Expanding The Breadth Of Drug Discovery Through Experimental And Computational Techniques

Nicole Esther Kerstetter, University of Pennsylvania

Abstract

Drug discovery and design is extremely important in today’s world in order to treat andcure diseases which have encountered resistance and those that have not yet had a cure. Different methodologies and techniques are necessary to discover these new therapeutics. The goal of this thesis is to explore a variety of drug discovery methods and their effectiveness in designing new drugs and to expand on their capabilities. Firstly, solvent mapping was modified for use in Reverse Micelle NMR (RM NMR) to determine its effectiveness in locating small molecule hotspots on protein surfaces. The results of this study showed weakly binding molecules productively interacting with the smooth surface of Interleukin-1β, a drug target for inflammatory diseases. Their binding location and ability varied among the molecules with varying functional groups. These small molecules may act as building blocks for future therapeutics. Secondly, a RM NMR Fragment Based Drug Discovery method was used to explore the binding of hydrophilic fragments typically missed in standard screenings on Interleukin-1β. A ‘hit’ rate of ~10% of the total library was achieved, with fragments binding as weakly as 200 mM, surpassing the detection limit of current screening methods. These hits comprised a total surface coverage of two-thirds, showing it’s possible to effectively screen proteins without tight-binding pockets. Additionally, the surface coverage illustrates it is possible to design an inhibitor based on most desired surface locations. Lastly, Molecular Dynamics and docking simulations were used to uncover the binding pose and mechanism of action of the GTx-560, a potential drug for Aldo-Keto Reductase 1C3 (AKR1C3), an important prostate cancer target. GTx-560 was found to bind in the opposite orientation of the ‘cognate pose’, with the trifluorobenzene ring in the oxyanion hole of the binding pocket. Pump-probe molecular dynamics determined allosteric effects from GTx-560 biding may contribute to the loss of coactivator function in AKR1C3 by preventing its binding to the Androgen Receptor. The work presented here shows the future of drug discovery and how modifications to existing methods can open the door to a wider range of drug targets as well as the molecules used in drug design.