April 15, 2013 (Vol. 33, No. 8)
RNAi research has capitalized on the ability to design and deliver small interfering RNAs (siRNAs) or short hairpin RNAs (shRNAs) to knock down expression of target genes with strong selectivity and specificity.
These genetic tools have applications in many research areas to study the genetic mechanisms and functions of various pathologies, leading to the potential for therapeutic applications.
Life Technologies has developed tools for the design and research applications of siRNAs, explains senior staff scientist Alexander Vlassov, Ph.D. The Silencer® Select platform includes thousands of siRNAs with LNA chemical modifications and optimal designs to target all human, mouse, and rat genes in highly potent and specific fashion. Customers can also submit their own sequences to design custom siRNAs for unusual targets, or if they want to test additional siRNA designs for a particular target. Silencer Select siRNAs are a good option for in vitro experiments, says Dr. Vlassov.
The Ambion® In Vivo siRNAs, developed specifically for in vivo applications based on the parental Silencer Select platform, have proven to be highly stable when injected into the bloodstream of animal models, he continues. Although in contact with nucleases found in body fluids, these siRNAs do not readily degrade due to stabilizing chemical modifications.
The half-life of Ambion In Vivo siRNAs in the bloodstream is more than 24 hours, while regular siRNAs can degrade in a matter of minutes, Dr. Vlassov notes. By using these stable siRNAs, researchers can achieve prolonged effects and perform fewer animal injections. He explains that with a single injection at a low dose, the duration of siRNA-mediated knockdown can last for up to one month.
Along with siRNAs, Life Technologies has also developed tools for miRNA research. Similar to siRNAs, miRNAs downregulate gene expression, but usually do so by repressing translation instead of inducing mRNA cleavage.
miRNA mimics are 18–25 bp double-stranded synthetic molecules (with limited chemical modifications) that mimic endogenous miRNAs and are used to increase their intracellular levels, while miRNA inhibitors are single-stranded oligonucleotides (with heavy chemical modifications) that bind to a miRNA of interest and reduce its intracellular levels.
When using an miRNA mimic or inhibitor, readouts of knockdown or derepression, respectively, of mRNA or protein levels of the miRNA’s targets can confirm the efficiency of the mimic or inhibitor. Life Technologies’ offers mirVana™ products to target miR-122 in vivo, and miR-1 and let-7 in vitro, along with a number of reporter assays.
Dr. Vlassov says that the library of miRNA mimics and inhibitors is frequently updated to match the latest version of miRBase, the online miRNA database. Additionally, he notes that “due to the natural complexity of miRNA pathways, there may be a risk of off-target effects because miRNAs do not have perfect homology to their targets, each miRNA can have hundreds of targets, and every mRNA can be targeted by multiple miRNAs. It’s a complicated story, but our products are still highly specific despite these challenges.”
Life Technologies provides a panel of lipid-based delivery reagents, such as Lipofectamine® RNAiMAX and Invivofectamine® 2.0, for transfection in cell culture or animal models, respectively. While efficient delivery of siRNA or miRNA cargo into cells is usually straightforward, in vivo systemic delivery can be quite challenging and presents more barriers to overcome, such as the potential for some material to become bound to blood proteins or be filtered out by the kidneys before reaching the organ of interest. Currently, the Invivofectamine 2.0 reagent can effectively deliver siRNA and miRNA cargo into liver cells in animal models, says Dr. Vlassov.
“We would like to develop separate Invivofectamine reagents to transfect siRNAs into the lungs, brain, spleen, and kidney, as it remains difficult to consistently and effectively transport siRNA cargo to these tissues via tail vein injection.”
Issues of siRNA specificity have been identified in a number of studies over the last few years, in which siRNA delivery results in phenotypes that can only be explained by deregulating genes that were not intended for targeting by the siRNA, explains Michael Hannus, Ph.D., who works at the University of Regensburg and collaborates with Intana Bioscience. Dr. Hannus indicates that these off-target effects can be quite extensive, as one siRNA design can potentially target 8,000 human genes on average.
Reducing Off-Target Effects
“Our approach was to design multiple siRNAs that all have the same ‘on’ target but all have different ‘off’ targets,” Dr. Hannus remarks. “With a sufficiently complex siRNA pool, the on-target gene knockdown effect will remain strong while the off-target effects of each individual siRNA are diluted with an increasing number of siRNA designs contained by the pool.”
By spiking in a control siRNA previously published as demonstrating strong off-target effects, Dr. Hannus and his collaborators confirmed that the number of deregulated genes was dramatically reduced by increasing the complexity of the siRNA pool, as measured on an Affymetrix genomic microarray. Furthermore, the researchers discovered that a pool of approximately 30 siRNAs provided sufficient complexity necessary to substantially dilute any off-target effects.
According to Dr. Hannus, “Our technology provides complex but defined siRNA pools. Using an algorithm, we select a collection of optimal sequences to compose the pool.” The research team developed a novel enzymatic approach to generate siRNA pools of defined sequences via in vitro transcription and a specific nuclease. All siRNAs are precisely 21 nucleotides long to provide maximum efficiency and avoid any possible effects from the interferon response.
Dr. Hannus and colleagues are currently working on expanding their collection of siRNA pools, termed “siPools.” The goal is to assemble a genome-wide human siPool library within the next two years. Furthermore, as the enzymatic approach is far cheaper than chemically synthesizing siRNAs, large amounts of siPools can be produced at a low cost, offering an attractive alternative for in vivo RNAi applications.
Polymer-Based Delivery of siRNAs
As an alternative to cationic lipid-based delivery, Dynamic Polyconjugates™ (DPCs) are polymer-based vehicles for delivery of siRNAs into specific tissues or cell types of interest to promote efficient and stable knockdown, explains Dave Rozema, Ph.D., vp of chemistry at Arrowhead Research.
DPCs are 5–20 nm in size and are composed of a polymer to which targeting ligands are reversibly attached. The DPCs are guided to the cell of interest by the targeting ligand and are taken up by the endosome, which is lysed by the polymer to release the siRNA cargo into the cytoplasm. Dr. Rozema noted that using polymer-based technology allows for easy control of the size of the resulting formulation, whereas nucleic acid–lipid complexes tend to have a larger size that may make delivery more difficult.
When DPCs were first developed by Arrowhead, the siRNA cargo was covalently attached to the polymer (hence the term “polyconjugate”), precluding the need for electrostatics to hold the two together, as is the case for lipid-based delivery.
“We have recently discovered that it is not necessary to attach the two if the siRNA and polymer each have their own targeting groups; they will still meet up in the targeted tissue by co-injection without being physically attached to one another,” Dr. Rozema explains. “There may be a marginal difference in efficacy with respect to the siRNA dose, but what we gain is a huge simplicity in the formulation because we can make the siRNA and polymer separately, so for manufacturing and analytics it becomes a much easier task.”
Arrowhead is pursuing this style of DPC formulation with intravenous delivery of siRNA-mediated knockdown of hepatitis B virus, in which the modified polymers and siRNA cargo are targeted to hepatocytes. Arrowhead is also researching DPC designs for subcutaneous liver delivery, as well as identifying high-capacity, rapidly internalizing ligands for targeting to other tissues.
Dr. Rozema emphasizes that “the research team has designed polymers that are biodegradable and can be easily cleared, as well as concentrating on controlled polymerizations to allow for repeatable and scalable processing of a tight distribution of polymer sizes and compositions.”
DPC delivery of siRNA cargo allows for a long-term, sustainable effect, claims Dr. Rozema. He says delivery with one monthly injection is enough to induce 90% knockdown of the target gene in mice; in nonhuman primates, over 80% knockdown for three months has been achieved.
Screening for Gene Candidates Using shRNAs
While these technological advances in siRNA synthesis and delivery have optimized the effectiveness of siRNAs in RNAi-mediated therapeutics, shRNAs are also important tools for RNAi research that can overcome some of the difficulties presented by using siRNAs.
By exploiting the cell’s endogenous RNAi processing machinery, shRNA constructs allow for potent, sustainable knockdown using low copy numbers that can result in fewer off-target effects. siRNA molecules must overcome barriers like circulating nucleases and endosome acidity, but lentiviral transduction of shRNA vectors removes these potential hindrances.
Another advantage of shRNA libraries is efficient transduction of nontransfectable cell types, which can be a limitation for siRNA libraries. Additionally, shRNAs can be integrated into the host genome and are only administered once, while siRNAs remain in the cytosol and usually have a more transient effect.
Annaleen Vermeulen, Ph.D., senior scientist and R&D manager at Thermo Fisher Scientific, explains that to set up a successful shRNA screen, a pooled library is used containing multiple shRNAs, which are then packaged into lentiviral particles and transduced into the control and experimental populations such that there is one shRNA per cell on average. For example, the experimental sample may be treated with a cancer drug in order to look for enrichment in cell death in the presence of the drug and the shRNA, identifying genes that are responsible for cell death sensitization.
In the past, Thermo Fisher Scientific has created genome-wide libraries with multiple shRNAs targeting each gene, but they now have also made smaller libraries that target gene families, such as the kinase library, while still maintaining high shRNA coverage. These smaller libraries require fewer cells to be transduced, which can be particularly useful when working with rare cell lines that are hard to isolate.
“From publications and from internal work, we noticed that researchers often identify only a few strong hits from these screens that are really useful for understanding their biology. We decided that if reproducibility of these screens could be improved, we could increase the dynamic range of these screens, so that there would be more candidate genes to follow up on,” Dr. Vermeulen states.
She and colleagues systematically examined the pooled shRNA screen reproducibility for different experimental parameters, including shRNA representation at the transduction step, the PCR amplification step, and data analysis by microarray or next-generation sequencing (NGS).
With one shRNA per cell on average, it was essential to have at least 500 to 1,000 cells representing each shRNA to provide reproducible data and confidence in the hits that were identified. Furthermore, maintaining shRNA representation during amplification also proved to be critical in obtaining reproducible data. While there was overlap between the results obtained using an NGS or microarray readout, NGS had a much greater dynamic range, the hits were inherently more reproducible, and there were fewer false positives.
“We have shown that to have high reproducibility and confidence in the significance of your hits from a pooled shRNA screen, it is critical to have the highest fold shRNA representation that is reasonable for your experiment, as well as optimal amplification conditions,” Dr. Vermeulen concludes. By optimizing these factors, Thermo Fisher Scientific has developed a complete workflow for pooled shRNA screening combined with NGS analysis.
Cellecta is developing RNAi dropout viability and rescue screens, says Paul Diehl, Ph.D., director of marketing and business development. Dropout viability screens identify essential genes for cell viability, such as genes required for cancer cell proliferation, or genes that increase cell sensitivity to a particular drug. Rescue screens, on the other hand, identify genes required for apoptosis or for cell sensitivity to a drug.
Dr. Diehl explains that Cellecta uses a synthesized pool of oligonucleotides to make a library of shRNA-expression constructs instead of making individual plasmids. This approach allows them to generate an entire library within three months. The screening process utilizes a complex library of 27,000–55,000 barcoded shRNA constructs per pool with typically 5–10 shRNAs per gene. The barcodes enable each shRNA construct to be identified by PCR and high-throughput screening.
In a dropout viability screen, by comparing the barcodes present in the surviving cell population with the original library’s barcode distribution, shRNAs interfering with or slowing down cell growth can be identified, since they are depleted in the final population. It is the gene targets of these deleted shRNAs that are likely essential for cell viability. Conversely, in a rescue screen, shRNAs that are enriched in the final population compared to the original library interfere with genes that produce the lethal response in the presence of the apoptosis inducer.
Dr. Diehl also notes that while shRNA-specific library barcoding identifies the particular shRNAs present in the cell population, Cellecta has developed a new approach that includes a second barcode on the lentiviral vector, called a clonal barcode. This variation provides an additional layer of specificity.
“When you assess the shRNAs in surviving cells, you can not only identify the total distribution of shRNAs but also look at how many separated clones produced the population with a specific shRNA. Basically, you can actually track the fate of each individual cell from the original population in which the library was introduced,” Dr. Diehl says.
“This can help identify if all the cells with a specific shRNA grew or if there was one weird clone skewing the results, so you can separate external and spurious effects that alter growth rates from those produced by the presence of the shRNA itself.” Cellecta has moved this technology into a xenograft model to identify, for example, which genes are required for tumor growth in vivo.
Dr. Diehl has worked with a dropout screen identifying genes essential for viability in leukemic cells, as well as a rescue screen identifying genes essential for Fas-induced apoptosis. The results from the dropout screen identified many general viability genes, but there were no major hits common across different blood lines, he says.
The rescue screen identified canonical genes like Fas, Fadd, and Bid, as well as genes not normally associated with Fas-induced apoptosis. With collaborators at Roswell Park Cancer Center, Cellecta’s research team treated mice with the Fas ligand to induce apoptosis of hepatic cells, and they were able to validate that synthetic siRNA genes identified in the in vitro screen, as well as chemical inhibitors targeting the corresponding proteins, were both able to protect the mice from Fas-induced hepatic failure.
According to Dr. Diehl, “Cellecta is continuing to develop both our shRNA and high-throughput barcode tracking technologies with a plan to move into applications with more direct clinical and diagnostic relevance.”