Asperosaponin VI Ameliorates The CMS-Induced Depressive-Like Behaviors By Inducing A Neuroprotective Microglial Phenotype To Restore Hippocampal Synaptic Plasticity Via PPAR-γ Pathway

Xue Jiang Guizhou University of Traditional Chinese Medicine Saini Yi Guizhou University of Traditional Chinese Medicine Qin Liu Guizhou University of Traditional Chinese Medicine Dapeng Su Guizhou University of Traditional Chinese Medicine Liangyuan Li Guizhou University of Traditional Chinese Medicine Chenghong Xiao Guizhou University of Traditional Chinese Medicine Jinqiang Zhang (  552450374@qq.com ) Guizhou University of Traditional Chinese Medicine https://orcid.org/0000-0003-1563-827X


Introduction
Major depressive disorder (MDD) is a pervasive neuropsychiatric disorder and a signi cant contributor to the global burden of disease [1]. MDD has heterogeneous causes and clinical manifestations, which has impeded an understanding of its pathogenesis and design of effective treatments [2,3]. Antidepressants based on monoamine neurotransmitters can reduce symptoms, but they are effective in only one-third to half of patients [4][5][6]. Even when effective, current drugs take 3-6 weeks to begin working, and they often produce adverse effects [7]. Therefore, developing more effective antidepressants with fewer side effects is urgently needed.
MDD is associated with neuroin ammation, such as the increased concentrations of pro-in ammatory cytokines [8,9]. Such neuroin ammation can be driven by microglia, the innate immune cells of the central nervous system [10,11]. When continuously stimulated by immune responses, microglia can adopt a pro-in ammatory phenotype, secreting pro-in ammatory cytokines and neurotoxic substances, which damage synapses, induce apoptosis, inhibit neurogenesis, and eventually lead to depression symptoms [12][13][14]. Conversely, certain signals can induce microglia to adopt a protective phenotype, in which scavenger receptors on their surface recognize metabolic wastes and nerve cell debris, which the microglia phagocytose [15,16]. The microglia also releases anti-in ammatory cytokines and neurotrophic factors that protect and repair neurons, ultimately alleviating depression symptoms [17,18]. Therefore, regulating the phenotype of activated microglia is an attractive strategy for treating depression.
The triterpenoid saponin asperosaponin VI, one of the active components of the traditional Chinese medicine Radix Dipsaci, exerts anti-osteoporotic and anti-in ammatory effects, and it can pass through the blood-brain barrier to protect neurons and improve neurological diseases [19][20][21][22]. In previous work, we showed that asperosaponin VI, delivered at a dose of 40 mg/kg, can inhibit neuroin ammatory responses by hippocampal microglia and mitigate depression-like behaviors induced by lipolysaccharide in mice [23]. These ndings suggested that asperosaponin VI exerts its antidepressant effects at least in part by regulating hippocampal microglial function.
To verify and extend these previous ndings, we examined the effects of asperosaponin VI on microglial phenotype and behaviors of mice exposed to chronic mild stress (CMS), a classical model of MDD. We also examined the potential involvement of the PPAR-γ pathway in mediating the effects of asperosaponin VI.

Animals
Male 8-week-old C57BL/6 mice were purchased from Changsha Tianqin Biotechnology (Changsha, China) and allowed to acclimate for one week prior to experiments. All mice were housed individually under a standard 12-h light-to-dark cycle in temperature-and humidity-controlled rooms throughout the experiments. Mice were randomly assigned to experimental or control groups and an observer blinded to treatment conditions performed the behavioral tests and collected and analyzed the data. All experiments were approved by the Institutional Animal Care and Use Committee of the Guizhou University of Traditional Chinese Medicine.

Chronic mild stress (CMS)
Mice 9 weeks old at the start of experiments were caged individually and subjected to CMS as described [17] for three weeks. Each day, animals were exposed to three of the following stressors in random order: empty water bottles (12 h), food deprivation (12 h), tail clipping (10 min), restraint ( The SPT was performed as described [17]. Mice were individually housed, deprived of food and water for 12 h, and then given access to 1% sucrose solution (A) and water (B) for 2 h. The bottle positions were switched daily to avoid a side bias. The sucrose preference was calculated each week for each mouse using the formula: 100 × [VolA / (VolA + VolB)]. Sucrose consumption was normalized to the body weight of each mouse.

Tail suspension test (TST)
Each mouse was individually suspended by applying adhesive tape to the tip of the tail and connecting the tape to a ledge 30 cm above the cage oor. The animal was recorded for 6 min using a high-de nition camera. An observer masked to treatment conditions recorded the latency between suspension and rst abandonment of struggle as well as the time spent immobile during the 6-min period.

Forced swimming test (FST)
At 24 h before the test, mice were placed individually in a glass cylinder of height 25 cm and diameter 15 cm that was lled with water at 26°C to a depth of 15 cm. The next day, the mice were placed again in the same situation for 6 min and recorded using a high-de nition camera. An observer blinded to treatment conditions recorded the time spent immobile during the last 4 min.

Open eld test (OFT)
Mice were placed into an open eld (50 × 50 cm) and allowed to explore freely for 15 min. Total distance and time spent in the center (25 × 25 cm) were determined using video-tracking software (OFT100, Taimeng Tech. Chengdu, China).

Novelty-suppressed feeding test (NSFT)
Mice were deprived of food and water for 12 h before the test, then each mouse was placed for 5 min in a rectangular chamber (40 × 40 × 30 cm) containing a sugar pill in the center of the chamber. The time it took for a mouse to pick up the sugar with its forelimb was recorded as latency using a camera system.

RNA extraction and real-time PCR
At the end of experiments, mice were sacri ced, the whole brain was removed, and the hippocampus and cortex were isolated and placed into separate enzyme-free 1.5-mL microcentrifuge tubes. Total RNA was extracted separately from the hippocampus and cortex using Trizol (Invitrogen Life Technologies, Shanghai, China), then reverse-transcribed into cDNA using the high-capacity cDNA conversion kit (Takara, Tokyo, Japan) in strict accordance with the manufacturer's instructions. RT-PCR reaction mixture contains 1 µL of template cDNA, 5 µL MasterMix, and 1 µL primer (Sangon Biotech, Sichuan, China); add DEPC water to a total reaction volume of 10 µL. The PCR was performed in a CFX 96 system (Bio-Rad, Hercules, California, USA) using the following steps: pre-denaturation at 95°C for 30 s, denaturation at 95°C for 5 s, then 39 cycles of annealing at the appropriate temperature for 34 s, followed by extension.
Each sample was analyzed in three replicates. Expression level was determined using the 2 −ΔΔCt method with reference to the β-actin gene. The primers of each gene were listed in Supplementary Table 1. 2.6 Immunocytochemistry Whole brain from mice was perfused and xed in 4% paraformaldehyde for 48 h, dehydrated, frozen, cut into thin sections, thoroughly cleaned with 0.5% Triton X-100 for 15 min, blocked with 10% donkey serum for 1 h, and incubated overnight at 4°C with the following primary antibodies (Abcam, Cambridge, UK): mouse anti-Iba1 (1:400), mouse anti-Gfap (1:500), rabbit anti-iNOS (1:100), rabbit anti-Arg-1 (1:200), and rabbit anti-NeuN (1:200). On the next day, slices were washed three times with phosphate-buffered saline (PBS) and incubated in the dark for 2 h with secondary antibodies. DAPI was added to stain the nuclei, and slices were observed under a uorescence microscope (Olympus BX 51, Japan). Images were imported into Image J software (version 1.45J; National Institutes of Health, Bethesda, MD), and an intensity threshold was de ned to differentiate positive staining from background.

Enzyme-linked immunosorbent assay (ELISA)
Hippocampus and cortex from mice were placed into separate 1.5-mL microcentrifuge tubes and completely homogenized. The concentration of total protein in the supernatant was determined using the BCA kit (Boster, Shanghai, China), then aliquots of the supernatants were diluted to the same total protein concentration. These diluted samples were assayed for interleukin (IL)-1β, IL-10, brain-derived neurotrophic factor (BDNF) and tumor necrosis factor (Tnf)-α using a commercial ELISA kit (Boster, Shanghai, China) in strict accordance with the manufacturer's instructions.

Western blotting
The hippocampus and cortex of the mice were lysed by RIPA lysis buffer (Solarbio, Beijing, China) then centrifuged at 1000×g for 30 min. The concentration of total protein was measured by the BCA method. Equal amount of protein was resolved using 12% SDS polyacrylamide gel. Fractionated proteins were transferred onto PVDF membranes at 300 mA for 30 min, then the membrane was washed in TBST, blocked in skim milk for 30 min, and incubated overnight on a shaker at 4°C with primary antibody. The membrane was again washed three times with TBST, incubated with secondary antibody for 30 min, washed three times with TBST, and Bands were visualized using the BM Chemiluminescence Western Blotting Kit (Roche Diagnostics GmbH, Mannheim, Germany).. Membranes were analyzed using the ChemiDoc Touch system (Bio-Rad, Hercules, California, USA), and band intensity was quanti ed using Alpha software (version 1.45J; National Institutes of Health, Bethesda, MD, USA).

Statistical analysis
All statistical analyses were performed using GraphPad Prism software (version 8.0, SPSS Inc., Chicago, USA). Data were presented as mean ± SEM. Pairwise comparisons were assessed for signi cance using Student's two-tailed t-test, and comparisons among three or more values were assessed using one-or two-way ANOVA and Tukey's post hoc tests. Levels of signi cance are marked in gures as * p < 0.05, signi cant; ** p < 0.01, very signi cant; and *** p < 0.001, highly signi cant.

Asperosaponin VI ameliorates depression-like behaviors induced by CMS in mice
CMS for three weeks reduced sucrose preference (Fig. 1A), which was partially reversed by a subsequent 3-week treatment with asperosaponin VI or imipramine (Fig. 1B). Analysis at the level of individual mice showed that the 3-week treatment with asperosaponin VI improved sucrose preference in nearly 90% of CMS mice, while imipramine improved it in just over 60% (Fig. 1C).
CMS shortened latency and led to longer time spent immobile in the TST (Fig. 1D) and FST (Fig. 1E). Asperosaponin VI, but not imipramine, prolonged latency in the FST, while both compounds shortened the time spent immobile (Fig. 1F). Neither compound affected the distance travelled in the OFT (Fig. 1G), indicating that asperosaponin VI does not affect nerve transmission. Asperosaponin VI signi cantly improved feeding latency in the NSFT, and it partially reversed the weight loss induced by CMS (Fig. 1H).
3.2 Asperosaponin VI induces hippocampal microglia to switch from a pro-in ammatory to neuroprotective phenotype after CMS CMS increased expression of the microglial marker Iba1 and the marker of microglial activation Cd11b in the hippocampus, as well as expression of Cd11b in cortex ( Fig. 2A). Microscopy of tissue slices showed that CMS induced soma enlargement, thickening and shortening of processes and loss of branching in hippocampal microglia ( Fig. 2B and 2C). These ndings indicate stress-induced microglial activation.
Asperosaponin VI did not cause obvious changes in morphology or CD11b expression in microglia in the hippocampus or cortex, nor did it reverse the increase in Iba1 + area, increase in cell number or decrease in microglial branching induced by CMS ( Fig. 2A-2C). Similarly, the compound did not alter the CMS-induced radial morphology of astrocytes in the hippocampus and cortex (Fig. 2D-2F).
These results suggest that asperosaponin VI does not inhibit CMS-induced activation of hippocampal microglia; instead, it may regulate the type of microglial activation to in uence depression-like behaviors. Consistent with this hypothesis, we found that CMS increased the proportions of pro-in ammatory (iNOS + -Iba1 + ) microglia and decreased the proportions of anti-in ammatory (Arg-1 + -Iba1 + ) in hippocampus but not in cortex ( Fig. 3A and 3B). Asperosaponin VI partially reversed these changes.
3.3 Asperosaponin VI acts via the PPAR-γ pathway to exert its anti-in ammatory and antidepressant effects Since PPAR-γ signaling plays a key role in anti-in ammatory microglial phenotypes, we asked whether asperosaponin VI acts via such signaling to exert its "phenotype switching" effect. Indeed, CMS reduced expression of PPAR-γ-1 and PPAR-γ-2 as well as levels of phosphorylated PPAR-γ in hippocampus, which asperosaponin VI partially reversed (Fig. 4A). PPAR-γ localized in cytoplasm and nucleus of microglia in the hippocampus of mice that were exposed to CMS and then treated with asperosaponin VI (Fig. 4B).
To con rm the role of PPAR-γ in mediating the anti-in ammatory effects of asperosaponin VI, we repeated the above experiments in the presence of the PPAR-γ antagonist GW9662 (Fig. 4C), which effectively blocked the PPAR-γ pathway in hippocampus (Fig. 4D). Such blockade abolished the antidepressant effects of asperosaponin VI in the SPT, TST and FST (Fig. 4E), as well as its ability to increase numbers of Arg-1 + microglia and decrease numbers of iNOS + microglia in the hippocampus of CMS mice (Fig. 5A and 5B). In contrast, blockade did not alter the morphology or Cd11b expression of hippocampal microglia in CMS mice.
Blockade of PPAR-γ signaling also abolished the ability of asperosaponin VI to suppress proin ammatory cytokines and elevate anti-in ammatory cytokines in the hippocampus of CMS mice (Fig. 5E-5I). These results suggest that asperosaponin VI exerts its antidepressant and anti-in ammatory effects via the PPAR-γ signaling pathway.

Asperosaponin VI protects hippocampal synaptic plasticity from CMS
Dysfunctional microglia can communicate abnormally with neurons, which disrupts synaptic function and may help to explain MDD behaviors. In our mice, CMS reduced hippocampal expression of Cx3cl1 and its receptor Cx3cr1 as well as expression of CD200 and its receptor CD200R; these pairs mediate communication between neurons and microglia. Asperosaponin VI partially reversed these CMS-induced changes, while GW9662 abolished the effects of asperosaponin VI (Fig. 6A).
CMS downregulated PSD95, CamKII α and CamKII β as well as decreased levels of phosphorylated GluA 2 in hippocampus, but did not affect the number of NeuN + cells in the hippocampal dentate gyrus, all of which suggest inhibition of synaptic plasticity but not apoptosis of mature neurons (Fig. 6B and 6C). These changes in synaptic plasticity were partially reversed by asperosaponin VI, but not in the presence of GW9662 (Fig. 6C and 6D). Levels of PPAR-γ positively correlated with levels of PSD95, CamKII α, CamKII β and phosphorylated GluA 2 (Fig. 6E). These results suggest that asperosaponin VI requires PPAR-γ to induce neuroprotective microglia and repair CMS-induced damage to synaptic plasticity in hippocampus.

Discussion
We showed in previous work that asperosaponin VI inhibits NF-κB signaling to mitigate lipopolysaccharide-induced depression-like behaviors in mice by reducing microglia-mediated acute neuroin ammation [23]. Here, using a classical animal model of depression, we showed that asperosaponin VI induces a PPAR-γ-dependent neuroprotective microglial phenotype that mitigates depression-like behaviors induced by CMS, which is associated with restoration of hippocampal synaptic function. Our work extends the list of conditions where asperosaponin VI can exert therapeutic antiin ammatory and neuroprotective effects in the brain, a list that already includes Alzheimer's disease and optic nerve damage [19,24].
Depression usually manifests as diverse debilitating symptoms, including hopelessness and anhedonia [25]. Anhedonia, a core symptom of MDD, can be assessed in the SPT [26]. In addition to the SPT, we used the FST and TST to assess passive stress-coping behavioral despair [27,28]. As expected, CMS caused depression-like behaviors in all these tests, which subsequent asperosaponin VI treatment improved, to an even greater extent than the classic monoamine antidepressant imipramine. In fact, asperosaponin VI but not imipramine partially restored weight loss caused by CMS, suggesting that the former may lack serious side effects at the dose of 40 mg/kg. We believe that these results indicate genuine antidepressant effects of asperosaponin VI, because the compound did not signi cantly alter performance in the OFT.
Our previous research indicated that the dysregulation of pro-and anti-in ammatory cytokines plays a crucial role in depression [8]. In the present study, CMS upregulated the pro-in ammatory IL-1β, IL-6, iNOS and Tnf-α in hippocampus of mice, and asperosaponin VI reversed these changes while also upregulating the anti-in ammatory cytokines Arg-1, IL-10, and Tgf-β as well as BDNF. These results establish a link between the neuroprotective and anti-in ammatory effects of asperosaponin VI. Stress or immunostimulation has been shown to induce neuroin ammation, which appears to involve microglial activation, particularly in the hippocampus [29,30]. Consistent with these previous studies, we found here that CMS caused morphological changes in hippocampal microglia indicative of microglial activation. Asperosaponin VI did not reduce the extent of overall microglial activation in hippocampus of CMS mice; instead, it altered the type of such activation, from a pro-in ammatory to neuroprotective phenotype. The decrease in proportion of pro-in ammatory microglia translates to lower production of pro-in ammatory cytokines and neurotoxic products (such as nitric oxide and quinolinic acid) [31], which increase neuropathic pain and inhibit hippocampal neurogenesis, contributing to cognitive de cits and depression-like behaviors [10,32,33].
We found that the anti-in ammatory effects of asperosaponin VI are mediated by PPAR-γ, a liganddependent transcription factor belonging to the nuclear hormone receptor superfamily [34]. PPAR-γ regulates the expression of anti-in ammatory cytokines [35], and the PPAR-γ agonists pioglitazone or rosiglitazone can switch activated microglia cells from a pro-in ammatory to anti-in ammatory state [36,37]. Our previous research showed that asperosaponin VI acts via PPAR-γ to switch activated microglia from a pro-in ammatory to anti-in ammatory phenotype in vitro [38]. In present study, we further demonstrated that asperosaponin VI acts via PPAR-γ to induce a neuroprotective phenotype in hippocampal microglia of CMS-exposed mice and mitigate depressive-like mouse behaviors. Conversely, blocking the PPAR-γ signaling pathway abolished the neuroprotective microglia in induced by asperosaponin VI in hippocampus of CMS-exposed mice, as well as the antidepressant effect of asperosaponin VI. Thus, we speculated that asperosaponin VI exerts its antidepressant and antiin ammatory effects via the PPAR-γ signaling pathway to regulating the phenotype of microglia.
How proin ammatory microglia lead to depression is a hot topic of current research. A great deal of researches showed the dysfunctional microglia can lead to abnormal neuron-microglia communication and disrupt synaptic function [39][40][41]. In this study, we found CMS-induced decrease in the intercommunicating molecules between neuron and microglia (Cx3cl1 / Cx3cr1 and CD200 / CD200R) in hippocampus of CMS-exposed mice was reversed by ASA VI treatment. Meanwhile, the PSD-95, CamKII α and CamKII β as well as decreased levels of phosphorylated GluA 2 in hippocampus of CMS-exposed mice, these are thought to be crucial for morphological maturation and synaptic development of hippocampal neurons [42][43][44][45][46][47], were partially reversed by asperosaponin VI via PPAR-γ-dependent pathway. These results reveal for the rst time the role of asperosaponin VI in maintaining normal communication between neurons and microglia as well as in repairing CMS-induced damage to synaptic plasticity in hippocampus.

Conclusion
In summary, our experiments in mice suggest that CMS induces depression-like behaviors by inducing a pro-in ammatory microglial phenotype that disrupts neuron-microglia communication and synaptic function. Asperosaponin VI ameliorates the effects of CMS by inducing, via PPAR-γ, a neuroprotective microglial phenotype that partially restores hippocampal synaptic function (Fig. 7). These ndings may provide further insights into the pathogenesis of depression and the development of natural antidepressants. Our study provides further evidence for asperosaponin VI as a potential antidepressant and a reference for research on depression.

Declarations
Ethics approval and consent to participate The animal study was reviewed and approved by Guizhou University of Traditional Chinese Medicine.

Consent for publication
All authors agree to the publication of this manuscript.

Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Competing interests
The authors declare no con icts of interest related to this research.       Mechanism by which asperosaponin VI improve CMS-induced depression CMS in mice induces a proin ammatory microglial phenotype, disrupting neuron-microglia communication and synaptic function in hippocampus, ultimately leading to depression-like behaviors. Asperosaponin VI may ameliorate the effects of CMS by inducing microglia to adopt a PPAR-γ-dependent neuroprotective phenotype.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.