Agent-based model of resistant bacterial evolution in an heterogeneous medium

Abstract

In this chapter, an agent-based model was developed using NetLogo to study bacterial evolution and the process in which bacterial populations are able to acquire antibiotic resistance in an heterogeneous medium with an escalated antibiotic gradient. Molecular interactions between wild-type (WT) E. coli and the antibiotic trimethoprim (TMP) were abstracted and conceptually represented to correctly model the complex biological system and interactions. The antibiotic gradient was spatially separated in a plate, starting from 1 × minimal inhibitory concentration (MIC) of WT E. coli and scaling by 10-fold up to 1000 × MIC. A set of experiments was performed under the same initial conditions, varying between five different dnaQ alleles, to measure the mean antibiotic resistance, bacterial population, deaths caused by antibiotic and deaths caused by the lack of nutrients to determine if the simulations had biological meaning. One hundred simulations were carried out for every allele. Each run of the model had different, but similar results, meaning that the inherent variability was eliminated and the overall behavior of the model was properly characterized. Tukey tests were performed to measure significant differences between results from the different alleles. A typical bacterial growth curve with well defined phases was obtained. Bacteria were able to acquire antibiotic resistance and migrate from the borders towards the center of the plate, where the concentration of TMP was 1000 × MIC. Those results alongside the graphical environment gives enough prove that the model is well implemented to study the bacterial evolution and the process in which they are able to develop antibiotic resistance.