The main advantage of fragment-based drug discovery (FBDD) is that the starting point, despite having weak potency, often has high efficiency of interaction with biological target. Optimization of the fragment by adding atoms ideally preserves this initial efficiency, to yield a tractable lead with superior properties. One potential pitfall is that modern hit optimization strategies, particularly those involving molecular modeling, can be deficient in the way they handle these quite small molecular fragments.
In an extension to our existing scaffold hopping methodology we show how a highly valuable FBDD task – fragment growing or linking – can be conducted computationally with high probability of success by using Spark, Cresset’s scaffold hopping and bioisostere replacement software. The use of databases of structures derived from real compounds gives a high degree of diversity combined with reasonable probability of chemical synthesis. Creation of final ‘product’ molecules, which are fully energy minimized before computational scoring, ensures that only linking fragments that can truly work are progressed into the final results.
Uracil DNA glycosylase (UDG) is a potentially interesting target for both cancer and anti-viral therapies. In the body it is important as a DNA repair protein. During DNA replication uracil can be mis-incorporated into DNA or formed enzymatically through the action of cytosine deaminase. Uracil removal is an ongoing process necessary for maintaining DNA integrity. UDG actively targets uracil containing DNA (Figure 1), senses uracil and clips the N-glycosidic linkage (Figure 2). Inhibition of UDG has potentially useful clinical consequences through degrading replication efficiency in cells loaded with viral or cancer genes. A recent effort to produce synthetic inhibitors of this protein1 relied on an active fragment tethering approach, which yielded some interesting albeit weakly potent bis-oxime linked active ligands. We describe an alternative to this, using a molecular modeling technique, which demonstrates the utility of Spark for the efficient elaboration of tractable fragment growing or linking chemistry.