A Network Pharmacology Study of Chinese Medicine QiShenYiQi to Reveal Its Underlying Multi-Compound, Multi-Target, Multi-Pathway Mode of Action

Chinese medicine is a complex system guided by traditional Chinese medicine (TCM) theories, which has proven to be especially effective in treating chronic and complex diseases. However, the underlying modes of action (MOA) are not always systematically investigated. Herein, a systematic study was designed to elucidate the multi-compound, multi-target and multi-pathway MOA of a Chinese medicine, QiShenYiQi (QSYQ), on myocardial infarction. QSYQ is composed of Astragalus membranaceus (Huangqi), Salvia miltiorrhiza (Danshen), Panax notoginseng (Sanqi), and Dalbergia odorifera (Jiangxiang). Male Sprague Dawley rat model of myocardial infarction were administered QSYQ intragastrically for 7 days while the control group was not treated. The differentially expressed genes (DEGs) were identified from myocardial infarction rat model treated with QSYQ, followed by constructing a cardiovascular disease (CVD)-related multilevel compound-target-pathway network connecting main compounds to those DEGs supported by literature evidences and the pathways that are functionally enriched in ArrayTrack. 55 potential targets of QSYQ were identified, of which 14 were confirmed in CVD-related literatures with experimental supporting evidences. Furthermore, three sesquiterpene components of QSYQ, Trans-nerolidol, (3S,6S,7R)-3,7,11-trimethyl-3,6-epoxy-1,10-dodecadien-7-ol and (3S,6R,7R)-3,7,11-trimethyl-3,6-epoxy-1,10-dodecadien-7-ol from Dalbergia odorifera T. Chen, were validated experimentally in this study. Their anti-inflammatory effects and potential targets including extracellular signal-regulated kinase-1/2, peroxisome proliferator-activated receptor-gamma and heme oxygenase-1 were identified. Finally, through a three-level compound-target-pathway network with experimental analysis, our study depicts a complex MOA of QSYQ on myocardial infarction.


Introduction
Chinese medicine, an important branch of the healthcare system in China, has gradually gained popularity both at home and abroad [1]. Chinese medicine, a complex system with guidance from traditional Chinese medicine (TCM) theories, has proven to be especially effective to treat chronic and complex diseases. However, the underlying mechanisms of action (MOA) are rarely investigated systematically.
One of the consensus is that TCM produces its efficacy in a holistic way [2], taking the advantage of multi-target therapy being more effective in combating polygenic diseases than monotherapies [3]. With the advancements in understanding of pathobiology of human diseases, people find that most diseases are not simply caused by one single factor [4]. It is especially true for the complex multifactorial chronic diseases such as cardiovascular disease (CVD), cancer and diabetes [5,6]. The TCM treatments can be visualized as a complexity against complexity paradigm between multi-target therapy such as TCM and the complex biological networks of human diseases. The development of analytical tools such as systems biology [7], network biology [8] and network pharmacology [9,10] provide an opportunity to unravel the complex and holistic mechanisms of TCM in treating complex diseases. In today's post-genomic era, it has become much easier to obtain huge amounts of data related to the effect of drugs on transcriptome of target tissues. Data analysis becomes critical to identify the MOA of drugs. Recently, network models and network analysis of genomic data has been proven useful to translate the microarray information and identify potential targets [11,12] and enriched pathways involved in the MOA [13][14][15]. Tools such as Cytoscape have often been used to construct and analyze networks [16].
In this study, the differentially expressed genes (DEGs) were identified from myocardial infarction (MI) rat model treated with QSYQ, followed by constructing a CVD related compoundtarget-pathway network linking main compounds to those DEGs supported with literature evidences and the pathways that are functionally enriched in ArrayTrack. The associations of DEGs with CVD were evaluated based on information in database CHD@ZJU of its version 1.0 (http://tcm.zju.edu.cn/chd/) developed by our group [22]. We also manually collected the information of targets reported for 9 compounds (i.e. Ast, Cal, For, DSS, PCA, RSA, Rg1, Rb1 and R1) from literatures in PubMed. The judging criteria of a gene being a target is that a CVD related DEG can be found directly influenced by a compound of QSYQ in literature. Furthermore, as there were no reports on the targets of ENL, SDL and RDL, the potential targets of these three sesquiterpenes were proposed based upon gene expression data and further experimentally validated in this study. Finally, we constructed the compound-target-pathway network on anti-MI of QSYQ to decipher the underlying multi-compound, multi-target and multi-pathway mechanism.

Preparation of QiShenYiQi Extracts
QSYQ was prepared from four herbs, Radix Astragalus membranaceus (Chinse name Huangqi), Radix Salvia miltiorrrhiza (Chinse name Danshen), Panax notoginseng (Chinse name Sanqi), and Dalbergia odorifera T. Chen (Chinse name Jiangxiang). Briefly, crude drug Huangqi were crushed, for Huangqi, we added water and extracted with refluxing twice, then the extract were filtrated on gauzes. Filtrate was concentrated on a rotary evaporator. In the alcohol sedimentation step, we slowly added 95% ethanol to the filtrate to make alcohol concentration to about 70%. Cooled the solution at room temperature and the supernatant was filtered through cotton gauzes. After recycling the alcohol, the residue was concentrated again to form Huangqi extract. As for crude drug Danshen and Sanqi, they were extracted together with the same procedure as that of Huangqi. For Jiangxiang, we boiled crushed crude drug Jiangxiang with refluxing, and essential oil was extracted in an extraction apparatus. Roots of Sanqi, the trunk and dry heartwood of root from Jiangxiang have been used.

Animal Study and Gene Expression Data Analysis
Male Sprague-Dawley rats (,250 g) were obtained from Zhejiang Experimental Animal Center. The Animal Ethic Review Committees of Zhejiang University approved all experimental procedures.
The rats were maintained in air-conditioned rooms (temperature, 22-26uC; humidity, 40%-70%) with a 12-h light-dark cycle, and were acclimatized to the setting. To provide comfortable conditions for rats, each cage contained 5 rats, and the bedding material sawdust was renewed every 2 days. Rats were free to get fresh foods and water, and the water was renewed everyday. All surgery was performed under 10% chloral hydrate anesthesia (300 mg/kg), and all efforts were made to minimize suffering. Rat myocardial infarction was made by occlusion of left anterior descending coronary artery (LAD) according to Yamaguchi [23,24]. Rats post-operation were laid flat on an electric blanket to help maintain the body temperature of rats. And for those rats which were really weak post-operation, we helped them to recover heart beating with chest cardiac massage and recover breath by connecting the animals with rodent ventilator. For the LAD ligation operation, the mortality was about 40%. Six surviving rats after LAD ligation were randomly divided into myocardial infarction group (MI) and QSYQ-treated group. Three shamoperated rats were treated in the same ways as the rats in MI group but without LAD ligation, which were served as control group (Control). In clinical, QSYQ pills are administered orally administrated to patients with unstable angina and heart failure to prevent acute coronary syndrome. In accordance with clinical recommendations, QSYQ was given intragastrically (i.g.) at a dose of 105.6 mg/kg once a day. Rats in control, and MI groups were administered with saline. Rats were sacrificed after 7 days of i.g. administration under 10% chloral hydrate anesthesia (300 mg/ kg). Rats were fixed on a flat plate, and blood samples were collected by abdominal aortic method, then the chest was carefully dissected and the heart was cut quickly. Three tissue samples on the border between infarct and non-infarct area were dissected from left ventricles of each group. The tissue samples were stored at 280uC refrigerator.
Total RNA was extracted using TRIzol Reagent (Invitrogen) and purified using RNeasy Mini kit (QIAGEN), following manufacturers' protocols. RNA quality was evaluated using an Agilent 2100 Bioanalyzer and electrophoresis in 2% (w/v) agarose gels. Only RNA with RNA integrity number (RIN) greater than 7.0 and 28S/18S ratio greater than 0.7 was used for microarray analyses. Whole genome microarray analysis was performed using Affymetrix rat Genome 230 2.0 array. Briefly, total RNA were amplified, labeled, and purified using GeneChip 39IVT Express Kit (Affymetrix) according to the manufacturer's instructions to obtain biotin labeled cRNA. Array hybridization and wash were performed using GeneChip Hybridization, Wash and Stain Kit in Hybridization Oven 645 (Affymetrix) and Fluidics Station 450 (Affymetrix) following the manufacturer's instructions. Slides were scanned by GeneChip Scanner 3000 and Command Console Software 3.1 with default settings. Spike-in control transcripts were monitored to verify hybridization integrity. The dataset as CEL files was deposited to GEO and the access number is GSE54134. Significant genes were identified through an algorithm developed in-house [25].
As mentioned above, we validated the CVD-related DEGs by manually checking published literatures in PubMed (by April 10 th , 2013). The keywords used to search papers in PubMed were: The literatures related to pharmacology and activity of Ast, Cal, For, DSS, PCA, RSA, Rg1, Rb1 and R1 were manually read. A DEG was considered as a potential target of a compound if it was found to be affected directly by the compound in a paper. For each paper, the information of target, PubMed ID, and CVD relevancy were recorded. Detailed information on the references used in this study can be found in Table S1.

Experimental Target Validation
CVD-related DEGs were validated by manually checking published literature in PubMed (by April 10 th , 2013). However, there were no previous studies reporting the targets of ENL, SDL and RDL. These three sesquiterpenes were isolated from volatile oils from Jiangxiang, which was reported to have anti-inflammatory activities, and was also used as a penetration enhancer [26]. Extracellular signal-regulated kinase-1/2 (ERK1/2), peroxisome proliferator-activated receptor-gamma (PPARc) and heme oxygenase-1 (HO-1), which play important roles in inflammation, were identified as the potential targets of ENL and SDL. The antiinflammatory activity of these targets was validated using LPSstimulated RAW264.7 macrophage cells as models.

Cell Viability Analysis by MTT Assay
Murine macrophage cell line RAW264.7 was maintained in DMEM medium supplemented with 10% heat-inactivated fetal bovine serum at 37uC in 5% CO 2 . To determine the effect of ENL and SDL on cell viability, RAW264.7 cells were seeded in 96-well plates at concentrations of 5000 cells/well. Following overnight incubation, cells were treated with several concentrations of ENL, RDL and SDL for 24 h. Cell viability was determined by the MTT assay. For MTT assay, the media were replaced with fresh DMEM medium that containing 10% FBS, penicillin/streptomycin, and MTT (0.5 mg/ml). After 4 h of incubation at 37uC, the media were removed, and replaced with 150 mL/well of DMSO. After mixing at 400 rpm in Thermomixer comfort (Eppendorf) at Indo stands for indomethacin and was used as positive control. Each group was compared with model, and * stands for p,0.05, ** stands for p,0.01. Each assay was conducted in triplicate and repeated three times. ENL is trans-nerolidol; RDL is (3S,6R,7R)-3,7,11-trimethyl-3,6-epoxy-1,10-dodecadien-7-ol; SDL is (3S, 6S, 7R)-3,7,11trimethyl-3,6-epoxy-1,10-dodecadien-7-ol. doi:10.1371/journal.pone.0095004.g001 37uC for 10 min, absorbance at 550 nm was measured on a Bio-Tek micro-plate reader. Each assay was conducted in triplicate and repeated three times.

Nitric Oxide Assay
RAW 264.7 cells (4610 4 cells in 96-well microplate) were stimulated with 200 ng/ml LPS in 37uC CO 2 incubator for 24 h. Nitric oxide (NO) production was detected by measuring nitrite levels in the culture media using the Griess reagent according to the manufacturer's instructions (Beyotime Institute of Biotechnology, China). Absorbance was measured at 550 nm.
Effects of U0126, T0070907, and Pio on Nitric Oxide Secretion U0126, a selective ERK1/2 phosphorylation inhibitor, was used to incubate with RAW264.7 macrophages together with LPS (200 ng/ml) or compounds with different concentrations plus LPS (200 ng/ml). PPARc phosphorylation inhibitor T0070907 and PPARc agonist Pio were also used in the same way as that of U0126. The concentration of U0126, T0070907 and Pio were used at 10 mM. NO production was determined using Griess Reagent. Each assay was conducted in triplicate and repeated three times, and outliers were excluded.

Western Blot Analysis
For western blot assay, RAW264.7 macrophages were seeded into 60-mm dishes, cells were cultured in the 37uC CO 2 incubator overnight. The second day, cells were treated with different compounds. The cells were collected after 24 h treatment. Cells were gently rinsed with ice-cold PBS. Then, 300 mL of cold RIPA buffer (RIPA lysis buffer, Cat No. P0013B, Beyotime Institute of Biotechnology, Haimen, Jiangsu, China) plus PMSF (Cat No. ST506, Beyotime Institute of Biotechnology, Haimen, Jiangsu, China), complete Mini (Cat No. 04693124001, Roche) and PhosSTOP (Cat No. 04906845001, Roche) were added to each dish. All steps of protein extraction were carried out on ice. Cells were lysed completely within 10 min, and the protein lysate was transferred to 1.5 ml centrifuge tubes, then the tubes were centrifuged at 12000 rpm for 10 min at 4uC. The supernatant in each tube was carefully collected without disturbing the sediments. Protein concentrations in the collected supernatants were determined using BCA protein assay kit (

Network Construction and Network Analysis
The DEGs related to CVD and the main compounds were used as input in ArrayTrack v.3.5.0 to carry out pathway enrichment analysis, in which the MI or CVD associated pathways were selected for further analysis. Then, we constructed compoundtarget-pathway network and compound-pathway network in Cytoscape [16]. A compound and a target were connected with a solid line if the link was reported in at least one paper ($1 literature) when they were; and the edge between target and pathway was constructed using the relationship found in Array-Track v.3.5.0 pathway enrichment analysis [27].

Statistical Analysis
All experiments were performed three times in triplicate. Results are expressed as mean 6 SD. Two-tailed unpaired Student's t tests were used to compare two distinct groups. When more than three groups were compared, One-Way ANOVA was performed between different treatment groups using PASW Statistics 18 (SPSS Inc., USA). A p value ,0.05 was considered to be statistically significant.

DEGs Analysis
312 genes (442 probe sets) in CHD@ZJU V1.0 database were found to be differentially expressed, out of which 83 genes with reverse rate (RR).0.5 and fold change (FC).1.1 after the QSYQ treatment were selected for further analysis (Table S2). Combined with the targets information about the compounds collected manually from published papers, 52 genes were included in this study, with additional information about compound, target, pathway, category and Fisher P-value ( Table S3). The number of targets and the pathways of the compounds are listed in Table 1.

Effect of ENL, RDL and SDL on RAW264.7 Macrophages Viability
After 24 h co-incubation with different concentrations (50, 20, 10, 5 mM) of ENL, RDL and SDL, MTT assay results showed that there was no obvious cytotoxicity of the three compounds (Fig. 1B) at these concentrations. Concentrations #50 mM were used for subsequent experiments.

Effects of ENL and SDL on NO Production in LPSstimulated RAW264.7 Macrophages
Two sesquiterpene compounds (ENL and SDL) from the volatile oil of Dalbergia odorifera T. Chen were able to suppress NO secretion from LPS-induced RAW264.7 macrophages. Both ENL and SDL significantly inhibited the NO secretion by LPSstimulated RAW264.7 cells at 50 mM (p,0.01 for ENL, p,0.05 for SDL compared to model) (Fig. 1C). However, RDL at all concentrations had no obvious inhibition effect on NO secretion by LPS-stimulated RAW264.7 cells. Indomethacin at 25 mM was used as the positive control. More details of indomethacin, NEL and SDL on NO production can be found in Fig. 1C.
Effects of U0126, T0070907, and Pio on Nitric Oxide Secretion U0126, an antagonist of ERK1/2 phosphorylation, Pio, a PPARc agonist, and T0070907, a PPARc phosphorylation inhibitor, were able to inhibit NO secretion by LPS-induced RAW264.7 macrophages. So these three tool drugs were used to validate the effects of ENL and SDL. 200 ng/ml LPS significantly induced NO secretion ( 11 p,0.01, vs control by student's t test) (Fig. 3). Combined-treatment with either of the three drugs, U0126, T0070907, and Pio, significantly strengthened the effect of ENL on suppressing NO secretion except 20 mM ENL plus Pio (Fig. 3A, B, C). U0126 also significantly enhanced the effect of SDL on inhibition of NO production (##p,0.01, vs. the group with SDL treated alone) (Fig. 3D).

Network Construction and Pathway Analysis
Combining the relationships in Table S3 and the relationships from the experimental validation part, that is the potential targets such as PPARG (targets of ENL and SDL), HMOX1 (targets of ENL, RDL and SDL), MAPK3 and MAPK1 (targets of ENL and SDL) were also used to construct the network (Fig. 4). Then the paired relationships information (link between compound and target genes, link between target gene and pathways) was imported to Cytoscape software, and a compound-target-pathway network was constructed (Fig. 4). The three-level network consisted of 84 nodes (12 compounds, 55 genes and 17 pathways) and 230 edges. The detailed information of the compound names, gene names, pathway names and their category, and Fisher P-values can be found in Table S3. Through the visualization and network analysis of the relationships among compound, target gene and pathway in the constructed network, 52 DEGs were affected by at least one compound, in which 14 genes (DUSP1, EPO, FGF1, FLT1, FN1, HIF1A, LDHA, LDHB, PLAT, PRKCB, RELA, VEGFA and VWF) were the potential targets of the compounds based on the CVD-related literatures in PubMed. Among the 52 genes, RELA was influenced by 9 compounds. RELA is associated with cell growth and death, immune system, and endocrine system. IL6, PTGS2 and VEGFA were affected by 6 compounds, which maybe related with inflammation and angiogenesis. GPX1, HIF1A, IL1B, LDHA, LDHB and PPARG were affected by 5 compounds. FN1 and PLAT were influenced by 4 compounds. While five targets were influenced by 3 compounds, nine targets were influenced by 2 compounds, and the rest twenty six were affected by only one compound.
Using the 52 DEGs, 16 KEGG pathways were enriched in the ArrayTrack v.3.5.0 pathway enrichment analysis ( Table 2), in which apoptosis signaling pathway is a critical pathway in regulating cell death. Most of the pathways, such as Cytokinecytokine receptor interaction, NOD-like receptor signaling pathway, PPAR signaling pathway, Toll-like receptor signaling pathway, MAPK signaling pathway, Hematopoietic cell lineage, Cytosolic DNA-sensing pathway, Complement and coagulation cascades and Arachidonic acid metabolism, are either directly involved in immune and inflammation or play important roles there. In addition, there are several energy metabolism associated pathways including PPAR signaling pathway, Adipocytokine signaling pathway, Pyruvate metabolism. The Glutathione metabolism is antioxidant related, Focal adhesion plays role in cell motility and structure, VEGF signaling pathway has function in vasodilation, Complement and coagulation cascades is related with coagulation, and mTOR signaling pathway plays a role in hypoxia induced angiogenesis.

Discussion
Based on the results of microarray data and published literature information available in PubMed, a compound-target-pathway network (Fig 4) was constructed for the multi-compound-multitarget-multi-pathway mode of action (MOA) of the Chinese medicine QSYQ. The cardioprotective or anti-MI function of QSYQ is complex, involving modulation of groups of targets and a number of associated pathways. Cardiomyocytes apoptosis occurs promptly following acute myocardial infarction (AMI), and apoptosis pathway (including the mitochondrial death pathway) plays a pivotal role in this pathological process [28][29][30]. Inflammation accompanies from the initial stage of AMI and participates in the development of ischemia injury, one of the mechanisms is post-AMI hypoxia elicited inflammatory response [31]. Among the 16 pathways, nine are immune and inflammation related, including Cytokine-cytokine receptor interaction, NODlike receptor signaling pathway [32,33], PPAR signaling pathway [34][35][36], Toll-like receptor signaling pathway [37,38], MAPK signaling pathway [39], Hematopoietic cell lineage, Cytosolic DNA-sensing pathway [40], Complement and coagulation cascades [41,42] and Arachidonic acid metabolism [43]. Compounds from Dalbergia odorifera suppressed NO production in our in vitro target validation experiments. Anti-inflammation effect of QSYQ has been reported. QSYQ can reduce high sensitivity C reactive protein level in atherosclerotic rabbit to prevent the progress of atherosclerosis by inhibiting inflammatory reaction [44]. Also, it was reported that QSYQ can inhibit inflammatory effect through reducing TNF-a and IL-6 levels in rat after acute myocardial infarction [45,46]. In clinical settings, QSYQ can improve cardiac function through anti-inflammatory effects by reducing TNF-a and IL-6 levels both in AMI and chronic heart failure patients [47,48]. QSYQ could decrease the levels of high sensitivity C reactive protein, plasminogen activator inhibitor-1 and endothelin-1 in acute coronary syndrome patients [49].
Myocardial energy metabolism is another important event during and post-AMI, including fatty acid and glucose utilization. Acetyl-CoA is common end product of fatty acid, glucose and pyruvate to supply ATP. PPAR signaling pathway may prevent ischemic insult involving anti-inflammatory effect and promoting glucose utilization [36]. Adiponectin is a key adipocytokine having beneficial effect to facilitate fatty acid beta-oxidation and increase glucose utilization [50].
The potential targets of those 9 main compounds from Huangqi, Danshen, Sanqi were validated using CVD-ralated literatures, and there was no reference reporting the targets of two main sesquiterpenes from Jiangxiang. In our experimental study, we investigated the anti-inflammatory effects of two main sesquiterpenes in volatile oils of Jiangxiang, i.e. ENL and SDL. SDL was newly identified from Jiangxiang in our previous study [21]. To the best of our knowledge, the pharmacological activities of ENL and SDL, and their potential protein targets have not been previously reported. We used nitric oxide assay, Western blot technology and small molecular compounds specific to the proteins of interest to validate the identified targets. As discussed above, ENL and SDL could suppress NO production in LPSstimulated RAW264.7 macrophages.
We propose herein that the three sesquiterpene compounds (especially ENL and SDL) from volatile oil of Jiangxiang may suppress LPS-stimulated NO production from RAW264.7 macrophages at least partially via inhibition of ERK1/2 and PPARc phosphorylation, and increasing expression of PPARc and HO-1, in which NF-kB, ROS and iNOS maybe involved (Fig. S1.).
Chinese medicine treats diseases in a holistic way to make body recovery from imbalanced state by modulating a complex MOA including multiple pathways. Therefore, rodent animal models mimicking human disease are more suitable to evaluate the therapeutic effects of TCM and discover its MOA than cell culture. In our study, the underlying MOA of a clinical effective Chinese medicine QSYQ is systematically investigated combining mRNA expression profiles at molecular level on a rat myocardial infarction model. Using rat model is more convenient and easy to . Multi-compound-multi-target-multi-pathway network of TCM QSYQ on treating myocardial infarction. There were 52 differentially expressed genes which were validated in published literatures. Node size of targets and pathways are proportional to the number of compounds associated with the node. This network depicted a clear three-level structure of the mode of action of main compounds of QSYQ, that is multi-compound regulated multi-pathways through modulating of groups of genes to show their pharmacological and clinical effects. Compound was linked to targets, while target were linked to pathways, a certain compound linked indirect link to pathways through certain targets. Diamond represents compounds; Circle represents target genes; Vee represents CVD related pathways. The edge went through compound to target, and target to pathway. doi:10.1371/journal.pone.0095004.g004 perform research to get rich information about the drug investigated both at phenotype and molecular base. On the other hand, cell culture model can be used to mimic a certain biological step of the drug investigated. Cell culture experiment has its merit as it is much easier to manipulate and to set investigation conditions. The results from animal experiments can be confirmed and studied in depth within cell culture experiments. However the results from animal studies should be carefully interpreted and concluded to explain the potential MOA of drug in treating human disease.
For the additional predicted targets as shown in Fig. 4, it is also important to validate them for a better understanding of the underlying MOA of QSYQ. Cell culture models play an important role in target validation. For example, inflammation related targets validation can be done with LPS-stimulated macrophage RAW264.7 cell culture model. TNF-a injured endothelial cell models are also used for anti-inflammatory related targets validation. The other two important processes mentioned in our study, apoptosis and energy metabolism can be studied with H9C2 cell line and primary neonatal rat heart cells to investigate cardioprotective effects and mechanisms of active compounds and drugs. Hypoxia or hypoxia-reoxygenation conditions are used to mimic in vivo ischemia or ischemia-reperfusion conditions. In addition, siRNA technology and target specific small molecular tool drugs could help to further confirm the validation results. Secondly, the targets validated at the in vitro experiments need further in vivo experiments confirmation. Rat MI model can be used to investigate the contributions of several targets to the anti-MI effects of compounds. The mRNA and protein expression profiles of specific targets after the intervention of single compounds or compounds combinations are tested using RT-PCR and western blot technologies. Finally, transgenetic mice models (both knock out and knock in) if available would provide more conclusive information for validating predicted targets with regard to cardioprotective effects.
The approach described in this paper provided an example on discovering the multi-compound, multi-target, multi-pathway MOA of a clinically effective Chinese medicine. A few steps have to be followed to ensure successful results. First, it is important to choose a clinically effective Chinese medicine with appropriate animal model available. Second, there should be reasonable understanding of the chemical compositions of the Chinese medicine. Third, in vitro experiments using cell culture models should be applied to test the pharmacological protective effects of potential compounds. Animal models (usually rodents) mimicking human disease are also used to confirm the effectiveness of potential compounds. Through these investigations, we understand the efficacious material basis of the Chinese medicine. Finally, network pharmacology analysis can be applied to integrate all these information to gain a full understanding of the MOA of the Chinese medicine.
In conclusion, the potential targets of 12 major compounds of QSYQ were identified by integrating gene expression data analysis and literature mining. These compounds may regulate multiple pathways through modulation of groups of genes to exert their anti-MI effects, involving biological processes of anti-apoptosis, anti-inflammation, antioxidant, anti-coagulation, energy utilization facilitation and angiogenesis promotion. Potential targets of 9 main compounds were validated by CVD-related references, while potential targets of three sesquiterpene compounds were further experimentally validated for their roles in mediation of inflammation. Figure S1 Proposed molecular mechanisms of the anti-inflammatory effects of two sesquiterpene compounds from volatile oil of Dalbergia odorifera T. Chen. ENL is proposed as an inhibitor of ERK1/2 phosphorylation and PPARc phosphorylation, an agonist of PPARc, via which ENL could suppress NO production from LPS-stimulated RAW264.7 macrophages. SDL is also an antagonist of ERK1/2 phosphorylation, through which SDL may attenuate NO secretion. HO-1 may also participated in the regulation process of these two naturally occurring products in anti-inflammation. ENL is trans-nerolidol; SDL is (3S, 6S, 7R)-3,7,11-trimethyl-3,6-epoxy-1,10-dodecadien-7-ol. (TIF)