Network Pharmacology Approach to Elucidate Possible Action Mechanisms ofHippophae Rhamnoides Linn for Treating High Altitude Disease

Every year, large numbers of people are ascending to high altitudes for the purposes of pleasure and work. After ascent to high altitude (≥2500 m), the inability of the human body to adapt to the hypobaric and hypoxia environment can induce tissue hypoxia, then a series of high-altitude diseases (HAD) including Acute Mountain Sickness (AMS), High Altitude Pulmonary Edema (HAPE), and High-Altitude Cerebral Edema (HACE) would develop. These dis eases can develop at any time from several hours to 5 days and can range in severity from mild with minimal effect on the planned travel itinerary to life-threatening illness [1, 2]. Therefore, it is needed to develop new drug of the treatment of HAS. Traditional Chinese Medicine (TCM), which embraces centuries of knowledge and practical experience, has been used to treat many complex diseases in China [3-5]. In the highlands of our country, TCM show great poten- tial in the prevention and treatment of HAD because of their low price, rich resources and low side effect. Among them, Hippophae Rhamnoides Linn. (HRL) has been widely used in the treatment of HAS [6-7]. However, the underlying mechanisms of these anti-hypoxia effects remain unclear because of the complex compositions. Generally, the complex components in TCM exert their pharmacological effect through a multi-target and multi-pathway, which can hardly be elucidated by traditional methods. Recent developments in computational methods for screening digital compound libraries for multi-targeting drugs has resulted in a new branch of bioinformatics called network pharmacology. Network pharmacology is a integrity, synergy and dynamics analysis method based on disease, gene, protein target and drug interaction network. Up to now, it has been widely used to study TCM [8-12]. The purpose of the study is to screen the bioactive components in HRL and elucidate its molecular mechanisms in the treatment of HAS by network pharmacology. Finally, molecular docking were performed to validate the binding activities between the ingredients and anti-HAD targets of HRL.

Active compounds and corresponding target collection. Firstly, the chemical components of HRL were searched using TCMSP (http://tcmspw.com/tcmsp.php). Secondly, DL ≥ 0.18 and OB ≥ 30% were set as the thresholds to screen the active compounds via ADME analysis [13]. Thirdly, we screened the targets related to the active compounds using the TCMSP databases and obtained the Gene ID of the targets from the uniport (https://www.uniprot. org/) database.

Candidate target collection

The genes of targets associated with “high altitude sickness” and “high mountain sickness” disease names were collected through Gene Cards (https://www.genecards.org/), OMIM (https://www. omim.org/), Drug Bank database (https://www.drugbank.ca) and PharmGkb database (https://www.pharmgkb.org/). The Venny 2.1 (http://bioinfogp.cnb.csic.es/tools/venny/index.html) online tool was used to map the active ingredient target and the disease target of HAS and to draw Venn diagrams. After removing the duplicate genes, the overlapping genes of targets related to HAS were collected as the candidate targets.

Creating a PPI network

Common gene targets of HRL and HAS were imported into the STRING database (https://string-db.org/) to create a PPI network using the species qualified Homo sapiens, a confidence level of 0.9 and hidden disconnected nodes. Subsequently, the PPI network was visualized and further analyzed using Cytoscape software 3.7.2. The topological property of nodes in the interaction network were assessed by calculating six parameters with the CytoNCA plug-in: The six parameters which were “Degree Centrality (DC),” “Betweenness Centrality (BC),” “Closeness Centrality (CC),” “Eigenvector Centrality (EC),” “Network Centrality (NC),” and “Local Average Connectivity (LAC)” were used to measure the importance of nodes in the network. A node with high DC, BC, CC, EC, NC and LAC values means that it plays a highly important role in the network. Based on the results of topological property analysis of the above PPI, targets above the median were selected as the core targets.

GO and KEGG pathway enrichment

Potential targets were screened by the KEGG pathway and GO analysis (biological process, molecular function and cellular component) using the Metascape (https://metascape.org/) database. P-values (P<0.05) were considered as statistically significant, with a smaller P-value indicating a more significant correlation. The results of the analysis were selected the top 10 items with the highest enrichment and displayed them in the form of bubble graphs using website (http://www.bioinformatics. com.cn/).

Construction of the disease-target-compound network

To analyze the association among the candidate targets, the active compounds, a component-target network was obtained using Cytoscape 3.7.2 software. In the network, the nodes of different colors and shapes represented different targets and active compounds. Then, the core compounds were obtained through the component-target network.

Molecular docking

Core targets were obtained from the PPI network for molecular docking. With the help of the NCBI database (https://www.ncbi. nlm. nih.gov/pubmed/), the 2D structures of the ligand molecules were obtained and tranformated to 3D structures by chemoffice software and stored as a Mol2 file. The PDB-ID of the core targets ware also accessed. First, ligands for docking can be prepared through PyMOL selections including water deletion and the extraction of original ligands, ions and solvents. Then proteins and ligands were saved through Autodock Tools in PDBQT formats and were docked by perl software. Vina software was used to validate network pharmacology by molecular docking. Affinity (kcal/mol) is described by the minimum free energy which represents the degree of docking coincidence of molecules. The lower the minimum free energy is, the better the binding of ligands to receptor proteins is.

Active compounds and targets HRL

After screening was performed for OB ≥ 30% and DL ≥ 0.18, 33 active compounds and 423 related targets of the active compounds were searched via the TCMSP databases. Detailed information about these ingredients is provided in Table 1. Next, these targets were transformed into gene names and gene IDs via the Uniprot database, null and repetitive targets were deleted, and 182 effective active ingredient targets were obtained.

Candidate targets

By means of the four available resources, namely, the Gene cards, OMIM, Drug Bank and PharmGkb databases, 394 target genes were obtained. After merging HAS-related targets and active compound targets by venny website, 48 overlapping targets were recognized as candidate targets (see Figure 1).

PPI network analysis

Based on 48 candidate targets, PPI network was constructed by importing the gene ID of the candidate targets to the String database. Next, cytoscape 3.7.2 was used to visualize the PPI network. Based on the results of topological property analysis of the PPI, targets above the median were selected as the core targets (AKT1, MAPK14, CAT, TP53, AR, HIF-1A, VEGFA, NR3C1, CXCL8), and visualized by CytoNCA plug-in (see Figure 2).

GO enrichment analysis

To reveal the pharmacological mechanisms of HRL in HAS, 48 overlapping targets were put into the Metascape database for annotation GO enrichment analysis including biological process, molecular function and cellular component. The top 10 GO analysis results were screened, with P < 0.01 serving as the threshold, as shown in Figure 3. Among these terms, the biological processes were mostly related to the response to lipopolysaccharide, response to nutrient levels, reactive oxygen species metabolic process, monocarboxylic acid metabolic process, response to peptide, response to oxygen levels, positive regulation of cell migration, cellular response to organic cyclic compound, extrinsic apoptotic signaling pathway, regulation of reactive oxygen species metabolic process. These results help to elucidate the biological function changes in the body after treatment with HRL.

KEGG pathway enrichment analysis

KEGG pathway enrichment analysis of the 48 candidate targets was obtained through the Metascape database. Based on the “count” values and P ≤ 0.05, a total of 16 pathways were obtained, and the top 10 pathways are shown as the core pathways in Figure 4. The results indicate that the mechanisms of HRL against HAS were related to pathways in cancer (hsa05200), fluid shear stress and atherosclerosis (hsa05418), HIF-1 signaling pathway (hsa04066), VEGF signaling pathway (hsa04370), complement and coagulation cascades, and so forth.

Compound-target network analysis

The former results showed that 16 compounds and 48 targets may be the bio-active substances and the pharmaceutical targets for HRL in the treatment of AMS. Based on the compounds and predicted targets, we constructed a network using Cytoscape 3.7.2. As presented in Figure 5, the yellow nodes represent active compounds, and the blue nodes represent potential targets. The edges represent the interaction between them, and the node sizes are proportional to the node degrees. The network indicated the potential relationships between the compounds and the targets, thereby revealing the potential pharmacological mechanisms of HRL for the treatment of HAS. Through topological analysis, we selected the compounds with the highest degree value as the core compounds. As shown in Figure 5, quercetin, kaempferol and pelargonidin, are the core compounds in HRL (degree=40), (degree=16) and (degree= 8), respectively.

Docking results analysis

Based on the PPI network, we selected 9 core target genes for molecular docking. Affinity (kcal/mol) was used to value the score for the molecular docking, and an affinity < −7 indicated a stronger binding activity [14]. In Table 2, the affinity scores (< −7) of the former three ingredients obtained in 3.6 and targets are shown, which meant the components could act on the targets. The results indicated that the molecular docking results were consistent with the PPI screening results, and the reliability of hub gene was verified by molecular docking. The docking results of HRL to the HAS targets were shown in Table 2, and Figure 6 was presented to show the combination of quercetin and six targets including AKT1, AR, CXCL8, HIF-1A, VEGFA and TP53 in a 3D graph.

In this study, there were 33 active ingredients (16 key active ingredients) in HRL, and 48 targets were screened out for HAS treatment. Through enrichment analysis of GO biological processes and KEGG signaling pathways, the therapeutic mechanisms of HRL on HAS may primarily involve the following effects: response to lipopolysaccharide, peptide, response to nutrient levels, regulate reactive oxygen species metabolic process, act on HIF-1 signaling pathway and VEGF signaling pathway, and extrinsic apoptotic signaling pathway. Moreover, it was preliminarily predicted that HRL may regulate the core targets (AKT1, MAPK14, CAT, TP53, AR, HIF- 1A, VEGFA, NR3C1, and CXCL8) through quercetin, kaempferol and pelargonidin, which have been validated by molecular docking. To date, some of the predicted targets such as HIF-1A, VEGFA, NR3C1 have been confirmed by published literatures [15-19], moreover, Dexamethasone, which acts on NR3C1 has been used in clinical practice of HAS treatment [20, 21]. As for AR (androgen receptor), which several components including pelargonidin, quercetin, and kaempferol could bind tightly to it (Table. 2) and many researchers have reported the correlation between HAS and sex [22, 23], however, the internal mechanism of HRL acting on AR needed further research. This study not only explored the potential molecular mechanisms regarding the anti-hypoxia effect of active components in HRL, which laid the foundation for further investigation, but also presented novel clues for the development and utilization of HRL resources.

Not applicable.

None.

No conflict of interest.

1. Anna HK, Jolanta KZ, Bogumił L (2016) High altitude illness. Przegl Epidemiol 70 (3): 490-499.
2. Andrew M L, Erik RS, Peter B (2017) Acute high-altitude sickness. Eur Respir Rev 26(160096): 1-14.
3. Ga ZC, San ZJ, Nan JT, Luo SD, Zha XDZ, et al. (2019) Research Status and Exploration of Prevention and Treatment of Plateau Disease with Tibetan Medicine. Mod Chin Med May 21(5): 694-698.
4. Jin WJ, Ma HP, Yang SP, Jing LL, Fan PC (2020) Research progress of anti-hypoxia traditional chinese medicine. Med & Pharm J Chin PLA 32(9): 113-117.
5. Ma YZ, Wu Q, Ji BM, Yuan HL, Zhao WJ, et al. (2020) Anti-hypoxia efficacy and mechanism of traditional Chinese medicine Kang-Gao-Yuan-Bing-Fang. Medical Journal of Air Force 36(3): 226-230.
6. Zhang XH, Xi AQ, A XR, Lv XM, Zhao SX (1996) Therapeutic effect of compound tianji capsule in patients wit in high altitude polycythemia. Jounrnal of high-altitude medicine 6(4): 51-54.
7. Ma HY, Chen JY, Zhang LS (2011) Protective effects of Tianji capsule on vascular endothelial cells from oxidative injury induced by hydrogen peroxide in vitro. J Sichuan Univ (Med Sci Edi) 42 (1): 44-7.
8. Shi B, Cui QH, Feng Zhenlong, Li JX (2019) Mining potential drugs for prevention and treatment of acute mountain sickness based on bioinformatics approach. Chin J Clin Healthc 22 (4): 503-507.
9. Hopkins AL (2008) Network pharmacology: the next paradigm in drug discovery. Nat Chem Biol 4: 682-690.
10. Christopher D, Peter H (2017) Advances in the Prevention and Treatment of High-Altitude Illness. Emerg Med Clin N Am 35: 241-260.
11. Liu ZH, Sun XB (2012) Network pharmacology: new opportunity for the modernization of traditional Chinese medicine. Acta Pharmaceutica Sinica 47 (6): 696-703.
12. Liu AL, Du GH (2010) Network pharmacology: new guidelines for drug discovery. Acta Pharmaceutica Sinica 45(12): 1472-1477.
13. Wang, Y, Zheng, C, Huang, C, Li, Y, Chen, X, et al. (2015) Systems pharmacology dissecting holistic medicine for treatment of complex diseases: an example using cardiocerebrovascular diseases treated by TCM. Evid. Based Compl. Alternat 980190: 1-19.
14. Hsin KY, Ghosh S, Kitano H (2013) Combining machine learning systems and multiple docking simulation packages to improve docking prediction reliability for network pharmacology 8(12): e83922.
15. Kang Li, He CH (2019) Gastric Mucosal Lesions in Tibetans with High-Altitude Polycythemia Show Increased HIF-1A Expression and ROS Production. BioMed Research International. Hindawi. BioMed Research International pp.1-11.
16. Liu HH (2019) The expression of VHL/HIF signaling pathway in the erythroid progenitor cells with chronic mountain sickness. National Medical Journal of China 99(34): 2670-2674.
17. Yang YZ, Du H, Li YH, Guan W, Tang F, et al. (2019) NR3C1 gene polymorphisms are associated with high- altitude pulmonary edema in Han Chinese. Journal of Physiological Anthropology 38(4): 1-8.
18. Du H, Zhao J, Su ZH, Liu YN, Yang YZ (2018) Sequencing the exons of human glucocorticoid receptor (NR3C1) gene in Han Chinese with high-altitude pulmonary edema. Journal of Physiological Anthropology 37(7): 1-5.
19. Zhang JH, Yang Shen, Chuan Liu, Jie Yang, Yuan-Qi Yang, et al. (2020) EPAS1 and VEGFA gene variants are related to the symptoms of acute mountain sickness in Chinese Han population: a cross-sectional study. Military Medical Research 7(35): 1-12.
20. Jiang XL, Xu XL, Sun HG, Huang H (2014) Drugs for high altitude pulmonary hypertension: A Systematic review. Chinese General Practice 17(31): 3729-3733.
21. Ma J, Fan PC, Zhang Q, Ma HP, Wang R, et al. (2013) Clinical pharmacology in the prophylaxis and treatment of high-altitude illness. Journal of Pharmaceutical Practice 31(4): 246-250.
22. Shen Y, Yang YQ, Liu C, Yang J, Zhang JH, et al. (2020) Association between physiological responses after exercise at low altitude and acute mountain sickness upon ascent is sex dependent. Military Medical Research 7(53): 1-9.
23. Hou YP, Wu JL, Tan C, Chen Y, Guo R, et al. (2019) Sex-based differences in the prevalence of acute mountain sickness: a meta-analysis. Military Medical Research 6(38): 1-12.