by Clara Hong, John Scheuring, and Saurab Faruque
Most antibiotics require multiple mutations to develop high levels of resistance. In order to better understand the pathways to sequential mutations that lead to resistance, an experimental design that adjusts to the changing levels of evolution in bacterial populations is required. Previous experiments failed to explore genetic pathways to resistance because they focused on exposing the bacterial populations to fixed drug concentrations. However, this experiment focuses on regulating drug concentrations to ensure continuous stress on the evolving population. The morbidostat assists in this process by setting a constant growth rate, with an antibiotic inhibitor. The inhibitor is continuously adjusted in response to two conditions.
Once the optical density (OD) is greater than the threshold and the growth rate exceeds the dilution rate, the morbidostat increases the concentration of the inhibitor. To clarify, the OD indicates the density of cell volume in a given culture based on light absorbance measurements. The device maintains the cultures in the mutantselection window (MSW) — the range of drug concentration that is below the level of total bacterial decimation and above bacterial prosperity. The MSW changes based on the evolution rate: as resistance grows in a cell culture, the MSW will increase along with it. By monitoring and calculating the growth rate and dilution rate, the device is controlling the drug concentration input in accordance with the MSW. In this particular experiment, bacterial growth was inhibited by 50%, a number which is kept constant by adjustment of the drug concentration.
In the experiment, three antibiotic drugs were tested on drugsensitive E.coli. Five isogenic (identical in genotype) populations were grown in parallel for result reproducibility. As a result, a total of fifteen evolving populations inhibited by chloramphenicol, doxycycline, and trimethoprim were grown. The experiment followed the populations by measuring their growth rate over the drug concentration. Based on the graph built from the resulting data of drug concentration over time, resistance was shown to increase dramatically. However, while resistance to chloramphenicol and doxycycline was graphed smoothly over time, trimethoprim showed a stepwise graph. The results suggest that resistance to trimethoprim derives from mutations in smaller regions in the gene which would explain the frequent gap in time where no change occurred till the next occurrence of mutations. To test this hypothesis, the bacterial populations underwent wholegenome sequencing.
There were several interesting aspects of the sequencing result for doxycycline and chloramphenicol. Although the two drugs target ribosomes, no mutations on ribosomal genes were found. This absence of mutation may signify a negative epistasis —when two or more mutations result in a smaller effect than an individual mutation — or cost in growth rate. Another finding showed positive cross resistance and similarity in growth patterns for the populations grown under both antibiotics. Despite their phenotypic similarities, the populations exhibited differing mutations. This result suggests that because of a large target size, the number of sites at which loss of function mutations may occur, there are numerous ways to resist stress created by chloramphenicol or doxycycline.
The sequencing result for Trimethoprim showed a very limited target size focused on amplifying the DHFR gene. This gene has been found to be important for cell growth and exponentiation. Due to the possibility of an antibiotic successfully inhibiting one copy of the gene, amplification provides another copy to act as a backup and secures this vital gene. Trimethoprim is a drug that specifically inhibits the DHFR gene and because of this specificity, evolution occurred slowly in a stepwise manner since certain mutations must occur in a sequential order in this target size for higher resistance.
Through the use of the morbidostat device, a successful experiment comparing different resistance evolutionary pathways was conducted. The device’s ability to change drug input based on observations of culture growth allows for better prediction in multidrug resistance evolution over an extended time. Through the device we have seen two different resistance pathways and its significance in the molecular scale. The results show that further experimentation can be done using the morbidostat to identify other paths to resistance, understand the reasoning for such behavior, and apply the results to our own reallife environment.
Toprak, E et al. “Evolutionary Paths to Antibiotic Resistance under Dynamically Sustained Drug Selection.” Nature Genetics 44:101-105 (2012).