May 15, 2016 (Vol. 36, No. 10)
Richard A. A. Stein M.D., Ph.D.
RNAi Is Working on Tactics to Avoid siRNA Degradation and Improve Targeting and Delivery
Initially observed in plants and the nematode Caenorhabditis elegans, and subsequently in all major eukaryotic species, RNA interference (RNAi) has been recognized as a post-translational mechanism for the silencing of specific genes. RNAi is instigated by double-stranded RNA (dsRNA) molecules, and it exploits the base sequences of dsRNA, or rather the base sequences of the molecules derived from dsRNA, to silence genes in a sequence-specific manner.
RNAi evolved as a way to protect host genomes from parasitic nucleotide sequences, such as those arising from viral infections. But RNAi is not just a natural mechanism. It is also a contrivance, a research tool or, potentially, a therapeutic modality. The RNAi pathway provides a new framework to artificially introduce dsRNA into organisms to silence specific genes based on sequence complementarity.
Actually, these interventions may skip one of the natural pathway’s steps, the segmenting of long dsRNAs into short dsRNAs called short interfering RNAs (siRNAs). This segmenting process, which depends on an enzyme called Dicer, can be circumvented by introducing siRNAs directly. Then the siRNAs can be “unwound” by the enzyme Argonaute, resulting in two single-stranded RNAs (ssRNAs), a passenger strand and a guide strand. The passenger strand is degraded, but the guide strand is incorporated into the RNA-induced silencing complex (RISC). Making use of the guide strand’s base sequence, RISC can hybridize with mRNA that has a complementary base sequence. Ultimately, the target mRNA is subjected to endonucleolytic cleavage.
“Tricked Out” siRNA Termini
RNAi has been used to knockdown fusion proteins. “Because we were very limited in the sequence space that we could use, we explored alternate approaches to increase siRNA targeting potency,” says W. Mark Saltzman, Ph.D., professor of biomedical and chemical engineering at Yale University. In a recent effort to optimize the potency of siRNA targeting, Dr. Saltzman and colleagues modified the termini of two siRNAs that had been designed to target two distinct fusion oncogenes, BCR-ABL and TMPRSS2-ERG.
One of the optimization strategies involved a G-U wobble pair at the 5′ end of the siRNA guide strand to create a protruding functional group as a result of non-Watson-Crick base-pairing. This facilitates interaction with Argonaute, and is thought to enhance recognition and interaction with RISC.
The other approach involved the introduction, based on thermodynamic calculations, of a highly destabilizing terminal asymmetry at the 3′ end of the sense strand, to enhance guide-strand selection by the Argonaute protein.
“These approaches demonstrate the possibility of optimizing siRNA design when the choices for a sequence are limited,” asserts Dr. Saltzman.
The half-lives of the proteins encoded by the oncogenes are very different, approximately 40 hours for BCR-ABL, and 1 hour for TMPRSS2-ERG, and this experimental setting provided an excellent framework to better understand the way in which protein stability impacts the effectiveness of siRNA targeting.
“We found that getting substantial knockdown at the protein level using siRNA depends on the half-life of the protein in the cell,” reports Dr. Saltzman. Dr. Saltzman and colleagues revealed that even though the optimized targeting sequence at 1 nM led to a large decrease in BCR-ABL mRNA levels, no detectable changes in protein levels could be seen. This indicated that for proteins with very long half-lives, knocking down siRNAs for a period of several days might not have a significant effect on protein levels.
“This means that we need to accomplish not only a strong but also a very sustained protein knockdown in order to have an effect,” explains Dr. Saltzman. To improve the knockdown of long-lived cellular proteins, Dr. Saltzman’s team developed a nontoxic poly(lactic-coglycolic acid) nanoparticle delivery system that releases siRNA over several days and leads to more sustained cellular protein knockdown. Dr. Saltzman’s team also showed that in an in vitro system using leukemic cells, this approach can lead to robust cell death.
“Delivery remains a key consideration,” notes Dr. Saltzman. “Most of the translational work has used targets in the liver because this organ is the easiest to deliver siRNA to, but the challenge is to show that this strategy can work in other organs as well.”
“We are developing novel chemistry that improves siRNA targeting,” says Xianbin Yang, Ph.D., principal investigator and director of research and development at AM Biotech. Dr. Yang and colleagues hope to address low stability and suboptimal potency at target sites. These are two of the major challenges that impact the prospects of clinical translation in siRNA-based therapies.
To improve siRNA-mediated silencing, Dr. Yang and colleagues previously designed siRNA molecules containing phosphorodithioate (PS2) substitutions, in which the two nonbridging phosphate oxygen atoms are replaced by sulfur atoms. These modified siRNAs showed increased silencing, as compared to their wild-type counterparts, as a result of the increased serum stability of the siRNA molecules.
More recently, building on this initial work, Dr. Yang and colleagues incorporated an additional chemical modification, the addition of a 2′-O-methyl group on the PS2-modified siRNA molecule (MePS2). This modification, the investigators revealed, further enhanced silencing activity. As compared to the PS2 modification, MePS2-modified siRNAs exhibited increased resistance to degradation and stability, and a further improvement in silencing, which was explained by the increased siRNA loading on the RNA-induced silencing complex.
Dr. Yang and colleagues proceeded to validate this work in a mouse model of disease. They showed that their chemically modified siRNA was therapeutically helpful in targeting GRAMD1B, a protein involved in ovarian cancer chemoresistance.
Because siRNA molecules cannot cross cellular membranes, a strategy of critical importance for successful siRNA-based targeting in vivo is the use of aptamers for delivering siRNA. Developing the chemistry of aptamer’s affinity is a fundamental determinant of the diagnostic and therapeutic utility of aptamer-based siRNA delivery strategies.
“We can currently achieve aptamer binding in the picomolar range,” declares Dr. Yang. “Recently, we solved the co-crystal structure to support the reasoning behind the increase in affinity.”
Lipid Nanoparticle siRNA Delivery
“Using siRNA technologies, we have demonstrated, as a proof of concept, that we can protect nonhuman primates that had been infected with Ebola virus, if the therapeutic intervention is performed three days after exposure,” says Thomas W. Geisbert, Ph.D., professor of microbiology and immunology at The University of Texas Medical Branch at Galveston.
The Makona strain of the Ebola virus was isolated in the most recent outbreak that started in 2014 in Guinea. Using lipid-nanoparticle-encapsulated siRNAs, Dr. Geisbert and colleagues targeted this strain and showed that, when initiated three days after exposure, the intervention was able to protect 100% of the viremic and clinically ill rhesus monkeys. Animals that were treated developed only mild clinical signs and symptoms, and showed reduced mortality and decreased viremic loads.
“We want to know how far we can wait to administer this treatment, and at what stage of the disease intervention is too late,” explains Dr. Geisbert. For this work, Dr. Geisbert and colleagues used siRNA-Ebola3, a cocktail that contains siVP35-3 and siLpol-3, both of which are active against the virus. The FDA has cleared the use of this siRNA cocktail encapsulated in lipid nanoparticles in Ebola virus-infected patients, and the approach is currently being evaluated for use in infected patients in Sierra Leone.
“We would also like to use this technology in combination with other technologies,” informs Dr. Geisbert. An important lesson from the HIV therapeutics field is the strength of combining multiple drugs, an approach that can target distinct facets of the viral lifecycle while reducing the risk of adverse effects.
“This provided a multipronged approach to hit the virus from different angles and exert a synergistic effect,” continues Dr. Geisbert. “We are looking at whether we can use a combination between siRNA and other approaches, such as monoclonal antibodies, to attack the virus at multiple vulnerable points to obtain enhanced benefits.”
Although antibodies have been historically used in antiviral therapeutics, generating them can be time-consuming and, particularly for epidemics caused by newly emerging or re-emerging viruses, delays stemming from optimization, manufacturing, and scale-up may be critical. One of the promising aspects of siRNA technologies is the possibility of developing medical therapies as soon as a viral sequence is known.
“Scale-up is always going to be an important problem,” advises Dr. Geisbert. “But the technology can be adjusted relatively quickly.”
“From the standpoint of their clinical use, siRNAs have seen both positive and negative developments over the past decade,” says Marco S. Weinberg, Ph.D., assistant professor of molecular and experimental medicine at the Scripps Research Institute. Because siRNAs have to be delivered into every target cell, where they act to suppress all RNA molecule copies, siRNA-based silencing has been more challenging in acute as compared to chronic infections.
“The burden of infection is so much higher in acute infections. As a result, the stoichiometry necessary to suppress all RNA molecules is not achievable,” explains Dr. Weinberg. “Nonetheless, chronic infections, in which viruses are at a reasonable copy number and are limited to specific tissues, represent an area of tremendous potential for this technology.”
One of the major challenges is that siRNAs cannot be delivered into cells as naked RNAs. In an attempt to overcome this challenge, Dr. Weinberg and colleagues entered into a collaboration with John Rossi, Ph.D., a molecular biologist at City of Hope. Together, these researchers developed aptamers that can help siRNAs become internalized into cells.
Among several aptamer candidates that Dr. Weinberg and colleagues examined, one inhibited HIV-1 replication with a half-maximal inhibitory concentration in the nanomolar range. In addition to delivering the siRNA intracellularly, the aptamer also blocked the CCR5 receptor needed for the cellular entry of the virus, thus providing a dual inhibitory mechanism.
Previous work from Dr. Rossi’s laboratory combined the use of anti-HIV gp120 aptamers and anti-HIV siRNAs to selectively deliver siRNAs to HIV-infected cells and inhibit viral entry. “The approach in the current study is different, in that we can additionally target cells that are still naïve to the infection,” explains Dr. Weinberg. “This provides, therefore, a therapeutic as well as a prophylactic approach.”
One of the key differences between DNA and RNA viruses is that RNA viruses lack proofreading. As a result, they are susceptible to a massive amount of random mutations that could compromise the efficiency of silencing. “That is why silencing approaches have to be very carefully designed when targeting such as virus,” notes Dr. Weinberg.
On the other hand, unlike for small molecule drugs, siRNA-based targeting presents the advantage that it is governed by the same mechanistic principles irrespective of the protein that is being ultimately targeted. “Another advantage is that multiple complexes can be targeted together to prevent the virus from escaping from the drug,” continues Dr. Weinberg.
An additional obstacle in using siRNA therapeutics for viral infections is that certain viruses are able to establish a latent state after internalization. “So far we have not succeeded in addressing latency with siRNAs or with other drugs, and HIV is no exception,” admits Dr. Weinberg. “Once the virus is latent, siRNA is useless.”
This opens the need to continuously treat an infection for the lifetime of a patient. “Solutions that resolve challenges surrounding viral latency would not necessarily fall into the ballpark of what siRNA can do at this stage,” concludes Dr. Weinberg.