A Potential Autophagy-Related-Gene Based Signature in Patients with Preeclampsia.

OBJECTIVE

Preeclampsia (PE) is a significant cause of maternal and offspring mortality and morbidity. The purpose of this study is to identify the potential diagnostic signatures of autophagy-related genes (ATGs) in pregnancies with preeclampsia.

METHODS

The expression profile of mRNA was obtained from GSE75010 (placenta samples) and GSE48424 dataset (blood samples). The potential differentially expressed ATGs of PE were screened by R software. The gene-ontology (GO) enrichment analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, correlation analysis, and protein-protein interactions (PPI) were applied for the differentially expressed ATGs. The diagnostic markers of PE were then screened based on least absolute shrinkage and selection operator (LASSO) logistic regression and support vector machine-recursive feature elimination (SVM-RFE). Receiver operating characteristic (ROC) analysis was used to investigate the predictive value of these diagnostic markers. Target miRNAs were predicted based on the miRDB, DIANA-micro T, Targetscan, and miRWalk databases, and were further validated in GSE84260.

RESULTS

A total of 20 differentially expressed ATGs were identified between PE and healthy pregnancies. Functional analysis of differentially expressed ATGs indicated several enriched terms related to autophagy, apoptosis, angiogenesis, inflammation, immune response, hypoxia-inducible factor 1 (HIF-1), forkhead box O (FoxO) and AMP-activated protein kinase (AMPK) signaling pathway. A total of 12 ATGs were recognized based on LASSO and SVM-RFE, which made an excellent distinction in both the placenta tissues (area under the curve [AUC] = 0.903) and the blood samples (AUC = 0.972). Furthermore, four feature ATGs (leptin [], ERO1-like [], phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit beta [], and mitogen-activated protein kinase 8 []) were screened and also shown an excellent diagnostic efficacy (AUC = 0.869 in placenta samples, and AUC = 0.914 in blood samples). Additionally, 81 target miRNAs were predicted according to the 4 feature ATGs. After evaluating the miRNA expression pattern of GSE84260, 11 miRNAs were selected. Finally, a miRNA-mRNA regulatory network was constructed, which may participate in the development of PE.

CONCLUSIONS

We established an autophagy-related-gene based signature that may predict pregnancies with PE. And we also constructed a miRNA-mRNA regulatory network, which may deepen our understanding of the molecular mechanism underlying the development of PE.