Heteroresistance (HR) is a special type of antibiotic resistance that involves the occurrence of a subpopulation of bacteria that are more resistant than the main subpopulation. While these resistant bacteria are present at very low frequencies (10^-7), presence of antibiotics can lead to their rapid enrichment in a population. The HR phenotype is widely observed in various bacteria and against several clinically relevant antibiotics. Given their inherent instability and low frequency in the population, they are particularly difficult to detect and treat efficiently.

         In our group at Uppsala University, we have been studying the mechanisms underlying HR and seeking ways to combat its spread. The most common mechanism responsible for conferring HR phenotype are tandem amplification of resistance genes, which when present in a single copy are not sufficient to confer resistance, but when present in multiple copies confer high-level resistance. Recent research by our group showed that a majority of the detected heteroresistance in Gram-negative clinical isolates stems from unstable gene amplifications¹. Since gene amplifications incur a substantial fitness cost, these amplifications are transient and rapidly lost from the population in absence of selection further impairing their detection².

            The magnitude of the fitness cost associated with gene amplification is one of the main factors that determines whether they can be stably maintained in the population or not. We, therefore, asked how and by which mechanisms HR populations can ameliorate the fitness cost associated with tandem gene amplifications. These questions are pivotal as they will influence not just the future evolutionary path of such strains, but also hold significant implications for the detection and clinical management of infections caused by heteroresistant bacteria.

            In this study, we investigated the predominant genetic mechanisms that can offset the fitness burden associated with gene amplifications in four clinical isolates of Gram-negative bacteria exhibiting the HR phenotype to three different antibiotics. When subjected to increasing concentrations of their respective antibiotic several fold above their Minimum Inhibitory Concentration (MIC) of the main susceptible population, resistant mutants were enriched very rapidly. These mutants, apart from being highly resistant as indicated by their very high MICs, were also accompanied by significant amplification of resistance gene copies and decreased growth rates. To address our question of how the fitness cost is compensated, these mutants were evolved for several generations in presence of the same antibiotic on which it was selected. We found that the bacteria could rapidly acquire less costly chromosomal resistance mutations, which together with a reduction in the gene amplifications of the resistance gene, resulted in a fitness increase while maintaining high levels of resistance. This is important as it shows that transient gene amplification acts as an intermediate and facilitator in the evolution of stable resistance.

         At the outset of our project, we anticipated the possibility of observing chromosomal restructuring, such as the deletion of costly genes within amplified units. However, this mechanism was not detected in our study due to the high copy numbers present. Thus, a deletion occurring in a single copy within a highly amplified array would likely result in only a marginal fitness improvement, and the deletion of costly genes across all amplified units would necessitate a prolonged stepwise process, with each step associated with minimal fitness gains.

         The instability and transient nature of gene amplifications pose significant challenges in detecting HR bacteria. Given that the stability and fitness cost of gene amplifications dictate the presence of the resistance phenotype within the population, it becomes crucial to ascertain the recombination rates and fitness costs, as well as the potential impact of compensatory mutations on them. By employing a deterministic model, we showed that the loss rates are driven mainly by the fitness of the populations and are not affected by the chromosomal mutations.

         Our study holds particular significance for the clinical setting, as compensatory mechanisms that mitigate the pronounced fitness impacts could promote the emergence of stably resistant bacteria originating from HR isolates. This underscores the necessity for improved detection and surveillance techniques, alongside treatments aimed at more effectively eliminating HR bacteria.

References

1. Nicoloff, H., Hjort, K., Levin, B. R. & Andersson, D. I. The high prevalence of antibiotic heteroresistance in pathogenic bacteria is mainly caused by gene amplification. Nat. Microbiol. 4, 504–514 (2019).
2. Pereira, C., Larsson, J., Hjort, K., Elf, J. & Andersson, D. I. The highly dynamic nature of bacterial heteroresistance impairs its clinical detection. Commun. Biol. 4, 1–12 (2021).