Search tips
Search criteria 


Logo of dnaMary Ann Liebert, Inc.Mary Ann Liebert, Inc.JournalsSearchAlerts
DNA and Cell Biology
DNA Cell Biol. 2010 September; 29(9): 473–485.
PMCID: PMC2931542

Age- and Diet-Specific Effects of Variation at S6 Kinase on Life History, Metabolic, and Immune Response Traits in Drosophila melanogaster


Life history theory hypothesizes that genetically based variation in life history traits results from alleles that alter age-specific patterns of energy allocation among the competing demands of reproduction, storage, and maintenance. Despite the important role that alleles with age-specific effects must play in life history evolution, few naturally occurring alleles with age-specific effects on life history traits have been identified. A recent mapping study identified S6 kinase (S6k) as a candidate gene affecting lipid storage in Drosophila. S6k is in the target of rapamycin pathway, which regulates cell growth in response to nutrient availability and has also been implicated to influence many life history traits from fecundity to life span. In this article, we used quantitative complementation tests to examine the effect of allelic variation at S6k on a range of phenotypes associated with metabolism and fitness in an age-, diet-, and sex-specific manner. We found that alleles of S6k have pleiotropic effects on total protein levels, glycogen storage, life span, and the immune response and demonstrate that these allelic effects are age, diet, and sex specific. As many of the genes in the target of rapamycin pathway are evolutionarily conserved, our data suggest that genes in this pathway could play a pivotal role in life history evolution in a wide range of taxa.


Most phenotypic traits of organisms are quantitative, exhibiting continuous rather than categorical variation in populations. This variation is the product of genetic variation at many loci and their interaction with the environment (Tanksley, 1993). A fundamental goal of quantitative genetics is to determine the genetic basis of variation in quantitative traits and to understand the processes that maintain this variation. This is essential, not only for practical reasons such as assessment of disease risk in human medicine and selective breeding in agriculture, but also for understanding of the evolution of phenotypic traits.

Among the most important set of quantitative traits from an evolutionary perspective are life history traits. As these traits determine the age schedules of birth and death, they are major components of fitness, and so primary targets of natural selection. Life history theory posits that differential allocation of energy among the competing demands of growth, development, reproduction, energy storage, and somatic maintenance and repair lead to trade-offs among traits (Stearns, 1991). Thus, genes controlling how energy is used and allocated should play key roles in the evolutionary divergence of life history strategies between species and contribute significantly to variation in fitness within populations. Although it is clear that genes influence energy allocation, we know very little about the genes controlling this process. As such, identification of the genes that contribute to natural variation in life history variation is a major goal of evolutionary biology.

One particular area of life history research in which age-specific genetic effects play a central role is the study of the genetic basis of longevity. A central assumption of the two leading evolutionary theories of aging, mutation accumulation (Medawar, 1952), and antagonistic pleiotropy (Williams, 1966), is that senescence evolves through the action of alleles that have age-specific phenotypic effects on fitness. Under the mutation accumulation model, alleles that are neutral with respect to fitness, but deleterious at later ages, can accumulate in a population over time. An alternative, but not mutually exclusive, theory is the antagonistic pleiotropy model of senescence (Williams, 1957). Under this model, an allele with positive effects on fitness at early age but deleterious effects at later age will accumulate in the population over time. Under both models, accumulation of alleles with late-acting deleterious give rise to senescence and limit life span. Both of these theories rely on the fact that selection against deleterious alleles gets progressively weaker after the onset of reproduction (Hamilton, 1966). This is because individuals that reproduce before the age at onset of the deleterious effects will have already passed their alleles on to the next generation, thus making it difficult for natural selection to remove them. Although the assumption that alleles with age-specific effects are central to life history theory and evolutionary models of senescence, few studies have attempted to characterize the age-specific effects of alleles at defined loci on traits associated with fitness.

Environmental influences also have a strong effect on life history traits in general (Yang et al., 2008) and senescence in particular, both directly (e.g., Brakefield et al., 2005; Masoro, 2005; Valenzano et al., 2006) and indirectly by modifying the genetic influences on life history traits (Hoffmann and Merila, 1999; Vieira et al., 2000; Charmantier and Garant, 2005). Because environmental variation is ubiquitous in natural populations, it is also important to evaluate how age-specific allelic effects on fitness traits are modulated by ecologically relevant environmental variation. Such knowledge would provide a more complete picture of the genetic basis of variation in life history traits and aging because the relative influence of genes contributing to the phenotype may vary depending on the environmental conditions experienced. An efficient approach to this problem is to first identify the genes that contribute to phenotypic variation in traits known to exhibit age-related changes, and then characterize the age-specific effect of alleles at these loci under different environmental conditions.

In an earlier study, De Luca et al. (2005) carried out a quantitative trait loci (QTL) mapping experiment with Drosophila melanogaster to identify chromosomal regions (QTL) producing variation in triacylglycerol levels, an important storage lipid in all organisms (McManus and Travis, 1998; Meex et al., 2009; Petschnigg et al., 2009). One candidate gene identified by this study was S6 kinase (S6k), a key component of the target of rapamycin (TOR) signaling pathway. Notably, S6k was also identified as a candidate gene contributing to variation in life span by a previous QTL study that used the same set of lines (Nuzhdin et al., 1997). Genes regulating the TOR pathway are ideal candidates to test for their influence on age-specific metabolism, as the TOR pathway regulates protein synthesis and glucose homeostasis in cells in response to nutrient availability (Arsham and Neufeld, 2006). Downstream components of this pathway are also influenced by insulin and insulin growth factor (IGF) signaling, and mutations in genes in the TOR pathway including TOR and insulin receptor have been shown to influence life span, reproduction, and metabolism (Grewal, 2009).

TOR itself is a kinase that responds to nutrient availability by phosphorylating and activating S6k. S6k is a critical gene in this pathway as it phosphorylates a number of potential substrates, including ribosomal protein s6 (Rp6) and elongation initiation factor 4B. The exact regulatory control of translation by the S6K protein is not known, but experimental evidence suggests that phosphorylation of Rp6 by S6K downregulates protein synthesis, whereas phosphorylation of elongation initiation factor 4B upregulates protein synthesis (Ruvinsky and Meyuhas, 2006; Ma and Blenis, 2009). In this way, S6k may act to fine-tune the rate of protein production.

In this article we use quantitative complementation tests (Pasyukova et al., 2000, 2004; De Luca and Leips, 2007) to test the influence of natural alleles of S6k age-specific fecundity, immune response, total protein, glycogen and triacylglycerol (TAG) levels, and life span. We chose these traits as they should reflect energy investment in reproduction, maintenance and repair, and storage, all key elements that should compete for acquired energy. Because we wished to examine the effect of environmental variation on the allelic effects of this gene, we reared flies on one of two diets, a regular diet of cornmeal molasses and yeast (STD diet) and under dietary restriction, which is identical to the regular diet but with a 60% reduction in yeast. Yeast is the primary source of protein and cholesterol for this species, and therefore reduction in yeast should have far-reaching effects on many phenotypic traits. We also used an independent mutant stock of S6k to verify the influence of this gene on the immune response as well as to assess mechanistically how this gene might affect immunity. To do this we used a modification of a standard phagocytosis assay to examine the role of this gene on the ability of blood cells to engulf bacteria.

Materials and Methods

We carried out quantitative complementation tests (Pasyukova et al., 2000) to evaluate the effects of alternative alleles of S6k on live weight, life span, and the following age-specific traits: glycogen content, triacylglycerol levels, total protein, fecundity, and immune response.


For the quantitative complementation test we used two isogenic laboratory strains, Oregon and 2b (described in detail in Hughes and Leips, 2006), which were the parental strains of the recombinant inbred lines used in the initial QTL mapping experiment for lipid storage (De Luca et al., 2005).

We used two mutant stocks of S6k for the experiments described in this study. For the quantitative complementation test we used Bloomington Stock P{PZ}S6k07084 ry506/TM3, ryRK Sb1 Ser1 (Stock #11713), which has a P-element insert in the 5′ region (~48 bp upstream of the first ATG site in the coding region). For the phagocytosis experiments we used the w1118 stock from the Bloomington Drosophila Stock Center (#6326) and a hypomorphic S6k mutant, PBac{WH}S6kf07622 stock from the Exelixis Drosophila Stock Collection at the Harvard Medical School. The w1118 stock was used as the control in the phagocytosis experiment since the mutant strain was derived from w1118 and should be genetically identical except for the piggyBac insertion (Thibault et al., 2004). Unlike the stock with the P-element insert that was used in the quantitative complementation test (#11713), this mutant stock is viable as a homozygote. The PBac stock has the lepidopteran transposon piggyBac inserted into the 3′ UTR section of S6k.

Crossing design for quantitative complementation tests

Ten males of each parental strain, 2b and Ore, were collected as virgins and mated with 20 females of the S6k mutant strain separately in 500 mL population bottles. The opening of each bottle was covered by egg laying chamber (Petri dish containing 2% agar brushed with food and sprinkled with yeast) and inverted. From each dish, 50 first to second instar larvae were collected and placed in vials containing 10 mL of standard fly food to standardize larval density.

Virgin offspring were collected and sorted into one of four genotypes: Ore/Balancer, Ore/S6k mutant, 2b/Balancer, and 2b/S6k mutant. Flies with the balancer chromosome were distinguishable because they had the Stubble bristle mutation, which is on the balancer chromosome. Virgin males and females of each genotype were separated and placed in vials containing either the STD diet or yeast-restricted diet (DR), 5–6 to a vial.

For all phenotypes for which we made age-specific measurements, we measured early and late age traits on different individuals because in all but the fecundity assay, the assays themselves were destructive. Individuals held until 4 weeks were transferred onto new food at least once a week until the phenotypic measurements were made.

Live weight, total protein, glycogen, and TAG

Three replicate groups of 20 flies per age, sex, diet, and genotype were anesthetized, weighed, and homogenized using the homogenization protocol described in Clark and Keith (1988). Briefly, adults were homogenized on ice using 25 μL of homogenization buffer (0.01 M KH2PO4, 1 mM EDTA pH 7.4). The homogenates were centrifuged in a microcentrifuge at 2000 rpm for 2 min at 4°C. The lipid layer on the surface was resuspended with the supernatant in 1.5 mL tubes. TAG levels were assayed at the Clinical Nutrition Research Center at University of Alabama at Birmingham using the Vitros DT60 II reader (Johnson and Johnson Clinical Diagnostics, Rochester, NY) and Vitros TRIG DT slides. The Vitros TRIG DT Slide is a multilayered film that contains all the reagents necessary to determine TAG levels in the homogenate. The analyzer calculates the amount of TAGs in the sample by measuring the amount of light emitted from a colored dye produced in the last of a series of chemical reactions that start with the conversion of TAG molecules into glycerol and fatty acids. A precision test was carried out at each run using quality-control materials at various concentrations of TAGs. Each sample was assayed twice and the mean used in the analysis. Total proteins were measured for each homogenate using a standard Lowry protein assay.

Glycogen content was determined separately for the thoraces and the abdomens of flies. Glycogen was measured using a modification of the protocol described in Clark and Keith (1988). Briefly, aliquots of 1.67 μL of homogenate were added to 250 μL of a reagent containing 0.1 U/mL of amyloglucosidase, 5 U/mL of glucose oxidase, 1 U/mL of peroxidase, and 0.04 mg/mL of O-dianisidine dihydrochloride. After a 30-min incubation period at 37°C, OD540 was measured. Concentration of glycogen was determined from glucose and glycogen standards run with each replicate. Each sample was assayed twice and the mean used in the analysis. Previous studies have shown that this protocol accurately reflects glycogen concentration and that endogenous glucose present in the flies contributes only negligibly to the results (Clark and Keith, 1988). Total proteins were also measured for each sample.

Virgin fecundity

There were two reasons to measure virgin fecundity in this study. The first was that the eggs of Drosophila contain a significant amount of glycogen (Gutzeit et al., 1994) and we wanted to be able to assess the extent to which variation in glycogen levels might be explained by variation in fecundity. The second was that age-specific fecundity often exhibits a trade-off with life span (Rose, 1991), although this is typically observed in mated flies, and we wanted to examine the potential for trade-offs among these traits in our study. Thirty-five females per genotype were maintained separately in vials containing one of the two food treatments. We counted eggs laid by individual females over 10 consecutive days beginning at 1 and 4 weeks of age. Two groups of females were used: one group was used to obtain 1 week fecundity data, whereas the second group was used to determine 4 week fecundity. At the end of each egg counting period females were frozen for thorax measurements. Thoraces were measured using a stereomicroscope fitted with an ocular micrometer. Females that laid no eggs during this time period were not included in the final analysis.

Immune response assay—bacterial clearance

One measurement of the immune response was the rate of clearance of a bacterial load of Escherichia coli by flies at each age. We used a procedure modified from McKean and Nunney (2001), along with a control of sham-infected flies, injected with only Ringer's solution. A streptomycin-resistant strain of E. coli, strain HB101, was suspended in Drosophila Ringer's solution and diluted until an OD600 reading of 1.0 ± 0.1 was reached (Eppendorf BioPhotometer [Hauppage, NY], Beckman DU-65 Spectrophotometer [Brea, CA]). Three microliters of the bacterial or Ringer's solution was loaded into pulled glass capillary needles and connected to a microinjector (Eppendorf FemtoJet, Westbury, NY). The microinjector was set at a compensation pressure of 0.49 psi, injection pressure of 0.49 psi, and injection time of 0.2 s.

Flies were injected in the abdomen under CO2 anesthesia. Twenty-four hours after injection each surviving fly was placed into 250 μL of 20% glycerol solution in a 1.5 mL centrifuge tube and homogenized with a pestle and motorized homogenizer (Fisher Scientific, Pittsburgh, PA). We plated 25 μL of this solution as well as two other serial diluted aliquots of this solution to obtain plates with countable colonies. Solutions were plated onto LB agar plates containing 100 μg/mL streptomycin and incubated for 15–18 h at 37°C. Bacterial colonies were manually counted, and this colony count served as the immune response phenotype for that replicate. We obtained colony counts from a minimum of 30 flies per genotype per age sex and dietary condition. Sham-infected flies were also homogenized and plated as above. We observed no bacterial growth on plates containing homogenates of sham-injected flies.

Lifespan assay

Male and female flies were placed five or six to a single-sex vial containing one of the two diets. Vials were monitored every other day for dead flies and placed on new food once a week.

Immune response assay—phagocytosis assay

We modified a newly developed procedure developed by Dr. Louisa Wu (University of Maryland College Park) for measuring phagocytosis in larval blood cells, summarized here. We injected third instar larvae of the w1118 control and S6k stock (PBac{WH}S6kf07622) with heat-killed fluorescent bacteria (E. coli labeled with tetramethylrhodamine conjugate, Invitrogen #E2862 [Carlsbad, CA]). Individual larvae were held for 30 min and then bled onto microscope slides. Blood cells were fixed in 4% formaldehyde for 10 min. We then applied 5 μL of Oregon Green 488 phalloidin in 200 μL phosphate-buffered saline + 1% bovine serum albumin for 20 min. Blood cells were stained with phalloidin because this toxin binds actin at the cell cortex and so made it possible for us to distinguish cell boundaries. Cells were washed two to three times in phosphate-buffered saline and then we added Prolong + DAPI (Invitrogen # P36930). The Prolong was added as an antifade agent, and the nuclear stain DAPI was used to ensure that we were counting bacteria in viable cells. We counted the number of engulfed bacteria per blood cell using the 40× objective of a Zeiss Axio Imager Z1 microscope with AxioVision software. Blood cell counts were obtained from seven mutant larvae and six controls, with the number of cells counted per individual ranging from 8 to 60. This range was due to the fact that the total number of countable blood cells varied among individuals. We also counted the total number of cells containing bacteria between the control and mutant strains per larvae but found no difference between the genotypes in this measure of phagocytosis (data not shown).

Statistical analyses

Main effects of sex, diet, and age

We used either analysis of variance (ANOVA) or analysis of covariance (ANCOVA) (detailed for each trait below) to assess the main effects of diet, sex, and age on each phenotype where appropriate. For these analyses we used reduced models, pooling data from all genotypes. All analyses were carried out in SAS using Proc GLM (SAS V9.1; Cary, NC).

Live weight

We used ANOVA to test for the effects sex, diet, and age on live weight using the model y = c + s + a + d plus all interactions, where y = live weight of individual flies in milligrams, c is a constant, s = sex (male or female), a = age (1 or 4 weeks posteclosion), and d = diet (STD food or DR). All effects in this model were fixed effects.

Total protein, glycogen, and triacylglycerol

We analyzed the main effects of sex, diet, and age on these traits in ANCOVA. For total protein content we used the model y = c + lw + s + a + d + all interactions + error, where y = total protein, lw = live weight, a covariate to correct for differences in body size and all other variables as above. We analyzed glycogen content in the abdomen and thorax separately and only analyzed triacylglycerol in the abdomen. For glycogen and triacylglycerol we used the model y = c + p + s + a + d + all interactions + error, where y = glycogen or triacylglycerol and p = tissue-specific protein content, a covariate to account for differences in amounts of extracted tissue. In each full model tested, none of the interactions involving covariates (live weight or total protein) and the main effects were significant, and thus these interaction terms were dropped in the final analysis. Total protein and glycogen data were transformed to natural logs before analysis to satisfy assumptions of ANOVA.

Immune response—bacterial clearance

We used ANOVA to test for the main effects of sex, age, and diet on the ability to clear infection using the model y = c + s + a + d + all interactions + error, where y = colony count 24 h after infection and all other parameters as above. We used the natural log of the colony count for individual to satisfy assumptions of ANOVA.

Virgin fecundity

We used an ANCOVA to test the effect of age and diet on fecundity using the model y = c + tl + a + d + all interactions + error, where y = the total eggs laid per female over a 10-day period, and tl = thorax length. In the full model, none of the interactions between thorax length and the main effects were significant, so interaction effects involving live weight were dropped from the model in the final analysis.

Life span

We used mixed model ANOVA to test for the fixed main effects of sex and diet on lifespan using the model y = c + s + d + s*d + v(s d) + error, where y = lifespan in days, and v(s d) is the random effect of vial nested within sex and food.

Quantitative complementation tests

All complementation analyses were done separately for each sex, age, and diet combination. The general model for each trait was y = c + line + geno + line*geno + error, where line was 2b or Ore, geno = mutant or balancer, and line*geno was the interaction term. In the quantitative complementation test significant phenotypic differences between the Ore/S6k and 2b/S6k genotypes can arise from two mechanisms, by differences in the additive effects of the Ore and 2b alleles at the S6k locus on the trait and due to the phenotypic effects of allelic differences between the parental strains outside of the locus being tested (Pasyukova et al., 2000). To control for the effects of allelic differences between the Ore and 2b strains on the trait elsewhere in the genome, we compare the trait values of the Ore/Balancer and 2b/Balancer genotypes with those of the Ore/S6k and 2b/S6k mutants using ANOVA. In this analysis a significant interaction term would indicate a failure to complement. In addition, to use this test to support the interpretation that alternative alleles at S6k produce variation in the phenotype tested, the difference in the average phenotypes of each line over the deficiency (the Ore/Def–2b/Def) must be greater than the difference between the average phenotypes of each line over the balancer (Ore/Bal–2b/Bal).

Quantitative complementation tests—triacylglycerol

For TAG analyses we used the model: y = c + p + line + geno + all interactions + error. Terms in the model are as described above except that y = triacylglycerol level (assessed from the abdomen only) from 5 individuals pooled from each treatment combination. We also included total protein (p) as a covariate in the model. Any interactions involving the covariate that were not significant were dropped from the model in the final analysis.

Quantitative complementation tests—glycogen

For glycogen analyses we used the model y = c + p + line + geno + all interactions + error. Terms in the model are as described above except that y = glycogen level (assessed separately from the abdomen and thorax) from five individuals pooled from each treatment combination. We also include total protein (p) as a covariate in the model. Any interactions involving the covariate that were not significant were dropped from the model for the final analysis.

Quantitative complementation tests—total protein

For total protein analysis we used the model y = c + lw + line + geno + all interactions + error. Terms in the model are as described above except that in this model y = total protein level (whole body protein obtained from five individuals pooled from each treatment combination per replicate). We include live weight as a covariate in the model. Any interactions involving the covariate that were not significant were dropped from the model for the final analysis

Quantitative complementation tests—immune response—bacterial clearance

For the bacterial clearance analysis we used the model y = c + line + geno + all interactions + error. Terms in the model are as described above except that in this model y = the total number of colony forming units in an individual fly 24 h after artificial infection.

Quantitative complementation tests—virgin female fecundity

For analysis of fecundity we used the model y = c + tl + line + geno + all interactions + error. Terms in the model are as described above except that in this model y = the total number of eggs laid in a 9-day period per female. We also include thorax length (tl) as a measure of body size as a covariate in the model. We did not use weight, as this variable could be confounded with egg load. We checked for significant interactions between covariate and the main effects. Any interactions involving the covariate that were not significant were dropped from the model in the final analysis.

Immune response—cellular phagocytosis assay

For analysis of the effects of S6k on phagocytic ability, we carried out a nested ANOVA using the model y = c + line + ind(line) + error, where y = the transformed number of bacteria per cell (transformed to the square root of the number per cell + 1 square root of the number per cell to satisfy assumptions of ANOVA, Snedecor and Cochran, 1980), and ind(line) = individuals nested within line. This was necessary as we measured several cells per individual, and thus individual cell counts were not independent of individuals.


Live weight

Main effects of age/sex/diet

We found significant differences in live weight between the sexes and between ages. Males were significantly smaller than females (the average live weight for males was 0.77 mg, the average weight of females was 1.03 mg, F1,396 = 477.4, p < 0.0001), and older flies were significantly lighter than younger flies (average weight of 1-week-old flies was 0.91 mg, the average weight of 4-week-old flies was 0.89 mg, F1,396 = 6.51, p = 0.006). Neither diet nor any interactions among the main effects significantly affected live weight.

Quantitative complementation test: effects of variation in S6k

We found no effects of variation at S6k on the live weight of males or females at any age or on any diet.

Total protein

Main effects of age/sex/diet

Total protein was significantly influenced by both age (F1,395 = 568, p < 0.0001) and diet (F1,395 = 89, p < 0.0001), with age having the most dramatic effect. Total protein decreased by 60% from week 1 to 4 (the least squared mean corrected for differences in live weight for 1-week-old flies was 54.7 μg/fly, and for 4-week-old flies was 22.1 μg/fly), and the effect of age explained 41% of the variation in the data. Reduction in protein content with age in Drosophila has previously been reported (Johnson and Butterworth, 1985). We suspect that the magnitude of decline in total protein observed in our study was due to age-associated decrease in muscle mass, although this remains to be confirmed. The effect of diet was also substantial as flies on DR had 33% less total protein than those on the STD diet (the least squared mean for flies on the STD diet was 46.1 μg/fly, and for those on DR was 30.7 μg/fly). As yeast is the primary source of protein for Drosophila, the effect of DR on total protein levels is predictable. Indeed, in a study of the effect of dietary composition on metabolic traits in Drosophila, Skorupa et al. (2008) report that flies maintained on a diet of high yeast also had a high protein body composition compared with diets with reduced yeast levels. Additional effects on protein level could be due to different feeding rates of flies on the different diets, but this remains to be evaluated in future studies. After accounting for differences in live weight, there was no difference between males and females in total protein and there were no significant interactions among the main effects.

Quantitative complementation test: effects of variation in S6k

Variation in S6k influenced total protein levels, but the allelic effects at this locus were age, sex, and diet specific, with significant effects only identified in 4-week-old females on DR (F1,30 = 5.0, p = 0.04, Fig. 1). In this case, females with the 2b allele had 20% more protein than female flies with the Ore allele.

FIG. 1.
Results of complementation test for total protein. Results are for females on the yeast-restricted diet (DR) at 4 weeks of age. Values shown are the means of each genotype ± one standard error.


Main effects of age/sex/diet

Abdominal glycogen: Abdominal glycogen differed significantly between sexes, ages, and diets, and we found significant interactions among all of the main effects except for the sex-by-diet interaction (Table 1). On average, females had 24% more abdominal glycogen than males (the least squared mean for males was 3.61 μg/fly, and for females 4.75 μg/fly). This difference between sexes may be due to a variety of factors, including sex-specific differences in metabolic rates, energy acquisition, or the presence of eggs in females (Gutzeit et al., 1994). Older flies had 61% less abdominal glycogen than younger flies (the least squared mean for 1-week-old flies was 6.0 μg/fly, and for 4-week-old flies was 2.4 μg/fly). This cannot be explained by reduced egg load in females with age as female fecundity actually slightly increased with age (see Results for virgin female fecundity below). Glycogen is also a major component of the cells of the fat body, but age-related decline in glycogen as a proportion of the cellular constituents of fat body cells has only been previously reported in males (Johnson and Butterworth, 1985). Age-related decreases in glycogen in this study may be explained by any number of factors, including increased metabolic rates with age, a shift in the allocation of energy away from storage or reduced energy acquisition with age, or age-related decreases in feeding. These possibilities remain to be examined. Flies on the STD diet had 20% more glycogen than those on DR (the least squared mean for flies on the STD diet was 4.65 μg/fly, and for flies on DR was 3.73 μg/fly), which is to be expected given the higher nutritional value of the STD diet. The sex-by-age interaction resulted from the fact that abdominal glycogen in females declined by a greater percentage with age (43%) than it did in males (34%, data not shown). The age-by-diet interaction was due to the fact that the drop in abdominal glycogen with age was much greater for flies on the DR (58%) than it was for those on the STD diet (15%). The interaction that explained the largest percent of the variation in the data was the sex-by-age-by-diet interaction, and was driven largely by the fact that males at 1 week of age had 40% more abdominal glycogen on the DR than they did on the STD diet (Table 2). This pattern was reversed in older males and it matched the pattern exhibited by females at all ages, being that females had higher glycogen on the STD diet.

Table 1.
Results from Analysis of Covariance on Abdominal Glycogen Levels (with Protein as the Covariate)
Table 2.
Least Squared Means of Abdominal Glycogen Levels

Thoracic glycogen: All main effects and their interactions influenced thoracic glycogen levels (Table 3). The main effects of sex, age, and diet on thoracic glycogen were similar to that of abdominal glycogen: females had 17% more glycogen than males (the least squared mean for males was 0.95 μg/fly, and for females was 1.14 μg/fly), older flies had 65% less glycogen than younger flies (the least squared mean for 1-week-old flies was 1.55 μg/fly, and for 4-week-old flies was 0.54 μg/fly) and flies on standard food had 15% more glycogen than those on the RD (the least squared mean for flies on the STD diet was 1.08 μg/fly, and for those on DR was 1.04 μg/fly). The interaction effects explaining the greatest amount of variation in glycogen storage were the age-by-diet and the sex-by-age-by-diet interactions, together explaining 26% of the phenotypic variation. As was the case for abdominal glycogen, these interactions are largely due to the sex-specific influence of diet on glycogen storage in young flies, specifically young males increasing the level of glycogen stored under DR, with females exhibiting the opposite pattern (Table 4). As discussed above relative to the effects of age on glycogen storage, there are many possible explanations for the sex-, diet-, and age-specific interactions affecting glycogen storage, and these need to be investigated in future experiments.

Table 3.
Results from Analysis of Covariance on Thoracic Glycogen Levels
Table 4.
Least Squared Means of Thoracic Glycogen Levels

Quantitative complementation test: effects of variation in S6k

Abdominal glycogen: Allelic variation at S6k affected female glycogen levels in an age-dependent manner (Table 5). Variation in S6k influenced abdominal glycogen only in 4-week-old females, and on both diets, females with the 2b allele had significantly more glycogen than females with the Ore allele (Fig. 2) (STD diet: F1,31 = 9.14, p = 0.01; DR: F1,30 = 7.33, p = 0.01). Variation in S6k had no effect on glycogen storage in the abdomens of males at any age on any diet.

FIG. 2.
Results of complementation test for female abdominal glycogen on the DR at 4 weeks of age. Values shown are the means of each genotype ± one standard error.
Table 5.
Age-, Diet-, and Tissue-Specific Influences of Variation at S6 Kinase on Glycogen in Females

Thoracic glycogen: Variation at S6k also affected female thoracic glycogen levels, but in contrast to the effect on abdominal glycogen, variation in S6k only affected young females and the allelic effects were reversed from that seen in the abdomen. In the thorax, female flies with Ore alleles had significantly more glycogen than those with 2b alleles on both diets (Fig. 3, STD: F1,31 = 5.60, p = 0.02; DR: F1,19 = 4.59, p = 0.04). Variation at S6k also affected glycogen levels in male thoraces, but only when young and under DR. In this case, males with the Ore allele had 36% more glycogen in the thorax than males with the 2b allele (F1,31 = 10.66, p = 0.003).

FIG. 3.
Results of complementation test for female thoracic glycogen on the DR at 1 week of age. Values shown are the means of each genotype ± one standard error.


Main effects of age/sex/diet

Sex, age, and diet all significantly affected TAG levels, but age had the largest effect, explaining 14% of the total phenotypic variation (Table 6). Males had 3% more TAG than females (the least squared mean of males was 8.18 μg/fly, and for females was 7.91 μg/fly). Older flies had 9% more TAG (the least squared mean for 1-week-old flies was 7.65 μg/fly, and for 4-week-old flies was 8.43 μg/fly). Age-related increase in TAG levels are common in mammals (e.g., Tucker and Turcotte, 2003) and are one of the hallmarks of metabolic syndrome in aged humans (Kong et al., 2006; Penninx et al., 2009). Flies on DR had 2% less TAG than those on the STD diet (the least squared mean for flies on the STD diet was 8.12 μg/fly, and for those on DR was 7.96 μg/fly). Although the sex-by-age effect was significant, it explained very little of the phenotypic variation. Both males and females had higher TAG levels at 4 weeks, but females had a slightly higher increase with age compared with males (12% vs. 11% based on least squared means after accounting for differences in body weight).

Table 6.
Results from Analysis of Covariance on Abdominal TAG Levels

Quantitative complementation test: effects of variation in S6k

We found no effects of variation at S6k on TAG levels in males or females at any age or on any diet.

Virgin fecundity

Main effects of age/diet

Both diet and age significantly affected virgin fecundity, and there was a significant age-by-diet interaction (Table 7). Females at 1 week of age actually laid fewer eggs than those at 4 weeks (the least squared mean for 1-week-old females was 1.0 eggs/day, and for 4-week-old females was 2.4 eggs/day). Flies on DR had fewer eggs per day than those on the STD diet (the least squared mean for flies on the STD diet was 2.6 eggs/day, and for those on DR it was 0.9 eggs/day). The age-by-food interaction resulted from the fact that females on the STD diet had a 70% increase in the number of eggs laid per day from weeks 1 to 4, whereas females in the DR treatment increased their daily egg production by 45% from week 1 to 4.

Table 7.
Results from Analysis of Variance on Main Effects of Diet and Age on Virgin Fecundity

Quantitative complementation test: effects of variation in S6k

Variation at S6k had no effect on virgin fecundity at any age on either diet.

Life span

Main effects of sex/diet

Averaged over all genotypes, there was no significant difference in life span between sexes or between diets, or any significant effect of the interaction between sex and diet on life span.

Quantitative complementation test: effects of variation in S6k

Variation in S6k did have a significant influence on life span, but only in one sex on one diet. Female flies with the Ore allele lived 24% longer than those with the 2b allele but only on the STD diet (F1,119 = 4.64, p = 0.03, Fig. 4). Variation at S6k had no effect on the life span of males on either diet.

FIG. 4.
Results of complementation test of bacterial clearance ability (A) Males on the regular diet at 4 weeks of age. (B) Females on the regular diet at 1 week of age. (C) Females on the DR at 4 weeks of age. All values shown represent the means of each genotype ± one ...

Immune response—bacterial clearance ability

Main effects of age/sex/diet

Both sex and diet had significant influences on the ability to clear bacterial infection (Table 8). The clearance ability of males was 10% greater than that of females (the least squared mean of the natural log of colony forming units in males was 5.10, whereas that of females was females 5.69). The relative ability of males to clear the infection compared to females is even more impressive when one considers that males are approximately 25% smaller than females (based on our live weight measurements). As such, male flies are receiving a significantly higher infection load per body mass than females. Although we do not know the physiological basis of this difference, we do not think that this result is due to differences in bacterial growth between the sexes. In our experience with females, E. coli does not grow in flies after infection (unpublished data).

Table 8.
Results from Analysis of Variance on Bacterial Clearance Ability

Flies on DR had a 31% greater ability to clear infection than those on the STD diet (the least squared mean of the natural log of colony forming units for flies on the STD diet was 6.38, and for flies on DR 4.40) with differences due to diet explaining 33% of the total variation. Although there was no main effect of age, there was a significant sex-by-age interaction. This interaction was due to the fact that the clearance ability of males decreased by 5% with age, whereas the clearance ability of females actually increased. There was also a significant sex-by-diet interaction. In this case, although clearance ability of both males and females was improved on the restricted diet, dietary restriction was more beneficial to males (increasing clearance ability by 40%) compared with females (increased clearance ability by 22%).

Quantitative complementation test: effects of variation in S6k

Variation in S6k influenced the immune response of both sexes, but in an age- and diet-specific manner. S6k only affected male immune response at 4 weeks of age on the STD diet (Fig. 4A, F1,117 = 7.36, p = 0.007). In this case, males with the 2b allele had a higher bacterial clearance rate than males with the Ore allele. Variation at S6k influenced bacterial clearance rate of females under two conditions, at early age on the STD diet (F1,120 = 5.58, p = 0.02) and at late age on the restricted diet (F1,138 = 5.94, p = 0.02). In both cases, females with the 2b allele had improved clearance ability compared to those with the Ore allele (Fig. 4B, C).

Confirmation of effects of S6k on immune response and phagocytic ability

To independently assess the role of S6k on the immune response, we measured bacterial clearance ability of female flies with a hypomorphic mutation of this gene and females of a control strain, which were genetically identical except for the mutation in S6k. We used these same lines also to measure the phagocytic ability of blood cells of mutant and control strains using a phagocytosis assay on larval blood cells and heat-killed bacteria. Bacterial clearance ability of the mutants on the STD diet was reduced by 40% compared with controls (F1,52 = 16.47, p < 0.0002). There was also a significant effect of S6k on the phagocytic ability of blood cells (F1,435 = 18, p < 0.0001). Surprisingly, the average number of bacteria per blood cell from the mutant flies was 12% greater than the number found in blood cells of the control strain (Fig. 5). A second replicate of this experiment yielded quantitatively similar results (not shown).

FIG. 5.
Comparison of the phagocytic ability of blood cells of the S6k mutant and w1118 control strain. Values shown are averages ± one standard error. S6k, S6 kinase.


We used quantitative complementation tests (Pasyukova et al., 2000, 2004; De Luca and Leips, 2007) to test the age- and diet-specific effects of variation at S6k on life history, and metabolic and immune response traits in two laboratory strains, Ore and 2b, of D. melanogaster. One objective of the study was to investigate whether genetically based trade-offs between traits were regulated by this locus. In this case, if a trade-off existed we would have expected one allele of S6k to have a positive effect on one fitness trait but a negative effect on another trait, and this was not observed for most traits (Table 9). However, one interesting finding was the apparent trade-off between bacterial clearance ability at 1 week of age and thoracic glycogen storage. Mounting a successful immune response requires a significant energy investment (e.g., Demas et al., 1997; Moret and Schmid-Hempel, 2000; Bonneaud et al., 2003). There are a number of studies indicating an evolutionarily conserved link between immune function and insulin/IGF signaling (Diangelo et al., 2009; Zeyda and Stulnig, 2009), which in turn regulates glycogen levels. In addition, S6k plays an essential role in pathways that render cells unresponsive to insulin (Manning, 2004). Based on these observations, we speculate that genotypic differences in the efficiency of the immune response between Ore and 2b may reflect differences in allocation of glucose between traits associated with survival. Of course, given the number of traits we examined overall, the genetically based trade-off we observed here may be spurious. Future experiments are needed to see if this trade-off is maintained and if so to provide a mechanistic basis for the association between these traits.

Table 9.
Summary of Significant Phenotypic Effects of Variation at S6 Kinase from the Quantitative Complementation Tests

For the most part, all significant allelic effects appeared to be maintained for a given trait, although it is interesting that flies with the Ore allele had higher thoracic glycogen, whereas flies with the 2b allele had higher abdominal glycogen. Although eggs contain a significant amount of glycogen (Gutzeit et al., 1994) the higher abdominal glycogen in older flies with the 2b allele was not due to higher egg loads (at least as measured by the number of eggs laid) as we found no effect of variation at S6k on virgin egg production. We also found that natural variants of S6k have extensive pleiotropic effects on metabolic traits, immune response, and life span. Given the important role of S6k in regulating protein translation, it is not surprising to observe extensive pleiotropic effects, and indeed studies of S6k mutants have found effects on cell and body size (Shima et al., 1998; Montagne et al., 1999), cell survival (Harada et al., 2001), autophagy (Scott et al., 2004), egg production (Hansen et al., 2005; Terashima and Bownes, 2005), hunger-driven behaviors (Wu et al., 2005), life span (Kapahi et al., 2004), and phagocytosis (Ganesan et al., 2004; Stroschein-Stevenson, 2006). It is important to note, however, that S6k may act on a number of different substrates (Ruvinsky and Meyuhas, 2006), in addition to the Rp6, and thus the phenotypic effects of variation in this gene may be mediated through its effects on other processes.

We have also demonstrated that the allelic effects of variation in this gene are modified by diet and age and are not consistent across sexes. Although such context-dependent allelic effects are often found (Leips and Mackay, 2000; Vieira et al., 2000) the age-specific nature of these allelic effects have not been investigated. The TOR pathway is involved in nutrient sensing and variation in nutrient levels are known to rapidly affect transcript levels of S6k (Arsic and Guerin, 2008). As such, genetic variation at this locus might produce differences in the degree to which S6k is expressed. Alternatively, the activity of S6K depends on how efficiently it becomes phosphorylated by kinases upstream in the TOR pathway (Long et al., 2004). If alternative alleles produce functionally different S6K proteins, this could influence the activity of S6K, altering the efficiency of phosphorylation of its targets. Exactly how the context-dependent effects of these natural alleles manifest themselves is an important but unresolved issue.

Our finding that variation in S6k affected glycogen levels but not TAG is interesting given the fact that the original mapping study identified S6k as a gene regulating TAG levels. Although there are many potential explanations for the lack of congruence between the findings of this study and that of De Luca et al. (2005), the most likely is that S6k was not one of the genes that contributed to phenotypic variation in TAG in the earlier QTL mapping study. This is not necessarily unexpected as QTL studies identify regions of the genome affecting phenotypic traits, not the genes themselves. There were a large number of genes within the confidence interval of the original QTL, and further fine mapping experiments will be necessary to identify the actual genes giving rise to variation in TAG in this QTL. We did, however, confirm that variation in S6k contributes to variation in life span as suggested by the earlier QTL study of Nuzhdin et al. (1997).

We were able to confirm the effects of variation at S6k on bacterial clearance ability using an independent set of lines, so it is clear that this is an important gene regulating age-specific immunity. Interestingly, although the S6k mutant flies had a poorer bacterial clearance ability compared with the control, their phagocytic ability was significantly improved over that of the control. Immune response to bacterial infection in flies involves two processes: phagocytosis and antimicrobial peptide (AMP) production (Lemaitre and Hoffmann, 2007). Phagocytosis is executed by adult hemocytes (plasmatocytes), which are macrophage-like blood cells (Crozatier and Meister, 2007), whereas fat body tissue produces the bulk of AMPs (although other tissues also contribute) (Brennan and Anderson, 2004). Given the central role of S6k in protein translation (Ma and Blenis, 2009) it is conceivable that variation in S6k could influence both the production of AMPs as well as phagocytosis. Activation of the mammalian homolog of S6k (p70S6 kinase) has been shown to play a direct role in phagocytosis in mouse macrophages (Ganesan et al., 2004), presumably by influencing the actin cytoskeleton (Berven et al., 2004), remodeling of which is important for phagocytosis. The fact that flies with the hypomorphic mutation of S6k had an improved phagocytic ability may have more to do with an indirect effect of this mutation. S6k plays a central role in protein translation in response to nutrient availability, so dysregulation of this gene or reduced functional expression of this gene may mimic the effects of reduced nutrient availability to the cell. Mammalian macrophages reared in nutrient-limiting conditions in cell culture have enhanced phagocytic ability, which is presumably related to the enhanced autophagy that cells undergo when facing nutrient stress (Martinet et al., 2009). Extending these observations to our study, one hypothesis is that flies with the mutation in S6k have reduced production of AMPs but enhanced phagocytic ability. As the difference in phagocytic abilities between control and mutant flies was small, relative to the difference in bacterial clearance ability between these genotypes, it could be that the reduced production of AMPs offset the benefits of enhanced phagocytosis. Of course, because our clearance phenotype is measured 24 h after infection, the relative influence of phagocytosis, production of AMPs, as well as other aspects of the fly physiology that might influence the immune response are confounded. Recent studies have shown an important interaction between the processes of phagocytosis and antimicrobial protein production in invertebrates, suggesting that phagocytosis is most important immediately after infection while the production of AMPs playing a role only after phagocytic cells are saturated (Elrod-Erickson et al., 2000; Haine et al., 2008). It should also be noted that the bacterial clearance assay was carried out on adults, but the phagocytosis assay was carried out using larval blood cells. Although we expect the phagocytic ability of larval and adult blood cells to be similar, given the fact that adult blood cells appear to be derived from larval blood cells (Minakhina and Steward, 2010), we cannot be certain of this without further study. Exactly how allelic variation in S6k influences the various traits that determine immunocompetence remains to be determined.

In summary, we have shown that natural alleles of S6k have extensive pleiotropic effects on traits related to fitness, and that these effects are highly dependent on the sex of the individual, the diet they are reared on, as well as age. A note of caution should be added with respect to this conclusion. Several of the quantitative complementation tests were only marginally significant; given that we tested a number of traits in different combinations of sex, age, and diet, we might expect to identify some significant relationships based on chance alone. However, for some traits we found a repeated association between variation at S6k and phenotypic variation in different treatment combinations (e.g., glycogen and the immune response assays), which greatly strengthens our conclusions for those traits. In addition, we were able to independently validate the influence of S6k on immune response traits. Future research is necessary to validate the effects of this gene on the other focal traits in this study.

An additional caveat is that we have not identified the actual causal polymorphism producing these different effects nor do we know for certain that alternative alleles of S6k with functional effects on the phenotypes evaluated are segregating in natural populations. The alleles tested here represent fixed differences in the S6k gene between two laboratory strains. Given the size of this gene (12 kb), there are likely to be many genetic differences between the tested strains at this locus alone, and thus it would be difficult to separate the effects of different variants with only two strains. Identification of the actual polymorphisms affecting these traits will require other approaches such as candidate gene association tests using a large number of individuals from a natural population or whole genome association tests, which may soon be possible in Drosophila using the Drosophila Genetic Reference Panel (

This is the first study that we are aware of to identify age- and diet-specific effects of natural alleles on fitness traits. As such, a great deal of future research is necessary to determine the mechanisms by which alleles exert environmental and age-specific effects on phenotypes. Such knowledge would enhance our understanding of the genetic basis of variation in life history traits and aging and provide a significant contribution to our understanding of phenotypic evolution.


Thanks to Allen Mosenkis and Ankita Tandon, who helped with life span assay. We also thank two anonymous reviewers, whose suggestions significantly improved the article. This work was supported by an NIH Grant 5R01HL80812 and NSF Grant DEB-0349856.

Disclosure Statement

No competing financial interests exist.


  • Arsham A.M. Neufeld T.P. Thinking globally and acting locally with TOR. Curr Opin Cell Biol. 2006;18:589–597. [PubMed]
  • Arsic D. Guerin P.M. Nutrient content of diet affects the signaling activity of the insulin/target of rapamycin/p70 S6 kinase pathway in the African malaria mosquito Anopheles gambiae. J Insect Physiol. 2008;54:1226–1235. [PubMed]
  • Berven L.A. Willard F.S. Crouch M.F. Role of the p70(S6K) pathway in regulating the actin cytoskeleton and cell migration. Exp Cell Res. 2004;296:183–195. [PubMed]
  • Bonneaud C. Mazuc J. Gonzalez G. Haussy C. Chastel O. Faivre B. Sorci G. Assessing the cost of mounting an immune response. Am Nat. 2003;161:367–379. [PubMed]
  • Brakefield P.M. Gems D. Cowen T. Christensen K. Grubeck-Loebenstein B. Keller L. Oeppen J. Rodriguez-Pena A. Stazi M.A. Tatar M. Westendorp R.G. What are the effects of maternal and pre-adult environments on ageing in humans, and are there lessons from animal models? Mech Ageing Dev. 2005;126:431–438. [PubMed]
  • Brennan C.A. Anderson K.V. Drosophila: the genetics of innate immune recognition and response. Annu Rev Immunol. 2004;22:457–483. [PubMed]
  • Charmantier A. Garant D. Environmental quality and evolutionary potential: lessons from wild populations. Proc R Soc B Biol Sci. 2005;272:1415–1425. [PMC free article] [PubMed]
  • Clark A.G. Keith L.E. Variation among extracted lines of Drosophila melanogaster in triacylglycerol and carbohydrate storage. Genetics. 1988;119:595–607. [PubMed]
  • Crozatier M. Meister M. Drosophila haematopoiesis. Cell Microbiol. 2007;9:1117–1126. [PubMed]
  • De Luca M. Yi N. Allison D.B. Leips J. Ruden D.M. Mapping quantitative trait loci affecting variation in Drosophila triacylglycerol storage. Obes Res. 2005;13:1596–1605. [PubMed]
  • De Luca M. Leips J. Mapping genetic polymorphisms affecting natural variation in Drosophila longevity. Methods Mol Biol. 2007;371:307–320. [PubMed]
  • Demas G.E. Chefer V. Talan M.I. Nelson R.J. Metabolic costs of mounting an antigen-stimulated immune response in adult and aged C57BL/6J mice. Am J Physiol. 1997;273:R1631–R1637. [PubMed]
  • Diangelo J.R. Bland M.L. Bambina S. Cherry S. Birnbaum M.J. The immune response attenuates growth and nutrient storage in Drosophila by reducing insulin signaling. Proc Natl Acad Sci USA. 2009;106:20853–20858. [PubMed]
  • Elrod-Erickson M. Mishra S. Schneider D. Interactions between the cellular and humoral immune responses in Drosophila. Curr Biol. 2000;10:781–784. [PubMed]
  • Ganesan L.P. Wei G. Pengal R.A. Moldovan L. Moldovan N. Ostrowski M.C. Tridandapani S. The serine/threonine kinase Akt promotes Fc gamma receptor-mediated phagocytosis in murine macrophages through the activation of p70S6 kinase. J Biol Chem. 2004;279:54416–54425. [PubMed]
  • Grewal S.S. Insulin/TOR signaling in growth and homeostasis: a view from the fly world. Int J Biochem Cell Biol. 2009;41:1006–1010. [PubMed]
  • Gutzeit H.O. Zissler D. Grau V. Liphardt M. Heinrich U.R. Glycogen stores in mature ovarian follicles and young embryos of Drosophila: ultrastructural changes and some biochemical correlates. Eur J Cell Biol. 1994;63:52–60. [PubMed]
  • Haine E.R. Moret Y. Siva-Jothy M.T. Rolff J. Antimicrobial defense and persistent infection in insects. Science. 2008;322:1257–1259. [PubMed]
  • Hamilton W.D. The moulding of senescence by natural selection. J Theor Biol. 1966;12:12–45. [PubMed]
  • Hansen I.A. Attardo G.M. Roy S.G. Raikhel A.S. Target of rapamycin-dependent activation of S6 kinase is a central step in the transduction of nutritional signals during egg development in a mosquito. J Biol Chem. 2005;280:20565–20572. [PubMed]
  • Harada H. Andersen J.S. Mann M. Terada N. Korsmeyer S.J. p70S6 kinase signals cell survival as well as growth, inactivating the pro-apoptotic molecule BAD. Proc Natl Acad Sci USA. 2001;98:9666–9670. [PubMed]
  • Hoffmann A.A. Merila J. Heritable variation and evolution under favourable and unfavourable conditions. Trends Ecol Evol. 1999;14:96–101. [PubMed]
  • Hughes K.A. Leips J. Quantitative trait locus analysis of male mating success and sperm competition in Drosophila melanogaster. Evolution. 2006;60:1427–1434. [PubMed]
  • Johnson M.B. Butterworth F.M. Maturation and aging of adult fat body and oenocytes in Drosophila as revealed by light microscopy morphometry. J Morphol. 1985;184:51–59. [PubMed]
  • Kapahi P. Zid B.M. Harper T. Koslover D. Sapin V. Benzer S. Regulation of lifespan in Drosophila by modulation of genes in the TOR signaling pathway. Curr Biol. 2004;14:885–890. [PMC free article] [PubMed]
  • Kong A.P. Chan N.N. Chan J.C. The role of adipocytokines and neurohormonal dysregulation in metabolic syndrome. Curr Diabetes Rev. 2006;2:397–407. [PubMed]
  • Leips J. Mackay T.F. Quantitative trait loci for life span in Drosophila melanogaster: interactions with genetic background and larval density. Genetics. 2000;155:1773–1788. [PubMed]
  • Lemaitre B. Hoffmann J. The host defense of Drosophila melanogaster. Annu Rev Immunol. 2007;25:697–743. [PubMed]
  • Long X. Muller F. Avruch J. TOR action in mammalian cells and in Caenorhabditis elegans. Curr Top Microbiol Immunol. 2004;279:115–138. [PubMed]
  • Ma X.J.M. Blenis J. Molecular mechanisms of mTOR-mediated translational control. Nat Rev Mol Cell Biol. 2009;10:307–318. [PubMed]
  • Manning B.D. Balancing Akt with S6K: implications for both metabolic diseases and tumorigenesis. J Cell Biol. 2004;167:399–403. [PMC free article] [PubMed]
  • Martinet W. Schrijvers D.M. Timmermans J.P. Herman A.G. De Meyer G.R. Phagocytosis of bacteria is enhanced in macrophages undergoing nutrient deprivation. Febs J. 2009;276:2227–2240. [PubMed]
  • Masoro E.J. Overview of caloric restriction and ageing. Mech Ageing Dev. 2005;126:913–922. [PubMed]
  • McKean K.A. Nunney L. Increased sexual activity reduces male immune function in Drosophila melanogaster. Proc Natl Acad Sci USA. 2001;98:7904–7909. [PubMed]
  • McManus M.G. Travis J. Effects of temperature and salinity on the life history of the sailfin molly (Pisces: Poeciliidae): lipid storage and reproductive allocation. Oecologia. 1998;114:317–325.
  • Medawar P.B. An Unsolved Problem of Biology. H.K. Lewis; London: 1952.
  • Meex R.C.R. Schrauwen P. Hesselink M.K.C. Modulation of myocellular fat stores: lipid droplet dynamics in health and disease. Am J Physiol Regul Integr Comp Physiol. 2009;297:R913–R924. [PubMed]
  • Minakhina S. Steward R. Hematopoietic stem cells in Drosophila. Development. 2010;137:27–31. [PubMed]
  • Montagne J. Stewart M.J. Stocker H. Hafen E. Kozma S.C. Thomas G. Drosophila s6 kinase: a regulator of cell size. Science. 1999;285:2126–2129. [PubMed]
  • Moret Y. Schmid-Hempel P. Survival for immunity: the price of immune system activation for bumblebee workers. Science. 2000;290:1166–1168. [PubMed]
  • Nuzhdin S.V. Pasyukova E.G. Dilda C.L. Zeng Z.B. Mackay T.F. Sex-specific quantitative trait loci affecting longevity in Drosophila melanogaster. Proc Natl Acad Sci USA. 1997;94:9734–9739. [PubMed]
  • Pasyukova E.G. Vieira C. Mackay T.F. Deficiency mapping of quantitative trait loci affecting longevity in Drosophila melanogaster. Genetics. 2000;156:1129–1146. [PubMed]
  • Pasyukova E.G. Roshina N.V. Mackay T.F.C. Shuttle craft: a candidate quantitative trait gene for Drosophila lifespan. Aging Cell. 2004;3:297–307. [PubMed]
  • Penninx B.W. Nicklas B.J. Newman A.B. Harris T.B. Goodpaster B.H. Satterfield S. de Rekeneire N. Yaffe K. Pahor M. Kritchevsky S.B. Metabolic syndrome and physical decline in older persons: results from the health, aging and body composition study. J Gerontol A Biol Sci Med Sci. 2009;64:96–102. [PMC free article] [PubMed]
  • Petschnigg J. Wolinski H. Kolb D. Zellnig G. Kurat C.F. Natter K. Kohlwein S.D. Good fat, essential cellular requirements for triacylglycerol synthesis to maintain membrane homeostasis in Yeast. J Biol Chem. 2009;284:30981–30993. [PMC free article] [PubMed]
  • Rose M.R. Evolutionary Biology of Aging. Oxford University Press; New York: 1991.
  • Ruvinsky I. Meyuhas O. Ribosomal protein S6 phosphorylation: from protein synthesis to cell size. Trends Biochem Sci. 2006;31:342–348. [PubMed]
  • Scott R.C. Schuldiner O. Neufeld T.P. Role and regulation of starvation-induced autophagy in the Drosophila fat body. Dev Cell. 2004;7:167–178. [PubMed]
  • Shima H. Pende M. Chen Y. Fumagalli S. Thomas G. Kozma S.C. Disruption of the p70(s6k)/p85(s6k) gene reveals a small mouse phenotype and a new functional S6 kinase. Embo J. 1998;17:6649–6659. [PubMed]
  • Skorupa D.A. Dervisefendic A. Zwiener J. Pletcher S.D. Dietary composition specifies consumption, obesity, and lifespan in Drosophila melanogaster. Aging Cell. 2008;7:478–490. [PMC free article] [PubMed]
  • Snedecor G.W. Cochran W.G. Statistical Methods. Iowa State University Press; Ames: 1980.
  • Stearns S.C. The Evolution of Life Histories. Oxford University Press [USA]; New York: 1991.
  • Stroschein-Stevenson S.L. Foley E. O'Farrell P.H. Johnson A.D. Identification of Drosophila gene products required for phagocytosis of Candida albicans. PLoS Biol. 2006;4:e4. [PMC free article] [PubMed]
  • Tanksley S.D. Mapping polygenes. Annu Rev Genet. 1993;27:205–233. [PubMed]
  • Terashima J. Bownes M. A microarray analysis of genes involved in relating egg production to nutritional intake in Drosophila melanogaster. Cell Death Differ. 2005;12:429–440. [PubMed]
  • Thibault S.T. Singer M.A. Miyazaki W.Y. Milash B. Dompe N.A. Singh C.M. Buchholz R. Demsky M. Fawcett R. Francis-Lang H.L. Ryner L. Cheung L.M. Chong A. Erickson C. Fisher W.W. Greer K. Hartouni S.R. Howie E. Jakkula L. Joo D. Killpack K. Laufer A. Mazzotta J. Smith R.D. Stevens L.M. Stuber C. Tan L.R. Ventura R. Woo A. Zakrajsek I. Zhao L. Chen F. Swimmer C. Kopczynski C. Duyk G. Winberg M.L. Margolis J. A complementary transposon tool kit for Drosophila melanogaster using P and piggyBac. Nat Genet. 2004;36:283–287. [PubMed]
  • Tucker M.Z. Turcotte L.P. Aging is associated with elevated muscle triglyceride content and increased insulin-stimulated fatty acid uptake. Am J Physiol Endocrinol Metab. 2003;285:E827–E835. [PubMed]
  • Valenzano D.R. Terzibasi E. Cattaneo A. Domenici L. Cellerino A. Temperature affects longevity and age-related locomotor and cognitive decay in the short-lived fish Nothobranchius furzeri. Aging Cell. 2006;5:275–278. [PubMed]
  • Vieira C. Pasyukova E.G. Zeng Z.B. Hackett J.B. Lyman R.F. Mackay T.F. Genotype-environment interaction for quantitative trait loci affecting life span in Drosophila melanogaster. Genetics. 2000;154:213–227. [PubMed]
  • Williams G.C. Evolution 11. (1957). Pleiotropy, natural selection and the evolution of senescence. Evolution. 1957;11:398–411.
  • Wu Q. Zhang Y. Xu J. Shen P. Regulation of hunger-driven behaviors by neural ribosomal S6 kinase in Drosophila. Proc Natl Acad Sci. 2005;102:13289–13294. [PubMed]
  • Yang L.H. Bastow J.L. Spence K.O. Wright A.N. What can we learn from resource pulses? Ecology. 2008;89:621–634. [PubMed]
  • Zeyda M. Stulnig T.M. Obesity, inflammation, and insulin resistance—a mini-review. Gerontology. 2009;55:379–386. [PubMed]

Articles from DNA and Cell Biology are provided here courtesy of Mary Ann Liebert, Inc.