Salmonella enterica serovar Dublin (S. Dublin) is a zoonotic infection that can be transmitted from cattle to humans through consumption of contaminated milk and milk products. Outbreaks of human infections by S. Dublin have been reported in several countries including high-income countries. A high proportion of S. Dublin cases in humans are associated with systemic illness. The genetic basis of virulence and invasiveness of S. Dublin is not well characterized. The aim of this study is to characterize the invasome of S. Dublin that enable the bacteria to cause systemic illness by the invasion of the bloodstream.
A set of S. Dublin submitted to Centre National de Référence des Salmonella, Institut Pasteur were selected for whole genome sequencing (WGS). The set of isolates included 22 human invasive isolates; 19 isolates from blood and 2 isolates from urine in addition to one isolate from pus. For comparison, we included 6 clinical non-invasive isolates from stool and 7 veterinary isolates. We also included the original S. Dublin isolate isolated from the stool of a patient in Dublin, Ireland (WS247) in 1929 giving the name of Dublin serovar. Furthermore, the reference S. Dublin isolate; SARB13 isolated in France from cattle in 1982 was also included in this study. Genomic DNA was prepared for Illumina pair-end (PE) sequencing using the Illumina NexteraXT® and the libraries were sequenced using an Illumina platform and MiSeq Control Software 18.104.22.168. All isolates were pair-end sequenced using 100bp PE libraries. Reads were assembled using Velvet and the best possible assembly with the highest N50 value was
annotated using RAST server. BLASTn was used for the alignment of virulence genes and genomic regions. Single nucleotide polymorphisms (SNPs) were identified using samtools mpileup. The best-fit model for nucleotides substitution was determined by jModelTest then a maximum likelihood (ML) phylogeny based on SNPs was constructed by MEGA6 software using 1000 bootstrap replicates. Acquired antimicrobial resistance genes were determined using ResFinder available from Center for Genomic Epidemiology https://cge.cbs.dtu.dk//services/all.php.
WGS revealed several mobile genetic elements (MGEs) that potentially enable the bacteria to cause invasive disease in humans including Gifsy-2 prophage, the novel pathogenicity island ST313-GI that harbours the putative virulence gene st313-td and Salmonella pathogenicity islands; SPI-6 harbouring T6SS, SPI-7 harbouring the Vi antigen and SPI-19 harbouring T6SS and the virulence plasmid. All S. Dublin isolates harbour Gifsy-2 prophage, SPI-6 and SPI-19 however the virulence plasmid, the pathogenicity island ST313-GI and the Vi antigen harboured by SPI-7 were absent from some S. Dublin invasive isolates indicating that they are not are the main virulence determinants in S. Dublin. Interestingly, two invasive human isolates from blood were resistant to multiple antibiotics; aminoglycosides, beta-lactam, sulphonamides, trimethoprim and phenicol indicating the emergence to multi-drug resistance in S. Dublin. The phylogenetic SNP analysis of S. Dublin isolates showed that invasive and gastroenteritis isolates were intermixed as SNPs were randomly distributed around the chromosome of S. Dublin.
WGS revealed several virulence factors that form the bacterial invasome and enable the bacteria to cause systemic illness in humans including Gifsy-2 prophage, T6SSSPI-6 and T6SSSPI-19 however no genomic markers were detected that differentiate among invasive and non-invasive isolates suggesting that host factors and immune response play a significant role in the disease outcome.