The gut microbiome is key for proper human development and for maintenance of health. However,
microbiome integrity can be severely impacted by consumption of antimicrobial xenobiotics that target some of
its microbial members. Such selection pressure can trigger shifts in its species composition and ultimately
culminate in microbiome dysbiosis that is detrimental to health. Recent drug screens revealed that a quarter of
host-targeted drugs, that are not prescribed as antibiotics, have potent antimicrobial activity at physiological
concentrations. Yet, while the bacterial toxicity of many host-targeted drugs is widely appreciated, the targets
of these drugs in bacterial cells remain mostly unknown. Given the scale of this phenomenon, involving
hundreds of widely prescribed drugs, there is a critical need to develop a high-throughput approach for
identifying the pathways and processes underlying bacterial toxicity. Such systematic understanding will reveal
which host-targeted drugs resemble known categories of antibiotic drugs and if prolonged treatment can
unintentionally contribute to multi-drug antibiotic resistance. Moreover, classification of host-targeted drugs by
mechanism of bacterial toxicity can uncover potentially new targets for by yet-to-be developed antibiotics.
Our overall goal is to investigate the bacterial pathways and processes underlying the toxicity of host-targeted
drugs and reveal the mechanisms that can propel evolved resistance against them. We will use Escherichia
coli as a model system and will leverage on a pooled genetic screening approach we developed to map all
single-gene knockouts that modulate toxicity across multiple drugs. We will screen 117 host-targeted drugs
that we already identified as inhibitory for E. coli and a complementary set of 142 standard antibiotics (with
known targets). Taken together, these systematic measurements of drug sensitivity (3,680 knockout strains X
269 drugs) will allow us to computationally infer a network graph mapping drug-drug resemblance by profiles of
knockout strain sensitivity. This network will be used to annotate the bacterial targets of host-targeted drugs by
both the identity of screen hits and by the drug similarity to well-annotated antibiotics. We will complement the
screen with lab evolution experiments and will identify the resistance mechanisms that emerge under drug
selection. This approach will expand our mechanistic understanding of drug targets by uncovering additional
resistance mechanisms against the same drugs (e.g., gain-of-function mutations). Lastly, we will investigate
collateral drug resistance in evolved strains and natural E. coli isolates and will evaluate if adaptation to host-
targeted drugs poses a risk by unintentionally selecting for multi-drug and antibiotics resistance. The proposed
work will, for the first time, systematically map the mechanisms underlying bacterial sensitivity to host-targeted
drugs and will uncover if chronic administration of specific host-targeted drugs increases the likelihood for
emergence of cross-resistance against specific antibiotics or multi-drug resistance.