The target validation step of the drug discovery process aims to prove that a given target is directly involved in disease and can be used for development of a therapeutic drug. Validation of a target requires collection and comprehension of a wide range of in vitro and in vivo data. Current validation methods are tailored for a particular gene and require individualized, time-consuming, and often expensive studies. With a steady stream of poorly characterized candidates coming from functional genomics efforts, target validation has become a bottleneck in drug development.
The failure, however, to correctly determine the biological relevance of the target could result in large accumulated costs due to high-attrition rates on the later stages of drug development. Thus, technologies that streamline target prioritization efforts are of immense value to the drug development community. Particularly valuable are the technologies that filter the candidates in a high-throughput, cost-effective manner by answering questions that are critical to the decision-making process. RNA interference, SpliceArrays, and bioluminescent imaging are some of the most promising players in the field.
RNAi-based technologies have steadily gained popularity by addressing the critical needs of target validation, such as high-predictive value, high-throughput, relatively low-cost, ability to knockdown several genes at the same time, and short experimental times. RNAi acts by binding to target RNA and initiating an enzymatic RNA-degradation cascade. Even though gene expression is not eliminated completely, the reduction in protein translation may result in enough phenotypical changes to make a conclusion about the targets physiological relevance.
We have quickly found out that in order to make a decision about a target, we did not need a complete silencing, 40 60% silencing was enough to disrupt critical angiogenesis processes, says Gabriel Kremmidiotis, Ph.D., vp of cancer research, Bionomics (www.bionomics. com.au). The companys Angene pipeline serves for identification and validation of angiogenesis targets.
Using subtractive hybridization, 160 genes were selected initially. Bionomics used a bioinformatics approach to narrow the pipeline down to 50 genes that could be amenable for small molecules or antibody targeting. Next, the company employed an siRNA approach to determine which genes elicit relevant phenotypic changes in endothelial cell behavior, including proliferation, cell migration, and capillary tube formation.
It was critical for us to filter the genes in the most effective way, states Dr. Kremmidiotis. siRNA shaved years off our target validation process. It does not matter if we missed some potential targets, as long as we end up with four to five good targets. The 20 genes that passed interrogation with siRNA, were subjected to a barrage of more laborious validation methods, such as tissue profiling, QPCR, immunohistochemistry, and in vivo capillary tube formation in a breast tumor model. Bionomics now plans to enter the first round of compound screening for one of its targets, p73RhoGAP.
Bionomics used retroviral vectors for siRNA delivery into endothelial cells. Some cell lines are resistant to penetration by chemical methods, explains Andrew Farmer, Ph.D., director of molecular and cellular biology, Clontech (www.clontech.com). Viruses offer the ability to precisely titrate siRNA delivery to each cell. Therefore, you can establish the rate of target inhibition with greater precision. Integration of a fluorescent marker in a viral construct enables monitoring of the rate of infection, as well as separation of infected from noninfected cells. Viral constructs can be also used for delivery of RNAi in vivo.
In vivo RNAi Delivery
Validation of a target gene in vitro is currently viewed as only a preliminary step in the drug development funnel. The results based on cell culture, even using multiple cell lines, are often not reproducible in the whole organism. To quickly filter out unsuccessful candidates, more and more companies turned to the early in vivo experiments. Knocking down a target gene in vivo allows elucidation of the pathway upstream and downstream of the target, of the systemic physiological effect of gene inhibition, or of the potential side effects of the chemical antagonists.
RNAi again emerges as a powerful tool for in vivo work, providing the ability to turn off the gene in an adult animal. The RNAi that is used in vitro can be used again in vivo, if the delivery challenges are overcome. Access to the target cells is a key issue for in vivo delivery of the nucleic acids, says James Hagstrom, Ph.D., vp of scientific operations at Mirus Bio (www.mirusbio.com).
Nucleic acids (NAs) have to withstand degradation in the blood stream and cross the endothelial barrier. Earlier observations showed that rapid injection of NAs into mouse tail vein using a large volume (12 mL) of isotonic solution results in high levels of hepatocytes transfection. Even though the exact basis of the phenomenon is not understood, it is presumed that NAs leak through the holes in endothelial lining of blood vessels supplying the liver.
Liver is a key organ in many metabolic pathways and a central organ for synthesis and secretion of a variety of proteins, continues Dr. Hagstrom. Knocking down gene expression in 1040% of hepatocytes provides tremendous insight into target gene effect on the whole organism. It also enables collection of ADME/Tox data that is critical for early pass/fail decision on drug candidates.
Hydrodynamic tail vein siRNA delivery has been successfully used for inhibition of apoptotic events in mouse livers by inhibiting either caspase 8 or FAS receptor. Mirus in vivo delivery portfolio includes TransIT Hydrodynamic Delivery Solutions and Pathway IV (hydrodynamic delivery into muscle cells). The latest addition is TransIT In Vivo Gene Delivery System, the polymer-based formulation. When mixed with NAs, the disulfide polymer forms stable, nonaggregating, 100-nm particles, preferentially targeting liver hepatocytes. The future NA delivery products include nanoparticle formulations containing liver targeting peptides and endosomolytic polymers.
Maximizing Relevant Results
With the advent of RNAi, the line between target identification and validation became less clear, suggests David Dorris, Ph.D, vp of RNAi Technologies, Ambion (www.ambion.com). Greater emphasis on systems biology leads to large-scale experiments, elucidating multiple genes in the several pathways. RNAi makes it possible to test 400 to 500 target genes in a matrix format. Chemically synthesized siRNA is a versatile tool with choices ranging from a single molecule to whole genome-based libraries.
Offerings of predesigned siRNAs, custom siRNA services, and delivery mechanisms have exploded (Table). Frost and Sullivan predicts further growth of RNAi applications specifically for target validation, generating over $140 million, or nearly one-half of total RNAi market revenues in 2010.
Interpretation of the complex biological results becomes even more difficult when variability in the quality of the reagents is added to a complicated experimental design. Do you measure the effect of your gene or rather artifacts of your experimental system? continues Dr. Dorris. If you have clear controls for a given phenotype, you may be able to differentiate between a real event and a side effect caused by the quality of the reagents. The solution, however, is to minimize variables in the experimental design by consistently using the same reagents in every experiment.
According to the company, Ambions Silencer In Vivo Ready siRNAs contain no toxic products, salts, or pyrogens and can be produced on the large-scale that is suitable for animal experiments. Many individual siRNAs have been prevalidated experimentally to reduce gene expression by at least 70%. We advocate the use of individual siRNAs as opposed to pooled siRNAs for target validation experiments. You eliminate yet another variable, speeding up data interpretation and reducing deconvolution time, says Dr. Dorris.
While in vivo RNAi silencing is still limited by inefficiencies of delivery methods, genetically modified animals, such as knockout mice, has served the purpose for many years. Even though knocking out some of the genes results in embryonic lethality, many knockout models have been successfully used for drug discovery and target validation.
The ability to study the gene effect on multiple biochemical pathways over a long period of time makes these models particularly attractive. Our goal is to drastically reduce the number of experimental animals required for target discovery and validation processes, says Michael Sterns, vp of Xenogen (www. xenogen.com).
First, we subject our genetically modified animals to a comprehensive panel of over 65 bioassays and challenge tests (Serial Phenotyping Compression Technology). The results of such studies give our clients tremendous insight into primary and secondary target effects in a number of therapeutically relevant areas, such as obesity, inflammation, hypertension, pain, and others. Our models help to predict the efficacy and side effects of a therapeutic compound before it is even administered.
Next, we use light to track disease progression or drug effect, continues Dr. Sterns. Xenogens VivoVision Solutions develops transgenic mice and cell lines incorporating disease-specific genetic markers tagged with luciferase. By using biophotonic imaging to track gene activation, the scientists are able to collect and interpret in vivo data in real time. The same animal can be used over multiple time points, adds Dr. Sterns. Not only does this eliminate statistical variability, inherent to assessments by histopathology, but also drastically reduces the number of experimental animals required for the study.
In a study by Sugen (www.sugen.com), bioluminescent imaging was used to demonstrate the efficacy of an experimental drug, directed against metastatic bone cancer. Athymic mice were injected with a metastatic breast cancer cell line carrying the luciferase gene. Tumor progression with and without drug treatment was monitored in individual mice in real time using Xenogens IVIS Imaging system. Bioluminescent imaging data supported the drugs efficacy in inhibiting the growth of metastasis and consequently, considerably sped up drug development efforts.
Most of modern gene discovery methods based on DNA microarrays are inherently biased toward genes with the polyA tail at the 3 end. However, the current estimates of gene transcripts are four to five times higher than the number of genes. Alternative splicing may account for some of this diversity. A growing number of examples, including apoptosis, beta-thalassemia, familial dysautonomia, FGF signaling, and spinal muscular atrophy, demonstrate that RNA isoforms have considerable biological effect. Thus, characterization of the RNA isoforms may soon become an integral part of the target validation process.
In one recent example, the correction of a splicing event gives hope to children suffering from spinal muscular atrophy (SMA). Homozygous mutation obliterates function of the survival motor neuron gene, SMN1, leading to degeneration of motor neurons of the spinal cord, causing atrophy of muscles and eventual death. Due to aberrant splicing, the backup gene, SMN2, produces only 10% of the full-length transcripts.
Treatment with valproic acid and some other histone deacetylase inhibitor drugs, is able to correct splicing and increase the amount of full-length SMN2 transcripts by several fold. The multicenter clinical trial is currently under way to assess safety and efficacy of valproic acid treatment for children with SMA.
ExonHit Therapeutics (www.exonhit.com) utilizes a database of publicly available sequences to identify alternatively spliced forms. Using this information, the company generates a series of specific exon body and junction probes and organizes them on the SpliceArrays.
By comparing RNA from several samples, the SpliceArrays are able to monitor expression of splice variants, including skipping of individual exons. The SpliceArrays are arranged by gene families (GPCR, ion channels, nuclear receptors, etc.) and contain genes that are known to produce alternatively splice transcripts, as well as genes that do not splice.
The dogma of one gene producing only one protein has been replaced by one gene producing many transcripts or proteins. Alternative splicing could result in transcripts with stop codons. These transcripts were thought to be quickly degraded. It now seems that that this degradation process is a secondary event, and many of the alternative transcripts can participate in protein translation, impacting various signaling pathways, says Richard Einstein, Ph.D., vp of research. There is little data relating to the expression patterns of splice variants. We only now can begin to catalog gene-specific splicing events and how many proteins actually result from them.
The SpliceArrays are not only able to identify alternative gene products but also provide information about differential expression and regulation of the isoforms under various conditions, their involvement in pathways, and their significance in pathology, adds Dr. Einstein.
In collaboration with Allergan (www.allergan.com), ExonHit was able to identify altered signaling pathways related to neuropathic pain. A new orally active small molecule is now being developed and has already shown a good profile in in vivo preclinical models.