IT HAS been hailed as the new gene therapy, but the comparison may be getting a bit too close for comfort. The promising technique of RNA interference, or RNAi, now has its own safety scare: one variant of the method can cause fatal liver damage in mice.
Gene therapy aims to cure diseases by giving cells healthy copies of defective genes. However, it is difficult to deliver therapeutic genes to a sufficiently high number of cells, and the field has been beset by safety scares 鈥� including three cases of leukaemia among children treated for severe immune deficiency.
RNAi turns the approach on its head by shutting down the genes that are causing disease instead of adding new copies. Clinical trials are under way using RNAi to turn off a gene that causes blood vessels to grow across the retina, producing a form of blindness called macular degeneration. Meanwhile, a trial on healthy volunteers has tested an RNAi therapy designed to silence a key gene of the respiratory syncytial virus, which can cause pneumonia.
Advertisement
RNAi uses short stretches of RNA, just over 20 bases long, to bind to and block messenger RNAs 鈥� the intermediate molecules between genes and the proteins they code for. In the existing clinical trials, these 鈥渟mall interfering RNAs鈥�, or siRNAs, are given to patients by injection. However, it is hard to get siRNAs into cells this way, and they are soon broken down in the body, so it may be necessary to engineer viruses that can enter cells and pump out siRNAs over long periods.
One way this can be done is by making viruses produce longer RNA pieces that fold in two like a hairpin. These get snipped up into siRNAs by enzymes already inside the cell. It is this approach that has proved fatal. Mark Kay鈥檚 team at Stanford University in California has been studying mice treated with viruses engineered to produce a range of RNA hairpins. Many of the viruses caused severe liver damage, killing some mice (Nature, vol 441, p 537).
鈥淢any of the viruses that produced RNA hairpins caused severe liver damage, killing some of the mice鈥�
Kay believes the RNA hairpins disrupted essential gene regulatory mechanisms, which snip up naturally occurring 鈥渕icroRNA鈥� to produce an RNAi-like effect. The hairpins seemed to monopolise the cellular machinery needed to process microRNA.
The good news is that not all hairpins have this effect 鈥� 13 of the 49 Kay tested were safe. Even so, the results will make regulators cautious about approving clinical trials based on viruses that produce hairpin RNA. And Kay鈥檚 team has found no way of predicting which hairpins are dangerous. 鈥淧eople are just going to have to do toxicity studies,鈥� he says.
Kay鈥檚 findings also do not apply if siRNAs are used directly, stresses John Maraganore, president of Alnylam Pharmaceuticals in Cambridge, Massachusetts. Alnylam is developing the respiratory syncytial virus therapy and other RNAi treatments, and its studies suggest that siRNAs are safe. Other researchers have raised concerns, though, after finding that adding siRNAs to cell cultures can silence genes other than the intended target.
