Closed Genome Sequence of Salmonella enterica Serovar Richmond Strain CFSAN000191, Obtained with Nanopore Sequencing

Here we report the genome sequence of Salmonella enterica serovar Richmond strain CFSAN000191, isolated from tilapia from Thailand in 2005. The genome was determined by a combination of long-read and short-read sequencing.

ABSTRACT

Here we report the genome sequence of Salmonella enterica serovar Richmond strain CFSAN000191, isolated from tilapia from Thailand in 2005. The genome was determined by a combination of long-read and short-read sequencing. This strain was used for source tracking in a 2012 Salmonella enterica serovar Bareilly foodborne outbreak in the United States.

ANNOUNCEMENT

Salmonella enterica is one of the most important bacterial enteric pathogens and is implicated in foodborne illnesses worldwide (1). Among the many outbreaks of Salmonella infection occurring every year in the United States, there was an outbreak involving Salmonella enterica serovar Bareilly in 2012 that was linked to the consumption of a contaminated product containing raw yellowfin tuna (2). Single-nucleotide polymorphism (SNP) phylogenetic analysis using the whole-genome sequence (WGS) generated from 100 S. Bareilly strains (encompassing outbreak-related and non-outbreak-related strains) showed that the patients in the United States became infected with an S. Bareilly strain isolated from scraped tuna that was imported from a fishery in India for use in the production of spicy tuna sushi. The complete genome sequence of one representative strain (CFSAN000189) from that outbreak was obtained by PacBio sequencing (GenBank accession numbers CP006053 and CP006054) (2).

The genome of another Salmonella strain (CFSAN000191, reported as S. Bareilly) was sequenced to understand more about the diversity between S. Bareilly and Salmonella enterica serovar Richmond and to be used for future outbreak investigations. The strain was grown overnight in Luria-Bertani (LB) medium at 35°C, and the DNA was extracted with a DNeasy blood and tissue kit (Qiagen). The long reads for each strain were generated with MinION sequencing (Nanopore, Oxford, United Kingdom). The sequencing library was prepared with the RAD004 rapid sequencing kit. The sequencing library contained DNA fragmented randomly by a transposase present in the fragmentation mix of the RAD004 kit, rendering fragments >30 kb. This library was run in a FLO-MIN106 (R9.4.1) flow cell, according to the manufacturer’s instructions, for 48 h. The sequencing output was 3 Gb (340,000 reads, but only reads above 5 kb were used for the downstream analyses, 132,158 reads) for an estimated average genome coverage of 360×. The short-read whole-genome sequence (WGS) for strain CFSAN000191 generated previously at a genome average coverage of 50× (2) was retrieved from the NCBI (SRA accession number SRR498369). The final genome sequence was achieved using the pipeline described previously (3). Briefly, the genome sequence was obtained with de novo assembly using Nanopore data and default settings in the Canu program v1.7 (4). A second assembly was generated using a SPAdes (5) hybrid assembly (with default settings) using both Nanopore and MiSeq data generated for the strain. The resulting assemblies from Canu were error corrected using Pilon (6) and the MiSeq data. The final corrected assembly (FA) was generated by comparing the SPAdes hybrid and Canu-polished assemblies using Mauve (7). The two assemblies agreed in synteny and size, and therefore the SPAdes hybrid assembly was used as the FA. The FA consisted of a single contig of 4,726,630 bp (chromosome). The FA sequence was annotated using the NCBI Prokaryotic Genome Automatic Annotation Pipeline (PGAAP, https://www.ncbi.nlm.nih.gov/genome/annotation_prok). In silico multilocus sequence typing (MLST) analyses (https://enterobase.warwick.ac.uk/species/index/senterica) showed that CFSAN000191 belonged to sequence type 909 (ST909). In silico serotyping using SeqSero (8) (http://www.denglab.info/SeqSero), a tool to infer the serovar from the genes that determine antigenic structure, showed the strain to be S. Richmond (7:y:1,2) and not S. Bareilly (7:y:1,5), as initially suggested (2). The S. Bareilly and S. Richmond strains both belong to ST909. The GC content was 52.2%, similar to that of other Salmonella strains. Whole-genome SNP analysis, performed as described previously (9) but using the CFSAN000189 genome sequence as reference and 101 other S. Bareilly and S. Richmond genome sequences available at NCBI and used in a previous report (2), showed that the FA closed genome of this strain (CFSAN000191) was indistinguishable from the de novo MiSeq assembly for the same strain (JMMH00000000).

Data availability.

The GenBank accession number for this genome sequence is CP032622, and the SRA accession number for the Nanopore run is SRR7941349.