Back

Conjugation structures plasmid populations through host-lineage restriction

Matlock, W.; MacLean, R. C.

2026-02-19 microbiology
10.64898/2026.02.19.706745 bioRxiv
Show abstract

Conjugation mediates plasmid transfer between bacterial species, driving the horizontal spread of traits like antibiotic resistance. However, genomic and experimental evidence indicates that many conjugative plasmids are restricted to particular host lineages despite broad theoretical host ranges. Using 4,281 plasmids from an epidemiologically coherent sample of 1,739 Escherichia coli bloodstream infection isolates, we quantify the host-lineage associations of 30 plasmid backbones and assess how these associations structure plasmid co-occurrence. To achieve this, we develop a novel Bayesian modelling framework that separates genuine backbone-backbone associations from patterns arising due to shared host ancestry or abundance. First, we find that conjugative backbones exhibit stronger host-lineage restriction than mobilisable backbones, and that restriction increases with plasmid size independently of mobility class. Comparison with global plasmid diversity shows that the most host-lineage restricted backbones remain restricted beyond the studied population, whereas the least restricted backbones span a mean of seven host species. Next, after accounting for host phylogeny and abundance, we find that two thirds of backbone pairs show no strong association or avoidance; however, backbones sharing host lineages co-occur more frequently than expected. Lastly, we identify a clique of strongly associated mobilisable backbones that appear to exploit a shared set of lineage-restricted conjugative partners. A mathematical model demonstrates that host-lineage restriction of conjugative backbones, together with specificity in conjugative-mobilisable transfer, is sufficient to generate clustering among mobilisable plasmids. Collectively, our findings reframe conjugation as a mechanism that promotes within-lineage persistence and shapes plasmid community structure, with potentially important consequences for the accumulation of resistance and virulence determinants.

Matching journals

The top 7 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.