Connecting TDG and active DNA demethylation with chromatin dynamics in mouse embryonic stem cells

Genetic stability is achieved through a sophisticated cellular DNA damage response that identifies and repairs various types of DNA lesions. Small DNA base lesions are processed by Base Excision Repair (BER), starting with the action of a DNA glycosylase that removes a specific damaged base from the DNA backbone. Thymine DNA glycosylase (TDG) is one of the eleven mammalian DNA glycosylases identified and was first discovered by its ability to recognize G•T mismatches, initiating BER by T excision. The latter finding that this DNA glycosylase, unlike all others, is essential for mouse embryogenesis led to the discovery of an epigenetic role of TDG-mediated BER, specifically in active DNA demethylation. According to the current model, active DNA demethylation starts with the engagement of TET enzymes at methylated cytosine in the context of CpG dinucleotides, which sequentially oxidize the 5-methylcytosine (5mC) to 5-hydroxymethylcytosine (5hmC), 5-formylcytosine (5fC), and 5-carboxylcytosine (5caC). The latter two oxidation states are then recognized by TDG as base damage and excised, starting the BER process that restores the unmethylated cytosine. While this process is mechanistically well understood, it does not explain by itself how TDG contributes to the regulation of chromatin plasticity as implicated by the severe phenotype of a TDG defect on stem cell functionality. In the recent years, studies focused on understanding how TET proteins network with epigenetic regulators to shape chromatin dynamics. For instance, it was shown that TET proteins interact with several protein complexes, including Polycomb Repressive Complex 2 (PRC2) and the SIN3A histone deacetylase (SIN3A/HDAC), to regulate transcription (Neri et al., 2013; Stolz et al., 2022; Wu et al., 2011), or can recruit OGT to activate the H3K4-methyltransferase SET1/COMPASS complex (Deplus et al., 2013; Vella et al., 2013). In contrast, the exploration of the TDG network within the chromatin landscape has been relatively neglected.
In my Ph.D. research, I took advantage of the BIOID2-MS method to investigate the functional space within which TDG operates in mouse embryonic stem cells (mESCs) and HEK293T cells. This approach allowed us to generate a comprehensive view of TDG’s protein interaction network partners. Besides confirming expected interactions with TET1 and the BER protein APE1 for the first time in a hypothesis-free, unbiased manner, the TDG-BIOID2 screen identified protein interactions that connect TDG activity with four new functional spaces beyond DNA repair; chromatin organization and gene transcription, RNA-mediated functions and RNA binding, ribosome biogenesis and function, and chromosomal organization. Together with colleagues, I established that TDG forms a regulatory network with chromatin remodelers (such as RUVBL1/2 and SMARCA4) and factors from the H3K4 methyltransferase complex (HCFC1, TAF9). This observation explains the substantial dysregulation of histone marks previously observed in TDG-deficient cells (Cortazar et al., 2011; Garcia-Gomez et al., 2017). My results further demonstrate an as-yet undescribed partnership of TDG with RNA-orchestrated genomic processes, showing that this DNA glycosylase interacts with different RNA-binding proteins and different long-noncoding RNAs with a role in epigenetic regulation. Interestingly, TDG can induce active DNA demethylation in the DNA hybrid situation of an R-loop context, suggesting a function in gene regulation through noncoding RNAs-mediated targeting of DNA demethylation.
Moreover, I interrogated the newly discovered functional interaction of TDG with SMCHD1 in mESC. SMCHD1 is critical for silencing the DUX4 locus, promoting the exit from totipotency (2C stage) in early embryogenesis (Lemmers et al., 2012; Ruebel et al., 2019). Single-cell RNA sequencing of mESC populations identified a 2C-like subpopulation to be upregulated in TDG-deficient cells. This rare population arises spontaneously in mESC cultures, shares features of the 2C stage, and can potentially differentiate into trophectoderm. We elucidated the interplay of TDG and SMCHD1 in the control of pluri- and totipotency, a poorly understood process. This revealed that TDG acts upstream of SMCHD1 in the regulation of early 2C-associated genes, like Duxf3 and Zscan4, whereas SMCHD1 contributes to later stages, controlling Duxbl genes and trophectoderm genes (Gata3 and Cdx2).
In summary, my experimental research towards my Ph.D. thesis has provided a first comprehensive view of the functional space of TDG-dependent active DNA demethylation, connecting the associated DNA repair activity with mechanisms of chromatin organization and gene control. The novel insight into the relevant TDG interactome and the validation of the most significant of these findings, including interactions with histone modifiers and long noncoding RNAs, helps explain the complex phenotype of TDG deficiency on DNA methylation, chromatin modification and structure, developmental gene regulation and cell differentiation and provides suitable entry points into the investigation of the underlying mechanisms. The research presented in my Ph.D. thesis paves the way for a more granular understanding of the function of TDG-dependent active DNA demethylation in the cellular plasticity.