Replacement soaking for human tankyrase 2 enables studies on substrate analogues and inhibitors.

Abstract

Crystallography has facilitated the development of inhibitors in modern drug discovery. Complex structures are traditionally obtained by co-crystallization or soaking methods, but they are not always successful; co-crystallization may lead to selective crystallization of the apoprotein, and compounds cannot necessarily be soaked into preformed crystals. Successful endeavours in this field include human tankyrases TNKS1 and TNKS2, which are promising drug targets controlling a wide range of signalling events. In the case of TNKS2, apo crystals have a closed conformation in the donor loop, delineating the NAD⁺-binding active site, and compound soaking requires a large conformational change, which often damages the crystals and deteriorates the diffraction quality. This work describes a crystal form of TNKS2 that can be used efficiently in replacement soaking (also known as cross-soaking) when TNKS2 is co-crystallized with a low-affinity inhibitor. Replacement soaking is demonstrated with recently described nanomolar inhibitors as well as a nonhydrolysable analogue of NAD⁺, which all occupy the substrate-binding site. The obtained complex structures offer new insight into the binding modes of the inhibitors and the substrate NAD⁺, and comparison to previously published complex structures of closely related PARP proteins with NAD⁺ analogues reveals key characteristics of the NAD⁺-binding site of TNKS2. In cases where traditional methods fail to yield complex structures, the replacement soaking method provides an alternative method, which was demonstrated here to be necessary for elucidating the complex structures.