Introduction

Eat a big meal, and your body will erupt into an orgy of hormonal signals: liver talks to fat, which talks back to liver, while muscles, stomach, intestines, heart, brain, and pancreas all send their own urgent communiques. These messages unify the organs in a single, critical mission: separate the food's components as quickly as possible, and sort them for either storage or elimination.

If a single plate of food can do all that, what are we to make of long-term processes such as aging and weight gain, or chronic diseases such as atherosclerosis and osteoporosis? Physiologists have traditionally approached these questions by studying individual components, not because they failed to appreciate the system's interconnectedness, but because their tools weren't sophisticated enough to take a broader view.

That's finally changing. "There are so many advances, both in molecular biology and genetics ... that have shown that there are many functions for many organs that were not supposed to be functions of these organs, and also many connections between organs that are not the ones that I learned ... when I was in medical school," says Gerard Karsenty, chairman of genetics and development at Columbia University and co-organizer of the New York Academy of Sciences' first meeting on the evolving field of integrative physiology.

At the meeting, held May 14–16, 2008, at the NYAS conference center, researchers from several branches of medicine and biology discussed the data behind some of those surprising new connections, revealing and abetting a convergence between fields that previously kept to themselves. The gathering featured more than a dozen main presentations, two "data blitz" sessions of brief but intense talks, and a substantial poster area. A small sample of presentations, organized around two of the meeting's most common themes, provides a glimpse of the vast territory the event covered.

The problem of plenty

Nutrition is arguably the most ancient problem in biology, so it's little surprise that energy homeostasis—how the body processes food—has been a major focus of physiological research. These days, much of that discussion centers around an irony of modern life. After four billion years under the constant threat of starvation, evolution has produced a species that has overcome the problem all too well. Humans' success at feeding ourselves has been so comprehensive, at least in the developed world, that our ancient calorie-hoarding adaptations now work against us, driving a global epidemic of obesity.

Obesity research has revealed new connections between organ systems as well as critical regulatory genes.

Unfortunately, evolution has had no time to respond, and our bodies are handling this new development poorly. Long-term obesity brings with it a suite of complications that shorten life and call for expensive interventions. "The cost to U.S. society alone exceeds $170 billion a year, and of course it's growing with this epidemic," says Yale University's Gerald Shulman, who delivered the meeting's final keynote presentation.

Eating less and exercising more can prevent or reverse obesity, but that advice is easier given than followed. Consequently, researchers are trying to develop other treatments, and to understand the full range of molecular signals controlling the condition. This work has revealed new connections between organ systems, as well as critical regulatory genes that link lifespan to energy homeostasis. "The reason that these genes might be there is that this adaptive process has selected for genes ... that can recognize food scarcity, and translate it into these effects on basic life choices, reproduction versus hunkering down and slowing the aging process," says Massachusetts Institute of Technology biologist Leonard Guarente, the meeting's other keynote speaker.

Apoptosis and taxes

Besides being biologically interesting, the link between food intake and aging also connects the obesity epidemic to the other major health transition now sweeping the globe. As starvation and many infectious diseases yield to technology, more people are living long enough to encounter medical problems that used to be rare. Collectively known as the "diseases of aging," conditions such as heart disease, cancer, dementia, and osteoporosis are reaching pandemic prevalence.

The classical view of these diseases was that they were "post-evolutionary" events. Because they generally affect people long after their prime reproductive years, the theory went, there was no selective pressure for the body to develop defenses against them. Once we've lived long enough to reproduce, biology no longer cares what happens to us, so we simply fall apart.

Aging may be a programmed process.

More recently, researchers have come to view aging as a programmed process, in which highly evolved molecular signals make deliberate—if unfortunate—choices about life and death. Atherosclerosis, or hardening of the arteries, provides a good example of this phenomenon. "Everybody sitting here has atherosclerotic lesions, and most of those will never do you any harm from now until the day you die, but some of them will," says Columbia University's Ira Tabas. The distinction between benign plaques and deadly ones is chiefly a matter of apoptosis. Tabas and his colleagues have found that in some lesions, a combination of cellular stress and a "second hit" from another signaling pathway sends macrophages into a suicide pact. The dead cells then accumulate in the plaque, creating a necrotic core that worsens the lesion. Drugs that alter or block these signals might prevent or even reverse heart disease.

Though integrative physiology is still in its infancy, researchers are optimistic about its potential. For inspiration, Kitsis cites James Clerk Maxwell's famous equations, which unified electricity and magnetism a century ago. "I would like to think that sometime when I'm still around, we will have some connections between [physiological pathways] that explain mechanistically how these things interface with each other," says Kitsis. It's certainly food for thought.

Alan Dove Alan Dove is a science writer and reporter for Nature Medicine, Nature Biotechnology, and Bioscience Technology.

Diseases of Aging

Highlights

• Sirtuin proteins connect calorie intake, lifespan, metabolism, and many of the diseases of aging at the molecular level.
• Inappropriate apoptosis drives the formation of deadly atherosclerotic plaques.
• Estrogen may signal osteoclasts to eliminate themselves, preventing osteoporosis.

Something to look forward to

From cholera to measles to Yersinia, the gradual improvement of both infrastructure and medicine—at least in the developed world—have virtually eliminated most of the classical plagues. Instead, modern humans increasingly live long enough to have problems their ancestors seldom would have encountered: cancer, heart disease, osteoporosis, and dementia.

For most of the 20th century, scientists thought of these "diseases of aging" as a product of simple decay. Over time, the thinking went, the body simply accumulates so many insults that something is bound to fail. Things fall apart; the center cannot hold.

A handful of molecular signals could underlie multiple diseases.

That view seems to fit well with evolutionary theory, at least. Diseases of aging occur long after the prime reproductive years, so there would be little selective pressure to evolve defenses against them. More recently, though, data from systems as diverse as budding yeast and mice have revealed a more complicated picture.

When their food intake is restricted, organisms tend to live longer, indicating that they have some biological control over their own aging. Evolutionarily, it makes sense to reproduce fast and die young, making room for the offspring, as long as those offspring are likely to survive. When times are lean, though, living longer to wait for better conditions is probably a better strategy.

In the emerging field of integrative physiology, researchers are trying to understand how this basic choice works, and especially how it influences the diseases of aging. The work has revealed surprising connections between organ systems, and uncovered the tantalizing possibility that a handful of molecular signals could underlie multiple diseases.

Help for the hard-hearted

One of those signals is the histone deacetylase SIRT1, the human version of a gene originally discovered in yeast by Leonard Guarente and his colleagues. By linking overall metabolic activity to gene regulation, SIRT1 indirectly induces or inhibits a vast range of physiological activities, and it appears to be the key controller of the lifespan extension seen in calorie restriction experiments.

Since discovering SIRT1, Guarente's team has focused on tracing its many activities downstream to different diseases of aging. Looking at a mouse that lacks the SIRT1 gene, for example, the researchers found unusually low levels of HDL, or "good cholesterol" in the bloodstream, but high cholesterol in tissues. That suggested a role for SIRT1 in cardiovascular disease, a condition driven partly by the inexorable increase in cholesterol with age. The mouse phenotype also provided a testable hypothesis. "Maybe the process of reverse cholesterol transport ... from cells out of the body through HDL is defective in the SIRT1 knockout," says Guarente, who delivered one of the meeting's keynote presentations.

The investigators soon found that the primary transporter in this process, a protein called ABCA1, is sufficiently reduced in the mutant mice to account for the cholesterol accumulation. Additional experiments revealed that SIRT1 normally deacetylates a regulatory protein called LXR, which then regulates ABCA1 levels to control cholesterol transport. Drugs targeting this pathway might eventually be useful for treating or preventing cardiovascular disease.

While cholesterol is a major cause of cardiovascular disease, most patients are more concerned about the condition's effects, the most infamous of which is a myocardial infarction, or heart attack. When this occurs, the loss of blood flow, or ischemia, around a section of cardiac tissue causes the cells to die; if enough cells die, the patient goes with them. Besides killing ischemic tissue, though, a myocardial infarction also prompts adjacent cells with normal blood flow to initiate apoptosis, or programmed cell death, creating a zone of destruction much larger than the ischemia alone.

For Richard Kitsis, a leading apoptosis researcher who presented his results at the meeting, this collateral damage—or any other tissue pathology, for that matter—is no accident. "I don't believe there is accidental cell death," he says, adding that "If an anvil fell out of the ceiling and hit my head and killed me, I would say the cells in the top of my head that died by getting hit with the anvil will have actually participated in their own death by initiating a program.". Even if it is deliberate, though, excessive cell death in cardiac tissue is clearly a bad biological choice.

To figure out what causes cells to pick this unfortunate option, Kitsis and his colleagues have focused on an apoptosis regulator called ARC, which is induced in breast cancer and inhibited in mouse models of myocardial infarction. Mice lacking the gene for ARC are substantially less sensitive to breast tumors, but suffer much greater cardiac damage if they're given heart attacks. Mildly overexpressing ARC in cardiac tissue before inducing a myocardial infarction prevents much of the tissue damage. "There's a causative effect between the ARC going away and the [cell] death occurring. If you could keep the ARC around, maybe you could make infarcts smaller just using a natural product that the body makes," says Kitsis.

Because of its yin/yang role in infarctions and cancer, though, simply giving a patient a constant dose of an ARC-stimulating drug would probably be a bad idea—saving heart tissue might come at the expense of encouraging tumor cells. Instead, Kitsis envisions developing compounds that a patient with cardiovascular disease would take only at the first signs of a heart attack, limiting the size of the resulting infarct even before reaching the emergency room.

The second shooter

Not all atherosclerosis ends with a call to 9-1-1, though, as the disconcerting introduction of Ira Tabas's presentation made clear. "Everybody sitting here has atherosclerotic lesions, and most of those will never do you any harm from now until the day you die, but some of them will," says Tabas. He adds that "if you take a look to the left of you and the right of you, one of these people will die of this disease."

For atherosclerosis, the difference between harmless and deadly lies in the core of the plaque. Benign plaques have unremarkable centers, but the cores of bad plaques are full of necrotic tissue and dead macrophages. The macrophage graveyard appears to be a product of excessive apoptosis, and also inadequate phagocytosis, the mechanism that normally clears away dead cells. But what causes these two processes to go haywire in some plaques and not others?

Using cultured cells and mouse models, Tabas and his colleagues have found that stress on the endoplasmic reticulum (ER), the cellular organelle responsible for secreting proteins, is a major contributor to the problem. When the ER gets overloaded, it releases calcium into the cytosol, triggering multiple intracellular signaling pathways that try to alleviate the stress. Intracellular calcium release is a common occurrence in the cell, though, and cells don't spontaneously commit suicide. The key, says Tabas, is signal strength. "When it is participating in a death pathway, it is incredibly amplified, prolonged, et cetera, and that's what differentiates the death pathway from normal physiology," he says.

But even a roaring ER stress signal isn't enough to kill the cell by itself. In addition, the macrophages must suffer a "second hit," which tips the balance from adaptation and survival toward resignation and death. One of the most important second hits is a signal from Toll-like receptors, which have evolved as part of the innate immune response to recognize specific, ancient infectious diseases. In cultured cells, the researchers found that this signal is highly synergistic with ER stress—either signal alone causes modest cell death, but the two together stimulate a massive die-off.

Only some atherosclerotic plaques promote apoptosis.

Several studies have now confirmed that ER stress occurs routinely in bad atherosclerotic plaques, and these lesions also host a long list of plausible second hit signals. Genetically engineered mice with alterations in either signaling pathway are much more resistant to heart disease than wild-type controls.

The results certainly help explain how bad atherosclerotic plaques form, and even suggest a few good targets for drugs to reverse this pathology, but they also raise a question: why would macrophages, which form the backbone of the immune system, be so prone to killing themselves?

In a word, tuberculosis. This microbe, and a few other intracellular pathogens, can evade the normal immune response by hiding inside macrophages. If the infected macrophages undergo apoptosis, though, they will be endocytosed by roving phagocytic cells, which will then digest the infectious agent and present its antigens on their surfaces, blowing its cover. "One of the major characteristics of virulent TB is that it keeps macrophages alive," says Tabas. Avirulent strains of the mycobacterium are susceptible to macrophage apoptosis. Apparently, an adaptation that helped our ancestors survive infection now primes us for heart disease.

The leg bone's connected to the ... estrogen?

Besides heart disease, everyone who lives long enough can anticipate worrying about osteoporosis. The master regulator SIRT1 crops up in this condition, too, regulating osteoblastogenesis and osteoclastogenesis. Disrupting the balance between these two processes causes a range of bone diseases.

Osteoporosis, the most common skeletal disease in developed countries, generally involves either a surplus of osteoclast cells, which break down old bone tissue, or a deficit of osteoblasts, which build up the bone. Both problems occur in mice that lack the gene for SIRT1. Knocking out the gene only in the osteoclast cell lineage also causes osteoporosis, and produces mostrous, multinucleated osteoclasts. "This to me strongly suggests that SIRT1 is functioning as a restraint, a negative regulator of osteoclast differentiation, such that when you take it away, differentiation goes up markedly," says Guarente, whose lab did this work.

Estrogen and androgens also help regulate bone buildup and loss, which is why osteoporosis hits postmenopausal women particularly hard. As women's estrogen levels decline, their osteoclasts become more abundant, breaking down more bone than the osteoblasts generate. But what connects estrogen to the osteoclasts at the molecular level?

Shigeaki Kato of the University of Tokyo is studying that question, and at the meeting he presented results from an impressive series of mouse genetics studies on it. After a conventional deletion of the estrogen receptor revealed tantalizing but incomplete results, Kato and his colleagues realized that they needed another mouse strain. "We need a knockout in an androgen receptor and estrogen receptor to compare the sex hormone action," says Kato.

Deleting the androgen receptor is tricky, though. Males without this protein have testicular feminization: female anatomy, but with testes in place of ovaries. That makes the mice sterile, and therefore impossible to maintain as a strain. Instead, the researchers generated conditional knockouts using the popular Cre/Lox system, which allowed them to breed a line of parent animals normally, inducing the gene deletion in a new litter at will.

In males, androgens may be responsible for maintaining bone mass.

Inducing the androgen receptor deletion causes testicular feminization, as expected, and also makes the males become obese later in life. The males also develop osteoporosis. Females without the receptor, however, do not develop either problem. "In the normal condition in a male ... maybe androgen is important to maintain and attenuate bone mass loss, but in a female maybe estrogens are important," says Kato.

The researchers then developed several additional mouse lines, including ones that lacked either androgen or estrogen receptors only in osteoclasts. Taken together, the results help clarify the connections between hormones and bone density. "We think that female estrogen is a kind of a signal to ask mature osteoclasts to retire from their job ... suggesting that maybe estrogens induce the apoptosis and the shortening [of] the lifespan of mature osteoclasts," says Kato.

Viewed as a set, these presentations and others at the meeting reinforced the new idea that aging and its associated diseases are part of a regulated process. We may not know how to turn back the clock yet, but at least we're starting to understand some of its gears.