ERP006749 PRJEB7066 ena-STUDY-UNIVERSITY OF OXFORD-22-08-2014-12:00:57:344-845 Positive selection and compensatory adaptation interact to stabilize non-transmissible plasmids Plasmids are catalysts of bacterial evolution that permit adaptation to novel selective pressures, such as antibiotics. However, it remains challenging to understand how plasmids can persist over the long term as a result of fitness costs associated with plasmid carriage. Classical models predict that horizontal transfer is necessary for plasmids to persist, but whole genome sequencing has recently revealed that almost half of plasmids are non-transmissible. Here we use a combination of mathematical modelling and experimental evolution coupled to whole genome sequencing to investigate how a simple, non-transmissible plasmid, pNUK73, can be maintained in populations of the pathogenic bacterium P. aeruginosa. We find that positive selection for plasmid-encoded antibiotic resistance and compensatory adaptation allow pNUK73 to become rapidly stabilized. Compensatory adaptation by P. aeruginosa recovers the cost associated with plasmid carriage, but compensation alone is not sufficient to maintain the plasmid as a result of segregational loss of plasmids. Positive selection mediated by exposure to antibiotics is necessary to increase plasmid frequency and offset the effects of segregational loss. Crucially, we find that feedback occurs between these processes. Positive selection increases the efficacy of selection for compensatory adaptation by increasing the population size of plasmid-bearing lineages. Compensatory adaptation, in turn, increases the effect of positive selection on plasmid stability by slowing the rate at which the plasmid is lost between episodes of positive selection. Our study provides a new understanding of how plasmids can persist in bacterial populations, and it helps to explain why plasmid-mediated resistance can be maintained in bacterial populations after antibiotic use is stopped. Plasmids are catalysts of bacterial evolution that permit adaptation to novel selective pressures, such as antibiotics. However, it remains challenging to understand how plasmids can persist over the long term as a result of fitness costs associated with plasmid carriage. Classical models predict that horizontal transfer is necessary for plasmids to persist, but whole genome sequencing has recently revealed that almost half of plasmids are non-transmissible. Here we use a combination of mathematical modelling and experimental evolution coupled to whole genome sequencing to investigate how a simple, non-transmissible plasmid, pNUK73, can be maintained in populations of the pathogenic bacterium P. aeruginosa. We find that positive selection for plasmid-encoded antibiotic resistance and compensatory adaptation allow pNUK73 to become rapidly stabilized. Compensatory adaptation by P. aeruginosa recovers the cost associated with plasmid carriage, but compensation alone is not sufficient to maintain the plasmid as a result of segregational loss of plasmids. Positive selection mediated by exposure to antibiotics is necessary to increase plasmid frequency and offset the effects of segregational loss. Crucially, we find that feedback occurs between these processes. Positive selection increases the efficacy of selection for compensatory adaptation by increasing the population size of plasmid-bearing lineages. Compensatory adaptation, in turn, increases the effect of positive selection on plasmid stability by slowing the rate at which the plasmid is lost between episodes of positive selection. Our study provides a new understanding of how plasmids can persist in bacterial populations, and it helps to explain why plasmid-mediated resistance can be maintained in bacterial populations after antibiotic use is stopped.