| Download datasets and supplementary data files |
|---|
Summary
Determining the genetic architecture of complex traits is challenging because phenotypic variation arises from interactions between multiple, environmentally sensitive alleles. We quantified genomewide transcript abundance and phenotypes for six ecologically relevant traits in D. melanogaster wild-derived inbred lines. We observed 10,096 genetically variable transcripts and high heritabilities for all organismal phenotypes. The transcriptome is highly genetically inter-correlated, forming 241 transcriptional modules. Modules are enriched for transcripts in common pathways, gene ontology categories, tissue-specific expression, and transcription factor binding sites. The high transcriptional connectivity allows us to infer genetic networks and the function of predicted genes based on annotations of other genes in the network. Regressions of organismal phenotypes on transcript abundance implicate several hundred candidate genes that form modules of biologically meaningful correlated transcripts affecting each phenotype. Overlapping transcripts in modules associated with different traits provides insight into the molecular basis of pleiotropy between complex traits.
Full Article
Experiment design
We derived inbred lines from the Raleigh, USA population by 20 generations of full-sib mating. We used the C(2L)RM-P1, b1; C(2R)RM-SKIA, cn1bw1 compound autosome (CA) stock for fitness assays. P-element mutations and co-isogenic control lines were a gift of H. Bellen (Howard Hughes Medical Institute, Baylor College of Medicine). We reared flies on cornmeal-molasses-agar medium at 25 °C, 60–75% relative humidity and a 12-h light-dark cycle unless otherwise specified.
Organismal phenotypes.
For the starvation stress resistance group, we placed ten same-sex, 2-d-old flies in vials containing 1.5% agar and 5 ml water, and scored survival every eight hours (N = 5 vials/sex/line). For the chill coma recovery group, we placed 3- to 7-d-old flies in empty vials on ice for three hours, and recorded the time for each individual to right itself after transfer to room temperature (N = 20 flies/sex/line). For longevity, we placed five 1- to 2-d-old same-sex virgin flies into vials containing 5 ml medium, and recorded survival every two days (N = 5 vials/sex/line). For locomotor reactivity, we placed single 3- to 7-d-old flies into vials containing 5 ml medium. The following day, between 8 am and 12 pm, we mechanically disturbed each fly19, and recorded the total activity in the 45 s immediately following the disturbance. We obtained two replicate measurements of 20 flies/sex/replicate/line. For the copulation latency group, we aspirated pairs of 3- to 7-d-old virgin flies into vials containing 5 ml medium between 8 am and 12 pm, and recorded the number of minutes until initiation of copulation, for a maximum of 120 min (N = 24 pairs/line). For the reproductive fitness group, we used the competitive index technique45, 46. We reared all wild-type and CA parents in constant density (10 pairs) vials. We placed six 3- to 4-d-old virgin CA males and females and three 3- to 4-d-old wild-type males and females in a vial containing 10 ml medium, discarding the flies after 7 d. The competitive index was the ratio of the number of wild type to the total number of progeny emerging by day 17 (N = 20 replicate vials/line).
About cases
The raw microarray data are deposited in the ArrayExpress database (www.ebi.ac.uk/arrayexpress,) under accession number E-MEXP-1594
We have derived a population of 192 inbred lines by 20 generations of full sib inbreeding of isofemale lines collected from the Raleigh, NC population. A White Paper to obtain complete genome sequences of these lines (link to pdf of White Paper) has been approved by the National Institutes of Health National Human Genome Research Institute, using a combination of 454 XLR long read pyrosequencing and paired end Solexa short read (currently 45 bp) technology. The sequencing is being done at the Baylor College of Medicine Sequencing Center, in collaboration with Richard Gibbs and Stephen Richards.
About tissue
Whole body. 3- to 5-d-old flies from the inbred lines. All samples were frozen between 1 and 3 pm. We extracted RNA from two independent pools (25 flies/sex/line)We derived inbred lines from the Raleigh, USA population by 20 generations of full-sib mating. We used the C(2L)RM-P1, b1; C(2R)RM-SKIA, cn1bw1 compound autosome (CA) stock for fitness assays. P-element mutations and co-isogenic control lines were a gift of H. Bellen (Howard Hughes Medical Institute, Baylor College of Medicine). We reared flies on cornmeal-molasses-agar medium at 25 °C, 60–75% relative humidity and a 12-h light-dark cycle unless otherwise specified.
About platform
We used Affymetrix Drosophila 2.0 arrays to assess transcript profiles of 3- to 5-d-old flies from the inbred lines. All samples were frozen between 1 and 3 pm. We extracted RNA from two independent pools (25 flies/sex/line), and hybridized 10 g fragmented cRNA to each array. We randomized RNA extraction, labeling and array hybridization across all samples, and normalized the raw array data across sexes and lines using a median standardization.
Each transcript is represented by 14 perfect-match 25-bp oligonucleotides. To identify perfect-match probes with SFPs between the wild-derived lines and the strain used to design the array, we quantified the maximal degree to which the variation between lines could be reduced by partitioning the lines into two allelic classes. We computed the sum of squared deviations from each class mean and expressed their sum as a fraction of the total sum of squares. The smallest fraction across all bipartitions was used to score each probe. We identified 3,136 candidate SFPs with scores 0.1 (a tenfold reduction in the sum of squares). We validated polymorphisms in 20 of 21 of these SFPs by designing primers flanking the SFP and sequencing the PCR products (data not shown).
About data processing
Our measure of expression for each probe set was the median log2 signal intensity of perfect-match probes without SFPs. We used negative control probes to estimate the background intensity, and removed probes below this threshold.
Data may have been normalized prior to entry into GeneNetwork using 2z + 8 transform. This method simply stabilizes the variance of all array data sets, and resets the z score to 8 (rather than 0).
Contributors
Ayroles JF, Carbone MA, Stone EA, Jordan KW, Lyman RF, Magwire MM, Rollmann SM, Duncan LH, Lawrence F, Anholt RR, Mackay TF.
Data entry by Arthur Centeno, UTHSC.
Citation
Ayroles JF, Carbone MA, Stone EA, Jordan KW, Lyman RF, Magwire MM, Rollmann SM, Duncan LH, Lawrence F, Anholt RR, Mackay TF (2009) Systems genetics of complex traits in Drosophila melanogaster. Nat Genet 41(3):299-307. PMID: 19234471
Acknowledgment
Mackay laboratory: http://mackay.gnets.ncsu.edu/MackaySite/Homepage.html
Notes
Please see Mackay lab link: http://mackay.gnets.ncsu.edu/MackaySite/DGRP_files/The40ExpressionDataMatrix10096.txt
ArrayExpress: accession number E-MEXP-1594