Identification of Potential Transcriptional Biomarkers Differently Expressed in Both - and -Induced Sepsis via Integrated Analysis.

Sepsis is a critical, complex medical condition, and the major causative pathogens of sepsis are both () and (). Genome-wide studies identify differentially expressed genes for sepsis. However, the results for the identification of DEGs are inconsistent or discrepant among different studies because of heterogeneity of specimen sources, various data processing methods, or different backgrounds of the samples. To identify potential transcriptional biomarkers that are differently expressed in - and -induced sepsis, we have analyzed four microarray datasets from GEO database and integrated results with bioinformatics tools. 42 and 54 DEGs were identified in both and samples from any three different arrays, respectively. Hierarchical clustering revealed dramatic differences between control and sepsis samples. GO functional annotations suggested that DEGs in the group were mainly involved in the responses of both defense and immune regulation, but DEGs in the group were mainly related to the regulation of endopeptidase activity involved in the apoptotic signaling pathway. Although KEGG showed inflammatory bowel disease in the E. coli group, the KEGG pathway analysis showed that these DEGs were mainly involved in the tumor necrosis factor signaling pathway, fructose metabolism, and mannose metabolism in both - and -induced sepsis. Eight common genes were identified between sepsis patients with either or infection and controls in this study. All the candidate genes were further validated to be differentially expressed by an ex-vivo human blood model, and the relative expression of these genes was performed by qPCR. The qPCR results suggest that GK and PFKFB3 might contribute to the progression of -induced sepsis, and CEACAM1, TNFAIP6, PSTPIP2, SOCS3, and IL18RAP might be closely linked with -induced sepsis. These results provide new viewpoints for the pathogenesis of both sepsis and pathogen identification.