Decreased heat-shock resistance and down-regulation of Hsp70 expression with increasing age in adult Drosophila melanogaster


  • J. G. Sørensen,

    Corresponding author
    1. Department of Genetics and Ecology, University of Aarhus, Bldg. 540, Ny Munkegade, DK-8000 Aarhus C, Denmark
      †Author to whom correspondence should be addressed. E-mail:
    Search for more papers by this author
  • V. Loeschcke

    1. Department of Genetics and Ecology, University of Aarhus, Bldg. 540, Ny Munkegade, DK-8000 Aarhus C, Denmark
    Search for more papers by this author

†Author to whom correspondence should be addressed. E-mail:


1. Adult Drosophila melanogaster Meigen were examined for Hsp70 expression level and heat-shock resistance from 0 to 8 days of age – the latter being close to the mean life span in natural D. melanogaster populations.

2. Over the 8 days, both induced Hsp70 expression level and induced heat-shock resistance showed a linear decrease. Heat-shock resistance without prior hardening also decreased, but at a lower rate than both induced Hsp70 expression level and heat-shock resistance after hardening.

3. The results suggest that the decrease in induced Hsp70 expression level in D. melanogaster with age is regulatory and not due to a reduced ability to express Hsp70 as a result of senescence.

4. The changes in induced Hsp70 expression level with age are likely to be determined by natural selection and are evolutionarily adaptive. The decrease with age in stress resistance after hardening may then be considered a direct consequence of age-dependent reduction in expression levels of induced Hsp70.


Susceptibility to thermal stress increases as an organism ages (Hollingsworth & Bowler 1966; Miquel et al. 1976; Niedzwiecki, Kongpachith & Fleming 1991). Two kinds of observations have led to the idea that changes in the expression of heat-shock proteins (Hsps) could be related to the changes in resistance with age in Drosophila melanogaster Meigen. First, the level of expression of Hsps is related to the ability of the organism to resist a variety of abiotic stresses (e.g. Feder et al. 1996; Dahlgaard et al. 1998) and second, a correlation between high resistance to different abiotic stresses and increased longevity has been found in several studies (Rose et al. 1992; Martin, Austad & Johnson 1996; Tatar 1999; F.M. Norry & V. Loeschcke, unpublished data). In Drosophila most studies addressing age-dependent stress resistance and protein expression have used an experimental design where individuals a few days old were compared with rather old ones, often more than 30 days old (Miquel et al. 1976; Niedzwiecki et al. 1991; Schwarze, Weindruch & Aiken 1998). From an evolutionary point of view changes occurring in 1-month-old or older flies seem irrelevant for explaining any evolutionary processes, since Drosophila flies are thought to have a life span considerably less than a month in nature (Rosewell & Shorrocks 1987; Turelli & Hoffmann 1995).

Here we report the results of a study of the possible relationship between expression of the inducible Hsp70, heat-shock resistance and age. To our knowledge, this is the first study where these traits have been investigated together within the time frame that can be considered as close to the natural life span of D. melanogaster in nature (Rosewell & Shorrocks 1987). The inducible Hsp70 of Drosophila was an obvious candidate protein for these investigations. It is the major Hsp of Drosophila, being massively expressed after exposure to many different stresses (reviewed in Ananthan, Goldberg & Voellmy 1986) and has been shown to be important for thermal resistance after hardening in adults (Dahlgaard et al. 1998). The Hsp70 family consists of several proteins in Drosophila (Lindquist & Craig 1988) of which Hsp70 is purely inducible and not present at all in the absence of stress (Velazquez et al. 1983). By measuring the expression of the inducible Hsp70 only, we were able to avoid any confounding effects caused by differential changes in expression by different Hsp70 family members. A change in induced Hsp70 expression level has earlier been observed as an evolutionary response to selection for heat resistance (Bettencourt, Feder & Cavicchi 1999; Sørensen et al. 1999; Lansing, Justesen & Loeschcke 2000) and in wild-caught heat-adapted Drosophila populations (Sørensen, Dahlgaard & Loeschcke 2001).

Hsp70 expression levels after induction and heat-stress resistance both with and without prior heat hardening were measured. This allowed us to investigate whether the decrease in heat resistance with age is related to a possible change in the induced expression level of Hsp70. Alternative scenarios could be expected. In the first, resistance decreases slowly with age and the level of Hsp70 expression after induction is either constant or increases with age. As aged flies in general are more susceptible to stress than young flies, a hardening treatment would be a more severe stress for the aged flies. If this is so, induced Hsp70 expression level could increase with age, as induced Hsp70 expression has been shown to be proportional to stress intensity (DiDomenico, Bugaisky & Lindquist 1982). Owing to the higher expression, heat-stress resistance may be maintained until a certain age where increased expression of Hsp70 can no longer compensate for the general decreased physiological performance (Fleming, Reveillaud & Niedzwiecki 1992). The second scenario suggests that resistance and induced Hsp70 expression decrease jointly with age. One explanation for decreased resistance with age could be a decreased expression level of protective systems, including Hsps, which would be reflected in a decreased level of induced Hsp70 expression. This could be the result of a general decreased ability to express proteins in older flies so that reduced Hsp70 expression level could be viewed as a measure of metabolic capability. Alternatively, the energetically expensive Hsp system could be adaptively down-regulated, to conserve energy for other purposes, directly causing the decrease in stress resistance with age.

Materials and methods

Origin and maintenance of flies

Drosophila melanogaster were collected in October 1997 from two populations in Denmark, one located at Hov in Jutland and the other at Hvidovre on Zealand. The flies were maintained as isofemale lines until February 1998 (27 isofemale lines from Hov, and 30 isofemale lines from Hvidovre), when all the lines from both populations were mixed and maintained as one interbreeding population in high numbers.

Experimental design

Parental flies were set up in eight 200-ml bottles with standard medium (sugar, yeast, oatmeal and agar) under uncrowded conditions with 20 pairs per bottle laying eggs for 24 h. Flies were set up in bottles every second day over a 10-day period so that newly emerged flies could be collected every second day. Upon emergence virgin flies less than 6 h old were collected, sexed under CO2 anaesthesia and maintained at a density of 25 flies per vial. Flies were transferred to fresh food vials every other day. Vials with virgin flies aged 0 (less than 6 h), 2, 4, 6 and 8 days were thus available. All age groups were treated at the same time to reduce experimental variation. Vials from each age group were randomly assigned to the three different experiments: induced Hsp70 expression level, heat-shock resistance after hardening and heat-shock resistance without hardening.

Heat hardening

Hardening took place in glass vials with thin foam stoppers in the bottom to prevent flies from getting stuck on the glass. The top-stoppers were moistened with tap water and inserted fully to ensure nearly saturated humidity in order to avoid desiccating the flies. The vials were evenly spaced in racks that were put in preheated waterbaths, with the water level exceeding the top-stopper’s lower end. The hardening treatment was 35 ± 0·1 °C for 1 h. Flies were subsequently kept at 25 °C for 1 h before being either heat-shocked for testing of heat resistance or frozen at −70 °C for Hsp70 measurement.

Hsp70 expression

Flies assigned to Hsp70 measurement were homogenized and the level of Hsp70 was measured using the enzyme-linked immunosorbent assay (ELISA) protocol described in Dahlgaard et al. (1998) and Sørensen et al. (1999) applying a monoclonal antibody (7.FB), specific for the inducible Hsp70 of Drosophila. Hsp70 level was calculated from four replicate ELISA plates. One replicate sample of each treatment and sex was represented on each of the four plates. Variation between plates was corrected by adjusting plate grand means to the grand mean of plate one. For D. melanogaster the spectrometric response has been shown to be linear with the increase in Hsp70 within a broad range of concentrations (Dahlgaard et al. 1998; Lansing et al. 2000).

Heat-shock resistance

Two heat-shock resistance traits were measured. These were survival after heat-shock with or without prior heat hardening at 35 °C for 1 h. Heat-stress resistance after hardening was measured by exposure to 39 °C or 39·5 °C for 55 min. In the heat-stress resistance experiment with no prior hardening the vials were put directly to 38 °C for 55 min. Within an hour of heat shock, flies were transferred to fresh vials, and allowed to recover for 24 h before being scored. Flies were considered dead if they were unable to walk after being touched lightly with a brush.

Preferably the same conditions (temperatures and durations) should have been used for survival after heat-shock with and without prior heat hardening. This is not possible, though, as the temperatures used after hardening would give around 100% mortality in all age groups not hardened, and the temperature used for the non-hardened would give around 0% mortality in hardened flies.


The age-dependent changes in induced Hsp70 expression level and heat-stress resistance were analysed with GLM using age as a covariate. Analysis was omitted in the 39·5 °C heat-stress experiment as only two data points had values above zero. Since survival after heat-shock was calculated as proportions, the data were arcsin-square-root transformed to improve homogeneity of variances. All statistics were performed with the computer program package SPSS (SPSS 1998).


Hsp70 expression

The level of Hsp70 expression after hardening (35 °C for 1 h) decreased linearly over the 8-day period (Fig. 1; GLM: Females b = −0·029 ± 0·005, P < 0·001; Males b = −0·027 ± 0·005, P < 0·001). At day 8 the expression level had decreased to about 25% of that observed in newly emerged flies (less than 6 h old). No differences were found between males and females except in 4-day-old flies where the expression level of females exceeded that of males.

Figure 1.

Hsp70 expression after hardening in arbitrary units (means ± SE) in adult D. melanogaster females (open dots) and males (filled dots). The Hsp70 expression level after induction at 35 °C are given for flies from 0 to 8 days old.

Heat-shock resistance

After hardening (35 °C for 1 h), survival to a heat-shock of 39 °C for 55 min declined from day 2 to day 8 (Fig. 2). No differences were found between the sexes. Resistance was very high in 0- and 2-day-old flies with the survival being approximately 0·85 at both ages. Thereafter, survival decreased resulting in a survival below 0·10 for both sexes in 8-day-old flies. Over all ages the survival rate decreased linearly (GLM: Females b = −0·11 ± 0·009, p < 0·001; Males b = −0·12 ± 0·009, P < 0·001).

Figure 2.

Survival rate after hardening (35 °C) and heat shock at 39 °C (means ± SE) in adult D. melanogaster females (open dots) and males (filled dots) for flies from 0 and 8 days old. The Y-axis scale is arcsin-square-root transformed.

For the heat-shock at 39·5 °C, survival was approximately 0·90 in flies 0 day old and resistance decreased faster than after the heat-shock at 39 °C. In 2-day-old flies the survival had decreased to approximately 0·15 for males and to 0 for females and by day 4 no flies survived the stress treatment (Fig. 3). No measurements were made on flies older than 6 days of age, as no survival was expected after that time.

Figure 3.

Survival rate after hardening (35 °C) and heat shock at 39·5 °C (means ± SE) in adult D. melanogaster females (open dots) and males (filled dots) for flies from 0 and 8 days old. The Y-axis scale is arcsin-square-root transformed.

Without hardening, survival to a heat-shock at 38 °C for 55 min was approximately 0·95 in flies less than six hours old and approximately 0·90 in flies 2 days old for both sexes (Fig. 4). From day 4 to 8 survival seemed to have reached a plateau. In this period the survival of males (0·70) was slightly higher than that of females (0·50). Over all measurements the survival rate decreased linearly (GLM: Females b = −0·063 ± 0·01, P < 0·001; Males b = −0·037 ± 0·008, P < 0·001).

Figure 4.

Survival rate after heat-shock at 38 °C (means ± SE) without hardening in adult D. melanogaster females (open dots) and males (filled dots) for flies from 0 and 8 days old. The Y-axis scale is arcsin-square-root transformed.


Previous work has shown that heat-stress resistance of newly emerged adults can be very high compared with older flies (Krebs, Feder & Lee 1998; Stratman & Markow 1998; Feder 1999). This was also found in this study, where survival to a potentially lethal heat stress was almost 100% in all treatments of flies less than 6-hours-old. The mechanism causing this high resistance in freshly emerged flies is unknown. The high resistance was found in both heat-hardened and non-hardened flies, suggesting that the high heat-stress resistance is under developmental control. As pupae and freshly emerged adults are immobile or have limited mobility, behavioural avoidance is not possible and these life stages are therefore particularly vulnerable to environmental stress. Therefore high resistance could be especially important in these life stages (Krebs & Loeschcke 1995).

The observations made in this study suggest that heat-shock resistance with and without hardening are two independent traits. This is not surprising since Hsps are strongly involved in resistance after hardening and less so in resistance without hardening, as induced Hsp70 expression is limited at the extreme ‘high end’ of the temperature scale (Krebs 1999). It seems as if heat-shock resistance after hardening decreased with age faster than did resistance without hardening, probably due to decreased expression of inducible Hsp70.

Why should Hsp70 expression and thereby thermal resistance after hardening decrease so rapidly with age? The hsp70 genes are tightly regulated and not expressed unless induced by stress (DiDomenico et al. 1982; Velazquez et al. 1983). Hsp70 is important for stress resistance after hardening in adult Drosophila (Feder & Hofmann 1999), but costly in terms of energy expenditure and fitness (Krebs & Loeschcke 1994; Feder 1999). Therefore, the expression level of Hsp70 after induction is thought to be tightly regulated by selection (Bettencourt et al. 1999; Sørensen et al. 1999, 2001; Lansing et al. 2000). Once sexual maturity is reached, the Hsp system might be evolutionarily dispensable as increased fitness might be attained by spending less energy on protection and more on reproduction.

Some studies have addressed the expression of Hsps and stress resistance in relation to ageing (Niedzwiecki et al. 1991; Wheeler, Bieschcke & Tower 1995; Bonelli et al. 1999; Rao, Watson & Jones 1999; Wheeler, King & Tower 1999). Of the several Hsps studied, attention has especially focused on Hsp70, which is the major Hsp in Drosophila. No consistent results have been found in these studies, even though most studies detected a decreased expression level at very old ages. For instance, the induced expression level of Hsp70 in adult D. melanogaster was increased in 28-day-old individuals and decreased in 47-day-old individuals as compared to flies 1–4 days old (Niedzwiecki et al. 1991). The pattern of increased Hsp70 expression in aged organisms seems to contradict our results. We suggest that this can be explained by the mechanism regulating Hsp70 expression. First, there is the inducible expression after stress exposure (as measured in this study), which occurs in high quantities and is connected to thermal resistance. Second (as suggested by Wheeler et al. 1999) there is an expression that occurs in aged individuals without heat- or other stress exposures. The expression level of Hsp70 in old flies without stress induction is very low. This can be seen for example in Wheeler et al. (1999) who used 10–20 times more protein per sample to get detectable bands in old flies than we use for Western blotting with the same antibody in heat-treated younger flies. The age-induced expression of Hsp70 is probably caused by accumulation of abnormal protein (Wheeler et al. 1999). No such accumulation or effects of senescence are expected in flies between 0 and 8 days.

Niedzwiecki et al. (1991) found the heat resistance of D. melanogaster to be constant until the age of 35 days and decreasing thereafter, and in this study heat-stress resistance already decreased between day 0 and day 2. Different measures of heat resistance have yielded different results, suggesting that different measures of heat-stress resistance are at least partly independent (Hoffmann et al. 1997; Sørensen et al. 2001). However, both Niedzwiecki et al. (1991) and this study measured the same heat resistance trait: survival after exposure to high temperature. Perhaps the resistance level was relatively high for a long time in the study of Niedzwiecki et al. (1991) because the stress-temperature used in their treatment was only 37 °C. The age at which a certain decrease in survival after hardening is attained strongly depends on the stress intensity as seen in this study, where a large difference in survival was observed after stress at 39 °C or 39·5 °C. The more severe the stress the steeper is the decrease in stress resistance with age. By using a rather high stress level it was shown that a decrease in resistance with age occurred from the time of emergence. Moreover, a decrease in inducible Hsp70 with age seemed to follow closely the decrease in resistance after hardening, suggesting that a down-regulation of induced Hsp70 expression was involved in the stress resistance decrease.


We are grateful to Dr Susan Lindquist for kindly providing the 7.FB antibody and to J.S.F. Barker, M. Hercus, F. Norry, R. Bijlsma, R. Krebs, M. Holmstrup and two anonymous referees for helpful comments on the ms. We also thank Trine Gammelgaard for technical help, the Carlsberg Foundation (No. 990338) and the Danish National Research Council for financial support, and the Centre for Environmental Stress and Adaptation Research for their hospitality while the final version of the ms was written. V.L. also thanks the Institute for Advanced Study at La Trobe University for providing excellent working conditions.