Proceed With Care

Proceed With Care
May 25, 2006
Celia Henry Arnaud
Chemical and Engineering News
A new study sounds a cautionary note for one form of RNA interference (RNAi). Researchers at Stanford University show that sustained high doses of short hairpin RNA (shRNA) can lead to liver toxicity and death in mice (Nature 2006, 441, 537).

COURTESY OF MARK KAY

RUNNING INTERFERENCE Kay, Grimm, and colleagues found that some short hairpin RNAs are toxic.

In RNAi, short pieces of RNA are used to silence gene expression. This RNA can either be introduced directly into cells as short interfering RNA (siRNA) or be delivered via a viral vector that produces shRNA, which goes through additional processing in the microRNA pathway in the cell to give siRNA.
The Stanford researchers, led by postdoc Dirk Grimm and genetics professor Mark A. Kay, use 49 different shRNAs, which they deliver using an adeno-associated virus, to target hepatitis B and other genes in the livers of mice. Of these shRNAs, 36 caused liver damage and 23 killed the mice. The toxicity may be caused by overloading exportin-5, the protein in the microRNA pathway that transports the shRNA from the nucleus to the cytoplasm for further processing. Synthetic siRNAs enter the pathway at a later point, bypassing this process.
"This is the first study where somebody is able to get very high levels of a vector to express shRNA in every single cell of the target organ," Kay says. His team used a virus known to efficiently transfer RNA into cells and tweaked the construct such that it would produce extremely high doses of shRNA in cells. "We were probably overzealous trying to make the most robust method possible," he says.
John J. Rossi, a molecular biologist at the Beckman Research Institute of the City of Hope in Duarte, Calif., doesn't believe that this study should prevent the use of shRNA. "There's really no reason to be fearful of using shRNAs. They just have to be used with caution," he says. "You have to make sure that your shRNA is not perturbing the microRNA pathway." Kay's study serves as a reminder that "overexpression of any shRNA can be toxic to cells," Rossi says.
Irvin S.Y. Chen, a molecular biologist at the University of California at Los Angeles, has seen cases in which shRNAs that are not toxic in cancer cell lines turn out to be toxic in primary cell cultures. Such differences highlight the importance of selecting the right model system. "It's important to use a model system that's close to the human situation, ideally in primary cells or an animal model system," such as the one used here, Chen says.
Phillip D. Zamore, a biochemist at the University of Massachusetts Medical School, Worcester, thinks the study strengthens the case for using siRNAs rather than shRNA for RNAi. He points out that the nontoxic shRNAs in the study are the ones that look the most like siRNAs. "People will now use more biologically authentic methods to generate small RNAs in vivo," he predicts.
Comments: 0
Votes:40