Forward chemical genetics: progress and obstacles on the path to a new pharmacopoeia R Scott Lokey Forward chemical genetics is a new method to systematize the discovery and use of small molecules as tools for basic biological research. This approach requires three basic components: a library of compounds; an assay, in which the library is screened for a cellular or organismal phenotype; and a method to trace an active compound to its biological target. Bioactive compounds have traditionally been isolated from natural product extracts, although `diversity-oriented synthesis' and commercial compound collections are gaining in prominence. New techniques, such as image-based screening and the cytoblot method, have increased the throughput of phenotypic assays. Strategies are also being developed to streamline target identi®cation using molecular biological approaches. Addresses Department of Chemistry and Biochemistry, University of California Santa Cruz, Santa Cruz, CA 95064, USA e-mail: lokey@chemistry.ucsc.edu Current Opinion in Chemical Biology 2003, 7:91±96 This review comes from a themed issue on Proteomics and genomics Edited by Matt Bogyo and James Hurley 1367-5931/03/$ ± see front matter ß 2003 Elsevier Science Ltd. All rights reserved. DOI 10.1016/S1367-5931(02)00002-9 Abbreviation DOS diversity-oriented synthesis Introduction Within the past several years, a new ®eld has emerged at the interface between cell biology and synthetic organic chemistry that promises to systematize the discovery of small organic compounds as tools to probe biochemical pathways. Recent advances in diversity-oriented (i.e. combinatorial) synthetic methods, the availability of com- mercial compound libraries, and developments in high- throughput screening and laboratory automation, have made it possible to mine large compound collections for interesting biological activity. This ®eld has been given the name `forward chemical genetics' to highlight its similarity to forward genetics in biology [1,2]. Like mutants in classical genetics, small molecules are screened for their ability to induce a particular phenotype in cells or cellular extracts. Instead of deleting or impair- ing protein function at the genetic level, as in classical genetics, small molecules generally act by inhibiting (or activating) a particular protein or set of proteins directly. Tracing the inhibitor (or activator) back to its target protein can, in principle, provide a causal link between the target and its associated phenotype. Forward chem- ical genetics requires three components: one, a collection or `library' of compounds; two, a biological assay with a quanti®able phenotypic output; and three, a strategy for identifying the target(s) of active compounds (Figure 1). This review focuses on recent progress in each of these areas as well as some of the dif®culties that have been encountered with current methods. The idea of using small molecules to dissect biochemical pathways is not new. There are many examples in which tracking down the phenotypic effects of small molecules has led to important biological discoveries, going back to the role of colchicine in the discovery of the cytoskeletal protein tubulin [3]. Many small molecules, including the Hsp90 inhibitor geldanamycin [4], the Golgi disruptor Brefeldin A [5], and a G2 checkpoint kinase inhibitor [6], were initially identi®ed by their phenotypic effects. The targets of these compounds were ultimately identi®ed, providing important insights into their roles in various cellular processes. Chemical genetics is thus a reinvention of classical pharmacology; in the modern version, new synthetic methods and screening technologies are being combined in an effort to dramatically expand the current pharmacopoeia of bioactive reagents. Sources of small molecules Natural products Different sources of chemical diversity have been tapped for input into phenotypic screens. Among these, natural products have the longest track record. The classic method for identifying bioactive natural products, known as bioassay-guided puri®cation, is an iterative process in which samples are extracted with solvent and the result- ing crude extracts are assayed for biological activity. The cycle of puri®cation and screening is repeated until a pure, active compound is obtained. Although historically successful, this iterative approach has some limitations. Only those extracts that contain the most potent and/or abundant bioactive compounds will score in the ®rst round of screening, whereas low-abundant compounds of moderate bioactivity Ð although potentially interest- ing Ð may be overlooked. Also, cytotoxic compounds can mask subtle phenotypic effects of other compounds pre- sent in the crude mixture. Third, synergistic or cooper- ative effects between compounds in complex mixtures may lead to bioactivity that disappears upon further fractionation. A simple pre-fractionation of the crude extracts would go far toward alleviating these problems. Parallel preparative HPLC systems now exist that can 91 www.current-opinion.com Current Opinion in Chemical Biology 2003, 7:91±96