[Bioinformatics analysis of ventilator-induced lung injury genome microarray based on gene expression omnibus database and key gene verification].

OBJECTIVE

To investigate differential expression gene (DEG) in mice with ventilator-induced lung injury (VILI) by bioinformatics analysis, and to verify the key genes by reproducing the VILI mouse model.

METHODS

(1) Experiment 1 (bioinformatics analysis): the microarray dataset of GSE9368 and GSE11662 regarding VILI mice and those in the spontaneous breathing control group were downloaded from the gene expression omnibus (GEO) database. DEG obtained by R and Venn map was further used to obtain common DEG. DAVID online database was used to obtain gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis. Finally, the protein-protein interaction (PPI) analysis of common DEG was carried out by using Search Tool for the Retrieval of Interacting Genes Database (STRING) and the key genes were screened out by using CytoScape software, molecular complex detection (MCODE) analysis plug-in and CytoHubba plug-in with maximum cluster centrality (MCC), maximum neighbor connectivity (MNC) and degree. (2) Experiment 2 (related protein verification): VILI mouse model was reproduced by high tidal volume (20 mL/kg) ventilator. Spontaneous breathing control group was set up. Hematoxylin-eosin (HE) staining was performed to assess lung injury and the key genes screened in experiment 1 were verified by immunohistochemical staining.

RESULTS

(1) Experiment 1 results: a total of 114 DEG were screened from GSE9368 dataset, including 99 up-regulated genes and 15 down-regulated genes. A total of 258 DEG were screened from GSE11662 dataset, including 188 up-regulated genes and 70 down-regulated genes. Furthermore, 66 common DEG were obtained, including 61 up-regulated genes and 5 down-regulated genes. GO analysis showed that the common DEG were mainly involved in inflammatory response, immune response, leukocyte and neutrophil chemotaxis. KEGG analysis showed that the common DEG were involved cell adhesion, cytokine receptor interaction and tumor necrosis factor (TNF) signaling pathway. STRING and CytoScape analysis were used to construct gene PPI network diagram and important sub modules. And the CytoHubba plug-in with MCC, MNC and degree algorithms was used to perform topology analysis and then taken an intersection to obtain eight genes including suppressor of cytokine signaling 3 (SOCS3), interleukin-1β (IL-1β), matrix metalloproteinase-9 (MMP-9), integrin Itgam, CXC chemokine ligand 2 (CXCL2), CXC chemokine receptor 2 (CXCR2), Sell and CC chemokine receptor 1 (CCR1). (2) Experiment 2 results: a mouse model of high tidal volume VILI was reproduced. Compared with the spontaneous breathing control group, the lung tissue was injured slightly at 0 hour after the end of ventilation, and the lung tissue structure was significantly damaged at 6 hours after the end of ventilation, showing bleeding in alveolar cavity, significant increase and collapse of alveolar wall thickness, and infiltration of inflammatory cells. The top three genes from intersection and topological analysis including IL-1β, SOCS3 and MMP-9 were verified by immunohistochemical staining. The results showed that the expressions of IL-1β, SOCS3 and MMP-9 were gradually increased with time of ventilation, the differences were found at 6 hours as compared with those in the spontaneous breathing control group [IL-1β (integral A value): 8.40±2.67 vs. 5.10±0.94, SOCS3 (integral A value): 9.74±1.80 vs. 5.95±1.31, MMP-9 (integral A value): 11.45±6.20 vs. 5.36±1.28, all P < 0.05].

CONCLUSIONS

Bioinformatics analysis based on GSE9368 and GSE11662 data sets found that VILI is mainly related to inflammatory injury, cytokines and immune cell infiltration; IL-1β, SOCS3 and MMP-9 might be biomarkers of VILI.