Abstract
INTRODUCTION
Treatment failure linked to antimicrobial resistance (AMR) is one of the most important threats to human health. The genes that encode AMR have become highly mobile and, since the introduction and widespread use of antibiotics, have disseminated globally across a diverse set of bacteria. This has made many infections hard or impossible to treat. Extrachromosomal elements called plasmids are one of the principal mechanisms by which AMR genes have spread. Although we know that the AMR gene cassettes that plasmids carry have become more complex over time, often conferring resistance to multiple antibiotic classes, the evolution of the plasmids themselves during this antibiotic era remains poorly understood.
RATIONALE
To understand how plasmids have adapted to carry AMR genes, we sequenced plasmids from historical clinical samples taken between 1917 and 1954 (the Murray Collection) and combined this with all available sequence data from the public archive to trace their descendants among contemporary and currently circulating plasmids. Given this time span, we were able to explore the evolutionary trajectories and genetic changes that define plasmid adaptation during the antibiotic era.
RESULTS
By comparing our historical plasmids with their modern descendants by means of a genome similarity network, we defined close families of related plasmids. We found that they were divided into three categories reflecting distinct evolutionary trajectories: (i) ancestral plasmids that disappeared as intact backbones from modern public archives; (ii) those that survived to the modern day without substantial change, preserving a core genome; (iii) and those primarily formed through fusion of one or more ancestral plasmid(s), resulting in conspicuously large and complex modern plasmids. It is also clear that a relatively small group of modern plasmids carry almost all the AMR genes seen today, including those conferring resistance to first-line and last-resort antibiotics. These plasmids originated from ancestral backbones devoid of AMR genes and display characteristics linked to the ability to acquire and transmit these genes.
CONCLUSION
Our longitudinal view of plasmid evolution over the past 100 years revealed highly dynamic processes throughout their evolution to the selective pressure of antibiotic use, including gene turnover and genome fragmentation and fusion. A minority of plasmids already existing in pathogens became global multidrug-resistance vectors as humans industrialized the use of antibiotics, and their descendants now pose a high risk to human health. Their emergence involved both conservative gene acquisition into stable backbones and rapid evolution underpinned by genome recycling, fusion, and reassortment, generating complex diversity.
Treatment failure linked to antimicrobial resistance (AMR) is one of the most important threats to human health. The genes that encode AMR have become highly mobile and, since the introduction and widespread use of antibiotics, have disseminated globally across a diverse set of bacteria. This has made many infections hard or impossible to treat. Extrachromosomal elements called plasmids are one of the principal mechanisms by which AMR genes have spread. Although we know that the AMR gene cassettes that plasmids carry have become more complex over time, often conferring resistance to multiple antibiotic classes, the evolution of the plasmids themselves during this antibiotic era remains poorly understood.
RATIONALE
To understand how plasmids have adapted to carry AMR genes, we sequenced plasmids from historical clinical samples taken between 1917 and 1954 (the Murray Collection) and combined this with all available sequence data from the public archive to trace their descendants among contemporary and currently circulating plasmids. Given this time span, we were able to explore the evolutionary trajectories and genetic changes that define plasmid adaptation during the antibiotic era.
RESULTS
By comparing our historical plasmids with their modern descendants by means of a genome similarity network, we defined close families of related plasmids. We found that they were divided into three categories reflecting distinct evolutionary trajectories: (i) ancestral plasmids that disappeared as intact backbones from modern public archives; (ii) those that survived to the modern day without substantial change, preserving a core genome; (iii) and those primarily formed through fusion of one or more ancestral plasmid(s), resulting in conspicuously large and complex modern plasmids. It is also clear that a relatively small group of modern plasmids carry almost all the AMR genes seen today, including those conferring resistance to first-line and last-resort antibiotics. These plasmids originated from ancestral backbones devoid of AMR genes and display characteristics linked to the ability to acquire and transmit these genes.
CONCLUSION
Our longitudinal view of plasmid evolution over the past 100 years revealed highly dynamic processes throughout their evolution to the selective pressure of antibiotic use, including gene turnover and genome fragmentation and fusion. A minority of plasmids already existing in pathogens became global multidrug-resistance vectors as humans industrialized the use of antibiotics, and their descendants now pose a high risk to human health. Their emergence involved both conservative gene acquisition into stable backbones and rapid evolution underpinned by genome recycling, fusion, and reassortment, generating complex diversity.
| Original language | English |
|---|---|
| Article number | eadr1522 |
| Journal | Science (New York, N.Y.) |
| Volume | 390 |
| Issue number | 6777 |
| Early online date | 25 Sept 2025 |
| DOIs | |
| Publication status | Published - 4 Dec 2025 |
Data Availability Statement
All data are available in the main text or the supplementary materials. The whole-genome sequencing data of the Murray Collection are publicly accessible from the BioProject PRJEB3255. The accessions of the isolates in which plasmids were identified in this study are listed in data S1. Our integrated plasmid dataset is composed of publicly available sequences. The database from which the plasmids were retrieved, their GenBank accessions, and BioSample accessions when available are listed in data S1 and S11. The analysis code used in this study and pairwise comparison data were made available in GitHub and archived in a Zenodo repository (78). The Murray plasmid sequences and all network files for the networks presented in this study were also archived in Zenodo (79). The strains in the Murray Collection are available at the National Collection of Type Cultures (NCTC).Acknowledgements
We thank M. A. Beale for the helpful discussions during this study. We also thank the Parasites and Microbes Programme Informatics and Samples teams at the Sanger Institute for technical support.Funding
This research was funded in part by the Wellcome Trust (grant 220540/Z/20/A). A.C. was supported by an ESPOD EMBL-EBI/Wellcome Trust Sanger Institute Joint Post-Doctoral Fellowship (45521) and Sanger Institute core funding from the Wellcome Trust (grant 220540/Z/20/A).
| Funders | Funder number |
|---|---|
| The Wellcome Trust | 220540/Z/20/A |
| The Wellcome Trust | 220540/Z/20/A |
ASJC Scopus subject areas
- General
