Using bisulfite sequencing and Southern blotting

Using bisulfite sequencing and Southern blotting analyses, we demonstrated previously that the major and minor satellites of pericentromeric heterochromatin are hypomethylated only in germ cell lineages, whereas all somatic SGX523 are hypermethylated, and that this DNA methylation profile is established during primordial germ cell development (Yamagata et al., 2007). This finding was further supported by the analysis of MethylRO mouse sections because most somatic cells exhibited the typical puncta pattern of mCherry-MBD-NLS, whereas testicular germ cells had obviously weaker or no signals. Although many studies have documented extensively the active DNA demethylation during preimplantation development (Reik, 2007; Wu and Zhang, 2010), the pericentromeric regions, which constitute 3.5% of the whole mouse genome that contains hypermethylated repeat sequences (Lehnertz et al., 2003; Waterston et al., 2002), are already hypomethylated, which seems to be the key epigenetic feature that distinguishes germ cells from somatic cells. In addition to the use of the MethylRO model as a reporter mouse for live-cell imaging, we have extended the application of this bioresource to MeDIP analysis. Conventional 5mC-mediated MeDIP is performed in vitro by binding extracted DNA with an anti-5mC antibody, which may yield unwanted results that do not reflect the in vivo status. In contrast, because the mCherry-MBD-NLS probe is expressed endogenously, we believe that it can capture the “exact moment” of DNA methylation dynamics. The use of an anti-RFP antibody for MeDIP-seq, and the fact that this reporter mouse is conditional, will allow us to perform cell-type- or tissue-specific MeDIP analysis by crossing with tissue-specific FLP Tg mice. Theoretically, this will result in an extremely low background, which is not possible with the conventional anti-5mC antibody-based methods. In conclusion, we have provided multiple evidences to show that MethylRO mice can capture the dynamic changes of the DNA methylation status both in vitro and in vivo. In particular, MethylRO mice can be used not only in live-cell imaging analyses but also in RFP-mediated MeDIP-seq and cross-section observation analyses, which extend the applications of this bioresource. Hence, we believe that this mouse model will become a powerful tool as well as technique to study DNA methylation dynamics during development, differentiation, and in pathological processes that lead to diseases.
Experimental Procedures
Acknowledgments We would like to thank Drs. Frank Costantini and Junji Takeda for providing the pBigT and ROSA26-targeting vector plasmids, Drs. Jan Ellenberg, Tomoya Kitajima, and Miho Ohsugi for the EGFP-CENP-C expression plasmid, Dr. Cristina M. Cardoso for the EGFP-PCNA expression plasmid, and Dr. Hitoshi Niwa for the Oct4-EGFP knockin-targeting vector. We would also like to thank Ms. Masumi Fujikawa for the MethylRO mouse illustration and Mr. Masanaga Muto for help with cryosectioning. Finally, we express our greatest gratitude to Drs. Masaru Okabe and Masahito Ikawa for their encouragement, guidance, and support throughout this work. This study was supported in part by a Grant-in-Aid from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.
Introduction T cells undergo most of their development in the thymus, the primary lymphoid organ that regulates their differentiation and maturation from blood-borne bone-marrow-derived precursors and appropriate selection for the induction of self-tolerance (Anderson et al., 2007). Thymic function is critically dependent on thymic epithelial cells (TECs), the most abundant cellular constituent of the stromal microenvironment. TECs are classified as two morphologically and functionally distinct subsets based on their localization to the thymic cortex (cTECs) or medulla (mTECs). TEC development and identity require the forkhead-box transcription factor Foxn1, which, in the mouse, demarcates the prospective thymic primordium within the third pharyngeal pouch at embryonic day 11.5 (E11.5). Loss of Foxn1 results in a “nude” phenotype in mice and rats (Nehls et al., 1994, 1996) and in humans (Pignata et al., 1996), characterized by congenital hairlessness and defective TEC differentiation, the latter of which results in the absence of functional T cells and severe immunodeficiency.