PHILADELPHIA, PENNSYLVANIA—Tuberculosis (TB) is one of the biggest scourges there is among infectious diseases, killing nearly 2 million people a year, most in developing countries. It is also notoriously hard and slow to detect in places without top-flight health care systems. Yesterday, researchers reported at the American Chemical Society (ACS) meeting here that they’ve devised a simple new way to diagnose TB, and even distinguish living TB cells from dead ones, which could give doctors an easy way to see whether their anti-TB medications are working. The researchers are now putting their diagnostic through its paces with samples from patients in South Africa, which has one of the world’s highest incidences of TB, and they hope to launch a clinical trial of their test soon.
According to the World Health Organization, nearly 10 million people contract TB every year. That happens when someone infected with Mycobacterium tuberculosis coughs, or even speaks, and others nearby inhale tiny, bacteria-laden droplets. In wealthy countries, doctors take lung x-rays of patients suspected of having TB and test their sputum (mucus and spit) for DNA markers of the bacterium. But in developing countries, where there is limited access to these technologies, technicians often turn to a procedure known as the Ziehl–Neelsen (ZN) test.
Developed more than a century ago, the ZN test spritzes dye-laden liquids onto a sputum sample. After extensive processing, those dyes latch onto water-excluding “hydrophobic” compounds that are abundant in Mycobacteria membranes. But the test is slow to administer, not particularly sensitive, and gives many false positives, because many bacterial membranes contain hydrophobic compounds, says Carolyn Bertozzi, a chemist at Stanford University in Palo Alto, California.
Bertozzi and her colleagues wondered whether there wasn’t a better way to flag TB. Her team had spent more than a decade studying how different organisms, including pathogenic bacteria, attach a wide variety of sugar compounds to the proteins and fatty molecules called lipids that make up their cell membranes. In their early studies they found that unlike most other bacteria, and most other organisms, Mycobacterium tuberculosis and its close relatives use a type of sugar known as trehalose to carry out this construction. “We wondered whether we could use this to mark TB cells,” Bertozzi says.
The researchers designed a series of different trehalose sugars tagged with a fluorescent dye abbreviated DMN. The dye glows bright green when it absorbs light—but not always. If it’s surrounded by even the slightest bit of water it doesn’t shine. When water is excluded, as it is in the hydrophobic interior of a cell membrane, the dye lights up. Bertozzi’s hope was that if her team fed their dye-labeled sugars to the TB bugs, the microbes would take them up and append them to their cell membrane lipids, turning them green. That wouldn’t happen for dead TB cells that can’t take up the sugar, nor for the cells of most other organisms, making it possible to spot a live TB infection in highly diverse sputum samples.
In lab studies the researchers found just that. Bertozzi reported at the meeting that live Mycobacterium samples shone bright green under a fluorescence microscope, whereas those containing common bacteria, such as Escherichia coli or Staphylococcus aureus, didn’t fluoresce at all. What’s more, the live TB cells begin to glow in as little as 5 minutes after being fed the sugar, and turn bright green within 1 hour. ZN tests, by contrast, can take hours and often miss low levels of infection. Another standard test, wherein TB proteins are injected under the skin and doctors look for an immune response that shows a person has been infected, can take 3 days to produce a result.
Bertozzi and her colleagues also collaborated with Bavesh Kana, a biochemist at the University of the Witwatersrand in Johannesburg, South Africa, who provided sputum samples of patients suspected of having TB. Lab tests on those samples not only quickly flagged live TB, but showed a close match with more exacting DNA studies. Bertozzi says that she and her colleagues are hoping to launch a clinical trial of their diagnostic to test its success under real-world conditions.
“It’s pretty spectacular,” says Dale Boger, a chemist at the Scripps Research Institute in San Diego, California, who attended Bertozzi’s session. Boger adds, however, that even if TB is detected rapidly its resistance to current medications continues to be a major problem. Bertozzi notes that if the new test proves successful, it could help fight such resistance by allowing doctors to determine what medications are and aren’t working to defeat one of the most vexing and dangerous infectious diseases.