A bioinformatic approach to identify core genome difference between Salmonella Pullorum and Salmonella Enteritidis

Highlights

• Identified core genome difference between S. Pullorum and S. Enteritidis.

• Identified putative host-specificity factors in S. Pullorum.

• Provided a novel bioinformatics workflow to identify core genome differences between user-defined bacterial clades.

Abstract

S. Pullorum and S. Enteritidis are closely related in genetic terms, but they show very different pathogenicity and host range. S. Enteritidis infects many different hosts, usually causing acute gastroenteritis, while S. Pullorum is restricted to avian, where it causes systemic disease in young animals. The reason why they differ in host range and pathogenicity is unknown. The core-genome denotes those genes that are present in all strains within a clade, and in the present work, an automated bioinformatics workflow was developed and applied to identify core-genome differences between these two serovars with the aim to identify genome features associated with host specificity of S. Pullorum. Results showed that S. Pullorum unique coding sequences (CDS) were mainly concentrated in three regions not present in S. Enteritidis, suggesting that such CDS were taken up probably during the separation of the two types from their common ancestor. One of the unique regions encoded Pathogenicity Islands 19 (SPI-19), which encodes a type VI secretion system (T6SS). Single-nucleotide polymorphism (SNP) analysis identified 1791 conserved SNPs in coding sequences between the two serovars, including several SNPs located in a type IV secretion system (T4SS). Analyzing of 100 bp regions upstream of coding sequences identified 443 conserved SNPs between the two serovars, including SNP variations in type III secretion system effector (T3SE). In conclusion, this analysis has identified genetic features encoding putative factors controlling host-specificity in S. Pullorum. The novel bioinformatic workflow and associated scripts can directly be applied to other bacteria to uncover the genome difference between clades.

Introduction

The majority of Salmonella serovars are broad host range pathogens, however, a few serovars only infect one or a few host species (Uzzau et al., 2000). The underlying mechanism behind host specificity of these serovars remains unknown. Salmonella Pullorum (formal name S. enterica subspecies enterica serovar Gallinarum biovar Pullorum (S. Pullorum)) belongs to the host specific group as it only infects avian, where it causes systemic disease with high mortality in young chicken (Barrow and Freitas Neto, 2011; Shivaprasad, 2000). Studies have revealed that S. Pullorum is closely related to S. Enteritidis (formal name S. enterica serovars Enteritidis) (Langridge et al., 2015; Thomson et al., 2008), another Salmonella serovar, which commonly infects avian, but has a broad host range and mainly causes gastroenteritis (Rodrigue et al., 1990). The close evolutionary relationship and yet different host range and pathogenicity makes comparison of genomics and traits of S. Pullorum and S. Enteritidis a suitable approach to identify genome features that are associated with host-specificity in Salmonella.

The development towards host-specificity of Salmonella serovars has likely been driven by several mechanisms. Comparative genomics have already uncovered important genome features of host-specific serovars, most notably that each of them contain a number of specific genes, and that host adaptation has been accompanied by pseudogene formation in genes that apparently are not needed for the host-specific infection (Langridge et al., 2015; Thomson et al., 2008). Also, studies in other bacteria have indicated the importance of horizontal gene transfer for the evolution of microbial genomes (Pal et al., 2005; Popa et al., 2011; Treangen and Rocha, 2011) leading to regions which are specific to particular bacterial (sub)species or strains, termed the ‘mobilome’ (Dobrindt et al., 2004; Ou et al., 2007; Siguier et al., 2006).

A pan-genome is the entire set of genes for all strains within a clade, and the core-genome represents genes present in all strains (Medini et al., 2005; Tettelin et al., 2005; Vernikos et al., 2015). As the number of available genomes increases, so does the pan-genome of each serovar (Baddam et al., 2014; Laing et al., 2017), and pan-genome analysis tools have been established to increase the power of genome comparisons (Xiao et al., 2015). Core-genome, on the other hand, denoted the set of genes present in more than 90% of members of a clade. Pan- and core-genome analysis can provide a useful framework to determine the genomic diversity of the dataset at hand (Vernikos et al., 2015).

Apart from pathotype diversity associated with gene gain or loss, single-nucleotide polymorphism (SNP) may also contribute to host-specificity (Bekal et al., 2016; Yue et al., 2015; Yue and Schifferli, 2014). Previous research has reported that SNP mutations in coding sequences cause phenotypic difference in Salmonella (Hopkins and Threlfall, 2004; Thornbrough and Worley, 2012). Functionally important SNP mutation can also exist in non-coding regions (Hammarlof et al., 2018; Zaunbrecher et al., 2009), such as in promoter regions or in recognition motifs for regulators, usually located at the upstream of gene start codon (Haugen et al., 2008; McClure, 1985). Thus, it is important to be able to combine searches for unique genes with searches for SNPs that are likely to influence phenotype.

In the current study, we determined the pan-genome of S. Pullorum and S. Enteritidis, and based on this, we identified the shared core-genome of the two serovars, as well as the unique parts of the core-genome in each serovar. The aim was to discover putative host-specificity associated genes and SNPs in S. Pullorum. In order to do so, we designed a novel workflow to identify core-SNP and core-upstream-SNP between groups of strains. Our new genome-based analysis workflow (Corevar) with related scripts can also be applied to studies of other bacterial clades without additional adjustments.