152 SPECIES-SPECIFIC DIFFERENCES IN THE METHYLATION REPROGRAMMING DURING EARLY PRE-IMPLANTATION DEVELOPMENT

Abstract

Studies in mouse and human have shown extensive DNA methylation reprogramming in pre-implantation development followed by remethylation from implantation. However, the extent to which such reprogramming is conserved in mammals and the timing of demethylation and remethylation are unknown. As part of a major objective to characterise methylation dynamics in the bovine and porcine species from the oocyte to the blastocyst stage, we aimed here to compare the distribution of methylation at single-base resolution in both species at Day 7.5 of development. The DNA methylation profiles were obtained from individual blastocysts at Day 7.5 [pig: 3 in vivo, 3 in vitro; cow: 3 in vivo, 3 in vitro, 3 inner cell mass (ICM) and 3 trophoectoderm (TE) dissected from in vitro blastocysts] using the post-bisulphite adaptor tagging method and Illumina sequencing. For oocytes, data (GEO: GSE63330) from Schroeder et al. 2015 were analysed. Raw sequences were mapped, methylation calls made using Bismark and data analysis and visualisation was done within the SeqMonk platform. Gene expression profiles from individual blastocysts (3 pig, 3 cow) were obtained by RNA-seq. Annotated mRNA features were quantitated in SeqMonk and these were fed into DESeq2 for differential expression analysis (P < 0.05) as previously reported (Love et al. 2014 Genome Biol. 15, 550). Global methylation levels in whole blastocysts differed substantially between porcine and bovine embryos (in vivo: 12.33 ± 3.6 v. 28.33 ± 3.5%; in vitro: 15.02 ± 3.3 v. 24.41 ± 4.1%). In addition, the distribution of methylation differed: the pattern of cytosine methylated seemed random in the porcine genome, but was highly structured in the bovine genome, with methylation predominantly over gene bodies, resembling the profile previously reported in oocytes (Schroeder et al. 2015 PLoS Genet. 11, e1005442). Regarding correlation analysis, gene expression versus methylation were plotted. It suggested that gene body methylation reflected gene expression pattern in oocytes as well as in bovine blastocysts. Pair-wise comparison of isolated ICM and TE was filtered to require 5% change, and replicate set statistics were applied. This revealed very similar total and regional methylation levels in the 2 compartments, indicating that remethylation does not initiate preferentially in one compartment in bovine pre-implantation embryos. This confirms, from a viewpoint of the genome-wide DNA methylation, what has been observed in mouse for specific genes: the trophoblast-specific DNA methylation occurs after the segregation of the TE and ICM (Nakanishi et al. 2012 Epigenetics 7, 173–183). Our study is the first to provide whole genome methylation profiles from single blastocysts of economically important livestock species. Our data demonstrate that methylation reprogramming in early pre-implantation development is species specific. Knowledge of these specific patterns may have high importance when decisions are taken regarding the use of assisted reproductive technologies, cloning, or generation of transgenic animals.

This work was funded by AGL2015–66341-R (MINECO-FEDER), PRX14/00348 (MECD), 19595/EE/14 (F. Séneca).