Less Is Sometimes More: Roche Brain Shuttle Technology for Drug Delivery into CNS

After Roche and ISIS Pharmaceuticals announced last April that a co-development effort to apply Roche’s Brain Shuttle technology for the systemic delivery of antisense oligos would form part of their Huntington’s collaboration, I started to pay some attention to this technology.  This is because despite the already enormous promise of Oligonucleotide Therapeutics for CNS disorders (think of all the neurodegenerative diseases) with the direct administration of single-stranded oligonucleotides into the CNS, its value could be further enhanced with a systemic delivery approach such as the Brain Shuttle tech.

This is not only because intravenous administration methods would be preferable over the more invasive direct intra-CNS injections/infusions, but also because a systemic delivery approach promises a more uniform drug distribution and would minimize the importance of diffusion.  It is the limited diffusion of current RNAi delivery techs that is holding back RNAi Therapeutics in this important therapeutic area.

Monovalent antibody binding to transferrin receptor allows for efficient transcytosis

The fundamental principle behind Brain Shuttle is actually not that novel at all. Companies like Armagen have long attempted to target receptors such as the insulin receptor and the transferrin receptor on brain capillary endothelial cells as in normal physiology these receptors function to shuttle transferrin/iron and insulin across the notoriously recalcitrant blood-brain-barrier.

Based on newly published data by Roche scientists (Niewoehner et al. 2014), a conventional antibody approach does not work.  To wit, a conventional antibody consists of two binding sites (divalent) and such engagement apparently causes receptor trafficking to be re-directed to the degradative lysosome compartment of the brain endothelial cells, at least in the case of the transferrin receptor tested.  By contrast, when receptor targeting occurs via monovalent interaction, the normal receptor physiology, including transcytosis, is maintained.  Using this strategy, it was shown that the intra-parenchymal delivery, i.e. delivery into the brain proper, of an antibody against beta-amyloid (related to Alzheimer’s disease) that had been tethered to the monovalent transferrin receptor antibody was enhanced on the order of 50-fold.

The way how this research can be translated into RNAi Therapeutics is to simply append an RNAi trigger instead of the beta-amyloid antibody to the transferrin receptor antibody.  Or you could do away with proteins altogether and replace the transferrin receptor antibody with an aptamer-RNAi trigger combo (e.g. a Dicer-substrate in the spirit of tomorrows IPO by Dicerna Pharmaceuticals).  Hence, the irony of the ISIS-Roche delivery collaboration is that ISIS could- to put it just slightly hyperbolically- be shoveling its own grave by eroding the current advantage of phosphorothioate antisense over RNAi for gene knockdown in the CNS.

New Breakthrough in the Systemic Delivery of RNAi for the Brain

One often cited challenge for the wide application of RNAi in the treatment of disease is the systemic delivery of the RNAi inducing agent to the appropriate target cells. Systemic delivery is the ability to deliver RNAi to target cells following intravenous or even oral administration, whereas local, also called “Direct RNAi” approaches for eye, lung, and mucosal diseases appear to be less of a challenge.

Neurodegenerative diseases and other brain-related diseases are an attractive area for the application of an RNAi therapeutic, due to significant unmet medical needs and the existence of well validated gene targets. The delivery of RNAi to the CNS, however, is challenging due to the presence of the blood-brain barried (BBB) which makes it difficult for macromolecules to exit blood vessels and enter the CNS. First generation RNAi therapeutics for the CNS may therefore rely on direct injection or device implantation as is illustrated by the current Alnylam-Medtronic collaboration. Systemic delivery, however, would have the added advantage that a therapeutic would be well distributed throughout the brain due to the extensive vascularisation of the brain. Local administration strategies often suffer from the fact that the therapeutic agent is restricted to the site of application.

The paper published this Monday online in the leading scientific journal Nature by Kumar et al. shows that the ability of the rabies virus to cross the BBB can be conferred onto an siRNA if it was non-covalently linked to a 29 amino acid peptide from the rabies virus. Although the mechanism by which this peptide achieves this is unclear, the authors show that once in the CNS it binds to acetylcholine-receptors on neuronal cells and concomitantly delivers the siRNA load to those cells. A number of endogenously expressed genes were thereby targeted achieving 30-70% knockdown efficiencies, including a 70% reduction in SOD-1 activity.

SOD-1 when mutated is linked to the debilitating neuronal disease ALS (Amyotrophic Lateral Sclerosis; aka Lou Gehrig’s) for which there is no real treatment. Interestingly, preclinical therapeutic effects have previously been achieved with lentivirally delivered shRNAs. It is therefore hoped that ultimately RNAi may offer the first treatment alternative for ALS that addresses the underlying cause of the disease.

In another impressive demonstration of the system, almost all mice infected with an otherwise fatal dose of Japanese Encephalitis Virus were rescued when they received the rabies-peptide siRNA combination. This follows a report last year by the same group that single siRNAs could also protect against the related West Nile Virus.

This report demonstrates that progress in the delivery of RNAi Therapeutics can come from unexpected sources and illustrates the force and commitment with which this technology is moving forward. It is reasonable to assume that one by one, similar solutions and improvements to the RNAi delivery challenge will come for other difficult-to-target tissues. Further improvements to the present system should also be possible, as the authors suggest, for example through covalent conjugation of the siRNA to the peptide to avoid premature dissociation of the siRNA from the rabies peptide. It is also tempting to imagine that high-throughput approaches such as one involving peptide phage display libraries could yield siRNA-peptide combinations for other tissues of the body.

The key will be that the results can be reproduced in another laboratory, is easy-to-use, and similarly non-immunogenic in man as it is in mice. In that case, I would not be surprised to see the technology being licensed by a major RNAi Therapeutics company. The proximity of Harvard to Alnylam’s Cambridge, Mass., headquarters certainly makes this company a prime candidate for that. It should also be noted that one of the co-authors, Beverly Davidson from the University of Iowa, an acknowledged RNAi expert with experience in brain-related RNAi applications, had been associated with Sirna Therapeutics prior to their acquisition by Merck.