networkpharmacologystrategytoinvestigatethe...

Research ArticleNetwork Pharmacology Strategy to Investigate thePharmacological Mechanism of HuangQiXiXin Decoction onCough Variant Asthma and Evidence-Based MedicineApproach Validation

Qingqing Xia ,1 Mingtao Liu,2 Hui Li,2 Lijun Tian ,3 Jia Qi ,4 and Yufeng Zhang 1

1Department of Respiratory Medicine, Jiangyin Hospital of Traditional Chinese Medicine,Jiangyin Hospital Affiliated to Nanjing University of Chinese Medicine, Jiangyin, Jiangsu 214400, China2Department of Respiratory Medicine, Binzhou People’s Hospital, Binzhou, Shandong 256610, China3Department of Critical Care Medicine, Nantong -ird People’s Hospital, Nantong University, Nantong, Jiangsu 226001, China4Department of Pharmacy, Xin Hua Hospital Affiliated to Shanghai Jiao Tong University School of Medicine,Shanghai 200092, China

Correspondence should be addressed to LijunTian; [email protected], JiaQi; [email protected], andYufengZhang;[email protected]

Received 19 April 2020; Revised 11 September 2020; Accepted 24 September 2020; Published 30 October 2020

Academic Editor: Haroon Khan

Copyright © 2020 Qingqing Xia et al. 5is is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Objective. To investigate the pharmacological mechanism of HuangQiXiXin decoction (HQXXD) on cough variant asthma (CVA)and validate the clinical curative effect.Methods.5e active compounds and target genes of HQXXDwere searched using TCMSP.CVA-related target genes were obtained using the GeneCards database. 5e active target genes of HQXXD were compared withthe CVA-related target genes to identify candidate target genes of HQXXD acting on CVA. A medicine-compound-targetnetwork was constructed using Cytoscape 3.6.0 software, and a protein-protein interaction (PPI) network was constructed usingthe STRING database. Gene ontology (GO) function enrichment and Kyoto Encyclopedia of Genes and Genomes (KEGG)pathway enrichment analysis were performed using RGUI3.6.1 and Cytoscape 3.6.0. We searched the main database for ran-domized controlled trials of HQXXD for CVA. We assessed the quality of the included studies using the Cochrane Reviewers’Handbook. A meta-analysis of the clinical curative effect of HQXXD for CVA was conducted using the Cochrane Collaboration’sRevMan 5.3 software. Results. We screened out 48 active compounds and 217 active target genes of HQXXD from TCMSP. 5e217 active target genes of HQXXD were compared with the 1481 CVA-related target genes, and 132 candidate target genes forHQXXD acting on CVAwere identified.5emedicine-compound-target network and PPI network were constructed, and the keycompounds and key targets were selected. GO function enrichment and KEGG pathway enrichment analysis were performed.Meta-analysis showed that the total effective rate of the clinical curative effect was significantly higher in the experimental groupthan the control group. Conclusion. 5e pharmacological mechanism of HQXXD acting on CVA has been further determined,and the clinical curative effect of HQXXD on CVA is remarkable.

1. Introduction

Cough variant asthma (CVA) has the same pathogenesis asasthma and is characterized by persistent chronic airwayinflammation and airway hyperresponsiveness [1]. 5eclinical manifestations of CVA are chronic cough, which is

not always accompanied by wheezing. Chronic cough isoften the only symptom of CVA [2]. CVA is the mostcommon cause of chronic cough in China [3] and the secondmost common cause of chronic cough in Korea [4].

Inhalation of corticosteroids and long-acting β2-agonistsis recommended for the treatment of CVA [5, 6]. However,

HindawiEvidence-Based Complementary and Alternative MedicineVolume 2020, Article ID 3829092, 15 pageshttps://doi.org/10.1155/2020/3829092

mailto:[email protected]:[email protected]:[email protected]://orcid.org/0000-0001-5287-0639https://orcid.org/0000-0002-0306-0709https://orcid.org/0000-0002-4748-0959https://orcid.org/0000-0002-3154-7979https://creativecommons.org/licenses/by/4.0/https://doi.org/10.1155/2020/3829092

there is a high recurrence rate of the disease, and the long-term effect of these treatments is not ideal. Recently, Chineseherbal medicine has been used in the treatment of CVA[7, 8]. HuangQiXiXin decoction (HQXXD), added or re-duced from the traditional Chinese medicine (TCM) pre-scriptions Yupingfeng powder and Zhisou powder, mainlyconsists of Huangqi, Xixin, Jingjie, Fangfeng, Huangqin,Baizhu, Fuling, Chantui, Banxia, Baibu, and Gancao. It hasbeen reported that HQXXD is effective in the treatment ofCVA; however, our previous study showed that the totaleffective rate of HQXXD in the treatment of CVA was notsignificant compared with the control group, although therewas some evidence of a clinical effect [9].

In this study, we used network pharmacology and anevidence-based medicine approach to determine the phar-macological mechanism of HQXXD on CVA and validatethe clinical curative effect. First, we identified the activecompounds and target genes of the monarch and ministerherbs in HQXXD. Next, we used the CVA-related targetgenes and the active target genes of HQXXD to construct themedicine-compound-target network and protein-proteininteraction (PPI) network. Subsequently, gene ontology(GO) function enrichment and Kyoto Encyclopedia ofGenes and Genomes (KEGG) pathway enrichment wereperformed. Finally, we screened randomized controlledtrials (RCTs) that investigated the clinical curative effect ofHQXXD on CVA and performed a meta-analysis to validatethe clinical curative effect.

2. Materials and Methods

2.1. Screening of Active Compounds in HQXXD. Huangqi,Xixin, Jingjie, and Fangfeng are monarch and ministerherbs, which are regarded as the main active herbs inHQXXD, and have the Latin names Radix Astragali (RA),Herba cum Radix Asari (HRA), Herba Schizonepetae (HS),and Radix Saposhnikoviae (RS), respectively. We have de-termined that these four herbs are the main components inHQXXD.5e compounds in these four herbs were obtainedusing the Traditional Chinese Medicine Systems Pharma-cology Database and Analysis Platform (TCMSP) (http://tcmspw.com/tcmsp.php) [10]. We used the parameters oforal bioavailability (OB)≥ 30% and drug-likeness (DL)≥0.18 to screen for active compounds, and OB and DL arecommonly used in network pharmacology. OB relates to therate and extent of which the active component or compo-nents of the drug are absorbed into the body. DL indicatesthe degree to which a compound contains specific functionalgroups or has the same or similar physical characteristics tomost drugs.

2.2. Screening of the Target Genes of Active Compounds.We retrieved the corresponding target genes of activecompounds from TCMSP. After importing the target genesinto the UniProt knowledgebase (http://www.uniprot.org/)and setting the search format as “Homo sapiens,” the humanofficial gene symbols were identified, and the active targetgenes of HQXXD were screened out.

2.3. Acquisition of CVA-Related Target Genes and Identifi-cation of Candidate Target Genes of HQXXD Acting on CVA.Using “cough variant asthma” as the keyword in the Gen-eCards database (https://www.genecards.org/), we searchedand acquired CVA-related target genes. 5en, the activetarget genes of HQXXD were compared with the CVA-related target genes, and the intersecting genes were definedas the candidate target genes for HQXXD acting on CVA.

2.4. Construction of the Medicine-Compound-Target Networkand Selection of Key Compounds. We used Cytoscape 3.6.0software and its Network Analyzer tool function to constructand analyze a medicine-compound-target network [11].Nodes represented compounds and target genes, and therelationships between them were represented as edges. 5ekey compounds in HQXXD acting on CVA were selectedaccording to the degree of connection between the com-pound and the target gene.

2.5. Construction of the PPI Network and Selection of KeyTargets. 5e candidate target genes were introduced intothe STRING database (https://string-db.org/) to constructa PPI network of HQXXD acting on CVA. 5e researchspecies was defined as “Homo sapiens,” the lowest inter-action score was set to 0.4, and the rest of the parameterswere set to the default settings to obtain the PPI network.Using Cytoscape 3.6.0 software and its Network Analyzertool, the PPI network was subjected to topology analysis,and the key targets of HQXXD acting on CVA were thenselected according to the degree values of each target genein the PPI network.

2.6. GO Function Enrichment and KEGG Pathway EnrichmentAnalysis. RGUI3.6.1 and org.Hs.eg.db package were appliedto obtain the Entrez IDs of candidate target genes. 5en, weused RGUI, the clusterProfiler package, and Cytoscape toperform GO function enrichment analysis, which includedbiological process (BP), molecular function (MF), and cel-lular component (CC) analysis, and KEGG pathway en-richment analysis [12]. Adjusted P< 0.05 was considered asstatistically significant.

2.7. Validation of the Clinical Curative Effect of HQXXD onCVA

2.7.1. Screening of RCTs Investigating the Clinical CurativeEffect of HQXXD on CVA. We searched using the terms“HuangQiXiXin decoction” and “cough variant asthma” inPubMed, the Cochrane Central Register of Controlled Trials,the Chinese Biomedical Literature database, Chinese Na-tional Knowledge Infrastructure (CNKI), the ChongqingVIP database (CQVIP), and Wanfang Data from the es-tablishment of each database to January 15, 2020. When theterms were not included in the titles and abstracts, if theterms were present in the full text, the full-text paper wasalso screened.

2 Evidence-Based Complementary and Alternative Medicine

http://tcmspw.com/tcmsp.phphttp://tcmspw.com/tcmsp.phphttp://www.uniprot.org/https://www.genecards.org/https://string-db.org/

5e inclusion criteria included that the study wasdesigned as a RCT, the participants had symptoms that werein accordance with a diagnosis of CVA, HQXXDwas used inthe experimental group, the control group used conven-tional therapy without TCM therapy, and the clinical cu-rative effect was the main outcome measure. Exclusioncriteria included studies where there were incomplete dataon the clinical curative effect. For reports of the same study,the earliest published report was included and the rest wereexcluded.

2.7.2. Data Extraction and Quality Assessment. Two re-viewers independently extracted the information from theincluded studies. 5e main information included the firstauthor, year of publication, number of patients in eachgroup, methods of intervention in the experimental andcontrol groups, and outcome data of the clinical curativeeffect.

We used the Cochrane Reviewers’ Handbook forguidelines to assess the risk of bias. Seven criteria were used:random sequence generation; allocation concealment; pa-tient blinding; assessor blinding; incomplete outcome data;selective outcome reporting; and other risks of bias [13].

5ese main data were input into the Cochrane Col-laboration’s RevMan 5.3 software for meta-analysis to an-alyze the clinical efficacy of HQXXD on CVA.

2.8. Statistical Analysis. Some of the statistical analyses wereperformed automatically using the bioinformatic tools of theplatforms and software mentioned above. In GO functionenrichment and KEGG pathway enrichment, adjustedP< 0.05 was considered as statistically significant. WhenRevMan 5.3 software was used for meta-analysis, the totalefficiency rates were expressed as the odds ratio (OR) with95% confidence interval (CI). Heterogeneity was assessed bythe Q test (P value and I2), and P< 0.10 indicated hetero-geneity across studies. Studies with I2< 50% were consideredto have no heterogeneity and those with I2≥ 50% wereconsidered to have heterogeneity. If no heterogeneity wasdetected, the fixed effects model was used as the poolingmethod; otherwise, the random effect model was consideredto be the appropriate choice [14, 15]. P< 0.05 was consideredas statistically significant.

3. Results

3.1. ScreenedActiveCompounds inHQXXD. We obtained 87compounds from RA, 192 compounds from HRA, 159compounds from HS, and 173 compounds from RS inTCMSP (Supplementary files 1A, B, C, and D). Setting theconditions as OB≥ 30% and DL≥ 0.18, 20 active compoundsfrom RA, 8 active compounds from HRA, 11 active com-pounds from HS, and 18 active compounds from RS wereselected. Finally, with exclusion of duplications, 53 activecompounds from HQXXD were identified. 5e basic in-formation on the active compounds from HQXXD is shownin Table 1.

3.2. Screened Active Target Genes of HQXXD. 5e corre-sponding target genes of the 53 active compounds were alsoobtained in TCMSP (Supplementary file 2A). Using theUniProt knowledgebase, the corresponding gene symbolswere screened by setting the format as “Homo sapiens”(Supplementary file 2B (available here)). Four compounds(MOL000374, MOL000398, MOL000438, and MOL001506)did not have targets, and one compound (MOL000439) didnot have a compatible gene symbol. Finally, 217 active targetgenes of 48 active compounds from HQXXD were identified(Supplementary file 2C).

3.3. Acquired CVA-Related Target Genes and IdentifiedCandidate Target Genes of HQXXD Acting on CVA.Using “cough variant asthma” as the keyword in the Gen-eCards database, 1481 CVA-related target genes were ac-quired (Supplementary files 3A, B).

5e 217 active target genes of HQXXD were comparedwith the 1481 CVA-related target genes, and 132 candidatetarget genes of HQXXD acting on CVA were identified(Figure 1, Table 2).

3.4. Constructed Medicine-Compound-Target Network andSelected Key Compounds. 5e 132 overlapping candidatetarget genes corresponded to 47 active compounds(MOL000433 had correspondence to an overlapping can-didate target gene), including 15 compounds from RA, 8compounds from HRA, 10 compounds from HS, and 16compounds from RS (Supplementary file 4).

5en, a network of medicine-compound-target wasconstructed using the Cytoscape software and analyzed bythe Network Analyzer tool. 5ere were 184 nodes (4medicine nodes, 47 compound nodes, 132 target gene nodes,and 1 CVA node) and 653 edges in the network (Figure 2).5e top 14 compounds ranked by degree in the networkwere MOL000098, MOL000006, MOL000422, MOL000173,MOL000378, MOL000358, MOL000392, MOL000371,MOL000449, MOL000354, MOL000380, MOL000417,MOL001460, and MOL002962 (Table 3), which could beconsidered the key compounds in HQXXD acting on CVA.

3.5. Constructed PPI Network and Selected Key Targets.To further study the mechanism of HQXXD acting on CVA,132 overlapping target genes were mapped into the STRINGdatabase, and the PPI network was obtained. When we setthe lowest interaction score to 0.40, 132 target proteins in thenetwork had interactions, and 2373 edges represented theinteractions between the proteins (Supplementary file 5,Figure 3). 5e top 20 target genes ranked by degree in thenetwork are shown in Table 4, which can be considered thekey targets of HQXXD acting on CVA.

3.6.GOFunctionEnrichmentandKEGGPathwayEnrichmentAnalysis. 5e Entrez IDs of candidate target genes wereobtained using RGUI and org.Hs.eg.db (Table 2). 5en, GOfunction enrichment and KEGG pathway enrichmentanalysis were performed using RGUI and clusterProfiler.

Evidence-Based Complementary and Alternative Medicine 3

Table 1: Basic information on the active compounds in HQXXD.

Medicine Active compound CompoundIDOB(%) DL

RA Mairin MOL000211 55.38 0.78RA Jaranol MOL000239 50.83 0.29RA Hederagenin MOL000296 36.91 0.75

RA (3S,8S,9S,10R,13R,14S,17R)-10,13-Dimethyl-17-[(2R,5S)-5-propan-2-yloctan-2-yl]-2,3,4,7,8,9,11,12,14,15,16,17-dodecahydro-1H-cyclopenta[a]phenanthren-3-ol MOL000033 36.23 0.78

RA Isorhamnetin MOL000354 49.6 0.31RA 3,9-Di-O-methylnissolin MOL000371 53.74 0.48RA 5′-Hydroxyiso-muronulatol-2′,5′-di-O-glucoside MOL000374 41.72 0.69RA 7-O-Methylisomucronulatol MOL000378 74.69 0.3RA 9,10-Dimethoxypterocarpan-3-O-β-D-glucoside MOL000379 36.74 0.92RA (6aR,11aR)-9,10-Dimethoxy-6a,11a-dihydro-6H-benzofurano[s3,2-c]chromen-3-ol MOL000380 64.26 0.42RA Bifendate MOL000387 31.1 0.67RA Formononetin MOL000392 69.67 0.21RA Isoflavanone MOL000398 109.99 0.3RA Calycosin MOL000417 47.75 0.24RAHRA Kaempferol MOL000422 41.88 0.24

RA FA MOL000433 68.96 0.71RA (3R)-3-(2-Hydroxy-3,4-dimethoxyphenyl)chroman-7-ol MOL000438 67.67 0.26RA Isomucronulatol-7,2′-di-O-glucosiole MOL000439 49.28 0.62RA 1,7-Dihydroxy-3,9-dimethoxypterocarpene MOL000442 39.05 0.48RAHS Quercetin MOL000098 46.43 0.28

HRA 4,9-Dimethoxy-1-vinyl-b-carboline MOL012140 65.3 0.19HRA Caribine MOL012141 37.06 0.83HRA Cryptopin MOL001460 78.74 0.72HRA Sesamin MOL001558 56.55 0.83

HRA [(1S)-3-[(E)-But-2-enyl]-2-methyl-4-oxo-1-cyclopent-2-enyl](1R,3R)-3-[(E)-3-methoxy-2-methyl-3-oxoprop-1-enyl]-2,2-dimethylcyclopropane-1-carboxylate MOL002501 62.52 0.31

HRA (3S)-7-Hydroxy-3-(2,3,4-trimethoxyphenyl) chroman-4-one MOL002962 48.23 0.33HRA ZINC05223929 MOL009849 31.57 0.83HS Schizonepetoside B MOL011849 31.02 0.28HS Schkuhrin I MOL011856 54.45 0.52HS Diosmetin MOL002881 31.14 0.27HSRS Sitosterol MOL000359 36.91 0.75

HS 5,7-Dihydroxy-2-(3-hydroxy-4-methoxyphenyl)chroman-4-one MOL005100 47.74 0.27HS Luteolin MOL000006 36.16 0.25HSRS Beta-sitosterol MOL000358 36.91 0.75

HS Stigmasterol MOL000449 43.83 0.76HS Supraene MOL001506 33.55 0.42HS Campest-5-en-3beta-ol MOL005043 37.58 0.71

RS (2R,3R)-3-(4-Hydroxy-3-methoxy-phenyl)-5-methoxy-2-methylol-2,3-dihydropyrano [5,6-h][1,4]benzodioxin-9-one MOL000011 68.83 0.66

RS 11-Hydroxy-sec-o-beta-d-glucosylhamaudol_qt MOL011730 50.24 0.27RS Anomalin MOL011732 59.65 0.66RS Divaricatacid MOL011737 87 0.32RS Divaricatol MOL011740 31.65 0.38RS Ammidin MOL001941 34.55 0.22RS Ledebouriellol MOL011747 32.05 0.51RS Phelloptorin MOL011749 43.39 0.28RS 5-O-Methylvisamminol MOL011753 37.99 0.25RS Phellopterin MOL002644 40.19 0.28RS Wogonin MOL000173 30.68 0.23RS Mandenol MOL001494 42 0.19RS Isoimperatorin MOL001942 45.46 0.23RS Prangenidin MOL003588 36.31 0.22RS Methyl icosa-11,14-dienoate MOL007514 39.67 0.23RS Decursin MOL013077 39.27 0.38


GO BP function enrichment analysis showed that thecandidate target genes of HQXXD acting on CVA weresignificantly enriched in response to toxic substance, re-sponse to molecule of bacterial origin, response to lipo-polysaccharide, response to extracellular stimulus, responseto nutrient levels, response to metal ion, response to reactiveoxygen species, response to oxidative stress, cellular re-sponse to oxidative stress, response to antibiotic, and so on(Supplementary file 6A). 5e top 20 GO BP enrichmentsranked by the adjusted P value are shown in Figure 4(a).

GO MF function enrichment analysis showed that thecandidate target genes of HQXXD acting on CVA that weresignificantly enriched in protein heterodimerization activity,cytokine receptor binding, G protein-coupled amine re-ceptor activity, nuclear receptor activity, transcription factoractivity direct ligand-regulated sequence-specific DNAbinding, oxidoreductase activity acting on paired donorswith incorporation or reduction of molecular oxygen, hemebinding, tetrapyrrole binding, cytokine activity, ubiquitin-like protein ligase binding, and so on (Supplementary file6B). 5e top 20 GOMF enrichments ranked by the adjustedP value are shown in Figure 4(b).

GO CC function enrichment analysis showed that can-didate target genes of HQXXD acting on CVA were signifi-cantly enriched in membrane raft, membrane microdomain,membrane region, cyclin-dependent protein kinase holoen-zyme complex, mitochondrial outer membrane, serine/thre-onine protein kinase complex, integral component ofpresynaptic membrane, organelle outer membrane, outermembrane, protein kinase complex, and so on (Supplementaryfile 6C).5e top 20GOCC enrichments ranked by the adjustedP value are shown in Figure 4(c).

KEGG pathway enrichment analysis was further conductedfor determining candidate target genes of HQXXD acting onCVA. Candidate target genes were significantly enriched inadvanced glycation end products- (AGE-) receptor for the AGE(RAGE) signaling pathway in diabetic complications, fluid shearstress and atherosclerosis, hepatitis B, interleukin- (IL-) 17

85 132 1349

Active target genes of HQXXD

CVA-related target genes

Figure 1: Candidate target genes of HQXXD acting on CVA. 5e217 active target genes of HQXXD were compared with the 1481CVA-related target genes; then, 132 candidate target genes ofHQXXD acting on CVA were identified.

Table 2: Gene symbol and Entrez ID of candidate target genes.

Gene symbol Entrez IDPTGER3 5733MMP2 4313SLC6A2 6530ADRA2C 152HSPB1 3315PLAU 5328NOS2 4843CCNB1 891SLC6A4 6532IRF1 3659UGT1A1 54658ALOX5 240GJA1 2697COL1A1 1277BIRC5 332MAOA 4128PON1 5444CHEK1 1111BCL2 596ADRA2A 150CYP1A1 1543E2F1 1869HTR3A 3359CASP8 841PTGES 9536CHRM1 1128PPARG 5468CRP 1401GSTP1 2950CXCL8 3576SELE 6401FN1 2335AHR 196CHRNA2 1135NFE2L2 4780THBD 7056ECE1 1889MAPK14 1432ADRB1 153RAF1 5894EGF 1950IL1A 3552MPO 4353ADRA1B 147PCNA 5111CHUK 1147SPP1 6696PTGS2 5743SLPI 6590NCF1 653361CCND1 595ESR1 2099ADRB2 154POR 5447VEGFA 7422MYC 4609ADRA1D 146CCNA2 890ACHE 43MCL1 4170CCL2 6347


signaling pathway, tumor necrosis factor (TNF) signalingpathway, Kaposi sarcoma-associated herpesvirus infection,bladder cancer, prostate cancer, endocrine resistance, pancreaticcancer, and so on (Supplementary file 6D). 5e top 20 KEGGpathway enrichments ranked by the adjusted P value are shownin Figure 4(d).

Using the ClueGO plugin app and CluePedia plugin appof the Cytoscape software, the GO function enrichment andKEGG pathway enrichment from ClueGO can be displayedmore intuitively as shown in Figure 5.

3.7. Validation of the Clinical Curative Effect of HQXXD onCVA

3.7.1. Screened RCTs Investigating the Clinical Curative Effectof HQXXD on CVA. A total of 21 studies were retrievedthrough database searching. After removing duplication,nine studies were retained. A total of four irrelevant studieswere excluded after reading the title, abstract, and full text.Five RCTs [16–20] were included for further evaluation. 5eliterature screening process and results are shown inFigure 6.

3.7.2. Description of Included RCTs and Assessment of theMethodological Quality. Five eligible RCTs were identified.5e five RCTs were all conducted in China and included342 patients.5e five studies were all single-center studies.5e basic features of the included studies are outlined inTable 5.

Two RCTs [16, 19] employed adequate methods ofrandom sequence generation; none of the RCTs introducedallocation concealment; none of the RCTs describedblindness; all the RCTs had complete outcome data; and forall studies, we were unable to determine if these were se-lective reports (Figure 7).

3.7.3. Meta-Analysis. 5e five studies that compared thetotal effective rate of the clinical curative effect included atotal of 342 participants, 171 in the experimental groupsand 171 in the control groups. 5e five studies had ho-mogeneity (heterozygosity test, Chi2 �1.35, P � 0.85,I2 � 0%). When the fixed effect model was used to mergeOR values, the pooled OR was 2.86 (95% CI 1.37–5.95,Z � 2.80, P � 0.005). 5is indicated that the total effective

Table 2: Continued.

Gene symbol Entrez IDMMP1 4312STAT1 6772IL6 3569CASP3 836PARP1 142COL3A1 1281KDR 3791ABCG2 9429HMOX1 3162MMP3 4314PPARA 5465MAOB 4129CYP1A2 1544GSTM1 2944IL10 3586MAPK1 5594PLAT 5327MDM2 4193CXCL2 2920EGFR 1956NQO1 1728SOD1 6647IL2 3558ERBB3 2065ERBB2 2064IFNG 3458FOS 2353IL4 3565OPRD1 4985G6PD 2539TOP2A 7153MAPK8 5599ICAM1 3383CAV1 857BCL2L1 598CHEK2 11200RELA 5970HIF1A 3091NOS3 4846RB1 5925PGR 5241CHRM2 1129OPRM1 4988CXCL10 3627TYR 7299SERPINE1 5054VCAM1 7412CASP9 842ADH1B 125CDKN1A 1026NCOA2 10499AKT1 207IL1B 3553CHRM3 1131NFKBIA 4792IGFBP3 3486PTGS1 5742F3 2152NR1I2 8856JUN 3725ESR2 2100

Table 2: Continued.

Gene symbol Entrez IDBAX 581PRKCA 5578CD40LG 959CYP3A4 1576TP63 8626CYP1B1 1545PRSS1 5644LTA4H 4048MMP9 4318RXRB 6257


MOL002501

CVA

MOL009849

MOL000359 MOL011856

MOL001558

MOL001460

MOL012141

MOL002962 HRA

MOL000449

MOL005043

MOL000006MOL005100

MOL000358RA

PTGS1PRKCA IFNGSERPINE1 MAOAMOL011732 CASP3CHEK1

ICAM1 MCL1CDKN1AMOL011737 CHRM3 IL4ALOX5CXCL8

MMP1

MOL000011

ESR2UGT1A1MOL011730

BIRC5 CD40LGADH1BIGFBP3MOL012140

CHEK2 G6PDSOD1CASP8

OPRD1 ERBB3CCND1 PTGESHTR3A AKT1PON1

MAPK1SLPI IL1B MPOCHRM2 ADRB2 RAF1

THBD GSTP1PLAT TOP2APCNA CCL2MOL000211 PPARG

PLAU MMP9 ADRA1BSPP1 CAV1 HMOX1MOL011740 CYP1A1

PRSS1KDRPORESR1 MAPK8GJA1MOL001941 EGF

VCAM1 NOS2F3 CYP1A2COL1A1 CCNB1MOL002881

ADRA1D

MMP3 MDM2GSTM1 JUN IRF1 LTA4HCHRM1

BAX TP63 STAT1NQO1 NFKBIA CASP9MOL011747 ACHE

MOL000173

MOL001942

MOL011753

MOL013077MOL007514

MOL001494

MOL003588

MOL002644

HS

RXRB RB1EGFRHIF1AMOL011749 CHRNA2 PPARA FN1

MOL000417

MOL000422

MOL000442

MOL000098

MOL000392

MOL000380

MOL011849

MOL000387

MOL000239

CCNA2

PTGS2

IL10

PTGER3

CXCL10

NOS3

E2F1

MMP2

NR1I2

SLC6A4

NCOA2

IL1A

PARP1

NFE2L2

RSMOL000033

MOL000354

MOL000379

MOL000378

MOL000371

MOL000296

NCF1

CYP1B1

HSPB1

VEGFA

FOS

ABCG2

IL6

PGR

BCL2L1

MAOB

ADRA2A

ECE1

AHR

TYR

RELA

ERBB2

IL2

CHUK

SELE

ADRB1

MYC

CRP

MAPK14

CYP3A4

CXCL2

ADRA2C

OPRM1

COL3A1

SLC6A2

BCL2

Figure 2: Medicine-compound-target network.5ere were 184 nodes (4 medicine nodes, 47 compound nodes, 132 target gene nodes, and 1CVA node) and 653 edges in the network.

Table 3: Key compounds in HQXXD acting on CVA.

Compound ID Compound name Degree Medicine

MOL000098 Quercetin 96 RAHSMOL000006 Luteolin 41 HS

MOL000422 Kaempferol 37 RAHRAMOL000173 Wogonin 28 RSMOL000378 7-O-Methylisomucronulatol 23 RA

MOL000358 Beta-sitosterol 22 HSRSMOL000392 Formononetin 18 RAMOL000371 3,9-Di-O-methylnissolin 17 RAMOL000449 Stigmasterol 17 HSMOL000354 Isorhamnetin 16 RAMOL000380 (6aR,11aR)-9,10-Dimethoxy-6a,11a-dihydro-6H-benzofurano[3,2-c]chromen-3-ol 14 RAMOL000417 Calycosin 13 RAMOL001460 Cryptopin 13 HRAMOL002962 (3S)-7-Hydroxy-3-(2,3,4-trimethoxyphenyl) chroman-4-one 13 HRA


rate of the clinical curative effect was significantly higherin the experimental group compared with the controlgroup (Figure 8).

4. Discussion

CVA is characterized by wheezing cough, wind cough, andstubborn cough in TCM, which attributes qi deficiency andvigorous wind as themain causes of this disease and emphasizesthat the treatment of patients should be based on this aspect [21].

HQXXD is added or reduced from the TCMprescriptionYupingfeng powder and Zhisou powder [9]. RA, HRA, HS,and RS are monarch and minister herbs in HQXXD, whichare regarded as the main active herbs in HQXXD. RA isbelieved to tonify lung qi; HRA, HS, and RS are believed totonify the spleen and stomach and dispel wind dampness,which strengthen the effect of RA.

Chinese herbal medicines are complex and containmanycompounds. At present, most studies of TCM are limited toexplaining the mechanism of a certain compound and lack aholistic view. Currently, network pharmacology is widely

Figure 3:5e PPI network of HQXXD acting on CVA.When the lowest interaction score was set to 0.40, 132 target proteins in the networkhad an interaction and 2373 edges represented the interactions between the proteins.

Table 4: Key targets of HQXXD acting on CVA.

Key target DegreeAKT1 93IL6 91VEGFA 89JUN 81EGFR 79CXCL8 79CASP3 78MMP9 77PTGS2 77MAPK8 76MAPK1 76EGF 73MYC 73FOS 72IL1B 72ESR1 71CCND1 69CCL2 66FN1 64IL10 63


0 10 20 30 40

Reponse to toxic substance

Reponse to molecule of bacterial origin

Response to lipopolysaccharide

Response to extracellular stimulus

Response to nutrient levels

Response to metal ion

Response to reactive oxygen species

Response to oxidative stress

Cellular response to oxidative stress

Response to antibiotic

Reactive oxygen species metabolic process

Response to steroid hormone

Cellular response to reactive oxygen species

Response to xenobiotic stimulus

Cellular response to drug

Cellular response to xenobicotic stimulus

Response to mechanical stimulus

Regulation of the apoptotic signaling pathway

Response to nutrient

Response to oxygen levels

p adjust2.681962e – 332.829777e – 175.659554e – 175.489331e – 171.131191e – 16

BP

(a)

0 5 10 15 20

Protein heterodimerization activity

Cytokine receptor binding

G protein–coupled amine receptor activity

Nuclear receptor activity

Transcription factor activity, direct ligand-regulated sequence–specific DNA bindingOxidoreductase activity, acting on paired donors, with incorporation

or reduction of molecular oxygenHeme binding

Tetrapyrrole binding

Cytokine activity

Ubiquitin-like protein ligase binding

Cofactor binding

Steroid hormone receptor activity

Integrin binding

BH domain binding

Phosphatase binding

Ubiquitin protein ligase binding

Ammonium ion binding

DNA–binding transcription activator activity, RNA polymerase II–specific

Proximal promoter sequence–specific DNA binding

Kinase regulator activity

p adjust

2e – 05

4e – 05

6e – 05

MF

(b)

Figure 4: Continued.


used in the study of TCM and advocates whole and groupanalysis, which is consistent with the multicomponent,multitarget, and multipathway characteristics of TCM[22–24].

We identified the active compounds and target genes ofHQXXD from TCMSP. 5en, 132 candidate target genes ofHQXXD acting on CVA were identified, and the keycompounds of HQXXD acting on CVA were selected. 5e

0 5 10 15

Membrane ra�

Membrane microdomain

Membrane regionCyclin–dependent protein

kinase holoenzyme complexMitochondrial outer membrane

Serine/threonine protein kinase complex

Integral component of presynaptic membrane

Organelle outer membrane

Outer membrane

Protein kinase complex

Caveola

Integral component of postsynaptic membrane

Intrinsic component of presynaptic membrane

Intrinsic component of postsynaptic membrane

Extracellular matrix

Plasma membrane ra�

Integral component of synaptic membrane

Intrinsic component of synaptic membrane

Nuclear chromosome part

Vesicle lumen

p adjust

1e – 042e – 043e – 04

CC

(c)

0 10 20 30

AGE–RAGE signaling pathwayin diabetic complications

Fluid shear stress and atherosclerosis

Hepatitis B

IL–17 signaling pathway

TNF signaling pathway

Kaposi sarcoma–associated herpesvirus infection

Bladder cancer

Prostate cancer

Endocrine resistance

Pancreatic cancer

Human cytomegalovirus infection

Chagas disease (American trypanosomiasis)

Small cell lung cancer

Hepatitis C

C–type lectin receptor signaling pathway

Epstein–Barr virus infection

Proteoglycans in cancer

Toxoplasmosis

Relaxin signaling pathway

PI3K–Akt signaling pathway

p adjust

3.0e – 136.0e – 139.0e – 131.2e – 12

KEGG

(d)

Figure 4: Top 20 GO function and KEGG pathway enrichments. (a) 5e top 20 enriched BP functions of candidate target genes; (b) the top20 enriched MF activities of candidate target genes; (c) the top 20 enriched CC regions of candidate target genes; and (d) the top 20 KEGGpathway enrichments.


main key compounds were quercetin, luteolin, kaempferol,wogonin, 7-O-methylisomucronulatol, beta-sitosterol, for-mononetin, 3,9-di-O-methylnissolin, stigmasterol, and iso-rhamnetin. Quercetin has a potential protective andpreventive effect on the frequency of attacks and in con-trolling the most common symptoms of asthma [25];quercetin has both immunomodulatory and bronchodila-tory properties [26]. Luteolin has been shown to have anantiallergic effect in a murine model of allergic asthma andrhinitis [27]. Kaempferol may be a potent antiallergiccompound targeting the allergic asthma typical of airwayhyperplasia and hypertrophy [28]; kaempferol could sup-press eosinophil infiltration and airway inflammation inairway epithelial cells and in mice with allergic asthma [29].Wogonin could attenuate ovalbumin-induced airway in-flammation in a mouse model of asthma [30]. Beta-sitosterol

possesses antiasthmatic actions by inhibiting the cellularresponses and subsequent release/synthesis of 52 cytokines[31]. Stigmasterol has antiasthmatic properties and sup-pressive effects on the key features of allergen-inducedasthma [32]. 5e results of our study are in accordance withthese previous studies, which indicate that HQXXD has aneffect on CVA because of the action of these multiplecompounds.

PPI network analysis suggested that the action ofHQXXD on CVA is related to multiple targets. 5e keytargets were AKT1, IL6, VEGFA, JUN, EGFR, CXCL8,CASP3, MMP9, PTGS2, and MAPK8. Previous research hasindicated that AKT1 is related to pathway mechanisms inchildren with severe asthma [33]; IL6 polymorphisms arerelated to the effects of smoking on the risk of adult asthma[34]; VEGFA rs833069 SNPs are associated with asthma

Figure 5: GO function enrichment and KEGG pathway enrichment from ClueGO.


Records identifiedthrough database

searching(n = 20)

Records a�er duplicatesremoved(n = 9)

Records screened(n = 9)

Studies includedin quantitative

synthesis(meta-analysis)

(n = 5)

Records excluded(n = 4)

Additional recordsidentified through

other sources(n = 1)

Figure 6: Flow chart of the research selection process. A total of 21 studies were retrieved through database searching. After removingduplication, nine studies were retained. A total of four irrelevant studies were excluded after reading the title, abstract, and full text. FiveRCTs were included for further evaluation.

Table 5: Summary of RCTs investigating HQXXD in CVA.

Study year [ref] Country Sample size(experimental/control)Mean age (years)

(experimental/control) Experimental Control

Wang and Xie2009 [20] China 42 (21/21) 34.5± 2.7/33.2± 2.6 HQXXD Terbutaline tablet

Fan and Xie2014 [19] China 30 (30/30) 42.10± 8.27/38.2± 10.35 HQXXD

Procaterol hydrochloride tablet andalbuterol aerosol

Li 2015 [17] China 70 (35/35) 50± 1.5/49± 1.3 HQXXD Procaterol hydrochloride tablet andalbuterol aerosolWang and Xie2015 [18] China 90 (45/45) 41.12± 7.14/40.18± 9.35 HQXXD Procaterol hydrochloride tablet

Wei and Qin2019 [16] China 80 (40/40) 45.35± 3.88/43.25± 3.78 HQXXD

Ambroxol hydrochloride tablet andterbutaline sulfate tablet

Random sequence generation (selection bias)

Allocation concealment (selection bias)

Blinding of participants and personnel (performance bias)

Blinding of outcome assessment (detection bias)

Incomplete outcome data (attrition bias)

Selective reporting (reporting bias)

Other bias

Fan

and

Xie [

19]

Li [1

7]

Wan

g an

d Xi

e [20

]

Wan

g an

d Xi

e [18

]

Wi a

nd Q

in [1

6]

+ ? ? ? +

? ? ? ? ?

– – – – –

? ? ? ? ?

+ + + + +

+ + + + +

? ? ? ? ?

Figure 7: Risk of bias summary. Review authors’ judgments about each risk of bias item for each included study.


[35]; VEGFA is associated with the response to inhaledcorticosteroids in children with asthma [36]; smoking ag-gravates airway inflammation by activating the c-Jun aminoterminal kinase pathway in asthma [37]; EGFR activation-induced decreases in claudin 1 promote MUC5AC ex-pression and exacerbated asthma in mice [38]; the EGRF-dependent signaling pathway is associated with diseases,such as asthma [39]; neutrophils release CXCL8, NE, andMMP-9 in response to viral surrogates with R848-inducedCXCL8 release being specifically enhanced in neutrophils inasthma [40]. CASP3 may play a role in allergic asthma [41];the PTGS2 gene is associated with diisocyanate-inducedasthma [42]. 5e relationships between these key targets andCVA, some of which have been well studied, could be furtherstudied, and possible mechanisms that have not been pre-viously studied could be explored.

GO functional enrichment analysis suggested that GOBP for HQXXD in the treatment of CVA were enriched inresponse to toxic substances, molecules of bacterial origin,lipopolysaccharides, extracellular stimuli, nutrient levels,metal ions, reactive oxygen species, oxidative stress, anti-biotics, and cellular responses to oxidative stress. GO MFanalysis indicated that HQXXD in the treatment of CVAinvolved in protein heterodimerization activity, cytokinereceptor binding,G protein-coupled amine receptor activity,nuclear receptor activity, transcription factor activity, directligand-regulated sequence-specific DNA binding, oxidore-ductase activity acting on paired donors with incorporationor reduction of molecular oxygen, heme binding, tetra-pyrrole binding, cytokine activity, and ubiquitin-like proteinligase binding. GO CC analysis indicated that HQXXD inthe treatment of CVA were enriched in the membrane raft,membrane microdomain, membrane region, cyclin-de-pendent protein kinase holoenzyme complex, mitochondrialouter membrane, serine/threonine protein kinase complex,integral component of presynaptic membrane, organelleouter membrane, outer membrane, and protein kinasecomplex. 5ese regions are closely related to airway in-flammation, airway epithelial cell injury, airway hyper-responsiveness, and airway remodeling, which are closelyrelated to the pathological mechanisms of CVA [43–45].

KEGG enrichment pathway analysis showed thatmany pathways were closely related to the pathogenesis of

CVA. 5e main pathways included the AGE-RAGEsignaling pathway in diabetic complications, fluid shearstress and atherosclerosis, the IL-17 signaling pathway,TNF signaling pathway, and pathways involved in hep-atitis B, Kaposi sarcoma-associated herpes virus infec-tion, bladder cancer, prostate cancer, endocrineresistance, and pancreatic cancer. A relationship betweenthese pathways and CVA has also been found or pre-liminarily indicated. 5ese results indicated that HQXXDacts on CVA through multiple pathways. 5e mainpathways and main target genes in the pathways areworthy of further study.

5e pharmacological mechanisms of HQXXD acting onCVA have been investigated through a network pharma-cology approach, but there are some limitations with thisapproach. We identified active compounds and target genesof HQXXD from the TCMSP database and acquired CVA-related target genes from the GeneCards database. Althoughthese databases are the most comprehensive databasescurrently available, the data collection is still not necessarilycomprehensive, and the establishment of the screeningcriteria for active compounds cannot be completely accurate.In addition, we constructed a PPI network and performedGO function enrichment and the KEGG pathway enrich-ment analysis to investigate the pharmacological mecha-nisms of HQXXD acting on CVA. Further investigation ofthe key target genes and pathways through experimentalanalysis is required, which will be the focus of future re-search in our laboratory.

In addition, we used an evidence-based medicine ap-proach to perform a meta-analysis of the clinical curativeeffect of HQXXD on CVA. Our previous meta-analysisshowed that the total effective rate of HQXXD in thetreatment of CVA was not significant compared with thecontrol group [9]. We searched the main database andscreened RCTs investigating the clinical curative effect ofHQXXD in CVA to perform a meta-analysis, which indi-cated that the total effective rate of the clinical curative effectwas significantly higher in the experimental group than inthe control group. 5us, an evidence-based medicine ap-proach was used to indicate the efficacy of HQXXD in CVAfrom a clinical level, which was more meaningful than ex-perimental validation.

Experimental ControlEvents Total Events Total

28 30 25 3032 35 30 35

Study or subgroup

Fan and Xie [19]Li [17]

171Total (95% CI) 171

Totalevents 160 143Heterogeneity: chi2 = 1.35, df = 4 (P = 0.85); I2 = 0%Test for overall effect: Z = 2.80 (P = 0.005)

Weight(%) Year

18.228.1

20 21 16 21Wang and Xie [20] 8.3 6.25 [0.66, 59.03] 20092.80 [0.50,15.73] 20141.78 [0.39, 8.09] 2015

41 45 37 45Wang and Xie [18] 35.9 2.22 [0.62, 7.97] 2015Wang and Qin [16] 39 40 35 40 9.5 5.57 [0.62, 50.03] 2019

100.0 2.86 [1.37, 5.95]

0.005

Odds ratioM-H, fixed, 95% CI

Odds ratioM-H, fixed, 95% CI

2000.1 1 10Favours (control) Favours (experimental)

Figure 8: Forest plot of comparison: total effective rate of the clinical curative effect. When the fixed effect model was used to merge ORvalues, the pooled OR was 2.86 (95% CI 1.37–5.95, Z� 2.80, P � 0.005). 5is result indicated that the total effective rate of the clinicalcurative effect was significantly higher in the experimental group compared with the control group.


5. Conclusion

In conclusion, based on a network pharmacology approach,this study investigated the relationships between the activecompounds, target genes, and pathways, and further studiedthe multicomponent, multitarget, and multipathway char-acteristics of HQXXD acting on CVA. HQXXD acts on CVAvia the effect of multiple compounds related to multipletargets and through multiple pathways. Finally, we inves-tigated the clinical curative effect of HQXXD on CVA usingan evidence-based medicine approach, which indicated thatthe total effective rate of the clinical curative effect wassignificantly higher in the experimental group comparedwith the control group. In brief, the pharmacologicalmechanism of HQXXD acting on CVA has been investi-gated, and the clinical effect of HQXXD on CVA isremarkable.

Data Availability

5e data used to support this study are included within thesupplementary files.

Conflicts of Interest

5e authors declare that they have no conflicts of interest.

Authors’ Contributions

Qingqing Xia, Mingtao Liu, and Hui Li contributed equallyto this work. Qingqing Xia, Mingtao Liu, Hui Li, and YufengZhang conceived and designed the study and wrote themanuscript. Qingqing Xia, Mingtao Liu, Hui Li, and LijunTian were responsible for data collation and extraction andperformed the data analysis. Lijun Tian, Jia Qi, and YufengZhang performed supervision and project administration.All authors read and approved the final manuscript.

Acknowledgments

5e authors would like to acknowledge the web databaseplatform and software for data analysis. Victoria Muir,PhD, from Liwen Bianji, Edanz Group China (http://www.liwenbianji.cn/ac), edited the English text of adraft of this manuscript. 5is study was partially sup-ported by the Scientific Research Project from the HealthCommission of Wuxi (grant Q202055 to Xia), the Nan-tong Science and Technology Plan Project (grantMS12017004-2 to Tian), the National Natural ScienceFoundation of China (grant 81302768 to Qi), and Re-search Grant of Jiangyin Hospital of Traditional ChineseMedicine (grant to Zhang).

Supplementary Materials

Supplementary file 1: compounds of HQXXD in TCMSP.(A) 87 compounds of RA in TCMSP; (B) 192 compounds ofHRA in TCMSP; (C) 159 compounds of HS in TCMSP; (D)173 compounds of RS in TCMSP. Supplementary file 2:

target genes of HQXXD: (A) corresponding target genes ofactive compounds; (B) corresponding target gene symbols ofactive compounds; (C) 217 active target genes of HQXXD.Supplementary file 3: CVA-related target genes. (A) CVA-related target genes from GeneCards; (B) 1481 CVA-relatedtarget genes. Supplementary file 4: data of the medicine-compound-target network. Supplementary file 5: data of thePPI network. Supplementary file 6. GO function enrichmentand KEGG pathway enrichment analysis. (A) BP functionenrichment analysis; (B) MF function enrichment analysis;(C) CC function enrichment analysis; (D) KEGG pathwayenrichment analysis. (Supplementary Materials)

References

[1] H. Matsuoka, A. Niimi, H. Matsumoto et al., “Inflammatorysubtypes in cough-variant asthma: association with mainte-nance doses of inhaled corticosteroids,” Chest, vol. 138, no. 6,pp. 1418–1425, 2010.

[2] Y. Y. Koh, J. H. Jeong, Y. Park, and C. K. Kim, “Developmentof wheezing in patients with cough variant asthma during anincrease in airway responsiveness,” European RespiratoryJournal, vol. 14, no. 2, pp. 302–308, 1999.

[3] K. Lai and L. Long, “Current status and future directions ofchronic cough in China,” Lung, vol. 198, no. 1, pp. 23–29, 2020.

[4] T. J. An, J. W. Kim, E. Y. Choi et al., “Clinical characteristics ofchronic cough in Korea,” Tuberculosis and Respiratory Dis-eases (Seoul), vol. 83, no. 1, pp. 31–41, 2020.

[5] E. Tagaya, M. Kondo, S. Kirishi, M. Kawagoe, N. Kubota, andJ. Tamaoki, “Effects of regular treatment with combination ofsalmeterol/fluticasone propionate and salmeterol alone incough variant asthma,” Journal of Asthma, vol. 52, no. 5,pp. 512–518, 2015.

[6] W. Bao, Q. Chen, Y. Lin et al., “Efficacy of procaterol com-bined with inhaled budesonide for treatment of cough-variantasthma,” Respirology, vol. 18, no. 3, pp. 53–61, 2013.

[7] R. R. Fan, Z. H. Wen, D. W. Wang et al., “Chinese herbalmedicine for the treatment of cough variant asthma: a studyprotocol for a double-blind randomized controlled trial,”Trials, vol. 20, no. 1, pp. 1–8, 2019.

[8] Z.-d. Zhang, Y.-q. Deng, Y. Zhang, Y. Han, L. Lin, and E.-x. Chao,“TCM differential treatment of cough variant asthma,” Journal ofTraditional Chinese Medicine, vol. 30, no. 1, pp. 60–63, 2010.

[9] L. Chen, J. Qi, Y. F. Zhang, and W. L. Jiang, “Systematicreview of astragalus asarum decoction in the treatment ofcough variant asthma,”World Chinese Medicine, vol. 11, no. 1,pp. 155–158, 2016.

[10] J. Ru, P. Li, J. Wang et al., “TCMSP: a database of systemspharmacology for drug discovery from herbal medicines,”Journal of Cheminformatics, vol. 6, p. 13, 2014.

[11] P. Shannon, A. Markiel, O. Ozier et al., “Cytoscape: a softwareenvironment for integrated models of biomolecular interac-tion networks,” Genome Research, vol. 13, no. 11,pp. 2498–2504, 2003.

[12] G. Yu, L.-G. Wang, Y. Han, and Q.-Y. He, “clusterProfiler: anR package for comparing biological themes among geneclusters,” OMICS: A Journal of Integrative Biology, vol. 16,no. 5, pp. 284–287, 2012.

[13] J. P. T. Higgins, D. G. Altman, P. C. Gotzsche et al., “5ecochrane collaboration’s tool for assessing risk of bias in rand-omised trials,” BMJ, vol. 343, no. 2, Article ID d5928, 2011.


http://www.liwenbianji.cn/achttp://www.liwenbianji.cn/achttp://downloads.hindawi.com/journals/ecam/2020/3829092.f1.zip

[14] J. P. T. Higgins and S. G. 5ompson, “Quantifying hetero-geneity in a meta-analysis,” Statistics in Medicine, vol. 21,no. 11, pp. 1539–1558, 2002.

[15] J. P. T. Higgins, S. G.5ompson, J. J. Deeks, andD. G. Altman,“Measuring inconsistency in meta-analyses,” BMJ, vol. 327,no. 7414, pp. 557–560, 2003.

[16] X. L.Wei andQ. J. Qin, “Clinical efficacy and safety evaluationof Huangqi Xixin decoction in treating cough variantasthma,” World Latest Medicine Information, vol. 19, no. 38,pp. 164–166, 2019.

[17] W. H. Li, “Clinical observation of Huangqi Xixin decoction intreating cough variant asthma,” -e Medical Forum, vol. 19,no. 24, pp. 3388-3389, 2015.

[18] X. L. Wang and M. H. Xie, “Treatment of 45 cases of coughvariant asthma with Yiqi Shufeng,”Henan Traditional ChineseMedicine, vol. 35, no. 10, pp. 2443-2444, 2015.

[19] Y. L. Fan and M. H. Xie, “Clinical observation of AstragalusAsarum decoction in treating cough variant asthma (Qi de-ficiency and wind vigorous),” Journal of Emergency in Tra-ditional Chinese Medicine, vol. 23, no. 7, pp. 1353–1355, 2014.

[20] X. L. Wang and M. H. Xie, “Treatment of 21 cases of coughvariant asthma with Astragalus Asarum decoction,” Journal ofEmergency in Traditional Chinese Medicine, vol. 18, no. 9,pp. 1516-1517, 2009.

[21] D. Chen, F. Zhang, S. Tang et al., “A network-based systematicstudy for the mechanism of the treatment of Zhengs related tocough variant asthma,” Evidence-Based Complementary andAlternativeMedicine, vol. 2013, Article ID 595924, 15 pages, 2013.

[22] Q.Miao, P.-c.Wei, M.-r. Fan, and Y.-p. Zhang, “Clinical studyon treatment of cough variant asthma by Chinese medicine,”Chinese Journal of Integrative Medicine, vol. 19, no. 7,pp. 539–545, 2013.

[23] R. Zhang, X. Zhu, H. Bai, and K. Ning, “Network pharma-cology databases for traditional Chinese medicine: review andassessment,” Frontiers in Pharmacology, vol. 10, Article ID123, 2019.

[24] Y. Zhang, W. Jiang, Q. Xia, J. Qi, and M. Cao, “Pharmaco-logical mechanism of Astragalus and Angelica in the treat-ment of idiopathic pulmonary fibrosis based on networkpharmacology,” European Journal of Integrative Medicine,vol. 32, Article ID 101003, 2019.

[25] M. R. Cesarone, G. Belcaro, S. Hu et al., “Supplementaryprevention and management of asthma with quercetin phy-tosome: a pilot registry,” Minerva Medica, vol. 110, no. 6,pp. 524–529, 2019.

[26] T. T. Oliveira, K. M. Campos, A. T. Cerqueira-Lima et al.,“Potential therapeutic effect ofAllium cepa L. and quercetin ina murine model of Blomia tropicalis induced asthma,”DARU-Journal of Pharmaceutical Sciences, vol. 23, no. 1, p. 18, 2015.

[27] T. Y. Jang, A.-Y. Jung, T.-S. Kyung, D.-Y. Kim, J.-H. Hwang,and Y. H. Kim, “Anti-allergic effect of luteolin in mice withallergic asthma and rhinitis,” Central European Journal ofImmunology, vol. 1, no. 1, pp. 24–29, 2017.

[28] D. Shin, S.-H. Park, Y.-J. Choi et al., “Dietary compoundkaempferol inhibits airway thickening induced by allergicreaction in a bovine serum albumin-induced model ofasthma,” International Journal of Molecular Sciences, vol. 16,no. 12, pp. 29980–29995, 2015.

[29] J.-H. Gong, D. Shin, S.-Y. Han, J.-L. Kim, and Y.-H. Kang,“Kaempferol suppresses eosionphil infiltration and airwayinflammation in airway epithelial cells and in mice with al-lergic asthma,” -e Journal of Nutrition, vol. 142, no. 1,pp. 47–56, 2012.

[30] E. K. Ryu, T.-H. Kim, E. J. Jang et al., “Wogonin, a plantflavone from Scutellariae radix, attenuated ovalbumin-in-duced airway inflammation in mouse model of asthma via thesuppression of IL-4/STAT6 signaling,” Journal of ClinicalBiochemistry and Nutrition, vol. 57, no. 2, pp. 105–112, 2015.

[31] S. G. Mahajan and A. A. Mehta, “Suppression of ovalbumin-induced 52-driven airway inflammation by β-sitosterol in aGuinea pig model of asthma,” European Journal of Phar-macology, vol. 650, no. 1, pp. 458–464, 2011.

[32] A. O. Antwi, D. D. Obiri, and N. Osafo, “Stigmasterolmodulates allergic airway inflammation in Guinea pig modelof ovalbumin-induced asthma,” Mediators of Inflammation,vol. 2017, Article ID 2953930, 11 pages, 2017.

[33] Y. Kobayashi, C. Bossley, A. Gupta et al., “Passive smokingimpairs histone deacetylase-2 in children with severe asthma,”Chest, vol. 145, no. 2, pp. 305–312, 2014.

[34] T. K. Lajunen, J. J. K. Jaakkola, and M. S. Jaakkola, “IL6polymorphisms modify the effects of smoking on the risk ofadult asthma,” Journal of Allergy and Clinical Immunology,vol. 141, no. 2, pp. 799–802, 2018.

[35] T. 5ongngarm, A. Jameekornrak, N. Chaiyaratana et al.,“Effect of gene polymorphisms in ADAM33, TGFbeta1,VEGFA, and PLAUR on asthma in 5ai population,” AsianPacific Journal of Allergy and Immunology, 2019.

[36] Z. Wan, Y. Tang, Q. Song et al., “Gene polymorphisms inVEGFA and COL2A1 are associated with response to inhaledcorticosteroids in children with asthma,” Pharmacogenomics,vol. 20, no. 13, pp. 947–955, 2019.

[37] M.Guo, Y. Liu, X.Han et al., “Tobacco smoking aggravates airwayinflammation by upregulating endothelin-2 and activating thec-Jun amino terminal kinase pathway in asthma,” InternationalImmunopharmacology, vol. 77, Article ID 105916, 2019.

[38] Z. Jia, K. Bao, P. Wei et al., “EGFR activation-induced de-creases in claudin1 promote MUC5AC expression and ex-acerbate asthma in mice,” Mucosal Immunology, 2020.

[39] A. Z. El-Hashim, M. A. Khajah, R. S. Babyson et al., “Ang-(1-7)/MAS1 receptor axis inhibits allergic airway inflammationvia blockade of Src-mediated EGFR transactivation in aMurinemodel of asthma,” PLoS One, vol. 14, no. 11, Article IDe224163, 2019.

[40] F. S. M. Tang, D. Van Ly, K. Spann et al., “Differentialneutrophil activation in viral infections: enhanced TLR-7/8-mediated CXCL8 release in asthma,” Respirology, vol. 21,no. 1, pp. 172–179, 2016.

[41] X. Diao, J. Wang, H. Zhu, and B. He, “Overexpression ofprogrammed cell death 5 in a mouse model of ovalbumin-induced allergic asthma,” BMC Pulmonary Medicine, vol. 16,no. 1, p. 149, 2016.

[42] B. Yucesoy, M. L. Kashon, V. J. Johnson et al., “Geneticvariants inTNFα, TGFB1, PTGS1 and PTGS2 genes are as-sociated with diisocyanate-induced asthma,” Journal ofImmunotoxicology, vol. 13, no. 1, pp. 119–126, 2016.

[43] F. Yi, L. Han, B. Liu et al., “Determinants of response tobronchodilator in patients with cough variant asthma-arandomized, single-blinded, placebo-controlled study,” Pul-monary Pharmacology & -erapeutics, vol. 61, Article ID101903, 2020.

[44] H. Yuan, X. Liu, L. Li et al., “Clinical and pulmonary functionchanges in cough variant asthma with small airway disease,”Allergy, Asthma, and Clinical Immunology, vol. 15, p. 41, 2019.

[45] W. Sun and H.-Y. Liu, “Montelukast and budesonide forchildhood cough variant asthma,” Journal of the College ofPhysicians and Surgeons Pakistan, vol. 29, no. 4, pp. 345–348,2019.


networkpharmacologystrategytoinvestigatethe...

Documents