Journal of Heredity Advance Access originally published online on June 15, 2005
Journal of Heredity 2005 96(5):513-521; doi:10.1093/jhered/esi071
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Age-Specific Changes in Epistatic Effects on Mortality Rate in Drosophila melanogaster
From the Department of Genetics, Life Sciences, University of Georgia, Athens, GA 30602-7223. C. Spencer is currently at the Department of Zoology, University of British Columbia, 6270 University Blvd., Vancouver, BC V6T 1Z4, Canada
Address correspondence to Christine C. Spencer at the address above, or e-mail: spencer{at}zoology.ubc.ca.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Models for the evolution of senescence assume that genes with age-specific effects act independently of one another. Although recent empirical data show that longevity is influenced in part by interactions between genes, there are currently few data on whether epistasis influences age-specific components of mortality. To gauge if and how interactions affect age-specific traits, we incorporated the Drosophila visible marker mutations ebony, forked, and purple into seven wild-caught strains of D. melanogaster to examine gene x genetic background interactions. We found significant natural genetic variation for longevity and baseline mortality rates. Gene x genetic background interactions were prevalent not only for longevity but also for baseline mortality rates and age-specific mortality rates. We conclude that gene x genetic background epistasis is prevalent for aging-related traits and could play a significant role in the evolution of aging. These results suggest that future genetic models for the evolution of aging should incorporate the effects of epistasis.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Evolutionary studies of aging have largely focused on attempts to test predictions for the two predominant hypotheses for how senescence evolves. The first of these models, known as mutation accumulation (MA), was developed in 1946 by Sir Peter Medawar (1946, 1952). Medawar began with the assumption that some subset of genes exist that have age-specific effects on fitness. Any heritable, deleterious mutation that occurs in an early acting gene, such as a gene that regulates development, would be removed by natural selection. Heritable, deleterious mutations whose effects are confined to late ages would not be subject to these same selection pressures because their effects would only be revealed to natural selection after reproductive maturity, when the strength of selection begins to decline (Hamilton 1966). In the related antagonistic pleiotropy (AP) model, Williams (1957) proposed that alleles with late-acting genetic effects that lead to senescence would be favored by natural selection if they have early acting beneficial effects.
Theoreticians have developed mathematical models that provide explicit quantitative genetic predictions for the MA and AP theories of senescence (Charlesworth 1990, 2001; Charlesworth and Hughes 1996; Hamilton 1966; Rose 1985; Williams and Day 2003). However, these theories rest on a number of critical assumptions. First, both assume the existence of some subset of genes that acts only at certain ages or has pleiotropic actions specific to different ages. Only recently have studies looked for genes with age-specific effects. Two large-scale demographic studies have found evidence for age-specific effects of novel germ-line mutations on mortality rates in Drosophila melanogaster (Pletcher et al. 1998; Yampolsky et al. 2001). In addition to these demographic analyses, several researchers have used microarrays to compare gene expression levels at different ages. Pletcher et al. (2002) screened the D. melanogaster genome at 6–8 adult ages for differences in levels of gene expression between short-lived control flies and long-lived, calorie-restricted flies. Drosophila differentially expressed a minimum of 1,203 genes at various ages, although Rose and Long (2002) caution that this may be an overestimate. In a similar approach in Caenorhabditis elegans, Golden and Melov (2004) identified several stress-related genes that exhibit differential age-specific expression, whereas Landis and colleagues (2004) reported a small but consistent down-regulation in energy metabolism genes in adult Drosophila from early to late ages. In both D. melanogaster and C. elegans, gene expression in early adulthood is significantly higher than at later ages (McCarroll et al. 2004).
Although the aforementioned studies have addressed whether some subset of genes acts in an age-specific manner, a second and less recognized assumption also underlies MA and AP theories of senescence—that of additive gene action. Mathematical derivations of MA and AP typically assume that survival rates are influenced by a large number of genes with small, additive effects. They do not consider that individual genes can have major effects (e.g., Clancy et al. 2001; Lin et al. 1998; Parkes et al. 1998; Rogina et al. 2000; Tatar et al. 2001) and might interact with one another in a nonadditive fashion. In fact, the few studies that have looked explicitly for evidence of epistatic (nonadditive) effects on survival have detected genetic interactions. In quantitative trait locus (QTL) analyses, work on both flies (Leips and Mackay 2000) and worms (Shook et al. 1996) has found interactions between life span QTL. In a design similar to the one described here, Spencer et al. (2003) found significant gene x genotype interactions in the ability of superoxide dismutase overexpression to extend life span in Drosophila.
These studies provide evidence that genetic interactions influence life expectancy. However, life expectancy is a composite measure, integrating age-specific mortality rates over the life span of a cohort. As such, it does not tell us how or when in life epistatic interactions might influence mortality rates. Previous studies have not considered the possibility that epistatic interactions may influence mortality rates at some ages but not others.
With this in mind, we used D. melanogaster as a model system to examine the effects of interactions between three visible mutants and seven genetic backgrounds on age-specific mortality rates. This allowed us to test not only the hypothesis that nonadditive gene interactions influence mortality rates but also whether epistasis for mortality rates changes over the life span of a cohort. If a substantial portion of genetic variation for mortality rates results from epistatic interactions, then these interactions may shape the evolution of senescence in natural populations.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Flies and Crossing Design
In August 2000, we collected nonvirgin female D. melanogaster from a peach orchard in Watkinsville, Georgia. These females were used to establish isofemale lines that were subsequently inbred by full-sib mating for 14 generations, until the lines were on average 96% homozygous. These strains were the wild-caught, inbred controls for all subsequent strains. We crossed nine of these wild-caught inbred lines to three different Drosophila visible marker stocks carrying recessive visible mutations. The marker stocks were obtained from the Bloomington stock center and carried either ebony, forked, or purple, markers that visibly affect body color, bristle shape, and eye color, respectively. Though the fitness effects of the markers are largely unknown, they are not reported to reduce fecundity or male mating ability (Lindsley and Grell 1967). The forked (f) mutation is located on the X chromosome, and purple (pr) and ebony (e) are located on the second and third chromosomes, respectively.
Female offspring of the cross between the wild-caught inbred and the marker stocks were backcrossed to males from the inbred parental line for 10 generations. Every second generation during backcrossing, we mated the first filial offspring of the parental backcross to recover the recessive phenotype. After the mutations were 86–90% introgressed into the wild-caught inbred lines (Falconer and Mackay 1996), we established single mutation stocks in seven genetic backgrounds. Genes that are in tight linkage are unlikely to recombine after 10 backcrosses. On average, we expect 20 cM on either side of the focal mutation to be heterozygous for the inbred and mutant-line genotypes. Accordingly, we refer to these mutations as "mutation regions."
Crosses were conducted in 250-ml bottles, and stocks were maintained in 8 dram (25 x 95 mm) glass vials with 5 ml standard medium of cornmeal, agar, molasses, and yeast, with propionic acid added to prevent bacterial growth (Ashburner 1989). Flies were kept at 26°C while the crosses were conducted, and stocks were maintained at 24°C on a 12:12 dark/light cycle. Flies were handled at room temperature using CO2 anesthesia.
Estimating Life Expectancy and Mortality Parameters
To estimate life expectancy and mortality parameters, we collected 600 virgin male and 600 virgin female offspring for each mutation x genetic background combination and for each control genetic background. We established single-sex mortality cages, described shortly, with 200 flies per cage and three replicate cages per sex for every mutation x genetic background combination. This procedure generated seven lines x four mutation classes (e, f, pr, and control) x two sexes x three replicates for a total of 168 cages and 33,600 flies.
We estimated life expectancy in 32-oz. (0.95 L) plastic mortality cages (described in Promislow and Bugbee 2000) provided with standard cornmeal-molasses fly medium (Ashburner 1989). Cages were held at 24°C on a 12:12 dark/light cycle and were removed to room temperature for less than 2 h every 2 days to provide fresh fly medium and remove and count all dead flies. To estimate life expectancy and mortality parameters, we used the cage as the unit of replication.
Statistical Analyses
Mortality rates in both natural and lab populations are well described over most of the life span by the Gompertz model (Gompertz 1825; Pletcher et al. 2000; Promislow et al. 1999),
 | (1) |
The mixed model (MIXED procedure in SAS version 8.2; SAS Institute 2001) allowed us to analyze both fixed and random variables more flexibly than a standard general linear model. The mixed procedure was used to determine the effects of mutation and sex (fixed effects) and genetic background (random effect) on the dependent variables of life expectancy, baseline mortality (a), and rate of senescence (b). For the random effect (genetic background) and interaction terms involving the random effect, we ran a series of analyses that involved adding random terms to the model sequentially and comparing the model with the added term (full model) to the same model less the focal term (reduced model). We used a likelihood ratio test with the restricted maximum likelihood estimate (–2 x REML) for each full to reduced pair of models to determine if the model with the additional term provided a significantly better fit to the data. If the restricted maximum likelihood score, which follows a chi-square distribution, was significantly different with the additional term, then that term contributed significantly to the fit of the data to the model. Degrees of freedom were equal to the difference in the number of terms between the models. Tables 1, 2, and 3 list the random terms in the order they were considered for inclusion in the model. As an example, to test if the inclusion of sex x genetic background term provided a significantly better fit for the data over the reduced model where mean lifespan = sex + mutation + sex x mutation + genetic background, we would run both models and calculate the difference in restricted maximum likelihoods between the two models. If the difference were greater than 
then sex interacted significantly with genetic background as a predictor of mean lifespan.
|
|
|
Significant mutation x genetic background interactions would indicate gene x genetic background epistasis. To determine if the effects of epistasis were age-dependent, we tested for significant mutation x genetic background effects on the slope of the Gompertz curve. Such effects would suggest that the epistatic effects of these visible mutations change with age.
To estimate the mean squares correctly under all mixed model analyses described here, we calculated the denominator degrees of freedom using a Satterthwaite approximation, which can result in fractional degrees of freedom (SAS Institute 2001). To meet the assumptions of the mixed model, we natural log transformed the intercept (a) and angularly transformed the slope (b). For each sex, we used estimate statements in PROC MIXED to test for differences between each mutation line and its control. An estimate statement provides estimates of the variance components associated with each random variable, which we then tested against the control variance in a paired t test. To obtain an experimentwise 
= 0.05, we corrected for multiple comparisons using the Dunn-
idák method (Sokal and Rohlf 1995) with k = 44 comparisons and = 0.0012. To establish if the effects of each mutation were correlated across the sexes, we obtained Pearson's correlation coefficients using SPSS (1999).
As an alternative to the parametric Gompertz model, which fits mortality rates across all ages, we also tested for mutation x genetic background effects on mortality rates at different ages across the life span of each cohort. In this case, we estimated age-specific mortality, ln(µx), at three different ages, as described below. The advantage of this approach is that unlike the Gompertz model, it does not assume that genetic interactions change linearly with age. The disadvantage of this approach is that sampling error changes with age (Shaw et al. 1999). Statistical power is low at very early ages, when few flies die due to low mortality rates, and at late ages, when few flies remain. To address the problem of changes in sampling error with age, we have analyzed the data in two ways.
- By Age: We divided the number dying, dx, for each replicate cage into three parts corresponding to 3/8, 6/8, and 9/8 of the mean life expectancy for each genotype (mutation region x sex x genetic background). We then calculated the probability of survival to age x, Px, at each age and estimated ln(µx) at three ages. For instance, for ebony females from genetic background 9, the life expectancy was 60 days, so ln(µx) "by age" was estimated at ages 22, 44, and 66 days.
- By Sample: We divided the total number of flies from each replicate cage into four equal groups and discarded the fourth group because ln(µx) cannot be calculated for the zero values found at late ages in dx data. We calculated a single Px for each group and used these to estimate age-specific mortality, ln(µx). For the sample genotype ebony 9 females, the average ages across three replicates were 52, 60, and 72 days. Age was treated as a repeated measure using the repeated command in SAS PROC MIXED, again using the Satterthwaite approximation to calculate denominator degrees of freedom.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Among seven wild-caught inbred control lines, life span showed significant natural genetic variation among the genetic backgrounds and in the interaction between sex and genetic background (Figure 1, Table 1). Baseline mortality (ln(a)) showed significant natural genetic variation for genetic background and for its interaction with sex (Figure 1, Table 1). The slope of the mortality rate did not differ significantly between the sexes or among the genetic backgrounds but showed a significant interaction between these factors (Figure 1, Table 1). What is visually striking in Figure 2 is the greater 20-fold range in intercept of the Gompertz curve. The slope and intercept from the Gompertz curve were significantly negatively correlated with one another among control lines (R = –0.709, n = 161, P < 0.0001).
|
|
When we added single mutations into these genetic backgrounds, life span varied significantly with mutation region and sex x mutation region interaction, as well as with genetic background and all of its interaction terms (Table 2). For each mutation, the average change in male life span was positively and significantly correlated with the change in life span in females (Figure 3).
|
After correcting for multiple comparisons, we found that each of our major mutations significantly altered longevity in some but not all genetic backgrounds (Table 4, Figure 4). Counter to our expectations, almost half of these 22 significant changes in life span were increases. Among the 14 purple and ebony strains that differed significantly from their controls, the direction of effects on life span was the same for both purple and ebony, which implies an effect of genetic background (Table 4). In contrast to this, when forked significantly altered life span, its effect was to increase life span regardless of genetic background (Table 4).
|
|
To determine how underlying mortality forces caused life span to differ, we examined the age-independent (intercept) and age-dependent (slope) mortality parameters. In the presence of the marker mutations, the intercept showed a significant effect of mutation region, genetic background, and the interaction of genetic background with mutation region and with sex, which is evidence of gene x genotype epistasis (Table 2). In the presence of major mutations, the baseline mortality rate increased. After Dunn-
idák correction for multiple comparisons, four female and seven male mutation strains had a significantly greater age-independent mortality than the control (Figure 5).
|
The results for the analysis of slope showed mutation- and sex-specific variation as well as a significant effect of genetic background and its interaction with mutation (Table 2). Significant interaction terms in this analysis represent age-specific gene x genotype interactions. The rate of senescence decreased in the presence of a major mutation, compared to the control strains. This decrease was significant for five female and four male genotypes. Because the slope did not increase significantly over the control, we conclude that these particular mutations did not cause an accelerated rate of senescence as the flies aged (Figure 6).
|
What, if any, age-specific effect did each marker mutation have on mortality rates? If mortality rates exhibit age-specific epistasis for these data, we should observe significant interaction effects that include the factor age. Regardless of whether the data were grouped by sample size or by age, mortality rates showed no significant fixed-effect interactions with age (Table 3). After we included genetic background in the analysis, genetic background x age interactions showed two patterns: When the data were ordered by age, interactions were largely nonexistent, with genetic background x age the only significant interaction. When the data were organized by sample (accounting for sampling error at early and late ages), then all interaction effects that included genetic background x age explained the data significantly better than the corresponding reduced model (Table 3). Regardless of whether we grouped the data with equal age structure or equivalent sample size of dx, we observed significant interaction effects that incorporate age, a signature of age-specific epistasis.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In this experiment, we sought to determine if genetic interactions influence life expectancy, Gompertz mortality parameters, and age-specific mortality rates. Specifically, we tested (1) whether genetic background modified the effect on life span of three different major mutations, (2) if mutations altered life span via age-independent changes to mortality rate, and (3) if there were significant changes in age-specific mortality rates, providing evidence of age-specific genetic interactions. Our data showed that all three mutation regions altered life span and baseline mortality rates. The effect of each mutation region on life span and mortality rates depended upon genetic background and sex. In the presence of a major mutation, life span, baseline mortality rate (or frailty) and the rate of aging all showed strong evidence for gene x genotype epistatic interactions. Furthermore, when we analyzed mortality rates at three different ages, we found age x genotype and age x mutation x genotype epistasis. Considered together, these results provide strong evidence for age-specific change in gene x genetic background interactions, or age-specific epistasis. Below we discuss the implications of our results for theory on the evolution of aging.
Gene x Genotype Interactions in Wild-Caught Strains
The results presented here are based on our analysis of the effects of visible mutants on females and males from each of seven inbred control strains. In the absence of the mutants, the control strains demonstrated natural variation in life span and mortality parameters from a single, wild-caught population. Among seven wild-caught inbred lines, we found that most of the variation in longevity was due to changes in the frailty parameter (ln(a) in equation 1) rather than in the rate of aging (b in equation 1). This and many previous studies (e.g., Bell 1984; Dudycha and Tessier 1999; Geiger-Thornsberry and Mackay 2004; Snoke and Promislow 2003; Vieira et al. 2000) have demonstrated the substantial amount of genetic variation for longevity in natural populations. What remains to be discerned is whether in addition to genetic variation for aging-related traits, natural populations also harbor variation for the effects of modifier alleles on major mutations or "mutation regions." We tested this by adding mutations to these genetically variable strains.
By adding a single marker mutation of major effect into these wild-caught genetic backgrounds, we were able to detect significant additional variation in the form of gene x genotype (genetic background) epistasis. Most of this variation for longevity was due to epistatic interactions on the rate of aging (the slope b in equation 1) rather than the baseline frailty parameter (a in equation 1). These results add to the growing collection of studies that emphasize how critical it is to incorporate multiple genetic backgrounds into experimental designs that examine polygenic traits. Gene x genetic background interaction effects have been observed in a diversity of species and for many traits such as development (Gibson and van Helden 1997, Gibson et al. 1999), olfaction (Fedorowicz et al. 1998), and life span (Spencer et al. 2003) in Drosophila, as well as for viability and fecundity in Arabidopsis (Ungerer et al. 2003) and for the overexpression of growth hormone in trout (Devlin et al. 2001).
In this study, the nature of the gene x genotype epistasis could result from interactions between unspecified loci in the genome and either the marker locus itself or loci tightly linked to the marker. Because the marker loci confer major morphological changes, the effects of the marker are probably largely responsible for the observed fitness changes.
Although our a priori expectation was that the visible mutants would either reduce survival or leave it unchanged, in some cases, mutant strains had higher survival than controls strains. Although this could be due to differences in the cost of reproduction (not tested in this study) or the effects of modifier loci in the inbred backgrounds, there is an alternate explanation for the variation in fitness changes we observe: heterosis. We did not continue to inbreed our strains after we introgressed the visible marker into our inbred lines. Therefore, each line may have a slightly different cassette of heterozygous, introgressed DNA from the stock strain. These introgressed regions should interact differently with the inbred strains, each of which has a unique combination of alleles that might include different modifiers for introgressed mutation regions. Thus, the degree of heterosis might be unique for each strain. Heterosis or hybrid vigor caused by these heterozygous regions might differ in magnitude across backgrounds and could cause fitness differences.
There are two arguments against the effects of heterosis in these data. First, our results are consistent with other gene x genetic background studies, which find that genes with hitherto predictable effects suddenly yield unpredictable phenotypes when combined into different genetic backgrounds (Gibson and van Helden 1997; Spencer et al. 2003). Second, if heterosis were a significant factor, we might expect most mutant lines to have increased life span. However, decreases in life span occurred just as frequently as increases (Table 4).
Epistasis in Models on the Evolution of Senescence
Current mathematical models of the evolution of aging predict that variance in fitness traits should change with age differently depending on whether senescence is due to MA or AP (Charlesworth and Hughes 1996). According to Charlesworth and Hughes's (1996) model, if the accumulation of late-acting mutations contributes to the evolution of senescence, then additive and dominance genetic variance and inbreeding depression should increase with age in fitness-related traits. Alternatively, if antagonistically pleiotropic loci contribute to the evolution of aging, only additive genetic variance should increase with age. Models for both hypotheses assume additive effects for age-specific mutations on longevity, but this assumption has never been tested.
Contrary to the assumption of additivity, our data provide strong evidence for nonadditive gene x genetic background interactions for life span and mortality rate parameters for three Drosophila visible mutations. Taken in conjunction with previous evidence of epistatic effects on life span (Leips and Mackay 2000; Shook et al. 1996; Spencer et al. 2003), it seems apparent that life span has a complex genetic architecture that includes epistatic interactions.
Of particular importance is the fact that epistatic variation for mortality rate changed with age. Hermisson et al. (2003) have used quantitative genetic models to show that under some circumstances, epistatic variance may replace additive variance for fitness traits. Our results suggest an additional perspective: For fitness traits at different ages, the amount of epistatic genetic variance, relative to additive variance, may shape the way that age-specific mortality and fertility evolve.
In this study we demonstrated gene x genetic background epistasis for life span and mortality rate parameters. The importance of nonlinear interactions in the genetics of aging is further supported by recent work by Promislow (2004). In a study of the protein–protein interaction network in the yeast Saccharomyces cerevisiae, Promislow (2004) showed that proteins associated with senescence have significantly greater numbers of interactions than one would expect by chance. This work, in conjunction with our data on the effects of gene x genetic background epistasis, adds to the sparse but growing evidence that gene x genetic background interactions have powerful and unpredictable phenotypic effects in the evolution of senescence and in a variety of other contexts, from the expression of transgenes that escape from domesticated species to wild relatives (e.g., Devlin et al. 2001; Muir and Howard 2001) to the epidemiologic or pathologic display of genetic disease in humans (reviewed in Hardy et al. 2003).
Age-specific genetic interactions have not been considered in evolutionary theories of aging. Our results, combined with evidence of gene x genetic background interactions of the aging gene superoxide dismutase (Spencer et al. 2003), indicate that genetic interactions are a strong component of the genetic architecture of aging. We now need to develop appropriate models to determine the extent to which epistatic interactions can influence the evolution and maintenance of senescence in natural populations.
|
Acknowledgments |
|---|
We thank Tammy Haselkorn, Christi Howell, Amber Wright, and a slew of undergraduate workers for assistance with fly pushing. Members of the Promislow, Whitlock, and SOWD lab groups offered discussion and thoughtful comments. India, Andy Peters, Mike Whitlock, and SAS Technical Support staff provided statistical advice. This work was funded in part by National Institute on Aging grants AG21298 and AG14027 (to D.E.L.P.), the Ellison Medical Foundation, and an American Federation of Aging Research grant (to C.C.S.).
|
Footnotes |
|---|
Corresponding Editor: Stephen Schaeffer
Received November 1, 2004
Accepted March 23, 2005
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-
Ashburner M, 1989. Drosophila: a laboratory handbook Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Bell G, 1984. Evolutionary and nonevolutionary theories of senescence. Am Nat 124:600–603.[CrossRef]
Charlesworth B, 1990. Optimization models, quantitative genetics, and mutation. Evolution 44:520–538.[CrossRef]
Charlesworth B, 2001. Patterns of age-specific means and genetic variances of mortality rates predicted by the mutation-accumulation theory of ageing. J Theor Biol 210:47–65.[Medline]
Charlesworth B, and Hughes KA, 1996. Age-specific inbreeding depression and components of genetic variance in relation to the evolution of senescence. Proc Natl Acad Sci USA 93:6140–6145.
Clancy DJ, Gems D, Harshman LG, Oldham S, Stocker H, Hafen E, Leevers SJ, and Partridge L, 2001. Extension of life-span by loss of CHICO, a Drosophila insulin receptor substrate protein. Science 292:104–106.
Devlin RH, Biagi CA, Yesaki TY, Smailus DE, and Byatt JC, 2001. Growth of domesticated transgenic fish: a growth-hormone transgene boosts the size of wild but not domesticated trout. Nature 409:781–782.[CrossRef][Medline]
Dudycha JL, and Tessier AJ, 1999. Natural genetic variation of life span, reproduction, and juvenile growth in Daphnia. Evolution 53:1744–1756.[CrossRef]
Falconer DS, and Mackay TFC, 1996. Introduction to quantitative genetics Longman, Essex.
Fedorowicz GM, Fry JD, Anholt RRH, and Mackay TFC, 1998. Epistatic interactions between smell-impaired loci in Drosophila melanogaster. Genetics 148:1885–1891.
Geiger-Thornsberry GL, and Mackay TFC, 2004. Quantitative trait loci affecting natural variation in Drosophila longevity. Mech Ageing Dev 125:179–189.[CrossRef][Medline]
Gibson G, and van Helden S, 1997. Is function of the Drosophila homeotic gene Ultrabithorax canalized?. Genetics 147:1155–1168.[Abstract]
Gibson G, Wemple M, and van Helden S, 1999. Potential variance affecting homeotic Ultrabithorax and Antennapedia phenotypes in Drosophila melanogaster. Genetics 151:1081–1091.
Golden TR, and Melov S, 2004. Microarray analysis of gene expression with age in individual nematodes. Aging Cell 3:111–124.[CrossRef][Medline]
Gompertz B, 1825. On the nature of the function expressive of the law of human mortality. Philo Trans 27:513–519.
Hamilton WD, 1966. The moulding of senescence by natural selection. J Theor Biol 12:12–45.[CrossRef][Web of Science][Medline]
Hardy J, Singleton A, and Gwinn-Hardy K, 2003. Ethnic differences and disease phenotypes. Science 300:739–740.
Hermisson J, Hansen TF, and Wagner GP, 2003. Epistasis in polygenic traits and the evolution of genetic architecture under stabilizing selection. Am Nat 161:708–734.[CrossRef][Medline]
Landis GN, Abdueva D, Skvortsov D, Yang JD, Rabin BE, Carrick J, Tavare S, and Tower J, 2004. Similar gene expression patterns characterize aging and oxidative stress in Drosophila melanogaster. Proc Natl Acad Sci USA 101:7663–7668.
Leips J, and Mackay TFC, 2000. Quantitative trait loci for life span in Drosophila melanogaster: Interactions with genetic background and larval density. Genetics 155:1773–1788.
Lin YJ, Seroude L, and Benzer S, 1998. Extended life-span and stress resistance in the Drosophila mutant methuselah. Science 282:943–946.
Lindsley DL, and Grell EH, 1967. Genetic variations of Drosophila melanogaster Carnegie Institution of Washington, Washington, DC.
McCarroll SA, Murphy CT, Zou SG, Pletcher SD, Chin CS, Jan YN, Kenyon C, Bargmann CI, and Li H, 2004. Comparing genomic expression patterns across species identifies shared transcriptional profile in aging. Nature Genet 36:197–204.[CrossRef][Web of Science][Medline]
Medawar PB, 1946. Old age and natural death. Modern Quart 2:30–49.
Medawar PB, 1952. An unsolved problem in biology H. K. Lewis, London.
Muir WM, and Howard RD, 2001. Fitness components and ecological risk of transgenic release: A model using Japanese medaka (Oryzias latipes). Am Nat 158:1–16.[CrossRef]
Parkes TL, Elia AJ, Dickinson D, Hilliker AJ, Phillips JP, and Boulianne GL, 1998. Extension of Drosophila lifespan by overexpression of human SOD1 in motorneurons. Nature Genet 19:171–174.[CrossRef][Web of Science][Medline]
Pletcher SD, 1999. Model fitting and hypothesis testing for age-specific mortality data. J Evol Biol 12:430–439.[CrossRef][Web of Science]
Pletcher SD, Houle D, and Curtsinger JW, 1998. Age-specific properties of spontaneous mutations affecting mortality in Drosophila melanogaster. Genetics 148:287–303.
Pletcher SD, Khazaeli AA, and Curtsinger JW, 2000. Why do life spans differ? Partitioning mean longevity differences in terms of age-specific mortality parameters. J Gerontol Ser A-Biol Sci Med Sci 55:B381–B389.
Pletcher SD, Macdonald SJ, Marguerie R, Certa U, Stearns SC, Goldstein DB, and Partridge L, 2002. Genome-wide transcript profiles in aging and calorically restricted Drosophila melanogaster. Curr Biol 12:712–723.[CrossRef][Web of Science][Medline]
Promislow DEL, 2004. Protein networks, pleiotropy and the evolution of senescence. Proc R Soc Lond Ser B-Biol Sci 271:1225–1234.[Medline]
Promislow DEL, and Bugbee M, 2000. Direct and correlated responses to selection on age at physiological maturity in Drosophila simulans. J Evol Biol 13:955–966.[CrossRef]
Promislow DEL, Tatar M, Pletcher S, and Carey JR, 1999. Below threshold mortality: implications for studies in evolution, ecology and demography. J Evol Biol 12:314–328.[CrossRef]
Rogina B, Reenan RA, Nilsen SP, and Helfand SL, 2000. Extended life-span conferred by cotransporter gene mutations in Drosophila. Science 290:2137–2140.
Rose MR, 1985. Life history evolution with antagonistic pleiotropy and overlapping generations. Theor Popul Biol 28:342–358.[CrossRef][Web of Science]
Rose MR, and Long AD, 2002. Ageing: the many-headed monster. Curr Biol 12:R311–R312.[Medline]
SAS Institute,, 2001. SAS, version 8.2 SAS Institute, Cary, NC.
Shaw FH, Promislow DEL, Tatar M, Hughes KA, and Geyer CJ, 1999. Toward reconciling inferences concerning genetic variation in senescence in Drosophila melanogaster. Genetics 152:553–566.
Shook DR, Brooks A, and Johnson TE, 1996. Mapping quantitative trait loci affecting life history traits in the nematode Caenorhabditis elegans. Genetics 142:801–817.[Abstract]
Snoke MS, and Promislow DEL, 2003. Quantitative genetic tests of recent senescence theory: Age-specific mortality and male fertility in Drosophila melanogaster. Heredity 91:546–556.[CrossRef][Web of Science][Medline]
Sokal RR, and Rohlf FJ, 1995. Biometry W. H. Freeman, New York.
Spencer CC, Howell CE, Wright AR, and Promislow DEL, 2003. Testing an ‘aging gene’ in long-lived Drosophila strains: increased longevity depends on sex and genetic background. Aging Cell 2:123–130.[CrossRef][Web of Science][Medline]
SPSS I,, 1999. SPSS for Windows, 10.0.1 SPSS.
Tatar M, Kopelman A, Epstein D, Tu MP, Yin CM, and Garofalo RS, 2001. A mutant Drosophila insulin receptor homolog that extends life- span and impairs neuroendocrine function. Science 292:107–110.
Ungerer MC, Linder CR, and Rieseberg LH, 2003. Effects of genetic background on response to selection in experimental populations of Arabidopsis thaliana. Genetics 163:277–286.
Vieira C, Pasyukova EG, Zeng ZB, Hackett JB, Lyman RF, and Mackay TFC, 2000. Genotype–environment interaction for quantitative trait loci affecting life span in Drosophila melanogaster. Genetics 154:213–227.
Williams GC, 1957. Pleiotropy, natural selection, and the evolution of senescence. Evolution 11:398–411.[CrossRef][Web of Science]
Williams PD, and Day T, 2003. Antagonistic pleiotropy, mortality source interactions, and the evolutionary theory of senescence. Evolution 57:1478–1488.[Medline]
Yampolsky LY, Pearse LE, and Promislow DEL, 2001. Age-specific effects of novel mutations in Drosophila melanogaster. I. Mortality. Genetica 110:11–29.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. W. Fox, K. L. Scheibly, W. G. Wallin, L. J. Hitchcock, R. C. Stillwell, and B. P. Smith The Genetic Architecture of Life Span and Mortality Rates: Gender and Species Differences in Inbreeding Load of Two Seed-Feeding Beetles Genetics, October 1, 2006; 174(2): 763 - 773. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||