Report

Fitness Trade-Offs and Environmentally Induced Mutation Buffering in Isogenic C. elegans

See allHide authors and affiliations

Science  06 Jan 2012:
Vol. 335, Issue 6064, pp. 82-85
DOI: 10.1126/science.1213491

Abstract

Mutations often have consequences that vary across individuals. Here, we show that the stimulation of a stress response can reduce mutation penetrance in Caenorhabditis elegans. Moreover, this induced mutation buffering varies across isogenic individuals because of interindividual differences in stress signaling. This variation has important consequences in wild-type animals, producing some individuals with higher stress resistance but lower reproductive fitness and other individuals with lower stress resistance and higher reproductive fitness. This may be beneficial in an unpredictable environment, acting as a “bet-hedging” strategy to diversify risk. These results illustrate how transient environmental stimuli can induce protection against mutations, how environmental responses can underlie variable mutation buffering, and how a fitness trade-off may make variation in stress signaling advantageous.

A specific mutation can have different consequences in different individuals. For example, even in “Mendelian” human diseases, such as cystic fibrosis, an inherited mutation can result in severe disease in one individual but a milder phenotype in another (1). Incomplete penetrance is also observed in isogenic model organisms and is poorly understood (24).

Many mutations have outcomes that depend on the activity of molecular chaperones—proteins that aid the folding of other macromolecules (514). More generally, molecular mechanisms that promote environmental robustness (survival after environmental challenges) also tend to increase mutational robustness [the extent to which an organism’s phenotype remains constant in spite of mutation (1517)].

We investigated whether genetically increasing environmental stress resistance could modify mutation penetrance in the model organism Caenorhabditis elegans. We used a transgene to overexpress the transcription factor heat shock factor 1 (HSF-1), a master regulator of the environmental stress response. Transgenic animals are more resistant to a range of environmental challenges (18, 19) and show a delayed age-dependent reduction in protein-folding homeostasis (20). We crossed the hsf-1 transgenic animals with strains carrying diverse mutations that affect development but with outcomes that vary across individuals (table S1).

In 8 out of 11 tested cases, mutation penetrance was reduced in the transgenic animals (Fig. 1, fig. S1, and table S2). Protection was observed for mutations affecting both embryonic (Fig. 1A) and postembryonic (Fig. 1B) development. For example, embryonic lethality caused by a deletion in the intermediate filament protein gene ifb-1 reduced from 33% to 17% (48% of animals that would have died were protected, P = 5.7×10−12) (Fig. 1, fig. S1, and table S4). The buffered mutations are molecularly diverse and act in distinct pathways and tissues (table S1). Protection ranged from 18 to 88% in the different cases. All of the buffered mutations had temperature-sensitive outcomes, whereas those refractive to buffering did not, and they likely represent genetic nulls (tables S1 and S5).

Fig. 1

Genetic and environmental stimulation of mutation buffering during the development of C. elegans. Increased expression of HSF-1 reduces the penetrance of mutations acting early (A) or late (B) in development. Similarly, a 2-hour 35°C heat shock at the L1 stage of development reduces the penetrance of late-acting mutations (C). **P < 0.01, Fisher exact test; error bars indicate SEM. Phenotypes: Lvl, larval lethal; Emb, embryonic lethal; Bmd, body morphology defect; Muv, multivulva; Vul, vulvaless; Mig, male gonad migration defect; Mab, male abnormal (male tail ray defect); Dpy, dumpy; TF, transcription factor; EphR, ephrin receptor; IF, intermediate filament; ND, not determined. See also fig. S1, tables S1 to S5.

These observations suggest that, at least in C. elegans, a stimulated stress response can reduce the penetrance of partial loss-of-function mutations. We next tested whether the environmental stimulation of a stress response can have a similar effect. A mild environmental stimulus induces chaperone expression and promotes survival in subsequent environmental challenges, a response referred to as “hormesis” (21). We subjected animals to a transient heat shock as larvae to induce a stress response, allowed them to develop to adults, and examined the proportion of individuals affected by late-acting mutations. When a mutation was chaperone-dependent, a mild environmental challenge stimulated a reduction in penetrance (Fig. 1C). For example, an inactivating mutation in the zinc finger transcription factor LIN-29 caused abnormal male gonad migration in 46% of controls but only in 30% of animals receiving a prior heat stress (P = 5 ×10−4, a 35% reduction) (tables S3 and S4).

Beyond genetic differences, therefore, a second cause of variation in mutation outcome can be variation in the prior exposure of individuals to transient environmental stimuli (Fig. 1C and fig. S1, C to K). Although severe environmental stress can reduce mutation buffering (8, 9), transient environmental stimuli can promote it.

Isogenic individuals often show substantial variation in their response to a common environmental challenge (2124). We tested whether this interindividual variation also affects the outcome of mutations. We applied a transient heat shock and sorted animals according to their induction of one of the stress-responsive reporters hsp-16.2p::GFP (green fluorescent protein) or hsp-16.2p::mcherry (22). We found that animals in which a stronger stress response was induced had reduced mutation penetrance (Fig. 2, A and B; fig. S2; and table S6). Quantifying the response in individual animals confirmed this finding: Animals that had a stronger stress response were less likely to be affected by an inherited chaperone-dependent mutation (Fig. 2C and table S7).

Fig. 2

Interindividual variation in a stress response predicts variation in mutation outcome. (A) Animals received a 2-hour 35°C heat shock as L1 larvae and were sorted 1 day later into “high” (right worm) and “low” (left worm) populations, according to the induction of an hsp-16.2 chaperone promoter reporter. Histograms show representative GFP intensity distributions of sorted populations. Images on the right show the midbody region of lin-31(n1053); hsp-16.2::GFP animals. Asterisks mark pseudo-vulvae and arrowheads the vulva. (B) Mutation penetrance in the sorted populations for lin-31(n1053); hsp-16.2::GFP, lin-29(ga94); hsp-16.2::GFP, mab-19(bx83); hsp-16.2::mcherry, and vab-9(ju6); hsp-16.2::GFP (raw data in table S6). (C) GFP intensity in individual lin-31(n1053); hsp-16.2::GFP animals 12 hours after heat shock in larvae that did (right) and did not (left) ultimately develop an abnormal vulva phenotype. GFP expression levels are scaled between 0 and 1 in each panel.

Isogenic individuals that induce higher chaperone levels are therefore more resistant to the effects of certain inherited mutations. These individuals are also more resistant to heat stress and live longer (22). Why, therefore, is a strong stress response not induced in all individuals in a population? Mutations that increase stress resistance can also have a pleiotropic effect on fitness: Perturbing the insulin-like signaling pathway in C. elegans can reduce fecundity (25), and increased chaperone expression in Drosophila can be detrimental (26). We reasoned, therefore, that a fitness trade-off might occur across isogenic individuals, with those having a stronger stress response also incurring a fitness cost.

Using the hsp-16.2p::GFP reporter to separate individuals after a heat stress, we observed that, although individuals that had a stronger response were more resistant to a subsequent severe heat stress (fig. S3 and table S8), they had a significantly reduced early fecundity (fig. S3D and table S9). If conditions remain benign, therefore, individuals that have a stronger stress response incur an important fitness cost (fig. S3D and table S9).

In addition to HSF-1, a second key regulator of the stress response is the FOXO transcription factor DAF-16. DAF-16 relocalizes from the cytoplasm to the nucleus after a heat stress and, with HSF-1, induces the expression of target genes, such as hsp-16.2 (27, 28). Using a DAF-16::GFP fusion protein, we observed that the nuclear translocation of DAF-16 is rapid but that variation occurs across individuals in the time that DAF-16 remains in the nucleus (Fig. 3). For example, DAF-16 had been exported from the nucleus in half of animals ~100 min after a 3-hour heat shock (Fig. 3A). Sorting animals according to the nuclear residency time of DAF-16, we found that higher chaperone induction (fig. S4 and table S21), increased stress resistance (Fig. 3C, table S8), and reduced early fecundity (Fig. 3D and table S10) were all associated with prolonged DAF-16 signaling.

Fig. 3

A trade-off between the development of thermotolerance and reproductive fitness in an isogenic population. (A) Proportion of animals with nuclear DAF-16::GFP after a 2-hour 35°C heat stress at the L3 stage. (B) Two worms 50 min after heat shock: In the right individual (“long duration”), DAF-16::GFP is predominantly localized in the nucleus, whereas in the left (“short duration”) individual, it is also in the cytoplasm. (C) Survival after a 5-hour heat shock at 35°C for the first (short duration) and last (long duration) 10 to 20% of individuals in which DAF-16::GFP returns to the cytoplasm. (D) The number of embryos laid as adults. *P < 0.05, **P < 0.01, Fisher exact test for lethality, Wilcoxon rank sum test for fecundity (see also tables S8 and S10).

We next used a transcriptional reporter for an endogenously expressed chaperone, DAF-21 (Hsp90) to test whether variation in the stress response relates to preexisting variation in chaperone levels. Consistent with this idea, animals with higher daf-21p::mcherry expression before a heat shock developed greater thermotolerance (fig. S3G and table S8) and incurred a reproductive fitness cost after a mild heat stress (fig. S3H and table S11). Thus, interindividual variation in the response to a heat stress is, at least partially, due to preexisting molecular variation in a population.

We reasoned that this preexisting variation in chaperone expression might also affect the outcome of mutations. We used an RNA interference screen to define the individual chaperone dependence of different mutations: The results of this screen revealed that individual mutations differ in their chaperone dependence and provide a resource for future mechanistic work (fig. S5 and tables S13 to S20). In this screen, we identified the mutation lin-31(n1053) as strongly dependent on daf-21 (Hsp90) activity (fig. S5D). We tested whether endogenous variation in daf-21 expression predicted interindividual variation in the outcome of this mutation, by separating animals as larvae by their expression of the daf-21p::mcherry reporter and after the development of mutant phenotypes. Specifically for the lin-31(n1053) mutation, higher expression of the daf-21 reporter was associated with a reduced mutation penetrance (Fig. 4 and table S22). This illustrates that even in the absence of a macroscopic environmental perturbation there can be predictable interindividual variation in the capacity to buffer mutations.

Fig. 4

Interindividual variation in the expression of an endogenous chaperone during larval development predicts mutation outcome. lin-31(n1053);daf-21p::mcherry, lin-29(ga94);daf21p::mcherry, and vab-9(ju6);daf-21p::mcherry animals were sorted into high- and low-expressing populations as L4 larvae and scored for abnormal phenotypes as adults. Variation in the daf-21 reporter predicts variation in the outcome of the lin-31(n1053) mutation. **P = 5.7×10−4, Fisher exact test (table S22). Of these three mutations, only the penetrance of lin-31(n1053) is enhanced when daf-21 is inhibited by RNA interference (fig S5D).

In summary, we have shown that, after a mild environmental stimulus, a trade-off occurs in C. elegans between the development of stress resistance and reproductive fitness. Preexisting molecular variation in an isogenic population means that, after an environmental challenge, stress signaling is prolonged in some individuals, and these individuals develop increased stress resistance but a lower reproductive potential. C. elegans is a naturally self-fertilizing species and, in the wild, is likely to experience highly variable environmental conditions. In a dynamic and unpredictable environment, the generation of such phenotypic diversity can be beneficial, acting as a “bet-hedging” strategy to diversify risk (2933). We suggest that interindividual variation in stress signaling may therefore be beneficial, as it resolves a trade-off between the development of stress resistance and rapid reproduction.

We have also shown that an environmental stress response can stimulate genetic buffering and so protect individuals from inherited mutations. Variation in stress signaling, however, means that this induced capacity to buffer mutations differs predictably across individuals. The conservation of stress responses and their variability (21, 24) suggests that these concepts may apply quite widely, including perhaps in human genetic disease.

Supporting Online Material

www.sciencemag.org/cgi/content/full/science.1213491/DC1

Materials and Methods

Figs. S1 to S5

Tables S1 to S22

References (3448)

References and Notes

  1. Acknowledgments: We thank the European Research Council (ERC), ICREA, Spanish Ministry of Science and Innovation (MICINN) Plan Nacional grant BFU2008-00365, University and Research Grant Management Agency of Catalonia (AGAUR), ERASysBio+, ERA-NET SysBio+, and the EMBL-CRG Systems Biology Program for funding. O.C. was supported by a Beatriu de Pinós Fellowship and A.B by a Formación de Personal Investigador (FPI)–MICINN fellowship. Some nematode strains and bacterial clones were provided by T. Gidalevitz, C. Silva, A. Chisholm, A. Rougvie, and the Caenorhabditis Genetics Center, which is funded by the NIH National Center for Research Resources (NCRR). We thank R. Jelier for assistance with statistical analyses and M. Isalan and members of our laboratory for comments on the manuscript.
View Abstract

Navigate This Article