100 ASSESSMENT OF LOCUS-SPECIFIC DNA METHYLATION: OPTIMIZATION FOR BOVINE EMBRYOS

Abstract

Assessment of chromatin remodeling in early embryos is a major focus of studies today, and evaluation of DNA methylation at specific loci is one approach to study these epigenetic modifications. Our objective was to optimize the bisulfite sequencing methodology for use with very small cell numbers originating from pre-implantation embryos, making the process more time- and cost-efficient. The optimized steps include bisulfite conversion of small samples, bisulfite primer design, high-throughput plasmid DNA amplification, and preparation for sequencing. Methylation at 2 loci, Satellite I and Oct4, was investigated in bovine in vitro-produced (IVP) embryos collected at the 2-cell, 8-cell, and blastocyst stages. Bovine skin fibroblasts were first used to optimize the particular steps of the process. All reactions were run in duplicate and no-template negative and somatic cell-positive controls were treated alongside samples. Incorporating the use of Methyl Primer Express (Applied Biosystems, Foster City, CA), MacVector (Oxford Molecular Ltd., Campbell, CA), and Mfold software (Mathews DH et al. 1999 J. Mol. Biol. 288, 911–940; Zuker M 2003 Nucleic Acids Res. 31, 3406–3415) improved the specificity of bisulfite primers by exclusion of secondary or tertiary structures. The DNA from bisulfite treatment for 15 to 16 h was of better quality than DNA treated for 18 h. After initial PCR optimization, different cell concentrations were used to establish that detectable PCR products and subsequent methylation data could be obtained from DNA isolated from as few as 8 cells. Treating single blastocysts and pools of ten 8-cell and forty 2-cell embryos was sufficient for the entire scope of the experiment, allowing use of the same samples across all loci. After molecular cloning, plasmid DNA was amplified by 3 different methods and evaluated for efficiency: miniprep, TempliPhi (GE Healthcare, Piscataway, NJ), or 96-well glycerol stocks and automated TempliPhi format. Although TempliPhi alone was better than miniprep for small-scale experiments, it was the 96-well format that saved weeks of time and was most cost-effective. Sequencing was performed on a minimum of 8 clones/sample using ABI Prism sequencers (Applied Biosystems), and results were analyzed using Chromas Pro software (Technelysium Pty. Ltd., Helensvale, Australia). Percentage methylation of bovine IVP 2-cell, 8-cell, and blastocyst stage embryos for Satellite I was 25, 10, and 22%, respectively, and for Oct4 was 88, 88, and 79%, respectively. However, somatic cell methylation was 74% for Satellite I and 88% for Oct4, implying that Satellite I is demethylated during early embryo development, whereas Oct4 remains hypermethylated. In conclusion, these improved methods will benefit further studies of chromatin remodeling in early bovine pre-implantation embryos.

This project was supported by National Research Initiative Competitive Grant no. 2006-35203-16620 from the USDA Cooperative State Research, Education, and Extension Service.