Ordering the mob: Insights into replicon and MOB typing schemes from analysis of a curated dataset of publicly available plasmids

Abstract

Plasmid typing can provide insights into the epidemiology and transmission of plasmid-mediated antibiotic resistance. The principal plasmid typing schemes are replicon typing and MOB typing, which utilize variation in replication loci and relaxase proteins respectively. Previous studies investigating the proportion of plasmids assigned a type by these schemes ('typeability') have yielded conflicting results; moreover, thousands of plasmid sequences have been added to NCBI in recent years, without consistent annotation to indicate which sequences represent complete plasmids. Here, a curated dataset of complete Enterobacteriaceae plasmids from NCBI was compiled, and used to assess the typeability and concordance of in silico replicon and MOB typing schemes. Concordance was assessed at hierarchical replicon type resolutions, from replicon family-level to plasmid multilocus sequence type (pMLST)-level, where available. We found that 85% and 65% of the curated plasmids could be replicon and MOB typed, respectively. Overall, plasmid size and the number of resistance genes were significant independent predictors of replicon and MOB typing success. We found some degree of non-concordance between replicon families and MOB types, which was only partly resolved when partitioning plasmids into finer-resolution groups (replicon and pMLST types). In some cases, non-concordance was attributed to ambiguous boundaries between MOBP and MOBQ types; in other cases, backbone mosaicism was considered a more plausible explanation. β-lactamase resistance genes tended not to show fidelity to a particular plasmid type, though some previously reported associations were supported. Overall, replicon and MOB typing schemes are likely to continue playing an important role in plasmid analysis, but their performance is constrained by the diverse and dynamic nature of plasmid genomes.

Keywords: Antibiotic resistance; MOB typing; Plasmid database; Plasmid multilocus sequence typing; Replicon typing.

Fig. 1

Complete plasmid accessions added to NCBI since 2014. Where a hybrid short-read/long-read sequencing approach was used (as indicated by the accession metadata), this is represented as long-read only.

Fig. 2

(A) Cumulative number of plasmids added to NCBI from initial accession (7th November 1985) to sequence retrieval date (26th August 2016). Cumulative counts reflect all plasmids (green), as well as the subsets that were replicon typed (red) and MOB typed (blue). (B) Proportion of plasmids added to NCBI prior to a given date that could be typed by replicon typing (red) and MOB typing (blue). Grey shading around the lines represents a 95% binomial confidence interval, calculated by the Agresti-Coull method.

Fig. 3

Relationship between typing success and different plasmid characteristics. For plasmid size, bins are equally sized (kb) 1.3–6.1; 6.1–35; 35–68; 68–95; 95–137; 137–794. For number of resistance genes (all resistance classes included), bin ranges and sizes are: 0, n = 1089; 1, n = 311; 2, n = 134; 3–4, n = 192; 5–8, n = 183; 9–34, n = 194.

Fig. 4

Chord diagrams illustrate associations between replicon families and MOB types; circularly arranged sectors represent replicon families and MOB types, and scale bars indicate their relative sizes. Associations are indicated by intersecting chords. Replicon family sectors, coloured; MOB type sectors, grey. (A) Replicon family-MOB type associations amongst plasmids un-typed by one or both schemes. (B) Replicon family-MOB type associations amongst plasmids with both replicon and MOB type detected. Data on Col plasmids are deliberately not shown (see main text). Plasmid types are represented as the unique set of families detected on a plasmid, as described in Methods (i.e. IncF, IncF = IncF). Where multiple types of different replicon families are detected, data are not shown, but are presented in Fig. S6. Replicon family types detected in fewer than 10 plasmids are not shown except where a non-concordant pattern of association is observed (IncU) or where associated with a multi-type MOB type.

Fig. 5

Chord plots A and B illustrate associations between replicon and MOB types, for replicon types belonging to non-concordant replicon families. Where replicon types are shown to be non-concordant and a pMLST scheme is available, chord plots C and D show associations between pMLST types and MOB types; IncHI1 pMLST associations are not shown due to the small number of plasmids belonging to this subtype. Plasmid types are represented as the unique set of types detected on a plasmid, as described in Methods (i.e. IncFIA, IncFII, IncFII = IncFIA, IncFII). Replicon types and pMLST types detected in fewer than 5 plasmids are not shown except where a non-concordant pattern of association is observed, or where associated with a multi-type MOB type. Associations with untyped plasmids are not shown. Col family replicon types correspond to the replicon probe names in the PlasmidFinder database. IncF pMLST is based on the so-called FAB formula where sequence types are determined according to the allele for IncFII, IncFIA and IncFIB. As an example, the most common IncF pMLST type is F2:A-:B- which reflects detection of allele 2 for IncFII, and detection of no alleles for IncFIA and IncFIB.

Fig. 6

Associations between β-lactamase resistance genes identified and (A) MOB types; (B) replicon families. MOB types/replicon families detected in fewer than 10 plasmids are not shown.