117 Transcription readthrough in in vivo-developed bovine oocytes and preimplantation embryos
S. Ranjitkar A , M. Siri B , J. Sun B , G. Liu C and X. Tian AA University of Connecticut, Storrs, CT, USA
B Old Dominion University, Norfolk, VA, USA
C Animal Genomics and Improvement Laboratory, Beltsville, MD, USA
Reproduction, Fertility and Development 35(2) 185-186 https://doi.org/10.1071/RDv35n2Ab117
Published: 5 December 2022
© 2023 The Author(s) (or their employer(s)). Published by CSIRO Publishing on behalf of the IETS
As much as 87% of the genome is transcribed. Readthrough transcriptions, either failure to terminate beyond the transcription end site (TES), or aberrant transcription at other intergenic regions, are present in normal physiological conditions but are enhanced in the presence of stressors such as abnormal osmolarity and viral infection. Readthrough transcription is hypothesised to increase chromatin accessibility. More than 10,000 genes are transcribed during preimplantation bovine embryo development. These are likely regulated through the drastic changes in DNA methylation. Failure of transcription termination and intergenic transcription have not been characterised in early embryos or correlated with DNA methylation. We aimed to determine (1) whether embryogenesis possesses widespread readthrough transcription, and (2) if so, are they correlated with DNA methylation levels of intergenically transcribed regions? Automatic Readthrough Transcription Detection (ARTDeco; Roth et al., 2020; BMC Bioinformatics) detects downstream of gene (DoG) transcripts (4–15 kb in length after TES), readthrough transcripts (intergenic transcription starting10 kb downstream from TES;100 bp–15 kb in length), and read-in transcripts (intergenic transcription starting1 kb upstream and passing through the transcription start site [TSS]; 100 bp–15 kb in length). Using RNA-seq data of in vivo-developed oocytes, embryos at 2–16 cell, morula, and blastocyst stages (Jiang et al., 2014; BMC Genomics), we found abundant readthrough and read-in transcripts in the range of 25.5–66.7% of expressed genes (defined as transcript per million > 1) at different stages of embryo development. DoG transcripts were less copious, at an average of 10%. However, Pearson correlation revealed that DoG transcription across different stages of development was significantly correlated with gene expression (P < 0.05). Interestingly, the levels of expression for these intergenic transcripts were mostly higher than those of expressed genes (readthroughs: up to 3-fold, read-ins: up to 1-fold, DoGs: up to 7-fold). Furthermore, many transcripts of readthrough (1,504), read-in (1,045), and DoG (1,021) were common across all stages of preimplantation development, yet were differentially expressed (log fold change > 2; P < 0.05). We found gradual but unpatterned decreases in DNA methylation densities 10 kb up-to downstream of the intergenic transcribed regions. Similarly, no significant Pearson correlation (P < 0.05) was found between levels of intergenic transcription and DNA methylation 1 kb upstream of TSS (read-ins) or 1 kb downstream of TES (readthroughs and DoGs). In conclusion, we showed widespread intergenic transcription in forms of readthroughs and read-ins in bovine preimplantation embryos. They were not, however, correlated with DNA methylation in the investigated regions.