DBT affects sleep in both circadian and non-circadian neurons.

Sleep is a very important behavior observed in almost all animals. Importantly, sleep is subject to both circadian and homeostatic regulation. The circadian rhythm determines the daily alternation of the sleep-wake cycle, while homeostasis mediates the rise and dissipation of sleep pressure during the wake and sleep period. As an important kinase, dbt plays a central role in both circadian rhythms and development. We investigated the sleep patterns of several ethyl methanesulfonate-induced dbt mutants and discuss the possible reasons why different sleep phenotypes were shown in these mutants. In order to reduce DBT in all neurons in which it is expressed, CRISPR-Cas9 was used to produce flies that expressed GAL4 in frame with the dbt gene at its endogenous locus, and knock-down of DBT with this construct produced elevated sleep during the day and reduced sleep at night. Loss of sleep at night is mediated by dbt loss during the sleep/wake cycle in the adult, while the increased sleep during the day is produced by reductions in dbt during development and not by reductions in the adult. Additionally, using targeted RNA interference, we uncovered the contribution of dbt on sleep in different subsets of neurons in which dbt is normally expressed. Reduction of dbt in circadian neurons produced less sleep at night, while lower expression of dbt in noncircadian neurons produced increased sleep during the day. Importantly, independently of the types of neurons where dbt affects sleep, we demonstrate that the PER protein is involved in DBT mediated sleep regulation.