Gotcha! New test stalks diseases early

Chemists adapt a DNA test to measure proteins that mark HIV, cancer and more

PCR machine

Chemists have created a powerful new test for detecting disease. It uses a common DNA-copying procedure, called PCR, to screen for antibodies in tiny test-tubes containing small amounts of blood or saliva.


Diseases sneak up on us. They begin triggering changes in the body long before we feel sick. But a new method has the potential to spot telltale signs early, before a disease gains a foothold. That could help doctors diagnose — and treat — disorders before they do too much damage.

The new technique detects antibodies. These are substances made by the immune system. They stick to foreign material, such as germs or cancer cells. This essentially brands the foreigners so the body knows to attack those cells. But occasionally antibodies will brand the wrong cells, triggering their destruction. That can lead to what is known as autoimmune disease. Such conditions include a type of diabetes that can occur in children.

Good or bad, antibodies serve as disease markers. Before symptoms show, these proteins lurk in the blood. But most lab tests can’t detect them early in the onset of a disease. That’s because there’s too little of these markers for a test to spot.

Peter Robinson and Cheng-ting Tsai are graduate students who work in a chemistry lab at Stanford University in Palo Alto, Calif. They were thinking about this problem when they got a flash of insight.

For years, researchers have been using tests to study a person’s DNA. These methods can measure super-small amounts of genetic material using a process known as PCR. Those initials stand for polymerase (Puh-LIM-ur-ace) chain reaction. This process makes millions of copies of DNA fragments in just a few hours. It offers a fast, cheap way to diagnose disorders caused by faulty genes. It also can identify the DNA fingerprint of disease-causing germs.

But antibodies are proteins. And measuring them is more cumbersome than studying DNA. Some tests don’t work well because they force proteins into unnatural shapes. Other tests keep proteins in their normal shape but require expensive chemicals or special equipment.

An “aha!” moment came while Robinson was reading about HIV. This is the virus that attacks immune cells, causing acquired immune deficiency syndrome, or AIDS. The immune system makes antibodies in response to HIV. And most HIV tests check for antibodies to this virus.

Those tests don’t actually scout directly for HIV. Upon learning that, Robinson got the idea to adapt the PCR test to detect HIV antibodies. This idea, he says, “just kind of clicked.”

The innovation

Typically, antibodies are fished out of blood samples by adding molecular “bait.” That’s some compound that the antibody will bind to. Robinson and Tsai did something clever: They attached short pieces of DNA to the “bait.” They used two DNA bits. Let’s call them “A” and “B.” For PCR to work, the “A” and “B” strands need to be side-by-side — but this would hardly ever occur when the bait floats randomly in a solution. However, when antibodies glom onto the bait, the “A” and “B” fragments are much more likely to come close together — close enough for the PCR to work.

PCR makes huge numbers of copies of a desired bit of DNA. If it made lots of the bait DNA, scientists would realize which antibodies had been present, even if there had been only tiny amounts — as would occur in the early stages of a disease.

Still, it wasn’t clear if the technique would work on real samples from sick people. That’s the most challenging part of developing a new medical test, known as an assay, Tsai says. “It’s scary and exciting to look at the results. They tell us whether to keep pursuing the project.”

Initially the team tested its assay on a marker for thyroid cancer. (Not all cancers have a reliable disease marker, but this one does. And it was easy to get patient samples.) Compared with four existing tests for the same marker, the new one was 1,000 to 10,000 times more sensitive. That means the new procedure required only a thousandth to a ten-thousandth as much marker to make a positive identification.

This sensitivity was dramatic. Indeed, Robinson notes, “We were pretty surprised.” The team described its new advance in a paper posted online February 16 in ACS Central Science.

The new findings caught the attention of Mark Pandori. He is a microbiologist at the University of California, San Francisco. He also heads a county public-health lab. Each year, that lab screens thousands of people for HIV. Traditional methods measure HIV antibodies in blood. Nowadays, the lab also uses swab tests that scout for antibodies in saliva.

While more convenient and popular, oral HIV tests can be misleading. Antibodies in saliva are present at only a hundredth to one-thousandth the concentration that they can be found in blood. That means an oral test is only reliable if done at least 40 to 50 days after a person is infected, when lots of antibodies have developed. But people are most contagious 21 to 42 days (3 to 6 weeks) after getting infected, Pandori says. So during that important window of time, people can’t know they are HIV-positive because antibody counts are too low for common tests to pick up.

By comparison, blood tests can detect HIV antibodies just 20 to 21 days after a person first picks up the virus. With the new PCR-based test, “We’d have the sensitivity of the blood test with the convenience of oral fluid samples,” Pandori explains.

Next steps

The Stanford team is now working to validate its test with saliva from the public-health lab. If successful, the test could screen large populations of people for HIV infection at the most valuable time for starting treatment.

The researchers also are hoping to develop their assay for type 1 diabetes, the kind that can start during childhood. The immune systems in people with this disease make harmful antibodies. Those antibodies trigger the killing of cells in the pancreas. That’s the organ that makes the hormone insulin. (The other form of diabetes — known as type 2 — is not an autoimmune disease. People with type 2 diabetes make insulin, but their cells don’t use it as well as they should.)

Srinath Sanda is a pediatrician and researcher at the University of California, San Francisco. He sees a lot of kids with diabetes. He thinks the new assay looks promising. Having a fast, accurate assay could “help determine a patient’s type of diabetes and ensure they receive the right treatment,” he says.



This is one in a series presenting news on technology and innovation, made possible with generous support from the Lemelson Foundation.

Esther Landhuis is a freelance journalist in the San Francisco Bay Area. She worked on her high school newspaper and spent a decade studying biology before discovering a career that combines writing and science.

More Stories from Science News Explores on Chemistry