Abstract
Genomewide gene expression patterns were investigated in inbred and noninbred Drosophila melanogaster lines under benign and stressful (high temperature) environmental conditions in a highly replicated experiment using Affymetrix gene chips. We found that both heat-shock protein and metabolism genes are strongly affected by temperature stress and that genes involved in metabolism are differentially expressed in inbred compared with noninbred lines, and that this effect is accentuated after heat stress exposure. Furthermore we show that inbreeding and temperature stress cause increased between-line variance in gene expression patterns. We conclude that inbreeding and environmental stress both independently and synergistically affect gene expression patterns. Interactions between inbreeding and the environment are often observed at the phenotypic level and our results reveal some of the genes that are involved at the individual gene level. Our observation of several metabolism genes being differentially expressed in inbred lines and more so after exposure to temperature stress, together with lower fitness in the investigated inbred lines, supports the hypothesis that superiority of heterozygous individuals partly derives from increased metabolic efficiency.
INBRED populations may suffer greatly from changes in the environment that noninbred populations perceive as nonstressful, indicating that intrinsic and extrinsic stresses may interact. There is evidence that such interaction can lead to synergism between genetic and environmental stresses causing normally benign environmental conditions to become harmful (Jiménez et al. 1994; Bijlsma et al. 1999; Reed et al. 2002; Bijlsma and Loeschcke 2005). Knowledge of the consequences of the combined effects of genetic load and stressful environmental conditions at the physiological level is limited. Genomic and proteomic technologies are useful tools in this regard (Holloway et al. 2002; MacBeath 2002) and combining traditional population and quantitative genetics with modern high-throughput “omics” technologies will help to enable an understanding as to which genes are important for given traits under different environmental conditions.
Organisms have developed a series of mechanisms to cope with environmental stress, including increased expression of stress proteins and changes in metabolism and hormone concentrations (Pletcher et al. 2002; Sørensen et al. 2003, 2005; Landis et al. 2004). Recently it has been shown that some of the same mechanisms are also affected by intrinsic stresses, such as inbreeding and aging (Kristensen et al. 2002, 2005; Pletcher et al. 2002; Landis et al. 2004; Pedersen et al. 2005). The deleterious consequences of inbreeding, environmental stress, and the interactions between them are of serious concern in diverse fields of biological sciences, such as evolutionary genetics, conservation biology, and animal breeding. To detect the genes affected by the two types of stress, we investigated gene expression profiles of male Drosophila melanogaster in 5 noninbred lines and 10 replicate lines inbred to the same level but at two different rates, after exposure to nonstressful or stressful (heat stress) environmental conditions.
MATERIALS AND METHODS
Inbreeding procedure and maintenance of the lines:
Inbred (fast and slow rate) and noninbred lines were generated from a genetically diverse mass population of D. melanogaster (see Bubliy and Loeschcke 2005 for details). Lines with expected equivalent levels of inbreeding (F ≈ 0.67) were obtained by two different rates of inbreeding, either through five generations of full-sib mating (fast inbreeding) or by maintaining a population size of two pairs during nine generations (slow rate). Five independent inbred lines were generated for each of the two breeding regimes (see Kristensen et al. 2005 for details on the inbreeding procedures and maintenance of stocks). After reaching the desired level of inbreeding, all lines were flushed to sizes of minimum 1000 individuals (within two generations). Five noninbred lines, each founded by ∼1000 breeding individuals, were established at the time when the inbreeding procedures were initiated. The noninbred lines and the flushed inbred lines were each kept in 10 200-ml bottles and within each line flies from all the bottles were mixed in every generation prior to setting up the next generation. Throughout and following the inbreeding procedure all flies were maintained in one climate room (25° ± 0.2°, 12/12-hr light/dark cycle).
Sampling of flies and environmental test conditions:
Twenty, 25, and 30 parental pairs were set up for egg laying in bottles before being discarded 24 hr later, within the noninbred, slow-, and fast-inbred lines, respectively. The bottles were 200-ml bottles containing 35 ml of the standard oatmeal–sugar–yeast–agar medium. The level of larval density was moderate and within the range considered to be optimal for developing D. melanogaster cultures (Barker and Podger 1970) in all bottles. The number of flies emerging from the bottles was not significantly different across the three treatments (noninbred: 356 ± 6, fast inbreeding: 387 ± 11; slow inbreeding: 357 ± 28).
Forty male flies (<12 hr old) were collected from each line by sampling from randomly chosen bottles. Sampling was done by four people in one afternoon and within 3 hr. Flies from the different treatments were sampled in rotating order, so that the time of collection and the person collecting were randomized between the treatments. Twenty flies per line were transferred to each of two 5-ml plastic vials. The plastic vials had a construction that allowed for ventilation when partly open. One plastic vial per line was placed in a 36° water bath for 1 hr followed by 1-hr recovery at 25° before being frozen in liquid nitrogen. During the 1-hr recovery period the vials were partly opened. The other plastic vial (with 20 males) per line was exposed to normal temperatures (25°) for 1 hr with the vials closed, followed by a 1-hr period (still at 25°) with the vials partly open. Thereafter flies were frozen in liquid nitrogen (at exactly the same time as the flies exposed to 36°).
RNA purification and microarray processing:
Isolation and labeling of RNA and microarray processing were performed as described elsewhere (Kristensen et al. 2005; Sørensen et al. 2005). Total RNA from whole flies was isolated using Trizol Reagent (Invitrogen). A total of 5 μg RNA was labeled using the SuperScript Choice System (Life Technologies) according to the manufacturer's instructions, except for using an oligo-dT primer containing a T7 RNA polymerase promoter site. Biotin-labeled cRNA was prepared using the BioArray High Yield RNA Transcript Labeling Kit (Enzo). A total of 15 μg of cRNA was loaded onto the Affymetrix probe array cartridge (Drosophila Genome Array Version 1). Hybridization mixture from each line and environmental regime (heat stress and nonstress) was loaded to one array yielding a total of 30 probe arrays for the experiment.
Statistical analysis
The data were analyzed using programs developed in R (version 2.1.1), a programming language and development environment for statistical computing and graphics (http://www.r-project.org/). Preprocessing of expression values was performed using the Robust Multi-array Analysis algorithm (Irizarry et al. 2003; Wu et al. 2004). In this algorithm, raw intensity values are background corrected on the basis of a model using sequence information followed by quantile normalization and a robust multichip fit with median polish (Wu et al. 2004). This algorithm combines the strengths of stochastic model-based algorithms and physical models and has been shown to be superior in accuracy and precision to other normalization methods such as MAS, RMA, and PerfectMatch (Wu and Irizarry 2004).
Differential expression of functional groups of genes in inbred lines compared with outbred lines and/or after exposure to 36° for 1 hr was assessed using the method proposed by Goeman et al. (2004). This method is based on an empirical Bayesian generalized linear model in which the regression coefficients between expression data and the sample treatments (e.g., breeding regime) are the random variables. The method investigates whether samples with similar treatments tend to have similar gene expression patterns. For a group of genes to be significantly differently expressed many of them need to be correlated with the sample treatment. This approach represents a powerful tool for the dissection of complex changes in gene expression and is complementary to analyses at the level of individual genes. Groups of genes significantly differentially expressed were assigned to functional categories using the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway information (Kanehisa and Goto 2000) downloaded from the DAVID database (http://apps1.niaid.nih.gov/david/) and the gene ontology (GO) (Gene Ontology Consortium 2001) information obtained from NetAffymetrix. We analyzed 60 groups defined by KEGG and 385 groups within the GO category Biological Process. Analyses based on KEGG pathway information involved testing only groups with ≥10 genes, and when analyzing on the basis of GO information only groups with ≥3 genes were tested. The P-values were calculated on the basis of 10,000 permutations and adjusted for multiple testing by controlling the false-discovery rate at the 5% level (Benjamini and Hochberg 1995).
Differential expression of individual genes was assessed using linear modeling and empirical Bayes methods (Smyth 2004) as implemented in the package Linear Models for Microarray Analysis (Smyth 2005). The linear model includes the effects of breeding regime (fast and slow inbred and noninbred) and temperature. Specific treatment contrasts were tested on the basis of a robust t statistic in which the standard errors of the estimated log fold changes have been moderated across genes, i.e., shrunk toward a common value. As zero and one genes were significantly differentially expressed between the fast- and slow-inbred lines at 25° and 36°, respectively (see supplemental Table 1 at http://www.genetics.org/supplemental/), the two inbred regimes were pooled. Multiple testing was accounted for by controlling the false-discovery rate at 5% (Benjamini and Hochberg 1995). For each gene, the contrast noninbred vs. inbred was tested for differentially expressed gene transcripts at both temperatures. Furthermore, the contrasts noninbred 25° vs. noninbred 36° and inbred 25° vs. inbred 36° were tested for differentially expressed gene transcripts. Genes with a P-value <0.01 were considered differentially expressed. Genes were categorized on the basis of the Biological Process as defined by the GO (Gene Ontology Consortium 2001). To identify overrepresented GO categories in the list of differentially expressed genes we used a chi-square test as implemented in the Expression Analysis Systematic Explorer (EASE) on-line application (http://david.niaid.nih.gov/david) (Hosack et al. 2003).
The Affymetrix array (Drosophila Genome Array Version 1) contained 13,966 probe sets representing ∼13,000 unique genes. To exclude genes that could not be confidently detected in the data analysis, probe sets with less than three present (P) calls within at least one of the three treatments were excluded (a transcript product must be present on at least three chips within either the noninbred or the inbred treatments). The filtered gene set contained 9572 transcripts.
A Kolmogorov–Smirnov test (Conover 1971) was used to determine if the distribution of the within-gene variances in gene expression levels differed significantly between the two breeding treatments.
RESULTS AND DISCUSSION
To screen for functional groups of genes that were differentially expressed in response to heat stress and/or inbreeding, we undertook an analysis of transcript expression in male D. melanogaster. The results show that inbreeding and heat stress profoundly affect gene expression and that the two distinct types of stresses interact in their effects. The distinct transcript responses to heat will not be discussed in detail here. Instead we will focus on the effects of inbreeding and interactions between inbreeding and temperature stress on gene expression patterns.
Differential gene expression in inbred and noninbred populations at benign and stressful temperatures:
At nonstressful temperatures, 16 functional groups of genes were differentially expressed in inbred compared with noninbred lines on the basis of information from the KEGG pathway database (Table 1; supplemental Figure 1 at http://www.genetics.org/supplemental/), whereas no functional groups of genes were differentially expressed on the basis of the EASE score criteria grouping genes based on the biological process GO annotation information. With the conservative false-discovery rate utilized here (controlled at the 5% level) only 12 genes were differentially expressed in inbred compared with outbred lines at nonstressful temperatures (Table 2; supplemental Table 2 at http://www.genetics.org/supplemental/). These genes are Ady43A, VhaSFD, CG5966, CG15203, Rtnl1, CG6415, CG9080, CG14481, CG9619, CG8147, l(3)IX-14, and CG11089, which are all involved in metabolic processes in the cell.
TABLE 1.
KEGG pathways associated with breeding regime at benign (25°) and stressful (36°) temperatures
| Pathway ID | Pathway name | n | BH |
|---|---|---|---|
| NI 25° vs. I 25° | |||
| DME00251 | GLUTAMATE METABOLISM | 22 | 0.025 |
| DME00310 | LYSINE DEGRADATION | 24 | 0.025 |
| DME00380 | TRYPTOPHAN METABOLISM | 25 | 0.025 |
| DME00910 | NITROGEN METABOLISM | 18 | 0.025 |
| DME00230 | PURINE METABOLISM | 69 | 0.025 |
| DME00632 | BENZOATE DEGRADATION VIA COA LIGATION | 28 | 0.025 |
| DME00252 | ALANINE AND ASPARTATE METABOLISM | 17 | 0.025 |
| DME00650 | BUTANOATE METABOLISM | 23 | 0.025 |
| DME00630 | GLYOXYLATE AND DICARBOXYLATE METABOLISM | 12 | 0.031 |
| DME00561 | GLYCEROLIPID METABOLISM | 57 | 0.034 |
| DME00340 | HISTIDINE METABOLISM | 11 | 0.037 |
| DME00710 | CARBON FIXATION | 17 | 0.037 |
| DME00260 | GLYCINE | 22 | 0.045 |
| DME03050 | PROTEASOME | 42 | 0.045 |
| DME00071 | FATTY ACID METABOLISM | 26 | 0.046 |
| DME00051 | FRUCTOSE AND MANNOSE METABOLISM | 20 | 0.046 |
| NI 36° vs. I 36° | |||
| DME00230 | PURINE METABOLISM | 69 | 0 |
| DME00251 | GLUTAMATE METABOLISM | 22 | 0.008 |
| DME00310 | LYSINE DEGRADATION | 24 | 0.008 |
| DME00380 | TRYPTOPHAN METABOLISM | 25 | 0.008 |
| DME00051 | FRUCTOSE AND MANNOSE METABOLISM | 20 | 0.008 |
| DME00650 | BUTANOATE METABOLISM | 23 | 0.008 |
| DME00632 | BENZOATE DEGRADATION VIA COA LIGATION | 28 | 0.008 |
| DME00910 | NITROGEN METABOLISM | 18 | 0.008 |
| DME00710 | CARBON FIXATION | 17 | 0.008 |
| DME04070 | PHOSPHATIDYLINOSITOL SIGNALING SYSTEM | 24 | 0.009 |
| DME00620 | PYRUVATE METABOLISM | 30 | 0.009 |
| DME00562 | INOSITOL PHOSPHATE METABOLISM | 30 | 0.009 |
| DME00350 | TYROSINE METABOLISM | 19 | 0.009 |
| DME00630 | GLYOXYLATE AND DICARBOXYLATE METABOLISM | 12 | 0.009 |
| DME00260 | GLYCINE | 22 | 0.014 |
| DME00071 | FATTY ACID METABOLISM | 26 | 0.014 |
| DME00020 | Citrate cycle (TCA cycle) | 34 | 0.017 |
| DME00280 | VALINE | 27 | 0.018 |
| DME00561 | GLYCEROLIPID METABOLISM | 57 | 0.018 |
| DME00330 | ARGININE AND PROLINE METABOLISM | 26 | 0.018 |
| DME00450 | SELENOAMINO ACID METABOLISM | 11 | 0.018 |
| DME00010 | GLYCOLYSIS/GLUCONEOGENESIS | 38 | 0.018 |
| DME00720 | REDUCTIVE CARBOXYLATE CYCLE (CO2 FIXATION) | 14 | 0.021 |
| DME00252 | ALANINE AND ASPARTATE METABOLISM | 17 | 0.022 |
| DME00120 | BILE ACID BIOSYNTHESIS | 13 | 0.023 |
| DME00640 | PROPANOATE METABOLISM | 20 | 0.025 |
Groups of functionally related genes were tested for association to breeding regime using the global test. P-values were obtained from permutations and adjusted for multiple testing. Only significant (p < 0.05) groups are presented.
BH, Benjamini–Hochberg adjusted p-value; I, inbred; n, number of genes in the pathway; NI, noninbred.
TABLE 2.
Number of up- and downregulated genes in contrasts between noninbred and inbred lines at benign and stressful temperatures
| NI 25° vs. NI 36° | I 25° vs. I 36° | NI 25° vs. I 25° | NI 36° vs. I 36° | |
|---|---|---|---|---|
| NI 25° vs. NI 36° | 2326 | 998 | 7 | 1 |
| I 25° vs. I 36° | 783 | 3171 | 4 | 4 |
| NI 25° vs. I 25° | 3 | 2 | 12 | 0 |
| NI 36° vs. I 36° | 1 | 12 | 3 | 176 |
Number of up- and downregulated genes in the contrasts noninbred (NI) 25° vs. NI 36°, inbred (I) 25° vs. I 36°, NI 25° vs. I 25°, and NI 36° vs. I 36° is shown on the diagonal (underlined). The number of up- and downregulated overlapping genes between contrasts is shown in above and below the diagonal, respectively.
After exposure to 36° for 1 hr, 27 and 38 functional groups of genes were differentially expressed on the basis of KEGG information and the EASE score criteria, respectively (Tables 1 and 3; supplemental Figure 2 at http://www.genetics.org/supplemental/) and 176 individual genes were differentially expressed in inbred lines compared with outbred lines (Table 2; supplemental Table 2 at http://www.genetics.org/supplemental/). The higher number of individual genes being differentially expressed in inbred compared with outbred lines under stressful environmental conditions is a clear documentation of an interaction between the environment, investigated by exposure to different temperature regimes, and the genotype, represented by noninbred and inbred populations, at the transcriptional level. Thus we have identified candidate loci that are potentially important for explaining conditionally expressed inbreeding depression as suggested by several authors (Bijlsma et al. 1999; Keller and Waller 2002; Vermeulen and Bijlsma 2004a,b). The 176 genes that were differentially expressed between the noninbred and inbred populations after exposure to 36° are candidates for explaining inbreeding by environmental interactions at the phenotypic level and should be investigated further, e.g., by knockout and qPCR studies.
TABLE 3.
Number of genes within a given functional category, percentage of total number of genes differentially expressed within that category, and EASE scores for groups of genes being differentially expressed within the biological process GO system for the contrast noninbred 36° vs. inbred 36°
| Functional category | No. of genes | % | EASE scores |
|---|---|---|---|
| Carboxylic acid metabolism | 22 | 12.9 | 2.40E-9 |
| Organic acid metabolism | 22 | 12.9 | 2.40E-9 |
| Aromatic compound metabolism | 12 | 7.1 | 1.70E-6 |
| Amino acid metabolism | 15 | 8.8 | 1.74E-6 |
| Amino acid and derivative metabolism | 15 | 8.8 | 6.10E-6 |
| Amine metabolism | 15 | 8.8 | 8.93E-6 |
| Amino acid catabolism | 7 | 4.1 | 3.62E-5 |
| Amine catabolism | 7 | 4.1 | 5.07E-5 |
| Purine base metabolism | 7 | 4.1 | 6.27E-5 |
| Nucleobase metabolism | 7 | 4.1 | 4.71E-4 |
| Aromatic amino acid family metabolism | 4 | 2.4 | 1.18E-3 |
| Heterocycle metabolism | 8 | 4.7 | 1.19E-3 |
| Imp biosynthesis | 3 | 1.8 | 2.31E-3 |
| Imp metabolism | 3 | 1.8 | 2.31E-3 |
| Pyruvate metabolism | 3 | 1.8 | 5.40E-3 |
| Lipid metabolism | 14 | 8.2 | 7.62E-3 |
| Calcium-mediated signaling | 5 | 2.9 | 8.17E-3 |
| Purine nucleotide biosynthesis | 5 | 2.9 | 8.17E-3 |
| Fatty acid metabolism | 6 | 3.5 | 8.63E-3 |
| Purine nucleotide metabolism | 5 | 2.9 | 8.64E-3 |
| Amino acid biosynthesis | 5 | 2.9 | 0.0101 |
| Carbohydrate metabolism | 14 | 8.2 | 0.0102 |
| Amine biosynthesis | 5 | 2.9 | 0.0136 |
| Purine nucleoside monophosphate biosynthesis | 3 | 1.8 | 0.0150 |
| Purine nucleoside monophosphate metabolism | 3 | 1.8 | 0.0150 |
| Purine ribonucleoside monophosphate biosynthesis | 3 | 1.8 | 0.0150 |
| Purine ribonucleoside monophosphate metabolism | 3 | 1.8 | 0.0150 |
| Nucleoside monophosphate biosynthesis | 3 | 1.8 | 0.0191 |
| Nucleoside monophosphate metabolism | 3 | 1.8 | 0.0191 |
| Ribonucleoside monophosphate biosynthesis | 3 | 1.8 | 0.0191 |
| Ribonucleoside monophosphate metabolism | 3 | 1.8 | 0.0191 |
| Nucleotide biosynthesis | 5 | 2.9 | 0.0208 |
| Nucleotide metabolism | 6 | 3.5 | 0.0251 |
| Purine ribonucleotide biosynthesis | 4 | 2.4 | 0.0402 |
| Physiological process | 99 | 58.2 | 0.0408 |
| Purine ribonucleotide metabolism | 4 | 2.4 | 0.0419 |
| Ribonucleotide biosynthesis | 4 | 2.4 | 0.0436 |
| Ribonucleotide metabolism | 4 | 2.4 | 0.0454 |
The genes were grouped by category within the systems by the EASE application on the DAVID homepage (http://david.niaid.nih.gov/david/version2/index.htm). Categories with significant EASE scores (<0.05) are presented here. The test calculates the probability of detecting the actual detected number of genes in a category, by evaluating the proportion of genes in each gene list belonging to a category vs. the proportion of genes belonging to this category out of all known genes.
Major metabolic pathways involved in carbohydrate, energy, lipid, nucleotide, amino acid, and glycan metabolism are affected by inbreeding, and this is accentuated under temperature stress (Tables 1 and 3; supplemental Figures 1 and 2 at http://www.genetics.org/supplemental/). When sorting the KEGG pathways according to their P-values we observed a large overlap in the high-ranking pathways associated with inbreeding under both benign and stressful environmental conditions (Table 1). This is a clear indication that inbreeding affects the network of chemical reactions that are essential in forming ATP, NADPH, and building blocks for biosyntheses. Myrand et al. (2002) and Parsons (2004, 2005) have shown and discussed that higher resistance to environmental stress in heterozygous individuals can be explained by their lower basic metabolic needs, leaving more energy available for resisting stressful conditions. In this study we found that a higher number of functional groups and single genes were differentially expressed in inbred lines compared with outbred lines after exposure to 36°. Most of these genes are related to metabolic pathways (Tables 1 and 3; supplemental Figure 2 at http://www.genetics.org/supplemental/). We also found that the majority of genes being differentially expressed were upregulated in inbred lines compared with noninbred lines under temperature stress (supplemental Figure 2 and Table 2 at http://www.genetics.org/supplemental/). This supports the hypothesis that the energy requirements are higher in homozygous individuals exposed to heat stress and is indirect support for the hypothesis that better performance of heterozygous individuals partly derives from their reduced energy expenditure for maintenance metabolism (Koehn and Bayne 1989; Myrand et al. 2002; Parsons 2004, 2005). In further support of this interpretation are results showing that the inbred lines investigated here are less heat resistant and have lower productivity (estimated as eggs laid per unit time) compared to the noninbred lines (Pedersen et al. 2005). In summary, the gene expression data, the observed inbreeding depression (Pedersen et al. 2005), and the observations of reduced metabolic efficiency in homozygous individuals (Myrand et al. 2002) support the hypothesis that a lower metabolic efficiency in more homozygous individuals plays a key role in explaining inbreeding depression and interactions between inbreeding and the environment.
An alternative interpretation of the result is that differential expression (mainly an upregulation) of metabolism genes in inbred populations is a direct consequence of expression of mutant deleterious alleles at loci within distinct metabolic pathways leading to metabolic disorders. Opposing this interpretation, however, is first that drift would cause expression of different deleterious alleles in different populations. Given the relatively high number of biologically independent replicated inbred lines investigated in this study (10 inbred lines), the general pattern observed would then not be expected. Second, given the high complexity of chain reactions in a given metabolic pathway, even an important difference in efficiency at one enzymatic reaction level should be hardly detectable at the level of the whole genotype (Kacser and Burns 1981). Third, due to dominance the expression of most deleterious mutations would lead to a decreased instead of an increased expression of genes involved in metabolism (Kacser and Burns 1981). Associated with this latter interpretation (expression of mutant deleterious alleles in inbred lines) of the data, the observed interactions between inbreeding and temperature stress at the transcript level can also be explained by buffering mechanisms being compromised upon exposure to temperature stress and more so in inbred compared to control lines. Such decanalization could lead to the expression of cryptic genetic variation in the inbred lines. The evolutionary importance of this phenomenon has recently received much attention (Rutherford 2003; Gibson and Dworkin 2004; Hoffmann and McKenzie 2005). Our results may partly be explained in a decanalization framework. The observation that the within-gene variances increase across lines in the inbred lines (Figure 1) could be interpreted in favor hereof (as opposed to a directional change of gene expression in the inbred lines favored by the authors of this article). However, if decanalization was of major importance in relation to explaining the observed results we would expect that genes with high within-gene variance across lines would be overrepresented among the genes being differentially expressed with inbreeding, and this did not seem to be the case (results not shown). However, further studies should be devoted to investigating the underlying mechanisms behind the observed results.
Transcript responses to heat in noninbred and inbred populations:
Exposure to short-term temperature stress (1 hr at 36°) led to differential expression of 59 functional groups of genes on the basis of information from the KEGG database (supplemental Figure 3 and Table 3 at http://www.genetics.org/supplemental/) within the noninbred lines. Functional groups defined by Biological Process that were strongly responding to heat stress included: response to heat (GO:9408), protein complex assembly (GO:6461), protein folding (GO:6457), and response to stress (GO:6950). A total of 2326 individual genes (Table 2) were differentially expressed in the noninbred populations after exposure to 36°. The genes being differentially expressed after heat exposure are to a large extent the same genes as those detected by Sørensen et al. (2005). Sørensen et al. (2005) found 770 genes to be differentially expressed in D. melanogaster females exposed to mild heat stress for 1 hr. Of these genes 313 were also differentially expressed in our study and the majority (303) of these genes was changed in the same direction in both studies. In our study a larger number of genes were differentially expressed (2326), which may be due to differences in (1) the filtering process of genes and (2) the statistical methods used to assess differential expression in the two studies. The functional groups and the individual genes being upregulated in response to heat in our study represent well-known stress response genes such as molecular chaperones and genes involved in metabolism. Many groups of genes being downregulated after exposure to 36° are related to metabolism and it seems that most metabolic processes are downregulated in response to heat stress (supplemental Figure 3 at http://www.genetics.org/supplemental/).
Exposure to short-term temperature stress led to differential expression of 57 functional groups (KEGG annotation) and 3171 individual genes (Table 2; supplemental Table 3 and Figure 4 at http://www.genetics.org/supplemental/) within the inbred populations. There is a huge overlap between the genes being differentially expressed after exposure to heat in the noninbred populations and in the inbred populations (Table 2). Among the genes that are nonoverlapping but differentially expressed after heat exposure within either the inbred or noninbred populations more than one-third (66 of 176) are differentially expressed within the contrast noninbred 36° vs. inbred 36°. A larger number of genes were differentially expressed at 36° within inbred compared to noninbred populations at this temperature, and a large proportion of the nonoverlapping genes in the comparisons noninbred 25° vs. noninbred 36° and inbred 25° vs. inbred 36° were differentially expressed in the comparison noninbred 36° vs. inbred 36°. This indicates that there is a strong interaction between temperature stress and inbreeding in their effects on gene expression patterns and that we have identified the gene loci responsible for the interaction.
Temperature and inbreeding increases between line variance in gene expression:
Within both environments the within-gene variance in expression levels in the inbred lines was higher than that within the noninbred lines (noninbred 25° vs. inbred 25°: D = 0.9969, P < 2.2e-16; noninbred 36° vs. inbred 36°: D = 0.9972, P < 2.2e-16; see Figure 1). Genetic differentiation between lines is expected to increase with inbreeding (Falconer and Mackay 1996). This is one reason that higher variance in phenotypes between inbred compared to noninbred lines is commonly observed (Falconer and Mackay 1996). Our results showing higher within-gene variance in gene expression across replicate inbred lines compared to across control lines indicate that there may be a strong correlation between variance at transcript level and variance in phenotype (Figure 1).
Figure 1.—
Distribution plots of standard deviations for within-gene variances in gene expression levels for the noninbred and inbred lines at the two temperatures investigated.
It is also interesting to observe that exposure to temperature stress caused an increase in the within-gene variance in gene expression levels for both noninbred and inbred lines (noninbred 25° vs. noninbred 36°: D = 0.997, P < 2.2e-16; inbred 25° vs. inbred 36°: D = 0.9897, P < 2.2e-16; see Figure 1). Assuming a high correlation between gene expression levels and phenotype this environmental impact on the between-line variance is of interest from an evolutionary point of view as it emphasizes that population differentiation may occur at a faster rate under harsh environmental conditions independent of drift and inbreeding.
Conclusion:
In this study we found that the expression of heat-shock protein and metabolism genes is strongly affected by temperature stress exposure and that primarily genes involved in metabolism are affected by inbreeding. We also found that the effect of temperature stress on gene expression is accentuated in inbred compared to noninbred lines. On the basis of these results we conclude that inbreeding and environmental stress both independently and synergistically affect gene expression patterns and propose that the genes found to be differentially expressed in this article are candidate genes involved in coping with environmental and genetic stress. They should be investigated in more detail in future studies to gain knowledge of the specific role of these genes in different genetic backgrounds and under different environmental conditions.
Acknowledgments
We thank Rodrigo Laboriau, Morten Muhlig Nielsen, and Jesper Givskov Sørensen for fruitful discussions on various parts of this experiment; Doth Andersen and Bente Devantié for excellent technical assistance; Stuart Barker, Greg Gibson, Lawrence Harshman, Ary Hoffmann, Mick Madsen, and two anonymous reviewers for critical comments on earlier versions of the manuscript; and the Danish Natural Sciences Research Council (V.L.) and the Danish Agricultural and Veterinary Research Council (T.N.K. and P.S.) for financial support.
References
- Barker, J. S. F, and R. N. Podger, 1970. Interspecific competition between Drosophila melanogaster and Drosophila simulans—effects of larval density on viability, developmental period and adult body weight. Ecology 51: 170–189. [Google Scholar]
- Benjamini, Y., and Y. Hochberg, 1995. Controlling the false discovery rate—a practical and powerfull approach to multiple testing. J. R. Stat. Soc. Ser. B Stat. Methodol. 57: 289–300. [Google Scholar]
- Bijlsma, R., and V. Loeschcke, 2005. Environmental stress, adaptation and evolution: an overview. J. Evol. Biol. 18: 744–749. [DOI] [PubMed] [Google Scholar]
- Bijlsma, R., J. Bundgaard and W. F. van Putten, 1999. Environmental dependence of inbreeding depression and purging in Drosophila melanogaster. J. Evol. Biol. 12: 1125–1137. [Google Scholar]
- Bubliy, O. A., and V. Loeschcke, 2005. Correlated responses to selection for stress resistance and longevity in a laboratory population of Drosophila melanogaster. J. Evol. Biol. 18: 789–803. [DOI] [PubMed] [Google Scholar]
- Conover, W. J., 1971. Practical Nonparametric Statistics. John Wiley & Sons, New York.
- Falconer, D. S., and T. F. C. Mackay, 1996. Introduction to Quantitative Genetics, Ed. 4. Longman, Harlow, UK.
- Gene Ontology Consortium, 2001. Creating the gene ontology resource: design and implementation. Genome Res. 11: 1425–1433. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gibson, G., and I. Dworkin, 2004. Uncovering cryptic genetic variation. Nat. Rev. Genet. 5: 681–690. [DOI] [PubMed] [Google Scholar]
- Goeman, J. J., S. A. van de Geer, F. de Kort and H. C. van Houwelingen, 2004. A global test for groups of genes: testing association with a clinical outcome. Bioinformatics 20: 93–99. [DOI] [PubMed] [Google Scholar]
- Hoffmann, A. A, and J. A. McKenzie, 2005. Capacitators, stressors, phenotypic variability and evolutionary change, pp. 159–189 in Variation: A Central Concept in Biology, edited by B. Hallgrímsson, and B. Hall. Elsevier Academic Press, Burlington, MA.
- Holloway, A. J., R. K. van Laar, R. W. Tothill and D. D. L. Bowtell, 2002. Options available—from start to finish—for obtaining data from DNA microarrays. Nat. Genet. 32: 481–489. [DOI] [PubMed] [Google Scholar]
- Hosack, D. A., G. Dennis, B. T. Sherman, H. C. Lane and R. A. Lempicki, 2003. Identifying biological themes within lists of genes with EASE. Genome Biol. 4: 4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Irizarry, R. A., B. Hobbs, F. Collin, Y. D. Beazer-Barclay and K. J. Antonellis, 2003. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 4: 249–264. [DOI] [PubMed] [Google Scholar]
- Jiménez, J. A., K. A. Hughes, G. Alaks, L. Graham and R. C. Lacy, 1994. An experimental study of inbreeding depression in a natural habitat. Science 266: 271–273. [DOI] [PubMed] [Google Scholar]
- Kacser, H., and J. A. Burns, 1981. The molecular basis of dominance. Genetics 97: 639–666. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kanehisa, M., and S. Goto, 2000. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic Acids Res. 28: 27–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Keller, L. F., and D. M. Waller, 2002. Inbreeding effects in wild populations. Trends Ecol. Evol. 17: 230–241. [Google Scholar]
- Koehn, R. K., and B. L. Bayne, 1989. Towards a physiological and genetic understanding of the stress response. Biol. J. Linn. Soc. 37: 157–171. [Google Scholar]
- Kristensen, T. N., J. Dahlgaard and V. Loeschcke, 2002. Inbreeding affects Hsp70 expression in two species of Drosophila even at benign temperatures. Evol. Ecol. Res. 4: 1209–1216. [Google Scholar]
- Kristensen, T. N., P. Sørensen, M. Kruhøffer, K. S. Pedersen and V. Loeschcke, 2005. Genome wide analysis of inbreeding effects on gene expression in Drosophila melanogaster. Genetics 171: 157–167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Landis, G. N., D. Abdueva, D. Skvortsov, J. Yang, B. E. Rabin et al., 2004. Similar gene expression patterns characterize aging and oxidative stress in Drosophila melanogaster. Proc. Natl. Acad. Sci. USA 101: 7663–7668. [DOI] [PMC free article] [PubMed] [Google Scholar]
- MacBeath, G., 2002. Protein microarrays and proteomics. Nat. Genet. 32: 526–532. [DOI] [PubMed] [Google Scholar]
- Myrand, B., R. Tremblay and J.-M. Sévigny, 2002. Selection against blue mussels (Mytilus edulis L) homozygotes under various stressful conditions. J. Hered. 93: 238–248. [DOI] [PubMed] [Google Scholar]
- Parsons, P. A., 2004. From energy efficiency under stress to rapid development and a long life in natural populations. Biogerontology 5: 201–210. [DOI] [PubMed] [Google Scholar]
- Parsons, P. A., 2005. Environments and evolution: interactions between stress, resource inadequacy and energetic efficiency. Biol. Rev. 80: 589–610. [DOI] [PubMed] [Google Scholar]
- Pedersen, K. S., T. N. Kristensen and V. Loeschcke, 2005. Effects of inbreeding and rate of inbreeding in Drosophila melanogaster – Hsp70 expression and fitness. J. Evol. Biol. 18: 756–762. [DOI] [PubMed] [Google Scholar]
- Pletcher, S. D., S. J. Macdonald, R. Margurerie, U. Certa, S. C. Stearns et al., 2002. Genome-wide transcript profiles in aging and calorically restricted Drosophila melanogaster. Curr. Biol. 12: 712–723. [DOI] [PubMed] [Google Scholar]
- Reed, D. H., D. A. Briscoe and R. Frankham, 2002. Inbreeding and extinction: The effect of environmental stress and lineage. Conserv. Genet. 3: 301–307. [Google Scholar]
- Rutherford, S. L., 2003. Between genotype and phenotype: Protein chaperones and evolvability. Nat. Rev. Genet. 4: 263–274. [DOI] [PubMed] [Google Scholar]
- Smyth, G. K., 2004. Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Stat. Appl. Genet. Mol. Biol. 3: Article 3. [DOI] [PubMed]
- Smyth, G. K., 2005. Limma: linear models for microarray data, pp. 397–420 in Bioinformatics and Computational Biology Solutions Using R and Bioconductor, edited by R. Gentleman, V. Carey, S. Dudoit, R. Irizarry, and W. Huber. Springer, New York.
- Sørensen, J. G., T. N. Kristensen and V. Loeschcke, 2003. The evolutionary and ecological role of heat shock proteins. Ecol. Lett. 6: 1025–1037. [Google Scholar]
- Sørensen, J. G., M. M. Nielsen, M. Kruhøffer, J. Justesen and V. Loeschcke, 2005. Full genome gene expression analysis of the heat stress response in Drosophila melanogaster. Cell Stress Chaperon. 10: 144–160. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vermeulen, C. J., and R. Bijlsma, 2004. a Characterization of conditionally expressed mutants affecting age-specific survival in inbred lines of Drosophila melanogaster: lethal conditions and temperature sensitive periods. Genetics 167: 1241–1248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vermeulen, C. J., and R. Bijlsma, 2004. b Changes in mortality patterns and temperature dependency of lifespan in Drosophila melanogaster caused by inbreeding. Heredity 92: 275–281. [DOI] [PubMed] [Google Scholar]
- Wu, Z. J., and R. A. Irizarry, 2004. Preprocessing of oligonucleotide array data. Nat. Biotechnol. 22: 656–658. [DOI] [PubMed] [Google Scholar]
- Wu, Z. J., R. A. Irizarry, R. Gentleman, F. M. Murillo and F. Spencer, 2004. A model based background adjustment for oligonucleotide expression arrays. J. Am. Stat. Assoc. 99: 909–917. [Google Scholar]