Integrating metabolomics and network pharmacology to assess the effects of quercetin on lung inflammatory injury induced by human respiratory syncytial virus

Abstract

Quercetin (QR) has significant anti-respiratory syncytial virus (RSV) effects. However, its therapeutic mechanism has not been thoroughly explored. In this study, a lung inflammatory injury model caused by RSV was established in mice. Untargeted lung tissue metabolomics was used to identify differential metabolites and metabolic pathways. Network pharmacology was used to predict potential therapeutic targets of QR and analyze biological functions and pathways modulated by QR. By overlapping the results of the metabolomics and the network pharmacology analyses, the common targets of QR that were likely to be involved in the amelioration of RSV-induced lung inflammatory injury by QR were identified. Metabolomics analysis identified 52 differential metabolites and 244 corresponding targets, while network pharmacology analysis identified 126 potential targets of QR. By intersecting these 244 targets with the 126 targets, hypoxanthine-guanine phosphoribosyltransferase (HPRT1), thymidine phosphorylase (TYMP), lactoperoxidase (LPO), myeloperoxidase (MPO), and cytochrome P450 19A1 (CYP19A1) were identified as the common targets. The key targets, HPRT1, TYMP, LPO, and MPO, were components of purine metabolic pathways. The present study demonstrated that QR effectively ameliorated RSV-induced lung inflammatory injury in the established mouse model. Combining metabolomics and network pharmacology showed that the anti-RSV effect of QR was closely associated with purine metabolism pathways.

Figure 1

QR attenuated the weight loss in RSV-infected mice. (A) Experimental protocol. (B) Body weight changes of the mice. Significance: ## P < 0.01 vs control group; # P < 0.05 vs control group; ** P < 0.01 vs model group; * P < 0.05 vs model group.

Figure 2

QR mitigated RSV-induced pulmonary histopathological damage and inhibited virus replication. (A) Pathological changes in lung tissue induced by RSV. Scale bar, 100 μm. (B) RSV expression assessed via an immunofluorescence assay. Scale bar, 50 μm. (C–E) Lung injury scores according to the degree of lung damage. (F) RSV fluorescence intensity. (G,H) RSV-G and RSV-F mRNA levels. Data are presented as the mean ± standard error. Significance: ## P < 0.01 vs control group; # P < 0.05 vs control group; ** P < 0.01 vs model group; * P < 0.05 vs model group.

Figure 3

QR reduced the production of pro-inflammatory cytokines in lung tissues of RSV-challenged mice. (A–E) mRNA Levels of IL1β, IL2, IL6, TNF-α, and IFN-γ in lung tissue. (F–I) Levels of IL1β and IL6 in lung tissue and serum. Data are presented as the mean ± standard error. Significance: ## P < 0.01 vs control group; # P < 0.05 vs control group; ** P < 0.01 vs model group; * P < 0.05 vs model group.

Figure 4

Multivariate data analysis of lung tissue metabolites. (A,B) PCA score plot in the positive and negative ion modes (n = 6). (C,D) PLS-DA score plot in the positive and negative ion modes (n = 6). (E,F) PLS-DA permutation test graph in the positive and negative ion modes (n = 200).

Figure 5

Volcanic map of the identified differential metabolites. (A,B) In Mod and Con groups, 239 metabolites were identified. (C,D) In QR and Mod groups, 160 metabolites were identified. P values were all < 0.05.

Figure 6

Analysis of Metabolic Pathways. (A) Metabolic pathway enrichment analysis of differential metabolites. Node size is based on enrichment ratio; node color is based on P value. (B) Enrichment ratios of three important metabolic pathways.

Figure 7

Prediction of the potential therapeutic targets and PPI Network Analysis. (A) Venn diagram of the potential therapeutic targets. (B) PPI analysis. (C) Drug-Target-Disease network diagram.

Figure 8

Biological function analysis. (A) GO and (B) KEGG enrichment analysis of the potential therapeutic targets.

Figure 9

Integrated analysis of metabolomics and network pharmacology. (A) Venn diagram of the key targets. (B) The compound-reaction-enzyme-gene network of purine metabolism. The red hexagons, gray diamonds, green round rectangles, and purple circles represent active compounds, reactions, proteins, and genes, respectively. (C–H) The peak areas of metabolites. Data are presented as the mean ± standard error. Significance: ## P < 0.01 vs control group; # P < 0.05 vs control group; ** P < 0.01 vs model group; * P < 0.05 vs model group.

Figure 10

Molecular docking patterns of QR with key targets. (A) HPRT1-QR. (B) TYMP-QR. (C) LPO-QR. (D) MPO-QR. The crystal structure of key targets were obtained from RCSB Protein Data Bank (PDB, http://www.rcsb.org/). The PubChem (https://pubchem.ncbi.nlm.nih.gov/) was used to prepare the chemical structure of QR. The molecular docking was executed by AutoDock-Vina 1.1.2 (https://vina.scripps.edu/). The PyMOL 2.3.0 (https://pymol.org/2/) and BIOVIA Discovery Studio 2016 (http://www.discoverystudio.net/) were applied for results processing and visualization.

Figure 11

Effects of QR on the key targets. (A–D) mRNA Levels of HPRT1, TYMP, LPO, and MPO in lung tissue. Data are presented as the mean ± standard error. Significance: ## P < 0.01 vs control group; # P < 0.05 vs control group; ** P < 0.01 vs model group; * P < 0.05 vs model group.