January 15, 2008 (Vol. 28, No. 2)
A Cornucopia of siRNA- and RNAi-based Therapeutics in Development Show Promise
Researchers from academia and industry will present the latest developments in the application of interference RNA (RNAi) as a tool for gene silencing at “RNAi2008 Functions and Applications of Non-coding RNAs” to be held in March in Oxford, U.K. The meeting will focus on advances in the field and current opportunities and challenges in developing RNAi-based strategies for use in understanding the molecular mechanisms underlying disease processes, identifying new drug targets, and developing therapeutic agents capable of silencing disease-related genes.
siRNA Therapeutics
“Refined Animal Models for Optimizing Delivery of Functional siRNAs to Skin,” is the title of the presentation to be given by Roger Kaspar, Ph.D., on behalf of TransDerm (www.transderm.org). The skin disorder on which the company is focusing its initial product development effort is pachyonychia congenital (PC), a rare autosomal dominant disease in which a mutation in the gene for keratin 6a (K6a) causes painful skin lesions to form. The K6a N171 mutation is a single-nucleotide replacement mutation in which an adenine is present in the mutant gene form in place of a cytosine, resulting in an amino acid change.
Dr. Kaspar, CEO of TransDerm, will describe how the company’s TD101 therapeutic small interfering RNA (siRNA) specifically targets the N171K mutant form of the gene without affecting the wild-type gene.
To assess the activity of TD101 and optimize a controlled-dose delivery system, TranDerm is developing transgenic mouse models. In collaboration with Christopher Contag, Ph.D., and colleagues in the molecular imaging program at Stanford University, the company is using molecular imaging in transgenic mice to demonstrate the effectiveness of siRNA knockdown of reporter genes.
Two in vivo imaging-based approaches for validating RNAi activity have yielded quantitative and qualitative evidence of gene expression inhibition. In one method, the researchers intradermally coinjected a plasmid expressing the firefly reporter gene luciferase with siRNA molecules targeting the reporter gene mRNA into mouse paws. The results demonstrated potent inhibition of luciferase expression. Northern blot analysis supported luciferase mRNA inhibition as the mechanism of action.
The second method involved the construction of a mutant K6a/luciferase bicistronic reporter construct that was codelivered into mouse skin with K6a mutant-specific siRNAs. In vivo bioluminescence-based imaging showed weak luciferase activity in the coinjected mouse skin compared to control animals.
The company plans to use human skin explants derived from PC patient biopsies grafted onto immunocompromised, nude mice as a test stage to determine whether the RNAi activity seen in mouse skin will translate to similar efficacy in human skin samples. This work, being carried out with collaborators at the Ciemat Institute in Madrid, Spain, is still in the early stages.
TransDerm’s TD101 product is an unmodified siRNA intended to be administered directly to the PC lesions. Although intradermal injection offers a reliable delivery method, it is not a particularly patient-friendly option, especially given the need for repeated administration of the drug to treat new lesions. TransDerm is experimenting with two alternative siRNA delivery methods.
One is a topical, lipid-based formulation called GeneCreme. Its main limitation at present is how to ensure reliable uptake and dosing. The other approach is called the Soluble Tip Microneedle Array. Each array is composed of a grid of dissolvable, hollow protrusions into which siRNAs can be loaded. The microneedles penetrate the outer layer of the skin, the stratum corneum, where the tips are dislodged and remain. Time-release dispersion of the siRNA occurs as the tips dissolve.
“An ideal application would be to embed the array in a type of Band-aid device that the patient would apply and push down on to deliver the drug,” says Dr. Kaspar.
siRNA drugs intended for systemic administration require chemical modification or some sort of protective packaging to prevent rapid enzymatic degradation in the bloodstream, which would compromise their ability to reach the intended target and to exert a therapeutic effect.
Jørgen Kjems, Ph.D., a professor in the department of molecular biology at Aarhus University in Denmark, part of the Interdisciplinary Nanoscience Center (iNANO; www.inano.dk), is experimenting with three-stranded siRNAs as a means of optimizing the ability to introduce chemical modifications without changing the molecules’ activity.
Dr. Kjems begins with a small, double-stranded RNA, keeps one strand whole, and cleaves the second nonfunctional strand into two pieces. The result is a short internally segmented interfering RNA (sisiRNA). His group has demonstrated that these three-stranded constructs are amenable to a greater number of chemical modifications. With a double-stranded RNAi molecule about 20% of the nucleotides can be modified without losing activity, 100% of the nucleotides in an sisiRNA molecule can undergo chemical modification, allowing for more options to improve their pharmacodynamic properties.
This finding created an opportunity for evaluating a large variety of chemical modifications. Dr. Kjems has shown that the difference between an unmodified siRNA and a fully modified sisiRNA may be as great as a 100-fold increase in stability in serum.
This triple-stranded approach offers an additional benefit related to RNAi uptake and activity. When an siRNA is taken up by a cell, an intracellular complex selects one of the two RNA strands to retain. With the sisiRNA, the uncleaved strand—the active strand—is preferentially retained.
Delivering the Goods
In addition to developing a range of chemical modification strategies for improving in vivo delivery and therapeutic activity of the sisiRNAs, Dr. Kjems group is experimenting with methods of generating double-stranded siRNAs that can mimic three-stranded constructs, which enable chemical modification of every nucleotide.
Another approach being pursued in Dr. Kjems’ laboratory is the development of polymer-based siRNA encapsulation approaches. By mixing anionic siRNAs with chitosan—a natural, cationic, fully biodegradable polymer—they have been able to produce siRNA/chitosan particles 100–200 nm in size and have demonstrated knockdown of green fluorescent protein expression in the lungs of green mice carrying this gene.
Ongoing experiments in which these particles are injected into mice with rheumatoid arthritis show that the particles are attracted to the animals’ inflamed joints, where they deposit their therapeutic cargo. Other projects to evaluate particle-based siRNA delivery are focusing on knockdown of influenza virus, in collaboration with the Max Planck Institute, and on tumor inhibition.
RXi Pharmaceuticals (www.rxipharma.com) is developing RNAi therapeutics that will initially target neurological disorders (amyotrophic lateral sclerosis), metabolic diseases (obesity and type 2 diabetes), and cancer. The company is currently in a quiet period during preparations for a share distribution.
An agreement with TriLink Biotechnologies grants RXi an exclusive license to three RNAi chemistry technologies for use in developing its rxRNA™ compounds. RXi contends that its rxRNAs are up to 100 times more active than conventional siRNAs and are nuclease resistant. The licensed technologies include a patented preactivated carbonyl linker for coupling carrier molecules to RNA molecules during solid-phase synthesis to enable targeted delivery of an RNAi drug. The company also acquired the rights to patent-pending chimeric RNA-DNA duplexes, which are RNAi-like compounds composed of an RNA antisense strand and a modified DNA sense strand. When combined to form rxRNAs, these duplexes offer potential advantages in target specificity.
Tariq Rana, Ph.D., a founding scientist at RXi, developed nanotransporters designed to deliver RNAi compounds to target tissues. Dr. Rana and his colleagues published examples of successful delivery of RNAi molecules to the mouse liver and the ability to achieve specific gene knockdown with a dose of 1 mg RNAi/kg body weight, with no apparent immune stimulation.
Matthew Wood, Ph.D., university lecturer, and colleagues at the University of Oxford, U.K., are exploring the complexities of using RNAi to silence genes implicated in two neurodegenerative disorders, spinocerebellar ataxia and Parkinson’s disease. Both have a genetic component and a dominant pattern of inheritance (in some cases of Parkinson’s), making them good candidates for gene silencing to suppress expression of the mutant form of a gene without inhibiting the normal gene. If successful, this approach would leave a patient with, typically, sufficient normal protein levels.
Based on research that determined the optimal placement of a nucleotide mismatch in a short, 19 nucleotide siRNA molecule to be around position 16, in order to achieve maximum discrimination between a normal gene and the mutant form containing a single point mutation, Dr. Wood’s group created a series of short hairpin RNAs (shRNA) in which single nucleotide mismatch was placed stepwise at positions 9 through 17 in order to understand how to optimize shRNA design against mutant target genes. Using a luciferase assay, they demonstrated gene silencing and the ability of the siRNA to discriminate between the mutant and wild-type allele.
The researchers designed shRNAs carried in expression vectors to determine whether the results of experiments using siRNA molecules would be predictive of what occurs with full-length RNAs, as these expression vectors are ultimately what would be used to deliver an RNAi drug to patients. They were able to validate the effectiveness of the RNAi mechanism, induced by shRNAs targeting a point mutation, to silence the mutant gene and discriminate the target gene from the normal allele.
The mutation that causes SA-7, one form of spinocerebellar ataxia, is not a point mutation; it is a multinucleotide expansion, a repeated sequence similar to the mutation that causes Huntington’s disease, that would be difficult to target using RNAi.
In about 70% of patients, however, the presence of this mutation is associated with an SNP, and Dr. Wood’s group demonstrated that the mutated gene can be silenced by targeting this SNP. They screened a large collection of siRNA molecules with the target gene linked to luciferase to identify the one best able to silence the SA-7 gene and found that a short RNA sequence with a nucleotide complementary to the SNP situated at position 15 or 16 was most effective.
They went on to confirm this effect using shRNA expression vectors targeting the full-length gene. When they designed a dual reporter system that allowed them to study the effect of gene silencing when both the SA-7 mutant gene and the wild-type gene were expressed in the same cell, they found greatly enhanced gene discrimination compared to the results of assays in cells containing only the mutated or wild-type allele alone.
The results in cell-based assays were striking. Presence of the SA-7 gene in a cell causes the formation of protein aggregates, and when the SA-7 and wild-type genes are present, even normal proteins will be drawn into these aggregates. Silencing of the mutant gene eliminates aggregate formation, and the normal protein is able to distribute and function appropriately in the cell.
The next advance will be to validate the safety and efficacy of this RNAi in mouse models of SA-7. This disease is somewhat unique among neurodegenerative disorders in that it affects not only cells in the brain but also in the eye, causing retinal degeneration. The ability to deliver the drug to the eye and to study its effects on the disease phenotype in an easily accessible organ will facilitate preclinical and clinical testing. Dr. Wood hopes to take this treatment into clinical studies within 12 to 18 months.
Genome-wide Screening
Christopher Lord, Ph.D., staff scientist, and colleagues at the London-based Institute for Cancer Research, are using RNAi technology and high-throughput RNAi screens for target identification and validation in cancer drug discovery. RNAi screens have particular use for identifying drug targets in defined subsets of patients, according to Dr. Lord.
As an example, he describes his work in the identification of therapeutic targets that may be useful in women with breast or ovarian cancer who bear mutations in either the BRCA1 or BRCA2 genes. In 2005, Dr. Lord and colleagues used RNAi technology to demonstrate that cells with mutations in either BRCA1 or BRCA2 are much more sensitive to RNAi knockdown of a DNA repair enzyme, PARP, than are normal cells. This work led directly to a clinical trial of PARP inhibitors in patients.
“The PARP inhibitors are now in a Phase II trial,” says Dr. Lord. “We went very quickly from RNA inhibition, to small molecule inhibitors, to drug-like molecules, to trial.”
The use of siRNA to perform genome-wide screens for target identification and validation has advantages and disadvantages, according to Dr. Lord. RNAi is an attractive tool because it works well in a large variety of cell types and has the potential to knockdown every gene with a relatively high degree of selectivity. Furthermore, commercially available siRNA libraries have made the technology more accessible and user friendly.
Limitations to the application of siRNA screens in drug discovery include the cost of the libraries and the relatively short duration of silencing with transfected siRNA.
An alternative technique for performing gene silencing screens relies on plasmid or viral vectors that deliver shRNA precursors that are processed into siRNA molecules inside a cell. At present, available shRNA libraries are not optimally robust, in Dr. Lord’s view; their potential for recombination compromises their usefulness.
Opportunities for the future include generating more robust genome-wide plasmid libraries and lowering the cost of siRNA libraries. Dr. Lord also describes the advantages to be gained by combining plasmid-based RNAi screens with next-generation sequencing technologies such as Illumina’s Solexa or Applied Biosystems’ SOLiD sequencing systems. This approach could help solve the problems of both price and speed in genome-wide screens.
In her talk, Gwen Fewell, Ph.D., product manager for RNAi at Open Biosystems (www.openbiosystems.com), will describe how the company’s Expression Arrest™ microRNA-adapted shRNA (shRNAmir) libraries overcome some of the drawbacks of siRNA triggers and enable a range of RNAi applications based on the technology’s ability to achieve stable gene knockdown, to perform inducible RNAi experiments, to facilitate in vivo gene silencing, and to enable multiplexed whole-genome screening strategies.
Developed in collaboration with Greg Hannon, Ph.D., professor at Cold Spring Harbor Laboratory, and Steven Elledge, Ph.D., professor at Harvard Medical School, the shRNAmir technology incorporates triggers modeled after primary miRNA transcripts. This strategy allows for processing via the endogenous RNAi pathway, “which has been shown to produce increased and specific knockdown,” says Dr. Fewell.
Libraries targeting the human or mouse genomes include multiple shRNAmir’s per gene packaged in either lentiviral or retroviral vectors. The presence of a TurboGFP marker in the lentiviral vector enables tracking of shRNAmir expression in transfected or transduced cells.
In August, Open Biosystems added inducible TRIPZ lentiviral shRNAmir libraries targeting the human genome to its family of genome-wide, vector-based RNAi products. The TRIPZ lentiviral vector is engineered to be a Tet-On® system, and the shRNA is turned on in the presence of doxycycline. The addition of a TurboRFP marker allows visual tracking of inducible shRNAmir expression.
“Inducible shRNA takes RNAi to the next level,” says Dr. Fewell. “Researchers can now analyze in parallel the on-and-off states of a gene and study its interactions within complex pathways. It also allows researchers to examine the function of essential genes by turning them off temporarily without killing the cells. TRIPZ lentiviral shRNA makes it possible to analyze as well as validate the association between silencing of a target gene and a particular phenotype in the same experiment,” she adds.
Ongoing technology development at Open Biosystems is focusing on enhancing methods for genome-wide screening using multiplexed or pooled RNAi formats.
“Each hairpin expressing vector in our shRNAmir libraries has a unique 60 nucleotide molecular bar code that allows you to track the abundance of a given shRNA in a complex population,” Dr. Fewell explains. Users can pool thousands of different shRNAs, transduce a single population of cells, and perform positive or negative selection screens. The technology is applicable for a variety of biological effects associated with an identifiable and selectable phenotype such as cell survival, increased proliferation, adhesion, migration, or marker expression, reports Dr. Fewell.
“This multiplexed RNAi approach can be set up to study oncogenic pathways, mechanism of action studies, or the effects of drugs on cells sensitized by integrating specific shRNA,” notes Dr. Fewell.
A recent example of the application of multiplexed RNAi for genome-wide screening comes from the laboratory of Michael Green, Ph.D., and colleagues at the University of Massachusetts Medical School. They used a genome-wide pooled shRNAmir approach to identify a pathway of 28 genes required for Ras-mediated epigenetic silencing of the proapoptotic gene Fas. Activation of the Ras oncogene blocks Fas expression, thereby inhibiting Fas-induced apoptosis.