Prognostic implication of downregulated exosomal miRNAs in patients with sepsis: a cross-sectional study with bioinformatics analysis

Background: Despite the understanding of sepsis-induced extracellular vesicles (EVs), such as exosomes, and their role in intercellular communication during sepsis, little is known about EV contents such as microRNA (miRNA), which modulate important cellular processes contributing to sepsis in body fluids. This study aimed to analyze the differential expression of exosomal miRNAs in plasma samples collected from sepsis patients and healthy controls, and to identify potential miRNA regulatory pathways contributing to sepsis pathogenesis. Methods: Quantitative real-time PCR-based microarrays were used to profile plasma exosomal miRNA expression levels in 135 patients with sepsis and 11 healthy controls from an ongoing prospective registry of critically ill adult patients admitted to the intensive care unit. The identified exosomal miRNAs were tested in an external validation cohort (35 sepsis patients and 10 healthy controls). And then, functional enrichment analyses of gene ontology, KEGG pathway analysis, and protein–protein interaction network and cluster analyses were performed based on the potential target genes of the grouped miRNAs. Finally, to evaluate the performance of the identified exosomal miRNAs in predicting in-hospital and 90-day mortalities of sepsis patients, receiver operating characteristic curve (ROC) and Kaplan–Meier analyses were performed. Results: Compared with healthy controls, plasma exosomes from sepsis patients showed significant changes in 25 miRNAs; eight miRNAs were upregulated and 17 downregulated. Additionally, the levels of hsa-let-7f-5p, miR-331-3p miR-301a-3p, and miR-335-5p were significantly lower in sepsis patients than in healthy controls (p < 0.0001). These four miRNAs were confirmed in an external validation cohort. In addition, the most common pathway for these four miRNAs were PI3K-Akt and mitogen-activated protein kinase (MAPK) signaling pathways based on the KEGG analysis. The area under the ROC of hsa-let-7f-5p, miR-331-3p, miR-301a-3p, and miR-335-5p level for in-hospital mortality was 0.913, 0.931, 0.929, and 0.957, respectively (p < 0.001), as confirmed in an external validation cohort. Also, the Kaplan–Meier analysis showed a significant difference in 90-day mortality between sepsis patients with high and low miR-335-5p, miR-301a-3p, hsa-let-7f-5p, and miR-331-3p levels (p < 0.001, log-rank test). Conclusion: Among the differentially-expressed miRNAs detected in microarrays, the top four downregulated exosomal miRNAs (hsa-let-7f-5p, miR-331-3p miR-301a-3p, and miR-335-5p) were identified as independent prognostic factors for in-hospital and 90-day mortalities among sepsis patients. Bioinformatics analysis demonstrated that these four microRNAs might provide a significant contribution to sepsis pathogenesis through PI3K-Akt and MAPK signaling pathway.