Construction of a NETosis-related gene signature for predicting the prognostic status of sepsis patients.

BACKGROUND

Sepsis is a common traumatic complication of response disorder of the body to infection. Some studies have found that NETosis may be associated with the progression of sepsis.

METHODS

Data of the sepsis samples were acquired from Gene Expression Omnibus (GEO) database. Gene set enrichment score was calculated using single-sample gene set enrichment analysis (ssGSEA). Weighted gene co-expression network analysis (WGCNA), protein-protein interaction (PPI) networks analysis, and stepwise multivariable regression analysis were performed to identify NETosis-associated genes for sepsis prognosis. To assess the infiltration of immune cells, the ESTIMATE and CIBERPSORT algorithms were used. Functional enrichment analysis was conducted in the clusterProfiler package.

RESULTS

Different programmed death pathways were abnormally activated in sepsis patients as compared to normal samples. We screened five important NETosis associated genes, namely, CEACAM8, PGLYRP1, MAPK14, S100A12, and LCN2. These genes were significantly positively correlated with entotic cell death and ferroptosis and negatively correlated with autophagy. A clinical prognostic model based on riskscore was established using the five genes. The ROC curves of the model at 7 days, 14 days and 21 days all had high AUC values, indicating a strong stability of the model. Patients with high riskscore had lower survival rate than those with low riskscore. After the development of a nomogram, calibration curve and decision curve evaluation also showed a strong prediction performance and reliability of the model. As for clinicopathological features, older patients and female patients had a relatively high riskscore. The riskscore was significantly positively correlated with cell cycle-related pathways and significantly negatively correlated with inflammatory pathways.

CONCLUSION

We screened five NETosis-associated genes that affected sepsis prognosis, and then established a riskscore model that can accurately evaluate the prognosis and survival for sepsis patients. Our research may be helpful for the diagnosis and clinical treatment of sepsis.