Chemical component analysis of the traditional Chinese medicine Guipi Tang and its effects on major depressive disorder at molecular level

Abstract

Ethnopharmacological relevance

Guipi Tang (GPT) is a widely used traditional Chinese medicine that is used to treat major depressive disorder. However, the molecular mechanisms of its effects remain unclear.

Aim of the study

This study aimed to investigate the antidepressant-like effects of GPT and explore its underlying molecular mechanisms.

Materials and methods

Male Sprague–Dawley rats were subjected to a chronic unpredictable mild stress (CUMS) procedure and treated with various doses of GPT, with fluoxetine treatment as a positive control. Behavioural tests (including sucrose preference test, novelty-suppressed feeding test, open-field test and forced swim test), terminal deoxynucleotidyl transferase dUTP nick end labeling and enzyme-linked immunosorbent assay were conducted. The levels of Bax, Bcl-2, cleaved caspase-3, PI3K, p-PI3K, AKT, p-AKT, BDNF, TrkB and CREB or p-CREB were assessed at the protein level using western blotting or immunofluorescence.

Results

GPT consists of mainly known drugs, such as liquiritin and ginsenosides. It reversed depressive behaviours and decreased cell apoptosis in the hippocampi of CUMS rats. It significantly upregulated the protein level of Bax, p-Akt, p-PI3K, BDNF, TrkB and p-CREB and downregulated the level of cleaved caspase-3 and Bcl-2.

Conclusions

GPT had anti-depressive activity as indicated by the amelioration of depression-like behaviour and the inhibition of hippocampal neuronal apoptosis in CUMS rats. This inhibition was mediated partly by modulating the PI3K/Akt and/or BDNF/TrkB/CREB pathway, in which, glycosides, the main components of GPT, might be involved.

Graphical abstract

Highlights

• •Traditional Chinese medicine GPT reverses the depressive-like behavior.
• •GPT inhibits cell apoptosis in the hippocampi.
• •GPT reduced serum corticosterone in rats with chronic unpredictable mild stress.
• •PI3K/Akt and/or BDNF/TrkB/CREB axis participate in GPT anti-depression activity.

Akt; Apoptosis; BDNF; Chronic unpredictable mild stress; Depression; Guipi Tang.

1. Introduction

Major depressive disorder (MDD) is a psychological disease characterised by a pervasive and persistent low mood, amotivation, anhedonia, low energy and/or fatigue and has a high prevalence. The World Health Organization has reported that depression is the second leading cause of illness burden (Mathers and Loncar, 2006; Murray and Lopez, 1997). Over the past few decades, major classical antidepressants, such as tricyclic antidepressants and monoamine oxidase inhibitors, have played an important role in the treatment of depression. However, a meta-analysis of these data showed that <50% of patients with depression benefit from these agents (Entsuah et al., 2001). Therefore, there is a long way to go in the development of antidepressants.

Compared with matched healthy individuals, patients with multiple depressive episodes have reduced hippocampal volume (MacQueen et al., 2003). Hippocampal volume reduction generally occurs in patients who have MDD for >2 years or in those who have experienced >1 episode of the disease (McKinnon et al., 2009). Human post-mortem studies have found convincing evidence for apoptosis in many regions in the brain of patients with depression, such as the entorhinal cortex, subiculum, dentate gyrus, CA1 and CA4 (Lucassen et al., 2001). Recently, the potential effects of regulatory apoptosis in depression or some neurodegenerative diseases have been widely investigated (Ghaffari-Nasab et al., 2021; Park et al., 2020). Therefore, apoptosis in the hippocampus can be a prime focus area for research into the mechanisms of antidepressant drugs. Studies have shown that the loss of mitochondrial membrane potential, upregulation of the antiapoptotic proteins Bcl-2 and Bcl-xL and activation of the phosphatidylinositol 3-kinase (PI3K)/Akt signal are involved in the anti-apoptotic mechanisms of fluoxetine (Yang et al., 2017).

For hundreds of years, Guipi Tang (GPT), a traditional Chinese herbal formula prepared from 12 herbs, has been commonly used in China and Japan for the treatment of poor memory, insomnia, fatigue, palpitations, neurosis and anaemia (Araki et al., 2021). The major chemical ingredients of GPT include liquiritin, nodakenin and glycyrrhizin, which are all able to cross the blood–brain barrier (BBB) as itself or its main metabolite (Seo et al., 2014; Lee et al., 2015; Qin et al., 2022; Song et al., 2017; Tabuchi et al., 2012). It is reported that treating rats with GPT improves learning and memory and, at the cellular level, promotes cell proliferation in the hippocampal dentate gyrus (Oh et al., 2005). According to a previous report, the newly proliferated cells stimulated by GPT may differentiate into neurons (Oh et al., 2005). Smaller hippocampal volume is reported in major depressive disorder (MDD), and fewer granule neurons in unmedicated-MDD without fewer neuronal progenitor cells suggests a cell maturation or survival defect, perhaps related to MDD duration (Boldrini et al., 2013). A study demonstrated that GPT exerts no acute or genotoxic effects on rats and is hence safe (Park et al., 2014). Clinically, GPT improves symptoms of anxiety and life quality in elderly depression patients, which was later proven by a meta-analysis of randomized controlled trials (Sheng et al., 2017; Li et al., 2014). In addition, use a combination of GPT and antidepressants significantly ameliorate the symptoms of depression and increase the rates of effectiveness and recovery compared with antidepressant therapy alone (Sheng et al., 2017). In theory of traditional Chinese medicine, the deficiency of heart and spleen, which is often a key characteristic in MDD, can be nourished by GPT (Chen et al., 2022). However, the underlying molecular mechanisms of the anti-depressive effects of GPT remain unclear. Therefore, in the present study, we aimed to examine the potential anti-depressive effects of GPT and its underlying molecular mechanisms.

2. Materials and methods

2.1. Animals

A total of 72 male Sprague–Dawley rats weighing 160–180 g were obtained from the Experimental Animal Centre of the Shanghai University of Traditional Chinese Medicine. The rats were divided into two groups, namely, control (n = 12) and treatment (n = 60) groups, according to their body weight and baseline sucrose intake values. All animals were housed at 22 ± 2 °C, with a relative humidity of 55 ± 5% and a 12-h light/dark cycle (lights on at 7:00 a.m. and lights off at 7:00 p.m.), and had free access to food and water. All experimental protocols or animal handling procedures were approved by the Animal Care and Use Committee (ACUC) of the Shanghai University of Traditional Chinese Medicine (approval number: SZY201612002) and performed according to the Guide for the Care and Use of Laboratory Animals (National Research Council (US) Institute for Laboratory Animal Research, 1996).

2.2. Preparation of GPT

The 12 herbs that comprise GPT are shown in Table 1. All herbs obtained from the Zhongshan Hospital Affiliated with Fudan University were confirmed by the National Institutes for Food and Drug Control according to the Pharmacopeia of China (Chinese Pharmacopoeia Commission, 2015). The manufacturing processes and quality control methods were based on a previous study conducted in our centre (Zhang et al., 2017). The herbs were immersed and boiled in distilled water, in a proportion of 1 kg herbs: 1 L water, for 30 min and 1 h, respectively. The first decoction was obtained after the suspension was filtered via gauze, and the second and third decoctions were obtained after a repeated process. Thereafter, the suspension from the three decoctions was mixed, and dehydrated alcohol (final concentration [v/v]: 75% alcohol) was added to the mixture. After 24 h, the suspension was collected, and all liquid mixed with 75% alcohol was blended and centrifuged at 2000 × g for 20 min. The alcohol was removed using a rotary evaporator. The final concentration of GPT decoction was reconstituted with distilled water to 1 g/mL (w/v), and the decoction was stored at −20 °C until further use.

Table 1.

Composition of Guipi-tang.

Chinese name	Latin name	Family	The part used	Proportion
Bai Zhu	Atractylodes macrocephala Koidz.	Omposite	Root and Rhizome	3.0 g
Ren Shen	Panax ginseng C. A. Mey.	Araliaceae	Root and Rhizome	3.0 g
Huang Qi	Astragalus membranaceus Bunge	Leguminosae	Root	3.0 g
Dang Gui	Angelica sinensis (Oliv.) Diels	Umbelliferae	Root	3.0 g
Gan Cao	Glycyrrhiza uralensis Fisch.	Leguminosae	Root and Rhizome	1.0 g
Fu Ling	Poria cocos Wolf	Polyporaceae	Root	3.0 g
Yuan Zhi	Polygala tenuifolia Willd	Polygalaceae	Root	3.0 g
Suan Zaoren	Ziziphus jujuba Mill.	Rhamnaceae	Seed	3.0 g
Mu Xiang	Aucklandia lappa Decne.	Composite	Root	1.5 g
Long Yan Rou	Dimocarpus longan Lour.	Sapindaceae	Aril	3.0 g
Sheng Jiang	Zingiber officinale Rosc.	Zingiberaceae	Root	1.5 g
Da Zao	Ziziphus jujube Mill.	Rhamnaceae	Fruit	2.0 g

Note: All the plant names have been checked with the latest version in The Plant List (http://www.theplantlist.org).

2.3. Groups and drug administration

Rats were divided into two matched groups according to their body weight and baseline sucrose intake values. Baseline sucrose intake was established through a sucrose preference test (SPT) (described in a later section). Rats in the control group (n = 12) were given pure water, and those in the treatment group (n = 60) were subjected to a CUMS procedure for 28 days. Subsequently, the rats were divided into five groups based on their new sucrose preference index: CUMS (model, pure water), CUMS + low-dose GPT (oral GPT-L, 3.15 g/kg/day), CUMS + middle-dose GPT (oral GPT-M, 6.3 g/kg/day), CUMS + high-dose GPT (oral GPT-H, 12.6 g/kg/day) and CUMS + fluoxetine (oral FLU, 10 mg/kg) groups. The dose required for animals was converted from the dose recommended for humans, and the body surface area was normalised. The common clinical daily dose of GPT for humans is 30 g/60 kg (Yim et al., 2014); therefore, the dose (Dose_rat) required for a rat (body weight, 200 g) is calculated as (30 g/60 kg) × 6.3 = 3.15 g/kg. We selected 3.15, 6.3 and 12.6 g/kg/day as the low, medium and high dosages for rats.

2.4. Chronic unpredictable mild stress procedure

The CUMS procedure was performed as described in a previous study (Papp, 2012) with some modifications. Briefly, rats in the control group (no stress) were housed together (5 rats per cage, cage size: 55 × 40 × 20 cm) with free access to food and water, whereas stressed animals were housed separately (1 rat per cage, cage size: 29 × 18 × 16 cm) and subjected to the CUMS procedure for 8 consecutive weeks. The CUMS procedure consisted of the following stressors: food and water deprivation, soiled cages, restraint, cage tilting (45°), stroboscopic illumination, white noise, reversal of light/dark cycle, swimming in 4 °C cold water for 5 min, crowding, pair-housing, water deprivation followed by exposure to an empty water bottle and food deprivation followed by restricted food. A part of the protocol is shown in Table 2. To prevent habituation and ensure the unpredictability of the stressors, all stressors were randomly scheduled over a 1-week period and repeated throughout the 8-week experiment (Yi et al., 2014a). The body weight and sugar consumption of the rats were recorded weekly.

Table 2.

Main protocol for CUMS procedure.

	Stressors	Time
Day 1	Food and water deprivation	24 h (9:00–9:00)
Day 2	Inaccessible food	1 h (9:00–10:00) 1 h (9:00–10:00)
	Exposure to an empty bottle
	Pair-housing	8 h (10:00–18:00)
Day 3	Forced swimming (4 °C)	5 min (10:00)
Day 4	Cage-tilting 45°	8 h (10:00–18:00)
Day 5	Restraint 2–6 h	4 h (9:00–13:00)
	Stroboscopic illumination	15 h (18:00–9:00)
Day 6	Soiled Cage	20 h (13:00–9:00)
Day 7	Crowing	6 h (10:00–16:00)

2.5. Sucrose preference test

Because anhedonia is a core symptom of depression, the preference for sucrose was used as a standard of success for the animal model establishment. SPT was used to operationally define anhedonia and was performed as described previously (Konkle et al., 2003) with minor modifications before and after performing the CUMS procedure. The rats were first acclimatised to a 1% w/v sucrose solution and were subsequently exposed for 24 h to two bottles of 1% sucrose solution followed by 24-h exposure to one bottle of 1% sucrose solution and another bottle of tap water. After adaptation, the rats were deprived of water for 24 h and placed in individual cages for 1 h before the test. SPT was conducted by housing the rats in individual cages with free access to two bottles containing either 1% sucrose solution or tap water. After 6 h, the volumes of consumed sucrose solution and tap water were recorded, and sucrose preference was calculated as follows: sucrose preference (%) = sucrose consumption/(water consumption + sucrose consumption) × 100%.

2.6. Novelty-suppressed feeding test

The novelty-suppressed feeding test (NSFT) was performed as described in a previous study (Furmaga et al., 2011) with minor modifications. Briefly, a plastic box (70 × 70 × 45 cm) that was painted black was prepared in a room with dim lighting, and a single pellet of food was placed on a white paper platform in the centre of the box. The experiment was performed after the rats were deprived of food and water for 24 h. The rats were transferred to the room 2 h before the test and were subsequently individually placed in the corner of the apparatus. Latency from the time when the rats were placed in the corner to the moment the rats started to eat food was recorded within 5 min. Thereafter, the rats were immediately transferred to the home cage for the food consumption test. A pre-weighed pellet was kept in the cage, and its weight was measured after 10 min.

2.7. Open-field test

The open-field test (OFT) was used to detect spontaneous locomotion activity, exploration and anxiety-related behaviours of the rats. It was performed as previously described (Christoffel et al., 2012; Monteggia et al., 2007; Asakawa et al., 2016) with minor modifications. Rats were placed in the test room to adapt to the environment 1 h before beginning the experiment. Each rat was individually placed in the centre of the quadrangular arena (68 × 68 cm, with 45-cm-height walls) under dim lighting. A video tracking system (Ethovision 3.0, Noldus, Leesburg, Virginia) was used to record latency to enter the arena, the time spent in the arena, the distance moved and the frequency of crossing the arena. In addition, time spent in a delineated ‘periphery zone’, a delineated ‘non-periphery zone’ and a delineated ‘centre zone’ (32 × 32 cm) was recorded during the 5-min test. Rat droppings were removed, and the quadrangular arena was cleaned with alcohol before each animal test.

2.8. Forced swim test

In the forced swim test (FST), the rats were forced to individually swim in a vertical behavioural test cylinder (diameter, 20 cm; height 45 cm) filled with fresh water and maintained at 24 ± 1 °C. Water in the cylinder was changed for every rat and maintained at a height of 30 cm, which was sufficient to prevent the rats from touching the bottom with their tail. The rats were allowed to swim for 15 min as a training period 24 h before commencement of the test. During the test, the rats were observed for 4 min after vigorous activity of 2 min, and the whole 6-min process was videotaped for subsequent quantification. The rats were considered immobile if they remained floating on the surface with their front paws making movements that were necessary to keep them afloat (Tonissaar et al., 2008).

2.9. Enzyme-linked immunosorbent assay

After FST, the rats were deeply anaesthetised, and blood was collected from their abdominal aortas. An enzyme-linked immunosorbent assay (ELISA) kit (AR E-8100, Labor Diagnostika Nord GmbH & Co.) was used to evaluate the amount of serum corticosterone (CORT). The blood samples were centrifuged at 1008 g for 15 min at 4 °C. The calibrator and serum (10 μL) samples were transferred into a 96-well plate, followed by the addition of an incubation buffer (100 μL) and an enzyme conjugate (50 μL). The plate was incubated for 2 h at room temperature on a microplate mixer (>600 rpm). Thereafter, the wells were rinsed 4 times with diluted wash solution (300 μL/well) after discarding their contents. The substrate solution (200 μL) was added to the wells and incubated for 30 min in the dark, and the stop solution (50 μL) was added to each well to terminate the reaction. Finally, the absorbance was determined at 450 nm within 15 min.

2.10. Terminal deoxynucleotidyl transferase dUTP nick end labeling

Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) was performed according to the protocol of the In Situ Cell Death Detecting Kit (Roche Diagnostics GmbH, Penzberg, Germany) to evaluate apoptosis by labelling DNA fragmentation. Briefly, 0.3% H₂O₂ was used to block tissue sections in methanol, and the sections were subsequently washed thrice with PBS (pH 7.4) for 5 min. Thereafter, the sections were incubated in 20 μg/mL of proteinase K and PBS (pH 7.4) for 20 min at 37 °C and washed with PBS (5 times for 3 min each). The TUNEL reaction mixture was added to the specimen and incubated for 2 h at 37 °C. After rinsing thrice with PBS (5 min each time), the sections were stained with DAPI (Beyotime) for 5 min, and the slides were sealed with coverslips using Permount.

2.11. Western blotting

After 24 h of FST, the rats were sacrificed via cardiac perfusion with normal saline, and the hippocampus was dissected. Hippocampal tissues were then frozen in liquid nitrogen and stored at −80 °C. The tissues were prepared with protease inhibitors (Beyotime, Haimen, Jiangsu, China) and phenylmethylsulfonyl fluoride (Beyotime, Haimen, Jiangsu, China) in a lysis buffer and centrifuged at 13,500 g for 20 min at 4 °C, and the supernatant was collected. A BCA kit (Beyotime) was used to detect protein concentration. Thereafter, protein samples at appropriate concentrations were subjected to electrophoresis on polyacrylamide gels and subsequently transferred onto polyvinylidene fluoride membranes (PVDF, Millipore, Bedford, MA, USA). After blocking the membranes with 5% non-fat milk solution that was mixed with milk, they were incubated with tris-buffered saline and 0.1% Tween-20 (TBST) for 1 h. Subsequently, the membranes were incubated with primary antibodies at 4 °C overnight (1:1000; Bcl-2 [#3498], Bax [#14796], phospho-PI3K [#4228], PI3K [#4257], Akt and phospho-Akt [#8205], cleaved caspase-3 [#9661], TrkB [#4603], CREB [#4820] and p-CREB [#9198] from Cell Signaling Technology [Beverly, MA, USA] and brain-derived neurotrophic factor (BDNF) (ab108319) from Abcam [Cambridge, MA, USA]). The membranes were rinsed with TBST for 10 min thrice and then incubated with horseradish peroxidase-conjugated secondary anti-rabbit antibody (Beyotime, China) for 2 h. Finally, the membranes were incubated with a chemiluminescence reagent (Millipore) and detected using a FluorChem FC2 gel imaging system (Alpha Innotech, Santa Clara, CA, USA). Bands were densitometrically analysed using the Quantity One software (Bio-Rad Laboratories, Hercules, CA, USA).

2.12. Immunofluorescence

The rat brain tissues were fixed in 4% paraformaldehyde (pH 7.4) at 4 °C for 24 h and subsequently dehydrated in a graded series of sucrose at 20 and 30%. Frozen sections of the brain were blocked with normal goat serum for 20 min at room temperature and incubated with anti-BDNF IgG (1:500, ab108319, Abcam) overnight at 4 °C. They were washed thrice with PBS (5 min each time) and allowed to react with rhodamine-conjugated goat anti-rabbit IgG for 1 h. The brain sections were counter-stained with DAPI for 10 min at 37 °C. BDNF-positive cells were observed using an immunofluorescence microscope (Olympus DP71, Olympus Co., Tokyo, Japan). Five randomly selected fields were used to count the average number of positive cells in the regions.

2.13. Compositional analysis of Guipi Tang with UPLC-HRMS

2.13.1. Chromatographic conditions

The Agilent 1290 Infinity II liquid chromatography system was used for ultra-performance liquid chromatography (UPLC) analysis, including a gG4220A binary pump, G4226A autosampler, G1316C column thermostat and G4212A DAD. The Waters ACQUITY UPLC® HSS T3 column (2.1 × 100 mm, 1.8 μm) was used for chromatographic separation. Aqueous formic acid (0.1%, A)-Formic acid acetonitrile (0.1%, B) was used as the mobile phase. The following procedure was used for gradient elution: 0–10 min, 0% B; 10–35 min, 0–15% B; 35–75 min, 15–40% B; 75–88 min, 40–95% B; 88–90 min, 95% B; 90–90.1 min, 95–0% B and 90.1–92 min, 0% B. The flow rate was set at 0.3 min/mL, the column temperature was 30 °C, the detection wavelength was 280 nm and the injection volume was 1 μL.

2.13.2. Mass spectrometry conditions

The AB Sciex Triple TOF® 4600 system was used for mass spectrometry, which contains an electrospray ion source and a quadrupole-time-of-flight tandem mass analyser. First-level (TOF-MS) and second-level (product ion) detection parameters were set respectively. A) TOF-MS parameters: Electrospray positive and negative ion source voltages were 5000 V and −4500 V, respectively; ion source temperature was 500 °C; the atomising gas used was N₂; atomising gas pressure (GS1) was 50 psi; heating gas pressure was (GS2) 50 psi; curtain gas pressure was 35 psi; declustering potential was 100 V; collision energy was 10 V and mass range (m/z) was 50–1700. B) Product ion parameters: The range of the secondary mass spectrum (m/z) was 50–1250, declustering potential was 100 V, collision energy was ±40 eV, collision voltage swing was 20 eV, ion release delay was 30 ​m ​s and ion release width was 15 ​m ​s.

2.14. Statistical analysis

All data were collected and presented as mean ± standard error of the mean (SEM). The Student’s t-test was used for between-group comparisons. Behavioural tests were analysed using two-way analysis of variance (ANOVA) followed by the least significant difference (LSD) post-hoc test, and one-way ANOVA with LSD post-hoc analysis was performed for other multiple comparisons. Data graphics were made using GraphPad Prism (version 7.0a, GraphPad Software Inc., San Diego, CA, USA). A P-value <0.05 was considered statistically significant.

3. Results

3.1. Composition of GPT

In order to further reveal the association between GPT and its anti-depression effects, the main composition of GPT was analysed. After UPLC-Q-TOF/MS was performed, 36 compounds were identified based on multi-stage mass spectrometry information and a high-resolution mass spectrometry database for natural products (Table 3, Supplementary Figures S1–3). Among these compounds, sucrose, uridine, maltohexaose, polygalaxanthone III, liquiritin, calycosin-7-O-glucoside, polygalaxanthone VIII, liquiritinapioside, spinosyn, ginsenoside Rg1, ginsenoside Re, glycyrrhizic acid and licoricesaponin K2 were some main compounds of GPT (Table 3).

Table 3.

Composition of Guipi-Tang identified with UPLC-HRMS.

No.	Time min	Adduct ion	m/z Actual value	m/z Theoretical value	ppm	Molecular formula	Molecular weight	Chinese name	English name	MS/MS data	Affiliation
1∗	1.31	[M-H]-	341.1111	341.1089	3.7	C12H22O11	342.30	蔗糖	Sucrose	341.1071; 179.0556; 119.0349	/
2∗	4.18	[M-H]-	243.0632	243.0623	6.3	C9H12N2O6	244.20	尿苷	Uridine	200.0597; 152.0383; 124.0418; 110.0275	/
3	9.20	[M + FA-H]-	873.2749	873.2729	4.1	C30H52O26	828.72	麦芽五糖	Maltopentaose	873.2886; 827.2669; 647.1971; 485.1450; 341.1121	/
4	11.01	[M-H]-	282.0861	282.0844	3.6	C10H13N5O5	283.24	鸟苷	Guanosine	282.0869; 150.0432; 133.0164; 108.0207	/
5∗	15.49	[M + FA-H]-	1035.3315	1035.3257	7.5	C36H62O31	990.86	麦芽六糖	Maltohexaose	989.3179; 827.2705; 665.2177; 341.1103	/
6	19.48	[M-H]-	329.0894	329.0878	3.0	C14H18O9	330.29	1-O-香草酰-β-D-葡萄糖苷	1-O-vanilloyl-beta-D-glucose	167.0353; 152.0117; 123.0462; 108.0225	/
7	26.42	[M-H]-	353.0877	353.0878	2.5	C16H18O9	354.31	绿原酸	Chlorogenic acid	353.0878; 191.0564; 161.0275	Saussurea lappa
8	27.60	[M-H]-	353.0883	353.0878	4.2	C16H18O9	354.31	隐绿原酸	Cryptochlorogenic acid	353.0864; 191.0576; 179.0359; 173.0465; 135.0464	Saussurea lappa
9	27.74	[M-H]-	517.1599	517.1563	5.8	C22H30O14	518.47	西伯利亚远志糖A5	Sibiricose A5	517.1608; 341.1072; 175.0424	Polygala tenuifolia
10	29.08	[M-H]-	547.1672	547.1668	4.5	C23H32O15	548.49	西伯利亚远志糖A6	Sibiricose A6	547.1756;341.1108;223.0655;205.0524	Polygala tenuifolia
11	31.53	[M]+	342.1673	342.1700	-6.4	C20H24NO4	341.40	木兰花碱	Magnoflorine	342.1681; 297.1102; 265.0837	Saussurea lappa
12∗	37.67	[M + H]+	569.1475	569.1501	-6.7	C25H28O15	568.48	远志口山酮III	Polygalaxanthone III	569.1466; 419.0960; 317.0617; 287.0523	Polygala tenuifolia formula
13∗	37.68	[M-H]-	417.1206	417.1191	5.0	C21H22O9	418.39	甘草苷	Liquiritin	417.1211; 255.0675; 135.0095; 119.0508	Honey-roasted licorice
14∗	37.68	[M + H]+	447.1250	447.1286	-6.9	C22H22O10	446.40	毛蕊异黄酮苷	Calycosin-7-O-glucoside	285.0750; 270.0502	Astragalus
15∗	38.29	[M + H]+	569.1460	569.1501	-5.8	C25H28O15	568.48	远志口山酮VIII	Polygalaxanthone VIII	569.1475; 437.1058; 419.0944; 317.0640; 287.0519	Polygala tenuifolia formula
16∗	38.35	[M-H]-	549.1628	549.1614	5.3	C26H30O13	550.51	芹糖甘草苷	Liquiritin apioside	549.1640; 417.1241; 255.0675; 135.0091	Honey-roasted licorice
17∗	38.77	[M-H]-	607.1718	607.1668	6.7	C28H32O15	608.54	斯皮诺素	Spinosin	607.1711; 427.1022; 307.0619	Saussurea lappa
18	44.30	[M-H]-	753.2318	753.2248	5.4	C34H42O19	754.69	3,6′-二芥子酰基蔗糖	3,6′-Disinapoyl sucrose	753.2292; 649.1778; 547.1690; 529.1588; 223.0635; 205.0530	Polygala tenuifolia formula
19	45.05	[M-H]-	515.1210	515.1195	5.0	C25H24O12	516.45	异绿原酸C	Isochlorogenic acid C	515.1313;353.0892;191.0568;173.0487	Saussurea lappa
20	47.08	[M-H]-	681.2042	681.2036	5.1	C31H38O17	682.62	细叶远志苷A	Tenuifoliside A	681.2008; 443.1167; 281.0633	Polygala tenuifolia formula
21∗	49.72	[M + FA-H]-	845.4939	845.4904	6.7	C42H72O14	801.01	人参皂苷 Rg1	Ginsenoside Rg1	845.4991; 799.4902; 637.4341; 475.3781	Ginseng
22∗	50.02	[M + FA-H]-	991.5512	991.5483	6.3	C48H82O18	947.15	人参皂苷 Re	Ginsenoside Re	991.5583; 945.5489; 799.4909; 783.4899	Ginseng
23	61.96	[M-H]-	679.3771	679.3699	7.2	C36H56O12	680.82	细叶远志皂苷	Tenuifolin	679.3709; 425.3075	Polygala tenuifolia formula
24	62.14	[M-H]-	837.3957	837.3914	6.3	C42H62O17	838.93	甘草皂苷 Q2	Licorice saponin Q2	837.3866; 351.0535; 193.0353	Honey-roasted licorice
25	64.30	[M-H]-	1153.6168	1153.6011	9.8	C54H92O23	1109.29	人参皂苷Rb1	Ginsenoside Rb1	1153.6151; 1107.6005; 945.5422; 783.5051	Ginseng
26	65.77	[M-H]-	837.3941	837.3914	6.5	C42H62O17	838.93	甘草皂苷 P2	Licoricesaponin P2	837.3870; 643.3316; 485.3397; 351.0549	Honey-roasted licorice
27	68.47	[M-H]-	837.3969	837.3914	7.0	C42H62O17	838.93	甘草皂苷 G2	Licorice saponin G2	837.3973; 351.0717; 193.0388	Honey-roasted licorice
28∗	69.40	[M-H]-	821.4056	821.3965	8.8	C42H62O16	822.93	甘草酸	Glycyrrhizic acid	821.3992; 645.3643; 351.0581	Honey-roasted licorice
29∗	72.37	[M-H]-	821.4076	821.3965	7.3	C42H62O16	822.93	甘草皂苷 K2	Licorice saponin K2	821.3949;351.0558;193.0382	Honey-roasted licorice
30	75.54	[M + H]+	191.1056	191.1067	-2.4	C12H14O2	190.24	E-丁基苯酞	E-butylphthalide	191.1054; 145.1029; 117.0675	Angelica
31	78.25	[M + H]+	191.1065	191.1067	-2.4	C12H14O2	166.22	藁本内酯	Ligustilide	191.1106; 145.1013; 115.0507	Angelica
32	79.18	[M + FA-H]-	913.4829	913.4802	6.4	C45H72O16	869.04	黄芪皂苷I	Astragaloside I	913.4862; 867.4777; 825.4803; 807.4620	Honey-roasted licorice
33	79.19	[M + H]+	233.1533	233.1536	-3.5	C15H20O2	232.32	白术内酯II	Atractylenolide II	233.1520; 215.1363; 187.1475; 161.1309	Fried atractylodes with honey bran
34	79.19	[M + H]+	233.1533	233.1536	-3.5	C15H20O2	232.32	木香烃内酯	Costunolide	233.1521; 215.1366; 187.1477; 161.1310	Saussurea lappa
35	80.43	[M + H]+	231.1374	231.1380	-1.5	C15H18O2	230.30	白术内酯I	Atractylenolide I	213.1303; 185.1313; 165.0674; 157.0997; 143.0847; 129.0688	Fried atractylodes with honey bran
36	80.43	[M + H]+	231.1374	231.1380	-1.5	C15H18O2	230.30	去氢木香内酯	Dehydrocostus lactone	213.1311; 185.1311; 165.0681; 157.0999; 143.0848; 129.0675	Saussurea lappa

^∗The main component with higher content.

3.2. GPT alleviated CUMS-induced depression-like behaviours in rats

The effects of GPT on depression-like behaviour in rats were investigated using behavioural tests including SPT, FST, OFT and NSFT. In SPT, the sucrose intake values were significantly lower in the model group than in the control group after 3 weeks of CUMS induction (p < 0.01; F = 3.157, p = 0.12, Figure 1A). After 4 weeks of drug intervention, differences in sucrose intake values between the treatment and model groups were analysed using two-way ANOVA. Sucrose intake was significantly higher in the GPT-H, GPT-M, GPT-L and FLU groups than in the model group (F = 8.48, p < 0.01, Figure 1B), indicating that GPT improved CUMS-induced anhedonia in rats with depression-like behaviour. In NSFT, two-way ANOVA was performed to analyse the latency for different groups. As shown in Figure 1C, latency in the model group was significantly longer than that in the control group (p < 0.01). However, latency in the GPT-H, GPT-M and FLU groups (134.57 ± 18.60 s) was significantly lower than that in the model group (F = 4.53; p < 0.05, p < 0.05, p < 0.01, respectively, Figure 1C). No significant differences in home-cage sucrose intake were observed among the six groups (F = 0.99, p > 0.05, Figure 1D), which was consistent with the results of a previous study (Yi et al., 2014b). These results suggested that GPT could reduce latency in the time required by CUMS rats to take the first bite of food during NSFT, and the effects of GPT may not result from a general increase in feeding.

Figure 1.

GPT treatment exerts antidepressant-like effects in the sucrose preference test (SPT) and novelty-suppressed feeding test (NSFT). (A) Baseline sucrose intake before drug intervention in rats in the control and model groups. (B) The GPT and FLU groups had significantly higher sucrose intake than the model group. (C) Time spent in NSFT. (D) Food consumption in NSFT (∗p < 0.05 and ∗∗p < 0.01 versus the model group; ^#p < 0.05 and ^##p < 0.01 versus the control group).

In OFT, the movements of rats were recorded using a video tracking system, and the results are shown in Figure 2A. Exposure to the CUMS procedure significantly reduced the time spent by rats in the central and non-peripheral zones and decreased the frequency of rats crossing the centre. As shown in Figure 2B, the time spent by rats in the centre was significantly shorter in the model group than in the control group (p < 0.01). However, the time spent by rats in the centre was significantly longer in the GPT-H, GPT-M, GPT-L and FLU groups (F = 8.10, p < 0.01). The time spent by rats in the non-peripheral zone was significantly shorter in the model group than in the control group (p < 0.01) and longer in the GPT-H and GPT-M groups than in the model group (F = 3.49, p < 0.05, Figure 2B). The frequency of crossing the centre was lower in rats exposed to the CUMS procedure (4.88 ± 0.91) than in the control group and significantly higher in the GPT-H and GPT-M groups than in the model group (F = 4.11, p < 0.05 or p < 0.01, Figure 2C). Although differences in the frequency of crossing the non-peripheral zone were not statistically significant, a trend similar to that for the frequency of crossing the centre was observed in GPT groups, compared with the control group, and a tendency towards a higher frequency of crossing the non-peripheral zone was observed in the GPT and FLU groups (p > 0.05, Figure 2C).

Figure 2.

Effects of GPT on CUMS rats in the open field test and forced swim test. (A) Typical tracks of the movements of rats. (B) Time spent in the central and non-peripheral areas. (C) Frequency of crossing the central and non-peripheral areas. In each group, n = 10 (∗p < 0.05 and ∗∗p < 0.01 versus the model group; ^#p < 0.05 and ^##p < 0.01 versus the control group).

In FST, the immobility duration of rats was recorded as a measure of behavioural despair (Lucki, 1997). As depicted in Figure 2D, the immobility time was significantly longer in the model group than in the control group (p < 0.01). Chronic (8 weeks) treatment with GPT at moderate and high doses reduced the immobility duration when compared with the model group (F = 8.79, p < 0.01 or p < 0.05). A similar effect was also observed in the GPT-L group. The results of FST indicated that GPT decreased the immobility duration and exhibited antidepressant-like effects.

3.3. GPT increased the blood corticosterone levels in CUMS rats

As shown in Figure 3A, the post-hoc analysis indicated that CUMS led to an evident increase in the serum CORT levels (p < 0.05). The serum CORT levels were significantly lower in the GPT-H group than in the model group (F = 2.16, p < 0.05) but did not significantly decrease in the GPT-M, GPT-L and FLU groups as compared with the model group (p > 0.05).

Figure 3.

Effects of GPT on blood CORT levels and neuronal apoptosis in the hippocampus. (A) Blood CORT levels were evaluated using ELISA. (B) Photomicrographs of neuronal apoptosis in the hippocampus (400 × magnification). The emergence of green and blue represents TUNEL-positive cells. (C) Positive cells were counted. (D) The percentage of apoptotic neurons was calculated; n = 5 for each group (∗p < 0.05 and ∗∗p < 0.01 versus the model group; ^#p < 0.05 and ^##p < 0.01 versus the control group).

3.4. GPT reduced neuronal apoptosis in the hippocampi of CUMS rats

TUNEL staining was used to evaluate neuronal apoptosis in the hippocampus. As demonstrated in Figure 3B, TUNEL-positive neuronal cells were frequently observed in the hippocampi of rats exposed to CUMS; however, in the control group, only TUNEL-negative neurons were observed. After GPT treatment, the number of cells decreased in a dose-dependent manner. Moreover, as depicted in Figure 3C and D, the proportion of TUNEL-positive neuronal cells was higher in the model group than in the control group (p < 0.01) but was lower in the FLU, GPT-H and GPT-M groups (F = 57.83, p < 0.01) than in the model group. The cell apoptotic rates in the GPT-L group relative to those in the model group showed no significant differences. These results suggested that GPT inhibited CUMS-induced cell apoptosis in the hippocampus in a dose-dependent manner.

3.5. GPT activated the PI3K/Akt pathway and regulated the levels of apoptotic proteins

As shown in Figure. 4A and B, based on the one-way ANOVA test, a marked difference was found in the expression of p-PI3K/PI3K and p-AKT/AKT in the hippocampi between the model and control groups. Treatment with FLU and GPT significantly upregulated the phosphorylation of PI3K and p-AKT in the hippocampus (p < 0.05 or 0.01, Figure 4A; p < 0.05, Figure 4B, respectively) when compared to the model group. The increased levels of p-PI3K and p-AKT were not attributable to an increase in total PI3K or AKT levels. However, the GPT treatment (3.15 g/kg) did not have a significant effect on protein expression or phosphorylation of PI3K or AKT (p > 0.05, Figure 4A and B).

Figure 4.

Effects of GPT on phosphorylation levels of PI3K/AKT and protein levels of Bax, Bcl-2 and cleaved caspase-3. (A) Semi-quantitative analysis of PI3K phosphorylation detected using western blotting. (B) Quantitative analysis of AKT phosphorylation via western blotting. (C) The expression of Bcl-2 and Bax. (D) The expression of cleaved caspase-3. GAPDH was used as the internal control; n = 6 for each group (∗p < 0.05 and ∗∗p < 0.01 versus the model group; ^#p < 0.05 and ^##p < 0.01 versus the control group).

The expression of the classical apoptosis-related proteins Bax and Bcl-2 in the hippocampus of rats was measured via western blotting (Figure 4C). Rats in the model group had increased Bax expression and decreased Bcl-2 expression in the hippocampus as compared with rats in the control group. However, rats in both FLU and GPT treatment groups had increased Bcl-2 expression and decreased Bax expression in the hippocampus as compared with rats in the model group. Changes in the Bax-to-Bcl-2 ratio are more sensitive and valid than the level of individual proteins; therefore, the Bax/Bcl-2 ratio in the hippocampi of rats was also assessed. The hippocampal Bax-to-Bcl-2 ratio was higher in the model group than in the control group (p < 0.01) but significantly decreased after FLU or GPT treatment (p < 0.01). Moreover, the Bax/Bcl-2 ratio in the hippocampus of rats treated with 6.3-g/kg GPT (GPT-M group) was the lowest.

As shown in Figure 4D, the activity of caspase-3 in the hippocampus was significantly higher (with higher levels of cleaved caspase-3) in CUMS rats than in the control group (p < 0.01). However, treatment with FLU (10 mg/kg) and GPT (6.3 g/kg, 12.6 g/kg) decreased the expression of cleaved caspase-3 (p < 0.01) in a dose-dependent manner. GPT (3.15 g/kg) treatment did not affect the activity of caspase-3.

3.6. GPT induced BDNF/TrkB/CREB signalling

The expression of BDNF in hippocampi was evaluated through immunofluorescence (Figure 5A) and western blotting (Figure 5B). The results showed that BDNF was significantly (p < 0.01) repressed in rats exposed to CUMS (Figure 5A and B) when compared with rats in the control group. Subsequently, and the level of the key molecules, including TrkB and p-CREB, in the downstream signals, which was measured with western blotting, also showed a similar trend (p < 0.05; Figure 5C and D). However, the expression of BDNF and TrkB and phosphorylation of CREB were almost restored in rats treated with GPT when compared with rats in the model group (p < 0.05; Figure 5A–D).

Figure 5.

Effects of GPT on the expression of BDNF, TrkB, CREB and p-CREB in the hippocampus of the brain. (A) Immunofluorescence staining was performed to qualitatively analyse BDNF expression in the hippocampus. (B–D) Western blotting was used to semi-quantitatively measure the relative level of BDNF, TrkB, CREB and p-CREB in the hippocampus (∗p < 0.05 and ∗∗p < 0.01 versus the model group; ^#p < 0.05 and ^##p < 0.01 versus the control group).

4. Discussion

In the present study, we investigated the antidepressant effects of GPT and found that GPT alleviated depression-like behaviours and reduced neuronal apoptosis. In addition, we tried to reveal the underlying molecular mechanisms of these antidepressant effects and demonstrated that treatment with certain doses of GPT had significant antidepressant efficacy and inhibited unpredictable chronic mild stress-induced neuronal apoptosis in the hippocampi of rats. After investigating the ability of the main active components in GPT to pass through BBB, we found that uridine, liquiritin, liquiritin apioside, ginsenoside Rg1, ginsenoside Re, and glycyrrhizic acid are able to penetrate BBB, and therefore get access to directly interact with hippocampal cells (Cansev, 2006; Zheng et al., 2018; Qin et al., 2022; Li et al., 2015; Shi et al., 2013; Tabuchi et al., 2012).

Consistent with previous studies (Bambico et al., 2015), we found that sucrose preference was significantly reduced in CUMS-exposed (8 weeks) rats, which was corrected via GPT treatment. This finding indicated that GPT could correct the inability to experience pleasure in CUMS-exposed rats. As a measure of anxiety-like behaviour, NSFT is used to examine depression-like behaviour because it is sensitive to chronic antidepressants (Gordon and Hen, 2004; Yi et al., 2010). The first feeding latency is an indication of increased anxiety levels in NSFT. In accordance with a previous study (Li et al., 2016), we found a significant decrease in the first feeding latency in rats treated with GPT. Total food consumption between groups showed no differences, indicating that the effects of GPT did not result from a general increase in feeding (Yi et al., 2014b). Administration of GPT also exerted antidepressant-like effects by reducing the immobility duration in FST with a corresponding increase in escape-oriented behaviours. FST has been interpreted as a model of ‘behavioural despair’ and used to examine antidepressant-like behaviour in numerous pharmacological and genetic rodent models (Cryan et al., 2002). Time spent in an open and bright arena is a key index to assess anxiety-like behaviour in OFT because rats avoid entering such spaces (Monteggia et al., 2007). The distance of movement in the route from the centre can reflect the ability to explore a new environment and the level of anxiety of rats. Rats exposed to the CUMS procedure spent significantly less time in the centre and non-peripheral zone and crossed the centre less frequently. These changes were corrected by the administration of GPT and FLU, thus demonstrating the effects of GPT on depression-related anxious behaviours. These results indicated that GPT could reverse depression-like behaviours and exert antidepressant effects in rats.

Regarding the level of CORT, no significant difference was observed between the model and treatment groups. This may be because CORT secretion is greatly affected by the time of day. In this study, blood was collected regardless of the time of day, it may have introduced a high level of variability in CORT measurements. Despite the non-significance after middle or low dose of GPT, a reduction tendency can be obviously seen from the results (Figure 3A). The seral corticosterone reduction may be caused by these molecules in GPT: sucrose, liquiritin, ginsenoside Rg1, and ginsenoside Re. The seral corticosterone reduction may be caused by these molecules in GPT: sucrose, liquiritin, ginsenoside Rg1, and ginsenoside Re. It is shown that sucrose intake lowers cold exposure-induced corticosterone secretion (Bell et al., 2002). Liquiritin is able to restrain the reduction of BDNF/TrkB signal, which is suppressed in depressed patients or animals after corticosterone treatment (Wu et al., 2018). It is possible that liquiritin causes a reactivation of BDNF-mediated signal and then negatively feedback to the regulation of corticosterone, leading to its decrease. Ginsenoside Rg1 can also inhibit the increase of plasma level of corticosterone, which might promote the recovery of hypothalamic-pituitary-adrenal (HPA) axis and hypothalamic-pituitary-gonadal (HPG) axis (Mou et al., 2017). Ginsenoside Re significantly increases BDNF mRNA level like Ginsenoside Rg1 does (Lee et al., 2020a, 2020b).

The abovementioned findings revealed that GPT exerted antidepressant effects in CUMS-exposed rats. To further examine the antidepressant mechanisms of GPT, we conducted a more in-depth study. Apoptosis has been proposed as a mechanism leading to stress-related depression-like behaviour. Without appropriate trophic support or activity, newly generated neurons in adulthood die via apoptosis (Deckwerth et al., 1996). For instance, enhanced apoptosis was found in the brain structures of maternally separated rats (Lee et al., 2001). In addition, previous post-mortem studies have revealed increased hippocampal apoptosis in the entorhinal cortex and hippocampal subfields (Lucassen et al., 2001). The antidepressant tianeptine reduced apoptosis in the dentate gyrus and temporal cortex, indicating that antidepressant drugs may influence cell viability through trophic/protective properties (Lucassen et al., 2004). In addition, the antidepressant fluoxetine prevents apoptosis in the dentate gyrus of maternally separated rats (Lee et al., 2001) and rats subjected to CUMS (Yang et al., 2017). In this study, apoptosis was observed in the hippocampus of rats exposed to CUMS and was significantly reduced in rats treated with GPT and fluoxetine. This finding indicated that GPT might exert antidepressant effects by reducing hippocampal neuronal apoptosis. To further verify this finding, we examined the apoptosis-related proteins Bax, Bcl-2 and caspase-3.

Caspase-3 activity is considered a definite marker of apoptosis (Jarskog et al., 2004). By stimulating NMDA receptors, caspase-3 can be transiently activated via the mitochondrial pathway in long-term depressed mice (Li et al., 2010). Consistent with a previous study (Yang et al., 2017), we found that the level of cleaved caspase-3 was significantly increased in the hippocampi of rats with depression-like behaviour, and this increase could be adjusted via GPT treatment. Bax is required for trophic factor deprivation-induced neuronal death (Deckwerth et al., 1996). The Bax/Bcl-2 ratio is important for cell survival because it critically controls apoptosis. The increased Bax/Bcl-2 ratio causes the mitochondria to release cytochrome C into the cytoplasm to trigger subsequent apoptotic signals (Raisova et al., 2001). In this study, we observed an increased Bax/Bcl-2 ratio after CUMS exposure, and this finding is consistent with that reported previously, which indicated that CUMS exposure resulted in increased expression of the proapoptotic protein Bax and the activation of its downstream caspases (Zhang et al., 2017). However, GPT suppressed the expression of Bax protein but promoted the expression of anti-apoptotic Bcl-2 and inhibited the cleavage of caspase-3. Therefore, these results supported the hypothesis that cell apoptosis is involved in the antidepressant effects of GPT.

The potential mechanisms underlying the antidepressant effects of GPT on rats exposed to CUMS should be further clarified via a neuronal anti-apoptotic approach. Studies have demonstrated that the PI3K/Akt pathway, which is important for neuroprotection and anti-apoptotic function, regulates survival through anti-apoptotic activation in the brain and hippocampal neurons (Shehata et al., 2012; Yang et al., 2017; Lu et al., 2006). Upon stimulation by growth and survival factors, receptor tyrosine kinases (RTKs) mediate the activation of the PI3K/Akt pathway, resulting in the phosphorylation of BAD. Subsequently, p-Bad (Ser136) binds to the secluded 14-3-3 proteins in the cytosol and blocks the activated signal of caspase-3 to induce apoptosis (Yu et al., 2016). Consistent with a previous study, CUMS exposure decreased the activation levels of PI3K/AKT (Yang et al., 2017), whereas treatment with GPT and fluoxetine significantly upregulated the levels of PI3K/AKT. Altogether, we demonstrated for the first time that GPT inhibited hippocampal apoptosis through the PI3K/AKT pathway in CUMS rats.

BDNF has antiapoptotic and antidepressant effects, and its level is decreased in depression (Dwivedi, 2009). In this study, we confirmed this decrease in the model group and showed that the decrease was reversed by GPT; in addition, the expression of TrkB and activation of CREB were also restored. These targets have attracted the attention of many researchers in the context of depression (Ye et al., 2017; Lee et al., 2020a, 2020b). Furthermore, reactivation of BDNF may result in better therapeutic effects. However, studies have also shown that BDNF has both pro- and anti-depressant effects, which depend on different regions of the brain (Miyanishi and Nitta, 2021). In the hippocampus and frontal cortex, BDNF expression is reduced after depression, and BDNF upregulation stimulates neurogenesis in the hippocampus and restores depression-like behaviour induced by chronic stress (Miyanishi and Nitta, 2021). In this study, BDNF was extracted from the hippocampus. The expression of BDNF restored owing to GPT treatment might benefit patients. However, a previous study demonstrated that various other physiological changes are also involved in the mechanisms of neuronal apoptosis in depressed rats. Umbelliferone can reduce neuronal apoptosis by blocking neuroinflammation induced by high levels of the inflammatory cytokines IL-1, IL-6 and TNF-α (Qin et al., 2017). The upregulated expression of BDNF and activated cascades of the BDNF/MAPK/ERK pathway/Bcl-2 are also neuroprotective against hippocampal neuronal apoptosis (Peng et al., 2008).

To better understand the association between GPT and its anti-depression effects, the composition of GPT was revealed. As a result, 13 main compounds were screened from the identified 36 compounds, which mainly belonged to glucosides, and some of them have been reported to have anti-depression effects, such as liquiritin (Zhao et al., 2008), ginsenoside Rg1 (Fan et al., 2018) and ginsenoside Re (Lee et al., 2012). In these anti-depression effects, defence against oxidative stress, suppression of glial activation and/or modulation of the central noradrenergic system may be involved. In addition, non-glycosides may also be involved. Clinically, glycyrrhizic acid can adjunctively treat depression through anti-inflammation (Cao et al., 2020).

In summary, we demonstrated that GPT exhibited antidepressant effects on depression-like behaviour and cell apoptosis in the hippocampi of CUMS rats, which may involve hippocampal neuronal apoptosis regulated by the PI3K/Akt pathway and/or other pathways mediated by BDNF/TrkB/CREB. These findings provide further insights into the function of GPT as a potential antidepressant agent and support the clinical application of GPT with a more detailed and validated potential molecular basis.

Declarations

Author contribution statement

Tingting Li, Xiangting Li, Jingsi Zhang: Performed the experiments; Analyzed and interpreted the data; Wrote the paper.

Zhonghai Yu, Fan Gong, Jun Wang, Haiyan Tang: Analyzed and interpreted the data.

Jun Xiang, Wen Zhang: Contributed reagents, materials, analysis tools or data.

Dingfang Cai: Conceived and designed the experiments.

Funding statement

This work was supported by Si Ming Clinical Investigation [SGKJLC-202024], National Natural Science Foundation of China [81703857 and 81903970], The Special Fund for Standardized Training of Resident Physicians in Zhongshan Hospital [008].

Data availability statement

Data will be made available on request.

Declaration of interest’s statement

The authors declare no conflict of interest.

Additional information

No additional information is available for this paper.

Appendix A. Supplementary data

The following is the supplementary data related to this article: