SEARCH

SEARCH BY CITATION

Keywords:

  • chill coma;
  • cold tolerance;
  • cross-tolerance;
  • desiccation;
  • heat shock;
  • high temperature;
  • laboratory selection;
  • longevity;
  • rapid cold hardening;
  • starvation

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References

Abstract  Cross tolerance, whereby tolerance to one environmental stress is correlated with tolerance to other stressors, is thought to be widespread in insects. We used lines of Drosophila melanogaster Meigen (Diptera: Drosophilidae) selected for survival at a 1-h exposure to −5°C to examine the extent to which this selection results in increased tolerance to other stresses, including high and low temperatures, desiccation and starvation. While selection improved tolerance to acute cold exposure and survival at −5°C, there was little effect of selection regime on tolerance to other stressors. There was no correlation between tolerances to any of the stressors, suggesting different mechanisms of tolerance. This supports arguments that correlations between stress tolerances during selection experiments with D. melanogaster may be coincidental. The magnitude of heat-hardening was apparently constrained by basal tolerance among lines, but the magnitude of the rapid cold-hardening response was not correlated with basal cold tolerance, implying that the relationship between inducible and basal tolerances differs at high and low temperatures.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References

In many cases, insects that tolerate a particularly harsh environment may have significantly increased tolerance to a number of stressors. For example, the freeze-tolerant gall fly Eurosta solidaginis is not only extremely cold-tolerant, but also very tolerant of desiccation (Ramløv & Lee, 2000). Cross-tolerance – where tolerance to one environmental stressor results in increased tolerance to another – is thought to be widespread in insects, particularly with regards to low temperatures and desiccation, longevity and starvation tolerance and desiccation and high temperatures (Bayley et al., 2001; Harshman & Hoffmann, 2000; Ring & Danks, 1994; Vermeulen & Loeschcke, 2007; Wu et al., 2002). This could arise from the similarity in mechanisms governing tolerance to multiple stressors (e.g., maintenance of membrane or protein conformation; regulation of metabolic reserves [Gibbs et al., 1997]), or because genetic loci associated with tolerance of different stressors are linked (Phelan et al., 2003). Conversely, conflicting mechanisms for different types of stress tolerance may lead to trade-offs; for example, low and high temperature tolerances may require opposing levels of membrane fluidity or for them to be maintained (Kellett et al., 2005; Overgaard et al., 2006).

Experimental tests of cross-tolerance in insects have generally focused on Drosophila melanogaster Meigen (Diptera: Drosophilidae) (e.g., Gibbs, 1999, but see Sinclair, 2000 and Bayley et al., 2001). The ease of rearing and artificially selecting Drosophila allows tests of cross-tolerance by selecting for tolerance of one stress and examining its influence on tolerance of others. Discordance in results between apparently similar selection experiments cast some doubt on the usefulness of artificial selection experiments (Harshman & Hoffmann, 2000). However, these discordances may in themselves be informative. For example, Bubliy and Loeschcke (2005) conducted reciprocal experiments examining correlations between tolerance to several environmental stressors, and found a correlation between cold and desiccation resistance in D. melanogaster. However, in spite of good a priori reasons for expecting a correlation between these tolerances (Ring & Danks, 1994), Sinclair et al. (2007b) found no increase in cold tolerance in desiccation-selected D. melanogaster. One explanation for this is that while increased cold tolerance may result in increased desiccation tolerance, the reverse does not apply. This may be because the primary factors associated with desiccation tolerance are not advantageous at low temperatures, but the primary factors associated with cold tolerance are useful in desiccating conditions (but are not normally targets during desiccation selection). Desiccation and selection result in different gene expression profiles in D. melanogaster (Sinclair et al., 2007a). In the case of desiccation and cold, further cross-tolerance experiments (beginning with flies selected for cold tolerance) not only provide a further test of the relationship between selected traits, but allow the resolution of outstanding questions about the nature, as well as the direction, of relationships in cross-tolerance (in this case desiccation and cold).

There are a number of ways to measure tolerance to environmental stress – for example, Sinclair and Roberts (2005) itemized 27 different metrics from the literature, each called ‘cold tolerance’. The relationships between these different metrics are not always clear – continuing with the cold example, measures of acute cold tolerance (survival of a few hours at a low sub-zero temperature, such as −5°C) may be measuring a different set of physiological responses or tolerances than measures of chill coma recovery (Sinclair & Roberts, 2005). However, the relationships among different measures of cold tolerance have not often been examined. As well as examining cross-tolerances between different environmental stressors, variation among lines in selection experiments can be used to determine the correspondence between different metrics of response to the same environmental stress. This approach may therefore reveal the extent to which the mechanisms underlying the different metrics are shared. There was no correlation between high temperature knockdown resistance and high temperature mortality within D. birchii and D. serrata (Berrigan & Hoffmann, 1998) or D. melanogaster (Folk et al., 2006; Hoffmann et al., 1997). By contrast, Hori and Kimura (1998) reported that cold knockdown and lethal temperature are correlated among eight species in the D. melanogaster group, and Berrigan (2000) reported that high temperature LT50 (the lethal temperature at which 50% of individuals are killed by a stress), knockdown time and knockdown temperature were correlated among six species of Drosophila. Finally, Anderson et al. (2005) found that selection for chill coma recovery resulted in enhanced survival at −2°C in D. melanogaster, suggesting at least some relationship exists between these two measures of ‘cold tolerance’.

Previous experiments that have used artificial selection to explore cross-tolerance between tolerance to cold and other environmental stressors have used long-term, mild, cold exposure (Bubliy & Loeschcke, 2005), chill coma onset or recovery (Anderson et al., 2005), or have examined cold tolerance responses of lines selected for other traits, like desiccation (Sinclair et al., 2007b). Here we used lines of flies selected for more than 70 generations for tolerance of −5°C for 1 h – with and without a hardening pre-treatment – and assayed the ability to survive (or recover from) a variety of environmental and other stresses. Because the selection treatments generated between-line variation in most traits, this design allowed us not only to examine the effect of selection on other traits, but also to examine the extent to which tolerances are correlated between environmental stresses.

Using these selected lines, we were able to test the following hypotheses:

  • (i) 
    Constitutive acute cold tolerance is part of a connected suite of stress tolerances. As a result, we predict that selection for increased cold tolerance should result in increased tolerance to other stressors, particularly desiccation, starvation and stresses associated with ageing.
  • (ii) 
    Inducible cold tolerance (such as that imparted by rapid cold hardening [RCH, Lee et al., 1987]) results from upregulation of pathways not associated with basal tolerances. As a result we predict that a selection regime that includes a pre-treatment should have no effect on basal tolerances to cold or other stressors.
  • (iii) 
    Upper and lower thermal tolerances require conflicting changes to membrane composition and fluidity. As a result, we predict a trade-off between basal tolerances to high and low temperatures across lines of flies.
  • (iv) 
    After Kellett et al. (2005), we hypothesize that hardening capacity should be constrained by basal thermal tolerance at both high and low temperatures. Thus, we predict that, among lines, there will be a negative relationship between LT50 and magnitude of LT50 change with heat hardening (i.e., more basally heat-tolerant flies show reduced magnitude of hardening), and a positive relationship between acute cold tolerance (as LT50) and magnitude of hardening (i.e., improved basal cold tolerance results in a reduced RCH response).
  • (v) 
    Different measures of ‘cold tolerance’ are correlated. We used three different measures of cold tolerance: temperature survived for a fixed acute exposure, time survived at a fixed low temperature and chill coma recovery time after 4 h at 0°C. If most measures of ‘cold tolerance’ are equivalent, we predict that selection for acute cold tolerance should result in an increase in other measures of cold tolerance as well, and that there would be positive correlations between all three measures among lines of flies.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References

Base population

Selected lines of Drosophila melanogaster were developed in, and obtained from, the Chippindale Laboratory at Queen's University (Kingston, ON, Canada) and were originally derived from the LHM population (described in Chippindale & Rice, 2001) and subdivided (to form subgroups 1, 2 and 3) in January of 2004. These subgroups were each divided into four selection regimes (see below) and artificially selected every generation from generations 1 through 44, every other generation for generations 44 through 65 and every generation again from generation 65 onwards (upon transfer to the University of Western Ontario). The flies were reared at 25°C (12: 12 L: D) and 50% relative humidity (RH) on a molasses-cornmeal-yeast (12: 12: 5 v/w/w) medium.

Selection regimes

Selection was conducted on the 11th day of the lifecycle during which 20 vials from each line (≈40 adult flies in each) were transferred to empty 35-mL vials and a foam stopper was pushed down to restrict the flies to the bottom 25 mm of the vial. Acclimation (A) lines were exposed for 2 h at 0°C, followed by 1 h at −5°C. Zero (Z) lines were exposed to 2 h at 0°C (as for the A lines), but with no subsequent cold exposure. Freeze (F) flies were exposed for 1 h at −5°C without any pretreatment, and the control group (C) remained at 25°C. Thus, we selected for acute tolerance to −5°C with (A), and without (F) a pretreatment known to induce RCH (Czajka & Lee, 1990), and controls for handling (C) and the RCH pretreatment (Z). After the selection treatment, all flies were returned to the rearing temperature (25°C) and transferred to 3.75 L population cages, where the survivors were given 48 h to recover before eggs were collected to establish the next generation. Eggs were placed into vials with approximately 10 mL of food medium at a density of 75–100 eggs per vial. The F1 line was lost in August 2005, giving a current total of 11 lines (3 each of A, C and Z, 2 of F).

Prior to use in experiments, populations of 500–2000 flies from each line were removed from the selection regime for one generation to avoid maternal effects (Crill et al., 1996). The resulting offspring were thus two generations removed from selection when used in experiments. Experiments were conducted on flies derived from generations 70 (chill coma recovery and longevity), 71 (acute low-temperature tolerance), 72 (desiccation and starvation tolerance), 73 (survival at −5°C) and 75 (high-temperature tolerance). Following CO2 anesthesia and segregation, flies were given a 48 h recovery period (Nilson et al., 2006) and used in experiments at 4 days of adult age.

Body size and water content were assessed gravimetrically using a Mettler UMX-5 microbalance. Frozen flies were weighed (± 1 μg) before (fresh mass), after overnight drying at 60°C (dry mass) and water content calculated as (fresh mass − dry mass).

Acute low-temperature tolerance

Non-virgin, 4-day-old flies were segregated into groups of 10 individuals per vial (n= 3 vials per sex/line/temperature/pre-treatment combination, 924 vials total). For each of seven experimental exposures (2 h at 0, −1, −2, −3, −4, −5, and −6°C), half the vials were exposed to a RCH pretreatment (2 h in an ice-water bath at 0°C) followed by transfer to the test temperature, while the other half were used to assess basal cold tolerance (immediate transfer to the test temperature). The vials, in racks, were held by lead weights in a pre-cooled plastic container with 50 mm deep solution of industrial grade calcium chloride (Sifto Road Salt, Mississauga, ON, Canada). The plastic container was placed in a Sanyo MR-153 incubator (Sanyo Scientific, Bensenville, IL, US) set to maintain the temperature of the solution at the experimental temperature as monitored by a thermocouple. The entire space in which the flies were confined was below the surface of the salt solution, which served to buffer any temperature variations in the incubator. Following the treatment, flies were transferred into 6-well tissue culture plates with a 1 cm3 piece of food to assess recovery. Survival was assessed after 24 h. Flies were considered ‘alive’ if they could stand and walk in a co-ordinated fashion.

Survival time at −5°C

Survival time at −5°C was assessed in the same manner as acute cold tolerance, with the exception of exposure temperatures and times. Vials of flies were exposed to −5°C for eight time periods from 20 to 160 min (n= 3 vials per sex/line/temperature/pre-treatment combination, 1056 vials total). Half of the vials were exposed to an RCH pre-treatment followed by immediate transfer to −5°C, while the other half were placed directly into the −5°C test condition. Survival was assessed after 24 h.

High-temperature tolerance

Non-virgin flies were segregated into groups of 10 individuals per vial (n= 3 vials per sex/line/temperature/pre-treatment combination, 1056 vials total). For each of the eight experimental temperatures (37, 38, 38.5, 39, 39.5, 40, 41 and 42°C) half of the vials were exposed to 36.5°C for 1 h to elicit a heat shock response followed by 1 h at 25°C to allow transcription of heat shock proteins (Pardue et al., 1992), while the other half were used to estimate basal heat tolerance. All of the vials were then placed at the experimental temperature for 1 h. Exposures were conducted with vials inverted to prevent flies from adhering to the food. Following exposure, vials were moved to 25°C and survival was assessed after 24 h as above.

Chill coma recovery time

Chill coma recovery was assessed as per Nilson et al. (2006). Virgin flies were segregated into groups of 10 individuals per vial (n= 10 vials per sex/line combination). Four days after segregation, vials of flies were transferred into empty polypropylene cryovials (2 mL, VWR International, Mississauga, ON, Canada) and placed into resealable bags in an ice bath at 0°C for 4 h. Following the exposure, flies were quickly transferred from cryovials to recovery dishes with a 1 cm3 piece of food containing twice the amount of agar as in the standard food medium to prevent flies in chill coma from sticking to the food. Recovery – the ability of the fly to right itself and stand – was assessed every 60 s until all of the flies had recovered, and the time required for 80% of each group of 10 flies to recover was used in analyses (Nilson et al., 2006).

Desiccation resistance

Experimental flies were placed under CO2 anesthesia and segregated into clean, empty 35-mL vials (n= 10 individuals per sex/line combination). Desiccation resistance was determined as described in Gefen et al. (2006). Briefly, a short foam stopper was pushed into the vial to restrict the 10 individuals to the bottom half of the empty vial. Two grams of silica gel (ScholAR Chemistry, West Henrietta, NY, US, 6–12 mesh) was added above the stopper of each vial, and the top of the vial was sealed with Parafilm (Alcan Inc., Neenah, WI, US). Vials were then placed in an incubator at 25°C. Mortality – a fly that was unable to right itself – was assessed hourly until all flies were dead.

Starvation resistance

Starvation resistance was determined using the method of Bubliy and Loeschcke (2005). Experimental flies were anesthetized using CO2 and individually segregated into vials containing 10 mL of 2% agar to prevent desiccation (n= 10 individuals per sex/line combination). The vials were then placed into an incubator at 25°C, 50% RH. The presence of the agar likely raised the humidity inside the vial, and also provided a source of ingestible water (but not food) for the flies. Mortality was assessed every 4 h until the first death was observed, and then hourly until all flies were dead.

Longevity

Virgin flies were placed under CO2 anesthesia within 8 h of eclosion and were individually segregated into vials containing 10 mL food medium (n= 10 vials per sex/line combination). The vials were kept in an incubator at 25°C, 50% RH (12: 12 L: D). Mortality was assessed daily and surviving individuals were transferred without anesthesia into fresh food vials bi-weekly.

Statistical analyses

Because of the loss of the F1 line, we had reduced power to detect differences between selection regimes per se. As a result, we perform all analyses looking for differences between lines, utilizing post hoc tests to search for consistent patterns in tolerances between lines that may arise from the selection regimes. Lower lethal temperature, survival at −5°C and heat tolerance data were compared among lines, sexes and pre-treatments using generalized linear models with a probit link function and binomial error distribution using PROC GENMOD in SAS (SAS/STAT v. 9.1, SAS Institute Inc., Cary, NC, US). Although measures of centrality (e.g. LT50) are presented in graphs, statistical comparisons reflect differences among models, not among those values. Comparisons of dry mass, longevity, starvation resistance, chill coma recovery time and desiccation resistance among lines, sexes and treatments were made using a nested general linear model and Tukey's post hoc tests using PROC GLM on SAS.

Correlations between stress tolerances among lines were examined using LT50 for acute cold tolerance, survival at −5°C and heat tolerance, and means for other experiments. Mean dry mass and mean body water content were used to examine correlations between body size and water content and stress tolerance. Correlation between magnitude of hardening (difference in LT50 of pre-treated vs. controls) and basal tolerance (LT50 of controls) was also examined. Correlations were performed separately for each sex using Spearman's rank correlation (PROC CORR on SAS). False Discovery Rate correction (PROC MULTTEST on SAS) was applied to each correlation matrix (Benjamini & Hochberg, 1995). Correlations were subjected to a power analysis in R Statistica Software (R version 2.4.1, 2006) to determine the number of samples (selected lines) required for a significant correlation at 0.8 power.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References

Body size

None of the selection regimes resulted in a consistent change in body size; however, flies from sub-group 1 were significantly larger than those from sub-group 3 (F2,198= 8.17, P= 0.0004). In addition, flies from lines Z2 and F3 were significantly smaller than those from most other lines (F8,198= 6.17, P < 0.0001).

Thermal tolerances

Male flies had significantly lower LT50 for acute cold tolerance than female flies (Wald χ2= 401.40, df = 1, P < 0.001, Fig. 1A, D), and a rapid cold hardening pretreatment resulted in improved LT50 of both sexes in almost all lines (Wald χ2= 101.54, df = 1, P < 0.001, Fig. 1A, D). Although LT50 differed among lines in both males and females (males: Wald χ2= 63.22, df = 10, P < 0.001; females: Wald χ2= 123.94, df = 10, P < 0.001), LT50s of cold-selected male flies were not substantially lower than most other lines. Males of Z3 (controls exposed to the RCH pre-treatment) had higher basal LT50s than other lines (i.e., were less cold-hardy), and cold-hardened males of lines C1, Z3, Z2 and A2 were less cold-hardy than their RCH pretreated counterparts in other lines, accounting for a significant line × pretreatment interaction (Wald χ2= 48.99, df = 10, P < 0.001). Among females, the two cold-selected lines (F2 and F3) were significantly more cold-hardy than all other lines with the exception of A1, although their RCH-induced cold tolerance was not significantly different from a larger number of other lines. Thus, the selection regime appears to have enhanced acute low-temperature tolerance more in females than in males.

image

Figure 1. Lower lethal temperature (A, D) survival at −5°C (B, E) and heat tolerance (C, F) of male (A–C) and female (D–F) Drosophila melanogaster, with (grey) and without (black) a hardening-pretreatment. Bars designate the lethal temperature at which 50% of individuals are killed by a stress (LT50) and number of minutes at which 50% of individuals survive (survival at −5°C) ± standard error. Statistical comparisons of survival were performed between models, rather than a direct comparison of LT50. Stars indicate a significant effect of rapid cold hardening or heat pretreatment within each line/sex combination. Bars within each treatment that share a letter are not significantly different. Selection treatments are as follows: C1–3: Control; Z1–3: Exposure to 0°C for 1 h; A1–3: Tolerance for 1 h at −5°C following 1 h hardening at 0°C; F2, F3: Exposure to −5°C for 1 h.

Download figure to PowerPoint

Male flies survived longer than female flies at −5°C (Fig. 1B, E; Wald χ2= 342.81, df = 1, P < 0.001), and an RCH pretreatment resulted in a significant decrease in survival time at −5°C (Fig. 1B, E; Wald χ2= 111.07, df = 1, P < 0.001), although the significance of the latter was not consistent across all lines. Lines F2 and F3 (cold-selected) were the most tolerant of −5°C in females, with the exception of line A1, which did not differ significantly from the F lines in either sex, and line A2, which was similar to F3 in females. By contrast, males of F2 and F3 did not differ significantly from several other lines among the other treatments (Fig. 1B). In addition, the magnitude of the selection response was much greater in females than males: Males of F2 and F3 had an LT50 8.6 and 7.7 min greater at −5°C (respectively) when compared with the control (C) lines within the same group, whereas female F lines survived for 29.1 and 24.2 min longer at −5°C than their control counterparts (Fig. 1E).

Female flies were significantly more tolerant of high temperatures than males (Wald χ2= 250.37, df = 1, P < 0.001). Pretreatment at 36.5°C significantly increased high-temperature survival in all lines of males, and most lines of females (Wald χ2= 293.41, df = 1, P < 0.001; Fig. 1C, F). There were a number of significant differences in high-temperature tolerance among lines (Fig. 1C, F; males: Wald χ2= 102.97, df = 10, P < 0.001; females: Wald χ2= 239.27, df = 10, P < 0.001). However, there were no consistent patterns associated with selection treatment that suggest that there was an effect of selection on high-temperature tolerance.

Female flies of all lines took significantly longer to recover from cold exposure than their male counterparts (Table 1, Fig. 2A, E). Male flies in A2 had significantly shorter chill coma recovery times than others, (Fig. 2A), and females from lines F2 and C2 took longer to recover than the other females (Fig. 2E), accounting for a significant effect of line. However, there was no evidence that any of the selection regimes consistently changed chill coma recovery (Table 1, Fig. 2A, E).

Table 1.  Results of general linear models of the effect of selection and sex on chill coma recovery (time to 80% recovery) and mean starvation and desiccation resistance and longevity in adult Drosophila melanogaster.
  dfF-ratioP
  1. Subgroups are the three subgroups (1, 2, 3) from the initial subdivision; Regime the selection treatments. Significant effects are highlighted in bold type.

Chill coma recoverySubgroup  21.28 0.281
Regime (subgroup)  82.41 0.017
Sex  14.240.041
Subgroup × sex  21.71 0.182
Regime (subgroup) × sex  81.10 0.362
Error198  
Desiccation resistanceSubgroup  20.75 0.474
Regime (subgroup)  81.83 0.074
Sex  1384.27<0.001
Subgroup × sex  25.150.007
Regime (subgroup) × sex  81.70 0.101
Error195  
Starvation resistanceSubgroup  27.78<0.001
Regime (subgroup)  83.55<0.001
Sex  114.31<0.001
Subgroup × sex  210.06<0.001
Regime (subgroup) × sex  84.05<0.001
Error195  
LongevitySubgroup  22.74 0.067
Regime (subgroup)  81.36 0.215
Sex  10.01 0.933
Subgroup × sex  21.19 0.307
Regime (subgroup) × sex  82.150.034
Error180  
image

Figure 2. Chill coma recovery time (A, E), longevity (B, F), starvation resistance (C, G) and desiccation resistance (D, H) of male (A–D) and female (E–H) Drosophila melanogaster. Bars designate the mean time to 80% recovery (± standard error, A) and mean number of days (B) or hours (C, D) survived (± standard error). Bars with differing letters are statistically significantly different (Tukey's HSD P < 0.05); absence of letters across a graph indicates a lack of statistical differences. Selection treatments are as follows: C1–3: Control; Z1–3: Exposure to 0°C for 1 h; A1–3: Tolerance for 1 h at −5°C following 1 h hardening at 0°C; F2, F3: Exposure to −5°C for 1 h.

Download figure to PowerPoint

Desiccation resistance, starvation resistance and longevity

Female flies of all lines were consistently more resistant to desiccation than male flies (Table 1). A significant subgroup × sex interaction probably reflects differences in the relative positions of sub-groups between the sexes, but there was no consistent pattern indicative of an influence of selection (Table 1, Fig. 2D, H). There were no significant effects of line on desiccation tolerance.

Male flies had greater resistance to starvation than females across all lines (Table 1), although all significant differences among lines and selection treatments were in female flies (Fig. 2C, G), which is the likely explanation for the significant interaction effects (Table 1). Flies from sub-group 3 were less starvation-resistant than those from the other sub-groups. There were significant differences between lines, although the pattern of the differences (spread across all of the selection treatments and sub-groups) suggests that the selection regimes had little effect on starvation resistance (Table 1, Fig. 2G).

There was a significant regime × sex interaction in longevity of the selected lines of flies. This may have been driven by the differences that only appear within the females, which in turn were driven by the significantly shorter lifespan of the cold-selected F2 flies compared to the others (Table 1, Fig. 2B, F).

Correlations between stress tolerances

After false discovery rate correction, there were no significant correlations between estimates of stress tolerance among lines of flies within either sex (Table 2). However, power analysis indicates that a relatively modest increase in sample size (1–8 more lines) would have rendered some trends toward correlation among stressors significant in the female flies. Specifically, reciprocal correlations among acute cold tolerance, survival at −5°C and longevity approached significance, with increased acute cold tolerance (LT50) associated with increased survival time at −5°C and both increased acute cold tolerance and increased survival time at −5°C are associated with decreased longevity (Table 2).

Table 2.  Spearman ranked correlation coefficients of stress tolerances among 11 lines of male (bottom left) and female (upper right) adult Drosophila melanogaster.
inline image

Correlation between magnitude of hardening and basal tolerance

Basal acute cold tolerance was not correlated with magnitude of the RCH response in either male or female flies (male: rs=−0.464, P= 0.150; female: rs=−0.152 P= 0.655). Similarly, survival time at −5°C was not correlated with the magnitude of change in survival time after an RCH pretreatment in either males or females (male: rs= 0.112, P= 0.743; female: rs=−0.033, P= 0.924). The magnitude of heat-hardening was significantly negatively correlated with basal high temperature tolerance in females (rs=−0.846, P= 0.001), and nearly so in males (rs=−0.557, P= 0.075). A power analysis indicates that a sample size of 22 lines would be required to render the correlation significant in males.

Correlation between body size and stress tolerances

Body size (measured as dry mass) was significantly, positively correlated with longevity in male flies (rs= 0.836, P= 0.0182), but not in females (rs= 0.100, P= 0.770). No other measures of stress tolerance were significantly correlated with dry mass in either sex (P > 0.05 in all cases). There was also a significant negative correlation between water content and acute cold tolerance in male flies (such that more cold-tolerant flies have a higher water content; rs=−0.772, P= 0.0371), but not in females (rs=−0.300, P= 0.370) or between water content and any other measure of stress tolerance (P > 0.05 in all cases).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References

Selection over 70 generations for tolerance of acute cold exposure (−5°C for 1 h) resulted in a moderate increase in tolerance to exposure to −5°C in males, and a more substantial increase in cold tolerance in females. However, cold selection had very little effect on tolerance to other stressors, including those associated with low temperatures, or desiccation, which has been previously linked to cold tolerance. The selection pressure after the first few generations was likely relatively mild for the F lines, as the intensity (i.e. temperature) was not increased to account for increased tolerance of the selected populations (although exposure to −5°C without pretreatment still resulted in high mortality for the other lines Fig. 1B, E). In D. melanogaster larvae, RCH pretreatment improved both acute and chronic cold tolerance (Rajamohan & Sinclair, 2008), but in the present study, RCH pretreatment resulted in a uniform decrease in survival time at −5°C. This may result from cumulative chilling injury starting when the flies are first exposed to 0°C for the RCH pretreatment. Under these circumstances, the A lines, which had an RCH-inducing pretreatment prior to cold exposure, would be expected to have shown improved cold tolerance as a result of their selection regime. That this was not the case may imply that the mechanisms underlying RCH are not easily subject to selection in Drosophila (e.g., there may be limited variation at the loci associated with RCH).

We hypothesized that selection for acute cold tolerance would result in an increase in tolerance to other stressors. While we did observe significant effects of selection regime on tolerance to starvation and longevity of female flies, in both cases certain cold-selected lines had reduced longevity or starvation tolerance, and there was weak evidence of these trade-offs only in the correlation between tolerance to −5°C and longevity (Table 2). In addition, reduced starvation resistance appeared in several different selection regimes and lineages, which suggests that this effect was not a consistent consequence of the cold selection regime. In contrast to our results, Kristensen et al. (2007) found a slight increase in heat tolerance of cold-selected flies and vice versa, Anderson et al. (2005) found that chill-coma selected lines did not have enhanced high-temperature tolerance; while Bubliy & Loeschcke (2005) found that selection for chronic cold tolerance increased starvation resistance as well. Nevertheless, other studies have observed a lack of direct cross-tolerance in response to selection in Drosophila and have suggested that a general relationship between tolerances to environmental stressors may not exist (Harshman & Hoffmann, 2000). Phelan et al. (2003) show that, in many cases, correlated tolerances to environmental stress disappear with time during artificial selection. They suggest that this may be due to trade-offs, or due to chance linkage that is gradually broken down as the number of generations increase. This may be supported by the observation of Kristensen et al. (2007) that cross-tolerances remain when selection is relaxed, implying that selection is no longer acting to break up remaining linkage groups. Norry et al. (2007) find that several quantitative trait loci associated with cold tolerance are linked to genes associated with other environmental stresses. Thus, there may also have been trade-offs with other inadvertent (or controlled-for) traits selected for in the present design. For example, all the groups were exposed to starvation during the selection regime, and the rearing protocol effectively selected all groups for fecundity at 14 days of age (Archer et al., 2003). Thus, these additional selective pressures equally may have selected against linked traits associated with other environmental stressors.

An alternative explanation for the observed lack of cross-tolerance is that if there is a general stress response that elicits cross-tolerance, it exists because of generalities in the inducible response system, not in the basal tolerances. Thus, selective improvement of basal tolerances will not result in detectable cross-tolerance. For example, Sørensen et al. (2007) found that heat tolerance and longevity are only correlated in D. melanogaster when a heat shock protein-inducing pretreatment is applied. Similarly, in studies of hormesis, Le Bourg (2005, 2007) has demonstrated that stress tolerance (in particular longevity) is improved in adults after earlier exposures to other mild stresses. Together, these examples suggest that, for example, a relationship between cold and desiccation tolerance may only be apparent after exposure to desiccation or cold stress. However, we do note that there are apparently few similarities in the expression of candidate genes in response to desiccation or cold (Sinclair et al., 2007a), although there are few studies of responses to desiccation in Drosophila to expand upon this observation.

Although energy reserves have been linked to tolerance to environmental stress in Drosophila (Chippindale et al., 1998), we only found a relationship between body size (usually correlated with energy reserves) and longevity in males. Similarly, Terblanche et al. (2008) find little relationship between low temperature tolerance and lipid reserves in tsetse flies. If body size (and perhaps, therefore energy reserves) is not related to cold tolerance, then perhaps the selection treatments did not generate sufficient directional change in body size to allow us to detect correlations with other stress tolerances. Desiccation tolerance and cold tolerance are thought to be linked in insects (Ring & Danks, 1994), although there is only limited evidence for this link in D. melanogaster (Bubliy & Loeschcke, 2005; Sinclair et al., 2007a,b). Several species utilise dehydration as part of a cold-tolerance strategy, perhaps to concentrate cryoprotectants (Holmstrup et al., 2002; Ramløv & Lee, 2000), but this is in the reverse direction to the relationship we observed, and water content thus is unlikely to have been a mechanism for changing cold tolerance in D. melanogaster.

We hypothesized that there would be a trade-off between high- and low-temperature tolerances, as previously observed in D. melanogaster along a latitudinal gradient (Hoffmann et al., 2002), and which might be expected from a quantitative trait locus association between high- and low-knockdown temperatures (Norry et al., 2007). Among our selected lines, we found no correlational evidence for a trade-off. In addition, the high-temperature tolerance of the cold-selected flies was not significantly different from other lines in a fashion that implies that the selected lines are significantly less heat-hardy. These observations are consistent with the assertion that high- and low-temperature tolerances may be decoupled in insects (Addo-Bediako et al., 2000). However, we do note that our metrics of both high-temperature tolerance (survival at a 1 h exposure) and low-temperature tolerances are different from those used by previous authors (who primarily use knockdown), further emphasizing the importance of methodology in determining relationships between traits (see also Folk et al., 2007). We found that males were more cold-hardy, while females were more heat-hardy (Fig. 1); sex differences have been noted before for high-temperature tolerances in D. melanogaster (e.g. Folk et al., 2006). The sex differences we observed are consistent with a trade-off between high- and low-temperature tolerances, and it is possible that these differences may derive from an ecological trade-off in thermal habitat utilization in wild adult D. melanogaster.

Kellett et al. (2005) demonstrated a trade-off between basal tolerance to high temperatures and hardening capacity in Drosophila species (including D. melanogaster), a result that has also been observed in other ectotherms (Stillman, 2003), and which has significant implications for species responses to climate change. We also observed a significant trade-off at high temperatures in females, and a trend that approached significance in males, such that lines with greater basal tolerance to high temperatures exhibited reduced hardening capacity. However, we observed no such relationship between basal and induced low-temperature tolerance (either acute cold tolerance or tolerance at the selected temperature of −5°C). This suggests that at low temperatures there is no trade-off between hardening capacity and basal tolerance. This contrast between high- and low-temperature responses suggests that the link between hardening and basal differences may be different in high- and low-temperature tolerances (see also Sinclair & Roberts, 2005), and may be associated with a general decoupling of high- and low-temperature tolerances in insects (Addo-Bediako et al., 2000). Of equal ecological relevance may be the relationship between basal tolerance and longer-term acclimation responses, which likely have different mechanisms and do not necessarily follow the same patterns as short-term hardening (Rako & Hoffmann, 2006).

Rapid cold hardening does not seem to be associated with new gene products, but rather with membrane remodeling (Overgaard et al., 2005, 2006), and prevention of cold-induced apoptosis (Yi et al., 2007). It is unclear whether prevention of apoptosis is an important component of changes in basal cold tolerance, but it is conceivable that this inducible response is not modified constitutively, meaning that basal cold tolerance and magnitude of hardening would remain independent. Changes in basal cold tolerance (at least among species) seem to also be associated with membrane remodeling. However, it may be that the remodeling mechanisms involved in basal and induced cold tolerance differ. For example, fatty acid chains change with a cold-hardening pretreatment (Overgaard et al., 2005), but these changes are not apparent with longer-term acclimation in Chymomyza costata (Kostal et al., 2003). By contrast, high-temperature mortality in D. melanogaster is associated strongly with protein unfolding, aggregation and degradation, and changes to both basal and inducible high-temperature tolerance are associated with regulation of the heat shock response (Feder & Hofmann, 1999).

Our final hypothesis was that different measures of cold tolerance (in this case, acute low-temperature tolerance, survival at −5°C and chill coma recovery) would be correlated. While there was a near-significant correlation between acute cold tolerance and survival at −5°C in female flies, there was no evidence of this in males, nor was there any other evidence (either from comparison of selected lines or from other correlations) to suggest a tight relationship between measures of cold tolerance. We (anecdotally) observed that A-selected flies had reduced locomotion and an inability to fly for a day or more after a cold exposure, although their ability to stand and walk was unaffected. A different metric of performance might have revealed a significant impact of A-selection on these flies. Sinclair and Roberts (2005) pointed out that the relationship between different metrics of cold tolerance, particularly between acute cold shock tolerance, chronic cold tolerance and chill coma onset (knockdown) and recovery is uncertain, and our data suggest that they should not be treated as equivalent, a finding that is in contrast to Anderson et al. (2005) who found a correlation between chill coma recovery (selected over c. 30 generations) and tolerance to a prolonged exposure to −2°C. The reasons for this contrast are not immediately clear, although we speculate that this may reflect similarities in the mechanisms of survival at −2°C and chill coma recovery (perhaps ion homeostasis, Kostal et al., 2006), whereas cold injury at lower temperatures could be a result of other influences like membrane phase transitions (Ramløv, 2000). Folk et al. (2007) provide similar evidence for a lack of correlation among measures of high-temperature tolerance, which further emphasizes the importance of methodology in population-level investigations of cold tolerance in Drosophila.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References

Using variation among lines of flies selected for low-temperature tolerance (and their controls), we found little evidence of cross-tolerance between environmental stresses, nor did we find evidence for a trade-off between high- and low-temperature tolerances in these selected lines. We confirmed that basal tolerances constrain hardening capacity in the context of high temperatures, but at low temperatures, found that selection regimes that included a hardening pretreatment did not improve basal cold tolerance, and that basal tolerance did not constrain hardening capacity at low temperatures. These observations imply that not only are rapid cold-hardening and the heat shock response very different in their mechanisms, but that their relationships to basal tolerances to the same stress differ.

Finally, we found only limited correlation between different measures of ‘cold tolerance’, reinforcing the importance of selecting the correct metric (and cautiously interpreting comparisons between them) in comparing thermal tolerance between species and studies.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References

Thanks to Heather Tarnowski and members of the Sinclair Lab for their assistance in fly management, and to Katie Marshall and Stephanie Bedhomme for critical comments on an earlier version of the manuscript. We are grateful to Adam Chippindale and Virginia Walker of Queen's University, Kingston, ON, Canada, who shared the selected fly lines with us. This research was supported by NSERC Undergraduate Student Research Awards to JPW and HAM, and an NSERC Discovery Grant and grants from the Canadian Foundation for Innovation and Ontario Research Foundation to BJS.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Conclusions
  8. Acknowledgments
  9. References
  • Addo-Bediako, A., Chown, S.L. and Gaston, K.J. (2000) Thermal tolerance, climatic variability and latitude. Proceedings of the Royal Society of London B, 267, 739745.
  • Anderson, A.R., Hoffmann, A.A. and McKechnie, S.W. (2005) Response to selection for rapid chill-coma recovery in Drosophila melanogaster: physiology and life-history traits. Genetical Research, 85, 1522.
  • Archer, M.A., Phelan, J.P., Beckman, K.A. and Rose, M.R. (2003) Breakdown in correlations during laboratory evolution. II. Selection on stress resistance in Drosophila populations. Evolution, 57, 536543.
  • Bayley, M., Petersen, S.O., Knigge, T., Kohler, H.R. and Holmstrup, M. (2001) Drought acclimation confers cold tolerance in the soil collembolan Folsomia candida. Journal of Insect Physiology, 47, 11971204.
  • Benjamini, Y. and Hochberg, Y. (1995) Controlling the false discovery rate – a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society B, 57, 289300.
  • Berrigan, D. (2000) Correlations between measures of thermal stress resistance within and between species. Oikos, 89, 301304.
  • Berrigan, D. and Hoffmann, A.A. (1998) Correlations between measures of heat resistance and acclimation in two species of Drosophila and their hybrids. Biological Journal of the Linnean Society, 64, 449462.
  • Bubliy, O.A. and Loeschcke, V. (2005) Correlated responses to selection for stress resistance and longevity in a laboratory population of Drosophila melanogaster. Journal of Evolutionary Biology, 18, 789803.
  • Chippindale, A.K., Gibbs, A.G., Sheik, M., Yee, K.J., Djawdan, M., Bradley, T.J. and Rose, M.R. (1998) Resource acquisition and the evolution of stress resistance in Drosophila melanogaster. Evolution, 52, 13421352.
  • Chippindale, A.K. and Rice, W.R. (2001) Y chromosome polymorphism is a strong determinant of male fitness in Drosophila melanogaster. Proceedings of the National Academy of Sciences of the United States of America, 98, 56775682.
  • Crill, W.D., Huey, R.B. and Gilchrist, G.W. (1996) Within- and between-generation effects of temperature on the morphology and physiology of Drosophila melanogaster. Evolution, 50, 12051218.
  • Czajka, M.C. and Lee, R.E., Jr. (1990) A rapid cold-hardening response protecting against cold shock injury in Drosophila melanogaster. Journal of Experimental Biology, 148, 245254.
  • Feder, M.E. and Hofmann, G.E. (1999) Heat-shock proteins, molecular chaperones, and the stress response: Evolutionary and ecological physiology. Annual Review of Physiology, 61, 243282.
  • Folk, D.G., Hoekstra, L.A. and Gilchrist, G.W. (2007) Critical thermal maxima in knockdown-selected Drosophila: are thermal endpoints correlated? Journal of Experimental Biology, 210, 26492656.
  • Folk, D.G., Zwollo, P., Rand, D.M. and Gilchrist, G.W. (2006) Selection on knockdown performance in Drosophila melanogaster impacts thermotolerance and heat-shock response differently in females and males. Journal of Experimental Biology, 209, 39643973.
  • Gefen, E., Marlon, A.J. and Gibbs, A.G. (2006) Selection for desiccation resistance in adult Drosophila melanogaster affects larval development and metabolite accumulation. Journal of Experimental Biology, 209, 32933300.
  • Gibbs, A.G. (1999) Laboratory selection for the comparative physiologist. Journal of Experimental Biology, 202, 27092718.
  • Gibbs, A.G., Chippindale, A.K. and Rose, M.R. (1997) Physiological mechanisms of evolved desiccation resistance in Drosophila melanogaster. Journal of Experimental Biology, 200, 18211832.
  • Harshman, L.G. and Hoffmann, A.A. (2000) Laboratory selection experiments using Drosophila: what do they really tell us? Trends in Ecology and Evolution, 15, 3236.
  • Hoffmann, A.A., Anderson, A. and Hallas, R. (2002) Opposing clines for high and low temperature resistance in Drosophila melanogaster. Ecology Letters, 5, 614618.
  • Hoffmann, A.A., Dagher, H., Hercus, M. and Berrigan, D. (1997) Comparing different measures of heat resistance in selected lines of Drosophila melanogaster. Journal of Insect Physiology, 43, 393405.
  • Holmstrup, M., Bayley, M. and Ramløv, H. (2002) Supercool or dehydrate? An experimental analysis of overwintering strategies in small permeable Arctic invertebrates. Proceedings of the National Academy of Sciences of the United States of America, 99, 57165720.
  • Hori, Y. and Kimura, M.T. (1998) Relationship between cold stupor and cold tolerance in Drosophila (Diptera: Drosophilidae). Environmental Entomology, 27, 12971302.
  • Kellett, M., Hoffmann, A.A. and McKechnie, S.W. (2005) Hardening capacity in the Drosophila melanogaster species group is constrained by basal thermotolerance. Functional Ecology, 19, 853858.
  • Kostal, V., Berkova, P. and Simek, P. (2003) Remodelling of the membrane phospholipids during transition to diapause and cold-acclimation in the larvae of Chymomyza costata (Drosophilidae). Comparative Biochemistry and Physiology B, 135, 407419.
  • Kostal, V., Yanagimoto, M. and Bastl, J. (2006) Chilling-injury and disturbance of ion homeostasis in the coxal muscle of the tropical cockroach (Nauphoeta cinerea). Comparative Biochemistry and Physiology B, 143, 171179.
  • Kristensen, T.N., Loeschcke, V. and Hoffmann, A.A. (2007) Can artificially selected phenotypes influence a component of field fitness? Thermal selection and fly performance under thermal extremes. Proceedings of the Royal Society B, 274, 771778.
  • Le Bourg, E. (2005) Hormetic protection of Drosophila melanogaster middle-aged male flies from heat stress by mildly stressing them at young age. Naturwissenschaften, 92, 293296.
  • Le Bourg, E. (2007) Hormetic effects of repeated exposures to cold at young age on longevity, aging and resistance to heat or cold shocks in Drosophila melanogaster. Biogerontology, 8, 431444.
  • Lee, R.E., Jr., Chen, C.-P. and Denlinger, D.L. (1987) A rapid cold-hardening process in insects. Science, 238, 14151417.
  • Nilson, T.N., Sinclair, B.J. and Roberts, S.P. (2006) The effects of carbon dioxide anesthesia and anoxia on rapid cold-hardening and chill coma recovery in Drosophila melanogaster. Journal of Insect Physiology, 52, 10271033.
  • Norry, F.M., Gomez, F.H. and Loeschcke, V. (2007) Knockdown resistance to heat stress and slow recovery from chill coma are genetically associated in a quantitative trait locus region of chromosome 2 in Drosophila melanogaster. Molecular Ecology, 16, 32743284.
  • Overgaard, J., Sørensen, J.G., Petersen, S.O., Loeschcke, V. and Holmstrup, M. (2005) Changes in membrane lipid composition following rapid cold hardening in Drosophila melanogaster. Journal of Insect Physiology, 51, 11731182.
  • Overgaard, J., Sørensen, J.G., Petersen, S.O., Loeschcke, V. and Holmstrup, M. (2006) Reorganization of membrane lipids during fast and slow cold hardening in Drosophila melanogaster. Physiological Entomology, 31, 328335.
  • Pardue, M.L., Ballinger, D.G. and Hogan, N.C. (1992) The heat-shock response – cells coping with transient stress. Annals of the New York Academy of Sciences, 663, 125138.
  • Phelan, J.P., Archer, M.A., Beckman, K.A., Chippindale, A.K., Nusbaum, T.J. and Rose, M.R. (2003) Breakdown in correlations during laboratory evolution. I. Comparative analyses of Drosophila populations. Evolution, 57, 527533.
  • Rajamohan, A. and Sinclair, B.J. (2008) Short-term hardening effects on survival of acute and chronic cold exposure by Drosophila melanogaster larvae. Journal of Insect Physiology, 54, 708718.
  • Rako, L. and Hoffmann, A.A. (2006) Complexity of the cold acclimation response in Drosophila melanogaster. Journal of Insect Physiology, 52, 94104.
  • Ramløv, H. (2000) Aspects of natural cold tolerance in ectothermic animals. Human Reproduction, 15(Suppl. 5), 2546.
  • Ramløv, H. and Lee, R.E., Jr. (2000) Extreme resistance to desiccation in overwintering larvae of the gall fly Eurosta solidaginis (Diptera, Tephritidae). Journal of Experimental Biology, 203, 783789.
  • Ring, R.A. and Danks, H.V. (1994) Desiccation and cryoprotection: Overlapping adaptations. CryoLetters, 15, 181190.
  • Sinclair, B.J. (2000) Water relations of the freeze-tolerant New Zealand alpine cockroach Celatoblatta quinquemaculata (Dictyoptera: Blattidae). Journal of Insect Physiology, 46, 869876.
  • Sinclair, B.J., Gibbs, A.G. and Roberts, S.P. (2007a) Gene transcription during exposure to, and recovery from, cold and desiccation stress in Drosophila melanogaster. Insect Molecular Biology, 16, 435443.
  • Sinclair, B.J., Nelson, S., Nilson, T.L., Roberts, S.P. and Gibbs, A.G. (2007b) The effect of selection for dessication resistance on cold tolerance of Drosophila melanogaster. Physiological Entomology, 32, 322327.
  • Sinclair, B.J. and Roberts, S.P. (2005) Acclimation, shock and hardening in the cold. Journal of Thermal Biology, 30, 557562.
  • Sørensen, J.G., Kristensen, T.N., Kristensen, K.V. and Loeschcke, V. (2007) Sex specific effects of heat induced hormesis in Hsf-deficient Drosophila melanogaster. Experimental Gerontology, 42, 11231129.
  • Stillman, J.H. (2003) Acclimation capacity underlies susceptibility to climate change. Science, 301, 6565.
  • Terblanche, J.S., Clusella-Trullas, S., Deere, J.A. and Chown, S.L. (2008) Thermal tolerance in a south-east African population of the tsetse fly Glossina pallidipes (Diptera, Glossinidae): Implications for forecasting climate change impacts. Journal of Insect Physiology, 54, 114127.
  • Vermeulen, C.J. and Loeschcke, V. (2007) Longevity and the stress response in Drosophila. Experimental Gerontology, 42, 153159.
  • Wu, B.S., Lee, J.K., Thompson, K.M., Walker, V.K., Moyes, C.D. and Robertson, R.M. (2002) Anoxia induces thermotolerance in the locust flight system. Journal of Experimental Biology, 205, 815827.
  • Yi, S.X., Moore, C.W. and Lee, R.E. (2007) Rapid cold-hardening protects Drosophila melanogaster from cold-induced apoptosis. Apoptosis, 12, 11831193.