Steroid Control of Longevity in Drosophila melanogaster

See allHide authors and affiliations

Science  28 Feb 2003:
Vol. 299, Issue 5611, pp. 1407-1410
DOI: 10.1126/science.1080539


Ecdysone, the major steroid hormone ofDrosophila melanogaster, is known for its role in development and reproduction. Flies that are heterozygous for mutations of the ecdysone receptor exhibit increases in life-span and resistance to various stresses, with no apparent deficit in fertility or activity. A mutant involved in the biosynthesis of ecdysone displays similar effects, which are suppressed by feeding ecdysone to the flies. These observations demonstrate the importance of the ecdysone hormonal pathway, a new player in regulating longevity.

In humans, changes in steroid hormones occur during aging (1), but whether those changes are a cause or an effect of aging remains unclear. To investigate the role of steroids in the aging process, we used genetics to manipulate a steroid hormone in adult Drosophila flies.

Steroid hormones in insects are ecdysteroids, and the major form in Drosophila is ecdysone. Its active metabolite, 20-OH-ecdysone, is important in developmental transitions and metamorphosis in Drosophila melanogaster (2). Ecdysone is also involved in oogenesis in the adult fly, but other functions are not known (2, 3). 20-OH-ecdysone circulates and binds to a heterodimeric nuclear receptor consisting of an ecdysone receptor (EcR) and Ultraspiracle (USP), a homolog of the retinoid X receptor (RXR) (2). In the absence of ecdysone, the EcR-USP heterodimer is thought to form a complex with one or more corepressor proteins (N-CoR and SMRT), which bind to chromosomal histone deacetylases (Sin3A/Rpd3) (2, 4). When the ligand is bound to the receptor, the complex binds instead to coactivators that recruit histone acetyltransferases, thus activating the transcription of various genes, including transcription factors (5), chaperones (6), apoptosis genes (7), and catalase (8). We investigated the role of ecdysone during adulthood by studying flies with mutations in EcR (9) and in a gene involved in ecdysone biosynthesis, DTS-3 (10).

The EcR gene encodes three isoforms. We first studied a mutant, EcRV559fs , which has a 37–base pair deletion in the predicted ligand-binding domain in a region common to the three isoforms (9). It is homozygous lethal during development but adult viable as a heterozygote.EcRV559fs/+ flies lived longer than the controls (Fig. 1); male and female average life-spans increased by 40 to 50%. This was true for heterozygous offspring of crosses between two independentcinnabar brown (cn bw) backgrounds: the parental line from the Bender laboratory and a cn bw from our own laboratory stock. The same increase in longevity was observed, regardless of whether the male or female parents were mutant forEcRV559fs . Progeny of the two cn bwlines crossed with each other showed no differences in longevity from the two parental lines.

Figure 1

Extension of life-span in an ecdysone receptor heterozygous mutant. Survival curves for cn EcRV559fsbw/cn EcR + bw males and females, and cn EcR + bw controls raised at 25°C are shown. The cn, EcR, and bwmutations are all on the second chromosome. Male and female heterozygotes showed a 45% increase in mean life-span (P < 0.004; Wilcoxon rank test, one tail). Results are averages of four replicates of 40 flies per run ± SD. Flies were collected 2 to 3 days after adult eclosion to allow time for mating. All of the experiments were done using “standard” food (22).

Developmental time and weight of the adult flies inEcRV559fs/+ were equivalent to those of control flies (Fig. 2, A and B). However,EcRV559fs/+ flies showed increased resistance to three stresses: oxidative challenge, heat, and dry starvation (Fig. 2C). The EcRV559fs/+ were also more active than controls, as measured by their performance in fast phototaxis, a good indicator of stamina (Fig. 2D). The increase in longevity is thus not linked to a reduced activity level.

Figure 2

Phenotypic characteristics ofcn EcRV559fs bw/cn EcR + bw. (A) Developmental time. The cumulative emergence of adult progeny as a function of time ± SD is shown. Two crosses were made, each in triplicate: female and male cn EcR + bw crossed to each other, and femalecn EcR + bw crossed to cn EcRV559fs bw/SM6G males. cn EcR + bw progeny are represented as actual numbers divided by two, because the cross with the EcRmutant produces two different genotypes. (B) Weight. Young flies were collected as in Fig. 1, with four replicates for each cross. Each point corresponds to the average weight per fly ± SD. (C) Resistance to various stresses. Bars represent the percentage increase in mean life-span compared to parental cn bwcontrols ± SD (n = number of runs × number of flies per run). (D) Phototaxis. The performance index is the total number of positive responses in five trials, divided by five multiplied by the number of flies. Each bar shows the average performance index of 14 runs ± SD. For all tests, the cn EcRV559fs bw/cn EcR+ bw flies were in acn bw background; the controls were cn bw. Mated flies were collected as in Fig. 1.

Life-span can also be extended in Drosophila when the costs of reproduction are eliminated (17). Because ecdysone is required for normal oogenesis (2, 3) and spermatogenesis (12), reduced hormone function could increase survival through changes in reproduction. However, heterozygous EcR mutants showed no defect in oogenesis (3). Age-specific fecundity and fertility were greater in the mutant flies than in the controls (Fig. 3). Indeed, heterozygous mutant males seemed to induce greater fertility in their female mates. Thus, the increased life-span caused by reduced EcR activity did not lead to a loss in reproductive output.

Figure 3

Reproductive ability of EcRV559fs/+ flies. Data are expressed as mean ± SD. (A) Fecundity and fertility measured under conditions similar to those of the life-span experiments in Fig. 1. In curve 1, male cn EcRV559fs bw/cn EcR+ bw were crossed with female cn EcRV559fs bw/cn EcR + bw for 3 days, then separated. The comparison is to curve 2, which shows parental male cn EcR + bw crossed with female cn EcR + bw. (B andC) Fecundity and fertility measured in a different genetic background. (B) In curve 1, malew1118CS10 flies (w1118 was previously crossed 10 times with Canton-S) were crossed with female cn EcRV559fs bw/cn EcR + bw lifelong, as compared to curve 2, in which male w1118CS10 were crossed with female cn EcR + bw. (C) In curve 1, male cn EcRV559fs bw/cn EcR + bw were crossed with femalew1118CS10 flies lifelong and compared to curve 2, in which male cn EcR+ bw/cn EcR+ bw were crossed with femalew1118CS10 flies.w1118CS10 , crossed to itself lifelong (23), gave the same result as male cn EcR+ bw crossed with femalew1118CS10 flies.

Other mutations in the predicted ligand-binding domain (EcRA483T and EcRM554fs ), in the DNA binding domain (EcRF288Y andEcRC300Y ), in the Canton-S background, and in the cn bw background (9) also showed increased longevity (fig. S1, C to F). Note that the life-span of cn bw homozygotes is essentially the same as for wild-type Canton-S and for w1118 10 times backcrossed to Canton-S (fig. S1, C to F). The possibility of a deleterious background in thecn bw controls is also counterindicated by the fact that the heterozygous mutant female progeny of the cross between thermosensitiveEcRA483T and cn bw controls do not live longer than the cn bw controls at 25°C. In summary, all of the mutant alleles of EcR that were tested as heterozygotes, in the ligand-binding domain and in the DNA binding domain, showed increased longevity.

Because ecdysone receptor mutants resulted in increased longevity, we also investigated a mutant affecting the ecdysone biosynthetic pathway. DTS-3, which was induced in the wild-type Samarkand strain, is a dominant lethal during development at an elevated temperature (10). The gene encodes a protein with Krüppel Zn-finger domains (13) that is specifically involved in ecdysone biosynthesis, because its developmental defect can be rescued when flies are fed 20-OH-ecdysone (10,14). After a few days at restrictive temperature (29°C),DTS-3/+ female adults, but not males, have a 50% lower ecdysteroid titer and reduced fertility (10). FemaleDTS-3/+ adults showed a temperature-dependent increase in longevity (Fig. 4A). At 29°C, females showed an increase of mean life-span of 42%. Consistent with the earlier observation that the ecdysteroid titer is reduced only in females, males did not show significantly increased life-span at any of the temperatures tested (fig. S2A). To control for genetic background, we outcrossed DTS-3 twice with our w1118 line. Similar increases in longevity persisted (fig. S2B).

Figure 4

DTS-3, a temperature-sensitive mutant affecting ecdysone synthesis in females. (A) Increase in life-span for DTS-3/+ females at higher temperature. Crosses to study longevity, as in Fig. 1, were made at 25°C (a temperature permissive for development). (B andC) Effect of shifting DTS-3/+ females from high to low temperature at various times during adulthood. Bars represent averages ± SD of four runs of 40 flies for the 25° to 20°C experiments (B) and four runs of 25 flies for the 29° to 20°C experiments (C). Black bars, DTS-3/+; gray bars, Samarkand. The percentage increase in mean life-span is shown above the bars. (D) Feeding 20-OH-ecdysone toDTS-3/+. Survival curves at restrictive temperature (29°C) with different concentrations of 20-OH-ecdysone in the food are shown. Females were placed in vials with 3 ml of food supplemented with 10−3 to 10−5 M 20-OH-ecdysone (20-OH-E) in 1% ethanol. In all cases, flies were allowed to develop at 25°C (a temperature permissive for development), until 2 to 3 days of adulthood before testing; DTS-3/+ was compared to the Samarkand (Sam) wild-type control, because the mutant allele was generated in a Samarkand background. Mated flies were collected as inFig. 1.

An increase in resistance to dry starvation was also linked to the temperature at which the flies were maintained before testing. After being exposed at 29°C, DTS-3/+ females showed an increase in mean survival time at 25°C of about 33%, and the result was the same when the flies were outcrossed with the w1118 line (fig. S2C). Again, males showed no changes in resistance (Fig. S2C). For DTS-3/+ females, spending only the 10 to 11 first days of adulthood at restrictive temperatures (29° or 25°C) was sufficient to induce an increase in resistance (Fig. 4, B and C). More strikingly, if the DTS-3/+ females spent the same period at 25°C before being shifted to 20°C, they lived longer than if they had spent their entire lives at 20°C; after 20 days at 25°C, the subsequent life-span at 20°C was as if the flies had not aged during the time spent at the higher temperature.

When ecdysone was fed to the DTS-3/+ females throughout adulthood at 29°C, the increase in longevity was reduced, in a dose-sensitive manner (Fig. 4D). Feeding 20-OH-ecdysone to the heterozygous flies also reversed the increased survival under dry starvation conditions (fig. S2D). At the ecdysone concentrations used (10−5 to 10−3 M) in the standard food, there was no substantial effect on life-span in wild-type Samarkand females, which discounts a possible toxic effect of the hormone.

Our experiments show that two different components of a single pathway, the ligand and the receptor, can be manipulated to extend longevity. Hormones are known to have differential effects on males and females, and we also observed such differences with the ecdysone pathway. We have also shown, in one mutant studied in more detail, that fertility and activity levels can be raised, despite an increase in life-span.

Another hormonal pathway that is important for longevity regulation in diverse model organisms is the insulin pathway (15). Although the steroid and insulin pathways have very different roles, there is evidence that they do interact (16,17). In the fly, juvenile hormone (JH) is also involved in aging (18). Because USP has been proposed as a receptor for JH (19), steroids and JH might interact in regulating life-span.

The mutations in our study may be changing the balance between repression and activation of transcription of various target genes. Extension of longevity by gene silencing has been shown inDrosophila through mutation in Rpd3, another histone deacetylase interacting with the EcR-USP complex (20), or by feeding the flies phenylbutyrate, an inhibitor of histone deacetylase that induces one spectrum of genes while repressing others (21).

But which of the numerous genes that EcR regulates are crucial for the observed effects? Some of them, such as chaperones and catalase, are known to be important for longevity. Intrinsic levels of steroids may optimize certain functions but be detrimental to others, so life-span extension may require fine-tuning of the expression of appropriate combinations of genes.

Supporting Online Material

Materials and Methods

SOM Text

Figs. S1 and S2


  • * To whom correspondence should be addressed. E-mail: benzer{at}


Stay Connected to Science

Navigate This Article