247 MESSENGER RNA EXPRESSION PATTERNS OF DNA AND HISTONE METHYLTRANSFERASES IN PREIMPLANTATION DEVELOPMENT OF IN VIVO- AND IN VITRO-PRODUCED BOVINE EMBRYOS

Abstract

The effects of in vitro production (IVP) and/or somatic nuclear transfer on mRNA expression patterns have mostly been determined in morulae and blastocysts, i.e. after embryonic genome activation. Comparative data regarding mRNA expression patterns throughout the oviductal phase of pre-implantation development are scarce. Here we studied mRNA expression for genes related to DNA methylation and modification of histones which account for the major epigenetic reprogramming during development. Pertubated epigenetic reprogramming of the genome is a likely cause of developmental abnormalities and epigenetic diseases associated with assisted reproduction technologies. The objective of the present study was to compare mRNA expression of DNA methyltransferases Dnmt1, -3a, and -3b and histone methyltransferases SUV39-h1 and G9a between in vivo-derived bovine embryos and their IVP counterparts using a semiquantitative RT-PCR assay (Wrenzycki et al. 2002 Biol. Reprod. 66, 127-134) employing two embryos for each assay. In vivo-derived embryos were collected from 28 superovulated heifers by endoscopic flushing of oviducts (zygotes to 8-cell stages) (Besenfelder et al. 2001 Theriogenology 55, 837-845) or by uterine flushing (16-cell stages to blastocysts). Endoscopic flushing at different time points after AI (Days 1, 1.5, 2, 3, 4, and 4.5) yielded 31 zygotes; 15 two-cell, 5 three-cell, 13 four-cell, 1 five-cell, 2 six-cell, and 11 eight-cell embryos; 4 degenerated embryos; and 18 unfertilized ova. The recovery rate (corpora lutea counted per recovered embryos) was 58% and 62% for the endoscopic and uterine flushing, respectively. Differences in the relative abundance of each gene transcript between groups were tested using ANOVA with the main effects being origin (in vivo/in vitro) and developmental stage (zygote to blastocyst) and their interactions followed by multiple pairwise comparisons using a Tukey test (P < 0.05). Origin of embryos affected the relative abundance of transcripts for Dnmt1, Dnmt3a, and SUV39-h1, and developmental stage affected the relative abundance of transcripts for Dnmt1, -3a, -3b, SUV39-h1, and G9a. No interactive effects were observed for origin and developmental stage in the relative abundance of all transcripts. The relative abundance of Dnmt1 transcripts differed significantly between in vivo- and in vitro-produced morulae and blastocysts. For Dnmt3a, mRNA differences were determined between in vivo- and in vitro-produced 10-16-cell stages and morulae. Suv39-h1 transcripts differed significantly between in vivo- and in vitro-derived zygotes, 2-cell embryos, 8-cell embryos, 10-16-cell embryos, and blastocysts. The results suggest that IVP alters mRNA expression of genes related to epigenetic modifications very early in development, even before the embryonic genome has been activated.