Salmonella enterotoxin (Stn) regulates membrane composition and integrity

Abstract

The mechanism of action of Salmonella enterotoxin (Stn) as a virulence factor in disease is controversial. Studies of Stn have indicated both positive and negative effects on Salmonella virulence. In this study, we attempted to evaluate Stn function and its effects on Salmonella virulence. To investigate Stn function, we first performed in vitro and in vivo analysis using mammalian cells and a murine ileal loop model. In these systems, we did not observe differences in virulence phenotypes between wild-type Salmonella and an stn gene-deleted mutant. We next characterized the phenotypes and molecular properties of the mutant strain under various in vitro conditions. The proteomic profiles of the total cell membrane protein fraction differed between wild type and mutant in that there was an absence of a protein in the mutant strain, which was identified as OmpA. By far-western blotting, OmpA was found to interact directly with Stn. To verify this result, the morphology of Salmonella was examined by transmission electron microscopy, with OmpA localization being analyzed by immunogold labeling. Compared with wild-type Salmonella, the mutant strain had a different pole structure and a thin periplasmic space; OmpA was not seen in the mutant. These results indicate that Stn, via regulation of OmpA membrane localization, functions in the maintenance of membrane composition and integrity.

Fig. 1.

Roles of Stn in Salmonella virulence. (A) Invasion assay using HeLa cells. Invasion assay was performed at 37°C, 5% CO₂ for 1 hour at MOI=10. (B) Survival assay using human macrophages. Differentiated U937 cells were infected at MOI=10 for 10 minutes. Data are means ± s.d. of three independent experiments with assays in duplicate. (C) Enterotoxicity of Salmonella strains. Fluid accumulation was measured as the amount of fluid content (in ml) per length (in cm) of ligated murine intestinal loop. PBS was used as the control reaction of this experiment. The results are mean ± s.d. values in five experimental animals. Statistical significance was set at P<0.05 (*, vs PBS). WT, wild-type strain.

Fig. 2.

Regulation of OmpA localization by Stn. (A) SDS-PAGE of prepared protein fractions. Proteins were loaded on 10% gel in each lane as follows: 10 μg of whole cell protein fraction, 5 μg of secreted protein fraction, 10 μg of periplasmic protein fraction and 5 μg of total membrane protein fraction. Gel was stained with Coomassie Brilliant Blue R-250. M, protein marker; lane 1, wild type; lane 2, Δstn. (B,C) Transmission electron microscopic image of wild-type (B) and Δstn mutant (C) cells in ultrathin sections (magnification: 12,000×). Bacteria were cultured in LB medium at 37°C for 16 hours. (D–G) Immunogold labeling of OmpA using anti-OmpA antibody (magnification: 20,000×). Bacteria were cultured on LB agar plates at 37°C for 16 hours. D, wild-type strain; E, Δstn; F, ΔompA; G, wild type using normal mouse serum for control reaction. Arrows indicate the gold particles.

Fig. 3.

Quantification of mRNA levels associated with the membrane protein genes in Salmonella. Bacteria were cultured in LB medium until mid-log phase (OD₆₀₀=0.8). Fold change in the mRNA levels from the Δstn versus the parent strain was measured by quantitative real-time PCR. mRNA levels were normalized to 16S rRNA levels. Data are expressed as means ± s.d. values of five independent experiments.

Fig. 4.

Interaction between Stn and OmpA. OmpA (1 μg/reaction) was separated by SDS-PAGE and transferred to PVDF membranes, which were soaked in the presence of purified TF-tag (10 μg; A) or purified TF-Stn (10 μg; B), and also in the absence of probe (C) at 4°C for 16 hours. Interaction of TF-Stn and OmpA was detected with anti-TF antibody. Asterisks indicate the non-specific signals with anti-TF antibody and arrow is the specific signal that formed as a result of the OmpA-Stn complex. Panel C was performed as a control reaction for the verification of antibody quality. TF, trigger factor-tag used as a negative control.

Fig. 5.

Sequence alignment of partial Stn and each of the A subunits of CT and LT. Amino acid sequences of Stn, CT (CT-A) and LT (LT-A) were derived from S. typhimurium strain Q1, Vibrio cholerae El Tor strain N16961, and enterotoxigenic E. coli strain H10407, respectively (Chopra et al., 1994; Heidelberg et al., 2000; Crossman et al., 2010). Asterisks represent the identical amino acid residues in Stn compared with those of CT and LT. Box and italic letters indicate the conserved motif and catalytic amino acid for ADP-ribosyltransferase of CT and LT (Laing et al., 2011).