Coexistence of blaOXA-58 and tet(X) on a Novel Plasmid in Acinetobacter sp. From Pig in Shanghai, China

The purpose of this study was to characterize the complete sequence of a novel plasmid carrying tigecycline resistance gene tet(X) and carbapenemase gene blaOXA-58 from a swine Acinetobacter sp. strain SH19PTT10. Minimal inhibitory concentration (MIC) was performed using microbroth dilution method. The isolate SH19PTT10 was highly resistant (16 mg/L) to tigecycline, and also exhibited resistance to ampicillin, streptomycin, tetracycline, chloramphenicol, florfenicol, ciprofloxacin, and sulfamethoxazole/trimethoprim. Although SH19PTT10 harbored blaOXA-58, it was susceptible to cefotaxime and meropenem. The genome sequence of SH19PTT10 was determined using PacBio single-molecule real-time sequencing. Plasmid pYUSHP10-1 had a size of 174,032 bp and showed partial homology to several plasmids found in Acinetobacter isolates. It contained two repA genes, putative toxin-antitoxin systems (HipA/HipB, RelE/RelB, and BrnT/BrnA), partitioning genes (parA and parB), and heavy metal resistance-associated genes (copA/copB, nrp, and czcA/czcD) but the transfer region or proteins was not found. pYUSHP10-1 carried 16 resistance genes, mainly clustered in two mosaic multiresistance regions (MRRs). The first MRR contained sul3, qacI-aadA1-clmA1-aadA2-blaCARB-2-dfrA16 cassette, aac(3)-IId, and blaOXA-58. The blaOXA-58 gene was associated with ISAba3, as previously described. The second MRR is the tet(X) region (ISAcsp12-aph(3')-Ia-IS26-ΔxerD-tet(X)-res-ISCR2-sul2) related to the corresponding region in other tet(X)-bearing plasmids. The pdif sites, as well as mobile elements, play an important role in mobilization of DNA modules and plasmid evolution. Coexistence of numerous resistance genes on a single plasmid may contribute to the dissemination of these genes under pressure posed by different agents, which may explain the presence of clinically crucial resistance genes tet(X) and blaOXA-58 in livestock. Thus, rational drug use and continued surveillance of tet(X) and blaOXA-58 in livestock are warranted.

Tigecycline, belonging to the novel glycylcycline class, has been a last-resort antibiotic to treat serious infections caused by extensively drug-resistant Gram-negative bacteria, including Acinetobacter (Noskin, 2005;Asif et al., 2018). However, novel plasmid-mediated high-level tigecycline resistance genes tet(X) [former name tet(X3)~tet(X5)] 2 have been identified in Acinetobacter isolates from animals and humans in China in 2019 (He et al., 2019;Wang et al., 2019). The emergence and dissemination of tet(X) will impair the efficacy of tigecycline in clinical treatment, thus would pose a significant threat to public health. The co-location of tet(X) and carbapenem resistance gene bla NDM-1 was previously described in Acinetobacter isolates from animals (duck, goose, and cow) and the environment (soil and sewage; Cui et al., 2020;He et al., 2020). In this study, we aimed to determine and analyze the complete sequence of a single plasmid bearing tet(X) [formerly designated as tet(X3), GenBank accession no. MK134375] and bla OXA-58 obtained from a swine Acinetobacter sp. strain in Shanghai, China, providing insights into the genetic structures of the plasmid and these genes.

Bacterial Strain and tet(X) Detection
In September 2019, one strain SH19PTT10 was isolated from the feces sample of a pig by Tryptic Soy Agar plate containing tigecycline (2 mg/L) from a pig farm located in Shanghai, China and was identified using 16S ribosomal RNA (rRNA) gene sequencing (Kim et al., 2010). The presence of tet (X) [former name tet(X3), tet(X4), and tet(X5)] was detected by PCR and sequencing (He et al., 2019;Wang et al., 2019).

Conjugation/Transformation Experiments
Conjugation experiments were performed using streptomycinresistant E. coli C600 as the recipient strain, as previously described (Chen et al., 2007). Transconjugants were selected using 2 mg/L tigecycline and 3,000 mg/L streptomycin. Transformation was carried out by heat-shock and electroporation using E. coli DH5α and Acinetobacter baumannii ATCC 19606. Transformants were selected by 2 mg/L tigecycline.

Whole Genome Sequencing and Analysis
The whole genome of SH19PTT10 was extracted and sequenced using PacBio single-molecule real-time sequencing (RSII platform, Pacific Biosciences, Menlo Park, CA, United States). Raw sequence data were introduced into the non-hybrid hierarchical genome assembly process (HGAP version 4). The 16S rRNA gene sequences of SH19PTT10 and other representatives across the genus Acinetobacter were aligned using ClustalW, and a phylogenetic tree was constructed by the neighbor joining algorithm using MEGA 7.0 (Kumar et al., 2008). The plasmid sequence was analyzed and annotated using the RAST server (Aziz et al., 2008), ResFinder 3.2 (Zankari et al., 2012), ISfinder (Siguier et al., 2006), PlasmidFinder (Carattoli et al., 2014), BLAST 4 , and Gene Construction Kit 4.5 (Textco BioSoftware, Inc., Raleigh, NC, United States). The replicase genes (rep) of plasmids from SH19PTT10 were assigned to a group according to the typing scheme for plasmids in A. baumannii (Bertini et al., 2010). Plasmid pYUSHP10-1 was compared with other plasmids using BLASTn and BRIG (Alikhan et al., 2011).

Nucleotide Sequence Accession Number
The complete sequence of pYUSHP10-1 was deposited in GenBank under the accession number MT107270.

The pdif Sites of pYUSHP10-1
Recently, numerous plasmids from various Acinetobacter species have been identified to contain 28-bp pdif site, consisting of 11-bp inversely-oriented binding sites for the XerC and XerD recombinases separated by a spacer of 6 bp (Supplementary Table S2; D' Andrea et al., 2009;Merino et al., 2010;Blackwell and Hall, 2017;Cameranesi et al., 2018). We also found 17 pdif sites on pYUSHP10-1 (Figure 1; Supplementary Tables  S2 and S3). The modules flanked by inversely-oriented pdif sites either with the XerC sites internal (D/C and C/D) or the XerD sites internal (C/D and D/C) are identified as dif module and may be able to mobilize mediated by pdif sites using XerC-XerD recombination system (D' Andrea et al., 2009;Merino et al., 2010;Blackwell and Hall, 2017;Cameranesi et al., 2018). For example, one module in pYUSHP10-1 consisting of ISAjo2like and putative toxin-antitoxin system BrnT/BrnA, was surrounded by inversely-oriented pdif sites (XerD/C and XerC/D; Supplementary Figure S2). This dif module was adjacent to a hypothetical protein and a further dif module carrying another putative toxin-antitoxin system encoding ParE toxin and helixturn-helix family protein (Supplementary Figure S2). Interestingly, the pdif sites flanking two dif modules were also  surrounding the hypothetical protein in the opposite orientation, making it being a putative dif module that may be mobile (Supplementary Figure S2). Different dif module combinations were found in some Acinetobacter plasmids such as pABIR (EU294228), pJ9-3 (CP041590), and pM131-2 (JX101647), indicating that the presence of the pdif sites facilitated the mobilization of discrete DNA segment through multiple events (Supplementary Figure S2). Additional dif modules identified in pYUSHP10-1 mostly include toxin-antitoxin systems, and copA/copB were also found in dif module (Figure 1).

pYUSHP10-1 Carrying Resistance Genes and dif Modules
pYUSHP10-1 contained 16 resistance genes, which were mainly clustered in two mosaic multiresistance regions (MRRs; Figure 1). The first MRR carried an approximately 32.2-kb segment consisting of three parts. The first of these (~11.6 kb) was bounded at both ends by IS26 and comprised a truncated mefB (encoding macrolide efflux protein), sul3 (sulfonamide resistance), and an incomplete class 1 integron with ΔintI1 and the qacI-aadA1-clmA1-aadA2-bla CARB-2 -dfrA16 cassette array (Figure 2). This fragment showed 99.9% identity to those of plasmids from E. coli such as pMB5876 (MK070495) and pMRSN346355_67.9 (CP018123), suggesting that pYUSHP10-1 may capture this segment from E. coli plasmids by IS26-mediated transposition or homologous recombination (Figure 2). The second part (3,296 bp) contained an open reading frame (ORF) encoding AAA family ATPase, aminoglycoside resistance gene aac(3)-IId, and mobile element ISAba14 (Figure 2). It was highly similar (>99%) to the corresponding regions of multiple plasmids found in Acinetobacter and Enterobacteriaceae isolates, such as Acinetobacter plasmids p225n_1 and pC54_001 and E. coli plasmid pEC2-NDM-3 (KC999035). Interestingly, the pdif site adjacent to ISAba14 and the pdif site upstream of IS26 were in inverse orientation, suggesting that the acquisition of ~16.5 kb segment including two hypothetical proteins and the first and second parts of MRR I in pYUSHP10-1 was possibly mediated by the pdif sites via site-specific recombination (Figure 2).
The last part of MRR I (17,301 bp) was highly similar to those of Acinetobacter plasmids p225n_1 (one single nucleotide polymorphism), p19110F47-2, pC54_001, and pAP43-OXA58-NDM1 with additional deletions (Figure 2). The primary component of this fragment is the bla OXA-58 region. As previously described (Fu et al., 2014;Feng et al., 2016;Klotz et al., 2017), bla OXA-58 was flanked by two copies of ISAba3 with opposite orientation, although the upstream ISAba3 was incomplete in Frontiers in Microbiology | www.frontiersin.org 5 September 2020 | Volume 11 | Article 578020 pYUSHP10-1 (Figure 2). OXA-58 shows weak activity against the carbapenems and is unable to hydrolyze some cephalosporin such as ceftazidime and cefotaxime (Poirel et al., 2005). The insertion of other IS elements, such as ISAba825, ISOur1, IS1008, and IS1006 into the upstream ISAba3 may provide a promoter to enhance the expression of OXA-58 and leading to the carbapenem resistance (Fu et al., 2014;Narciso et al., 2017;Chen et al., 2019;Matos et al., 2019). In this study, we did not observe insertion of IS element into ISAba3, suggesting the lacking of a putative promoter for bla  overexpression, thus the strain SH19PTT10 showed susceptibility to cefotaxime (4 mg/L) and meropenem (0.5 mg/L). The bla OXA-58 region has been frequently bracketed by two Re27 sequences, recognized as pdif sites later (D' Andrea et al., 2009), which have been shown to mediate the mobilization of bla OXA-58 (Poirel and Nordmann, 2006). In this study, two pdif sites in inverse orientation (XerD-XerC and XerC-XerD) were identified flanking the 2,257-bp segment containing ISAba3bla OXA-58 -ΔISAba3 arrangement in pYUSHP10-1, as observed in numerous Acinetobacter plasmids (Figure 2). The downstream ISAba3 was commonly followed by the putative transcriptional regulator gene araC1 and threonine efflux protein gene lysE in Acinetobacter isolates (Figure 2; Fu et al., 2014;Feng et al., 2016;Matos et al., 2019). An additional pdif site was found adjacent to lysE (Figure 2), which may account for the mobilization of this segment together with the bla OXA-58 dif module.
However, the conserved structure araC1-lysE was not found in pYUSHP10-1. A truncated araC1 was identified in p225n_1, pAP43-OXA58-NDM1, and pC54_001, possibly due to homologous recombination between araC1 and the downstream putative gene encoding site-specific integrase, which shared a sequence of 6 bp (TAAGTT) that might be the site of recombination (Figure 2). The same ΔaraC1 was also observed in p19110F47-2, which truncated by an incomplete ISAba20 (Figure 2).
The horizontal transfer of tet(X) is likely to be associated with ISCR2 (He et al., 2019). The generation of a circular molecule [ISCR2-xerD-tet(X)-res] by recombination between the 5 https://www-is.biotoul.fr two copies of ISCR2 in the same orientation could lead to the insertion of the tet(X) module (He et al., 2019). Although two copies of ISCR2 were only intact in p34AB, the formation of different but related tet(X) structures may have resulted from additional molecular events mediated by mobile elements (e.g., IS26 and Tn5393) via transposition or homologous recombination and have the potential to evolve diverse tet(X) genetic contexts.
pYUSHP10-1 also carried other resistance genes, including macrolide resistance genes mph(E)/msr(E), florfenicol resistance gene floR, and streptomycin resistance genes strA/strB (Figure 1). As previously described (Blackwell and Hall, 2017), a 2,950-bp segment including the macrolide resistance genes mph(E)/msr(E) was surrounded by two pdif sites in inverse orientation (Figure 1), further suggesting that the pdif sites may mediate mph(E)/msr(E) mobilization in Acinetobacter plasmids. Furthermore, many intact or truncated insertion sequences were present in the backbone of pYUSHP10-1, such as ISAba14 (Figure 1). These IS elements may enable this plasmid to capture more genes and evolve through IS-mediated recombination events.

CONCLUSION
We report the isolation and genetic characterization of an Acinetobacter sp. strain exhibited multiresistance phenotype, due to the acquisition of a novel plasmid carrying 16 resistance genes, including tigecycline resistance gene tet(X) and carbepenemase gene bla  . The Acinetobacter sp. may serve as an important reservoir of antimicrobial resistance genes. Coexistence of numerous resistance genes on a single plasmid may facilitate its dissemination and persistence under different selection pressure, which may explain the presence of clinically crucial antibiotic resistance genes tet(X) and bla OXA-58 in livestock. Additionally, at least two different mechanisms, site-specific