Placental microRNA expression associates with birthweight through control of adipokines: results from two independent cohorts

Abstract

MicroRNAs are non-coding RNAs that regulate gene expression post-transcriptionally. In the placenta, the master regulator of foetal growth and development, microRNAs shape the basic processes of trophoblast biology and specific microRNA have been associated with foetal growth. To comprehensively assess the role of microRNAs in placental function and foetal development, we have performed small RNA sequencing to profile placental microRNAs from two independent mother-infant cohorts: the Rhode Island Child Health Study (n = 225) and the New Hampshire Birth Cohort Study (n = 317). We modelled microRNA counts on infant birthweight percentile (BWP) in each cohort, while accounting for race, sex, parity, and technical factors, using negative binomial generalized linear models. We identified microRNAs that were differentially expressed (DEmiRs) with BWP at false discovery rate (FDR) less than 0.05 in both cohorts. hsa-miR-532-5p (miR-532) was positively associated with BWP in both cohorts. By integrating parallel whole transcriptome and small RNA sequencing in the RICHS cohort, we identified putative targets of miR-532. These targets are enriched for pathways involved in adipogenesis, adipocytokine signalling, energy metabolism, and hypoxia response, and included Leptin, which we further demonstrated to have a decreasing expression with increasing BWP, particularly in male infants. Overall, we have shown a robust and reproducible association of miR-532 with BWP, which could influence BWP through regulation of adipocytokines Leptin and Adiponectin.

Keywords: adipokines; adiponectin; birthweight; hsa-miR-532-5p; leptin; microRNA; placenta; placental microRNA.

Figure 1.

microRNA association with birthweight percentile. Volcano plots illustrate the results of the differential expression analyses for the New Hampshire Birth Cohort Study (round points; (a) and the Rhode Island child health study (square points; b), respectively. On the y and x axes, -log₁₀(p-values) in the association of each microRNA with BWP and effect estimates, or the log₂ fold change in each microRNA per one percent change in BWP, are shown respectively. The horizontal-dashed lines represent the Bonferroni threshold (log₁₀[0.05/M], where M is the number of microRNAs tested). (a) 7 microRNAs are significantly (FDR < 0.05) associated with BWP in NHBCS (grey filled shapes), with two microRNAs significant after Bonferroni correction. (b) 11 microRNAs are significantly (FDR < 0.05) associated with BWP in RICHS (grey filled shapes), with three microRNAs significant after Bonferroni correction. (c) Wald statistics for the association of each microRNA with BWP from the cohort-level differential expression analyses were plotted and labelled to indicate if they were: consistent in direction of effect (empty triangle), significant in RICHS (square), or significant in NHBCS (circle)

Figure 2.

miR-532 associates with birthweight group. Estimates of log₂ fold change for SGA and LGA, both relative to AGA, illustrate a significant reduction of miR-532 in placentae from SGA newborns and an apparent increase in miR-532 in LGA newborns for NHBCS and RICHS. Birthweight groups, small (SGA) and large (LGA) for gestational age are represented by dashed and solid lines, respectively

Figure 3.

miR-532 target transcript abundance association with LGA or SGA, relative to AGA. Log₂ fold change of miR-532 putative target transcript abundance in placentae from SGA and LGA newborns, relative to AGA newborns

Figure 4.

Pathway analysis of miR-532 putative targets. Select consensus path DB results are presented in a bubble plot. Dot colour represents -log₁₀(FDR). Dot size indicates the percentage of pathway members that are represented by miR-532 targets. Pathway sources are indicated on the right y-axis

Figure 5.

Sex-specific association of birthweight percentile and group with placental leptin transcript abundance in RICHS. (a) log₂ fold change of LEP abundance in placentae from SGA and LGA newborns, relative to AGA newborns. (b) Estimates of log₂ fold change per one BWP. Colour represents sample stratification: black (all placentae), red (female placentae) and blue (male placentae)