytocin (OT) and its receptors (OTR) play critical roles in controlling stress, energy metabolism, and social attachment (Onaka et al. 2012). OTR presents in presynaptic and postsynaptic membranes and regulates neuronal excitability (Bakos et al. 2018). OT bonds OTR and modulates synaptic transmission onto CRH neurons, by which OT suppresses CRH neuron excitability (Jamieson et al. 2017). The OT-OTR signaling in the HIP, the OT-OTR signaling regulates neuronal excitability and synaptic plasticity (Lin and Hsu 2018). Upon OTR activation in the CA1, pyramidal neurons substantially increase spontaneous inhibitory postsynaptic currents (Maniezzi et al. 2019). The OTR activation in the CA2 results in pyramidal cells depolarizing and fire bursts of action potentials (Tirko et al. 2018). The CA3 region receives the direct projections from the OT neurons in the hypothalamic paraventricular nucleus (PVN). The released OT bonds the OTR expressed in CA3 pyramidal neurons to control the hippocampal neurogenesis (Lin et al. 2017). The maturation of newly generated granule cells in the DG can be impaired by the conditional delete of OTR in hippocampal excitatory neurons (Lin et al. 2017).

Experiences can shape avoidance behaviors in humans and rodents. The intense stress of long-term maternal deprivation (MD) strengthens CRH and weakens the OT system (Bardo et al. 2021), thereby disrupting the balance and impairing inhibitory avoidance in adulthood (Vivinetto et al. 2013). However, mild stress of repeated maternal separation of daily 20 min decreases the OTR ligand binding in the HIP at postnatal day (PND) 8, and this effect is confined to the dHIP and does not persist into adulthood (Noonan et al. 1994). Environmental enrichment (EE), as a condition of prosperity, supplies the rodents with various sensory, motor, recognition, memory, and sociality stimulations. EE increases OT neurons in the parvocellular PVN (Wei et al. 2021) and improves avoidance impairment induced by MD (Vivinetto et al. 2013). However, it is still unclear how MD and EE affect the hippocampal neuron morphology modulated by CRH-CRHR1 signaling and OT-OTR system in the dHIP.

In the present study, we hypothesized that experiences affected avoidance behaviors and neuronal morphology in the HIP via disrupting the balance of CRHR1 and OTR, which modulated neurogenesis and maturation. To demonstrate the hypothesis, we used a male BALB/c mouse model of MD (PND 3–21) to induce avoidance behaviors deficit and abnormal neurogenesis in the dHIP. Meanwhile, we selected post-weaning EE (PND22–110) to remedy that. Next, we checked the mouse models’ differences in neuron morphology and synaptic connection marked by synaptophysin (SYP) and postsynaptic density protein-95 (PSD95) in the dHIP regions. Thirdly, we explored CRH, CRHR1, and OTR levels in the HIP and their correlations with the parameters of cell numbers, Nissl body levels, and Pearson’s R of SYP and PSD95 colocalization. The findings indicated that experiences affected avoidance behaviors and the neuronal morphology in the HIP. MD increased CRH, CRHR1, and OTR levels in the CA1. EE probably decreased CRH levels in the CA1 and increased CRHR1 levels in the CA3. EE did not affect the OTR level but could decrease the OTR levels in the CA1 of the MD mice. In the MD mice, it appeared that the hippocampal neurons had more dendrites in the region where both CRHR1 and OTR levels were increased. The findings may provide new insight into the neuronal morphology regulation and the individual hippocampal difference driven by experiences.

Ethical statement

The experiment protocol was approved by the Policies Governing the Use of Live Vertebrate Animals of the Lanzhou University (License number: jcyxy20191010). The animal experiments were performed under the guidance of the National Institutes of Health guide for the care and use of laboratory animals (NIH Publications No. 8023, revised 1978). We calculated the sample size and did our best to minimize the animal number for use and their suffering.

Experimental animals

The male BALB/c pups used in the study were obtained from the mating of one male and two females in each cage. The female and male BALB/c mice for mating were obtained from the Experimental Center of Lanzhou University. Animals were kept in Plexiglas cages (17 cm × 24 cm × 14 cm) with enough food and purified water in an animal house. The animal house conditions were 12 h light/12 h dark cycle, the temperature of 22 ± 2°C, the humidity of 30%–40%, and the noise level lower than 40 dB. The gestation day was determined by recording the day of the mating and checking the vaginal plug. The delivery day was estimated by observing the dam’s behaviors, the vaginal orifice, and nipples. During the gestation, the dams were fed with special food for gestation.

Experimental design

The effects of MD and post-weaning EE were studied via a two × two between-group experiment. The design yielded four groups: non-deprivation (ND), four hours per day of MD on the PND3–21, EE, and non-EE (NE) from PND22–110. We assigned the dams randomly into four groups before delivery. About eight dams in each group were kept for the study. The dams before parturition were housed alone in a cage. After two days of delivery, pups were culled to 4–5 pups per litter. Dams of MD were transferred to another room daily between 11 am to 3 pm, and pups were left in the house cages. The dams of ND were handled for 2 min per day on the PND3−21. All pups were weaned on the PND21. Six male pups were housed in a cage (17 cm × 24 cm × 14 cm) with fresh bedding and enough food and water under NE conditions. The EE rearing condition was designed as six mice were reared in an environment-enriched cage. The EE cages (43 cm × 33 cm × 20 cm) were more extensive than the NE cages and filled with various toys, including balls, tunnels, swings, confetti, and colored cloths. The toys were changed weekly. The positions of the toys were changed every three days. To reduce variation, the behavioral tests were conducted during the light phase. On the PND110, 82 male offspring from 33 litters finished the behavioral tests and were included in the study.

Behavioral tests

Passive avoidance test (PAT)

The apparatus consisted of a lit chamber and a dark chamber (15.5 cm [L] × 14.0 cm [W] × 21.0 cm [H] each) connected by a door (3.7 cm [L] × 4.0 cm [H]). The light of the lit chamber was on during testing. A 24 V current was applied to the dark chamber via a grid. A mouse was allowed to explore two chambers without light or shock for 5 min during training. The following day, the mouse was introduced to the lit chamber and habituated for 30 s before testing. During the five minutes of testing, the door was opened, and the mouse was allowed to enter the dark chamber with an electric grid. The mouse was traced by a video-tracking system (TM-vision, Chengdu Techman Software Co., Ltd., Chengdu, China). The time spent entering the dark chamber before leaving the lit chamber was recorded as the latency. The mouse received a foot shock and returned to the lit chamber. When a mouse entered the dark chamber again, the wrong times would be increased by one time. After 24 h and one week, the tests were repeated to test the memory of the mice. The tests were conducted by a trained investigator who did not know the experimental design.

Step-on test (SOT)

The apparatus consisted of eight chambers. Each chamber was 15.5 cm (L) × 14.0 cm (W) × 39.8 cm (H) in size and had a metal grid floor and escape platform (4.5 cm in diameter and 4.5 cm in height). A 24 V current was applied via a metal grid during testing. The habituation was performed on the first day. Each mouse was allowed to explore the chamber for 5 min. The next day, the mice were introduced to the chambers individually. After 30 s of habituation, the current was applied, and the activities of the mice in 5 min were recorded. We recorded the time for escaping to the platform for the first time and on the grid floor. When the mice jumped off the platform, an error would be recorded as one time. Only individuals that stepped onto the platform were used for the analysis. After 24 h and one week, the tests were repeated to evaluate the memory of the mice. The tests were conducted by a trained investigator who did not know the experimental design.

Nissl staining

Mice were anesthetized and perfused intracardially with 0.9% saline for 10 min, followed by 4% paraformaldehyde (PFA) for 10 min. Brains were collected and transferred to 4% PFA solution for 24 h, 20% sucrose for 24 h, and 30% sucrose for three days, successively. Then, the brains were sliced coronally into 25 µm sections and preserved in a protectant solution at −20℃ until used. The sections containing similar dHIP morphology were selected according to locations of bregma anterior-posterior −1.855 mm to −2.255 mm among the entire brain sections. The selected sections were mounted on gelatin-coated slides for the staining. After the sections were dried at room temperature for 24 h, the slides were immersed in 75% ethanol, 85% ethanol, 95% ethanol, 100% ethanol Ⅰ, 100% ethanol Ⅱ, 95% ethanol, 85% ethanol, 75% ethanol, and 50% ethanol, successively. Next, the slices were stained with 0.5% Cresyl violet (Solaribio, Shanghai, China) for 20 min after the slices were incubated in distilled water for 1 min. Finally, the slides were transferred to 95% ethanol, 100% ethanol, 100% ethanol, xylene Ⅰ, and xylene Ⅱ, successively before they were covered with neutral balsam. The tests were conducted by a trained investigator who did not know the experimental design.

Golgi-Cox staining

Brains were collected through perfusion intracardially with 0.9% saline for 10 minutes and 4% PFA for 5 minutes. Each brain was tied with a cotton thread, marked with a label, and immediately immersed in a Golgi-Cox solution (5 ml per brain) prepared one month ago. The Golgi–Cox solution was prepared according to the previous study (Bayram-Weston et al. 2016). After two days, the brains were transferred to a new Golgi-Cox solution. The impregnation was finished in an amber bottle at room temperature for 14 days. The brains were transferred to 0.1 M PBS for 6 h, 20% sucrose for 24 h, and 30% sucrose for 24 h in turns. Then, the brains were sectioned into 60 µm-thick slices coronally on a freezing microtome after freezing for half an hour. All sections of each brain were collected orderly and mounted on the gelatin-coated slides. The slides were dried at room temperature in the dark for 48 h. Then, the slices were blackened with 12% ammonia for 60 min to visualize the staining. A fixative agent was used to fix the staining for half an hour. Finally, dehydration, cleaning, and sealing were performed conventionally.

Antibody Characterization

The anti-OT antibody (Abcam, Cat# ab212193, RRID: AB_2895534) was a recombinant monoclonal antibody of rabbit and recognizes bands of about 12 kDa on western blots of mouse brain tissue lysate (Figs. s1a and s1b). The anti-OTR antibody (Abcam, Cat# ab217212, RRID: AB_2904223) was raised against a synthetic peptide within mouse OTR aa 332-388 conjugated to keyhole limpet hemocyanin (manufacturer’s datasheet). The anti-OTR antibody recognizes a single band of about 52 kDa on western blots of mouse brain lysates (Fig. s1c). The rabbit raised the anti-CRH/CRF antibody (Proteintech, Cat# 10944-1-AP, RRID: AB_2084279) against a CRH/CRF fusion protein Ag1371 (manufacturer’s datasheet), and the observed molecular weight was about 35 kDa on western blots of mouse brain lysates (Fig. s1d). The anti-CRHR1 antibody (Thermo Fisher Scientific, Cat# 702611, RRID: AB_2734802) was raised in rabbits against a synthetic peptide corresponding to human CRHR1 [aa425-aa444] (manufacturer’s datasheet), and the observed molecular weight was between 50 kDa and 70 kDa on western blots of mouse brain lysates (Fig. s1e). The anti-SYP antibody (Abcam, Cat# ab32594, RRID: AB_778204) was raised against a synthetic peptide corresponding to amino acids 253-272 of rat synaptophysin and detected a band of approximately 34 kDa on western blots mouse brain lysates (Fig. s1f). The anti-PSD95 antibody (Millipore, Cat# MAB1596, RRID: AB_2092365) was raised in mice against recombinant rat PSD95 (manufacturer’s datasheet), and the observed molecular weight was about 100 kDa on western blots of mouse brain lysates (Fig. s1g). The peroxidase-conjugated goat anti-rabbit lgG (ZSGB-Bio, Cat# ZB-2301, RRID: AB_2747412) and the FITC-labeled goat anti-rabbit IgG (H+L) antibody (ZSGB-Bio, Cat# ZF-0311, RRID: AB_2571576) were polyclonal antibodies targeted at rabbit IgG (H+L). The goat anti-mouse IgG coupled with CY3 antibody (1:200, AMSBIO, Cat# BA1031, RRID: AB_10890402) was a polyclonal antibody targeted at mouse IgG. The antibodies were bought and verified in our lab.

Immunohistochemistry

Mice were perfused intracardially with 0.9% saline for 10 min and 4% PFA for 10 min. Brains were collected and fixed in 4% PFA for 24 h before being transferred to 20% sucrose solution. The brains were immersed in 30% sucrose before sectioning. All slices (25 µm) of each brain were sectioned coronally and stored in a protectant solution before immunohistochemistry. The slices of each animal at the primarily exact anterior-posterior location and exhibited similar dHIP morphology were selected for each trial for each primary antibody. The selected slices were incubated in 4% PFA for 20 min after being washed with phosphate-buffered saline (PBS) three times. Then, the slices were incubated in 0.3% H₂O₂ for 20 min. After washing, the antigen retrieval was performed to the process that slices were impregnated in 0.01 M citrate buffer (pH 6.0) at 95℃ for 10 min. After cooling and washing, 10% normal goat serum was used to block the slices for 60 min. Next, slices were incubated in primary antibodies, namely, rabbit anti-OT (1:1000, Abcam, Cat# ab212193, RRID: AB_2895534), rabbit anti-OTR (1:500, Abcam, Cat# ab217212, RRID: AB_2904223), anti-CRH (1:1000, Proteintech, Cat# 10944-1-AP, RRID: AB_2084279), and anti-CRHR1 (1:800, Thermo Fisher Scientific, Cat# 702611, RRID: AB_2734802) at 4℃ for 24 h. The negative control was added the dilution buffer for the primary antibodies. The sections were then incubated with peroxidase-conjugated goat anti-rabbit lgG (1:200, ZSGB-Bio, Cat# ZB-2301, RRID: RRID: AB_2747412) at 37°C for one hour after they were washed with PBS for 15 min. After washing, the sections were stained with DAB (3,3'-diaminobenzidine) substrate solution for 3–10 min, and then were attached to the gelatin-coated slides. The next day, dehydration and cleaning were performed. Finally, the slides were sealed with neutral balsam and dried out for three days.

Immunofluorescence

PSD95 and SYP colocalization, which indicated the synaptic connections, was visualized through immunofluorescence. Briefly, the sections were, in turns, washed three times, immersed in 4% PFA for 20 min, rewashed three times, and blocked with a blocking buffer (10% normal goat serum, 0.1% Triton X-100 in 0.01 M PBS) for 60 min. Subsequently, the sections were incubated in the primary antibody of PSD95 (1:1500, Millipore, Cat# MAB1596, RRID: AB_2092365) at 4°C for 18 h. The next day, the sections were incubated in the secondary antibody of goat anti-mouse IgG coupled with CY3 (1:200, AMSBIO, Cat# BA1031, RRID: AB_10890402) at 37°C for one hour and then incubated in the primary antibody SYP (1:500, Abcam, Cat# ab32594, RRID: AB_778204) at 4℃ for 18 h. On the third day, the sections were incubated in the secondary antibody of goat anti-rabbit IgG coupled with fluorescein isothiocyanate (1:200, ZSGB-Bio, Cat# ZF-0311, RRID: AB_2571576) at 37°C for 60 min and then stained with DAPI solution (1:1000, Solarbio, Cat# C0060) for 10 min. Finally, the slices were mounted on the glass slides and sealed with 50% glycerol.

Images acquiring and quantification

The slides were labeled and imaged at room temperature using a microscope with a CCD camera (BX53, Olympus, Tokyo, Japan) and a cellSens software (Olympus cellSens Software, RRID: SCR_014551) followed the principle of optical fractionator (West et al. 1991). The Nissl staining and immunochemistry images were captured at the bright field with a uniform exposure time and a magnification of 40× in a lit room, followed by the measurement by the Image J software (Rawak Software Inc., Stuttgart, Germany, RRID: SCR_003070). For immunofluorescence, the images were captured at the uniform exposure time under the stereotyped fluorescence pattern with an objective len at a magnification of 20× and without a filter in a dark room, followed by the analysis of the SYP and PSD95 colocalization using a Coloc 2 plugin in Fiji software (http://fiji.sc/Fiji, Fiji, RRID: SCR_002285).

For Sholl analysis, the neurons (about 18 neurons per group) were taken according to the principle of optical fractionator (West et al. 1991) at the magnification of 20× by a CCD camera of the BX53 light microscope (BX53, Olympus, Tokyo, Japan). Briefly, the sections containing dHIP (bregma anterior-posterior −1.855 mm to −2.255 mm) among the whole brain sections were located by the hippocampal morphology (Paxinos and Franklin 2008). The divisions of the dHIP, including DG and CA1–3, were located according to the stained neuronal morphology. To overcome the influence of structural variability in the neurons among each region, we only analyzed the neurons located in the regions of interest in the dHIP. The study included one neuron in each region of the dHIP on every 60 µm-thick slice.

The plane was moved from the top to the bottom of the sections during imaging to observe neurons’ dendritic trees. Only the neurons with intact dendritic trees were included in the study. The plane was adjusted to obtain all visible branches of the neurons as far as possible. Reconstruction was performed if necessary. The individual neuron was obtained using the cutting plugin in adobe photoshop CC software for the subsequent analysis. During cutting, the branches of the tested neuron were distinguished carefully. This method overcame the influences of the fine cracks in the Golgi-stained sections. After the individual neuron was obtained, Sholl analysis was conducted to evaluate the dendritic distribution and the dendrite number by the Sholl Analysis plugin in Fiji software (http://fiji.sc/Fiji, Fiji, RRID: SCR_002285).

Statistical analysis

The sample size was determined by G*power 3.1 Version 3.1.9.6 software (Heinrich Heine University Düsseldorf, German, RRID: SCR_013726) before the experiments by assuming P-value <0.05 as acceptable, the study power of 80%, and calculating the standard deviation of the results of the pretest as well as considering of 10% drop-out rate. The formal experiments and the statistics were performed in a double-blind method. The biological replicates in all experimental groups were shown in the figures. The technical replicates depended on the investigations. The mice for behavioral tests and the Golgi-Cox staining were tested for once. The slices for immunochemistry and immunofluorescence were replicated twice. The variance homogeneity of the data was checked by Shapiro-Wilk’s test. When the data were normally distributed with no outliers, two-way analysis of variance (ANOVA) in the GraphPad Prism software (GraphPad Software, Inc., La Jolla, CA, USA, RRID: SCR_002798) was used to check the effects of MD and EE, if needed, Tukey’s multiple comparisons test was used to discover the differences between two groups of mice. If the sample size is small and the data are skewed or contain outliers, the two-way nonparametric ANOVA (Scheirer-Ray-Hare Test) in the SPSS software Version 23 (IBM SPSS Statistics, RRID: SCR_019096) was used to analyze the data. Data in the figures were displayed as mean ± standard deviation (SD) or median and 95% confidence interval (CI). Nonparametric Spearman correlation analysis was used to discover the correlation between two groups of mice. The sample size and the statistical method used in each image were indicated in its figure illustrations. A P value of < 0.05 was considered statistically significant.

EE alleviated avoidance behaviors impaired by MD in male BALB/c mice

The experimental timeline is shown in Fig. 1a. The hippocampal regions of CA1–3 and DG sampled for microscopy were shown in Fig. 1b. The PAT and SOT schematic drawings are shown in Fig. 1c and 1d. In the PAT, because the mice prefer darkness, the mice entered the darkrooms immediately after the sliding doors were open; because the grids in the darkrooms were applied an electric current, the mice returned to the lit rooms after receiving the foot shocks. Thus, the longer latency to enter the darkroom represents the better avoidance. A two-way nonparametric ANOVA of the data at 0 h showed that EE increased the spending time in the lit room before entering the darkroom (F _{(1, 44)}= 6.966, P = 0.011, Figs. 1e and f), suggesting that EE improved immediate avoidance. MD did not affect the spending time in the lit room (F _{(1, 44)}= 0.052, P = 0.820, Fig. 1f). The results at 24 h showed that MD decreased, but EE increased the spending time in the lit room before entering the darkroom (MD, F _{(1, 44)}= 7.676, P = 0.008; EE, F _{(1, 44)}= 22.696, P < 0.001; Fig. 1g) and MD + EE mice spent less time in the lit room before entering the darkroom than ND + EE mice (P < 0.001), suggesting that MD impaired the mid-term avoidance memory and the effect could not be improved by EE in the PAT. A two-way nonparametric ANOVA of the data at 1 w showed that EE increased the spending time in the lit room before entering the darkroom (F _{(1, 44)}= 6.892, P = 0.012, Fig. 1h), suggesting that EE improved long-term avoidance memory. A Kruskal-Wallis post hoc test showed that MD + EE mice spent less time in the lit room than ND + EE mice (P < 0.001, Fig. 1h), suggesting that EE improved the long-term avoidance memory, but the combination of MD and EE could not.

In the SOT, once the switch was open, the mice stepped on the platform immediately to avoid the electric shock from the grid with a 24V current. Thus, if ignoring the ability to jump, the longer latency to step on the platform represents the worse avoidance. A two-way nonparametric ANOVA of the data at 0 h showed that MD increased and EE decreased the latency to step on the platform (EE, F _{(1, 44)}= 12.835, P = 0.002; MD, F _{(1, 44)}= 15.488, P < 0.001; Fig. 1i), suggesting that EE improved immediate avoidance, but MD impaired. A Kruskal-Wallis post hoc test showed that MD + EE mice spent less time stepping on the platform than MD + NE mice (P < 0.001, Fig. 1i), suggesting that EE could restore the avoidance impairment of the SOT induced by MD. At 24 h, two-way nonparametric ANOVA showed that MD and EE had interaction effects (F _{(1, 44)}= 9.072, P = 0.004, Fig. 1j), the post hoc test suggested that MD impaired the mid-term avoidance memory of NE mice (P = 0.003) and EE could alleviate that (P = 0.027, Fig. 1j). A two-way nonparametric ANOVA of the data at 1 w showed that MD increased the latency to step on the platform (F _{(1, 44)}= 9.314, P = 0.004, Fig. 1k) and had interaction with EE (F _{(1, 44)}= 5.186, P = 0.028). The Kruskal-Wallis test showed that MD + NE mice spent more time stepping on the platform than ND + NE mice (P = 0.001, Fig. 1k), suggesting that MD impaired long-term avoidance memory. The Kruskal-Wallis test also showed that MD + EE mice spent similar time stepping on the platform as ND + EE mice (P = 0.625, Fig. 1k), suggesting that EE could alleviate long-term avoidance memory impairment induced by MD. Together, these results of SOT showed that MD impaired the avoidance behaviors and EE improved.

EE increased cells in the CA1–3 of the dHIP and the interaction of MD and EE normalized in the CA1 but not CA3

DAPI is a fluorescent probe that forms a DAPI-nucleic acid complex by attaching DNA (Kapuscinski 1995). We performed DAPI staining to evaluate the nuclei containing DNA in the dHIP regions (Figs. 2a–d). Because most eukaryotes have only one nucleus, the area% of DAPI staining can indicate the cell number in the mouse dHIP. A two-way ANOVA of the data in the CA1 (Fig. 2a) showed that, probably, EE (F_{(1, 20)} = 3.542, P = 0.074) increased nuclei in the CA1 and meanwhile had interaction with MD (F_{(1, 20)} = 7.068, P = 0.015). A Tukey’s test showed that compared to ND + NE, the treatments of ND + EE (P = 0.021) and, probably, MD + NE (P = 0.083) increased cells in the CA1 (Fig. 2a). But, MD + NE mice had similar cells in the CA1 as MD + EE mice (P = 0.946). These results showed that EE increased cells in the CA1 but the interaction of MD and EE normalized the cells in the CA1 as the control mice. The two-way ANOVA of the DAPI staining in the CA2 showed that EE probably increased cells in the CA2 (F_{(1, 20)} = 4.203, P = 0.053, Fig. 2b), MD did not affect cell numbers in the CA2, but the interaction of MD and EE normalized the cell number (F_{(1, 20)} = 7.208, P = 0.014, Fig. 2b). The two-way ANOVA of the DAPI staining in the CA3 showed that EE increased cells in the CA3 (F_{(1, 20)}= 5.776, P = 0.026, Fig. 2c), MD did not affect cell numbers in the CA3, and EE had not interaction with MD (F_{(1, 20)} = 0.912, P = 0.352, Fig. 2c). The two-way ANOVA of the DAPI staining in the DG showed that neither MD nor EE affect cell numbers in the DG (MD, F_{(1, 20)} = 0.031, P = 0.861; EE, F_{(1, 20)} = 1.293, P = 0.269, Fig. 2c). Together, these results suggested that EE increased in the CA1–3 (Figs. 2e1–4), and the interaction of MD and EE normalized the cells in the CA1 but not CA3.

EE could restore the Nissl bodies increased by MD in the CA1 but not the CA3 and DG

The Nissl body constructs endoplasmic reticulum and fine granules (Palay and Palade 1955), representing the biosynthetic function of the neuron. We performed Nissl staining of Cresyl violet to evaluate the neuronal biosynthetic function in the regions of the dHIP (Figs. 3a–d). We calculated the integrated density/µm² of Cresyl violet staining to compare the difference in Nissl bodies levels. The two-way ANOVA of the data in the CA1 revealed that EE (F _{(1, 20)} = 5.049, P = 0.036, Fig. 3a) and its interaction with MD (F_{(1, 20)} = 6.034, P = 0.023, Fig. 3a) decreased Nissl bodies. The Tukey’s test showed that MD + NE mice had higher Nissl bodies levels in the CA1 than ND + NE mice (P = 0.035, Fig. 3a) and MD + EE mice (P = 0.016, Fig. 3a), suggesting that EE could reduce the Nissl bodies increased by MD in the CA1. The two-way ANOVA of the data in the CA2 showed that neither MD nor EE affected Nissl bodies levels in the CA2 (Fig. 3b). The two-way ANOVA of the data in the CA3 showed that EE (F _{(1, 20)} = 4.640, P = 0.043, Fig. 3c) and, probably, MD (F _{(1, 18)} = 3.602, P = 0.072, Fig. 3c) increased Nissl bodies levels in the CA3. The Tukey’s test showed that ND + EE mice had higher Nissl bodies levels in the CA2 than ND + NE mice (P = 0.01, Fig. 3c) and MD + EE mice (P = 0.015), but MD + EE mice had similar Nissl bodies levels as MD + NE mice (P = 0.958), suggesting that EE could not alter the Nissl bodies levels in the CA3 of the MD mice. The data of the DG showed that MD increased Nissl bodies levels (F _{(1, 20)} = 6.960, P = 0.016, Fig. 3d), and the MD + EE mice had similar Nissl bodies levels as MD + NE mice (P = 0.346, Fig. 3d), suggesting that MD increased Nissl bodies in the DG and EE could not restore that. Together, these results showed that for Nissl bodies, MD probably increased in the CA1, CA3, and DG, EE increased in the CA3 and decreased in the CA1, and EE could restore the Nissl bodies altered by MD in the CA1 but not the CA3 and DG (Figs. 3e1–4). These results revealed that experiences such as MD and EE might affect social behaviors via altering the neuronal protein synthesis in the dHIP.

MD decreased synaptic connections in the CA2–3 and DG, and EE restored that in the CA3 but not CA2 and DG

To observe the synaptic connections in the dHIP, we performed immunofluorescence staining of SYP and PSD95. SYP is a presynaptic membrane marker (Wiedenmann and Franke 1985), and PSD95 is a marker of the postsynaptic membrane (Kim et al. 2007). We calculated the Pearson’s R values for SYP and PSD95 colocalization to determine the synaptic connection levels. A two-way ANOVA of the data in the CA1 showed that neither MD nor EE altered the synaptic connections (MD, F_{(1, 20)} = 3.522., P = 0.075, Fig. 4a). In the CA2, whether EE or not, MD downregulated the colocalization of PSD95 and SYP (F _{(1, 19)} = 9.757, P = 0.005, Fig. 4b), suggesting that EE could not restore the synaptic connection’s downregulation induced by MD in the CA2. A two-way nonparametric ANOVA of the data in the CA3 showed that EE increased but MD decreased the synaptic connections in the CA3 (EE, F _{(1, 20)}= 16.364, P = 0.001; MD, F _{(1, 20)} = 9.205, P = 0.007; Fig. 4c). The Kruskal-Wallis test showed that MD + EE mice had higher synaptic connection levels than MD + NE mice (P = 0.021, Fig. 4c), which had lower synaptic connection levels than ND + NE mice (P = 0.005, Fig. 4c), suggesting that EE could restore the synaptic connection downregulation induced by MD in the CA3. The two-way ANOVA of the data from the DG region showed that MD reduced synaptic connection levels (F _{(1, 19)}= 9.248, P = 0.007, Fig. 4d), and EE did not affect (F _{(1, 19)}= 0.005, P = 0.942, Fig. 4d). Tukey’s test showed that MD + EE mice probably had lower synaptic connection levels than ND + EE mice (P = 0.0501, Fig. 4d), suggesting that MD affected the effect of EE to increase synaptic connections in the DG. Together, the abovementioned results showed MD decreased synaptic connections in the CA2–3 and DG, EE increased synaptic connection levels in the CA3, and thus, EE restored synaptic connection downregulation induced by MD in the CA3 but not CA2 and DG (Figs. 4e1–4).

EE restored the dendritic branches alteration induced by MD in the CA1 but not CA3

We conducted Golgi-Cox staining to exhibit neuron morphology in the dHIP. Using Sholl analysis, we evaluated the dendritic branches number of the pyramidal neurons in the CA1, CA2, and CA3 and granule cells in the DG. The intersections numbers in Sholl analysis represented the dendritic branches complexity of the neurons. A two-way nonparametric ANOVA showed that MD probably increased but EE decreased dendritic branches of pyramidal neurons in the CA1 (EE, F _{(1, 68)} = 44.095, P < 0.001; MD, F _{(1, 68)} = 3.306, P = 0.073; Fig. 5a). The Kruskal-Wallis test showed that MD + NE mice had more dendritic branches than ND + NE mice (P = 0.008, Fig. 5a) and MD + EE mice (P = 0.001), suggesting that EE could restore the dendritic branches alteration induced by MD in the CA1. A two-way nonparametric ANOVA of the total intersections of CA2 pyramidal neurons showed that neither MD nor EE altered branches of pyramidal neurons in the CA2 (MD, F_{(1, 68)} = 0.002, P = 0.969; EE, F_{(1, 68)} = 1.653, P = 0.203; Fig. 5b). A two-way ANOVA of the total intersections of CA3 pyramidal neurons showed that EE increased but MD decreased the branches of the CA3 pyramidal neurons (EE, F _{(1, 68)} = 8.712, P < 0.004; MD, F _{(1, 68)} = 20.80, P < 0.001, Fig. 5c). The Tukey’s test showed that ND +EE mice had more dendritic branches than ND + NE mice (P = 0.007) and MD + EE mice (P < 0.001), suggesting that EE increased the branches of pyramidal neurons in the CA3 but could not restore the effect of MD on the branches of pyramidal neurons in the CA3. A two-way nonparametric ANOVA of total intersections of granular cells in the DG showed that neither MD nor EE altered the dendritic branches of granular cells in the CA3 (EE, F _{(1, 68)} = 0.327, P = 0.569; MD, F _{(1, 68)} = 2.473, P = 0.120, Fig. 5d). Together, these abovementioned results indicated that MD decreased dendritic branches in the CA3 and, probably, increased dendritic branches in the CA1, and oppositely, EE increased dendritic branches in the CA3 and decreased dendritic branches in the CA1. However, EE restored the dendritic branches alteration induced by MD in the CA1 but not CA3 (Figs. 5e1–4).

MD increased CRHR1 levels in the CA1 and decreased the CRHR1 levels in the CA2. The effects of EE on the CRHR1 levels depend on whether MD or not

Next, we checked CRHR1 and CRH levels in the dHIP regions using immunohistochemistry to find out the mechanism for the neuronal morphology alteration. We calculated the integrated density/µm² of immunochemistry staining to evaluate the protein levels among the groups. The analysis of the CA1’s data showed that EE probably decreased CRH protein levels (F _{(1, 20)} = 3.278, P = 0.085, Fig. 6a) and MD elevated CRHR1 protein levels in the CA1 (F _{(1, 20)} = 5.387, P = 0.031, Fig. 6b). The post hoc tests showed that MD + NE mice had higher levels of CRH than ND + NE mice (P = 0.027, Fig. 6a) and lower levels of CRH than MD + EE mice (P = 0.0018, Fig. 6a), suggesting that EE could decrease the CRH levels increased by MD. However, EE could not affect the CRHR1 levels increased by MD in the CA1 (F _{(1, 20)} = 1.327, P = 0.255, Fig. 6b). The analysis of the CA2’s data showed MD did not affect CRH levels (F _{(1, 20)} = 1.721, P = 0.204, Fig. 6c) but decreased the CRHR1 levels (F _{(1, 18)} = 9.502, P = 0.0064, Fig. 6d), and EE did not affect both levels of CRH and CRHR1, and meanwhile, the interaction effect of MD and EE was to decrease the levels of CRH (F _{(1, 20)} = 25.31, P < 0.001, Fig. 6c) and CRHR1 (F _{(1, 20)} = 15.996, P = 0.001, Fig. 6d) in the CA2. The post hoc tests showed that compared to the ND + NE mice, both MD + NE (P = 0.001) and ND + EE mice (P = 0.015, Fig. 6c) had higher CRH levels in the CA2. Besides CRH, ND + EE mice also had higher CRHR1 levels in the CA2 (P = 0.045, Fig. 6d). MD + EE mice had lower CRH levels than MD + NE mice (P = 0.006, Fig. 6c) and, probably, CRHR1 levels (P = 0.070, Fig. 6d) than ND + EE mice, suggesting that EE decreased the CRH levels increased by MD and MD could affect the effect of EE to increase the CRHR1 levels in the CA2. A two-way ANOVA of the CA3’s data showed the interaction effect of MD and EE was to decrease the levels of CRH (F _{(1, 20)} = 13.83, P = 0.001, Fig. 6e) and CRHR1 (F _{(1, 20)} = 40.74, P < 0.001, Fig. 6f) in the CA3. The Tukey’s test showed that MD + NE mice had higher CRHR1 levels and, probably, CRH levels than ND + NE mice (P = 0.001, Fig. 6f; P = 0.099, Fig. 6e) and MD + EE mice (P = 0.001, Fig. 6f; P = 0.042, Fig. 6e), suggesting that MD increased CRH-CRHR1 signaling of the CA3 and EE could restore. The Tukey’s test also showed that ND + EE mice had higher CRH levels than MD + EE mice (P = 0.050, Fig. 6e), and higher CRHR1 levels than ND + NE mice (P = 0.003, Fig. 6f) and MD + EE mice (P = 0.001, Fig. 6f), suggesting that MD could affect the effects of EE to increase CRH and CRHR1 levels in the CA3. A two-way ANOVA of the DG’s data showed that neither MD nor EE altered the CRH and CRHR1 levels, but they had the interaction effect on the levels of CRH (F _{(1, 20)} = 7.593, P = 0.012, Fig. 6g) and CRHR1 levels (F _{(1, 20)} = 6.533, P = 0.019, Fig. 6h). Tukey’s tests showed that MD + EE mice tended to have lower CRHR1 levels than MD + NE mice (P = 0.060, Fig. 6h), suggesting that EE was beneficial for decreasing CRHR1 levels in the MD mice. Taken together, for CRHR1 levels, MD increased in the CA1 and CA3 but decreased in the CA2. EE increased CRH levels in the CA1. For MD mice, the EE treatment decreased CRH levels of the CA1–3 but only decreased CRHR1 of the CA3 region.

To explore the effects of CRH-CRHR1 signaling to modulate the neuron morphology in the dHIP, we performed the nonparametric Spearman correlations between CRHR1 levels and the tested parameters of neuron morphology. Without considering the effects of the groups, the results showed that CRHR1 protein levels were positively correlated with CRH protein levels in all tested regions of the dHIP (CA1, n = 22 mice, r = 0.581, P = 0.005, Fig. s2a1; CA2, n = 22 mice, r = 0.429, P = 0.047, Fig. s2a2; CA3, n = 24 mice, r = 0.628, P = 0.001, Fig. s2a3; DG, n = 24 mice, r = 0.534, P = 0.007, Fig. s2a4). The CRHR1 levels were positively correlated with %area of DAPI staining in the CA1 (n = 23 mice, r = 0.433, P = 0.039, Fig. s2b1) and CA2 (n = 21 mice, r = 0.407, P = 0.067, Fig. s2b2) but not CA3 (Fig. s2b3) and DG (Fig. s2b4). The CRHR1 levels were positively correlated with integrated density/µm²of cresyl violet-staining in the CA3 (n = 24 mice, r = 0.567, P = 0.004, Fig. s2c3) but not CA1 (Fig. s2c1), CA2 (Fig. s2c2), and DG (Fig. s2c4). The CRHR1 levels were correlated with the Pearson’s R of SYP and PSD95 colocalization positively in the CA1 (n = 23 mice, r = 0.426, P = 0.043, Fig. s2d1), CA2 (n = 20 mice, r = 0.458, P = 0.042, Fig. s2d2) , and DG (n = 20 mice, r = 0.515, P = 0.020, Fig. s2d4) , and negatively in the CA3 (n = 20 mice, r = −0.438, P = 0.053, Fig. s2d3). Together, the results suggested that CRHR1 levels were positively correlated with CRH protein levels in all regions of the dHIP. The results also suggested that CRHR1 levels were associated with synaptic connection levels in all regions of the dHIP (positively in the CA1, CA2, and DG, negatively in the CA3), Nissl body levels in the CA3, and cell nuclei levels in the CA1 and CA2.

MD elevated the OTR level in the CA1 region of the dHIP

Accumulating evidence shows that reciprocal regulation of the CRH and OT systems balances stress (Dabrowska et al. 2011). We checked the OTR level associated with stress, energy metabolism, and social attachment (Onaka et al. 2012). A two-way ANOVA of the data in the CA1 showed that MD and EE had interaction effects (F _{(1, 20)} = 10.76, P = 0.004, Fig. 7a). The Tukey’s test showed that MD + NE mice had higher OTR levels in the CA1 than ND + NE mice (P = 0.008, Fig. 7a); the MD + EE mice had lower OTR levels than MD + NE mice (P = 0.041, Fig. 7a), suggesting the EE could decrease OTR levels increased by MD in the CA1. A two-way ANOVA of the data of the CA2 showed that neither MD nor EE altered the OTR levels in the CA2, but they probably interacted with each other (F _{(1, 20)} = 3.443, P = 0.078, Fig. 7b). In the CA3 and DG, the two-way nonparametric ANOVA showed that neither MD nor EE affected the OTR levels, but MD and EE probably had interaction effects (CA3, F _{(1, 20)} = 6.961, P = 0.016, Fig. 7c; DG, F _{(1, 20)} = 3.137, P = 0.092, Fig. 7d). The above results showed EE restored the OTR levels increased by MD in the CA1. Both MD and EE did not affect the OTR levels of the CA2–3 and DG.

To ensure the OTR level in the dHIP, we checked the OTR level in the lateral entorhinal area (ENTL). Fig. s3a shows the relationship between the ENTL and HIP. Layers Ⅲ and Ⅴ of the ENTL are closely associated with the CA1 in the HIP, and Layer Ⅱ is linked to the CA3 and DG (Deng et al. 2010). Fig. s3b shows the representative images of OTR immunochemistry. A two-way ANOVA of the total OTR levels in the ENTL showed that neither MD nor EE affected the total OTR levels, but MD and EE had interaction effects (F _{(1, 20)} = 9.302, P = 0.006, Fig. s3c). However, Tukey’s test did not show any differences. Next, we analyzed the OTR levels in Layer Ⅱ, Ⅲ–Ⅳ, and Ⅴ–Ⅵ of the ENTL (Fig. s3d–f). The results of Layer Ⅱ showed that neither MD nor EE probably affect OTR level (F _{(1, 20)} = 3.180, P = 0.090, Fig. s3d). The two-way ANOVA of Layer Ⅲ–Ⅳ also showed that neither MD nor EE affect the OTR levels (Fig. s3e). The results of Layer Ⅴ–Ⅵ showed that MD probably increased the OTR levels (F _{(1, 19)} = 3.951, P = 0.062, Fig. s3f), and meanwhile, MD and EE interacted with each other (F _{(1, 19)} = 6.344, P = 0.021, Fig. s3f). The Tukey’s test showed that MD + NE mice had higher OTR levels in the Layer Ⅴ–Ⅵ than ND + NE mice (P = 0.027, Fig. s3f) and, probably, MD + EE mice (P = 0.066, Fig. s3f), suggesting that EE could decrease OTR levels in the Layer Ⅴ–Ⅵ increased by MD. Thus, the results reassured that MD elevated the OTR level in the CA1 region, which EE could alleviate.

Next, we performed the nonparametric Spearman correlation analysis to show the relationships between CRHR1 levels and OTR levels and the parameters of neuronal morphology. The result showed that the CRHR1 levels correlated with OTR levels in the CA1 (n = 20 mice, r = 0.483, P = 0.031, Fig. s4a1) and CA3 (n = 21 mice, r = 0.436, P = 0.048, Fig. s4a3), but not DG (Fig. s4a4) and CA2 (Fig. s4a2). The OTR levels were probably correlated with the %area of DAPI in the CA1 (n = 21 mice, r = 0.400, P = 0.072, Fig. s4b1), CA2 (n = 21 mice, r = 0.487, P = 0.025, Fig. s4b2), and DG (n = 20 mice, r = 0.550, P = 0.012, Fig. s4b4), but not CA3 (Fig. s4b3). The OTR levels were correlated with integrated density/µm²of cresyl violet-staining positively in the CA1 (n = 22 mice, r = 0.654, P = 0.001, Fig. s4c1) and, probably, negatively in the CA2 (n = 21 mice, r = −0.395, P = 0.077, Fig. s4c2), but not CA3 (Fig. s4c3) and DG (Fig. s4c4). The OTR levels were correlated with the Pearson’s R of SYP and PSD95 colocalization positively in the CA2 (n = 20 mice, r = 0.462, P = 0.040, Fig. s4d2) and DG (n = 20 mice, r = 0.484, P = 0.031, Fig. s4d4), negatively in the CA3 (n = 21 mice, r = −0.579, P = 0.006, Fig. s4d3). In the CA1 of the ND + NE mice, the OTR levels and Pearson’s R of SYP and PSD95 colocalizations were negatively correlated (n = 5 mice, r = −0.975, P = 0.033, Fig. s4d1). Together, these results indicated that the OTR levels positively correlated with CRHR1 levels in the CA1 and CA3, cells number in the CA2, Nissl bodies in the CA1, and synaptic connections in CA2 and DG (Fig. 8).

Further, to explore the balances of CRHR1 and OTR levels, we compared the ratios of CRHR1 to OTR among groups in the tested dHIP regions. A two-way ANOVA of the CA1 showed that MD and EE had interaction effects of elevating the OTR levels of the CA1 (F_{(1, 20)} = 6.142, P = 0.022, Fig. 7e1). However, Tukey’s test did not show any differences between groups. The two-way nonparametric ANOVAs showed that neither MD nor EE affect the ratios of CRHR1 to OTR in the CA2 and CA3 (Fig. 7e2 and Fig. 7e3). A two-way ANOVA of the DG’s data showed that EE decreased the ratios of CRHR1 to OTR (F _{(1, 20)} = 4.87, P = 0.039, Fig. 7e4), but no difference was found in the following Tukey’s test, suggesting that EE disrupted the balance of the CRHR1 and OTR in the DG. Together, EE alone decreased the ratios of CRHR1 levels to OTR levels in the DG, MD and EE had interaction to elevate the ratios in the CA1.

Here, we used MD as adversity and EE as prosperity to study the effect of experiences on avoidance behaviors and the hippocampal neuron morphology modulated by the CRHR1 and OTR systems. The results showed that MD impaired the avoidance behaviors in the male BALB/c mice, and EE improved that. EE increased in the CA1–3 of the dHIP and the interaction of MD and EE normalized the cells in the CA1 but not CA3. MD probably increased Nissl bodies in the CA1, CA3, and DG, and EE could restore the Nissl body levels altered by MD in the CA1 but not the CA3 and DG. MD decreased synaptic connections in the CA2, CA3, and DG, and EE restored synaptic connection’s downregulation induced by MD in the CA3 but not CA2 and DG. MD decreased dendritic branches in the CA3 and, probably, increased dendritic branches in the CA1, and EE restored the dendritic branches alteration induced by MD in the CA1 but not CA3. For CRHR1 levels, MD increased in the CA1 and decreased in the CA2, and the EE treatment alleviated CRH levels in the CA1, CA2, and CA3 of the MD mice, but only decreased CRHR1 of the CA3. MD increased OTR levels in the CA1 and EE could restore. MD did not affect the OTR levels of the CA3 and DG, and meanwhile, EE could not alter that in the MD mice. The abovementioned results showed that MD and EE could affect avoidance behaviors, neuronal morphology in the dHIP, and the levels of the CRHR1 and OTR. Because the correlation analysis showed that CRHR1 and OTR levels were associated with synaptic connections, Nissl bodies, and cell nuclei in the dHIP, MD and EE might affect the neuronal morphology via modulating the CRHR1 and OTR levels and their balance.

Information is processed in distinct regions of the HIP via a classic tri-synaptic circuit through which signals are transferred through two parallel pathways, namely, the entorhinal cortex (EC)-CA2-CA1 pathway and the EC-DG-CA3 circuit (Bartesaghi et al. 2006; Hitti and Siegelbaum 2014). Both CA1 and CA3 are involved in the memory for sequential non-spatial events that compose unique experiences (Farovik et al. 2010). CA1 lesions lead to difficulties in performing temporal but not spatial tasks (Gilbert et al. 2001) or a preference for spatial location opposite to normal animals (Hunsaker and Kesner 2008). CA3 encodes an episodic or event memory involving a space component (Kesner 2007; Rolls 2007), and CA3 lesion results in impaired spatiotemporal processing (Hunsaker and Kesner 2008). CA2 is an essential hub for social memory, and the inactivation of pyramidal neurons in the CA2 causes a loss of ability to remember a conspecific mouse (Hitti and Siegelbaum 2014). DG is essential for the episodic engram’s programming, retrieval, and discrimination (Hainmueller and Bartos 2020). DG lesions lead to difficulties in performing tasks with high levels of spatial interference (Gilbert et al. 2001; Hunsaker and Kesner 2008). In our study, MD increased Nissl bodies and neuronal complexity in the CA1, decreased synaptic connections in the CA2 and CA3, and increased Nissl bodies in the DG, suggesting that MD could affect the functions of the all tested regions of the dHIP. EE restored the cells and synaptic connections alterations induced by MD in the CA3, and Nissl bodies and neuron complexity alterations induced by MD in the CA1 but could not restore the alterations in the CA3, suggesting that EE partially restored the hippocampal impairments induced by MD.

Stress promotes secretion of the CRH and activates CRHR1 located on pyramidal cell dendrites in the HIP (Ivy et al. 2010). Inhibitory interneurons release CRH within the HIP; the peptide may be necessary and sufficient to initiate the changes to hippocampal neurons associated with early-life stress (Maras and Baram 2012). The CRH-CRHR1 signaling activation models hippocampal structure and function in a dose- and time-dependent manner (Maras and Baram 2012). Low-short duration of CRH-CRHR1 signaling activation enhances hippocampal function, whereas more extended CRH-CRHR1 signaling activation correlates with lower complexity and dendritic length of the hippocampal neurons (Maras and Baram 2012). The CRHR1 antagonist prevents dendritic atrophy and long-term potentiation attenuation in CA1 Schaffer collateral synapses (Ivy et al. 2010). Hippocampal commissural/associational (C/A) pathways establish their synaptic connections on CA3 pyramidal cell dendrites during the prenatal (roughly days 0–5) and infancy (second postnatal week) stages (Maras and Baram 2012). The C/A fibers have limited capacity to make synaptic contacts if the original connections are disrupted beyond the 3rd postnatal week (Maras and Baram 2012). In the study, long-term MD increased CRHR1 levels in the CA1 in the adult male mice, and meanwhile, MD increased pyramidal neuron branches in the CA1 but decreased branches of pyramidal neurons in the CA3. Post-weaning EE of 90 d restored the altered intersections induced by MD in the CA1 but not CA3. These results indicated that the CA3 had limited capacity to recover if the structure was impaired by long-term MD during the neonatal days. These results also suggested that the CA1 was more plastic to experience than CA3. However, it was unclear why long-term MD increases pyramidal neuron branches in the CA1? So, we checked the OTR levels associated with hippocampal neurogenesis (Lin et al. 2017).

OTR expressed on the glutamatergic neurons modulates dendrite branching and functions via Gq/11-coupled OTR (Ripamonti et al. 2017). Inhibitory hippocampal neurons are less affected by OT because of low OTR expression (Ripamonti et al. 2017). Activation of the CRHR1 increases the release of local OT (Klampfl et al. 2018), which can reduce evoked GABAergic inhibition at many synapses (Mitre et al. 2018). Our previous study showed that MD decreased OT neurons in the PVH (Wei et al. 2021), and in the study, MD enhanced CRH-CRHR1 signaling in the CA1. The result was following the available evidence that the changes in behavioral risk induced by early life adversity are associated with a strengthening of CRH systems and the weakening of OT systems (Bardo et al. 2021). The decreased OT levels in the magnocellular PVH can result in the OTR increase in the brain nucleus, such as the basolateral amygdala; the increased OTR parallels the increased neuronal complexity (Wei et al. 2021). This might be why MD increased the branches of the pyramidal neurons in the CA1. In the same way, EE increased OT neurons in the parvocellular PVH (Wei et al. 2021), and in the study, EE probably decreased CRH in the CA1. However, EE did not disrupt the balance of CRHR1 and OTR in the CA1, meaning that the OTR levels paralleled the CRHR1 in the CA1 of the EE mice. This might be why EE did not increase but decreased the complexity of pyramidal neurons in the CA1. Besides, EE alone decreased the ratios of CRHR1 levels to OTR levels in the DG; this may be one of the reasons that EE increased pyramidal branches in the CA3.

The viral tracing method demonstrates that local CRH inhibitory interneurons export extensive presynaptic inputs onto the new neurons, revealing the neuronal modulatory effects of CRH for synapse connections in the adult brain (Garcia et al. 2014). CRHR1 is primarily responsible for mediating the synaptic actions of CRH on principal hippocampal cells (Chen et al. 2012). Consistently, the present study showed that CRHR1 levels correlated with synaptic connections in all tested regions of the dHIP. OTR presents in the presynaptic and postsynaptic membranes and regulates neuronal excitability (Bakos et al. 2018). The correlation analysis showed that OTR levels positively correlated with CRHR1 levels in the CA1 and CA3 but negatively correlated with CRHR1 levels in the CA2, indicating that OTR might play a role in modulating synaptic connections. Besides synaptic connections, OTR plays a role in controlling neurogenesis (Lin et al. 2017) and energy metabolism (Chaves et al. 2013). In the study, we found that OTR levels correlated with synaptic connections in the CA2, CA3, and DG, and in addition, nuclei containing DNA in the CA1, CA2, and DG, and Nissl bodies in the CA1 and CA2. Many neuropsychiatric disorders, including autism, Alzheimer’s disease, schizophrenia, mood disorders, and posttraumatic stress disorder, had OT system abnormalities (Marazziti and Catena Dell’osso 2008; Cochran et al. 2013). Lack of OTRs leads to reversal learning impairment (Chini et al. 2014). In the study, MD induced avoidance impairment and increased OTR level in the CA1; EE improved avoidance and reduced the ratios of CRHR1 to OTR level in the DG, indicating that the OT system may correlate with avoidance behaviors driven by experiences. Thus, besides of CRH-CRHR1 system, OT-OTR signaling is a promising system for drug development to alleviate neuropsychiatric disorders with abnormal avoidance behaviors and hippocampal neuronal morphology.

Moreover, the correlation analysis indicated that OTR levels positively correlated with CRHR1 levels in the CA1 and CA3 but negatively correlated with CRHR1 levels in the CA2, suggesting that CA2 was a unique area different from CA1 and CA3. The hippocampal CA2 region is essential for social memory (Hitti and Siegelbaum 2014). In the CA2, compared with the ND + NE mice, the CRH levels increased, but the CRHR1 levels were unchanged in the MD mice. The finding may explain that MD induced social deficits and avoidance impairment.

Experiences cause neurobiological alterations in the OT or CRH-containing neural circuits of children and adults, resulting in increased risk or protection of psychopathology (Heim et al. 1997; Heim and Nemeroff 2001). Our study revealed how early-life MD and post-weaning EE affected the OT signaling and CRH system and resulted in hippocampal neuron morphology and avoidance behavior alterations in male BALB/c mice. The dHIP’s regions showed elective vulnerability to the different experiences. The alterations in dHIP’s regions induced by early life stress can be partially recovered by late-life environmental improvement. The result may help understand avoidance memory and experience-related mental disorders. Future research is needed to explore the effects of early environment and experiences on hippocampal neuron functions.

Acknowledgments

This work was supported by the National Natural Science Foundation of China [No. 81570725 and 81870949] to Yu-Hong Jing.

Author contributions

FW and YJ were responsible for experimental design and revision. FW performed the experiments and wrote the manuscript. XD, BM, WL, YC, and TZ assisted FW with the analysis of Golgi-Cox staining. YZ, LZ (Lang Zhang) and LZ (Long Zhao) assisted FW with the mouse husbandry work. LZ (Long Zhao) and YJ assisted FW with the data collection and analysis.

Conflict of interest

The authors declare no conflict of interest.

Bagosi Z, Balangó B, Pintér D, et al (2015) The effects of CRF and urocortins on the hippocampal glutamate release. Neurochem Int 90:67–71. https://doi.org/https://doi.org/10.1016/j.neuint.2015.07.015
Bakos J, Srancikova A, Havranek T, Bacova Z (2018) Molecular mechanisms of oxytocin signaling at the synaptic connection. Neural Plast 2018:4864107. https://doi.org/10.1155/2018/4864107
Bardo MT, Hammerslag LR, Malone SG (2021) Effect of early life social adversity on drug abuse vulnerability: Focus on corticotropin-releasing factor and oxytocin. Neuropharmacology 191:108567. https://doi.org/10.1016/j.neuropharm.2021.108567
Bartesaghi R, Migliore M, Gessi T (2006) Input–output relations in the entorhinal cortex–dentate–hippocampal system: Evidence for a non-linear transfer of signals. Neuroscience 142:247–265. https://doi.org/10.1016/j.neuroscience.2006.06.001
Bayram-Weston Z, Olsen E, Harrison DJ, et al (2016) Optimising Golgi-Cox staining for use with perfusion-fixed brain tissue validated in the zQ175 mouse model of Huntington’s disease. J Neurosci Methods. https://doi.org/10.1016/j.jneumeth.2015.09.033
Chaves VE, Tilelli CQ, Brito NA, Brito MN (2013) Role of oxytocin in energy metabolism. Peptides 45:9–14. https://doi.org/10.1016/j.peptides.2013.04.010
Chen Y, Andres AL, Frotscher M, Baram TZ (2012) Tuning synaptic transmission in the hippocampus by stress: the CRH system. Front Cell Neurosci 6:13. https://doi.org/10.3389/fncel.2012.00013
Chen Y, Bender RA, Brunson KL, et al (2004) Modulation of dendritic differentiation by corticotropin-releasing factor in the developing hippocampus. Proc Natl Acad Sci U S A 101:15782–7. https://doi.org/10.1073/pnas.0403975101
Chini B, Leonzino M, Braida D, Sala M (2014) Learning about oxytocin: pharmacologic and behavioral issues. Biol Psychiatry 76:360–6. https://doi.org/10.1016/j.biopsych.2013.08.029
Cimadevilla JM, Fenton AA, Bures J (2000) Functional inactivation of dorsal hippocampus impairs active place avoidance in rats. Neurosci Lett 285:53–56. https://doi.org/https://doi.org/10.1016/S0304-3940(00)01019-3
Cochran DM, Fallon D, Hill M, Frazier JA (2013) The role of oxytocin in psychiatric disorders: a review of biological and therapeutic research findings. Harv Rev Psychiatry 21:219–47. https://doi.org/10.1097/HRP.0b013e3182a75b7d
Dabrowska J, Hazra R, Ahern TH, et al (2011) Neuroanatomical evidence for reciprocal regulation of the corticotrophin-releasing factor and oxytocin systems in the hypothalamus and the bed nucleus of the stria terminalis of the rat: Implications for balancing stress and affect. Psychoneuroendocrinology 36:1312–1326. https://doi.org/10.1016/j.psyneuen.2011.03.003
Deng W, Aimone JB, Gage FH (2010) New neurons and new memories: how does adult hippocampal neurogenesis affect learning and memory? Nat Rev Neurosci 11:339–350. https://doi.org/10.1038/nrn2822
Farovik A, Dupont LM, Eichenbaum H (2010) Distinct roles for dorsal CA3 and CA1 in memory for sequential nonspatial events. Learn Mem 17:12–17. https://doi.org/10.1101/lm.1616209
Garcia I, Quast KB, Huang L, et al (2014) Local CRH Signaling Promotes Synaptogenesis and Circuit Integration of Adult-Born Neurons. Dev Cell 30:645–659. https://doi.org/https://doi.org/10.1016/j.devcel.2014.07.001
Gilbert PE, Kesner RP, Lee I (2001) Dissociating hippocampal subregions: A double dissociation between dentate gyrus and CA1. Hippocampus 11:626–636. https://doi.org/10.1002/hipo.1077
Hainmueller T, Bartos M (2020) Dentate gyrus circuits for encoding, retrieval and discrimination of episodic memories. Nat Rev Neurosci 21:153–168. https://doi.org/10.1038/s41583-019-0260-z
Heim C, Nemeroff CB (2001) The role of childhood trauma in the neurobiology of mood and anxiety disorders: preclinical and clinical studies. Biol Psychiatry 49:1023–1039. https://doi.org/10.1016/S0006-3223(01)01157-X
Heim C, Owens MJ, Plotsky PM, Nemeroff CB (1997) Persistent changes in corticotropin-releasing factor systems due to early life stress: relationship to the pathophysiology of major depression and post-traumatic stress disorder. Psychopharmacol Bull 33:185–92
Hitti FL, Siegelbaum SA (2014) The hippocampal CA2 region is essential for social memory. Nature 508:88–92. https://doi.org/10.1038/nature13028
Hunsaker MR, Kesner RP (2008) Evaluating the differential roles of the dorsal dentate gyrus, dorsal CA3, and dorsal CA1 during a temporal ordering for spatial locations task. Hippocampus 18:955–64. https://doi.org/10.1002/hipo.20455
Ivy AS, Rex CS, Chen Y, et al (2010) Hippocampal dysfunction and cognitive impairments provoked by chronic early-life stress involve excessive activation of CRH receptors. J Neurosci 30:13005–13015. https://doi.org/10.1523/JNEUROSCI.1784-10.2010
Jamieson BB, Nair BB, Iremonger KJ (2017) Regulation of hypothalamic corticotropin-releasing hormone neurone excitability by oxytocin. J Neuroendocrinol 29:e12532. https://doi.org/10.1111/jne.12532
Kapuscinski J (1995) DAPI: a DNA-specific fluorescent probe. Biotech Histochem Off Publ Biol Stain Comm 70:220–233. https://doi.org/10.3109/10520299509108199
Kesner RP (2007) Behavioral functions of the CA3 subregion of the hippocampus. Learn Mem 14:771–81. https://doi.org/10.1101/lm.688207
Kim MJ, Futai K, Jo J, et al (2007) Synaptic accumulation of PSD-95 and synaptic function regulated by phosphorylation of serine-295 of PSD-95. Neuron 56:488–502. https://doi.org/10.1016/j.neuron.2007.09.007
Klampfl SM, Schramm MM, Gaßner BM, et al (2018) Maternal stress and the MPOA: Activation of CRF receptor 1 impairs maternal behavior and triggers local oxytocin release in lactating rats. Neuropharmacology 133:440–450. https://doi.org/10.1016/j.neuropharm.2018.02.019
Lin Y-T, Chen C-C, Huang C-C, et al (2017) Oxytocin stimulates hippocampal neurogenesis via oxytocin receptor expressed in CA3 pyramidal neurons. Nat Commun 8:537. https://doi.org/10.1038/s41467-017-00675-5
Lin Y-T, Hsu K-S (2018) Oxytocin receptor signaling in the hippocampus: Role in regulating neuronal excitability, network oscillatory activity, synaptic plasticity and social memory. Prog Neurobiol 171:1–14. https://doi.org/10.1016/j.pneurobio.2018.10.003
Maniezzi C, Talpo F, Spaiardi P, et al (2019) Oxytocin increases phasic and tonic GABAergic transmission in CA1 region of mouse hippocampus. Front Cell Neurosci 13:178. https://doi.org/10.3389/fncel.2019.00178
Maras PM, Baram TZ (2012) Sculpting the hippocampus from within: stress, spines, and CRH. Trends Neurosci 35:315–324. https://doi.org/10.1016/j.tins.2012.01.005
Marazziti D, Catena Dell’osso M (2008) The role of oxytocin in neuropsychiatric disorders. Curr Med Chem 15:698–704. https://doi.org/10.2174/092986708783885291
Meena P, Sharma D, Jain A, et al (2012) Physical and environmental stressors induces region specific neuronal damage in hippocampal subfields of mice. Proc Natl Acad Sci India Sect B Biol Sci 82:291–297. https://doi.org/10.1007/s40011-012-0022-4
Mitre M, Minder J, Morina EX, et al (2018) oxytocin modulation of neural circuits. Curr Top Behav Neurosci 35:31–53. https://doi.org/10.1007/7854_2017_7
Noonan LR, Caldwell JD, Li L, et al (1994) Neonatal stress transiently alters the development of hippocampal oxytocin receptors. Brain Res Dev Brain Res 80:115–120. https://doi.org/10.1016/0165-3806(94)90094-9
Oleksiak CR, Ramanathan KR, Miles OW, et al (2021) Ventral hippocampus mediates the context-dependence of two-way signaled avoidance in male rats. Neurobiol Learn Mem 183:107458. https://doi.org/https://doi.org/10.1016/j.nlm.2021.107458
Onaka T, Takayanagi Y, Yoshida M (2012) Roles of oxytocin neurones in the control of stress, energy metabolism, and social behaviour. J Neuroendocrinol 24:587–598. https://doi.org/10.1111/j.1365-2826.2012.02300.x
Palay SL, Palade GE (1955) The fine structure of neurons. J Biophys Biochem Cytol 1:69–88. https://doi.org/10.1083/jcb.1.1.69
Paxinos G, Franklin K (2008) The mouse brain in stereotaxic coordinates, compact | 978-0-12-374244-5 | Elsevier
Quinn JJ, Wied HM, Ma QD, et al (2008) Dorsal hippocampus involvement in delay fear conditioning depends upon the strength of the tone-footshock association. Hippocampus 18:640—654. https://doi.org/10.1002/hipo.20424
Ripamonti S, Ambrozkiewicz MC, Guzzi F, et al (2017) Transient oxytocin signaling primes the development and function of excitatory hippocampal neurons. Elife 6:e22466. https://doi.org/10.7554/eLife.22466
Rolls ET (2007) An attractor network in the hippocampus: theory and neurophysiology. Learn Mem 14:714–31. https://doi.org/10.1101/lm.631207
Teicher MH, Anderson CM, Polcari A (2012) Childhood maltreatment is associated with reduced volume in the hippocampal subfields CA3, dentate gyrus, and subiculum. Proc Natl Acad Sci 109:E563–E572. https://doi.org/10.1073/pnas.1115396109
Tirko NN, Eyring KW, Carcea I, et al (2018) Oxytocin transforms firing mode of CA2 hippocampal neurons. Neuron 100:593-608.e3. https://doi.org/https://doi.org/10.1016/j.neuron.2018.09.008
Trusel M, Nuno-Perez A, Lecca S, et al (2019) Punishment-predictive cues guide avoidance through potentiation of hypothalamus-to-habenula synapses. Neuron 102:120-127.e4. https://doi.org/10.1016/j.neuron.2019.01.025
Vivinetto AL, Suárez MM, Rivarola MA (2013) Neurobiological effects of neonatal maternal separation and post-weaning environmental enrichment. Behav Brain Res 240:110–8. https://doi.org/10.1016/j.bbr.2012.11.014
Wei F, Li W, Ma B, et al (2021) Experiences affect social behaviors via altering neuronal morphology and oxytocin system. Psychoneuroendocrinology 129:105247. https://doi.org/10.1016/j.psyneuen.2021.105247
West MJ, Slomianka L, Gundersen HJG (1991) Unbiased stereological estimation of the total number of neurons in the subdivisions of the rat hippocampus using the optical fractionator. Anat Rec 231:482–497. https://doi.org/10.1002/ar.1092310411
Wiedenmann B, Franke WW (1985) Identification and localization of synaptophysin, an integral membrane glycoprotein of Mr 38,000 characteristic of presynaptic vesicles. Cell 41:1017–28. https://doi.org/10.1016/s0092-8674(85)80082-9
Yazgan I, Hanson JL, Bates JE, et al (2021) Cumulative early childhood adversity and later antisocial behavior: The mediating role of passive avoidance. Dev Psychopathol 33:340–350. https://doi.org/10.1017/S0954579419001809

Download PDF

Reviews received at journal
28 Mar, 2022
Reviewers invited by journal
28 Mar, 2022
Editor invited by journal
23 Mar, 2022
Editor assigned by journal
23 Mar, 2022
First submitted to journal
21 Mar, 2022

You are reading this latest preprint version

Experiences Shape Hippocampal Neuron Morphology and the Balance of CRHR1 and OTR

Status:

Version 1

Abstract

Figures

Introduction

Methods And Materials

Golgi-Cox staining

Results

EE alleviated avoidance behaviors impaired by MD in male BALB/c mice

Discussion

Conclusion

Declarations

References

Supplementary Files

Status:

Version 1