Cellular Warriors at the Battle of the Bulge

See allHide authors and affiliations

Science  07 Feb 2003:
Vol. 299, Issue 5608, pp. 846-849
DOI: 10.1126/science.299.5608.846

Researchers are picking apart the molecular signals the body uses to regulate its weight—work that may lead to new antiobesity drugs

When Mae West said that too much of a good thing can be wonderful, she wasn't talking about food. Over the past 50 years or so, food availability has soared, at least in the developed world, and too much food has turned out to be far from a good thing. That order of fries you buy in your local fast-food restaurant isn't the only thing that's become supersized in the United States: So has the national waistline.

By current estimates, 30% of U.S. adults are obese—roughly double the percentage 20 years ago—and another 35% are overweight. Children and adolescents haven't been immune to this obesity epidemic; 15% are too fat. All this excess poundage is much more than an aesthetic issue; obesity is a major risk factor for such life-threatening diseases as type II diabetes (Science, 26 April 2002, p. 686), heart attack, stroke, and some types of cancer, including breast and colon cancers. Indeed, some 300,000 people die of obesity-related diseases every year in the United States alone.

But there's a glimmer of hope. Researchers have learned a great deal about how the body regulates its weight. “We know that there are physiological systems in place that seem to be involved in maintaining weight,” says obesity researcher Jeffrey Flier of Beth Israel Deaconess Medical Center in Boston.

One of these systems is primarily concerned with short-term weight regulation—how often and how much we eat on a given day—and the other with longer-term regulation. Over the past few years, scientists have identified numerous components of each. Recently, for example, two peptide hormones produced by the digestive tract, known as ghrelin and PYY, have been linked to short-term feeding behaviors, whereas leptin, and to a lesser extent, insulin, are key to weight maintenance over months and years.

Obesity researchers have also made progress toward understanding how these hormones exert their effects. Among other things, they've found that certain brain regions, such as the arcuate nucleus, play a critical role in integrating the hormones' activities, sending signals that tell the body to adjust its food intake and energy expenditure. “A coherent wiring diagram can now be drawn” showing how these hormones work, says leptin discoverer Jeffrey Friedman of Rockefeller University in New York City.

The pharmaceutical industry has been having great difficulty coming up with antiobesity drugs that are both safe and effective (see p. 849), but the wealth of information now being gained should provide several new drug targets.

In one regard, though, the message coming out of this work is depressing for people who want to lose weight: The body's weight-control systems have apparently been designed to protect more against weight loss than weight gain (see Friedman Viewpoint on p. 856). That undoubtedly reflects human evolutionary history in which, until very recently, food scarcity, not overabundance, was the danger. As geneticist Rudolph Leibel of Columbia University College of Physicians and Surgeons in New York City says, “you can bemoan the fact that we're set up this way, but it's what's gotten us here.”

Appetite controllers.

The body produces hormones that act through the brain to regulate short- and long-term appetite and also the body's metabolism. The diagram shows the sources of several of the hormones now under intensive investigation.


Long-term savings

The discovery that helped kick off the current surge of obesity research was the Friedman team's identification of leptin in 1994. Interest in the area “exploded” as a result, says geneticist Stephen O'Rahilly of the University of Cambridge, U.K.: “It was the first [antiobesity] hormone you could get your hands on.”

Friedman and his colleagues found leptin by tracing the gene at fault in a mutant strain of extremely obese mice and went on to show that they could cure the animals' obesity by treating them with the hormone. It produced weight loss by decreasing the animals' appetite while at the same time revving up their metabolic rates. As humans have their own version of the leptin gene, the results grabbed everyone's attention. Would leptin prove to be the “magic bullet” that would cure the ever-growing human obesity problem?

Those hopes were soon dashed. Some rare cases of human obesity are caused by defects in leptin production. O'Rahilly, Sadaf Farooqi, also at Cambridge, and their colleagues have recently used leptin to treat three children who were extremely obese because they don't make the hormone. The children's weights quickly dropped, mainly because they ate much less than before, the researchers reported in the October 2002 Journal of Clinical Investigation. “Although leptin deficiencies are rare, they are treatable,” O'Rahilly says.

But unlike such patients and the mutant mice, most obese humans turned out to have higher than normal blood levels of leptin, which is produced by fat cells. For reasons not yet understood, they are resistant to its actions. That's one reason why many obesity researchers now think that leptin's main role is protecting against weight loss in times of deprivation rather than against weight gain in times of plenty.

When a person's fat stores shrink, so does leptin production. In response, appetite increases while metabolism decreases. But the converse does not happen. Beyond a certain point, increased leptin production does little to inhibit appetite or increase metabolism. “The system is designed to defend itself against starving to death and not being able to reproduce,” Leibel says.

Although leptin is not very effective for treating garden varieties of human obesity, its discovery did open the door to a better understanding of the body's weight-control mechanisms. Shortly after the Friedman group discovered the hormone, a team led by Louis Tartaglia of Millennium Pharmaceuticals in Cambridge, Massachusetts, found the gene for the receptor through which leptin exerts its effects. From there, researchers were able to show that the neurons of the arcuate nucleus, which was already known to be involved in appetite regulation, carry relatively large amounts of the receptor and might thus be prime targets for leptin in the brain.

Since then, numerous labs have traced the neuronal pathways through which leptin works in the brain and have shown that other hormones involved in weight control often work through the same pathways. Particularly important is the arcuate nucleus, which Friedman describes as the “master center: the seat of both the short-term and long-term [weight-regulatory] systems.”

The arcuate nucleus, which is located in the hypothalamus, contains two major types of neurons with opposing actions. Activation of one type, which produces peptide neurotransmitters called neuropeptide Y (NPY) and agouti-related peptide (AgRP), stimulates appetite while reducing metabolism. In contrast, activation of the other type, known as POMC/CART neurons, causes the release of α-melanocyte-stimulating hormone α-MSH), which inhibits eating.

Central command centers.

The arcuate nucleus (ARC) of the brain contains two sets of neurons with opposing effects. Activation of the AgRP/NPY neurons increases appetite and metabolism, whereas activation of the POMC/CART neurons has the opposite effect. These neurons connect with second-order neurons in other brain centers, and from there the signals are transmitted through the nucleus tratus solitarius (NTS) to the body. Many appetite-regulating hormones work through the ARC, although they may have direct effects on the NTS and other brain centers as well.


When fat stores and leptin levels are declining, the NPY/AgRP neurons are activated and the POMC neurons are inhibited, leading to weight gain. Conversely, at least in nonresistant animals, increasing fat stores and leptin levels lead to inhibition of the NPY/AgRP neurons and activation of the POMC neurons, resulting in weight loss. The NPY/AgRP and POMC/CART neurons then send their signals through certain other brain centers to the nucleus tractus solitarius of the brain stem, and from there to the rest of the body.

One indication of the importance of these circuits for weight control comes from O'Rahilly and Farooqi and also from Philippe Froguel's team at the Institute of Biology in Lille, France. Although very few cases of human obesity have been linked to mutations in either the leptin or leptin receptor genes, these researchers showed a few years ago that mutations in the receptor through which α-MSH exerts its appetite-inhibiting effects are much more common, accounting for perhaps 5% of severe obesity cases. Researchers are now looking for compounds that can beef up activation of the receptor in obese people who don't carry such mutations.

Additional targets for potential antiobesity drugs come from work in which researchers have been pinning down the mechanisms by which leptin turns up the body's metabolism. It apparently does this at least partly by altering the pathways through which fat is metabolized. For example, in work reported in Nature in January 2002, Barbara Kahn's team at Beth Israel found that leptin activates a so-called kinase enzyme in muscle that inhibits acetyl coenzyme A carboxylase, an enzyme that catalyzes a key step in fat synthesis.

As a result, the building blocks that would otherwise go into fat formation are shifted into a pathway that oxidizes them, providing energy for muscle cells. “We didn't say [in the paper] that this causes leanness,” Kahn says, “but it probably does. If an animal oxidizes its fatty acids instead of storing them, it is going to be leaner.”

Results from Friedman and his colleagues suggest that something similar may happen in the liver, although there a different enzyme, stearoyl-CoA desaturase-1 (SCD-1), is involved. The Rockefeller team has evidence that leptin exerts its antiobesity effects by turning down the activity of the SCD-1 gene. They found that they could protect leptin-deficient mice from obesity by inactivating the SCD-1 gene. The animals also had much higher metabolic rates than ordinary leptin- deficient mice, and their livers stored less fat. In some way the researchers don't yet understand, Friedman says, turning down SCD-1 activity fosters fat metabolism.

Leptin's effects.

Because of a gene defect, the boy doesn't make leptin, but treatment with the hormone, begun when he was 3.5 years old (top), brought his weight down to normal levels, as shown at age 8.


Insulin revival

Although leptin has received the lion's share of attention as an appetite and metabolism regulator, there are other players, and some of them also work through the arcuate nucleus. For example, some 25 years ago, Daniel Porte of the University of California, San Diego, and Stephen Woods of the University of Cincinnati suggested that the hormone insulin acts through the brain to regulate weight. Interest in the idea waned somewhat after the discovery of leptin, but recent work is reviving it. “Insulin is definitely having a comeback,” Flier says.

Some of this evidence comes from Ronald Kahn (no relation to Barbara Kahn) and colleagues at the Joslin Diabetes Center in Boston. They stymied insulin action in the brains of mice by knocking out the insulin receptors located there and found that the animals overate and became fat (Science, 22 September 2000, p. 2122).

Insulin receptors occur throughout the brain, but other work has tied the hormone's appetite-suppressing action directly to the arcuate nucleus. Insulin infused into the brain near the arcuate nucleus inhibits production of the appetite-stimulating NPY, researchers such as Michael Schwartz of the University of Washington, Seattle, have found. And when Luciano Rossetti's team at Albert Einstein College of Medicine in New York City inhibited production of the insulin receptor specifically in the arcuate nucleus of mice, the animals immediately increased their food intake, the team reported in the June 2002 issue of Nature Neuroscience. As Schwartz puts it, “as long as the brain has normal insulin sensitivity, you eat less and lose weight.” He and others note, however, that insulin's effects in this regard aren't as strong as leptin's.

Related studies may lead to antiobesity drugs that could circumvent obese people's resistance to the hormones' effects. For instance, in experiments described in the April 2002 issue of Developmental Cell, Barbara Kahn, Benjamin Neel, also at Beth Israel, and their colleagues knocked out an enzyme called protein tyrosine phosphatase 1B (PTP1B) in mice. The animals gained much less weight when fed a high-calorie diet than did normal controls. This apparently happens because the enzyme inhibits leptin and insulin signaling in the hypothalamus and other brain areas. Thus, it may be possible to bolster the hormones' effects with a PTP1B inhibitor. Barbara Kahn says the enzyme is a “terrific drug target. Other than being lean,” she says, “the mice are pretty normal.”

Long-term control.

Leptin levels help the body regulate weight.


Short-term appetite control

In addition to getting a handle on how the body regulates appetite and metabolism over the long haul, obesity researchers are gaining a better understanding of how it controls appetite on a daily basis. Several years ago, they identified cholecystokinin, a peptide released into the bloodstream by the intestine, as a “satiety hormone”—one that tells us when we've had enough to eat. Two recently identified appetite-regulating hormones are now attracting attention, both scientifically and from the drug-development point of view. These are ghrelin, an appetite stimulant, and PYY, a suppressant.

Kenji Kanagawa, Masayasu Kojima, and colleagues at the National Cardiovascular Center Research Institute in Osaka, Japan, discovered ghrelin, a peptide produced by the stomach, about 3 years ago. They found that it causes the release of growth hormone by the pituitary gland. About a year later, however, Matthias Tschöp, then at Lilly Research Laboratories in Indianapolis, Indiana, and his colleagues discovered that the hormone has another function as well: It's a potent appetite stimulator.

This discovery helped clear up a major mystery in appetite research, says David Cummings of the Veterans Affairs Puget Sound Health Care System in Seattle, Washington. He points out that people generally want to eat at specific times of day. We want lunch, say, around noon. The trigger for that urge wasn't known; it comes upon us even when there's no food around to stimulate appetite and, Cummings notes, it seemed highly unlikely that leptin could be a “meal initiator,” because fat stores don't drop between breakfast and lunch. But ghrelin seems to fit the bill.

For example, when Stephen Bloom of the Imperial College Faculty of Medicine in London and his colleagues injected ghrelin into human volunteers, it had “an amazingly powerful” effect in increasing the amount of food they subsequently ate, he says. In addition, Cummings and his colleagues, including Puget Sound's Brent Wisse, found that ghrelin levels rose an hour or two before a meal and went down to trough levels afterward—“exactly what was predicted” for a meal initiator, Cummings says. Ghrelin may stimulate appetite by working through the arcuate nucleus, as researchers have found that it activates the NPY/AgRP neurons there.

Central command.

In the arcuate nucleus, NPY/AgRP neurons (green) and POMC/CART neurons (red) fight for control of feeding behavior.


Although ghrelin is part of the short-term appetite-control system, it can, if overproduced, lead to obesity. Prader-Willi syndrome is an inherited condition that causes its victims to be extremely obese—so much so, Cummings says, that they often die before age 30 of obesity- related diseases. The Seattle team found that the patients have what Cummings describes as the “highest ghrelin levels ever measured in any humans,” although the increased ghrelin production is apparently an indirect effect of the other chromosomal abnormalities underlying the disease.

Prader-Willi syndrome is rare, and most obese humans tend to have lower ghrelin levels than people of normal weight, but there is another way in which the hormone may contribute to obesity. The Cummings team reported in the 23 May issue of The New England Journal of Medicine that ghrelin production increased in people who had lost weight through dieting. The hormone may thus be part of the mechanism that undermines a dieter's ability to shed pounds.

Not every form of weight loss causes ghrelin production to go up, however. An operation called gastric bypass, which involves taking a small portion of the upper stomach and reconnecting it to the small intestine, seems to be an effective way of treating extreme obesity. Cummings and his colleagues have found that, for reasons not yet understood, ghrelin levels go down—and stay down—in people who have undergone the surgery. This might be why they don't try to compensate for their smaller stomachs by eating more frequently.

Meals have to be terminated as well as initiated. And recent work by Bloom's group, in collaboration with that of Roger Cone of Oregon Health and Science University in Portland, shows that PYY has an important role to play in that regard. The researchers reported in the 8 August 2002 issue of Nature that infusions of the hormone lead to decreased eating by mice, rats, and human volunteers. The hormone acts in the arcuate nucleus, in this case inhibiting the activity of the appetite-stimulating NPY/AgRP neurons and stimulating the appetite-suppressive POMC cells.

Whether this recently accumulated knowledge about the body's weight-control systems will pay off in better antiobesity treatments remains to be seen. But if it does, both epidemiology and a new experimental study suggest that the reward may be large: a longer life. In addition to lowering one's risk of deadly obesity-related diseases, calorie restriction can extend the life-spans of organisms ranging from the fruit fly to rodents.

Exactly why this is so is not clear. But in the 24 January issue of Science (p. 572), Ronald Kahn, with Barbara Kahn and Matthias Blüher, also at the Joslin Diabetes Center, report that leanness alone may be all it takes. In earlier work, the Boston workers had genetically modified mice so that their fat cells do not make insulin receptors. Those animals have 50% to 70% less fat than unaltered mice, but they otherwise appear healthy.

In the new work, the researchers found that the median life span of the modified mice increased from about 30 months to 33.5 months. Because these animals, despite their leanness, actually eat more than normal mice, the Boston group concludes that the decreased fat tissue produced by calorie restriction, rather than the sparse food intake itself, is what's important for greater longevity. Given how hard it is to lose weight, keeping a life-span perspective in mind might help us resist adding new pounds.

View Abstract

Stay Connected to Science

Navigate This Article