Portal de Periódicos da CAPES

RNA-Based Therapeutics: Ready for Delivery?

2009; Cell Press; Volume: 136; Issue: 4 Linguagem: Inglês

10.1016/j.cell.2009.02.010

1097-4172

Laura Bonetta,

CRISPR and Genetic Engineering

Delivery of RNA-based therapeutics to specific tissues for treating a variety of diseases faces many hurdles. But with several clinical trials and a slew of studies in animal models underway, the future of RNA-based therapeutics seems bright. Delivery of RNA-based therapeutics to specific tissues for treating a variety of diseases faces many hurdles. But with several clinical trials and a slew of studies in animal models underway, the future of RNA-based therapeutics seems bright. In 1998, the US Food and Drug Administration (FDA) approved the first RNA-based drug for the treatment of a human disease. The drug, fomivirsen, is an antisense oligonucleotide (oligo) that blocks synthesis of a key protein of cytomegalovirus. It is used to treat an inflammation of the eye caused by the virus. Ten years later, although only one other RNA-based drug (an RNA aptamer) has made it to market, the arsenal of potential molecules and approaches continues to expand. It now includes a new generation of antisense oligos, aptamers, ribozymes, RNA decoys, splice-site targeted oligos, small-interfering RNAs (siRNAs), short hairpin RNAs (shRNAs), and microRNAs (miRNAs) (see Review by T.A. Cooper, L. Wan, and G. Dreyfuss in this issue of Cell). Although they all target RNA, these compounds work by distinct mechanisms. However, they share similar hurdles to clinical application—the biggest of which is efficient delivery to the desired tissue. Antisense oligos like fomivirsen work by binding to a complementary sequence in an mRNA, thus preventing production of the protein encoded by that mRNA. When antisense oligos were first introduced as possible therapies, they were chemically modified to make the RNA more stable. "In spite of modifications the oligos still were not very effective [in knocking down gene expression] and had other unexpected effects," says Ryszard Kole, senior vice president at AVI Biopharma (Corvallis, OR). However, advances in RNA chemistry have led to "second-generation" antisense oligos with different modifications and improved activities. Fomivirsen's developer, Isis Pharmaceuticals (Carlsbad, CA), in collaboration with Genzyme (Cambridge, MA), recently launched one phase II and three phase III clinical trials of mipomersen. This second-generation antisense RNA drug reduces the production of apoB-100 (a protein critical for the synthesis and transport of "bad cholesterol") in patients with hypercholesterolemia. In addition, Isis has several other compounds in phase I trials including an antisense drug that targets C-reactive protein implicated in rheumatoid arthritis. According to Isis president, Stanley Crooke, many of the issues that plagued the original antisense drugs are no longer major challenges. "The oligos we use can be injected in the bloodstream and delivered to several tissues where they can persist for 14 to 30 days, long enough to see a therapeutic response," he says. However, not all tissues are amenable to delivery. "These oligos don't cross the blood-brain barrier, they don't get into the heart and don't get into muscle," says Crooke. When administered subcutaneously, the Isis second-generation antisense drugs become localized in a variety of tissues including kidney, liver, and fat. As a result, clinical trials have focused primarily on targeting diseases that involve systemic delivery to these tissues or, alternatively, tissues that can be targeted by local oligo delivery (such as the eye or lung). Several ongoing clinical trials are testing RNA oligos that work by RNA interference (RNAi). These oligos are short double-stranded pieces of RNA, called siRNAs, that use the cellular machinery—and in particular the RNA-induced silencing complex (RISC)—to degrade specific mRNAs before they can be translated into proteins. In December 2008, Opko Health (Miami, FL) announced that it had completed enrollment of a phase II trial of the siRNA bevasiranib to treat age-related ("wet") macular degeneration (AMD) caused by the abnormal growth of blood vessels behind the retina leading to vision loss. Bevasiranib targets the mRNA for vascular endothelial growth factor (VEGF), a protein that promotes blood vessel growth. The trial came under scrutiny when Jayakrishna Ambati at the University of Kentucky reported last year that many siRNAs, including bevasiranib, suppress blood vessel formation by activating an immune receptor called TLR3, suggesting that bevasiranib's effect might be, at least in part, nonspecific. Other companies are also conducting clinical trials of siRNAs for treating AMD. Interestingly, the only other FDA-approved RNA-based drug is pegaptanib, an RNA oligo (aptamer) that blocks VEGF activity. Although approved for treating AMD, pegaptanib suffered from competition from a monoclonal antibody against VEGF made by Genentech. But RNA therapeutics are also targeting tissues beyond the eye. Alnylam Pharmaceuticals (Cambridge, MA) initiated a phase II trial of an inhaled siRNA that targets lung epithelial cells to treat respiratory disease caused by syncytial virus infection. The company has two other drugs expected to go into clinical trials soon that will test systemic delivery of siRNAs to treat high blood cholesterol and liver tumors. "In a very short time, Alnylam has shown that you could have systemic delivery, first in mice, then in monkeys, and the FDA just cleared the path for human trials," says Thomas Tuschl, at The Rockefeller University and a member of Alnylam's scientific advisory board. Tuschl acknowledges that delivery of siRNAs to certain tissues is still a challenge. "Many studies are focusing on the mechanism of endocytosis, how RNAs either alone or coupled with other molecules are taken up by cells," he says. "Another interesting development is that you can force single-stranded RNA to go into RISC and act like siRNA. There are studies emerging to indicate that using single-stranded RNA may have some advantages." Another class of small RNAs called miRNAs are produced inside cells. Each miRNA is thought to fine tune the expression of several mammalian genes. By using antisense oligos (called antagomirs) that target specific miRNAs, "we can regulate whole pathways or networks of genes. It is a whole different way of making drugs," says Peter Linsley, CSO of Regulus Therapeutics, a joint venture between Isis and Alnylam. A big challenge in developing miRNA-based drugs is to identify which of the more than 8000 known miRNAs (http://microrna.sanger.ac.uk/sequences) to target, but there are some good candidates. In December 2008, German scientists in collaboration with Regulus reported that the miRNA miR-21 is overexpressed in human heart and that targeting this miRNA with an antagomir prevented heart failure in mouse models of heart disease. Another focus of the company is miR-122, an miRNA essential for the replication of hepatitis C virus. "The virus has less of a chance to develop resistance to a miRNA-based drug since the site targeted by this miRNA has been conserved. It should not contain residues that mutate readily," says Linsley. Meanwhile, John Rossi's group at City of Hope in Duarte, California and Benitec Inc. (Melbourne, Australia) are tackling HIV using another small RNA, an shRNA that silences gene expression through RNAi. Rossi and his team have engineered a viral vector to carry this shRNA and deliver it to hematopoietic stem cells where it is incorporated into DNA. The shRNA is expressed forming a double-stranded hairpin structure that is cleaved by the cellular machinery into an siRNA. Rossi's shRNA targets two critical HIV genes; meanwhile, the viral vector transporting the shRNA contains an RNA decoy that sequesters an HIV protein necessary for transcription of the viral genome. The viral vector also carries an RNA enzyme, called a ribozyme, which inactivates a cellular protein needed for HIV infection. Although the ribozyme and RNA decoy are not as powerful as RNAi, Rossi says his group "did not want to use three different shRNAs because we did not want to overwhelm the cell machinery for RNAi. Instead, we used three different approaches." As Rossi starts to use his triple-punch therapy in a pilot feasibility study in four HIV patients, one question is whether a cell's RNAi system can be so overburdened that it eventually shuts down. A challenge with shRNAs, as Mark Kay, director of human gene therapy at Stanford University, discovered, is that they can be toxic to cells when they are overexpressed. The solution is to use weak promoters to drive sufficient (but not too powerful) shRNA expression for a therapeutic effect. "If you are trying to target cholesterol or lipid, a 50%–80% reduction in gene expression is a great therapeutic effect," says Kay. To circumvent the problem of toxicity with shRNAs, gene therapy researcher Beverly Davidson at the University of Iowa placed the shRNAs into artificial miRNA-based expression systems, before insertion into an adenovirus-associated viral (AAV) vector. AAV has a natural attraction for neurons, and Davidson has used it to deliver shRNAs targeting the huntingtin gene (mutated in Huntington's disease) into the brains of mice with this neurodegenerative disorder. Her group, having shown that RNAi is therapeutic in two distinct mouse models of Huntington's disease, is now gearing up for studies in nonhuman primates. Phillip Zamore at the University of Massachusetts Medical School is also attempting to treat Huntington's disease with RNAi but without viral vectors. In 2007, Zamore and his collaborator Neil Aronin showed that injecting mice with Huntington's disease with an siRNA targeting huntingtin improved disease symptoms. He is now developing siRNAs that specifically target the huntingtin allele carrying the mutation but not the normal allele. Zamore says that there is at least one polymorphism in the huntingtin gene that segregates with the mutation in a majority of Huntington's disease patients. Thus, it may be possible to design an RNA therapeutic that targets the mutant allele of huntingtin in a large number of Huntington's patients, rather than having to design a different siRNA molecule for every patient. The siRNA and shRNA strategies each have their own advantages and disadvantages. For example, siRNA treatment can be stopped at any time, but shRNA delivered with a viral vector remains in the tissue for the lifetime of the patient. But without the vector, delivery can be more challenging. "It is not clear that siRNAs are taken up by neurons or other cells in the brain, such as oligodendrocytes or astrocytes. So they may not be getting to the cells needed for therapy," says Davidson. "Alternatively, we may find that therapeutic benefit requires targeting all cell types. Until more studies are done in rodent models and non-human primates, we cannot know if that is needed or possible." "There are two considerations when choosing a target for RNA therapies," says Rossi. "The first is the disease. Are there no other treatments for it? And AMD is a good example. The other consideration is can we get the RNA triggers to the tissues affected by the disease? The liver, for example, is a good target, while tissues like the pancreas are more difficult." That does not mean that tissues for which delivery is more challenging cannot be accessed. Swamy Manjunath at Harvard Medical School has used a short peptide derived from the rabies virus glycoprotein (which binds specifically to the acetylcholine receptor of neurons) to deliver siRNAs to mouse brain after intravenous administration. Once inside nerve cells, the siRNAs target Japanese encephalitis virus preventing the virus from causing fatal encephalitis in these animals. Studies like these are showing promising results, but delivery continues to be a problem for the actual development of drugs that will be broadly beneficial to patients and that are of interest to large pharmaceutical companies. "I do see that delivery will continue to improve and the safety of the delivery systems will continue to improve, but right now we only see incremental changes. We would like to see a breakthrough," says Alan Sachs, vice president of RNA therapeutics at Merck Co., Inc. (Whitehouse Station, NJ). In 2006, Merck acquired Sirna Therapeutics making a strong commitment to pursue the development of RNA-based therapies. "We know the technology works quite well. The remarkable thing about siRNA is that it takes advantage of a natural mechanism to provide a very potent way of inhibiting RNA expression," says Sachs. "But the question still is whether we can deliver this drug safely." Delivery systems for siRNAs include peptides, polymers, and lipids bound to the RNA molecule, and each is associated with some toxicity. "Every drug has some toxicity at high enough doses," explains Sachs. "We have to be able to develop compounds that can deliver siRNAs in sufficient amounts to have a high knockdown effect, but with a sufficient margin of safety." Compared to delivery, other issues that need to be considered when designing siRNA-based therapies do not represent major challenges. "Delivery is by far the biggest problem," says Zamore. "The other potential problems are vastly overstated." One concern is off-target effects, whereby a particular siRNA will bind to mRNA targets other than the one it was originally designed against. But such nonspecific effects can be controlled by adding chemical modifications at certain positions in the siRNA. "Off-target effects are not as much of a problem today," says Linsley. "To some extent you can predict what the off-target effects might be for a particular siRNA but they differ between primates and mice. You never really know what the problems might be until you go into humans." Similarly, the immune response caused by siRNAs reported in some studies can be controlled by chemically modifying the siRNA. However, stimulating the immune response may be, in some cases, beneficial. "We use all the capabilities of an RNA molecule for functional activities," says Gunther Hartmann of the University of Bonn, Germany. Hartmann has shown that an siRNA could decrease the expression of the Bcl2 gene, which is highly expressed in cancer cells and protects cells from undergoing apoptosis. By adding a triphosphate to the 5′ end of this siRNA, Hartmann made this molecule look like a virus to the immune system. The immune response against the siRNA worked synergistically with its silencing effect on Bcl2 to provoke large-scale death of lung tumor cells in a mouse model of metastatic cancer. "Effective tumor therapy requires different activities so people think combination therapy is the way to go," says Hartmann. "We have combined two activities in one molecule." One application for which siRNAs cannot be used, but synthetic oligos are proving potentially valuable, is the modification of splicing to generate a new mRNA variant. Oligos used in this strategy are chemically modified so that they are not recognized by the cellular RNAi or degradation machinery. AVI BioPharma's Kole uses morpholinos—oligos with standard nucleic acid bases but bound to morpholine rings instead of deoxyribose or ribose sugars and linked through phosphorodiamidate groups instead of phosphates—coupled to arginine-rich cell-penetrating peptides in a variety of disease models. Kole has used these peptide-linked morpholinos to restore hemoglobin expression in a mouse model of the genetic blood disease, thalassemia. Mutations that cause thalassemia induce abnormal splicing of the β-globin gene encoding a hemoglobin subunit. A morpholino that blocks an aberrant splice site in a mutant form of β-globin results in production of the correctly spliced β-globin mRNA. Modifying aberrant splicing potentially could correct the defect in several other genetic diseases that cannot be reversed by gene therapy. One such disease is Duchenne Muscular Dystrophy (DMD) caused by mutations in a 2.4 Mb gene encoding the protein dystrophin, a critical structural component of muscle cells. The mutations result in prematurely truncated, nonfunctional dystrophins. Annemieke Aartsma-Rus and colleagues at Leiden University use 2′-O-methyl-modified ribose oligos that cause the skipping of certain exons in the dystrophin mRNA to restore a normal reading frame. This strategy results in the production of dystrophin proteins that are missing an internal exon but are partly functional, leading to a less severe phenotype. This group has published a study in four DMD patients showing that injection of a 2′-O-methyl-modified ribose oligo into muscle restored dystrophin synthesis in that muscle. The Dutch company Prosensa Therapeutics is now conducting a clinical trial to test systemic delivery of the same modified oligo in DMD patients. "The muscles are leaky because they do not produce dystrophin so the uptake is much higher than in normal muscle. So the disease is helping us with delivery," says Aartsma-Rus. "In mice we get enough oligos in muscle to have a beneficial effect." In January 2009, the MDEX Consortium led by Francesco Muntoni at University College London, UK announced the results of a phase I clinical trial using a morpholino oligo to induce dystrophin production by exon skipping in the foot muscles of four boys with DMD. The group has now started a trial of systemic delivery of the same oligo to DMD patients. Because bacteria do not have a mechanism for RNAi, siRNAs cannot be used to silence bacterial genes. However, Nobel Laureate Sidney Altman at Yale University has developed a strategy using peptide-linked morpholinos to inactivate bacterial genes. These oligos hybridize to the bacterial mRNA of choice to form specific complexes that then induce degradation of the target mRNA by the bacterial enzyme RNase P. "We have shown we can do it with Bacillus anthracis, Yersinia pestis, and other pathogenic bacteria," says Altman. "We can inactivate them very quickly." A newer, and still somewhat controversial, RNA-based molecule with possible therapeutic applications is antigene RNA (agRNA). AgRNAs are short double-stranded RNA molecules that do not target mRNAs but rather target gene promoters. These molecules either activate or inhibit gene expression in mammalian cells through an unknown mechanism. However, David Corey, at the University of Texas Southwestern Medical Center in Dallas, has shown that agRNAs specific for the progesterone receptor gene bind to antisense transcripts that overlap with the promoter for this gene. These agRNAs recruit proteins and trigger formation of a large protein complex that includes transcription factors close to the progesterone receptor gene promoter. Somehow, having a bulky complex so close to the promoter has an effect on expression of this gene. "The most obvious application of agRNAs is that you can use them to upregulate gene expression. There is currently no general way to do that with standard RNA technologies," says Corey. Understanding how RNA-based therapies work is key if these approaches are to eventually succeed in the clinic. Results of several ongoing clinical trials will show whether RNA-based therapeutics are set to become major players in the treatment of disease. "I think they will play a huge role in the therapeutic arena, but people have to be patient. If we proceed on a reasonable path, the payoff could be great," says Kay. "People have to realize that it takes time, but it is not from a lack of enthusiasm."