Stressor interaction networks suggest antibiotic resistance co-opted from stress responses to temperature.

Environmental factors like temperature, pressure, and pH partly shaped the evolution of life. As life progressed, new stressors (e.g., poisons and antibiotics) arose as part of an arms race among organisms. Here we ask if cells co-opted existing mechanisms to respond to new stressors, or whether new responses evolved de novo. We use a network-clustering approach based purely on phenotypic growth measurements and interactions among the effects of stressors on population growth. We apply this method to two types of stressors-temperature and antibiotics-to discover the extent to which their cellular responses overlap in Escherichia coli. Our clustering reveals that responses to low and high temperatures are clearly separated, and each is grouped with responses to antibiotics that have similar effects to cold or heat, respectively. As further support, we use a library of transcriptional fluorescent reporters to confirm heat-shock and cold-shock genes are induced by antibiotics. We also show strains evolved at high temperatures are more sensitive to antibiotics that mimic the effects of cold. Taken together, our results strongly suggest that temperature stress responses have been co-opted to deal with antibiotic stress.

Variation in Mutant Prevention Concentrations.

Objectives:Understanding how phenotypic traits vary has been a longstanding goal of evolutionary biologists. When examining antibiotic-resistance in bacteria, it is generally understood that the minimum inhibitory concentration (MIC) has minimal variation specific to each bacterial strain-antibiotic combination. However, there is a less studied resistance trait, the mutant prevention concentration (MPC), which measures the MIC of the most resistant sub-population. Whether and how MPC varies has been poorly understood. Here, we ask a simple, yet important question: How much does the MPC vary, within a single strain-antibiotic association? Using a Staphylococcus species and five antibiotics from five different antibiotic classes-ciprofloxacin, doxycycline, gentamicin, nitrofurantoin, and oxacillin-we examined the frequency of resistance for a wide range of concentrations per antibiotic, and measured the repeatability of the MPC, the lowest amount of antibiotic that would ensure no surviving cells in a 1010 population of bacteria. Results: We found a wide variation within the MPC and distributions that were rarely normal. When antibiotic resistance evolved, the distribution of the MPC changed, with all distributions becoming wider and some multi-modal. Conclusion: Unlike the MIC, there is high variability in the MPC for a given bacterial strain-antibiotic combination.