ABSTRACT: Transfer of frozen-thawed embryos leads to sex-specific DNA hypermethylation in both human and mouse placentas. : In humans, frozen ET led to placentas with significant alterations in DNA methylation with most genes demonstrating hypermethylation compared to placentas from fresh ET (4402 CpGs; 1600 genes; p-value < 0.05 in two–tailed unpaired t tests, and a mean methylation difference > 0.05). When compared to control samples, both frozen and fresh samples showed significant differences in methylation.(Frozen vs Control: 5600 CpGs, 2775 genes; Fresh vs Control: 4096 CpGS; 1914 genes; p-value < 0.05, mean methylation difference > 0.05). Paired analysis showed similar trends, despite controlling for maternal environment. Sex specific analysis revealed that these changes were primarily driven by male placentas, as seen by a larger number of CpGs that were differentially methylated, and a larger proportion of genes with at least 2 differentially methylated CpG sites in male placentas as compared to female placentas. (Males: 15572 CpGs, 4399 genes; Females: 7441 CpGs, 3070 genes; p-value < 0.05, mean methylation difference > 0.05). In order to isolate the effects of vitrification we utilized the mouse model, controlling for the effect of the maternal hormonal milieu. In the first genome-wide profiling of DNA methylation in a mouse IVF model, we found that trends in DNA methylation differences paralleled those seen in human placentas. Using similar significance cutoffs as for human samples, and BumpHunter analysis that identifies differentially methylated regions to compensate for a lower number of samples, placentas derived from frozen embryos were predominantly hypermethylated compared to placentas after fresh ET (589 DMRs, 445 genes). Both frozen and fresh embryo transfer samples demonstrated perturbations compared to control samples (Frozen vs Control: 1787 DMRs, 1319 genes; Fresh vs Control: 1119 DMRs, 808 genes). Consistent with the sex-specific trends we observed in human placentas, differences between samples derived from frozen compared to fresh embryo transfer were driven by changes in male placentas (Males: 1069 DMRs, 798 genes; Females:14 DMRs, 11 genes) . When considering the effect of IVF as a whole, murine placentas were predominantly hypomethylated, replicating existing human genome-wide methylation data. As seen previously, IVF led to changes in placental weight, placental morphology and microvessel density in mice, but no persistent changes were seen after embryo vitrification alone. Sexually dimorphic epigenetic changes could indicate differential susceptibility of male and female embryos to IVF-associated perturbations. This observation highlights the importance of sex-specific evaluation of the incidence of adverse outcomes. Similarities between changes seen in mouse and human samples underscore the suitability of the mouse model in evaluating the effect of ART on the epigenetic landscape. This convergence is especially valuable, given limited access to human tissues and the ability to isolate specific interventions in the mouse model.
