RNA-based molecules have redefined the universe of tractable targets by putting virtually anything that is gene encoded within reach of a disease-modifying agent. This redefinition has launched RNA as the biotech industry's third drug modality.

The emergence of noncoding RNAs presents yet further opportunities for the modality, but a rudimentary understanding of the bioactivities of these molecules, the delivery hurdles and the difficulties of target selection and validation still surround this new therapeutic class.

A think tank convened by SciBX that comprised academic, biotech, pharma and VC stakeholders surveyed the state of the noncoding RNA space. The key opinion leaders laid out a road map for developing and implementing RNA therapeutics and for exploiting the molecules-which include long noncoding RNAs (lncRNAs) and microRNAs-as disease targets.

"With small molecules, there is an undruggable or at least difficult to drug set of targets. A second modality, of course, is biologics-antibodies and so forth. But there are still a lot of targets that we would like to go after that we cannot drug with those approaches. And that is where RNA and oligonucleotides come in. In principle, there is no undruggable target anymore," said panelist Aimee Jackson. She is director of target development at miRagen Therapeutics Inc., which is developing anti-miRNAs.

For the near term, the think tank recommended that drug developers prioritize diseases for which there is solid knowledge of the human genetic underpinning and focus on strategically selecting RNA targets based on how accessible they are in humans. In particular, to avoid clinical setbacks caused by poor target selection, the key opinion leaders emphasized the importance of identifying reliable biomarkers and developing target engagement assays for clinical studies.

For the long term, the panel outlined activities that will be needed to gain a comprehensive understanding of the biology of emerging classes of noncoding RNAs. In addition, it pressed for new delivery agents such as conjugates or particle formulations that expand the repertoire of accessible tissues and improve intracellular release of therapeutics following endocytosis.

Thus, for the second straight year, endocytosis was called out as a black box that impedes drug delivery. The 2012 SciBX Summit, which focused on macrocycles and constrained peptides, also stressed the need for translational studies on endocytosis.1

In addition to Jackson, the think tank consisted of David Corey, Jean-François Formela, Art Krieg, John Maraganore and Laura Sepp-Lorenzino.

Corey is a professor of pharmacology and biochemistry at The University of Texas Southwestern Medical Center and has been exploring oligonucleotide-based therapeutics and RNA targets for more than 20 years. Formela is a partner in the life sciences group at Atlas Venture and a cofounder of RaNA Therapeutics Inc., which is developing molecules to upregulate gene expression by targeting lncRNA. At the time of the meeting, Krieg was CEO of RaNA. Most recently, he was SVP and CSO at Sarepta Therapeutics Inc., which develops RNA-based therapeutics targeting mature or precursor mRNAs to turn gene expression on or off. He departed that position earlier this week. Maraganore is CEO of Alnylam Pharmaceuticals Inc., which develops therapeutics based on RNAi. Sepp-Lorenzino is VP and an entrepreneur in residence at Alnylam. At the time of the summit, she was executive director of RNA therapeutics discovery biology at Merck & Co. Inc.

According to Formela, "The conjunction of the maturation of the modality, the explosion of the biology and the focus of the industry and of the capital market on trying to solve orphan disease has created a perfect opportunity for investors."

Maraganore said that numerous advances have minimized some of the early negatives associated with RNA therapeutics, including undesirable toxicity or immune system activation.

Formela added that in terms of toxicology, therapeutic RNAs are becoming more like small molecules. "If we end up with a platform that is just as good-maybe better in terms of time to development of a candidate because the design on the front end is much more rational than the design of a small molecule-I think it will be a very attractive platform," he said.

RNA buzz

There has been a steady increase in investment in oligonucleotide-based therapeutics over the past three years, including a surge early this year (see "2014 first-quarter industry activity in RNA therapeutics").2

More than $1 billion was invested in the RNA space during 1Q14, headlined by Alnylam receiving a $700 million equity investment from the Genzyme Corp. unit of Sanofi that expanded the strategic alliance between the companies.

Genzyme obtained global rights to commercialize Alnylam's lead compound, patisiran, outside the U.S. and Western Europe. The expansion added three additional molecules from Alnylam's pipeline to the alliance. Until 2020, Genzyme has the option to co-develop and commercialize outside the U.S. and Western Europe all rare genetic disease therapeutics in Alnylam's pipeline.

Other big deals include Alnylam's acquisition of Sirna Therapeutics Inc. from Merck for $75 million up front plus milestones and Moderna Therapeutics Inc.'s $100 million deal with Alexion Pharmaceuticals Inc. to develop messenger RNA-based therapeutics in rare disease.

On the finance side, Dicerna Pharmaceuticals Inc. raised $90 million in a January IPO.

Also that month, the NIH announced a data-sharing initiative with Life Technologies Corp. The goal is to release 65,000 siRNA sequences targeting more than 20,000 human genes. New data generated from the initiative will be added to PubChem on an ongoing basis, making the database a resource for the RNA community.

In fact, investment in basic research and resources for the RNA community has become a high priority for government funding agencies even in the face of shrinking budgets.

The NIH made its first large-scale investment in RNA biology in 2003 when the National Human Genome Research Institute launched a public research consortium called ENCODE (the ENCyclopedia Of DNA Elements) to identify all functional elements in the human genome. The pilot and scale-up phases of ENCODE cost about $185 million. In 2012, the National Human Genome Research Institute announced the third phase of the project and said that it could invest up to $119 million through 2016.

No code

The discovery that the majority of the genome is expressed as noncoding RNA was a major-and unexpected-finding from ENCODE.

"I think that was the big bang," said Formela. "If 80% of the genome is transcribed and biologically active, frankly, it does not take a very smart VC or PhD to figure out that there will be a huge amount of biology that is unprecedented that you can potentially affect."

"Within the last 10 years, it's become clear that most of the genome is transcribed. Most of that transcription is likely to be noise produced at low volumes. And it's not likely to have a biological effect. But what of the transcription, especially around genes' enhancer regions; what kind of effects can that have?" asked Corey.

The panel emphasized that drug developers in the noncoding RNA space can take advantage of existing technologies used for coding RNAs such as those that induce silencing or exon skipping.

"In terms of the platform chemistry, we have a lot of tools at our disposal," said Jackson. "Chemical modifications to the RNA backbone, to the ribose moiety as well as to the base of the RNA enable us to dial in a lot of characteristics for affinity, for stability in the body and for duration of action within the body. We are starting to appreciate just how drug-like we can make these small RNAs by combining the chemical modifications to the various moieties of the RNA structure."

"There is no further generation of chemistry that is going to be needed," said Krieg. "The tools that we have now are the chemistries that have come about from well over 20 years-probably closer to 30 years-of work from chemists. Right now, today, we know that with single-stranded approaches we can get into multiple organs in the body by simple subcutaneous administration. Double-stranded approaches have made huge strides."

Maraganore added that chemical modifications-together with new screening tools-have also addressed toxicology concerns such as activation of toll-like receptors or intracellular RNA sensors.

"There is a huge difference in potency, pharmacology and, you could argue, toxicity, between some of the first-generation chemistry and some of the latest-generation chemistry," added Formela. "In terms of binding specificity and potency pharmacology, a lot of progress has been made. You can now tune those features. Maybe not 100%, but you get a good sense of how quickly you can get to a development candidate, which is a very important first step-and frankly a requirement-to make an investment."

Talking targets

Although the chemistry for noncoding RNA therapeutics is on solid ground, the best way to find good targets is open to debate, and their importance cannot be overstated.

"Target selection is the single most difficult step that an early stage company takes," said Formela. "And I personally feel that it's probably the largest competitive factor or success factor in getting there soon enough that you don't fall out of favor with investors. It's extremely important."

Step one, said Sepp-Lorenzino, is making sure a target is expressed in tissue types amenable to RNA therapeutics. The liver is the clear front-runner here, and additional tissue types include muscle, eye, CNS, lymphocytes, skin, kidney and tumors (see "Accessible human tissues").

Within those tissues, well-validated protein targets that need to be upregulated to achieve a therapeutic effect are particularly attractive. And at RaNA, Krieg said, scientists further home in on the nucleotide regions adjacent to a target's gene to locate potential regulatory noncoding RNA sequences.

Panelists also said that phenotypic approaches should be pursued for identifying therapeutically relevant noncoding RNAs.

At miRagen, Jackson said, scientists "are doing more phenotypic-based approaches in vivo and bypassing in vitro altogether." She added that because expression of both miRNAs and the mRNAs they regulate varies by tissue and cell type, and also varies in the disease setting, phenotypic screening can provide the most physiologically relevant information for target selection.

Ultimately, using biomarkers to assess target engagement in humans will be essential to tie target modulation to impact on disease. Such biomarkers are important tools for predicting success in clinical trials.

At Alnylam, "we have to have a point-of-care marker that we can read out in Phase I, whether it is a secreted protein where you can measure a decrease in the protein levels in the blood or some other indirect but related type of factor," said Maraganore. In the absence of a reliable biomarker, it is difficult if not impossible to distinguish between failure to hit a target and a bad target.

For example, said Maraganore, "we've been using 5ʹ-RACE, which is a PCR-based technique, to confirm that we get cleavage at the target mRNA. We initially did that in humans with biopsy samples in a liver cancer study, but we've since been able to do that by capturing cleaved RNA in exosomes that are circulating in plasma. Literally, we can detect RNAi in liver from a plasma sample."

The emergence of lncRNAs

Although miRNAs already form the basis of a number of preclinical programs, lncRNAs are a step behind as less is known-and much is debated-about how they affect biology.3

For lncRNAs to gain acceptance as a therapeutic class, researchers will need to find molecules that can impact the course of a disease, uncover the mechanism of action and generate animal models that recapitulate the human biology of both the lncRNA and the disease.

Early on, said Krieg, "we discovered that it's possible to upregulate gene expression by targeting what we now think are lncRNAs in enhancer regions of various genes."

He added, "When you look at the expression patterns of the lncRNAs, they're very selectively expressed. It's just a single layer of cells in one tissue, often the CNS, or a single point in time. There's an extraordinary level of biologic complexity there, and we're just scratching the surface of how we can target that for drug development."

According to Corey, one possibility is that lncRNA "remains associated with chromosomal DNA and acts as a scaffold for proteins that can then affect transcription or splicing. In a second mechanism, the RNA is produced in sufficient quantity that it can work either by transiting to another location in the genome or by producing miRNAs."

Whichever is right, "the frontier of targeting lncRNAs themselves as disease targets is going to take a little bit more time," said Maraganore. "It is going to take work from people in academia to figure out the biological function of those molecules."

Nevertheless, Corey proposed that mechanism elucidation take a backseat to target discovery. Confirming that a lncRNA impacts a disease phenotype can be grounds to move forward. "You have to make sure that you have the interesting molecule first because understanding the mechanism could take a pretty long time and require some investment-and you want to make sure that it's worth that investment up front," he said.

Despite the limited understanding of lncRNAs and their mechanisms, at least three companies have been founded to pursue lncRNAs as targets.

Curna Inc. was founded in 2008 by Claes Wahlestedt to develop antagoNAT technology for upregulating expression of therapeutically relevant proteins. At the time, Wahlestedt was a professor of molecular therapeutics and adjunct professor of molecular and integrative neuroscience at Scripps Florida.

AntagoNATs are natural antisense transcript antagonists against endogenous antisense lncRNAs that inhibit mRNA translation. Opko Health Inc. acquired Curna in 2011 for $10 million in cash.

MiNA Therapeutics Ltd. is developing small activating RNAs (saRNAs) that target regulatory DNA regions upstream of therapeutic genes to increase their expression. saRNAs have been reported to work via multiple mechanisms, including disrupting the activity of repressive lncRNAs.4,5 The company was cofounded in 2008 by John Rossi, chair of molecular and cellular biology and dean at City of Hope.

RaNA was cofounded in 2011 by Jeannie Lee, a professor of genetics and pathology at Harvard Medical School and Massachusetts General Hospital and a Howard Hughes Medical Institute investigator.

The biotech is developing therapeutics to disrupt lncRNAs that recruit polycomb repressive complex 2 (PRC2) to promoters of therapeutically relevant genes and thereby specifically upregulate that gene's expression. The company is focused on spinal muscular atrophy (SMA) and Friedreich's ataxia.

According to Krieg, target selection at RaNA is pretty straightforward. Researchers perform deep RNA sequencing in multiple cell types to define lncRNAs expressed in diseased and normal states. Next, they use protein-RNA interaction assays such as RNA immunoprecipitation sequencing (RIP-seq) and crosslinking immunoprecipitation sequencing (CLIP-seq), which involve immunoprecipitation of chromatin regulatory proteins with associated noncoding RNAs. Deep sequencing of the associated RNAs generates a transcriptome for that tissue. RaNa is doing that in multiple tissues, Krieg said.

"Companies like RaNA can effectively use lncRNAs as a strategy for doing something that is validated, which is to upregulate target gene expression," said Maraganore.

Academics also are starting to puzzle out therapeutic targets from the large pools of lncRNAs.

In 2013, a group from The Johns Hopkins University School of Medicine reported a lncRNA associated with potassium channel Kv1.2 (KCNA2) that might repress the gene's expression. In rats, the lncRNA is upregulated in neurons in response to damage.6 Although the lncRNA is expressed in humans, its relationship to KCNA2 and the mechanism by which it regulates gene expression still need to be determined before it can be considered a therapeutic target.7

Also in 2013, a team from the University of California, San Diego and The University of Texas MD Anderson Cancer Center reported two lncRNA targets that functionally contribute to prostate cancer growth.8 The group is developing oligonucleotides targeting the lncRNAs for further preclinical studies.9

Looking out for orphans

To build a successful therapeutic program, the panel recommended focusing on orphan indications that have been genetically validated in humans in which using RNA could provide an advantage over competing modalities.

"We look for a therapeutic area or an indication where there is not anything else that can be developed" said Krieg. "There are lots of targets, but we do not want to compete with other modalities."

He added, "We like orphan disease indications not because it is the only area where we can develop therapeutics, but you get a few extra years of exclusivity and clinical trials are smaller and easier to run for a company like RaNA."

RaNA is not the only company opting for the orphan route (see "Orphan drug designations"). The most popular therapeutic areas for orphan programs are neurology, cancer, ophthalmology, musculoskeletal disease and metabolic diseases, largely because these provide targets in accessible tissues.

Although orphan diseases provide strategic benefits, the panel emphasized that having a target with strong genetic validation data should also be a key consideration when selecting an indication.

"I also put a lot of stock in genetic validation," Jackson said. "There is a lot of information available now on dysregulation of miRNAs, but dysregulation does not tell you enough. I put stock in either genetic or pharmacological validation that modulating the activity of the miRNA is actually driving the phenotype."

Maraganore agreed but raised the bar a notch. "The target has to have a human genetic foundation. Even a mouse knockout or transgenic is not enough. We want to have loss-of-function or gain-of-function human mutations that validate the target, essentially taking the biological risk as close to zero as you can possibly get in drug development."

Animal house

Identifying animal systems that faithfully model lncRNAs in human disease is going to be difficult. Noncoding RNAs are not well conserved across species, which will hamper the validation of lncRNAs as therapeutic targets.

That obstacle will affect assessment of both efficacy and toxicity.

"It's going to be a challenge for those thinking about drug development with lncRNAs as targets to think about how you make sure that your in vivo pharmacology in a mouse might translate ultimately to a human," said Maraganore.

"For safety studies you need to have one species, most likely the primate, where there will be some conservation and you can answer pharmacotoxicology questions," said Sepp-Lorenzino.

This challenge could also extend to miRNAs. "They are conserved, but whether the network in the mouse is the same as the network in the human is going to have to be worked out," said Maraganore.

According to Corey, questions about sequence conservation "emphasize how little we know. We do not even know how big a problem it is ultimately going to be."

Because many disease models are designed to mimic loss of function or gene knockout in humans, Formela said that it is important to show that a gene locus is "regulated in the same way-or at least in a similar way-in animal models and in human cells."

"For some diseases there already are humanized models," said Krieg. "For example, in spinal muscular atrophy, there are several different mouse models. For Friedreich's ataxia, the same thing. More and more foundations now realize if they want therapeutics to be developed for their disease, having a humanized mouse model will help."


Getting nucleic acid therapeutics into many types of tissue remains a significant challenge for the entire field of RNA therapeutics. Although chemical modifications stabilize RNAs for circulation, rerouting the molecules away from being cleared by the reticuloendothelial system and getting them across cell membranes remains difficult.

"We knew from day one that delivery was something that had to be solved. It is a key issue whether you are targeting a coding RNA, a lncRNA or an miRNA. Delivering to the target is the key, number-one issue," said Maraganore.

The panel debated whether advances in nucleic acid chemistry alone could overcome the delivery issue.

"If you have a potent enough oligonucleotide for the target that you are pursuing, then even though half of the total dose is going to wind up in the liver and kidney, enough gets into muscle, intestine and other organs that you may want to target that you can still have a biologic activity in those organs," said Krieg.

"You can mask the charge to make it less negative and more hydrophobic," added Sepp-Lorenzino. "The issue with this is learning how to layer on all these chemical modifications while retaining activity."

Krieg agreed. "It is easier targeting outside of the liver with neutral backbone chemistries," he said. "They do not get sucked up as much by the liver, and you can put targeting ligands or peptides on them that will improve uptake in other tissues. Ultimately, I think the field will use different approaches for different targets and indications."

Maraganore said that with conjugate delivery systems, which use a ligand to deliver the RNA to cells expressing a specific receptor, "we are seeing target gene knockout in the liver at tissue concentrations that are below a microgram [of drug] per gram of tissue. The single-stranded approaches without a targeting [strategy] have to go to hundreds of micrograms per gram of tissue. They often get bystander organs, like the kidney, that have 10-fold higher levels of therapeutic in that tissue."

Krieg was not convinced that a delivery vehicle is necessary-he thinks the best chemistries are up to the task.

"One of the key lessons, perhaps, from the Prosensa failure in the DMD [Duchenne muscular dystrophy] trial may be selecting a chemistry that will give you the best therapeutic index. The particular chemistry they used was one of the earliest chemistries. It is significantly less potent-it has less binding affinity-compared with the newer chemistries," he said.

Prosensa Holding N.V.'s drisapersen antisense molecule failed in a Phase III trial in Duchenne muscular dystrophy in 2013.

"Even if you are targeting muscle, your levels in muscle are going to be very low proportional to your levels in liver and kidney. And so you are off-target from a tissue perspective. Accumulation of that xenobiotic is ultimately going to limit your dose effectiveness in the muscle," said Maraganore.

New chemistries are clearly in the works. Isis Pharmaceuticals Inc. recently published its generation 2.5 chemistry, which uses constrained ethyl bicyclic nucleic acids. For one test lncRNA, the new chemistry produced better target knockdown in various tissues than the 2.0 chemistry.10

Maraganore agreed that the chemistries have come a long way but said that getting the right tissue exposure is still a challenge. "The lower your tissue exposure for any given pharmacologic effect, the better your safety window is going to be. You cannot rely on simple backbone modifications to be a targeting mechanism."

Sepp-Lorenzino and Formela noted that oligonucleotides alone might not gain access to all tissues.

"Even for antisense and for single-stranded oligos, you can get into what we call permissible tissues, but delivery is extremely inefficient," said Sepp-Lorenzino.

She described two classes of targeting strategies that could help improve efficiency and expand the number of permissible tissues. "One is to encapsulate the nucleic acid into something else, mask the charge and that particle is what mediates biodistribution and uptake. The other, building on the conjugate approach, is to take your oligo and impart the drug-like properties that you will need to have-absorption, biodistribution and uptake."

The first strategy includes lipid-based delivery systems that many companies are pursuing (see "Delivery platforms"). Several lipid-formulated oligonucleotide therapeutics have advanced to clinical trials, but target tissues thus far are limited to the liver, blood cells and solid tumors.

According to Maraganore, the lipid nanoparticle (LNP) approach works nicely in the liver. "It is a rational mechanism of uptake. LNPs will recruit apolipoprotein E (APOE) to bind to them, and the APOE-bound particles are taken up by the low-density lipoprotein receptor (LDLR) on hepatocytes."

Improvements to the therapeutic index of molecules delivered via LNPs in hepatocytes have come from Alnylam and Arrowhead Research Corp.

Alnylam recently reported lipopeptide nanoparticles (LPNs) with improved efficacy over other lipid-based formulations, including LNPs. In nonhuman primates, 0.3 mg/kg siRNA delivered in LPNs led to a 95% reduction in TTR mRNA.11 By comparison, the same dose of the company's LNP-formulated clinical candidate patisiran produced an 80% reduction.

Lipid-based delivery also has limitations. "It will never work subcutaneously because the lipids are proinflammatory," said Maraganore. "So you have to use i.v. infusion."

He added, "Lipid-based approaches also have a relatively narrow therapeutic index. These molecules consistently show toxicities in primates and rodents at the single-digit milligram per kilogram level, so you have to operate within effective dose levels that give you a 5- to 10-fold margin. You have to have potent molecules."

"When we were optimizing LNPs, everything worked in rodents, and yet it was very difficult to get translation from the rodent to a primate," Sepp-Lorenzino said. "We ended up doing a lot of work in nonhuman primates. Then we went backward to try to understand what were the differences between the two systems."

"We had the same experience. The good news is when you optimized your formulation approach in the primate, it still worked in the rodent, so you did not lose the rodent as a good model," said Maraganore. "It was the lipid-to-siRNA ratio that ultimately was important."

Silence Therapeutics plc has developed an siRNA-lipoplex technology called AtuPLEX to deliver siRNA to tumor vasculature. The technology relies on synthetic cationic lipids with additional lipid components to improve multiple aspects of delivery. In nonhuman primates, 0.3 mg/kg Atu027-a lipoplexed siRNA targeting protein kinase N3 (PKN3)-was the minimal effective dose in target knockdown assays in lung and liver.12 Data from nonhuman primate studies in cancer have not yet been reported.

The second delivery strategy highlighted by Sepp-Lorenzino is the conjugate approach. This is exemplified by Alnylam's N-acetylgalactosamine (GalNAc)-siRNA conjugates that consist of siRNAs linked to GalNAc residues for hepatocyte-specific delivery in the liver. GalNAc conjugates are recognized and taken into the cell by asialoglycoprotein receptor 1 (ASGR1; CLEC4H1).

Arrowhead's dynamic polyconjugates (DPCs) combine the two delivery approaches, including nonlipid-based nanoparticles conjugated to a targeting ligand that delivers siRNA. Arrowhead's liver-targeting DPC incorporates a cholesterol-siRNA conjugate. In nonhuman primates, about 0.2 mg/kg siRNA formulated in DPC led to 99% knockdown of a target gene in the liver.

Given the impact that the GalNAc-siRNA conjugate approach has had on Alnylam's pipeline, Maraganore recommended that scientists discover comparable receptor-ligand pairs to expand the list of accessible tissues. "The future will be around novel receptor-ligand pairs in different parts of the body as a way of affecting delivery of oligos to those sites," he said. Importantly, he provided a list of attributes of the GalNAc-ASGR1 system that scientists should consider in their search for new ligand-receptor pairs for accessing additional tissues (see Box 1, "Attributes of the GalNAc-ASGR1 system").

Beyond the GalNAc-siRNA conjugate system, there was not a lot of enthusiasm for what can be achieved with conjugates. Thus, the panel discussed additional ways to increase the dose that reaches the target.

One strategy was to pursue local routes of administration tailored to the target tissue using, for example, topical application for the skin and eyes, intrathecal injection for the CNS, inhalation for the lung or intravitreal injection for the eyes.

"I think there is going to be more focus on local delivery approaches combined with techniques such as conjugates to get more specific cellular uptake and to avoid accumulation-based toxicities. That is where the delivery field is going at the moment," said Jackson.

Ex vivo

As an outside-the-box approach to solve the delivery hurdle, the panel briefly discussed the use of nucleic acids ex vivo.

"Obviously, there are many ex vivo applications that one could consider. The challenge from a business perspective is ultimately that it becomes cell therapy, or it becomes a different type of model, versus what all of us ideally like to focus on, which is a vial with drug that is used as the therapeutic," said Maraganore. "At Alnylam we spent a lot of time thinking about ex vivo, but frankly couldn't get too excited about it versus focusing on in vivo and making in vivo work."

Krieg agreed. "We can just go in vivo with the technologies that we have now. We've seen from the many antisense companies, Alnylam and other RNAi companies that these molecules work well in vivo in multiple tissues. We just want to go in vivo as quickly as possible."

The one exception Krieg noted was using ex vivo as an option to deliver a lncRNA as the therapeutic. "If we did decide that we wanted to use a lncRNA as a therapeutic itself, that would be much more difficult to deliver in vivo compared with an oligonucleotide. In that case, ex vivo therapy could make sense. Otherwise, the uptake of single-stranded oligonucleotides just by subcutaneous injection is good enough."

Despite the panel's muted enthusiasm for ex vivo approaches, several companies are pursuing the strategy.

Benitec Biopharma Ltd. has two RNA-based ex vivo therapeutic programs. The first, Cal-1, is partnered with Calimmune Inc. and is in Phase I/II testing for HIV/AIDS. Cal-1 is a lentiviral vector that expresses an shRNA to downregulate CC chemokine receptor 5 (CCR5; CD195)-the HIV receptor on T cells-and C46 to inhibit viral entry. After modification, the T cells should be resistant to viral infection and are reintroduced into the patient.

Benitec also is developing Dendrarna, IL-12 shRNA-expressing dendritic cells, in partnership with Medistem Inc. to treat rheumatoid arthritis (RA). Dendrarna is in preclinical testing.

Marina Biotech Inc.'s CEQ508 relies on ex vivo-modified Escherichia coli to deliver shRNA targeting b-catenin (CTNNB1) mRNA to the intestine. CEQ508 is in Phase Ib/IIa testing for familial adenomatous polyposis and has orphan designation from the FDA.

Apeiron Biologics AG is pursuing an ex vivo knockdown therapy for casitas B cell lymphoma-b (CBL-B). Their strategy is to use RNAi that encodes an E3 ubiquitin ligase ex vivo in blood cells to activate the antitumor activity of NK and T cells. The activated cells will then be returned to patients. The biotech hopes to start Phase I testing this year.13


In addition to using delivery to increase the diversity of accessible cell and tissue types, the panel said that delivery technologies can help increase the amount of drug that reaches intracellular target RNA. Indeed, panelists called for dedicated efforts to uncover ways of enabling escape of delivered therapeutics from endosomes or lysosomes.

"The evidence indicates that at least 95%, maybe 98%, of the administered oligonucleotide is in the endosomal-lysosomal compartment and does not get into the cytoplasm," said Krieg.

"The percentage of your dose that ends up binding to your transcript is tiny," added Sepp-Lorenzino.

In vitro studies have shown that for siRNAs delivered via LNPs, only 1%-2% of the molecules escape endosomes, and they preferentially escape from early, moderately acidic vesicles.14

Sepp-Lorenzino said that at least three approaches could improve efficiency. Two involve coadministration or conjugation of an agent that enhances cytosolic delivery. A third involves chemistry that shields the molecule's negative charge.

She highlighted Arrowhead's strategy as an example of how to improve endosomal escape by coadministration. Arrowhead's DPCs co-deliver a therapeutic siRNA along with a targeted, lytic polymer that breaks open intracellular vesicles.

For a conjugation approach, Sepp-Lorenzino highlighted ongoing work at Merck. "We have been looking at peptides that can enhance endosomal escape or cytosolic delivery. We have looked at peptides that are tolerated, that induce escape from the right compartment and do not cause uncontrolled complexation of siRNAs. I think it is a valid approach," she said.

She did not discuss specific outcomes from the work but did say that "some of these vesicles may be more amenable to letting go of their contents. If you go after the early endosome, maybe it will be safer than disrupting a lysosome, where you are going to have lysosomal enzymes causing damage."

There are multiple caveats to approaches that rely on additional agents to improve cytosolic delivery. "Every time you add something to your formulation, whether it is a coadministration or a conjugate, it adds complexity, whether it is manufacturing complexity or toxicological complexity," said Sepp-Lorenzino.

Maraganore added, "As soon as you start introducing peptides, I think you start involving aspects of immunity that might be limiting."

The final strategy Sepp-Lorenzino noted that companies are pursuing is reducing the charge on the RNA itself.

Krieg suggested some of this work should be undertaken collaboratively. "If we can succeed in [improving] that endosome release step, that would mean a huge increase in the potency of our oligonucleotides," he said. "I think overcoming the hurdle of endosomal escape for the RNA therapeutics community would be a perfect kind of precompetitive area for investigation."

The panelists agreed that supporting research to move science past a barrier that limits drug delivery for multiple therapeutic platforms should be a priority for funding agencies focused on advancing translational medicine. They highlighted efforts from scientists such as Marino Zerial, who is looking at cytosolic release of macromolecules, as an example of the type of work that will be needed.

Zerial, director of the Max Planck Institute of Molecular Cell Biology and Genetics who also has served as a scientific advisor to Alnylam and Merck, agreed that "cytosolic delivery is a major unsolved problem for any biologic."

He added that the problem has been completely neglected by funding agencies, and he is surprised by the success companies have had in such a short time given how little is known about the biological pathways that affect cytosolic delivery.

"We don't understand how this process works yet," he said. "It will take a collaborative effort from experts in multiple disciplines including chemistry, cell biology, physics and physiology."

Zerial said that any potential solution will need to be rooted in biology. In addition, he said that there is a fundamental disconnect between the assays most researchers rely upon and those that are needed. "The problem cannot be solved without biophysical techniques that provide a level of resolution that matches the biology. Simple light microscopy is not enough. The experiments that are needed require state-of-the art know-how and the best cell biological methods."

The panel acknowledged a need for both a step-by-step approach to the problem and new model systems.

"There are many differences between trafficking in cell culture and in vivo or even in primary cells," said Sepp-Lorenzino. "We are able to deliver whatever to HeLa cells. Can we say this is going to be very important for human disease? No; we are skipping 10,000 steps in between."

Raising the bar in preclinical research

The panel repeatedly emphasized the importance of robust validation, establishing both that modulating the target can affect disease and that a given therapeutic candidate acts specifically through the target (see Box 2, "Considerations for compelling target and therapeutic validation").

Although reproducibility issues affect every aspect of preclinical research,15,16 the RNA field needs to pay particularly close attention to target specificity, aptamer or immune-mediated effects and applying rigorous statistical analyses to resulting data.

"Our number-one priority is reproducibility," said Formela. "We always reproduce anything we are interested in."

For example, a large number of controls are needed to demonstrate sequence-specific, on-target activity. According to Corey, "You need experiments with mismatched and scrambled controls to show that it is not just some artifact of introducing something into cells. You also need to rule out effects like immune responses."

"You have to keep a very high index of suspicion for aptamer-like effects or immune-like effects of your nucleic acid," added Krieg.

"Not a lot is known about the mechanism of how these RNAs work. There is a lot of controversy out there. There are a lot of ideas. There are also a lot of papers that do not stand up very well to scrutiny," said Corey.

However, having the right people on board is crucial for successfully balancing the different perspectives.

Formela said, "Particularly in the field of oligonucleotides, companies start mostly with RNA people and not necessarily disease people. To really hit the sweet spot, you need to have the right mix of people who are agnostic to the modality but will understand why your particular modality might be very well fitted for their disease. Target selection and target-product profile early on will be the most important drivers of success."

Donner, A. SciBX 7(28); doi:10.1038/scibx.2014.815
Published online July 24, 2014


1.   Kotz, J. SciBX 5(45); doi:10.1038/scibx.2012.1176

2.   Donner, A. SciBX 6(41); doi:10.1038/scibx.2013.1151

3.   Wahlestedt, C. Nat. Rev. Drug Discov. 12, 433-446 (2013)

4.   Donner, A. SciBX 7(5); doi:10.1038/scibx.2014.136

5.   Reebye, V. et al. Hepatology 59, 216-227 (2014)

6.   Zhao, X. et al. Nat. Neurosci. 16, 1024-1031 (2013)

7.   Fishburn, C.S. SciBX 6(28); doi:10.1038/scibx.2013.711

8.   Yang, L. et al. Nature 500, 598-602 (2013)

9.   Cain, C. SciBX 6(34); doi:10.1038/scibx.2013.914

10. Hung, G. et al. Nucleic Acid Ther. 23, 369-378 (2013)

11. Dong, Y. et al. Proc. Natl. Acad. Sci. USA 111, 3955-3960 (2014)

12. Aleku, M. et al. Cancer Res. 68, 9788-9798 (2008)

13. Donner, A. SciBX 7(10); doi:10.1038/scibx.2014.274

14. Gilleron, J. et al. Nat. Biotechnol. 31, 638-646 (2013)

15. Haas, M.J. SciBX 7(3); doi:10.1038/scibx.2014.76

16. Fishburn, C.S. SciBX 7(10); doi:10.1038/scibx.2014.275


Alexion Pharmaceuticals Inc. (NASDAQ:ALXN), Cheshire, Conn.

Alnylam Pharmaceuticals Inc. (NASDAQ:ALNY), Cambridge, Mass.

Apeiron Biologics AG, Vienna, Austria

Arrowhead Research Corp. (NASDAQ:ARWR), Pasadena, Calif.

Atlas Venture, Cambridge, Mass.

Benitec Biopharma Ltd. (ASX:BLT), Sydney, New South Wales, Australia

Calimmune Inc., Tucson, Ariz.

City of Hope, Duarte, Calif.

Dicerna Pharmaceuticals Inc. (NASDAQ:DRNA), Watertown, Mass.

Genzyme Corp., Cambridge, Mass.

Harvard Medical School, Boston, Mass.

Howard Hughes Medical Institute, Chevy Chase, Md.

Isis Pharmaceuticals Inc. (NASDAQ:ISIS), Carlsbad, Calif.

The Johns Hopkins University School of Medicine, Baltimore, Md.

Life Technologies Corp., Carlsbad, Calif.

Marina Biotech Inc. (Pink:MRNA), Bothell, Wash.

Massachusetts General Hospital, Cambridge, Mass.

Max Planck Institute of Molecular Cell Biology and Genetics, Dresden, Germany

Medistem Inc., San Diego, Calif.

Merck & Co. Inc. (NYSE:MRK), Whitehouse Station, N.J.

MiNA Therapeutics Ltd., London, U.K.

miRagen Therapeutics Inc., Boulder, Colo.

Moderna Therapeutics Inc., Cambridge, Mass.

National Human Genome Research Institute, Bethesda, Md.

National Institutes of Health, Bethesda, Md.

Opko Health Inc. (NYSE:OPK; Tel Aviv:OPK), Miami, Fla.

Prosensa Holding N.V. (NASDAQ:RNA), Leiden, the Netherlands

RaNA Therapeutics Inc., Cambridge, Mass.

Sanofi (Euronext:SAN; NYSE:SNY), Paris, France

Sarepta Therapeutics Inc. (NASDAQ:SRPT), Cambridge, Mass.

Scripps Florida, Jupiter, Fla.

Silence Therapeutics plc (LSE:SLN), London, U.K.

University of California, San Diego, La Jolla, Calif.

The University of Texas MD Anderson Cancer Center, Houston, Texas

The University of Texas Southwestern Medical Center, Dallas, Texas