Logo: to the web site of Uppsala University

The global rise of antimicrobial resistance, particularly among WHO critical priority Gram-negative pathogens such as Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacterales, present a critical threat to modern medicine. This problem is further compounded by a decades-long decline in antibiotic discovery, underscoring the urgent need for innovative screening strategies and the identification of previously unexploited targets. To address this challenge, this thesis employs DNA-encoded chemical library (DECL) technology to discover inhibitors of UDP-2,3-diacylglucosamine diphosphatase (LpxH), a key enzyme of the Raetz pathway associated with lipid A biosynthesis. Through a multidisciplinary research program integrating combinatorial chemistry, computational modelling, in vitro assays and structure-based drug design, three complementary DECL discovery campaigns were pursued. The first focused on the synthesis of a three-cycle, 403,200-member peptide-like DECL. Principal component analysis was used to expand the building block selection to maximize chemical diversity. Affinity selections against human Carbonic Anhydrase II validated the performance of the library through the enrichment of primary sulfonamide pharmacophores. Unfortunately, screening against multiple LpxH orthologs failed to identify enriched binders, revealing the inherent limitations of flexible, linear peptide-like scaffolds for occupying the active site of LpxH. To overcome these structural limitations, a scaffold-free 1.3-million-member library designed to increase molecular rigidity and three-dimensional space diversity was synthesized. Affinity selections against LpxH identified a novel, target-specific chemotype based on a central thiophene core, which was subsequently investigated using molecular docking to establish a plausible binding mode and guide further validation and optimization. Finally, a third effort employed a 7.2-billion-member commercial DECL to discover both known and previously unexplored LpxH inhibitors. Through off-DNA synthesis, in vitro assays and structure-based drug design, these hits were optimized into compounds with improved biochemical and antibacterial activity. Target engagement was further validated using engineered strains, confirming LpxH as the mechanism of action. Together, these findings demonstrate the potential of DECL technology for the discovery of novel inhibitors against challenging Gram-negative antibacterial targets.