Introduction
The age-1 mutant under stress
Role of daf-2 inhibition in longevity
Competitive fitness of daf-2 mutants
The daf-2 mutant requires daf-16
Translation to human health
Conclusion & future research
References

The insulin/IGF-1 pathway in aging and fecundity

Final paper for Professor Stephen Stearns, EEB 535: Evolution and Medicine
Peter M. Gayed, Yale University

This online essay is best viewed with a maximized browser window.

6464 words
11 figures
22 references

Introduction
Three major cellular and physiological systems regulate the lifespan of an organism, and each appears to exert its effects independently of the others. The first is the mitochondrial electron transport chain; the next and perhaps least understood (though first identified in 1934) is dietary restriction; and the last and best studied is the insulin and insulin-like growth factor 1 (IGF-1) signaling pathway, or IIS pathway, which is the main focus of this paper.
McCay CM, Crowell MF. Prolonging the Life Span. The Scientific Monthly. 1934;39(5):405-414.

In this essay, I hope to explore the biological basis of longevity and its relationship with fertility. Specifically, I will review the primary literature on longevity and fertility in a long-lived C. elegans mutant known as daf-2. daf-2 encodes the C. elegans orthologue of the mammalian insulin/IGF-1 receptor studied in humans and mice. Two research groups in particular have focused extensively on the daf-2 mutant. The first group, at the University of California, San Francisco, is headed by Cynthia Kenyon, who first identified the long-lived daf-2 mutant. Her laboratory was also the first to develop a manipulable model of daf-2 in C. elegans using small interfering RNA (siRNA) molecules, and primarily employs genetic methods to study longevity. The second group, also located in California, at the Buck Institute for Age Research, is led by Gordon Lithgow. The Lithgow group conducts evolutionary as well as genetic research, which evaluates the fitness costs associated with increased longevity. Work from both groups will be analyzed in detail and placed in the context of evolutionary as well as molecular thinking.
Although aging as a cellular phenomenon proper is not the focus of this paper, the analysis of the primary literature presented here will allow us to make a few insights about how cellular and molecular correlates of aging and reproductive modules might function. In addition, by reviewing the original data, I hope to suggest several experiments that might better tease out the relationship between longevity and reproductive fitness. Finally, I offer potential applications of this information in human health.
Fitness of the age-1 mutant under starved conditions
review of data. In a 2000 study, the Lithgow group at the University of Manchester showed that a mutation in the age-1 gene (the first discovered long-lived C. elegans mutant and a gene which encodes a necessary kinase of the IIS pathway) resulted in an 80 percent increase in lifespan without any effect on fecundity. This study, however, had been performed under plentiful lab conditions, where nutrition and roaming space were not limited. When the group repeated the experiment under "stressed conditions" that were thought to more closely resemble conditions in nature, the age-1 mutants revealed a remarkable decline in fitness as measured by allele frequencies.

To begin, the age-1 mutant strain was grown with wild type (WT) C. elegans on E. coli supplemented agar at 20 degrees Celsius. At this temperature, both the mutant and WT strains progress from larva to adults. Neither enter the dauer phase, a state of diapause in which growth and development are suspended during environmentally harsh conditions. Only hermaphrodite worms who reproduce by self-fertilization were used to prevent sexual reproduction and mixing of gene pools between the strains.

After each generation of progeny, the investigators collected 100 random eggs and transferred them to an identical agar plate kept 7 degrees warmer. At the higher temperature, the more "permissive" age-1 strain gives rise to dauers while the WT continues to yield adults. The ratio of dauers-to-adults, therefore, would represent the frequency of the age-1 allele. After ten generations, it was shown that the frequency of age-1 either remained the same or increased (figure 1). In each arm of the experiment, there was no consistent decline in age-1 frequency, suggesting that the long-lived mutant is not less fit than WT.

Walker DW, McColl G, Jenkins NL, Harris J, Lithgow GJ. Evolution of lifespan in C. elegans. Nature. 2000 May 18;405(6784):296-7.
figure 1. At 20 degrees Celsius, the frequency of the age-1 allele (here, labeled as hx546) either remained constant or increased after ten generations in the presence of WT competitors. The experiment was conducted at three different initial allele frequencies (0.1, green; 0.5, orange; 0.9, blue) tested in duplicate.

However, when the researchers introduced starvation cycles, the age-1 allele declined from an initial frequency of 0.5 to a mean of 0.06 (figure 2).

figure 2. Under starvation conditions, the frequency of the age-1 allele (again labeled hx546) reliably declined. Each trial began with an initial frequency of 0.5 and was re-assessed for six starvation cycles.
limitations. There are some limitations to the design used in this study, however. Each starvation cycle was carried out by allowing the worms to exhaust their shared food source (E. coli). This took about five days. The investigators then waited another four days (starvation period) before collecting eggs to transfer onto a new plate and repeating the cycle. Thus, each cycle lasted a total of nine days. Although the starvation cycles might be more similar to something seen under natural conditions, the transfer of eggs alone (and not any of the adults) precludes the study of fitness between the two populations in a natural environment. That is, the true contribution of long-lived adults could only be studied if these individuals were allowed to contribute progeny throughout their lifetime, viz., at an age where they survive but their WT competitors do not. If the duration of the study were extended beyond the normal lifespan of the WT worm (roughly 30 days), it is possible that the long-lived mutants might predominate since they maintain reproductive output similar to the WT.
Although the ideas of J. B. S. Haldane and Sir Peter Medawar would predict that early life traits are significantly more important than traits expressed in late age, Annette Baudisch at the Max Planck Institute extended the mathematical models of W. D. Hamilton to show that the strength of selection does not necessarily decrease with age. The contribution of long-lived mutants toward the end of life, therefore, might contribute to fitness of the population more than originally predicted.
Baudisch A. Hamilton's indicators of the force of selection. Proc Natl Acad Sci U S A. 2005 Jun 7;102(23):8263-8.
future directions. Nevertheless, the Walker study provides good evidence of antagonistic pleiotropy, but it does so under seemingly austere starvation cycles. Although the authors comment that their 9-day starvation model is "an environment thought to mimic natural conditions," they do not include their rationale (Why nine days and not 12? Why nine and not three?) or point to any references which have validated or tested this model.

To better support their conclusions (and also to provide a valid model which other investigators could employ), variations on the starvation model could be trialed. Parameters such as length of starvation, E. coli density, and agar plate size could be varied. It would be interesting to know how a range of "stress conditions" might affect the fitness of the age-1 mutant. Moreover, a descriptive study of the soil in which C. elegans is found might allow researchers to more closely simulate natural conditions in the lab.

daf-2 inhibition in adulthood doubles lifespan without effect on fertility
review of data. Dillin et al. performed a clever set of experiments to assess the role of daf-2 on lifespan and fertility at different stages of the worm's development. Because a daf-2 knockout would not suffice in this case, the group instead used RNA interference (RNAi), a highly conserved mode of gene inhibition at the level of transcription.

The group devised a system that would allow them to precisely control the actions of an RNA-interfering molecule against daf-2 mRNA. They developed E. coli which stably expressed a vector containing daf-2 double stranded RNA (dsRNA). When C. elegans feeds on these modified E. coli, the daf-2 dsRNA gains entry into the worm and is processed by a conserved enzyme known as Dicer. Processing of the dsRNA by Dicer yields a mature RNAi molecule which targets daf-2 mRNA for degradation.

Consistent with the phenotypes expected in daf-2 genetic mutants, C. elegans treated with the modified E. coli enter the dauer phase when grown at 27 degrees Celsius. Thus, by manipulating food source, the researchers could also manipulate when and for how long the worms were subject to daf-2 knockdown. This conditional knockout model proves useful for studying the role of daf-2 at different time points in the worm's development.

The researchers revealed that inhibition of IIS during adulthood was required for long life. In one part of their study, they transferred adult worms that had been fed the modified E. coli to plates containing E. coli with a dsRNA-containing vector for the Dicer gene (dcr-1). Since Dicer is required for the maturation of all RNAi molecules, introduction of dsRNA specific to the dcr-1 gene should abolish the activity of any RNAi targeting daf-2. When compared with controls, the worms which had been fed E. coli containing daf-2 dsRNA lived no longer, suggesting that continual knockdown of daf-2 is required for the long-lived phenotype.

limitations. Although it is not mentioned in the paper, knock down of dcr-1 by RNAi would lead to global inhibition of all RNAi molecules, not just those targeting daf-2. Thus, nothing specific can be said about the role of daf-2, as the reduction in lifespan may have been the result of large-scale gene dysregulation.

Despite this oversight, the paper goes on to provide compelling evidence that inhibition of daf-2 by RNAi during adulthood increases lifespan without effect on total number of progeny produced: After their development into adults, worms were transferred to plates containing daf-2 dsRNA E. coli on the first, second, third, fourth, or fifth days of adulthood (figure 3). Interestingly, the life-prolonging effect of daf-2 inhibition was greatest when initiated on the first day of adulthood and steadily diminished with delayed treatment. Initiation of RNAi on the fifth day of adulthood (or after) led to no difference in lifespan between the treatment and control groups.

Dillin A, Crawford DK, Kenyon C. Timing requirements for insulin/IGF-1 signaling in C. elegans. Science. 2002 Oct 25;298(5594):830-4.
figure 3. Graphs in the left column show the fraction of worms alive (y-axis) at time points 0 to 70 days (x-axis). Graphs in the right column display the number of progeny produced during adulthood, assessed every 12 hours for a total of five days. The labels in the top-right corner indicate the day of adulthood in which the worms were transferred to plates containing E. coli with an empty vector (control) or E. coli with the daf-2 dsRNA. Blue, control worms; Red, worms receiving daf-2 dsRNA E. coli.
Taken together, the major results of this paper suggest that (1) in adulthood, intact signaling through the IIS pathway limits longevity and (2) that the developmental program which abbreviates lifespan is established within the first four days of adulthood. After this point, inhibition of daf-2 can no longer "rescue" the long-lived phenotype.

review of data. To explore the possible mechanistic basis of daf-2, and to corroborate findings that the inhibition of the daf-2 pathway promotes longevity by neutralizing reactive oxygen species (see Gems et al., Genetics 150, 1998; Lithgow et al., Results Probl. Cell. Differ. 29, 2000; and Johnson et al., Exp. Gerontol. 36, 2001), investigators performed a survival test which subjected worms to 0.4 M paraquat, a symmetrical quaternary ammonium compound, also used as an herbicide, which generates reactive oxygen species (ROS). Control worms were compared to worms that either received daf-2 RNAi at hatching or on the first day of adulthood. Consistent with their longevity data, daf-2 RNAi initiated on the first day of adulthood exhibited the greatest survival, followed by daf-2 RNAi at hatching (figure 4). Survival rates for control worms were less than 15 percent at five hours, compared with the 40-53 percent survival rates seen with RNAi treatment (differences statistically significant between control and RNAi-treated worms, but not between the two arms of RNAi treatment).

figure 4. Comparison of worm survival rates among control, RNAi treatment at hatching, and RNAi treatment on the first day of adulthood when placed in 0.4 M paraquat, performed in duplicate. First trial (first rows) contained worm sample size n ≥ 25; second trial (second rows) contained n ≥ 40.
Testing the competitive fitness of daf-2 mutants
review of data. In the Proceedings of the Royal Society B, Jenkins et al. extended and further refined the results of Dillin et al. Because Dillin and colleagues only assessed changes in longevity conferred by daf-2 knockdown and not competitive fitness, Jenkins et al. grew genetic daf-2 mutants in the presence of WT worms. (They did not, however, compare WT to RNAi-inhibited worms, and this will become important later.)

Using the experimental design from the aforementioned study on age-1 mutants by Walker (also from the Lithgow group), Jenkins et al. competed daf-2 mutants with WT controls under both "standard" laboratory and starved conditions. As with the Walker study, starvation cycles lasted nine days total, five days over which food sources were depleted plus four more days of starvation. Allele frequencies were evaluated after each generation under standard conditions and after each starvation cycle. In the standard arm, an exponential fall in the daf-2 allele was observed, with extinction occurring by the fourth generation (figure 5).

Jenkins NL, McColl G, Lithgow GJ. Fitness cost of extended lifespan in Caenorhabditis elegans. Proc Biol Sci. 2004 December 7;271(1556):2523–2526.
figure 5. Mean frequency of the daf-2 allele when mutants were allowed to compete with WT type worms. Figure represents data collected from six independent trials where 50 eggs each from mutant and WT worms were initially plated and allowed to develop. Squares represent expected frequencies from a mathematical model of fitness which only accounted for differences in the number of progeny produced (and not other factors, such as generation time or egg-to-adult viability, which might reduce overall fitness).

More impressive was the rapid extinction of the daf-2 allele under starved conditions, which occurred by the third cycle (figure 6).

figure 6. Frequency of the daf-2 allele when mutants were allowed to compete with WT type worms under starvation conditions. Figure shows data points from five independent trials.

In addition to original experiments, Jenkins and colleagues also included a re-analysis of the raw data from the 2002 Dillin study. They showed that RNAi treatment initiated at hatching resulted in reductions in the number of progeny produced during the earliest stages of adulthood, namely, within the first 36 hours (figure 7).

figure 7. Mean number of progeny produced during the first three and half days of adulthood. Worms treated with daf-2 RNAi at hatching (empty bars) showed statistically significant reductions in total progeny produced during the second and third 12-hour intervals of adulthood when compared to WT (filled bars).
This new finding is particularly relevant as delays or increases in minimum generation time significantly impact the fitness of C. elegans, as would be predicted by the ideas of J. B. S. Haldane and Sir Peter Medawar. Although competitive fitness was not experimentally tested with RNAi-at-hatching versus control worms, Jenkins et al. used the re-analyzed data to compute relative fitness measures based on the progeny produced by RNAi-treated versus control worms. Considering only the first 24 hours of adulthood, a fitness ratio of roughly 0.77 was assigned to the RNAi-treated worms (average of 72 progeny produced by RNAi-treated worms was divided by an average of 93 progeny produced by control worms to yield relative fitness ratio ≈ 0.77). Using their mathematical model of fitness, the group computed the expected frequency of the daf-2 allele after several generations, finding an exponential decline that led to extinction as early as the 13th generation (figure 8).
Hodgkin J, Barnes TM. More is not better: brood size and population growth in a self-fertilizing nematode. Proc Biol Sci. 1991 Oct 22;246(1315):19-24.
figure 8. Decline in the frequency of the daf-2 allele as predicted by a mathematical model based on varying levels of relative fitness. The earliest extinction point is seen at the 13th generation and the latest is observed just beyond 40 generations.
Although initiation of RNAi at hatching reduced fertility in early adulthood, Jenkins and colleagues did not detect any reduction in fertility in worms treated with RNAi on the first day of adulthood. The near doubling of lifespan in these worms was not associated with decreased fecundity.

limitations. Although their original data revealed a clear reduction in the competitive fitness of the daf-2 mutant, the Jenkins findings cannot be compared one-to-one to the findings of the Dillin study. While the Dillin group used RNAi-directed therapy specific to the daf-2 gene, the studies performed by Jenkins used a screened daf-2 allelic mutant. Loss of the allele could be accompanied by the loss of other important loci or regulatory elements, which precludes a precise comparison between the findings of each group. In other words, the daf-2 strain might be missing other genes, associated with fertility pathways or not, that would result in decreased fitness.

At the time of their publication, no specific daf-2-/- (transgenic knockout) was available for study. The development of such a strain, however, would allow for direct competitive fitness tests between long-lived and control worms.

Taken together, the data from the Dillin et al. and Jenkins et al. studies suggest that the daf-2 pathway determines the reproductive program of C. elegans prior to adulthood, but that its life-prolonging effects are fully mediated during adult stages, and are independent of its role during larval development. More precisely, inhibition of daf-2 beyond the fifth day of adulthood does not endow the worm with any increase in longevity, providing good evidence that intact daf-2 abbreviates the lifespan of C. elegans by modulating programs during the first four days of adult life. Moreover, the strength of this pathway appears to be temporally determined: If RNAi is used to knockdown daf-2 on day one of adulthood, the maximum increase in lifespan is observed; if RNAi is initiated on day two, lifespan is increased, but slightly less so than on day one, and so on. Initiation on day five or beyond results in lifespans no different than controls.

future directions. Notable is the requirement of daf-2 inhibition throughout the lifespan of the adult worm. This suggests that the daf-2 gene is a constituitive inhibitor of longevity. What remains unknown is how daf-2 exerts its effects mechanistically. A number of possibilities exist. Because the gene encodes for the insulin/IGF-1 receptor, the most obvious explanation is that activation of this pathway inhibits a longevity program in the worm. When daf-2 is inhibited by genetic mutation or RNAi treatment, this longevity program is disinhibited, resulting in long life.

Another possibility is more consistent with the trade-off theory of longevity and fecundity. In this case, a finite energy reserve is either used for growth and reproduction (when daf-2 is activated) or diverted to a longevity/preservation module under conditions not favorable for growth and reproduction. This model could explain the discrepancies observed between the Dillin and Jenkins study: When energy supply is abundant (that is, "standard" lab conditions), the daf-2 mutant may not be confronted with an energetic dilemma, and therefore maintains normal fecundity while also achieving long life. Under starvation conditions, however, the available energy becomes scarce, but the daf-2 mutant remains "overcommitted" to a longevity program (said another way, the daf-2 mutant is unable to appropriately arbitrate energy use by the two opposing programs). Because the daf-2 pathway cannot be mobilized in the daf-2 mutant (genetic or RNAi-treated), decreasing energy availability will result in decreasing fecundity.
Longevity of daf-2 mutants is dependent on the presence of daf-16
review of data. Additional evidence in support of the energetic antagonism between longevity and reproduction was provided by Hsin and Kenyon after experiments in which microlaser ablation of germline precursor cells lead to statistically significant increases in lifespan (greater than 90 percent). At hatching, the reproductive system of C. elegans is defined by four cells (figure 9). The outer cells, labeled Z1 and Z4, give rise to the somatic gonad, while the inner cells, Z2 and Z3, are the precursors of all germline cells. In their experiments, the greatest increase in longevity was observed when Z2 and Z3 were eliminated at hatching (figure 10b), suggesting that the maturation of the germline is associated with an energy expenditure that limits longevity.

To corroborate their findings, the group repeated the experiment on wild strains of C. elegans collected from Wisconsin, Germany, and Australia. Among the four strains, germline ablation resulted in an average 60 percent increase in lifespan (figure 10c-e), suggesting that this phenomenon was not particular to laboratory-housed C. elegans.

Hsin H, Kenyon C. Signals from the reproductive system regulate the lifespan of C. elegans. Nature. 1999 May 27;399(6734):362-6
figure 9. Schematic showing the anatomical organization of the reproductive system in C. elegans. Z1 and Z4 give rise to the gonad while Z2 and Z3 form all sperm and oocytes.
figure 10. Germline ablation of Z2 and Z3 result in increased lifespans. b, Ablation of Z2 and Z3 in the laboratory strain of C. elegans (labeled N2) results in a near doubling of lifespan (red circle). Germline ablation of Z1 and Z4 only (blue triangle) or of Z1 through Z4 (blue diamond) results in lifespans no different from unmanipulated controls (black squares). c, d, and e, Repeating the Z2/Z3 ablation experiment with wild strains from Wisconsin (AB3), Germany (CB3191), and Australia (RC301) yields similar results, with a 60 percent mean increase in lifespan.

In an interesting next experiment, Hsin eliminated all four cells of the hatchling's reproductive system and found that the longevity phenotype was abolished—the worms only lived as long as unmanipulated controls (figure 10b). If elimination of germline precursors somehow diverted energy toward a longevity module, why would the added ablation of somatic gonad precursors extinguish this effect? Rather than a model of energetic antagonism, Hsin et al. proposed that the germline produces diffusible molecular signals which restrict lifespan.

A more precise alternative model not proposed in the paper, however, might be that the somatic gonads produce a longevity signal which the germline cells normally suppress. This would fit with their experimental data (figure 11): When germline cells only are eliminated, the longevity signal is disinhibited and lifespan is doubled (first row); when only somatic gonads are eliminated, the longevity signal is lost (second row); when all cells of the reproductive system are ablated (third row), no longevity signal can be produced, but neither is the lifespan shortened, suggesting that germline cells produce a signal which regulates the somatic gonads rather than producing a "life-abbreviating" signal which acts independently.

figure 11. Alternative molecular model for longevity signaling in the reproductive system. Somatic gonad precursor cells Z1 and Z4 produce a longevity signal which is inhibited by dominant signals (red lines) from germline precursors Z2 and Z3. Figure created with Google Drawings; original available here.
Why such a system might have evolved remains to be explained, but one possibility—and I will admit that this is me purely speculating—is that the somatic gonads in C. elegans, analogous to gonadal Sertoli and Leydig cells found in mammals, produce steroid-based sex hormones which can promote male sex drive. If, for instance, a male worm produced fewer germline cells and hence carried fewer sperm on average than his competitors, he would benefit from a greater drive to mate. This enhanced drive could be mediated by increased sex hormone production as a result of the overall reduced inhibitory signal emanating from the germline. What's more, if this were coupled with an increase in the longevity signal, a male worm might be able to compensate for decreased sperm load by living longer and mating more frequently.
Matyash V et al. Sterol-derived hormone(s) controls entry into diapause in Caenorhabditis elegans by consecutive activation of DAF-12 and DAF-16. PLoS Biol. 2004 Oct;2(10):e280
review of data. Finally, Hsin et al. set out to determine if daf-16, another gene implicated in long-lived mutants, might play a role in longevity, as mediated by the worm's reproductive system. daf-16 encodes a forkhead domain transcription factor known as FOXO, which is known to be under regulation by the IIS pathway.
Ogg, S. et al. The Fork head transcription factor DAF-16 transduces insulin-like metabolic and longevity signals in C. elegans. Nature. 1997 Oct 30;389(6654):994-9.
Though the exact (and no doubt numerous) effector functions of daf-16 are unknown, kinase signaling via the insulin/IGF-1 pathway results in phosphorylation of the transcription factor, precluding its translocation to the nucleus. (This is a common mechanism by which transcription factors are retained in the cytosol and prevented from exerting downstream effects.) daf-16 non-null mutants have functional daf-16 molecules which are either resistant to phosphorylation or somehow obviate the typical requirements for nuclear translocation (e.g., by carry a nuclear localization sequence with greater affinity for the nuclear pore complex). Whatever the exact mechanism, daf-16 non-null mutants show greater nuclear compartmentalization associated with increased lifespan.
Kenyon C. A conserved regulatory system for aging. Cell. 2001 Apr 20;105(2):165-8
Consequently, daf-16 null mutants have normal lifespans. Elimination or inhibition of the daf-2 pathway in these mutants does not result in any increase in lifespan, suggesting that the longevity observed in daf-2-inhibited worms requires daf-16. To test whether daf-16 is also required in gonad-mediated longevity, the Hsin group ablated the germline precursors Z2 and Z3 in daf-16 null worms and discovered no change in lifespan. Thus, in WT worms it appears as if the germline somehow inhibits the action of daf-16, and that this inhibition is conveyed by the daf-2 pathway.
future directions. Considered together, the results of the Walker, Dillin, Jenkins, and Hsin studies suggest an interesting connection to Conrad Waddington's concept of the epigenetic landscape. Because daf-2 has unique roles in the larval and adult phases (to ensure optimal fecundity in the former and promote longevity in the latter), there may be strict constraints which prevent the natural emergence of a C. elegans variant with long life and preserved fertility: A spontaneous mutant with depressed daf-2 signaling would result in a long-lived adult with decreased reproductive fitness (as observed in worms treated with RNAi at hatching). Thus, in the wild, this mutant would be outcompeted. If, however, a spontaneous mutant was able to differentially regulate daf-2 signaling, such that it was maintained in larval stages but silenced in adulthood (thereby simulating the RNAi-initiated-at-adulthood experiment), the mutant would maximize longevity without sacrificing fertility.
Principles of Evolutionary Medicine. Chapter 4, Evolution and development. Gluckman, Beedle, and Hanson, eds. Oxford: 2009.
Whether such a species exists in nature is unknown. Moreover, it is also not known if this daf-2 "differential" mutant would suffer costs predicted by antagonistic pleiotropy or whether such a mutant would be outcompeted by WT worms for reasons outside of reproduction and longevity, but this would be an important and interesting topic of research. Without doubt, experimentally engineering or testing this kind of mutant would be difficult. And presumably, the Jenkins group did not perform a competitive assay using the RNAi-at-adulthood model organism because they could not devise a method to feed one group of worms RNAi-expressing E. coli without also exposing controls. However, promising models now exist which could be used for this kind of experiment. One such technique is based on the Cre-ER recombinase system. In this system, a modified estrogen receptor (ER) is conjugated to the Cre recombinase, an enzyme which recognizes two loxP DNA sequences and excises any genes found between them. In the presence of tamoxifen, Cre-ER undergoes a conformational change which permits its entry into the nucleus. If a Cre-ER–expressing C. elegans mutant could be engineered with loxP sequences flanking the daf-2 gene, administration of tamoxifen would allow for the precise temporal excision of daf-2. Because WT worms would not be affected by tamoxifen, the two groups could be directly competed.
Hayashi S, McMahon AP. Efficient recombination in diverse tissues by a tamoxifen-inducible form of Cre. Dev Biol. 2002 Apr 15;244(2):305-18.
Translation to human health
Understanding the mechanistic basis of longevity-promoting daf-2 inhibition could lead to new and interesting biological or biomedical applications. Many commentators, and journalists especially, jump to the translational manipulation of daf-2 in humans to confer long life. Even one of the foremost investigators in this "new science of aging" is prone to invoking the Fountain of Youth metaphor. But evolutionary work has clearly shown the potential costs on reproduction, growth, and overall fitness of a species when longevity is maximized.

resistance against insult. Though the benefit of daf-2 manipulation at the organismal or species level remains disputed, the longevity-enhanced phenotype and its correlate stress resistance might be fruitfully applied to particular cell types within an individual, if not to the individual as a whole. For example, might so called permanent cells of the human body—neurons and cardiac myocytes, cells which typically show little to no "reproductive" (read: proliferative) potential—benefit from the stress-resistant phenotypes endowed by IIS inhibition? The major vulnerability of these cells in human disease is that once they are damaged, they can do little more than form a scar, healing the wound but permanently impairing the function of the organ. In the case of neurons, any degree of cellular loss will lead to significant functional decline. Because these permanent cells are differentiated and have lost their proliferative potential (nor do they require this potential for their physiological roles), any increase in stress resistance might make them more robust.
Furthermore, inhibition of IIS might also be tried in the so called "stable" cellular populations that comprise organs like the liver, kidney, and pancreas. While these tissues retain the capacity to proliferate, by adulthood they typically exist in the quiescent G0 phase of the cell cycle, performing necessary metabolic duties without the need to "turn over" or proliferate extensively. In particular, adult patients with chronic diseases of the kidney (e.g., diabetic nephropathy, polycystic kidney), liver (e.g., hepatitis, hemochromatosis), and pancreas (e.g., insulitis, recurrent pancreatitis) might benefit from therapies which promote cellular maintenance and repair over tissue growth and proliferation. Admittedly, the use of these therapies in diseases of childhood might not apply, as the IIS pathway is necessary for normal growth.
Robbins Pathological Basis of Disease. Chapter 3, Tissue Repair: Regeneration, Healing, and Fibrosis. Kumar ed. 8th ed. Saunders: 2007.
protein aggregating diseases. More promising might be the use of IIS inhibition in human diseases in which cumulative oxidative damage plays a role. Many neurodegenerative diseases, for example, may be the result of protein aggregation which occurs after covalent modification by reactive oxygen species (ROS). Moreover, sporadic neurodegenerative diseases become more prevalent with increasing age, though it is unclear whether this association is explained by higher rates of free radical generation, decreased clearance of aggregated protein, or a combination of both. Nevertheless, modifying the IIS pathway might improve the durability of cells against misfolded proteins or perhaps slow the rate of ROS generation enough to allow clearance before aggregation.
Ross CA, Poirier MA. Protein aggregation and neurodegenerative disease. Nat Med. 2004 Jul;10 Suppl:S10-7.

Clarke G et al. A one-hit model of cell death in inherited neuronal degenerations. Nature. 2000 Jul 13;406(6792):195-9.
Indeed, in a C. elegans model of Huntington disease, where the causative polyglutamic Huntingtin protein was overexpressed, mutants of the IIS pathway show significant reductions in the appearance of protein aggregates and in the incidence of the neuromuscular paralysis which characterizes the disease.
Morley JF et al. The threshold for polyglutamine-expansion protein aggregation and cellular toxicity is dynamic and influenced by aging in C. elegans. Proc Natl Acad Sci U S A. 2002 Aug 6;99(16):10417-22.
Promisingly, the role of the IIS pathway in cellular robustness and longevity appears to be conserved in higher mammals, including mice and humans. Mice carrying only one functional copy of the IGF-1 receptor, the mammalian orthologue of daf-2, show extended lifespans associated with increased resistance against the free radical generator paraquat (in this case, mice were subjected to serial injections of paraquat and monitored for survival). Moreover, it was shown that these mice remain as active, fertile, and large as their WT counterparts.
Holzenberger M. IGF-1 receptor regulates lifespan and resistance to oxidative stress in mice. Nature. 2003 Jan 9;421(6919):182-7.
In another fascinating study, a group from the King's College London crossed a "humanized" mouse model of Alzheimer's disease, which overexpresses the human amyloid beta precursor protein (Aβ), with a mouse line that lacks the insulin receptor gene Irs2. The resultant Alzheimer/Irs2-/- mice were then compared to age-matched Alzheimer mouse controls. At 12 months, Alzheimer/Irs2-/- mice not only exhibited significantly fewer (and smaller) neuronal plaques, but they also demonstrated better memory on fear conditioning tests, providing evidence that loss of Irs2 signaling protects against histological damage as well as functional decline. Interestingly, protected mice had levels of soluble Aβ similar to their unprotected counterparts, lending credence to the idea that Irs2 plays a role in the clearance of aggregated proteins.
Killick R et al. Deletion of Irs2 reduces amyloid deposition and rescues behavioural deficits in APP transgenic mice. Biochem Biophys Res Commun. 2009 Aug 14;386(1):257-62.
one hundred birthdays. In humans, mutations in the IGF-1 receptor have been correlated with centenarianism in Ashkenazi Jews. Unique mutations of FOXO3a, the mammalian orthologue of daf-16, have been identified in both Japanese-Hawaiian and German centenarians. Finally, variants of the human gene IRS2 have been associated with longevity in an Italian cohort.

Of course, inhibition of IIS might result in deficits not yet appreciated, and this is certainly something that would need to be assessed, but manipulation of daf-2-like pathways offers a novel means to protect against insults seen in human disease, especially those occurring in old age.
Dillin A, Cohen E. Ageing and protein aggregation-mediated disorders: from invertebrates to mammals. Philos Trans R Soc Lond B Biol Sci. 2011 Jan 12;366(1561):94-8.
Conclusions & future research
It is clear that aging and reproduction are closely intertwined. Molecular correlates of these two evolutionarily determined processes exist and, moreover, can be manipulated both in the lab and in the environment.
Flatt T, Schmidt PS. Integrating evolutionary and molecular genetics of aging. Biochim Biophys Acta. 2009 Oct;1790(10):951-62.
mechanisms of regulation. One important direction for future research is to characterize the mechanistic deficit of daf-2 mutants. While impairment in the insulin/IGF-1 receptor and downstream signaling may seem most likely, a variety of regulatory mechanisms could be at play. In addition to the encoded receptor, daf-2 or daf-2-associated loci might produce small interfering RNA (siRNA), the endogenous equivalent to the synthetic RNAi molecules used in the Dillin study. Expression of the daf-2 gene itself might also produce siRNA which target molecules required for longevity. There is also the possibility that the daf-2 locus exerts its effects via cis- or trans-regulatory genetic elements. In this scenario, daf-2 promoters might recruit cofactors to form a regulatory complex which can inhibit genes located on the same (cis-) or different (trans-) chromosomes.

Discovery of regulatory mechanisms employed by daf-2 would offer an entirely new avenue of research in which evolutionary trade-offs could receive molecular correlates. Moreover, if regulatory enhancer or suppressor elements did play a role in the longevity/fecundity trade-off, their elucidation might offer a better working model for this and other trade-off theories (not to mention a working model which might yield answers more quickly than the often tedious protein-protein, signaling pathway research currently in favor!).

evolution and molecular biology. The work presented here challenges some facets of classical evolutionary thinking. It has become clear that single genes can have large effects on the lifespan of an organism. Though the strength of selection remains strongest in genes which govern early life, there are evolved and conserved pathways, such as the IIS pathway, which appear to direct both early and late life traits in powerful ways. Moreover, the role of IIS in aging may have co-evolved with its role in reproduction and growth, characteristics which strongly influence the fitness of an organism. And while reproduction and longevity are tightly associated, there is mounting experimental evidence that they can be uncoupled, and that these two processes do not necessarily "compete" for a finite energy reserve. Still, a major area of research that has been relatively neglected by molecular biologists and geneticists is the natural variation of these traits outside of the laboratory.
Though many laboratry organisms appear to exhibit trade-offs between longevity and reproduction, as Linda Partridge notes in a 2005 review in Cell, "even one genuine exception to the rule that lifespan can be extended only by reducing fecundity implies that there can be no obligate trade-off between them." To support the validity of studies which imply the absence of a trade-off, molecular biologists and geneticists will need to employ a range of environmental conditions in which to grow their mutants. If the starvation cycling used by Jenkins et al. and Walker et al. were perhaps "too harsh," the standard conditions which laboratory animals enjoy may be too soft. Because natural conditions are varied and at times unpredictable, scientists would do well to test their animals in a spectrum of environments.

As Thomas Flatt and Daniel Promislow point out in a 2007 commentary in Science, although there are several lifespan-increasing mutations that negatively affect fertility, there are many more that appear to have no observable effect on reproductive or overall fitness. Though the competitive fitness studies performed by Walker et al. and Jenkins et al. (which placed age-1 and daf-2 mutants, respectively, against WTs) showed striking extinction rates among mutants, there remained factors which could not be controlled (discussion above). Whether such uncontrolled factors might account for the drastic impairments in fitness seems unlikely. Nevertheless, given the intriguing results of the RNAi-started-at-adulthood experiments performed by Dillin et al. (discussion above), the development of an experimental model which allows for precise temporal control of the daf-2 pathway in one set of worms while directly competing them against a WT set would provide more definitive conclusions concerning fitness in IIS mutants.

medawar to molecules. Sir Peter Medawar's mutation accumulation theory implies that a large number of mutations in a variety of genes would contribute to the process of aging, yet unique mutations in single genes appear to confer long life, and many of these genes converge onto the conserved IIS pathway, suggesting that aging is indeed under regulation of a few strong alleles rather than many weak ones. While the mutation accumulation theory is being questioned, George Williams's theory of antagonistic pleiotropy has been borne out in numerous studies, though several other publications, including the ones presented here, have demonstrated that some trade-offs can be uncoupled. Still, the mutants in these experiments have been trialed in relatively few conditions, often at the extremes of what might be observed in the wild. Moreover, the natural variation of these traits and the variations of the IIS pathway, in particular, await further characterization.
Flatt T, Promislow DE. Still pondering an age-old question. Science. 2007 Nov 23;318(5854):1255-6.
References
A hyperlinked list of all references in this paper can be found on my delicious.com social bookmarking account here: http://www.delicious.com/pgayed/essay.EvoMed. For convenience, each bookmark directly links to the PubMed abstract of the paper.