Article Figures & Data
Figures
- FIG 1
Integrative elements of the R. equi macrolide resistance plasmid pRErm46. (A) Genetic structure of the 6.9-kb transposon TnRErm46 carrying the macrolide resistance gene erm(46). The ISRe46 transposase is a novel member of the IS481 family. Its closest homolog is ISRae1 from Rhodococcus aetherivorans (amino acid identity, 82% [274/333]). TMP is a 25-kDa Gap-like (TauE/SafE superfamily) putative transmembrane protein with a possible role in small-molecule transport/export. See text and Table S1 for other TnRErm46 components. DR, direct repeats (shaded) at the junction with genomic DNA and adjacent inverted repeats (IR), which comprise the CTAG sequence targeted by TnRErm46. This sequence provides the TAG stop codons for the ISRe46 transposase on the right end and for some target genes—whose function is thus not interrupted (e.g., parB on the PAM 2287 pRErm46 plasmid)—on the left end. The left IR provides the TGA stop codon of the transposon’s distal hypothetical gene (HP). Stop codons are boxed. (B) Genetic structure of pRErm46’s class 1 integron (C1I). The black rectangles flanking IS6100 represent the 14-bp terminal inverted repeats with the sequence GGCTCTGTTGCAAA. The TnRErm46 insertion within the C1I cassette gene aadA9, unique to PAM 2287’s pRErm46, is indicated. The aadA9 gene is uninterrupted in most MLSr isolates where pRErm46 was detected, typically within an ≈4.4-kb contig covering the intI1-aadA9-qacE-sul1-orf5-tniΔ C1I sequence. PAM 2287’s pRErm46 carries two additional TnRErm46 copies (see Fig. 3 and Table S2). nt, nucleotide.
- FIG 2
Identification of erm(46)-associated plasmid DNA (pRErm46). Presence and absence of the macrolide resistance gene erm(46) or the virulence plasmid is indicated by plus and minus signs, respectively. The macrolide-susceptible, virulence plasmid-harboring R. equi 103S strain was used as a control. (A) Agarose gel of plasmid preparations of the indicated representative strains. M, molecular size marker (Promega 1-kb DNA ladder; top band is 10 kb). The arrowhead indicates the position of the virulence plasmid and the (pRErm46) resistance plasmid band; both migrate at a theoretical height of ≈15 kb due to supercoiling. The additional faster-migrating band in the plasmid preparations of PAM 2287 and cognate PAM 2350 transconjugant corresponds to a cryptic plasmid that appears to be inconsistently transferred. (B) Corresponding Southern blots probed with erm(46), IS6100, and vapA (pVAPA virulence plasmid) probes (see text for details).
- FIG 3
Genetic structure of pRErm46 from R. equi PAM 2287 (reference sequence, GenBank accession no. KY494640). Highlighted in color are the class 1 integron (yellow) and the TnRErm46 transposon (three copies, in red; underlined, TnRErm46 insertion shared by all pRErm46 plasmids, presumably the original site from which secondary transpositions took place). The macrolide resistance erm(46) gene within TnRErm46 is in black. Excluding the integrative elements, the pRErm46 backbone (in gray) is a conjugative replicon of 56.7 kb. pRErm46 (PAM 2287) sequence annotation is in Table S1.
- FIG 4
TnRErm46 insertions identified in pRErm46, the R. equi chromosome, and pVAPA virulence plasmid from the 18 MLSr equine isolates analyzed in this study (cumulative). Numbers correspond to the insertion sites as per the sequence coordinates of the reference R. equi 103S chromosome (GenBank accession no. FN563149) (21), 103S pVAPA (pVAPA1037, GenBank accession no. AM947677) (8), and PAM 2287 pRErm46 (GenBank accession KY494640 [this study]). In parentheses is the number of instances in which a particular insertion was detected. In pRErm46, 16/16 corresponds to 100% of isolates in which pRErm46 sequences were detected in the genome assemblies. One of the 20 chromosomal insertion sites was in a resistant clone-specific region. The accessory genome accounts for ≈20% of an R. equi isolate’s gene content (34).
- FIG 5
Fate of TnRErm46 macrolide resistance transposon upon acquisition by an equine R. equi isolate. (Top left) Formation of pRErm46 via transposition of TnRErm46 into a Corynebacteriales conjugative replicon; (bottom left) pRErm46 is conjugally transferred from a hypothetical Corynebacteriales donor to R. equi. The highly mobile TnRErm46 element can transpose from its original location (position 32567, common to all R. equi pRErm46 plasmids [see text and Fig. 4]) to other genome sites, resulting in several possible scenarios, as follows. (a) transposition within pRErm46 and onto the chromosome but not the virulence plasmid; (b) same as in panel a plus transposition to pVAPA; (c) same as in panel b, with subsequent loss of pRErm46; (d) same as in panel a, with subsequent loss of pRErm46. The pervasive colonization of the R. equi genome by TnRErm46 leads to stabilization of macrolide resistance in the host strain, while potential lateral transfer of TnRErm46 may occur via pRErm46, the pVAPA virulence plasmid, or both (solid red arrows). Options a, b, and d, in addition to TnRErm46 remaining in single copy at its original location in pRErm46 (isolate PAM 2275), have been verified in the 18 MLSr equine isolates analyzed (see Table S2).
- FIG 6
Clonal spread of TnRErm46-mediated macrolide resistance. Phylogenetic tree of 46 R. equi isolates based on core genome SNP analysis using Parsnp (58). The genomes analyzed are from 18 MLSr and 6 control MLSs equine isolates from different U.S. states (highlighted in red and blue, respectively) plus 22 isolates representative of the genomic diversity of the species, including the reference genome from strain 103S (DSM 104936 = NCTC 13926; GenBank accession no. FN563149 [21]) and the type strain DSM 20307 (= ATCC 6939; GenBank accession no. LWTX00000000 [34]). Numbers in the nodes indicate bootstrap values for 1,000 replicates. The tree graph was constructed with FigTree (http://tree.bio.ed.ac.uk/software/figtree/). See also Fig. S2.
- FIG 7
Acquisition of pRErm46 is fitness neutral. pRErm46 was conjugally transferred from the MDR clone (PAM 2287) to a different chromosomal lineage (strain 103S [see Fig. 6]). PAM 2350, pRErm46 103S transconjugant. PAM 2266, isogenic recipient 103S bacteria without pRErm46 (virulence plasmid-cured 103S with a chromosomal Hygr marker for counterselection of pRErm46 donor strain in mating experiments). (A) Growth curves in rich complex medium (LB or BHI) and chemically defined medium (mREMM [see Materials and Methods]). (B) Growth rate (μ) and maximum growth (A) values from experiments in panel A. (C) Competitive growth experiments. No significant differences between the isogenic pair of bacteria with and without pRErm46 were observed (all P values ≥ 0.7, one-way ANOVA followed by Šidák post hoc comparison).
Supplemental Material
TABLE S2
MLSr and control MLSs equine isolates from the United States analyzed in this study with indication of identified TnRErm46 insertion sites. Download Table S2, PDF file, 0.1 MB.
Copyright © 2019 Álvarez-Narváez et al.
This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.
TABLE S1
Annotation of pRErm46 from PAM 2287 (see Fig. 3 for the genetic structure of the plasmid). Download Table S1, PDF file, 0.2 MB.
Copyright © 2019 Álvarez-Narváez et al.
This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.
FIG S1
TnRErm46 insertions in pVAPA from PAM 2351 transconjugant (original donor, MLSr equine isolate PAM 2287 [Table S2]). PAM 2351 has four TnRErm46 copies as determined by SMRT (PacBio) sequencing. One of them is immediately at the left of the original (position 55130) insertion and resulted from a secondary transposition into position 49488 downstream of pVAPA_0361 in the plasmid partitioning region. According to the plasmid scaffold assembly, this transposition was associated with the deletion of the left region of the original copy of TnRErm46 (encoding the hypothetical protein, TetR-like regulator, and membrane transporter) plus an adjacent 5,684-bp plasmid segment encompassing ORFs pVAPA_0370 to _0410. Another TnRErm46 copy was inserted at position 62049 at the 3′ end of ORF pVAPA_480 (orf4/virR virulence regulator) in the plasmid vap PAI (M. Letek, A. A. Ocampo-Sosa, M. Sanders, U. Fogarty, T. Buckley, D. P. Leadon, P. Gonzalez, M. Scortti, W. G. Meijer, J. Parkhill, S. Bentley, and J. A. Vázquez-Boland JA, J Bacteriol 190:5797–5805, 2008, https://doi.org/10.1128/JB.00468-08), the transposon’s DR providing the stop codon for the gene, thus not obviously affecting its functionality. Interestingly, this insertion was accompanied by a duplication of a portion of the vap PAI from pVAPA_0430 (lsr2) to pVAPA_480 (orf4/virR), where a third additional TnRErm46 insertion took place. Download FIG S1, PDF file, 0.3 MB.
Copyright © 2019 Álvarez-Narváez et al.
This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.
TABLE S3
Representative R. equi isolates used in this study. Genome assemblies were previously reported (E. Anastasi, I. MacArthur, M. Scortti, S. Alvarez, S. Giguere, and J. A. Vazquez-Boland, Genome Biol Evol 8:3140–3148, 2016, https://doi.org/10.1093/gbe/evw222). Download Table S3, PDF file, 0.1 MB.
Copyright © 2019 Álvarez-Narváez et al.
This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.
FIG S2
SNP analysis of the MDR R. equi 2287 clonal isolates (in red) and control susceptible equine isolates (in blue) against the prototype 2287 genome (highlighted in cyan). The reference R. equi genome strain 103S (M. Letek, P. González, I. Macarthur, H. Rodríguez, T. C. Freeman, A. Valero-Rello, M. Blanco, T. Buckley, I. Cherevach, R. Fahey, A. Hapeshi, J. Holdstock, D. Leadon, J. Navas, A. Ocampo, M. A. Quail, M. Sanders, M. M. Scortti, J. F. Prescott, U. Fogarty, W. G. Meijer, J. Parkhill, S. D. Bentley, and J. A. Vázquez-Boland, PLoS Genet 6:e1001145, 2010, https://doi.org/10.1371/journal.pgen.1001145) is boxed in cyan. See Fig. 6 and the text. SNPs are represented as purple vertical lines, and phylogenetic relationships are shown in a neighbor-joining tree. Note that the clonal isolates differ by only a few SNPs (43 to 102) compared to the >25,000-SNP difference between the representative macrolide-susceptible isolates. SNP visualization was achieved by the Gingr program of genomes aligned with Parsnp in the Harvest suite (T. J. Treangen, B. D. Ondov, S. Koren, and A. M. Phillippy, Genome Biol 15:524, 2014, https://doi.org/10.1186/s13059-014-0524-x). Download FIG S2, PDF file, 0.3 MB.
Copyright © 2019 Álvarez-Narváez et al.
This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.
TABLE S4
Main oligonucleotides used in this study. Download Table S4, PDF file, 0.1 MB.
Copyright © 2019 Álvarez-Narváez et al.
This content is distributed under the terms of the Creative Commons Attribution 4.0 International license.
