DNA methylation is an epigenetic modification found in every branch of life. An essential enzyme for the maintenance of DNA methylation patterns in mammals is DNA methyltransferase 1 (DNMT1). Its recruitment is regulated through its large N-terminus, which contains six annotated domains. Although most of these have been assigned a function, we are still lacking a holistic understanding of the enzyme's spatio-temporal regulation. Interestingly, a large segment of the N-terminus is devoid of any known domain and appears to be disordered in its sequence. Over the past years, such disordered sequences have increasingly gained attention, due to their role in forming biomolecular condensates through liquid-liquid phase separation (LLPS). These liquid compartments offer specific environmental conditions distinct from the surrounding that can enhance protein recruitment and function.
In this work, we explore a potential role for the intrinsically disordered domain (IDR) in the recruitment of DNMT1. Taking an evolutionary approach, we uncover that structural features of the region that are key for IDR function are highly conserved. Moreover, we find conserved biochemical signatures compatible with a role in LLPS. Using a reconstitution assay and an opto-genetic approach in cells, we for the first time show that the DNMT1 IDR is capable of undergoing LLPS in vitro and in vivo. In addition, we define a novel region of interest (ROI) of about 120 amino acids in the IDR that appears to have been inserted in the ancestor of eutherian mammals. Although the ROI has a distinct biochemical signature, we find no effect on the LLPS behavior of the IDR. Therefore, we discuss other potential roles of the ROI related to DNA methylation, for example, imprinting.
Finally, we lay the foundation for investigating a biological function of the IDR and establish a system for screening DNMT1 mutant phenotypes in mouse embryonic stem cells. Swift depletion of the endogenous protein is enabled by degron-mediated degradation, while our optimized construct design and efficient derivation strategy ensure the robust expression of the large transgenes. In combination with different methods for DNA methylation read-out, this system can now be used to study the role of the IDR and ROI in maintaining the steady-state level of DNA methylation against mechanisms of passive and active demethylation, but also for studying phenotypes affecting the efficiency of DNMT1 recruitment in the future.