Genetic basis of variation in high sugar-induced diabetes-associated traits and development delay in Drosophila
Authors: Xuan Zhuang 1; Fabio Morgante 2; Abaranjitha Muniyasamy 1; Michael Ludwig 3; Soo Young Park 3; Yang Li 3; Matthew Stephens 3; Graeme Bell 3; Martin Kreitman 3
Affiliations: 1) Department of Biological Sciences, University of Arkansas, Fayetteville, AR; 2) Clemson Center for Human Genetics, Clement University, Clemson, SC; 3) Department of Ecology and Evolution, University of Chicago, Chicago, IL
Keywords: o. genotype-by-environment interaction; n. diabetes
Drosophila is a well-established model for investigating complex traits and many human diseases. It provides powerful tools for dissecting the contributions of both genes and environment on development and metabolism. We developed a Drosophila model to study high sugar diet (HSD) induced Type 2 diabetes (T2D) associated traits in large populations with different genetic backgrounds. We examined HSD induced phenotypes in larvae and in adults across a subset of Drosophila Genetic Reference Panel (DGRP). Flies on HSD display an increase in whole body glucose and glycogen level, a decrease in developmental rate, survivorship, body weight, and longevity, compared with flies under low sugar diet (LSD). The examined DGRP lines display a continuous and wide range of these phenotypes and large broad-sense heritability, suggesting great potential for quantitative trait loci (QTL) mapping. In the meanwhile, we developed a unique experimental system for genetic mapping named Drosophila Recombinant Populations (DRPs), which are consisted of 16 outbred advanced intercross populations (AIPs), each founded with 8 inbred DGRP lines. The DRPs provide about 70,000 genotypically distinct flies that allow us to apply the HSD induced T2D model to each individual to investigate the genetic architecture of these complex traits. We used one of the HSD induced traits, namely developmental delay, to perform a bulk segregant mapping analysis of extreme phenotypes. We developed a computational pipeline with a Hidden Markov model (HMM) to impute the whole genome of each fly based on their low-coverage sequenced genomes and the available high-coverage founder genomes. Genome-wide association studies (GWAS) identified 76, 2009, and 373 polymorphisms at p<10−5 for the analysis of HSD, LSD, and G x E (gene by diet) respectively. Gene ontology (GO) analysis indicates predominant enrichment for genes involved in some hormones and ketones cellular metabolic processes among HSD candidates, and reproduction and mediator complex among LSD candidates. RNAi validation studies of the top candidates (Cyp9b2 and CG15088 from GxE analysis, wun and Ten-m from HSD analysis) indicated their moderate but still significant effects in increasing the development rate on HSD. This study provides a broad survey of diabetes associated traits in HSD induced T2D flies, and further dissects the genetic architecture of development time on diets of different sugar concentrations, and interaction of gene and diet.