Plasmid evolution using long read genome sequencing of Escherichia coli bacteraemia isolates

Background: Plasmids are prevalent in E. coli and play a pivotal role in evolution of its antibiotic resistance (ABR), which is one of the major threats to human health in developed countries. Multiple plasmids, from a range of incompatibility groups, contain various combinations of antibiotic resistance genes. The study of plasmid evolution demands long read sequencing as extraction of the whole circular plasmids from the short read sequences is not possible due to the presence of repetitive regions.

Material/Methods: 162 Blood culture bacteria isolates were collected from Scotland for years 2013 and 2015 and sequenced using Illumina Hi-Seq sequencing platform. From 162, eight isolates were sequenced using the PacBio platform, selected on the basis of higher number of ABR genes and higher plasmid replicon types. Unicycler assembly was used to align short and long read sequences together for complete bacterial genomes and plasmids. These complete plasmid sequences were then compared with 155 publically available E. coli plasmids from EMBL database ranging from year 2000 to 2015.

Results: Within our 8 isolates [ST131 (n=5) and ST69 (n=3)], 21 plasmids (12 Col and 9 IncF) were identified, of which 7 Col plasmids were novel and were not detected by the pMLST typing method. Most of the ABR and virulence genes of the isolates were present in these plasmids. Two out of 3 ESBL carrying IncF plasmids have lost the CTXM-15 genes which have integrated into the bacterial genome. IncF plasmids in ST69 strains have acquired a portion of the Salmonella plasmid pUO-StVR2 that has multiple ABR and virulence genes (Figure 1). Highly toxic Col plasmids are prevalent in ST69 community strains compared to ST131 hospital strains. Phylogenetic analysis shows the evolutionary relationship within these plasmids, revealing specific disease transmission patterns.

Conclusions: PacBio sequencing of E. coli isolates identified novel plasmids, which were not identifiable using plasmid typing methods. Whole plasmid genome sequencing shows evidence of strain-to-strain transmission within a specific geographical location.