Highlight. The structure of the new infrared protein, and a shot of it glowing inside a living mouse (inset).

X. Shu et al., Science, 8 May 2009, p. 804

Fluorescent Proteins Go Invisible

Fluorescent proteins have been lighting up biology labs for more than a decade, glowing in response to everything from gene expression inside cells to the presence of anthrax and other biowarfare agents. But the most useful fluorescent protein might be the one you never see. Researchers report that they've made a fluorescent protein that emits infrared light. Because infrared can pass through tissue more easily than visible light can, the advance should allow researchers to trace individual molecules throughout the bodies of mice and other small, live animals.

The new protein is a modified version of one found in the bacterium Deinococcus radiodurans, renowned for its ability to survive huge doses of radiation. Previous groups have noted that one of its proteins, known as a phytochrome, absorbs deep-red light at the far end of the visible spectrum. It uses that energy to send a signal to the cell to turn on particular genes.

Could the energy instead be used to create infrared light? Roger Tsien, a biochemist at the University of California, San Diego, who shared last year's Nobel Prize in chemistry for his work on fluorescent proteins (ScienceNOW, 8 October 2008), and his colleagues decided to find out. They reworked the protein's genetic code, chopping off the portion that carries out the biochemical signaling. The effort produced infrared fluorescent proteins, but they only shone weakly. So the researchers went through several rounds of mutating the modified phytochrome gene, selecting the strongest ones. It worked, producing a protein that shone four times as bright as the original, the team reports in tomorrow's issue of Science. The researchers also inserted the gene for the new protein into an adenovirus that infects mouse liver. Five days after injecting the virus into mice's tail veins, they spotted infrared fluorescence from the animals's livers (see picture).

The new work "is an important step in the right direction," says John Frangioni, a fluorescence imaging expert at Beth Israel Deaconess Medical Center and Harvard Medical School in Boston, Massachusetts. Researchers might use the technology to probe how diseases such as cancer progress at the molecular level, he says. But Frangioni notes that much of the light from the new protein can still be blocked by animal tissue, so researchers will have to find infrared proteins that emit at longer wavelengths. That could be possible, says Tsien, because genetic databases already harbor more than 1500 bacteria proteins resembling phytochromes. So the race will continue to see if one of those can shine in biology's sweet spot.