WGS and plasmid sequencing

Genomic DNAs were sheared to 600 bp segments with a focused-ultrasonicator S220 (Covaris) to prepare a short-read sequencing library. DNA was sheared to 8,000 bp segments for long-read sequencing with g-TUBE (Covaris). The short-read library was prepared with dual-indexed, 300 bp paired-end reads using MiSeq Reagent v3 kits (600 cycles) and sequenced with an MiSeq instrument (Illumina). The long-read library was prepared using Ligation Sequencing 1D kits and sequenced with an MinION sequencer (Oxford Nanopore Technologies). Genomic assembly of both short and long reads was performed with Unicycler v0.4.4 [12].

The genome sequences were annotated using the DNA Data Bank of Japan (DDBJ) Fast Annotation and Submission Tool (DFAST, https://dfast.nig.ac.jp/) as described by Tanizawa [13]. In the plasmid analysis, antibiotic resistance genes were identified using ResFinder 3.1 [14], plasmid incompatibility replicon typing was performed with PlasmidFinder 2.0 [15] and multilocus sequence typing (MLST) was performed with MLST 2.0 [16], on the Center for Genomic Epidemiology website (http://www.genomicepidemiology.org/). The bacterial species was identified using the ANI calculator (http://enve-omics.ce.gatech.edu/ani/) [17]. The species identification cut off was 95%–96% of the and the subspecies cutoff was 98% compared with the genomic sequences of the type strain [5, 18, 19]. The insertion sequence element was identified by IS finder (https://www-is.biotoul.fr/index.php). The integron ID was retrieved from the INTEGRALL database (http://integrall.bio.ua.pt/) [20]. Circular representations of plasmid sequences were visualized using the BLAST Ring Image Generator (BRIG, http://brig.sourceforge.net) [21]. The plasmid structure was generated in Easyfig 2.2.3 (https://mjsull.github.io/Easyfig/) [22].