Resistance DNA connected patients, different bacteria and hospital drains

What were we trying to find out?

When the same type of antibiotic resistance keeps appearing in a hospital, what is actually spreading? It might be one bacterial strain moving between people. It might instead be mobile DNA carrying resistance genes between otherwise different bacteria. Or several unrelated resistant bacteria might simply be appearing at the same time. This retrospective research study examined 13 carbapenem-resistant bacterial isolates collected from ten patients and three hospital-drain samples at TUM Hospital in Munich.

What did we find?

• Routine identification did not reveal the full bacterial diversity. Genome sequencing resolved several different bacterial species among isolates that routine diagnostics had grouped more broadly. These included Citrobacter freundii, Citrobacter portucalensis, Citrobacter youngae, Citrobacter farmerii and Enterobacter mori.
• One Citrobacter freundii lineage persisted over a surprisingly long period. A cluster of C. freundii ST91 was found in patient and drain samples. Two isolates collected in March 2025 were genetically identical, and they differed by only one single-nucleotide change from an ST91 isolate collected 12 months earlier in the same hospital room. The data were consistent with persistence from an incompletely sampled reservoir, although the study could not establish exactly where that reservoir was.
• The resistance plasmid crossed bacterial and environmental boundaries. Closely related IncN plasmids carrying KPC-2 were detected across four bacterial taxa and in both patient and environmental isolates over about eight months. This provided evidence consistent with movement of the resistance plasmid between bacterial species and between patient-associated and environmental reservoirs.

What does it mean?

Tracking only the bacterial strain can miss part of the story. The chromosome helps us understand which bacteria are closely related. Plasmids can answer a different question: whether mobile resistance DNA is being shared across bacteria that may belong to different species. In this study, both processes were occurring at the same time. A bacterial lineage appeared to persist within the hospital environment, while closely related resistance plasmids were found across different bacterial hosts and sampling sources. Long-read sequencing was particularly useful because it allowed the resistance genes to be placed onto reconstructed plasmids rather than simply showing that the genes were present. The findings also show why environmental samples can be informative during research investigations of persistent antimicrobial resistance. Hospital drains may contain bacteria and mobile DNA that are relevant to the wider epidemiological picture, although identifying a related organism or plasmid in a drain does not by itself establish the direction of transmission.

What does it not show?

This was a small retrospective study of 13 isolates from one hospital ward, so the observed patterns should not automatically be generalised to other hospitals or bacterial populations. Plasmids are also difficult to use as simple transmission clocks. They can gain, lose and rearrange DNA, so there is no direct equivalent of a fixed single-nucleotide distance that proves when one plasmid was transferred from one bacterium to another. The study therefore combined sequence and structural comparisons, but plasmid relationships still need to be interpreted alongside the epidemiological context. The study also could not prove the direction of transfer between patients, bacteria and drains. In particular, cultures from the drain in the room associated with the long-term ST91 cluster remained negative, so the precise reservoir responsible for that persistence was not identified.

Credit

This international collaborative study was led by Ela Sauerborn and Lara Urban and brought together researchers and clinical scientists from TUM Hospital and Technical University of Munich, Helmholtz Munich, PHF Science, the London School of Hygiene & Tropical Medicine, the University of Zurich and RIFCON. I contributed to the genomic analysis as one member of the wider author team, with PHF Science listed as my institutional affiliation. Full author contributions and affiliations are given in the published paper.