Based on the three lead structures, substructure and similarity searches were conducted and the molecules obtained were docked and scored in SeeSAR. Furthermore, SeeSAR’s FastGrow function and included libraries were used on different bonds of the lead structures. This resulted in thousands of molecules which were docked in a standard docking as well as in template-based dockings and afterwards scored in SeeSAR. One of the three lead structures was bound in the binding site of a natural interaction partner of the target protein. For this molecule many promising new compounds were obtained. They were all docked and scored. For scoring, criteria in SeeSAR were selected e.g. no inter or intra molecular clashes, a molecular weight smaller than 500 g/mol or a log(P) ≤ 5. Further, SeeSAR’s intuitive colour-code visualized the results very well and made the evaluation of the compounds easier. Using this, the number of promising molecules was narrowed down to 30 molecules. Unfortunately, their synthesis could not be accomplished. In SeeSAR’s editor mode derivatives of these molecules were created, that also gave rise to new interactions in-silico, e.g. a halogen-pi-interaction. The synthesis and validation of these molecules is still pending. Furthermore, the similarity search based on this lead structure led also to 2000 molecules. They were also docked and scored. With the selected criteria and SeeSAR’s visualization 14 molecules were chosen that could serve as new lead structures. These are commercially available and will be analysed further. They will be used in soaking experiments to validate them crystallographically as binders. Furthermore, they will be used in thermal shift assays and microscale thermophoresis. For the other two lead structures that are binding in the dimer interface of the target protein two strategies were followed – stabilisation and destabilisation of the homodimer formed by the target protein. Many molecules were obtained via the FastGrow function. The compounds were docked and scored but none of them interacted well with the target protein. This meant the stabilisation strategy was tricky which is probably attributed to sterical hindrance due to space limitations in the binding site. The destabilisation strategy led unfortunately also to no promising compounds.
After 1 year, Johanna has achieved the following goals:
- The most important goal was to obtain enlarged molecules based on the lead structure that binds in the binding site of one of the natural interaction partners of the target protein. With SeeSAR’s FastGrow function and its included fragment libraries thousands of enlarged molecules were obtained. FastGrow was undertaken on different bonds of the molecule. Obtained molecules were docked and scored in SeeSAR. For the FastGrow via the amine function of the molecule hundreds of molecules seemed promising and they were narrowed down to 30 molecules. SeeSAR’s intuitive colour-code made this task easier. These 30 molecules needed to be synthesized for experimental validation. Unfortunately, the synthesis couldn’t be accomplished so derivatives of the molecules were created in SeeSAR’s editor mode and then docked and scored. The synthesis of the molecules and their experimental validation remains to be done.
- The next important goal was to find alternative lead structures in case the synthesis of the above mentioned molecules would turn out as complicated. For the lead structure that binds in the binding site of one of the natural interaction partners of the target protein compounds obtained in a similarity search were docked and scored. This resulted in 14 molecules. These are commercially available and will be analysed further. They will be used in soaking experiments for a crystallographic validation. Furthermore, they will be used in thermal shift assays and microscale thermophoresis assays. If they can be validated as binding molecules they will be used as lead structures and be further developed for example with SeeSAR’s FastGrow function.
- Another goal was to find out if the two lead structures that bind in the dimer interface of the target protein could be used to stabilise or destabilise the dimer which could both lead to its loss of function. For this FastGrow in SeeSAR was applied on various bonds of the molecules. Obtained compounds were docked and scored. None of the molecules seemed to interact well with the protein so that the strategy of stabilising did not seem to be reached with these molecules. For the destabilisation and therefore a dissociation of the dimer, molecules that were breaking protein-protein interactions were needed. Therefore, FastGrow was applied to the molecules with each of the monomers of the protein, so the binding site was modified. Molecules obtained so far were docked and scored. Based on their interactions and estimated affinities they do not seem to be promising in following the destabilisation strategy. This knowledge gained led the focus to the third lead structure.