To Explore the Network Pharmacology and Molecular Docking Mechanism of Chaihu Shugan Powder with the “Same Treatment for Different Diseases” for Insomnia and Depression Based on the COVID-19 Pandemic

Liang Wang Medical School of Chinese PLA Peng Wang Medical School of Chinese PLA Yingfan Chen Medical School of Chinese PLA Chen Li Medical School of Chinese PLA Xuelin Wang Medical School of Chinese PLA Mingwei Li Medical School of Chinese PLA Zhenxian Luan Medical School of Chinese PLA Yin Zhang Chinese PLA General Hospital Shaodan Li (  lsd301@126.com ) Chinese PLA General Hospital Minghui Yang Chinese PLA General Hospital


Introduction
During the global pandemic of COVID-19, economic data in various countries fell, and unemployment rates rose. During the epidemic, many people need to be isolated at home and face outstanding problems, such as work restrictions, academic delays and nancial burdens [1]. Many surveys [2][3][4][5] have shown that medical staff are at the front line of the ght against the epidemic and under high-intensity work pressure, and they are more likely to suffer from diseases such as depression and sleep disorders. Insomnia, as a common subjective sleep disorder, is mainly manifested as having di culty falling asleep, sleep maintenance disorder, continuous early awakening, decreased sleep quality, and total sleep time reduction, accompanied by daytime dysfunction [6]. Meanwhile, depression, as a widespread psychological disorder, is characterized by signi cant and lasting depression, reduced mobility, delayed thinking and cognitive functions with boring, helpless, incompetent, powerless, hopeless and worthless feelings, affecting up to 20% of the global population [7]. Studies have found that the relationship between sleep and depression is two-way, complex and obvious [8]. With the continuous development of the social economy, the prevalence and comorbidities of insomnia and depression are becoming increasingly common [9], and they often occur at the same time [10]. In particular, depression and insomnia caused by the impact of the COVID-19 epidemic have become increasingly prominent, and timely treatment cannot be waited for them [11]. Due to various adverse reactions in the clinical use of Western medicine, many studies have shown that the use of complementary and alternative medicine (CAM) in mental illnesses such as insomnia and depression is a common phenomenon [12]. In addition, compared with pure Western medicine treatment, TCM alone or in combination with Western medicine has the characteristics of quick onset, high cure rate, and fewer side effects [13].
For thousands of years, TCM and its compounds have been used to treat mental illnesses such as insomnia and depression [14]. The earliest record of CHSGP was found in "Jingyue Quanshu" and described it as a classic TCM prescription developed from Sini Soup. Its main components include Chaihu, Baishao, Chuanxiong, Chenpi, Xiangfu, Zhike, and Gancao. The principal active ingredients of CHSGP are monoterpene glycosides, glucose gallic acid, phenolic compounds, lactones, avonoids and triterpene saponins [15]. CHSGP has the effects of soothing the liver, promoting qi, activating blood and relieving pain and is often used to treat depression, insomnia, anxiety, chronic gastritwas, hepatitis and other diseases [16]. As the name suggests, "Same Treatment for Different Diseases" means different diseases can be treated the same. The theory of TCM classi es insomnia and depression into the categories of "insomnia" and "stagnation syndrome", respectively. The classi cation often involves multiple organs, such as the liver, spleen, lung, and heart, and can often be treated for liver depression and qi stagnation. This is the fundamental principle of treatment in the "Same Treatment for Different Diseases" of CHSGP. CHSGP matches the menstrual ow of insomnia and depression and can better exert its therapeutic effect. Pharmacological studies have found that the antidepressant pharmacological properties of CHSGP also have a certain therapeutic effect on insomnia and mainly regulate the neuroendocrine immune network, exerting anti-in ammatory and antioxidative stress effects [17].
Due to its multicomponent and multitarget characteristics, TCM makes in-depth research di cult. Network pharmacology explores the material basis of TCM treatment of diseases from a systemic and holwastic perspective by constructing a "drug-target-gene-disease" network and molecular mechanism [18]. This nding is consistent with the overall concept of TCM and Same Treatment for Different Diseases. It provides new ideas, new methods and new ways for TCM mining and analysis and shows the visualization and relevance of data. Its fast and accurate data mining classi cation and positioning method has been widely recognized [19,20]. Therefore, this study constructed a "component-target-pathway-disease" association network of CHSGP for insomnia and depression with "Same Treatment for Different Diseases" through network pharmacology and molecular docking. The mechanism of action of multiple targets and pathways provides a certain theoretical basis for clinical applications in treating insomnia-depression during the COVID-19 pandemic.

2.1.Research ideas
Under the context of the COVID-19 pandemic, this study was based on network pharmacology and molecular docking to explore the research ideas in the mechanism of CHSGP "treating different diseases with the same treatment" insomnia and depression. The experimental ow is shown in Figure 1.

2.2.Screening the effective ingredients and targets of CHSGP
Through the TCMSP database [21] (https://tcmspw.com/tcmsp.php), the pharmacological analysis platform of the Chinese medicine system, OB≥30%, DL≥0.18, was used as the ingredient screening criteria to search for Chaihu, Baishao, Chuanxiong, Chenpi, Xiangfu, Zhike, and Gancao. The components of the included compounds were predicted and extracted through the TCMSP database. All targets were corrected by the UniProt database [22](https://www.uniprot.org/), and nonhuman targets were removed to obtain "o cial genes symbol" of all targets. According to the theory of TCM, the effective active ingredient targets of TCM have the characteristics of selective distribution in the body, which is basically consistent with the visceral relationship of the corresponding meridian homing. A Venn diagram was used to show the intersection of the meridian targets of TCM.

Construction and analysis of the protein interaction network (PPI)
We inputed the common targets of drugs and diseases into the String database [23] (https://string-db.org/cgi/input.pl), used the multiple proteins function of the database to construct a PPI network, and set the biological species as "Homo sapiens" at the same time. Finally, a combined score >0.4 was used as the screening standard to obtain the PPI network model of the intersection target.

Topological analysis and cluster analysis
In this study, the screened PPI network information was imported into Cytoscape 3.8.0 for network topology analysis, and common key targets were screened based on the degree value and betweenness centrality. We imported the constructed PPI network into Cytoscape 3.8.0, analyzed gene clusters and screened core genes through MCODE module analysis [24].

Heatmaps of key target genes expressed in organs and tissues
We inputed the key target genes selected by topological analysis into the BioGPS database (http://biogps.org/#goto=welcome) and set "current layout" as the "default layout", downloaded the organ and tissue expression data corresponding to the rst 20 key target genes. We standardized and visualized the data and plotted the expression heatmaps of the top 20 key target genes in major organs and tissues.

2.6.Visualization of the component-disease-target network and screening of key TCM components
To better understand the complex interactions between components, diseases and corresponding targets, we built a component-disease-target network diagram based on the included components, disease treatments, and targets and introduced Cytoscape 3.8.0 and drew a visual network diagram. We imported the component-disease-target network diagram into Cytoscape 3.8.0 for topological analysis and sorted the components by degree. The higher the degree value, the more important the component (we selected the component greater than the average degree value as the key component for subsequent follow-up research).

2.7.GO enrichment analysis and KEGG pathways analysis
The common targets of drugs and diseases were analyzed for GO enrichment and KEGG pathway analysis. GO enrichment included biological process (BP), molecular function (MF), and cell component (CC) enrichment. Quoting the String database and ltering the items with the corrected P value <0.05. At the same time, R 4.0.3 software was used to install and reference clusterPro ler, enrichplot, and ggplot2 packages and draw histograms and bubble charts. We used the Sangerbox tool to visually analyze the drug-disease KEGG signaling pathway and draw a chord diagram.

2.8.Composition-disease-Pathway-Target Network Construction
We imported the component-disease-pathway-target network le into Cytoscape 3.8.0 to draw the path network diagram, which more intuitively showed the characteristics of the multicomponent-multitarget effect of the TCM active ingredients in the treatment of diseases.

2.9.Molecular docking veri cation
We imported the constructed component-disease-target network diagram to screen the key component compounds, such as kaempferol, luteolin, quercetin, beta-sitosterol, and 7-methoxy-2-methyl iso avone, into the PubChem database (https://pubchem.ncbi.nlm.nih.gov/), queried the MOL2 format, and imported Schrodinger software to establish a database, which was prepared as a ligand molecular database for molecular docking [25]. We downloaded the target protein crystal structure from the protein database (https://www.rcsb.org/), processed it with the Maestro11.9 platform, processed the protein with Schrodinger's Protein Preparation Wizard [26], and prepared the protein receptor [27]. All compounds were prepared according to the default settings of the LigPre module [28]. Finally, molecular docking and screening were carried out by the SP method, and Pymol2.1 was used for visual analysis.

Inquiry and analysis of TCM ingredients and target information
Using the TCMSP database and setting OB≥30% and DL≥0.18 as the screening conditions, the components of the included compounds were predicted by the TCMSP database. All targets were calibrated by the UniProt database, and nonhuman targets were removed. After summarizing and deleting the duplicates, we got 17 compound components and 179 targets of Chaihu, 13 compound components and 86 targets of Baishao, 5 compound components and 65 targets of Chenpi,7 compound components and 30 targets of Chuanxiong, Gancao has 93 compound components and 216 targets, Xiangfu has 18 compound components and 210 targets, Zhike has 5 compound components and 88 targets. The analysis of Guijing in CHSGP showed that Chaihu, Baishao, Chuanxiong, and Xiangfu belonged to the liver channel with 18 common targets; Baishao, Xiangfu, Zhike, Chenpi, and Gancao belonged to the spleen channel with 27 common targets; Chaihu, Chenpi, and Gancao belonged to the lung channel with 39 common targets, as shown in Table 1. The above data showed that each Chinese medicine of CHSGS mainly belonged to the liver, spleen, and lung meridians, which were also in line with the location and pathogenesis of insomnia and depression and re ected the connotation of "Same Treatment for Different Diseases", as shown in Figure 2.

Venn diagram of common targets in drugs and diseases
We inputed the selected drug targets and disease targets into Venn diagram production software (Venny 2.1) and obtained 113 common targets, which were used as common predictive targets for the following pathway enrichment analysis, as shown in Figure 3.

Construction and analysis of PPI network
There were 113 nodes, 1365 edges, and an average degree of 24.2 in the network. Figure 4A shows the PPI network diagram exported from the STRING website, and Figure 4B shows the PPI network diagram drawn by Cytoscape software. The color and size of the nodes in Figure 4B were adjusted according to the degree value. The larger the color, the redder the color and the greater the degree value. Thickness from thick to thin represents the edge betweenness from large to small.

Topological analysis
We imported the PPI network into Cystoscape3.8.0 [29], used the Network Analyzer tool to perform topology analysis, and sorted by degree; the greater the degree value, the greater the role of the node in the network graph. We selected genes with scores greater than the average score as key targets. A total of 42 key targets were screened, and the top 20 targets were listed as AKT1, IL6, IL1B, CASP3, MAPK3, PPARG, MMP9, CXCL8, IL10, HSP90AA1, ESR1, FOS, HIFIA, CREB1, NOS3, HMOX1, MMP2, CCND1, ERBB2 and CASP8. We used R 4.0.3 to draw pictures of the rst top target points, where the abscissa was the degree value of each target point, as shown in Figure 5.

MCODE cluster analysis
We imported the constructed PPI network into Cytoscape 3.8.0 and opened the MCODE module [24] to analyze gene clusters and screened core targets to obtain 7 gene clusters and 7 core genes (FOS, GABRA2, GSK3B, PON1, APOB, CYP1B1, ADRA1A). The speci c results are shown in Table   2. To better understand the complex interactions among ingredients, diseases and corresponding targets, we constructed an "ingredient-disease-target" network based on the included ingredients, the treatment of diseases and the action targets. We screened 119 core components, such as MOL000098 (quercetin), MOL000422 (kaempferol), MOL000358 (β-sitosterol), MOL003896 (7-methoxy-2-methyl iso avone), and MOL000006 (luteolin), and 113 core targets, such as AKT1, IL6, IL1B, CASP3, and MAPK3. We imported the component-disease-target network into Cytoscape 3.8.0 for topological analysis. The degree was sorted according to the ingredients. The higher the degree, the more important the ingredient, and the results are shown in Table 2. Table 3 shows that the top 5 compounds with a median value of the network were MOL000098 (quercetin), MOL000422 (kaempferol), MOL000358 (β-sitosterol), MOL003896 (7-methoxy-2-methyl) iso avones), and MOL000006 (luteolin), which could provide small drug compounds for subsequent molecular docking.  Figure 6A, 6B, 6C and 6D.

Construction component-disease-pathway-target network and chord diagram visualization
We imported the component-disease-pathway-target network le into Cytoscape 3.8.0 to draw the path network diagram. More intuitively, the characteristics of the multicomponent-multitarget effect of the TCM active ingredients in the treatment of diseases were revealed, as shown in Figure 7 (blue is the compound, yellow is the target of the Chinese medicine on the disease, green is the top 20 most signi cant pathways, red is the disease that is the disease, and purple is the Chinese medicine). The KEGG signaling pathway of drug-disease was visually analyzed using the Sangerbox tool, and it was found that the top 20 signaling pathways signi cantly enriched by KEGG involved 85 genes, as shown in Figure 8.

Molecular docking
We used Pymol2.1 software to visualize the complex formed by the compounds and proteins after molecular docking (select the compound with the most negative binding energy score for each target), which obtained the binding mode of the compound and the protein, as shown in Table 4.
According to the binding mode, the amino acid residues where the compound binds to the protein pocket could be clearly seen; for example, kaempferol binds to AKT1. The active amino acid residues included VAL-164, GLU-228, ALA-230, ASP-292, MET-281, GLU-234, etc. Kaempferol is a avonoid compound that contains multiple hydroxyl groups and can interact with the active groups of amino acids to form hydrogen bonds. For example, it could interact with the active groups of GLU-228, ASP-292, ALA-230, and GLU-234. The strong hydrogen bond interaction, with an average hydrogen bond distance of 2.0 Å, was much smaller than the traditional hydrogen bond of 3.5 Å, which played an important role in stabilizing small molecule ligands. In summary, kaempferol, luteolin, quercetin, and 7-methoxy-2-methyl iso avone compounds matched well with the ve target protein targets, could form stable complexes, and had a good relationship with the protein, which also indirectly veri ed these compounds. It had a regulatory effect on AKT1, IL1B, IL-6, FOS, GSK3B, GABRA and other targets. When the binding energy is less than 0, it is considered that the ligand and receptor can bind freely, and the lower the binding energy is, the stronger the a nity. These results indicated that the molecular docking results were consistent with the screening results of network pharmacology, which preliminarily veri ed the reliability of network pharmacology, as shown in Figure 9.

Discussion
Under the current COVID-19 pandemic, many people with depression and insomnia may be related to environmental mutations and social anxiety. The study found that the symptom levels of depression, anxiety and insomnia in the quarantine area were signi cantly higher than those in the nonquarantine area, and they were at higher risk of depression, anxiety and insomnia, especially the severity of depression [30]. TCM believes that the main pathogenesis of insomnia is the imbalance of Qi and the inability of Yang to enter Yin. The Yellow Emperor's Internal Classic states, "If you are unable to make decisions, or if your emotions are not smooth, the liver qi will be stagnant, and the qi axis will not turn, but if you want to stretch, you will be inward. Disturb the soul and cause insomnia", and depression is due to stagnation of liver qi, so liver-regulating qi is the basic rule for the treatment of insomnia and depression, suggesting that qi-regulating drugs are potential drugs for the treatment of depression.
CHSGP is composed of 7 herbs: Chaihu, Xiangfu, Chuanxiong, Chenpi, Zhike, Baishao, and Gancao. Saikosaponin, the main chemical component of Chaihu, has sedative, antipyretic, immune enhancement, anti-in ammatory, antiviral, liver, and antitumor effects. The α-germanone of Xiangfu has an antidepressant-like effect [31], and Jiawei Xiangfu Decoction combined with acupuncture can signi cantly improve the sleep state of menopausal insomnia patients with liver stagnation and qi stagnation [32]. The main active ingredients of Chuanxiong have both a sedative effect [26] and the potential to improve depression. The antidepressant mechanism of Chenpi and its main active ingredients, chuanchenein, hesperidin and naringenin, is related to the improvement of neurobiochemical, neuroendocrine and neurotrophic systems [33]. Zhike exerts antidepressant effects mainly through its monoaminergic mechanism and prokinetic effects [34]. Fructus aurantii can enhance gastrointestinal motility by changing the expression level of 5-HT in the gastrointestinal tract of rats [35], suggesting that it can affect sleep to a certain extent. Paeoni orin has neuroprotective and antidepressant biological activities, and it can treat related insomnia diseases through adenosine A1R to play a hypnotic effect [36]. Regulating the calmodulin/calmodulin-dependent protein kinase II (CaM/CaMKII) pathway and its downstream signaling molecules play an important role in the treatment and alleviation of affective disorders [37]. Glycyrrhizin can reduce IL-6 cytokines in brain tissue, and it is also a cytokine closely related to sleep/deprivation. IL-6 can inhibit the synthesis of IL-1 and reduce IL-6 cytokines, and then IL-1 increases synthesis, which in turn enhances the patient's nonrapid eye movement sleep time [38].
The results of the study showed that topological analysis identi ed the key components of quercetin, kaempferol, β-sitosterol, 7-methoxy-2-methyl iso avones, luteolin, naringenin and other key components in CHSGP as the disease. It was the material basis for the treatment of insomnia and depression with the same treatment of different diseases. It had the largest number of targets and was also its core compound. Studies have shown that quercetin enhanced the hypnotic activity of pentobarbital in a dose-dependent manner by prolonging sleep time [39] and reducing anxiety and depression-like behaviors, immune dysfunction and brain oxidative stress [40], related to its improvement of neurotrophic function and antiin ammatory effects [41]. Kaempferol has a wide range of pharmacological activities, which can reduce oxidative stress and proin ammatory cytokines, inhibit vascular endothelial in ammation [42], and have an effective antidepressant effect [26,43]. Studies have found [44] that some Chinese herbal extracts contain β-sitosterol, which has anti-in ammatory, sedative and hypnotic activities [45]. Luteolin is a widely used avonoid compound that has anti-anxiety and sleep-promoting effects [46]. Naringenin can activate Sirt1, enhance antioxidant capacity by reducing oxidative stress, and effectively improve in ammation and dopamine levels [47], thereby alleviating the pain caused by chronic sleep deprivation [48].
According to the drug-disease target intersection, CHSGP can treat insomnia and depression at the same time through 113 target proteins in different diseases. According to the degree value and the betweenness centrality, PPI analysis network topology analysis screened AKT1 and IL-6, and 42 core targets, such as IL1B, CASP3, and MAPK3, were mainly related to immunity, in ammation, oxidative stress, and neovascularization. AKT1 is involved in the regulation of processes such as metabolism, proliferation, growth and angiogenesis [49]. IL-6 is one of the most biologically active cytokines and can be called a sleepy mediator. Its circadian rhythm pattern re ects the steady-state drive of sleep [50], and increased IL-6 activity may cause depression by activating the hypothalamus-pituitary-adrenal axis or affecting neurotransmitter metabolism [51]. IL-1B protein is an important mediator of the in ammatory response, is involved in a variety of cell activities, such as cell proliferation, differentiation, and apoptosis, and is related to the diagnosis, speci c symptoms, and antidepressant treatment response of major depression [1]. Caspase-3 (CASP3) belongs to the cysteine aspartic protease (Caspase) family of proteases and plays an irreplaceable role in the apoptosis pathway. Hippocampal apoptosis caused by sleep deprivation includes the number of apoptotic cells, caspase-3 activation, and Bax and Bcl-2 regulation [52]. MAPK3, called mitogen-activated protein kinase 3, is an important part of the MAP kinase signal transduction pathway and plays an important role in the cellular response cascade caused by extracellular stimulation. Another study [53] showed that the expression of the AKT1, MAPK3 and IL-6 genes in anxiety patients often increased signi cantly. The MCODE cluster analysis in the PPI network screened 7 core genes, including FOS, GABRA2, GSK3B, PON1, APOB, CYP1B1, and ADRA1A. FOS, also known as c-Fos, is a proto-oncogene that participates in cell growth, differentiation, information transmission, animal learning and memory and maintains wakefulness. Studies have shown that [54] FOS protein is involved in the normal differentiation, growth, learning, memory and other processes of cells, and it is more highly expressed in the brain endothelium, hippocampus, and striatum. GABA is found only in the nervous tissues of animals. It is an important inhibitory neurotransmitter and plays a role in the development process. It participates in a variety of metabolic activities and has high physiological activity. The GABRA2 gene encodes the GABAA-2 subunit, which is an ionotropic receptor related to anxiety, depression, and other behavioral disorders (including drug dependence and schizophrenia). Studies have found that [55] a higher expression level of α2 may have antidepressant effects through the signal transduction of GABAAR containing α2.
PPI network and GO analysis showed that the strong interactions of these target genes affected the synaptic membrane and postsynaptic membrane by participating in biological processes such as drug response, regulation of blood vessel diameter, and oxidative stress response, which included the molecular functions of synaptic membrane, postsynaptic membrane and its components, and which involved cell functions such as neurotransmitter receptor activity, postsynaptic neurotransmitter receptor activity, G protein coupled amine receptor activity, etc. The molecular docking results showed that the molecular docking in compounds (kaempferol, luteolin, quercetin, 7-methoxy-2-methyl iso avone and beta-sitosterol) and ve target proteins (AKT1, IL1B, IL-6, FOS, GSK3B, and GABRA) matched well, and the binding energy was less than -6 kcal/mol, which could form stable complexes and indirectly veri ed that these compounds had a regulatory effect on targets such as AKT1.
This study is aimed at the high incidence of global depression and insomnia under the current situation of the COVID-19 epidemic, and it is di cult to nd effective prevention methods in clinical practice. At the same time, the network pharmacology of CHSGP provides us with an opportunity to explore the "Same Treatment for Different Diseases" potential pharmacological mechanism of CHSGP in treating insomnia and depression caused by the COVID-19 epidemic. Existing network pharmacology studies have screened 121 active ingredients and 15 depression-related targets of CHSGP from the database, mainly involving the regulation of neurotransmitters (serotonin, dopamine and adrenaline), the regulation of in ammatory mediators of TRP channels, calcium signaling pathway, cyclic adenosine monophosphate signaling pathway, and neural active ligand-receptor interaction, which are the channels through which signals play an antidepressant effect. It can have an antidepressant effect by improving neuronal plasticity, growth, transfer conditions and gene expression in neuronal cells and increasing the expression of gap junction proteins [16]. Animal experiments [62] con rmed that CHSGP can improve the depression-like behavior of rats exposed to chronic unpredictable mild stress (CUMS), possibly by inhibiting CHOP and caspase-12-mediated apoptosis of rat hippocampal cells. Although the research has some ndings, it still has many limitations. First, due to incomplete information, the role of some CHSGP compounds in "Same Treatment for Different Diseases" may be ignored. Second, it is di cult to screen speci c targets and key active ingredients directly from the complex cognitive CHSGP. Third, although CHSGP has antidepressant and sleep-improving effects in previous clinical studies and animal model experiments, its mechanism of action is still unclear, and further analysis and excavation are needed in subsequent clinical and animal experiments.
Therefore, it is necessary to further study the role and speci c targets and pathways of the important single components of the CHSGP in the process of treating insomnia and depression with the "Same Treatment for Different Diseases".

Conclusions
In summary, this study used network pharmacology to nd that the mechanism of CHSGP for insomnia and depression caused by the COVID-19 epidemic with "Same Treatment for Different Diseases" might be related to core targets such as AKT1, IL-6, IL1B, FOS, CASP3, MAPK3, and KEGG signaling pathways such as the AGE-RAGE pathway, neuroactive ligand-receptor interaction, and IL-17 signaling pathway. The results of molecular docking revealed that kaempferol, luteolin, quercetin and other compounds matched well with AKT1, IL1B, IL-6, FOS, GSK3B, GABRA and other target protein molecules. The overall analysis and veri cation of the relationship among CHSGP, insomnia and depression were the "Same Treatment for Different Diseases" relationship. In molecular biology research, in-depth pharmacological research and clinical application exploration of insomnia and depression provide a reference. However, this research is not perfect enough. It is necessary to carry out relevant clinical and animal experimental studies in the complex environmental background of the COVID-19 epidemic to verify its clinical e cacy and mechanism of action of CHSGP for insomnia and depression with the "Same Treatment for Different Diseases". Figure 1 The experimental ow of this study.