Fragment-based lead discovery

The concept of a 'target-rich, lead-poor' pipeline in drug discovery, and widespread concern about the attrition rate of chemical compounds in (pre)clinical development, are together fuelling the search for better quality hits and chemical lead series.A particular approach to lead identification for drug discovery, which offers a number of attractive features compared with high-throughput screening, involves the selection, screening and optimization of 'fragments'. Fragments typically have Mr = 120–250 and binding affinities in the range mM–30 μM. However, the weak absolute potency of fragments belies their high efficiency as ligands, because fragments are extremely potent for their size.Generally, fragments are identified using a biophysical screening method, most commonly NMR or protein crystallography supported by a conventional enzyme bioassay. Consequently, information about the structure of the fragment–protein binding interaction is generated as part of the screening.This structural information means that it is possible to incorporate a large element of design in optimizing the fragment into a high-affinity lead, either by growing additional binding groups or joining two fragments together. As a result, fragments can be optimized into nanomolar leads via the synthesis of significantly fewer compounds than in traditional approaches.Furthermore, starting the chemical optimization stage with a low-molecular-mass fragment is likely to produce leads whose Mr is still within the range desired for lead-likeness.This review discusses fragment-based lead discovery, and focuses on the output of this new approach by collating published examples from 25 protein targets. These targets are primarily enzymes and the screening techniques used include X-ray crystallography, nuclear magnetic resonance (NMR) spectroscopy, in vitro bioassays and mass spectrometry.