Insulin/IGF-1 Signaling
The insulin/IGF-1-like signaling (IIS) pathway is an evolutionarily conserved endocrine signaling pathway that regulates food storage and growth [73]. Across taxa, reductions in insulin signaling and/or responsiveness are used as organismal signals for various types of stress – for example, IIS is involved in the entry into dauer in C. elegans. IIS pathway activity also influences lifespan in multiple different organisms, including yeast, nematodes, flies, and mammals [74-76], and it has been proposed that the regulation of longevity via the IIS pathway may have evolved to enable animals to survive harsh, stressful conditions [75]. In C. elegans, the gene daf-2 encodes a homolog of the mammalian insulin and insulin growth factor (IGF-1) receptors [7]. When ligand-bound, DAF-2 triggers the IIS pathway [77] by activating AGE-1, a homolog of p11, the catalytic subunit of human phosphatidylinositol 3-kinase (PI3K) [6]. (Of note: allelic variation in PI3K appears to affect human longevity, as determined by comparing the representation of genotypes in younger and older age groups [76, 78]). Once activated by DAF-2, AGE-1 catalyzes the formation of the second messenger phosphatidylinositol-3,4,5-triphosphate (PIP3), which in turn binds and activates AKT-1 and AKT-2, homologs of the mammalian serine/threonine kinase Akt/PKB [6, 79, 80]. Activated AKT proteins form a complex with SGK-1, a homolog of the serum- and glucocorticoid-inducible kinase SGK. All three kinases in this complex can directly phosphorylate the FOXO-family transcription factor DAF-16, constraining it to the cytoplasm [81-83]. Absent IIS activity, dephosphorylated DAF-16 enters the nucleus and regulates the transcription of a variety of genes regulating lifespan, dauer formation, stress resistance, development, and metabolism, among others [84]. IIS activity is antagonized by DAF-18, a homolog of the human tumor suppressor PTEN. DAF-18 limits AKT-1 and AKT-2 activity by dephosphorylating PIP3 [5, 85]. In the nucleus, DAF-16 is further regulated by several different proteins, including the transcription factor HSF-1 and the SMEK ortholog SMK-1 [86-90].
Certain mutant daf-2 alleles confer two-fold increased longevity and resistance to various stresses, including heat shock, oxidative stress, heavy-metal stress, and infection [12, 18, 72, 91, 92]. These mutants express higher levels of antioxidant enzymes such as catalase and SOD [92]. Additionally, daf-2 mutants often constitutively enter the dauer state, and some alleles also confer other pleiotropic traits including reduced adult motility, abnormal adult morphology, and reduced brood size [93]. Mutants in daf-2 also have an increase in autophagy, a catabolic process through which proteins and organelles are delivered to lysosomes for degradation. The autophagy gene bec-1, the homolog of mammalian beclin 1, is required for the increased lifespan of daf-2 mutant animals [94], suggesting that autophagy is an essential lifespan-prolonging component of the stress responses typically held in check by IIS activity.
Many mutations in IIS pathway genes downstream of daf-2 also modulate C. elegans lifespan and stress resistance. age-1(hx546) animals, which have a reduction-of-function point mutation in the age-1 gene, exhibit a 40-110% extension in mean lifespan, and higher tolerance to thermal stress [12, 14, 15, 34, 95]. Furthermore, the relative activity of catalase and SOD in these animals, as compared to wild-type, increases with age in a manner that correlates with an age-dependent increase in the mutant animals' relative stress resistance [14, 34, 95]. Null mutants for age-1, in which the AGE-1 protein is truncated upstream of its kinase domain, have a mean lifespan about ten times that of wild-type animals [15]. Compared with the weaker age-1(hx546) allele, these animals are significantly more tolerant to oxidative stress, but also slightly less tolerant to thermal stress [15]. It is also worth noting that classic lifespan-extending mutants, including daf-2 and age-1 nematodes, exhibit heightened resistance to various bacterial pathogens [18].
Mutations in akt-1 and akt-2 have a surprisingly small effect on C. elegans lifespan. akt-1(mg306) mutants, which have a point mutation in the akt-1 gene, live only around 10% longer than wild-type [96, 97]. The mutants akt-1(ok525) and akt-2(ok393) lack most of the kinase domain necessary for Akt/PKB activity, yet do not exhibit extended lifespan or increased stress resistance, while akt-2(ok393) mutants subjected to akt-1(RNAi) have only a 19% lifespan extension as well as a weak increase in resistance against high temperatures and oxidative stress [79, 81]. On the other hand, sgk-1(RNAi) significantly increases lifespan as well as resistance to heat and oxidative stress, suggesting that SGK-1 may be more important for the regulation of stress response and lifespan, whereas AKT-1 and AKT-2 are more important in regulating dauer formation [81].
Mutations in daf-18, an inhibitor of the IIS pathway, decrease C. elegans lifespan [5, 98]. A mutation in daf-18 partially suppresses the increased lifespan of daf-2 animals [98], and daf-18 also necessary for the nuclear localization of daf-16 in daf-2 animals [99].
Interestingly, daf-18 mutants subjected to hormetic levels of thermal stress do not have longer lifespans [100]. Though other C. elegans genes have been found to be required for various forms of hormesis (the radiation-sensitive mutants rad-1 and rad-2 mutants do not exhibit increased resistance to X-ray irradiation subsequent to pre-treatment with oxidative stress [48]), daf-18 appears to be the only such gene that does not also directly disrupt stress response mechanisms. Further, it has been proposed that VAB-1, an Eph receptor tyrosine kinase, diminishes DAF-18 activity by phosphorylation [101]. Both vab-1 mutants and C. elegans overexpressing daf-18 live longer and are more sensitive to dauer conditions than wild-type [101], but the potential regulation of DAF-18 by VAB-1 remains to be investigated in the context of lifespan extension and stress tolerance.
In all cases mentioned above, the increased lifespan and stress resistance of IIS pathway mutants is dependent on daf-16 [15, 81, 93, 98]. Null mutants in daf-16 have shortened lifespan, and their germ cells are hypersensitive to ionizing radiation [88, 102]. Interestingly, daf-16 mutants exhibit similar levels of resistance to UV, oxidative stress, and pathogenic stress as compared to wild-type animals [18, 103], although daf-16 animals are less thermotolerant than wild-type [104], and C. elegans which overexpress daf-16 are more resistant to P. aeruginosa [105].
Since DAF-16 is thought to be the main target of the IIS pathway, it has long been of interest to identify the genes which it regulates. The stress-response genes sod-3, which codes for a mitochondrial MnSOD, and mtl-1, which codes for a metallothionein homolog, were among the first clear direct DAF-16 targets [72, 92]. Since then, a series of studies has revealed many antioxidant, metabolic, heat-shock, and antibacterial genes to be targeted by DAF-16 [106-109]. Furthermore, a DNA microarray study aimed at finding DAF-16 targets also uncovered two putative DAF-16 binding sites based on the over-representation of these sequences upstream of target genes [107]. One of these sequences has also been shown to be bound by DAF-16 in vitro [110].
In addition to IIS, other longevity pathways also converge on DAF-16. One such pathway involves the sirtuins, a highly conserved family of NAD+-dependent protein deacetylases that includes SIR-2.1 in C. elegans, Sir2 in yeast, and the SIRT proteins in mammals. Among other functions, sirtuins deacetylate histones, thereby playing a role in the regulation of chromatin state [111-113]. The sirtuins are known to regulate stress response and cell senescence – for instance, Sir2α in yeast represses stress-induced, p53-mediated apoptosis [40, 114-116]. In C. elegans, SIR-2.1 activates DAF-16, and sir-2.1 mutants are short-lived and stress sensitive, while overexpression of sir-2.1 increases lifespan by 50% in a daf-16 dependent manner [86, 87]. Following heat shock, SIR-2.1, which is localized to the nucleus, physically interacts with DAF-16, and this interaction depends on PAR-5 and FTT-2, which are homologs of 14-3-3 proteins, small acidic proteins that bind phosphoserine and phosphothreonine residues in particular sequence contexts [117]. Sir-2.1 is however not required for the longevity of IIS pathway mutants, leading to the suggestion that SIR-2.1 and the 14-3-3-like proteins act in a stress-dependent manner to activate DAF-16 and increase lifespan, which is nevertheless independent of IIS [117].
Another pathway converging on DAF-16 involves a member of the JNK family. The mammalian mitogen-activated c-Jun N-terminal kinases (JNKs) can be activated by cytokines or environmental stress and are known to regulate processes including development and cell survival [118]. JNK-1, the C. elegans homolog of mammalian JNK, is thought to regulate lifespan in parallel to the IIS pathway, by regulating the subcellular localization of DAF-16 [119]. Whereas phosphorylation of DAF-16 by the AKT and SGK kinases constrains it to the cytoplasm, phosphorylation of DAF-16 by JNK-1 causes it to translocate to the nucleus [119]. Consistent with this observation, jnk-1 mutants have shorter lifespan than wild-type, and C. elegans overexpressing jnk-1 exhibit extended lifespan in a manner independent of IIS but dependent on daf-16 [119]. The putative transcription factor LIN-14, which is required for normal lifespan (see below), has also been found to regulate DAF-16, though the mechanism of this regulation remains unknown [89].
Nuclear localization of DAF-16 is neither necessary nor sufficient for lifespan extension [99]: many additional regulators act on this transcription factor once it has translocated to the nucleus, and translocation is not increased in certain daf-16-dependent longevity mutants. When mutations are introduced at the AKT phosphorylation sites on DAF-16, the transcription factor localizes to the nucleus, but the animals expressing this mutant DAF-16 do not constitutively become dauers and do not have extended lifespan [83]. The transcription factor HSF-1, which regulates the heat-shock response in C. elegans, is thought to act together in the nucleus with DAF-16 to activate the transcription of certain genes, including small heat-shock proteins [88]. Similarly, the SMEK ortholog SMK-1 appears to regulate the transcriptional specificity of DAF-16 already localized to the nucleus [90]. The mammalian protein SMEK1 is phosphorylated in response to stress and is involved in the regulation of glucose metabolism under varying conditions, such as fasting [120]. It is thus conceivable that SMK-1 in C. elegans also modulates DAF-16 specificity as a response to certain stressors. In fact, the smk-1 gene in C. elegans is required for daf-2 mutants to exhibit extended lifespan, as well as resistance to pathogenic, UV, and oxidative stress. However, reduced activity of smk-1 does not affect the thermotolerance of daf-2 nematodes [90]. In addition, the C. elegans homolog of the highly conserved protein host cell factor 1 (HCF-1) is localized to the nucleus and physically associates with DAF-16 [121, 122]. Mammalian HCF-1 plays key roles in cell-cycle progression and has been shown to act primarily by binding to transcription factors and by assembling protein complexes [121, 122]. Mutations in the C. elegans gene hcf-1 result in a 40% increase in lifespan and greater resistance to oxidative stress, but not to heat stress [121]. It has been proposed that HCF-1 negatively regulates DAF-16 by forming a complex with DAF-16 in the nucleus and thus limiting access of DAF-16 to its target promoters [121].
Conversely, some results have indicated that DAF-16 function may not be limited to the nucleus [123]. First, though the longevity of age-1 mutants depends on daf-16, these animals do not have higher nuclear localization of DAF-16 [124]. Next, either knockout of or RNAi against cep-1, a transcription factor that regulates the germline response to DNA damage and is an ortholog of the tumor-suppressor p53, increases longevity in C. elegans in a daf-16-dependent manner, yet RNAi against cep-1 does not alter DAF-16 nuclear localization [125]. Further, cep-1 mutants, though long-lived, are not resistant to heat, oxidative, UV, or pathogenic stress [125]. A related result involves the conserved gene smg-1, which encodes a serine-threonine kinase that functions in nonsense-mediated mRNA decay. Inactivation of smg-1 by mutation or RNAi confers increased lifespan and resistance to oxidative stress in a daf-18-, daf-16-, and cep-1-dependent manner, yet in these animals DAF-16 is not localized to the nucleus to any degree beyond that in controls [126]. Finally, nuclear localization of DAF-16 in wild-type C. elegans occurs in response to starvation, heat shock, and oxidation (by treatment with juglone), but very little or not at all in response to UV irradiation [124, 127].
The effect of the IIS pathway on C. elegans lifespan seems to involve communication between cells and between tissues. Some daf-2 mosaic animals exhibit longer lifespans than wild-type, suggesting that daf-2 acts cell nonautonomously to modulate C. elegans lifespan (if daf-2 acted cell autonomously, then the daf-2(+) tissues of the mosaic would be expected to die earlier and thus kill the animal, resulting in a normal rather than a lengthened lifespan) [128]. One potential explanation for the cell-nonautonomy of daf-2 is that the IIS pathway activates the expression of the insulin/IGF-1 homolog INS-7, which may potentially activate the IIS pathway in other cells [107]. Other potential signaling molecules regulated by DAF-2 and DAF-16, such as the putative secreted protein encoded by scl-1, may also contribute to the cell non-autonomy of daf-2 [107]. Counterintuitively, tissue-specific studies suggest that DAF-2 and DAF-16 act in different tissues to influence C. elegans lifespan: DAF-2 functions mostly in the nervous system to modulate C. elegans longevity [129], whereas rescue of daf-16 in the intestine is the most effective in restoring the lifespan and thermotolerance of daf-2;daf-16 mutants [130]. To explain this discrepancy in tissue-of-action, it has been proposed that daf-2(+) neurons may produce a signal that decreases DAF-16 function in intestinal cells; however an assay using the sod-3::gfp reporter suggests that DAF-16 activity in these animals remains high in the intestine [130]. Experiments on this subject remain difficult to interpret.
A more recently identified target of the IIS pathway is SKN-1, a transcription factor that orchestrates the oxidative stress response in adult C. elegans, independent of DAF-16 activity, by activating the expression of genes coding for enzymes including catalases, SODs, and some glutathione S-transferases [131-133]. skn-1 mutants are short-lived and exhibit sensitivity to oxidative stress, and overexpression of SKN-1 in intestinal cells increases lifespan and oxidative stress resistance [131-133]. Mutations in skn-1 also suppress the longevity and oxidative stress resistance of daf-2 mutants [131]. Recently, it has been shown that IIS directly inhibits SKN-1 via phosphorylation by AKT-1, AKT-2, and SGK-1, whereas in IIS mutants, SKN-1 accumulates in the nucleus of intestinal cells and activates expression of genes involved in the oxidative stress response [131, 132]. Consistent with SKN-1 regulation of stress resistance, RNAi of two genes upregulated by SKN-1, nlp-7 and cup-4, leads to reduction in lifespan and oxidative stress resistance [134]. RNAi of ent-1, which is down-regulated by SKN-1, increases oxidative stress resistance without affecting longevity [134].
Many IIS pathway genes were initially identified in C. elegans based on their regulation of dauer formation. Dauer larvae are stress resistant, and because their growth is arrested they can have significantly lengthened lifespans, though nematodes that have exited dauer have a normal lifespan from that point on. In many respects, dauers and long-lived IIS mutant animals appear to have many similarities. However, the longevity of IIS pathway mutants can be separated from the effects on dauer formation. “Class 2” daf-2 mutants exhibit dauer-like traits as adults, while “class 1” mutants do not, yet both classes of mutants have extended lifespan [75, 93]. Longevity extension and the appearance of dauer-like traits can be separated by manipulating the timing of daf-2 expression, as demonstrated by an elegant work using RNAi against daf-2 [135]. Initiating daf-2(RNAi) in adulthood confers essentially the same degree of longevity and stress resistance as initiating daf-2(RNAi) at hatching; conversely, knockdown of daf-2 during development only does not have extended lifespan at 20°C, which suggests that the daf-2 pathway acts during adulthood, but not during development, to influence adult lifespan [135]. Further, the temporal requirements of longevity by IIS are also separate from those of other longevity pathways, suggesting that these pathways increase lifespan independently: reductions in mitochondrial respiration only increase lifespan if enacted early in the nematode's life, while dietary restriction extends longevity regardless of the timing [123].
The IIS pathway mutants reflect a correlation between longevity and stress resistance. Many of the genes downstream of daf-2 and daf-16 have clear roles in stress resistance, although it is not as clear how they influence aging. However, inconsistencies exist in the correlation between longevity and stress resistance. Moreover, current results give only an incomplete understanding of their association, and further study of the lifespan and stress response of various IIS pathway mutants is necessary.
Mitochondrial respiration
Mitochondrial respiration defects caused by mutation or RNAi can often extend C. elegans lifespan [8, 106, 109, 151-154], but the influence of these defects on stress resistance varies. For instance, isp-1 encodes the iron sulfur protein of complex III in the electron-transport chain in C. elegans, and a mutation in this gene yields significantly increased lifespan and resistance against paraquat [8, 9]. However, though RNAi inactivation of components of the electron-transport chain often increases resistance to H2O2 and high temperatures, it also increases sensitivity to paraquat [152]. Similarly, RNAi against cchl-1, the C. elegans Cytochrome c heme lyase – an enzyme that covalently attaches heme to cytochrome c, which is essential to the electron-transport chain [155] – confers longer lifespan, but not resistance to paraquat or H2O2 [123, 152].
The mechanisms behind the extended lifespan of nematodes with mitochondrial dysfunction are not well understood. Since ROS are produced as byproducts of mitochondrial respiration, one possible explanation is that mutations in mitochondrial respiration genes reduce ROS production. In support of this hypothesis, the long-lived isp-1 mutant has decreased oxygen consumption. The leucyl-tRNA synthetase gene lrs-2 is needed for the proper translation of genes encoded by the mitochondrial genome; lrs-2 mutations, which are thought to result in low electron-transport chain activity, extend C. elegans lifespan [106, 109, 152]. However, several observations controvert the theory that reduced ROS accounts for the longevity of mitochondrial mutants. For one, lifespan extension by mutations in mitochondrial respiration genes does not appear to correlate with decreases in lifelong oxidative stress as measured by whole-animal protein carbonyl content [156]. Moreover, at least one mitochondrial respiration mutant, clk-1, does not exhibit reduced respiration and has normal or slightly elevated levels of ATP, although it does have increased lifespan and reduced behavioral rates [123, 151, 154, 157]. (The clk-1 gene encodes a hydroxylase necessary for the biosynthesis of ubiquinone, which shuttles electrons as part of the mitochondrial respiratory chain [158].) Furthermore, not all disabling mutations in mitochondrial respiration lengthen lifespan. Mutants of mev-1, which codes for a subunit of complex II, have reduced mitochondrial respiratory rates and a thermotolerance comparable to wild-type, but also have a 30% decrease in lifespan [104, 159]. However, mev-1 mutants express only half as much SOD as wild-type and are hypersensitive to paraquat [160], which makes it difficult to clearly interpret lifespan results from this strain. Together, these results suggest that the longevity of C. elegans mitochondrial respiration mutants cannot be explained solely by lower rates of oxygen consumption and ROS production.
Given the many observations inconsistent with the theory of reduced ROS, alternative hypotheses have been put forth to explain the longevity of mitochondrial respiration mutants. It has been suggested that some mitochondrial mutations may slightly increase ROS levels, resulting in lifespan extension by a hormetic effect [106, 161, 162]. Another proposal is that reduced respiration early in life may stimulate a regulated response that subsequently extends C. elegans lifespan [75, 123, 154]. Both ideas hold that reduced respiration may cause a type of stress, and the response of the nematode's cells to this stress might have a life-extending effect. Consistent with this theory, inhibition of certain mitochondrial respiration genes by mutation or RNAi leads to increased expression of certain cell-protective and metabolic genes, and greater abundance of mitochondrial DNA [163]. Further, RNAi against the potential signaling genes fstr-1 and fstr-2 reduces lifespan, increases behavioral rates, and suppresses gene expression changes in clk-1 mutants [163]. These results suggest that fstr-1 and fstr-2 may be part of a pathway that triggers a life-extending response to reduced respiration in clk-1 mutants [163]. However, these genes are not required for the longevity of all other mitochondrial mutants [163].
In most cases, lifespan extension by mutations in mitochondrial respiration genes does not depend on daf-2 and daf-16, and is therefore independent of the IIS pathway [8, 106, 109, 151, 152, 154]. However, qm50, a point mutation in isp-1, does not extend the lifespan of daf-2 mutants, and its effect on lifespan is independent of daf-16 [8, 75].