Antibody-RNAi Trigger Conjugates Show Signs of Life

Despite jettisoning RNAi Therapeutics 4 years ago, the more innovative arm of Roche, Genentech, has continued to dabble in the technology.  In particular, it has been interested in applying its monoclonal antibody know-how, including antibody-drug conjugates (ADCs) to the delivery of RNAi Therapeutics.  A recent publication by Cuellar and colleagues provides insights into these efforts.

THIOMABs for the creation of drug-like conjugates

While antibodies have a relatively long history in the delivery of RNAi Therapeutics, some of the early findings were generally quite difficult to replicate.  Part of the problem may have been the fact the early efforts involved structurally ill-defined non-covalent protein-nucleic acid complexes held together by charge-charge interactions.

To get around this issue, Genentech applied their THIOMAB platform which allows for the covalent addition of therapeutic payloads at defined cysteine residues.  This process yielded THIOMAB-siRNA conjugates with one, or more often two RNAi triggers per monoclonal antibody.  To help visualize them, also for pharmacokinetic considerations, think of the monoclonal antibody part being ~13x bigger/heavier than each RNAi trigger.

Of note, the RNAi triggers were partially (~50% of residues) modified with 2’-O-methyls and 2’-F for stability (siSTABLE from Dharmacon).

Not all receptors are created equal

One of the reasons why I am picking out this paper for discussion is the thoroughness of the research presented.  For example, the Genentech researchers selected not just one, but seven distinct cell surface receptors for which various THIOMAB-siRNA were created which in turn were be tested in a number of settings.  The receptors were partly chosen to capture a range of cellular trafficking behaviors such as rapid lysosomal uptake, recycling receptors, and slow-turnover receptors.

Importantly, the research validated a critical rationale for the use of antibody-RNAi trigger conjugates namely uptake of the RNAi triggers that is dependent on the presence of the cognate receptor and covalent linkage to the antibody.

For all the conjugates (ARCs), the bulk was shown to accumulate in lysosomes regardless of presumed uptake kinetics.  Interestingly, despite the seemingly shared uptake pathway, if not dynamics, only two of the seven receptors were associated with gene silencing (TENB2 and NaPi2b, but not e.g. Her2). 

The degree of silencing was much weaker compared to when the same conjugates were lipofected with ~50% silencing starting to be seen at 10nM.  This was followed by a shallow dose-response plateauing at 70-80% silencing around 500nM.  Based on the data presented, it seems that limiting amounts of receptors were responsible for this. 

Selective accumulation in mouse tumor model

A highlight of the publication to me was the investigation of THIOMAB-siRNAs in a mouse tumor model.  Although you may consider a ~30% silencing (i.e. 70% expression of normal remaining) when you cherrypick tumor areas of most efficient delivery a somewhat disappointing outcome as was reported in this case, important lessons can be learned from that.

Firstly, THIOMAB-siRNAs only accumulated in the tumor when the tumor expressed the cognate receptor.  This is unlike nanoparticulate RNAi delivery to tumors which relies on a passive process of accumulation (the EPR effect).  This opens up the prospect of RNAi drugs with ‘cleaner’ delivery profiles with an increased margin of safety.

Selective accumulation for these conjugates also suggests that the ~180kDa macromolecules were able to relatively rapidly exchange between tumor interstitium and blood circulation despite their size.  Consequently, other conjugates in a similar size range or below, including Arrowhead’s DPCs, should be amenable to such ‘active targeting’ as well.

Nevertheless, despite this apparent agility, imaging techniques showed that tumor delivery was largely restricted to areas next to the vasculature.  This limitation in fact is what is also seen with nanoparticles which rely on the EPR effect.  So while the active targeting capability is an important step forward, tumor penetration issues remain to be solved.

Path forward

Despite the arguably underwhelming in vitro and in vivo knockdown results if you just look at the numbers, I am optimistic that the studies have further supported that systemic RNAi delivery can be applied beyond the liver, vascular endothelial cells, certain cells in the kidney and phagocytes.

What is missing in that particular piece of research making the conjugates that make it into the cells count.  Given the lysosomal accumulation, an obvious strategy to unlocking the true potential of ARCs would be to apply endosomal escape chemistries, especially masked chemistries such as in the DPCs by Arrowhead Research (actually specifically referenced by the authors).

In almost the same vein, the sparse chemical modification used mean that potency improvements will be gained if heavier modification is applied even without adding endosomal escape chemistries.  This is also because some of the data are consistent with a model whereby the THIOMAB-siRNA conjugates get broken apart in late endolysosomal compartments and it is from that population of freed RNAi triggers that escape into the cytoplasm may occur (note: this does not exclude spontaneous endolysosomal rupture as an additional escape pathway).  If RNAi triggers were made metabolically more stable through more extensive modification, the amount of RNAi triggers available for such escape would obviously be larger.

All that is lacking then is the issue of tumor penetration.  I could imagine, however, that some self-delivering and/or lipophilic strategies could be quite useful here.  The Genentech researchers actually speculate that the RNAi trigger may have been the culprit in limiting the claimed natural ability of monoclonal antibodies to more deeply penetrate tumors.  This suggests that charge masking may overcome this limitation.  

I am open to this idea, but based on my understanding of the literature, tumor penetration is a challenging issue for basically all therapeutic modalities, monoclonal antibodies and 'even' small molecules, so let's not blame double-strandedness for all pharmacologic problems.

With interest in RNA Therapeutics growing by the day, expect lines of delivery research such as this one to be picked up and pushed to the next level.

PS: for antibody-mediated RNA delivery, Avidity NanoMedicines and the Roche-ISIS collaboration on CNS delivery should also be worth watching.

January RNA Therapeutics Deal Frenzy Kicked Off

As is now tradition for Januaries in RNA Therapeutics, expect this month to be filled with mouthwatering business development announcements.

Without a doubt, privately held messenger RNA Therapeutics company Moderna Therapeutics stole the show today by announcing a $450M financing involving 'Viking Global Investors LP, Invus, RA Capital Management, and Wellington Management Company, LLP, as well as existing investors AstraZeneca and Alexion Pharmaceuticals'.

Biggest private financing in biotech history EVER.

With $800M in cash, close to a billion USD raised since starting up 2 years ago or so, the biggest challenge will be how to efficiently put the capital to good use and move into the clinic in the not-too-distant future.  Before that happens, we should be seeing one of the hottest biotech IPOs in 2015.

Before Moderna announced their financing, ISIS Pharmaceuticals was the talk of the day with their GI-related autoimmune deal with Johnson&Johnson.  Not only does the early development deal provide ISIS with another nice junk of upfront cash, this time $35M, and a gazillion in potential biotech milestone payments and royalties, jn practical terms I expect the research to be particularly useful to ISIS as it develops oral delivery for systemic applications (think lipid franchise for which the new CEO will be officially enthroned tomorrow).  

I believe that the deal was primarily driven by JNJ's interest in the space following the whopping $710M upfront deal by Celgene last year when it acquired rights to an antisense compound developed by oligonucleotide therapeutics nobody Nogra Pharma for IBD applications.  The high upfront payment indicates that competitive bidding was involved and it is likely that JNJ lost out.

Remember, Big Pharma moves in herds.

The good news for JNJ is that if the first-generation phosphorothioate DNA is indeed working by an on-target mechanism (which I doubt), then a competing compound employing much more advanced chemistry and design by ISIS Pharmaceuticals will be infinitely more potent and beat Celgene in the market.  And for $35M and $800M in milestones if everything goes according to plan, it's a comparative steal.

For ISIS Pharmaceuticals it is just another step to broaden their expertise and applications of their platform.  Moreover, the upfront payment will help the company to remain cashflow neutral to slightly positive until 2017, at which time I expect drug sales to be finally taking off.