The role and interaction of imprinted genes in human fetal growth

Abstract

Identifying the genetic input for fetal growth will help to understand common, serious complications of pregnancy such as fetal growth restriction. Genomic imprinting is an epigenetic process that silences one parental allele, resulting in monoallelic expression. Imprinted genes are important in mammalian fetal growth and development. Evidence has emerged showing that genes that are paternally expressed promote fetal growth, whereas maternally expressed genes suppress growth. We have assessed whether the expression levels of key imprinted genes correlate with fetal growth parameters during pregnancy, either early in gestation, using chorionic villus samples (CVS), or in term placenta. We have found that the expression of paternally expressing insulin-like growth factor 2 (IGF2), its receptor IGF2R, and the IGF2/IGF1R ratio in CVS tissues significantly correlate with crown-rump length and birthweight, whereas term placenta expression shows no correlation. For the maternally expressing pleckstrin homology-like domain family A, member 2 (PHLDA2), there is no correlation early in pregnancy in CVS but a highly significant negative relationship in term placenta. Analysis of the control of imprinted expression of PHLDA2 gave rise to a maternally and compounded grand-maternally controlled genetic effect with a birthweight increase of 93/155 g, respectively, when one copy of the PHLDA2 promoter variant is inherited. Expression of the growth factor receptor-bound protein 10 (GRB10) in term placenta is significantly negatively correlated with head circumference. Analysis of the paternally expressing delta-like 1 homologue (DLK1) shows that the paternal transmission of type 1 diabetes protective G allele of rs941576 single nucleotide polymorphism (SNP) results in significantly reduced birth weight (-132 g). In conclusion, we have found that the expression of key imprinted genes show a strong correlation with fetal growth and that for both genetic and genomics data analyses, it is important not to overlook parent-of-origin effects.

Keywords: birth weight; chorionic villus sampling; fetal growth restriction; genomic imprinting; placenta; type 1 diabetes.

Figure 1.

Correlation between imprinted gene expression in CVS and CRL. Expression levels of each gene relative to the L19 endogenous control gene were correlated to crown–rump length (CRL: mm) using a multiple linear regression model adjusted for maternal BMI, baby's sex, parity, gestational age when CRL was measured and maternal smoking habit. Positive correlations with CRL and (a) IGF2 expression (r = 0.77; p = 0.004), (b) IGF2R expression (r = 0.76; p = 0.03), (c) IGF2/IGF1R ratio (r = 0.74; p = 0.03) and (d) H19 expression (r = 0.74; p = 0.04) were observed. (Online version in colour.)

Figure 2.

Negative correlation between GRB10 term placental expression and head circumference. The expression level of GRB10 relative to the L19 housekeeping gene was correlated to head circumference (cm) using a multiple linear regression model adjusted for baby's sex, parity, gestational age at birth, maternal weight and smoking habits. GRB10 expression values in logarithmic scale was used. Significant negative association was observed for GRB10 term placenta expression and head circumference (r =−0.35; p = 0.04). (Online version in colour.)

Figure 3.

The association between paternal A/G SNP rs941576 at the DLK1 locus and fetal growth. Partial residual plots illustrating the correlation between paternal inheritance of the A or G allele and (a) birthweight (g), (b) head circumference (cm) and (c) placental weight (g), corrected for baby's sex, parity, gestational age, maternal weight and smoking habit in the multiple regression model. In comparison to the A allele, paternal G allele inheritance is associated with significantly reduced birthweight (p = 0.01, 95% CI−232 to −32) and head circumference (p = 0.01, 95% CI −0.85 to −0.11) but not with placental weight (p = 0.98, 95% CI −35 to 35). Paternal A/G, paternal transmission of A/G SNP rs941576; A, paternal transmission of the A allele; G, paternal transmission of the G allele; BW, birthweight; HC, head circumference; PW, placental weight. (Online version in colour.)

Figure 4.

Current hypothesis on the association between paternal G SNP rs941576 and fetal growth. Solid lines indicate results from this study and dotted lines indicate published data [41–43]. Paternal inheritance of the G allele is associated with an average reduction in birthweight by 132 g (p = 0.01). The paternal G allele is also correlated with reduction in DLK1 expression although not significantly (p = 0.47). There was a trend of positive association between DLK1 expression in term placenta and birthweight (p = 0.07). Our hypothesis suggests that the paternal G allele reduces DLK1 expression which causes reduction in birthweight and risk of type 1 diabetes. (Online version in colour.)