The identification of natural variants that contribute to a long circadian period in Drosophila melanogaster

National Heart, Lung, And Blood Institute

Investigators

Linked publications, trials & patents

Abstract

I. Our preliminary data suggests that differences in gene expression in the canonical circadian clock genes do not explain the long circadian period we observed in both male and female flies. We are therefore in the process of identifying potential candidate genes using RNA-Seq. Transcriptome profiles have been established for males and females of the long period line, and compared to flies having a normal circadian period. We measured gene expression across twelve time points in order to identify differences in cycling across a normal 24-hour day. We have frozen three replicates of 150 heads for each time point, sex, and genotype under dark:dark conditions. We have sequenced the RNA of these 144 samples. II. To identify genes with differences between the long period line and the normal period line, we first used a linear model to discern which genes were differentially expressed with respect to genotype and time. We identified 7,754 genes that were differentially expressed between the two genotypes; 2,704 that were differentially expressed among the different time points; and 27 genes that were differentially expressed for the genotype-by-time interaction term. Those genes that were differentially expressed with respect to time were analyzed further using both JTK-Cycle and RAIN in order to determine which of these genes may cycle in a circadian fashion. JTK-Cycle detected 343 genes and RAIN detected 1,286 genes that were putatively cycling. III. We narrowed the putative candidate genes further by identifying genes that overlapped between the transcriptomic analyses and our previous GWAS identifying 268 candidate genes for circadian rhythm phenotypes. We found 166 candidate genes common to both methods. From this, we tested 43 available Minos insertion mutations for circadian rhythm phenotypes. We found 11 mutations that significantly affected circadian rhythms. Three of these mutations had large (greater than 1 hour) effects, increasing circadian period over the control. We performed genetic rescue of one of the mutants. We published these findings, and the RNA-Seq data associated with this study were submitted to the Gene Expression Omnibus. IV. We are using a CRISPR strategy to perturb significant polymorphic variants in the long period line. This strategy will enable us to verify the importance of these polymorphisms. As a first step we will use polycistronic CRISPR to create small deletions within the candidate genes identified in III above. These mutants will be created in the genetic background of the long period line. We will determine whether the different combinations of these mutations can rescue the long circadian period.