Technology Trends: MicroRNA Inhibitors and Single-Strand RNAi

There have been developments in the areas of microRNA inhibition and single-strand RNA-mediated RNAi that might have strategic implications for delivery technologies and RNAi Therapeutics, respectively.

MicroRNA inhibition: naked antisense no more?

Currently, all development-stage anti-miR programs to my knowledge envisage the use of unformulated phosphorothioated antisense molecules with various high-affinity modifications such as LNA/LNA-type conformationally restricted nucleotides or 2’F and 2’MOE. To some degree, antisense and certain microRNA companies are making a living out of advertising that, unlike (most) RNAi Therapeutics, no intravenous administration was required.

At the same time, it is becoming clear that more complex structures such as Dharmacon’s miRIDIAN hairpin microRNA inhibitors or the tough decoys (also the synthetic versions that were newly developed in collaboration with Japanese RNAi behemoth Kyowa Hakko: Haraguchi et al. 2012) are considerably more potent on a per molecule basis. Because of their structural complexity, however, they would require delivery formulations for therapeutic use. It remains to be seen how often such formulations would need to be applied, but the early research by Haraguchi in tissue culture shows that the anti-miR effect with these structures can be relatively long-lived. Nevertheless, the in vivo pharmacology of these structured anti-miRs remains to be better explored, but I could imagine that especially for antiviral or oncology applications, the more rapid onset of action and the potentially improved targeting due to the delivery technology could yield positive surprises.

Single-strand RNAi Therapeutics: Stable 5’ phosphate and 2'F

A little more than a year after ISIS and Alnylam ended their ssRNAi Therapeutics collaboration (for which I believe Alnylam had greatly overpaid), ISIS and Merck have made progress in the area.

It had been well known based on particularly protein structural work that the 5’ phosphate modification in the guide strand is important for incorporation in the RNAi effector complex RISC. There has also been corresponding early evidence in ssRNAi research (Martinez et al 2002) that ssRNAs with a 5’ phosphate are more efficient inducers of RNAi, albeit at much lower efficicay compared to dsRNAi triggers. Notably, in the case of double-strand RNA-induced RNAi, prior 5’ phosphorylation is not necessary as this is efficiently accomplished inside the cells.

Based on work by ISIS presented at the Keystone conference in January it now seems that lability of the 5’ phosphate is partly responsible for the reduced efficacy of ssRNAi. Using a new 5’ phosphate mimic (and some other lipophilic modification strategies) that is much more stable than the natural counterpart, it was now possible, following subcutaneous administration, to achieve solid knockdown in rodents. However, the cumulative dose required to get there is still very high, in the range of 2^nd generation RNaseH antisense.

ssRNAi work just published by Merck (Haringsma et al. 2012) confirmed that the 5’ phosphate modification was important, but more importantly worked out the beneficial role of the sticky 2’-fluoro (2’F) modification in ssRNAi. The impressive efficacy-enhancing activity of the 2’-F modification (for which, I believe, Alnylam holds rights to an important patent via a license from ISIS, but a modification that has also raised genotoxicity concerns) was found to apply to both tissue culture and animal settings [Note: an earlier version mistakenly stated that the IP belonged to Alnylam]. Unlike the ISIS work, however, Merck studied ssRNAi in mice using SNALP-like liposomal delivery. It therefore seems as if Merck would need to extra work on adding chemistries such as phosphorothioates or conjugates so that its ssRNAi technology can be used in the ‘naked’ form. This is because the prospect of using ssRNAi in the naked versus formulated form would really be the only motivation for pursuing ssRNAi.