Regulation of C. elegans Life-Span by Insulinlike Signaling in the Nervous System

See allHide authors and affiliations

Science  06 Oct 2000:
Vol. 290, Issue 5489, pp. 147-150
DOI: 10.1126/science.290.5489.147


An insulinlike signaling pathway controlsCaenorhabditis elegans aging, metabolism, and development. Mutations in the daf-2 insulin receptor–like gene or the downstream age-1 phosphoinositide 3-kinase gene extend adult life-span by two- to threefold. To identify tissues where this pathway regulates aging and metabolism, we restored daf-2pathway signaling to only neurons, muscle, or intestine. Insulinlike signaling in neurons alone was sufficient to specify wild-type life-span, but muscle or intestinal signaling was not. However, restoring daf-2 pathway signaling to muscle rescued metabolic defects, thus decoupling regulation of life-span and metabolism. These findings point to the nervous system as a central regulator of animal longevity.

Each species has a characteristic life-span, ranging from 10 days for the nematode Caenorhabditis elegans to 80 years for humans. Despite these vast differences in life-span, shared features of aging in diverse species support the existence of a common mechanism for life-span determination (1). Reductions in caloric intake, insulin/insulinlike growth factor–I (IGF-I) signaling, and free radical levels can lengthen the life-span of animals as divergent as nematodes,Drosophila, and mammals (1–3). Mutations that decrease C. elegans daf-2 insulin/IGF-I–like receptor orage-1 phosphoinositide 3-kinase signaling result in severalfold extension of adult life-span (2, 4,5) and increased accumulation of fat (2,6–8). Null mutations in daf-2 orage-1 cause constitutive arrest at the dauer larval stage; dauer larvae have slowed metabolic rates, store large amounts of fat, express high levels of antioxidant enzymes such as catalase and superoxide dismutase (SOD), and live longer than reproductive adults (9). One reasonable hypothesis is that free radicals generated as by-products of metabolism damage cellular components (10). The lower level of free radicals in daf-2insulinlike signaling mutants is essential for life-span extension: The life-span extension in a daf-2 mutant requires the activity of a cytosolic catalase ctl-1 (11).

The cells where daf-2 pathway signaling is required for signaling normal life-span are not known. Insulinlike signaling may regulate metabolism and free radical production directly in aging skin or muscle, or these pathways may act in key signaling centers that then coordinately control the senescence of the entire organism. In addition, it is not clear whether insulin/IGF-I regulation of life-span is simply coregulated with metabolism or whether the metabolic shifts are mechanistically connected to the life-span regulation. Several components of the daf-2 pathway, such as akt-1,pdk-1, and daf-16, are widely expressed throughout development (12–14). Studies ofdaf-2 genetic mosaic animals showed that animals lackingdaf-2 activity from the entire AB cell lineage, which generates nearly all of the hypodermis and nervous system and half of the pharynx, have extended life-spans (15). However, mosaic animals lacking daf-2 activity from blastomere daughters of AB, which generate about half of the hypodermis, nervous system, and pharynx, did not show extended life-spans. These studies showed thatdaf-2 can act nonautonomously to regulate life-span but did not assign daf-2 longevity control to particular cell types.

To define the cell type(s) from which the daf-2 insulinlike signaling pathway functions to control C. elegans life-span, metabolism, and development, we restored daf-2 pathway function to restricted cell types by using distinct promoters to express daf-2 or age-1 cDNAs in either neurons, intestine, or muscle cells of a daf-2 or age-1mutant (16–22). Long life-span, metabolic changes, and dauer arrest were tested in these transgenic animals (Table 1). Because regulation of longevity may require gene activity over the entire life of the animal, the expression of green fluorescent protein (GFP) fusions to these promoters was confirmed to continue in aged animals (23).

Table 1

Phenotypes of animals with cell-type–restricteddaf-2 pathway signaling.

View this table:

The long life-span of daf-2 and age-1 mutants was rescued by neuronal expression of daf-2 or age-1, respectively, with the pan-neuronal unc-14 promoter (16, 24). Neuronally restrictedage-1 expression fully restored wild-type adult life-span to an age-1(mg44) null mutant (Fig. 1). This rescue is comparable to the positive control, ubiquitous expression ofage-1 from the dpy-30 promoter in the age-1mutant (17, 25). Neuronally restricteddaf-2 expression from the unc-14 promoter also rescued the long life-span of daf-2(e1370) mutants, although not as completely as the comparable age-1 rescued animals, but to the same extent as ubiquitous daf-2 expression from the dpy-30 promoter (Fig. 2). The longdaf-2(e1370) life-span is also rescued when daf-2is expressed from the unc-119 promoter, another neuron-specific promoter (18).

Figure 1

(A to E) Rescue ofage-1(−) long life-span by cell-type–restrictedage-1 activity. Life-span curves of populations of animals with cell-type–restricted age-1 activity assayed at 25.5°C. The blue line is wild type, the orange line is long-lived (m+z−) age-1(mg44) adults, and the magenta lines are independent lines of transgenic age-1(mg44) animals, as indicated. Results are cumulative from ≥two independent experiments with ≥50 animals per trial.

Figure 2

(A to D) Rescue ofdaf-2(−) long life-span by cell-type–restricteddaf-2 activity. Life-span curves of populations with cell-type–restricted daf-2 activity, as in Fig. 1.

Animals with age-1 expression restricted to a smaller set of neurons were also examined. The promoter for the mechanosensory neuron-specific beta-tubulin mec-7 was used to expressage-1 in about 10 neurons, including the six touch neurons (19). age-1 activity in these neurons showed little or no rescue of the long life-span phenotype (Fig. 1), indicating that this neural type or this small number of neurons does not contribute in a major way to longevity control.

In contrast to neuronal expression of daf-2 andage-1, restoration of daf-2 pathway activity to muscles from the promoter for muscle myosin, unc-54, was not sufficient to rescue the long life-span of daf-2 orage-1 mutants (Figs. 1 and 2 and Table 1) (20). Similarly, expression of daf-2 or age-1 in the intestine, the major site of fat storage, from the ges-1promoter does not rescue life-span as efficiently as neural expression of these genes (21). Intestinally restricteddaf-2 expression showed weak rescue of the long life-span ofdaf-2(e1370), whereas intestinal age-1 expression did not rescue the long life-span of age-1(mg44) mutants. The lack of longevity rescue was observed in multiple transgenic lines for both daf-2 and age-1. In addition, the muscle or intestinal age-1 and intestinal daf-2transgenes expressed sufficient gene activities to partially rescue dauer arrest phenotypes, showing that the fusion genes were functional.

The aging and metabolic outputs of daf-2 pathway signaling are separable. Restoring age-1 function ubiquitously to the nervous system or to muscle rescued the metabolic defects ofage-1 mutants (Table 1) (26). Paradoxically, given that the intestine is the major fat storage depot, expression of age-1 in the intestine only weakly rescued the metabolic defects. Ubiquitous and neuronal, but not intestinal or muscle, daf-2 expression reduced the level of fat accumulation in daf-2 mutants. The rescue of the metabolic phenotype was highly correlated with the rescue of dauer arrest by these transgenes (see below), suggesting that the metabolic rescue may be a consequence of dauer arrest rescue, or vice versa.

An important finding is that rescue of metabolic defects indaf-2 pathway mutants is not correlated with rescue of long life-span. Shifting metabolism away from fat accumulation, by restoringage-1 activity to muscle or intestine, is not sufficient to induce a short life-span. Because intestine and muscle are major sites of metabolic storage and activity, it is significant that they are not the major organs of longevity control. Rather, the lack ofdaf-2 pathway signaling in the nervous system of these chimeric animals may induce their long life-span.

The dauer arrest phenotype of daf-2 pathway mutants was rescued most effectively by restoring signaling to neurons (Table 1) (27). Expression ofage-1 in muscle, intestine, or themec-7–expressing neurons also rescued dauer arrest, but less efficiently than pan-neuronal expression. However, expression ofdaf-2 in muscle, unlike of age-1, did not rescue dauer arrest.

The conclusion that it is the expression of age-1 ordaf-2 within the nervous system that rescues aging depends on the unc-14 or unc-119 promoters driving expression only in neurons. One measure of specificity is that GFP fusions of these promoters show only expression in the expected tissues at all stages tested (23). However, weak expression below the detection limit of GFP in other cell types is possible. The phenotypes of age-1 and daf-2 allelic series show that the highest gene activities are needed for life-span regulation and less is needed for regulation of metabolism and dauer arrest (for example, maternally contributed age-1 activity can rescue both metabolism and dauer arrest, but not the longevity phenotype). Substantial age-1 and daf-2 gene activity is probably required to allow such potent longevity rescue in the nervous system, suggesting that weak promoter promiscuity is not a problem. Expression level differences between the neuronal, muscle, and intestinal promoters also do not appear to account for more potent life-span rescue by transgenes expressed from neuronal-specific promoters. We observed high levels of GFP expression from the muscle-specific unc-54 promoter, relative to the other promoters used (28). Consistent with this observation,unc-54 is more abundant in the 100,000 sequence C. elegans expressed sequence tag (EST) database, which contains 90unc-54 ESTs compared with 14 unc-14 ESTs and 2ges-1 ESTs. Thus, it is the activation of thedaf-2 pathway in the nervous system in particular, rather than high expression levels in any tissue, that rescues the longevity extension of daf-2 pathway mutations.

Genetic mosaic analyses of daf-2 support the interpretation that daf-2 signaling from the nervous system controls longevity. Wild-type life-span required daf-2 pathway activity in the AB blastomere descendents, which include nearly all of the nervous system as well as much of the ectoderm and half of the pharynx (15). Thus, although those studies could not mapdaf-2 pathway longevity control specifically to neurons, they are consistent with the results of the transgenic approach reported here. It may be important that the highest DAF-2 abundance revealed by antibodies to DAF-2 is in the nerve ring (29).

The more potent regulation of longevity by neuronal daf-2pathway signaling could represent distinct outputs from some or all neurons or, simply, that neuronal promoters restore daf-2pathway activity to more cells than muscle or intestinal promoters. The adult hermaphrodite nematode contains 302 neuronal cells, 95 body-wall muscle cells, and 20 intestinal cells. Although neurons constitute the largest number of cells, the total mass of neurons, which are smaller than nematode muscle or intestinal cells, is considerably less than the mass of muscle or intestinal cells. Further analysis of animals withdaf-2 pathway signaling restored to restricted neuronal subtypes should elucidate whether C. elegans life-span is controlled by a specific set of neurons or, alternatively, by a quorum of neurons that can be of any neuronal subtype. Although mammalian insulin signaling in the nervous system has not yet been examined for longevity control, there is evidence that insulin signaling in neurons and neuroendocrine cells controls feeding and metabolism (30,31).

Expression of daf-2 pathway genes in muscle, intestine, or the mec-7–expressing neurons can regulate dauer arrest and metabolism but not life-span. The daf-2pathway-mediated regulation of dauer arrest and metabolism can be decoupled from life-span regulation, and these represent distinct outputs of the daf-2 insulinlike signaling pathway.daf-2 pathway signaling in neurons may result in the production of a senescence-inducing neuroendocrine output that is not produced in muscle or intestine. Intestinal and muscle cells may contribute dauer and metabolic regulatory signals. The somatic gonad has been shown to affect life-span through the daf-2 pathway (32). The life-span signals from the somatic gonad may act to regulate neuronal daf-2 pathway activity. C. elegans life-span is also extended 1.5-fold when daf-2activity was lost from the EMS lineage, which contributes the intestine, some pharyngeal cells, the somatic gonad, and the sex muscles, suggesting that daf-2 signaling in one or several of these cell types is also necessary for normal aging (15). Our results also point to a minor role of intestinal daf-2pathway signaling in aging.

How does daf-2 signaling from neurons control life-span?C. elegans dauer larvae express high levels of the free radical–scavenging enzymes, catalase and SOD (9). The expression of catalase and Mn-SOD is transcriptionally regulated by DAF-16, the major target of daf-2 pathway signaling (11, 12, 33). Furthermore, mutations in ctl-1 cytosolic catalase reduce the life-span ofdaf-2 mutants, showing that ctl-1, and possibly other free radical–scavenging enzymes, are required for long life-span (11). Neurons may be particularly sensitive to free radical damage during aging. In fact, overexpression of Cu/Zn SOD in only motorneurons can extend Drosophila life-span by 48% (3).

We propose that neuronal DAF-2 activity maintains relatively low levels of free radical–scavenging enzymes, such as SOD-3 and CTL-1, by antagonizing the DAF-16 transcription factor. Loss of DAF-2 activity from neurons, relieving the negative regulation of DAF-16, induces higher expression levels of these free radical–scavenging enzymes, thereby protecting neurons from oxidative damage. By this model, neuronal daf-2 signaling might regulate an organism's life-span by controlling the integrity of specific neurons that secrete neuroendocrine signals, some of which may regulate the life-span of target tissues in the organism. Our results, together with those from Drosophila, suggest that oxidative damage to neurons may be a primary determinant of life-span.

  • * These authors contributed equally to this work.

  • To whom correspondence should be addressed: E-mail: ruvkun{at}

  • Present address: Department of Pathology, Harvard Medical School, Boston, MA 02114, USA.


View Abstract

Navigate This Article