0011 ARTIFICIAL SELECTION FOR LONG AND SHORT SLEEP DURATION IN DROSOPHILA MAPS TO BROAD DEVELOPMENTAL AND SIGNALING PATHWAY GENES

Abstract Introduction: Although many candidate genes have been identified for sleep, the extent to which naturally occurring polymorphisms can alter sleep duration is unknown. Here we applied an artificial selection scheme to a natural population of flies in order to determine how far night sleep duration could be driven up or down. Methods: We created an outbred population from ten Drosophila Genetic Reference Panel lines with variable sleep duration. From this outbred population, we created two short and two long-sleeping populations. We measured sleep in 100 males and 100 females of each population every generation, choosing the most extreme 25 sleepers of each sex as parents for the next generation. We maintained two additional populations as unselected controls. We used whole-genome sequences across seven generations to identify allelic variants responsible for the phenotypic changes. Results: After 13 generations of selection, long sleepers averaged 9.97 hours more nightly sleep than short sleepers. Mean 24-hour sleep was 198.9 ± 9.5 minutes in short sleepers, comparable to severe short-sleeping flies engineered with single-gene mutations; yet these flies were created from combinations of naturally occurring alleles. Selection for night sleep altered other sleep traits such as day sleep duration and night average bout length, but the numbers of night bouts were reduced in both long and short sleepers, increasing the consolidation of sleep. Flies from these populations also had normal lifespan, suggesting that there is little physiological consequence to being a natural long or short sleeper. Whole-genome sequences revealed thousands of changes in polymorphic allele frequencies between any two generations of selection. However, regressing allele frequency change across generations reduced the number of candidate polymorphisms to 126, implicating 80 candidate genes. Many of these genes could be connected via known physical and genetic interactions with several conserved developmental and signaling pathways such as the EGFR and Wnt pathways. Conclusion: Sleep duration can be driven to very high or low levels, suggesting that extreme short or long sleepers may exist in nature. Highly conserved broad developmental and signaling pathways influence natural variation in sleep duration. Support (If Any): Intramural Research Program, National Institutes of Health, NHLBI.