Prediction of active ingredients and potential mechanisms of Alisma decoction against atherosclerosis:A study based on UHPLC-QOrbitrap-HRMS and network pharmacology

Yi-Ming Bi,Wei Zhang,Yi-Gui Lai,Ying-Chang Fu,Kong-Zheng Li,Qiang Wang,Xue-Feng Jiang,Hong-Feng Liang,and Hui-Jie Fan

Abstract—Background:Atherosclerosis (AS) has been a potentially life-threatening disease worldwide.Alisma decoction (AD) is a famous Chinese formula in treating AS.However,the active components and potential mechanisms remain unknown.Methods:In this study,the active compositions of AD were analyzed by ultrahigh-performance liquid chromatography quadrupole-orbitrap high-resolution mass spectrometry (UHPLC-Q-Orbitrap-HRMS).The putative targets were predicted and interpreted by a series of bio-informative tools such as Venn,STRING,Network Analyzer,Cluster Marker,and DAVID.Finally,the results were validated by molecular docking using AutoDock.Results:The results showed that alisol A,alisol B,23-o-Acetylalisol B,atractylenolide Ⅲ and atractylenolide Ⅱ were the major compositions in AD.Thirty targets mainly distributed in three gene clusters were obtained after topological and cluster analysis.KEGG and GO analyses showed that genes in cluster 1 mainly participated in adipocytokine signaling pathway and played an important role in lipid metabolism,while genes in cluster 2 had regulatory actions via cAMP signaling pathway and may protect vascular function from AS,and genes in cluster 3 were closely related to inflammation.Furthermore,RXRA,AGT and CXCL8,the central gene of cluster 1,2 and 3,were verified to be strong binding with some major compositions in AD.Conclusion:The possible mechanisms of AD in the treatment of AS may be closely correlated with the regulation of lipid metabolism,vascular physiology and inflammation.Our study provided a new insight into the anti-AS effect of AD,but more pharmacological experiments should be performed for verification in the future.

Key words—Alisma decoction,Atherosclerosis,Network pharmacology,UHPLC-Q-Orbitrap-HRMS,Molecular docking

INTRODUCTION

Atherosclerosis (AS) is one of the most common chronic diseases worldwide.It has been reported that about 2 of 3 men and 1 in 2 women after age 40 often suffer from atherosclerosis,and clinical manifestations such as cardiovascular diseases and peripheral arterial diseases frequently occur in patients with atherosclerosis [1].The pathophysiology of AS is complicated,but in generally,this disease is induced by lipoproteins deposition and subsequently inflammations in the vessel walls [2].In brief,atherosclerotic plaques form at specific sites of vessel walls,and gradually develop with chronic inflammation induced by continued hypercholesterolemia and hypertension,which would aggravate stenosis or thrombosis that block the vessel and further lead to hypoxia and organ dysfunction.All these may lead to life-threatening events such as myocardial infarction (MI) and stroke.Nowadays,available treatments for AS mainly include drugs lowering lipoproteins such as statins and ezetimibe.However,the curative effects of statins therapy for primary prevention in elderly adults with AS are equivocal [3],and some patients are intolerant to statins [4].As a result,the risk of atherosclerotic vascular diseases remains high.More effective options with less side effects should be sought for AS prevention and treatment.

Alisma decoction (AD),also known as Zexie decoction,is a classical traditional Chinese herbal preparation firstly prescribed in the ancient Chinese medical scripture“Synopsis of Prescriptions of the Golden Chamber (Jin Gui Yao Lve)”,which is composed of two herbs,includingRhizoma Alismati(Ze-Xie in Chinese) andRhizoma Atractylodis Macrocephalae(Bai-Zhu in Chinese).Generally,the method of AD extraction is to decoct the medicinal herbs mixed with adequate water.Nowadays,AD has been widely used in a variety of diseases,including hypertension,hyperlipidemia,and cerebrovascular diseases.Some modern pharmacological articles demonstrated that AD showed its superior effects on hepatoprotection,antiinflammation,immunomodulation and anti-atherosclerotic activities [5,6].Besides,the combination uses of Ze-Xie and Bai-Zhu could reduce the levels of total cholesterol(TC),triglyceride (TG) and low-density lipoprotein (LDL),and inhibit lipids deposition and intimal hyperplasia of common carotid in ApoE-/-mice with AS [7].These findings suggested a potential pharmacological action of AD against AS,but the active compositions and the detailed mechanisms remain unknown.On the other hand,the current experimental methods are too simple to explore the whole effects of traditional Chinese formula [8].More novel analytical techniques for Traditional Chinese Medicine (TCM) research are expected.

As a modern methodology,ultrahigh-performance liquid chromatography coupled to quadrupole-Orbitrap highresolution mass spectrometry (UHPLC-Q-Orbitrap-HRMS)has been widely used in separation and identification of constituents in Chinese herbs and even in drug-containing serum [9,10].With higher sensitivity and specificity,this methodology allows an exact quantification based on mass measurement [11].On the other hand,network pharmacology is another modern methodology for TCM researches because it could predict the therapeutic targets of Chinese herbs and further forecast the promising pharmacological actions [12].The combination uses of UHPLC-Q-Orbitrap-HRMS and network pharmacology offers an effective mean of high throughput screening,which is convenient to study the complex traditional Chinese formulas.

MATERIALS AND METHODS

Reagents

Ze-Xie and Bai-Zhu were purchased from Zisun Herbal Pharmaceutical (Guangzhou,GuangDong,China).Alisol B and atractylenolide Ⅱ (purity ≥ 98%) were obtained from Chengdu Mansite Biotechnology (Chengdu,Sichuan,China).Formic acid was obtained from Aladdin Bio-Chem Technology (Shanghai,China).Acetonitrile and methanol were purchased from Thermo Fisher Scientific (New York,USA).Ultrapure water was obtained from an EPED-E1-10TJ system (Nanjing,China).

Ultrahigh-performance liquid chromatography quadrupole-orbitrap high-resolution mass spectrometry(UHPLC-Q-Orbitrap-HRMS) analysis

To obtain the active compositions of AD,Ze-Xie (15 g)and Bai-Zhu (6 g) were boiled and the extracts of these two herbs were filtrated using a 0.22 mm membrane.Then,the active compositions of the fresh AD extract were analyzed by UHPLC-Q-Orbitrap-HRMS equipped with an UHPLC system (UltiMate 3000 RS,Thermo Fisher Scientific,USA)coupled to Q Exactive Orbitrap (QE,Thermo Fisher Scientific,USA).The chromatographic separation analysis was conducted on a RP-C18 column (150×2.1 mm,1.8 μm)with a flow rate of 0.3 mL/min at 35 ℃.A mobile phase system including 0.1% formic acid-water (A) and 0.1%formic acid-acetonitrile (B) was built in detail:0-1 min,2%B;2 min,20% B;10 min,50% B;15 min,80% B;20-25 min,95% B;26-30 min,2% B.The following as parameters was used in the QE mass spectrometer:full mass resolution as 70000,dd-MS2 resolution as 17500,scan range as 150.0-2000.0 m/z,spray voltage as 3.8 kV(positive) or -3.1 kV (negative),respectively.The reference standards,alisol B and atractylenolide Ⅱ were also detected.Data were obtained and analyzed using Compound Discover 2.1 and Xcalibur 4.0.27 software.The compositions were obtained based on reference standards,and mass spectral libraries such as mzCloud,mzVault and ChemSpiderand which can provide the information required to confidently identify unknown compounds based on multi-stage and multidimensional mass fragment patterns.Finally,the active compositions were identified using mzCloud Best Match.In our study,compositions with mzCloud Best Match >70 were applied in further researches [13].

Computational target fishing and analysis

The putative targets of AD were predicated using TCMSP and BATMAN (http://bionet.ncpsb.org/batmantcm/).The target with a given cutoff“Score cutoff ≥ 30”was collected.The targets related to AS were collected from GeneCards (https://www.genecards.org/) and OMIM(http://www.omim.org/).All targets whose species are“Homo sapiens”were corrected to the official symbols using UniProt database (http://www.uniprot.org/).Targets of AD and AS were imported to a website tool Venn(http://bioinformatics.psb.ugent.be/webtools/Venn/) to screen the common targets.Then,the interactions of these common targets were investigated by another website tool STRING (https://string-db.org/). A protein-protein functional interactions (PPI) network with a minimum interaction score >0.9 was built via STRING.To find the key targets relevant to AD in the treatment of AS,a tool named“NetworkAnalyzer”in Cytoscape v3.7.2 was used to calculate the number of adjacent nodes of targets (also known as degree in the network) in PPI network [14].Targets with a degree value twice as much as the average one were selected as the key targets.

Target classification and analysis

Different genes complete different tasks cooperatively,which is described as gene regulatory network.In this study,a plug-in named“ClusterMarker”in Cytoscape v3.7.2 was used to classify multiple common genes into different networks.These networks may regulate some specific physiopathological progresses in AS development.The physiological effects of different networks were annotated using DAVID (https://david.ncifcrf.gov/).The Gene Ontology (GO) enrichment analysis mainly interpret biological process (BP) and Kyoto Encyclopedia of Genes and Genomes (KEGG) were also used to describe the features of these networks.Genes anticipating in different networks and their functional analyses were visualized by R software.The threshold value false discovery rate (FDR)<0.05 were set as statistically significant.Moreover,the key target with greatest degree value,distributed in different gene regulatory networks,would be considered as the seed in the network and further studied with molecular docking.

Molecular docking

AutoDock Vina software (http://vina.scripps.edu/) was engaged in investigating the active composition-gene connection in this study.The 3D structures files with pdb format of targets were downloaded from the RCSB PDB database (http://www.rcsb.org/),and these crystal structures of targets and the potential active ingredients were prepared by AutodockTools to separate local ligands,remove water,add hydrogen,regulate distribution of atoms prior to performing the docking process,respectively.The best models between active ingredients and targets were identified as those exhibiting lowest binding-energy.Docking conformations with a binding-energy <-7 kcal were regarded strong bio-active and visualized using PyMOL (https://pymol.org/2/).

RESULTS

Identification of active compositions

In this study,a novel analytical technology coupled UHPLC-Q-Orbitrap-HRMS was used to screen the active compounds of AD.At the same time,this technique can collect compositions with 30 different channels,and these channels can be used to distinguish compounds in a coelution peak.Though the total ion chromatograms are difficult to distinguish some compounds,these compounds can be collected with different channels.As a result,a total of 28 active compounds had a mzCloud Best Match score greater than 70 were putatively identified by comparing against database mzCloud.The active compounds of ADwere displayed as different peaks in typical total ion chromatograms (TICs).As shown in Figure 1A and B,20 chemicals were identified in positive ion and 8 chemicals were identified in negative ion.The result indicated that the fresh AD solution mainly consisted of triterpenoids,lactones,phenolic acids and so on.Detailed information about compositions,including name,molecular formula,molecular weight,retention time (T/Rmin),ion mode and mzCloud Best Match score are presented in Table 1.

Figure 1 The TICs of AD by UHPLC-Q-Orbitrap-HRMS analysis.(A) TICs in positive ion mode.(B) TICs in negative ion mode.

Figure 2 Determination and PPI network analysis.

Table 1 Components identification in AD

Target determination and PPI network analysis

The chemical names and the pubchem IDs were imported to TCMSP and BATMAN to predict the putative targets.To increase the accuracy of the target prediction,a total of 839 targets with score ≥ 30 were screened.Using the keyword“atherosclerosis”,the number of diseasetargets obtained from GeneCards and OMIM was 4481 and 550,respectively.Moreover,after removing duplication,only 461 drug-targets and 4942 disease-targets were identified.To explore the overlapping targets of AD and AS,a useful tool Venn was used and 240 overlapping targets were obtained.The gene-gene interactions play a critical role in atherosclerosis and anti-atherosclerosis.The gene neighborhood,gene fusions and gene co-occurrence of 240 genes were measured by STRING.With this database,an interactive functional association PPI network composed of 193 nodes and 663 edges was built,in which the nodes are defined as targets and edges are defined as gene interactions.To obtain the key targets,we used NetworkAnalyzer to calculate the degree of genes,and the result showed that the average value of 193 genes was 6.37,and 30 targets with a value greater than 13 were selected as key targets.The key portion that was predicted to interact with each other occupy about 15% of the whole PPI network and is assembled via a new network labeled with deeper orange.

The PPI network have been clustered to identify stable complexes,but it is difficult to analyze different gene clusters from multiple data.ClusterMarker is a computational tool capable of integrating data of diverse types and facilitating recognition of biologically meaningful patterns,which was used to determine the gene regulatory network in this study.Targets in the PPI network were divided into 18 clusters,and among these clusters,cluster 1,2,3 and 4 involved 56,32,30 and 20 targets,respectively.Cluster 1,2 and 3 also contained most of the key targets,which may collaboratively interplay with each other and may act through alliance actions in the treatment of AS.The PPI network,the key genes and gene clusters are shown in Figure 2.

Figure 3 KEGG and GO pathway analysis.(A to D) KEGG pathway and GO enrichment analysis of cluster 1 to 4,respectively.

GO and KEGG pathway analysis for gene clusters

Different gene cluster favors different biologically meaningful patterns.A comprehensive set of functional annotation tools can be found in DAVID database which provides some useful tips for investigators to explore biological meaning behind large list of genes,such as enriched biological themes,particularly KEGG pathway maps and GO terms.KEGG pathway analysis showed that targets in different clusters were primarily distributed in pathways related to AS.According to the FDR value,the representative pathways are shown in R software.Of these,the cluster 1 mainly included adipocytokine signaling pathway (FDR=8.87 E-12),insulin resistance (FDR=4.21 E-11),AMPK signaling pathway (FDR=1.54 E-10).Genes in cluster 2 mainly participated in neuroactive ligand-receptor interaction (FDR=8.17 E-11),cAMP signaling pathway (FDR=1.65 E-07) and calcium signaling pathway (FDR=1.37 E-05),while genes in cluster 3 focused on arginine biosynthesis (FDR=4.96 E-05),some infection diseases and autoimmune diseases.cluster 4 is related to amino acid metabolism,including Tyrosine,Tryptophan and Phenylalanine.

Additionally,we analyzed the role of 1,2,3 and 4 cluster in GO,and obtained 277,76,90 and 47 GO biological progress (BP) entries,respectively.These GO terms were primarily involved in metabolism (glucose,fatty acid,bile acid and bile salt and amino acid),vascular physiology (smooth muscle contraction,cytosolic calcium ion concentration and blood pressure),inflammatory response (nitric oxide biosynthetic process,chemokine biosynthetic process,response to lipopolysaccharide and acute-phase response),and oxidation-reduction process.The KEGG and GO terms of cluster 1,2,3,and 4 are shown in Figure3.

Target and composition interaction analysis

To further validate the interactions between active compounds and key targets and improve the accuracy of network pharmacology,three targets (RXRA,AGT,and CXCL8) with highest degree (24,34 and 24) in different clusters and five active compounds (23-o-acetylalisol B,alisol A,alisol B,atractylenolide Ⅱ and atractylenolide Ⅲ)obtained from network pharmacology method were used to perform molecular docking.The binding modes of the components and targets with hydrogen or hydrophobic bonds were predicted.The conformations with a binding energy <-7 kcal were displayed via PyMOL.The 3D structures of RXRA and AGT were downloaded in PDB database directly.However,the structure of IL-8,encoded by CXCL8,in complex with its ligand has not been determined.Given that this target is closely related to the regulation of NF-κB [15],we used the structure of NFκB to explore the target component interactions.A common hydrophobic core consisted of LIE268,PHE313,LIE345,HIS435,LEU436 and PHE439 was formed within RXRA and active ingredients (atractylenolide Ⅱ and atractylenolide Ⅲ).23-O-acetylalisol B,alisol A,alisol B,atractylenolide Ⅱ and atractylenolide Ⅲ also interacted with AGT on a stable hydrophobic core,and some of them contacted AGT residues via the hydrogen bonds.Additionally,23-o-acetylalisol B and atractylenolide Ⅲinvolved some common interplay with ARG302 of NFκB.The detailed information are shown in Table 2.

Table 2 The binding energy of key targets with ingredients in AD

DISCUSSION

Atherosclerosis (AS) is one of the most common disease driven by disturbed flow,lipid disorders and inflammations,for which the current therapeutic strategies are not satisfying [16,17].Gradually,AD has become an effective agent for treatment of AS5.However,as seen with most other Chinese formulas,the shortage of content determination and pharmacodynamics limits the clinical practices of AD.Thus,in the present study,the active compositions and potential mechanisms of AD against AS was revealed by mass spectroscopy,high-performance liquid chromatography and network pharmacology.

Accordingly,UHPLC-Q-Orbitrap-HRMS indicated that the extract from AD contained multiple active compositions,including alisol A,alisol B,23-o-Acetylalisol B,atractylenolide Ⅱ and atractylenolide Ⅲ.Most of these chemicals can basically reflect the overall composition characteristics of AD,and provide a quality control standard of it [18].The biological activities of these chemicals were reported previously.Alisol A and alisol B are two common compounds acting on lipoprotein lipase(LPL) and acyl coenzyme a-cholesterol acyltransferase(ACAT),which may regulate lipid metabolism in human body [19,20].Wang et al.reported that alisol A remarkably inhibits the formation of arterial plaques and reduces IL-6 expression in ApoE-/-mice,where the expression of peroxisome proliferator-activated receptor α(PPARα) and AMPK signaling pathway are activated,but NF-κB pathway is inhibited [21].Yu et al.,suggested that 23-o-Acetylalisol B inhibits the absorption of dietary lipids in the jejunum by affecting the lipid transport,resulting in atherosclerotic plaque area and lipid accumulation reduction in ovariectomized ApoE-/-mice [22].As the bioactive compositions of Bai-shu,atractylenolide Ⅱ and atractylenolide Ⅲ exert their anti-inflammation and immunoregulatory effects in various AS-related diseases[23,24].Besides,it has reported that atractylenolide Ⅲ is a mediator regulating energy metabolism in skeletal muscle cells and has a beneficial effect on obesity [25].

The results of network pharmacology showed that AD may act on different stage of AS via regulating different gene clusters.Many targets in cluster 1 were predicted to play an important role in the treatment of AS because most of them participated in energy metabolism,especially in lipid disorders.The result of targets and functional enrichment analyses suggested that adipocytokine signaling pathway,insulin resistance and AMPK signaling pathway were commonly associated with the regulation actions of AD.All of these pathways are closely related to energy homeostasis [26].Adipocytokine consisting of leptin and adiponectin,is an important regulator of energy intake in humans and rodents.Besides,the increased number of adipocytes is positively correlated with leptin secretion,and negatively correlated with adiponectin production [27].Leptin is a peptide hormone produced by white adipose tissue,which shows its ability to atherogenic actions by inducing endothelial dysfunction,inflammatory reaction,and endarterium hyperplasia [28].The atherosclerotic lesions in the leptin-deficient mice remained as fatty streaks,while those in the control group were mainly fibrous plaques,indicating that leptin is one of the molecules contributing to the development of AS [29].Adiponectin is another peptide produced by adipocytes and endothelial cells,and the supplement of adiponectin has insulin-sensitizing,anti-atherogenic,and anti-inflammatory effects,which is considered to be the therapeutic target for AS [30].

Moreover,some targets in cluster 2 were participated in pathways regulating vascular physiology,such as cAMP signaling pathway.cAMP is one of the most common second messengers,and it not only regulates intracellular calcium-handling proteins and the contractile machinery,but also some transcription factors and transporters in our body.Some reports suggested that increased vascular smooth muscle cell (VSMC) proliferation contributes towards stenosis in AS [31],while the activation of cAMP inhibits VSMC proliferation in human aortic and coronary[32].Besides,cAMP and its downstream molecules may play a key role in endothelial function because it is critically required for nitric oxide release [33].

Our study also showed some targets in cluster 3 achieved their pharmacological actions via regulating inflammatory pathways.It has been reported that exposure of endothelial cells to disturbed flow has been shown to activate inflammatory signaling,including NF-κB signaling pathway,which inducing the expression of chemokines such as CXCL8,IL-6 and further promoting leukocyte adhesion [34].AS is an inflammatory disease,and the interplay among lipid metabolism,vascular physiology and inflammation at multiple levels may exacerbate the development of AS.Our study suggested that genes mediated these progresses may be potential molecular targets for the treatment of AS.

Thus,molecular docking was performed to validate our prediction.The potential targets with the highest degree in the different clusters of AD include RXRA,AGT and CXCL8.RXRA encodes retinoid X receptor alpha,and is a member of adipocytokine signaling pathway.In adipocyte and hepatocyte,RXRA and PPAR merge together to activate transcription of genes involved in adipocyte differentiation,inflammation,and energy homeostasis [35].AD can relieve lipid deposition in foam cells as it decreased the ox-LDL-induced expression of inflammatory cytokines by regulating LXRA pathway.Most importantly,the activated LXRA forms active DNA binding complexes with RXR or PPAR,which leads to gene regulation [36].As the center of cluster 2,AGT encodes angiotensinoge,which is the precursor of all the angiotensin peptides.To investigate the role of AGT in AS,Lu et al.inhibited ATG expression in the hypercholesterolemic mice,and the result showed that the mice without AGT expression in circulation had lower systolic blood pressure and less AS lesions [37].A review summarized that AD can act on renin-angiotensin aldosterone system,protect endothelial function,inhibit vascular remodeling and regulate K+-Na+transporters,which is consistent with our predictions [38].CXCL8 encodes C-X-C motif chemokine ligand 8,which is a chemokine and regulated by NF-κB.Our results showed that atractylenolide Ⅲ was verified to be strong combination with RXRA,AGT and CXCL8,while atractylenolide Ⅱ had strong binding with RXRA and AGT,and 23-o-Acetylalisol B has strong binding with AGT and CXCL8.These finding were consistent with previous studies that AD may be a multi-targets agent that exerts anti-atherosclerotic action by inhibiting inflammatory cytokine production,protecting vascular endothelial function and regulating lipid disorder [21,23].

CONCLUSION

In the present study,the active components and the anti-AS effects of AD were determined.Based on this comprehensive bioinformatic analysis,the regulation of lipid metabolism,vascular physiology and inflammation may be the most relevant mechanisms of anti-AS effects of AD.Alisol A,alisol B,23-o-Acetylalisol B,atractylenolideⅡ and atractylenolide Ⅲ may be the potential active ingredients of AD,because they could act on key genes such as RXRA,AGT and CXCL8 to regulate multipathways in AS.However,the pharmacological studies on the mechanisms and targets may be not enough in our study.Based on the findings from network pharmacology and molecular docking,future studies should be undertaken to assess the effects in terms of specific targets at the molecular levels via some experiments

ACKNOWLEDGMENT

This work was supported by National Natural Science Foundation of China,Grant/Award Numbers:82 004103;Natural Science Foundation of Guangdong Province,Grant/Award Numbers:2019A1515 011560;Open fund of Key Laboratory of Ministry of Education for TCM Viscera-State Theory and Applications,Liaoning University of Traditional Chinese Medicine,Grant/Award Numbers:zyzx1901;Scientific Research Fund of People's Hospital of Yangjiang,Grant/Award Numbers:G2020001 and G2 020002;Scientific Research Project of Guangdong Provincial Bureau of Traditional Chinese Medicine:Grant/Award Numbers:20201392,20211436 and 20 212269;Science and Technology Program of Yangjiang Science and Technology Bureau,Grant/Award Numbers:2019.50.25,Science and Technology Program of Yangjiang Health Bureau,Grant/Award Numbers:2019050.

The datasets mentioned and/or analyzed in the current study are available from the corresponding author on reasonable request.

Competing interests:The authors declare that they have no conflict of interest.

Citation:Bi YM,Zhang W,Lai YG,Fu YC,Li KZ,Wang Q,et al.Prediction of active ingredients and potential mechanisms of Alisma decoction against atherosclerosis:A study based on UHPLC-Q-Orbitrap-HRMS and network pharmacology.Drug Combination Therapy.2021;3(4):19.doi:10.53388/DCT2021110504.

Executive editor:Jin-Feng Liu.

Submitted:11 August 2021,Accepted:02 November 2021,Online:03 November 2021

? 2021 By Authors.Published by TMR Publishing Group Limited.This is an open access article under the CC-BY license (http://creativecommons.org/licenses/BY/4.0/)