Introduction: The heterogeneity of sepsis presents significant challenges for early diagnosis and effective treatment. In this study, a proteomic examination of serum extracellular vesicles (EVs) was applied to a large cohort of patients with varying sepsis severity, aiming to identify potential biomarkers and uncover molecular subtypes of the condition. Methods: Peripheral blood samples from patients with sepsis (Sepsis-only, n=100), sepsis complicated with ARDS (S-ARDS group, n=50), sepsis complicated with AKI (S-AKI group, n=50), sepsis complicated with both ARDS and AKI (S-ARDS&AKI group, n=50), and healthy controls (n=50) were collected. The successful extraction of EVs was confirmed by transmission electron microscopy, nanoparticle tracking analysis, and western blot assays. Proteins were identified by mass spectrometry. Integrated bioinformatics analyses were performed to explore the vital biomarkers and molecular subtypes of sepsis. Results: EVs were successfully extracted in different clinical groups. A total of 4,754 proteins were detected in 279 samples after quality control. The results revealed that the expression level of CD81, FLOT1, FLOT2, HSP90AA1, and HSP90AB1 were gradually up-regulated during the progression of sepsis and may act as the promising biomarker for predicting the onset and severity of sepsis. Consensus clustering divided Sepsis-only group into four molecular subtypes with distinct clinical severity. Finally, the levels of metabolism-related proteins were associated with the severity and prognosis of sepsis. Conclusions: This study utilized EVs proteomics on large-scale cohorts to establish molecular subtypes of sepsis, uncovering the inter-individual heterogeneity of the condition. Additionally, metabolic dysregulation was found to be closely linked to disease severity and patient survival. Targeting metabolic disorders may, therefore, represent a promising therapeutic strategy for sepsis.
Bi et al. (Sun,) studied this question.