Wherever it's buried in the body, a disease leaves traces in the blood—or so the thinking goes. But finding these biomarkers, which can help catch the disease early on, has been an exercise in futility, with one promising candidate after another losing its luster once it receives scrutiny. A team of chemists and other researchers now propose a new way to pick up biomarkers with a blood test: by screening for antibodies that the body makes in response to particular diseases. So far, the group has reported results for only a small number of Alzheimer's disease patients. But they are hopeful that the approach will hold up and could be used for everything from lupus to cancer.
There are two common strategies for finding biomarkers in the blood or elsewhere in the body. The first is to focus on what's known about a disease—for example, looking for deterioration in certain brain regions in Alzheimer's disease. The second is essentially a fishing expedition, in which researchers compare, say, protein patterns from patients who have a particular disease with patterns from people who don't have it. Both methods have run into roadblocks. In the case of Alzheimer's disease, PET scans, MRIs, and spinal taps that remove some of the fluid that surrounds the brain and spinal cord have had some success in early diagnosis, but they are expensive or invasive. Even though there's no treatment for Alzheimer's disease, researchers are keen to find easy-to-use and reliable biomarkers, which would help ensure that people in Alzheimer's clinical trials really do have the disease and would make early treatment possible if it eventually becomes available.
Chemist Thomas Kodadek of The Scripps Research Institute in Jupiter, Florida, and his then-postdoc, M. Muralidhar Reddy, considered biomarkers that harness a classic feature of disease biology: antibodies, proteins that the immune system churns out in the presence of invaders. Scientists don't know that every disease elicits antibodies, but some diseases certainly do. "Antibodies are the rocks of the protein world," not easily damaged when studied in the lab, says Kodadek. This made the idea of measuring them in blood appealing.
Many other researchers had considered the value of antibodies as biomarkers, but they were stymied by the common strategy used to test for them. To know which antibodies to look for, you have to determine which molecules stimulate the immune system to produce the antibodies. This requires "a phenomenal understanding of early disease progression," says Kodadek, something we don't have for most diseases.
Kodadek, Reddy, and their colleagues took a different tack. They turned to "libraries" of thousands of peptoids, molecules that are developed into drugs that have myriad different structures. The thinking was that some peptoids, just by chance, would bind to whatever antibodies might be there—just as a few of several thousand randomly chosen shoes might fit your feet—allowing the researchers to determine whether people with a disease had an abundance of certain antibodies that healthy people lacked. In mice with a version of multiple sclerosis, a library of 4600 peptoids helped identify three antibodies, which the team used to diagnose the disease in other mice.
In six people with Alzheimer's disease, the same strategy, using 15,000 peptoids, picked up two antibodies found at high levels. The antibodies were also abundant in the blood from an additional 16 Alzheimer's patients. But these proteins were uncommon in the blood of a handful of people with Parkinson's disease or lupus.
The antibodies were also prevalent in two of the 16 healthy controls. Their presence could suggest that the biomarkers aren't specific to Alzheimer's disease. Or it could suggest that these two women—a 75-year-old and a 65-year-old—have early Alzheimer's disease. "We favor the latter hypothesis, but this cannot be concluded with certainty," the authors report  in the 7 January issue of Cell .
"It's an innovative idea," says Norman Relkin, a neurologist and neuroscientist at Weill Cornell Medical College in New York City. "You're looking for immune responses which may be disease-specific." Still, Relkin says that for Alzheimer's disease at least, it will take a lot more work to make the test reliable, especially because even with healthy aging, "the immune system tends to produce more dysfunctional antibodies."
Although the findings need to be replicated in more people to make sure the antibodies are specific to Alzheimer's disease, the paper "looks like a very thorough job," says Kaj Blennow, a neurochemist at Sahlgrenska University Hospital in Gothenburg, Sweden. A blood test would have big advantages over current Alzheimer's diagnostics, he says.
Kodadek, Reddy, and others have formed a company, Opko Health Inc., to further develop the technology; Reddy is the chief scientific officer. "As you can tell, I'm pretty excited about this," says Kodadek. But he's trying to keep his enthusiasm in check. "There's a long history of biomarkers hitting the graveyard," and although he's hopeful that won't happen here, "the point is, you never know."