Genetic determinants of pOXA-48 plasmid maintenance and propagation in Escherichia coli

Abstract

Conjugative plasmids are the main drivers of antibiotic resistance dissemination contributing to the emergence and extensive spread of multidrug resistance clinical bacterial pathogens. pOXA-48 plasmids, belonging to the IncL group, emerge as the primary vehicle for carbapenem resistance in Enterobacteriaceae. Despite the problematic prevalence of pOXA-48, most research focus on epidemiology and genomics, leaving gaps in our understanding of the mechanisms behind its propagation. In this study, we use a transposon sequencing approach to identify genetic elements critical for plasmid stability, replication, and conjugative transfer. Our results identify a novel type I toxin-antitoxin system, uncharacterized essential maintenance factors, and components of the type IV secretion system and regulatory elements crucial for conjugation. This study advances our understanding of pOXA-48 biology, providing key insights into the genetic factors underlying its successful maintenance and spread in bacterial populations.

Fig. 1. Circular genetic map of the pOXA-48 plasmid.

The outer and inner arrows represent the coding DNA sequences (CDS) on the forward and reverse strands, respectively. Each CDS is color-coded according to its predicted function, as indicated in the figure legend. The categories include transposon elements (Tn1999), insertion sequences (IS), genes involved in plasmid maintenance/establishment, replication, transfer, and others, as well as genes with unknown function.

Fig. 2. Identification of genes required for replication and maintenance of pOXA-48 plasmid.

a Tn-Seq analysis of transposon insertion abundance across the pOXA-48 plasmid represented as Log1p(reads/gene). The mean of reads (filled line) with SD (dotted lines) are represented. Genes are color-coded based on their predicted function. b Read coverage of transposon insertions around the repCBA locus. The schematic below shows the genomic context of these replication-related genes, highlighting regulatory elements, including the antisense RNA molecule RNAI and the origin of replication (oriV). c Tn-Seq read distribution for the korC gene and its neighboring region. The putative binding sites of KorC are shown below, along with the predicted promoter regions and the orf24’s codon start. d Read counts for the transposon insertions across the resD, parA, parB, nuc, orf19, orf20, and orf21 genes. e Stability assay of pOXA-48 plasmid derivatives with deletions of parA, parB, and nuc over 40 generations. The percentage of cells retaining the plasmid was measured at different time points, and the mean and SD of three independent clones are represented. f Effect of the orf19 deletion on cell growth was monitored by measuring the optical density at 600 nm (OD₆₀₀). The mean and SD of three independent clones are shown. g Stability assay of pOXA-48 plasmid deleted of orf19 at time 0 and after 40 generations. The percentage of cells retaining the plasmid is estimated from three independent clones (white dots).

Fig. 3. Identification of a type I toxin-antitoxin encoded by the pOXA-48 plasmid.

a Tn-Seq data showing read counts for transposon insertions across the agrB/dqlB region of the pOXA-48 plasmid. The schematic below illustrates the genomic organization of the predicted toxin-antitoxin system, with agrB in green encoding the antitoxin (A) and dqlB in gray the toxin (T). b Spot assays evaluating the toxicity of DqlB and the protective effect of agrB. Growth of E. coli (−/−; LY4078), E. coli producing agrB (LY4080), DqlB (LY4079) or agrB and DqlB (LY4081) is shown under conditions with either glucose or arabinose. c Schematic representation of the Targeted Antibacterial Plasmid (TAP) system used to investigate the role of agrB and DqlB. Donor cells carry a TAP-OXA degrading the pOXA-48 plasmid, leading to cell death or a TAP-OXA-A targeting the pOXA-48 plasmid but producing the antitoxin gene to counteract the toxin activity and rescue the transconjugant cell. The histograms show the number of viable transconjugants (CFU/mL) after acquisition of the different TAPs. Donors TAP-nsp (LY1369), TAP-OXA48 (LY1522), TAP-OXA48-PemI (LY1549), TAP-OXA48-agrB (LY4038); Recipients (MS388). Mean and SD are calculated from six independent experiments. P-value significance from Tukey’s multiple comparisons statistical analysis is indicated by ns: not significant, *** (p = 0.0002), **** (p < 0.0001). d Histograms showing the proportion of Inc plasmids containing a dqlB-like homolog.

Fig. 4. Identification of genes essential for conjugation and enriched in transposon insertion sites.

Log₂(FC) analysis of transposon insertion abundance comparing the output library to the input library. Threshold of 2 and −2 are indicated by the dotted lines. Genes are color-coded based on their predicted function.

Fig. 5. Impact of orf38 and orf36.1 deletions on conjugation frequency.

Histograms showing the frequency of transfer after 4 h of mating for the wild-type (wt), orf38 and orf36.1 deletion mutants, as well as their complementation with pOrf38 and pOrf36.1 plasmids producing respectively Orf38 and Orf36.1. Data are represented as mean and SD from (n) independent experiments (white circles). p-value significance from Kruskal-Wallis test corrected with Dunn’s multiple comparisons test is indicated by ns (not significant), * (p = 0.023, pOrf38; p = 0.0461, orf36.1), ** (p = 0.0013). # indicates two replicates with a detection limit of transconjugants below 10⁻⁸. Donors LY1844, LY3303, LY3697, LY3739, LY3747; Recipient LY945.

Fig. 6. Analysis of phenotypes associated with deletion of orf32, ssb, orf33, excA, or repC genes.

a Conjugative transfer frequency after 150 min of mating for the wild-type (wt) (LY1844) and mutants, orf32 (LY4148) and complemented strain with pOrf32 plasmid (LY4192), ssb (LY4149), orf33 (LY4153), excA (LY3260), and repC (LY3259) and complemented strain with pRepC plasmid (LY3606). Recipient LY945. Data are represented as mean and SD from at least six independent experiments (gray circles). Comparisons test is indicated by ns (not significant), *(p = 0.0133), **(p = 0.0095). b Plasmid copy number (PCN) for each mutant and repC complemented strain (repC pRepC) from stationary phase culture. The ratio of the mutant PCN to the wild-type PCN is represented. Data are represented as mean and SD from (n) independent experiments (gray circles). Comparisons test is indicated by ns (not significant), *(p = 0.044) c Minimum inhibitory concentration (MIC) of imipenem inhibiting the growth of the wild-type (wt), repC mutant, and repC complementated strain (repC pRepC). Data are represented as mean and SD from three independent experiments. Comparisons test is indicated by ns (not significant), *(p = 0.0463). d Histogram showing the exclusion levels of wild-type pOXA-48 transfer from donor to the recipients cells carrying wild-type pOXA-48 (wt), excA mutant with or without pExcA complementation plasmid. Data are represented as the mean and SD from (n) independent experiments (white circles). Comparisons test is indicated by ns (not significant), **(p = 0.0081). e PCN for excA mutant and complemented strain (excA pExcA) derived from colonies scratched from LB agar plates (solid). The ratio of the mutant PCN to the wild-type PCN are represented. Data are represented as mean and SD from seven independent experiments (gray circles). Comparison test is indicated by ns (not significant), **(p = 0.0011). f Conjugative transfer frequency after 60 min of mating for the wild-type (wt) and excA mutant, and complemented strain with pExcA derived from colonies scratched from LB agar plates (solid). Data are represented as mean and SD from (n) independent experiments (gray circles). Comparisons test is indicated by ns (not significant), ****(p < 0.0001). a–e p-value significance from the Kruskal-Wallis test corrected with Dunn’s multiple comparisons test.