Genetic and Genomic Response to Selection for Food Consumption in Drosophila melanogaster

Abstract

Food consumption is an essential component of animal fitness; however, excessive food intake in humans increases risk for many diseases. The roles of neuroendocrine feedback loops, food sensing modalities, and physiological state in regulating food intake are well understood, but not the genetic basis underlying variation in food consumption. Here, we applied ten generations of artificial selection for high and low food consumption in replicate populations of Drosophila melanogaster. The phenotypic response to selection was highly asymmetric, with significant responses only for increased food consumption and minimal correlated responses in body mass and composition. We assessed the molecular correlates of selection responses by DNA and RNA sequencing of the selection lines. The high and low selection lines had variants with significantly divergent allele frequencies within or near 2,081 genes and 3,526 differentially expressed genes in one or both sexes. A total of 519 genes were both genetically divergent and differentially expressed between the divergent selection lines. We performed functional analyses of the effects of RNAi suppression of gene expression and induced mutations for 27 of these candidate genes that have human orthologs and the strongest statistical support, and confirmed that 25 (93%) affected the mean and/or variance of food consumption.

Genetics of alcohol consumption in Drosophila melanogaster

“…We used mutational analyses to functionally implicate 86% of tested candidate genes in ethanol consumption in at least one sex, and to separate ethanol intake from sucrose intake per se . This level of validation represents an empirical false discovery rate similar to that obtained from validation studies on other phenotypes in GWA studies of the DGRP or extreme QTL mapping studies of DGRP-derived AIP (Garlapow et al , 2016, Morozova et al , 2015, Shorter et al , 2015, Swarup et al , 2013). …”

“…Some may also represent false positives resulting from genetic drift. Furthermore, since we only measured individual flies in the AIP for consumption of ethanol, we could not disentangle ethanol consumption from sucrose consumption, which is also genetically variable in the DGRP (Garlapow et al , 2016). …”

“…Failure to confirm some of the genes that harbor alleles with differential allele frequencies could be due to false positives or SNP location. For example, klg was previously associated with sucrose consumption (Garlapow et al , 2016) and ethanol sensitivity (Berger et al , 2008b, Morozova et al , 2015) and harbors 12 SNPs that vary across the 37 founder DGRP lines, but only one SNP ( 3R _18749072_SNP) showed significant association with ethanol consumption. This SNP is located in both a klg intron and upstream of the promotor of CG6600 .…”

“…Previous studies showed that many candidate genes affect the within-genotype environmental variance of sucrose consumption (Garlapow et al , 2016, Garlapow et al , 2015). Therefore, we estimated the effects of the candidate genes on the variance of ethanol and sucrose consumption relative to the appropriate control (Table S3) (Morgante et al , 2015).…”

“…To measure consumption of ethanol-supplemented food, we used a modified Capillary Feeding (CAFÉ) assay (Ja et al , 2007; Garlapow et al , 2016). At least 24 hours prior to the assay, we placed individual 3–7 day old virgin flies in vials containing culture medium following CO 2 anesthesia.…”