Background: The placenta-specific miRNA profile is subject to changes due to developmental or pathological processes. As this profile is partially reflected in the maternal circulation, placental miRNAs are attractive as potential markers for pregnancy complications. However, the quantification of miRNAs in plasma and placenta is subject to a number of specific challenges. The lack of consistency and reproducibility between the studies raises concerns that the results are extensively influenced by experimental conditions.

Goal: This study aimed at a systematic evaluation of the factors that may interfere with the accurate assessment of the expression of circulating miRNAs. miRNAs were extracted from 35 plasma samples obtained from pregnant women at different gestational ages and 17 term placentas. 11 miRNAs were studied using quantitative stem-loop RT-PCR, including five endogenous control candidates (let-7b, miR-16, miR-23a, miR-92, RNU48), one ‘spike-in’ miRNA control (ath-miR-159a), and four candidate biomarkers (mir-210, miR-411, miR-518b, miR-720). We systematically evaluated key technical and biological confounders, including variation in technical replicates for plasma, impact of pre-processing delay time up to 48 hours, as well as miRNA expression across placentae (total 9 samples at different locations/depths per each placenta), and expression variation of candidate endogenous controls in complicated and non-complicated pregnancies.

Results: All candidate controls but RNU48 were represented in plasma. Expression variation between extraction replicates in plasma was low for all miRNAs studied (SD min=0.19, max=0.27 cycles), demonstrating uniform performance of extraction protocols and assays. miRNAs revealed comparable levels of expression across placentae and high, but not uniform stability upon time, with a significant decrease in levels of miR-720, p=0.006, in placenta and increasing levels in plasma for miR-23a, ^p<0.0001, probably due to hemolysis. Candidate controls did not show differences in expression in plasma or placentae between normal and complicated pregnancies, and were prioritized according to their stability and representation. The top candidates will be used to access the expression of the candidate biomarker miRNAs.

Conclusion: Strict experimental handling and the use of exogenous and endogenous controls for normalization are critical for the reliable quantification of miRNA expression levels.