Prediction of drug targets in human pathogens

Abstract

The rapid emergence of resistant bacteria makes today's antibiotics more and more ineffective, consequently increasing the need for new pharmacological targets and novel classes of antibacterial drugs. The identification of new and druggable targets in bacteria is a criticalendeavour in pharmaceutical research of novel antibiotics to fight infectious agents. In this PhD project, a novel and innovative model was developed to predict potential drug targets in human pathogens, with the Enterobacteriaceae family chosen as case study pathogen group. This new model combines the singular value decomposition technique with biological filters comprised of a set of protein properties associated with bacterial drug targets and similarity to E. coli protein-coding essential genes. Using this new approach, 99 potential drug target proteins were identified. These proteins participate in eight different functions and areprotein-coding essential genes or similar to protein-coding essential genes of E. coli (strain K12). Hence disruption of the activities of these proteins is critical for cells leading to cell/pathogen death. Among the retrieved candidates, some are from bacteria with described drug-resistance. Additionally, since the identified drug target candidates have no similarity to the human proteome, the chance of occurrence of adverse effects (or at least known ones) on humans is significantly decreased. The tetrahydrodipicolinate N-succinyltransferase (DapD) enzyme was selected for target confirmation, due to its role in the biosynthesis of lysine and mDAP. This lysine biosynthetic pathway generates the intermediate mDAP, which is a building block in the cell wall synthesis in Gram-negative and many Gram-positive bacteria. To the best of our knowledge, this is the first attempt to rationalize the search of compounds to investigate the relevance of DapD as a new pharmacological target. To find target modulators, we built structure-based pharmacophore hypotheses that were later used to screen a virtual library of compounds. Post-screening filtersapplied were based on physicochemical and topological similarity to known Gram-negative antibiotics and allowed the retrieval of more fine tuned compounds. Screening hits passing all filters were docked into the catalytic groove of DapD and 8 of the most promising compounds were purchased from commercial sources to be tested in vitro. Two ligands were distinguished for inhibiting Escherichia coli (ATCC® 25922TM) growth in the antimicrobial susceptibility te (AST). These results provide a positive indication that these ligands can act as Gram-negative growth inhibiting compounds. LpxA, another drug target candidate predicted by application of the new model that is here described for identification of drug targets in human pathogens, was explored by other researcher that claimed to have identified inhibitors for this target. These compounds showed significant inhibition of LpxA activity and growth inhibitory activity against a panel of bacteria, including E. coli. These results validate LpxA as an effective drug target. The rational followed to identify potentially effective ligands and the described in vitro results for inhibition of both DapD and LpxA drug target candidates, establish the proof of concept forthe proposed novel and innovative model to predict potential drug targets in human pathogens. The drug target candidates identified here are worth to be further explored for the treatment Enterobacteriaceae causing diseases. All together, our results validate the model developed for drug target identification, which can potentially be applied to any other group of pathogens. This model paves the way for the discovery of novel drug targets for therapeutic intervention.