insights & perspectives · (“full feeding”), the organism may then be able to reverse the...

.

DOI 10.1002/bies

University of New SouEcology Research CenSydney, New South W

*Corresponding authMargo I. AdlerE-mail: margo.adler@g

Bioessays 36: 439

Insights & Perspectives

Hypotheses

Why do the well-fed appear todie young?

A new evolutionary hypothesis for the effect of dietary restriction on lifespan

Margo I. Adler* and Russell Bonduriansky

Dietary restriction (DR) famously extends lifespan and reduces fecundity across

a diverse range of species. A prominent hypothesis suggests that these life-

history responses evolved as a survival-enhancing strategy whereby resources

are redirected from reproduction to somatic maintenance, enabling organisms

to weather periods of resource scarcity. We argue that this hypothesis is

inconsistent with recent evidence and at odds with the ecology of natural

populations. We consider a wealth of molecular, medical, and evolutionary

research, and conclude that the lifespan extension effect of DR is likely to be a

laboratory artifact: in contrast with captivity, most animals living in natural

environmentsmay fail to achieve lifespan extension under DR.What, then, is the

evolutionary significance of the suite of responses that extend lifespan in the

laboratory? We suggest that these responses represent a highly conserved

nutrient recycling mechanism that enables organisms to maximize immediate

reproductive output under conditions of resource scarcity.

aging; autophagy; caloric restriction;

reproduction; somatic maintenance

Keywords:

dietary restriction; lifespan extension;

Introduction

Dietary restriction (DR), defined as areduction in nutrients without severemalnutrition, has been widely reportedto extend lifespan and reduce fecundityin eukaryotes ranging from yeast toprimates [1–4]. DR is the best known

.201300165

th Wales, Evolution andtre and School of BEES,ales, Australia

or:

mail.com

–450,� 2014 WILEY Pe

and most reproducible environmentalintervention demonstrated in laboratoryanimals to extend lifespan, reduceaging rate, and protect against diseaseand degeneration [5, 6], and as such, itis the topic of considerable investigationacross medical, molecular, evolution-ary, and ecological disciplines. But

Abbreviations:BAT, brown adipose tissue; DR, dietary restric-tion; IGF-1, insulin-like growth factor-1; IIS,insulin/IGF-1 signaling; TOR, target of rapamycin.

riodicals, Inc.

whether this effect is really as universalas believed (e.g. [7]) – as well as thespecifics of how DR extends lifespan,and the perplexing evolutionary ques-tion of why it happens at all – continuesto elude researchers.

One major evolutionary hypothesishas dominated the literature for deca-des: DR’s highly conserved effects onthe life history represent an evolvedstrategy to increase survival undernutrient scarcity [8, 9] (and reviewedin [10]). According to this hypothesis,selection has favored a re-allocation ofnutrients from reproduction to somaticmaintenance and repair, hence increas-ing chances of surviving the famine(Fig. 1) (e.g. [11, 12] and reviewed in [9,13–15]). Once nutrients become plentiful(“full feeding”), the organism maythen be able to reverse the nutrient re-allocation, and resume reproduction.We will refer to this as the “adaptiveresource re-allocation hypothesis”.

Nearly a century of laboratoryresearch has yielded an impressiveamount of experimental data on DR,and a rapidly increasing understandingof the molecular and physiologicalbases of its effects. Yet, we are still along way from understanding the eco-logical and evolutionary implications ofDR, and what applications, if any, itmay have for human health. In thispaper, we critically re-examine theadaptive resource re-allocation hypoth-esis. We challenge its two main asser-tions: (i) the mechanistic explanationthat a re-allocation of resources under-lies lifespan extension and (ii) the

www.bioessays-journal.com 439

Figure 1. The adaptive resource re-allocation hypothesis. Differential investment of resour-ces is generally thought to underlie the life-history responses to dietary restriction. (a) Theadaptive resource re-allocation hypothesis states that under full feeding, most resources areinvested in reproduction and few allocated to somatic maintenance or protection, with theresult that fully fed organisms neglect survival in order to maximize fecundity. (b) However,under DR, a scarcity of nutrients is thought to trigger an adaptive re-allocation of resourcesfrom reproduction to the soma, enhancing survival. (c) Under full feeding, a high reproductiverate entails high somatic costs (i.e. “costs of reproduction”, see text), which can furtherreduce survival. (d) Since reproduction is reduced under DR, costs of reproduction arereduced.

M. I. Adler and R. Bonduriansky Insights & Perspectives.....Hypotheses

evolutionary hypothesis that the life-span extension response to DR evolvedunder selection favoring survival overreproduction in times of resource scarci-ty. We then suggest a new framework,whereby DR’s effects are interpretedas part of a suite of facultative physio-logical responses that enables organismsto maximize immediate reproductiveoutput in times of famine as well asplenty.

The extended lifespanresponse to dietaryrestriction may be alaboratory artifact

Most wild animals are unlikelyto benefit from postponingreproduction

A key question in assessing the plausi-bility of the adaptive resource re-alloca-tion hypothesis is what kills animals in

440

the wild and in the laboratory. Lifespanextension under DR appears to beachieved in the lab, at least in part,through the reduced incidence of cancerand other old-age pathologies [16, 17],and thus a reduced rate of intrinsicaging. For example, extended lifespanin mouse studies is usually (but notalways) correlated with reduced tumorincidence [18]. Reduced rates of cancermay drive lifespan extension in otherDR laboratory animals, including spe-cies that are not typically consideredcancer victims. For example wild-typelaboratory flies have been shown todevelop cancer late in life in the testesand gut, two areas with mitotic activitythroughout life [19]. Yet, organisms inthe wild are typically subject to muchhigher rates of extrinsic (background)mortality (e.g. frompredation or infection)than laboratory populations [20, 21], and(with the exception of some long-livedspecies such as large vertebrates) fewindividuals in natural populations livelong enough to develop cancer and other

Bioessays 36: 4

old-age pathologies [22, 23]. As a result,we argue that a strategy of increasinginvestment to somatic maintenancewould do little to increase survival inthe wild for the great majority of taxain which a lifespan extension effect ofDR has been demonstrated (as repre-sented in Fig. 2). The adaptive resourcere-allocation hypothesis (Fig. 1) thusappears to be at odds with the ecologyof many of the species in which the life-history effects of DR have been demon-strated, such as yeast, nematode worms,flies, and rodents.

Although longitudinal field data onsmall-bodied, short-lived animals arescarce, the available evidence suggeststhat high background mortality rateswould impose a large cost on any delayin reproduction [20, 24–27]. Likewise,most short-lived organisms are unlikelyto have evolved strategies for survivingpredictable (e.g. seasonal) periods ofresource shortage as adults, ofteninvesting heavily instead in a single,brief reproductive period. In specieswhere adults do persist over seasonsunsuitable for reproduction, they tendto do so in a fully quiescent state (e.g.diapause). More typically, short-livedorganisms weather periods unsuitablefor reproduction as eggs or juve-niles [28], and may prolong the juvenilephase when resources are scarce, asobserved in the “dauer” response offree-living nematode larvae [2]. Thesituation may be quite different in

39–450,� 2014 WILEY Periodicals, Inc.

Figure 2. Investing in somatic maintenance is unlikely to pay off when the risk of extrinsicmortality is high.

..... Insights & Perspectives M. I. Adler and R. BondurianskyHypotheses

long-lived species. During sporadic orcyclical (e.g. seasonal) periods of nutri-ent scarcity, organisms subject to rela-tively low rates of extrinsic mortality,such as large-bodied vertebrates (in-cluding humans), could potentiallybenefit from increasing their survivalprospects at the cost of immediatereproduction (as discussed in [29])(al-though, as we show in the next section,it is not clear whether DRwould actuallyenhance survival). More typical species– including the ancestral eukaryotes inwhich the highly conserved suite ofphysiological responses to DR evolved –would be unlikely to benefit from thehypothesized strategy of shutting downthe reproductive system to “wait out”unpredictable periods of resource short-age during the reproductive seasonbecause their chances of surviving toreap the benefits would be minimal.Importantly, this suggests that theadaptive resource re-allocation hypoth-esis cannot account for the evolution ofresponses to DR.

Dietary restriction may reduce,not prolong, survival in the wild

The adaptive resource re-allocationhypothesis faces an even bigger prob-lem: DR reduces capacity to respond toenvironmental challenges, and maytherefore actually reduce survival pros-pects for animals in the wild. Onesituation in which DR would likelyentail a lifespan cost is exposure topathogens and parasites. While numer-

Bioessays 36: 439–450,� 2014 WILEY Pe

ous studies have reported a reducedage-related decline in immune functionunder DR, most studies have notmeasured actual susceptibility to intact,fully functional pathogens [10, 30].But simply maintaining the immunesystem in a sterile environment entailsrelatively low energetic costs as com-pared to mounting an immune responseagainst a real infection. Indeed, whensubjected to intact pathogens, DRanimals tend to show a reducedimmune response, and are more sus-ceptible to bacterial infection, viruses,and gut macroparasites [10, 30]. Forexample one study found that agedDR mice inoculated with influenzavirus were unable to regain body masspost-infection, and died sooner thanfully fed mice [31]. Impairment of theimmune response under DR is in linewith the fact that the drug rapamycin,which mimics key molecular effects ofDR, is used in humans as an immuneinhibitor [32] (but see [33]). A negativeeffect on immune function and recov-ery after infection is also consistentwith findings of a recent meta-analysison the lifespan extension effect ofDR [4], which showed that this effectis weaker in non-model organisms(an interesting recent example ofwhich is provided by [7]). As theauthors of the meta-analysis [4] pointout, model species are generally cul-tured in more sterile conditions thannon-model species, and DR’s life-extending effects on the latter maybe reduced by their increased vulner-ability to infection.

riodicals, Inc.

DR also appears to reduce capacity toavoid and respond to injury. Studies onrodents [34, 35] suggest that DR impairswound healing – an important survivalfunction for animals in natural environ-ments that must recover quickly frominjury in order to evade predators andlocate and compete for food. DR has alsobeen shown to result in bone thinningand osteoporosis in humans [34], andreduced bone mineral content in mon-keys [7], suggesting that vertebratesfacing DR in the wild may be morevulnerable to fractures. Likewise, DRreduces muscle mass in rats, andalthough it attenuates muscle mass losswith age, the initial decrease has beensuggested to severely compromise sta-mina and strength [34]. DR may thushinder an animal’s ability to escapefrom or fight off predators or compet-itors. Such costs may be less importantif DR animals reduce activity rates andtherefore predator/competitor encoun-ter rates, but the available evidenceis equivocal [10]. Indeed, some studiessuggest that DR actually increasesactivity levels, perhaps as a strategy tolocate food [36].

Likewise, there is evidence that coldtolerance is compromised under DR.In mammals and birds, maintaining arelatively constant body temperature isa basic physiological function. Whenexposed to cold temperatures, thedesirable body temperature is achievedthrough an energy-costly process acti-vated in brown adipose tissue (BAT)known as non-shivering thermogene-sis [37]. DR may appear to increase coldtolerance because DR animals saveenergy by tolerating a lower bodytemperature, associated with increasedBAT activation, and because DR reducesage-related decline in BAT function [37].However, as in the case of immunity,simply maintaining a functional levelmay be much less energy-costly thanmounting a response to a challengingcold-exposure. Indeed, DR mammalsexposed to cold have been shown toprogressively lose body weight [38] or tohave a reduced rate of BAT hypertrophyand resultant higher mortality [39, 40].Humans practicing DR report increasedcold sensitivity, attributed to the lowerthreshold to initiate thermogenesis,reduced fat stores and reduced shiver-ing thermogenesis due to muscle massloss, making DR humans more

441


vulnerable to hypothermia and thus athigher risk for stroke, myocardial fibril-lation, and death [34]. Indeed, a studyon rats from the 1930s – one of theearliest reports of the DR effect – notedthat half the animals in the DR groupdied as a result of laboratory heatingsystem failures, while all of the fully fedanimals survived the temperaturedrops [41]. Thus, small-bodied, homo-iothermic DR animals may experienceincreased mortality under natural con-ditions, where cold-exposure is a con-stant threat [9, 42].

Laboratory conditions mayexaggerate the survival costs offull feeding

While most studies have found that anyextension in lifespan is accompanied bya reduction in fecundity, this trade-off isnot always observed [2, 14]. The appar-ent absence of such a trade-off suggeststhat the standard “full-feeding” diets onwhich laboratory animals are main-tained are not optimized for theirphysiological needs, and thus imposesurvival costs with no correspondingfecundity benefit [43, 44]. Lifespan“extension” on certain DR regimensmay therefore simply reflect reducedintake of a poorly balanced diet. Forexample, Lee et al. [45] report that, atsome protein:carbohydrate ratios, in-creased total nutrient intake reducesboth fecundity and longevity in Dro-sophila melanogaster. Such an effect issuggested by the results of Grandisonet al. [13], who found that adding onlythe amino acid methionine to thestandard Drosophila DR diet resultedin increased fecundity without reducedlongevity. When additional essentialamino acids were added to this DRþmethionine diet, however, fecundityincreased above the levels of thestandard “fully fed” animals, whilelongevity was reduced. These resultsshow that the standard full-feedinglaboratory diet reduces longevity asmuch as it does per unit gain infecundity simply because it is nutrition-ally unbalanced, and is therefore aninappropriate “control” diet againstwhich to compare DR animals. Unbal-anced laboratory diets could increaseboth immediate (age-independent) andage-dependent mortality [4]. This sug-

442

gests that, relative to captivity, themortality costs associated with fullfeeding may be lower in natural pop-ulations, where animals have greateropportunity to optimize their diet byselecting among a variety of foods.

The poor nutrient balance of labo-ratory diets raises an important caveatin interpreting DR studies: becausethere is no universally agreed uponDR regimen for any species, DR effectsvary, and results are difficult to gener-alize and compare across studies andspecies [44]. Different DR regimens mayalso elicit lifespan extension in differentways, as has been shown for C. elegans,in which different genetic pathways areactivated depending on which DR regi-men is used [46]. One example ofmethodology confounding interpreta-tion of results comes from a study byJa et al. [47], showing that a commonlyused DR regimen in Drosophila resultedin dehydration of fully fed flies, andthat providing additional water abol-ished the lifespan extension effect. Thesame study also confirmed results fromLee et al. [45] that protein:carbohydrateratio, not total nutrient intake, is themain determinant of diet-dependentlifespan in Drosophila. This illustratesthe value of a “nutritional geometry”approach to DR research [48], andsuggests that results from older studiesshould be interpreted with caution.

Moreover, some authors havesuggested that strong selection forrapid growth and large litter/clutchsizes in laboratory cultures may favor“gluttony”, which might amplify themortality costs of ad libitum feeding inlab-adapted animals [18, 49]. This mightexplain Harper et al.’s [18] finding thatthe grand-offspring of wild mice didnot exhibit a longevity response to DR.This interpretation is also consistentwith the considerable variation in thestrength and even sign of the DR effecton longevity among lab-adapted mousestrains [50], and the finding that DReffects are generally weaker in non-model animals, which are likely to beless lab-adapted [4].

Finally, the longevity-reducingeffects of full feeding may be exaggerat-ed in the lab because of the relativelysedentary lifestyles of captive animals.He et al. [51] found that exercise inducesautophagy after feeding in mice, andtherefore acts as a potent inhibitor of

Bioessays 36: 4

diet-induced cellular damage, as dis-cussed below. Animals in the wild arelikely to get far more exercise than theircaptive counterparts, and may thereforeavoid some of the costs of full feeding.

Resource re-allocation isunlikely to underlie thelife-history responses toDR

Signals, not nutrients per se,control the lifespan response

In contrast to the hypothesis that a re-allocation of nutrients from reproduc-tion to somatic functions promotessurvival under DR, empirical evidencestrongly suggests that signaling, even inthe absence of nutrient manipulation, issufficient to trigger the responses oflifespan and aging to nutrient availabil-ity. Recent research shows that the suiteof life-history responses to nutrientavailability is largely controlled by theactivity of highly conserved nutrient-responsive pathways (reviewed in [1–3,6]). Two of the best-described pathwaysthat have been implicated in regulatingthe life-history responses to nutrientavailability are insulin/IGF-1 signaling(IIS) and target of rapamycin (TOR).Both the IIS and TOR pathways areactivated by plentiful nutrients, andwhen switched on, they result inorganism-wide changes in gene expres-sion as well as in cellular growth andregulatory activity [2, 52]. Some of themost important and relevant down-stream effects of the combined activityof the IIS and TOR pathways includeup-regulation of cellular growth andproliferation rates (directly increasingreproductive rates) as well as down-regulation (inhibition) of cellular recy-cling, repair, and error-preventionmechanisms, most notably autophagyand apoptosis [5] (Fig. 3Ba,b).

Deactivation of nutrient-responsivepathways or of some of their down-stream effects, through mutations af-fecting receptors, or infusionwith signalinhibitors such as the TOR-inhibitingdrug rapamycin, has been found tomimic the life-history responses to DR inmultiple species: extending lifespan,reducing the incidence, and delayingthe onset of cancers and age-related


Figure 3. Nutrient-responsive pathways control multiple physiological and life-history responses. A: Dietary restriction. (a) Under DR ornutrient-signal inhibition (e.g. administration of rapamycin), nutrient-responsive pathways are inhibited and thus do not obstruct autophagyand apoptosis. (b) Autophagy and apoptosis inhibit cell growth rate, precluding high rates of reproduction and somatic responses likely to beimportant for wild survival. (c) Autophagy and apoptosis enhance cellular repair, reducing intrinsic aging by lowering rates of cancer and old-age pathologies, which bolsters survival in the lab. It is unclear how wild survival would be affected. B: Full feeding. (a) Full feeding orinfusion with signal agonists, such as Insulin-like growth factor-1 (IGF-1) or growth hormone (GH), activate nutrient-responsive pathways,which inhibit autophagy and apoptosis, precluding cellular repair and leading to a high rate of intrinsic aging. (b) The inhibition of autophagyand apoptosis allows cell growth and proliferation to operate at a high level, enabling a high rate of reproduction (costs of which maycontribute to intrinsic aging), and also enhancing somatic responses that are likely to bolster wild survival. (c) High intrinsic aging rate reducessurvival in the lab, but effects in the wild are unknown.

..... Insights & Perspectives M. I. Adler and R. Bonduriansky

443Bioessays 36: 439–450,� 2014 WILEY Periodicals, Inc.

Hypotheses


pathologies, and reducing fecundity,even when the animal is fully fed [2, 5,14] (Fig. 3A). Intriguingly, the lifespanextension effect of DR in Drosophila ispartially reversed in flies exposed toodorants from live yeast, demonstratingthat the lifespan response to nutrients iscontrolled in part by signaling ratherthan ingestion [53]. Conversely, infusingDR animals with hormones such asgrowth hormone (GH) or insulin-likegrowth factor-1 (IGF-1), which activatethe downstream targets of the IISpathway, has been shown to reversethe beneficial effects of DR on lifespanand aging [5] (Fig. 3B) – fecundity is not,of course, increased in hormone-infusedDR animals, since actual nutrients arenecessary for reproduction. If lifespanextension (or shortening) can beachieved independently of nutrients,this suggests that the resource re-allocation explanation for the extensionof lifespan under DR may be misguided.A study in D. melanogaster, which usedstable isotopes to track nutrient invest-ment in reproductive and somatictissues, bolsters this conclusion byshowing that fully fed flies invest moreresources than their long-lived DRcounterparts into reproductive andsomatic tissues, suggesting that thelonger lifespans of the DR flies cannotbe explained by increased somaticinvestment [15]. Moreover, reducedreproductive rate in DR animals is notnecessarily indicative of resource re-allocation. Compared to fully fed ani-mals, DR animals may reproduce less asa straightforward consequence of re-source limitation.

Costs of reproduction matter,but cannot be the whole story

Given that fully fed animals typicallyreproduce more than DR animals, costsof reproduction could contribute to thedifference in lifespan between fully fedand DR animals (as considered by,e.g. [14, 54]) (see Figs. 1 and 3). In thiscontext, costs of reproduction refer todirect somatic wear-and-tear associatedwith production of gametes or nourish-ment of young [54], and do not includecosts of mating, such as exposure totoxic substances in semen [55, 56],which may not be reduced in DRanimals if nutritional state is unrelated

444

to mating rate. While costs of reproduc-tion clearly have the potential todecrease lifespan (Fig. 1c andFig. 3Bb), there is strong evidence, atleast in flies, that reproduction itself isunnecessary for the observed reductionin lifespan under full feeding or throughinterventions that activate pathwayssuch as IIS and TOR independently ofnutrients. Mair et al. [57] preventedfemale D. melanogaster from reproduc-ing through removal of the ovaries or amutation that blocks vitellogenesis(yolk formation), yet these sterile fliesstill experienced lifespan reductionsunder full feeding in comparison toDR. Moreover, Bjedov et al. [58] foundthat rapamycin extends lifespan andreduces reproduction in wild-type D.melanogaster females, but also extendslifespan in sterile mutants. The samestudy reported that rapamycin reducesreproduction in a dose-dependent man-ner, while different doses of the drughad roughly the same effect on lifespan,suggesting that reduced fecundity can-not be the sole cause of the lifespanextension [58]. The lack of a straightfor-ward trade-off between survival andfecundity is also demonstrated by theresults of Lee et al. [45]. Thus, whilecosts of reproduction may contribute tothe negative association between nutri-ent availability and longevity, they arenot the sole cause of this association.

An alternative hypothesisfor the life-historyresponses to dietaryrestriction

Above, we showed that extended life-span under DR does not appear to resultfrom a re-allocation of resources fromreproduction to somatic maintenance,and that such a re-allocation would inany case be unlikely to increase fitness –or even prolong life – for most animalsin the wild. The evidence outlined abovethus casts doubt on the long-held ideathat DR’s effects reflect an evolvedresponse geared to increased survival.Why, then, did these responses evolvein ancestral eukaryotes, and why arethey so highly conserved across diverseeukaryote lineages? To answer thesequestions, we suggest that it is neces-sary to refocus attention from the life-

Bioessays 36: 4

history outcomes observed in the non-natural environment of the laboratory,to the underlying physiology thatappears to trigger these outcomes.

At the heart of all organismalresponses to nutrient availability arethe highly conserved nutrient-respon-sive pathways, including IIS and TOR,which have been linked to lifespanextension in diverse species such asyeast, flies, nematode worms, andmice [2]. These pathways initiate acascade of physiological responses, anumber of which have been suggestedto act separately or in concert tomodulate the lifespan response tonutrient availability. Below we examineone of the key responses – the modula-tion of cellular recycling and repairmechanisms, including autophagy andapoptosis – and consider why thisresponse may have evolved in naturalpopulations. Although the physiologi-cal response to DR is highly complexand multi-faceted, we focus on theseparticular processes because they areexceptionally well studied, and providea foundation for an alternative evolu-tionary hypothesis of DR. In the follow-ing sections, we construct this newhypothesis as a series of propositions.

Autophagy and apoptosis areinhibited in fully fed animals tomaximize growth

Autophagy and apoptosis are two of thebest described cellular recycling andrepair mechanisms that respond plasti-cally to nutrient availability, but manyothers are likely to play a role as well.Autophagy is an intra-cellular processwhereby portions of the cell are seques-tered, broken down, and recycled [59],promoting protection and survival ofthe cell [60] and aiding in the preven-tion of neurodegeneration and can-cer [61]. Apoptosis (i.e. programmedcell death) is a systemic process re-quired for normal organismal function-ing that also removes cells that arecancerous or damaged by disease [62].Apoptotic cells are dismantled fromwithin and then recycled [63]. Whenan animal is fully fed, its nutrient-responsive pathways are activated, andthese pathways in turn inhibit cellularrecycling and repair mechanisms, in-cluding autophagy and apoptosis [5]


..... Insights & Perspectives M. I. Adler and R. BondurianskyHypotheses

(Fig. 3Ba). In contrast, when nutrient-responsive pathways are down-regulat-ed under DR, these cellular mechanismsare dis-inhibited (i.e. up-regulated) [59](Fig. 3Aa). But why has selectionfavored the up- or down-regulation ofthese processes in response to nutrientavailability?

The likely answer is that thesemechanisms entail a trade-off: althoughthey can protect, repair, and recycleaging and damaged cells, resulting inlong-term benefits to the organism(detailed below), they also limit therate of cellular growth and prolifera-tion [64, 65], which in turn limits theimmediate reproductive rate [66](Fig. 3Ab). By increasing its cell growthrate, the animal can increase its conver-sion rate of nutrients into reproductiveproducts such as eggs and, in mam-mals, nourishment of young and devel-opment of the endometrium [67]. This isalso true for milk production, as cellgrowth and proliferation are necessaryfor initiating and maintaining lactation(e.g. [68, 69]). Increasing the rate ofcellular growth and proliferationappears to be beneficial not only forincreasing reproductive rate but also foran animal’s ability to respond toenvironmental challenges, throughfunctions such as increased cold toler-ance, wound healing, and immunefunction (Fig. 3Bb). We suggest, there-fore, that autophagy and apoptosis areinhibited under full feeding as anefficient way of allowing the animal toincrease its cell growth rate to takeimmediate advantage of availablenutrients. Although this may result inlatent damage, standard evolutionarytheory predicts that selection willstrongly favor reproduction early in life,even if it entails late-life costs [70].Moreover, as we argued above, thosecosts will be negligible for many small,short-lived species with high rates ofextrinsic mortality.

Figure 4. Adaptive responses to DR make reproduction more attainable. Responses to DRlower the baseline level of resource intake needed for basic functions, enabling reproductiveinvestment (gray arrows) at a lower nutrient level. This is achieved in part through theup-regulation of cellular recycling mechanisms, such as autophagy and apoptosis, whichreduce nutrient requirements by decreasing the size and number of cells that must bemaintained, and which recycle stored nutrients to supplement incoming resources.

Autophagy and apoptosis aredis-inhibited under DR toincrease efficiency

We suggest that selection would favorthe dis-inhibition (i.e. up-regulation) ofcellular recycling mechanisms underDR because they allow animals to makemore efficient use of limited resources,


possibly allowing for some immediatereproduction. Autophagy frees upstored nutrients in cells, a function thathas been suggested to allow the organ-ism to operate with lower resourceintake [71, 72], and apoptosis recycleswhole cells and reduces cell number [63,73], allowing the organism to functionmore efficiently and with reduced cellmass to maintain. One interestingexample of this form of nutrient recy-cling is observed in the cockroachNauphoeta cinerea, in which femalescan initiate apoptosis, even in someoocytes, when nutrients are too scarceto produce a full clutch – in this case,when the females were held understarvation (e.g. [74]). Of course, survivalis a prerequisite of reproduction, and abaseline level of nutrients must beprovided before reproduction will bepossible. DR responses appear to lowerthis baseline, making immediate repro-duction more attainable when nutrientsare scarce (Fig. 4).

Processes such as autophagy andapoptosis could be viewed as mecha-nisms of differential resource allocationbecause, under DR, stored resources arerecycled and put to use for survival orreproduction. However, this form of

riodicals, Inc.

differential allocation differs fundamen-tally from the form envisioned under theadaptive resource re-allocation hypothe-sis because it does not involve sacrificingreproduction for the sake of somaticmaintenance. Instead, we suggest thatit represents an alternate, “making-the-best-of-a-bad-situation” resource-usemode: the organism makes more efficientuse of incoming resources, but with aslower conversion rate that is more-than-sufficient to accommodate the lowerresource intake rate under DR. That is,a low rate of cell growth and proliferationwill not be as costly to an animal that hasfew nutrients to convert to new growthanyway. This slower nutrient conversionrate would, however, entail a cost if morenutrients were available, explaining whyIIS andTOR inhibition in fully fed animalsreduces fecundity [2], and would verylikely increase susceptibility to environ-mental challenges that necessitate a cellgrowth response.

Cellular recycling mechanismsaffect lifespan as a side effect

There is strong empirical evidencethat cellular recycling and repair

445

Figure 5. Responses to nutrient availability entail trade-offs. Levels of autophagy andapoptosis are inversely correlated with nutrient availability, with resulting effects in theopposite direction on cell growth rate, reproductive rate, and environmental-stress tolerance(e.g. wound healing, cold tolerance, and immune function). Laboratory survival trades off withthese functions. Effects on wild survival are unknown, but may be weak or even opposite toeffects in the lab.


mechanisms play a key role in theresponse of lifespan to nutrient avail-ability in laboratory animals (Fig. 3).The clearest evidence exists for autoph-agy, which has been shown to berequired for DR to extend lifespan, asgenetic interventions that blockautophagy preclude the lifespan exten-sion effect of DR and TOR inhibition [71,75], as well as that of IIS inhibition [76].More than just extending lifespan, up-regulation of autophagy under DR hasbeen shown to mediate numerous anti-aging effects across model species,including reduced loss of muscle func-tion and reduced lipofuscin levels inthe heart; blocking autophagy preventsthese beneficial effects [59]. The re-sponse to DR, mediated by mechanismssuch as autophagy, thus appears tocomprise a suite of beneficial physio-logical responses. But the fact thatthese are all at least partly dependenton autophagy suggests that the com-mon mechanism of reduced cellulardamage has numerous phenotypicmanifestations, and this suggests asimple link among the various anti-aging phenotypes triggered by DR inthe lab. In other words, while DR isoften represented as adaptively initiat-ing a complex array of organismalresponses to ensure survival untilresources return, we suggest a simplerscenario: DR up-regulates cellular re-pair and recycling, which, in turn,down-regulates cellular growth andproliferation, and slows aging bothsystemically and within individualorgans and tissues.

Although the role of apoptosis inthe lifespan response to nutrient avail-ability is less clear, abundant evidencesuggests that its dis-inhibition underDR may also promote lifespan exten-sion in the laboratory. It is wellestablished that inhibition of apoptosisis a central cause of tumor develop-ment leading to cancer [77], and whilethe precise effects of DR on apoptosisare unknown, the higher level ofapoptosis enabled when nutrient-re-sponsive pathways are down-regulatedmay contribute to the reduced agingrate [17]. However, as we have notedabove, the survival-promoting effectsof autophagy and apoptosis are unlike-ly to be realized in natural populationsof short-lived animals (Fig. 3Ac,Bc andFig. 5).

446

Responses to DR are highlyconserved because theypromote immediatereproduction

The evidence outlined above suggests aplausible evolutionary explanation forthe highly conserved responses to DR.Under DR, autophagy and apoptosis areup-regulated (i.e. dis-inhibited) to freeup stored nutrients in the cells andmake the animal function more effi-ciently, potentially enabling some im-mediate reproduction. Because highrates of autophagy and apoptosis re-duce the intrinsic aging rate, survival isextended in the laboratory (but proba-bly not in the wild) as a secondaryconsequence (Fig. 3Ac). We thereforesuggest that immediate reproductiveoutput, not survival to reproduce later,has been the key target of selection inthe evolution of physiological responsesto DR. In animals lacking parental care,these responses may allow males tocontinue to pursue matings and allowfemales to produce at least some eggs.Similarly, in animals with extendedparental care, these responses mayallow for the production of at leastsome viable offspring, which may beable to compensate for poor earlynutrition if conditions improve lateron. Moreover, if poor conditions persist,femalesmay recycle their investment, as

Bioessays 36: 4

in the case of small mammals in whichinfanticide is commonly practiced un-der nutrient scarcity [78, 79]. A positiveeffect on immediate reproductive capac-ity offers a plausible explanation for theevolutionary origin and conservationof physiological responses to DR ineukaryotes.

While offering an alternative tothe problematic “adaptive resource re-allocation hypothesis”, our hypothesisalso furnishes an evolutionary basisfor a key premise of the recent “hyper-function” or “bloated soma” theory[6, 80]. This theory suggests that bio-synthetic processes that promote cellproliferation and tissue growth in earlylife – thereby enhancing reproductivecapacity – continue non-adaptively intolate life, resulting in damaging cellularhypertrophy (cell mass growth) andhyperplasia (organ or tissue growthdue to increased cell number). Accord-ing to the bloated soma theory, suchprocesses continue despite causingsomatic damage and ageing becausethere is little selection to stop them; butbecause DR down-regulates nutrient-responsive pathways that controlgrowth and reproduction, it reducesthe rate of biosynthesis and thus therate of aging [6]. However, the bloatedsoma theory fails to explain why selec-tion would favor down-regulation ofbiosynthesis under DR. Our hypothesis


Table 1. The propositions that make up our alternative hypothesis for the life-history effects of DR observed in the laboratoryentail a number of specific predictions. Strong evidence exists in support of some of these predictions, while others need to betested, and we suggest possible experiments.

Proposition Prediction(s) Available evidence/suggested experiments

Lifespan is extended under DR in partbecause cellular recycling mechanisms(e.g. autophagy, apoptosis) areup-regulated

Blocking some or all of thesemechanisms would precludethe lifespan extensionresponse to DR

A few studies have inhibited genes necessary forautophagy in C. elegans held under DR, and found thatautophagy is required for DR to extend lifespan[71, 75, 76]. Some authors have suggested that apopto-sis may also be important in the lifespan extension effectof DR [17], but as far as we are aware, a clear causal linkhas not yet been established

Selection favors the (otherwiseperplexing) inhibition of key cellularrecycling mechanisms, such asautophagy and apoptosis, under fullfeeding because they reduce cellgrowth rate. This would prevent theanimal from taking advantage ofavailable nutrients by reducing:(1) Reproductive rate

(2) Ability to respond to somaticchallenges

Up-regulating autophagy andapoptosis under full feedingwould:(1) Decrease reproductive rate

(2) Reduce ability to respond toenvironmental challenges.

(1) A number of studies have blocked nutrient-responsivepathways while fully feeding the animal, through mutationsand signal inhibitors. These studies have thus indirectlyincreased autophagy and apoptosis in fully fed animals(although pleiotropic effects of blocking these pathwayscould make results difficult to interpret). Many of thesestudies have reported reduced reproduction [2, 58]

(2) In terms of effects on the ability to cope withenvironmental hazards, it would be very interesting to usethe method of blocking nutrient-responsive pathways infully fed animals to measure survival, compared to a fullyfed control group, when exposed to challenges such ascold temperatures, injuries, and live pathogens

The benign environment of thelaboratory masks the survivalcosts of DR

(1) Making the laboratoryenvironment more realistic,in terms of environmentalchallenges, would negate orpossibly reverse the beneficialeffects of DR on lifespan

(1) DR animals exposed to cold temperatures [38–40],injuries [34, 35], and live pathogens [10, 30] in thelaboratory have demonstrated reduced ability to respondto these challenges, and sometimes experiencedincreased mortality as a result

(2) DR in the wild would reduce,rather than prolong, survival(at least for populations withrelatively high rates of extrinsicmortality)

(2) To the best of our knowledge, no study has yetmanaged to dietary restrict wild animals, due to numerousmethodological challenges. Holding wild animals under DRcould potentially be achieved by manipulating the animals’ability to consume or process food, or by manipulatingfood availability in the wild environment. If suchmanipulation were successful, animals could be trackedand survival recorded. A related, but less direct, tactic isto regularly capture marked wild animals and feedsupplementary nutrients to one group, while thenon-supplemented group would be the control,representing something closer to DR. Such studies havetended to find supplemental feeding has no effect onsurvival or that it increases survival; very few haveobserved that survival decreases in response tosupplemental feeding in the wild [18, 29, 81]

DR extends lifespan in the laboratoryby reducing intrinsic aging (old-agepathologies and degeneration), butthese benefits would be unlikely totranslate to the harsher environment ofthe wild, where animals (particularlysmall, short-lived species) do not oftenlive long enough to develop old-agepathologies

(1) DR animals in the laboratorywill have delayed or reducedincidence of cancer and otherold-age pathologies comparedto their fully fed counterparts,accounting for their longerlifespans

(2) In small, short-lived species,most animals will die in the wilddue to extrinsic causes beforeold-age pathologies can playmuch of a role

(1) DR is well known in laboratory animals to reduce theoccurrence and delay the onset of cancers and numerousother pathologies and deterioration associated with oldage [2, 5]. However, these discoveries are generally madeby killing the animal prematurely to detect tumors andother signs of pathology, precluding an assessment ofmortality factors. More work is needed on causes of deathin DR vs. fully fed animals. One useful experiment might beto create numerous replicates of paired animals – eachanimal in the pair treated identically, except one would befully fed and the other DR. When the fully fed animal dies,an autopsy could be performed to detect possible causesof death. The paired DR animal would be killed at the samepoint to examine whether the same pathologies are absentor reduced

(Continued )

..... Insights & Perspectives M. I. Adler and R. Bonduriansky

447Bioessays 36: 439–450,� 2014 WILEY Periodicals, Inc.

Hypotheses

Table 1. (Continued)

Proposition Prediction(s) Available evidence/suggested experiments

(2) Wild animals have been shown to live a significantlyshorter time than their laboratory counterparts [20, 21],and predation has been identified as the main cause ofmortality in some natural populations [82–85]. However,more studies specifically examining causes of death inwild animals, using methods such as autopsies oncarcasses, tracking of individuals to determine predation,and recording environmental conditions at time of death,are needed

Up-regulating cellular recyclingmechanisms such as autophagyand apoptosis is beneficial under DRbecause these mechanisms allow theanimal to make more efficient use ofscarce resources, potentially allowingfor some immediate reproductionthat would not otherwise have beenpossible

Blocking autophagy and/orapoptosis under DR wouldreduce reproduction comparedto DR controls

We are not aware of any studies that have tested this idea,but the methods are available for doing so, at least forautophagy. The methodology used in the studies that haveinhibited genes necessary for autophagy, and measuredeffects in DR animals on lifespan compared to wild-type DRcontrols [71, 75, 76], could be used to measure effects onreproduction. Such an experiment might be extended usinga nutritional geometry approach, in order to reveal a widerrange of nutritional effects


complements the bloated soma theoryby suggesting a selective advantagefor the modulation of cellular growthmechanisms under DR: at least in short-lived animals, organismal responses tovarying levels of nutrient abundancefunction to maximize immediate repro-ductive output.

Conclusions and outlook

In summary, we have shown that theadaptive resource re-allocation hypo-thesis is inconsistent with both thephysiology of DR’s effects, and withthe ecology of natural populations. Thetypical effects of DR – reduced repro-duction and increased longevity – donot appear to result from changes innutrient allocation. Rather, effects onlifespan and aging are mediated bycellular signaling pathways, and repro-duction is reduced under DR becausethere are simply fewer resources avail-able to support it and because cellulargrowth rate is reduced. Moreover,although DR tends to increase longevityin the protected environment of thelab, it is unlikely to do so in naturalpopulations because DR reduces capac-ity to respond to a variety of environ-mental challenges. Because mostspecies of animals (and probably theearly eukaryotes in which the physio-logical responses to DR evolved) aresubject to a high extrinsic mortality rate

448

in the wild, they would be unlikely tobenefit from a strategy of postponingreproduction in any case. We suggest analternative, more plausible evolutionaryhypothesis: the highly conserved phys-iological responses to DR, including theup-regulation of autophagy and apo-ptosis, represent a nutrient-recycling/resource-efficiency mode that may al-low for some immediate reproduction intimes of famine.

The key propositions of our hypo-thesis can be tested experimentally(Table 1). Perhaps most importantly,there is a pressing need for betterunderstanding of differences betweenlaboratory and natural environments inselective pressures and the consequen-ces of variation in diet. One of the keygoals in seeking to understand why andhow DR extends lifespan and whetherthis may be relevant outside of thelaboratory is to discover what usuallykills animals in the lab and the wild. Ifanimals such as rodents, insects, andnematodes regularly die of predation,injury, or exposure to parasites orextreme temperatures, and if DRincreases susceptibility to these mortal-ity causes, then the potential for DR toextend lifespan in natural populationsis low. Conversely, if susceptibility toold-age pathologies has a substantiveeffect on fitness in natural populationsof such organisms, and DR substantiallyreduces susceptibility to extrinsic sour-ces of mortality such as predation (e.g.

Bioessays 36: 4

by reducing activity rates), then thiswould lend credence to the idea thatincreasing investment in survival undernutrient scarcity could be adaptiveacross diverse eukaryotic taxa. Thesequestions can be addressed throughstudies on DR’s effects in naturalpopulations, or through laboratoryexperiments that simulate the harshconditions experienced bywild animals.

Finally, although we have arguedthat increased survival under DR isunlikely to be realized in naturalpopulations, the fact that we can elicitthis response in benign environmentscreates enormous potential for applica-tions to human health. Nonetheless,we must be aware that some of thefactors that limit or negate the potentialbenefits of DR in natural populationsmay also apply to humans (see [34] fora review of known costs in humans).Any intervention that reduces the ca-pacity for reproduction, immune re-sponse, wound healing, and otherfunctions, may be undesirable. Forexample rapamycin has often beenconsidered as a human anti-aging agentbecause of its ability to inhibit activa-tion of the TOR pathway withoutrestricting nutrients [10], but it isessential to know what functions wouldsuffer as a side-effect. An enhancedintegration of insights from evolution-ary, molecular, and medical researchwill be crucial in elucidating thefunctions and underlying mechanisms


..... Insights & Perspectives M. I. Adler and R. BondurianskyHyp

of life-history responses to nutrientavailability, as well as the potentialfor DR, or substances that elicit itseffects, as promoters of healthy longev-ity in humans.

otheses

AcknowledgmentsWe thank Mark Brown, Jonny Anomaly,Angela Crean, Edith Aloise King, Mi-chael Garratt, David Sinclair, AndreasFlouris, Marcos Vidal, and Lindsay Wufor their expertise and insightful discus-sion and comments, Bernice andMortonHunt for copy-editing the manuscript,and Barbara Adler for drawing thecartoon in Fig. 2. We are also verygrateful for the thorough and thoughtfulreviewer and editorial comments wereceived in revising the manuscript.This work was funded through a PhDscholarship to M.A. from the Universityof New SouthWales and the Evolution &Ecology Research Centre, and an Aus-tralian Research Council Research Fel-lowship and Discovery grant to R.B.

References

1. MairW,Dillin A. 2008. Aging and survival: thegenetics of life extension by dietary restric-tion. Annu Rev Biochem 77: 727–54.

2. Kenyon CJ. 2010. The genetics of ageing.Nature 464: 504–12.

3. Fontana L, Partridge L, Longo VD. 2010.Extending healthy life span – yeast to humans.Science 328: 321–6.

4. Nakagawa S, Lagisz M, Hector KL, Spen-cer HG. 2012. Comparative and meta-analyt-ic insights into life extension via dietaryrestriction. Aging Cell 11: 401–9.

5. Longo VD, Fontana L. 2010. Calorie restric-tion and cancer prevention: metabolic andmolecular mechanisms. Trends PharmocolSci 31: 89–98.

6. Gems D, Partridge L. 2013. Genetics oflongevity in model organisms: debates andparadigm shifts.Annu Rev Physiol 75: 621–44.

7. Mattison JA,Roth GS,Beasley TM, TilmontEM, et al. 2012. Impact of caloric restrictionon health and survival in rhesusmonkeys fromthe NIA study. Nature 489: 318–21.

8. Holliday R. 1989. Food, reproduction andlongevity – is the extended lifespan of calorie-restricted animals an evolutionary adapta-tion? BioEssays 10: 125–7.

9. Kirkwood TBL, Shanley DP. 2005. Foodrestriction, evolution and ageing. Mech Age-ing Dev 126: 1011–6.

10. Speakman JR, Mitchell SE. 2011. Caloricrestriction. Mol Aspects Med 32: 159–221.

11. de Jong G, van Noordwijk AJ. 1992.Acquisition and allocation of resources:genetic (co)variances, selection, and lifehistories. Am Nat 139: 749–70.

12. Kirkwood TBL, Rose MR. 1991. Evolution ofsenescence: late survival sacrificed for repro-


duction. Philos Trans R Soc Lond B Biol Sci332: 15–24.

13. GrandisonRC,PiperMDW,PartridgeL.2009.Amino-acid imbalance explains extension oflifespan by dietary restriction in Drosophila.Nature 462: 1061–4.

14. Partridge L,Gems D,Withers DJ. 2005. Sexand death: what is the connection? Cell 120:461–72.

15. O’Brien DM, Min KJ, Larsen T, Tatar M.2008. Use of stable isotopes to examine howdietary restriction extends Drosophila life-span. Curr Biol 18: R155–6.

16. Speakman JR, Hambly C. 2007. Starving forlife: what animal studies can and cannot tellus about the use of caloric restriction toprolong human lifespan. J Nutr 137: 1078–86.

17. Zhang Y, Herman B. 2002. Ageing andapoptosis. Mech Ageing Dev 123: 245–60.

18. Harper JM, Leathers CW, Austad SN. 2006.Does caloric restriction extend life in wildmice? Aging Cell 5: 441–9.

19. Salomon RN, Jackson FR. 2008. Tumors oftestis and midgut in aging flies. Landes Biosci2: 265–8.

20. Kawasaki N, Brassil CE, Brooks RC,Bonduriansky R. 2008. Environmentaleffects on the expression of life span andaging: an extreme contrast between wild andcaptive cohorts of Telostylinus angusticollis(Diptera: Neriidae). Am Nat 172: 346–57.

21. Van Voorhies WA, Fuchs J, Thomas S.2005. The longevity of Caenorhabditis ele-gans in soil. Biol Lett 1: 247–9.

22. Comfort A. 1979. The Biology of Senes-cence. 3rd edn. New York: Elsevier.

23. Kirkwood TBL,Austad SN. 2000.Why doweage? Nature 408: 233–8.

24. Norrdahl K, Korpimaki E. 1995. Mortalityfactors in a cyclic vole population. Proc BiolSci 261: 49–53.

25. Zajitschek F, Brassil CE, Bonduriansky R,Brooks RC. 2009. Sex effects on life spanand senescence in the wild when dates ofbirth and death are unknown. Ecology 90:1698–707.

26. Sherratt TN, Laird RA, Hassall C, Lowe CD,et al. 2010. Empirical evidence of senescencein adult damselflies (Odonata: Zygoptera). JAnim Ecol 79: 1034–44.

27. Bonduriansky R, Brassil CE. 2002. Rapidand costly ageing in wild male flies. Nature420: 377.

28. Stearns SC, Koella JC. 1986. The evolutionof phenotypic plasticity in life-history traits:predictions of reaction norms for age and sizeat maturity. Evolution 40: 893–913.

29. Le Bourg E. 2010. Predicting whetherdietary restriction would increase longevityin species not tested so far. Ageing Res Rev9: 289–97.

30. Kristan DM. 2008. Calorie restriction andsusceptibility to intact pathogens. Age 30:147–56.

31. Gardner EM. 2005. Caloric restrictiondecreases survival of aged mice in responseto primary influenza infection. J Gerontol BiolSci 60A: 688–94.

32. Abraham RT, Wiederrecht GJ. 1996. Immu-nopharmacology of rapamycin. Annu RevImmunol 14: 483–510.

33. Araki K, Turner AP, Shaffer VO, GangappaS, et al. 2010. mTOR regulatesmemory CD8 Tcell differentiation. Nature 460: 108–12.

34. Dirks AJ, Leeuwenburg C. 2006. Caloricrestriction in humans: potential pitfalls andhealth concerns. Mech Ageing Dev 127: 1–7.

riodicals, Inc.

35. Hunt ND, Li GD, Zhu M, Levette A, et al.2012. Effect of calorie restriction and refeed-ing on skin wound healing in the rat. Age 34:1453–8.

36. Shanley DP, Kirkwood TBL. 2000. Calorierestriction and aging: a life-history analysis.Evolution 54: 740–50.

37. Carrillo AE, Flouris AD. 2011. Caloricrestriction and longevity: effects of reducedbody temperature. Ageing Res Rev 10:153–62.

38. Puerta ML, Abelenda M. 1987. Cold accli-mation in food-restricted rats.CompBiochemPhysiol 87A: 31–3.

39. Johnson TS, Murray S, Young JB, Lands-berg L. 1982. Restricted food intake limitsbrown adipose tissue hypertrophy in coldexposure. Life Sci 30: 1423–6.

40. Gordon CJ. 1993. Temperature Regulation inLaboratory Rodents. New York: CambridgeUniversity Press.

41. McCay CM, Maynard LA, Sperling G,Barnes LL. 1939. Retarded growth, life span,ultimate body size and age changes in thealbino rat after feeding diets restricted incalories. J Nutr 18: 1–13.

42. Martin B, Mattson MP, Maudsley S. 2006.Caloric restriction and intermittent fasting:two potential diets for successful brain aging.Ageing Res Rev 5: 332–53.

43. Simpson SJ, Raubenheimer D. 2007. Calo-ric restriction and aging revisited: the need fora geometric analysis of the nutritional basesof aging. J Gerontol A Biol Sci Med Sci 62:707–13.

44. Piper MDW, Partridge L. 2007. Dietaryrestriction in Drosophila: delayed aging orexperimental artefact? PLoS Genet 3: e57.

45. LeeKP,Simpson SJ,Clissold FJ,Brooks R,et al. 2008. Lifespan and reproduction inDrosophila: new insights from nutritionalgeometry. Proc Natl Acad Sci USA 105:2498–503.

46. Greer EL, Brunet A. 2009. Different dietaryrestriction regimens extend lifespan by bothindepedent and overlapping genetic path-ways in C. elegans. Aging Cell 8: 113–27.

47. Ja WW, Carvalho GB, Zid BM, Mak EM,et al. 2009. Water- and nutrient-dependenteffects of dietary restriction on Drosophilalifespan. Proc Natl Acad Sci USA 106:18633–7.

48. Bruce KD, Hoxha S, Carvalho GB, YamadaR, et al. 2013. High carbohydrate-low proteinconsumption maximizes Drosophila lifespan.Exp Gerontol 48: 1129–35.

49. Austad SN, Kristan DM. 2003. Are micecalorically restricted in nature? Aging Cell 2:201–7.

50. Liao C-Y, Rikke BA, Johnson TE, Diaz V,et al. 2010. Genetic variation in the murinelifespan response to dietary restriction: fromthe extension of life to life shortening. AgingCell 9: 92–5.

51. He C, Bassik MC, Moresi V, Sun K, et al.2012. Exercise-induced BCL2-regulatedautophagy is required for muscle glucosehomeostasis. Nature 481: 511–5.

52. Emlen DJ, Warren IA, Johns A, Dworkin I,et al. 2012. A mechanism of extreme growthand reliable signaling in sexually selectedornaments and weapons. Science 337:860–4.

53. Libert S, Zwiener J, Chu X, VanVoorhies W,et al. 2007. Regulation of Drosophila life spanby olfaction and food-derived odors. Science315: 1133–7.

449


54. Tatar M, Carey JR. 1995. Nutrition mediatesreproductive trade-offs with age-specificmortality in the beetle Callosobruchus mac-ulatus. Ecology 76: 2066–73.

55. Fowler K,Partridge L. 1989. A cost of matingin female fruitflies. Nature 338: 760–1.

56. ChapmanT,Liddle LF,Kalb JM,WolfnerMF,et al. 1995. Cost of mating in Drosophilamelanogaster females is mediated by maleaccessory gland products. Nature 373: 241–4.

57. MairW,Sgro CM, Johnson AP,ChapmanT,et al. 2004. Lifespan extension by dietaryrestriction in female Drosophila melanogasteris not caused by a reduction in vitellogenesisor ovarian activity. Exp Gerontol 39: 1011–9.

58. Bjedov I, Toivonen JM, Kerr F, Slack C,et al. 2010.Mechanisms of life span extensionby rapamycin in the fruit fly Drosophilamelanogaster. Cell Metab 11: 35–46.

59. Rubinsztein DC, Marino G, Kroemer G.2011. Autophagy and aging.Cell 146: 682–95.

60. Madeo F, Tavernarakis N, Kroemer G.2010. Can autophagy promote longevity?Nat Cell Biol 12: 842–6.

61. Mizushima N, Klionsky DJ. 2007. Proteinturnover via autophagy: implications formetabolism. Annu Rev Nutr 27: 19–40.

62. Elmore S. 2007. Apoptosis: a review of pro-grammed cell death. Toxicol Pathol 35: 495–516.

63. Taylor RC, Cullen SP, Martin SJ. 2008.Apoptosis: controlled demolition at the cellu-lar level. Nat Rev Mol Cell Biol 9: 231–41.

64. Scott RC, Juhasz G, Neufeld TP. 2007.Direct induction of autophagy by Atg1 inhibitscell growth and induces apoptotic cell death.Curr Biol 17: 1–11.

65. Vellai T, Bicsak B, Toth ML, Takacs-VellaiK, et al. 2008. Regulation of cell growth byautophagy. Autophagy 4: 507–9.

450

66. Narbonne P, Roy R. 2006. Regulation ofgermline stem cell proliferation downstreamof nutrient sensing. Cell Div 1: 29.

67. Bronson FH. 1989.Mammalian ReproductiveBiology. Chicago and London: The Universityof Chicago Press.

68. Capuco AV, Wood DL, Baldwin R, McLeodM, et al. 2001. Mammary cell number,proliferation, and apoptosis during a bovinelactation: relation to milk production andeffect of bST. J Dairy Sci 84: 2177–87.

69. Traurig HH. 1967. Cell proliferation in themammary gland during late pregnancy andlactation. Anat Rec 157: 489–503.

70. Williams GC. 1966. Natural selection, thecosts of reproduction, and a refinement ofLack’s principle. Am Nat 100: 687–90.

71. Hansen M, Chandra A, Mitic L, Onken B,et al. 2008. A role for autophagy in theextension of lifespan by dietary restriction inC. elegans. PLoS Genet 4: e24.

72. He C, Klionsky DJ. 2009. Regulating mech-anisms and signaling pathways of autophagy.Annu Rev Genet 43: 67–93.

73. James SJ, Muskhelishvili L, Gaylor DW,Turturro A, et al. 1998. Upregulation ofapoptosis with dietary restriction: implica-tions for carcinogenesis and aging. EnvironHealth Perspect 106: 307–12.

74. Barrett ELB, Preziosi RF,Moore AJ,MoorePJ. 2008. Effects of mating delay andnutritional signals on resource recycling in acyclically breeding cockroach. J Insect Phys-iol 54: 25–31.

75. Jia K, Levine B. 2007. Autophagy is requiredfor dietary restriction-mediated life spanextension in C. elegans. Autophagy 3: 597–9.

76. Melendez A, Talloczy Z, Seaman M,Eskelinen E-L, et al. 2003. Autophagy genes

Bioessays 36: 4

are essential for dauer development and life-span extension in C. elegans. Science 301:1387–91.

77. Evan GI, Vousden KH. 2001. Proliferation,cell cycle and apoptosis in cancer. Nature411: 342–8.

78. Schneider JE, Wade GN. 1989. Effects ofmaternal diet, body weight and body compo-sition on infanticide in Syrian hamsters.Physiol Behav 46: 815–21.

79. Sabau RM, Ferkin MH. 2013. Food restric-tion affects thematernal behavior provided byfemale meadow voles (Microtus pennsylvani-cus). J Mammal 94: 1068–76.

80. BlagosklonnyMV. 2008. Aging: ROS or TOR.Cell Cycle 7: 3344–54.

81. Boutin S. 1990. Food supplementationexperiments with terrestrial vertebrates: pat-terns, problems, and the future. Can J Zool68: 203–20.

82. Jepsen N, Aarestrup K, Økland F, Ras-mussen G. 1998. Survival of radio-taggedAtlantic salmon (Salmo salar L.) and trout(Salmo trutta L.) smolts passing a reservoirduring seaward migration. Hydrobiologia371–372: 347–53.

83. Seip DR, Cichowski DB. 1996. Populationecology of caribou in British Columbia.Rangifer 9: 73–80.

84. Fincke OM. 1982. Lifetime mating success ina natural population of the damselfly, Enal-lagma hageni (Walsh) (Odonata: Coenagrio-nidae). Behav Ecol Sociobiol 10: 293–302.

85. Schekkerman H, Teunissen W, OosterveldE. 2009. Mortality of black-tailed godwitLimosa limosa and Northern Lapwing Vanellusvanellus chicks in wet grasslands: influenceof predation and agriculture. J Ornithol 150:133–45.


insights & perspectives · (“full feeding”), the organism may then be able to reverse the...

Documents