Recombinant Salmonella expressing SspH2-EscI fusion protein limits its colonization in mice

Abstract

Background: Activation of inflammasome contributes to the clearance of intracellular bacteria. C-terminus of E. coli EscI protein can activate NLRC4 (NLR family, CARD domain containing-4) inflammasome in macrophages. The purpose of this study was to determine if activation of NLRC4 inflammasome by EscI can reduce the colonization of Salmonella in mice.

Results: A recombinant S. typhimurium strain expressing fusion protein of the N-terminal SspH2 (a Salmonella type III secretion system 2 effector) and C-terminal EscI was constructed and designated as X4550(pYA3334-SspH2-EscI). In vitro assay showed that X4550(pYA3334-SspH2-EscI) significantly enhanced IL-1β and IL-18 secretion (P < 0.05) and pyroptotic cell death of mouse peritoneal macrophages, compared with those infected with control strain, X4550(pYA3334-SspH2). In vivo studies showed that colonization of X4550(pYA3334-SspH2-EscI) in both spleen and liver were significantly lower than that of X4550(pYA3334-SspH2) (P < 0.05). The bacterial counts of X4550(pYA3334-SspH2-EscI) in mice decreased, while those of X4550(pYA3334-SspH2) increased over the time after infection. Additionally, X4550(pYA3334-SspH2-EscI) induced a less pathological alteration in spleen and liver than X4550(pYA3334-SspH2).

Conclusion: Fusion protein SspH2-EscI may be translocated into macrophages and activate NLRC4 inflammasome, which limits Salmonella colonization in spleen and liver of mice.

Keywords: Colonization; Inflammasome; Mice; Salmonella; SspH2-EscI.

Fig. 1

Construction of recombinant bacteria. a PCR products. M, Marker; 1, 5’-terminal of sspH2; 2, 3’-terminal of escI. b Recombinant plasmid pMD20 T-SspH2-EscI digested with Nco I and Sal I. M1-M2, Marker; 1, pMD20 T-SspH2-EscI. c Recombinant plasmid pYA3334-SspH2-EscI digested with Nco I and Sal I. M1-M2, Marker; 1, pYA3334-SspH2-EscI. d Growing curve of recombinant bacteria

Fig. 2

In vitro infection of mouse peritoneal macrophages. C57BL/6 mouse peritoneal macrophages seeded on 96-well plates were pre-stimulated with 1 μg/ml E. coli lipopolysaccharide to induce the expression of pro-IL-1β for 3 h. The freshly cultured X4550(pYA3334-SspH2-EscI), X4550(pYA3334-SspH2) and X4550(pYA3334) were then added with the desired multiplicity of infection (MOI). The cell plate was centrifuged to enhance the contact of bacteria with the cells and the infected cells were then incubated for 30 min. The supernatants containing uninfected bacteria were replaced with RPMI 1640 complete medium (100 μl/well) containing 100 U/ml penicillin, 100 μg/ml streptomycin, and 1 μg/ml LPS prior to the start of the subsequent incubation. The uninfected cells were used as control. a Supernatant IL-1β and IL-18 levels at 4 h post-infection (hpi) with MOI 10, 50 and 100; b LDH release at 1, 3, 5 hpi with MOI 100 and 24 hpi with MOI 10, 50 and 100; c Intracellular caspase-1 activation at 1 h hpi with MOI 100; d Cell morphology at 24 hpi with MOI 100 (a, uninfection; b, infected with X4550(pYA3334); c, infected with X4550(pYA3334-SspH2); d, infected with X4550(pYA3334-SspH2-EscI), arrows show pyroptotic cell death); e The count of cells and intracellular bacteria at 24 hpi with MOI 100. The data shown are representative of three replicate experiments

Fig. 3

In vivo infection of mice. Six-week-old C57BL/6 mice were intravenously injected with either freshly collected X4550(pYA3334), X4550(pYA3334-SspH2) or X4550(pYA3334-SspH2-EscI), 1× 10⁶ cfu/mouse. Several days later, the weight of spleen (a), the bacterial colonization (b) in spleen and liver, and the contents of IL-6 and TNF in serum (c) were counted. Three weeks post-infection, the spleen and liver (d) of the mice were stained with hematoxylin-eosin for pathological assay, all scale bars represent 50 μm. Five mice were used in each treatment. The data shown are representative of three replicate experiments