DNA methylation restricts coordinated germline and neural fates in embryonic stem cell differentiation.

As embryonic stem cells (ESCs) transition from naive to primed pluripotency during early mammalian development, they acquire high DNA methylation levels. During this transition, the germline is specified and undergoes genome-wide DNA demethylation, while emergence of the three somatic germ layers is preceded by acquisition of somatic DNA methylation levels in the primed epiblast. DNA methylation is essential for embryogenesis, but the point at which it becomes critical during differentiation and whether all lineages equally depend on it is unclear. Here, using culture modeling of cellular transitions, we found that DNA methylation-free mouse ESCs with triple DNA methyltransferase knockout (TKO) progressed through the continuum of pluripotency states but demonstrated skewed differentiation abilities toward neural versus other somatic lineages. More saliently, TKO ESCs were fully competent for establishing primordial germ cell-like cells, even showing temporally extended and self-sustained capacity for the germline fate. By mapping chromatin states, we found that neural and germline lineages are linked by a similar enhancer dynamic upon exit from the naive state, defined by common sets of transcription factors, including methyl-sensitive ones, that fail to be decommissioned in the absence of DNA methylation. We propose that DNA methylation controls the temporality of a coordinated neural-germline axis of the preferred differentiation route during early development.

DNMT3A-dependent DNA methylation is required for spermatogonial stem cells to commit to spermatogenesis.

DNA methylation plays a critical role in spermatogenesis, as evidenced by the male sterility of DNA methyltransferase (DNMT) mutant mice. Here, we report a division of labor in the establishment of the methylation landscape of male germ cells and its functions in spermatogenesis. Although DNMT3C is essential for preventing retrotransposons from interfering with meiosis, DNMT3A broadly methylates the genome (with the exception of DNMT3C-dependent retrotransposons) and controls spermatogonial stem cell (SSC) plasticity. By reconstructing developmental trajectories through single-cell RNA sequencing and profiling chromatin states, we found that Dnmt3A mutant SSCs can only self-renew and no longer differentiate in association with spurious enhancer activation that enforces an irreversible stem cell gene program. Our findings therefore highlight a key function of DNA methylation in male fertility: the epigenetic programming of SSC commitment to differentiation and lifelong spermatogenesis supply.