Kristallstrukturen mit kleinen Sondenmolekülen: Ausleuchten von Bindetaschen, Startpunkt für ein Fragment-basiertes Wirkstoffdesign und Erhalten von Phaseninformationen zur Strukturlösung

This thesis addresses a variety of different approaches in which structure based drug design plays an important role. Starting with de novo structure determination of macromolecules, followed by fragment screening via X-ray crystallography and finally the structure based lead discovery against malaria and tuberculosis. The emphasis of this thesis was to find small, soluble molecules which mark not only the position of functional groups but also the putative scaffold position, i.e. an aromatic ring system, to create a protein-based pharmacophore. Furthermore, the active sites mapped by this experimental approach can be used as a reference to develop and validate in silico methods to calculate hotspots of binding, such as implemented in the program DrugScore. In this study a selection of different small and highly soluble molecules with a molecular weight of 67 – 297 g/mol were either soaked or co-crystallized in varied concentrations into crystals of different protein targets and subsequently screened by X-ray crystallography. Complexes of TLN, PKA, DXI and IspD with phenol, aniline, urea, N-methylurea and 1,2-propanediol could be determined. Furthermore, it could be evidenced that the positions of the functional groups and ring systems of these molecules were well in agreement with most in silico hotspot predictions calculated by the program DrugScore. The relevance of the observed probe molecule poses for drug binding could be confirmed by superimposing much larger ligands for which complex crystal structures have been determined. It was possible to match not only the position of the functional groups by the found molecular probes but also parts of the scaffold of the larger ligands. The fact that small molecular probes can still be discovered as kind of original seeds from which an entire inhibitor can be grown, opens a very promising and prospective option using the binding of such small probes as a starting point of a fragment-like de novo design. Furthermore, molecular probes such as phenol, aniline, urea, N-methylurea and 1,2-propanediol seemed to be suitable to be applied rather general to many proteins to perform an experimental binding site mapping. Particularly, if they are used in parallel and are further supported by computer tools such as DrugScore to fully map the binding pocket for a subsequent inhibitor design. In collaboration with the group of Prof. Joel L. Sussman in Rehovot, Israel it was possible to incorporate xenon under pressure into crystals of TLN and TcAChE and perform successfully phase determinations via SAD, with in-house data. In addition, further enzymes as EP, TGT and Sap2 were also derivatize with xenon. Unfortunately, only the 3D-structure of EP could be determined additionally to TLN and TcAChE by the SAD method with in-house data. Nevertheless, with the help of TLN, EP and TcAChE as model systems it could be demonstrated on one hand that single (TLN & EP) or double (TcAChE) bound xenon atom is sufficient to produce interpretable density maps, using various programs and program combinations for processing, scaling, phasing and model building and on the other that xenon SAD phasing, using data collected on a home source, may provide an effective, efficient and low-cost alternative to MAD phasing at synchrotron sources. The last topic of this thesis was to discover a new lead structure against malaria and tuberculosis by inhibiting IspD, an enzyme of the non-mevalonat pathway. On the basis of the X-ray structure of IspD in complex with 1,2-propanediol it was attempt to create a new lead by linking this core fragment with an other substrate analogous fragment. Therefore a thermofluoro assay and X-ray crystallography was used for screening purposes. Even the 3D-structure of IspD was known new crystallographic conditions needed to be found, which allow exhaustingly screening via soaking. In the development of new reproducible crystallization conditions and by performing the thermofluoro assay a high amount of soluble protein was required. Therefore the existing purification protocol was been amended and the yield rate could be enhanced from 134 mg up to 203 mg soluble protein out of 4 l bacteria culture. By finding new crystallization terms over 1000 different conditions have been tested. Finally, the diffraction pattern of IspD crystals could be improved from a starting resolution of 4 Å up to 2.34 Å using an in house X-ray source. Although six different fragments were found in the thermofluoro assay, but up to now it was not possible to detect one of these potential fragments in the binding site of IspD. Nevertheless, it was possible to deposit the crystal structure of IspD in complex with 1,2-propanediol in the PDB. This crystal structure is the first non substrate analogous complex structure of IspD in the PBD.