Deciphering the Structural Diversity and Classification of the Mobile Tigecycline Resistance Gene tet(X)-Bearing Plasmidome among Bacteria

Tigecycline is an expanded-spectrum tetracycline used as a last-resort antimicrobial for treating infections caused by superbugs such as carbapenemase-producing or colistin-resistant pathogens. Emergence of the plasmid-mediated mobile tigecycline resistance gene tet(X4) created a great public health concern. However, the diversity of tet(X4)-bearing plasmids and bacteria remains largely uninvestigated. To cover this knowledge gap, we comprehensively identified and characterized the tet(X)-bearing plasmidome in different sources using advanced sequencing technologies for the first time. The huge diversity of tet(X4)-bearing mobile elements demonstrates the high level of transmissibility of the tet(X4) gene among bacteria. It is crucial to enhance stringent surveillance of tet(X) genes in animal and human pathogens globally.

KEYWORDS tigecycline resistance, tet(X) variants, tandem repeats, resistance plasmidome, Nanopore sequencing A ntibiotics, considered a major breakthrough of modern medicine, are drugs utilized to treat bacterial infections in humans and animals (1,2). However, the emergence of antibiotic resistance (AR) is becoming a great threat to human and animal health worldwide, to a great extent resulting from misuse, abuse, and overuse of antibiotics (2). Furthermore, environmental contamination with antibiotic resistance genes (ARG) is a contributor to AR among pathogens (3,4). Owing to the complex facilitators of AR transmission in humans, animals, and environments, the One Health approach was proposed to tackle the expanding AR problem globally (5). The presence of multidrug-resistant (MDR) bacteria in animal fecal microbiotas as a factor in accelerating their transmission has been recognized, with a focus on the resistome derived from animal and environmental microbiomes (6). The structural diversity of resistance plasmids derived from fecal microbiotas has scarcely been investigated.
Tetracycline antibiotics have been a fundamental antibacterial agent for more than 6 decades and are widely used in clinical settings and animal sectors because of (i) their broad-spectrum activity against Gram-positive, Gram-negative, and atypical bacteria, (ii) their low cost, (iii) and their ability to be administered orally and intravenously (7)(8)(9)(10). Tetracyclines exert antimicrobial activity by inhibiting bacterial protein synthesis through binding to the 30S bacterial ribosome subunit (11). Due to the extensive usage of tetracyclines, resistance has emerged among commensal bacteria and pathogens via two major mechanisms, including efflux pumps and ribosome protection, and is now widespread (9,12). To counter tetracycline resistance, tigecycline, a semisynthetic glycylcycline derivative of tetracycline, was approved for clinical use in 2005 (9). With the emergence and spread of carbapenemase-producing Enterobacteriaceae (CRE) and colistin-resistant Enterobacteriaceae (13,14), tigecycline is regarded as the last-resort antibiotic to treat severe infections caused by MDR pathogens. Although tigecycline can evade bacterial tetracycline resistance posed by drug efflux pumps and ribosomal protection (15), resistance to tigecycline has been reported and is caused by upregulation of efflux pumps or mutations (16)(17)(18)(19)(20). Another tigecycline resistance mechanism can be conferred by a flavin-dependent monooxygenase, Tet(X), which can inactivate tigecycline enzymatically (21,22). Although tet(X) was first discovered in the obligate anaerobe Bacteroides fragilis (23), the emergence of tet(X) and its variants in clinical pathogenic microbiotas constitutes a potential public health risk (24)(25)(26). Environmental microbiota analyses indicated that tet(X) existed in environmental bacteria such as Flavobacterium and Bacteroides (27)(28)(29), which implied that environmental microbiotas may be one source of tet(X). However, the horizontal gene transfer (HGT) of tet(X) among microbiotas, especially those most closely related to human health, was not investigated in detail.
Recently, two reports highlighted the emergence of plasmid-mediated tet(X3) and tet(X4) conferring high-level tigecycline resistance in bacteria of animal, food, and human origins (30,31). This indicates that HGT of mobile tigecycline resistance tet(X) genes among pathogens is becoming a real threat, which is further demonstrated by the emergence of tet(X) genes from different sources, novel tet(X5) in Acinetobacter baumannii and tet(X6) in Proteus spp. (32)(33)(34)(35)(36)(37)(38). Various tet(X4)-bearing plasmids have been characterized (30,31,35,39,40), but the diversity and polymorphism of tet(X4)bearing genetic structures, especially in the tigecycline resistome of animal fecal microbiotas and surroundings, was not investigated systematically. In this study, we probed the diversity and polymorphism of tet(X4)-bearing plasmids genomically and phenotypically, with the perspective of the resistance plasmidome. The findings imply that the mobile tigecycline resistance gene tet(X4) is present extensively in fecal commensals and environments. that plasmid reorganizations may occur. The genetic basis of the tet(X4)-bearing plasmids and plasmid reorganizations was investigated further.
A wide variety of tet(X4)-harboring plasmids. According to S1-PFGE, a total of 27 tet(X4)-harboring plasmids with different sizes were selected and sequenced with the long-read Nanopore MinION platform, and these plasmids ranged from 12 to 294 kb in length (see Table S3). The majority of tet(X4)-harboring plasmids were de novo assembled completely by combining short-read and long-read data, and a few of them were de novo assembled based on available long-read data. Sixteen plasmids could be successfully transferred into the recipient E. coli C600. These 27 tet(X4)-harboring plasmids were categorized into nine different replicon types. Eight plasmids, ranging from 97 to 129 kb, belonged to IncFIB(K)/IncFIA(HI1)/IncX1 group of hybrid plasmids. Four plasmids, ranging from 112 to 136 kb, were classified as IncFIItype plasmids. Three plasmids, ranging from 31 to 50 kb, were classified as IncX1-type plasmids. The rest of the plasmids ranged from 12 to 294 kb and were classified as IncQ1 type, IncA/C2 type, IncFIB type, IncFIA(HI1)/IncHI1A/IncHI1B(R27) hybrid type, IncHI1B(R27)/IncFIA(HI1)/IncHI1A/IncX1 hybrid type, and IncFIB(K)/IncFIA(HI1)/IncHI1A/ IncHI1B(R27) hybrid type (from small to large) (see Table S3). In total, nine types of tet(X4)-harboring plasmids were detected in the same slaughterhouse. These types of plasmids were more abundant than the existing 15 tet(X4)-bearing plasmids of different replicons in the NCBI database (see Table S4), which indicated that the microbiome in slaughterhouse had become a reservoir of tet(X4). In addition, most of these tet(X4)-harboring plasmids were conjugative, showing that the plasmid-mediated trait greatly enhanced transmissibility of the tigecycline resistance gene tet(X4).
To investigate the characterization of all available tet(X4)-bearing plasmids, 15 tet(X4)-bearing plasmids from the NCBI nonredundant (nr) database were retrieved as of 1 December 2019 (see Table S4). These were analyzed with the 27 tet(X4)-bearing plasmids sequenced in this study to determine the distribution of resistance genes, replicons, and insertion sequences among the 42 plasmids (Fig. 2). It was found that ISCR2 was associated with tet(X4) in all plasmids, no matter what the plasmid types were. This observation was consistent with the idea that ISCR2 could mediate the generation of the circular form, facilitating transmission of tet(X4) between different genetic contexts (30,43). ISCR2 had already been reported to be associated with various  Table S4.
Characterization of tet(X)-Bearing Plasmids resistance genes, such as sul2 and floR (44). The resistance gene floR, conferring resistance to florfenicol, which is often used in animal settings, was the second most closely associated gene with tet(X4), being absent from only four plasmids, i.e., three IncQ plasmids and one IncA/C2 plasmid (Fig. 2a). To the best of our knowledge, the IncA/C2 plasmid pRF173-1_87k_tetX characterized in this study is a novel tet(X4)bearing plasmid. It is 87,445 bp in size and carries genes related to plasmid replicon, maintenance, conjugative elements, and resistance, including erm (42) and tet(X4) flanked by ISCR2 and IS26. Online BLASTn analysis showed that it was most similar to the online IncA/C plasmids found in Salmonella spp. and other Enterobacteriaceae at 88% coverage with more than 99% identity (Fig. 3). IncA/C plasmids are large, present in low copy numbers, conjugative, and associated with the emergence of multidrug resistance in enteric pathogens of humans and animals (45). ISCR elements, including ISCR2, are key players in IncA/C plasmid evolution (46), and the emergence of the tet(X4)-bearing IncA/C2 plasmid may be mediated by ISCR2 and the circular intermediate ISCR2-tet(X4)-abh detected previously (30). The occurrence of tet(X4) in this classical MDR plasmid would draw wide attention, and surveillance on tet(X4)-bearing IncA/C2 plasmids in environmental microbiotas and pathogens should be performed. The smallest tet(X4)-bearing plasmid identified in this study was pRF25-1_12k_tetX _flye, belonging to IncQ1, which was most similar to pLHM10-1-p6, with the genetic structure ISCR2-catD-tet(X4)-oISCR2 (see Fig. S1a) (31). IncQ plasmids are a large group of small mobilized plasmids (5.1 to 14.2 kb) with wide host ranges carrying various resistance genes (47,48). Their mobilization takes place with the help of other conjugative plasmids, which was proved by the cotransmission of pRF25-1_12k_tetX_flye and pRF25-1_147k_flye, an IncFIB/IncX1 hybrid conjugative plasmid harboring mph(A), aac(3)-VIa, aadA1, tet(M), qnrS1, bla TEM-1A , and floR. IncQ plasmids were found distributed in environmental microbiomes such as wastewater (49), provoking the concern that tet(X4)-bearing IncQ plasmids may exist in other environmental microbiomes.
Three IncX1 plasmids were also found among these 27 plasmids. In the NCBI database, some mcr-harboring plasmids and tet(X4)-harboring plasmids belonging to the IncX1 type and sharing similar backbones were found (39, 50) (Fig. 4). The tet(X4)-harboring IncX1 plasmids ranged from 31 to 59 kb and were characterized by two variable regions, including a multidrug resistance region and a type IV secretion system (T4SS) gene cluster containing different virB genes (Fig. 4). Two plasmids, pRB3-1_31k_tetX and pRF45-1_31k_tetX, with deletion of the T4SS gene cluster lost the conjugative ability. Owing to the existence of tet(X4)-harboring and mcr-1-harboring IncX1 plasmids, the co-occurrence of tet(X4) and mcr-1 in the same conjugative IncX1 plasmid should be put under close surveillance for risk assessment.
Four IncFII plasmids (pRF58-1_un_136k_tetX_flye, pRF2-1_117k_tetX_flye, pRF65-1_ 113k_tetX_flye, and pRF71-1_112k_tetX_flye) were identified in these 27 plasmids (see Table S3). BLASTn analysis of the four plasmids in NCBI database showed that they had 95% nucleotide identity at more than 70% coverage to p47EC (MK134376.1), which was the first identified tet(X4)-bearing plasmid in E. coli (30) (see Fig. S1d). Meanwhile, another IncFIB type plasmid, pRS3-1_136k_tetX_flye, also shared a similar backbone with p47EC (see Fig. S1d). Even though the backbones of the five plasmids were similar to that of p47EC (all of them belonged to the major IncF incompatibility type), they belonged to different IncF replicons compared with the plasmid p47EC (IncFIB).
Two tet(X4)-positive strains were detected in some samples. Interestingly, two different tet(X)-positive strains were detected in the same sample for seven samples (see Table S5). These strains were designated RB3-1, RB3-2, RS3-1, RS3-2, RW8-1, RW8-2, RF148-1, RF148-2, RF14-1, RF14-2, RF108-1, RF108-2, RF45-1, and RF45-2 and were divided into seven groups according to the sources. A single sample containing multiple tet(X4)-positive isolates suggested that tet(X4) had spread in the same micro-biota. We also found that the tet(X4)-harboring plasmids carried by these strains were diverse (see Table S5). Meanwhile, WGS analysis showed that strains RF108-1 and RF108-2 are similar phylogenetically to each other and that RF45-1 and RF45-2 have a common ancestor (Fig. 1). Copy number variations of the tet(X4) gene were found between tet(X4)-harboring plasmids in strains RF45-1 and RF45-2 (see Table S3). A helper plasmid was detected in strain RF108-2 but absent from strain RF108-1, and it played an important role in conjugation (see Table S6). Meanwhile, three repeats of tet(X4) were also found in pRF108-1_107k_tetX_flye, but only one tet(X4) copy was found in pRF108-2_97k_tetX, indicating that polymorphism of tandem repeats happened during bacterial division of the same clone. The resistome in a single sample was investigated previously (6, 28), but the tet(X4)-bearing plasmidome was not studied. Our initial attempt to recover different tet(X4)-bearing plasmids in a single microbiota was successful and proved the complex transmission routes of tet(X4) among environmental and fecal microbiota. This method is limited because of the number (only two in this study) of tet(X4)-bearing strains isolated from one sample; more colonies harboring tet(X4) should be recovered and analyzed phenotypically and genomically, and a long-read metagenomics method could be performed to analyze the polymorphism of tet(X4)-bearing structures in single-microbiota samples.
Co-occurrence of two different tet(X) variants in a single sample. Within the seven samples positive for two tet(X)-bearing isolates, strains RF14-1 and RF14-2 belonged to E. coli and Providencia rettgeri, respectively (see Table S5). A tet(X4) variant carried by an IncX1 plasmid, with a single nucleotide mutation compared with the tet(X4) gene (30), was found in strain RF14-1. This single-base substitution had no effect on the function of the tet(X4) gene, which was confirmed by measuring the MICs of E. coli harboring the cloning vector via the TA cloning method. Also, the IncX1-type plasmid carrying the tet(X4)-like gene could be successfully transferred into E. coli C600. The other two IncX1 plasmids carrying tet(X4) isolated in this study have no transferability (see Table S3). Further comparative analysis of two types of IncX1 plasmids was performed. The plasmids were subsequently analyzed using the Web-based tool oriTfinder to identify the vital elements related to conjugation (51), which showed that pRF14-1_50k_tetX had a complete module pertaining to conjugation which was lacking in pRB3-1_31k_tetX and pRF45-1_31k_tetX (Fig. 4). In Providencia rettgeri strain RF14-2, a novel tet(X) variant, designated tet(X6), was detected in an SXT/R391 integrative and conjugative element (ICE). The function of tet(X6), i.e., the ability to confer tigecycline resistance, was confirmed by the TA cloning method. It was homologous to tet(X4) (87.3%) and other tet(X) variants (36). Coincidentally, R391 was also first discovered in a Providencia rettgeri clinical isolate (52) and was later classified as the SXT/R391 family (53)(54)(55). The ICE in strain RF14-2 integrated into the 5= end of the gene prfC, which was a typical characteristic of all members of the SXT/R391 family (55,56). This novel SXT/R391 ICE was designated ICEPreChnRF14-2 according to the nomenclature system (54). Structural analysis showed that the tet(X6)-containing region was integrated into variable region III of the ICE (Fig. 5a). In addition, the ICE in RF14-2 had a structure similar to that of ICEPmiFra1 in Proteus mirabilis in the NCBI database (Fig. 5a). The tet(X6)-bearing structure was in the variable region III of the ICE, sharing a similar structure with other ICEs with different accessory regions characterized by ISCR2 (Fig. 5b). Although the prevalence of the tet(X6)-bearing ICE was low compared with the high incidence of tet(X4)-bearing plasmids among tet(X)-bearing strains in this study, the existence of a tet(X6)-bearing ICE located on the chromosome implied that tet(X) transmission occurred via multiple routes. The co-occurrence of different tet(X) variants in different genetic backgrounds of diverse bacteria highlighted the complex tet(X) evolution in microbiotas.
The diversity of tet(X4)-harboring contexts and tandem repeats. All the tet(X4)harboring genetic contexts from tet(X4)-bearing plasmids in the online database and in this study were analyzed and categorized into four major groups (Fig. 6a). ISCR2-tet(X4)-abh was the prominent structure observed in tet(X4)-bearing plasmids (see Table S5). The second type (G2), which had the reverse gene arrangement compared with the first type (G1), was divided into two subtypes depending on the presence of oISCR2 (Fig. 6a). The third type (G3), which had the conserved structure abh-tet(X4)-ISCR2-virD2-floR, was categorized into three subtypes with different genes-IS26, oISCR2, and ISCR2-in the upstream region. The last genetic structure type (G4) had the longest genetic region, abh-tet(X4)-ISCR2-yheScat-zitR-ISCR2-virD2-floR. The first structure type was mainly distributed in small plasmids, including IncQ and ColE2-like plasmids (see Table S4) (31,57). Distribution of other tet(X4)-bearing structures existed in plasmids of different replicons and chromosomes (30,34), without direct relationship to the tet(X4)-bearing structures.
Although Nanopore long-read data were generated to perform de novo assembly of tet(X4)-bearing plasmids, it was still impossible to obtain complete plasmid sequences for some tet(X4)-bearing plasmids, even with different assembly strategies (see Ta-ble S3). After single-plasmid-molecule analysis had been performed as described previously (58), all the uncompleted plasmids were found to harbor multiple copies of tet(X4). For tet(X4)-bearing plasmids with low copy numbers of tet(X4), Nanopore long reads could cover the repeat region, and complete plasmid sequences could be finished, such as those for pRF108-1_107k_tetX_flye and pRW8-2_117k_tetX_flye. However, for tet(X4)-bearing plasmids with tet(X4) tandem repeat regions larger than long reads, or with heterogenous numbers of tet(X4) tandem repeats, accurate copy numbers were difficult to obtain, resulting in failure of plasmid assembly. This demonstrated that the repeat regions containing tet(X4) were diverse and in a polymorphic state. To summarize, four kinds of repeat regions were detected, which ranged from 4.6 to 20 kb in length (Fig. 6b). They were widely distributed in different types of plasmids. The most common repeat structure was abh-tet(X4)-ISCR2 in 4,606 bp, which was the reported tet(X4)-bearing circular intermediate (4,608 bp) (30,43). This circular intermediate may play an important role in the generation of tet(X4) tandem repeat structures. An IncQ tet(X4)-bearing plasmid, pRB3-2_un_11k_tetX_flye in RB3-2, similar to pLHM10-1-p6 (31), was found in tandem repeats of the whole plasmid, and there could be as many as five repeats after analysis of all the long reads (see Fig. S2). This phenomenon was similar to the reported tet(X4)-bearing ColE2-like plasmid p16EC-9K, which was also observed in a polymorphic state of tandem plasmid repeats (57), and this may benefit the transmission of tet(X4) (59). Two additional large-repeat structures, IS26-aadA2lnu(F)-IS26-abh-tet(X4)-ISCR2-floR-tet(A)-oIS26 in 14 kb and oISCR2-erm(42)-ISCR2-abhtet(X4)-ISCR2-floR-IS26 in 20 kb, were found in RF45-2 and RF15-1, respectively. Certain long reads harboring tandem repeats of tet(X4) were illustrated to infer the complex structures (see Fig. S2). The 20-kb genetic structure was the longest tet(X4)-bearing tandem repeat, implying the complex structures of tet(X4) among plasmids. Although the multiple repeats of tet(X4) were common in the samples, the MICs of tetracycline, including tigecycline, were not affected significantly. The reason for the frequent occurrence of tet(X4) repeats in natural isolates, compared with the low prevalence of duplications of other resistance genes, warrants further investigations. Recently, the polymorphism of tet(X4), especially the tandem repeats of the ISCR2tet(X4) structure, probably generated by rolling-circle transposition, in E. coli was reported (57). Similarly, ISCR1-qnrB6 tandem repeats following a complex class I integron were also observed in the plasmidome of Salmonella and were resolved by single Characterization of tet(X)-Bearing Plasmids long-read analysis (58). Furthermore, the relative copy numbers of ISCR2 to tet(X4) based on WGS data were analyzed, and the results showed that ISCR2 copy numbers were equal to or higher than that of tet(X4) (Fig. 1), enhancing the idea that ISCR2 is a pivotal element facilitating tet(X4) transmission. The observation of multiple copy numbers of tet(X4) in this study suggests that ISCR elements may play important roles in the enrichment of resistance genes in microbiota.
Reorganizations of tet(X4)-bearing plasmids. Among the conjugative plasmids, the tet(X4)-bearing plasmids of six transconjugants showed different plasmid sizes compared with that of their parental strains after conjugation (Fig. 7a). From the results of previous studies (50,60,61), we speculated that the plasmids of these six transconjugants were formed via recombination in the process of conjugation. To probe the molecular mechanism of plasmid reorganization among the six tet(X4)-bearing strains, plasmid DNA was extracted from the six transconjugants, and long-read sequencing was performed. All the plasmids were finished in complete and circular forms except RF15-1 and TRF15-1 (see Table S6). Linear comparison between plasmid pTRW7-1_317k in transconjugant TRW7-1 and its parental plasmids in RW7-1 showed that pTRW7-1_317k was derived from the fusion of pRW7-1_235k_tetX and pRW7-1_81k by homologous recombination of IS26 (Fig. 7b). In TRF108-2, plasmid pTRF108-2_171k was composed of pRF108-2_97k_tetX and pRF8-1_74kb in donor strain RF8-1 generated by homologous recombination of IS26-ΔTnAs1 (Fig. 7c). The plasmid pTRW8-1_368k in TRW8-1 was formed by homologous recombination of pRW8-1_246k and pRW8-1_122k_tetX through the genetic structure IS26-mph(A)-orf-orf-ΔIS6100 (Fig. 7d). Similarly, plasmid pTRF10-1_388k was generated by homologous recombination of pRF10-1_269kb and pRF10-1_119k_tetX via the common region IS26-orf-sul3-orf-orf-aadA1-cmlA1-aadA2 (Fig. 7e). Distinctively, the generation of pTRF52-1-389kb was generated after interplasmid (pRF52-1_119k_tetX and pRF52-1_269kb) transposition via IS26 replicative transposition, resulting in duplications of IS26 and the target sequence in the cointegrate pTRF52-1_389kb (Fig. 7f). The role of IS26 in reorganizing plasmids by replicative transposition was recognized in other plasmids among MDR bacteria (62). This was the first report of tet(X4)-bearing plasmid reorganization by replicative transposition. For the plasmid reorganization in TRF15-1, the plasmids were not successfully assembled even with the Nanopore long-read data, which implies that possible underlying complex structures exist, and this warrants further investigation.
The tet(X4)-bearing plasmid reorganization would incorporate more resistance genes and replicon genes in the novel large MDR hybrid or cointegrate plasmids, which expanded the host range, causing a severe public health concern. Although the reorganizations were observed during conjugation under laboratory conditions, the prevalence of such MDR hybrid or cointegrate plasmids in the natural environment, with a focus on the MDR plasmidome among different environments, should be investigated.
Conclusions. The data presented in this study expand the understanding of diversity of tet(X4)-bearing plasmids, tet(X4)-bearing genetic contexts, and complex tet(X4)-bearing plasmid reorganization and highlight the wide distribution of various tet(X4)-bearing structures in different E. coli clones. Identification of tet(X6)bearing ICEs in the same tet(X4)-positive microbiota spotlighted the existence of coevolution of multiple tet(X) variants. Nanopore long reads significantly enhanced the characterization of polymorphism of tet(X4)-bearing plasmids from the perspective of single-plasmid-molecule analysis. The importance of plasmids and ISCR2 in facilitating the transmission of tet(X) was confirmed. In summary, this work demonstrates the significant role of the tet(X)-bearing plasmidome in the swine slaughterhouse for tet(X) transmission along the pork production chain, and stringent surveillance of tet(X)-bearing microbiotas of animals, humans, and the environment should be conducted to evaluate the risk posed by the emerging plasmid-mediated tigecycline-resistant tet(X) variants.  Characterization of tet(X)-Bearing Plasmids low temperature during rapid transfer to our lab for further processing. Solid and liquid samples (1 g each) or cotton swabs (surface samples) were incubated in 5 ml LB broth supplemented with tigecycline (2 mg/liter) for 6 h to enrich the tigecycline-resistant microbiota. The enriched cultures were streaked onto MacConkey agar plates containing tigecycline (2 mg/liter) to screen single colonies, and colonies with different morphology characteristics were further purified and stored in LB broth with 15% glycerol at Ϫ80°C. Genomic DNA of samples was prepared by the boiling method. The presence of tet(X) was checked using PCR with the primers tet(X)-F (5=-TGA ACC TGG TAA GAA GAA GTG-3=) and tet(X)-R (5=-CAG ACA ATA TCA AAG CAT CCA-3=) (581 bp). The PCR amplicons were subsequently sequenced by Sanger sequencing. 16S rRNA gene sequencing was performed to confirm species identification of the tet(X)positive isolates using universal primers (16S-F, 5=-AGA GTT TGA TCA TGG CTC-3=; 16S-R, 5=-GGT TAC CTT GTT ACG ACT T-3=).

MATERIALS AND METHODS
Antimicrobial susceptibility testing. The MICs of colistin and tigecycline were determined by the broth microdilution method in accordance with Clinical and Laboratory Standards Institute (CLSI) guidelines (63) and were interpreted according to the European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines, with the resistance breakpoint at Ͼ2 mg/liter for tigecycline (http:// www.eucast.org/clinical_breakpoints/) (64). The MICs of other antimicrobials for the tet(X)-positive isolates were measured by the agar microdilution method and interpreted in accordance with the CLSI standard (63). E. coli ATCC 25922 was used as the quality control.
Filter mating assay and S1-PFGE. To investigate the transferability of tet(X)-carrying genetic elements, a conjugation assay with a filter mating method (65) was carried out using tet(X)-positive strains as the donor strains and rifampin-resistant E. coli C600 as the recipient (1:4). Transconjugants were selected on LB agar plates supplemented with rifampin (300 mg/liter) and tigecycline (2 mg/liter). The transconjugants harboring tet(X4) were confirmed by PCR and antimicrobial susceptibility testing as described above. To characterize the tet(X)-bearing plasmid profiles, tet(X)positive strains and their transconjugants were digested with S1 nuclease (Takara, Osaka, Japan) followed by pulsed-field gel electrophoresis (PFGE) with the CHEF Mapper XA system (Bio-Rad, CA). The Salmonella Braenderup H9812 standard strain restricted with XbaI was used as the molecular marker.
Genome extractions, plasmid extractions, and high-throughput sequencing. Genomic DNA of the tigecycline-resistant strains was extracted using the TIANamp bacterial DNA kit (TianGen, Beijing, China), following the manufacturer's instruction. The plasmids of strains were extracted using the Qiagen plasmid midi-kit (Qiagen, Germany) after overnight culture in 100 ml LB broth. The genomic DNA of selected strains with different resistance phenotypes was subjected to short-read sequencing (2 ϫ 150 bp) with the Illumina HiSeq 2500 platform. Subsequently, genomic DNA of certain strains and plasmid DNA were sequenced with the Oxford Nanopore Technologies MinION long-read platform with the RBK004 barcoding library preparation kit and MinION R9.4.1 flow cells to obtain the complete sequences, as described previously (66,67).
Bioinformatics analysis. Short-read Illumina raw sequences of 27 strains were separately assembled using SPAdes (68), and contigs less than 500 bp were discarded. Multilocus sequence typing (MLST) of strains was performed using the mlst tool (https://github.com/tseemann/mlst) based on assembled contigs. The draft genomes were annotated using the software Prokka (69). The phylogenetic tree of these E. coli strains was constructed using Roary and FastTree based on SNPs of core genomes (70,71). The Flye long-read assembly tool was used to perform de novo assembly of Nanopore long-read MinION sequences of plasmids DNA and genomic DNA (72). The draft plasmid and tet(X4)-bearing structures were analyzed using the BLASTn program against the nr database. Genomic DNA with short-read Illumina and long-read Nanopore data were used to perform de novo assembly with the hybrid strategy as described previously (67). The high-quality complete genome sequences were annotated using RAST (http://rast.nmpdr.org/) automatically and modified manually. Plasmid replicons, insertion sequences, and antimicrobial resistance determinants were determined using online tools (https://cge.cbs.dtu.dk/ services/). BRIG and Easyfig were used to generate the genetic comparison figures (73,74). All tet(X4)bearing plasmids available in the NCBI nr database were downloaded for further analysis. The diversity and heterogeneous status of the tet(X4)-bearing plasmidome were investigated based on the Nanopore single-molecule analysis method (58).
Data availability. The tet(X4)-bearing plasmids generated in this study were deposited in the National Center for Biotechnology Information database (see Table S4). The assembled plasmid sequences with only Nanopore data and single long reads analyzed individually were deposited in the figshare database (https://figshare.com/s/6077f70a0ec952ee2796) for reference. Other data that support the findings of this research are available upon request.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only.