Researchers at the University of California, San Francisco, or UCSF, have developed an imaging tool that could help doctors locate and visualize bacterial infections in the body and to rule out other common causes of inflammation.
The research team reported in Scientific Reports that scans made with the imaging technique known as positron emission tomography, or PET, detected infections in mice caused by either of the two broad groups of bacteria, gram-negative and gram-positive, without generating a signal from other causes of inflammation.
To perform PET imaging, doctors inject patients with small doses of "radiotracers" that bind to particular proteins or accumulate in tumors, inflamed areas, and other problem spots. The most commonly used tracer, a sugar-like molecule called FDG, accumulates in infected areas, but also follows immune cells to germ-free inflammation sites and tumors.
The treatment for a sterile inflammation is the last thing doctors would want for a patient with an infection.
Other tracers, like radiolabeled antibodies that attach to particular bacteria, could easily miss many infectious strains, and can also emit a stronger signal from dead bacteria, which often have ruptured and spilled their contents, than from intact, live ones.
The search for a better radiotracer brought Michael Ohliger, an assistant professor of radiology and biomedical imaging at UCSF and one of the paper's senior authors, together with fellow UCSF researchers David Wilson and Oren Rosenberg, the paper's other senior authors.
For them, the ideal molecule would detect only live bacteria, rather than bind to living or dead cells indiscriminately; it had play an active part in their growth. And it couldn't be a substance used by human cells, because then every cell in the body would "light up" on a PET scan.
One group of molecules that fit the bill was the D-amino acids, which bacteria take up from their environment to build their protective cell walls. These molecules are mirror images of the L-amino acids, which all organisms use to build proteins. But human cells make much smaller use of the D variety.
The team reasoned that a radiolabeled D-amino acid would zero in on bacteria. They settled upon D-methionine, a minor component of the bacterial cell walls that they found gives a strong signal when radiolabeled. To probe D-methionine's capabilities, they injected infectious bacteria - both the gram-negative Staphylococcus aureus and the gram-positive Escherichia coli - into mice.
When they later injected D-methionine molecules tagged with a single radioactive Carbon-11 atom into the mice, PET scanning showed the radiotracer accumulating at both kinds of injection sites.
If D-methionine-based PET imaging were approved for use in humans, it would let doctors facing challenging diagnoses find and treat infections much more quickly. The method could also give greater certainty to doctors when prescribing antibiotics, which, if overused, can promote resistant bacterial strains.
