Prediction of the targets of the main components in blood after oral administration of Xanthii Fructus: a network pharmacology study

Xanthii Fructus (XF), a famous traditional Chinese medicine (TCM), has been widely used in the treatment of rhinitis and other diseases. However, the targets of the main XF components found in the blood after oral administration of XF extract are still unclear. In the current study, a feasible systems pharmacology method was developed to predict these targets. In accordance with our previous research, XF components were selected including cleomiscosin A, myristic acid, succinic acid, xanthosine, sitostenone, emodin, apigenin, and chrysophanol. Three components, namely emodin, apigenin, and chrysophanol, failed to be detected with target proteins, thus the other five components, namely cleomiscosin A, myristic acid, succinic acid, xanthosine and sitostenone, were eventually chosen for further systematic analysis. Ninety-nine target proteins and fifty-two pathways were found after a series of analyses. The frequency of some target proteins was much higher than that of others; high frequencies were obtained for P15086, P07360, P07195, MAOM_HUMAN (P23368), P35558, P35520, ACE_HUMAN (P12821), C1S_HUMAN (P09871), PH4H_HUMAN (P00439), FPPS_HUMAN (P14324), P50613, P12724, IMPA1_HUMAN (P29218), HXK1_HUMAN (P19367), P14061, and MCR_HUMAN (P08235). The frequency of eight pathways was also high, including Generic Transcription Pathway, RNA Polymerase II Transcription, Metabolism, Metabolism of steroids, Gene expression (Transcription), Cellular responses to stress, Platelet activation, signaling and aggregation, Signaling by Receptor Tyrosine Kinases, and Cellular Senescence. This study identified a common pathway – the Metabolism pathway – for all five XF components. We successfully developed a network pharmacology method to predict the potential targets of the main XF components absorbed in serum after oral administration of XF extract.

of TCM modernization. In a network pharmacology study, drug-drug networks can be constructed based on the similarities in the structures and efficacies of different drugs. In the process of TCM modernization, some researchers have achieved good initial results in exploring the essential properties of TCMs and revealing their comprehensive overall effects on multi-pathways, multi-targets and multi-components via the research ideas of network pharmacology. [8][9][10][11] Xanthii Fructus (XF) is the ripe fruit of Xanthium sibiricum Patr. XF is used for the treatment of cramping, numbness of the limbs, ulcers, sinusitis, catarrhs, and pruritus, for its function in smoothing nasal orices and eliminating wind-dampness. 12 In modern clinic application, XF is commonly used for the treatment of rhinitis. Particularly when combined with Magnoliae os, mint and other Chinese medicines, XF has enhanced effects in curing chronic rhinitis, allergic rhinitis and other rhinitis. 13

Screening active ingredients
In our previous study (unpublished), components such as myristic acid, succinic acid, xanthosine, emodin, apigenin, and chrysophanol were identied from serum samples aer oral administration of XF extracts. Components such as cleomiscosin A and sitostenone were ltered using the traditional Chinese medicine systems pharmacology (TcmSP™) database, and the parameters were set as follows: oral bioavailability (OB) $ 30%, drug-likeness (DL) $ 0.18. The structures of the components mentioned above are shown in Fig. 1.

Prediction of active component targets
Firstly, the MDL SD (*.sdf) type les of the above active ingredients were searched using the PubMed database. Secondly, targets, including information like the target name, matching value, target protein abbreviation, function, disease and applicable results related to the modied compound, could be predicted by importing each component le in *sdf format into the PharmMapper database. The top 20 high-matching targets, by value, were used as the TCM target proteins related to the components. The targets were then searched for in the UniProt database to identify human-related target codes.

Pathway comments and analysis
The retrieved target protein information was analyzed using the Reactome database to obtain the result of the related pathway "pathwayIdexByPathway_kegg". A pathway was selected as reliable when its P value was less than 0.01.

Drug-target-pathway relationship
The predicted targets of ve chemical components of XF, namely cleomiscosin A, myristic acid, succinic acid, xanthosine and sitostenone, were recorded in excel tables titled as 'component-protein' and 'protein-pathway'. The tables were imported into Cytoscape soware to construct the main effect components of the XF-target-pathway network. The network was mainly composed of three types of nodes: effect component, protein and pathway. The effect components and their related target proteins, and the proteins and their related pathways were all side-linked. When the target protein of the effect component was the same as the target protein of the pathway, the effect component was side-linked to the pathway.

Pathway analysis of potential target proteins
The potential pathway information for the ve effect components in XF is shown in Table 2.

Main effect component-target protein-pathway network construction for XF
An effect component-target-pathway network model was established using Cytoscape soware, and the relationship between the 5 components, 99 targets and 52 pathways is shown in Fig. 3. There were complex network relationships between the effect components of XF and their targets, as well as the targets and pathways.

Discussion
The PharmMapper database can be used to search for potential targets based on small active molecules. This database uses a pharmacophore matching method to obtain drug point information by rapidly searching four major databases. This database is based on 7000 pharmacophore models and can cover most clinical indications. According to the network pharmacological prediction of the ve components in XF, all ve components can be connected with the same pathway via the same target, and also can be connected with the same pathways with different targets. Different components can produce the same effect through different ways, and also can offer multi-target synergy.
Interestingly, this predicted common pathway is consistent with the result we got from the metabolic pathway analysis experiment (unpublished), which indicates that this result is reliable although it still requires further verication.

Conclusion
In this paper, a network pharmacology method has been successfully developed to predict the potential targets of the main components absorbed in serum aer oral administration of XF extract. When considered alongside our previous antiallergic rhinitis metabolomics study, the predicted potential targets and the role of the pathways were considered to have a certain degree of accuracy. This article has established a "multi component-multi target-multi pathway" network model for TCM research, and started to unravel the multidimensional regulatory action of XF, which may provide a reference and basis for studying the molecular mechanism of XF.

Conflicts of interest
The authors have declared no conicts of interest.