Combinatorial quantification of 5mC and 5hmC at individual CpG dyads and the transcriptome in single cells reveals modulators of DNA methylation maintenance fidelity

Transmission of 5-methylcytosine (5mC) from one cell generation to the next plays a key role in regulating cellular identity in mammalian development and diseases. While recent work has shown that the activity of DNMT1, the protein responsible for the stable inheritance of 5mC from mother to daughter cells, is imprecise; it remains unclear how the fidelity of DNMT1 is tuned in different genomic and cell state contexts. Here we describe Dyad-seq, a method that combines enzymatic detection of modified cytosines with nucleobase conversion techniques to quantify the genome-wide methylation status of cytosines at the resolution of individual CpG dinucleotides. We find that the fidelity of DNMT1-mediated maintenance methylation is directly related to the local density of DNA methylation, and for genomic regions that are lowly methylated, histone modifications can dramatically alter the maintenance methylation activity. Further, to gain deeper insights into the methylation and demethylation turnover dynamics, we extended Dyad-seq to quantify all combinations of 5mC and 5-hydroxymethylcytosine (5hmC) at individual CpG dyads to show that TET proteins preferentially hydroxymethylate only one of the two 5mC sites in a symmetrically methylated CpG dyad rather than sequentially convert both 5mC to 5hmC. To understand how cell state transitions impact DNMT1-mediated maintenance methylation, we scaled the method down and combined it with the measurement of mRNA to simultaneously quantify genome-wide methylation levels, maintenance methylation fidelity and the transcriptome from the same cell (scDyad&T-seq). Applying scDyad&T-seq to mouse embryonic stem cells transitioning from serum to 2i conditions, we observe dramatic and heterogenous demethylation and the emergence of transcriptionally distinct subpopulations that are closely linked to the cell-to-cell variability in loss of DNMT1-mediated maintenance methylation activity, with regions of the genome that escape 5mC reprogramming retaining high levels of maintenance methylation fidelity. Overall, our results demonstrate that while distinct cell states can substantially impact the genome-wide activity of the DNA methylation maintenance machinery, locally there exists an intrinsic relationship between DNA methylation density, histone modifications and DNMT1-mediated maintenance methylation fidelity that is independent of cell state.