DURHAM -- A new world of medical discovery lies in the microscopic odds and ends that circulate in blood and other bodily fluids.
These tiny molecules are called metabolites. Byproducts and components of the body's energy factory, they can signal with great accuracy the onset of diseases, including ailments such as heart disease and diabetes.
Finding the biological markers and studying their pathways are the bases of the emerging science of metabolomics (meh-TA-boh-LOM-ics) Hailed as a noninvasive way to pinpoint when food, drugs, environmental pollutants and genes begin causing harm, metabolomic science is already producing tantalizing results.
This year, a team of Duke researchers reported that metabolic markers circulating in blood could identify people with cardiovascular disease. Not only that, they could also predict who was likely to suffer a heart attack or other life-threatening event - something no stress test or expensive catheterization can now do.
Two other Duke teams found that diabetes isn't just a problem of sugar metabolism, but that protein and fat also play a role in triggering the disease in people who are obese.
And scientists at RTI International, a private research group in Research Triangle Park, have found that certain metabolites correlate to early signs of liver damage from drugs such as acetaminophen, fostering hope that a simple blood or urine test might one day predict who will develop problems.
More findings are on the way. Last month RTI announced that it is committing $2 million to a new center that will include metabolomics research. And a Duke team, led by Dr. Svati Shah, a cardiologist, recently received a $3.5 million federal stimulus grant to continue metabolomic studies in heart disease.
"It speaks to the promise of this field," Shah said.
Originally a genetics researcher, Shah said she is fascinated by the link between genes and metabolism. But where genes provide a rough blueprint predicting a person's risk of disease, metabolites reveal the here-and-now evidence that problems are afoot.
"It's a strong fingerprint of a person's condition," explained Christopher Newgard, director of the Sarah W. Stedman Nutrition and Metabolism Center, where Duke's metabolomics research is concentrated.
Mapping 6,500 metabolites
Newgard oversees a 35,000-square-foot laboratory housed in a former biotechnology company's office complex north of downtown, far from the Duke University campus. He collaborates with scientists throughout the university, running analyses using sophisticated instruments that measure the masses and concentrations of molecules.
Because the science is relatively new, there's much to be done.
Like a road map of the United States, a giant wall chart hangs in Newgard's lab. It's a mind-boggling jumble of the estimated 6,500 individual metabolites produced in the human body.
Mapping these metabolites has been a mission of scientists in Canada, who announced in 2007 that they had accounted for 2,200 of the molecules in an effort not unlike the Human Genome Project that decoded DNA.
By knowing healthy, normal metabolic interactions, Newgard said, scientists can detect early, subtle changes that eventually lead to illness. One of the most promising features of the science is that these disruptions in metabolic processes are evident in blood, urine and saliva - easily accessed body fluids. As a result, the science could lead to simple screenings for toxic drug reactions, pollution exposures, diseases and genetic abnormalities.
Newgard said his mission is not simply to confirm a diagnosis, but to predict it and understand how it came about.
"What we want to stand for is new insights into biology," he said. "We've been doing this for six years, and for three years, it was pretty dry. Now the information is starting to come out."
Fats, proteins and diabetes
In diabetes studies that he co-wrote last year, Newgard said metabolic components offered a surprising finding about the role of fats and proteins in the development of diabetes. In that common disease, the blood sugar glucose is the standard metabolite everyone focuses on. A product of carbohydrate metabolism, glucose is elevated in a diabetic's blood because the body's cells can't take it in. All that excess glucose eventually damages organs.
But Newgard and the Duke teams found another metabolic marker in human studies that could offer a new explanation for how obese people develop type 2 diabetes, the most common form of the disease and one often associated with poor lifestyle choices. The new marker was associated with protein metabolism, not the usual carbohydrate culprit, and was highly accurate in predicting insulin resistance.
In addition, a group led by Deborah Muoio, an assistant professor of medicine, found that high levels of fat created a metabolic blockage that impaired the body's ability to switch between glucose and fat for fuel. Combined, the findings indicate that high-calorie, fatty foods - a common feature of the typical American diet - may trigger insulin resistance in obese people.
Some of those same metabolites were subsequently found to be associated with heart disease.
Dr. William Kraus, a Duke cardiologist who worked with Shah and Newgard on the heart disease study, said the researchers were amazed at how well the metabolic markers predicted severe cardiovascular disease.
Predicting heart-attack risk
They studied blood samples of patients who had undergone traditional screenings years ago at Duke, including catheterizations that hunt for blockages. Some of those people turned out to have no heart disease, and others had significant problems.
When the metabolomic team analyzed the patients' blood samples, they found that tell-tale metabolite clusters could indicate who was healthy and who had disease. In addition, one blood marker clearly predicted which patients would end up having heart attacks.
"This is really the holy grail in cardiology," Kraus said. "You can do a catheterization on somebody and see they have obstruction. But you can't predict who is going to have a heart attack."
The group is now working to determine whether the metabolic clusters are simply signals of disease, or somehow instigate it.
Shah said the possibilities of the new science are dramatic, particularly if a blood test can replace expensive procedures - and save lives.
"We live in a society where patients want an answer," she said. "Chest pain is a very common complaint. If we could come up with markers that are used either alone, or in combination with stress testing, and tell a patient with a greater degree of certainty that you don't have blockage, we can avoid a heart catheterization."
Detecting liver damage
A similar hope motivates the work of researchers at RTI, who have been using metabolomic science to detect liver damage that can occur with certain drugs. Therapies for tuberculosis or epilepsy - and even the main ingredient in Tylenol - can be highly toxic to the liver in some patients. But it's not clear which patients might develop problems.
Scientists at RTI have been observing metabolites in the urine of rats dosed with one of five drugs they have studied, including acetaminophen and a therapy for tuberculosis. .
Susan Sumner, head of the metabolomics and obesity research program at RTI's new Consortium for Molecular Epidemiology, Genomics, Environment, and Health, said they have discovered metabolic signatures that suggest liver disease coming on.
"If we can go and validate this marker profile now for clinical use, all these people on drugs can be monitored," Sumner said, adding that doctors could then change treatment in time to avert permanent damage.
She said the research institute's new center will focus additional metabolomic studies on problems of obesity, working with doctors at East Carolina University's Brody School of Medicine.
"Obesity is fundamentally an energy imbalance," she said. And that disease state, like any other, starts with a few molecules gone awry.
