A novel phage element of Salmonella enterica serovar Enteritidis P125109 contributes to accelerated type III secretion system 2-dependent early inflammation kinetics in a mouse colitis model

Abstract

Salmonella enterica subsp. I serovar Enteritidis exhibits type III secretion system 2 (TTSS2)-dependent early colonization and inflammation kinetics faster than those of closely related S. enterica serovar Typhimurium. To investigate the accelerated TTSS-2-dependent pathogenic potential of S. Enteritidis, we focused on its genome. Results of a previously published comparative genomic study revealed the presence of mutually exclusive genes in both serovars. In this study, we investigated the roles of six S. Enteritidis-specific genes in vivo by using differential fluorescence induction (DFI) through putative gene-specific promoters. The promoter construct associated with the gene locus SEN1140 induced green fluorescent protein (GFP) expression in the gut lumen, lamina propria, mesenteric lymph nodes, and related systemic organs. To further investigate the potential role of SEN1140, we compared a SEN1140 deletion mutant with S. Typhimurium in a TTSS1-deficient background. Interestingly, the S. Enteritidis mutant lacking SEN1140 did not show the unique TTSS-2-dependent early colonization and inflammation kinetic phenotype of S. Typhimurium. Consistent with this result, complementation of SEN1140 restored the TTSS-2-dependent accelerated inflammatory potential of S. Enteritidis. This report presents a suitable screening strategy that uses a combination of DFI, fluorescence-activated cell sorting, quantitative PCR, and wild-type isogenic tagged-strain techniques to explore the unique roles of S. Enteritidis-specific genes in bacterial pathogenesis.

Fig 1

Recombinant promoter construct-bearing Salmonella strains colonize equally in vivo. Freshly grown cultures of seven differentially WITS of S. Enteritidis wild-type strain M1525 harboring different promoter constructs integrated into a GFP reporter plasmid were mixed in equal proportions (1:1:1:1:1:1:1) to create a mixed-inoculum pool. (A) Individual WITS tags from three independent replicates of the inoculum pool were quantified by qPCR. All WITS showed an average equal proportion of 11 to 19% of the inoculum pool. (B) At day 1 p.i., all seven strains were found to have colonized well and to be shed in equal ratios in the feces. All strains showed corresponding relative WITS ratios of about 14.28%, which represents the theoretical average WITS ratio (shown by the dashed line) of individual WITS in a pool of seven strains.

Fig 2

The GFP-positive fraction was represented mainly by one promoter construct. A mixed-inoculum pool (1:1:1:1:1:1:1) of WITS of S. Enteritidis having specific promoter constructs was used to infected a group of five streptomycin-pretreated C57BL/6 mice, whose cecal contents were analyzed at day 2 p.i. (A) Cecum samples were collected and suspended in 500 μl PBST solution. The suspension was diluted 1:100, passed through a 50-μm sieve, and FACS sorted to collect GFP_Pos and GFP_Neg bacterial fractions from the third and fourth quadrants, respectively, during FACS sorting. The FACS was calibrated with a negative-control strain bearing plasmid pM968. (B) Relative WITS counts of GFP_Pos (●) and GFP_Neg (○) fractions of cecal contents. Collected GFP fractions were enriched in LB medium supplemented with an appropriate antibiotic. Genomic DNA was isolated from the culture, and qPCR was performed for WITS. Most of the GFP_Pos fraction was represented by strain M1525/pM2155. *, P < 0.001 (Kruskal-Wallis test). (C) Statistics of FACS-sorted GFP_Pos and GFP_Neg fractions obtained from cecal contents on a logarithmic scale showing higher expression of GFP through promoter construct pM2155 than the positive control pM975 in the mouse cecum.

Fig 3

Construct pM2155 is expressed at systemic sites in vivo. All of the constructs were tested in C57BL/6 for GFP expression at the systemic sites monitored after cecal colonization. (A) Cecal PFA sections were stained for immunofluorescence microscopy of the lamina propria. Shown are the counts of all of the GFP-expressing bacteria in the lamina propria. *, P < 0.05 (unpaired t test with Welch's correction). (B) Immunofluorescence staining of the lamina propria (i), spleen (ii), liver (iii), and MLN (iv) showing GFP-expressing strain M1525/pM2155. Bacteria expressing GFP at the host tissue sites are shown in the insets of all immunofluorescence images. Bars, 15 μm.

Fig 4

Strain Z290 colonizes efficiently in vivo. Different groups of streptomycin-pretreated C57BL/6 mice were infected with S. Enteritidis strain M1511 (SPI-1 deficient), S. Typhimurium strain SB566 (SPI-1 deficient), and S. enteritidis strain Z290 (SPI-1 deficient, SEN1140::cat). The cecum (A), MLN (B), spleen (C), and liver (D) were isolated at day 2 p.i., and the CFU in each organ were enumerated. P > 0.05 for Z290 versus the other strains.

Fig 5

Deletion of the SEN1140 gene lowers inflammation in the absence of TTSS1. The mouse groups shown in Fig. 4 were also analyzed for secretion of lipocalin 2 as a marker of cecal inflammation. (A) Fecal samples were collected from each group of mice at 12 and 24 h p.i. and suspended in sterile PBS, and the lipocalin 2 concentrations in the supernatants of fecal samples were estimated by ELISA. (B) Cecum sections from each group of mice were examined to determine pathoscores. (C) PFA sections of mouse cecum were stained for immunofluorescence microscopy, and the number of CFU of each strain expressing GFP in the lamina propria was determined. (D to F) H&E-stained representative cecum sections from each group of mice showing induced cecal inflammation: D, S. Enteritidis strain M1511 (SPI-1 deficient); E, S. Typhimurium strain SB566 (SPI-1 deficient; F, S. Enteritidis strain Z290 (SPI-1 deficient, SEN1140::cat). L, lumen; Lp, lamina propria; S, submucosal edema. Bars, 200 μm.