First, we have selected a group of pathogens belonging to the Enterobacteriaceae family isolated from human patients from the Hospital Universitario Ramón y Cajal where the plasmid pOXA48 has been found.
Then, we have measured the effects of the plasmid in those bacteria. We have found, that the presence of the plasmid affects both to the resistance to antibiotics and to the physiology of the cell. The bacteria bearing pOXA48 become resistant to b-lactam antibiotics but in the absence of the antibiotic the plasmid impose a cost, reducing the growth of the bacteria bearing them. Additionally, we estimated the plasmid copy number of the pOXA48 in the cells, showing that the plasmid copy number is higher in Klebsiella pneumoniae than in Escherichia coli
Once we have studied the effects of the plasmid in the cells, we have propagated them in controlled environments in the laboratory for 150 generations, allowing the bacteria and the plasmid to coevolve together. In this experiment, called experimental evolution, we have propagated 5 clones of Escherichia coli and 7 clones of Klebsiella pneumoniae both in the presence and in the absence of antibiotics. Then, we sequenced the genome of more than 150 evolved populations to understand the evolutionary dynamics of plasmid/host adaptation. Once the experiment was done, we have focused our attention in 4 main different aspects that are crucial for the biology of the plasmid: i) has the cost of the pOXA48 being compensated after the coevolution with the different clones? ii), what are the mutations driving plasmid/host adaptation?, iii) has the plasmid changed/evolved? and iv) does the plasmid/host adaptation is strain or species-specific or is that adaptation driven by common mechanisms?
i) Our results showed that the initial cost of the plasmid was compensated after 150 generations of coevolution with the host independently of the strains, and of the presence or absence of antibiotic in the environment.
ii) The cost of plasmids is usually associated with the plasmid copy number that the plasmid has in the cell. Therefore, we found that when the cells bearing the plasmid were propagated in the absence of antibiotics, the plasmid copy number was reduced probably reducing the cost of pOXA48. However, in the presence of antibiotics, the plasmid copy number was the same that at the beginning of the experiment.
iii) We found several mutations in the plasmid at the end of the experiment. Some of them were encoded in the machinery that the plasmid uses to spread among different cells. Some other were found affecting the proteins involved in the replication of the plasmid. The most relevant mutations that we found are some that inactivate the carbapenemase, and therefore, making the cell sensitive to antibiotics again.
iv) By sequencing the whole-genome of the evolving populations, we found that the mechanisms driving plasmid/host adaptation are both species-specific and strain-specific. In Escherichia coli, we found that the vast majority of adaptive mutations wer single nucleotide mutations in the chromosome of the bacteria. Further, those mutations varied among the different clones. However, in Klebsiella pneumoniae, the presence of the plasmid induced the expression of some genetic elements known as insertion sequences. The insertion sequences are mobile genetic elements capable of "jump" to different parts of the genome. We have found that the solely presence of pOXA48 induced the movement of those insertion sequences, and that the adaptive mutations are caused by the movements of those insertion sequences and not by single nucleotide polymorphisms. As in Escherichia coli, each different Klebsiella pneumoniae clone adapted to the plasmid in a different way.
We have presented the results of REPLAY in the meeting organised by the International Society for Plasmid Biology and Other Mobile Genetic Elements (Tolouse, France 2022). We are drafting the manuscript and it will be submitted during 2023 to a high impact open access journal. Once the paper is submitted, we will announce it in our twitter pages and in our websites. All the data generated by REPLAY and the code used for the analysis of the results has been deposited in
https://github.com/jorgEVOplasmids/expev_pOXA-48(s’ouvre dans une nouvelle fenêtre)
