The idea of delaying aging with a simple intervention is daunting considering the number of organs and processes in play. Yet studies using animal models have shown that life span is indeed malleable, that it can be manipulated by genetics or the environment and further that studies with model systems have converged on just three main pathways—insulin signaling, sirtuins and mTOR (mammalian target of rapamycin). In addition to there being an improbably small number of them, they also intersect physiologically around the issue of cell energetics3 (Fig. 3). These pathways are all engaged in networks that sense the environment and respond to stress. Periods of starvation, for example, lead to insulin signaling; sirtuins respond to metabolic and genotoxic stress through multiple pathways; and mTOR senses nutrients and amino acids, increasing resistance to environment stresses.

Figure 3: Molecular targets for caloric restriction and interventions against premature aging. (Reprinted with permission from ref. 3.) Full size image

The central role of insulin. When asked what the secret of long life is, Eline Slagboom of Leiden University Medical Center (the Netherlands) responds without a moment's hesitation, “insulin sensitivity.” Insulin signaling's role in longevity first came to light in 1993 in nematode studies, which not only put the spotlight on this pathway, but also demonstrated unexpectedly that a simple genetic change could alter life span. Cynthia Kenyon (at the University of California, San Francisco, at the time, and now Calico's senior scientific advisor) showed that mutating a single gene doubled nematode life span, a feat that has been reproduced repeatedly and even extended to as much as tenfold4. Kenyon's mutants, which not only had a longer life span but also a longer health span, needed a controller gene, daf-16, which did the actual work of extending life in the worms. Similar long-lived mutant worms had been discovered a few years earlier by the University of Colorado's Tom Johnson (age-1) and independently by Michael Klass, a post-doc at the time and also at the University of Colorado, Boulder. Sequence analysis first done by Gary Ruvkin's laboratory, now at Massachusetts General Hospital in Boston, showed that these genes were all part of the insulin signaling pathway—daf-2 and age-1 resembled insulin-like growth factor 1 (IGF-1) and phosphatidylinositol kinase, respectively.

A connection with human biology was made when sequence analysis of daf-16 revealed a startling resemblance to the gene encoding the highly conserved Forkhead transcription factor (FOXO). Human FOXO, like its worm counterpart, directs the synthesis of proteins in the insulin signaling pathway. A particular FOXO, FOXO3a, is found in long-lived humans, and variations in the gene lengthen life in Okinawan and Ashkenazi Jewish centenarians. Two copies of FOXO give you a twofold chance to live long; with three copies of the gene, you have a threefold chance to live past 100.

In study after study, low IGF levels are associated with longevity, according to mouse geneticist Andrzej Bartke of the Southern Illinois School of Medicine (Springfield), who won the first Methuselah Prize—a largely ceremonial prize awarded by Aubrey de Grey and David Gobel's foundation of the same name—for creating a mouse with twice the normal mouse life span. Dwarf mice, which live up to 70% longer than control mice, have low blood glucose and low insulin. (The latter finding is contrary to expectations; low insulin levels would be expected to raise circulating glucose levels.) In fact, Bartke found that testing insulin sensitivity in dwarf mice, which is done by injecting insulin and measuring blood glucose, proved tricky. “We have to use low levels of insulin, otherwise the blood sugar goes down so low that they literally faint, and we have to give them a quick shot of glucose to bring them back,” he says. “It is very dramatic.”

A correlation between low IGF-1 and longevity exists for humans as well. Barzilai, who has been studying a cohort of long-lived Ashkenazi Jews, finds that low IGF levels predicts longevity in people over 90 years of age. In addition, a population of Ecuadorian dwarfs, the Laron dwarf, which has low IGF-1 levels (secondary to a mutation in the growth hormone receptor) are protected from age-related maladies. Jaime Guevara-Aguirre of the Institute of Endocrinology, Metabolism, and Reproduction (Quito, Ecuador) had been studying a population of 99 Laron dwarfs since 1998, and in 2011, he, along with Valter Longo of the University of Southern California (Los Angeles), reported that Larons had a lower incidence of diabetes and cancer compared to controls. (However, they did not live longer, dying mainly of causes unrelated to age, like alcoholism and convulsive disorders.)

Partridge's group is taking advantage of the genetic systems in flies and mice to dissect the molecular underpinnings of insulin signaling. Looking at how reduced insulin signaling rescues declining neural circuitry in flies, which recapitulate the neurodegeneration seen in older humans, she finds that endocytic recycling plays a role, which may be a specific response to insulin signaling or a more general response to nutrient shortages. Genetic manipulations have also allowed her to pick apart the pathways involved with the multiple phenotypic changes that accompany reduced insulin signaling in flies—small body size, reduced or delayed reproduction, growth inhibition, increased stress resistance and, of course, increased life span.

Uncertainty around sirtuins. Not long after the worm experiments were published, Leonard Guarente's group at the Massachusetts Institute of Technology (MIT) unearthed a family of genes that extended life span in yeast, which they termed sirtuins (for silent information regulator), so-called because of their role in gene silencing. Increasing the copy number of one in particular, sir-2, increased the life span of Saccharomyces cerevisiae by 30%. This highly conserved family of genes exists throughout the animal kingdom; mammals have at least seven sirtuins whose protein products function variously as protein-modifying enzymes. Sir-2 encodes NAD+-dependent lysine deacetylase.

The sirtuin story took a slight detour into the commercial sector, following the Geron debacle (Box 1); venture capitalists thought they had (once again) in their grasp the secret to longevity and rushed in to extend these findings to humans (Box 3). And if extending life was not exciting enough, David Sinclair, a former post-doc of Guarente's, announced in 2003 that a compound found in red wine, resveratrol, induces sirtuins, answering the age-old question of why the French can eat lots of rich food and alcohol and live long and happy lives. This was followed by several years of bickering among researchers, many of whom had been part of Guarente's original team at MIT5.

Box 3: Gene follies The discovery of longevity genes in invertebrates in the 1990s unleashed a small tsunami of startup activity. According to Peter DiStefano, CEO and still president of the now-defunct Elixir Pharmaceuticals, it was a heady time for aging research. “Aging was at that time what the microbiome is today. It was on NPR [National Public Radio], in the New York Times, you couldn't go a Sunday without there being an aging article,” he says. Kenyon's paper had galvanized the field4, providing unexpected proof that life span is subject to modification by seemingly simple genetic changes. “The '93 Nature paper was the key—it pointed [out] how it could be done,” says Cindy Bayley, a venture capitalist at Arch Ventures who put together the deal that led to the first of a small cadre of genomics-based aging companies that emerged in the late 1990s. Around 1999, Bayley, who had been following the field since Kenyon's work, met Guarente, and brought the two scientists together, coming up with an initial funding round of $8.5 million from Bayley's own firm, along with several others, to start a company, Elixir Pharmaceuticals. And with intellectual property in house for the daf-2 and sir-2 genes, and for a third gene called indy (for 'I'm not dead yet'), discovered in Drosophila by Stephen Helfand at the University of Connecticut, they set about to find ways to manipulate the pathways that these genes control, all of which seem to extend the life span of model organisms. Elixir was established with the idea, at least to Guarente, of screening for small molecules that could activate sir-2. Even so, Guarente's vision was not shared by those running the company, who felt that the biology around activating an enzyme is too complicated. Another problem with the original Elixir game plan was that aging is not an indication recognized by the FDA, or one that can be approached in human clinical trials. So, when the time came to raise additional funds from investors, the management felt that they needed a tractable disease. DiStefano made the argument that aging and metabolic diseases, particularly diabetes, are a lot alike. “If you look at diabetes, it looks like accelerated aging—if you look at the targets, look at the physiology, the pathophysiology, everything. I don't think we got any pushback from anybody there.” Accordingly, the company looked for molecules to in-license, which as far as Guarente was concerned was the first step in the wrong direction. Guarente ultimately left Elixir to join another aging company that had been founded several years after Elixir, by his former post-doc David Sinclair and a local serial entrepreneur Christoph Westphal. Sinclair shared Guarente's vision of screening for activators of sirtuins, and Sirtris jumped in to fill the gap left by Elixir when it pivoted away from sirtuins, according to Guarente. By all measures, Elixir was a successful company. By 2009, they had pulled in close to $100 million in private funds, they had molecules in various stages of clinical development with partners. Yet, just when they needed access to large amounts of capital to move their clinical programs to the next stage, the economy collapsed. The company exists only on paper now, although several of their programs are being taken forward by partners. “I like to say we were a successful failure,” says DiStefano. Without money to maintain their intellectual property, much of it has been lost. Meanwhile, Sinclair and company moved ahead with their program. Founded in 2004 with a bevy of Cambridge, Massachusetts, luminaries on its scientific advisory board, the company made headlines with its discovery in 2003 that sir-1 is activated by resveratrol, a component of red wine. This led to many years of arguing among researchers over the connection between resveratrol and sirtuins, with some claiming the result was an artifact, others reporting difficulty in reproducing the company's findings. Notwithstanding all the sturm und drang, the company floated a successful $69-million IPO (initial public offering) in 2007 (pre-economic melt-down) on the basis of preclinical data and was bought by GlaxoSmithKline (GSK) a year later for close to three-quarters of a billion dollars. The purchase by GSK hardly put an end to the soap opera that surrounded sirtuins and resveratrol. The high-profile deal and the amount of money involved for a company without a product was cause for something between amazement and consternation in the blogosphere, especially after GSK closed down the Cambridge site (Sirtris had been allowed to operate independently in Cambridge for several years after GSK acquired it) and brought the program in-house. Today, the head of GSK's 12-employee sirtuin development performance unit, Jim Ellis, says that they have positive clinical results with one of the Sirtris molecules in psoriasis, but that they are looking for molecules with better properties to take forward in psoriasis and other inflammatory disorders. In other words, they are starting over.

And while all that was going on, work on sirtuins in both the public and private sector continued to uncover intriguing connections to nutrient sensing. This family of proteins deacetylates numerous substrates (histones, transcription factors, mitochondrial proteins, metabolic proteins and dozens of others); they also extend life span not just in yeast, but also in worms, flies and mice. According to Guarente, one critical activity linked to caloric restriction is the activation of mitochondrial proteins by sirtuins, which shifts the balance of metabolism and energy production toward the more efficient oxidative phosphorylation pathway.

Recent work by Johan Auwerx of École Polytechnique in Lausanne and Guarente showed that the levels of NAD+, a co-factor for some sirtuins, decline with age in worms, and restoring levels of NAD+ pharmacologically, increases life span via FOXO signaling. In similar experiments with mice, Sinclair's group at Harvard has recently shown that two-year-old mice, injected with a precursor to NAD+ for just a week, are all but indistinguishable from six-month-old mice in various muscular feats. Sinclair, who participated in the early round of startups back in the late 1990s, founded a new company in 2008, Metro Biotech, to take this finding into human testing.

Long before IGF-1 and sirtuins came on the scene, caloric restriction was known to extend life span. The earliest life extension protocol for any model species was caloric restriction (CR) in rodents described in the 1930s by Clive McCay, a nutritionist at Cornell University (Ithaca, NY, USA). (The first description of caloric restriction actually goes back much farther. In the fifteenth century, a 35-year-old Venetian nobleman named Luigi Cornaro, who was in poor health, restricted his diet to 12 ounces of solid food per day and is said to have lived to over 100). McCay showed that reducing caloric intake by 40% can extend life and delay the onset of age-related pathologies6. The basic observation has been replicated repeatedly and remains the most robust and reproducible of life-extending protocols. Caloric restriction works in many species; however, the data in nonhuman primates and humans are not so clear cut.

Although the molecular mechanism behind caloric restriction still remains somewhat mysterious—recent evidence suggests restricting single amino acids may work just as well—connections between caloric restriction and both sirtuins and insulin signaling have been demonstrated in multiple model systems, including humans. Caloric restriction induces sirtuin expression in mice and humans, and knockout experiments ablates some of the beneficial effects of caloric restriction, including increased life span in mice. Mice and humans who adopt a restricted diet have lower levels of IGF-1 and blood glucose.

At the center with mTOR. mTOR is a threonine serine kinase that functions as an amino acid and nutrient sensor, thereby coordinating stress and survival responses. It does so by integrating information from multiple upstream pathways, among them the insulin signaling pathway. When nutrients are plentiful, mTOR stimulates growth and inhibits autophagy, among other things. Inhibiting mTOR with rapamycin extends life span in multiple species, from yeast to mouse, presumably by shifting metabolism away from growth toward maintenance.

Mouse geneticist Richard Miller, at the University of Michigan, led one of three teams that described the positive effects of rapamycin on life span in mice, the first study done in a mammal, as part of the National Institute on Aging (NIA)'s Interventions Testing Program (ITP). Since 2004, the ITP has been coordinating the testing of compounds for effects on life span in mice, which is carried out independently by three laboratories in the United States—the University of Michigan, the Jackson Labs and University of Texas at San Antonio. The effect on life span, which was more pronounced in female than male mice, was observed even when rapamycin was administered to middle-aged mice (600 days).