Aim 1. We created an outbred population of flies from the five longest-sleeping and five shortest-sleeping DGRP lines. We performed a full diallel cross of these lines, and then randomly mated flies from the F1 generation of this cross to create the first generation of outbred flies. We continued the random mating procedure for 22 generations, allowing SNPs from the long and short-sleeping flies to recombine. We chose 96 representative SNPs from the genome-wide association study that segregate in the outbred population and developed Taqman assays for each SNP.
Aim 2. From the outbred population, we created six selection populations. Two populations were selected for short sleep, two for long sleep, and the remaining two populations were unselected controls. We measured sleep in each population every generation. We chose a subset of flies to create the subsequent generation as appropriate (for example, the shortest-sleeping flies from the population selected for short sleep become the parents for the next generation). We ascertained the effectiveness of the selection procedure by plotting the means of each selection population and comparing them to the controls. We used high-throughput genotyping methods to genotype single flies prior to selection and at various generations during the selection procedure. We determined the SNP allele frequencies in single flies of the outbred population prior to selection, and we will compare these frequencies that of SNPs in subsequent generations. We will associate SNP genotypes of single flies with their corresponding sleep. Significant SNPs denote genomic regions responsive to the selection procedure.
Aim 3. For each generation of selection, we froze groups of flies, separated by each selection population and sex. Genomic DNA has been extracted from a subset of these samples and sequenced in order to determine the frequency of all SNP alleles present in the genome at each generation. We are have applied two different methods of analysis in order to determine how allele frequencies change in each selection population and over time. These methods will be compared in order to determine which method, if any, are the most appropriate. This analysis will pinpoint those SNPs most responsive to the artificial selection procedure. This data was reported in a manuscript we published in 2017.
Aim 4. Polymorphic variants may influence gene expression, and the signal of gene expression may be detectable across generations. We froze groups of flies of each sex and selection population for each generation. Total RNA has been extracted from these samples, and poly-A+ RNA sequenced. We have aligned the sequence data and estimated read counts for each gene. We have used our previously validated RNA-Seq pipeline to conduct an analysis of differential expression among selection schemes, generations, and sexes. We applied a Gaussian process model to genes significant for this analysis in order to identify putatively interacting genes. We are in the process of validating the candidate genes we identified.
Aim 5. Although artificial selection necessarily results in some inbreeding, genetic variation is still present in the selection populations that could cause behavioral phenotypes to revert once selection is relaxed. We have therefore created 10 inbred lines for each selection population via full-sib mating for 20 generations. This procedure resulted in 39 inbred lines: 19 long-sleeping lines, and 20 short-sleeping lines. We refer to these inbred lines as the Sleep Inbred Panel. These lines have been fully sequenced. This is a community resource, and the fly stocks are available at the Bloomington, Indiana Stock Center. The sequences are available for download through the SRA. A manuscript describing the construction of the Sleep Inbred Panel and the sequence data was published in 2018 in G3: The Sleep Inbred Panel, a collection of inbred Drosophila melanogaster with extreme long and short sleep duration.
Aim 6. Previous studies of mutations that alter sleep phenotypes showed that flies with extreme sleep have deficits in other important characteristics such as locomotion, lifespan, and learning and memory. We will measure relevant life history and fitness traits in the selection lines and compare them to the performance of unselected controls. Decreased performance in these measures may be a consequence of a long- or short-sleeping phenotype, or they may reflect a shared genetic architecture between sleep and the trait. We have already examined lifespan and egg-to-adult viability for these populations. Our current goal is to examine neural correlates of sleep in orderto determine whether extreme long- or short-sleeping flies have differences in cognitive functioning or brain structure. We have measured learning and memory as well as neural correlates of sleep in the inbred lines mentioned in Aim 5 above. The results have been submitted for publication.
Aim 7. One of the greatest challenges in functional genomics is to understand how polymorphic variants actually alter sleep phenotypes. CRISPR technology makes it possible to replace one allelic variant with its alternate. We are using the extreme long and short sleepers that we created in Aim 5 for this project (the Sleep Inbred Panel, SIP). We are in the process of replacing the long-sleeper allele from three polymorphisms in a long-sleeping line of the SIP with the short-sleeper alleles. In addition, we have developed a high-throughput method using CRISPR to screen the 126 candidate polymorphisms (80 genes) for long and short sleep that we identified in the artificial selection experiment. This experiment is ongoing.
| Serrano Negron, Yazmin L; Hansen, Nancy F; Harbison, Susan T (2018) The Sleep Inbred Panel, a Collection of Inbred Drosophila melanogaster with Extreme Long and Short Sleep Duration. G3 (Bethesda) 8:2865-2873 |
| Lee, Hangnoh; Cho, Dong-Yeon; Wojtowicz, Damian et al. (2017) Dosage-Dependent Expression Variation Suppressed on the Drosophila Male X Chromosome. G3 (Bethesda) : |
| Harbison, Susan T; Serrano Negron, Yazmin L; Hansen, Nancy F et al. (2017) Selection for long and short sleep duration in Drosophila melanogaster reveals the complex genetic network underlying natural variation in sleep. PLoS Genet 13:e1007098 |
| Harbison, Susan T; McCoy, Lenovia J; Mackay, Trudy F C (2013) Genome-wide association study of sleep in Drosophila melanogaster. BMC Genomics 14:281 |
