Investigation on the Mechanism of Qubi Formula in Treating Psoriasis Based on Network Pharmacology

Objective To elucidate the pharmacological mechanisms of Qubi Formula (QBF), a traditional Chinese medicine (TCM) formula which has been demonstrated as an effective therapy for psoriasis in China. Methods The Traditional Chinese Medicine Systems Pharmacology (TCMSP) database, BATMAN-TCM database, and literature search were used to excavate the pharmacologically active ingredients of QBF and to predict the potential targets. Psoriasis-related targets were obtained from Therapeutic Target Database (TTD), DrugBank database (DBD), MalaCards database, and DisGeNET database. Then, we established the network concerning the interactions of potential targets of QBF with well-known psoriasis-related targets by using protein-protein interaction (PPI) data in String database. Afterwards, topological parameters (including DNMC, Degree, Closeness, and Betweenness) were calculated to excavate the core targets of Qubi Formula in treating psoriasis (main targets in the PPI network). Cytoscape was used to construct the ingredients-targets core network for Qubi Formula in treating psoriasis, and ClueGO was used to perform GO-BP and KEGG pathway enrichment analysis on these core targets. Results The ingredient-target-disease core network of QBF in treating psoriasis was screened to contain 175 active ingredients, which corresponded to 27 core targets. Additionally, enrichment analysis suggested that targets of QBF in treating psoriasis were mainly clustered into multiple biological processes (associated with nuclear translocation of proteins, cellular response to multiple stimuli (immunoinflammatory factors, oxidative stress, and nutrient substance), lymphocyte activation, regulation of cyclase activity, cell-cell adhesion, and cell death) and related pathways (VEGF, JAK-STAT, TLRs, NF-κB, and lymphocyte differentiation-related pathways), indicating the underlying mechanisms of QBF on psoriasis. Conclusion In this work, we have successfully illuminated that Qubi Formula could relieve a wide variety of pathological factors (such as inflammatory infiltration and abnormal angiogenesis) of psoriasis in a “multicompound, multitarget, and multipathway” manner by using network pharmacology. Moreover, our present outcomes might shed light on the further clinical application of QBF on psoriasis treatment.


Introduction
Psoriasis is a common and frequently occurring disease in dermatology, which is characterized by easy diagnosis and difficult treatment, as well as recurrent disease course [1]. An epidemiological survey shows that the overall prevalence of psoriasis is approximately 0.5% in China [2]. e pathogenesis of psoriasis is still not completely clarified. Present studies have suggested that autoimmune disorders, dysfunction in various inflammatory signal transduction pathways, abnormal expression of psoriatic susceptibility gene, and obesity might be involved in the pathogenesis of psoriasis [3][4][5][6][7][8][9]. e unknown pathogenesis has brought difficulties to the treatment, without curative approaches against psoriasis at present. Traditional Chinese medicine (TCM) has unique advantages in treating psoriasis, which can play therapeutic roles through multiple targets and multiple pathways, corresponding to the dysfunction of various pathways underlying the pathogenesis of psoriasis [10][11][12][13]. However, TCM also has defects. Due to the unclear component entering the blood through the compound TCM, the mechanism of action is not completely clear, which restricts the further standardization and internationalization of TCM for psoriasis.
Conventional researches on the mechanism of one traditional Chinese medicine mostly follow the model of "one drug-one target-one disease," which cannot reflect the characteristics of TCM (multicompound, multitarget, and multipathway). Herein, in this study, a comprehensive approach [14] (a combination of multiple network-based computational and algorithm-based approaches) was utilized, by combining prediction of active compounds based on multiple pharmacokinetic parameters, excavation of diverse drug targets, and network analysis from a macroscopic perspective, aiming at the illumination of the underlying mechanisms of QBF on psoriasis and providing ideas for subsequent research. [15] (http:// bionet.Ncpsb.Org/batman-tcm/index.Php/Home/Index/ index) is a bioinformatics analysis tool for analyzing pharmacological active ingredients of Chinese medicines. In order to obtain the information about the ingredients of QBF, "SHUI NIU JIAO, DI HUANG, CHI SHAO, MU DAN PI, ZI CAO, JIN YIN HUA, LIAN QIAO, BAN LAN GEN, and GAN CAO" were used as key words to search in the BATMAN database, giving rise to a total of 572 compounds.

Screening of Potential Pharmacological Active Ingredients and Targets of Qubi Formula. BATMAN-TCM
e Traditional Chinese Medicine Systems Pharmacology (TCMSP) database [16] (http://tcmspw.com/tcmsp.php) is a unique platform which can provide pharmacokinetic properties (involving oral bioavailability, drug-likeness, aqueous solubility, etc.) and potential targets for natural compounds. We searched the above 572 compounds in the TCMSP platform to obtain their pharmacokinetic parameters. After obtaining the pharmacokinetic information on the 572 compounds, the reference was screened according to oral bioavailability (OB) ≥30% and drug-likeness (DL) ≥0.18 (mean value for all molecules within the DrugBank database), aiming to screen potential pharmacologically active ingredients of QBF [17]. In this study, these cut-off values utilized helped to efficiently and maximally collect data from QBF using the least components, and the pharmacokinetic data reported may account for this [18]. OB [19], defined as the distribution degree of an oral dose of drug into bloodstream, is one of the most requisite premises in terms of oral drug discovery as well as clinical application. Additionally, drug-likeness, which is defined as a qualitative concept for assessment of the structural similarity of compounds with clinical therapeutics in the DrugBank database, is determined early after drug discovery [20]. In addition, through literature review, certain compounds with the OB < 30 or DL < 0.18 but with extensive pharmaceutical activities (such as oleic acid [21] and arnebinol [22]), or those with relatively higher contents (like jioglutin and alkannin) or those used for the quality identification of single herb in the Pharmacopoeia (goitrin), were also added as potential pharmacologically active ingredients of QBF.
Apart from assisting in exploring the active ingredients of TCM, TCMSP databases could also predict the potential targets of compounds based on SysDT model, HIT database, reverse molecular docking, etc. [23].

Collection of Known Psoriasis-Related Targets, Psoriasis.
In order to obtain the known psoriasis-related targets, "psoriasis" was used as key word to search in erapeutic Target Database (TTD) [24], DrugBank database [25], MalaCards database [26], and DisGeNET database [27]. After searching in the DisGeNET database, results were sorted by the disease specificity index (DSI), following by the removal of targets lower than the median of DSI obtained from all the known psoriasis-related genes. Additionally, the drugs with abnormal status in TTD and DrugBank database and their corresponding targets were also taken out. e detailed information of these known psoriasis-related was summarized in Table S1 after redundancy deletion.

Excavation of Core Targets of QBF for Treating Psoriasis
and Construction of Core Network on Active Ingredients-Targets. First, targets obtained from the above two steps (potential targets of QBF and known psoriasis-related targets) were standardized in the UniProt database [28] by selecting the species "Homo sapiens," aiming to acquire the single universal gene names. en, both potential targets of QBF and known psoriasis-related targets were uploaded to the online Wayne diagram tool (http://bioinfogp.cnb.csic.es/ tools/venny/index.html, Version 2.1.0) for mapping; that is, targets from these two sets were intersected to obtain the candidate targets of QBF for treating psoriasis.
Subsequently, the candidate targets were imported into the String database [29] to obtain protein-protein interactions (PPIs) by setting the minimum value of the combined score at 0.400 and the species as "Homo sapiens." e topological parameters (DMNC, Degree, Closeness, and Betweenness) of each target (node) in the network were calculated using the cytoHubba plugin [30].
e median values of these four parameters of all nodes were used as a screening condition. Nodes with all the four parameter greater than the median values were considered as the main hubs that played core roles in the PPI network, that is, the core targets of QBF for psoriasis. Finally, Cytoscape was used to construct the core network of active ingredients-targets.

Enrichment Analysis of Core Targets.
e ClueGO [31] plugin from Cytoscape software, integrative GO-biological process (BP), and KEGG database were applied to perform enrichment analysis on the core targets, and species was selected as "Homo" in the ClueGO interface. All the core targets were sequentially imported, followed by enrichment analysis. κ value was defaulted at 0.4 and p was set at ≤0.05, which were used as the screening conditions for plotting enrichment analysis.

Screening of Potential Pharmacologically Active Ingredients and eir Targets of Qubi Formula.
e search through the BATMAN-TCM database revealed a total of 572 compounds of QBF. Accumulative efforts have been made to clarify the therapeutic mechanisms of TCM, however, with sluggish progress on the molecular level. Due to the unavailable effective methods specifically developed for the identification of the active compounds in medicinal herbs, OB screening combined with drug-likeness assessment may be a feasible strategy. In this study, OB ≥ 30% and DL ≥ 0.18 were used as the screening conditions. en, a total of 202 possible compounds with proper values of above two parameters were collected for potential pharmacologically active ingredients from the herbal constituents of QBF. In addition, among the compounds that were screened out, we found another 35 compounds through the literature research of PubMed. Although these 35 compounds did not meet the screening conditions of OB and DL, they were reported to use a wide range of pharmacological activities and were thus included in potential pharmacologically active ingredients. Finally, the active ingredients of SNJ, GC, CS, DH, MDP, ZC, JYH, LQ, and BLG were 6, 97, 26, 6, 12, 17, 30, 27, and 51, respectively. e proportions of the active ingredients in the sovereign drug (Jun), minister herbs (Chen), assistant herbs (Zuo), and messenger herbs (Shi) in all the 175 ingredients were 3.4%, 30.86%, 53.14%, and 55.43%, respectively. Among them, some compounds were widely present in multiple herbs of QBF, such as oleic acid, methyl linolenate, sitosterol, and paeoniflorin. e basic information of the potential pharmacologically active ingredients of Qubi Formula is shown in Table 1.
Subsequently, we explored the potential targets of the 237 potential pharmacologically active ingredients by excavating TCMSP databases, which yielded to 939 targets (shown in Table S2). e numbers of potential targets linked by SNJ, GC, CS, DH, MDP, ZC, JYH, LQ, and BLG were 668, 234, 148, 49, 207, 90, 219, 229, and 111, respectively. e proportions of the potential targets of the sovereign drug (Jun), minister herbs (Chen), assistant herbs (Zuo), and messenger herbs (Shi) in all the 939 targets were 71.14%, 11.93%, 30.35%, and 24.92%, respectively. Although the number of targets correlated with each herb of QBF is different, significant overlaps were observed in the nine herbs, which was suggestive of the congenerous or antergic roles of the various components in QBF via the regulation of similar targets.
In order to holistically and systemically obtain comprehensive understanding of the ingredient-target network in QBF, a network map was constructed by using Cytoscape, including 6273 edges and 1212 nodes ( Figure 1). To be specific, the node degree indicated the number of target or edge correlated with the node according to topological analysis. A total of 142 ingredients were found in the as-established network to have the median of ≥18 degrees. Of them, quercetin, arginine, and oleic acids kaempferol and luteolin acted on 166, 96, and 54 targets, respectively, which were subsequently considered as the critical pharmacologically active ingredients of QBF.

Excavation of the Core Targets of Qubi Formula in Treating
Psoriasis. Psoriasis has been recognized as the polygenic disorder. In addition, the investigation of the interactions between genes as well as gene and environment could be used to reveal the pathogenesis of psoriasis. After the targets with abnormal status from TTD and DrugBank database and DSI < 0.535 (the median of DSI) from DisGeNET database were excluded, we collected 605 targets (Table S1) associated with psoriasis from the four accessible resources. Notably, 104 of the identified potential targets of the QBF were also the well-recognized psoriasis disease-(or therapeutic drugs) related targets (Table S3 and Figure 2(a)). And these 104 targets were defined as the candidate targets for QBF in treating psoriasis.
Subsequently, to further select the core targets of QBF in treating psoriasis, the String database was used to construct the PPI network of the above targets ( Figure 2(b)), followed by the calculation of the topological parameters (DMNC, Degree, Closeness, and Betweenness) of each node in the network using the cytoHubba plugin. e median values of these four parameters of all nodes were used as a screening condition. Nodes with all the four parameters greater than the median values were considered as the main hubs that played core roles in the PPI network. As a result, 27 targets ( Table 2 and Table S4) were screened from the 104 candidate targets based on the values of topological parameters ( Figure 2(c)), that is the core targets of QBF in treating psoriasis.

Construction of Active Ingredients-Targets Core Network for Qubi Formula in Treating Psoriasis.
In order to further understand the "multicompound and multitarget" mechanism of Qubi Formula in treating psoriasis, we searched for the potential ingredients of Qubi Formula which could affect the 27 core targets based on the relationship between the ingredients and their targets, followed by construction of the core network on active ingredients-targets (Figure 3(a)) using Cytoscape software and the statistical analysis of the degree of each node in the network. As shown in Figure 3(b), the degrees of active ingredients ranged from 1 to 18 in the core network, with the median of 2, indicating that more than half of the compounds acted on more than one target, while the degree of the target ranged from 1 to 160, with the median of 2. Among all the active ingredients, the top 4 were quercetin, luteolin, kaempferol, and wogonin in terms of Degree. Previous preclinical studies have confirmed that all of them could delay the progression of psoriasis and relieve symptoms in multiple animal models [32][33][34]. Among all the targets, the top three were PTGS2, MAPK14, and NOS3 in terms of Degree, and all of them have been demonstrated to play important roles in the pathogenesis of psoriasis, such as inflammatory infiltration, abnormal differentiation of keratinocytes, and oxidative stress injury [35,36].  Figure 1: Construction of the QBF compound-potential target network. e compound-potential target network was constructed by linking the candidate compounds and their potential targets of the 9 herbs, which are constituents of QBF. e nodes representing candidate compounds are shown as polychrome square and the targets are indicated by orange circle.  FASLG  ALOX5  IRF1  GATA3  CXCL11  CD40LG   CXCL10  SLPI  SNAI1  LCN2  IL1A  IFNG  FAS  IL2  CCND1  IL4   TLR4  BIRC5  JAK2 MAPK14  VCAM1  MMP9  NFKB1  IGF2  KDR  ICAM1  NFKBIA  IL10 CXCL8  PTGES  RELA  PTEN  SELE  STAT3  MMP2  EGF  MMP1  BCL2L1  JUN  IL1B  MAPK8  BAX  IL6  TNF  MMP3  CCL2  ARG1  TP53  TGFB1  VEGFA  CRH  EGFR  HSP90AA1  CRP  MAPK1  PTGS2  PLAU  DPP4  PCNA  IGF1  PRKCA  MAPK3  INS  BCL2  NOS2  ALB  SPP1  PTPN1  LEP  PPARG  EDN1 BDNF   HMOX1  LRRK2  HIF1A  SOD2  FABP5  ADIPOQ  NOS3  APOE  RXRA  GPT   CAT  PPARA  AHR  RXRB  SLC6A4  SOD1 VDR RARS Evidence-Based Complementary and Alternative Medicine

Enrichment Analysis of the Core Targets of Qubi Formula in Treating Psoriasis.
In order to further understand the mechanism of "multitarget and multipathway" of Qubi Formula in treating psoriasis, ClueGO plugin was used to perform enrichment analysis of GO-PB and KEGG on core targets and to excavate the biological processes and signaling pathways regulated by Qubi Formula in treating psoriasis. ese 27 core targets were involved in several biological process, mainly including nuclear translocation of proteins, cellular response to multiple stimuli (immunoinflammatory factors, oxidative stress, and nutrient substance), lymphocyte activation, regulation of cyclase activity, cell-cell adhesion, and cell death (Figure 4(a)). Moreover, according to the pvalues of enriched pathways and their correlation with psoriasis, we were most interested in the following five representative signal pathways including VEGF, JAK-STAT, TLRs, NF-κB, and lymphocyte differentiation-related pathway (Figure 4(b) and Table 3).

Discussion
Qubi Formula is an experience prescription for psoriasis at the Department of Dermatology in our hospital. It is especially suitable for patients with blood-heat subtype of psoriasis, with radiated skin lesions throughout the whole body, redness, obvious scales, itching and burning, red tongue, and yellow fur. In this formula, SNJ, DH, JYH, and BLG can clear heat, cool blood, and remove toxic materials. MDP and CS are responsible for cooling blood and removing blood stasis. GC is in charge of detoxifying and reconciling medicine. e therapeutic effects of Qubi Formula in the clinical treatment of psoriasis are significant. However, the active ingredients and potential targets of Qubi Formula are unclear, which hinders the further development and application of the prescription.
Network pharmacology is a new strategy for drug design and development based on the rapid development of systematical biology and multidirectional pharmacology. is concept was first proposed by Hopkins AL in 2007, which was switched from previous "disease-single target-single    drug" model of new drug development to the "diseasemultitarget-multidrug" model. is idea coincides with the "holistic view" of TCM. erefore, the application of the network pharmacology method can provide certain research ideas for discovering the mechanism of Qubi Formula in treating psoriasis.
Enrichment analysis of GO-BP and KEGG on the core targets further suggests that Qubi Formula could intervene in psoriasis through multiple biological processes by acting on several signaling pathways, involving VEGF, JAK-STAT, TLRs, NF-κB, and lymphocyte differentiation-related pathways. As shown in the Figure 4(a), these five signaling pathways cross-talk effects in the network. VEGF signal pathway has been demonstrated to not only induce pathological angiogenesis in psoriatic lesions by regulating the proliferation and differentiation of endothelial cells, but also aggravate the inflammatory response via increasing vascular permeability to promote inflammatory cell infiltration [57,58]. Moreover, as it is well known that psoriasis is an inflammatory disease mediated by T lymphocytes, abnormal differentiation of T lymphocytes (especially 1 and 17 cells) and excessive secretion of proinflammatory factors (such as ILs) are closely related to the progression of the disease [59]. In this study, we have also demonstrated that there are diverse critical signaling pathways related to T lymphocytes differentiation and proinflammatory factors production (TLRs, JAK-STAT, and NF-κB) regulated by QBF on psoriasis therapy [60][61][62].
Our team has previously confirmed the safety and effectiveness of Qubi Formula in treating psoriasis through clinical observations. Based on the results of this study, we speculate that the regulatory role of Qubi Formula on psoriasis is not unilateral, but is directly or indirectly involved in the comprehensive treatment of the four major pathological factors of psoriasis through multiple signaling pathways associated with immunoinflammatory response, metabolism, and abnormal angiogenesis. Despite the valuable discoveries, there are still certain limitations. In the present study concerning the network pharmacological analysis on Qubi Formula, only the interactions between the components of QBF and the psoriasis-related targets were considered, but the interactions between the ingredients, the dosage of each ingredient, and the effects of the different processing methods of medicinal materials were neglected. erefore, the obtained results must be verified by further experiments.

Conclusion
e present study illustrates the systemic "multicompound and multitarget" efficacy of QBF against psoriasis. Moreover, this study also provided a theoretical basis to determine the synergistic effects of TCM in treating diseases and the role of systematic network pharmacology in elucidating the potential mechanisms of action of TCMs. However, as this study was based on data mining and data analysis, further validation studies should be undertaken.

Data Availability
e data used to support the findings of this study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that they have no conflicts of interest regarding the publication of this paper.