Salmonella Typhimurium Invalidated for the Three Currently Known Invasion Factors Keeps Its Ability to Invade Several Cell Models

Abstract

To establish an infection, Salmonella has to interact with eukaryotic cells. Invasion of non-phagocytic cells (i.e., epithelial, fibroblast and endothelial cells) involves either a trigger or a zipper mechanism mediated by the T3SS-1 or the invasin Rck, respectively. Another outer membrane protein, PagN, was also implicated in the invasion. However, other unknown invasion factors have been previously suggested. Our goal was to evaluate the invasion capability of a Salmonella Typhimurium strain invalidated for the three known invasion factors. Non-phagocytic cell lines of several animal origins were tested in a gentamicin protection assay. In most cells, we observed a drastic decrease in the invasion rate between the wild-type and the triple mutant. However, in five cell lines, the triple mutant invaded cells at a similarly high level to the wild-type, suggesting the existence of unidentified invasion factors. For the wild-type and the triple mutant, scanning-electron microscopy, confocal imaging and use of biochemical inhibitors confirmed their cellular uptake and showed a zipper-like mechanism of internalization involving both clathrin- and non-clathrin-dependent pathways. Despite a functional T3SS-1, the wild-type bacteria seemed to use the same entry route as the mutant in our cell model. All together, these results demonstrate the existence of unknown Salmonella invasion factors, which require further characterization.

Keywords: T3SS-1; T3SS-1 independent invasion; cell models; entry mechanism; trigger; zipper.

Figure 1

S. Typhimurium 14028 inactivated for the three known invasion factors remains invasive for several eukaryotic cell lines. Invasion abilities of S. Typhimurium 14028 and the STM-3Δ mutant were compared using gentamicin protection assays performed on different cell lines: epithelial cell lines (white bars), fibroblasts (green bars) and endothelial cell lines (brown bars). Bacteria (MOI = 10) were deposited on cells for 1.5 h followed by the addition of gentamicin (100 μg/mL) for 1.5 h. Empty bars represent the number of intracellular wild-type bacteria. Hatched bars represent the number of intracellular mutant bacteria. The results correspond to the mean ± SEM of at least three independent experiments performed in duplicate and expressed in log CFU/well. Cell lines were classified according to their dependence or not of T3SS-1 for invasion. Statistical analyses, using asymptomatic two-sample Fisher-Pitman permutation tests, were performed on intracellular bacterial counts between S. Typhimurium 14028 and STM-3Δ mutant for each cell type. Significance was ^*P < 0.05 and ^** P < 0.01.

Figure 2

Significant T3SS-1 independent invasion is also observed even after a short infection time. Gentamicin protection assays were performed on HeLa and AML-12 cells. Bacteria (MOI = 10) were deposited on cells for 15 min followed by the addition of gentamicin (100 μg/mL) for 1.5 h. Experiments were performed three times in duplicate and expressed as a percentage of relative infection. Data were analyzed using asymptotic two-sample Fisher-Pitman permutation tests. Significance was ^**P < 0.01. Statistical analyses also showed no significantly differences between the invasion rates of the STM-3Δ and the STM-ΔinvA mutants (p = 0.1969 in the HeLa cells vs. p = 0.3514 in the AML-12 cells).

Figure 3

Confocal microscopy reveals the intracellular location of S. Typhimurium inactivated for the three known invasion factors in AML-12 cells. One projection of Z-stack sections, obtained by confocal microscopy (Leica SP8), of AML-12 cells infected for 1.5 h at 37°C (MOI = 100) with either the S. Typhimurium 14028 (A) or the STM-3Δ mutant (B) that expressed DsRed and were then processed by immunofluorescence. Intracellular bacteria were red due to DsRed expression or red and green (white arrowheads) when extracellular due to DsRed expression staining with a Salmonella O-antigen antibody revealed with goat anti-rabbit Alexa 488 (green) before permeabilization. Cell nuclei (DAPI) are shown in blue and actin in white (Phalloidin Alexa 647). Dotted lines indicate the position of intracellular bacteria in the Z-stack. Scale bar: 10 μm.

Figure 4

Scanning electron microscopy reveals weak cell surface rearrangements assuming a zipper-like mechanism for both the S. Typhimurium wild-type and its triple mutant. Cells seeded on glass coverslips at 70% of confluence were infected with either the S. Typhimurium 14028 (A,B) or the STM-3Δ mutant (C,D) at MOI = 100 then washed in PBS and processed for scanning-electron microscopy. Cells were then prepared for scanning electron microscopy. We selected representative observations of AML-12 cell surface rearrangements induced by S. Typhimurium 14028 and STM-3Δ infection. They showed filipodia contact with Salmonella (A−60 min and C−90 min) with weak membrane invagination surrounding Salmonella and engulfment (B−150 min and D−150 min). Bars: 1 μm.

Figure 5

The T3SS-1 apparatus of the S. Typhimurium 14028 strain is functional in AML-12 cells. S. Typhimurium 14028 or ΔinvA expressing SopD effector-TEM-1 β-lactamase were used to infect AML-12 cells for 1.5 h. Following infection, cells were loaded with CCF4-AM for 2 h and visualized for green or blue fluorescence through microscopy. Green fluorescence indicates that CCF4-AM was loaded and the presence of blue cells due to cleaved CCF4-AM revealed translocation. AML-12 cells were non-infected (A), or infected with S. Typhimurium 14028 carrying the empty vector pCX340 (B), S. Typhimurium 14028 expressing SopD effector-TEM-1 in pCX340 (C) and STM-ΔinvA expressing SopD effector-TEM-1 in pCX340 (D).

Figure 6

Amiloride affected the invasion of S. Typhimurium 14028 strain in HeLa cells T3SS-1 dependent-invasion. Invasion abilities of the S. Typhimurium 14028 and STM-3Δ mutant were compared using gentamicin protection assays performed on HeLa cells, AML-12 cells and IPEC-1 cells in the presence or not of amiloride. On HeLa monolayers, the single mutant 1invA was added as a control. Cells were pre-treated with amiloride (1 mM) (black bars) or the culture medium (gray bars) for 30 min. Bacteria (MOI = 10) were deposited on cells for 1.5 h followed by the addition of gentamicin (100 μg/mL) for 1.5 h. The numbers of internalized bacteria were determined. Data are the mean ± SEM in duplicate and repeated at least twice for each strain. Statistical analyses, using asymptomatic two-sample Fisher-Pitman permutation tests, were performed on intracellular bacterial counts between the infection in presence of culture medium (gray) or amiloride (black). Significance was *P < 0.05.

Figure 7

Both clathrin- and non-clathrin-dependent pathways are used by the S. Typhimurium 14028 strain and its triple mutant during interaction with AML-12 cells. Invasion abilities of the S. Typhimurium 14028 and STM-3Δ mutant were compared using gentamicin protection assays performed on AML-12 cells in the presence of different drugs. AML-12 cells were pre-treated with inhibitors at the indicated concentrations or with the appropriate solvent used for the dilution of the drugs for 30 min with amiloride (culture medium), filipin (ethanol) and genistein (DMSO), for 1 h with CPZ (culture medium) and MDC (DMSO). Bacteria (MOI = 10) were deposited on cells for 1.5 h followed by the addition of gentamicin (100 μg/mL) for 1.5 h. The dilution-effect of each drug was performed and viability of the bacteria was checked in the presence of all dilutions of the drugs used. The numbers of internalized bacteria were determined and expressed relative to values obtained for cells treated with control-diluted reagent arbitrarily set at 100%. Data are the mean ± SEM in duplicate and repeated at least twice for each strain. Data were analyzed using asymptotic two-sample Fisher-Pitman permutation tests. Significance was ^***p < 0.001, ^**p < 0.01, ^*p < 0.05. (A). represents the results obtained for the wild-type bacteria. (B). represents the results obtained for the STM-3Δ.