Abstract
Conventional antibiotic drug discovery selects leads based on bacterial growth inhibition. This approach is ineffective against growth-arrested persister cells. When the treatment stops, persister cells revert to normal cells, causing the infection to relapse. To address the challenge of persistent infections, a paradigm shift in antibiotic development is needed to identify new leads that can eradicate dormant cells. Based on our foundational study, we recently proposed a set of principles for developing new persister killing agents. Here, we report the discovery of new leads that are effective against persister cells using a tailored chemoinformatic clustering algorithm based on these principles. We focused on persister penetration using a small compound library that has known antimicrobial activities against normal cells. Experimental testing of eleven compounds identified from clustering led to the discovery of five new compounds that can effectively penetrate and kill persister cells of Escherichia coli HM22. The top leads were further tested and also found active against persister cells of Pseudomonas aeruginosa and uropathogenic E. coli (UPEC), as well as UPEC biofilms and biofilm-associated persister cells. This rather high yield demonstrates the potential of this new rational approach in identifying effective agents against dormant cells, a root cause of persistent infections that is largely missed in conventional screening.