Antibiotic resistance is a growing concern to public health. New treatment strategies may alleviate the situation by slowing down the evolution of resistance. Here, we evaluated sequential treatment protocols using two fully independent laboratory-controlled evolution experiments with the human pathogen Pseudomonas aeruginosa PA14 and two pairs of clinically relevant antibiotics (doripenem/ciprofloxacin and cefsulodin/gentamicin). Our results consistently show that the sequential application of two antibiotics decelerates resistance evolution relative to monotherapy. Sequential treatment enhanced population extinction although we applied antibiotics at sub-lethal dosage. In both experiments, we identified an order-effect of the antibiotics used in the sequential protocol, leading to significant variation in the long-term efficacy of the tested protocols. These variations appear to be caused by asymmetric evolutionary constraints, whereby adaptation to one drug slowed down adaptation to the other drug, but not vice versa. An understanding of such asymmetric constraints may help future development of evolutionary robust treatments against infectious disease.
Data for dose response curves
This file contains the data for the dose-response curves as shown in supplementary Fig. S2 of the paper. The columns indicate the following:
$date: the data was collected on two days;
$treatment: specifies antibiotic treatment;
$ID: a number from "0" to "10" that serves as ID for the ascending antibiotic concentrations;
$conc_ng_ml: concentration of antibiotic in ng/ml;
$replicate: name of technical replicate;
$od_600: optical density at 600 nm read after 12h of incubation at 37°C.
Data_Roemhild_2015a.csv
Data for both evolution experiments
This file contains the data from the two evolution experiments, as summarized in Figs. 1,2,3,5,S3,S5, and S6 of the paper. The columns show the following:
$well: position of population in 96-well plate;
$season: number of transfers within the evolution experiment;
$time: time point of od measurement in hours; od was measured in 15 min intervals;
$od_600: optical density at 600 nm;
$ID: a number from "1" to "10" that serves as ID for treatment groups as specified in the paper plus "11" for the evolving nodrug references;
$treatment: treatment type, including monotherapy (mono), regular changes of antibiotics (regular), and random changes of antibiotics (random);
$drug: antibiotic used;
$exp: number of evolution experiment.
Data2_Roemhild_2015a.csv
Data for cross-resistance test
This file contains the data for the cross-resistance test (Fig. 4).
$evoexp: number of evolution experiment or the ancestral population
$evoexp_season: season of the evolution experiment after which populations were streaked on agar plates for characterization
$evoexp_drug: antibiotic present in $evoexp_season and on the agar plates
$evoexp_well: position of population in 96-well plate during evolution experiment
$clone_id: name of the five cfus that were isolated from each $evoexp_well
$test_drug: antibiotic present in the cross-resistance test
$ic: the inhibitory concentration aimed for with the applied concentration; if the value is "E" these are wells with antibiotic but without the bacteria
$concentration_ng_ml: concentration of the antibiotic in the cross-resistance test in ng/ml
$od_600: optical density at 600 nm after 12h of incubation at 37°C
Datafile3_Roemhild_2015_Corrigendum.txt