Predicting bacterial fitness cost associated with drug resistance.

Objectives: It has been proposed that antimicrobial resistance could be associated with a fitness cost in bacteria, which is often determined by competition experiments between isogenic strains (wild-type and mutant). However, this conventional approach is time consuming and labour intensive. An alternative method was developed to assess the fitness cost in drug-resistant bacteria.

Methods: Time-growth studies were performed with approximately 1 × 10(5) cfu/mL of Acinetobacter baumannii or Pseudomonas aeruginosa at baseline. Serial samples were obtained to quantify the bacterial burden over 24 h. The growth rates (K(g)) of isogenic strains (antibiotic susceptible and resistant) were determined individually and used to predict their relative abundance in a co-culture over an extended period of time. The predicted difference between the two strains was subsequently validated by in vitro growth competition experiments.

Results: The growth rates of A. baumannii were not significantly different in different strengths of growth medium. The difference in bacterial burden observed in competition studies was in general agreement with the predicted difference based on K(g) values, suggesting good predicting ability of the mathematical model.

Conclusions: The proposed mathematical model was found to be reasonable in characterizing bacterial growth and predicting the fitness cost of resistance. This simple method appears robust in the assessment of fitness cost associated with drug resistance and warrants further investigations.