• Open Access

Delayed behavioural aging and altered mortality in Drosophilaβ integrin mutants


Mike Grotewiel, Michigan State University, S-306 Plant Biology Building, 178 Wilson Road, East Lansing, MI 48824–1312, USA. Tel.: +1 517 3535554; fax: +1 517 4327120; e-mail: grotewie@msu.edu


The genetic basis for aging is being intensely investigated in a variety of model systems. Much of the focus in Drosophila has been on the molecular–genetic determinants of lifespan, whereas the molecular–genetic basis for age-related functional declines has been less vigorously explored. We evaluated behavioural aging and lifespan in flies harbouring loss-of-function mutations in myospheroid, the gene that encodes βPS, a β integrin. Integrins are adhesion molecules that regulate a number of cellular processes and developmental events. Their role in aging, however, has received limited attention. We report here that age-related declines in locomotor activity are ameliorated and that mean lifespan is increased in myospheroid mutants. The delayed functional senescence and altered mortality in myospheroid flies are independent of changes in body size, reproduction or stress resistance. Our data indicate that functional senescence and age-dependent mortality are influenced by β integrins in Drosophila.


Several laboratories have made substantial progress toward understanding molecular–genetic determinants of lifespan in Drosophila. Mutations in methuselah (mth) (Lin et al., 1998), Insulin-like receptor (InR) (Tatar et al., 2001), chico (Clancy et al., 2001), Indy (Rogina et al., 2000) and the ecdysone receptor (Simon et al., 2003) cause substantial increases in maximum lifespan of the fly. Ectopic expression of superoxide dismutase (Orr & Sohal, 1994; Sun & Tower, 1999; Sun et al., 2002) and peptide methionine sulphoxide reductase (Ruan et al., 2002) also extend lifespan in Drosophila. Similarly, flies fed 4-phenylbutyrate, a histone deacetylase inhibitor, exhibit increased maximum lifespan (Kang et al., 2002). Thus, loss-of-function mutations, ectopic expression of repair systems and drug treatment can extend lifespan in Drosophila. Intriguingly, the severity of age-related locomotor declines is reduced in flies overexpressing superoxide dismutase with catalase (Orr & Sohal, 1994) or methionine sulphoxide reductase (Ruan et al., 2002), and in flies fed 4-phenylbutyrate (Kang et al., 2002). These studies indicate that ectopic gene expression and drug treatment can protect flies from age-related declines in locomotor activity. Mutation of mth, however, does not appear to confer protection from age-related behavioural declines (Cook-Wiens & Grotewiel, 2002). It is unclear therefore whether loss-of-function mutations can protect Drosophila from functional senescence.

In this report we describe our studies on the role of a β integrin in functional senescence and longevity in Drosophila. Integrins are cell surface receptors that mediate a number of biological processes including development, inflammation, tumour metastasis and wound healing, (Hynes, 1992). In Drosophila, two β integrins, βPS (Leptin et al., 1989) and βν (Yee & Hynes, 1993), have been identified. βPS, encoded by the X-linked myospheroid (mys) gene (MacKrell et al., 1988), appears to be the major integrin β subunit in flies. βPS is important for a number of developmental events including attachment of the somatic musculature to the embryonic body wall and maintenance of the structural integrity of the dorsal epithelium during embryogenesis (MacKrell et al., 1988). Additionally, condensation of the larval nervous system (de la Pompa et al., 1989), development of the wing (Brower & Jaffe, 1989) and development of the eye (Zusman et al., 1990) also depend on βPS. Whereas the role of βPS in embryonic and larval development has been studied extensively, its function in adults remains largely unexplored.

We find that loss-of-function mutations in mys protect flies from age-related defects in locomotor behaviour. Additionally, mortality during the first half of the fly lifespan is decreased in mys mutants, causing significant increases in mean lifespan. Our results indicate that normal aging in Drosophila depends on β integrin function and establish that loss-of-function mutations can retard functional senescence.


As part of our ongoing analysis of integrins and behaviour, we noticed that aged flies harbouring mutations in mys seemed more vigorous than control animals. To test explicitly the possible role of the mys gene product (βPS) in aging, we assessed functional senescence and longevity in a series of mys mutants. We focused on three alleles of mys: mysxG, mysnj and mysts2. mysxG is a recessive lethal allele that acts as a genetic null in complementation tests for viability (Bunch et al., 1992) (data not shown). This allele harbours a small deletion in exon five that causes a frame-shift and premature truncation of the predicted βPS protein (Jannuzi et al., 2002) (Fig. 1A). Additionally, mRNA expression from this allele was significantly reduced on Northern blots (Fig. 1B). mysnj acts as a strong hypomorphic allele with diminished viability when in trans to mysxG (Bunch et al., 1992) (data not shown). This allele also expressed reduced amounts of mys mRNA (Fig. 1B), although it appears to encode a wild type protein (D. Brower, personal communication). mysts2 acts as a weak hypomorphic allele with good homozygous viability. Viability of mysts2/mysxG transheterozygotes, however, is reduced at elevated temperatures (Bunch et al., 1992) (data not shown). This allele contains a point mutation causing a G→D substitution at amino acid 347 in the extracellular domain of βPS (D. Brower, personal communication) (Fig. 1A). This collection of molecular and classical genetic studies confirms that these three alleles are loss-of-function mutations in mys.

Figure 1.

The mys locus and alleles used. (A) Schematic of the mys gene. Exons are represented by boxes; introns are represented by single lines that connect exons. Filled boxes represent protein coding regions. The mysxG allele is a 113-bp deletion (represented as a bar above the gene schematic) within exon 5 that causes a frame shift with misincorporation of 18 amino acids and premature truncation of βPS (Bunch et al., 1992; Jannuzi et al., 2002). The mysts2 allele (weak hypomorph) is a point mutation that causes a G→D substitution at amino acid 347 (D. Brower, personal communication). The molecular lesions in mysts2 and mysxG were confirmed by sequencing of PCR products (data not shown). (B) Northern analysis of mys gene expression. Adult whole body RNA from 0.5-week-old control (lane 1), mysnj/+ (lane 2) and mysxG/+ (lane 3) was hybridized with probes derived from mys or rp49 (loading control) cDNAs. The mys and rp49 probes hybridized to single bands at 4400 and 600 nucleotides, respectively. Quantification revealed that mys RNA expression was reduced 15% in mysnj/+ and 40% in mysxG/+. Data are representative of two independent experiments.

To explore the role of βPS in functional senescence, we assessed control flies and mys mutants as they aged using a negative geotaxis assay. Behavioural performance in these assays was recorded as the net vertical distance walked by flies after they were tapped to the bottom of a plastic cylinder (Cook-Wiens & Grotewiel, 2002). The ability of wild-type flies to perform this task decreases markedly with age (Miquel et al., 1976; Orr & Sohal, 1994; Sun & Tower, 1999; Cook-Wiens & Grotewiel, 2002), making it a useful measure of age-related functional decline.

mys is an essential gene on the X chromosome. Consequently, we focused on female flies in our experiments because more genotypes that are viable are possible in females than in males. Control females and females heterozygous for mysxG (null) performed well in negative geotaxis assays at a young age (0.5 weeks) (Fig. 2A). Geotaxis in control flies decreased dramatically with age as expected. By contrast, mysxG/+ flies exhibited significantly elevated geotaxis performance relative to age-matched control animals (Fig. 2A). Similarly, whereas females heterozygous for mysnj (strong hypomorph, Fig. 2B) and females homozygous for mysts2 (weak hypomorph, data not shown) performed indistinguishably from control flies at 0.5 weeks of age, these mutants performed significantly better at greater ages (Fig. 2B). Thus, age-dependent loss of geotaxis was ameliorated in three independent alleles of mys. These data indicate that βPS, the mys gene product, plays a role in senescence of locomotor skills in Drosophila.

Figure 2.

Delayed locomotor senescence in mys mutants. Negative geotaxis was assessed in female mys mutants as a function of age. Overall, locomotor activity was significantly greater in mysxG/+ (A) and mysnj/+ (B) mutants than in control animals (two-way anova, P < 0.0001 for effect of age and genotype, n = 17–25 per genotype at each age). Locomotor activity was indistinguishable in 0.5-week-old control and mys flies of either genotype (Bonferroni's multiple comparison test, ns). By contrast, age-specific geotaxis was significantly better in mysxG/+ (A; 2.5 to 8.5 weeks old) and mysnj/+ (B; 2.5 to 6.5 weeks old) than in control flies (*Bonferroni's multiple comparison test, P < 0.05). Data are mean ± SEM. Pilot studies indicated that young homozygous females harbouring mysnj exhibited defects in geotaxis (data not shown), confounding their use in tests for age-related declines in locomotor skills.

To determine whether the delayed functional senescence in mys animals is associated with altered lifespan, we assessed survival of mys mutants under normal housing conditions at 25 °C. Mean lifespan was increased 20% in females heterozygous for the mysxG allele (Fig. 3A). Females heterozygous for mysnj exhibited a comparable increase in mean lifespan (Fig. 3A). Mean lifespan was also increased 20% in mysts2 homozygous females (Fig. 3B). Maximum lifespan was extended 5–10% in these three genotypes, but this change was not statistically significant. Consistent with the increases in mean lifespan, mortality during the first several weeks of life was reduced and the onset of age-dependent mortality appeared to be delayed in mys mutants (Fig. 3C,D). The increase in age-specific mortality, however, appeared to be greater in mys mutants than in control animals. The effects of mys mutations on survival are consistent with trade-offs between enhanced early adult fitness and exacerbated mortality later in life. Together, our behavioural and survival analyses indicate that the beneficial effects of mys mutations are manifested largely during the first half of the fly lifespan.

Figure 3.

Increased mean lifespan and altered mortality in mys mutants. Survival of females with the indicated genotypes was assessed at 25 °C. Genotype had a significant effect on survival of mysxG/+ and mysnj/+ females (A) and mysts2 homozygous females (B) (log-rank test, chi-square statistic; P < 0.001). Mean lifespan was increased 20% in these three mys genotypes. Data are representative of 1–5 experiments (nine in total) that began with 200–300 flies per genotype in each experiment. The control flies for the experiment in panel A had a relatively high level of age-independent mortality; mysxG/+ flies, however, exhibited a similar pattern of mean lifespan extension in three independent experiments in which control animals had a more rectangular survival curve (data not shown), confirming that the effects of mys mutations on lifespan are not artefacts associated with unusually high age-independent mortality in the controls. Females homozygous for mysnj had no increase in mean lifespan (data not shown). (C,D) Mortality in mys mutants. Mortality was determined from life tables for the experiments shown in panels A and B. Data are presented as log mortality for mysxG/+ and mysnj/+ (C) and mysts2 (D).

Many genetic manipulations that extend lifespan in C. elegans (Kenyon, 2001) and Drosophila (Lin et al., 1998; Luckinbill, 1998; Parkes et al., 1998; Arking et al., 2000) enhance survival under stressful environmental conditions. To determine whether the delayed functional senescence and altered mortality in mys mutants were associated with changes in stress resistance, we subjected control and mys mutant females to a battery of stress tests. Survival of mys mutants in the presence of paraquat (a free-radical generator) (Fig. 4A,B), in conditions of starvation (Fig. 4C,D) and under desiccation (Fig. 4E,F) were largely unaltered. Survival of heat stress (37 °C) and hyperoxia was similarly unchanged in mys mutants (data not shown). These studies demonstrate that mys mutants do not have increased resistance to stress, indicating that the delayed functional senescence and altered mortality in mys animals are not due to enhanced protection from environmental stressors experienced under normal housing conditions.

Figure 4.

Stress resistance in mys mutants. Survival of flies in the presence of paraquat (A,B), under starvation (C,D) and under desiccation (E,F) was assessed in 0.5-week-old animals of the indicated genotypes. Stress resistance in mys flies was largely indistinguishable from that of control animals. In some experiments, the mysts2 flies exhibited a decreased resistance to paraquat and desiccation, but this result was not consistently obtained. Data are representative of 2–5 experiments with each stressor that began with 80–100 flies per genotype in each experiment.

Decreases in body mass and female reproduction are associated with extended lifespan in Drosophila (Clancy et al., 2001; Tatar et al., 2001), although these changes do not appear to be required for enhanced longevity (Tatar et al., 2003). We assessed these parameters to determine whether the delayed functional declines and altered mortality in mys animals were associated with reductions in body size or reproductive status. We found that control flies and mys mutants had indistinguishable whole body masses (Fig. 5A). We also found that the mys mutants had similar rates of egg-laying (Fig. 5B) and laid similar (mysnj/+) or elevated (mysxG/+ and mysts2) total numbers of eggs (Fig. 5C) as compared with control animals. The changes in functional senescence and mortality in mys mutants therefore are not secondary to diminutive body size or reduced reproduction.

Figure 5.

Body mass and egg-laying in mys mutants. (A) Genotype had no effect on whole body mass in 0.5-week-old flies [one-way anova, ns, n = 40 (control), 15 (mysnj/+), 15 (mysxG/+), and 25 (mysts2)]. Data (mean ± SEM) are compiled from three independent experiments. (B) The aggregate weekly rates of egg-laying derived from egg counts at 12-h intervals were not substantively altered in mys mutants. (C) Overall, genotype had a significant effect on the total number of eggs laid during the first 10 weeks of adulthood [one-way anova, P = 0.0075, n = 15 (15 vials with three flies each)]. Relative to control flies, the total number of eggs laid was not changed in mysnj/+ flies and was significantly increased in mysxG/+ and mysts2 animals (Dunnett's multiple comparison test, P < 0.05). Data (mean ± SEM) are representative of three independent experiments.


Our studies reveal that senescence of locomotor function is delayed and age-specific mortality is reduced in mys mutants, indicating that β integrins influence these key aspects of aging in Drosophila. The underlying changes in physiological status occur in the absence of altered stress resistance and do not correlate with diminished body mass or reduced egg-laying. Thus, altered aging in mys mutants does not appear to be related to enhanced protection from stress or to obvious trade-offs in body size or reproduction. As integrins are found throughout the animal kingdom (Hynes, 1992), we speculate that they might be part of a conserved genetic pathway that impacts aging in animals. Thus, it would be interesting to determine whether mutations in β integrin genes in C. elegans or mice produce similar changes in functional senescence and mortality.

Mutations in mys have rather complex effects on mortality that result in increased mean lifespan, but subtle changes in maximum lifespan. Age-independent mortality is consistently lower and the onset of age-dependent mortality is delayed in mys animals. Curiously, mys mutants exhibit an increase in the rate of age-dependent mortality after its onset. The beneficial effects of mys mutations during the first few weeks of life therefore are followed by exacerbated age-dependent mortality. The mys mutants appear to trade early adult fitness for greater age-dependent mortality during later life.

Our studies show that protection from age-related functional declines can occur via loss-of-function mutations in an endogenous gene. This indicates that the mys gene is required for the normal progression of age-related locomotor declines in Drosophila. We speculate that other genes have similar influences on age-related behavioural changes. For example, it would be interesting to determine whether mutations in InR, chico and Indy might also retard behavioural aging in flies.

Flies are protected from age-related declines in geotaxis by simultaneous overexpression of Cu2+/Zn2+-superoxide dismutase and catalase (Orr & Sohal, 1994) and when fed 4-phenylbutyrate, a histone deacetylase inhibitor (Kang et al., 2002). Similarly, ectopic expression of peptide methionine sulphoxide reductase confers protection from physical decline as measured in a related behavioural assay (Ruan et al., 2002). These transgenic and drug-treated flies are resistant to environmental stressors (Kang et al., 2002; Ruan et al., 2002) or exhibit reduced amounts of oxidatively damaged proteins as they age (Orr & Sohal, 1994), suggesting that the preservation of physical status in these animals is related to protection from oxidative damage or stress. Our studies show that retarded functional senescence can occur in the absence of overt changes in resistance to stress, including oxidative stress. Thus, protection from age-related locomotor declines in flies does not appear to be dependent on enhanced protection from environmental stress, consistent with our previous studies on the longevity/stress-resistant mutant mth (Cook-Wiens & Grotewiel, 2002). Our interpretation of these results is that oxidative stress might not be a causal factor in age-related behavioural declines in flies. Interestingly, chico mutant flies exhibit extended lifespan without enhanced stress survival (Clancy et al., 2001). Thus, it is possible experimentally to uncouple stress survival from both delayed behavioural aging and lifespan extension in flies, suggesting that augmentation of neither of these aging parameters is dependent on enhanced stress resistance.

Other interpretations of our findings are possible. For example, behavioural aging and stress survival might not correlate because the tissues involved in behavioural aging under normal housing conditions are distinct from those that support survival under conditions of explicit environmental stresses. This possibility would allow for enhanced stress resistance in key behavioural tissues but not in tissues that mediate stress survival, leading to the uncoupling seen in our studies. Additionally, it is possible that behavioural aging and stress survival are supported by the same tissues, but are differentially sensitive to stress. In this case, the level of stress resistance conferred on a tissue might be sufficient to ameliorate age-related behavioural changes, but inadequate to alter stress survival significantly. Nevertheless, our studies suggest that behavioural aging in flies might not be related to damage due to stress.

Loss-of-function mutations in mth (Lin et al., 1998), InR (Tatar et al., 2001), chico (Clancy et al., 2001), Indy (Rogina et al., 2000) and the ecdysone receptor pathway (Simon et al., 2003) extend lifespan in Drosophila. Although it is possible that mys functions within one of these pathways, the aging phenotype of mys flies is distinct from that of previously characterized mutants. In contrast to mth flies (Lin et al., 1998; Cook-Wiens & Grotewiel, 2002), mys mutants exhibit delayed locomotor senescence and normal stress resistance. Additionally, mys mutants have normal body mass and reproduction, whereas InR (Tatar et al., 2001) and chico (Clancy et al., 2001) flies are dwarf and lay few eggs. Furthermore, maximal lifespan is dramatically extended in mth, InR, chico, Indy and ecdysone receptor mutants, whereas maximum lifespan is only subtly changed in mys flies. Thus, the aging phenotype of mys flies is unique, suggesting that integrins might represent a novel pathway to influence aging.

The molecular–genetic basis for aging is being intensely studied in a variety of whole-animal model systems. The bulk of the work in these studies has focused on understanding the regulation of lifespan, whereas functional senescence has received comparatively little attention. Our studies showing that declines in locomotor function are substantially ameliorated by mutations in mys are a step toward filling the broad gap between our understanding of age-related functional declines and that of lifespan determination.

Experimental procedures

Fly stocks and fly husbandry

All flies used were grown on a standard sucrose–cornmeal–agar medium at 25 °C/65% relative humidity under a 12 : 12-h light–dark cycle. The control strain for these studies was a line harbouring the w1118 allele backcrossed to Canton-S for 10 generations; this is also known as w[cs] (Cook-Wiens & Grotewiel, 2002). All mys alleles (kindly provided by N. Brown, Cambridge University, and D. Brower, University of Arizona) were backcrossed to this control strain for six generations and then maintained as homozygous stocks (mysts2 and mysnj) or as a balanced stock over FM6 (mysxG). The mysnj and mysxG designations are synonymous with mysnj42 and mysxG43, respectively (Lindsley & Zimm, 1992). For all studies, bottles were seeded with parents such that larval crowding was avoided and that comparable numbers of adults emerged in all bottles. For geotaxis and longevity studies, 1- to 3-day-old adult flies were anaesthetized briefly with CO2, gender separated and then housed in fresh food vials at 25 flies per vial. Flies for negative geotaxis and longevity studies were transferred to new food vials every 3–4 days throughout the experiments according to a prescribed schedule.

Molecular analyses

Northern analyses were performed on total RNA isolated from whole 1- to 3-day-old adult females using standard techniques (Stoltzfus et al., 2003). The mys probe was generated by labelling a mys+ cDNA (kindly provided by D. Brower, University of Arizona). Autoradiographs were quantified using a Personal Molecular Imager FX and Quantity One v4.2.2 software (Bio-Rad, Hercules, CA, USA). The molecular lesions in mysts2 (D. Brower, personal communication) and mysxG (Bunch et al., 1992; Jannuzi et al., 2002) were confirmed by PCR amplification using primers 5′-ATTGGGCGGTGTGATTGC-3′ and 5′-TTGTCCTTCATCTCCACCG-3′ (mysts2) or primers 5′-GCCAGGAGTCCAACGATACA-3′ and 5′-CTTGTAGATGGGATTCTCGC-3′ (mysxG) followed by DNA sequencing.

Negative geotaxis assays

Negative geotaxis was assessed as previously described (Cook-Wiens & Grotewiel, 2002; Stoltzfus et al., 2003). Flies of the appropriate ages were briefly anaesthetized with CO2, sorted singly into food vials, and allowed to recover overnight (∼18 h) at 25 °C/65% relative humidity. Negative geotaxis behaviour was recorded as the net distance (cm) climbed by individual flies in a vertical polystyrene tube during a 10-s test that began immediately after being tapped to the bottom of the tube. The geotaxis scores of non-performing flies (flies that failed to climb) were excluded from the data reported in the main text. Although the number of non-performing flies was not significantly different between control and mys mutants (data not shown), there was a trend for the controls to have more non-performing animals at advanced ages than did mys animals. Data were analysed using Prism 3.0 (GraphPad Software, San Diego, CA, USA).

Longevity studies

Female adult flies (1–3 days old) were collected as above and then housed at 25 flies per vial under a 12 : 12-h light–dark cycle and 65% relative humidity/25 °C. Surviving flies were counted every 3–4 days after being transferred to fresh food vials. Flies that escaped, became trapped in the food or were physically damaged during the experiments were censored. Statistical analyses on survival data were performed using JMP (SAS Institute Inc., Cary, NC, USA). Analysis of mortality was performed as described (Lin et al., 1998) using a five-cell moving average to smooth the mortality data prior to log transformation. Each experiment started with 200–300 flies per genotype.

Analyses of stress resistance

For all stress resistance studies, 1- to 3-day-old adults were briefly anaesthetized with CO2, collected into food vials and allowed to recover overnight (∼18 h) at 25 °C/65% relative humidity. Flies for the paraquat resistance experiment were subsequently deprived of food and water for 6 h and then transferred to food vials that were empty except for a disc of filter paper saturated with a 5% sucrose solution containing 20 mm paraquat (Lin et al., 1998; Zou et al., 2000; Cook-Wiens & Grotewiel, 2002; Ruan et al., 2002). Flies in paraquat vials were housed at room temperature in a humidified atmosphere and assessed for survival every few hours until all flies were dead. To evaluate resistance to starvation, flies were collected as above, transferred to vials containing 1% agar without sucrose, housed at 25 °C/65% relative humidity and assessed for survival every few hours until all flies were dead (Clark & Fucito, 1998; Lin et al., 1998; Harshman et al., 1999; Clancy et al., 2001). Desiccation resistance was performed on flies collected as above and then housed at 25 °C in empty food vials with 2.5 cm of desiccant above the cotton plug in each vial (Rose et al., 1992; Force et al., 1995; Harshman et al., 1999). Resistance to thermal stress (Clark & Fucito, 1998; Lin et al., 1998) and hyperoxia (Orr & Sohal, 1993; Mockett et al., 1999, 2001) were performed essentially as described. Survival data were analysed using JMP (SAS Institute Inc.).

Determination of whole body mass

One- to 3-day-old flies were collected with brief CO2 anaesthesia, placed into 1.5-mL tubes and weighed in groups of 10. Data were analysed using Prism 3.0 (GraphPad Software).

Egg-laying studies

Three virgin adult females (1–2 days old) were placed together in each food vial and mated to control males for 3 days. On day 3, the flies were anaesthetized, the males were discarded and the females transferred to new food vials. The number of eggs was counted every 12 h throughout the experiment including during the 3 days of mating. Neither the genotype of the male nor the length of time the males were available to the females significantly affected the relative number of eggs laid by different genotypes (data not shown). Data were collected as the number of eggs per female per unit time for individual vials and then compiled to generate weekly egg-laying rates or the total number of eggs laid during the experiment. Data were analysed using Prism 3.0 (GraphPad Software).

Supplementary material

Life tables from survival analyses are available from http://www.blackwellpublishing.com/products/journals/suppmat/ACE/ACE060/ACE060sm.htm.


We thank Chris Garth for expert technical assistance, Danny Brower, Nick Brown and Ron Davis for fly stocks, Scott Pletcher for advice on mortality analyses, and Danny Brower, Laurent Seroude and Sarah Elsea for helpful comments on the manuscript. This work was supported by N.I.H. grants MH60787 and AG21199 to M.S.G.