Network pharmacology analysis of a patented Chinese herbal medicine for alleviating anxiety disorder in in vitro fertilization-embryo transfer

Abstract

Objective

Qu's formula 3 (QUF3) is a patented Chinese herbal medicine used to alleviate anxiety disorders during in vitro fertilization-embryo transfer (IVF-ET). This study aimed to identify the potential active constituents and molecular mechanisms of action of QUF3 in alleviating anxiety disorders during IVF-ET.

Methods

The active constituents of QUF3 were identified from the Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform and literatures. Potential targets of anxiety disorder and IVF-ET were identified using GeneCards, Online Mendelian Inheritance in Man, and the UniProt Database. Protein-protein interaction (PPI) network, gene ontology (GO), and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed to identify the potential mechanisms. Molecular docking and molecular dynamics (MD) simulations were performed to visualize and verify the results.

Results

Quercetin, sophoranol, luteolin, kaempferol, and neurotoxin inhibitors were identified as the TOP 5 active constituents of QUF3. Forty common targets were shared among QUF3, anxiety disorders, and IVF-ET. Tumour necrosis factor, interleukin-6, vascular endothelial growth factor A, epidermal growth factor, interleukin-1B, cellular tumour antigen p53, matrix metalloproteinase-9, and oestrogen receptor were identified as the TOP 8 potential targets through PPI analysis. A total of 697 biological processes, 20 cellular components, and 54 molecular functions were identified. Further, 91 KEGG pathways were revealed to be enriched. The TOP 5 active constituents were verified to have good binding activity with the TOP 8 potential targets using molecular docking and MD simulations.

Conclusions

The mechanism of QUF3 in alleviating anxiety disorders in patients undergoing IVF-ET may be related to the interleukin-17 and tumour necrosis factor signalling pathways, inhibiting inflammatory responses and antioxidants, which may provide a solid foundation for the clinical application and further study of QUF3.

Graphical abstract

1. Introduction

In vitro fertilization-embryo transfer (IVF-ET) is an important treatment modality for infertile couples to achieve pregnancy; however, this method has some limitations, such as high cost and uncertain effects, which may cause anxiety among couples undergoing IVF-ET.^1,2 Anti-anxiety drugs used in this condition should not cause adverse effects on the would-be mother and the embryo/foetus and should not alter the woman's hormone profile, especially the pituitary hormones, oestrogen and progesterone. At different stages of ovulation induction, egg retrieval, and embryo transfer, many different difficulties may cause conception failure.³ Several studies have shown that infertile women undergoing IVF-ET are prone to psychological disorders such as anxiety and depression, which may have adverse effects on pregnancy outcomes.4, 5, 6

According to the theory of traditional Chinese medicine (TCM), the kidney is the primary organ that controls the reproductive ability in humans, and infertility is mainly related to kidney qi deficiency and chong-ren.⁷ Recent studies have shown that Chinese herbal medicine (CHM) has obvious benefits in alleviating the anxiety in IVF-ET patients.8, 9, 10 Qu's formula 3 (QUF3) is a patented Chinese herbal medicine for alleviating anxiety disorder associated with IVF-ET and is owned by the corresponding author's team (Chinese National Invention Patent: CN103877186A). It contains Poria Cocos (Fushen), Pseudostellariae Radix (Taizishen), Cuscutae Semen (Tusizi), Fructus Ligustri Lucidi (Nvzhenzi), Ziziphi Spinosae Semen (Suanzaoren), Herba Taxilli (Sangjisheng) and Platycladi Semen (Baiziren), with a percentage of 2:4:3:4:2:3:2 (100 g per dose). The Drug can be converted into capsules, decoctions, powders, or granules for oral administration. The clinical effects of QUF3 are remarkable; however, the underlying mechanisms have not been investigated.

The purpose of this study was to explore the treatment mechanisms of QUF3 in anxiety disorders in patients undergoing IVF-ET through a combination of network pharmacology, molecular docking analysis and molecular dynamics (MD) simulations.

2. Materials and methods

2.1. Screening the active constituents and potential targets of QUF3

The Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform (TCMSP; https://old.tcmsp-e.com/index.php)¹¹ was used to identify the active constituents of the seven herbs. Oral bioavailability ≥30% and drug-like ≥0.18 were used as the screening criteria.¹² Since Poria Cocos (Fushen) was not retrieved in TCMSP, its active constituents were obtained through the Hua-Yuan network (https://www.chemsrc.com/) and the relevant literature. The molecular structures of the corresponding constituents of Poria Cocos were obtained from the PubChem Database (https://pubchem.ncbi.nlm.nih.gov)¹³ and saved as Canonical SMILES. The obtained SMILES files were imported into the Swiss Target Prediction Database (http://www.swisstargetprediction.ch/),¹⁴ and “Homo sapiens” was used as the research species to obtain the targets of the constituents. The corresponding targets of the active constituents of other herbs were identified using TCMSP. Finally, UniProKBt in the Universal Protein Database (UniProt, http://www.Unitprot.org/)¹⁵ was used to match the official names of the target genes.

2.2. Anxiety disorder and IVF-ET related-targets predictions

“Anxiety disorder”, “in vitro fertilization-embryo transfer”, “IVF-ET” were used as the keywords to obtain potential targets of disease through GeneCards Database (https://www.genecards.org/), human online Mendelian genetic Database (OMIM, https://www.omim.org/), and Drugbank Database (https://www.drugbank.ca/). After integrating all collected disease targets and removing duplicates, we summarised the potential targets of the disease.

2.3. Network visualization

The active constituents of QUF3 and its potential targets were imported into Cytoscape software (Version 3.7.2, USA)¹⁶ to build a “herb-constituent-target” (H–C–T) network. “CytoNCA” was used to study and calculate network topology parameters, with particular emphasis on three important topological characteristics: degree centrality (DC), betweenness centrality (BC), and closeness centrality (CC). The DC of a node was used to define the topological characteristics of a single node and represented the total number of routes connected by a node and other nodes in the network. The ratio of all paths to the number of shortest paths travelling through a node was calculated using BC. The more this node contributed to preserving network connectivity, the higher the ratio. The reciprocal of the total distance from one node to all other nodes was calculated using the CC. The closer a node was to the centre, the greater was the value. Finally, the edge thickness represented the BC value between the two nodes, and the node size and colour were adjusted to match the magnitude of the DC value. According to the “Degree” value, the TOP 5 active constituents of H–C–T network were selected as the small molecule ligands for molecular docking.

2.4. Protein-protein interaction (PPI) network construction

The intersection targets of the QUF3 active constituents and anxiety disorder with IVF-ET were obtained and imported into STRING Database (http://stringdb.org/) to construct PPI network.¹⁷ The study species was set as “Homo sapiens”, the minimum interaction threshold was set to “medium confidence” (>0.4), and the rest were kept at the default. Finally, the PPI results were imported into the Cytoscape software (Version 3.7.2, USA) for visualization. After calculations using the “cytohubba” in Cytoscape, the key potential targets were obtained according to the “Degree” values.

2.5. Gene ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses

Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were performed using Metascape (http://metascape.org/).¹⁸ Intersection targets were entered into the gene list and species were set as “H. sapiens”, with P < 0.01 as the threshold for significant enrichment. The results were saved and visually analysed using an online drawing website (http://www.bioinformatics.com.cn/).

2.6. Molecular docking analysis

To screen the interacting activity and predict the most likely binding mode between the active constituents and the potential targets, which were obtained according to the “Degree” values as reported by Wang et,al.,¹⁹ the AutoDockTools 1.5.7²⁰ and PyMOL software²¹ (Version 2.1, USA) were used to calculate and visualize the molecular docking analysis. The 3D structures of the target proteins were searched using the Protein Data Bank Database²² (PDB, https://www.rcsb.org/). Mol2 files of the TOP five active constituents of QUF3 were downloaded from the TCMSP Database. The lowest binding energy for each docking event was used as the final result.

2.7. Molecular dynamics (MD) simulations

Molecular dynamics (MD) simulations were performed using the GROMACS 2020.3 software^23,24 (https://pubmed.ncbi.nlm.nih.gov/35815556/, https://pubmed.ncbi.nlm.nih.gov/33939584/). The amber99sb-ildn force field and general Amber force field (GAFF) were used to generate the parameters and topologies of the proteins and ligands, respectively. The simulation box size was optimised such that the distance between each atom of the protein and the box was greater than 1.0 nm. The box was filled with water molecules with a density of 1. To make the simulation system electrically neutral, water molecules were replaced with Cl− and Na⁺ ions. Following the steepest descent method, an energy optimisation of 5.0 × 10⁴ steps was performed to minimise the energy consumption of the entire system, and finally to reduce the unreasonable contact or atom overlap in the entire system. After energy minimisation, a first-phase equilibration was performed using an NVT ensemble at 300 K for 100 ps to stabilise the temperature of the system. Second-phase equilibration was simulated using an NPT ensemble at 1 bar and 100 ps. The primary objective of the simulation was to optimise the interaction between the target protein, solvent, and ions such that the simulation system was fully pre-equilibrated. All MD simulations were performed for 100 ns under an isothermal and isostatic ensemble at a temperature of 300 K and a pressure of 1 atm. The temperature and pressure were controlled by the V-rescale and Parrinello-Rahman methods, respectively, and the temperature and pressure coupling constants were 0.1 and 0.5 ps, respectively. The Lennard-Jones function was used to calculate the van der Waals force, and the nonbond truncation distance was set to 1.4 nm. The bond lengths of all the atoms were constrained using the LINCS algorithm. The long-range electrostatic interactions were calculated using the particle mesh-Ewald method with a Fourier spacing 0.16 nm.

3. Results

3.1. Screening the active constituents and potential targets of QUF3

After searching the TCMSP database and literature, we obtained 53 active constituents and 545 potential targets. Twelve in Poria Cocos (Fushen), eight in Pseudostellariae Radix (Taizishen), eleven in Cuscutae Semen (Tusizi), thirteen in Fructus Ligustri Lucidi (Nvzhenzi), nine in Ziziphi Spinosae Semen (Suanzaoren), two in Herba Taxilli (Sangjisheng), and five in Platycladi Semen (Baiziren). The details regarding the active constituents of QUF3 are listed in Table 1.

Table 1.

Information of the active ingredients of QUF3.

Herb	Active ingredients
Poria Cocos (Fushen)	Neurotoxin Inhibitor, Pseurotin D, Anti-neurodegeneration agent 1, D-erythro-Sphingosine, (R)-(−)-N-3-Benzylnirvanol, N-Acetyl-2-O-methyl-a-neuraminic Acid Methyl Ester 4,7,8,9-Tetraacetate,Ergosterol peroxide, 8-Hydroxypinoresinol, terpineols, Camaldulenic acid, Tormentic acid, Ursocholic acid
Pseudostellariae Radix (Taizishen)	Supraene, Acacetin, Linarin, Beta-sitosterol, Luteolin, Taraxerol, Schottenol, 1-Monolinolein
Cuscutae Semen (Tusizi)	Sesamin, NSC63551, Isorhamnetin, Beta-sitosterol, Kaempferol, Campest-5-en-3beta-ol, Isofucosterol, Matrine, Sophranol, CLR, Quercetin
Fructus Ligustri Lucidi (Nvzhenzi)	Beta-sitosterol, Kaempferol, Taxifolin, Lucidumoside D, Lucidumoside D_qt, (20S)-24-ene-3β, 20-diol-3-acetate, Eriodictyol, Syringaresinol Diglucoside_qt, Lucidusculine, Olitoriside, Olitoriside_qt, Luteolin, Quercetin
Ziziphi Spinosae Semen (Suanzaoren)	Ceanothic acid, (S)-Coclaurine, Daucosterol, Jujuboside A_qt, Phytosterol, Sanjoinenine, Swertisin, Zizyphusine, Mairin
Herba Taxilli (Sangjisheng)	Sitosterol, Quercetin
Platycladi Semen (Baiziren)	Arachidonic acid, 11,14-eicosadienoic acid, Sitosterol, Dihomolinolenic acid, 5Z-eicosenoic acid

3.2. Target search of anxiety disorders and IVF-ET

We searched and selected related disease targets using “anxiety disorders” as keywords. In the GeneCards database, the maximum value of Score was 103.57, the minimum value was 0.20, and the median was 3.40; therefore, a target Score >3.40 was set as the potential target. Overall, 3241 predicted targets were retrieved from GeneCards, 616 from OMIM, and 40 from DrugBank. In total, 2909 target genes closely related to anxiety disorders were identified after eliminating duplicate genes. The search with “IVF-ET” and “in-vitro fertilization and embryo transfer” revealed 139 targets from the GeneCards database and 559 targets from the OMIM database. After removing the duplicates, 341 potential therapeutic targets were identified. Potential targets of drugs and diseases were entered into an online drawing website (http://www.bioinformatics.com.cn/) and a Venn map (Fig. 1) was created. Forty intersecting genes were identified.

Fig. 1.

Venn diagram of the intersection genes of the active constituents of QUF3, the anxiety disorder-related targets and IVF-ET related targets.

3.3. Construction of H–C–T and “constituent-target-pathway” (C–T–P) networks

As shown in Fig. 2, the H–C–T network contained 602 nodes (6 TCM nodes, 51 compound nodes, and 545 predicted target nodes) and 1135 edges. Each edge represents the relationship between QUF3, its active constituents, and target genes. The “Degree” value represents the number of routes connected between the network and the nodes, and plays a key role in the nodes with large “Degree” value in the network relationship. The TOP 10 active constituents with Top 10-Degree values are shown in Table 2. Information regarding the nodes in H–C–T network is presented in Table 3.

Fig. 2.

“Herb-constituent-target” (H–C–T) network of QUF3.

Note: FS: Poria Cocos (Fushen); TZS: Pseudostellariae Radix (Taizishen); TSZ: Cuscutae Seme (Tusizi); NZZ: Fructus Ligustri Lucidi (Nvzhenzi); CZR: Ziziphi Spinosae Semen (Suanzaoren); BZR: Platycladi Seme (Baiziren); A1: luteolin; A2: beta-sitosterol; B1: kaempferol; B2: quercetin; C: sitosterol.

Table 2.

Core active ingredients in the TOP 10 of Degree values.

	Ingredients	Degree	OB%	DL	Drug
B2	Quercetin	92	46.43	0.28	TSZ NZZ SJS
TSZ7	Sophranol	45	55.42	0.68	TSZ
A1	Luteolin	37	36.16	0.25	TZS NZZ
B1	Kaempferol	37	41.88	0.24	TSZ NZZ
FS1	Neurotoxin Inhibitor	36	–	–	FS
FS2	Pseurotin D	34	–	–	FS
FS3	Anti-neurodegeneration agent 1	34	–	–	FS
A2	Beta-sitosterol	29	36.91	0.75	TZS TSZ NZZ
FS7	Ergosterol peroxide	29	44.39	0.82	FS

Note: FS: Poria Cocos(Fushen); TZS: Pseudostellariae Radix(Taizishen); TSZ:Cuscutae Seme(Tusizi); NZZ: Fructus Ligustri Lucidi(Nvzhenzi); CZR: Ziziphi Spinosae Semen(Suanzaoren); SJS: Herba Taxilli (Sangjisheng).

Table 3.

The information of nodes in “Herb-constituent-target” (H–C–T) network.

Node	MOL ID	Molecular name	OB	DL	Origin
BZR1	MOL001439	Arachidonic acid	45.57	0.2	Platycladi Semen (Baiziren)
BZR2	MOL002211	11,14-Eicosadienoic acid	39.99	0.2	Platycladi Semen (Baiziren)
BZR3	MOL003927	Dihomolinolenic acid	44.11	0.2	Platycladi Semen (Baiziren)
BZR4	MOL008153	5Z-eicosenoic acid	30.7	0.2	Platycladi Semen (Baiziren)
CZR1	MOL001521	Ceanothic acid	33.41	0.77	Ziziphi Spinosae Semen (Suanzaoren)
CZR2	MOL001522	(S)-Coclaurine	42.35	0.24	Ziziphi Spinosae Semen (Suanzaoren)
CZR3	MOL001525	Daucosterol	36.91	0.75	Ziziphi Spinosae Semen (Suanzaoren)
CZR4	MOL001527	Jujuboside A_qt	34.96	0.62	Ziziphi Spinosae Semen (Suanzaoren)
CZR5	MOL001532	Phytosterol	36.91	0.75	Ziziphi Spinosae Semen (Suanzaoren)
CZR6	MOL001539	Sanjoinenine	67.28	0.79	Ziziphi Spinosae Semen (Suanzaoren)
CZR7	MOL001542	Swertisin	31.83	0.75	Ziziphi Spinosae Semen (Suanzaoren)
CZR8	MOL001546	Zizyphusine	41.53	0.55	Ziziphi Spinosae Semen (Suanzaoren)
CZR9	MOL000211	Mairin	55.38	0.78	Ziziphi Spinosae Semen (Suanzaoren)
FS1	–	Neurotoxin Inhibitor	–	–	Poria Cocos (Fushen)
FS2	–	Pseurotin D	–	–	Poria Cocos (Fushen)
FS3	–	Anti-neurodegeneration agent 1	–	–	Poria Cocos (Fushen)
FS4	–	D-erythro-Sphingosine	–	–	Poria Cocos (Fushen)
FS5	–	(R)-(−)-N-3-Benzylnirvanol	–	–	Poria Cocos (Fushen)
FS6	–	N-Acetyl-2-O-methyl-a-neuraminic Acid Methyl Ester 4,7,8,9-Tetraacetate	–	–	Poria Cocos (Fushen)
FS7	–	Ergosterol peroxide	–	–	Poria Cocos (Fushen)
FS8	–	8-Hydroxypinoresinol	–	–	Poria Cocos (Fushen)
FS9	–	terpineols	–	–	Poria Cocos (Fushen)
FS10	–	Camaldulenic acid	–	–	Poria Cocos (Fushen)
FS11	–	Tormentic acid	–	–	Poria Cocos (Fushen)
FS12	–	Ursocholic acid	–	–	Poria Cocos (Fushen)
NZZ1	MOL004576	Taxifolin	57.84	0.27	Fructus Ligustri Lucidi (Nvzhenzi)
NZZ2	MOL005147	Lucidumoside D_qt	54.41	0.47	Fructus Ligustri Lucidi (Nvzhenzi)
NZZ3	MOL005169	(20S)-24-ene-3β,20-diol-3-acetate	40.23	0.82	Fructus Ligustri Lucidi (Nvzhenzi)
NZZ4	MOL005190	Eriodictyol	71.79	0.24	Fructus Ligustri Lucidi (Nvzhenzi)
NZZ5	MOL005195	Syringaresinol diglucoside_qt	83.12	0.8	Fructus Ligustri Lucidi (Nvzhenzi)
NZZ6	MOL005211	Olitoriside	65.45	0.23	Fructus Ligustri Lucidi (Nvzhenzi)
NZZ7	MOL005212	Olitoriside_qt	103.23	0.78	Fructus Ligustri Lucidi (Nvzhenzi)
TSZ1	MOL001558	Sesamin	56.55	0.83	Cuscutae Semen (Tusizi)
TSZ2	MOL000184	NSC63551	39.25	0.76	Cuscutae Semen (Tusizi)
TSZ3	MOL000354	Isorhamnetin	49.6	0.31	Cuscutae Semen (Tusizi)
TSZ4	MOL005043	Campest-5-en-3beta-ol	37.58	0.71	Cuscutae Semen (Tusizi)
TSZ5	MOL005440	Isofucosterol	43.78	0.76	Cuscutae Semen (Tusizi)
TSZ6	MOL005944	Matrine	63.77	0.25	Cuscutae Semen (Tusizi)
TSZ7	MOL006649	Sophranol	55.42	0.28	Cuscutae Semen (Tusizi)
TSZ8	MOL000953	CLR	37.87	0.68	Cuscutae Semen (Tusizi)
TZS1	MOL001790	Linarin	39.84	0.71	Pseudostellariae Radix (Taizishen)
TZS2	MOL006554	Taraxerol	38.4	0.77	Pseudostellariae Radix (Taizishen)
TZS3	MOL006756	Schottenol	37.42	0.75	Pseudostellariae Radix (Taizishen)
TZS4	MOL002464	1-Monolinolein	37.18	0.3	Pseudostellariae Radix (Taizishen)
TZS5	MOL001689	Acacetin	34.97	0.24	Pseudostellariae Radix (Taizishen)
TZS6	MOL001506	Supraene	33.55	0.42	Pseudostellariae Radix (Taizishen)
A1	MOL000006	Luteolin	36.16	0.25	Pseudostellariae Radix (Taizishen), Fructus Ligustri Lucidi (Nvzhenzi)
A2	MOL000358	Beta-sitosterol	36.91	0.75	Pseudostellariae Radix (Taizishen), Fructus Ligustri Lucidi (Nvzhenzi), Cuscutae Semen (Tusizi)
B1	MOL000422	kaempferol	41.88	0.24	Cuscutae Semen (Tusizi), Fructus Ligustri Lucidi (Nvzhenzi)
B2	MOL000098	quercetin	46.43	0.28	Cuscutae Semen (Tusizi), Fructus Ligustri Lucidi (Nvzhenzi), Herba Taxilli (Sangjisheng)
C1	MOL000359	sitosterol	36.91	0.75	Herba Taxilli (Sangjisheng), Platycladi Semen (Baiziren)

Note: OB:oral bioavailability; DL:drug-like; Since Poria Cocos (Fushen) was not retrieved in TCMSP, its active ingredients were obtained through the Hua-Yuan network (https://www.chemsrc.com/) and the relevant literatures. MOL ID, OB and DL were unavailable.

CytoScape3.7.2 was used to construct the core C–T–P network (Fig. 3). The network contained 10 chemical constituents, 394 targets, and 20 important pathways. According to the ranking of “Degree” values, the chemical constituents of QUF3 in alleviating anxiety in IVF-ET were mainly quercetin, sophoranol, luteolin, beta-sitosterol, neurotoxin inhibitor, and ceanothic acid. Targets of these active components include interleukin-6 (IL-6), tumour necrosis factor (TNF), interleukin-1B (IL1B), C–C motif chemokine 2, Caspase-8.

Fig. 3.

“Constituent-target-pathway” (C–T–P) network of QUF3.

Note: Octagons are the core constituents of QUF3. Diamond represents the key target. The inverted triangles represent pathways. The area of a node represents its “Degree”, and the larger the area, the more important is the node.

3.4. PPI network

A PPI network was constructed using 40 intersection targets (Fig. 4), and 39 nodes and 310 edges were obtained for the analysis. Higher “Degree” values indicate greater importance of the node. As shown in Fig. 4, the larger the node, the darker the colour, and the more important the gene. According to the “Degree” value and colour of the nodes, the TOP 8 key genes were: TNF, IL-6, vascular endothelial growth factor A (VEGFA), epidermal growth factor (EGF), IL1B, cellular tumour antigen p53 (TP53), matrix metalloproteinase-9 (MMP9), and Oestrogen receptor (ESR1).

Fig. 4.

PPI network.

Note: The node size is proportional to its “Degree”. Nodes are coloured in a radial pattern from blue to red, with the corresponding “Degree” values increasing.

3.5. GO and KEGG analysis

GO enrichment and KEGG pathway analyses of the 40 intersection genes were performed using Metascape, with P < 0.01 as a significant enrichment restriction. GO enrichment analysis included 771 GO function items, including 697 biological processes (BP), 20 cell components (CC), and 54 molecular functions (MF). The TOP 10 significantly enriched terms in the BP, MF, and CC categories are shown in Fig. 5. BP was mainly enriched in response to radiation (GO:0009314), cellular response to abiotic stimuli (GO:0071214), cellular response to environmental stimuli (GO:0104004), cellular response to light stimuli, and cellular response to UV (GO:0034644). CC was mainly enriched in platelet alpha granules (GO:0031091), cytoplasmic vesicle lumen (GO:0060205), vesicle lumen (GO:0031983), and external encapsulating structures (GO:0030312). MF was mainly enriched in cytokine receptor binding (GO:0005126), signalling receptor regulator activity (GO:0030545), growth factor receptor binding (GO:0070851), and oestrogen response element binding (GO:0034056).

Fig. 5.

GO enrichment analysis.

A total of 91 KEGG pathway enrichment items were identified, and the TOP 20 items were screened, as shown in Fig. 6. It mainly involved the advanced glycation end products (AGE) and their receptor (RAGE) signalling pathway in diabetic complications, lipid and atherosclerosis, rheumatoid arthritis, human cytomegalovirus infection, interleukin-17(IL-17) signalling pathway, and TNF signalling pathway.

Fig. 6.

KEGG enrichment analysis.

3.6. Molecular docking analysis

TNF, IL6, VEGFA, EGF, IL1B, TP53, MMP9, and ESR1 were selected as potential therapeutic targets for molecular docking with the TOP five active constituents: quercetin (B2), sophoranol (TSZ7), luteolin (A1), kaempferol (B1), and a neurotoxin inhibitor (FS1). The binding energy less than 0 indicated that the ligand and receptor can spontaneously dock, the binding energy less than −5.0 kcal/mol indicated well docking activity, and the binding energy less than −7.0 kcal/mol indicated strong docking activity.²⁵ The results show that quercetin, sophoranol, luteolin, kaempferol, and neurotoxin inhibitors have strong affinity to TNF, IL6, IL1B, TP53, MMP9 as well as ESR1 as shown in Table 4. There were six molecular docking pairs with binding energies less than −8 kcal/mol, namely, TNF-FS1, TNF-TSZ7, MMP9-A1, MMP9-B1, MMP9-B2, MMP9-TSZ7, and MMP9-TSZ7, which formed multiple structurally stable hydrogen bonds. Taking the results with binding energies less than −8 kcal/mol as an example, the docking results were visualised using PyMOL software (Fig. 7). As shown in Fig. 7F, TSZ7 formed four hydrogen bonds with ARG-239 in MMP9 with bond lengths of 2.3, 2.4, 1.9 and 1.9, respectively. As shown in Fig. 7A, FS1 bound best to TNF. It can form hydrogen bonds with SER-99 and four hydrogen bonds with PHE-103, with bond lengths of 2.9, 2.2, 2.2, 3.5, and 2.7.

Table 4.

Molecule docking energy of core ingredients of QUF3 and key target.

	Resolution	FS1	A1	B1	B2	TSZ7
TNF (PDB ID: 5M2J)	2.70A	−9.59	−5.54	−7.9	−7.51	−8.23
IL6 (PDB ID: 1ALU)	2.50A	−7.8	−4.67	−6.61	−5.73	−6.08
VEGFA (PDB ID: 4KZN)	1.71A	−6.68	−5.41	−5.77	−5.29	−6.04
IL1B (PDB ID: 31BI)	1.89A	−7.88	−6.37	−7.06	−7.23	−6.95
TP53 (PDB ID: 8A92)	1.37A	−7.83	−6.84	−6.86	−7.3	−6.53
MMP9 (PDB ID: 4XCT)	1.30A	−6.88	−9.42	−9.64	−9.41	−11.13
ESR1 (PDB ID: 1X7R)	2.00A	−7.83	−7.5	−3.96	−3.79	−5.5
EGF (PDB ID: 1NQL)	2.80A	−6.68	−5.15	−5.53	−5.65	−6.25

Note: B2:quercetin; TSZ7: sophranol; A1: luteolin; B1: kaempferol; FS1: neurotoxin inhibitor.

Fig. 7.

Docking display of QUF3 active constituents and key targets.

A, TNF-FS1; B, TNF-TSZ7; C, MMP9-A1; D, MMP9-B1; E, MMP9-B2; F, MMP9-TSZ7.

3.7. The molecular dynamics (MD) simulations

The six molecular docking pairs with binding energies less than −8 kcal/mol were used for further analysis. As shown in Fig. 8A, the protein ligand reaches equilibrium after 70 ns. There was no significant change in the solvent-accessible surface area and radius of gyration values of the protein-ligand complexes during all composite simulation processes, indicating the stable binding of protein ligands (Fig. 8B,C). To investigate the interactions between proteins and ligands, we first conducted hydrogen bonding analysis of the protein ligands, as shown in Fig. 8D. The average hydrogen bond numbers of MMP9-A1, MMP9-B1, MMP9-B2, MMP9-TSZ7, TNF-FS1, and TNF-TSZ7 were 1.01, 0.74, 1.91, 1.50, 1.69, and 0.89, respectively, indicating the presence of hydrogen bond interactions between the proteins and ligands. To better explain the interaction energy between proteins and ligands, gmx-Mmpbsa (https://jerkwin.github.io/gmxtool/) was used. The script determines the binding energies of all the protein-ligand complexes during the equilibrium phase. The binding energies between the proteins and ligands are shown in Table 5. In the protein-ligand complex system, the binding free energies of MMP9-A1, MMP9-B1, MMP9-B2, MMP9-TSZ7, TNF-FS1, and TNF-TSZ7 were negative, with negative values of −118.920, −129.227, −46.478, −108.129, −210.974, and −219.147 kJ/mol, respectively, indicating that the protein-ligand can stably bind. The main interaction energy is the van der Waals interaction energy.

Fig. 8.

Molecualr dynamics (MD) simulation of QUF3 active constituents and key targets.

Note: A: Simulation of Protein Ligand Complex Process: Changes in the root-mean-square deviation (RMSD) of protein over time; B: Simulation of Protein Ligand Complex Process: Changes in the solvent-accessible surface area (SASA) of protein over time; C: Simulation of Protein Ligand Complex Process: Changes in the radius of gyration (Rg) of protein over time; D: Simulation of Protein Ligand Complex Process: Changes in the hydrogen bonds of protein over time.

Table 5.

MMPBSA analysis of protein and ligand.

Energy	MMP9-A1	MMP9-B1	MMP9-B2	MMP9-TSZ7	TNF-FS1	TNF-TSZ7
Van der Waals Energy (KJ/mol)	−180.565	−196.986	−126.668	−181.726	−173.177	−137.395
Electrostatic energy (kJ/mol)	−28.948	−19.014	−113.992	−42.975	−240.173	−256.306
Polar solvation energy (KJ/mol)	96.461	97.013	163.526	116.147	192.468	168.855
Nonpolar solvation Energy (KJ/mol)	−17.754	−17.439	−14.481	−15.607	−17.310	−14.540
Total Binding Energy (KJ/mol)	−135.806	−136.425	−91.615	−124.162	−238.192	−239.386
TΔS(KJ/mol)	16.886	7.199	45.137	16.033	27.218	20.239
Total Binding Free Energy (KJ/mol)	−118.920	−129.227	−46.478	−108.129	−210.974	−219.147

4. Discussion

It is estimated that 25–60% of couples with infertility receiving IVF-ET are more susceptible to anxiety than normal fertile couples.²⁶ Anxiety caused by IVF-ET has a negative impact on patients’ quality of life, mental health, and pregnancy outcomes.²⁷ Therefore, alleviating the anxiety of infertile patients undergoing IVF-ET is essential for their mental and reproductive health. Compared with anxiolytic drugs such as fluoxetine (Prozac), sertraline (Zoloft), paroxetine (Paxil), and citalopram (Celexa), which have gastrointestinal effects, somatic dependence, anxiety rebound, and other problems,^28,29 the therapeutic effect of TCM in relieving anxiety in IVF-ET is gentler and more effective.^30,31

Studies have shown that selective serotonin reuptake inhibitor-based antidepressants may adversely affect fertility in both men and women.³² Use of antidepressants during pregnancy is associated with severe congenital malformations, and a cohort study has showed that citalopram could significantly increase the risk of congenital malformations, and that antidepressants inhibiting serotonin reuptake effects had an increased risk of certain organ-specific defects.³³ Studies have reported that TCM has advantages in treating anxiety during IVF-ET.34, 35, 36 Acupuncture, ear point pressure beans, music therapy, mindfulness yoga therapy are the traditional ways of treating anxiety during pregnancy in TCM. TCM believes that anxiety and depression are mostly related to liver dysfunction. Huifang used the Kai-Yu-Zhong-Yu decoction, a classic gynaecological prescription of TCM, to treat patients during IVF-ET, and improve the natural pregnancy and clinical pregnancy rate.³⁷ We have compared the active components of the Kai-Yu-Zhong-Yu decoction with QUF3 and found that there were many different chemical constituent and molecules. This study is the first to focus on the molecular mechanism of Chinese herbal Medicine in the treatment of anxiety in patients undergoing IVF-ET.

Pharmacological studies have shown that quercetin and kaempferol have anxiolytic effects.^38,39 Quercetin is believed to have antioxidant, anti-inflammatory, immunoprotective, and anticancer effects. It was found to alleviate anxiety-like behaviour and monoamine oxidase activity in mice by regulating oxidative stress.⁴⁰ Some researchers reviewed that, reducing the level of reactive oxygen species (ROS) may help alleviate the level of anxiety.⁴¹ Experiments showed that quercetin alleviated anxiety caused by methamphetamine by reducing the level of ROS, mitochondrial membrane potential, increasing oxygen consumption rate, and adenosine triphosphate production.⁴² Kaempferol is a flavonoid antioxidant that suppresses the production of reactive oxygen species, regulates endogenous antioxidants such as superoxide dismutase and glutathione, suppresses the formation and aggregation of β-amyloid (β-A), and regulates brain-derived neurotrophic factor to protect the brain.⁴³ Modern experiments have shown that luteolin can improve the anxiety and depression-like symptoms caused by chronic pain.⁴⁴ Luteolin can also lead to an increase in norepinephrine and decreased serotonin levels in the medial prefrontal cortex and hippocampus, showing antidepressant and anxiolytic effects.⁴⁵ In this study, the main active constituents of QUF3 included quercetin, luteolin, sophoranol, kaempferol, and neurotoxin inhibitors, which is consistent with the results of some previous studies.

There are 40 common targets of QUF3, anxiety disorders, and IVF-ET that were regarded as potential therapeutic targets for QUF3 for alleviating anxiety disorders in IVF-ET patients. The PPI network results indicated that the associations among the targets were close. The TOP eight potential targets were TNF, IL6, VEGFA, EGF, IL1B, TP53, MMP9, and ESR1, which were partly consistent the results of previous studies. TNF is a pro-inflammatory cytokine involved in multiple cellular responses, such as apoptosis and proliferation, and has been viewed as a target for neuropathic pain.⁴⁶ TNF levels change significantly in the cells, plasma, and serum of severely depressed patients,^47,48 while TNF signalling also plays an important role in the state of anxiety.⁴⁹ This is partly in line with the results of molecular docking analysis and MD simulations. It has been shown that the excessive production of cytokines is a key factor leading to inflammation in the nervous system, and the pathological state of anxiety and depression is crucial.^50,51 IL-6 and IL-1B are important pro-inflammatory cytokines, and the inflammatory response is also closely related in stress-related neuropsychiatric disorders.^52,53 The levels of TNF-αand IL-6 are increased in the foetal cord blood of pregnant women with high anxiety levels, which can also negatively affect the newborn.⁵⁴ VEGF regulates vascular endothelial cells, and studies have found that VEGF is widely involved in the regulation of anxiety, depression, and other behaviors.⁵⁵ VEGF is thought to have antidepressant-like effects, and VEGF signalling mediates cell proliferation of antidepressants through Flk-1 and causes behavioural changes in chronic antidepressant models.⁵⁶ Warner-Schmidt reported that in a rat model of depression, VEGF expression in the hippocampus was elevated and was regulated by selective serotonin reuptake inhibitors or selective serotonin/norepinephrine reuptake inhibitors, which were used for antidepressant treatment.⁵⁷ EGF promotes the growth and development of the nervous system.⁵⁸ Some studies suggest that serum EGF levels are significantly negatively correlated with cognitive dysfunction⁵⁹ and suggest that it may be a suitable biomarker for cognitive, emotional, and social disorders in the diagnosis and prognosis of different mental disorders.⁶⁰ TP53, also known as p53 gene, is widely distributed in the hippocampus, thalamus, and cerebellum, and is involved in the regulation of the transcription and expression of genes related to neuronal differentiation and apoptosis.⁶¹ In SIRT1/p53 pathway, SIRT1 was increased by CHM intervention, and the acetylation of p53 downstream in SIRT1 signalling pathway was downregulated to exert antidepressant-like effect.⁶² Matrix MMP9 has been implicated with several neurological diseases. Treatment with MMP9 inhibitors reduces anxiety, and the knockdown of MMP9 in mice not only improves social interaction and social cognition but also reduces anxiety in mice.⁶³ The results of this study proved that the active constituents of QUF3 have good binding energy with MMP9. Oestrogen is believed to play a key role in anxiety. ESR1 variants are associated with a risk of anxiety disorders.⁶¹ Several studies have shown that ESR1 may play a role in the predisposition to mood disorders, especially in women.^64,65

KEGG enrichment analysis revealed that the pathways with high enrichment were the AGE-RAGE signalling pathway in diabetic complications, IL-17 signalling pathway, and TNF signalling pathway. Previous trials have shown that Meningeal γδ T cells associated through neuronal IL-17a signalling to regulate anxiety-like behaviour. IL17a-/-mice had significantly more open arms than the control group, and injection of IL-17a neutralising antibody into the cerebrospinal fluid of mice reduced anxiety-like behaviour.⁶⁶ Acanthopanacis Cortex can improve depressive behaviour and regulate the levels of neurotransmitters, neurotrophies, and pro-inflammatory cytokines. The possible mechanism is mediated by neuroinflammation modulation through the IL-17 signalling cascade.⁶⁷ The TNF signalling pathway is a classical inflammation-related pathway that plays an important role in causing cognitive and memory impairment in the brain.⁶⁸ Liu et al. used sleep deprivation to produce an anxious state in mice and confirmed that key genes enriched in TNF signalling are altered during acute paradoxical sleep deprivation.⁶⁹ In the C-T-P network, each chemical constituent corresponds to multiple targets and pathways, further proving that QUF3 acts through the synergistic action of multiple constituents and targets to treat anxiety disorders.

5. Limitations

This study had some limitations. First, molecular docking and MD simulations were used to verify the binding activity between the active constituents and potential targets. Second, this was an in silico study, and further validation studies, such as biochemical and biophysical studies and in vivo or in vitro experiments, are needed in the future to validate these results. Third, the clinical efficacy and safety of QUF3 should be evaluated. Finally, this study focused only on the molecular mechanisms of QUF3; future experiments should be refined to evaluate the interactions between multiple drug constituents to provide a reference for clinicians.

6. Conclusion

In this study, a network pharmacology strategy was used to explore the mechanism by which patented CHM-QUF3 alleviates anxiety disorders in patients undergoing IVF-ET. Quercetin, sophoranol, luteolin, kaempferol, and neurotoxin inhibitors are the active compounds of QUF3. TNF, IL6, VEGFA, EGF, IL1B, TP53, MMP9, and ESR1 are potential therapeutic targets that provide an effective basis for clinical practice. However, biochemical and biophysical experiments, as well as well-designed clinical studies, should be performed to verify these results.

Funding

This research was supported by Zhejiang Provincial Natural Science Foundation of China under Grant No. LZ21H270001.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Abbreviations

QUF3

Qu's formula 3

IVF-ET

in vitro fertilization-embryo transfer

TCMSP

Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform

OMIM

Online Mendelian Inheritance in Man

PPI

protein-protein interaction

GO

gene ontology

KEGG

Kyoto encyclopedia of genes and genomes

MD

molecular dynamics

TNF

tumour necrosis factor

IL6

interleukin-6

VEGFA

vascular endothelial growth factor A

EGF

epidermal growth factor

IL1B

interleukin-1B

TP53

cellular tumour antigen p53

MMP9

matrix metalloproteinase-9

ESR1

estrogen receptor

TCM

Traditional Chinese medicine

CHM

Chinese herbal medicine

OB

oral bioavailability

DL

drug-like

DC

degree

BC

betweenness centrality

CC

closeness centrality

PDB

Protein Data Bank Database

GAFF

general Amber force field

CCL2

C–C motif chemokine 2

CASP8

caspase-8

BP

biological processes

CC

cell component

MF

molecular functions

IL-17

interleukin-17

AGE

advanced glycation end products

RAGE

advanced glycation end products -their receptors

SASA

solvent-accessible surface areas

Rg

radius of gyration

RMSD

root-mean-square deviation

SSRIs

selective serotonin reuptake inhibitor-based antidepressants

ROS

reactive oxygen species

MMP

mitochondrial membrane potential

OCR

oxygen consumption rate

ATP

adenosine triphosphate

β-A

β-amyloid

BDNF

brain-derived neurotrophic factor

SIRT1

sirtuin 1

CSF

cerebrospinal fluid

AC

acanthopanacis cortex

PSD

paradoxical sleep deprivation