Read our COVID-19 research and news.

Bacteria with rewired genetic circuitry detect diabetes and cancer.

Bacteria with rewired genetic circuitry detect diabetes and cancer.

Chris Bickel/Science

Engineered bacteria detect cancer and diabetes in urine

Most of us think of bacteria as the enemy, but each of our bodies harbors trillions of microbes, most of them beneficial or benign. Now, you can add two new friendlies to the list. This week, two groups of synthetic biologists seeking to repurpose living microbes for human benefit report genetically modifying bacteria to detect cancer in mice and diabetes in humans.

Clinicians have sought to exploit microbes for more than a century. Beginning in 1891, an American bone surgeon named William Coley injected more than 1000 patients with bacterial colonies in hopes that they would shrink inoperable tumors. The treatment sometimes worked, in part because the microbes preferentially seek out tumor tissue, which is rich in nutrients yet has few immune cells to knock out any pathogens. But the results were uneven, and with the rise of radiation and chemotherapy, the approach fell out of favor. More recently, synthetic biologists have begun to modify bacteria to fight cancer and other diseases—engineering them to secrete toxins inside tumors, for example. A couple of these therapies have even made it into clinical trials, though none have been approved yet.

Far less effort has been directed at using bacteria as a test for disease. Sangeeta Bhatia, a biomedical engineer at the Massachusetts Institute of Technology (MIT) in Cambridge, and her colleagues previously worked on cancer detection using metal nanoparticles. In the presence of a tumor, the particles would release snippets of proteins called peptides that could be detected in the urine. Unfortunately, Bhatia says, the signal was often too weak to serve as a clear indicator of disease. Bhatia’s team then realized that bacteria offered a potentially superior option. The researchers knew that microbes with a taste for tumor often penetrate the masses as they grow and replicate. So Bhatia’s group joined up with a team led by Jeff Hasty, a bioengineer at the University of California, San Diego, to reprogram bacteria that could be fed to mice and, in the presence of cancer, would produce a luminescent signal with a simple urine test.

They started with a harmless strain of bacteria called Escherichia coli Nissle 1917, which is commonly added to yogurt and other foods as a probiotic to promote digestive health. First, they fed the bacteria to mice and confirmed that the microbes crossed the gut and colonized tumors in the liver. They engineered the bacteria to produce a naturally occurring enzyme called LacZ when they encountered a tumor. Next, the researchers injected mice with compounds that were precursors for light emitters. These were two-part molecules made up of a sugar linked to luciferin, a luminescent molecule. When bound together, the pair doesn’t emit light, but LacZ acts like a pair of scissors that cuts the two apart. So, in mice that had liver cancer populated by E. coli, the LacZ produced by the microbes released the luminescent compound, which was then excreted in the animals’ urine, turning those samples from yellow to red. What's more, Bhatia and her colleagues report in the current issue of Science Translational Medicine this week, while conventional imaging techniques struggle to detect liver tumors smaller than 1 square centimeter, this approach was able to flag tumors as small as 1 square millimeter.

In a separate study also reported in the current issue of Science Translational Medicine, researchers led by structural biochemist Jerome Bonnet of the University of Montpellier in France followed a related strategy to detect a key sign of diabetes, namely elevated glucose in the urine of human patients. The researchers added genetic circuitry to the bacteria so that they produced a large amount of a red fluorescent protein once the concentration of glucose in their surroundings reached a certain level. In this case, however, the team’s strain of E. coli wasn’t injected into people first, rather simply added to urine samples, where they produced a color change. For now, this approach isn’t any better than a standard glucose meter. But because the detection scheme can be repurposed to detect other targets, it could serve as a platform for a broad array of future diagnostics.

“They are both nice advances for the field,” says Jim Collins, a synthetic biologist at MIT. But he cautions that both approaches remain years away from being approved for clinical use. Tim Lu, also a synthetic biologist at MIT, agrees. “Taken together this pair of papers demonstrates that synthetic biology will be useful not only for therapeutics but diagnostics as well.” That might just give bacteria a good reputation after all.