Researchers are synthesizing fewer, more diverse structures, while other investigators tout the potential of siRNA libraries
Patrick McGee
Senior Editor
When it emerged in the early 1990s, combinatorial chemistry, or combichem as it came to be known, was seen by many as a breakthrough tool for lead generation. Yet over time, researchers realized that maintaining quality and purity has been a problem, as has the fact that very few compounds from combichem libraries have demonstrated therapeutic potential. Those issues have made the field suspect in the eyes of some, but researchers have been moving ahead in generating small-molecule libraries and still others have been focusing on synthesizing small interfering RNA (siRNA) libraries.
“A lot of the early combichem efforts focused around what we would call a pharmacophore, a core structure, and then they would create a whole lot of compounds around that. You’d build very large collections of compounds, but there wasn’t much true diversity in it,” says William Marshall, PhD, executive vice president of research and operations at Dharmacon Inc., Lafayette, Colo., a provider of RNA oligo-dependent applications and technologies.
“People are now doing fewer total syntheses, but doing them around more skeletons so that they get a better coverage of chemical space,” Marshall says. “I think that is really what the whole idea is, it’s just cut back from just synthesizing huge numbers of compounds that may have fairly similar chemical structures and do more diverse structures and fewer of them. Then you focus your combichem efforts once you find leads to fill the space around that particular diversity foci you found that was interesting.”
There are two approaches to library synthesis, says Peter Linsley, PhD, executive director of research at Rosetta Inpharmatics, a wholly owned subsidiary of Merck & Co. Inc. founded in 1996 to design DNA microarray gene expression technologies. The vector-based approach shows promise, but it is at an earlier stage of development and is “plagued” by false positives and false negatives, he says. The other approach focuses on synthesis. “I think everyone would say that synthesis would be the preferable way to [build libraries] because it is cleaner and you know exactly what you’ve got, but the problem has been cost.” Linsley says that the cost of synthetic duplexes has come down tremendously over the last year or so.
Academic centers
In addition to dropping prices, the field could also benefit from advanced techniques coming out of academic centers. The National Institutes of Health plans to spend $20 million over the next five years to fund four Centers of Excellence in Chemical Methodologies and Library Development (CMLD), which are charged with developing new techniques for synthesis, separation, purification, and analysis with the goal of accelerating the creation of new libraries. The four centers are at Harvard University, Boston University, the University of Pittsburgh, and the University of Kansas.
“In each one of the centers, there are several projects going on, and most are synthetic projects or deal with synthetic methodologies,” says John Schwab, PhD, program director in the Division of Pharmacology, Physiology and Biological Chemistry, National Institute of General Medical Sciences, Bethesda, Md. “Many of them are focused on developing particular scaffolds inspired by natural products with identifiable biological activity. Our grantees are excited about collaborating with one another to mix and match those scaffolds, in some cases to make hybrid compounds combining two known biologically active scaffolds.” Schwab said researchers at the centers are also using new reaction methodologies from one center on a scaffold being developed at another center.
Exciting synergy
“What is particularly exciting to us, aside from the spirit of collaboration, is the obvious synergy and the ability to reach into unanticipated pockets of diversity.” Schwab says the perceived lack of impact of combinatorial chemistry and libraries may not be on the mark. He describes the field as “very, very young” with methodologies and strategies that are still being rapidly developed. “Industry, which has been a prime practitioner of high-throughput synthesis, does not have the resources to invest in the development of new methodologies. As a result, there was actually relatively limited diversity being generated, and people in industry were talking about how they were essentially investigating . . . the same pockets of diversity space, the same libraries.”
Schwab thinks the problem may not be due to the chemistry, but rather to the design of the high-throughput screens which are integral to the process. “Pharmaceutical companies make strategic choices about what kinds of targets they consider to be acceptable from a financial point of view, and it’s possible that their decisions have restricted the potential for drug discovery.”
John Porco, PhD, describes the research being done at the four centers as “grassroots” work that could ultimately benefit pharmaceutical companies. “I think the best thing we can do is develop new chemistries,” says Porco, assistant professor of chemistry and director of the CMLD at Boston University. Porco says that because they are usually working on deadline, researchers at pharmaceutical companies are often not able to develop novel chemistry. “[In an academic setting] we have greater liberty to take more time to develop the types of new chemical reactions that we do. In a pharmaceutical company, you’re up against a deadline to get a drug or a candidate or some compound out the door.”
Last June, CMLD launched an online resource that outlines nearly 50 procedures for synthesizing complex molecules, and additional protocols will be added as they are developed. The publicly accessible, Internet-based notebook can be searched using substructures, keywords, yields, or physical and analytical properties associated with a given compound. “Our mission is to facilitate the use of these compounds in the biological community. The compounds that we make are sent to a list of collaborators who are, at this moment, exclusively academic scientists who screen our compounds,” Porco says. “It’s tool building,” Schwab says of the efforts of the four CMLD centers. “As better tools become available, practitioners will be able to do what they need to do more effectively, so these tools will be broadly available.”
Industry and academia
Marshall says the strengths of industry and academia are complementary. “For most of the innovations in terms of true high-speed synthesis and full diversity coverage, full development of that will occur in the industry. Academia will develop those key, novel chemical synthesis methods and different high-yielding reactions and new approaches to things that could then be put into a broader array of more of the crank-through synthesis,” which will then produce large numbers of new entities.
One area in which the CMLD centers are not working is the synthesis of siRNA libraries. When RNA interference (RNAi) emerged in the late 1990s, Dharmacon provided RNAi-related products to researchers. It has since become a resource for those investigating the mechanisms of siRNA-induced gene knockdown and applying RNAi to human biotherapeutics. “We’ve developed genome-wide collections of siRNAs. What we did was to look very systematically at the way you do a good reverse genetics experiment, and RNAi is the first really compelling, reproducible, robust technology to knock down genes of any type. We then developed predictive algorithms that allow us to make siRNA molecules,” Marshall says.
By allowing researchers to open up current bottlenecks in target validation and by providing a potential new way to think about screening, Marshall says siRNA libraries can shorten the drug discovery pipeline. “You can screen for target identification and validation in a single step.” Marshall says that instead of having researchers perform a microarray to find candidate genes that have to be looked at individually, siRNA collections allow researchers to simply look for a good cell assay and screen by gene.
Marshall says Dharmacon’s siRNA collections range from 120 important genes that regulate the cell cycle to bigger collections like the druggable genome, which consists of 7,400 different genes. Abbott Laboratories and Bayer have both purchased about 4,000 to 5,000 different gene inhibitors. “What they’re doing is, they’re going in and straining for novel gene function. You can actually validate genes,” says Marshall. “Let’s say you have some indication that the gene is important in the disease state; now you can use siRNA to validate that your suspicion was correct.”
Filed Under: Drug Discovery