Dying cells in a tumor (upper left) or damaged tissue release DNA fragments that course through blood and that contain markers that point to their origin in the body.

Dying cells in a tumor (upper left) or damaged tissue release DNA fragments that course through blood and that contain markers that point to their origin in the body.

Chris Bickel/Science

Tests of blood-borne DNA pinpoint tissue damage

In every living person, even the healthiest, death abounds. Cells throughout a person die naturally all the time, shedding fragments of DNA into the blood. When injuries or illness damage specific parts of a body, cell death generates even more of this so-called circulating DNA. Several research teams are now developing ways to trace it to the tissue from which it originated, hoping to detect early stages of a disease or monitor its progression.

“Noninvasive measurement of cell death is a super exciting area with endless applications,” says developmental biologist Yuval Dor of the Hebrew University of Jerusalem. Dor’s own group reports this week that its technique for tracing the origin of circulating DNA detected the expected type of cell death in people with pancreatic cancer, type 1 diabetes, multiple sclerosis, and brain injuries. Two more groups have reported preliminary but encouraging results for other cancers. 

The published work does no more than prove the concept, says Alain Thierry, director of research at the National Institute of Health and Medical Research in Paris. The circulating DNA techniques still need to be vetted on a large scale, and they face competition from other diagnostic tests that rely on blood-borne hormones, metabolites, and other nongenetic molecules. Still, Thierry says, “having information from circulating DNA from a blood test could be very powerful” for detecting or tracking diseases. 

A few assays based on circulating DNA are already in use. Since 2011, doctors can order a test for Down syndrome that analyzes DNA shed by a fetus into a pregnant woman’s blood. Cancer researchers monitor circulating DNA carrying cancer-related mutations to follow patients’ responses to targeted therapies and are working on similar tests for early signs of cancer. But those tests cannot identify the DNA’s origins. For many other diseases that boost tissue-specific cell death, circulating DNA had little to offer.

One potential solution is to look at chemicals called methyl groups, which cells attach to certain genes to prevent them from being transcribed into proteins. Each cell type, whether muscle, neuron, or blood, has a distinct DNA methylation pattern. In October 2015, a group led by Dennis Lo of The Chinese University of Hong Kong reported in the Proceedings of the National Academy of Sciences (PNAS) that they could detect those distinctive patterns in circulating DNA. In liver cancer patients, for example, the team saw a rise in DNA traceable to dying liver cells. Their technique, however, relied on whole genome analysis of methylation that costs upward of $1000 per sample.

In PNAS this week, a team led by Dor, DNA methylation expert Ruth Shemer at the Hebrew University of Jerusalem, and endocrinologist Benjamin Glaser from the Hadassah Medical Center in Jerusalem offers a potentially simpler, and cheaper, approach. The group drew on known methylation signatures for various tissues to pick out methylated sites along the genome that provide unique fingerprints. They report that by scanning circulating DNA for those hotspots in people recently diagnosed with type 1 diabetes, they detected dying β cells—the insulin-producing cells in the pancreas—in all of the 11 people tested. Their test also revealed evidence of dying β cells in the blood of people who had received islet cell transplants—a potential sign of immune rejection of the islets. 

Immunobiologist Kevan Herold of Yale University says the result bodes well for a circulating DNA test that could be used to screen people at high risk for type 1 diabetes before their pancreas is so damaged that blood sugar levels rise. “The hope is that we could intervene at an early stage and try to prevent progression,” he says. 

Circulating DNA can also flag other conditions, Dor and his colleagues showed. In blood from 14 out of 19 patients with relapsing multiple sclerosis, they detected circulating DNA shed by brain cells called oligodendrocytes, which the disease destroys. The team could also identify dying brain cells in blood samples from people who had brain damage due to head trauma or cardiac arrest. For pancreatic cancer, the team saw signs of pancreatic cell death in the blood of about half of 42 patients; the group even distinguished cancer from pancreatitis—which also raised cell-free DNA levels in seven of 10 patients—by incorporating a test for a cancer mutation. 

Methylation isn’t the only way to trace circulating DNA. In January, a team at the University of Washington, Seattle, described in Cell a test that relies on tissue-specific differences in how DNA is packaged in structures called nucleosomes. In three out of five people with advanced cancers, the team could use those nucleosome fingerprints to trace circulating DNA to the cancerous tissue

Researchers now plan to test the nucleosome and methylation approaches in larger groups of people. Clinical biochemist Eleftherios Diamandis of the University of Toronto in Canada predicts that at least for initially diagnosing cancer, the tests may face an obstacle. In the early stages of cancer—just when the tests would be most useful—there may be little or no circulating tumor DNA in a person’s blood. “I am not sure if even the most fascinating technical developments will solve the issue of abundance,” Diamandis says. Still, some were also once skeptical that Down syndrome could be reliably picked up from fetal DNA in a pregnant woman’s blood, and now thousands of women a year use the screening tool.