The role of insulin/IGF-like signaling in C. elegans longevity and aging

Aging studies in mammals are challenging owing to relatively long life spans and a limited set of genetic tools that are capable of assessing the genome-wide changes associated with disease. By contrast, C. elegans lives for only 2–3 weeks and is easily propagated in culture, facilitating rapid genetic analyses. Genetic resources for C. elegans include hundreds of publicly available mutant strains; full-genome RNA interference (RNAi) libraries, which allow efficient knockdown of gene expression by simply feeding worms with bacteria expressing double-stranded RNAs (dsRNAs); gene-deletion strains, which are being generated for each of the ∼19,000 genes; and >2000 promoter::gfp strains, which allow characterization of cellular and sub-cellular gene expression patterns. Moreover, all of the 959 lineage-invariant somatic cells of C. elegans are post-mitotic, making the worm an attractive model for aging in non-dividing cells such as mammalian muscle cells and neurons that depend on maintenance instead of cell turnover for their longevity. The mitotically dividing C. elegans germline, which also ages (Garigan et al., 2002), provides an excellent model for the study of age-related reproductive decline (Hughes et al., 2007). Although signals from the germline influence life span (Hsin and Kenyon, 1999), and life span and reproductive span are linked (Luo et al., 2009), post-mitotic somatic cell integrity limits the overall life span of the animal (Herndon et al., 2002).

Importantly, C. elegans shares many similarities with mammals and humans, including aging and functional senescence (Fig. 1). Aging worms experience a decline in mobility (Herndon et al., 2002; Huang et al., 2004; Hsu et al., 2009), chemotaxis (Murakami and Murakami, 2005) and reproductive capacity (Hughes et al., 2007), and become increasingly susceptible to lethal infections (Garigan et al., 2002; Garsin et al., 2003). Exogenous expression of proteins such as amyloid beta (1–42) [Aβ(1–42)] and polyglutamine (polyQ) also make the worm a model for late-onset diseases such as Alzheimer’s disease and Huntington’s disease (Link, 1995; Faber et al., 1999). These features make C. elegans an excellent system to study conserved aging pathways and age-related diseases in mammals.

Over 20 years ago it was discovered that mutations in daf-2 and age-1 double the life span of worms (Friedman and Johnson, 1988; Kenyon et al., 1993). Subsequent cloning of these genes uncovered the importance of the insulin/insulin-like growth factor-1 (IGF-1) signaling (IIS) pathway in aging regulation. daf-2 encodes the only insulin/IGF-1 receptor expressed in worms, and age-1 is the catalytic subunit of the downstream phosphoinositide 3-kinase (PI3K) (Morris et al., 1996; Kimura et al., 1997). In addition to the overall life span extension, worms bearing either of these mutations are highly resistant to oxidative stress (Honda and Honda, 1999), hypoxia (Scott et al., 2002), heat stress (Lithgow et al., 1995), heavy metals (Barsyte et al., 2001) and bacterial pathogens (Garsin et al., 2003). These findings support the notion that aging is a genetically regulated process and that mutations in longevity-linked genes delay the onset of stochastic decline (Herndon et al., 2002).

The IIS pathway also acts as a food and stress sensor during development. When food is abundant, worms develop rapidly and uninterrupted through four larval stages to reach adulthood. If worms develop in hot, food-limited or overcrowded conditions, they enter an alternative long-lived larval state called dauer in which reproductive maturity is delayed and stress resistance increases. Favorable conditions promote IIS-regulated exit from dauer, and development to reproductive adults (Cassada and Russell, 1975). In fact, daf-2 mutants were originally characterized by a prominent dauer-constitutive phenotype (Riddle et al., 1981).

Both the long-lived and dauer phenotypes of daf-2 worms are dependent on the downstream forkhead transcription factor DAF-16 (Riddle et al., 1981; Kenyon et al., 1993). Briefly, signaling through DAF-2 activates PI3K, which leads to the phosphorylation of DAF-16 and its inactivation by nuclear exclusion. In the absence of IIS or in daf-2 or age-1 mutants, DAF-16 enters the nucleus and enacts a transcriptional program that doubles worm life span (Lin et al., 1997; Ogg et al., 1997; Lee et al., 2001). Loss of daf-2 or age-1 also slows the age-related declines described above (Herndon et al., 2002; Huang et al., 2004; Garigan et al., 2002; Garsin et al., 2003); thus, the IIS pathway regulates longevity through its modulation of aging processes. This inverse correlation between aging phenotypes and survival allows the study of aging through evaluation of life span.

IIS is an evolutionarily conserved pathway that regulates life span across many species. Drosophila melanogaster, like C. elegans, has a single insulin-like receptor (dInR) that, when mutated, extends life span in a manner that is dependent on its daf-16 homolog, dFOXO (also known as foxo) (Tatar et al., 2001). Overexpression of dFOXO exclusively in the fat body (Hwangbo et al., 2004), or mutation of the insulin receptor substrate protein chico, similarly extends life span (Clancy et al., 2001). Mammals encode several daf-2 homologs [IGF-1 receptor (IGF-1R), insulin receptor (IR)-A and IR-B] that display even greater complexity in their ability to form multiple homodimer and heterodimer pairs (Benyoucef et al., 2007). Despite these differences in insulin receptor expression, the functional consequences are similar, as reduced IIS extends life span in multiple mammalian species. Longevity in dogs is inversely proportional to body size (Greer et al., 2007). Interestingly, a single nucleotide polymorphism in IGF-1 is associated with small size, suggesting that IIS influences aging in dogs (Sutter et al., 2007). Heterozygous IGF-1R knockout mice are long-lived (Holzenberger et al., 2003), and adipose tissue-specific insulin receptor knockout (FIRKO) mice also exhibit increases in life span (Bluher et al., 2003). Further evidence comes from studies of the klotho (Kl) gene, which appears to regulate insulin sensitivity. Klotho knockout mice exhibit symptoms of enhanced aging, whereas overexpression of klotho results in repressed IIS and, consequently, life span extension (Kuroo et al., 1997; Kurosu et al., 2005).

Recent studies suggest that IIS and aging are coordinated processes in humans as well. Variants of klotho in humans (KL-VS) are associated with age (Arking et al., 2002). Moreover, genetic variations in IIS pathway components are linked to long life (Pawlikowska et al., 2009). Mutations in a human daf-16 homolog, FOXO3a, are linked to increased longevity (Willcox et al., 2008; Anselmi et al., 2009; Flachsbart et al., 2009). Long-lived men who are homozygous for the FOXO3a GG genotype also display greater insulin sensitivity (Willcox et al., 2008). In addition, IGF-1R mutations are highly represented in populations of centenarians (Suh et al., 2008). The highly conserved functional nature of IIS between worms, flies and mammals, including humans, demonstrates that findings from C. elegans can contribute greatly to our understanding of aging in higher organisms.

Understanding the molecular nature of aging requires identification of the transcriptional and cellular changes that occur as a consequence of IIS and, in particular, DAF-16/FOXO activity. Genome-wide analyses of DAF-16 targets in C. elegans have contributed to our knowledge of how life extension is achieved in the worm. It is likely that some of these genes will also influence mammalian life span, with the goal being to identify targets for the treatment of aging and age-associated diseases.

Several approaches have been used to identify DAF-16 downstream targets, including bioinformatics, chromatin immunoprecipitation (chIP) and serial analysis of gene expression (SAGE) (Lee et al., 2003; Halaschek-Wiener et al., 2005; Oh et al., 2005). However, the greatest wealth of data was generated using unbiased microarray analyses. A comparison of daf-2 and daf-2/daf-16 RNAi-treated or mutant worms revealed that both activation and repression of many genes is required for longevity (Murphy et al., 2003). In fact, no single DAF-16 target that has been tested can recapitulate the long-lived phenotype of daf-2 worms (Murphy et al., 2003), suggesting that many disparate cellular pathways are regulated in concert to extend life span (Murphy et al., 2003; McElwee et al., 2004).

Microarray analyses identified several genes that were expected to be involved in stress response, including sod-3 (manganese superoxide dismutase), mtl-1 (metallothionein), ctl-1 (catalase) and hsp-16.2 (heat shock protein), which were upregulated in daf-2 mutants, as well as several novel categories of gene that had not been previously associated with life span extension (Murphy et al., 2003; McElwee et al., 2004). Pathogen resistance factors, cytochrome P450s, and various steroid, lipid and carbohydrate metabolism genes are all upregulated by DAF-16 (Murphy et al., 2003). Several other gene classes are conversely repressed by DAF-16, including some involved in neuronal signaling, apolipoprotein binding, RNA binding, DNA replication and protein turnover (Murphy et al., 2003). These genes antagonize life span and are accordingly downregulated in daf-2 mutants. Many uncharacterized life span-enhancing or - shortening genes were also revealed in these studies, suggesting that additional pathways influencing longevity remain undiscovered (Murphy et al., 2003).

It is currently unknown whether the DAF-16/FOXO-dependent transcriptional changes that have been observed in worms are recapitulated in Drosophila and mammals. Although some longevity mechanisms will prove specific to C. elegans, several similarities in known FOXO targets across species suggest that the life span-extending pathways identified in worms will extend to many organisms.

C. elegans encodes ∼40 insulin-like molecules (INS) that are expressed primarily in neurons, although a few are also expressed in the intestine (Pierce et al., 2001; Li et al., 2003; Murphy et al., 2003; Murphy et al., 2007). Worm insulins act as either DAF-2 agonists (such as INS-7 and DAF-28) or antagonists (INS-1 and INS-18), and it has been proposed that INS act in opposing pairs to regulate IIS in response to environmental signals (Murphy et al., 2007). Insulins are secreted from tissues and have both autocrine and paracrine effects that modulate DAF-2 signaling. Recent evidence also suggests that INSs have endocrine effects on IIS that influence the coordination of aging throughout the body (Murphy et al., 2007). For example, laser ablation of sensory taste neurons extends life span in a daf-16-dependent manner, indicating that environmental signals in one tissue influence the state of aging elsewhere (Alcedo and Kenyon, 2004).

In a current model of IIS endocrine signaling, environmental cues trigger the release of INS and other hormones from sensory neurons that subsequently regulate IIS in peripheral tissues. The central nervous system (CNS), intestine and gonad respond by producing secondary hormones that modulate DAF-16 activity in the entire worm. daf-2 and age-1 expression are required in the nervous system for non-cell-autonomous regulation of IIS. Mosaic worms that have lost daf-2 activity solely in a subset of neurons are long-lived (Apfeld and Kenyon, 1998). Neuronal expression of wild-type age-1 or daf-2 in their respective mutant backgrounds is both necessary and sufficient to restore normal life span of long-lived mutants (Wolkow et al., 2000).

The tissue-specific requirement for daf-16 is markedly different. Neuronal daf-16 expression in a daf-16;daf-2 background resulted in a small (5–20%) extension of life span, whereas daf-16 rescue in the intestine, which serves as the worm’s adipose tissue, partially recapitulated the long-lived phenotype (50–60%) (Libina et al., 2003). However, daf-2 mutants expressing only non-intestinal daf-16 also exhibited substantially extended life span (70–80% over background) (Libina et al., 2003). Furthermore, reconstitution of both neuronal and intestinal daf-16 in daf-16;age-1 mutants produces an additive effect on life span, suggesting that multiple sources of DAF-16 activity regulate aging (Iser et al., 2007).

Similarly, adipose tissue and the nervous system play a role in aging regulation in mice and flies. Fat-specific insulin receptor knockout (FIRKO) mice and flies expressing dFOXO in the fat body have increased longevity (Bluher et al., 2003; Hwangbo et al., 2004). Mice lacking neuronal IGF-1 receptors or insulin receptor substrate-2 (Irs2) are similarly long-lived (Taguchi et al., 2007; Kappeler et al., 2008).

Observations that daf-16 expression in the worm intestine affects DAF-16 activity in distant tissues, and the discovery that DAF-16 regulates the expression of insulin-like peptides, led to the discovery of a positive feedback loop (Libina et al., 2003; Murphy et al., 2003). RNAi knockdown of the DAF-2 agonist ins-7 extends life span (Murphy et al., 2003). In a process termed FOXO-to-FOXO signaling, DAF-16 represses expression of ins-7 in the worm intestine. Lower INS-7 levels reduce DAF-2 activity in the entire worm, propagating DAF-2/DAF-16 signaling (Murphy et al., 2007). Endocrine regulation of aging is probably not unique to C. elegans. Several lines of evidence suggest that Drosophila neuronally derived insulin-like peptides (dILPs) act in peripheral tissues to influence IIS and aging (Hwangbo et al., 2004; Broughton et al., 2005; Bauer et al., 2007; Geminard et al., 2009). Moreover, mice with defects in growth hormone and reduced IGF-1 signaling are long-lived (Flurkey et al., 2001; Holzenberger et al., 2003).

The study of C. elegans provides a unique opportunity to model the complex genetics and biochemistry of several age-associated diseases in addition to the study of overall longevity. Systems have been developed for the study of Alzheimer’s disease [human Aβ(1–42), Tau] (Link, 1995; Kraemer et al., 2003), Parkinson’s disease (α-synuclein) (Lakso et al., 2003), Huntington’s disease (polyQ) (Faber et al., 1999), amyotrophic lateral sclerosis (ALS) (sod-1) (Wang et al., 2009), Hutchinson-Gilford progeria syndrome (nuclear lamin) (Haithcock et al., 2005), cancer (ionization-induced apoptosis) (Bergamaschi et al., 2003) and diabetes (Schulz et al., 2007; Lee, 2009). Mutations in IIS appear to ameliorate the toxic effects of Aβ(1–42) and polyQ (Morley et al., 2002; Hsu et al., 2003; Florez-McClure et al., 2007). These findings demonstrate promise for identifying genes that influence disease outcomes. As additional disease models are established and specific genetic targets are identified, drug screens will facilitate the discovery of novel therapeutics. Such screens have successfully uncovered drugs that extend life span in worms (Petrascheck et al., 2007).

There is a continual search for genetic factors that influence both overall aging and the onset and progression of age-related diseases. The linkage of IIS to aging using the simple worm model C. elegans has laid the foundation for detailed molecular studies aimed at understanding and modulating longevity. Using disease models and drug screens, new findings in the worm can be tested in mammals with the goal of improving the extent and quality of life in humans. Such an achievement would have profound health, social and economic impacts on society.