Targeted Delivery Strategies Coming to the Fore

Hand in hand with a rapidly expanding understanding of the biological mechanisms of RNAi Therapeutics delivery, we are hearing more and more about ligand-guided targeted delivery. Just today, mdRNA announced notice of allowance for a patent application on the identification of a peptide specifically binding to the cell surface of hepatocellular carcinoma cells, and Alnylam has recently started to talk about their discoveries on ApoE-dependent and Apo-E independent SNALP cellular uptake pathways paving the way towards new targeted SNALP delivery strategies (Systemic RNAi Delivery Roundtable). As it increasingly looks like Calando’s transferrin-targeted CALAA-01 will not remain the only targeted RNAi Therapeutics formulation in the clinic for long as targeted delivery is set to provide the next push towards more potent and safer RNAi Therapeutics delivery, I will try and briefly explain the rationale behind targeted delivery and some of the challenges that need to be overcome.

Success in targeted delivery is measured by either preferential uptake of the siRNA in the target tissue compared to non-target tissues or achieving lower efficacious dosages by taking advantage of particularly productive receptor-mediated uptake pathways, preferably both. The detailed pharmacological effects are not only determined by the ligand, but also by where it is attached to. Steps that may be affected can include biodistribution, cellular uptake once at the target tissue, or the avoidance of certain cells and tissues.

SNALP delivery for example may benefit from targeting ligands by reducing uptake by macrophages or relying on non-specific charge-charge interactions for cellular uptake, both of which can be safety liabilities. For the liver, de-targeting from macrophages may be even more important than active targeting. The DPCs by Mirus (now Roche) were particularly exciting here in that the data suggested that the right presentation of ligands (simple sugars in this case) on the particle surface may eliminate the unspecific uptake of a delivery platform and instead re-target it to new cell types. Equally exciting data by Alnylam suggests that it should be possible to greatly limit the ‘non-specific’, mainly ApoE-mediated uptake of ionizable SNALPs by shielding their surface and then re-direct them by adding new ligands on their surface. By then further increasing their circulation times through creating very stable particles (e.g. by increasing the stability of the stealth shield), a delivery platform may then also be applicable to new therapeutic application fields by increasing the chances that a ligand recognizes receptors in distal tissues.

Lipoplexes such as Silence Therapeutics’ Atuplexes may also benefit from targeting ligands. Since the interaction of immune cells with blood endothelia is very well studied this may e.g. allow it to be targeted to specific endothelia such as the blood-brain-barrier.

Targeting ligands are already part of many siRNA-conjugate approaches. Achieving endosomal release in addition to cellular uptake is a big challenge for this area of delivery, and it will be interesting to see whether in fact those receptors that prove effective for nanoparticle delivery may be the types of receptors to be avoided for siRNA-conjugates in favor of channeling them into more non-specific pathways.

There are, of course, also challenges associated with targeted delivery. One is to identify suitable ligand-receptor interactions as endosomal maturation processes can differ greatly, e.g. in the degree and rate of acidification and receptor recycling, which imposes new types of pharmacokinetic demands on a delivery system.

A systematic effort to discover the best receptors may be to screen a panel of siRNA-nanoparticles containing (single-chain/nanobody-type) antibodies on their surface and that are targeted to a wide array of cell surface receptors and then select those with the best silencing results. The most promising receptors, hopefully patentable, may be pursued then either with the antibodies themselves or alternative, smaller ligands. New ligands to given receptors may be discovered through panning peptide display libraries against that receptor, something e.g. that mdRNA does with their trp-cage peptide libraries.

Avoiding adaptive immunity, especially to novel designer ligands is another added challenge for targeted delivery. And finally, when all these questions have been answered, the not-so-trivial task is to find formulation methods that allow for clinical and commercial scale-up of the more complex particles. At the end of the day, however, it is those platforms for which a detailed mechanistic basis has been established and those teams that have turned formulation into an art that will succeed. It can be done.

Money from the Sidelines Moving into RNAi Therapeutics Again

The recent signs that the appetite for RNAi Therapeutics by Big Pharma is returning coupled with accumulating evidence that RNAi can be triggered in Man with some of today’s delivery technologies, is finally translating into investors buying into the promise of RNAi Therapeutics again. Today’s almost doubling in the shares of Arrowhead Research (ticker: ARWR) on volume of over 40% of outstanding on the heels of a Nature publication showing that their RONDEL delivery system was able to induce an RNAi mechanism of action in solid tumor tissues in real patients, is quite impressive evidence for this. The incredible performance of RXi Pharmaceuticals (ticker: RXII) which has made much out of their ‘self-delivering siRNAs’ and who are also positioning themselves as an RNAi trigger alternative to Alnylam another one.

If the trend were to indeed continue, this time around I would expect much of the action to be in the limited RNAi delivery technologies that have made it into the clinic. Except for the $125M acquisition of Mirus Bio by Roche, I have long felt that the RNAi Therapeutics marketplace and investment community has never really reflected the importance of delivery in making RNAi Therapeutics a reality. In addition to being the critically enabling factor for RNAi Therapeutics, such IP should also be the most valuable kind.

The argument for delivery-related investments is based on the expectation that any Big Pharma that is seriously considering RNAi Therapeutics would first want to secure access to technologies that have shown potential as clinically viable platforms. This would allow them to gain familiarity with RNAi Therapeutics and, by being on its cutting edge, assemble know-how and lay down further IP to help position them as leaders in the field, an opportunity that was missed in the case of recombinant proteins and monoclonal antibodies.

A scarcity of technologies that have made it into the clinic providing some evidence of RNAi efficacy in the absence of show-stopping toxicities should further increase their value. At the moment these are SNALPs, RONDEL, Atuplexes, a lentivirus and tkRNAi bacteria. You will not be surprised to hear that based on the strong and abundant pre-clinical, including non-human primate data, highest quality science, modularity as a platform, well-defined structures and scalable manufacturing, and the proven ability to assist partners in bringing their RNAi triggers into the clinic, I am particularly fond of SNALPs and Tekmira. I cannot believe that companies like Novartis or Pfizer do not consider it a risk that access to Tekmira could be lost due to an outright acquisition of the company.

In times of cost-cutting, it is important for companies that want to partner or sell their technologies to keep the necessary talent to make technology transfer possible. Such a group of people would also enhance the attraction to acquire a company as an important pillar of the broader RNAi Therapeutics efforts of a Big Pharma, similar to what Alnylam Europe was for Roche or Coley for Pfizer. Losing such capabilities could quickly render a technology stale as Targeted Genetics seems to be running the risk.

All the ingredients of a small acquisition wave of RNAi Therapeutics technologies and companies are therefore in place: a limited supply of enabling technologies and scientific talent, capital markets that make it difficult, if not impossible for small RNAi Therapeutics companies to realize the full value of these technologies on their own, and big pharmaceuticals that will consider the recent results by Tekmira and Calando/Arrowhead as important de-risking events and feel the urgency to act now.

CALAA-01 Results Indicate RNAi Therapeutics to Revolutionize Solid Cancer Treatment

In yet another important proof point that RNAi Therapeutics is slowly but surely morphing into a therapeutic reality, Calando just reported in the highly prestigious journal Nature an interim analysis of their phase I clinical study of CALAA-01 demonstrating an RNAi mechanism of action in solid cancer patients. The results are the most direct demonstration of RNAi in Man thus far, and, at least equally significant, show that the Enhanced Permeability and Retention (EPR) effect is indeed about to turn into the Achilles' Heel of solid cancers in the era of nanoparticle siRNA delivery.

Solid cancers are considered by many the most important near-term value driver for RNAi Therapeutics. The fact that Calando’s RONDEL nanoparticle delivery technology has now been shown to accumulate in solid cancer tissues is not only an important de-risking event for CALAA-01 itself, but it immediately suggest that RONDEL, and also quite likely other nanoparticle-siRNA delivery technologies, can facilitate the targeting of essentially any gene in solid cancers. It is this flexibility of the platform that makes RNAi Therapeutics ideally suited to satisfy the demands of a genetically such heterogeneous disease group as cancer.

Nanoparticle-siRNA delivery such as RONDEL is currently the most promising approach for treating solid cancers with RNAi Therapeutics. All of these rely on the particles being able to circulate sufficiently long in the blood stream and being small enough so that they can passively accumulate through EPR in the solid tumors as these lack an effective way of draining them through lymphatics. While the absolute amount of RONDEL particles that accumulated in the tumors remains to be determined, the Nature study clearly showed that a decent amount of them does accumulate there at ~0.6mg/kg.

While particle stability and ability to accumulate through EPR is a critical first step where we should also expect to see important differentiation between the technologies, once there, the nanoparticles then have to be taken up by the cancer cells and moreover mediate the cytoplasmic release of the RNAi trigger into the cytoplasm. Once again, RONDEL meets this challenge as it was possible to detect cleaved target mRNA characteristic of an RNAi mechanism of action. Although the so called RACE assay used for this analysis is not quantitative, in my experience, you have to get sufficiently robust RNAi to cleanly detect such products. Consistent with this notion, well over 50% target knockdown was detected for CALAA-01, although the limited number of samples (3) and the reliance on often only historical negative controls leaves open the question of the exact knockdown potency of CALAA-01.

It is in fact the pioneering work with RONDEL-siRNA delivery from which this pharmacokinetic model has been derived and that is now informing quite a bit our thinking of how to design siRNA-nanoparticles for solid cancer applications. For example, it was work on CALAA-01 that showed that the utility of the (transferrin) targeting ligand is in enhancing the cellular uptake of the nanoparticles, not so much in concentrating them in the tumors in the first place, and should be influential in the selection of targeting ligands being contemplated now for all types of delivery technologies, including SNALPs.


Summary of CALAA-01 results


CALAA-01 was also the first serious RNAi Therapeutics candidate for solid cancers that had entered the clinic in 2008. For a more thorough molecular background on this program, please refer to a previous entry (here) that was part of a 3-part collaborative series with Tobias Wolfram on the first RNAi Therapeutics solid cancer candidates. Briefly, CALAA-01 is a self-assembled nanoparticle system based on a cationic co-polymer containing cyclodextrin complexed with siRNA targeting RRM2, a gene involved in DNA replication, and that like most of the advanced nanoparticle-siRNA systems is stabilized through pegylation (some of these carry the transferrin targeting ligand). While Tobias and I had cautioned about a relative lack of preclinical in vivo characterization of CALAA-01, particularly not having directly demonstrated that CALAA-01 can knock down genes via RNAi in solid tumors following intravenous administration, it seems that the gamble of skipping some of the pre-clinical proof-of-concept has paid off for the investigators and Arrowhead Research (of which Calando is a subsidiary).

It was also a clever move that, amongst other tumor types, a number of melanoma patients were enrolled as such tumors can be easily accessed through biopsy and thus represent good material for molecular analysis of a clinical candidate. 3 of these volunteered to be biopsied. Despite the limited sample size, a number of conclusions can be drawn from the data: 1) RONDEL-siRNAs accumulated in the tumor tissues in a dose-dependent manner; 2) consistent with this, in the patient that received the highest dose, an RNAi mechanism of action was unambiguously confirmed by 5’ RACE; 3) this effect was long-lasting (>4 weeks) and accompanied by mRNA knockdown and disease stabilization; 4) RONDEL was safe and well tolerated in a study that involved repeat administrations and dose-limited toxicity has yet to be reached (right now at ~0.6mg/kg with dose escalation ongoing).

Clearly, these are highly encouraging results and call for the continuation of the phase I studies and further investigation of the modular RONDEL delivery platform. Importantly for shareholders, it should put Arrowhead in a better position to fulfill their stated goal of monetizing RONDEL either through licensing or outright sale. Part of the value would derive from the pioneering position of the RONDEL technology and from representing a complementary approach to lipid-based delivery which is currently dominating RNAi Therapeutics development programs.

Implications beyond CALAA-01 and RONDEL

Taking a step back, the CALAA-01 results add to the increasing amount of clinical data that first of all lay to rest concerns that there is a prohibitive fundamental (acute) toxicity issue for RNAi Therapeutics per se, and secondly provide evidence that therapeutic reality is within the reach of current systemic RNAi delivery technologies. They come on the heels of SNALP-ApoB results that showed preliminary signs of RNAi efficacy at a similar dose level while ongoing improvements in SNALP efficacy continue to widen its therapeutic index, and continuing dose escalations for both Alnylam’s VSP-02 and Silence Therapeutics’ Atu-027 solid cancer programs which can be interpreted as confirming the tolerability of some of the important delivery technologies in the space (SNALP and Atuplex).

With such progress, urgency to invest in RNAi Therapeutics, particularly in enabling delivery technologies should return. As the transition into the clinic is fraught with uncertainties for new chemistries (safety, efficacy, and manufacturing scale-up), it is particularly the less than a handful of delivery technologies that have made it there for which a land-grab might erupt as is also indicated by the fact that Tekmira now counts 8 RNAi-related pharmaceutical collaborators (both early and advanced stage), mdRNA at least 3 such early-stage collaborators, and a rumor-driven rise in the shares of RXi Pharmaceuticals possibly due to their mysterious claims on ‘self-delivering’ siRNAs. With the publication of the results that are about to garner widespread attention, Arrowhead Research/Calando should, of course, also get a few enquiries for RONDEL as it resumes its search for partners. Considering the generally weak share prices of RNAi Therapeutics companies, all the ingredients seem to be in place to see related deal activities, with a significant portion of the value being driven by the solid cancer opportunity for RNAi Therapeutics.

Update May 29, 2010: More details about the study, including the safety profile was published recently in an abstract for the upcoming 2010 ASCO cancer conference:

Systemic delivery of siRNA via targeted nanoparticles in patients with cancer: Results from a first-in-class phase I clinical trial

Author(s): A. Ribas and colleagues

Background: Systemically delivered small interfering RNA (siRNA) would allow targeting oncogenic molecules beyond current approaches. We report on the first siRNA trial with a targeted nanoparticle delivery system. Methods: Open-label, dose- escalation trial in pts with solid refractory cancers receiving 4 i.v. infusions (30 min) on d 1, 3, 8, and 10 of 21-d cycles. CALAA-01 nanoparticles consisted of a cyclodextrin-based polymer, transferrin protein (hTf) targeting ligand, polyethylene glycol (PEG) for stability, and siRNA against ribonucleotide reductase M2 (RRM2). The 70 nm particles were designed to minimize renal clearance and allow tumor vasculature permeation with binding to tumor hTf receptors (TfR). Primary endpoints (safety, MTD determination) were based on the first cycle. Results: 15 pts accrued to 5 dose levels (3, 9, 18, 24, 30 mg/m²), median age 62 (range 53-85). Most common histologies: GI (4), melanoma (3). Dose escalation progressed with no DLTs. Most common treatment-related AEs: fatigue (47%), fever/chills (33%), allergic (33%), constipation (33%), nausea/vomiting (20%), all g1-2. 1 pt had g3 anemia and 2 pts had g2 thrombocytopenia, all shortly after CALAA-01 infusions with rapid recovery. 1 pt had possibly-related sinus bradycardia (g2). No objective tumor responses were seen; 1 pt at highest dose had stable metastatic melanoma for 4 mo, a change from prior course. Biopsies in 3 pts (melanoma) showed particles in tumors. At 30 mg/m² RRM2 knockdown (mRNA and protein) was seen with confirmation of mechanism by specific cleavage sequence (RACE-PCR). TfR was not downregulated in cancer cells.Conclusions: Systemic delivery of siRNA via targeted nanoparticles is safe and can induce specific, siRNA-mediated gene silencing. This approach could be expanded to any currently undruggable cancer therapy target.

The Pfizer-Tekmira Deal: More than Meets the Eye?

In yet another sign that interest in RNAi Therapeutics Big Pharma is picking up again, Tekmira today announced the initiation of a research collaboration with Pfizer focused on its SNALP siRNA delivery technology. This trend follows a number of proof-of-concept studies over the last year not only in non-human primates, but increasingly also in Man, which in aggregate are allaying much of the concerns caused by the findings that non-specific, innate immune responses had been responsible for a number of early RNAi in vivo study results (see study and review by Tekmira scientists).

In Tekmira, Pfizer is certainly choosing a leader in the development of RNAi into a clinical reality, and despite the early stage of this relationship, this could harbor the seeds of much more to come. This is not only because the two research teams should be a good cultural fit, but also because Tekmira could become a central piece in Pfizer’s strategy of RNAi Therapeutics as a platform. Sure, Pfizer has been active with a number of deals in the space, including an eye disease collaboration with Quark and a DNA-directed RNAi approach for HCV with Tacere, but following its acquisition of Coley as the launching pad for RNAi Therapeutics, it has yet to decide on where it will get the fundamental IP from in terms of RNAi triggers and also has not committed yet to any particular delivery technology. Given this delay, I have therefore come to believe that Pfizer may pursue a strategy that includes a concerted move in both trigger and delivery.

It is no secret that there is a certain tension in what Tekmira feels it deserves from Alnylam, and what Alnylam is willing to pay. Clearly, SNALP delivery must have accounted for a significant part of the financials Alnylam achieved in the Roche and Takeda platform alliances. One important corporate development goal of Tekmira is therefore to create enough know-how and IP that Alnylam does not control so that Alnylam cannot take it any more for granted, and with more options create best shareholder value. Pfizer could be Tekmira’s white knight because it is also fair to assume, as evidenced by the persistent patent oppositions by Pfizer against Alnylam especially in Europe, that Pfizer would prefer not to pay Alnylam $300M upfront for a platform license. What better way to gain leverage over Alnylam than by building a relationship with Tekmira on whose technology Alnylam has build so much of its pipeline and business development?

The main reason why Tekmira does not sport an RXi-like $100M market cap, despite vastly superior enabling technology and financials, is because it does not claim to have proprietary RNAi triggers. One could therefore argue that if somebody like Pfizer decided that it did not need Alnylam’s RNAi trigger IP, combining Tekmira’s SNALP technology with an RNAi trigger workaround solution (e.g. blunt-ended siRNA’s depending on the Tuschl outcome) would create synergies that would easily justify a Mirus-type price tag for Tekmira. Given the recent impressive share price performance of RXII, I would not want to exclude that RXi Pharmaceuticals could feature in this equation.

One way of interpreting today’s news is therefore as yet another important validation of SNALP technology by a Big Pharma, all the more important since with Pfizer SNALP’s potential technology risks have passed the most critical smell test, innate immune activation and liver toxicity (Coley and Tekmira’s precursor Inex used to compete on TLR therapeutics), but with modest immediate financial implications. Behind hit, however, could be the beginning of the end of Tekmira as we know it. With the platform adoption option date coming up, Novartis may also want to have a say in this.

MicroRNAs Moonlighting as Molecular Decoys

A reader of this blog asked about the impact on our understanding of microRNA biology of a paper that was just published in Cell and shows microRNA-328 to function both as a traditional RNA silencing agent and molecular decoy for an RNA binding protein (Eiring et al. [2010]. miR-328 functions as an RNA decoy to modulate hnRNP E2 regulation of mRNA translation in leukemic blasts). The paper demonstrates that miR-328 binds tightly to an RNA binding protein, hnRNP E2, so that it is no more available to act as a suppressor of C/EBPalpha translation through the recognition of a miR328-related sequence element in the C/EBPalpha mRNA. This is a critical event for triggering the blast crisis stage of chronic myelogenous leukemia. The finding that a microRNA can act as a competitive inhibitor of sequence-specific regulators of RNA processing, in addition to its well known function in RNA silencing, not only necessitates a re-evaluation of the biological network that microRNAs operate in, but raises the prospect for new (small) RNA Therapeutics approaches.

Discovery of miR-328 as a competitor for hnRNP E2 binding

HnRNP E2, a protein that belongs to a class of proteins originally described to be highly abundan proteins in the nucleus that bind to RNA, has been known repress the translation of C/EBPalpha by binding to a C-rich element of that mRNA. Since C/EBPalpha is critical for myeloid differentiation of blood stem cells, hnRNP E2 ensures that a proper ratio of precursor to differentiated cells is maintained. A remaining puzzle in the understanding of this regulation, however, was the observation that myeloid precursor cell proliferation could still occur in certain situations when hnRNP E2 levels are high without evidence for a post-translational modification of hnRNP E2 that might account for its loss of C/EBPalpha mRNA-repression activity.

Since small non-coding RNAs, particularly microRNAs, are implicated in cell proliferation and differentiation as well as in related disease such as cancer, the authors of the paper hypothesized that it is a microRNA that negatively modulated hnRNP E2-mediated translation repression activity by competitively binding to it, and that the loss of such regulation is responsible for the differentiation deficit in the blast crisis stage of chronic myelogenous leukemia (as opposed to the more indolent chronic phase of the disease). I have to admit that postulating a microRNA decoy activity would have seemed like a risky move at this point, but by following through with microRNA expression profiling, they indeed were able to come up with a microRNA, miR-328, that fulfilled the criteria for such an activity: its diminished abundance/loss during blast crisis CML and a C-rich element akin to the one found in C/EBPalpha mRNA.

Sure enough, a battery of binding experiments demonstrated that miR-328 was an efficient competitor for hnRNP E2 binding and that this freed C/EBPalpha mRNA which was associated with its increased translation into protein and myeloid differentiation. The reason why miR-328 was essentially lost in (blast crisis) CML was due to over-active signaling downstream of the BCR-ABL mutant protein which is characteristic for CML.

Therapeutic RBP Decoys?

The finding that a microRNA may act as a sequence-specific molecular decoy for an RNA-binding protein raises the possibility that microRNA decoy mimicry or inhibition strategies could be therapeutic. While much more work remains to be done to validate miR-328 as a good candidate for microRNA decoy mimicry in CML by restoring myeloid differentation, it will be important to find out whether other microRNAs live similar double-lives. Though not trivial, a careful analysis of the direct interactions of microRNAs with other RNA binding proteins (RBPs) might yield additional candidates. Such therapeutic approaches would not even have to be limited to microRNAs but could encompass all kinds of natural and synthetic RNA-binding protein decoys. In fact, an RNA decoy mimicking the TAR element of HIV RNA is part of the triple (shRNA, ribozyme, decoy)-agent developed by the City of Hope and Benitec, and DNA transcription factor decoy approaches have also been conceived. Of course, such therapeutic approaches could piggy-back on many of the advances already made for other RNA therapeutic modalities before, including delivery and nucleic acid chemistries. This also means various start-up and business development opportunities for existing companies (e.g. out-licensing of established delivery technologies).

Beyond that, however, I would caution that such decoys should be much harder to develop as a platform since each RBP-RNA relationship would have to be looked at in detail in order to understand for example which exact sequence and structural motif is involved and where in the cell one has to intervene. In addition, although some would argue that it is actually a potential advantage, since most RBPs bind to multiple RNAs, a therapeutic approach has to take into account the complex systems effect that it could trigger, somewhat reminiscent of microRNA Therapeutics. By contrast, once the genetic relationship is understood, the goal for RNAi Therapeutics is always the same: get the RNAi trigger into the cytoplasm of the target cell to surgically remove one gene at a time. It is this basic simplicity that has always been at the bottom of my fascination for RNAi Therapeutics.

While Alnylam Focuses Suit on Whitehead and UMass, a New Tuschl Loophole Approach Gains Traction

Not a day goes by in RNAi Therapeutics land without hearing the sounds from the RNAi trigger IP battle grounds. Today is no exception…

Wading through the court documents from the Alnylam-Max Planck suit on how the Whitehead prosecuted the Tuschl I (T-I) patent application (aka the 'Tuschl Tussle'), it struck me that while it was Whitehead, their hired patent counsel, and UMass that apparently conceived of the strategy of how to incorporate data from the T-II patent into T-I against the interests of Alnylam and Max Planck, the MIT seemed to merely go along unwittingly. As a result, MIT not only became a victim in that they were deprived of the benefits from the therapeutic agreement they were part of with Max Planck and the Whitehead, which to my knowledge anticipated equal sharing of the profits from therapeutic licenses derived from the combined T-I and T-II estate, but also because they ended up as defendants in the suit.

Perhaps realizing this, Alnylam announced today that it will leave the MIT off the hook in return that they will be bound by any ruling in favor of Alnylam/Max Planck. It could also help drive a wedge between the former co-defendants as the MIT will now feel less fearful about telling their side of the story which could turn out to be quite revelatory. The MIT should have every reason to be unhappy with the Whitehead and the way T-I was handled. The argument that Whitehead, and implicitly, the MIT, aided UMass for political reasons, thereby leaving therapeutic licensing money on the table, never really flew with me since a) institutions just don’t give away money like this, and b) I could never imagine a research institution of the stature of Whitehead as being nationalistic and therefore anti-Max Planck.

While the Tuschl Tussle is going on, there is another Tuschl loophole movement in the field that appears to be gaining some traction. It initially started with mdRNA’s claims that the mere inclusion of a single ‘funny-looking’ nucleotide into an siRNA, unlocked nucleic acids (UNAs), would help it get around the Tuschls. It long cited evidence by outside patent counsel that supposedly supported their belief that they had freedom-to-operate in the RNAi trigger space. The news that Quark Pharmaceuticals had just initiated dosing for their 5^th (!) clinical RNAi program for a neuroprotective agent of the eye, it reminded me of their similar claims about ‘proprietary siRNAs’. To find out about the nature of these claims, I looked up what potentially applicable patent applications they had filed, and came up with the following main claim:

"1. A compound having structure (IX) set forth below:

(IX) 5' (N)x - Z 3' (antisense strand)

3' Z'-(N')y- z" 5' (sense strand) wherein each of N and N' is a ribonucleotide which may be unmodified or modified, or an unconventional moiety; wherein each of (N)x and (N')y is an oligonucleotide in which each consecutive N or N' is joined to the next N or N' by a covalent bond; wherein Z and Z' may be present or absent, but if present is independently 1-5 consecutive nucleotides covalently attached at the 3' terminus of the strand in which it is present; wherein z" may be present or absent, but if present is a capping moiety covalently attached at the 5' terminus of (N')y; wherein x =18 to 27; wherein y =18 to 27; wherein (N)x comprises modified and unmodified ribonucleotides, each modified ribonucleotide having a 2'-O-methyl on its sugar, wherein N at the 3' terminus of (N)x is a modified ribonucleotide, (N)x comprises at least five alternating modified ribonucleotides beginning at the 3' end and at least nine modified ribonucleotides in total and each remaining N is an unmodified ribonucleotide; wherein in (N')y at least one unconventional moiety is present, which unconventional moiety may be an abasic ribose moiety, an abasic deoxyribose moiety, a modified or unmodified deoxyribonucleotide, a mirror nucleotide, and a nucleotide joined to an adjacent nucleotide by a 2 '-5' internucleotide phosphate bond; and wherein the sequence of (N)x is substantially complementary to the sequence of (N')y; and the sequence of (N')y is substantially identical to the sequence of an mRNA encoded by a target gene."

As the red highlight shows, the ‘proprietary’ claim of the Quark siRNAs rests on the sense/passenger strand similarly containing at least one ‘funny-looking’ nucleotide. I was disappointed, however, that the specifications did not explain why this should have any unique advantages. Since in addition to novelty (that's where they will likely try to focus their arguments on), a patent has to also fulfill the demands of non-obviousness and utility, I have my serious doubts that this and the usiRNA patent applications will stand up to closer scrutiny, and if they did get by any patent offices, Alnylam would ultimately challenge them. It suggests, however, that while patent protection for these RNAi triggers is unlikely, a number of players increasingly view this strategy as a way to at least gain independence from the Tuschls. The exact reasoning behind this is mysterious to me, since these modifications should already be explicitly covered by the Tuschls when they refer to 'nucleoside analogues' which usiRNAs are. And even if the Tuschls did not explicitly mention them, it would appear obvious that an siRNA is an siRNA whether it contains a limited number of ‘funny’ nucleotides or not.

As I said, I have yet to hear a convincing argument by the companies what their belief is based on. Just stating that they believe so is not enough to convince investors and potential partners. I am always willing to listen to such arguments.

Sense-induced Trans-Silencing: Combining the Best of RNAi and Antisense

No, this is not about single-strand RNAi (ssRNAi), nor an anti-antisense rant. This is about my (probably) last publication as a wet-bench scientist [Haussecker et al (2010): Human tRNA-derived small RNAs in the global regulation of RNA silencing], where my co-authors and myself stumbled across a new method of inducing RNAi that combines elements of DNA-directed RNAi (ddRNAi) and single-strand oligo therapeutics. We christened the technology sense-induced trans-silencing (SITS).

It all begins with the discovery of a new type of small RNA

The story starts with our investigations into the zoo of small RNAs that exists in our cells. We had originally reported on capped small RNAs that are generated in the course of hepatitis delta virus replication. In molecular biology, viruses often make for good model systems for studying general gene regulation mechanisms, and so we set out to discover the non-viral cellular counterparts of the HDV small RNAs. Long story short, instead of discovering the cellular counterparts of the HDV small RNAs (these in fact do exist and have since been widely reported by others to be generated during normal transcription initiation of many cellular genes, e.g. this one), we, in the spirit of curiosity-driven basic science, soon homed into a quite peculiar population of small RNAs that were derived from tRNAs.

It turns out that we were not the first to observe them. In fact many others had cloned them in their small RNA sequencing studies. They were, however, never fully investigated because scientists were apprehensive that they were mere degradation products of highly abundant RNAs. Maybe because we were able to visualize them through enzymatic means before Northern Blotting and to us looked like bona fide small silencing RNAs with 5’phosphates (tRNA degradation products are not 5’P), we continued to study their interaction with components of the RNAi pathway. Importantly, the tRNA-derived small RNAs (tsRNAs) were bound by Argonaute proteins, the effectors of RNAi.

A control experiment yields an unexpected result

Since tsRNAs interacted with Argonaute proteins, we went on with the standard assays to show that they had gene silencing capacity: luciferase reporters into which tsRNA target sites had been cloned. As a control, we co-transfected the luciferase reporters with oligonucleotides that were antisense to the tsRNAs (and therefore sense to the target gene), fully expecting that this would block any tsRNA silencing activity. To our surprise, instead of relief from gene silencing, the oligonucleotides actually triggered gene silencing. Based on subsequent mechanistic studies, it seems that while single-stranded 5’-phosphorylated small RNAs preferentially incorporate into Argonautes 3 and 4 which are not good silencers, converting them into double-stranded RNAs through the addition of complementary oligonucleotides now made them the preferred substrate for the most potent gene silencers among the Argonautes: Argonaute 2. Because the oligonucleotide that induces the silencing is sense with respect to the target gene, we termed the method Sense-Induced Gene Silencing, or SITS.

Sense-Induced Trans-Silencing- The Method

To effect SITS, the guide RNA is engineered into a tRNA expression vector and introduced into the target cells. By itself, this has little effect on target gene expression. However, once you introduce an oligo that is complementary to the guide RNA, silencing is triggered. This means that it is possible to tap into the inherently potent RNAi gene silencing pathway without the need for introducing double-stranded RNA which, in the absence of formulation, is harder to do compared to single-stranded oligos, for example as is largely practiced in antisense. The differentiating factor from antisense technology, however, is that the potency of RNAi on a per molecule basis is significantly higher, as will the general pharmacology be quite different for SITS.

	RNAi	antisense	SITS
silencing	good	challenging	good
delivery	challenging	good	good

You may ask, why not do straight-forward DNA-directed RNAi. I agree that DNA-directed RNAi is extremely promising. One drawback of most ddRNAi methods, however, is that it is a hit-and-run approach with little ability to regulate the degree of silencing after the DNA has been introduced. In the case of SITS, the activity of the guide strand that is derived from the DNA vector is only turned on as long as there is sense oligo present. Regulated ddRNAi is not new per se, but this method distinguishes itself in that it does not require a potentially immunogenic (foreign) protein such as in tetracycline-controlled systems.

Another question (that also a patent examiner might ask) is that in a way this is adding sense and antisense and thereby reconstituting active siRNAs. In a way this is a correct description of the mechanism, but to my knowledge, the temporally separate addition of passenger (‘sense’) and guide oligos (‘antisense’) has not been demonstrated to trigger successful gene silencing. Our method therefore is not only unexpected because the attempt to antagonize a 5’-phosphorylated small RNA did not inhibit gene silencing, but actually induced it, but also in light of such failures. The fact that it does silence could be due to the curious re-direction of the tsRNAs from Ago3/4 to Ago2 that occurs in the SITS setting. Also, unlike is the case of single-strand RNAi where the transfection of a single-stranded guide RNA is sufficient to trigger silencing, tRNA-derived small RNAs curiously do not do the same even when present in high concentrations. Moreover, I am not aware of attempts to provide passenger and guides through a combination of DNA-directed and synthetic means.

Potential uses of SITS

I probably would not be credible if I claimed that SITS was the preferred embodiment for all RNAi Therapeutics approaches. Like ddRNAi versus synthetic siRNA therapeutics, SITS should have its niches. Applications that might benefit from SITS include those that are also considered acceptable for gene therapies. In case you have not noticed, gene therapies really have come a long way in that there are now more and more clinical success stories, albeit often for orphan diseases (orphan diseases are becoming more and more popular in the drug development industry). Neurodegenerative disease, including those of the eyes could be one primary field of use, especially since AAV and lentiviruses have been shown to be well tolerated there and allow for long-term gene expression. Huntington’s Disease comes to mind where shRNA toxicity has been a concern as well as the question whether too much knockdown of the Huntington gene could be detrimental, too. With SITS, you would first introduce an AAV or lentivirus in the limited area affected by the disease and then via one of the pumps administer from time to time the sense oligo for measured gene knockdown. If there are side-effects, you could titrate down the dose. Furthermore, you may not have to get the vector into all the target cells, but where it does, it should be able to promote consistent gene silencing meaning that a critical number of cells could be kept alive for a long time enough to keep the patient functional.

Commercial aspects

Of course, without patents and, tied to it, funding a technology like this will never make it off the ground. If you are interested, you can contact either myself or the Stanford Office of Technology Licensing. One aspect that has kept investors away from gene therapies is a difficult business model that needs to justify $1M price tags for often just one round of treatment. Amsterdam Molecular Therapeutics for example thinks it can do so for their promising lead candidate for a lipoprotein lipase deficiency that may soon be commercialized based on the fact that you get an almost life-long benefit from one round of treatment, as compared to the $150K annually for some of Genzyme’s drugs that need to be taken continually. In the case of SITS, the cost for gene therapy could again be spread as the initial vector administration would be followed by the routine administrations of synthetic oligonucleotides.

And last but not least, back to tsRNA biology

If they are not junk, then what do tsRNAs do? To answer this question and based on the observation that tsRNAs by themselves do not appear to have robust trans-silencing capacity, we tested whether dialing up or down the abundance of tsRNAs would affect the global activity of other classes of small RNAs such as microRNAs or siRNAs because of competition for shared proteins. We in fact observed that when the abundance of tsRNAs was increased, microRNA activity was decreased. Since tRNA and supposedly tsRNA abundance is correlated with cell proliferation, tsRNAs may be part of the answer to the long-standing question why microRNAs, which biology often uses to determine the differentiated, non-cycling state, are globally down-regulated in cancer.

So that’s it from me as a bench scientist. Having been a bench scientist is a unique opportunity to follow and test your ideas. I was very fortunate that in the labs that I worked in, science for science sake, the ultimate expression of freedom, was possible. In an age that tries to quantify and regulate every aspect of our lives, this is sometimes difficult to maintain. It is therefore important to keep in mind, also as policy makers, that the saying about the best inventions being born out of serendipity is not a platitude, it is a reality.