Title: REFiL - A systematic approach to the rapid elaboration of fragments into leads

Abstract

Fragment-based drug design (FBDD) has been widely adopted in both industry and academia and has proven to be a robust approach to identify small molecules that bind to a range of protein targets. In most cases, biophysical binding assays represent the mainstay for screening fragment libraries to identify suitable starting points for elaboration. However, the sensitivity of these assays means that the hit rates observed in fragment screens can be relatively high, and large numbers of hits are typically found in the primary screen. Moreover, it is common for fragment hits to bind very weakly to the target protein. The range of binding affinities observed in the primary screen varies significantly between different proteins, but in our experience, it is not uncommon for the highest affinity fragments from primary screens to bind with K_D ~ 1 mM. In such cases many of the hits from the primary screen bind with affinities that cannot be measured accurately – which makes ranking the hits challenging – particularly if metrics such as ligand efficiency are being considered. Even in cases where the fragment hits bind with higher and measurable affinities, selecting the “best” fragments to advance into further development is not always trivial.

These represent significant challenges that are faced in almost all FBDD campaigns.

• What is the most reliable way to select fragment hits for chemical elaboration?
• What is the most efficient way to elaborate the initial fragment hit into a more potent lead molecule?

To provide a more robust and systematic approach to this early phase of FBDD we have developed a workflow termed REFiL (Rapid Elaboration of Fragments into Leads). This workflow employs a combination of chemoinformatic design and biophysical screening approaches to validate and rank fragment hits, and to identify suitable vectors for chemical elaboration. Subsequently, it employs microscale parallel synthesis of targeted libraries that are designed to sample the chemical space around elaboration vectors.

In this presentation I will describe the strategy that we have implemented combining biophysical binding assays centred largely around NMR spectroscopy, surface plasmon resonance and X-ray crystallography with microscale parallel chemistry to enable a systematic and rapid approach whereby weakly-binding fragment hits can be elaborated into more potent ligands.