Chronic restraint stress induces depression-like behaviors and alterations in the afferent projections of medial prefrontal cortex from multiple brain regions in mice

Introduction: The medial prefrontal cortex (mPFC) forms output pathways through projection neurons, inversely receiving adjacent and long-range inputs from other brain regions. However, how afferent neurons of mPFC are affected by chronic stress needs to be clarified. In this study, the effects of chronic restraint stress (CRS) on the distribution density of mPFC dendrites/dendritic spines and the projections from the cortex and subcortical brain regions to the mPFC were investigated. Methods: In the present study, C57BL/6 J transgenic (Thy1-YFP-H) mice were subjected to CRS to establish an animal model of depression. The infralimbic (IL) of mPFC was selected as the injection site of retrograde AAV using stereotactic technique. The effects of CRS on dendrites/dendritic spines and afferent neurons of the mPFC IL were investigaed by quantitatively assessing the distribution density of green fluorescent (YFP


Introduction
According to a report released by the World Health Organization in 2021 W. H.O (2023), there are about 280 million people suffering from depression worldwide, and it has become a common mental disorder.However, the pathogenesis of depression is unclear.
The monoamine hypothesis has been used as the foundational model for the development of numerous clinical antidepressants currently in use.However, the administration of these medications often necessitates a delay of several weeks before alleviating depressive symptoms, and only one-third of patients experience relief.Another one-third must switch to a different or combined drug regimen to achieve effectiveness, while approximately one-third remain unresponsive to existing medications (Trivedi et al., 2006, Souery et al., 2011).Currently, research on the pathogenesis of depression has shifted its focus towards neural network and circuit models to elucidate how key nodes within the limbic system and other systems interact to produce persistent alterations in affective states (Timothy Spellman, 2020).
In mice, the mPFC is composed of multiple regions including the anterior cingulate (ACC), prelimbic (PL), and infralimbic (IL).Studies have shown that excitotoxin-induced ACC lesions can cause depressivelike behavior in rats in the FST (Bissiere S et al., 2006).Chronic restraint stress (CRS) resulted in decreased apical dendritic arbors in the rat ACC (Liston et al., 2006), while chronic (unpredictable) mild stress (C(U)MS) increased GABA levels in the ACC of rats (Perrine et al., 2014).In stress-induced depression, PL and IL often act as upstream regulators of many cortical and subcortical regions (Lee and Han, 2019).In rodents, chronic stress significantly reduced the firing rate of PL neurons (Lee et al., 2011, Friedman et al., 2017), but increased activity in IL neurons (Lee et al., 2011).C(U) MS for 21 days caused significant synaptic loss in the PL (Ota et al., 2014, Xu et al., 2016).The roles of these distinct regions in cognition gradually shift along the dorsoventral axis, from decision making and action to motivation and emotion (Anastasiades and Carter, 2021).
In the present study, the IL was selected as the injection site of retrotracing virus through stereotaxic technology.The fluorescence protein YFP (green), widely distributed in the soma, axons, and dendrites of excitatory neurons in many brain regions in the transgenic mice (Feng et al., 2000), and retro-tracing virus recombinant fluorescent protein (red) were used to evaluate the effects of chronic restraint stress on the mPFC dentrites/dendritic spines and afferent neurons in other brain regions to the mPFC, in order to provide experimental evidence for further exploring the pathogenesis of neural circuits in depression.

Animals and ethical approval of the study
Adult male B6.Cg-TgN (Thy-YFP-H) 2Jrs (20-25 g) were kindly granted from prof.Jiangning Zhou's laboratory in University of Science and Technology of China.In these transgenic mice, a fluorescent protein (eYFP+) is constitutively expressed under the Thy1 gene promoter and distributed in the neuronal soma, axons, and dendrites in many brain regions including mPFC (Feng et al., 2000).Mice were housed at 21 ± 2 • C, with a light cycle between 08:00 h and 20:00 h.Food and water were provided ad libitum.Take a few of the tails of mice and identify transgenic mice by PCR.Primers, F: ACAGACACACACCCAGGACA; R: CGGTGGTGCAGA TGAACTT.Animals were treated according to the Guidelines of the Regulations of Experimental Animal Administration issued by the State Committee of Science and Technology of the People's Republic of China on November 14, 1988.All animal experiments were conducted with the approval of the Animal Use and Care Committee of Anhui Normal University (AHNU-ET2021025).

Retrograde tracing
Mice were anaesthetized with an intraperitoneal injection of 1% Sodium Pentobarbital (83 mg/kg) before they were placed in a mouse stereotaxic head holder (Anhui Zhenghua Biologic Apparatus Facilities Co., Ltd., Huaibei, China).After a skin incision, a hole was made on the skull over the precuneiform nucleus with a fine drill followed by the injection of 20-40 nl of PT-0100 rAAV-hSyn-mCherry-WPRE-hGH pA (PT-0100, Brain VTA Co., Ltd., Wuhan, China) solution into the M.-J.Ge et al.
precuneiform nucleus (AP 1.78, ML 0.31, DV − 2.65; 6 mice) with a 5 μl Hamilton syringe.The syringe was left in place for 15 min before it was withdrawn.The control group received normal saline injections into the precuneiform nucleus.At the end of the procedure, the skin was sutured and topical Tetracycline Hydrochloride Ointment (Northwest Synthetic Pharmaceutical Factory, Lanzhou, China) was smeared on the incision.Buprenorphine Hydrochloride (Tianjin Institute of Pharmaceutical Research Pharmaceutical Co., Ltd, Tianjin, China) was injected subcutaneously to relieve postoperative pain.The retrograde AAV injection of all mice was done by one experimenter, and then these mice were randomly divided into CON and CRS groups.

Chronic restraint stress
Three weeks after virus injection, the mice were randomly divided into the control group (CON, n = 6) and chronic restraint stress group (CRS, n = 6).The chronic restraint stress used in this study was performed as previously described (Seo et al., 2012, Seo et al., 2017).In brief, mice were individually placed headfirst into well ventilated 50 ML polypropylene conical tubes, which were then plugged with a 3-cm-long middle tube, and finally tied with a cap of the 50 ML tube.The mice were restrained for 2 h every day for 2 weeks.After each session of restraint, the animals were returned to their home environment, in which they were housed in pairs in normal plastic cages with free access to food and water.An outline of the study procedures and time intervals is given in Fig. 1.

Behavioral tests
Mice were brought to the test room 30 min before the start of each behavioral test and remained in the same room through the test.All behavioral tests were performed during the light cycle under lighting condition in 70 lx by experimenters blind to treatment information.

Sucrose preference test (SPT)
The SPT was performed as described previously (Willner et al., 1987, Zhu et al., 2018).Briefly, before the testing, mice were acclimated with two bottles of a 2% sucrose solution after 24 h food and water deprivation.One of the 2% sucrose bottles was replaced with tap water for the next 24 h.On the day of testing, mice housed in individual behavioral cages were given two bottles with 60 ML tap water and 60 ML 2% sucrose solution respectively over a 24 h period.The position of the two bottles was changed to avoid location preference after 12 h of testing.The percentage (sucrose intake / sucrose intake plus water intake) was represented as the sucrose preference.

Open field test (OFT)
The OFT was used to evaluate animal locomotor activity and emotional response (Walsh and Cummins, 1976).The apparatus was opaque and black all around, including a square 80×80 (cm) surrounded by a height of 40 cm walls, and the bottom was divided into 25 quadrates of 16×16 cm by a white line.In a quiet environment, each mouse was placed in the center area, which was defined as 9 squares in the center, and the behaviors were recorded for 5 min.The animals were placed in the laboratory room 30 min prior to testing to adapt to the environment.The device was cleaned with 70% ethanol thoroughly after each trial.The activity represented as number of movements and the number of line crossing were recorded by behavior video analysis system (Huaibei Zhenghua Biological Equipment Co., Ltd., Huaibei, China).

Novelty-suppressed feeding test (NSFT)
NSFT was performed by following the procedure of previous literature (Seo et al., 2017).Briefly, after 24 h food deprivation (water was provided ad libitum), mice were assayed by novelty-supressed feeding (NSF) test.At the end of this time, a single 2 × 2 cm oval food pellet was placed on a circular piece of white filter paper (150 mm diameter) positioned in the center of the open field (40 × 40 × 40 cm).Each mouse was placed in a corner of the open field.The latency to first bite the laboratory chow pellet and consumption over 15 min were recorded.Immediately after the mouse began to eat the chow, the tested animal was placed in its home cage alone with a weighed piece of chow for 30 min.At the end of this period, the amount of food consumed was determined by weighing the piece of chow.The open field was cleaned with 70% ethanol thoroughly after each trial.

Forced swimming test (FST)
The FST is considered to reflect the extent of despair of a rodent that it experiences (Porsolt and Jalfre, 1977) and is widely used in the evaluation of animal models of depression (Li et al., 2011).In this study, the mice were individually placed in a round glass cylinder (30 cm height × 16 cm diameter) with water at 25 ± 1 • C and a depth of 14 cm.The FST lasted for 12 min and the lasted 5 min was recorded as test session.The immovable time in 5 min was video-recorded for later analysis.After each session, the mice were removed from the water, well dried with paper towels, and returned to their clean home cages.

Tail suspension test (TST)
As previously described (Seo et al., 2012), mice were suspended individually by their tails from a metal rod fixed 50 cm above the surface of a table in a test room.The tip of the tail was fixed using adhesive tape.The subject mouse was suspended by the tail for 6 min and the immobility time was counted for the last 5 min.

Tissue preparation
Following the behavioral tests, the mice were deeply anesthetized with 1% Sodium Pentobarbital and perfused (via a transcardial approach) with 0.9% saline followed by euthanasia via decapitation.The brain tissues of mice were isolated and placed in 4% paraformaldehyde solution for 1 hour, then incubated in 20% and 30% sucrose solutions until brain tissues were sunk and stored in a − 20 degree refrigerator.These brain tissues were used for frozen sections using a Leica cryostat (Leica Biosystems Inc., Bufalo Grove, IL, USA), and for further image analysis and quantification of neurons or/and dendrites or/and dendritic spines.

Image analysis and quantification of neurons/dendrites
When comparing count between groups, no less than three sections were selected per mouse, and no less than three fields of view (×60 objective) were randomly selected for each statistical region in each section.In the IL of mPFC (between Bregma +2.22 mm and +1.98 mm), YFP-positive dendritic density statistics included three regions: the basal dendritic region, the apical proximal dendritic region (<50μm from the center of the neuronal body) and the apical distal dendritic region (>150μm from the center of the neuronal body) of pyramidal cells.Image analysis and quantification of the histological sections were performed by one author who was not aware of the experimental information.Images were captured under Laser scanning confocal microscope (OLYMPUS FV1000) and analyzed using image analysis software (ImageJ).The background was subtracted with a rolling value of 15, converted to 8-bit deep images and binarized using a determined threshold value (reduce noise 5, particles 2 -~).The percentage of the field of view covered by neurons or/and dendrites was measured for comparison between CON and CRS in the distribution density of neurons or/and dendrites (Di Cunto et al., 2017).

Quantification of dendritic spines
Analysis of the density and classification of dendritic spines were mainly based on the methods described in the previous literatures (Sacktor et al., 2013, Foster Olive et al., 2018).Concisely, at the basal dendrites, apical proximal dendrites and apical distal dendrites in mPFC IL layer 2/3, secondary or tertiary dendritic segments of pyramidal neurons were selected for analysis.The number of neurons used for analysis in each section is not less than 10, and the total length of the dendritic segments is greater than 300 microns.Z-stacks of dendrites (up to 80 microns total on Z-axis; optical section thickness = 0.5 mm, i.e., 160 images per stack) were obtained at 60×6 magnification on an OLYMPUS FV1000.IMARIS 7.5 (Andor Technology; Belfast, Northern Ireland) was used to reconstruct z-stacks into 3D models for analysis.Using customized settings based on spine parameters as previously described (Zhang et al., 2016, Foster Olive et al., 2018, Xu et al., 2021), IMARIS Filament Tracer module was used to detect, quantify, and characterize spine structures (Table 1).

Statistical analyses
In this study, all data were analyzed with SPSS v22.0 software (IBM, New York, NY, USA) and expressed as the mean ± standard error of the mean (SEM).Using the Levene test for homogeneity of variance, data with a P value greater than 0.05 were analyzed parametrically, and data with a P value less than 0.05 were analyzed non-parametrically.Parametric data were analyzed using Independent-Samples T test.Nonparametric data were analyzed using the Mann-Whitney U test.P < 0.05 was considered to be statistically significant.The charts are all plotted by GraphPad Prism 8.0.

Chronic restraint stress induced depression-like behaviors
Independent-Samples T test showed that there was a significant decreased sucrose consumption in SPT (P < 0.05) (Fig. 2A), an increase of the latency time in NSFT (P < 0.05) (Fig. 2B), a downregulated activity time in the center regions in OFT (P < 0.05) (Fig. 2E) and an extended immobility time in FST (P < 0.05) (Fig. 2G) in CRS when comparison with control mice.No significant difference was found in the immobility time in TST between CON and CRS (P > 0.05) (Fig. 2F).

Chronic restraint stress resulted in decrease of neurons and loss of dendrites/dendritic spines in mPFC
The injection site of retrograde AAV in the infralimbic cortex (IL) (AP 1.78 mm, ML 0.31 mm, DV 2.75 mm) of mPFC was shown in Fig. 3A.
In the ipsilateral mPFC, the neurons labeled by the retrograde tracing virus were widely found in distinct regions, including the area 1 of cingulate cortex (cg1), area 2 of cingulate cortex (cg2), prelimbic cortex (PrL), IL, medial orbital cortex (MO), and dorsal peduncular cortex (DP) (Fig. 2B).In these regions, the green fluorescent protein YFP was selectively labeled in the pyramidal cells of layer 5 (L5), and the red fluorescent protein was distributed in the neurons found mainly in the L2/3, L5, and L6, as well as rarely in the L1 (Fig. 3G).However, the pyramidal neurons co-labeled with two fluorescent proteins can be rarely seen (Fig. 3C-E).Chronic restraint tress induced a general decrease in the distribution density of red fluorescence positive neurons in multiple subregions of mPFC, including Cg1, Cg2, PrL, IL, and DP (Table 2).Mann-Whitney U test displayed that there was a significant difference in the distribution density of red fluorescence labeled neurons between CON and CRS in Cg1 (P < 0.01) and Cg2 (P < 0.001) (Fig. 3J-M).
In the contralateral mPFC, the red fluorescence labeled neurons, from the structural levels, were mainly distributed in the L2/L3, L5, and L6, and from the perspective of regional composition, were located in the Cg1, Cg2, PrL, IL, MO, and DP, while there was a lower distribution density of red fluorescence positive neurons in these layers or regions when compared with that of the ipsilateral mPFC (Table 2).In addition, chronic restraint stress resulted in a decline in the distribution density of red fluorescence positive neurons in Cg1 and Cg2.There was a significant difference in the distribution density of red fluorescence positive neurons between CON and CRS in the IL (P < 0.05) (Fig. 3F).
Quantitative analysis of YFP fluorescent protein positive dendrites/ dendritic spines in the IL of mPFC disclosed that chronic restraint stress led to a reduced distribution density of dendrites in the basal dendrites (P < 0.0001) (Fig. 4B) and apical proximal (P < 0.0001) (Fig. 4C) and distal (P < 0.001) (Fig. 4D) dendrites, and a significant loss of dendritic spines in the apical distal dendrites (P < 0.05) (Fig. 4I) as well as an obvious downregulation in the density of stubby-type spines in the apical distal dendrites (P < 0.05) (Fig. 4T).However, there was no significant difference between CON and CRS in the distribution density and type of dendritic spines in the basal and apical proximal dendrites as well as in the densities of other three dendritic spine types, filopodia, long thin, and mushroom, in the apical distal dendrites (Fig. 4G, H, J-S,  and U).

Chronic restraint stress reduced afferent projections of the mPFC from olfactory cortex
There were retrograde tracing positive neurons found in multiple regions of olfactory cortex in a degree from dense to very scarce in both ipsilateral and contralateral olfactory cortex (Fig. 5A, C, D and E) (Table 2).In these regions, Independent-Samples T test revealed that chronic restraint stress significantly reduced the distribution density of neurons labeled by red fluorescence in the ipsilateral DTT (P < 0.05)

Chronic restraint stress led to reduction of afferent projections of the mPFC from other cortices
In both the ipsilateral and contralateral cortex, the red fluorescence positive neurons in a degree from dense to very scarce were detected in multiple cortices, including the dorsal part of agranular insular cortex (AID), the ventral part of agranular insular cortex (AIV), the primary auditory cortex (Au1), the caudal interstitial nucleus (Cl), the ventral area of secondary auditory cortex (AuV), the dysgranular insular cortex (DI), the ectorhinal cortex (Ect), the granular insular cortex (GI), the primary motor cortex (M1), the secondary motor cortex (M2), the perirhinal cortex (PRh), the retrosplenial agranular cortex (RSA), the retrosplenial granular cortex (RSG), the primary somatosensory cortex (S1), the secondary somatosensory cortex (S2), the temporal association cortex (TeA), the primary visual cortex (V1), the lateral area of secondary visual cortex (V2L), the mediolateral area of secondary visual cortex (V2ML), and the mediomedial area of secondary visual cortex (V2MM) (Fig. 5A, B, F and I) (Table 2).
In these cortices, almost all of red fluorescence positive neurons were also co-labeled by YFP (Fig. 6C-E).
In some of these cortices, the distribution density of red fluorescent positive neurons in CRS presented a decrease in a degree compared with CON (Table 2).Mann-Whitney U test indicated that there was a significant difference between CON and CRS in the distribution density of neurons labeled by red fluorescent protein in the AID (P < 0.01) (Fig. 6G) and AIV (P < 0.05) (Fig. 6H) of ipsilateral cortex, as well as in the AID (P < 0.001) (Fig. 6J) and AIV (P < 0.0001) (Fig. 6K) of contralateral cortex.
Both ipsilateral and contralateral amygdala presented retrograde tracing red fluorescence positive neurons in a degree from very dense to scarce (Table 2).
Independent-Samples T test showed that chronic restraint stress caused a significant upregulation in the distribution density of red fluorescence positive neurons distributed in the anterior part (BLA) (P < 0.05) (Fig. 7A, B and G) and posterior part (BLP) (P < 0.05) (Fig. 7C, D  and H) of basolateral amygdaloid nucleus, and the posterior part of basomedial amygdaloid nucleus (BMP) (P < 0.01) (Fig. 7I) in the ipsilateral amygdala, as well as in the BLA of the contralateral amygdala (P < 0.05) (Fig. 7E, F and J).

Chronic restraint stress downregulated afferent projections of the mPFC from septum/basal forebrain
Scarce or very scarce neurons labeled by red fluorescence were detected in distinct subnucleus of the septum/basal forebrain in both ipsilateral and contralateral cortex (Fig. 9A and B) (Table 2).

Chronic restraint stress decreased afferent projections of the mPFC from hypothalamus
In multiple regions of both ipsilateral and contralateral hypothalamus, there were neurons labeled by red fluorescence ranging from moderate to very scarce (Fig. 10A and C) (Table 2).
Independent-Samples T test unveiled that chronic restraint stress reduced the distribution density of red fluorescence labeled neurons in the dorsomedial hypothalamic nucleus (DM) of both ipsilateral (P < 0.05) (Fig. 10B) and contralateral (P < 0.05) (Fig. 10D) hypothalamus.

(caption on next page)
M.-J.Ge et al.

Chronic restraint stress begot an afferent projection decline of the mPFC from thalamus
The subnucleus of thalamus presented red fluorescence positive neurons in a degree from dense to very scarce (Fig. 11A-F) (Table 2).
In the contralateral thalamus, scarce or very scarce density of red fluorescence positive neurons was found in the subnucleus (Table 2).Independent-Samples T test showed that there was a significant difference in the distribution density of red fluorescence positive neurons between CON and CRS in the VM (P < 0.001) (Fig. 11O) and VRe (P < 0.05) (Fig. 11P).

Chronic restraint stress did not induce a significant change in the afferent projections of mPFC from the midbrain and brainstem
In both ipsilateral and contralateral midbrain and brainstem, scarce and very scarce red fluorescence positive neurons were observed in multiple regions, including the periaqueductal gray (PAG), dorsal raphe nucleus (DR), paramedian raphe nucleus (PMnR), oral part of pontine reticular nucleus (PnO), pedunculopontine tegmental nucleus (PPTg), parvicellular part of red nucleus (RPC), compact part of substantia nigra (SNC), reticular part of substantia nigra (SNR), and ventral tegmental area (VTA) (Table 2).In these regions, however, no significant difference was found between CON and CRS.

Discussion
In the present study, chronic restraint stress induced depression-like behaviors, including a decrease of sucrose preference in SPT, a decline of activity time in the center regions in OFT, and an increase of the latency time in NSFT and immobility time in FST.Meanwhile, chronic restraint stress also caused a downregulation in the distribution density of retrograde labeled neurons in the Cg1 and Cg2 of ipsilateral mPFC, and in the IL of contralateral mPFC, and a general loss of dendrites and dendritic spines in mPFC.In addition, chronic restraint stress resulted in a reduction in the afferent projections of mPFC from multiple cortices such as the AID, AIV, vHIP, and some areas of the olfactory cortex, as well as several septum/basal forebrain, hypothalamic, thalamic, and midbrain and brainstem subnuclei, while an increase in the afferent projections of mPFC from the BLA, BLP and BMP of the amygdala.Here we used the schematic diagrams to outline the effects of CRS on the neural pathways of mPFC subregions (Fig. 12.) and different brain regions (Fig. 13.) projecting to the IL of mPFC.

Effects of CRS on the local circuit in mPFC
In the past decades, although a large number of literatures have been accumulated, broad agreement on the subdivision of rodent mPFC and whether there are strict functional-anatomical analogues between primate and rodent mPFC (TM, 1995, Laubach et al., 2018, Jacobs and Moghaddam, 2021) remains elusive (Howland et al., 2022).One of the most important reasons for the controversy of mPFC subdivision may be derived from the confusion of acronyms.For example, the prelimbic cortex has always had three different acronyms, PL, PrL, and Cg3, used simultaneously in different literatures (Laubach et al., 2018).In this study, we defined the range of mPFC between Bragma 4.68 and -0.12 referring to Palomero Gallagher and Zilles (2015) (Palomero-Gallagher and Zilles, 2015), and designated the mouse mPFC subdivisions as Cg1, Cg2, PrL, IL, MO, and DP, respectively, according to the nomenclature of Paxinos and Franklin (Paxinos, 2001) and general views (Heidbreder andGroenewegen, 2003, Palomero-Gallagher andZilles, 2015).
In comparison with the classic sensory cortex that is divided into six main layers, L1 to L6, the mPFC lacks a canonical thalamo-recipient L4, but has similar cell types including intratelencephalic (IT) cells, pyramidal tract (PT) cells, and corticothalamic (CT) cells (Anastasiades and Carter, 2021).Our results showed that morphologically, green fluorescence labeled neurons should be the PT cells located in L5 (Feng G, 2000), while red fluorescence positive neurons should be the IT cells distributed in L2/3, L5, and L6, and the CT cells detected in L5 and L6 (Harris and Shepherd, 2015) as well as the two fluorescent co-labeled PT cells very rarely found in L5, suggesting that the afferent projections of IL in the local circuit of mPFC mainly came from the IT and CT cells and rarely from the axon lateral branches of PT cells.
Several meta-analyses have consistently revealed that compared with the healthy control group, the volume of mPFC in patients with major depressive disorder (MDD) is significantly reduced (Campbell et al., 2004, Videbech, 2004, Koolschijn et al., 2009, McKinnon et al., 2009, Cole et al., 2011, Kempton et al., 2011, Arnone et al., 2012, Bora et al., 2012, Du et al., 2012, Lai, 2013, Arnone et al., 2016, Wise et al., 2016, Belleau et al., 2019).Moreover, individuals with MDD and an anxiety disorder showed more pronounced mPFC reductions than those with just one disorder (van Tol MJ, 2010).Autopsy studies of patients with MDD indicated that only the volume of neurons in mPFC decreased, but the number of pyramidal neurons did not decrease (ajkowska and Miguel-Hidalgo, 1999).Our results suggest that the dendrite/dendritic spine loss widely occurred in the basal and apical dendrite areas may be the structural basis for the reduction of mPFC volume in MDD, and that the effect of CRS on the synaptic plasticity of mPFC may be due to the loss of both presynaptic projective neurons/axon lateral branches and postsynaptic dendrites/dendritic spines.

Effects of CRS on the inputs of olfactory cortex to mPFC
A large number of clinical studies have shown that many patients with neurological or psychiatric diseases, such as depression, express olfactory dysfunction (Bhattarai et al., 2022).Retrograde tracing from the mPFC showed that the afferent neurons of mPFC were mainly located in the anterior olfactory nucleus (AON) and taenia tecta (TT) (Hoover WB, 2007, DeNardo et al., 2015, Biskamp et al., 2017, Moberly et al., 2018), and these neurons were glutamatergic neurons (Moberly et al., 2018).The present study disclosed similar results that retrograde tracing labeled neurons were distributed mainly in DTT of the ipsilateral Fig. 4. Effects of chronic restraint stress on the density of dendrites/dendritic spines of the mPFC.(A) showed the YFP labeled neurons of mPFC at the Bregma 1.78 mm.The densities of dendrites were calculated from the basal (B), proximal (C), and distal (D) areas, respectively.A slice view (E) acquired by laser scanning confocal microscope (FV1000, 60 × 6 for objective magnification) and its trace (F) at the apical proximal dendritic segments labeled by transgenic YFP protein in CON.The densities of dendritic spines were calculated from the basal (G), proximal (H), and distal (I) dendrites, respectively.The density statistics of dendritic spines (four types of dendritic spines: filopodia, long thin, stubby, and mushroom) in basal dendrites (J, K, L, M), proximal dendrites (N, O, P, Q), and distal dendrites (R, S, T, U), respectively.The three white boxes in (A) indicate the sampling locations of dendrite statistics from basal area, proximal area, and distal area, respectively.The bar is 100 μm in (A).Data expressed as the means ± SEM (n = 6).*P < 0.05, ***P < 0.001, ****P < 0.0001.
M.-J.Ge et al.   olfactory cortex and very scarcely in other subregions of olfactory cortex, and a significant decreased input of the IL from DEn in both ipsilateral and contralateral olfactory cortex, suggesting a potential mechanism of olfactory dysfunction through the reduced projections mentioned above.

Effects of CRS on the inputs of other cortices to mPFC
Accumulated literatures have reported that virtually all regions of the mouse isocortex send projections to mPFC.A major proportion of the projection regions comes from the agranular forebrain cortex adjacent to the mPFC.In sensory cortex, the projections to mPFC mainly derive from S1, but less from V1 and Au1 (Ahrlund-Richter et al., 2019).Using rabies virus trans-synaptic tracing, DeNardo and colleagues found that the input to mPFC L5 originated from multiple prefrontal cortices in mice including agranular insula (AI), motor cortex, and contralateral mPFC (DeNardo et al., 2015).In rats, retrograde tracing injected by fluorogold (FG) at IL showed that the mPFC received inputs from other cortices including frontal polar cortex (lateral and medial divisions), retrosplenial cortex, agranular insular cortex (dorsal, posterior and ventral divisions), perirhinal cortex, entorhinal cortex (lateral and medial divisions), and ectorhinal cortex (Hoover and Vertes, 2007).
Our results revealed that in mice, although retrogradely labeled neurons were distributed in virtually all cortices of the isocortex, the CRS induced decrease in the distribution density of retrogradely labeled neurons presented only in AID and AIV of both ipsilateral and contralateral cortex, suggesting a limited effect of CRS on the inputs of other cortices to mPFC.

Effects of CRS on the inputs of amygdala to mPFC
A large number of studies have proved that there are reciprocally connections between mPFC and amygdala (H, 2000, Chang et al., 2013, Ruff, 2014, DeNardo et al., 2015, Janak and Tye, 2015, Chen and Hong, 2018, Wittmann et al., 2018, Ährlund-Richter et al. 2019, Gangopadhyay et al., 2021, Bhattarai et al., 2022).These connections are found in humans, nonhuman primates, and rodents, showing evolutionarily conservatism (Gangopadhyay et al., 2021).Injection with FG at rat IL retrogradely traces the FG labeled neurons in multiple amygdala subdivisions including the basolateral nucleus of amygdala (BLA),  (F, G, and H) respectively showed the neurons with green, red, and their merged fluorescence in the CA1 of CON and CRS at the Bregma − 1.46.The comparative schematic diagrams of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS were from the ipsilateral (I) and contralateral (J) hippocampus at Bregma − 2.46 mm, respectively.(K), (L), (M), and (N) represented the effects of CRS on the density of red fluorescence positive neurons in the ipsilateral vDG, vCA1, vCA2, and vCA3, respectively.(O), (P), (Q), and (R) represented the effects of CRS on the density of red fluorescence positive neurons in the contralateral vDG, vCA1, vCA2, and vCA3, respectively.Data expressed as the means ± SEM (n = 6).*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.Fig. 9. Effects of chronic restraint stress on the density of neurons in septum/basal forebrain projecting to the mPFC.The comparative schematic diagrams of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS were from the septum/basal forebrain at Bregma − 0.10 mm (A) and the AcbC at Bregma 0.98 mm (B), respectively.(C), (D), (E), (F), and (G) showed the effects of CRS on the density of red fluorescence positive neurons in the ipsilateral LSI, the contralateral LSI, the ipsilateral LSD, the contralateral LSD, and the ipsilateral AcbC, respectively.Data expressed as the means ± SEM (n = 6).*P < 0.05, **P < 0.01, ***P < 0.001.posterior of basomedial nucleus of amygdala (BMP), amygdalo-piriform transition zone (TR), cortical nucleus of amygdala (COA), lateral nucleus of amygdala (LA), and posterior nucleus of amygdala (PA) (Hoover and Vertes, 2007).In mouse, the brain-wide maps generated by a retrograde trans-synaptic rabies virus system demonstrated that the excitatory neurons and four inhibitory interneuron subtypes, SST43, PV, VIP, and SOM INs, in the mPFC receives pronounced projections from the BLA.In addition, other amygdala subdivisions also send projections directly to the mPFC (Ahrlund-Richter et al., 2019).In layer-specific input patterns, more minor inputs of L5 mPFC neurons originated from the BLA (DeNardo et al., 2015).Recent studies have identified the inputs of mPFC from the medial amygdala nucleus (Pardo-Bellver et al., 2012, Ahrlund-Richter et al., 2019), as well as from the GABAergic projections in the centromedial amygdala (Seo et al., 2016, Ahrlund-Richter et al., 2019, Sun et al., 2019).
The bidirectional communications between mPFC and BLA are important for multiple emotional states (Burgos-Robles et al., 2017, Bloodgood et al., 2018), especially for mediating synergistic interactions to enable goal-directed behaviors based on affective and reward-related information (Gangopadhyay et al., 2021).By using functional magnetic resonance imaging (fMRI), Sun et al. found that the BLA-mPFC functional connectivity was increased in obsessive-compulsive disorder (OCD) mice (Sun et al., 2019).The activity of BLA projecting to the prelimbic cortex was increased during auditory fear conditioning, and enhanced responses of prelimbic cortex to shock-predicting cues were observed after learning (Burgos-Robles et al., 2009, Calhoon andTye, 2015).Some studies demonstrated that BLA inputs are particularly enriched in L2 of PL (Little andCarter, 2013, Wang et al., 2016), and CRS can cause the dysregulation occurring in BLA projection neurons receiving mono-directional inputs from dorsomedial prefrontal cortex (dmPFC→BLA) rather than those reciprocally connected with dmPFC (dmPFC↔BLA) (Liu et al., 2020).Together with our results, that CRS induced an increased input to the IL of mPFC from BLA, BLP, and BMP, suggests a possible hypothesis that the CRS induced dysregulation occurring in dmPFC→BLA and vmPFC↔BLA, but not in BLA→dmPFC.

Effects of CRS on the inputs of hippocampus to mPFC
Intensive investigation revealed that direct and mutual information processing between the mPFC and hippocampus plays a role in mnemonic encoding and emotional regulation (Spellman et al., 2015, McGarry and Carter, 2017, Chao et al., 2020).A direct monosynaptic projection to the mPFC mainly comes from ventral hippocampus (vHPC) including CA1 and subiculum, and rarely from dorsal hippocampus, in both mice (Liu andCarter, 2018, Chao et al., 2020) and rats (Hoover and Vertes, 2007).Liu and Carter found that the direct projection from the vHPC to the mPFC primarily enters the IL L5, followed by the IL L2/3 and PL L5, and less enters the PL L2/3 and L6 (Liu and Carter, 2018).Moreover, the neurons in mPFC receiving hippocampal inputs include glutamatergic pyramidal neurons and GABAergic interneurons (Chao et al., 2020).Recent a study found that there are two parallel circuits of hippocampal projection to PFC located in the superficial or deep pyramidal layers of the CA1/subiculum border (Hu et al., 2023).
Basic and clinical studies revealed that a significant reduced volume is observed in some brain regions such as mPFC and hippocampus, which are associated with regulating mood and cognition (Jin et al., 2016).Another investigation demonstrated that the hippocampal volume significantly decreases in the depressed patients with recurrent  Our results showed that reduced inputs of hippocampus to mPFC mainly originates from vHPC in CRS mice, suggesting that depression is closely related to reduced hippocampal, especially ventral hippocampal projections to the mPFC.

Effects of CRS on the inputs of septum/basal forebrain to mPFC
Structurally, there is a reciprocal projection between mPFC and the LSD (Risold and Swanson, 1997, Chiba and Nakano, 2001, Vertes, 2004), which preferentially innervates somatostatin-expressing (SST+) GABAergic neurons in the mouse mPFC (Sun et al., 2019) and has been considered as a component of antidepressant actions (Contreras and Marván, 1988, Contreras et al., 1990, Sheehan et al., 2004, Contreras et al., 2018).The vertical and horizontal diagonal band of Broca (VDB/HDB) nuclei also maintains reciprocal projections with the mPFC (Chaves-Coira et al., 2018).In rats, injections of FG in the ventral regions of IL resulted in a considerable amount of labeled neurons in all rostro-caudal levels of HDB/VDB (Chaves-Coira et al., 2018) and in the claustrum, an appropriate number of labeled neurons in the taenia tecta and endopiriform nucleus, and a rare number of labeled neurons in the septum and substantia innominate (Hoover and Vertes, 2007).Using monosynaptic rabies viral tracers in combination with fluorescence micro-optical sectioning tomography, the inputs from the nucleus of the diagonal band to SST+ neurons in the mPFC of mice were identified more than those to parvalbumin-expressing (PV+) and vasoactive intestinal peptide-expressing (VIP+) neurons.In three subtypes of GABAergic neurons of IL area, PV+ neurons received the smallest input from the substantia innominate, while all three types of GABAergic neurons in the mPFC were directly innervated by cholinergic neurons in the basal forebrain (Sun et al., 2019).
In addition, the research on the connection between mPFC and AcbC mainly focuses on the structure and function of the projection of mPFC to AcbC.Amygdala stimulation and Pavlovian conditioned stimuli resulted in selective activation of mPFC projection neurons to AcbC (McGinty and Grace, 2008).IL inactivation caused a general reduction in conditioned suppression following discriminative conditioning (Piantadosi et al., 2020).Acute optogenetic stimulation of the mPFC-NAc increased object recognition (Papp et al., 2023).After acute sleep deprivation, there was selective reduction of the glutamate release probability at the mPFC-to-NAc synapses (Liu et al., 2016).
Our results showed that injection with the retrograde AAV in IL only generated scarce or very scarce labeled neurons in the septum/basal forebrain, suggesting that the projection of these brain areas to mPFC don't undertake a critical effect of CRS.

Effects of CRS on the inputs of hypothalamus to mPFC
Previous studies have shown that in rats, few nerve fibers project directly from the lateral hypothalamus (LH) to the mPFC (Date Y et al., 1999, Fadel, 2002).Another research by injection with the retrograde tracer, FG, into the IL of the mPFC revealed that labeling was considerably less intense in the periaqueductal gray (PAG) and supramammillary nucleus (SUM) (Hoover and Vertes, 2007).However, recent studies have shown that the mPFC-LH pathway plays a critical role in many aspects.Recording in vivo the population calcium activity revealed that the glutamatergic mPFC-LH pathway was activated in response to acute stressors and was inhibited during food consumption (Clarke et al., 2023).MPFC cells that project to the LH promote dominance behavior during reward competition (Padilla-Coreano et al., 2022).
Our results indicated that in mice, projections to the mPFC was moderate from PH, and scare from DM and LH, while significantly decreased projections from DM to the mPFC originated from response to the CRS, suggesting a possible role of DM-mPFC pathway in the formation of CRS-induced depression-like behaviors.

Effects of CRS on the inputs of thalamus to mPFC
The thalamus forms a diverse and complex set of cytoarchitectonically, chemoarchitectonically, and hodologically distinct nuclei and has tight connectional and functional relationships with the cortex (Vertes et al., 2015).Inputs from thalamus to the mPFC are closely related to various cognitive functions, including learning, attention, working memory, and arousal (Parnaudeau et al., 2013, Bolkan et al., 2017, Schmitt et al., 2017, Honjoh et al., 2018).The axonal terminals of focal thalamic projections were intensive in the middle layers, and sparser diffuse in the layer I, of the mPFC (Mair et al., 2022).These  thalamo-cortical projections can activate mPFC excitatory networks and feedforward inhibition (HJ, 1988, Kuroda et al., 1996, Kuroda and Murakami et al., 1998, Xiao et al., 2009, Collins et al., 2018).
Although the inputs to mPFC come from multiple thalamic nuclei including VM, most of the current works has still focused on the MD thalamus (Collins et al., 2018, Ährlund-Richter et al. 2019, Sun et al., 2019, Anastasiades et al., 2021).Recent evidence showed that MD enhances cortical connectivity and regulates the signal processing properties of mPFC neurons through distinct thalamic-cortical pathways, which compensate for the uncertainty associated with low signal or high noise (Schmitt et al., 2017, Mukherjee et al., 2021).There are two prominent axonal bands identified in the L1 and L3 of mPFC from MD inputs (Anastasiades and Carter, 2021).Among the three types of neocortical excitatory cells (IT, PT and CT neurons) (Harris and Shepherd, 2015), the IT cells of mPFC L3 are strongly activated by the inputs of MD, which can induce these IT cells to produce action potentials (Collins et al., 2018).Similarly, there is a dense axonal band in the L1 of mPFC from VM inputs, which mainly engages the distal apical dendrites of PT cells (Collins et al., 2018, Anastasiades et al., 2021).This thalamo-cortical dendritic targeted connection mode can facilitate the activation of PT cells, thus enhancing the output of mPFC (Anastasiades et al., 2021).
In addition, in L6 or more deeper layer of mPFC, there is no obvious axonal band from thalamic inputs (Crandall et al., 2017, Collins et al., 2018).The stimulation of anteromedial thalamic nucleus (AM) neurons projecting to the mPFC reinforces goal-directed behavior and activates dopamine neurons (Yang et al., 2022).On the contrary, the mPFC projection neurons targeting the lateral MD also regulate adaptive control to flexibly optimize behavioral responses in goal-directed behavior (Bruinsma et al., 2022).Postsynaptic thalamic neurons process synaptic inputs from mPFC, highlighting the mechanisms that potentially amplify distinct pathways underlying cognitive control of behavior, while chemogenetic silencing of mPFC projections to lateral and medial mediodorsal thalamus subregions oppositely regulates cognitive control (de Kloet et al., 2021).
In rats, injections with the FG into the IL of mPFC give rise to the retrograde labeling at a heavy density in the nucleus reuniens of thalamus (RE), rhomboid nucleus of thalamus (RH), paratenial nucleus of thalamus (PT), paraventricular nucleus of thalamus (PV), and MD, at a moderate density in the interanteromedial nucleus of thalamu (IAM)s, and at a light density in the intermediodorsal necleus of thalamus (IMD) and nucleus pontis oralis (RPO) (Hoover and Vertes, 2007).Our results showed that in mice, the thalamus neurons labeled by retrograde AAV injected into IL were distributed densely in AMV, Rt, and VM, moderately in AM, PC, and VRe, and scarcely in MD and Re, as well as were reduced in these subnuclei of CRS, suggesting that decreased inputs of these subnuclei of thalamus to mPFC may play an important role in CRS-induced depression.
Locus coeruleus (LC) on the PFC have a mutual excitatory influence (Sara and Bouret, 2012).The NE neurons of LC in the brainstem are the main source of noradrenergic projections to the mPFC (Hui and Beier, 2022).The mPFC state can be rapidly modulated by LC phasic activity that follows a direct electrical stimulation (Marzo et al., 2014).LC neurons innervating the mPFC undergo less excitable in response to stress and appear to promote exploration (Borodovitsyna et al., 2020).
It is well known that neurotransmitter 5-HT can regulate multiple brain functions such as emotion (Cools et al., 2008, Dayan and Huys, 2009, Cohen et al., 2015).The raphe nucleus of the brain stem, especially the dorsal raphe nucleus (DRN), is the main distribution areas of 5-HT neurons.The 5-HT neurons in these nuclei sent extensive and widespread axonal projections to the entire forebrain including the mPFC (Moore and Jones, 1978).Injection in the mPFC of retrograde virus leads to a strong fluorescence signal distributed in the raphe (Geddes et al., 2016).In male rats, chronic unpredictable stress (CUS) induces a depression-like behavior accompanied by decreased 5HTergic innervation of the mPFC, while pretreatment with MK801, the NMDA receptor inhibitor, blockes the decrease in the projection of DRN to the mPFC (Natarajan et al., 2017).
Previous studies showed that FG injection into the IL of rat mPFC retrogradely labeled the neurons in multiple subareas of the midbrain and brainstem, including distribution with densely in the DRN and LDT, moderately in the VTA, CLi, MR, and LC, and sparsely in the PAG, SNc, NI, and RPO (Hoover and Vertes, 2007).The retrograde tracing revealed that in mice, LC neurons projecting to mPFC were more numerous in females than that in males (Cardenas et al., 2021), which was similar to our results that retrograde labeling in a degree from scarce to very scarce was found in the midbrain and brainstem projecting to the mPFC in the male mice.

Conclusion
In conclusion, chronic restraint stress caused a significant loss of dendrites/dendritic spines in the IL of mPFC and a obvious reduction in input neurons of multiple cortices to the mPFC as well as an increase in input neurons of amygdala to the IL of mPFC, suggesting a possible mechanism that CRS leads to disturbances in synaptic plasticity by affecting multiple inputs to the IL of mPFC, ultimately causing depression-like behaviors.

Declaration of Competing Interest
The authors declare that there is no conflict of interest.experimental methods, Ms. Wen-Qing Mei, Yan Zhang, Xiu-Ling Zhang, Xin Chen, Yue Cheng, Yu-Xuan Wang, and Min Wu for help in the experiment, and Mr. Hao Yan and Ms. Xiao-Yan Ma for help in the confocal photomicrographs.This work was supported by the National Natural Science Foundation of China (No. 30470537), The University Synergy Innovation Program of Anhui Province (GXXT-2020-030), The Natural Science Foundation of the Department of Education, Anhui Province (ZD2008006-1), Anhui Provincial Key Laboratory of the Conservation and Exploitation of Biological Resources, Anhui Provincial Key Laboratory of Molecular Enzymology and Mechanism of Major Diseases, Key Laboratory of Biomedicine in Gene Diseases and Health of Anhui Higher Education Institutes and the Innovation Team of Scientific Research Platform in Anhui Universities.

Fig. 1 .
Fig. 1.Schematic diagram of the mouse experiment process.CRS, chronic restraint stress; behavioral tests include the sucrose preference test (SPT), noveltysuppressed feeding test (NSFT), open-field test (OPT), tail-suspension test TST), and forced swim test (FST).The mice were sacrificed after the fourteenth week for the next experiment.

Fig. 2 .
Fig. 2. Effects of CRS on the behavior performance of mice.Chronic restraint stress of 2 h daily for consecutive 2 weeks produced a decline in the sucrose consumption in SPT (A), an extended latency to feed in NSFT (B), a decrease in the time spent in the center area in OFT (E), and a prolonged immobility time in FST (G), but did not result in a significant increase in the immobility time in TST (F).(C) and (D) represented the travel traces of mice in OFT from the CON and CRS, respectively.CON, control; CRS, chronic restraint stress; SPT, sucrose preference test; NSFT, novelty-suppressed feeding test; OFT, open field test; TST, tail suspension test; FST, forced swim test.Data expressed as the means ± SEM (n = 6).*P < 0.05.

Fig. 3 .
Fig. 3. Effects of chronic restraint stress on afferent neurons into the IL from other regions of the mPFC.(A) showed the sites in CON and CRS (indicated by the arrows) injected by the retrograde AAV in the IL of mPFC (AP 1.78 mm, ML 0.31 mm, DV − 2.75 mm) at Bregma 1.78 mm, and five grades of dot density, ranging from very dense (++++) to very scarce (+/-), set in the schematic diagram to reflect stereology observations (Table 2).(B) showed a comparative schematic diagram of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS in the ipsilateral mPFC at Bregma 1.78 mm.The green (C), red (D), and yellow (E) fluorescence were derived from the YFP gene, the retrograde AAV, and their merged fluorescence, respectively.(F) indicated the effects of CRS on the density of red fluorescent protein-positive neurons in the contralateral IL. (G) showed the layers of mPFC, including L1, L2/3, L5 and L6.(I) showed the contralateral IL indicated by the red box in (H) at Bregma 1.78 mm from CON and CRS.(K) showed the ipsilateral Cg1 and Cg2 indicated by the red box in (J) at Bregma 1.18 mm from CON and CRS.(L) and (M) indicated the effects of CRS on the density of red fluorescent protein-positive neurons in the ipsilateral Cg1 and Cg2, respectively.In (E), the right arrow indicated the neuron labeled by only red fluorescence, and the left arrow indicated the neuron co-labeled by green and red fluorescence.Data expressed as the means ± SEM (n = 6).*P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 5 .
Fig. 5. Effects of chronic restraint stress on the projections to the mPFC from olfactory cortex.The comparative schematic diagrams of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS were from the ipsilateral (A) and contralateral (C) olfactory cortex at Bregma 2.46 mm, and from the ipsilateral (E) and contralateral (D) DEn at Bregma 0.98 mm, respectively.The comparative schematic pictures of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS were from the ipsilateral DTT at Bregma 2.46 mm (B) and the ipsilateral DEn at Bregma 0.98 mm (F), respectively.(G), (H), (I), and (J) showed the effects of CRS on the density of red fluorescence positive neurons in the ipsilateral DTT, contralateral DTT, ipsilateral DEn, and contralateral DEn, respectively.Data expressed as the means ± SEM (n = 6).*P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 6 .
Fig. 6.Effects of chronic restraint stress on the density of neurons in other cortices projecting to the mPFC.The comparative schematic diagrams of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS were from the ipsilateral other cortices at Bregma 0.98 mm (A) and at Bregma − 2.18 mm (B), and from the contralateral other cortices at Bregma 0.98 mm (I), respectively.The green (C), red (D), and yellow (E) fluorescence were derived from the YFP gene, the retrograde AAV, and their merged fluorescence, respectively.(F) showed a comparative schematic picture of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS in the ipsilateral AI at the Bregma 0.98 mm.(G) and (H) showed the effects of CRS on the density of red fluorescence positive neurons in the ipsilateral AID and AIV, respectively.(J) and (K) represented the effects of CRS on the density of red fluorescence positive neurons in the contralateral AID and AIV, respectively.Data expressed as the means ± SEM (n = 6).*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 7 .
Fig. 7. Effects of chronic restraint stress on the density of neurons in amygdala projecting to the mPFC.The comparative schematic diagrams (A) and pictures (B) of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS were from the ipsilateral BLA at Bregma − 1.58 mm.The comparative schematic diagrams (C) and pictures (D) of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS were from the ipsilateral BLP at Bregma − 2.18 mm.The comparative schematic diagrams (E) and pictures (F) of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS were from the contralateral BLA at Bregma − 1.58 mm.(G), (H), (I), and (J) represented the effects of CRS on the density of red fluorescence positive neurons in the ipsilateral BLA, BLP, and BMP, and contralateral BLA, respectively.Data expressed as the means ± SEM (n = 6).*P < 0.05, **P < 0.01.

Fig. 8 .
Fig. 8. Effects of chronic restraint stress on the density of neurons in hippocampus projecting to the mPFC.(A) and (B) showed the schematic diagrams of the distribution of retrograde AAV fluorescent protein-positive neurons in the dorsal hippocampus at Bregma − 1.46 mm and in the ventral hippocampus at Bregma − 2.46 mm, respectively.(C, D, and E) and (F, G, and H) respectively showed the neurons with green, red, and their merged fluorescence in the CA1 of CON and CRS at the Bregma − 1.46.The comparative schematic diagrams of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS were from the ipsilateral (I) and contralateral (J) hippocampus at Bregma − 2.46 mm, respectively.(K), (L), (M), and (N) represented the effects of CRS on the density of red fluorescence positive neurons in the ipsilateral vDG, vCA1, vCA2, and vCA3, respectively.(O), (P), (Q), and (R) represented the effects of CRS on the density of red fluorescence positive neurons in the contralateral vDG, vCA1, vCA2, and vCA3, respectively.Data expressed as the means ± SEM (n = 6).*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 10 .
Fig. 10.Effects of chronic restraint stress on the density of neurons in hypothalamus projecting to the mPFC.The comparative schematic diagrams of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS were from the ipsilateral (A) and contralateral (C) hypothalamus at Bregma − 1.70 mm, respectively.(B) and (D) showed the effects of CRS on the density of red fluorescence positive neurons in the ipsilateral and contralateral DM, respectively.Data expressed as the means ± SEM (n = 6).*P < 0.05。.

Fig. 11 .
Fig. 11.Effects of chronic restraint stress on the density of neurons in thalamus projecting to the mPFC.The comparative schematic diagrams of the distribution of retrograde AAV fluorescent protein-positive neurons between CON and CRS were from the thalamus at the Bregma − 0.70 (A), the ipsilateral thalamus at Bregma − 1.58 mm, and the contralateral thalamus at Bregma − 1.58 mm, respectively.The comparative pictures of the distribution of retrograde AAV fluorescent proteinpositive neurons between CON and CRS were from the ipsilateral AMV at the Bregma − 0.70 (B), the ipsilateral PC at the Bregma − 0.70 (D), and the ipsilateral thalamus at Bregma − 1.58 mm (F), respectively.(G), (I), (J), (H), (K), (L), (M), and (N) indicated the effects of CRS on the density of red fluorescence positive neurons in the ipsilateral AM, AMV, MD, PC, Re, Rt, VM, and VRe, respectively.(O) and and (P) showed the effects of CRS on the density of red fluorescence positive neurons in the contralateral VM and VRe, respectively.Data expressed as the means ± SEM (n = 6).*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 12 .
Fig. 12.The schematic diagram of effects of CRS on the local projections to the IL from the PrL, Cg1, Cg2, MO, and DP of mPFC.The dotted line indicates that chronic restraint stress caused a decrease in local projections to the IL.

Fig. 13 .
Fig. 13.The schematic diagram of effects of CRS on the projections to the IL from the multiple different brain regions.The dotted or continuous line indicates that chronic restraint stress respectively caused a decrease or an increase in projections to the IL.

Table 2
Distribution and statistics of brain regions projected to the medial prefrontal cortex.