Journal Club: Targeted Systemic Delivery of Stabilized Immunoliposomes to Leukocytes

In the latest issue of the premier scientific journal Science, Dan Peer and colleagues from Harvard Medical School report on the successful development of an RNAi delivery system that allows for targeted gene knockdown in leukocytes following systemic (intravenous) administration (Peer et al. Systemic Leukocyte-Directed siRNA Delivery Revealing Cyclin D1 as an Anti-Inflammatory Target. Science 319: 627). This research may be an important milestone towards opening up many more targets for systemic RNAi Therapeutics beyond the liver, lung, and tumors.

Until now, most systemic RNAi applications in pre-clinical animal models involved the passive delivery of various nanoparticle formulations which proved particularly effective in highly vascularized organs such as the liver. Other organs, however, are more difficult to reach mainly because most RNAi delivery formulations will either have been metabolized or excreted through the kidney, before they can reach the less accessible places in the body or enter some of the more difficult-to-transfect cell types of the blood. The present paper represents one of the more promising reports on achieving efficient gene knockdown in these organs through formulations that are actively targeted to the organ of interest through a range of peptides, antibodies, aptamers, and small molecules.

In order to combine high RNAi loading capacity, stability with consequently favorable circulation times, and cell targeting, the authors started with simple ~80nm (neutral) liposomes and added to their outside stabilizing hyaluronan (similar principle to the cationic liposome-based SNALPs- stable nucleic acids lipid particle- which typically carry stabilizing PEG on their outside). In a second step, they covalently attached an antibody specific for an integrin highly expressed on leukocytes that traffic to the gut and play a central role in autoimmune inflammatory bowel diseases such as Crohn’s. In a last step, the immunoliposome particles were loaded with the siRNA cargo that had been condensed with the highly basic protamine so as to achieve 80% loading efficiency (~4000 siRNAs per particle). That the siRNA-protamine condensation step should facilitate such efficient loading into the neutral liposomes is quite notable and may solve the poor nucleic acid loading typically associated with neutral liposomes (note: neutral liposomes may be advantageous over certain cationic liposomes in reducing unwanted interactions in the body).

The nanoparticles were then tested for gene silencing in the notoriously difficult-to-transfect leukocytes. Strikingly, sequence-specific, integrin-dependent, and antibody-dependent silencing was obtained both in vitro and in mice. The silencing efficiency was quite remarkable, typically ranging between 70-85%. When an siRNA was targeted to cyclin D1, this resulted in a reduced Th-1 response while the Th-2 response was unaffected. This is thought to be beneficial in treating chronic inflammatory diseases like Crohn’s, and sure enough, Cyclin D1 suppression by RNAi in a mouse model for intestinal inflammation almost completely reversed the disease phenotype. Hence, this paper not only demonstrates systemic gene silencing in leukocytes, but also validates Cyclin D1 as a potential drug target for Crohn’s and other inflammatory diseases.

Two aims underlie the targeted delivery paradigm: one is to deliver the RNAi trigger in sufficient amounts to the tissue of interest to achieve therapeutic levels of gene silencing; the other is to reduce potentially harmful exposure of non-target tissues to the siRNA, e.g. in instances where uptake of the siRNA in macrophages or dendritic cells may increase the risk of an unwanted immune response and render dosing less predictable, or where silencing of a gene itself in a non-target tissue may have adverse consequences. So far, most reports on targeted delivery satisfied mainly aim 1 while leaving aim 2 largely unaddressed. Quite impressively, the authors demonstrate in a biodistribution experiment truly targeted delivery by showing that the siRNA-nanoparticle were quantitatively re-directed to the gut in the mouse disease model (an almost 100-fold increase) and depleted from the blood and liver which would otherwise take up a good fraction of siRNA-nanoparticle studded with control antibody.

Of course, more time will have to be spent characterizing this system before it may move into the clinic. Safety is just one aspect that needs to be looked at in more detail, but at least in terms of cytokine activation and body weight was found to be satisfactory in the present study. Although some experiments involved repeat administration, it will further be important to determine whether there is immune recognition to any of the components of the particles which may interfere with repeat dosing. This also relates to the comparative complexity of these multifunctional particles which may also mean that the total costs of manufacturing them could considerably exceed that of the siRNA effector alone.

Surely, there will be a number of diseases for which such expenses are more than justified. I personally am not a great believer in blanketing large populations more or less indiscriminately with statins or aspirin, but rather like to see more targeted therapies that show clear benefits in well-defined patient populations. Since I expect many of the future RNAi Therapeutics to fall into the latter category and ultimately represent a bigger bang for the healthcare dollar, it is therefore important that the healthcare and patent systems facilitate rather than antagonize such innovative drug development efforts.

Journal Club: Alnylam and Collaborators Make Progress in Understanding and Optimising siRNA Uptake In Vivo

In yet another elegant paper, Alnylam and collaborators from the ETH in Zurich and the Rockefeller report this week in the journal Nature Biotech on the mechanism of siRNA uptake in vivo (Wolfrum et al.). This is significant, since a systematic approach to understanding siRNA delivery in vivo should most adequately address the delivery challenge that is considered by many to be the main barrier to the broad application of RNAi as a therapy. It is also a sign of the maturity of the RNAi field in general that no only are there now a variety of innovative delivery systems evaluated almost by the day, systems that show promise serve as leads for a detailed investigation of the underlying biology.

The study by Wolfrum and colleagues follows another high-profile publication 3 years ago (Soutschek et al.) where Alnylam scientists demonstrated gene silencing in mice following systemic administration of cholesterol-conjugated siRNAs. That study showed that although such siRNAs could silence genes particularly in the liver and gut, quite high amounts of siRNAs were needed (50mg/kg). By studying the uptake of the siRNA conjugates in these tissues, the authors not only hoped to understand why they functioned at all, but also to optimise their potency.

Efficient in vivo drug delivery requires favourable pharmacokinetics. Particularly, a drug has to be present in the blood for sufficient length of time so that it has a chance to accumulate in its target tissue. One reason for example why many experimental drugs fail is because they are rapidly excreted through the kidneys. This may often be prevented if the drug could interact with components of the blood such as the abundant lipoprotein particles.

Indeed, the authors find that siRNAs conjugated to cholesterol or other lipophilic molecules associated with the similarly greasy HDL and LDL lipoprotein particles. These would ferry them around in circulation and bring them into the proximity of cells that carry on their surface receptors for either HDL and/or LDL. Strikingly, pre-assembling the siRNA with purified HDL and LDL particles quite significantly increased the potency of the siRNAs. Furthermore, mice lacking either of the receptors for the lipoproteins were much less prone to gene silencing by the same pre-formulated siRNA particles.

In a further interesting twist, it was shown that siRNAs were not taken up by the cells as part of internalising lipoproteins, but that the siRNAs would take advantage of their proximity to the cell membrane during the docking, release, and re-docking process of their lipoprotein carriers with their receptors. Amazingly, through a combination of gene knockdown experimentation and blockage by antibodies, at least one of the actual entry routes for the siRNA was inferred to be the human homologue of the SID-1 gene that had earlier been shown to mediate systemic RNAi in the worm C. elegans.

Systemic RNAi describes the spread of an siRNA from one cell to another cell in the same or even different tissues. Systemic RNAi in worms and plants is associated with the amplification of RNAi, and both systemic RNAi as well as RNAi amplification were thought to have been lost during human evolution. It is therefore a surprise that SID-1 would still function in siRNA uptake, with demonstrated selectivity for siRNAs relative to other types of nucleic acids. This also raises the intriguing possibility that some sort of natural siRNA uptake should occur in humans.

Of more immediate importance, the present paper opens the door for the systematic screening of new lipophile-siRNA conjugates with improved association kinetics with lipoprotein particles, or even pre-formulation of such conjugates with lipoproteins or other natural or synthetic carriers of the blood. I look forward to what this line of investigation will yield next.