Mathematical modeling to predict the fitness cost associated with triclosan tolerance in Salmonella enterica serovars

Adaptation of Campylobacter field isolates to propionic acid and sorbic acid is associated with fitness costs

“…The proportions of all adapted variants were shown to decrease over time, as indicated by K c values greater than zero. One possible reason for the findings is that the tolerance-mediating adaptation was associated with a loss of bacterial fitness, resulting in wild-type isolates outcompeting their adapted variants as stated in previous studies (Zhang et al 2006;Luangtongkum et al 2012;Rensch et al 2015). Mutations conferring resistance to antibiotics often impose fitness costs, as many antibiotics target important biological functions in the cell and resistance to them may either disrupt those processes or lead to increased energy consumption (Melnyk et al 2015).…”

“…Statistical analysis of the changes in log ratios of the wild‐type isolates and their adapted variants was performed using a linear regression model in accordance with Rensch et al . (2015). Briefly, the log 10 differences ( LD ) were calculated by subtracting the log 10 concentrations of the adapted variants from the log 10 concentrations of the respective wild‐type isolates.…”

“…Furthermore, 95% uncertainty intervals were determined for the estimated parameter K c , for the predicted time‐dependent mean LD (referred to as ‘confidence intervals’) as well as for the predicted individual observations of LD for each Campylobacter isolate (referred as ‘prediction intervals’) (Rensch et al . 2015). The linear regression models were performed using the software R (R Core Team 2019).…”

“…Growth competition experiments were performed as described by Rensch et al (2015) with minor changes. For this purpose, the respective Campylobacter wild-type isolates were co-cultured together with their adapted variants.…”

“…The observed LD values were fitted to a linear regression model with intercept fixed as zero to set the concentrations of the wild-type isolates and their adapted variants equal at time zero. The estimated regression parameter K c can be considered as the difference between the growth rate constant of the wildtype isolate minus the growth rate constant of the adapted variant (Rensch et al 2015). Furthermore, 95% uncertainty intervals were determined for the estimated parameter K c , for the predicted time-dependent mean LD (referred to as 'confidence intervals') as well as for the predicted individual observations of LD for each Campylobacter isolate (referred as 'prediction intervals') (Rensch et al 2015).…”