Detection of Genetic Elements Carrying vanA in Vancomycin-Resistant Enterococcus saigonensis VE80 T Isolated from Retail Chicken Meat

Vancomycin, a glycopeptide antibiotic, is important for treating serious infections caused by enterococci; however, its therapeutic efficiency is limited by the emergence of vancomycin-resistant enterococci (VRE). A systematic review and meta-analysis revealed an increasing trend in the isolation of vancomycin-resistant Enterococcus faecium isolates from bloodstream infections in the Eastern Mediterranean, Southeast Asia, and the Western Pacific in the early 2010s (Shiadeh et al., 2019). Most recently, the extensive clinical use of vancomycin and the nosocomial emergence of VRE in hospitals in Vietnam were reported (Santona et al., 2018; Vu et al., 2019).

In our previous investigation of antibiotic-resistant Enterococcus strains in retail food materials in Ho Chi Minh City, Vietnam, Enterococcus saigonensis, a novel species of the genus Enterococcus, harboring vanA were isolated from chicken meat and liver in December 2012 (Harada et al., 2016). Although the use of vancomycin or other glycopeptide antibiotics in animal production in Vietnam is not well documented, a recent survey of commercial feed in Vietnam did not identify avoparcin (a glycopeptide analogue) used as a growth promotor in any of the formulations examined (Nhung et al., 2016; Van Cuong et al., 2016). Therefore, the isolation of E. saigonensis harboring the vanA gene cluster from retail meat in Vietnam was unexpected.

The genetic determinants of VanA resistance typically reside on a mobile DNA element such as Tn1546, indicating the potential of horizontal transfer of vancomycin resistance to enterococcal species (Arthur and Courvalin, 1993). Moreover, the heterogeneity among Tn1546 elements provides a compelling rationale to pursue the molecular characterization of the Tn1546 structure to provide epidemiological information about the spread of resistance determinants (Willems et al., 1999; Guardabassi and Dalsgaard, 2004; Biavasco et al., 2007). Therefore, we aimed to detect genetic elements carrying vanA in E. saigonensis type strain VE80^T (VE80^T) in this study.

E. saigonensis VE80^T ( = JCM 31193^T = CCUG 68827^T) was resistant to vancomycin (minimum inhibitory concentration [MIC] of >256 μg/mL) and sensitive to teicoplanin (MIC of 8 μg/mL) (Harada et al., 2016). To detect the presence of a Tn1546 (or Tn1546-related) element of the VRE isolate, we conducted polymerase chain reaction (PCR) and Sanger sequence analyses of total DNA extracted using an alkaline lysis method and primers listed in Supplementary Table S1. We used BLASTN to compare the element nucleotide sequence detected in this study with those in the GenBank database (www.ncbi.nlm.nih.gov/BLAST/). In addition, the obtained sequence data were compared with the reference sequences such as GenBank accession nos. M97297, AB563188, and AB663321, using the Clustal W algorithm and GENETYX-win software (Version 11; GENETYX, Tokyo, Japan). To determine the genomic location of the vanA gene cluster of VE80^T, we used S1-nuclease pulsed-field gel electrophoresis (PFGE) and Southern hybridization methods with a vanA probe as previously described (Hashimoto et al., 2000; Harada et al., 2012; Terada et al., 2019). PFGE plugs were prepared essentially as described previously (Stanley et al., 1995), with 1% SeaKem Gold Agarose (Lonza, Basel, Switzerland) instead of low-melting-point agarose. A probe for vanA gene cluster was labeled using a PCR DIG Probe Synthesis Kit (Roche Products, Welwyn Garden City, United Kingdom) with the following primers: vanA-7409F and vanA-7986R (Harada et al., 2012). To monitor the robustness of Southern hybridization, E. faecium FN1 (Fujita et al., 1998) was used as a positive control.

Whole-genome sequencing of VE80^T was performed using a combination of the Illumina MiSeq (San Diego, CA) and PacBio RS II (Pacific Biosciences, Menlo Park, CA) platforms. Genomic DNA of VE80^T was extracted from overnight brain heart infusion broth cultures using a PowerSoil DNA Isolation Kit (MO BIO Laboratories, Inc., Carlsbad, CA). PacBio sequence reads were assembled de novo using HGAP 3.0 of SMRTPipe v.1.85 (Pacific Biosciences). Miseq reads were mapped onto the assembled Pacbio contig using CLC Genomics Workbench version 9.5.3 (CLC bio/Qiagen) to correct sequencing errors. The genome sequences were automatically annotated using the DNA Data Bank of Japan (DDBJ) Fast Annotation and Submission Tool (https://dfast.nig.ac.jp/).

PCR and Sanger sequence analyses of a Tn1546-related element revealed that VE80^T harbored a transposase (orf1) and a resolvase (orf2) gene of Tn1546 and all vancomycin-resistance genes corresponding to those of Tn1546 (GenBank accession no. M97297). Furthermore, orf1, orf2, vanR, vanH, vanA, vanX, and vanZ of the VE80^T genome were 100% identical to the corresponding genes encoded by the prototype Tn1546 element (Fig. 1). However, the VE80^T vanS sequence differed from that of Tn1546 by three base changes at Tn1546 nucleotide positions 4796 (T→G), 4808 (G→C), and 4855 (A→T), leading to three deduced amino acid changes in VanS at positions 50 (Leu→Val), 54 (Glu→Gln), and 69 (Gln→His), respectively. The vanY sequence from VE80^T differed from that of vanY of Tn1546 by one base change at Tn1546 nucleotide position 9280 (G→C), leading to a single deduced amino acid change in VanY at position 77 (Val→Leu). One base change was detected at Tn1546 nucleotide position 5880 in the intergenic region between vanS and vanH of VE80^T.

OPEN IN VIEWER

FIG. 1.

Maps of the genomic organization of the prototype Tn1546 and locations of point mutations in Tn1546-related elements of Enterococcus saigonensis VE80^T. The position numbers and base substitutions of mutations are indicated (Arthur and Courvalin, 1993).

In hybridization analyses, the vanA probe hybridized with the ∼76-kb plasmid DNA of VE80^T (Supplementary Fig. S1), suggesting the vanA cluster was located on a plasmid in VE80^T. Whole-genome analyses revealed that the complete genomic sequence of VE 80^T (2,740,356 bp; a G + C content of 36.32%) comprised the plasmids pVE80-1, 76,008 bp; pVE80-2, 25,275 bp; and pVE80-3, 3347 bp. Genome annotation data revealed that the vanA gene cluster of VE80^T was located on pVE80-1. The nucleotide sequences of the vanA gene cluster and the genome sequence of VE80^T were deposited into the nucleotide sequence database of the DDBJ under accession nos. LC517537 and AP022822–AP022825, respectively.

Sequence comparisons further revealed that the sequence of the Tn1546-related element of VE80^T is 100% identical to that of plasmid pTW9 DNA of Enterococcus faecalis isolated from poultry in Taiwan (GenBank accession no. AB563188). Furthermore, the vanS sequence with three point mutations compared with Tn1546, which causes teicoplanin susceptible, was 100% identical to those of vanA-genotype E. faecalis and E. faecium isolated from broiler droppings in Japan (Hashimoto et al., 2000) or chicken meat in Taiwan (Lauderdale et al., 2002), as well as to those of Enterococcus cecorum isolated from retail chicken meat in Japan (GenBank accession no. AB663321; Harada et al., 2012). Of these vanA-genotype VRE isolated in Japan, E. faecalis GV2 possesses vanA gene clusters on a conjugative plasmid, pMG2 (Hashimoto et al., 2000). These data indicate that this element may be prevalent in poultry production farms in Asia through transposons and transferable plasmids.