RNA interference (RNAi), or post-transcriptional gene silencing, is proving to be a powerful method for reducing the expression of specific genes in many organisms — particularly in plants and invertebrates. If RNAi could be applied with equal facility to mammals, this could open the floodgates to studying gene function in cell lines and whole organisms, and could even give rise to therapeutic applications. Two recent papers report progress in both of these areas.

RNAi is a naturally occurring phenomenon and is mediated by double-stranded (ds) RNA. If the dsRNA matches the sequence of a transcript within a cell, it will direct the degradation of that transcript by endogenous nucleases. Any gene can therefore be downregulated by supplying dsRNA of the same sequence as the target transcript. This can be done by transfecting cells with dsRNA itself or with a construct that expresses a version of the target transcript containing an inverted repeat.

Although this method works extremely well in invertebrates, such as Caenorhabditis elegans, its use in mammalian cells has been less successful. Partly, this is because the presence of dsRNA in mammals activates a protein kinase, PKR, which is involved in a defensive response to viruses and leads to a generalized reduction of gene expression. Last year, it was shown that PKR can be circumvented by using shorter dsRNAs, of ∼21 nucleotides in length, which allowed the transient silencing of target genes in several mammalian cell lines.

Paddison et al. have extended this observation by showing that another way to get round PKR is to express an inhibitor of PKR function. Using this approach, they successfully carried out RNAi in somatic mouse cells, using longer dsRNA. In addition, they showed that stable silencing of target genes could be achieved in a mouse embryonal carcinoma cell line by expressing a transcript containing an inverted repeat. This work could pave the way towards alternative strategies for knocking out gene function in mice, as well as in somatic mouse cell lines.

The aim of the study by Caplen et al. was to see whether RNAi could be used to downregulate the expression of a known disease-causing gene. Specifically, they were studying the mutant form of the androgen receptor that causes spinal and bulbar muscular atrophy (SBMA). This disease — like Huntington disease and several other neurodegenerative disorders — is caused by the expansion of a sequence that encodes a stretch of glutamines. Caplen et al. showed that RNAi could be used to reduce the expression of the expanded repeat in a tissue culture model of SBMA in both Drosophila cells and human cells. In the human cells, RNAi also reduced the cytotoxic effects that are associated with the expression of these polyglutamine repeats.

Although these new studies extend the uses of RNAi in mammalian cells, some important technical issues remain to be addressed, such as cell-type specificity. For example, if we could understand why RNAi works better in some cell types than others, it might be possible to target RNAi to the cells that are affected by a given disease. Such a finding would herald some exciting therapeutic opportunities.