Main

'Let a cell teach you how to do RNA interference most effectively', could have been the motto behind collaborative efforts by the groups of Greg Hannon and Scott Lowe at Cold Spring Harbor Laboratory and Stephen Elledge at Harvard University to generate a second generation of shRNA libraries. Their goal was to develop an effective screening tool to identify genes essential for the regulation of tumor growth; a previous shRNA expression library had given good results, but it suffered from sub-optimal efficiency of knockdown.

The scientists adopted a strategy, first introduced by Brian Cullen at Duke University, which takes advantage of the cell's endogenous RNA interference (RNAi) machinery to process and cleave an shRNA embedded in a larger microRNA fold, resulting in more stable expression of small interfering RNAs (Fig. 1). Hannon and Elledge created large libraries of microRNA-based shRNA vectors targeting many of the human and mouse genes, and observed more efficient knockdown than with the previous shRNA library (Silva et al., 2005).

Figure 1: Cells process a hairpin RNA (red) embedded in a microRNA fold like endogenous microRNAs.
figure 1

The RNAi machinery of the cell cleaves the microRNA (at the sites indicated with arrows) leading to a robust expression of small interfering RNAs.

Full size image

An added bonus of the new library is the ease of cloning that allows quick insertion of any shRNA into the microRNA fold, together with a promoter of choice. Although the Hannon-Elledge library is driven by an RNA polymerase III promoter, a promoter the cell uses for small noncoding RNAs, microRNAs—like mRNAs—are transcribed from RNA polymerase II promoters. These promoters yield abundant and stable transcripts and, most importantly, can be made inducible.

Intrigued by the possibility of an inducible shRNA library, the groups of Elledge and Lowe, working independently, cloned some microRNA-shRNA constructs under the control of a tetracycline-regulated RNA polymerase II promoter and showed tight control of gene knockdown in cultured cells even when only a single copy of the shRNA vector was integrated into the host genome (Stegmaier et al., 2005; Dickins et al., 2005). In addition, Ross Dickins in the Lowe lab demonstrated that the growth of tumors formed by cells infected with an oncogene and an inducible shRNA vector targeting p53, depended on the induction of the shRNA. Shutting the shRNA off caused re-expression of p53 and shrinkage of the tumors. Dickins sees an important application for such a system in the identification of therapeutic targets in cancer. He says: “Experiments like ours can provide a rational basis for targeting particular components of pathways because the short hairpin RNA can mimic a targeted therapeutic.” If the knockdown of a gene causes a tumor to shrink, a drug targeting this gene will most likely have the same effect.

Learning from a cell's own RNAi machinery how to best regulate gene expression may well turn out to be a powerful tool to better understand and fight tumor cells.