Antibiotic compounds targeting bacterial RNA polymerase holoenzyme assembly and other protein-protein interaction inhibitiors identified with an in vivo-bret drug discovery platform

The research and development of a new drug is a lengthy and costly process, which requires considerable investment by the pharmaceutical industries, with a very low success rate and a constant need for innovative approaches. The so-called “reverse approach” (or target-based drug discovery) is based on the screening of small molecule libraries, to identify "hit compounds" capable of interacting and modulating the biological activity of the target of interest. In recent years, protein-protein interactions (PPI) have emerged as promising new targets, especially considering their key role in most cellular processes, under both physiological and pathological conditions. The aim of this thesis work is the development of an in vivo high-throughput platform based on the BRET (Bioluminescence Resonance Energy Transfer) technology as a tool to monitor PPI and to screen compound libraries for the identification of new potential PPI inhibitors. To this end, I reproduced various interactions of marked biomedical interest in the unicellular eukaryote Saccharomyces cerevisiae and monitored them using the yeast-BRET (yBRET) assay. The set up and validation of yBRET was carried out using the HDM2-p53 interaction, involved in cell-cycle control, as a proof of concept case study, given the availability of effective inhibitors targeting this particular PPI. The same yBRET technology was subsequently applied, and further optimized, for the study of the interaction between the ' subunit of bacterial RNA polymerase (RNAP) and the 70 specificity factor of Escherichia coli, as a new target for the discovery of antibiotics capable of interfering with transcription complex assembly, thus blocking pathogen viability and propagation. The screening of a subset of computationally pre-selected molecules plus a collection of small-molecule libraries of different origin was performed on this interaction, for a total of over 18,000 different compounds. Hit compounds emerging from yBRET were validated with a dedicated ELISA, to confirm the ability of the hits to disrupt the RNAP-70 interaction, and with the use of an in vitro transcription assay to verify their ability to inhibit RNAP activity. Finally, I demonstrated that the validated lead compounds are in fact endowed with antimicrobial activity against both Gram-positive (Bacillus subtilis) and Gram-negative (E. coli) bacteria, displaying growth inhibition starting from 200 μM. This led to the identification of new potential antibiotics, some of which have chemical structures similar to those of already known 70-' interaction inhibitors (a class of indole-containing compounds), while others have completely new scaffolds. I also applied the yBRET screening to other interactions of biomedical interest, such as receptor-ligand interactions involved in human immuno-modulatory pathway. I focused, in particular, on the extracellular domains of 2B4-CD48, whose immuno-inhibitory interaction is involved in T cell exhaustion, and CD40-CD40L, whose immuno-stimulatory interaction is associated to various autoimmune inflammatory pathologies of the central nervous system and in lymphomas. For the latter interaction, I developed a surface exposed version of yeast BRET (named, syBRET), which allows the expression of the interacting partner proteins on the yeast cell wall. This allowed the reconstruction of the pentameric CD40-CD40L complex, formed by the CD40L trimeric ligand associated with two monomers of CD40 receptor, each containing eight disulfide bridges. The syBRET was validated with Suramin, a known inhibitor of this interaction. Overall the yBRET methodology proved to be a fairly robust, versatile and high-throughput tool for PPI-orineted drug discovery.