Decoding the Deadly Dance: NETosis Genes Predict Neonatal Sepsis Fate

Abstract

Background: Neonatal sepsis, a systemic inflammatory response to infection, is a major cause of morbidity and mortality in newborns. Neutrophil extracellular trap formation (NETosis), while crucial for pathogen clearance, can contribute to organ dysfunction in sepsis. This study aimed to identify key NETosis-related genes for prognostication in neonatal sepsis. Methods: We analysed whole blood transcriptome datasets (GSE26440, GSE26378, GSE25504) from neonates with sepsis and controls. Differentially expressed NETosis genes (DE-NET genes) were identified, and a machine learning approach was used to select the most influential genes. A NET score model was constructed and validated using single-sample gene set enrichment analysis (ssGSEA). The model's performance was evaluated using ROC analysis. The interplay between key-NET genes and the complement-coagulation (CC) system was investigated. Clinical samples were also collected for validation . Results: Sixteen DE-NET genes were identified, and LASSO further refined these to 8 key-NET genes. The key-NET gene signature and NET score model showed excellent predictive performance (AUCs > 89%) in distinguishing survivors from non-survivors. Mediation analysis revealed that key-NET gene expression precedes and potentially drives complement-coagulation activation. Conclusions: We present an 8-gene prognostic model for risk stratification in neonatal sepsis, based on early blood transcript signatures in neonates. Our findings underscore the central role of NETosis in sepsis-induced coagulopathy, revealing potential therapeutic targets for intervention.