Researchers have performed the first genome-wide in vivo RNA interference screen in mice, using it to rapidly uncover new regulators of tumor growth in a model of squamous cell carcinoma and identify some potential therapeutic targets for the disease. The findings are published August 14 in Nature (1).
“The technology revolutionizes the speed at which we can now evaluate the functional significance of genes in mammalian epithelial tissues,” says senior author Elaine Fuchs, at The Rockefeller University in New York. “Moreover, our method enables us to target not only the skin, but also the epithelium of the eye, head and neck, breast, and reproductive tracts.”
In mammals, genome-wide RNAi screens—in which small pieces of RNA are used to systematically dampen or eliminate expression of thousands of genes—have until now been confined to a petri dish.
In 2010, Fuchs and her colleagues laid out a protocol for RNAi in living mice, knocking down two skin cancer-linked genes in the epithelial tissue of embryos and creating a new quantitative assay for measuring their effects on downstream genes (2). The group published details on that protocol earlier this year (3).
Their original goal, however, was to create an unbiased large-scale screen for genes regulating tumor growth. “We reasoned that if you could do this with 1 or 2 genes at a time, why not do it with 16,000 genes?” says co-author Slobodan Beronja, now an assistant member at the Fred Hutchinson Cancer Research Center, who conducted the research as a postdoc in Fuchs’ lab. But the process of pooling the lentiviral shRNAs was not simple, and by itself took three to four months.
In the new study, the team injected shRNAs targeting nearly 16,000 genes into pregnant mice whose embryos were 9.5 days old. After nine more days of development, the scientists removed the embryos and quantified the individual shRNAs through sequencing.
The group then compared the levels of each shRNA to those in an initial pool, to learn how each of the gene targets affected growth. If a given shRNA disrupted an essential mediator of growth, the cells and the shRNA found in those cells would be eliminated or significantly reduced by day 18.5 in the treated animals; conversely, if an shRNA disrupted a negative growth regulator, those cells would proliferate and the shRNA would be overrepresented.
The researchers identified 1800 genes necessary for normal growth and a set of about 160 unique genes as regulators of tumor growth in a mouse model of squamous cell carcinoma, including some that were already known and some unexpected, like Mllt6—a gene previously linked to leukemia but not SCC.
About 120 of those genes are potential targets for cancer therapy, Beronja says.
The screen used about 100 litters of mice. In theory, if the scientists could tell with certainty that each of the 120,000 epidermal cells they targeted contained only one shRNA, the experiment could use a single animal. However, the injected virus passively diffuses into cells, and the process isn’t perfect. Too much virus winds up infecting most cells, each one containing several shRNAs; too little virus doesn’t infect enough cells to include all the shRNAs in one embryo, Beronja says. That means there was a delicate balance to strike between the concentration of virus and the number of animals used in the study, he adds.
Still, now that the steps of the process have been worked out and the shRNA libraries are commercially available, the protocol should be relatively straightforward. The group has already trained others to implement their techniques, and is open to helping new groups get started. “This study was done in a single lab,” Beronja says. “We hope that this sends the message that this is not that complicated.”
In addition, although the new study focused on skin, the method could also be adapted to study brain, mammary gland, lungs, and other tissues. “If you can infect a skin cell you can do it in any cell,” Beronja says.
1. Beronja, S, Janki P, Heller E, Lien W, Keyes BE, Oshimori N, Fuchs E. 2013. RNAi screens in mice identify physiological regulators of oncogenic growth. doi:10.1038/nature12464 [Epub ahead of print]
2. Beronja S, Livshits G, Williams S, Fuchs E. 2010. Rapid functional dissection of genetic networks via tissue-specific transduction and RNAi in mouse embryos. Nature Medicine 16(7): 821-8277. doi: 10.1038/nm.2167
3. Beronja S, Fuchs E. 2013. RNAi-mediated gene function analysis in skin. Methods Mol Biol. 961:351-61. doi: 10.1007/978-1-62703-227-8_23.