Computational modelling of metal-mediated protein-ligand interactions

Although metalloproteins account for nearly half of all proteins in nature, computational
modelling of metal-mediated protein-ligand interactions is understudied and molecular
mechanics programs and force field parameters compatible to proteins and transition
metals are not readily available. Within this thesis, various computational approaches
were pursued towards the design of artificial metalloenzymes and the modelling of metal-
mediated protein ligand interactions. Four challenges were identified and addressed. The
first consisted of the identification of suitable protein scaffolds for the creation of artificial
facial-triad motifs. The second part focused on the development of reliable force field
parameters for the anchoring of sulfonamide bearing anchors within human carbonic
anhydrase 2. In order to reliably predict the geometry of catalytically relevant piano stool
artificial cofactors within host proteins, the third part aimed at developing force-field
parameters (using the Valbond formalism) for d6 -piano stool complexes. Finally, the
fourth step combined the above developments towards the reliable prediction of first and
second coordination sphere environments around artificial cofactors/inhibitors.