Structure-based design and combinatorial optimization of novel thymidylate synthase inhibitors

Protein structures have facilitated the discovery of new lead compounds for a large number of enzymes and receptors. The subsequent optimization of these leads has typically involved the one-by-one synthesis of variants, which is often slow. To speed up the "rational drug design cycle" we have combined structure-based methods with combinatorial technologies. We began with the structure of Thymidylate Synthase (TS) a well known target for the design of antiproliferative drugs. Using the computer program DOCK, we investigated the active site of TS with a database wof 180,000 commercially available compounds. The program generated a list of potential ligands for the enzyme from which we identified a new non-substrate analog TS inhibitor. This inhibitor had a Ki of 160 iiM. Using solid phase in parallel techniques, we synthesized a small library of compounds, analogs of the original lead, and tested them against Lactobacillus caset TS. A first round analog had a Ki of 1.5 jiM. A second round library is now being constructed to further optimize this series of compounds. Structure-based inhibitor discover)' has proven to be a useful technique for the discovery of novel lead compounds for drug design. Combinatorial chemistry is able to introduce focused diversity into a series of compounds efficiently. Our preliminary efforts to combine these techniques have allowed us to discover a novel lead, predict a binding site and rapidly optimize the affinity of the series of compounds.