Abstract
Background Mitochondrial dysfunction has been proven to be a characteristic change in the myocardium of patients with sepsis. In our previous research, we revealed that some mitochondrial dysfunctions occur synchronously in the peripheral blood of sepsis patients and affect mortality with inflammatory and other related genes. However, these mitochondrial dysfunctions are not described in detail. Whether mitochondrial dysfunction affects the mortality of sepsis patients as an independent risk factor still needs to be further validated.
Objects and methods In our study, we aimed to present the co-varied genes and pathways related to mitochondrial and aerobic respiratory function in myocardium and peripheral blood of sepsis patients, and to verify their effects regarding the mortality of sepsis. We applied weighted gene co-expression network analysis(WGCNA)to generate different modules from myocardium and blood datasets, and subsequent enrichment analysis was used to identify the mitochondrial-and aerobic respiratory-related modules. We obtained the co-varied differential expressed genes(DEGs)from the modules to separate sepsis patients into different subgroups and compare the survival rate between them. Machine learning algorithms were applied for mortality predictive model construction and validation.
Results Blue and magenta modules in blood and blue modules in the myocardium were identified as being related to mitochondrial and aerobic respiratory function. There was a strong overlap in gene expression and pathways between these modules, and DEGs from them separated sepsis patients into two groups, but there was no statistical difference in mortality between the different groups(p-value=0.078). However, models generated from these DEGs performed well in mortality prediction.
Conclusion Our research has found that some genes and pathways associated with mitochondrial aerobic respiratory dysfunction are generally altered in myocardium and peripheral blood, and the changes of these related genes can reflect the severity and mortality of sepsis. Therefore, we can expect the application prospect of these mitochondria-related genes as biomarkers of infectious cardiomyopathy.