The insulin/IGF-1 signalling (IIS) pathway connects nutrient levels to metabolism, growth and lifespan in eukaryotes ranging from yeasts to humans, including nematodes such as the genetic model organism Caenorhabditis elegans. The link between ageing and the IIS pathway has been thoroughly studied in C. elegans; upon reduced IIS signalling, a genetic survival program is activated resulting in a drastic lifespan extension. One of the components of this program is the upregulation of antioxidant activity but experiments failed to show a clear causal relation to longevity. However, oxidative damage, such as protein carbonyls, accumulates at a slower pace in long-lived C. elegans mutants with reduced IIS. This is probably not achieved by increased macroautophagy, a process that sequesters cellular components to be eliminated as protein turnover rates are slowed down in IIS mutants. The IIS mutant daf-2, bearing a mutation in the insulin/IGF-1 receptor, recapitulates the dauer survival program, including accumulation of fat and glycogen. Fat can be converted into glucose and glycogen via the glyoxylate shunt, a pathway absent in vertebrates. These carbohydrates can be used as substrates for trehalose synthesis, also absent in mammals. Trehalose, a non-reducing homodimer of glucose, stabilises intracellular components and is responsible for almost half of the lifespan extension in IIS mutants. Hence, the molecular mechanisms by which lifespan is extended under reduced IIS may differ substantially between phyla that have an active glyoxylate cycle and trehalose synthesis, such as ecdysozoans and fungi, and vertebrate species such as mammals.
The ageing process of free-living nematodes was first explored in 1970 by David Gershon using the vinegar eel Turbatrix aceti. By applying synchronisation and chemical sterilisation techniques, he established the first survival curves of worm cohorts and monitored age-specific loss of enzyme activities (Erlanger & Gershon, 1970; Gershon, 1970; Gershon & Gershon, 1970). A few years later, ageing in Caenorhabditis elegans was first described in studies covering survival at varying environmental conditions (Klass, 1977) and the non-ageing characteristics of the dauer diapause stage (Klass & Hirsh, 1976). Two studies in the early 1980s nucleated the field of the genetics of ageing in C. elegans. First, Tom Johnson showed that in C. elegans the genetic components of recombinant inbred lines dictate lifespan (Johnson & Wood, 1982), and, second, Michael Klass discovered the first C. elegans mutants with extended lifespan (Klass, 1983). The latter strains were not outcrossed and showed pleiotropic phenotypes and, therefore, the author assumed the worms lived longer due to a dietary restriction effect rather than by mutation in one specific gene that regulates ageing rate. Subsequent genetic analysis of these strains by Tom Johnson resulted in the discovery of age-1 (ageing alteration), the first mutated gene ever described that causes lifespan extension in a metazoan (Friedman & Johnson, 1988). With the discovery of a second longevity mutant, daf-2 (abnormal dauer formation), a few years later (Kenyon et al., 1993), the genetics of ageing field gained substantial momentum. The daf-2 mutant was not recovered from a genetic longevity screen but was isolated in an effort to analyse genetically dauer formation (Kenyon, 2011). In the Kenyon study, it was shown that lifespan extension in daf-2 mutants is fully dependent on daf-16 gene activity. A few years later, these genes were cloned, characterised and found to be part of an Insulin/Insulin Growth Factor 1-mediated signalling (IIS) pathway, with daf-2 coding for the Ins/IGF-1 receptor (Kimura et al., 1997), age-1 being a homologue of a phosphoinositide 3-kinase subunit (Morris et al., 1996) and daf-16 representing the C. elegans forkhead box O (FOXO) transcription factor (Lin et al., 1997; Ogg et al., 1997).
The IIS pathway
After the seminal discoveries of age-1, daf-2 and daf-16, other genes of the IIS pathway (Fig. 1) were identified and characterised (Ogg & Ruvkun, 1998; Paradis & Ruvkun, 1998; Paradis et al., 1999; Wolkow et al., 2002; Hertweck et al., 2004). Most IIS genes have homologues in species ranging from yeast to humans, indicating the strong evolutionary conservation of this pathway. Moreover, the IIS pathway seems to influence lifespan in many of these species, including humans (Kenyon, 2010). These discoveries elicited quite some optimism and many research groups have since focused on the IIS pathway and the mechanisms by which it extends C. elegans lifespan. The gene index in the abstract book of the 2015 International C. elegans Meeting (Los Angeles, CA, USA) leaves no doubt that daf-2 and daf-16 are, currently, by far the most studied C. elegans genes.
Although the IIS pathway generally regulates growth and metabolism in all eukaryotes, its constituents have diverged into networks that differ among phylogenetic groups (Papatheodorou et al., 2014; McGaugh et al., 2015). As a result, there is no one-to-one gene homology between the IIS components of C. elegans and humans. The worm genome encodes for 40 insulins, most of which have unidentified functions (Pierce et al., 2001; Cornils et al., 2011), whilst in humans only three related insulin-like peptides occur: insulin, IGF-1 and IGF-2. On the other hand, C. elegans has only one known IIS receptor, daf-2, whilst humans express several receptors, each controlling downstream pathways that trigger specific physiological functions (Clayton et al., 2011). In C. elegans, the FOXO transcription factor DAF-16 is the endpoint of the IIS cascade. A forkhead box (FOX) is an 80-100 amino-acid DNA-binding motif that is present in a wide variety of FOX proteins. These proteins are clustered in subfamilies ranging from A to S based on sequence homology and DAF-16 belongs to the O subfamily. A single daf-16 locus in C. elegans encodes five isoforms (Kwon et al., 2010; Chen et al., 2015), two of which influence lifespan, whilst in humans four separate genes make up the FOXO family. Also here, two FOXO forms have been linked to lifespan determination (Martins et al., 2016). Hence, despite their diversity, FOXO transcription factors seem to regulate lifespan in a wide range of eukaryotes, but likely via different downstream programs.
The FOXO/DAF-16 lifespan program
Being identified as a master switch in lifespan regulation, daf-16 has received much attention over the last two decades. The identification of genes under control of this transcription factor could reveal the molecular mechanism(s) of lifespan extension.
With the dawn of the omics era, transcriptomic (McElwee et al., 2003; Murphy et al., 2003; McElwee et al., 2004; Halaschek-Wiener et al., 2005), proteomic (Dong et al., 2007; Jones et al., 2010; Depuydt et al., 2013, 2014; Stout et al., 2013; Walther et al., 2015) and metabolomic (Fuchs et al., 2010) studies showed that IIS mutants undergo massive changes in gene expression and shifts in metabolic networks. DAF-16 targets were also determined by analysing DNA binding sites of this transcription factor with DamID and ChIP (Oh et al., 2006; Schuster et al., 2010). In an early microarray study, many of the differentially expressed genes of IIS mutants were tested for their influence on lifespan. Most of these genes only had a partial effect, suggesting that the total IIS lifespan extension is caused by the upregulation of a wide variety of genes, including genes involved in the cellular stress and antimicrobial response, as well as metabolic genes (Murphy et al., 2003). These differentially expressed genes largely overlap with the transcriptional pattern observed in C. elegans dauers (McElwee et al., 2004, 2006), which is not surprising as in the long-lived IIS mutants, the dauer program is probably activated heterochronically during adulthood.
The DAF-16/dauer longevity program: a struggle against free radicals?
In an early effort to characterise the physiology of the long-lived IIS mutant age-1, two researchers independently found that the antioxidant activity is increased in this strain (Larsen, 1993; Vanfleteren, 1993). This fits well with the free radical theory of ageing that states that free radicals, produced as a by-product of oxidative metabolism in the mitochondria, cause molecular damage that accumulates over time and underlies the functional decline that characterises ageing (Harman, 1956, 1972). In this reasoning, the high antioxidant levels in long-lived age-1 mutants would scavenge free radicals, preventing damage accumulation and lead to lifespan extension. High antioxidant activity was later also confirmed for the long-lived IIS mutant daf-2 (Houthoofd et al., 2003, 2005). Also in the transcriptomic studies mentioned above, it was confirmed that a broad antioxidant response is activated in IIS mutants (Murphy et al., 2003; Halaschek-Wiener et al., 2005). The expression and activity of the primary ROS scavengers superoxide dismutase (SOD) and catalase (CTL), in particular, are strongly upregulated. However, are antioxidant upregulation and lifespan extension causally related? An early test using RNAi knockdown of sod-3, ctl-1 and ctl-2 genes suggested that these genes are only marginally involved in IIS lifespan extension as knockdown resulted in a slight (10-15%) decrease of daf-2 lifespan (Murphy et al., 2003). Later studies using knockout mutants did not find any effect. Knocking down the two mitochondrial SODs as well as two cytosolic SODs did not shorten daf-2 lifespan, whilst knockdown of the extracellular SOD extended daf-2 lifespan even more (Doonan et al., 2008). Later it was shown that quintuple mutants, lacking all five C. elegans SODs, had a normal lifespan but were extremely sensitive to oxidative stress. This suggests that, under normal conditions, oxidative stress is not a lifespan determinant (Van Raamsdonk & Hekimi, 2012). It is still to be tested whether the lack of all five sod genes reduces the lifespan of long-lived IIS mutants.
Damage accumulation in IIS mutants
If antioxidant enzymes play no or only a minor role in IIS longevity, does oxidative damage accumulate in these worms at the same rate as in wild type controls? Several studies suggest this is not the case; IIS mutants seem to have low rates of oxidative damage accumulation. Protein carbonylation, a standard measure of oxidative protein damage, increases at slower rates in the age-1 mutant (Ishii et al., 2002) and in daf-2 mitochondria (Brys et al., 2007) compared to the wild type control. This decrease in the accrual of oxidative protein modifications in IIS mutants was confirmed in studies using mass spectrometry (Knoefler et al., 2012; Dhondt et al., 2016). This phenotype may be linked to increased antioxidant activity in IIS mutants but, again, a direct causal link to lifespan has not been shown.
Damage clearance and protein turnover
Oxidative damage may just be one type of damage that accumulates over time and its contribution to ageing may vary from species to species (Gladyshev, 2014). This can explain the limited effect of antioxidants to lifespan in the very short-lived C. elegans species. Here, more general processes that prevent damage accumulation may be of greater importance. Protein turnover has been put forward as a potential key player in ageing as it allows the maintenance of proteome quality over time (Gafni, 1990; Ryazanov & Nefsky, 2002; Tavernarakis & Driscoll, 2002). As protein turnover rates have been shown to decline strongly with age, it is comprehensible that ageing could be caused by a gradual self-reinforcing collapse of protein homeostasis (proteostasis).
The finding that autophagy is increased and genetically required for lifespan extension of IIS mutants supports this view (Melendez et al., 2003; Lapierre et al., 2013). Increased autophagy rates in IIS mutants may clear protein damage efficiently thereby postponing damage accumulation and ageing. However, not all studies are in accordance with this view. RNAi knockdown of 14 autophagy-related genes in daf-2 mutants showed that RNAi of only two genes caused lifespan shortening while the other 12 genes caused further extension of lifespan or had no effect (Hashimoto et al., 2009). Hence, the role of autophagy in IIS longevity is still unclear.
Recent findings cast further doubt on the importance of high bulk protein turnover rates in IIS longevity. Quantitative proteomics data showed a clear decrease in ribosomal subunits and translation factors in daf-2 mutants suggesting a decrease in protein synthesis rates in these worms (Depuydt et al., 2013; Stout et al., 2013). Active insulin and IGF pathways are known to support anabolic growth. Hence, a reduction-of-function mutation in these pathways, such as daf-2(e1370), is indeed expected to decrease anabolic processes such as protein synthesis. This was recently confirmed using classical 35S pulse chase labelling (Depuydt et al., 2016) and stable isotope labelling combined with mass spectrometry (Dhondt et al., 2016; Visscher et al., 2016). About half of the protein species show decreased turnover rates in the long-lived daf-2 mutants, while turnover of the other proteins remains unchanged (Dhondt et al., 2016). The IIS mutants do not seem to invest in the energy-consuming process of protein turnover to maintain proteostasis. Nevertheless, proteostasis is well maintained during adulthood in these mutants (Walther et al., 2015).
Investment in fat and glycogen synthesis
If antioxidants and protein turnover are not the (main) strategies to postpone molecular damage and ageing in IIS mutants, which processes do matter? The morphology of IIS mutants may provide a first clue: daf-2 mutants tend to accumulate large amounts of fat (Ogg et al., 1997; Depuydt et al., 2014) and glycogen (Frazier & Roth, 2009; Depuydt et al., 2014) in the intestine, hypodermis and to some extent in the body wall muscles, thereby phenocopying dauers (Fig. 2). Hence, IIS mutants seem to invest heavily in carbon storage in the form of carbohydrates and triglycerides. In dauers, accumulated fat serves as an energy source for long-term survival in the absence of feeding and its use is tightly controlled by intracellular energy sensors capable of switching lipases on and off (Narbonne & Roy, 2009). Glycogen accumulation in IIS mutants may seem linked to energy storage as well but it also relates to osmotic and anoxic stress resistance (Frazier & Roth, 2009; LaMacchia et al., 2015; Possik et al., 2015), which are important features in dauer survival.
The fat and glycogen storage phenotype is supported in proteomic studies. Many enzymes involved in fatty acid β-oxidation are upregulated in adult daf-2(e1370) mutants, while food uptake is drastically lowered, hinting at the use of stored fat as an energy source during adulthood in IIS mutants (Depuydt et al., 2014). In these mutants, most enzymes involved in carbohydrate metabolism are also strongly upregulated with glycogen synthase (GSY-1) belonging to the top most upregulated proteins. This agrees with the copious amounts of glycogen observed in the daf-2 intestine and hypodermis (Fig. 2).
Unlike in vertebrates, fat can be easily converted into glycogen in C. elegans (Fig. 3): fatty acids are degraded to acetyl-CoA by β-oxidation, which in turn can be fed to the glyoxylate cycle. The glyoxylate cycle is a shortcut of the tricarboxylic acid (TCA) cycle bypassing two decarboxylation steps. This is managed by a single polypeptide with two enzymatic functions: isocitrate lyase and malate synthase activity (Liu et al., 1995). In C. elegans, two isoforms of this multifunctional protein exist: the mitochondrial ICL-1 (Erkut et al., 2016) and the less characterised C08F11.14 (Frazier & Roth, 2009). When the glyoxylate cycle is activated, carbon is not completely oxidised and lost as CO2 but instead malate and succinate are synthesised that can be converted into oxaloacetate. The latter substrate can be fed into the gluconeogenesis pathway and the resulting glucose phosphate can finally be incorporated into glycogen. In short, fat can be converted into glycogen by glyoxylate cycle activity. It comes as no surprise that icl-1 is drastically upregulated in dauers (Holt & Riddle, 2003; Wang & Kim, 2003; Erkut et al., 2016) and IIS mutants (Murphy et al., 2003; Depuydt et al., 2014; Shen et al., 2014). Moreover, icl-1 activity is responsible for up to 45% of the lifespan extension observed in daf-2 mutants (Murphy et al., 2003; Shen et al., 2014). But how can the fat-to-glycogen conversion or glycogen itself support the longevity phenotype of IIS mutants?
Trehalose, a chemical chaperone that stabilises proteins and membranes
It is very likely that glycogen provides a fast and easily accessible source of glucose units for the synthesis of trehalose, a disaccharide with well known cytoprotective properties (Elbein et al., 2003; Perry & Wharton, 2011). In C. elegans dauers, trehalose is required for successful anhydrobiosis, an ametabolic desiccated state that allows individuals to survive sustained periods of drought stress (Erkut et al., 2011). Moreover, the glyoxylate shunt is required for anhydrobiotic survival as well in C. elegans, clearly linking fat-to-sugar conversion and trehalose metabolism (Erkut et al., 2016). Again this metabolic pattern is mirrored in IIS mutants; endogenous trehalose synthesis explains a considerable portion of the extended lifespan of daf-2 mutants. Exogenously added trehalose extends lifespan of wild-type worms while in daf-2 mutants lifespan is not further extended, which also indicates that longevity in daf-2 is (at least in part) caused by endogenous trehalose (Honda et al., 2010).
Using a recent C. elegans metabolic network model, trehalose production was predicted to be strongly supported by glyoxylate cycle activity under micro-aerobic conditions (Yilmaz & Walhout, 2016). However, in the absence of glyoxylate shunt activity, trehalose can still be produced in lower quantities, probably explaining the partial rescue of icl-1 mutation on daf-2 lifespan extension (Shen et al., 2014).
Finally, exogenously added trehalose increases protein stability, measured as the trichloroacetic acid-soluble protein fraction, in C. elegans. This fraction also correlates well with endogenous trehalose levels in C. elegans, indicating that intracellular trehalose acts as a potent protein stabiliser in this worm (Depuydt et al., 2016).
Hence, the protein and membrane stabilising properties of trehalose are likely key to IIS mutant longevity. The field of trehalose biology in C. elegans in underexplored and future work should reveal in which tissues it is synthesised, how it is transported between tissues, and whether certain tissue types benefit more from trehalose protection than others (with lifespan as a readout). Caenorhabditis elegans would also allow for easy genetic manipulation to increase artificially intracellular trehalose levels by over-expressing trehalose synthesis genes and/or knockdown of trehalases and test whether there is a beneficial effect on lifespan. Trehalose can act in concert with LEA proteins for its protective function (Chakrabortee et al., 2007). The C. elegans genome encodes a single lea gene homologue, known to be upregulated in daf-2 mutants (Dong et al., 2007; Depuydt et al., 2016) and awaiting functional characterisation.
Relevance to human biology
There is no doubt that C. elegans fuelled genetic ageing research in other models including vertebrate species. One of the most successful discoveries so far was the role of IIS signalling in lifespan determination, first revealed in C. elegans and later confirmed in many other species (Kenyon, 2010). Nevertheless, one should always bear in mind that, despite evolutionary conservation of many genes and pathways, C. elegans and humans differ widely in their molecular physiology as a result of differences in size, thermobiology and environmental challenges. Although reduced IIS pathway signalling may lead to broadly comparable phenotypes in worms and mammals, the means by which these characteristics are achieved may differ widely. IIS longevity in C. elegans is, at least in part, supported by metabolic pathways that are absent in mammals, such as glyoxylate shunting and trehalose synthesis (Fig. 3). Also the glutamate synthase homologue W07E11.1, which, like ICL-1, only occurs in bacteria, plants and nematodes, is highly upregulated in daf-2 mutants hinting at an important function in the IIS longevity phenotype (Depuydt et al., 2014). These metabolic patterns seem all part of an ancient genetic program that allows small organisms with limited mobility to undergo anhydrobiosis and survive repetitive drought/rehydration cycles (Alpert, 2006; Erkut et al., 2016). In C. elegans IIS mutants, this non-mammalian genetic program is switched on heterochronously in adults, making these worms stress resistant and long-lived in a way that cannot be directly achieved in humans. Nevertheless, detailed functional knowledge of this elaborate program may lead to future medical or pharmaceutical applications that may support human longevity or reduce frailty at advanced age.
We are grateful to Myriam Claeys for preparation of the electron microscopy samples. We are indebted to the Special Research Fund at Ghent University (BOF15/24j/013) and the Fund for Scientific Research – Flanders (G.04371.0N and G.0D64.14N) for financial support.
ChakraborteeS.BoschettiC.WaltonL.J.SarkarS.RubinszteinD.C.TunnacliffeA. (2007). Hydrophilic protein associated with desiccation tolerance exhibits broad protein stabilization function. Proceedings of the National Academy of Sciences of the United States of America 104, 18073-18078. DOI: 10.1073/pnas.0706964104
ChenA.T.GuoC.ItaniO.A.BudaitisB.G.WilliamsT.W.HopkinsC.E.MceachinR.C.PandeM.GrantA.R.YoshinaS. (2015). Longevity genes revealed by integrative analysis of isoform-specific daf-16/FoxO mutants of Caenorhabditis elegans. Genetics 201, 613-629. DOI: 10.1534/genetics.115.177998
DepuydtG.XieF.PetyukV.A.ShanmugamN.SmoldersA.DhondtI.BrewerH.M.CampD.G.2ndSmithR.D.BraeckmanB.P. (2013). Reduced insulin/insulin-like growth factor-1 signaling and dietary restriction inhibit translation but preserve muscle mass in Caenorhabditis elegans. Molecular and Cellular Proteomics 12, 3624-3639. DOI: 10.1074/mcp.M113.027383
DepuydtG.XieF.PetyukV.A.SmoldersA.BrewerH.M.CampD.G.2ndSmithR.D.BraeckmanB.P. (2014). LC-MS proteomics analysis of the insulin/IGF-1-deficient Caenorhabditis elegans daf-2(e1370) mutant reveals extensive restructuring of intermediary metabolism. Journal of Proteome Research 13, 1938-1956. DOI: 10.1021/pr401081b
DepuydtG.ShanmugamN.RasulovaM.DhondtI.BraeckmanB.P. (2016). Increased protein stability and decreased protein turnover in the Caenorhabditis elegans Ins/IGF-1 daf-2 mutant. Journals of Gerontology. A: Biological Sciences and Medical Sciences 71, 1553-1559. DOI: 10.1093/gerona/glv221
DoonanR.McElweeJ.J.MatthijssensF.WalkerG.A.HouthoofdK.BackP.MatscheskiA.VanfleterenJ.R.GemsD. (2008). Against the oxidative damage theory of aging: superoxide dismutases protect against oxidative stress but have little or no effect on life span in Caenorhabditis elegans. Genes and Development 22, 3236-3241. DOI: 10.1101/gad.504808
HouthoofdK.FidalgoM.A.HoogewijsD.BraeckmanB.P.LenaertsI.BrysK.MatthijssensF.De VreeseA.Van EygenS.MunozM.J. (2005). Metabolism, physiology and stress defense in three aging Ins/IGF-1 mutants of the nematode Caenorhabditis elegans. Aging Cell 4, 87-95. DOI: 10.1111/j.1474-9726.2005.00150.x
JonesL.M.StaffaK.PerallyS.LacourseE.J.BrophyP.M.HamiltonJ.V. (2010). Proteomic analyses of Caenorhabditis elegans dauer larvae and long-lived daf-2 mutants implicates a shared detoxification system in longevity assurance. Journal of Proteome Research 9, 2871-2881. DOI: 10.1021/pr9009639
LapierreL.R.De Magalhaes FilhoC.D.McquaryP.R.ChuC.C.VisvikisO.ChangJ.T.GelinoS.OngB.DavisA.E.IrazoquiJ.E. (2013). The TFEB orthologue HLH-30 regulates autophagy and modulates longevity in Caenorhabditis elegans. Nature Communications 4, 2267. DOI: 10.1038/ncomms3267
McElweeJ.J.SchusterE.BlancE.ThomasJ.H.GemsD. (2004). Shared transcriptional signature in Caenorhabditis elegans Dauer larvae and long-lived daf-2 mutants implicates detoxification system in longevity assurance. Journal of Biological Chemistry 279, 44533-44543. DOI: 10.1074/jbc.M406207200
McGaughS.E.BronikowskiA.M.KuoC.H.RedingD.M.AddisE.A.FlagelL.E.JanzenF.J.SchwartzT.S. (2015). Rapid molecular evolution across amniotes of the IIS/TOR network. Proceedings of the National Academy of Sciences of the United States of America 112, 7055-7060. DOI: 10.1073/pnas.1419659112
PierceS.B.CostaM.WisotzkeyR.DevadharS.HomburgerS.A.BuchmanA.R.FergusonK.C.HellerJ.PlattD.M.PasquinelliA.A. (2001). Regulation of DAF-2 receptor signaling by human insulin and ins-1, a member of the unusually large and diverse C. elegans insulin gene family. Genes and Development 15, 672-686. DOI: 10.1101/gad.867301
PossikE.AjisebutuA.ManteghiS.GingrasM.C.VijayaraghavanT.FlamandM.CoullB.SchmeisserK.DuchaineT.Van SteenselM. (2015). FLCN and AMPK confer resistance to hyperosmotic stress via remodeling of glycogen stores. PLoS Genetics 11, e1005520. DOI: 10.1371/journal.pgen.1005520
VisscherM.De HenauS.WildschutM.H.Van EsR.M.DhondtI.MichelsH.KemmerenP.NollenE.A.BraeckmanB.P.BurgeringB.M. (2016). Proteome-wide changes in protein turnover rates in C. elegans models of longevity and age-related disease. Cell Reports 16, 3041-3051. DOI: 10.1016/j.celrep.2016.08.025
WolkowC.A.MuñozM.J.RiddleD.L.RuvkunG. (2002). Insulin receptor substrate and p55 orthologous adaptor proteins function in the Caenorhabditis elegans daf-2/insulin-like signaling pathway. Journal of Biological Chemistry 277, 49591-49597. DOI: 10.1074/jbc.M207866200