Bioassay-based screening approaches search for hits with, for example, inhibitor constants at least in the low micromolar range. To get this much potency, a minimum number of interaction points between the molecule and the target protein are needed, which means that only larger compounds generate an initial hit.

Astex is building leads with X-ray crystallography. Credit: ASTEX TECHNOLOGY

Over the years, the compounds held by drug-discovery companies in their collections have been getting bigger, as screeners and medicinal chemists have chased the goal of potency. But this brings its own problems. “It is generally accepted that larger compounds lead to more late-stage attrition,” explains Harren Jhoti, co-founder and chief scientific officer of Astex Technology, a lead-discovery company based in Cambridge, UK. “Chemically refining large initial hits to work out which groups are important or not can be very time consuming.” The highly sensitive screening techniques that are now emerging, such as nuclear magnetic resonance and surface plasmon resonance, are providing the opportunity to search for lower-affinity hits, in the high micromolar or even low millimolar range. This opens up the possibility of screening for smaller starting compounds — 'chemical fragments' — which can then be built up into drug-like molecules by adding optimal functionalities, ideally using the structure of the target protein derived from X-ray crystallographic data. This, argues Jhoti, should allow for “a more directed path from hit to the lead than the hit-and-miss process of refining larger molecules”. Astex now has libraries of chemical fragments that bind in the low millimolar range, and already has examples of nanomolar lead compounds that were generated from millimolar hits.

Another company working with fragments is Graffinity Pharmaceuticals, in Heidelberg, Germany, which is taking advantage of the sensitivity of its optically based screening platform to test microarrays of chemical fragments. Technological challenges aside, another barrier to the development of fragment-based approaches is the generally held perception that high potency in binding to a target is a prerequisite for further development of a compound. “There is a change of mindset needed in order to convince a chemist that a millimolar hit is going to be of use,” says Jhoti.