Strukturbasierte Entwicklung und Charakterisierung von Inhibitoren der TGT,ein mögliches Ziel zur Therapie der Bakterienruhr

The target enzyme TGT is essential for the molecular pathogenesis of the inflammatory disease Shigellosis, which is caused by Shigella spp. Incorporation of modified nucleobases into tRNA are required to develop pathogenicity of Shigella. Therefore, TGT replaces the purine base guanine in the wobble position-34 of tRNAsAsn,Asp,His,Tyr by the modified base preQ1. The lin-benzoguanine inhibitors derived from purine bases bind in the active site of the base exchange reaction. By attaching substituents, the adjacent specificity pockets ribose 33 and ribose 34 can be addressed, respectively. In two studies, the water clusters of the named pockets were addressed and displaced by charge neutral monosaccharide substituents of the lin-benzoguanine inhibitors. Thereby, the furanosyl-substituents bind in comparable region as the phosphate group-35 of the natural tRNA substrate. This implies that furanosides might be a surrogate to target phosphate binding sites. As TGT modifies the natural tRNA substrate only as a homodimer, both, blocking of the active site and disturbing the dimeric quaternary structure of the TGT can be considered as inhibition mechanisms. The contact surface of the two TGT monomers is in close proximity to the ribose-34 pocket. Upon analyses of the co-crystal structures of the furanosyl-substituted inhibitors it became obvious, that by addressing the hydrophobic region of the side pocket, an adjacent loop helix motif was disturbed and lost its geometric integrity. This disturbance is associated with rearrangements of up to 20 amino acids. As this motif decisively contributes to the dimeric stability and rearrangements are only visible in co crystal structures, previously characterized ligands have been newly co-crystallized in an elaborate study, as only crystal structures obtained by the soaking method were available. The analysis of these co crystal structures revealed on the one hand different binding poses compared to the soaking structures and on the other hand confirmed further ligands that interfere with TGT interface. The dimer-disturbing effect of these ligands has been confirmed by native mass spectroscopy. One ligand caused a complete rearrangement of TGT monomer units (twisted dimer), leading to an arrangement obviously no longer catalytical active with twisted subunit packing. In a follow-up study, three inhibitors could be characterized, which also interfere with the dimerization of TGT. Two of these compounds co crystallized in both, the conventional and the twisted dimer, respectively. Another study followed a prodrug approach. As the lin-benzoguanine inhibitors do not diffuse through membranes, the scaffold was modified with different carbamate substituents. As a result, PAMPA measurements confirmed ligands of TGT that are membrane-permeable for the first time. Furthermore, from two carbamate derivatives co-crystal structures could be obtained, which due to surprising orientation of the substituents address the ribose-34/phosphate-35 pocket. In a kind of retrospective work the impact of amino acids flanking the guanine-34/preQ1 site on binding affinity of the examined ligands were studied. Based on benzimidazole derivatives, a cavity expansion, previously predicted by a MD-simulation, has been experimentally confirmed by co crystal structures for the first time. Both, hydrazide and guanidine groups attached to the benzimidazole scaffold stabilize a transient cavity opening, due to steric impact on Asp156 and solvation of Gln203 and Gly230. In future design cycles, guanidine-benzimidazoles can be derivatized to address the new pocket and thus exploit new affinity options. Another project followed the characterization of a nucleoside derivative, which inhibits as a putative transition state analogue the base exchange reaction of TGT. The developed deazaguanine-type inhibitor represents a new class of compounds and suggests a possible lead structure that could replace the established lin-benzoguanines in the future. In a large-scale project, nine crystallographers from the group of Prof. Klebe scanned the in-house fragment library by soaking crystal structures of the target enzyme Endothiapepsin. From 361 soaking crystal structures 71 complex structures (hit rate 20%) were obtained. Within the framework of this project, also a standardized refinement-protocol was developed, which optimizes and automatizes the identification of fragments in crystal structures. Based on the structural information of the fragments in the crystal structures new functional groups to address the catalytic dyad, as well as three hotspots remote from the active site were identified. In addition to unmasking false-positive hit from biophysical screening methods, it became obvious that biophysical methods frequently used to initially screen fragment libraries possibly reduces the number of accessible hits and thus the number of X-ray complex structures.