DNA methylation dynamics during zebra finch vocal development

Abstract: Like humans, songbirds are among the few animal groups that learn their vocalizations. They learn their song from a tutor early in life, analogous to human speech learning. In zebra finches, only the males sing, and song learning is restricted to a specific developmental time window, called the \u201ccritical period\u201d. This critical period consists of two overlapping phases: first, they memorize the song of an adult male tutor (sensory memorization phase), after which they actively start to vocalize themselves and gradually match their song to the memorized template (sensorimotor phase). We investigated which molecular mechanisms could explain why zebra finches can learn singing during, but no longer after the end of the critical period. Since the DNA sequence itself does not change during brain development, we hypothesized that epigenetic changes may be involved, as they determine when and where genes are expressed by altering the accessibility of the DNA. We focused on DNA methylation, an epigenetic modification that is known to play a role in learning, memory, and adaptation to the environment. Our results show that dynamic DNA methylation plays an important role in the brain regions controlling song learning and production, since these regions display very high levels of DNA methylation, as well as of the two antagonizing classes of enzymes that define the methylome: DNA methyltransferases (DNMT) catalyzing DNA methylation and ten-eleven translocation (TET) dioxygenases responsible for demethylation. DNA methylation levels were found to increase over the song learning period, specifically in these song control nuclei. In addition, we found many genes involved in brain plasticity to be regulated by DNA methylation. By the end of the critical period, the expression of DNMT and TET enzymes in the song control nuclei strongly decreased. A tutor song deprivation experiment showed that this decrease cannot explain the closing of the sensory phase of song learning. Whether it could explain the closing of the sensorimotor phase still remains to be investigated. Given the many parallels between vocal learning in songbirds and humans, unraveling the molecular mechanisms of vocal learning in songbirds could provide insights for the treatment of speech disorders. Moreover, the principles of critical period regulation seem largely generalizable between systems, meaning that a deeper understanding of the role of DNA methylation would provide new insights about how to re-induce youth-like neuroplasticity in the adult brain to improve recovery after brain injury, treat developmental disorders, or to enhance lifelong learning