Researchers at the RIKEN Center for Biosystems Dynamics Research (BDR) in Japan reported on the development of a robotic culture system that allowed them to experimentally evolve the common bacteria Escherichia coli, under pressure from a large number of individual antibiotics. Analysis of the resulting data enabling the investigators to identify mechanisms and constraints underlying evolved drug resistance. They suggest that the findings, published in Nature Communications, could feasibly help in the development of therapeutic drug strategies that minimize the likelihood that bacteria will develop resistance. In the team’s published paper, titled, “High-throughput laboratory evolution reveals evolutionary constraints in Escherichia coli,” research lead Tomoya Maeda, PhD, and colleagues, concluded, “… we demonstrate how our experimental system could provide a quantitative understanding of evolutionary constraints in adaptive evolution, leading to the basis for predicting and controlling antibiotic resistance … These findings bridge the genotypic, gene expression, and drug resistance gap, while contributing to a better understanding of evolutionary constraints for antibiotic resistance.”
Counteracting multidrug-resistant (MDR) bacteria is becoming a critical global challenge,” the authors wrote. It seems that every time new antibiotics are developed, novel, antibiotic-resistant bacteria emerge during clinical use, so “… alternative strategies for suppressing the emergence of resistant bacteria are actively being sought.” But to be able to do this, scientists will need a better understanding of how drug resistance evolves in bacteria. In fact this process is complicated, and involves numerous changes in genome sequences and cellular states, with studies investigating resistance evolution indicating that mechanisms for resistance are tightly interconnected. “Various mechanisms for antibiotic resistance have been identified, including the activation of efflux pumps, modifications of specific drug targets, and shifts in metabolic activities,” the team continued.
Understanding the constraints that impact on the evolution of antibiotic resistance will also be crucial for scientists to be able to predict and control drug resistance. However, a comprehensive study of resistance dynamics for large numbers of antibiotics has never been reported, the researchers noted. “Thus, elucidating evolutionary constraints are crucial for predicting and controlling the evolution of antibiotic resistance; however, despite its importance, a systematic investigation of evolutionary constraints for antibiotic resistance evolution is still lacking.”
As Maeda explained, “Laboratory evolution combined with genomic analyses is a promising approach for understanding antibiotic resistance dynamics. However, laboratory evolution is highly labor intensive, requiring serial transfer of cultures over a long period and a large number of parallel experiments.” Additionally, Maeda noted, identifying the genes that will allow resistance to antibiotics is not always easy because of the large number of genetic features that are contained within the data.
To overcome existing limitations, the team developed an automated robotic culture system that allowed them to perform high-throughput laboratory evolution of E. coli—under pressure from 95 different antibiotics – for more than 250 generations. With this new ability, they were able to quantify changes in the bacteria’s transcriptome—the set of all messenger RNAs and their transcripts, which is effectively a record of which genes are actually expressed.
“To analyze the expanded cross resistance/collateral sensitivity network, including both antibiotic and non-antibiotic stressors, while elucidating the molecular mechanisms associated with resistance acquisition, we choose a variety of antibacterial chemicals, including antibiotics with various mechanisms of action, and non-antibiotic toxic chemicals, against E. coli,” the scientists commented. “We quantify changes in the transcriptome, genomic sequence, and resistance profile in the evolved strains, producing a multiscale dataset for analyzing stress resistance.”
Their method produced resistance profiles for 192 of the evolved strains. The researchers also developed a machine-learning method or analyzing the large amount of data generated, allowing them to identify both novel and already recognized genes that contribute to the prediction of resistance evolution. “The analysis of reconstructed mutant strains also provided valuable information on the genetic basis of resistance acquisition,” they wrote. “These results demonstrated that the pattern of resistance acquisition observed in the evolved strains could be characterized by known resistance-conferring genes … Furthermore, the analysis identified genetic mechanisms for the stress resistance that have not been reported yet …”
“We found that E. coli‘s evolutionary dynamics is attributable to a relatively small number of intracellular states, indicating that it is likely equipped with only a limited number of strategies for antibiotic resistance,” said Maeda. By being able to quantify the constraints that affect evolution of antibiotic resistance in E. coli, the team hopes it will be possible to predict, and feasibly control, antibiotic resistance. “We believe that sharing our results in this manuscript, including identified mutations, transcriptome changes, and resistance profiles in the evolved strains, as well as phenotypic changes in the reconstructed mutants, will provide clues for future studies and contribute to the field of antibiotic resistance evolution,” they claimed.
For example, by using this new system, the researchers were able to test 2162 pairs of drug combinations and discovered 157 pairs that have the potential to suppress antibiotic resistance acquisition in E. coli. As Maeda noted, “We believe that our results can be applied to the development of alternative strategies for suppressing the emergence of drug-resistant bacteria.”