Delivering siRNA to animal disease models for validation of novel drug targets in vivo - siRNA

Intradigm (Rockville, Maryland, USA) has developed non-viral delivery systems for intratumoral administration of siRNA into several xenograft tumor models to knock down angiogenesis factors, resulting in marked tumor growth inhibition (17). Non-viral vectors have many advantages, especially for in vivo target validation over the viral vectors. Clearly, effective siRNA delivery is crucial for in vivo functional genomics (i.e., inhibition of gene expression in a significant number of cells so that the biological effect of a drug targeting the same protein can be emulated). Just as it is for viral vectors, an important requirement for non-viral vectors is that delivery itself doesn't cause significant background activity. Non-viral delivery methods tend to have much lower levels of biological effects, other than those from the gene, relative to viral vectors, which have many protein components. For this reason, we developed our method, which uses intratumoral administration of siRNA duplexes, increasing efficiency of nucleic acid delivery while avoiding unwanted background noise from the viral vectors.
Target Validation Using siRNA Delivery In Vivo
The large amounts of raw data generated by sequencing the genomes of humans and many other species have encouraged interest in developing and using functional genomics and proteomics methods for drug discovery. Drug targets are the key genes and proteins that more or less single-handedly control the underlying pathology of a disease (i.e., the "gatekeepers" of pathological processes [18]). Given the enormous pool of candidate drug targets and reliance upon high-throughput sorting, conventional target discovery must rely on simplified biological systems amenable to high-throughput assays. Typically, the preliminarily validated candidate targets identified from these simplified systems are used for drug discovery even though they are not fully validated for their ability to control disease pathology (19); that usually requires animal disease models to demonstrate efficacy.
To answer the question, "which genes or proteins are functioning in controlling the pathology" one concept has been realized--that pre-clinical animal models of disease, preferably clinically relevant disease models, can be developed for generation of data sets with enriched information. We hypothesized (9) that pre-clinical models showing efficacious processes can improve the quality of the data set for validation of the interested targets. Yet, better justification for investing in further drug discovery to these targets still will be warranted. The factors associated with pathological pathways that control disease dynamics can be recognized more easily by establishing genomics methods applied to active pre-clinical disease models that are undergoing changes.
This is quite distinct from the well accepted use of pathway analysis on tissue sections from either animal disease models or human clinical samples that clearly cannot provide information on disease dynamics.
To address the need to use pre-clinical disease models to validate gene functions, we have developed a concept using established animal xenograft cancer models to obtain the essential dynamic information. The concept is to trigger efficacy processes in the tumor models and then use the microarray differential display to identify the significant up- or down-regulated targets, followed by siRNA-mediated gene knockdown in the same model system for functional validation. The discovery and validation steps of the process are summarized in Figure 3.
Once the candidate genes are identified, the large number of candidates typically will require in vivo siRNA-based validation to reduce them. Validation of target genes and proteins is becoming focused on whether or not they have the ability to actually control an ongoing disease process and reverse the pathology or its symptoms. This is due largely to complex interactions of multiple cell types that result in disease pathology. Moreover, many proteins lie in certain pathways that only can be detected and validated in active disease tissues. For cancer, a typical process for candidate reduction requires a combination of bioinformatics analysis, cell pathway identification, high-throughput transgenic mice and other approaches to yield a reduced set of better candidates. Using siRNA delivery in the xenograft tumor model to knockdown the candidate targets represents an effective approach to validate the targets in vivo with disease-related pathological read-out, such as tumor growth inhibition. This approach is highly dependent upon the efficiency of siRNA delivery in vivo and the effects of siRNA-mediated silencing. The power of the method is that it is capable of generating data sets derived from induced disease dynamics (both target selection and validation were based on the tumor growth rates, which were perturbed by the nucleic add deliveries)--the key for therapeutic effects.
Contents identified for genes and proteins associated with efficacy represents the best information needed for functional validation of the candidate drug targets. In order for therapeutic intervention to be successful, the drug targets must be able to control the pathological process, or at least the symptoms that result. To fulfill this requirement, the search criteria and the resulting data set need to contain data relating to controlling pathological processes. To be effective, this method requires high efficiency of in vivo delivery of siRNA into the disease tissue of interest. Importantly, this approach provides insight into genes and proteins associated with the later stages of pathology because it can be applied to established disease tissues where those targets play critical biological roles.
Finally, selecting a broad versus a narrow pathway analysis method needs to be considered carefully, due to the complexity of each molecular pathway of disease. Narrower pathway analysis methods can be applied for specific pathological information. Conversely, genome-wide pathway analysis also can be used because the differential display of gene expression between control tumor and perturbed tumor is expected to yield only a small number of genes and proteins that are either up- or down-regulated. Regardless, the sensitivity of the chosen method needs to be high. This likely is true for another more fundamental reason--namely the levels of gene expression of the gatekeepers tend to be low relative to those of the overall gene population. The gatekeepers--genes and proteins receiving outside signals--generally initiate or block cascades of certain pathways, so their expression levels tend to be low in comparison to the members of the cascade, especially in the final stage of the cascade. A combination of sensitive, genome- or proteome-wide pathway analysis methods with gene perturbation can provide disease-controlling, focused discovery and validation (Figure 4).
Validation of Novel Targets for Antibody Drugs
The recent success of a number of new monoclonal antibody (mAb) therapies suggests a resurgence in the biotechnology industry that is set to be realized in years to come. Drugs such as Genentech's (South San Francisco, California, USA) Rituxan and Johnson & Johnson's (New Brunswick, New Jersey, USA) Remicade are helping to pave the way toward acceptance of these agents. Sales of monoclonal antibody drugs reaching $5 billion by the end of year 2003 highlighted the clinical efficacy of antibody therapy for serious long-term conditions such as arthritis and Crohn's disease (20). Moreover, the potential is quite vast, as more and more antibody products, such as Herceptin and Avastin (Genentech), are being evaluated for cancer indications.
Antibodies are emerging as a class of cancer therapeutics that can be developed rapidly once a good target is identified. In cancer, the tumorigenesis process is thought to be the result of abnormal overexpression of oncogenes, growth factors and mutant tumor suppressors, even though underexpression of other proteins also can play a critical role. Efforts to identify and validate tumorigenic targets have been focused mainly on those targets overexpressed in the tumor tissues and promoting tumorigenesis as a means to enable development of small-molecule and antibody anti-cancer drugs acting through an inhibitor mechanism (1). Although the mechanism of action of siRNA inhibitory effects knocking down expression of a gene target is quite different from that of antibody inhibitors blocking the function of the target protein, the biological effects on the down-stream factors generally are similar, due to the down-regulation of activities of the targeted proteins. Therefore, studies designed to reveal whether a gene target plays a tumorigenic role using siRNA-mediated knockdown will be able to identify proteins as potential monoclonal antibody drug targets, if the function of those protein can be blocked by antibody binding.
Intradigm has developed a unique target identification approach (Figure 3) named Efficacy-First[TM] discovery. This method utilizes efficient delivery of nucleic acid into a xenograft tumor model to induce phenotypic effects on tumor growth rate and then identify genes that are significantly up- or down-regulated, in correlation with the induced tumor growth rates. Using this method, we have selected a pool of gene targets markedly overexpressed in a tumor that demonstrated accelerated growth after treatment with the growth factor expression plasmid. To differentiate which of these gene targets might control tumor growth, we designed and administered siRNAs specifically targeting the mRNA sequences of the genes of interest by intratumoral delivery of the siRNA duplexes. After measuring tumor growth rates, following repeated administrations of siRNA duplexes, the handful of targets being knocked down and shown to induce tumor inhibition were validated as p