Tanshinone IIA as a therapy for PCOS via FOS/JUN/TP53 axis: Evidence from network pharmacology of Bajitian-Danshen pair

2.1 Reagents

Salvia miltiorrhiza (Danshen) granules and Eclipta prostrata (Bajitian) granules were obtained from Jiangyin Tianjiang Pharmaceutical Co., Ltd. Tan-IIA (MR 294.34, >99 % purity) was obtained from Shanghai YuanYe Biotechnology Corporation (Shanghai, China). Human insulin solution (cat. no. I9278) as well as glucose and sucrose assay kit (cat. no. MAK013) were from Sigma-Aldrich (St Louis, MO, USA)). Human chorionic gonadotropin (HCG; cat. no. NBP25954510) was from Novus Biologicals. ELISA kits for follicle-stimulating hormone (FSH) (cat. no. ab108641), LH (cat. no. ab108651), tumor necrosis factor (TNF)-α (cat. no. ab181421), testosterone (cat. no. ab178663) and the glucose oxidase assay kit (cat. no. ab138884) were obtained from Abcam. Pregnant mare serum gonadotrophin (PMSG) was obtained from Ningbo Sansheng Pharmaceutical Co., Ltd. (Ningbo, Zhejiang Province, China), while Bovine serum albumin (BSA; cat. no. SW3015) was from Solarbio Co., Ltd. (Beijing, China). The Dako Real™ Envision™ Detection System was commercially obtained from Dako (cat. no. K500711; Dako; Agilent Technologies, Inc., Santa Clara, CA, USA). Lipofectamine® RNAiMAX transfection reagent (cat. no. 13778150) was commercially acquired from Invitrogen (Shanghai, China). Phorbol 12-myristate 13-acetate (PMA) was commercially provided by Sigma (St. Louis, MO, USA).

2.2 Construction of a database of main active ingredients

The active ingredients of the Bajitian-Danshen pair were identified through the traditional Chinese medicine system pharmacology database and analysis platform (TCMSP, https://lsp.nwu.edu.cn/tcmsp.php). The main active ingredients of the Bajitian-Danshen pair were then selected according to the optimal toxicologic ADME rules reported in the literature (oral bioavailability (OB) ≥ 30 %; drug-like property (DL) ≥ 0.18).

2.3 Potential targets of drug and disease and target classification

All targets of Bajitian-Danshen active ingredients were extracted from the TCMSP database. All the names of targets were converted into gene symbols by using the Uniprot database (https://www.uniprot.org). The keyword “Polycystic Ovary Syndrome” was input into the OMIM database (https://omim.org/), GeneCards database (http://www.genecards.org/), PharmGkb database (https://www.pharmgkb.org/), DrugBank database (https://www.drugbank.ca/) and DisGeNET database (https://www.disgenet.org/) to find genes involved in PCOS pathogenesis. After that, the search results of the five databases were gathered, aggregated, and de-duplicated to obtain the PCOS-related gene list. The Venn diagrams were generated by using the online tool from the domain https://bioinformatics.psb.ugent.be/webtools/Venn/ to get the potential targets of Bajitian-Danshen pair in PCOS.

2.4 Compound-gene network and protein–protein interaction (PPI) network construction

The TCM compound regulatory network was constructed by using Cytoscape 3.8.0 (https://cytoscape.org/) to obtain the target relationship between the active ingredients and target genes. The protein–protein interaction (PPI) network was generated in STRING (https://string-db.org/). We selected “multiple proteins” in the database, inputted the list of potential targets of Bajitian and Danshen in PCOS with the species limited to “Homo sapiens” and confidence scores ≥ 0.4, and exported the PPI data. We analyzed the topology of the PPI network by using the CytoNCA plug-in in Cytoscape 3.8.0. We selected gene to carry out topology analysis on three parameters of degree centrality (DC), betweenness centrality (BC), and closeness centrality (CC) with the filter criteria limited to the top of 50 % according to the centrality of nodes to evaluate the importance of target genes.

2.5 Functional analysis of target genes

Functional enrichment analysis of targets and the visualization of enrichment terms were performed using the TCGABiolinks package based on the TCGAanalyze_EAcomplete and TCGAvisualize_EAbarplot functions. The terms with enrichment significance p value lower than 0.05 were considered significantly enriched.

2.6 Molecular docking

Based on the above findings, the active ingredient with the most significant number of targets was selected as the ligand, and its targets with the largest node degree in the PPI network was chosen as the receptor to perform molecular docking. We downloaded three-dimensional structures of the active ingredients from PubChem CID. We obtained protein structures of the targets using the PDB database and imported them into AutoDockVina and MGLtools to perform molecular structure processing and molecular docking.

2.7 Animals

This study included ten-week-old female Sprague-Dawley rats weighing 175.7–221.4 g (n = 50), which were bought from Shanghai Laboratory Animal Center, Co. Ltd (Shanghai, China). The animals were kept in a 25 ± 2 °C environment with 56 ± 13 % humidity in standard cages and a dark/light cycle of 12 h. Food and Tap water were accessible to rats ad libitum. The acclimation time for the rats in the environment before subsequent experiments was one week. Approval for animal experiments was obtained from the ethics commission of Shanghai Traditional Chinese Medicine Hospital (Shanghai, China) (approval no. 20190103). The experimental protocols agreed with the Chinese Ministry of Science and Technology standards for the Care and Use of Laboratory Animals.

2.8 Establishment PCOS rat model and drug treatments

For investigating the effect of Bajitian-Danshen, fifty rats were divided into five groups, comprising the control, PCOS, PCOS + Baj-Dan Low, PCOS + Baj-Dan Medium, and PCOS + Baj-Dan High. The low-dose group received a range of 0.25 g/kg/day, the medium-dose group received a range of 0.5 g/kg/day, while the high-dose group received a range of 1 g/kg/day.

For investigating the effect of Tan-II, seventy rats were randomly distributed into seven groups of 10 rats in each group: 1) Control group containing rats intragastrically administered with saline for 21 days; 2) PCOS group, composed of PCOS rats intragastrically administered with saline for 21 days; 3) PCOS + PMA group, composed of PCOS rats that received an intragastric gavage of saline and an intraperitoneal administration of 500 µg/kg PMA daily for 21 days; 4) PCOS + Tan-IIA Low group, composed of PCOS rats that received an intragastrical administration of 25 mg/kg/day Tan-IIA for 21 days; 5) PCOS + Tan-IIA Medium group, composed of PCOS rats that received an intragastrical administration of 50 mg/kg/day Tan-IIA for 21 days; 6) PCOS + Tan-IIA High group, composed of PCOS rats that received an intragastrical administration of 100 mg/kg/day Tan-IIA for 21 days; and 7) PCOS + PMA + Tan-IIA High group, containing PCOS rats intraperitoneally receiving 500 µg/kg PMA/day and intragastrically administered with 100 mg/kg/day Tan-IIA for 21 days. Tan-IIA was dissolved in normal saline.

For PCOS rat model establishment, 60 rats were subjected to daily administration of HCG (3.0 IU/day) and two doses of INS subcutaneously for a period of 22 days. From the 1st to the 11th day, the dose of INS was gradually increased from 0.5 to 3.0 U/day; the dose of INS was kept at 3.0 U/day from day 12 to day 22. The 10 rats in the control group were subcutaneously administered with the same volume of normal saline two times/day for 22 days. The first days of the experiments, 5 % dextrose solution was given as drinking instead of the conventional drinking water. The weight of rats taken each week while vaginal smear was, monitored each day until the 23rd day. The establishment of the PCOS model was confirmed by the loss of rat normal estrous cycle.

2.9 Determination of body mass index (BMI), Lee's index, and ovarian quotiety

At the end of the experiment, the length of rats from the anus to the nose was detected and used for calculating BMI, Lee's index and ovarian quotiety using the following formulas: ${\mathrm{L}\mathrm{e}\mathrm{e}}^{\prime }\mathrm{s}\mathrm{i}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{x}=\frac{\sqrt[3]{BodyWei\mathrm{g}\mathrm{t}\mathrm{h}}}{\mathrm{L}ength}\times 1000$ (Gu et al., 2020); $\mathrm{B}\mathrm{M}\mathrm{I}=\frac{\mathit{Weight}}{{\mathit{Length}}^2}$ ; $\mathrm{O}\mathrm{v}\mathrm{a}\mathrm{r}\mathrm{i}\mathrm{e}\mathrm{s}\mathrm{Q}\mathrm{u}\mathrm{o}\mathrm{t}\mathrm{i}\mathrm{e}\mathrm{t}\mathrm{y}=\frac{\mathit{OvarianWeigth}}{\mathit{BodyWeigth}(mg/100g)}$ (Yu et al., 2014). The unit of length was cm and that of the weight was g.

2.10 Specimen collection

After being fasted overnight the last day of the experiment, the rats were subjected to sodium pentobarbitone (55 mg/kg) injection, intraperitoneally. Next, the abdominal aorta blood (5 ml) was collected and centrifuged (1,500 × g for 20 min at 4 °C) to obtain the serum. For the collection of ovarian tissue, the rats were killed by injection of high dose pentobarbital (250 mg/kg). The death was confirmed by exsanguination and the ovarian tissue was collected after dissection. For gene expression analysis, immediate immersion of the ovaries in liquid nitrogen was conducted to guarantee optimal preservation. For other experiments, the collected ovaries were immediately weighted and conserved in a − 80 °C fridge samples; before each experiment, the tissue was fixed in 4 % paraformaldehyde at 25 °C for 24 h.

2.11 ELISA detection of hormones and TNF-α levels

ELISA kits were used for detecting the levels of LH (Invitrogen, Shanghai, China), testosterone (Abcam, #ab108666), FSH (Invitrogen, Shanghai, China) and TNF-α (Abcam, #ab181421) in the serum following the manuals provided by the manufacturers. The levels were calculated based on the absorbance at 450 nm that was detected spectrophotometrically.

2.12 Histomorphometry analysis

The ovarian tissues fixed as described above were treated with gradient concentrations of ethanol (50, 70, 80, 90, and 100 %) for 30 sec and subsequently washed with xylene. The 4-µm thick sections were stained with picric acid for 30 min at room temperature and then with hematoxylin-eosin for 15 min at room temperature. Histomorphometry and histological examination of the tissues were performed using a microscope (Leica DM5000 fluorescence, Leica Microsystems, Inc.).

2.13 TUNEL assay

After deparaffinization with Histo-Clear, sections (6 μm thickness) were permeabilized in 0.3 % Triton X-100. Apoptosis was then detected in the tissues using a specific TUNEL antibody, and the DeadEnd™Colorimetric TUNEL System (Promega, Madison, USA) was used for detection according to the manufacturer's recommendations. Olympus IX81 microscope (Olympus America Inc., Center Valley, PA) was used for microscopic analysis, and green fluorescence was used to identify TUNEL-positive cells; diamidino-2-phenylindole (DAPI) counterstaining was used to identify cell nuclei.

2.14 Immunofluorescence staining

Deparaffinized ovarian sections (6 μm thickness) were mounted on coated glass slides and then washed with PBS (pH 7.4), followed by two rinses with PBS containing 0.5 % TritonX-100. After blocking in IgG-free 2 % bovine serum albumin (BSA sigma) for 30 min, the slides were incubated overnight at 4 °C with anti-p53 antibody [pAb122] (ABCAM, catalog number: ab90363, dilution 1:500) and anti-c-Fos antibody [2H2] (ABCAM, catalog number: ab208942, dilution 1:500). The following day, the slides were incubated with a secondary antibody (goat anti-rabbit FITC-labeled) (Vector Labs., CA, USA, FI −1000). Slides were counterstained with Vectashield containing DAPI (VectorLabs., CA, USA, H-1200). Sections of ovaries were observed and imaged using a digital camera (Nikon DS -U2, Japan) in conjunction with an Olympus BX51 fluorescence microscope.

2.15 Isolation and culture of granulosa cells (GCs) from rat ovaries

To isolate GCs, six immature (19–24 days old) female rats were given a subcutaneous injection of PMSG (150 IU/kg), and their ovaries were harvested after 48 h. Ovarian GCs were then isolated by follicular puncture (23) and pooled. After filtering the cells and centrifugation at 200 × g at 4 °C for 5 min, the supernatant was removed. GCs were then used to prepare a cell suspension with Dulbecco's modified Eagles medium F12 (DMEM/F12 medium, Thermo Fisher Scientific, Inc.) containing 0.05 mg/ml gentamicin, selenium, transferrin, and INS. The isolated GCs were seeded into a 6-well cell culture plate (5 × 10⁴ cells/ml) in DMEM/F12 medium and incubated in a 5 % CO₂ atmosphere at 37 °C for 24 h in a humidified atmosphere.

2.16 Cell transfection

FOS small interfering (si) RNA was used to knock down FOS expression in GCs. AllStars scrambled siRNA were employed as negative controls (NC-siRNA). The siRNA sequences were as follows: si-FOS#1, 5′- GACCTAGGGAGGACCTTATCT-3′; si-FOS#2, 5′-GGCAAAGTAGAGCAGCTATCT-3′; and si-FOS#3, 5′- GACCTGTCTGGTTCCTTCTAT-3′. Lipofectamine® RNAiMAX (0.5 µl, 50 nM; Thermo Fisher Scientific, Inc.) was combined with siRNA (0.5 µl) for 20 min prior to experiments. GCs were cultured into a 6-well cell culture plate (3–5 × 104 cells/ml) until 90–95 % confluence; GCs were then assigned to the following groups: (i) normal, non-transfected cells; (ii) si-NC, transfected with NC-siRNA; (iii) si-FOS#1, transfected FOS siRNA-1; (iv) si-FOS#2, transfected with FOS siRNA-2; and (v) si-FOS#3, transfected with FOS siRNA-3. The next day, transfection efficiency was assessed by RT-qPCR and Western blotting.

2.17 Establishment of insulin resistance (IR) model of GCs and treatments

GCs were seeded in DMEM/F12 medium as described above. IR GCs were obtained by treating GCs with 1 µmol/L dexamethasone for three days. The culture medium was harvested to measure glucose levels using a commercial glucose and sucrose assay kit. GCs were allocated to five groups: i) Normal group, GCs without treatment; ii) IR group, INS-resistant GCs; iii) IR + NC group, IR-treated GCs that were transfected with si-NC; iv) IR + si-FOS, IR-treated GCs that were transfected with si-FOS; and v) IR + Tan-IIA group, IR-treated GCs that were subjected to 200 nM Tan-IIA treatment.

2.18 Viability assay

After treatment, the viability of the cells in each group was assessed by performing the MTT assay. Cells (1 × 10⁴ cells/well) were cultivated in 96-well plates for 48 h and subsequently added with 0.2 mg/ml MTT reagent (Sigma-Aldrich; Merck KGaA) for an additional 4 h incubation at 37 °C in 5 % CO₂. Later, formazan crystals were dissolved for 20 min using DMSO. Cell viability was gaged after detecting the absorbance at 490 nm using a microplate reader (Tecan Group, Ltd.).

2.19 RNA isolation and RT-qPCR

Total RNA from the GCs and rat ovary homogenates was isolated using the TRIzol® reagent and used for cDNA synthesis with GoScript™ Reverse Transcription System (Promega Corporation) for 60 min at 42 °C based on the protocol described by the manufacturer. SYBR-Green PCR master mix (Thermo Fisher Scientific, Inc.) was used for RT-qPCR amplification. Primer sequences were: PTGS2 forward: 5′- ACACTCTATCACTGGCATCC-3′, reverse 5′-GAAGGGACACCCTTTCACAT-3′; GADPH forward: 5′- ACGGCAAGTTCAACGGCACAG-3′, reverse 5′- GAAGACGCCAGTAGACTCCACGAC-3′; c-Fos forward: 5′- AGCATGGGCTCCCCTGTCA-3′, reverse 5′- GAGACCAGAGTGGGCTGCA-3′; TP53 forward: 5′-CCACCTTCTTTGTCCTGCCTG-3′, reverse 5′- GTCGGCTCCGACTATACCACTATC-3′; c-Jun forward: 5′- ACCTTCAACACCCCAGCCATG-3′, reverse 5′- GGCCATCTCTTGCTCGAAGTC −3′; MMP9 forward: 5′- AGGTGCCTCGGATGGTTATCG-3′, reverse 5′-TGCTTGCCCAGGAAGACGAA-3′; CASP3 forward: 5′-GGTATTGAGACAGACAGTGG-3′, reverse 5′- CATGGGATCTGTTTCTTTGC-3′. The thermal cycling conditions were activation at 50 °C, denaturation at 95 °C for 30 sec, 40 cycles of denaturation at 95 °C for 5 sec, and annealing at 60 °C for 30 sec. The 2^−ΔΔCq method was applied to quantify mRNA expression levels relative to GAPDH.

2.20 Western blot analysis

Cells and tissue homogenates were treated for 30 min with RIPA (cat. no. ab7937; Abcam). Next, after centrifugation (15 min, 11,500 × g, 4 °C), the BCA Protein Assay kit (cat. no. 23227; Thermo Fisher Scientific, Inc.) was employed for detecting total protein concentration. Protein (50 µg) was loaded for 12 % SDS-PAGE electrophoresis and transferred onto PVDF membranes. Afterwards, following blocking in 5 % at 25 °C for 2 h, the membranes were treated with primary antibodies against GAPDH (1:5,000; cat. no. ab8245; Abcam), anti-c-Fos antibody (1:1,000; cat. no. ab302667; Abcam), anti-PTGS2 antibody (1:1,000; cat. no. ab151571; Abcam), anti-c-Jun antibody (1:1,000; cat. no. ab40766; Abcam) and anti-p53 antibody (1:1,000; cat. no. ab90363; Abcam) overnight at 4 °C. Next, after washing with Tris-buffered saline added with 0.05 % Tween-20 (TBST), the membranes were treated for 1 h at 25 °C with the HRP-conjugated secondary antibody (1:5,000; cat. no. ab7097; Abcam). The ECL reagent (cat. no. WBKLS0050; EMD Millipore) was used for visualization, followed by densitometry analysis using the ImageJ software (version 1.47, National Institutes of Health).

2.21 Statistical analysis

One-way ANOVA followed by Bonferroni's post hoc test was used for detecting the statistical significance among three or more groups, while the T-test was used for analyzing the comparative difference among two groups using GraphPad Prism software version 8.0.1 (GraphPad Software, Inc.). P < 0.05 was considered for significant difference among groups.

3.1 Identification of hub bioactive ingredients of Bajitian-Danshen pair and their potential target genes in PCOS treatment

A total of 83 active ingredients of Bajitian (18 ingredients) and Danshen (65 ingredients) were retrieved following the ADME screening criteria (OB > 30 %, DL > 0.18) (Additional File S1). The targets of each active ingredient were predicted in the TCMSP database, and 95 targets of the Bajitian-Danshen pair were obtained. After combining the lists of genes from GeneCards, OMIM, PharmGkb, DrugBank, and DisGeNET databases, a total of 2,205 PCOS-related genes were predicted. The Venn diagram intersection of the PCOS-related genes and the target genes of Bajitian-Danshen allowed the identification of 58 common genes as the targets of Bajitian and Danshen combination in the treatment of PCOS (Fig. 1A). Based on the list of ingredients targeting these 58 common genes, we identified 16 Bajitian ingredients and 44 Danshen ingredients as potentially effective ingredients against PCOS (Additional File S2). The PPI network of the 58 target genes of the Bajitian-Danshen pair in PCOS treatment is depicted in Fig. 1B. The network comprised 57 connected nodes and 411 edges (Fig. 1B). The average number of neighbors was 14.421, with a diameter of 4 and a radius of 3 (Fig. 1B). MCODE analysis indicated the hub cluster of this PPI was composed of 21 genes (TP53, CASP3, PTGS2, MMP9, JUN, FOS, HSP90AA1, NFKBIA, CDKN1A, ESR1, PPARG, NOS3, GSK3B, CASP9, CDK2, TGFB1, PRKCA, EDN1, NOS2, AR, and PGR) as core genes in the PPI network (Fig. 1B). The functional enrichment analysis indicated that the target genes of the Bajitian-Danshen combination in PCOS were enriched in the biological processes of positive regulation of biosynthetic process (n = 13), positive regulation of nitrogen compound metabolic process (n = 11), positive regulation of apoptosis (n = 9), and positive regulation of programmed cell death (n = 9) (Fig. 1C). The most enriched cellular component terms were cell fraction (n = 10), neuron projection (n = 6), cell projection (n = 6), cell soma (n = 4), dendrite (n = 4), and integral to plasma membrane (n = 9), while the most enriched terms in the category of molecular function were adrenoceptor activity (n = 4), amine receptor activity (n = 5), alpha1 − adrenergic receptor activity (n = 3), and alpha − adrenergic receptor activity (n = 3) (Fig. 1C). The most important pathways potentially targeted by the Bajitian-Danshen ingredients in PCOS were Glucocorticoid Receptor Signaling (n = 19), Aryl Hydrocarbon Receptor Signaling (n = 14), molecular mechanisms of cancer (n = 15), Prostate Cancer Signaling (n = 10), IL-12 signaling and production of macrophages (11), and HIF1_Signaling (n = 10) (Fig. 1C). The ingredient-target network of the Bajitian-Danshen combination in PCOS treatment was generated and visualized using the CytoScape software (Fig. 2). The ingredient-target PPI network indicated Tan-IIA as the hub component of the Bajitian-Danshen combination in PCOS treatment (MOL007154, degree 27, betweenness centrality 0.2070204172530126, closeness centrality 0.45703124999999994) followed by beta-sitosterol (MOL000358, degree 24, betweenness centrality 0.15759452247335504, closeness centrality 0.46062992125984253), dihydrotanshinlactone (MOL007100, degree 24, betweenness centrality 0.04483259476537794, closeness centrality 0.45348837209302323), 1,2,5,6-tetrahydrotanshinone (MOL001601, degree 19, betweenness centrality 0.04779521728839308, closeness centrality 0.43333333333333335) that were also active in treating PCOS (Fig. 2). The node table indicating the network parameters was depicted in Additional File S3.

Close

Fig. 1

Identification of active ingredients of Bajitian and Danshen in PCOS treatment and the functional roles of their targets. A- Venn diagram indicating the intersection of PCOS-related genes and targets of Bajitian and Danshen. B- Protein-protein interaction network of targets of Bajitian and Danshen ingredients in PCOS treatment. C- Functional enrichment of targets of Bajitian and Danshen ingredients in PCOS treatment.

Export to PPT

Close

Fig. 2

Ingredient-gene interaction network. The ingredients of Bajitian and Danshen and their target genes were used for constructing the drug-gene interaction network.

Export to PPT

3.2 Protein-protein interaction network of Tan-IIA targets in PCOS and their functional enrichment analysis

Since Tan-IIA was found to be the ingredient with the highest number of targets in the treatment of PCOS, the 27 targets of this ingredient were used for protein–protein interaction network construction in the STRING database (Fig. 3A). The PPI network comprised 25 nodes and 78 edges with an average number of neighbors of 6.240 (Fig. 3A). The PPI network was imported into Cytoscape to identify the core targets and possible protein functional modules. Important targets of Tan-IIA were TP53, FOS, PTGS2, CASP3, MMP9, EDN1, NFKBIA, and JUN (Fig. 3A). These 27 Tan-IIA target genes were involved in the biological processes of positive regulation of the biosynthetic process (n = 8), drug metabolic process (n = 3), and homeostatic process (n = 6) (Fig. 3B). The prevalent cellular component terms were cell fraction (n = 7), microsome (n = 3), and vesicular fraction (n = 3) (Fig. 3B). The most enriched molecular function terms were vitamin D 24-hydroxylase activity (n = 2), oxidoreductase activity acting on paired donors (n = 5), peptide binding (n = 4), and oxygen binding (n = 3) (Fig. 3B). The most enriched target pathways of Tan-IIA target genes in PCOS were Aryl Hydrocarbon Receptor Signaling (n = 7), Glucocorticoid Receptor Signaling (n = 8), PPAR Signaling (n = 6), MIF Regulation of Innate Immunity (n = 5), Small Cell Lung Cancer Signaling (n = 5), inhibition of angiogenesis by TSP1 (n = 4), ILK Signaling (n = 6), IL-8 Signaling (n = 6), HIF1_signaling (n = 5), and molecular mechanisms of cancer (n = 7) (Fig. 3B).

Close

Fig. 3

Tan-IIA targets in PCOS treatment. A-Protein-Protein interaction network of Tan-IIA targets in PCOS treatment. Functional enrichment analysis of Tan-IIA targets in PCOS treatment. Tan-IIA = tanshinone IIA.

Export to PPT

3.3 Molecular docking analysis

To find the critical genes that Tan-IIA could target, we performed subsequent clustering of the PPI network of Tan-IIA targets to identify the different groups of interactive proteins using the clustering option in the STRING database. We found a set of 10 genes (CASP3, BCL2, DPP4, MMP9, TP53, PTGS2, CDKN1A, NFKBIA, FOS, and JUN) as credible targets of Tan-IIA targets in PCOS treatment (Fig. 4A). As shown in Fig. 4A, JUN and FOS were identified to pertain to the first cluster, CASP3, DPP4, BCL2, TP53, PTGS2, CDKN1A, and NFKBIA belonged to the second cluster, while MMP9 belonged to the third cluster. Thus, we subsequently verified the expression trends of CASP3, MMP9, TP53, PTGS2, FOS, and JUN in IR GCs and PCOS rat ovaries by RT-qPCR. As indicated in Fig. 4B and 4C, the expression levels of these genes were increased in IR GCs and PCOS rat ovaries, which indicated their roles in PCOS and the potential of Tan-IIA to revert their deleterious outcomes in PCOS.

Close

Fig. 4

Identification of key Tan-IIA targets in PCOS treatment. A- Hub clusters of key Tan-IIA targets in PCOS treatment obtained from the protein–protein interaction network. B- RT-qPCR validation of the mRNA expression of key Tan-IIA targets in PCOS treatment in IR GCs. C- RT-qPCR validation of the mRNA expression of key Tan-IIA targets in PCOS treatment in ovaries of PCOS rats. D- Molecular structure of Tan-IIA. D- Tridimensional structure of FOS. F-Docking of Tan-IIA and FOS. G- Binding affinity of Tan-IIA to FOS. Data are presented as the mean ± SEM. **P < 0.01, ***P < 0.001, ****P < 0.0001 vs. Control, Tan-IIA = tanshinone IIA, IR = Insulin-resistant, GCs = granulosa cells.

Export to PPT

Furthermore, we performed molecular docking and found that Tan-IIA (Fig. 4D) docked well to FOS (Fig. 4E) through several interactions with amino acid residues, as shown in Fig. 4F. The binding affinity between both molecules was −7.6 (Fig. 4G), indicating FOS is the most crucial target of Tan-IIA in PCOS.

3.4 Bajitian-danshen decoction counteracts PCOS and regulated predicted hub target genes

To explore the impact of the Bajitian-Danshen decoction on PCOS, PCOS rats were treated with low (PCOS + Baj-Dan Low), medium (PCOS + Baj-Dan Medium), and high (PCOS + Baj-Dan High) doses of the decoction. The results indicated that the Bajitian-Danshen decoction dose-dependently decreased PCOS-induced body weight gain (Fig. 5A), ovarian weight (Fig. 5B), and ovarian quotiety (Fig. 5C). Additionally, the decoction significantly decreased PCOS-induced increase in BMI (Fig. 5D), Lee’s index (Fig. 5E), LH (Fig. 5F), FSH (Fig. 5G), testosterone (Fig. 5H), and TNF-α (Fig. 5I). Furthermore, the results of HE staining showed that the control group had normal ovarian tissue morphology with various stages of follicles and multiple corpora lutea visible (Fig. 5J). On the other hand, the PCOS group exhibited numerous cystic expanded follicles and a significant decrease in the number of granulosa cells (Fig. 5J). After administering the Bajitian-Danshen decoction at different concentrations, there was a notable improvement in ovarian tissue morphology, with the high dose showing the most pronounced effects (Fig. 5J). The findings suggest that the herbal formulation can improve ovarian morphology, increase the number of ovarian granulosa cells, reduce the proliferation of follicle membrane cells and interstitial cells, and promote follicular development and ovulation (Fig. 5J). Moreover, we examined the effect of the Bajitian-Danshen decoction on the gene expression of the predicted hub targets, and the results indicated that the decoction dose-dependently counteracted PCOS-induced upregulations of CASP3, MMP9, TP53, PTGS2, FOS, and JUN (Fig. 5K). These results indicated that the Bajitian-Danshen decoction is effective in treating PCOS, and its effect might be driven by regulating the predicted targets.

Close

Fig. 5

Effects of Bajitian-Danshen decoction on (A) body weight, (B) weight of ovaries, (C) ovarian index, (D) BMI, (E) Lee’s index, (F) LH, (G) FSH, (H) testosterone, (I) TNF-α, (J) Ovary histopathological changes in rats with PCOS, and (K) the expression of CASP3, MMP9, TP53, PTGS2, FOS, and JUN. Data are presented as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 among the compared groups, ns = non-significant. Scale bar = 100 μm. Baj-Dan = Bajitian-Danshen.

Export to PPT

3.5 The therapy activity of Tan-IIA on PCOS might be driven via regulating FOS

To investigate the effect of Tan-IIA on the expression of FOS, the immunofluorescence assay was performed on the ovary tissue of rats in the different groups (Fig. 6). The positive ratio of FOS in the PCOS and PCOS + PMA groups was significantly increased compared to the control group (Fig. 6). Nevertheless, after Tan-IIA treatment, the positive ratio of FOS was significantly and dose-dependently decreased in the PCOS rats (Fig. 6). In addition, Tan-IIA significantly abolished the effect of PMA on the FOS-positive ratio in the PCOS rats (Fig. 6). These results indicated that the therapy activity of Tan-IIA on PCOS might be driven via regulating FOS.

Close

Fig. 6

Effect of Tan-IIA on FOS expression in PCOS rats. Immunofluorescence analysis of FOS in the ovary tissues of rats subjected to different treatments. Data are presented as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 among the compared groups, ns = non-significant, Tan-IIA = tanshinone IIA. Scale bar = 50 μm.

Export to PPT

3.6 Tan-IIA counteracts deleterious effects in rats with PCOS via FOS/JUN/TP53 signaling

In order to experimentally validate the therapeutic effect of Tan-IIA on PCOS and its potential regulation of FOS, a rat model of PCOS was established and treated with PMA, Tan-IIA at different doses, or the combined treatment of PMA and Tan-IIA. Compared to control rats, the weight of rats with PCOS was significantly increased (Fig. 7A). The body weights of rats in PCOS + Tan-IIA were dose-dependently decreased compared to the PCOS group (Fig. 7A). Moreover, treatment with PMA further increased the body weights of PCOS rats, but this effect was abolished by the treatment with a high dose of Tan-IIA (Fig. 7A). The weights of the ovaries of rats in the PCOS rats were significantly increased compared to control rats, and this effect was further increased by the treatment with PMA (Fig. 7B). On the contrary, the treatment of PCOS rats with Tan-IIA significantly decreased the ovarian weight in a dose-dependent manner (Fig. 7B). Moreover, the ovarian weights of rats were notably reduced in the PCOS + PMA + Tan-IIA High group compared to the PCOS + PMA group rats (Fig. 7B). Additionally, the quotiety of ovaries in the PCOS was remarkably increased comparatively to the control group, which was further promoted in the PCOS + PMA groups (Fig. 7C). However, compared to the PCOS groups, Tan-IIA decreased ovarian quotiety dose-dependently (Fig. 7C). In addition, as shown in Fig. 7D and 7E, BMI and Lee’s index were all increased in PCOS and this effect was further promoted by PMA treatment. Nonetheless, Tan-IIA dose-dependently decreased BMI and Lee’s index, and abrogated the effect of PMA on these variables (Fig. 7D and 7E). The effect of Tan-IIA on the serum levels of TNF-α, testosterone, LH, and FSH was investigated. The results indicated that the serum levels of TNF-α, testosterone, LH, and FSH in the PCOS group were remarkably increased compared to the control group (Fig. 7F-7I). Furthermore, TNF-α, LH, FSH, and testosterone levels in the PCOS rats treated with different doses of Tan-IIA were significantly decreased compared to the PCOS group (Fig. 7F-7I). Moreover, in the PCOS + PMA + Tan-IIA group, FSH, LH, TNF-α and testosterone levels were obviously diminished compared to the PCOS + PMA group (Fig. 7F-7I). To investigate whether the action of Tan-IIA on the histopathological changes of PCOS rats, the HE staining of ovary tissue from rats in different groups was performed (Fig. 7J). The microscopic images indicated that the follicles and corpora lutea in the ovaries of rats in the control group were in primordial follicles, primary follicles, and secondary follicles developmental stages (Fig. 7J). Serious histopathological changes were recorded in the PCOS and PCOS + PMA rats (Fig. 7J). Moreover, compared to the control rats, the number of large cystic follicles filled with fluid substantially increased while the GC layers deteriorated in the PCOS and PCOS + PMA groups (Fig. 7J). Moreover, a decreased number of corpora lutea was observed in both the PCOS and PCOS + PMA rats compared to the control rats (Fig. 7J). Treatment of PCOS rats with Tan-IIA at different doses obviously alleviated the pathological changes observed in the PCOS rats. In the PCOS + PMA + Tan-IIA treatment group, the pathological changes in ovarian tissue observed in the PCOS and PCOS + PMA rats were abrogated, with an increased number of GC layers and a decreased number of cystic follicles (Fig. 7J). These results indicated that Tan-IIA counteracts the pathological changes of ovaries and might regulate deleterious effects of PCOS, especially inflammation and hormonal disorders in PCOS, via regulating FOS.

Close

Fig. 7

Effects of Tan-IIA and PMA on (A) body weight, (B) weight of ovaries, (C) quotiety of ovaries, (D) BMI, (E) Lee’s index, (F) LH, (G) FSH, (H) Testosterone, (I) TNF-α, and (J) Ovary histopathological changes in rats with PCOS. Data are presented as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 among the compared groups, ns = non-significant, Tan-IIA = tanshinone IIA. Scale bar = 100 μm.

Export to PPT

3.7 Effect of Tan-IIA on cell apoptosis in vivo

We sought to verify the effect of Tan-IIA on cell apoptosis in vivo. The Tunel assay was performed on ovarian tissue sections. The results displayed that, compared to the control group, the level of apoptosis in the PCOS rats was significantly increased (Fig. 8). In addition, treatment with PMA promoted this effect (Fig. 8). Nevertheless, treatments of PCOS rats with various doses of Tan-IIA led to a significantly reduced level of apoptosis in the ovarian tissue compared to the PCOS group (Fig. 8). Furthermore, a high dose of Tan-IIA abolished the effect of PMA, as indicated in the PCOS + PMA group.

Close

Fig. 8

Effect of Tan-IIA on apoptosis in PCOS rats. Tunel immunostaining analysis of FOS in the ovary tissues of rats subjected to different treatments. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 among the compared groups. ns = non-significant, Tan-IIA = tanshinone IIA. Scale bar = 50 μm.

Export to PPT

3.8 Tan-IIA regulate the expression of TP53, PTGS2 and JUN in the ovarian tissues via FOS

The TP53, PTGS2 and JUN expression levels were detected by RT-qPCR (Fig. 9A) and western blotting (Fig. 9B). The results indicated that the TP53, PTGS2 and JUN expression levels in the PCOS and the PCOS + PMA groups were notably increased compared to the normal group (Fig. 9A and B). Furthermore, the TP53, PTGS2 and JUN expression in the PCOS + PMA + Tan-IIA High group was notably decreased comparatively to the PCOS + PMA group (Fig. 9A and B). Furthermore, immunofluorescence staining of TP53 indicated that the TP53-positive ratio was increased in the PCOS group (Fig. 9C). PMA treatment of PCOS rats further promoted the TP53-positive ratio (Fig. 9C). The positive ratio of TP53 was decreased by high-dose Tan-IIA treatment in comparison with the PCOS and PCOS + PMA groups (Fig. 9C). These results indicated that Tan-IIA regulates the expression level of TP53, PTGS2, and JUN via targeting FOS.

Close

Fig. 9

Effect of Tan-IIA on the FOS/JUN/TP53 signaling pathway. A- RT-qPCR detection of FOS, JUN, and TP53 in different treatment groups. B- Western blot detection of FOS, JUN, and TP53 in different treatment groups. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 among the compared groups. ns = non-significant, Tan-IIA = tanshinone IIA. Scale bar = 100 μm.

Export to PPT

3.9 Tan-IIA inhibits the proliferation and regulates FOS, TP53, PTGS2, and JUN in IR-treated GCs

To investigate the effect of Tan-IIA on GCs, these cells were transfected with three si-RNAs targeting FOS, and FOS mRNA and protein expression levels were analyzed by RT-qPCR and western blotting. Relatively to si-NC and untransfected cells, protein expression of FOS in the si-FOS#1, si-FOS#2, and si-FOS#3 groups was notably reduced, particularly in the si-FOS#2 group (Fig. 10A and B). Therefore, the FOS siRNA-2 was transfected in the GCs, and cell viability in different groups was determined by MTT assay (Fig. 10C). Cell viability in the IR and IR + si-NC groups was notably reduced compared to the normal group (Fig. 10C). The viability of IR-GCs treated with Tan-IIA or transfected with si-FOS was higher relative to untreated cells (Fig. 10C). In addition, TNF-α (Fig. 10D) and FOS (Fig. 10E) levels in the serum of IR and IR + si-NC rats were notably increased relative to the normal group. In addition, relative to the IR group, TNF-α and FOS levels in the IR + si-FOS and IR + Tan-IIA groups were notably reduced (Fig. 10D and 10E). These results suggested that Tan-IIA improved the viability of IR-GCs by regulating the production of FOS.

Close

Fig. 10

Tan-IIA regulates the viability of IR GCs via the FOS/JUN/TP53 axis. A- Western blotting confirmation of FOS silencing in GCs. B- Densitometry analysis of western blot band from the confirmation experiments. C- Effect of Tan-IIA on the viability of IR GCs. D- ELISA analysis of TNF-α. E- ELISA analysis of FOS. F- RT-qPCR analysis of the expression FOS, JUN and TP53 in different treatment groups. G- Western blot analysis of the expression FOS, JUN and TP53 in different treatment groups. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 among the compared groups or compared to control group, #P < 0.05 compared to IR, $P < 0.05 compared to IR + si-FOS, ns = non-significant, Tan-IIA = tanshinone IIA, IR = insulin-resistant, GCs = granulosa cells.

Export to PPT

TP53, PTGS2, and JUN expression levels were analyzed using RT-qPCR (Fig. 10F) and western blotting (Fig. 10G). The expression of TP53, PTGS2, and JUN in the IR and IR + si-NC groups were notably augmented relative to the normal group (Fig. 10F and G). Furthermore, TP53, PTGS2 and JUN expression levels in both IR + si-FOS and IR + Tan-IIA groups were notably lower relatively to the IR group (Fig. 10F and G). The effect of Tan-IIA treatment on FOS, TP53, PTGS2, and JUN expression levels was also notably greater relative to si-FOS transfection (Fig. 10F and G). These results indicated that Tan-IIA treatment downregulated the expression of FOS, TP53, PTGS2, and JUN.