Breaking Rules: The complex relationship between DNA methylation and X-chromosome inactivation in the human placenta

Background: The human placenta is distinct from most organs due to its uniquely low-methylated genome. Previous research has shown that DNA methylation (DNAme) is particularly depleted in the placenta at partially methylated domains and on the inactive X chromosome (Xi) in XX samples. While Xi DNAme is known to be critical for X-chromosome inactivation (XCI) in other tissues, its role in the placenta remains unclear. Understanding X-linked DNAme variation in the placenta may provide insights into XCI and implications for prenatal development and phenotypic sex differences. Methods: DNAme data were analyzed from over 350 human placental (chorionic villus) samples, along with samples from cord blood, amnion, chorion, and fetal somatic tissues. We characterized X chromosome DNAme variation in the placenta relative to sample variables including cell composition, maternal age, placental weight, and fetal birth weight, and compared these patterns to other tissues. We also evaluated the relationship between X-linked DNAme and previously reported XCI gene expression status in placenta. Results: Our findings confirm that the placenta exhibits significant depletion of DNAme on the Xi compared to other tissues. Additionally, we observe that X chromosome DNAme profiles in the placenta are influenced by cell composition, particularly trophoblast proportions, with minimal variation across gestation. Notably, low promoter DNAme is observed at most genes on the Xi, regardless of XCI status, challenging known associations in somatic tissues between low promoter DNAme and XCI escape. Conclusions: This study provides evidence that the human placenta has a distinct Xi DNAme landscape, which may inform our understanding of sex differences during prenatal development. Future research should explore the mechanisms underlying the placenta's unique X-linked DNAme profile and the factors involved in XCI maintenance. DNAme data for 28 chorionic villous (placental) samples, taken from the fetal-facing side of the placenta and combined before running on either of the Illumina Infinium HumanMethylation450 or MethylationEPIC arrays in Vancouver, Canada. DNAme data from raw IDAT files were read into R, and betas were extracted using the preprocessRaw function from minfi. Methylated intensity, unmethylated intensity, and detection p values were calculated for all probes, separately in the samples run on the 450K array (n=3) and the EPIC array (n=28). The raw data (meth/unmeth/detp) is available, as ae the full beta value matrices after using preprocessRaw for both 450K and EPIC samples. *************************************************************** IDAT files for human/patient samples were not submitted to GEO due to concerns about submitting personally identifiable SNP genotyping data for open access. ***************************************************************