Day 1 of Keystone RNAi Conference: New Small RNAs and Oral RNAi Delivery

More recently discovered types of small RNAs were the focus of the keynote session on the first night of the influential annual gathering of the RNAi community, the Keystone conference on RNAi, this year held in Whistler resort near Vancouver, Canada.

Following introductions by the scientific organizers Judy Lieberman (Harvard) and Phil Sharp (MIT), Craig Mello, Nobel laureate and faculty at the University of Massachusetts, gave an account of a variety of relatively recently discovered small RNA populations that are enriched in the worm germline and are characterized by specific structural features, such as 5’ triphosphorylations or 3’ methylations, and the Argonaute proteins (worms have over 20 of those, whereas humans only have 4) with which they are associated.

For many of these small RNAs, the biological function is still somewhat unclear and the field, as was also echoed by the second keynote address by Greg Hannon (Cold Spring Harbor) is busy using high-throughput sequencing techniques and bioinformatics to catalogue them. As an aside, with the speed sequencing technology is progressing, it could very well be that the future of microRNA diagnostics will entail the quantitative high-throughput sequencing of microRNAs rather than interrogating a limited number of defined microRNAs by microarray or PCR.

A particularly interesting small RNA population is one that is marked by 5’ triphosphorylation, unlike siRNAs that are 5’ monophosphorylated. It appears that in their biogenesis, a primary small RNA (siRNA or microRNA) cuts a target mRNA which then becomes the template for RNA-dependent RNA polymerases (RdRPs) that generate the 5’ triphosphorylated RNAs (aka secondary siRNAs). Surprisingly, the secondary small RNAs are the ones that carry out the majority of the gene silencing, while the primary small RNAs are present only in very small amounts and appear to function only in triggering the amplification of gene silencing.

It is, of course, now interesting to speculate whether synthetic small RNAs with various 5’ modifications could likewise function in human cells, possibly with different biological activities from siRNAs and microRNAs. However, in the case of the RdRP-dependent 5’ triphosphorylated small RNAs, one has to keep in mind that a dedicated RdRP appears to be absent in the human genome and that 5’ triphosphorylated RNAs in the cytoplasm may trigger unwanted cytokine responses. But maybe there will be some other type of modification around which one could build a therapeutic platform.

At the start of his presentation, Mello noted that one way to induce gene silencing in worms is by feeding them with bacteria expressing double-stranded RNAs (somewhat reminiscent of Cequent Pharmaceutical’s transkingdom RNAi approach). He finally came full circle at the end of his talk, when he showed some intriguing slides from colleagues at the University of Massachusetts, Michael Czech and Gary Ostroff, on the oral delivery of therapeutic RNAi. Starting with yeast ghosts consisting of essentially a shell of beta-glucans that they filled layer by layer with RNAs, including siRNAs, to create nanoparticles that would be taken up by the Peyer’s patches in the gut. Unfortunately this part of the presentation was quite brief, but it appears that they have succeeded in ameliorating inflammation in a mouse model by targeting TNF-alpha. It will now be important to demonstrate the generality of this phenomenon by targeting a number of other genes unrelated to immune responses. As this approach appears to target immune cells, it may at least initially be applicable for orally delivering RNAi for a number of immune related disorders.

This development could be particularly interesting for RXi given the close relationship between Mello and Czech and this new public company. Also taking into account that UMass RNAi efforts will be a major beneficiary of the Massachusetts life science initiative, it seems that RXi is finally showing some signs of scientific life.

RNAi by single-stranded RNAs

Before RNAi, there was antisense. Antisense strategies typically aim to suppress gene expression by either preventing translation initiation or degrading the messenger mRNA through an RNase H mechanism. As more mechanistic insight into the RNAi pathway was gained, it became clear that the final active, so called “guide” RNA that recognises and causes cleavage of the target RNA by RNAi is a single-stranded 19-23nt RNA bound to the catalytic RiSC endonuclease Argonaute 2 (in humans).

Sure enough, the provision of single-stranded RNAs was then shown by the Tuschl lab to be sufficient for inducing RNAi (Martinez et. al., 2002), although considerably higher concentrations were needed compared to the conventional double-stranded siRNAs in vitro. This work and potential applications thereof are the subject of a USPTO patent application by the Max Planck Society.

Nevertheless, ISIS Pharmaceuticals would claim that its antisense patents cover any therapy making use of a single-strand antisense mechanism, and in their interpretation this includes RNAi with single-stranded RNAs. ISIS’ leadership in nucleic acid technology such as nucleic acid modification and manufacturing technology is without doubt. This is also the reason why Alnylam, the leading RNAi company (of which Tuschl is a co-founder), has taken an exclusive license to a number of ISIS’ patents for the development of RNAi therapeutics, particularly for the use of modified siRNAs.

Although in tissue culture it is clear that double-stranded siRNAs are much more potent inducers of RNAi than single-stranded small RNAs, single-stranded RNAi molecules may have some utility for in vivo applications. In addition to the potential cost of goods advantage of a single-stranded versus a double-stranded approach, the delivery of single-strand RNA-like molecules to organs such as the liver may not even require as much as a formulation into a special delivery composition, as is suggested by the ability of single-stranded antisense compounds such as ISIS’ ApoB100-targeting 301012 to enter liver cells following subcutaneous injection. It is also possible, although not proven, that some of ISIS’ antisense compounds, particularly those with limited 5’ modifications, actually work through an RNAi mechanism.

The issue therefore relates to the value of a claim to a mechanism that was not even known to exist at the time when ISIS applied for patents relating to RNase-mediated gene suppression by an antisense mechanism involving ANY RNAse activity. While I think it is almost impossible that these claims will supersede the fundamental Fire-Mello and Tuschl II RNAi patents, they will probably carry some weight for the use of single-stranded RNAi molecules and ISIS is indeed pursuing such single-strand RNAi molecules for therapeutic purposes. Importantly, however, their utility in an in vivo setting still remains to be demonstrated.

The RNAi Patent Landscape

Patents are essential for drug development as they protect the significant investments that have to be made in order to obtain marketing approval from the relevant regulatory agencies. Common estimates put the cost of successfully developing a drug at around $1 billion. A whole technology platform such as RNAi would suffer and be unable to attract sufficient funding if the scientific discoverers and institutions failed to file patents and license them to organisations capable of taking RNAi to the clinic. Luckily, fundamental patents have been filed and have been either exclusively or non-exclusively licensed. What are they?

According to folklore, Fire and Mello, the discoverers of RNAi in worms, had to be “encouraged” by the NIH to spend the time to file a patent for RNAi. This resulted in the Fire and Mello patents that can be non-exclusively licensed by almost whoever wants to for a nominal fee. While their science was impeccable, commercially their patents suffer from the fact that originally the dsRNA inducer length was defined as 25bp and above. Additionally, even if efforts to bring this size down to the relevant 21 and upward size succeed, this will not give licensees automatically the freedom to operate as it does not claim the use as a therapeutic. For this, the Tuschl patent series are essential. These are based on Tuschl’s seminal and non-obvious discovery that small double-stranded RNAs (siRNAs) of 19-23 base-pair length are the mediators of RNAi gene silencing. It appears that Alnylam, of which Tuschl is a co-founder, can claim rights to most of the claims in the series, although the Tuschl I series can be claimed by Alnylam, Sirna Therapeutics (now a Merck subsidiary), and CytRX alike. This is because one of the four academic institutions involved in the licensing of Tuschl I decided to go it alone and license it to CytRX and and Sirna in addition to Alnylam. It seems strange to me that one institution alone can do this without the apparent support of the other parties, but I will leave this to the lawyers. The Kreutzer-Limmer patents, covering double-stranded RNAs for the purpose of gene inhibition, may also turn out to be an important piece of the puzzle and were acquired early on by Alnylam through its acquisition of Ribopharma AG.

There are a number of other pending and granted patents in the RNAi field, including siRNA manufacturing and modification patents that will be important in the actual drug development process. However, I view them as secondary albeit important and meritorious patents when measured on an innovation scale. Efforts are also being made to circumvent the need for prototype siRNAs by using either their biological precursors or even variants such as blunt-ended small double-stranded RNAs (apparently it works!), however their exact merit remains to be evaluated. In this context, I am encouraged by recent Supreme Court decisions that emphasise the importance of innovation and non-obviousness in a patent.

Note: The discussion did not cover DNA-directed RNAi that do not involve synthetic siRNAs.