Pooling resources to improve siRNAs

Since RNA interference was first reported in plants by Richard Jorgensen in 1990, and Andrew Fire and Craig Mello reported on gene silencing using RNAi in C. elegans in 1998, it has become the technique of choice for exploring gene function and interactions.

While its use as a therapeutic tool in humans is very much in its infancy, there is no doubt that the next decade will see a revolution in its applications on a genome-wide scale.

In the meantime, a few little issues need to be sorted out. One of these is off-target events mediated by short interfering RNAs (siRNA), in which knocking down or silencing genes causes unintended consequences elsewhere and thus false phenotypes.

Dharmacon's breakthrough was to find the source of these off-target events, give the siRNAs a little chemical nudge and much improve outcomes.

Normal siRNAs have 19 nucleotides of exact complementarity. Similar to the short areas of microRNAs (miRNAs), known as seed regions, siRNAs have a region that is six to seven nucleotides long, which binds in the untranslated region of related messages.

They then co-ordinately regulate the untranslated regions by causing translational repression when sequestering the mRNA in P-bodies. What Dharmacon discovered is the reason why off-target events occur: the siRNA acts like an miRNA.

"Off-targets events seem to be using the endogenous miRNA pathway," Dharmacon's Dr Devin Leake says. "The siRNAs are added to the cells and they co-opt or hijack the endogenous miRNA pathway.

"As a result it's going to participate not only in cleavage - which is the effect that we like in siRNA - but it's also going to participate in this other regulatory pathway that involves binding to the 3 prime untranslated regions (3'-UTR) of the transcript.

"The complex pulls the transcript into P-bodies - bodies found in the cytoplasm which then take part in nonsense-mediated decay. In the P-bodies you see not only the mRNA down-regulation from a nuclease called Xrn1 but you also see translational attenuation - it blocks the capping mechanism of the transcript and it interferes with translation."

Chemical modification

Leake was part of the team that discovered that chemically modifying the siRNAs could reduce off-target silencing. The Dharmacon team added 2'-O-methyl ribosyl to a particular region of the guide strand, which subsequently silenced most off-target transcripts.

In fact, Leake's team has seen up to a 90 per cent reduction in off-target events, which he describes as quite amazing.

"Initially, when we were looking at off-targeting we thought like many researchers did at the time that some of the sequence alignment tools, the bioinformatics tools such as BLAST and Smith-Waterman, would be effective in identifying and potentially avoiding off-targeting," he says.

"In Nature Methods in 2006, we described an experiment where we clearly showed that that is not the case."

Dharmacon developed a method combining bioinformatics, the chemical modifications and pooling strategies to minimise off-target effects. "If you look at model systems such as C. elegans, Drosophila, Arabidopsis, you can see that they take long double-stranded RNA, process it by the Dicer enzyme, and create high-complexity pools.

"This would be ideal if you could do it in mammalian cells, but the problem is when you introduce long dsRNA into mammalian cells it elicits an interferon response. The opportunity to have high-complexity pools doesn't exist unless you create them.

"What we've done is when the siRNAs that are designed with the algorithm we can pool a set of four siRNAs so these develop a low-complexity pool that has the critical attributes of larger pools. Those attributes are reliably high potency and enhanced specificity."

Collaborations

The RNAi R&D team at Thermo Fisher is working with a number of researchers throughout the world who are using RNAi for genome-scale screening.

One of these is Kaylene Simpson, an Australian who was part of the well-known collaboration, including Mark Shackleton, Geoff Lindeman and Jane Visvader at the Walter and Eliza Hall Institute in Melbourne, that saw a complete, functioning breast grown from a single mouse stem cell.

Simpson is now working with Professor Joan Brugge in her lab at Harvard Medical School.

"[Simpson] is looking at wound healing and has developed a high-throughput method of analysing wound healing," Leake says. "She is screening through siRNAs looking for a specific gene target that affects wound healing. She's a very talented researcher."

Leake himself has a pretty strong background in RNAi, having worked with Tom Blumenthal, chairman of the department of biochemistry and molecular genetics at the University of Colorado's Health Sciences Centre, and discoverer of trans-splicing and polycistronic mRNAs in C. elegans.

Leake joined Dharmacon in 2002 and was enticed to the Colorado-based company by the current chief scientific officer, Anastasia Khvorova, also well-known in the RNAi world.

The next few years are going to be very interesting as the technology improves, Leake says. "Delivery is still an issue - we've seen some significant progress in localised applications or dosing of siRNAs in vivo.

There are some reports of systemic delivery but typically those are delivering to the liver or kidney - any kind of filtration tissue.

"I'm hoping we will see real, significant progress in therapeutics. That's the big one."