Prediction of Early- and Late-Onset Pre-Eclampsia in the Preclinical Stage via Placenta-Specific Extracellular miRNA Profiling

Abstract

Pre-eclampsia (PE) is one of the severe complications of pregnancy in 3–8% of all cases and is one of the leading causes of maternal and perinatal mortality. The fundamental role in the pathogenesis of PE is assigned to maternal and/or placental factors, whereby the combination and manifestation of which determines the time of onset of the clinical symptoms of PE (before or after 34 weeks of gestation) and their severity. It is known that the expression level of miRNAs, the regulators of signaling cascades in the cell, depends on gestational age. In the present study, we focused on the identification of the placenta-specific miRNAs that differentiate between early- and late-onset pre-eclampsia (ePE and lPE) throughout pregnancy, from the first to the third trimester. A total of 67 patients were analyzed using small RNA deep sequencing and real-time quantitative PCR, which resulted in a core list of miRNAs (let-7b-5p, let-7d-3p, let-7f-5p, let-7i-5p, miR-22-5p, miR-451a, miR-1246, miR-30e-5p, miR-20a-5p, miR-1307-3p, and miR-320e), which in certain combinations can predict ePE or lPE with 100% sensitivity and 84–100% specificity in the 1st trimester of pregnancy. According to the literature data, these miRNA predictors of PE control trophoblast proliferation, invasion, migration, syncytialization, the endoplasmic reticulum unfolded protein response, immune tolerance, angiogenesis, and vascular integrity. The simultaneous detection of let-7d-3p, miR-451a, and miR-1307-3p, resistant to the repeated freezing/thawing of blood serum samples, in combination with biochemical (b-hCG and PAPP-A) and ultrasound (UAPI) parameters, allowed us to develop a universal model for the prediction of ePE and lPE onset (FPR = 15.7% and FNR = 9.5%), which was validated using a test cohort of 48 patients and demonstrated false-positive results in 26.7% of cases and false negatives in 5.6% of cases. For comparison, the use of the generally accepted Astraia program in the analysis of the test cohort of patients led to worse results: FPR = 62.1% and FNR = 33.3%.