Back

Pre-existing antibiotic tolerance facilitates plasmid-mediated carbapenem resistance evolution in clinical Klebsiella pneumoniae

zhang, W.; Zheng, B.; Zhou, M.; Zhang, R.; Xu, Y.; Liu, J.

2026-08-24 microbiology
10.64898/2026.08.24.746623 bioRxiv
Show abstract

Antibiotic tolerance enables bacteria to survive bactericidal antibiotic exposure and has been linked to resistance evolution in laboratory systems and individual infections, but its role in plasmid-mediated resistance evolution in clinical populations remains unclear. Here, we analyzed a longitudinal collection of more than 800 clinical Klebsiella pneumoniae isolates spanning 1997-2020. Among 779 minimum inhibitory concentration (MIC)-defined ertapenem-susceptible isolates, 137 (17.6%) displayed hidden ertapenem tolerance, defined by enhanced survival after 6 h at 30 times the isolate-specific ertapenem MIC, mostly without extended lag time or reduced growth rate. Tolerance was detected before local ertapenem introduction and was enriched among ertapenem-resistant isolates, supporting a population-level association between pre-existing tolerance and the emergence of carbapenem resistance. Genomic and plasmid-curing analyses separated plasmid-mediated carbapenem resistance from plasmid-independent antibiotic tolerance. Moreover, tolerant recipient backgrounds enhanced resistance plasmid acquisition, preserved viable recipients following antibiotic exposure and accelerated ceftazidime-avibactam resistance evolution. A phylogeny-guided variant-enrichment analysis further identified the uhpABC regulatory operon as a candidate tolerance-associated locus, and coordinated expression of the complete operon increased ertapenem survival. Together, these findings identify clinical antibiotic tolerance as a pre-existing, MIC-hidden phenotype that can facilitate plasmid-mediated carbapenem resistance evolution in K. pneumoniae.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.