Genome Biology 2005, 6:R36
comment reviews reports deposited research refereed research interactions information
Open Access
2005Harbisonet al.Volume 6, Issue 4, Article R36
Research
Quantitative genomics of starvation stress resistance in Drosophila
Susan T Harbison*†§, Sherman Chang‡, Kim P Kamdar‡ and
Trudy FC Mackay*†
Addresses: *Department of Genetics, North Carolina State University, Raleigh, NC 27695, USA. †WM Keck Center for Behavioral Biology, North
Carolina State University, Raleigh, NC 27695, USA. ‡The Torrey Mesa Research Institute, 3115 Merryfield Row, San Diego, CA 92121, USA.
§Current address: Department of Neuroscience, University of Pennsylvania Medical School, Philadelphia, PA 19104, USA.
Correspondence: Trudy FC Mackay. E-mail: trudy_mackay@ncsu.edu
© 2005 Harbison et al.; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Quantitative genomics of starvation stress resistance in Drosophila<p>The efficacy of transcriptional profiling for identifying networks of pleiotropic genes regulating complex traits was assessed. The tran-scriptional response to starvation stress in males and females of the Oregon-R and 2b <it>Drosophila</it> strains, as well as four recom-binant inbred lines derived from them, was shown to be different between the sexes and to involve approximately 25% of the genome.</p>
Abstract
Background: A major challenge of modern biology is to understand the networks of interacting
genes regulating complex traits, and the subset of these genes that affect naturally occurring
quantitative genetic variation. Previously, we used P-element mutagenesis and quantitative trait
locus (QTL) mapping in Drosophila to identify candidate genes affecting resistance to starvation
stress, and variation in resistance to starvation stress between the Oregon-R (Ore) and 2b strains.
Here, we tested the efficacy of whole-genome transcriptional profiling for identifying genes affecting
starvation stress resistance.
Results: We evaluated whole-genome transcript abundance for males and females of Ore, 2b, and
four recombinant inbred lines derived from them, under control and starved conditions. There
were significant differences in transcript abundance between the sexes for nearly 50% of the
genome, while the transcriptional response to starvation stress involved approximately 25% of the
genome. Nearly 50% of P-element insertions in 160 genes with altered transcript abundance during
starvation stress had mutational effects on starvation tolerance. Approximately 5% of the genome
exhibited genetic variation in transcript abundance, which was largely attributable to regulation by
unlinked genes. Genes exhibiting variation in transcript abundance among lines did not cluster
within starvation resistance QTLs, and none of the candidate genes affecting variation in starvation
resistance between Ore and 2b exhibited significant differences in transcript abundance between
lines.
Conclusions: Expression profiling is a powerful method for identifying networks of pleiotropic
genes regulating complex traits, but the relationship between variation in transcript abundance
among lines used to map QTLs and genes affecting variation in quantitative traits is complicated.
Background
Quantitative traits affecting morphology, physiology, behav-
ior, disease susceptibility and reproductive fitness are con-
trolled by multiple interacting genes whose effects are
conditional on the genetic, sexual and external environments
[1]. Advances in medicine, agriculture, and an understanding
of adaptive evolution depend on discovering the genes that
regulate these complex traits, and determining the genetic
Published: 24 March 2005
Genome Biology 2005, 6:R36 (doi:10.1186/gb-2005-6-4-r36)
Received: 24 August 2004
Revised: 22 December 2004
Accepted: 23 February 2005
The electronic version of this article is the complete one and can be
found online at http://genomebiology.com/2005/6/4/R36
R36.2 Genome Biology 2005, Volume 6, Issue 4, Article R36 Harbison et al. http://genomebiology.com/2005/6/4/R36
Genome Biology 2005, 6:R36
and molecular properties of alleles at loci that cause segregat-
ing genetic variation in natural populations. Assessing subtle
effects of induced mutations on quantitative trait phenotypes
in model organisms is a straightforward approach to identify
genes regulating complex traits [1-3]. However, the large
number of potential mutations to evaluate, the necessity to
induce mutations in a common inbred background, and the
level of replication required to detect subtle effects [1] all limit
the feasibility of systematic whole-genome mutagenesis
screens for complex traits in higher eukaryotes. Mapping
quantitative trait loci (QTLs) affecting variation in complex
traits to broad genomic regions by linkage to polymorphic
molecular markers is also straightforward. However, our abil-
ity to determine what genes in the QTL regions cause the trait
variation is hampered by the large number of recombinants
required for high-resolution mapping, and the small and
environmentally sensitive effects of QTL alleles [1,4].
There has been great excitement recently about the utility of
whole-genome transcriptional profiling to identify candidate
genes regulating complex traits, by assessing changes in gene
expression in the background of single mutations affecting
the trait [5,6], between lines selected for different phenotypic
values of the trait [7], and in response to environmental stress
and aging [8-12]. Transcript abundance is also a quantitative
trait for which there is considerable variation between wild-
type strains [11,13-17], and for which expression QTLs
(eQTLs) [18] have been mapped [15-17,19]. Thus, candidate
genes affecting variation in quantitative trait phenotypes are
those for which the map positions of trait QTL and eQTL coin-
cide [16,20].
Transcript profiling typically implicates hundreds to thou-
sands of genes in the regulation of quantitative traits and
associated with trait variation between strains; the majority
of these genes are computationally predicted genes that have
not been experimentally verified. To what extent do changes
in transcript abundance predicate effects of induced muta-
tions and allelic variants between strains on quantitative trait
phenotypes? It is encouraging that several studies have con-
firmed the phenotypic effects of mutations in genes impli-
cated by changes in expression [5-7]. However, limited
numbers of genes were tested, and their choice was not unbi-
ased. None of the candidate QTLs nominated by transcrip-
tional profiling has been validated according to the rigorous
standards necessary to prove that any candidate gene corre-
sponds to a QTL [1,4]. To begin to answer this question, we
need to compare gene-expression data with genes known to
affect the trait from independent mutagenesis and QTL map-
ping studies. This comparison has not been possible to date
because there are only a few complex traits for which the
genetic architecture is known at this level of detail, one of
which is resistance to starvation stress in Drosophila.
Previously, we used P-element mutagenesis in an isogenic
background to identify 383 candidate genes affecting starva-
tion tolerance in D. melanogaster [21]. Further, we mapped
QTLs affecting variation in starvation resistance between two
isogenic Drosophila strains, Oregon-R (Ore) and 2b [21], fol-
lowed by complementation tests to mutations to identify
twelve candidate genes affecting variation in starvation
resistance between these strains [21]. Here, we used Affyme-
trix Drosophila GeneChips to examine expression profiles of
two starvation-resistant and two starvation-sensitive recom-
binant inbred (RI) lines, as well as parental lines Ore and 2b,
under normal and starvation stress conditions. We used a sta-
tistically rigorous analysis to identify genes whose expression
was altered between the sexes, during starvation stress treat-
ment, between lines, and interactions between these main
effects. In the comparison of expression profiling with the P-
element mutagenesis performed previously, we found nearly
50% concordance between the effects of 160 P-element muta-
tions on starvation stress resistance and changes in gene
expression during starvation - 77 mutations with significant
effects also had significant changes in transcript abundance,
while 83 mutations did not affect the starvation resistance
phenotype, yet had significant changes in transcript level. We
identified 153 novel candidate genes for which there was var-
iation in gene expression between the lines and which co-
localized with starvation resistance QTLs. However, we did
not detect genetic variation in expression for any of the can-
didate genes identified by complementation tests. Our efforts
to associate genetic variation in expression with variation in
quantitative trait phenotypes is confounded by the observa-
tion of widespread regulation of transcript abundance by
unlinked genes, the difficulty in detecting rare transcripts
that may be expressed in only a few cell types at a particular
period of development, and genetic variation between QTL
alleles that is not regulated at the level of transcription.
Results
The sexually dimorphic transcriptome
Nearly one-half of the genome (6,569 probe sets) exhibited
significantly different transcript levels between the sexes
(P(Sex) < 0.001), with 3,965 probe sets upregulated in
females and 2,604 probe sets upregulated in males (the com-
plete list is given in Additional data file 1). The greatest differ-
ences in transcript abundance between the sexes were for
probe sets implicated in sex-specific functions: chorion, vitel-
line membrane, and yolk proteins involved in egg production
were upregulated in females; and accessory gland peptides,
male-specific RNAs, and protein ejaculatory bulb compo-
nents were upregulated in males. However, the probe sets
exhibiting sex dimorphism in expression fell into 28 biologi-
cal process and 41 molecular function Gene Ontology (GO)
categories; for most of these categories, differences in expres-
sion between the sexes was unexpected. We determined
which GO categories contained significantly different num-
bers of upregulated probe sets in males and females (Table 1).
Genes involved in the biological process categories of cell
communication, cell growth and/or maintenance,
http://genomebiology.com/2005/6/4/R36 Genome Biology 2005, Volume 6, Issue 4, Article R36 Harbison et al. R36.3
comment reviews reports refereed researchdeposited research interactions information
Genome Biology 2005, 6:R36
Table 1
Gene Ontology categories with sex-biased gene expression
Gene Ontology category Number of upregulated probe sets P-value*
Females Males
Biological process
Cell communication
Signal transduction 135 40 <0.0001
Cell growth and/or maintenance
Cell cycle 184 15 < 0.0001
Cell organization and biogenesis 207 65 < 0.0001
Transport 123 49 < 0.0001
Biosynthesis 238 43 < 0.0001
Catabolism 71 24 < 0.0001
Nucleic acid metabolism 374 28 < 0.0001
Phosphorous metabolism 147 60 <0.0001
Protein metabolism 495 113 < 0.0001
Development
Cell differentiation 33 11 7.41 × 10-4
Embryonic development 126 27 < 0.0001
Morphogenesis 200 50 < 0.0001
Pattern specification 76 9 <0.0001
Post-embryonic 50 11 < 0.0001
Gametogenesis 164 20 < 0.0001
Other development 84 17 < 0.0001
Cell death 25 5 1.54 × 10-4
Molecular function
Binding
DNA binding 310 46 < 0.0001
Nuclease 31 3 < 0.0001
RNA binding 180 38 < 0.0001
Translation factor 40 13 1.58 × 10-4
Nucleotide binding 187 68 < 0.0001
Protein binding
Cytoskeletal protein binding 89 43 < 0.0001
Transcription factor binding 28 3 < 0.0001
Enzymes
Hydrolase enzyme
Acting on acid anhydrides 177 94 < 0.0001
Acting on ester bonds 113 56 < 0.0001
Kinase enzyme 156 62 < 0.0001
Ligase enzyme 52 18 < 0.0001
Oxidoreductase enzyme 69 139 < 0.0001
Transferase enzyme 327 105 < 0.0001
Other enzymes 88 16 < 0.0001
Signal transducer
Signal transducer - receptor signaling protein 89 14 < 0.0001
R36.4 Genome Biology 2005, Volume 6, Issue 4, Article R36 Harbison et al. http://genomebiology.com/2005/6/4/R36
Genome Biology 2005, 6:R36
development, and cell death were upregulated more often in
females than in males. Genes involved in the molecular func-
tion categories of binding, most enzymes, signal transduc-
tion, structural molecules, and regulation of transcription
and translation were upregulated in females more often than
in males; however, genes encoding oxidoreductase enzymes,
carrier transporters and ion transporters were upregulated in
males more often than in females (Table 1).
The genomic distribution of sex-biased genes was not random
(Figure 1). There was a paucity of male-biased genes on the X
and fourth chromosomes, and an excess on chromosome 2R
(χ25 = 100.77; P < 0.0001). There was a deficit of female-
biased genes on chromosome 4, and an excess on chromo-
some 2R(χ25 = 29.18; P < 0.0001).
Transcriptional response to starvation stress
We found 3,451 probe sets with significantly different mean
transcript levels between the control and starved conditions
(P(treatment) < 0.001): 1,736 were downregulated (some by
as much as 40-fold) and 1,715 were upregulated (at most by
7.2-fold) during starvation (the complete list is available as
Additional data file 2). These probe sets fell into 24 biological
process and 25 molecular function GO categories. We deter-
mined which GO categories had a significantly different
number of up- and downregulated probe sets in response to
starvation stress. Genes affecting the biological processes of
protein and nucleic-acid metabolism (protein biosynthesis;
protein catabolism, folding, localization, modification, and
repair; biosynthesis of nucleic acid macromolecules and lip-
ids) were upregulated during starvation (Table 2). The
expression of genes in three molecular function categories
(nucleotide binding, hydrolases binding to acid anhydrides,
and ribosome structure) increased during starvation; while
defense/immunity proteins, peptidases, cuticle structural
proteins, and carrier transport proteins were downregulated
(Table 2).
The treatment × sex interaction term was significant (P <
0.001) for 817 probe sets, of which 715 had significant treat-
ment effects for one or both sexes in the separate sex analyses
(Additional data file 3). We categorized these 715 probe sets
as sex-specific if significant expression changes in response to
starvation occurred in one sex only; as sex-biased if expres-
Structural molecule
Ribosome structure 137 8 < 0.0001
Transcription regulator 199 35 < 0.0001
Translation regulator 42 13 < 0.0001
Transporter
Carrier transporter 82 143 < 0.0001
Ion transporter 30 70 < 0.0001
*Significant after Bonferroni correction.
Table 1 (Continued)
Gene Ontology categories with sex-biased gene expression
Chromosome locations of genes differentially expressed by sexFigure 1
Chromosome locations of genes differentially expressed by sex. (a)
Observed (magenta) and expected (blue) number of probe sets
upregulated in males. (b) Observed (magenta) and expected (blue)
numbers of probe sets upregulated in females.
Number
Chromosome
X2L2R3L3R4
Number
0
100
200
300
400
500
600
700
Chromosome
X2L2R3L3R4
0
200
400
600
800
1,000
1,200
(a)
(b)
http://genomebiology.com/2005/6/4/R36 Genome Biology 2005, Volume 6, Issue 4, Article R36 Harbison et al. R36.5
comment reviews reports refereed researchdeposited research interactions information
Genome Biology 2005, 6:R36
sion levels changed in the same direction in both sexes, but
were of different magnitude; or as sex-antagonistic if expres-
sion levels significantly changed in both sexes, but in opposite
directions (Figure 2a-c). Most probe sets exhibited sex-spe-
cific or sex-biased expression, with only two genes, CG14095
and Rpd3, meeting the sex-antagonistic criterion. More
probe sets exhibiting sex-specific or sex-biased expression
were downregulated (454) than upregulated (263) during
starvation. Starvation stress was accompanied by reduced
expression of genes involved in the developmental processes
of gametogenesis and sex determination as well as signal
transduction in females, and of genes involved in mechano-
sensory and reproductive behavior in males (Table 2).
Transcript abundance versus mutations
The genes represented by probe sets with significant treat-
ment and/or treatment × sex effects are candidate genes for
starvation resistance. Previously, we screened 933 co-iso-
genic single P-element insertion lines for their effect on star-
vation resistance [21]. Of these insertions, 383 had significant
effects on starvation resistance, while the remaining 550 did
not [21]. Of the 933 lines, we know the locations of the 385 of
the inserts and that genes tagged by these inserts are repre-
sented on the array. Thus, we can directly compare the extent
to which effects of P-element mutations on the starvation
phenotype correspond to changes in transcript abundance in
response to starvation. This comparison allows us to assess
the hypothesis that changes in transcript abundance can be
used to identify candidate genes with effects on phenotype, an
hypothesis implicit in previous microarray studies [5-7].
Overall, there was no statistical association between the phe-
notypic and transcript data (χ21 = 0.0006, P = 1). For 194
genes, there was agreement between the phenotype and the
expression level. Seventy-seven genes had significant differ-
ences in both transcript profile and mutant phenotypes, and
117 genes affected neither phenotype nor expression level
(Additional data file 4). There was disagreement between the
expression and phenotypic analyses for 191 genes (49.6%):
Table 2
Gene Ontology categories with increased or decreased gene expression during starvation
Gene Ontology category Number of probe sets P-value*
Upregulated Downregulated
Biological process
Cell growth and/or maintenance
Biosynthesis 119 31 < 0.0001
Protein metabolism 220 95 < 0.0001
Development 12 35 6.48 × 10-4†
Behavior 1 9 8.10 × 10-3‡
Molecular function
Binding
Nucleotide binding 76 38 3.36 × 10-4
Defense/immunity protein 3 18 6.55 × 10-4
Enzymes
Hydrolase
Acting on acid anhydrides 77 42 1.25 × 10-3
Peptidase 50 104 1.12 × 10-5
Structure
Cuticle structure 1 14 3.09 × 10-4
Ribosome structure 84 3 < 0.0001
Transporter
Carrier 46 84 8.05 × 10-4
Signal transducer 2 12 5.67 × 10-3†
*Significant after Bonferroni correction; †significant for females only; ‡significant for males only.