Rational design of small molecule modulators targeting the transcriptional function of the Salmonella Typhimurium response regulator PhoP

Pathogenic bacteria have acquired resistance against most clinical antibiotics. Infectious diseases caused by these bacteria are one of the biggest challenges in public health.  There is an urgent need to develop new mechanisms to treat infectious diseases. Traditional antibiotics are designed for their ability to kill or inhibit the growth of pathogens.  Both mechanisms induce strong selective antibiotic resistance pressure.  It has been suggested that interference with virulence factors without killing or inhibiting the growth of pathogens is a promising strategy for developing new antimicrobial drugs to combat antibiotic resistance. The PhoP/PhoQ two-component system consisting of a histidine kinase PhoQ and a cognate response regulator PhoP, is a major regulator of virulence function in the pathogen Salmonella enterica servovar Typhimurium and other gram-negative species. The PhoP protein is responsible for regulating the expression of multiple genes involved in controlling virulence, including biofilm formation and quorum sensing, intracellular survival, and resistance to antimicrobial peptides. Therefore, modulating the transcriptional function of the PhoP protein is an attractive target for identifying new antimicrobial agents. 

The main objective of this thesis was to study the druggability of the PhoP protein, and to identify small molecules to modulate the transcriptional function of PhoP as new type of antimicrobial drugs. Computer aided drug design approaches have played an important role in facilitating drug target identification and assessment, hit identification, hit-to-lead selection, and lead optimization.  In this thesis, computational methods and biological assays were integrated to rationalize and speed up the drug discovery process.  

In the first part of this thesis, we studied the druggability of the PhoP protein using molecular dynamics simulations, which can deliver insight into the dynamics and druggability of the PhoP protein. Since phosphorylation-mediated dimerization in the receiver domain of PhoP is essential for its transcriptional function, we hypothezise that disrupting or stabilizing of protein-protein interactions at the dimerization interface may inhibit or enhance the expression of PhoP-regulated genes. In this study, we performed molecular dynamics simulations on the active and inactive (dimeric and monomeric) PhoP regulatory domains, followed by binding free energy calculations, pocket-detecting screening, and a quantitative hot-spot analysis in order to assess the druggability of the protein. We identified two different druggable binding sites at the dimerization active site (the alpha4-beta5-alpha5 face) with energetic hot-spot areas, which can be used for structure-based drug design. These results showed that the dimer interface of the PhoP receiver domain is an attractive target and suitable for the design of small molecules.

To validate the putative binding pockets of the PhoP protein, we used a virtual screening protocol combined with pharmacophore-based and docking-based methods to identify small molecules that could potentially bind to the binding site. From the results of these virtual screenings, a limited number of compounds were selected, purchased from commercial vendors, and biologically validated using a gene reporter fusion assay. Based on these experimental results, chemical derivatives of the experimentally confirmed compounds were selected with expectation of improved activity. Finally, several novel scaffolds were proven to be able to modulate the transcriptional function of PhoP, either by enhancing or inhibiting expression of  PhoP-dependent genes. These novel structures can serve as starting points for developing virulence inhibitors as new antimicrobial drugs. 

In summary, we used molecular dynamics simulations to study the dynamics and protein-protein interactions of the PhoP protein, and to provide a computational druggability assessment. Based on the druggability studies of this target, we combined computational virtual screening methods with an experimental assay to identify the first hit compounds. These identified novel scaffolds can serve as starting points for developing antimicrobial drugs with new mechanisms that are less likely to induce antibiotic resistant phenotypes.