Dicer and double-stranded RNAs

Bio-Rad Laboratories' Steve Kulisch explains how synthetic 27-mer double-stranded RNAs can be designed to be processed by the Dicer endonuclease in a more predictable and efficient way.

RNA interference (RNAi) is an intrinsic cellular mechanism, conserved in most eukaryotes, that helps to regulate the expression of genes.

Researchers have exploited this natural mechanism by designing synthetic double-stranded RNAs (dsRNAs) for sequence-specific gene silencing to elucidate gene function. Such research has helped forge a rapid transition from discovery and research to potential therapeutic application.

Since it was first demonstrated that 19-23 nucleotide (nt) small interfering RNAs (siRNAs) are mediators of gene-specific silencing, design of siRNAs has sought to improve specificity and potency, which can reduce off-target effects.

While traditional synthetic siRNAs based on the 21- to 23-mer designs have been effective, increased understanding of the discrete events and enzymes involved in the RNAi pathway have recently led to significant improvements in the design of siRNAs, making them even more efficient tools for the induction, control and interpretation of gene silencing events in everyday research.

Researchers at City of Hope, a biomedical research and treatment centre in southern California, and Integrated DNA Technologies (IDT), the largest supplier of custom nucleic acids in the US, have developed of one such tool, Dicer-substrate siRNA, a highly potent mediator of RNAi.

RNAi overview

The RNAi pathway, part of a larger network that uses small RNA molecules as regulators of cellular signaling, relies on dsRNA as a trigger for sequence-specific gene silencing.

In this pathway, longer dsRNAs associate with Dicer endonuclease, a member of the RNase III family, which precisely cleaves the dsRNA into smaller functional siRNAs. These siRNAs then associate with an RNA-induced silencing complex (RISC), which targets any homologous mRNA for degradation.

It has recently been suggested that in addition to cleaving longer dsRNAs, Dicer endonuclease plays roles in loading processed dsRNA into RISC and in RISC assembly. This hypothesis has helped drive the development of a new class of siRNAs, termed Dicer-substrate siRNAs, that are highly potent mediators of gene-specific silencing.

While long (>30 nucleotide) dsRNAs have been used successfully to regulate gene expression in a number of eukaryotic organisms, they often activate intrinsic cellular immune responses that result in broad, nonspecific silencing when applied to mammalian systems. To prevent activation of these immune responses during RNAi experiments, researchers have generally used shorter (19-23 nt) dsRNAs.

More recent studies have demonstrated that dsRNAs 25-30 nt in length are even more powerful effectors of gene-specific silencing than 21-mers. Specifically, 25- to 30-mers are up to a hundred-fold more potent than 21-mer siRNAs targeting the same sequence.

This greater potency appears to depend on processing of the longer dsRNAs by Dicer, which cleaves the longer dsRNAs to produce 21-mers. When 27-mer siRNAs were selectively labeled with 6-carboxyfluorescein (6-FAM) to reduce cleavage by Dicer, a corresponding decrease in potency of siRNA was observed.

However, Dicer cleavage of 27-mers into specific 21-mers does not, by itself, explain the greater potency of the 27-mers. Since specific cleavage by Dicer is not sufficient to explain the increased potency, it has been hypothesized that providing Dicer with a substrate for cleavage (i.e., a 27-mer) improves the efficiency of the secondary role of Dicer - that of introducing siRNAs into RISC - and that this is responsible for the enhanced silencing by 27-mers.

Asymmetric design

Subsequent work on Dicer-substrate siRNAs has sought to improve their design to further increase the efficacy of silencing. General rules of siRNA duplex design, such as length, sequence preference and target accessibility, play a role in the relative potency of any given siRNA.

In addition to these rules, Dicer-substrate siRNAs can be designed to promote Dicer cleavage at a specific position to produce the most potent 21-mer product. Additional subtle design features can influence the dynamics of strand incorporation into RISC, promoting selective retention of the guide strand (the antisense strand, which is complementary to the target message) and leading to a significant impact on the performance of siRNAs in vitro.

By using electrospray ionisation mass spectrometry (ESI-MS) to analyse the Dicer products derived from a variety of 27-mers, researchers at IDT and City of Hope were able to identify structural features that encourage the production of a single, predictable, maximally active product.

Specifically, an asymmetric design that includes a two-base 3' overhang on one strand and the addition of two DNA residues to the 3' end of the other strand severely limits heterogeneity of the cleaved siRNA product - the blunt end is unfavourable for Dicer binding, and cleavage preferentially occurs 21-22 bases from the overhang.

Maximum potency is obtained when the two-base overhang is present on the antisense strand while the DNA bases are added to the sense strand. In this case, Dicer binds to the 5' end of the antisense strand, leading to preferential retention of this strand in RISC. These design features, which are incorporated into Bio-Rad's siLentMer Dicer-substrate siRNAs, ensure maximum potency in RNAi.

Benefits

Effective use of RNAi for research and therapeutics requires that non-specific effects be minimised. Non-specific effects may be related to sequence homology of an untargeted mRNA or activation of cellular responses - particularly induction of the interferon response, which can result in global translational arrest.

In addition, some data suggest that the RNAi machinery can be saturated, inhibiting the proper processing of precursors of microRNA (miRNA), leading to toxicity.

Non-specific effects and toxicity can be limited by using low concentrations of siRNA. For example, while full activation of the interferon pathway can be avoided by the modest use of siRNAs <30 nt in length, pro-inflammatory responses can be activated by higher concentrations of siRNA.

Dicer-substrate siRNAs, when used with an effective transfection reagent and protocol, are potent at concentrations as low as 100 pM, which minimises the potential for off-target effects.

To further reduce the potential for activating immune responses, Dicer-substrate siRNAs can be specifically designed to prevent the activation of pro-inflammatory cytokines (IFN-alpha and IFN-beta and the protein kinase R (PKR) pathway.

An additional benefit of Dicer-substrate siRNAs is longevity of silencing. When NIH 3T3 cells stably expressing enhanced Green Fluorescent Protein (eGFP) were transfected with 21-mers or 27-mers targeting the same site, eGFP suppression by the 21-mer lasted for about four days, while suppression by the 27-mer lasted up to 10 days.

While these results match observations made in other studies, some 21-mer siRNAs, described as hyperfunctional siRNAs, can produce comparable long-term silencing. Still, controlling the processing of 27-mers by Dicer to ensure the production of a single specific siRNA can allow lasting and consistently potent silencing at low concentrations and reduce the chance of off-target silencing events.

Steve Kulisch is marketing manager for gene transfer at Bio-Rad Laboratories in the US. References are available on request.