Expression patterns of microRNAs in the chorioamniotic membranes: a role for microRNAs in human pregnancy and parturition

Abstract

MicroRNAs (miRNAs) are involved in the post-transcriptional regulation of gene expression during development. This study was performed to determine gestational age-dependent changes in miRNA expression in the chorioamniotic membranes and to assess the significance of miRNAs in human pregnancy and parturition. The expression profile of 455 miRNAs was compared between patients at term without labour (TNL: n = 10), in labour (TL: n = 10), and preterm labour (PTL: n = 10) using microarrays. A total of 39 miRNAs were differentially expressed between term and preterm cases, of which 31 (79.5%) were down-regulated at term. Expression of ten miRNAs, including miR-338, differentially expressed between PTL and TL groups was decreased at term. Computational analyses using miRBase Targets have identified PLA2G4B, a phospholipase implicated in parturition, as a putative target of miR-338. Inhibition of endogenous miR-338 with anti-miR-338 increased the mRNA and protein expression of PLA2G4B in decidual cells. Luciferase assay with reporter constructs confirmed that the suppression of PLA2G4B occurs through binding of miR-338 to the 3UTR of PLA2G4B. Interestingly, the expression of Dicer, a key miRNA-processing enzyme, was markedly decreased at term, particularly with labour in the chorioamniotic membranes. Collectively, the novel findings reported herein strongly suggest that post-transcriptional regulation of genes by miRNAs, coupled with the changes of miRNA processing machinery in the chorioamniotic membranes, plays a role in pregnancy and parturition. Furthermore, the expression level of Dicer in the chorioamniotic membranes dichotomizes pathological preterm labour and physiological spontaneous labour at term.

Figure 1

Comparison of miRNA expression levels in the chorioamniotic membranes between preterm labor and term in labor cases. (A) Principal component analysis shows a segregation of the two groups of samples in the space of the first three principal components computed from normalized log ratios (sample/reference). (B) A heat map showing the log ratios of the 10 differentially expressed miRNAs between TL (term in labor) and PTL (preterm labor) cases, all of which were down-regulated at term (p < 0.05). Grey boxes represent missing values. (C) Confirmation of miR-338 down-regulation at term by real-time qRT-PCR (fold-change = 2.12; p < 0.05). (D) Confirmation of miR-449 down-regulation at term by qRT-PCR. (fold-change = 4.86; p < 0.05). (E) Confirmation of miR-136 down-regulation at term by qRT-PCR. (fold-change = 4.11; p < 0.05). (F) Confirmation of miR-199a* down-regulation at term by qRT-PCR (fold-change = 3.59; p < 0.05). Data are presented as mean ± SEM.

Figure 2

Microarray analysis of miRNA expression in the chorioamniotic membranes between preterm labor (PTL) and term no labor (TNL) cases. (A) An unsupervised analysis based on principal component analysis shows segregation of two a priori known groups of samples. (B) A heat map showing the log ratios of the 38 differentially expressed miRNAs between TNL and PTL cases. The chorioamniotic membranes at term without labor have decreased expression of 28 miRNAs and increased expression of 10 miRNAs. (p < 0.05). Grey boxes represent missing values.

Figure 3

Inhibition of PLA2G4B mRNA by miR-338 in decidual cells. (A) The cells were harvested by collagenase digestion of the decidua from a TNL patient who delivered a male baby. The cells have a normal female karyotype. (B) Transfection of the cells with anti-miR-338 induced a decrease in miR-338 expression (n=4; fold-change = 2.79, p < 0.05). (C) Anti-miR-338 transfection inversely increased PLA2G4B mRNA expression (n=5; fold-change = 3.72; p < 0.05). Data are presented as mean ± SEM and relative to the expression of a scrambled-sequence negative control. (D) PLA2G4B protein expression was also increased following anti-miR-338 transfection.

Figure 4

Suppression of PLA2G4B in decidual cells occurs through binding of miR-338 to PLA2G4B 3’UTR. (A) Target scanning of the PLA2G4B 3’UTR identified two putative miR-338 binding sites. The full-length PLA2G4B 3’UTR sequence was cloned in the downstream of the firefly luciferase gene in the reporter plasmid pMIR-REPORT™. (B) Transfection with the reporter plasmid containing the PLA2G4B 3’UTR decreased luciferase activity by 46% compared to a mock reporter plasmid. Renilla luciferase plasmid was used as a transfection control (n=3; p < 0.05). (C) Co-transfection of anti-miR-338 and a reporter plasmid with the PLA2G4B 3’UTR caused a 1.82 fold-increase in luciferase activity compared to cells co-transfected with anti-miR-338 and a mock reporter plasmid (n=3; p < 0.05).

Figure 5

Dicer expression in term and preterm chorioamniotic membranes. (A) Immunofluorescent staining showing ubiquitous expression in the cells comprising the chorioamniotic membranes. (B, C) The expression is also ubiquitous in the chorionic villi and umbilical cord, with more distinct immunoreactivity in the vascular endothelial cells and smooth muscle cells in both structures. (D) In the amnion, immunoblotting demonstrates decreased Dicer expression in TNL and TL cases compared to PTL cases. In the chorion-decidua, Dicer expression was substantially reduced in TL cases compared to PTL and TNL cases. The expression pattern of Dicer in the placenta is strikingly similar to those in the amnion and chorion. Of special note is that the molecular weights of Dicer in the chorioamniotic membranes and the placenta are different. The chorioamniotic membranes predominantly expressed a 219 kD (1,922 aa) full-length protein, while the placenta largely expressed 113 kD (997 aa) and 93 kD (820 aa) variants.