by John Hallock

May 12, 2005 | With the costs of populating drug pipelines rising sharply, drug companies are exploring new in vivo animal models to guide early pre-clinical drug development. Despite a plethora of available technologies to discern biological mechanisms, the relevance of such technologies is only as good as the physiological models to which they are applied. A complete picture of the biological interactions occurring in drug action and toxicity requires the examination of intact multicellular organisms.

One scientist leading the charge for emerging surrogate in vivo models in drug development is Randall Peterson, assistant professor of medicine at Massachusetts General Hospital and Harvard Medical School. After obtaining his Ph.D. from Harvard University in the laboratory of Stuart Schreiber, where he first demonstrated the feasibility of small-molecule screens in zebrafish, Peterson joined the lab of Mark Fishman (now the president of NIBR). Peterson is on the Scientific Advisory Board of Montigen Pharmaceuticals, and is a founder of Teleome Labs.

GETTING REAL: Simpler organisms permit in vivo work prevously restricted to in vitro says Peterson. photo by Michael Manning

Peterson’s group is currently using zebrafish for an entirely new application: pre-clinical drug development. This zebrafish model is allowing phenotype-based discovery of lead compounds that can suppress disease phenotypes (see Nature Biotechnology, May 2004), as well as new approaches for testing compound safety. Bio•IT World contributor John Hallock spoke with Peterson about issues surrounding the use of in vivo surrogate models in drug development and safety.

Q. What are some of the emerging in vivo models being used in drug development today?
Peterson: Much of the effort in developing novel in vivo models is focused on the high-end models — increasingly complex and expensive mammals like genetically modified mice, dogs, pigs, etc. These models are clearly very important, but I’m most excited about the low-end models such as the roundworm and the zebrafish that are beginning to enter into the drug development process. They aren’t as sophisticated, but they enable all sorts of previously unimagined applications because they are so inexpensive and easy to use. These models are bringing the advantages of in vivo studies to drug development steps that had been strictly limited to in vitro approaches.

What aspects of drug development do these new in vivo models address?
I think they address two issues: first, the difficulty in identifying optimal therapeutic targets for many diseases and, second, the issue of drug toxicity. [F]or many diseases, we don’t know what the optimal targets are, so it is difficult to use traditional target-based discovery approaches in the absence of a target. One solution is coming in the form of these emerging in vivo models that can be used in high-throughput chemical screens. By modeling diseases in these organisms, one can screen for compounds that reverse the disease phenotype and restore the animal to normal health, even when an effective therapeutic target has not been identified. Compounds discovered by this phenotype-based approach not only make interesting therapeutic lead compounds, but also may help identify and validate novel therapeutic targets.

How do these new models differ from traditional animal models in evaluating drug safety?
The capabilities of the zebrafish, for example, differ from those of the mammalian species in terms of their similarity to humans, but also in terms of cost and scale. Because of their high costs and difficulty, mammalian models are generally reserved for later preclinical stages of drug development. As a result, toxicities are only being discovered after significant time and resources have been invested in a drug candidate. Zebrafish toxicity assays are admittedly less sophisticated than traditional models, but they can be performed quickly, inexpensively, and with miniscule compound quantities. These advantages make it possible to derive preliminary toxicity data from the initial high-throughput screen, to prioritize initial lead compounds, or even to prescreen entire small-molecule libraries to remove potentially problematic compounds. The zebrafish is unlikely to eliminate the need for the more-sophisticated mammalian models, but it could greatly reduce the number of toxic compounds that are tested in mammals by identifying unsafe compounds much earlier.

What are the benefits of the systems biology approach for evaluating drug efficacy and safety?
The direct effect of a drug binding to its target is only a part of a drug’s biological activity. The effect of the drug binding to off-target molecules, the effects of its metabolites, and other pharmacodynamic factors all influence efficacy and safety. These secondary effects can lead to unexpected toxicities or can be beneficial. For example, beyond the cholesterol-lowering power of statins, some of their benefit appears to be due to off-target anti-inflammatory effects. Secondary effects are not easily measured in vitro, so you can’t know a compound’s true efficacy or safety until you test it in an intact animal. That is a strong impetus to involve animal models early in the process.

What are some of the key benefits surrounding the use of zebrafish specifically?
The zebrafish is a vertebrate, so it has most of the organ systems and physiological functions that humans have. Numerous zebrafish disease models have already been developed, the zebrafish genome has been sequenced, and genomic tools like microarrays and cDNA collections are well developed. Zebrafish are prolific and inexpensive to raise, and in embryonic and larval stages fit easily in the wells of 384-well plates. Another advantage is their transparency, which permits organ morphology and function to be monitored in the living organism using rapid, robust imaging tools. Finally, unlike many other model organisms, zebrafish absorb compounds readily from their surrounding environment; treatments can be performed simply by adding compounds to their water.

GLASS HOUSES: Angiograms highlighting the vasculature of a pair of mutant zebrafish. The defect in the upper fish has been suppressed by treatment with the small molecule GS4012. The background shows human endothelial cells induced to assemble into tubules by treatment with GS4012.

How consistent are pathologies and pathogenesis across species?
No animal model is a perfect surrogate for human pathogenesis, and one would expect similarities to decrease with phylogenetic distance. Therefore, emerging models like the zebrafish will in most cases be less predictive than mammalian models. That said, [our] data suggest there will be more similarities than differences between zebrafish and mammals. For example, several of the genetic mutations identified in zebrafish cause similar diseases in humans. Most of the drugs that cause cardiotoxicity in humans have a similar effect in zebrafish, and compounds that have been discovered by screening in zebrafish have similar effects in human cells. Time and experience will help us understand the limits of these newer models, but for now the remarkable thing is how much of the biology and pathology is conserved.

What should be done to bring newer models into mainstream drug development? Do you see any early adopters?
As with most technologies, they will only enter mainstream use as early adopters begin to have success with them. Among the large pharmaceutical companies, Novartis seems to be taking the lead, although a couple of the other large players appear to be following suit. Also, small companies like Phylonix, Zygogen, DanioLabs, and Montigen Pharmaceuticals are embracing these emerging in vivo models, either for their own discovery efforts or to provide larger companies with access to models.

How critical is academia in advancing chemical biology and drug discovery?
Academics are becoming increasingly interested in chemical biology and small-molecule discovery as a tool for fundamental biological research. In addition, affordable high-throughput screening technologies and chemical libraries are now available, and the top academic institutions have established core facilities for screening. As a result, large-scale screens that once seemed out of reach are now approachable for many basic and disease-oriented academic researchers. The influx of academics into small-molecule discovery is having two major effects. First, it is spurring new technological developments, especially in the area of phenotype-based screens. Second, it is bringing small-molecule discovery efforts to many diseases and targets that have been neglected by the industry.