DESPITE all the excitement it has generated, RNA interference might not be a magic bullet for treating diseases after all. Two reports have come to radically different conclusions about whether the technique targets only specific genes in human cells, or accidentally hits others as well.
RNA interference (RNAi) is the ability of plant and animal cells to block a specific gene by destroying the RNA copies that spool off it. Without these copies, the protein encoded by the gene cannot be produced. A cell鈥檚 RNAi machinery can be made to target a specific gene by adding 鈥渟mall interfering鈥� pieces of RNA, or siRNAs, that match part of the target gene鈥檚 sequence. The recent discovery that this works in mammals too has opened up an enormous range of possible uses in research and medicine.
One of the great attractions of the technique is its apparent specificity. Even a single change in an siRNA, which are about 20 鈥渓etters鈥� long, can prevent it silencing the target gene. But past experiments have not examined what effect siRNAs might have on other genes. Now two teams have done just that.
Advertisement
For their target, Patrick Brown鈥檚 team at Stanford University School of Medicine in California chose the green fluorescent protein gene from jellyfish, which they engineered into human cells. They attacked the gene with two siRNAs matching different sequences within the GFP gene. Then they used DNA chips to monitor the activity of 36,000 genes. Remarkably, they saw absolutely no sign of collateral damage (Proceedings of the National Academy of Sciences, DOI: 10.1073/pnas.1037853100).
If that was the end of the story, it would be champagne-popping time in RNAi labs everywhere. But last week Aimee Jackson and Steven Bartz at Rosetta Inpharmatics in Washington state published their results. They chose two human genes as their targets and hit one with eight different siRNAs and the other with 16. Given that turning off one human gene normally results in others being switched on or off too, they expected to see more than one gene affected.
But if blocking one specific gene triggered this domino effect, you would expect the siRNAs to affect the same group of genes every time. Instead, each different siRNA affected different and mostly non-overlapping sets of genes. This suggests that each siRNA hits a distinct group of innocent bystander genes as well as its intended target (Nature Biotechnology, DOI: 10.1038/nbt831).
The different results suggest some genes are hard to target without hitting bystander genes. Jackson says her team is already working on ways to minimise collateral damage. 鈥淲e know it is possible,鈥� comments Phillip Zamore of the University of Massachusetts Medical School in Worcester. 鈥淐ells have been using RNAi to control genes without harming themselves for millions of years.鈥� But developing siRNA-based drugs will be much harder if each siRNA has to be screened to ensure it does not knock out any vital genes in any of the hundreds of different human cell types. That champagne may have to stay on ice for some time.