Rinsenoside Rg1 and its involvement in Hippo–YAP signaling pathway alleviating symptoms of depressive-like behavior

Abstract

Ginsenoside Rg1 (G-Rg1) has potential antidepressant effects, but the underlying mechanism remains unclear. Presently, sixty 6-8 week-old male C57BL/6 mice were selected and randomly allocated to control, chronic restraint stress (CRS), CRS and low G-Rg1 administration (CRS + L-Rg1), CRS and high G-Rg1 administration (CRS + H-Rg1), and CRS and fluoxetine administration (CRS + FLX) groups. The component of anxiety in psychic processes and neuropathological changes occurring in dentate gyrus (DG) neurons were evaluated, where PC12 cells were assessed for the expression of G-Rg1. Both cell viability and apoptosis were analyzed. G-Rg1 (5 and 10 mg/kg/day) alleviated the behavioral manifestations of neuropathological processes revealed in DG neurons of CRS-induced mice. Western blotting analysis demonstrated the negative correlation of G-Rg1 level and that of Hipp-YAP signaling pathway components including p-YAP/YAP, p-MST1/MST1, and p-LATS1/LATS1, which were triggered by CRS. Combined therapy with G-Rg1 (10 mM) proved to have an inhibitory effect on PC12 cell viability and apoptosis compared to sole cort treatment. In addition, chronic G-Rg1 also reduced the protein expression levels of Hippo-YAP signaling pathway activated by corticosterone (Cort) including p-YAP/YAP, p-MST1/MST1, and p-LATS1/LATS1. The above mentioned improvements could be implemented due to XMU-MP-1 hampering the processes in Hippo-YAP signaling pathway. Importantly, the changes in synaptic plasticity and apoptosis were thoroughly investigated to determine the role of chronic G-Rg1 in the forementioned processes. In conclusion, chronic G-Rg1 played an important neuroprotective role in either CRS mice or Cort-treated cells associated with the inhibition of Hippo-YAP signaling pathway, which was the core part of decreasing neuronal apoptosis and enhancing synaptic plasticity.

Introduction

Depression is characterized by the diminished size of prefrontal cortex (PFC) and hippocampus, which are responsible for regulating mood and cognition. Furthermore, synaptic interactions have been lowered in numbers in these areas in individuals with depression¹. There are many hypotheses for the onset of depression, but a definitive pathogenesis to explain depression is lacking. Therefore, new therapeutic targets for depression need to be urgently discovered. Major depressive disorder (MDD) was suggested to be linked to a monoamine deficiency disorder, where low levels of monoamines in synaptic gaps were thought to cause symptoms of depression². Nonetheless, neuroplasticity hypothesis of depression has been developed implying cognition and mood improvements achieved by elevated neurogenesis as a downstream effect of antidepressants³.

Ginseng is a precious Chinese medicine with a long history of medicinal use in improving mental state and regulating neurological disorders including insomnia, depression, anxiety and neurasthenia⁴. Ginsenoside is considered to be the main active ingredient of ginseng, which has multiple effects on nervous, cardiovascular, and immune systems, among others⁵. There are many types of ginsenosides; among which, ginsenoside Rg1 (G-Rg1), a tetracyclic triterpenoid derivative with various pharmacological activities, enhances hippocampal neurogenesis, neuroplasticity that leads to the improvements in learning and memory, while also having anti-aging and anti-fatigue effects. Several studies have uncovered the antidepressant effects of G-Rg1 by inhibiting glial activation, synaptic defects, and apoptosis, oxidative stress and neuroinflammation⁶, and increasing the PKA, CREB phosphorylation, and expression of the brain-derived neurotrophic factor (BDNF)⁷. Consequently, G-Rg1 no doubt plays an ameliorative role in depression; however, no study has focused on the effects of G-Rg1 in improving the atrophy of hippocampus in stressed mice and its underlying mechanism.

The Hippo-YAP signaling pathway is a maintained kinase cascade playing a key role in cell growth, differentiation, proliferation, and apoptosis along with that of controlling tissue and organ developmental homeostasis and size. It is also involved in cardiovascular, neurological, and immune system diseases⁸. The integral parts of the pathway in mammals include Ste20-like kinase 1 (MST1) and MST2, homologs of the D. melanogaster Hpo kinase, large tumor suppressor kinase 1 (LATS1) and LATS2, adaptor proteins Salvador 1 (SAV1) (Sav in D. melanogaster), MOB1A and MOB1B (Mats in D. melanogaster), homologous transcriptional co-activator Yes-associated protein (YAP) and transcriptional co-activator with PDZ-binding motif (TAZ) (Yorkie in D. melanogaster), and TEAD transcription factors (TEAD1-TEAD4) (Scalloped in D. melanogaster)^9–12.

Mechanistically, the upstream components of the Hippo-YAP signaling pathway, including MAP4K, MST1/2, and its scaffolding protein SAV1, can phosphorylate LAST1/2 and their scaffolding protein MOB1A/B. The activated LATS1/2 in turn continues to phosphorylate the downstream YAP and TAZ, and the phosphorylated YAP/TAZ can connect to the 14-3-3 junction protein promoting their collection in cytoplasm. On the other hand, the phosphorylated YAP/TAZ can also be ubiquitylated in the cytoplasm for degradation, thus inhibiting their nuclear localization, a process that provides the facilitation of the Hippo-YAP signaling pathway¹³. YAP/TAZ drives the expression of downstream genes in regulating cell development and playing different physiological roles by binding to different transcription factors¹⁴. However, the uncontrolled activation of TAP/YAZ can lead to numerous pathological consequences¹⁵.

The disturbed functioning of the Hippo-YAP signaling pathway is strongly linked to neurodegenerative diseases¹⁶. The main characteristic of Alzheimer’s disease (AD) is the abnormal death of functional nerve cells, possibly stimulated by the Hippo-YAP signaling pathway¹⁷. In fact, the hippocampus appears atrophied with its size reduced in patients with depression induced by neuronal apoptosis¹⁸. However, the functions of this pathway in the pathogenesis of depression have not been well investigated.

Materials and methods

Animals and cells

Sixty adult male C57BL/6 mice used in this experiment were provided by the Experimental Animal Centre of North China University of Technology and were acclimatized for 1 week. They were randomly divided into the control (CON, n = 12), the chronic restraint stress model (CRS, n = 12), the G-Rg1 low-dose treatment (CRS + L-Rg1, 5 mg/kg/day, n = 12), the G-Rg1 high-dose treatment (CRS + H-Rg1, 10 mg/kg/day, n = 12), and the fluoxetine (FLX) treatment groups (CRS + FLX, n = 12). The CRS treatment was administered for 3 weeks in the model group, and the control was maintained in a standard environment. CRS treatment was performed by placing the mice in 50-mL conical tubes for 6 h per day for 21 days²⁴. All the experiments were carried out based on the Regulations of the Ministry of Health of the People’s Republic of China on Animal Management and received the approval from the Animal Research Ethics Committee of North China University of Technology (20230062). Highly differentiated rat adrenal pheochromocytoma cells (PC12) were used (ZQXZ-BIO, China).

Medication

Rg1 (15315, C42H72O14, purity ≥ 98%, Cayman, USA) at a dose of 20 mg/ml was dissolved in DMSO, and ginsenosides were intraperitoneally injected into the mice at 5 and 10 mg/kg/day in saline in the CRS + L-Rg1 and CRS + H-Rg1 groups, respectively, combined with CRS for 3 weeks²⁵. FLX hydrochloride (IF0390, C17H18F3NO·HCL, purity ≥ 98%, Solar bio, China) was dissolved in DMSO at 20 mg/ml, and the mice received intraperitoneal injections of FLX hydrochloride at a dose of 10 mg/kg/day for 3 weeks combined with CRS. Corticosterone (cort) solution (16063, C21H30O4, purity ≥ 98%, Cayman, USA) at a dose of 25 mg/ml was placed in DMSO and applied to PC12 cells at 200 µM for 24 h, while the same was done for XMU-MP-1 (22083, C17H16N6O3S2, purity ≥ 98%, Cayman, USA) at a dose of 2 mg/ml at 5 µM for 24 h²⁶.

CRS model

CRS was selected as a model for depression-like behavior in mice. After 1 week of acclimatization, mice requiring CRS treatment were confined in 50-ml cylindrical tubes, and these restraint tubes were surrounded by the holes for the mice to breathe. Mice were confined for 6 h per day for 21 consecutive days and kept in cages except during CRS²⁴.

Behavioral experiments

Sucrose preference test (SPT)

SPT was conducted to detect pleasure deficits in mice²⁷. The SPT method used in the present study was slightly modified from the technique used in our previous work¹⁹. Sucrose preference was determined by using the following formula: sucrose preference (%) = sucrose intake (g)/[sucrose intake (g) + water intake (g)] × 100%²⁸. SPTs were separately performed before the CRS procedure and once a week after CRS.

Forced swimming test (FST)

FST was applied to analyze behavioral desperation in mice. The FST method was carried out according to the protocol of our former study¹⁹.

Tail suspension test (TST)

TST was implemented to assess behavioral despair in mice. Mouse tails were glued to iron hooks and placed on a tail suspension for 6 min. The last 4 min of rest were recorded.

Open field test (OFT)

OFT was done to test anxiety-like behavior in mice. The experiment was performed by recording the behavioral activity of mice for 10 min using the Activity Monitor software program (MED-associates, USA).

Elevated plus maze (EPM)

EPM test was performed to monitor anxiety-like behavior in mice. This test was performed based on the method described in our previous study²⁰.

Morris water maze (MWM)

In the MWM experiment, mice were trained and their spatial learning and memory levels were evaluated. The MWM protocol was adopted based on a similar procedure depicted in our former research¹⁹. The performance of the mice was analyzed using the Smart 3.0 system (Pan lab, Spain).

Hematoxylin-eosin staining (HE)

After anesthetized by 30% urethane anesthesia (0.4 ml/100 g, i.p.), the mice were killed by cervical dislocation, then mouse brains were isolated and saturated with a 0.01-M phosphate-buffered saline (PBS) solution (pH = 7.2–7.4). Then, the brain of each mouse was taken, 4% paraformaldehyde-soaked, stored at 4 °C for at least 24 h, dried with a graded ethanol solution, and fixed in an optimal cutting temperature compound (tissue-tek, Sakura Finetek USA) for tissue segmentation. The Sect. (5 mm) underwent hematoxylin and eosin (H&E) staining and were analyzed under a Leica microscope (Wetzlar, Germany).

Immunofluorescence

In vivo, hippocampal slices were prepared with 4% paraformaldehyde for 30 min, rinsed with PBS for 5 min, and placed in 0.1% Triton X-100 at room temperature for 10 min. The cells were infested with mouse polyclonal antibodies against YAP1 (RT1664, 1:200 dilution, HUABIO, China) at 4 °C overnight. After the repeated rinsing with PBS, CY3-tagged goat anti-mouse IgG secondary antibodies (GB21301, 1:300 dilution, Service bio Co., Ltd. Wuhan, China) were applied to their surfaces for 1 h at room temperature, where 5 µg/ml DAPI (G1012, Service Bio Co., Ltd. Wuhan, China) was used for staining of nuclei. The images were acquired using a scanner (Panoramic MIDI, 3DHISTECH) or a standing fluorescence microscope (Nikon Eclipse C1, Japan).

In vitro, PC12 cells were washed three times with PBS and stored in 4% PFA for 15 min. After the application of 0.1% Triton X-100 in PBS composed of 10% goat serum albumin and preserved at room temperature for 45 min, cells were targeted by primary antibodies including mouse Anti-YAP (RT1664, 1:200 dilution, HUABIO, China), at 4 °C overnight, rinsed in PBS three times, and re-infested with secondary antibodies (goat anti-mouse IgG, S2001, 1:200 dilution, Report, China) at the same temperature for 1 h. After another cycle of PBS washing, the cells were prepared.

TUNEL staining

Brain samples were preserved in 4% paraformaldehyde for 24 h and sliced into paraffin-embedded sections, and the TUNEL-positive cells were evaluated with ApopTag^® Peroxidase In Situ Apoptosis Detection Kit (CHEMICON, USA) in accordance with the manufacturer’s instructions. Briefly, 4-µm sections were immersed in xylene to remove paraffin layer, and rehydration was provided by the preparation with the different portionswith graded concentrations of ethanol (100%, 95%, 90%, 80%, 70%), and then treated with proteinase K (10 µg/ml) for 15 min at room temperature to ensure the absence of occasional protein interactions. The sections were subsequently immersed in TUNEL reaction buffer for 1 h at 37 °C and then were prepared with 20-µl DAB substrate (ZSGB-BIO, China) for 2 min at room temperature.

Flow cytometry

Apoptosis was estimated in flow cytometry assay with the Annexin V-FITC Apoptosis Detection Kit I (BD, cat. no. 556547, 2020) based on the manufacturer’s protocol that was slightly modified according to the previously depicted procedure²¹. Apoptotic cells were identified with the Guava easyCyte™ Flow Cytometry (EMD Millipore Corp.).

Western blotting

The protocol of Western blotting was based on our former study¹⁹. The following primary antibodies were used as follows: Anti-YAP, RT1664, 1:1000 dilution, HUABIO, China; Phospho-YAP, ET1611-69, 1:1000 dilution, HUABIO, China; Anti-MST1, HA500043, 1:5000 dilution, HUABIO, China; Phospho-MST1, AF2367, 1:2000 concentration, Affinity Biosciences, China; Anti-LATS1, A17992, 1:1000 dilution, ABclonal, USA; Phospho-LATS1, AF7169, 1:2000 dilution, Affinity Biosciences, China; Anti-BDNF, ET1606-42, 1:2000 concentration, HUABIO, China; Anti-postsynaptic density 95 (PSD95), ET1602-20, 1:2000 dilution, HUABIO, China; Anti-NMDAR2A, ET1704-80, 1:500 dilution, HUABIO, China; Anti-Bax, ET1603-34, 1:5000 dilution, HUABIO, China; Anti-Bcl-2, ET1702-53, 1:2000 concentration, HUABIO, China; and β-actin, AC026, 1:100000 dilution, ABclonal, USA. Anti-mouse and anti-Rabbit IgG HRP conjugates, 1:5000 concentration, were incubated for 30 min at room temperature. Finally, the protein band intensity was identified with a digitalized chemiluminescence imaging system (CLINX-6300, China).

Statistical analyses

Statistical analysis was done with GraphPad Prism 8 program. The descriptive statistics is shown as mean ± SEM (Standard Error of the Mean). The MWM data in IT and RT were assessed with a two-way repeated-measures analysis of variance (ANOVA), with day × group as the between-subjects factor and the measurement session as the within-subjects. The other data were estimated by one-way ANOVA. Group and daily differences were analyzed post hoc using Bonferroni test. The significance level was considered as p < 0.05. Effect size (η²) was calculated.

Results

The hypothesis was tested by obtaining the evidence from the behavioral tests, histopathological observations, fluorescence staining, flow cytometry, and molecular measurements.

Effects of continuous G-Rg1 administration on the outcomes of sucrose test in stressed mice

In one-way ANOVA, the significant differences in sucrose preference were found between the five groups (Fig. 1A, F_{(4, 45)} = 24.33, p < 0.001, η² = 0.68). Post-hoc tests revealed that the values for this parameter were considerably lower in CRS group than in the control (p < 0.001 on the 21st). Nonetheless, they were significantly elevated after the treatment with G-Rg1 and fluoxetine in CRS + L-Rg1, CRS + H-Rg1, and CRS + FLX groups in comparison with that in CRS (p < 0.05, p < 0.001, and p < 0.001, on the 21st).

Fig. 1.

Effects of G-Rg1 administration on depressive-like behavior, anxiety-like behavior and cognitive impairment in stressed mice. (A) The percentage of sucrose consumption in the SPT. (B,C) Immobility time was determined in the FST and TST. (D) Ambulatory time zone in the OFT. (E) Total moving distance in the EPM. (F) Time in open arms in the EPM. (G) Entries in open arms in the EPM. (H) Number of platform area crossings in the SET stage. (I) Quadrant occupancy in the SET stage. (J) Number of platform area crossings in the RET stage. (K) Quadrant occupancy in the RET stage. (L) Mean escape latency was determined for each day in the IT (left) and the RT (right) stages among five groups. (M) Mean swimming speed in both the IT (left) and RT (right) stages. Data are presented as mean ± SEM (n = 10). *p < 0.05, **p < 0.01, ***p < 0.001, denote significant difference from the Control group; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001, denote a significant difference from the CRS group; n.s. not significant.

Effects of continuous G-Rg1 administration on FST and TST in stressed mice

One-way ANOVA showed that the durations of immobility differed significantly between the five groups in both FST and TST (Fig. 1B, C, F_{(4, 45)} = 11.81, p < 0.001, η² = 0.51; F_{(4, 45)} = 8.224, p < 0.001, η² = 0.42). Post hoc tests demonstrated that, in both FST and TST, these values in the mice in CRS group rose significantly compared with those of the control (p < 0.001, p < 0.001). Contrarily, the behavioral despair state of the mice differed significantly in the drug-treated group as compared with the CRS, as evidenced by a significant decrease in the immobility time in CRS + L-Rg1 (p < 0.001), CRS + H-Rg1 (p < 0.001), and CRS + FLX groups in FST (p < 0.001), and in the CRS + H-Rg1 (p < 0.001) and CRS + FLX (p < 0.001) groups in TST, whereas CRS + L-Rg1 compared with CRS groups did not exhibit the differences in the TST (p > 0.05). These data suggest that the G-Rg1 treatment significantly reversed the behavioral despair in affected by stress mice.

Effects of continuous G-Rg1 administration on OFT and EPM results in stressed mice

One-way ANOVA revealed the significant differences in ambulatory time in the central area among the five groups in OFT (F_{(4, 45)} = 12.26, p < 0.001, η² = 0.52). Post hoc tests demonstrated that the central region dwell time was significantly lower for CRS group than for control (Fig. 1D, p < 0.001). Compared with the former, CRS + L-Rg1, CRS + H-Rg1, and CRS + FLX groups exhibited longer central area dwell times (p < 0.001, p < 0.001, and p < 0.01).

In EPM, one-way ANOVA showed significant differences in total moving distance, open arm dwell time, and number of open arm entries among five groups in OFT (F_{(4, 45)} = 11.24, p < 0.001, η² = 0.50; F_{(4, 45)} = 16.00, p < 0.001, η² = 0.59; F_{(4, 45)} = 9.599, p < 0.001, η² = 0.46). Post-hoc tests showed that CRS group had significantly shorter total travel distance and open arm dwell time as well as lesser open arm entries than control (Fig. 1E–G, p < 0.001, p < 0.001, and p < 0.001). Interestingly, total distance (p < 0.001, p < 0.001, and p < 0.001), open arm dwell time (p < 0.01, p < 0.001, and p < 0.001), and open arm entries (p < 0.05, p < 0.05, and p < 0.001) were significantly greater for the CRS + L-Rg1, CRS + H-Rg1, and CRS + FLX groups than for the CRS group. These results suggest that G-Rg1 significantly improved the anxiety-like behavior in depressed mice.

Impact of continuous G-Rg1 administration on MWM results in stressed mice

The MWM test data for each day for the five groups are outlined in Fig. 1H–M. During the IT phase, escape latency decreased significantly with training (Fig. 2L, left). Two-way ANOVA demonstrated significant differences in time (F_{(4, 180)} = 44.41, p < 0.001, η² = 0.50) and groups (F_{(4, 45)} = 454.2, p < 0.001, η² = 0.98), except for the time × group interaction (F_{(16, 180)} = 2.661, p < 0.001, η² = 0.19). Post hoc analyses revealed significant differences in the escape latency between CRS and control groups (Fig. 1L, p < 0.001 on 2nd, 3rd, 4th, and 5th day) suggesting that CRS group found the platform for a longer period of time than the control group. Nonetheless, escape latency was significantly reduced in the CRS + L-Rg1 (Fig. 1L, p < 0.01 on 2nd day), CRS + H-Rg1 (Fig. 1L, p < 0.01 on 2nd and 3rd day; p < 0.001 on 4th, 5th day), and CRS + FLX (Fig. 1L, p < 0.001 on 2nd, 3rd, 4th, and 5th day) groups as compared with the CRS group.

Fig. 2.

Effects of G-Rg1 administration on neuronal damage in the hippocampal DG region of CRS mice. The representative photographs of HE stained hippocampus (Hippo) and dentate gyrus (DG) region in the five groups. Scale bar = 500 μm in the five pictures above, Scale = 200 μm in the five middle pictures, Scale bar = 100 μm in the five pictures below.

Learning flexibility was then measured after the movement of the hidden platform into the contralateral quadrant in the RT stage. The mean escape latency of each mouse was assessed after 2 days of training (Fig. 1L, right). Two-way ANOVA demonstrated significant differences in time (F_{(4, 72)} = 82.83, p < 0.001, η² = 0.82), groups (F_{(1, 18)} = 13.76, p < 0.01, η² = 0.43), and time × group interaction (F_{(4, 72)} = 3.306, p < 0.05, η² = 0.16), and post hoc multiple comparisons revealed that in the escape latency between CRS and control groups (Fig. 1L, right, p < 0.001 on 7th and 8th day) suggesting that the former had the platform for a longer period of time than the Control group. However, compared with the CRS group, the treatment effect of the drug-treated group was significant, with shorter escape latencies observed in the CRS + L-Rg1, CRS + H-Rg1, and CRS + FLX groups (Fig. 1L, right, p < 0.001 on 7th and 8th day) proposing that the G-Rg1-treated mice used spatial strategies to learn the platform position.

During the SET phase, the number of crossing platforms and the percentage of time spent in the N quadrant were examined. One-way ANOVA reflected no statistical differences among the groups in terms of the number of traversed platforms and dwell time in the N quadrant (Fig. 1H, I, F_{(4, 45)} = 4.497, p < 0.01, η² = 0.29; F_{(4, 45)} = 7.411, p < 0.001, η² = 0.40). Post hoc tests displayed a significant decrease in the number of platform crossings and time spent in the N quadrant in CRS group than in control (p < 0.01 and p < 0.001). Importantly, the platform crossings and N quadrant occupancy in the CRS + H-Rg1 and CRS + FLX groups were elevated compared to the values in CRS group (Fig. 1H, I, p < 0.05, p < 0.01; p < 0.05, and p < 0.01).

During the RET phase, platform crossings and percentage of time in the S quadrant were examined. One-way ANOVA showed a significant difference among the groups in the platform crossings and residence time in the N quadrant (Fig. 1J, K, F_{(4, 45)} = 5.535, p < 0.01, η² = 0.33; F_{(4, 45)} = 6.206, p < 0.001, η² = 0.36). Post hoc tests demonstrated that the platform crossings and S quadrant occupancy were significantly lowered in CRS group than in control (p < 0.05 and p < 0.001). However, the administration of H-Rg1 and FLX improved the values for these in stressed mice (Fig. 1J, K, p < 0.05, p < 0.05; p < 0.05, and p < 0.01).

Additionally, no statistical difference in swimming speed was detected among the five groups throughout the test (Fig. 1M, IT: F_{(16, 180)} = 0.5731, p > 0.05, η² = 0.05; RT: F_{(4, 72)} = 0.7978, p > 0.05, η² = 0.04).

Effects of continuous G-Rg1 administration on HE staining in stressed mice

The sections were stained with HE to identify the pathological conditions in the neurons in the DG area of the mouse hippocampus. As shown in Fig. 2, the granule cells were structurally intact and full and arranged normally in the hippocampal DG area of control group. Contrarily, many granule cells were damaged in the hippocampal DG area of CRS group and presented with ruptured nuclei, deep staining, and solid shrinkage. Interestingly, the cell damage in the hippocampal DG region of the G-Rg1-treated mice was significantly alleviated.

Effects of G-Rg1 on the expressions of synaptic plasticity- and apoptosis-related proteins in stressed mice

We examined the expression of NR2A, PSD95, and BDNF in hippocampus (Fig. 3A). One-way ANOVA displayed that protein expression levels of NR2A, PSD95, and BDNF differed significantly in the five groups (Fig. 3B–D, NR2A: F_{(4, 10)} = 6.071, p < 0.01, η² = 0.71; PSD95: F_{(4, 10)} = 8.010, p < 0.01, η² = 0.76; BDNF: F_{(4, 10)} = 7.738, p < 0.01, η² = 0.76). Post hoc tests reflected significantly decreased expression levels of NR2A, PSD95, and BDNF in CRS group (p < 0.05, p < 0.05, and p < 0.01). However, hippocampal synaptic plasticity was more improved in CRS + L-Rg1 group than in CRS group, accompanied by elevated PSD95 level (p < 0.05). Contrarily, the expression levels of NR2A and BDNF did not differ significantly (p > 0.05 and p > 0.05). Moreover, the NR2A, PSD95, and BDNF expressions in the CRS + H-Rg1 group were significantly more profound (p < 0.05, p < 0.01, and p < 0.05). Similarly, those of NR2A, PSD95, and BDNF were also significantly elevated in the CRS + FLX group (p < 0.05, p < 0.01, and p < 0.05).

Fig. 3.

Effects of G-Rg1 administration on synaptic plasticity, apoptosis and Hippo-YAP signaling pathway. (A, E,I) Results are immunoblots from single representative experiments. The expression values of NR2A (B), PSD95 (C), BDNF (D), Bax (F), Bcl-2 (G) were normalized with β-actin value. The expression values of p-YAP (J), p-MST1 (K), p-LATS1 (L) were normalized with YAP, MST1, LATS1 respectively. (H) The representative photographs of TUNEL stained dentate gyrus (DG) region in the five groups. Deep staining cells represent apoptotic cells. Scale bar = 50 μm in all images in the figure. Data are presented as mean ± SEM (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001, denote significant difference from the control group; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001, denote a significant difference from the CRS group. n.s. not significant. The proteins were run on different membranes and that the membranes were cropped before incubation with the primary antibody.

Concurrently, the expressions of pro-apoptotic factor Bax and anti-apoptotic factor Bcl-2 were revealed in the mouse hippocampus (Fig. 3E). The results of one-way ANOVA showed that the Bax and Bcl-2 protein levels in the hippocampus were significantly different among the five groups (Fig. 3F, G, Bax: F_{(4, 10)} = 13.20, p < 0.001, η² = 0.84; Bcl-2: F_{(4, 10)} = 9.607, p < 0.01, η² = 0.79). Bax promotes apoptosis, and the Bax level in the hippocampus of the CRS group was significantly more elevated than in control (p < 0.01). Meaningfully, the Bax expression was less prominent in CRS + L-Rg1, CRS + H-Rg1, and CRS + FLX groups, as opposed to CRS (L-Rg1: p < 0.05; H-Rg1: p < 0.001; FLX: p < 0.001). Inversely, Bcl-2 inhibits apoptosis, and the expression of Bcl-2 in the hippocampus of CRS group was significantly lower than in that of control (p < 0.001). However, L-Rg1, H-Rg1, and FLX significantly reversed the CRS-induced reduction of Bcl-2 (L-Rg1: p < 0.05; H-Rg1: p < 0.05; FLX: p < 0.05).

Furthermore, TUNEL staining revealed that cell apoptosis in the DG region was significantly more increased in the CRS group than in control (Fig. 3H). Contrarily, cell apoptosis in the CRS + L-Rg1, CRS + H-Rg1, and CRS + FLX groups was significantly improved in the former. These data propose that the antidepressant effect of G-Rg1 is closely associated with enhanced synaptic plasticity and reduced apoptosis.

Effects of G-Rg1 on the expression of Hippo-YAP signaling pathway proteins in stressed mice

One-way ANOVA displayed the significant differences in the expression levels of p-MST1/MST1, p-LATS1/LATS1, and p-YAP/YAP between the five groups (Fig. 3I–L, p-MST1/MST1: F_{(4, 10)} = 7.743, p < 0.01, η² = 0.76; p-LATS1/LATS1: F_{(4, 10)} = 15.48, p < 0.001, η² = 0.86; p-YAP/YAP: F_{(4, 10)} = 8.582, p < 0.01, η² = 0.77). The p-MST1/MST1, p-LATS1/LATS1, and p-YAP/YAP levels were significantly more upregulated in the hippocampus of CRS group than in control (p < 0.05, p < 0.001, and p < 0.01). Meanwhile, the p-MST1/MST1, p-LATS1/LATS1, and p-YAP/YAP levels were downregulated in CRS + L-Rg1, CRS + H-Rg1, and CRS + FLX groups compared with CRS (p < 0.05, p < 0.01, p < 0.01; p < 0.05, p < 0.01, p < 0.05; p < 0.01, p < 0.001, and p < 0.05, respectively).

Moreover, immunofluorescence staining reflected the significant differences in the expression of YAP proteins between the five groups (Fig. 4A, B, F_{(4, 10)} = 40.45, p < 0.001, η² = 0.94). The fluorescence intensity of YAP in the DG region of the CRS group was significantly lower than that in control (p < 0.001). Nevertheless, L-Rg1, H-Rg1, and FLX reversed the CRS-induced diminishment in YAP fluorescence intensity (p < 0.001, p < 0.01, and p < 0.01). Taken together, these outcomes have demonstrated that CRS activates the Hippo-YAP signaling pathway, whereas G-Rg1 significantly inhibits the Hippo-YAP signaling pathway in stressed mice.

Fig. 4.

Effects of G-Rg1 administration on YAP‑positive cells. DAPI (stained nucleus) is shown in blue, YAP staining is shown in red. Merged images show YAP staining and nuclear staining with DAPI in blue (A). Scale bar = 500 μm in the five pictures above. Scale bar = 100 μm in all figures below. Data are presented as mean ± SEM (n = 3). ***p < 0.001, denote significant difference from the control group; ^##p < 0.01, ^###p < 0.001, denote a significant difference from the CRS group (B).

Cort significantly decreased the viability of PC12 cells, whereas G-Rg1 opposed the effect of Cort by inhibiting Hippo-YAP signaling

As outlined in Fig. 5A, cort-provoked cell death occurred in a Cort-dependent manner in PC12 cells. Cell death happened at all Cort concentrations, with about 50% of the cells remaining viable at 200 µM acting for 24 h or 400 µM acting for 12 h (Fig. 5A). Accordingly, the dilution of 200 µM acting for 24 h was chosen for the following study. Figure 5B shows that the optimal therapeutic model of G-Rg1 was screened from 0.1-, 1-, 10, and 100-µM incubation for 24 h. Consequently, the CCK8 test revealed that G-Rg1 dramatically improved cell viability at 10 µM (Fig. 5B, F_{(5, 12)} = 37.26, p < 0.001, η² = 0.94); thus, the following experiments were conducted under 10-µM G-Rg1 exposure for 24 h.

Fig. 5.

Effects of G-Rg1, and XMU-MP-1 administration on the synaptic plasticity and apoptosis in PC12 cells. (A) Effect of Cort on PC12 cell viability. (B) Effect of G-Rg1 on the viability of PC12 cell. (C) Effect of Cort, G-Rg1 and XMU-MP-1 on PC12 cell viability. (D, H) Results are immunoblots from single representative experiments. The expression values of NR2A (E), PSD95 (F), BDNF (G), Bax (I), Bcl-2 (J) were normalized with β-actin value. (K–L) The staining cells with Annexin V- FITC were analyzed by flow cytometry to assess cell apoptosis. Data are presented as mean ± SEM (n = 3 for Western blotting, and n = 6 for flow cytometry). *p < 0.05, **p < 0.01, ***p < 0.001, denote significant difference from the control group; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001, denote a significant difference from the Cort group. n.s. not significant. The proteins were run on different membranes and that the membranes were cropped before incubation with the primary antibody.

XMU-MP-1, as a potent and selective MST1/2 inhibitor, inhibits the activation and facilitates the nuclear localization of the downstream effector YAP, which in turn promotes cell growth and tissue repair. The CCK8 results showed that Cort group has significantly more decreased PC12 cell viability than in control group (F_{(4, 10)} = 86.25, p < 0.001, η² = 0.97); whereas, XMU-MP-1 and G-Rg1 significantly reversed the Cort-induced viability reduction (Fig. 5C, p < 0.001 and p < 0.001). These outcomes propose that G-Rg1 reduces stress-induced damage, probably as an inhibitor of the Hippo-YAP signaling pathway, similar to XMU-MP-1.

Moreover, flow cytometry was applied to identify cell apoptosis (Fig. 5K, L). The results demonstrated that the early apoptosis ratio of PC12 cells was higher in Cort group than in control (p < 0.001). Fortunately, XMU-MP-1 and G-Rg1 significantly relieved this damage (p < 0.001 and p < 0.001).

Effects of G-Rg1, and XMU-MP-1 on the expressions of synaptic plasticity- and apoptosis-related proteins in PC12 cells

The expressions of NR2A, PSD95, and BDNF in the PC12 cells were checked (Fig. 5D). One-way ANOVA demonstrated that the protein levels of NR2A, PSD95, and BDNF differed significantly between the four groups (Fig. 5E–G: F_{(3, 8)} = 8.47, p < 0.01, η² = 0.76; F_{(3, 8)} = 8.365, p < 0.01, η² = 0.76; F_{(3, 8)} = 11.82, p < 0.01, η² = 0.82). The NR2A, PSD95, and BDNF levels were superior to those in Cort group than in control (p < 0.05, p < 0.05, and p < 0.01). However, the PSD95 and BDNF levels had significantly higher values in the XMU-MP-1 group (p < 0.05 and p < 0.05) than in Cort, except for NR2A (p > 0.05). Homoplastically, the protein expression levels of NR2A, PSD95, and BDNF in Cort + G-Rg1 group exceeded those in the opposite (p < 0.05, p < 0.05, and p < 0.05).

Meanwhile, Bax and Bcl-2 had the expression in PC12 cells (Fig. 5H). The results of one-way ANOVA revealed the significant differences in the protein levels of Bax and Bcl-2 among the four groups (Fig. 5I, J, F_{(3, 8)} = 8.49, p < 0.01, η² = 0.76; F_{(3, 8)} = 8.379, p < 0.01, η² = 0.76). The expression of Bax, which promoted cell apoptosis, was significantly more increased in Cort group than in control (p < 0.01). Contrarily, the Bax level was significantly more reduced in both Cort + XMU-MP-1 and Cort + G-Rg1 than in Cort (p < 0.05 and p < 0.05). The expression of Bcl-2, which hindered cell apoptosis, was significantly less pronounced in Cort group than in control (p < 0.05). Therefore, XMU-MP-1 and G-Rg1 significantly enhanced the expression of Bcl-2 in Cort-induced cells (p < 0.05 and p < 0.05).

Effects of G-Rg1, and XMU-MP-1 on Hippo-YAP signaling in PC12 cells

The expressions of the Hippo–YAP signaling pathway proteins encompassing p-MST1/MST1, p-LATS1/LATS1, and p-YAP/YAP (Fig. 6A) were detected. One-way ANOVA revealed that the levels of the above mentioned proteins differed significantly between the four groups (Fig. 6B–D, F_{(3, 8)} = 13.04, p < 0.01, η² = 0.83; F_{(3, 8)} = 17.46, p < 0.001, η² = 0.87; F_{(3, 8)} = 20.77, p < 0.001, η² = 0.89). The expressions of p-MST1/MST1, p-LATS1/LATS1, and p-YAP/YAP had significantly more amplified in Cort group than in control (p < 0.05, p < 0.01, and p < 0.001). Interestingly, the p-MST1/MST1, p-LATS1/LATS1, and p-YAP/YAP levels were significantly lower in XMU-MP-1 (p < 0.05, p < 0.01, and p < 0.01), and G-Rg1 (p < 0.05, p < 0.05, and p < 0.05) groups than in Cort group.

Fig. 6.

Effects of G-Rg1, and XMU-MP-1 administration on the Hippo-YAP signaling pathway in PC12 cells. (A) Results are immunoblots from single representative experiments. The expression values of p-YAP (B), p-MST1 (C), and p-LATS1 (D) were normalized with YAP, MST1, LATS1 value respectively. (E) Translocation of YAP proteins within and outside the nucleus. DAPI (stained nucleus) is shown in blue, YAP staining is shown in red. Merged images show YAP staining and nuclear staining with DAPI in blue. Scale bar = 25 μm in all images in the figure. Data are presented as mean ± SEM (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001, denote significant difference from the control group; ^#p < 0.05, ^##p < 0.01, ^###p < 0.001, denote a significant difference from the Cort group. n.s. not significant. The proteins were run on different membranes and that the membranes were cropped before incubation with the primary antibody.

YAP is phosphorylated when the Hippo-YAP signaling pathway is triggered, and the phosphorylated YAP (p-YAP) connects to the 14-3-3 protein in the cytoplasm and is degraded. In the present study, we detected the YAP translocation from the cytoplasm to the nucleus by immunofluorescence staining (Fig. 6E). Cort significantly inhibited the nuclear localization of YAP. Contrarily, XMU-MP-1 and G-Rg1 promoted the nuclear translocation of YAP.

Discussion

Our research provides in vivo and in vitro evidence that chronic G-Rg1 exerts a neuroprotective effect on a CRS mouse model and Cort-promoted PC12 cell depression model (Fig. 7). The behavioral disorders and neuropathological changes of the DG neurons in the CRS mouse model were considerably alleviated by continuous G-Rg1 administration. Furthermore, the cytotoxicity facilitated by cort in PC12 cells was also significantly lowered with G-Rg1 administration.

Fig. 7.

Working model of G-Rg1’s functions in protecting the synaptic plasticity and mitigating neuroapoptosis via the Hippo-YAP signaling pathway in the stress model. Hippo-YAP signaling pathway is activated in nerve cells of stressed mice. By inhibiting the activation of the Hippo-YAP signaling pathway, G-Rg1 facilitates the translocation of the YAP protein into the nucleus, thereby enhancing synaptic plasticity, reducing cellular apoptosis, and consequently alleviating the stress condition. Created in BioRender.com. https://app.biorender.com/illustrations/65794f3eb2971e95378d3ae4.

CRS and control groups have differ significantly in SPT. These outcomes are consistent with the findings of the previous studies^22,23. Thus, our data demonstrated that the CRS model modulated the depressive-like behavior in mice^24,25. Then, FST and TST are involved in the common technique for diagnosing the antidepressant effect of the drugs^20,26. The immobility time in both groups was regarded as an index of despair composing the core of depressive disorders¹⁹. Our data revealed that the immobility time in both of them was significantly decreased by the administration of G-Rg1 in CRS + L-Rg1 and CRS + H-Rg1 groups, as compared with CRS, proposing that chronic G-Rg1 treatment could effectively mitigate the animal’s depressive-like behavior, which adheres to the standards of a previous study²⁷.

The data acquired from the OFT test demonstrated that the time spent in the central areas was higher in the CRS + L-Rg1, and CRS + H-Rg1 groups than in CRS indicating that the anxiety symptoms were effectively alleviated by chronic G-Rg1 that was mostly compatible to a standard antidepressant drug, FLX. Furthermore, many studies have shown later that CRS promoted an anxiety-like behavior in the EPM test. The CRS-treated mice had more time in the closed side of the EPM²⁸. The data received in our study are based on those of the formerly published report, in which time and entries in open arms were lower in control group. However, this condition was significantly enhanced by prolonged administration of G-Rg1. This result is comparable to that of a previous study, in which a ginsenoside Rg-1, Rg-5, and Rk mixture significantly elevated the number of open arm entries and the time spent on the former²⁹.

To investigate the protective effect of G-Rg1 on cognition, MWM was done to check the training. The CRS group demonstrated memory dysfunction in the MWM. Importantly, our study data demonstrated that G-Rg1 can enhance the spatial memory ability of CRS mice. These findings are aligned with the results of a former research showing that G-Rg1 hinders cognitive disturbance and hippocampus senescence in a rat model of d-galactose-promoted aging³⁰.

Individuals with MDD who died of suicide had smaller anterior medial DG areas and fewer granule neurons (GNs) after death³¹; the DG cells in rats with chronic mild unpredictable stress are disorganized, with markedly malformed cells and coalesced nuclei³². This phenomenon in stressed mice has been confirmed in our previous studies²⁰. The present data demonstrated that cells in the DG region of CRS mice were deformed, and the apoptotic cells were exceeded in numbers. Interestingly, the treatment with G-Rg1 and FLX blocked the harmful effects of CRS on cell morphology in the DG zone. Subsequently, we examined the expression of synaptic plasticity- and apoptosis-related proteins to assess the underlying molecular mechanisms, by which G-Rg1 improves the hippocampal function.

Neuronal atrophy and loss of glutamatergic synaptic connections provoked by stress are key factors causing depressive symptoms³³. A previous postmortem study found a modified expression of NMDA receptor isoforms in the brain circuits of MDD patients and impaired NMDA receptor-mediated intracellular signaling pathways³⁴. Furthermore, a recent study revealed that CUS decreased the expressions of N-methyl-D-aspartate receptor 2 A (NR2A) and NR2B³⁵. PSD95 is an important postsynaptic scaffolding protein that is essential for morphological advancement and synaptic growth of hippocampal neurons³⁶. BDNF, a member of the neurotrophic factor family, is essential in neuronal morphology and synaptic plasticity, which serve as a basis for hippocampal circuit development and cognition³⁷. Currently, we found that the NR2A, PSD95, and BDNF levels were significantly decreased in CRS model as compared with control. Interestingly, the G-Rg1 and FLX treatments significantly enhanced the expression of these synaptic plasticity-connected proteins, which may be one of the potential mechanisms for resistance to depression-induced depressive symptoms and cognitive deficits. This is consistent with the findings of previous studies. Long-term G-Rg1 supplementation hinders the age-related cognitive decline by activating the expression of N-methyl-d-aspartate receptor subunit 1 (NR1) and PSD95 in C57BL/6J mice³⁸; Rg1 has antidepressant action via the facilitation of BDNF signaling pathway and elevated regulation of hippocampal neurogenesis³⁹.

The BCL-2 family of proteins plays a key role in apoptotic cell death⁴⁰. BCL-2-associated X protein (BAX) is a key factor in mitochondrial regulation of cell death, whereas the physiological role of BAX is important for tissue homeostasis, and the disturbances provoke aberrant cell death⁴¹. The present study indicated that the Bax expression was elevated, whereas Bcl-2 was diminished in the hippocampus of CRS group, as compared with control. Additionally, G-Rg1 and FLX could restore the abnormal level of apoptosis-related protein expression in stressed mice. Moreover, TUNEL staining revealed apoptotic vesicles in the hippocampus of the CRS group, whereas the apoptotic cells were reduced in the G-Rg1- and FLX-treated groups.

A previous meta-analysis of structural brain imaging studies has reported that the hippocampal volume is indeed decreased in patients with major depression⁴². The Hippo-YAP signaling pathway is the most important in monitoring organ development and plays a key role in size control in vivo¹⁰. Therefore, we speculated that the Hippo-YAP signaling pathway might be related to the reduction of the hippocampal volume in persons with depression. Our data demonstrated that the expressions of p-MST1/MST1, p-LATS1/LATS1, and p-YAP/YAP in the hippocampus of CRS group was significantly elevated; meanwhile, the fluorescence intensity of the effector molecule, YAP, was decreased in the nucleus, which proved that the Hippo-YAP signaling pathway was activated during an episode of CRS-induced depression. Meaningfully, the treatment with G-Rg1 and FLX improved this condition, which indicated that the activation of the Hippo-YAP signaling pathway was inhibited. A previous study showed that G-Rg1 alleviated hepatic ischemia-reperfusion injury in mice by adjusting ERα-regulating YAP expression⁴³. Logically, these results demonstrated that the improvement due to the treatment with G-Rg1 and FLX in CRS-induced depression is closely linked to the inhibition of Hippo-YAP signaling pathway.

Our previous study revealed that high concentrations (200–800 µM) of cort could activate cellular damage in PC12 cells²⁰. Our outcomes demonstrated that cort lowered the viability of PC12 cells that was dependent on dilution. More importantly, cort decreased the Hippo-YAP signaling pathway activity in PC12 cells. G-Rg1 and XMU-MP-1, a potent and selective inhibitor of MST1/2, significantly increased the viability of cort-treated cells by the CCK8 assay. The Western blot test showed that the NR2A, PSD95, BDNF, and Bcl-2 levels were significantly decreased in the Cort-induced cellular injury model, whereas the Bax level was significantly increased. Synchronously, the p-MST1/MST1, p-LATS1/LATS1, and p-YAP/YAP levels in the Hippo-YAP signaling pathway were also significantly elevated, which was comparable with the outcomes of the animal experiments. Similarly, the improvement effects of the G-Rg1 in vitro experiments were consistent with those of animal experiments. More importantly, the effects of XMU-MP-1 on the expression of synaptic plasticity-, apoptosis-, and Hippo–Yap-related proteins in Cort-induced PC12 cells were consistent with those of the G-Rg1 treatment. Furthermore, we observed that the YAP translocation inside and outside the nucleus by immunofluorescence staining in PC12 cells. Cort induced the retention in the cytoplasm and inhibited the nuclear localization of YAP, whereas G-Rg1 and XMU-MP-1 led to the recovery of the abovementioned abnormalities.

Conclusion

Our study identified the functions of the Hippo-YAP signaling pathway in the pathogenesis of depression in both the in vivo and in vitro studies, revealed the regulatory role of the Hippo-YAP signaling pathway in depression, and discovered a novel function of G-Rg1 in the prevention and treatment of depression, which may give new insights and set new targets for exploring the pathogenesis and treatment of depression.

Author contributions

Linyin Gao (First Author): Methodology, Software, Investigation, Formal Analysis, Data Curation, Writing-Original Draft; Jiarong Wang (Co-first Author): Methodology, Investigation, Formal Analysis; Xiuchang Liu (Co-first Author): Software, Investigation, Formal Analysis; Lei Wu: Resources; Ran Ding: Data Curation; Xuemei Han: Data Curation; Xindi Wang: Data Curation; Hao Ma: Data Curation; Jie Pan: Formal Analysis; Xiujun Zhang (Corresponding Author): Resources, Supervision.Haitao Wang (Corresponding Author): Resources, Supervision.Xueliang Shang (Corresponding Author): Conceptualization, Resources, Funding Acquisition, Supervision, Writing-Review & Editing.

Funding

This work was supported by grants from the STI2030-Major Projects (2021ZD0200700), the central government guides local science and technology development fund project (246Z2511G, 246Z2605G), the key project of Hebei Natural Science Foundation Chinese Medicine Joint Fund (H2022209080), the key research project of North China University of Science and Technology in 2023 (ZD-YG-202314-23), the research project of basic scientific research service fees in provincial universities of North China University of Technology (JQN2023039), Medical Science Research Project of Hebei Province in 2023 (20231585).

Data availability

All experimental data and materials have been presented in the manuscript.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

All the experiments were carried out based on the Regulations of the Ministry of Health of the People’s Republic of China on Animal Management and received the approval from the Animal Research Ethics Committee of North China University of Technology (20230062).

Consent for publication

All authors agreed to publish the manuscript in the journal of Molecular Neurobiology.

Consent to participate

We confirm that the manuscript has been read and approved by all named authors and that there are no other persons who satisfied the criteria for authorship but are not listed. We further confirm that the order of authors listed in the manuscript has been approved by all of us.