Adjunct Data Required
Amgen understands the value, and limitations, of RNAi analysis. “Genetic approaches are ideally and uniquely suited to identifying the genes essential for, say, cancer cell survival and proliferation,” says Kim Quon, Ph.D., principal scientist. But prior to the advent of RNAi technologies, he notes, this was not possible, “because the tools needed to efficiently apply loss-of-function genetics to human cancer cells did not exist.”
Amgen is conducting siRNA screens of panels of cancer cells lines. As Dr. Quon says, the approach sounds great, on paper.
“It is not so simple to implement in practice. False positives and false negatives abound, making very general phenotypes such as viability and proliferation particularly difficult to interpret.”
Dr. Quon’s research has focused on parsing out meaningful conclusions from the spectra of phenotypes that result from differential siRNA knockdown efficiencies. His research team has suggested that siRNAs be considered as series of “hypomorphic alleles” instead of knockout agents, and they have used this approach to develop a means of quantifying siRNA data.
“The long-term potential of RNAi is huge,” agrees Francesca Santini, Ph.D., cellular discovery biology leader at Merck. “The technology provides access to an enormous wealth of disease targets that were previously inaccessible with traditional small molecule drugs.”
Merck acknowledges the need of combined areas of expertise here, noting that it conducts adjunct substantiating studies throughout its RNAi drug development process. Safety assessments include microarray gene-expression studies for assessing off-target effects of siRNA drug candidates, and preclinical toxicology routinely involves monitoring of immune system activation.
Since its $1 billion-plus acquisition of Sirna Therapeutics in 2006, the company has made few announcements regarding its siRNA R&D, though it has presented data at meetings in recent years regarding identification of modulators of hypoxia-induced factor (HIF). Primary siRNA knockdown screens were coupled with gene-expression and cell-imaging assays and network-based analyses to more clearly elucidate specific HIF pathway activity.
“At Merck, we have a mid-term goal of using RNAi to improve decision-making and increase the probability of success of our traditional therapeutics. We are using RNAi to validate disease-specific targets and pathways, as well as generate in vitro disease models. In the preclinical setting, RNAi technology can be used to assess the effect of knocking down specific target genes in model systems. Data from such experiments will help us decide whether to pursue novel targets for small molecule or biologics development,” Dr. Santini says.
For compounds in later clinical development, the analytical potential of siRNA is further supportive. “In the clinic, RNAi could potentially be given to humans to measure the effect of knocking down target genes on well-established biomarkers such as blood glucose levels or LDL cholesterol. Obtaining human proof-of-concept for targets with RNAi could increase the probability of success of small molecule or biologic therapeutics once they reach the clinic.”
For certain treatment-recalcitrant viral infections, RNAi has been a particularly beneficial analytical tool. “To target HIV specifically, given the high sequence variation of HIV isolates, is always a challenge,” notes Renate Koenig, Ph.D., research assistant professor in the infectious and inflammatory disease center of the Burnham Institute for Medical Research. “Only siRNA has made it possible to knockdown each gene in the human genome, enabling us to gain a whole snapshot of the life cycle of HIV and an understanding of the virus’ pathways in detail.”
Through RNAi analysis, she adds, her team “determined several previously undescribed virus-host interactions that likely occur in concert to facilitate the early stages of HIV infection. For instance, nuclear import of the virus and integration of the virus seem to be coupled processes mediated by nuclear porins, karyopherins, and other soluble transport factors.”
Dr. Koenig and her colleagues are conducting whole-genome siRNA screening, with the goal of discovering new antiviral targets, for HIV and influenza virus. “It is the main research tool we are using, but it is inherently prone to false negatives and positives.”
Dr. Koenig’s team thus uses these approaches, in conjunction with other methodologies, to “rank” the data mined from the RNAi screens more effectively. The researchers combined their RNAi results with interrogation of “human interactome” databases, and assessment of protein-protein interactions, mRNA expression, and gene ontology.
The end result is a functional map that includes subnetworks of interacting pathogen-host factors, each of which may be a potential drug target. Such RNAi analyses can contribute substantially to greater efficacy of eventual pharmaceutical treatment regimens. For complex systemic infections such as HIV, for instance, eventual therapeutics based on RNAi may seemingly do well to target both host and pathogen pathways.
“Given the changeability of viral DNA, therapeutic targeting of virus genes is always a moving target. But to target host cellular proteins in addition to the viral proteins adds a fixed drug target,” Dr. Koenig notes. “Administration of multiple RNAi drugs, some targeting host processes and others viral processes, may be the most effective.”
Educated as a mathematician, Auguste Genovesio, Ph.D., head of the image-mining group at Institut Pasteur Korea, brings an understanding of the necessity of progressive data visualization in RNAi analysis. Citing the enormous variability of results in RNAi HIV screens (there may be as low as 7% overlap across results of different studies), he says single-gene knockdown studies by siRNA can be “a nightmare” in practice.