Bivalent drug repurposing and evolutionary escape modelling

Abstract

Antimicrobial resistance (AMR) poses a formidable global health challenge, challenging the efficacy of existing antibiotics and necessitating innovative solutions. In this project, an framework to combating AMR through bivalent drug targeting was explored. First, a framework for searching for proteins is described, leveraging the use of FATCAT to search for global and local similarities pairwise between various proteins. Here, two proteins in the folate pathway, folB and folX were found to have striking global similarity, and were chosen for further analysis. Ancestral sequence reconstruction was performed on the proteins to examine the plausibility of a common ancestor, and what such a sequence might look like. From there, a strategy for virtually screening compounds to find potential candidates that would bind to both folB and folX. A shortlist of compounds was found, at which point \emph{in vitro} experimental validation of the compounds on the proteins was conducted, encapsulating antibacterial growth measurements, IC50 determination, and recovery and overexpression experiments. Many compounds indicated promise in antifolate activity. Finally, evolutionary models were initialized on data previously available from experiments on folZ. While these models were less than successful, the modularity and ease of iteration makes for easy future improvements, given incoming data from evolution experiments. Ultimately, a promising workflow has been presented towards developing and/or repurposing therapeutics to combat AMR, providing a roadmap for the development of effective antimicrobial strategies amidst the evolving landscape of resistance.