Precise classification of antimicrobial resistance-associated IncP-2 megaplasmids for molecular epidemiological studies on Pseudomonas species

The incompatibility group P-2 (IncP-2) of plasmids defined by Bryan et al.¹ includes many antimicrobial resistance (AMR)-associated large plasmids (megaplasmids, ≳400 kb) found in Pseudomonas species,² including Pseudomonas aeruginosa. In a recently published article, Jiang et al.³ proposed a novel incompatibility group of plasmids, Inc_pRBL16, and listed 17 Inc_pRBL16 plasmids, including pRBL16 (accession no. CP015879), in Figure 1 of that article. The repA gene that encodes a replication initiation protein (RIP) for the Inc_pRBL16 plasmid, repA_pRBL16, was located between 303 463 and 304 650 nt in pRBL16. In Figure 2 of that article, Jiang et al.³ classified another set of 12 plasmids, which were previously proposed as IncP-2 plasmids based on the nucleotide sequences of the repA gene. The authors analysed five of them and reported that pSx1 (accession no. CP013115) and pCP017294 (PA83 plasmid unnamed1, accession no. CP017294) contain a single RIP gene, repA_IncP-2, whereas pOZ176 (accession no. KC543497), pTTS12 (accession no. CP009975), and pJB37 (accession no. KY494864) contain another RIP gene, repA_pRBL16 (repA_IncP-2 and repA_pRBL16 in pOZ176 were pOZ176_301 and pOZ176_183 genes, respectively), as the primary RIP gene, in addition to the auxiliary RIP gene repA_IncP-2.^3–5

We agree with the authors that they found two types of RIP genes in pRBL16 and pOZ176 (according to their nomenclature, repA_pRBL16 in pRBL16, and repA_pRBL16 and repA_IncP-2 in pOZ176³). However, we would like to highlight that they identified the true RIP gene of IncP-2 plasmids as repA_pRBL16, designated here as repP-2A, and not necessarily a novel replicon named repA_pRBL16. In addition, we propose that their repA_IncP-2 is not the primary RIP gene of IncP-2 plasmids. They also described that the above three plasmids (pOZ176, pTTS12, and pJB37) were misidentified as IncP-2 plasmids.^3–5 However, one of them, pOZ176, could not be stably maintained in the same bacterial cell with another IncP-2 plasmids in plasmid incompatibility tests, strongly indicating that pOZ176 is a member of the IncP-2 plasmids.⁶ Subsequently, Xiong et al.⁷ determined the complete nucleotide sequence of pOZ176 containing two RIP genes (pOZ176_183 and pOZ176_301 in accession no. KC543497), and proposed one of two repA genes (pOZ176_301, i.e. the auxiliary RIP gene in the plasmid) as a candidate RIP gene of IncP-2. Plasmids with this misidentified repA (pOZ176_301 gene), not repP-2A (pOZ176_183 gene), have been misrecognized as IncP-2 plasmids in some later studies, including ours.⁸

In this study, we determined the complete nucleotide sequence of the Pseudomonas aeruginosa plasmid Rms139 (accession no. LC653116),⁹ which has been classically identified as a member of IncP-2 through plasmid incompatibility tests.² This plasmid contains a sole RIP gene (repP-2A, located between 1 and 1188 nt in Rms139) whose nucleotide sequence showed 100% identity with that of repA_pRBL16 in pRBL16. Cazares et al.¹⁰ proposed the pBT2436-like family as a group of megaplasmids, including pBT2436 (accession no. CP039989), in Pseudomonas species. Of note, each of them contained a conserved RIP gene (FC629_32540 gene in pBT2436),¹⁰ showing high identity with repA_pRBL16 (92%–100% identity at the amino acid sequence level). This shows that repA_pRBL16 is the primary and true RIP gene (repP-2A) of IncP-2 plasmids. Therefore, the nucleotide-sequence-based classification of IncP-2 plasmids should be updated based on the sequence of repA_pRBL16.³

More recently, there have been several reports on AMR-associated IncP-2 megaplasmids in Pseudomonas species clinical isolates. Urbanowicz et al.¹¹ showed endemic spread of pBT2436-like megaplasmids carrying the carbapenemase gene bla_VIM-2 using 19 plasmids, including pPUV-1 (accession no. MT732179), in P. aeruginosa isolated in Poland, which formed a subgroup within a family of IncP-2 megaplasmids. Zhang et al.¹² showed that 16 IncP-2 megaplasmids [9 plasmids in P. aeruginosa isolated in China, including pHS17-127 (accession no. CP061377), and 7 plasmids in Pseudomonas species in the NCBI database] carry the carbapenemase gene bla_IMP-45 and this IncP-2 plasmid subgroup contributed to the worldwide spread of bla_IMP-45.

AMR genes are often carried on plasmids and spread among bacteria via conjugation. Precise classification of AMR-associated plasmids by phenotyping methods based on plasmid incompatibility and genotyping methods based on RIP sequences are crucial for molecular epidemiological studies on clinically relevant bacterial pathogens, including Pseudomonas species. Indeed, we confirmed that IncP-2 megaplasmids in recent clinical isolates of Pseudomonas species^3,10–12 actually contain the repP-2A gene and have accumulated a number of important AMR genes, such as carbapenemase genes (bla_IMP, bla_VIM, and bla_DIM), 16S ribosomal RNA methyltransferase genes conferring aminoglycoside resistance (armA), efflux pump genes conferring fluoroquinolone resistance (qnr), and efflux pump gene clusters conferring tigecycline resistance (tmexCD-toprJ) (Figure 1).

Figure 1.

The phylogeny tree constructed by the pipeline of Bactopia v1.7.1 (https://github.com/bactopia/bactopia) using nucleotide sequences of the indicated IncP-2 plasmids. Bar lengths represent the number of substitutions per nucleotide site. Plasmid names, bacterial species, sizes, replicon types and representative AMR genes (ARGs), including β-lactams (BL), aminoglycosides (AG), fluoroquinolones (FQ), and tigecycline (TGC) resistance genes, detected by Staramr v0.7.2 (https://github.com/phac-nml/staramr) with the custom nucleotide sequence database of plasmid replicons and ARGs, countries and years in which bacteria were isolated, accession numbers and references are shown. N.D., not detected.

Recent innovations in long-read sequencing technology and subsequent expansion of the plasmid nucleotide sequence database have enabled us to predict a novel classification of plasmid RIP genes without experimentally examining incompatibility. Therefore, it is important to keep updating the information on plasmid classification from the past for the future.

Funding

This work was supported by grants (JP21wm0325022 and JP21wm0225008 to M. Shintani; JP21fk0108093, JP21fk0108139, JP21fk0108133, JP21wm0325003, JP21wm0325022, JP21wm0225004, JP21wm0225008, JP21wm0325037, and JP21gm1610003 to M. Suzuki) from the Japan Agency for Medical Research and Development (AMED), grants (16H06279, 19H02869, and 19H05686 to M. Shintani; 20K10436 and JPMJCR20H1 to H. Suzuki; JP19H05679 and JP19H05686 to H. Nojiri; 20K07509 to M. Suzuki) from the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan, and grants (M. Shintani, H. Suzuki, and H. Nojiri) from Consortium for the Exploration of Microbial Functions of Ohsumi Frontier Science Foundation, Japan.

Transparency declarations

None to declare.