Aug. 19, 2015: Starting Off Right: Creating Better Chemical Probes
Creating a new therapeutic is like constructing a building. Both are highly complex, multi-year endeavors that require the contributions of many different disciplines. In each case, a solid foundation is critical: A building constructed on sand and a drug developed on faulty science will both fail.
A crucial part of the foundation for a new drug’s development comes from early tests of the scientific idea, when researchers use prototype drugs — termed “chemical probes” — in model testing systems. If either the chemical probe or the testing system is faulty, the drug’s development program will fail. I have written frequently about new testing systems NCATS is developing to more accurately predict a potential drug’s effects once it has entered the development pipeline, but I have not described our efforts to develop more accurate chemical probes more efficiently.
Chemical probes interacting with their molecular targets have often been referred to in a “key and lock” analogy, with probes being the keys to molecular locks in the body. But the molecular locksmiths — those trying to identify a chemical probe “key” to a particular target “lock” — are at two enormous disadvantages. First, the general principles governing which types of “keys” fit into which types of “locks” are not known, making testing of potential keys trial-and-error. And second, the number of potential chemical “keys” is functionally infinite. The number of potential “drug-like” chemical compounds is 1060 — that’s 10 followed by 60 zeroes — or more than the number of grains of sand on Earth. (For those of you who like nomenclature, that’s a novemdecillion.) Even with the robots at the NCATS Chemical Genomics Center (NCGC), only about 1 million (106) compounds can be tested, leaving the overwhelming majority of chemical space unexplored.
NCATS is taking multiple approaches to this important and exhilarating problem (for those of us brought up on Star Trek, chemical space exploration fires the imagination), and I will discuss these in the future. The end-goal of these efforts is to transform chemical probe (and eventually drug) identification from its current trial-and-error (mostly error) state into a predictive science, wherein a drug for any molecular disease target can be anticipated from informatics-driven computer models alone.
Partially because probe development is so difficult, it has recently become clear that many commonly used chemical probes are actually unreliable experimental tools, leading to faulty foundations for subsequent drug development. The probes’ effects aren’t strong enough, they interact with proteins other than the target, or their biological activity is misleading. Fortunately, several collaborative efforts are underway to address this problem, both within and outside of NCATS. This month, I’m a co-author on a Nature Chemical Biology commentary outlining the promise and issues with many existing chemical probes and offering potential solutions, including a newly developed wiki site called the Chemical Probes Portal. The scientific community can use the portal to disseminate reliable information about small molecules and to crowdsource information about the compounds’ properties and best uses. The goal is for researchers to visit the portal, get answers to their questions, and discover the best probes to use as well as how to use them to generate reliable, reproducible data.
Here at NCATS, the NCGC has developed and disseminated many chemical probes, working with disease-focused collaborators across the globe. The Chemical Probes Portal complements an existing public resource at NCATS, the Assay Guidance Manual, which disseminates information from experts around the world on the best methods to produce and validate chemical probes.
Through these and related efforts, NCATS is helping to put translational science on a solid foundation. And that will help get more treatments to more patients more quickly.
Christopher P. Austin, M.D.
National Center for Advancing Translational Sciences