Throughout embryonic development, cells undergo a series of lineage decisions, accompanied by morphological and functional changes, culminating in the formation of a complete organism. This intricate process is orchestrated by a complex interplay of diverse genetic and epigenetic mechanisms, including DNA methylation. After major changes shaping the somatic DNA methylome in the pre-implantation embryo, this modification remains globally stable, with local alterations occurring in a tissue-specific manner, often associated with putative genetic regulatory elements. However, in human pluripotent stem cells (hPSCs), thousands of highly methylated regions are targeted by DNA demethylases (TETs), whose local demethylation activity is counteracted by de novo methyltransferases (DNMT3s), resulting in a delicate balance referred to as DNA methylation turnover. What is the molecular mechanism and its functional role during pluripotency and developmental progression remains elusive. In my doctoral work, I combined experimental and analytical approaches to investigate the emergence and regulation of DNA methylation turnover during human pluripotency and early differentiation. I revealed that this dynamic mechanism substantially occurs at regions that undergo demethylation during in vitro three germ-layer differentiation, but that it is also active at genomic loci linked to mature lineage decisions. Importantly, I described the establishment of de novo DNA methylation turnover in transient progenitor populations for the first time, suggesting an extended regulative role of the DNA methylation turnover beyond pluripotency. Furthermore, I provide functional evidence that pluripotency-associated DNA methylation turnover regions have enhancer activity in differentiated cells, implying a potential functional regulatory role of the turnover. Regarding transposable elements, my analysis confirms that the DNA methylation turnover is highly target-specific. In particular, I reveal that the evolutionary young ERV1 LTR7up1/2 and the hominoid-specific ERVK LTR5-Hs subfamilies of the long terminal repeat (LTR) retrotransposons are prominently targeted by the DNA methylation turnover in hPSCs. Interestingly, specifically these subfamilies were previously shown to be bound by pluripotency factors, including NANOG, providing a possible underlying mechanism behind the turnover during pluripotency. Lastly, I generated various genetically modified hPSCs lines to experimentally dissect the functional role of TETs and DNMT3s at turnover targets. Thus, my work provides a valuable toolkit and an unexplored analytical angle into the target-specific regulation of DNA methylation turnover, emphasizing its potential role for human cell differentiation during embryonic development.
