Research Project
9310554
Project Summary/Abstract The proliferation of antibioticresistant pathogens has increased attention to the use of drug combinations to combat the evolution of resistance. However, while the phenotypic and genotypic evolutionary paths towards multidrug resistance can be constrained by tradeoffs between resistance to one drug and susceptibility to other drugs or the humanhost environment, evolution often circumvents these obstacles. Focusing on both E. coli and clinical species, we will unravel the potential and limits of such approaches for constraining evolution. In Aim 1, using a novel selection device, the MEGAplate, which follows multiple diversifying bacterial lineages as they migrate and evolve on large antibiotic gradient landscapes, we will comprehensively map the repertoire of multimutational paths to highlevel resistance to a range of antibiotics as well as to pairs of antibiotics presenting adaptive tradeoffs. Coupled with wholegenome sequencing and automated highthroughput phenotyping, this device will enable us to identify common and specific adaptive mechanisms, reveal the predictability and further evolutionary potential of each mutational step, and test whether channeling evolution towards ?quasi deadend? genotypes can impede longterm adaptation. In Aim 2, we focus on cycling drug pairs presenting reciprocal adaptive tradeoffs and examine vulnerabilities of the approach. Going ?beyond the average?, we will construct a deep library of singlestep mutants selected on one of several different antibiotics and test en masse their crossresistance to the other antibiotics. These results will identify rare ?escape? mutants which circumvent inherent tradeoffs between resistance to these drugs. Synthetically combining pairs of mutations which individually show resistance to one drug yet sensitivity to the other, we will reveal whether such mutations can interact nonadditively leading to resistance to both drugs. In Aim 3, we examine the effectiveness yet weaknesses of new ?selectioninverting? compounds which we recently found to act preferentially against bacteria expressing the tetracycline resistance efflux pump. We will use a microfluidic device to follow single cells while switching between tetracycline and the new compounds, thereby identifying the optimal regime for selection against the pump. We will then systematically mutate the tetracycline pump to identify mutations that escape the tradeoff. Finally, in Aim 4, we focus on the differences between in vitro and in vivo adaptive pathways to identify the evolutionary constraints imposed by the human host environment. We will use longitudinal isolates from clinical outbreaks to identify which of the resistant mutations observed in the lab also appear during pathogen evolution within the human body and which others are in vivo inaccessible. Comparing evolution of pathogens in immunocompetent versus immunosuppressed patients will point to evolutionary constraints imposed by innate immunity. In summary, our proposed research will reveal the genotypic and phenotypic constraints that govern evolutionary pathways towards multidrug resistance, both in the lab and during longterm infection in humans. Our longterm goal is to help design drug regimes that better prevent the emergence of resistance and to develop algorithms which based on the database of observed mutationalpaths will allow genomebased diagnostics of microbial infections that can both predict current resistance profile and anticipate its future evolution.
