Gene Co-Expression Analysis Identified Preserved and Survival-Related Modules in Severe Blunt Trauma, Burns, Sepsis, and Systemic Inflammatory Response Syndrome

Abstract

Jingrui Huo,1,* Lei Wang,2,* Yi Tian,2 Wenjie Sun,1 Guoan Zhang,1 Yan Zhang,1 Ying Liu,1 Jingjing Zhang,1 Xiaohui Yang,1 Yingfu Liu3 1Science and Technology Experiment Center, Cangzhou Medical College, Cangzhou, 061001, People’s Republic of China; 2Microbiology and Immunology Department, Cangzhou Medical College, Cangzhou, 061001, People’s Republic of China; 3Cangzhou Nanobody Technology Innovation Center, Cangzhou Medical College, Cangzhou, 061001, People’s Republic of China*These authors contributed equally to this workCorrespondence: Xiaohui YangScience and Technology Experiment Center, Cangzhou Medical College, Cangzhou, 061001, People’s Republic of ChinaTel [phone]Email [email protected] LiuCangzhou Nanobody Technology Innovation Center, Cangzhou Medical College, Cangzhou, 061001, People’s Republic of ChinaFax [phone]Email [email protected]: Severe trauma and burns accompanied by sepsis are associated with high morbidity and mortality. Little is known about the transcriptional similarity between trauma, burns, sepsis, and systemic inflammatory response syndrome (SIRS). Uncovering key genes and molecular networks is critical to understanding the biology of disease. Conventional analysis of gene changes (fold change) analysis is difficult for time-serial data such as post-injury blood transcriptome.Methods: Weighted gene co-expression network analysis (WGCNA) was applied to the trauma dataset to identify modules and hub genes. Module stability was tested by half sampling. Module preservations of burns, sepsis, and SIRS were calculated using trauma as reference. Module functional enrichment was analyzed in gProfiler server. Candidate drugs were screened using Connectivity Map based on hub genes. The modules were visualized in Cytoscape.Results: Seventeen modules were identified. The modules were robust to the exclusion of half the sample. They were involved in lymphocyte and platelet activation, erythrocyte differentiation, cell cycle, translation, and interferon signaling. In addition, highly connected hub genes were identified in each module, such as GUCY1B1, BCL11B, HMMR, and CEACAM6. High BCL11B (M13) or low CEACAM6 (M27) expression indicates better survival prognosis in sepsis patients regardless of endotype class and age. Network preservation in burns, sepsis, and SIRS showed a general similarity between trauma and burns. M4, M5, M13, M16, M20, and M27 were significantly associated with injury time in trauma and burns. High M13 (T cell activation), low M15 (cell cycle), and low M27 (neutrophil activation) indicate better survival of sepsis patients, regardless of endotype class and age. Moreover, the modules can efficiently separate patients with different diseases. Some modules and hub genes have good prognostic performance in sepsis. Based on the hub genes of each module, six candidate drugs were screened.Conclusion: This study first compared the gene co-expression modules in trauma, burns, sepsis, and SIRS. The identified modules are useful for disease prognosis and drug discovery. BCL11B and CEACAM6 may be promising biomarkers for sepsis risk assessment.Keywords: WGCNA, SIRS, Connectivity Map, hub gene

Keywords

• wgcna
• sirs
• connectivity map
• hub gene