By Chester A. Bisbee

Special To BioWorld Today

BOSTON -- The genomics revolution is transforming drug discovery and development, and as genome sequencing projects expand, a major challenge involves how to turn the huge accumulation of DNA information into commercially useful knowledge.

The solution may lie in high-throughput technology, whether it's the collection of sequence data and its analysis, the identification of drug targets and candidates, or the assays used to evaluate therapeutic activity.

Yet while many of the upstream processes, such as DNA sequencing and analysis, are conducted with high-throughput techniques, the more downstream assays used to identify gene functions and evaluate drug candidates that affect those functions remain slow and cumbersome.

But new assays based on developmental biology show great promise in turning identification and evaluation of drug candidates into high-throughput processes. Current knowledge in this area and potential ways of applying it to drug discovery were highlighted at a Nature Biotechnology conference titled "PharmacoGenesis: Postgenomic Drug Discovery Through Developmental Biology," held in Boston late last month.

Conservation of Functional Pathways

Efficient drug discovery using gene sequence data requires industry scientists to understand genes' functions. Fortunately for these scientists, said Gerald Rubin, "while the human genome contains approximately 70,000 genes, these genes appear to encode proteins for only approximately 1,000 critical biological pathways." Rubin is a Howard Hughes Investigator in the Department of Molecular and Cell Biology at the University of California at Berkeley.

Recent data show these proteins, their functions and interactions with other proteins are highly conserved, even between invertebrates and humans. As Rubin stated, "The take-home message is that functional gene modules are conserved throughout evolution. Even different pathways have significant similarities."

Several examples of this conservation have emerged from studies in invertebrate model systems, Rubin said. In one case, the TAM-1 gene involved in human leukemia has homologies to the notched gene of the fruit fly, Drosophila, and the lim gene of the nematode, Caenorhabditis elegans. As a result of this work, Rubin explained, "we now know these genes are involved in controlling differentiation and we know what receptors to be looking for to affect this process."

Using a human gene whose function is known in a developmentally based animal system is an important commercial goal because the model systems can be adapted to high-throughput assays.

As an example, Rubin showed that activated ras gene expression produces rough-looking eyes in fruit flies. "By making transgenics, you can examine large samples of complete genomes just by observing phenotypes under a dissecting microscope," he explained.

The power of this system comes from the ability to move freely between animal model systems, using eye phenotype in Drosophila and cell transformation assays in mouse 3T3 fibroblast cells, to study the control of ras gene expression.

Rubin added, "We have examined 185,000 [fruit fly] genomes in this model system in a search for modifiers of ras gene expression and have found 12 enhancers and 9 inhibitors."

Another challenge is how to use developmentally based systems to identify genes that control physiological and pathological pathways in disease. Given the confidence that gene pathways are evolutionarily conserved, the use of convenient model systems with animals not closely related to humans is increasing.

For example, zebrafish are being used to study cardiac development and hematopoiesis and to identify the genes involved in these processes. Zebrafish have advantages similar to Drosophila and C. elegans in that large numbers of the animals can be bred rapidly, making them adaptable to high-throughput technology.

In addition, since the zebrafish embryo is transparent, the phenotypes that develop from induced mutations can be easily seen and compared to ones relevant to humans.

Mark Fishman, chief of cardiology at Massachusetts General Hospital, in Boston, has developed a series of mutations in zebrafish that mimic a number of heart disease states in humans and behave as if they are due to a single gene defect.

"There appear to be relatively few informative phenotype mutations, and they appear to be conserved in vertebrates," he said, "suggesting that they are dissectable phenotypic modules."

Fishman has been able to identify what are apparently single-gene mutations that cause arrhythmias, which clinicians long have considered to be complex, probably multigene, defects.

Similarly, in studying hematopoiesis in zebrafish, Leonard Zon, a Howard Hughes Investigator at Children's Hospital, in Boston, has found thalassemia-like mutants that have no globin expression and fish that lack immune systems.

Zon said, "So much is known about the adult immune system function that the zebrafish is unlikely to contribute here. However, this system will help scientists to understand hematopoiesis at the initial developmental level."

While Zon has made progress in identifying genes involved in blood cell development, he and Fishman now face the task of cloning the genes responsible for the phenotype mutations they have observed. Once this process is complete, the sequence databases can be searched for homologous human genes and the hunt can begin for drug candidates that affect these genes.

Drug Discovery More Complex Than Thought

Charles Cohen, chief scientific officer at Creative BioMolecules, in Hopkinton, Mass., said, "The drug discovery pathway is even more complex than previously thought. Therefore, drug development will be that much more capital and infrastructure intensive than ever before."

A significant part of the solution to this problem may be developmentally-based model systems.

"Developmental biology will be the most effective for groups who are trying to find out how to restore function," observed Doros Platika, president and CEO of Ontogeny, in Cambridge, Mass. The greatest potential for commercial success, he said, will likely come in clinical applications for Alzheimer's disease, stroke, diabetes, osteoporosis, heart disorders and kidney failure. *