Skip to main content
NCBI home page
Infection and Immunity logoLink to Infection and Immunity
. 2014 Nov;82(11):4553–4562. doi: 10.1128/IAI.02297-14

Lipopolysaccharides Belonging to Different Salmonella Serovars Are Differentially Capable of Activating Toll-Like Receptor 4

Daniela Chessa a,✉, Luisella Spiga a, Nicola De Riu c, Paola Delaconi a, Vittorio Mazzarello a, Giulia Ganau a, Salvatore Rubino a,b
Editor: B A McCormick
PMCID: PMC4249323  PMID: 25135686

Abstract

Salmonella enterica subsp. enterica serovar (serotype) Abortusovis is a member of the Enterobacteriaceae. This serotype is naturally restricted to ovine species and does not infect humans. Limited information is available about the immune response of sheep to S. Abortusovis. S. Abortusovis, like Salmonella enterica subsp. enterica serovar Typhi, causes a systemic infection in which, under natural conditions, animals are not able to raise a rapid immune response. Failure to induce the appropriate response allows pathogens to reach the placenta and results in an abortion. Lipopolysaccharides (LPSs) are pathogen-associated molecular patterns (PAMPs) that are specific to bacteria and are not synthesized by the host. Toll-like receptors (TLRs) are a family of receptors that specifically recognize PAMPs. As a first step, we were able to identify the presence of Toll-like receptor 4 (TLR4) on the ovine placenta by using an immunohistochemistry technique. To our knowledge, this is the first work describing the interaction between S. Abortusovis LPS and TLR4. Experiments using an embryonic cell line (HEK293) transfected with human and ovine TLR4s showed a reduction of interleukin 8 (IL-8) production by S. Abortusovis and Salmonella enterica subsp. enterica serovar Paratyphi upon LPS stimulation compared to Salmonella enterica subsp. enterica serovar Typhimurium. Identical results were observed using heat-killed bacteria instead of LPS. Based on data obtained with TLR4 in vitro stimulation, we demonstrated that the serotype S. Abortusovis is able to successfully evade the immune system whereas S. Typhimurium and other serovars fail to do so.

INTRODUCTION

There are more than 2,500 known Salmonella serovars, and most of them cause gastroenteritis in humans, a localized infection characterized by acute intestinal inflammation, diarrhea, and fever (1, 2). Bacterial invasion and survival in host intestinal cells lead to the stimulation of the innate immune system, which in turn results in the massive intestinal inflammatory response that characterizes Salmonella-induced gastroenteritis. However, a few Salmonella serovars are host restricted and cause systemic infections that differ dramatically from gastroenteritis in their clinical presentation. One such serovar is Salmonella enterica subsp. enterica serovar Abortusovis, a sheep-adapted pathogen that does not infect humans. Abortion in sheep due to S. Abortusovis is an important health concern in the Mediterranean area, where the sheep industry has a significant economic impact (3). Infections have been reported in several countries where sheep farming is common (3–5). S. Abortusovis is commonly introduced into a flock by an infected sheep, which shows no overt disease manifestations but expels the pathogen through colostrum, milk, or feces (6). While adult sheep do not develop overt signs of disease, the fetus or newborn animals can develop lethal infections. In areas of endemicity, abortion occurs in 30% to 50% of sheep in a flock, generally during the first pregnancy and mainly during the last 2 months of gestation (a total of 5 months), through mechanisms that remain uncharacterized (3). After abortion, bacteria can be isolated from placenta and tissues (liver, spleen, brain, and stomach), which are the principal sites of multiplication.

Comparisons between adapted and nonadapted Salmonella serotypes suggest that the persistence of host-adapted S. Abortusovis serotypes could be based on circumvention of the immune-based protective response (7), with the ability to persist at systemic sites (8).

Host responses are frequently typical for groups of pathogens rather than being specific to individual pathogens, which suggests that some pathogens have different characteristics that are not recognized by the host immunity (9).

Translocation of bacteria from the intestinal lumen into the lamina propria is detected by the immune system through pattern recognition receptors (PRRs) (10) (Toll-like receptors [TLRs]) and the cytosolic nucleotide binding and oligomerization domain-like receptors (11). The PRR is able to recognize a microbe-associated molecular pattern (MAMP), but it is incapable of distinguishing bacteria from viruses or parasites (12).

Lipopolysaccharide (LPS) is a conserved MAMP localized in the outer membrane in Gram-negative bacteria. It is a potent agonist of the Toll-like receptor 4 (TLR4)-MD2-CD14 receptor complex (12), and it is considered the major bacterial antigen responsible for inducing the expression of proinflammatory molecules, such as tumor necrosis factor alpha (TNF-α), interleukin 1β (IL-1β), and IL-6, which allow the host to fight the intruder more efficiently.

Experimental evidence suggests that stimulation of TLR4 by LPS has an important role in the development of septic shock during Salmonella enterica subsp. enterica serovar Typhimurium infection (13–15).

As in the case of Salmonella enterica subsp. enterica serovar Typhi, the human-specific adapter serotype is not capable of eliciting a host response during in vivo infection. Interestingly, the LPS purified from S. Typhi elicits cytokine secretion in human monocytes at levels equal to those elicited by LPSs purified from nontyphoidal Salmonella serotypes (16). It is already known how S. Typhi evades recognition by TLR4 (17) and prevents the generation of a typical antibacterial response, which results in suppression of neutrophil recruitment in the intestinal mucosa (18) and weak induction of acute-phase responses during bacteremia (15).

Limited information is available about the immune response of sheep to S. Abortusovis. S. Abortusovis, like S. Typhi, causes a systemic infection in which, under natural conditions, animals are not able to mount the prompt immune response necessary to control the infection before the pathogens reach the uterus and cause abortions (19).

We have reported that the TLR4 pathway engaged by S. Typhimurium is not triggered by S. Abortusovis, suggesting that S. Abortusovis LPS might not be recognized by TLR4.

In order to decipher if the peculiar response to LPS induced by S. Abortusovis was due to malfunctions of three relevant virulence genes important for lipid A synthesis, we cloned the three genes—htrB, msbB, and ddg—and expressed them in S. Abortusovis.

MATERIALS AND METHODS

Bacterial strains and culture conditions.

The bacterial strains used in the present study are Salmonella Typhimurium (wild type) (20), a Salmonella Typhimurium msb mutant (21), Salmonella Abortusovis (22), and Salmonella enterica subsp. enterica serovar Paratyphi A (clinical isolates from Pakistan).

A mutation in msbB has been described previously (21) and was introduced into S. Abortusovis strain SS44 by P22 transduction to yield strain SSD1.

Cloning of the ddg, htrB, and msbB genes was performed with Escherichia coli strain DH5α (23).

The ddg, htrB, and msbB genes were amplified by PCR using the primers 5′-GAGCTCATGTTTCCTCAAAGCAAATT-3′ and 5′-GAATTCTCAGATGTAGAGTGACGCTT-3′ for ddg, 5′-GAGCTCATGACGAAG TTGCCTAAGTT-3′ and 5′-GAATTCTCAATAGCGCGACGGTACGC-3′ for htrB, and 5′-GAGCT CATGGAAACCAAAAAAAATAAT-3′ and 5′-CCCGGGTTATTTGATGGGATAAAGATC-3′ for msbB. The resulting 972-bp, 921-bp, and 972-bp PCR products were cloned into the vector pCR2.1 (Invitrogen), and amplification of the correct fragments was confirmed by sequence analysis. The htrB and ddg regions were cloned into the SacI and EcoRI sites, while msbB was cloned into the SacI and SmaI sites of the low-copy-number vector pWSK29 (24) to give rise to the plasmids pLS2, pLS3, and pLS4. Subsequently, the plasmids were introduced into S. Abortusovis by electroporation to yield strains SSD2, SSD3, and SSD4, respectively. The strains were cultured aerobically at 37°C in Luria-Bertani (LB) broth supplemented with antibiotics, as appropriate, at the following concentrations: carbenicillin, 100 mg/liter (LB Cb); kanamycin, 60 mg/liter (LB Km); and nalidixic acid, 50 mg/liter (LB Nal).

Construction of expression plasmids.

The TLR4 gene was PCR amplified from the chromosome of ovine placenta tissue, and the product was obtained using the primers 5′-AAGCTTATGATGGCGCGTGCCCGCCGGGCT-3′ and 5′-GAATTCGCGGCATTTACTTGTTAACTGA-3′. The resulting 2,599-kb PCR product was cloned into the vector pCR2.1 (Invitrogen), and amplification of the correct fragment was confirmed by sequence analysis. The TLR4 region was cloned at the HindIII and EcoRI sites of the high-copy-number mammalian expression vector pcDNA3.1(+) (Invitrogen) to give rise to the plasmid pLS1.

Transient transfection.

HEK293 cells were grown in 24- or 48-well tissue culture plates in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal calf serum (FCS) until 40% confluence was reached (∼24 h). The HEK293 cells were transiently transfected with a total of 250 ng of plasmid DNA. For the cotransfection experiments with pLS1 and pcDNA3.1, 100 ng of DNA for each plasmid was used. FuGene HD (Promega) was used as a transfection reagent at a lipid-to-DNA ratio of 5 to 1. As a control, nontransfected cells were used. After 48 h of incubation, the cells were infected with the appropriate bacterial strains (S. Typhimurium, S. Abortusovis, and S. Paratyphi A). Noninfected cells were used as a control.

LPS extraction conditions.

LPS was isolated from dried bacteria using hot phenol-water extraction (25–27). The crude LPS obtained was precipitated from the aqueous phase with 5 volumes of cold ethanol and separated by centrifugation (4°C and 8,000 × g for 20 min).

LPS Ultrapure from Salmonella Typhimurium, Salmonella Paratyphi A, and Salmonella Abortusovis was provided by LPS Bioscience, Orsay, France.

An LPS sample from each strain was loaded into a glycine–SDS-polyacrylamide gel for electrophoresis on a 15% acrylamide gel, followed by silver staining.

Tissue culture experiments.

Human embryonic kidney (HEK293) cells stably transfected with human TLR4-MD2-CD14 (HEK293-TLR4 cells) were purchased from Invivogen, San Diego, CA. The cells were maintained in DMEM with glucose (4.5 g/liter) and 10% FCS (Gibco) supplemented with blasticidin or hygromycin (Invivogen) according to the manufacturer's instructions. Cells were seeded (5 ×105 cells/well) in 24-well plates containing DMEM and 10% FCS and infected with 1 × 107 heat-killed bacteria/well for 90 min. To assess the response of the HEK293 cells to purified pathogen-associated molecular patterns (PAMPs), we added 10 μg of LPS from Salmonella enterica serotype Minnesota (LPS, 10 × μg/ml; Invivogen) and incubated the cells for 90 min. As a control, we used noninfected cells.

For gene expression analysis by real-time PCR, total RNA was extracted with TRI Reagent (Molecular Research Center) and processed according to the instructions of the manufacturer. Real-time PCR was performed using SYBR green (Promega) and the 7900HT Fast real-time PCR system. The fold change in mRNA levels was determined using the comparative threshold cycle (CT) method (Bio-Rad).

HEK293 cells stably transfected with human TLR4-MD2-CD14 were treated with 10 μg of LPS purified from S. Typhimurium, S. Abortusovis, and S. Paratyphi A, and also, 10 μg of LPS from the control S. enterica serotype Minnesota was added. All the LPSs were incubated for 90 min. As a control, we used nontreated cells. Next, we processed the cells for total RNA extractions as described above.

HEK293 cells transfected with human TLR4-MD2-CD14 were used to perform the blocking assay. We added 10 μg of LPS purified from S. Typhimurium in different wells, adding, in scalar concentrations, 10 μg, 20 μg, and 50 μg of the LPSs purified from S. Abortusovis and S. Paratyphi A. All the LPSs were incubated for 90 min. As a control, we used nontreated cells. Then, the cells were processed for total RNA extractions as described above.

HEK293 cells transiently transfected with ovine TLR4 were treated with 10 μg of the LPSs purified from S. Typhimurium, S. Abortusovis, and S. Paratyphi A, and as a control, we added 10 μg of LPS from S. enterica serotype Minnesota. All the LPSs were incubated for 90 min. As a control, we used nontreated cells. Finally, cells for total RNA extractions were processed as described above.

Cells transiently transfected with ovine TLR4 were seeded (5 × 105 cells/well) in 24-well plates containing DMEM and 10% FCS and infected with 1 × 107 heat-killed bacteria/well for 90 min. The experiment was performed as described above.

HEK293 cells transiently transfected with the ovine TLR4 were used to perform the blocking assay. The experiment was performed as described above.

The HEK293 cells transiently transfected with the ovine TLR4 were infected with 1 × 107 heat-killed bacteria/well of strains SSD2, SSD3, and SSD4. All the cells were incubated for 90 min. As a control, we used noninfected cells, and then, the cells were processed for total RNA extractions as described above.

Statistical analysis.

For statistical analysis of ratios (i.e., fold increases in IL-8 expression), data were transformed logarithmically to calculate geometric means. A parametric test (paired Student's t test) was used to calculate whether differences in the increases in IL-8 expression between the different LPSs used and between groups were statistically significant. A two-tailed P value of <0.05 was considered to be significant.

Fluorescence microscopy.

Sections of ovine placenta and human placenta were collected and deparaffinized by standard procedures, as previously described (28). Sections were blocked with PBS containing 2% bovine serum albumin (BSA) (blocking buffer) for 1 h.

Localization of the Toll-like receptor TLR4 in formalin-fixed, paraffin-embedded sections of ovine and human placentas was performed using TLR4 antibody (HTA125)- fluorescein isothiocyanate (FITC) (Santa Cruz, Inc.) The slides were incubated for 1 h at room temperature with the blocking buffer and then incubated with TLR4 antibody (1:250) for 1 h at room temperature in the dark before they were washed three times for 5 min in PBS plus gelatin (0.02%) and three times for 5 min each time in PBS. Excess liquid was removed from around the tissues, SlowFade Gold antifade reagent with DAPI (4′,6-diamidino-2-phenylindole) solution (Invitrogen) was applied to each slide, and then we applied the coverslip. The slides were examined by fluorescence microscopy.

To inhibit binding of TLR4 antibody with LPSs from S. Typhimurium, S. Abortusovis, and S. Paratyphi A, deparaffinized sections were blocked as described above. Approximately 10 μg of purified LPS in 0.3 ml of blocking buffer was added to slides and incubated for 1 h at room temperature. After three washes for 5 min each time in PBS plus gelatin (0.02%), a TLR4 antibody (1:250) was added in 0.5 ml of blocking buffer and incubated for 1 h in the dark at room temperature. Excess liquid was removed from around the tissues, and SlowFade Gold antifade reagent with DAPI solution (Invitrogen) was applied to each slide before the coverslip was placed on the slide. The slides were examined by fluorescence microscopy.

RESULTS

LPS analysis with SDS-PAGE.

LPS was isolated from dried bacteria using the hot phenol-water extraction method. SDS-PAGE analysis (Fig. 1) showed high heterogeneity, with “ladder-like” patterns of slowly migrating high-molecular-mass LPS species with O-polysaccharide chains of different lengths (smooth-type LPS), as well as fast-migrating bands of short-chain LPS molecular species with no O-polysaccharide chain attached to the core (rough-type LPS).

FIG 1.

FIG 1

Open in a new tab

SDS-PAGE analysis of LPSs. BP, LPS extracted from Bordetella pertussis; SAO, LPS from Salmonella Abortusovis; SPA, LPS from Salmonella Paratyphi A; STM, LPS from Salmonella Typhimurium; SGL, LPS from Shigella.

S. Abortusovis displayed a larger cluster of high-molecular-weight smooth-type LPS bands than S. Typhimurium. In contrast, the S. Paratyphi A electrophoretic profile displayed the presence of molecular species with shorter O-polysaccharide chains and the absence of clusters. These findings suggested that S. Abortusovis LPS is characterized by a very long O antigen, which is known to play an important role in colonization of host tissue (29–31).

Toll-like receptor 4 is expressed in both ovine placenta and human placenta.

TLR4 receptors are located on the surfaces of monocytes, macrophages, dendritic cells, and mast and myeloid cells on the intestinal epithelium. The receptor recognizes the LPS of Gram-negative bacteria, such as S. Abortusovis, and is therefore important for the activation of the innate immune system (3). In order to investigate the presence of TLR4 on ovine placenta tissue, we analyzed several samples kindly provided by the Laboratory of Pathology of the Istituto Zooprofilatico della Sardegna, Sassari, Italy.

The above-mentioned tissues were sampled both from a healthy pregnant sheep and from an S. Abortusovis-infected pregnant sheep.

After incubating the tissue samples with TLR4-FITC antibody, we noticed that TLR4 was present in the healthy control placenta samples and was widely distributed on the surface (Fig. 2B), while it was absent from the pathological tissues, probably mainly due to the pathological action exerted by the bacteria (data not shown).

FIG 2.

FIG 2

Open in a new tab

TLR4 distribution on ovine placenta tissue. (A) Binding of TLR4-FITC (green) antibody blocked with LPS of S. Typhimurium and nuclear stain (DAPI; blue). (B) Binding of TLR4-FITC antibody. (C) Binding of TLR4-FITC (green) antibody blocked with LPS of S. Paratyphi A and nuclear stain (DAPI; blue). (D) Binding of TLR4-FITC (green) antibody blocked with LPS of S. Abortusovis and nuclear stain (DAPI; blue). (E) Hematoxylin-eosin-stained section of the ovine placenta. A 10× objective was used for all panels.

Therefore, we hypothesized whether LPS belonging to S. Typhimurium, S. Abortusovis, or S. Paratyphi A could engage the TLR4 docking site, thus in turn avoiding specific antibody binding.

First, LPS Ultrapure isolated from S. Typhimurium was added to the placenta control tissue, and soon after that, we added the antibody specific for TLR4. Antibody binding to TLR4 was abrogated in the presence of S. Typhimurium LPS, as expected (Fig. 2A). Then, the experiment was repeated using purified LPSs from S. Abortusovis and from S. Paratyphi A; in these cases, as well, binding was abrogated (Fig. 2D and C, respectively). These results suggest that LPSs from S. Abortusovis, S. Paratyphi A, and S. Typhimurium are recognized by TLR4.

The same procedure was undertaken using human placenta tissue, which was donated by a patient who delivered a healthy baby at the University Hospital in Sassari, Italy. In this case, as well, all the LPS extracted was able to block recognition by the TLR4 antibody (Fig. 3).

FIG 3.

FIG 3

Open in a new tab

TLR4 distribution on human placenta tissue. (A) Binding of TLR4-FITC (green) antibody blocked with LPS of S. Typhimurium and nuclear stain (DAPI; blue). (B) Binding of TLR4-FITC (green) antibody. (C) Binding of TLR4-FITC (green) antibody blocked with LPS of S. Paratyphi A and nuclear stain (DAPI; blue). (D) Binding of TLR4-FITC (green) antibody blocked with LPS of S. Abortusovis and nuclear stain (DAPI; blue). (E) Hematoxylin-eosin-stained section of the ovine placenta. A 10× objective was used for all panels.

Taken together, these results demonstrate that LPSs isolated from the three above-mentioned Salmonella serotypes are able to engage TLR4 expressed on the surfaces of both human and ovine placentas.

Induction of TLR4 in vitro experiments.

HEK293-TLR4 cells were stimulated with heat-killed S. Abortusovis, S. Typhimurium, and S. Paratyphi A and with S. Typhimurium and S. Abortusovis with the msbB gene deleted (negative control). Heat-killed S. Abortusovis was added to the TLR4 cells, and expression of the IL-8 inflammatory chemokine was checked. S. Abortusovis and S. Paratyphi A were not capable of significantly triggering IL-8 expression, whereas S. Typhimurium did (P = 0.01), suggesting that the two host-specific serovars have LPSs that are able to interact with the receptor TLR4 present in these cells (Fig. 4A). These findings show for the first time that both S. Abortusovis and S. Paratyphi A, two host-specific serotypes, have in common the ability to hide themselves from the innate immune system.

FIG 4.

FIG 4

Open in a new tab

IL-8 expression elicited by S. Typhimurium, S. Abortusovis, and S. Paratyphi A or purified PAMPs in human embryonic kidney (HEK293) cells stably transfected with human TLR4. (A) IL-8 expression induced by purified PAMPs in HEK293 cells transfected with TLR4. The cells were incubated for 90 min with LPSs purified from S. Typhimurium, S. Abortusovis, and S. Paratyphi A (10 μg/well) and with an S. Minnesota LPS control (10 μg/well) prior to RNA extraction. (B) IL-8 expression induced by heat-killed S. Typhimurium, S. Abortusovis, and S. Paratyphi A in HEK293-TLR4 human cells infected with 1 ×107 bacteria/well for 90 min prior to RNA extraction. All data are shown as geometric means from three independent experiments ± standard errors.

To confirm these findings, LPSs were extracted from all the strains used in this study. As described in Materials and Methods, in order to determine LPS activity, we stimulated transfected cells with TLR4 and the three different LPSs, and the production of inflammatory cytokine (IL-8) was checked by real-time PCR.

IL-8 expression was induced in TLR4 cells by adding a commercially purified LPS preparation. The LPS extracted from S. Typhimurium was able to induce the production of IL-8, while the LPSs extracted from S. Abortusovis and S. Paratyphi A could not (P = 0.01). These results confirmed that IL-8 production in HEK293-TLR4 cells is due to the recognition of LPS by the TLR4 belonging to the TLR4-MD2-CD14 receptor complex. Notably, this interaction could not be observed upon LPS stimulation of S. Abortusovis and S. Paratyphi A (Fig. 4B).

Despite the fact that we were able to demonstrate that the LPSs purified from the two specific serotypes S. Abortusovis and S. Paratyphi A are engaged by TLR4, we could not detect any IL-8 production, suggesting that their LPSs may act differently.

Moreover, a blocking assay was set up in order to verify if the S. Abortusovis and S. Paratyphi A LPSs were able to compete with S. Typhimurium LPS for TLR4 binding. To achieve this goal, we stimulated the HEK293-TLR4 cells with 10 μg of S. Typhimurium LPS and added scalar concentrations of both S. Abortusovis and S. Paratyphi A LPSs.

Unexpectedly, the level of IL-8 produced by HEK293-TLR4 cell lines stimulated with S. Typhimurium LPS was reduced when either S. Abortusovis or S. Paratyphi A LPS was added (Fig. 5). This experiment showed that the three LPSs were competing for binding to the same receptor, but it also highlighted the fact that only the S. Typhimurium LPS is able to trigger the immune response.

FIG 5.

FIG 5

Open in a new tab

IL-8 expression elicited by LPSs of S. Typhimurium, S. Abortusovis, and S. Paratyphi A in human embryonic kidney (HEK293) cells stably transfected with human TLR4. The cells were incubated for 90 min with LPSs purified from S. Typhimurium (10 μg/well) plus S. Abortusovis or S. Paratyphi A (10 μg/well, 20 μg/well, and 50 μg/well). S. Minnesota LPS (10 μg/well) was used as a positive control prior to RNA extraction. ns, not significant. All data are shown as geometric means from three independent experiments ± standard errors.

Transient ovine TLR4 transfection and infections of HEK293 cells.

HEK293-TLR4 ovine cells were once again stimulated with heat-killed S. Abortusovis, S. Typhimurium, and S. Paratyphi A. Heat-killed S. Abortusovis was added to the TLR4 cells, and we measured the expression of the IL-8 chemokine. While neither S. Abortusovis nor S. Paratyphi A could trigger IL-8 expression, S. Typhimurium induced a relevant level of IL-8 expression (P = 0.005), further proving that the two host-specific serovars have LPSs that are able to interact with the ovine TLR4 receptor expressed by these cells (Fig. 6A). This seems to imply that this pattern recognition receptor in human and ovine cells is one of the key mechanisms used by S. Abortusovis and S. Paratyphi A to survive inside the host.

FIG 6.

FIG 6

Open in a new tab

IL-8 expression elicited by S. Typhimurium, S. Abortusovis, and S. Paratyphi A or purified PAMPs in human embryonic kidney (HEK293) cells transiently transfected with ovine TLR4. (A) IL-8 expression induced by heat-killed S. Typhimurium, S. Abortusovis, and S. Paratyphi A in HEK293-TLR4 human cells infected with 1 ×107 bacteria/well for 90 min prior to RNA extraction. (B) IL-8 expression induced by purified PAMPs. Cells were incubated for 90 min with LPSs purified from S. Typhimurium, S. Abortusovis, and S. Paratyphi A (10 μg/well) and S. Minnesota LPS (10 μg/well) prior to RNA extraction. All data are shown as geometric means from three independent experiments ± standard errors.

To validate this result, the LPSs extracted from all the strains were used once again to stimulate the above-mentioned transiently transfected (ovine TLR4) cell line.

Addition of a commercially purified LPS preparation induced IL-8 expression after the engagement of ovine TLR4. The LPS extracted from S. Typhimurium was able to induce the production of IL-8 (P = 0.001), while the LPSs extracted from S. Abortusovis and S. Paratyphi A could not. These results confirmed that IL-8 production in transiently transfected HEK293-TLR4 cells is due to the recognition of LPS by TLR4. Remarkably, this interaction could not be observed upon LPS stimulation of S. Abortusovis and S. Paratyphi A (Fig. 6B).

Even though we were able to demonstrate that the LPSs purified from the two different serotypes S. Abortusovis and S. Paratyphi A are able to bind TLR4, no IL-8 production was detected by real-time PCR, suggesting that their LPSs may act differently.

Following the same workflow, we added increasing concentrations of S. Abortusovis and S. Paratyphi A LPSs to HEK293-TLR4 ovine cells stimulated with a constant concentration of S. Typhimurium LPS (Fig. 7). The outcome of the experiment confirmed that the three LPSs were competing for the same binding site on the receptor and that only the S. Typhimurium LPS is able to trigger the immune response, inducing IL-8 secretion. By treating the HEK293 cells with our purified LPSs, we demonstrated that IL-8 production is indeed TLR4 dependent and not due to other impurities in the LPS preparations (Fig. 8).

FIG 7.

FIG 7

Open in a new tab

IL-8 expression elicited by LPSs of S. Typhimurium, S. Abortusovis, and S. Paratyphi A in human embryonic kidney (HEK293) cells transiently transfected with ovine TLR4. The cells were incubated for 90 min with LPSs purified from S. Typhimurium (10 μg/well) plus S. Abortusovis or S. Paratyphi A (10 μg/well, 20 μg/well, and 50 μg/well). S. Minnesota LPS (10 μg/well) was used as a positive control prior to RNA extraction. All data are shown as geometric means from three independent experiments ± standard errors.

FIG 8.

FIG 8

Open in a new tab

IL-8 expression elicited by LPSs of S. Typhimurium and S. Abortusovis in human embryonic kidney (HEK293) cells. The cells were incubated for 90 min with LPSs purified from S. Typhimurium (10 μg/well) and S. Abortusovis (10 μg/well) prior to RNA extraction. All data are shown as geometric means from three independent experiments ± standard errors.

Aiming to verify if the LPS extracted from S. Abortusovis was unable to stimulate IL-8 production due to the inactivation of three distinct (KDO)2-(lauroyl)-lipid IVA acyltransferase genes (htrB, msbB, and ddg), we cloned the htrB, msbB, and ddg genes from S. Typhimurium. S. Abortusovis transduction with the three genes was unable to restore LPS TLR4 signal transductions in the HEK293-TLR4 ovine cells used in this work (Fig. 9). The same experiment was performed using the HEK293-TLR4 human cell line with similar results (data not shown). Furthermore, we cloned and sequenced the htrB, msbB, and ddg genes from S. Abortusovis, and in silico analyses showed that there is 99.5%, 99.2%, and 99.1% sequence homology with same genes in S. Typhimurium. These findings were confirmed in a work recently published by our group (32).

FIG 9.

FIG 9

Open in a new tab

IL-8 expression elicited by S. Abortusovis wild type and with the plasmid pLS3-, pLS2-, and pLS4-expressed ddg, htrB, and msbB, respectively, in human embryonic kidney (HEK293) cells transiently transfected with ovine TLR4. The cells were infected with 1 ×107 bacteria/well for 90 min prior to RNA extraction. S. Minnesota LPS (10 μg/well) was used as a positive control prior to RNA extraction. All data are shown as geometric means from three independent experiments ± standard errors.

Our findings are intriguing, as they imply that a novel mechanism of immune evasion is used by S. Abortusovis.

DISCUSSION

LPS is a common structure belonging to enteric bacteria and one of the first targets recognized by the host immune system during infections. Aiming at deciphering why certain serovars of S. enterica are able to disseminate from the gastrointestinal (GI) tract to systemic sites in immunocompetent individuals while other serovars are able to colonize the GI tract, leading to an inflammatory response in the host, we studied the interactions between LPS and TLR4 using three different serovars. Until now, no “traditional” virulence factor has been identified that is present in all serovars capable of causing systemic disease, allowing them to be distinguished from the gastrointestinal serovars.

This strongly suggests that each serovar has acquired a unique set of genes that enable them to use distinct strategies to modulate the host immune response. Findings from this study suggest that S. Abortusovis LPS interacting with TLR4 (human and ovine) is unable to stimulate IL-8 production by a mechanism that does not imply the inactivation of three distinct (KDO)2-(lauroyl)-lipid IVA acyltransferase genes (htrB, msbB, and ddg) known to be relevant in LPS synthesis. Nonetheless, the findings reported in the present study are quite intriguing, as they imply that a novel mechanism of immune evasion is used by S. Abortusovis. Our goal was to better understand the pathogenic mechanisms of S. Abortusovis, a sheep-specific serovar that causes abortions.

Previous results obtained by our group using an ovine ileal loop assay showed that S. Abortusovis was less invasive than S. Typhimurium, Salmonella enterica subsp. enterica serovar Dublin, and Salmonella enterica subsp. enterica serovar Gallinarum and was comparable to an S. Typhimurium invH mutant.

Histological analysis of the ovine intestinal tissue used in this set of experiments showed that S. Abortusovis has the ability to cause no pathological changes with respect to the other serovars, highlighting unknown mechanisms used by this host-specific serovar to cause disease (33, 34). Thus, we decided to use a particular cell line to clarify the mechanism used by S. Abortusovis to avoid the host immune system.

Initially, we tried to localize TLR4 in tissues sampled from sheep placenta, and once this goal was achieved, we studied the interaction between TLR4 and the LPSs extracted from S. Typhimurium, S. Abortusovis, and S. Paratyphi A. The current literature reports that S. Typhimurium is able to activate cells expressing TLR4 and, conversely, that S. Typhi is not able to do this due to the presence of capsule (17).

LPSs were extracted from three Salmonella serovars and highly purified in order to obtain high-quality biological samples. Based on SDS-PAGE experiments, the three LPS samples were found to be of the smooth type, although no O chain cluster was clearly observed for S. Paratyphi A, in contrast to what was observed for the other two serovars.

Analyses of the three different O chains showed that the three serovars have O chains with different lengths; this difference may be highly important for bacterial adhesion, internalization, and virulence. Bravo et al. (35) reported that having O chains of different lengths represents an important virulence factor but does not interfere with TLR4 binding. This is in concordance with what was observed in the present study.

To confirm our first hypothesis, we used both the HEK293 cell line transfected with human TLR4-MD2-CD14 and the same cell line transiently transfected with ovine TLR4.

Interaction (TLR4-LPS binding plus IL-8 production) between TLR4 and purified LPS extracted from our bacteria was studied; moreover, we repeated the experiment, adding heat-killed bacteria instead of LPS. The results show that the LPSs of both S. Abortusovis and S. Paratyphi A reduced the expression of IL-8 in cell lines in the present study.

In addition, we demonstrated that, S. Typhimurium deficient in the msbB gene is not capable of inducing TLR-dependent IL-8 expression in host cells.

Several studies showed that TLR4 mutations are associated with increased sensitivity to various infections (36) and also highlighted the fact that TLR4 deficiency promotes infections caused by Gram-negative bacteria (37).

It was reported that Yersinia pestis modified LPS (38, 39) is able to activate TLR4, while the LPS extracted from the corresponding wild type was bound by the PRR but cannot activate it.

To our knowledge, this is the first study demonstrating that LPSs belonging to different Salmonella serovars are differentially capable of activating TLR4, even if all of them can engage it.

These different abilities explain why different Salmonella serotypes cause infections with diverse outcomes, and they may imply the existence of a yet-undiscovered pathogenic mechanism.

ACKNOWLEDGMENTS

We thank the Marie Curie People Programme, which supports our project.

We thank Giustina Casu Finlayson for helpful revision and Elena Uleri for help during transfection experiments.

Footnotes

Published ahead of print 18 August 2014

REFERENCES

  • 1.Kaufmann SH, Raupach B, Finlay BB. 2001. Introduction: microbiology and immunology: lessons learned from Salmonella. Microbes Infect. 3:1177–1181. 10.1016/S1286-4579(01)01498-8. [DOI] [PubMed] [Google Scholar]
  • 2.Wain J, Keddy KH, Hendriksen RS, Rubino S. 2013. Using next generation sequencing to tackle non-typhoidal Salmonella infections. J. Infect. Dev. Ctries. 7:1–5. 10.3855/jidc.3080. [DOI] [PubMed] [Google Scholar]
  • 3.Uzzau S, Brown DJ, Wallis T, Rubino S, Leori G, Bernard S, Casadesús J, Platt DJ, Olsen JE. 2000. Host adapted serotypes of Salmonella enterica. Epidemiol. Infect. 125:229–255. 10.1017/S0950268899004379. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Jack EJ. 1971. Salmonella abortion in sheep. Vet. Annu. 12:57–63. [Google Scholar]
  • 5.Pardon P, Sanchis R, Marly J, Lantier F, Pépin M, Popoff M. 1988. Ovine salmonellosis caused by Salmonella Abortusovis. Ann. Rech. Vet. 19:221–235. [PubMed] [Google Scholar]
  • 6.Nikbakht GH, Raffatellu M, Uzzau S, Tadjbakhsh H, Rubino S. 2002. IS200 fingerprinting of Salmonella enterica serotype Abortusovis strains isolated in Iran. Epidemiol. Infect. 128:333–336. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Montagne M, Martel A, Le Moual H. 2001. Characterization of the catalytic activities of the PhoQ histidine protein kinase of Salmonella enterica serovar Typhimurium. J. Bacteriol. 183:1787–1791. 10.1128/JB.183.5.1787-1791.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Uzzau S, Leori GS, Petruzzi V, Watson PR, Schianchi G, Bacciu D, Mazzarello V, Wallis TS, Rubino S. 2001. Salmonella enterica serovar-host specificity does not correlate with the magnitude of intestinal invasion in sheep. Infect. Immun. 69:3092–3099. 10.1128/IAI.69.5.3092-3099.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Tsolis RM, Young GM, Solnick JV, Bäumler AJ. 2008. From bench to bedside: stealth of enteroinvasive pathogens. Nat. Rev. Microbiol. 6:883–892. 10.1038/nrmicro2012. [DOI] [PubMed] [Google Scholar]
  • 10.Iwasaki A, Medzhitov R. 2004. Toll-like receptor control of the adaptive immune response. Nat. Immunol. 5:987–995. 10.1038/ni1112. [DOI] [PubMed] [Google Scholar]
  • 11.Franchi J, Marteau C, Crola da Silva C, Mitterrand M, André P, Kieda C. 2008. Cell model of inflammation. Biosci. Rep. 28:23–32. 10.1042/BSR20070012. [DOI] [PubMed] [Google Scholar]
  • 12.Hoebe K, Janssen E, Beutler B. 2004. The interface between innate and adaptive immunity. Nat. Immunol. 5:971–974. 10.1038/ni1004-971. [DOI] [PubMed] [Google Scholar]
  • 13.Khan SA, Everest P, Servos S, Foxwell N, Zähringer U, Brade H, Rietschel ET, Dougan G, Charles IG, Maskell DJ. 1998. A lethal role for lipid A in Salmonella infections. Mol. Microbiol. 29:571–579. 10.1046/j.1365-2958.1998.00952.x. [DOI] [PubMed] [Google Scholar]
  • 14.Engelberts I, von Asmuth EJ, van der Linden CJ, Buurman WA. 1991. The interrelation between TNF, IL-6, and PAF secretion induced by LPS in an in vivo and in vitro murine model. Lymphokine Cytokine Res. 10:127–131. [PubMed] [Google Scholar]
  • 15.Wilson RP, Raffatellu M, Chessa D, Winter SE, Tükel C, Bäumler AJ. 2008. The Vi-capsule prevents Toll-like receptor 4 recognition of Salmonella. Cell Microbiol. 10:876–890. 10.1111/j.1462-5822.2007.01090.x. [DOI] [PubMed] [Google Scholar]
  • 16.Bignold LP, Rogers SD, Siaw TM, Bahnisch J. 1991. Inhibition of chemotaxis of neutrophil leukocytes to interleukin-8 by endotoxins of various bacteria. Infect. Immun. 59:4255–4258. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Raffatellu M, Chessa D, Wilson RP, Dusold R, Rubino S, Baumler AJ. 2005. The Vi capsular antigen of Salmonella enterica serotype Typhi reduces Toll-like receptor-dependent interleukin-8 expression in the intestinal mucosa. Infect. Immun. 73:3367–3374. 10.1128/IAI.73.6.3367-3374.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Raffatellu M, Chessa D, Wilson RP, Tukel C, Akcelik M, Baumler AJ. 2006. Capsule-mediated immune evasion: a new hypothesis explaining aspects of typhoid fever pathogenesis. Infect. Immun. 74:19–27. 10.1128/IAI.74.1.19-27.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Cagiola M, Severi G, Forti K, Menichelli M, Papa P, De Giuseppe A, Pasquali P. 2007. Abortion due to Salmonella enterica serovar Abortusovis (S. Abortusovis) in ewes is associated to a lack of production of IFN-gamma and can be prevented by immunization with inactivated S. Abortusovis vaccine. Vet. Microbiol. 121:330–337. 10.1016/j.vetmic.2006.12.018. [DOI] [PubMed] [Google Scholar]
  • 20.Bäumler AJ, Tsolis RM, Heffron F. 1996. Contribution of fimbrial operons to attachment to and invasion of epithelial cell lines by Salmonella Typhimurium. Infect. Immun. 64:1862–1865. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Khan SA, Everest P, Servos S, Foxwell N, Zahringer U, Brade H, Rietschel ET, Dougan G, Charles IG, Maskell DJ. 1998. A lethal role for lipid A in Salmonella infections. Mol. Microbiol. 29:571–579. 10.1046/j.1365-2958.1998.00952.x. [DOI] [PubMed] [Google Scholar]
  • 22.Colombo M, Leori G, Rubino S, Barbato A, Cappuccinelli P. 1992. Phenotypic features and molecular characterization of plasmids in Salmonella Abortusovis. J. Gen. Microbiol. 138:725–731. 10.1099/00221287-138-4-725. [DOI] [Google Scholar]
  • 23.Grant SGN, Jessee J, Bloom FR, Hanahan D. 1990. Differential plasmid rescue from transgenic mouse DNAs into Escherichia coli methylation-restriction mutants. Proc. Natl. Acad. Sci. 87:4645–4649. 10.1073/pnas.87.12.4645. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Wang RF, Kushner SR. 1991. Construction of versatile low-copynumber vectors for cloning, sequencing and gene expression in Escherichia coli. Gene 100:195–199. 10.1016/0378-1119(91)90366-J. [DOI] [PubMed] [Google Scholar]
  • 25.Westphal O, Jann K. 1965. Bacterial lipopolysaccharide. Extraction with phenol-water and further applications of the procedure. Methods Carbohydr. Chem. 5:83–91. [Google Scholar]
  • 26.El Hamidi A, Tirsoaga A, Novikov A, Hussein A, Caroff M. 2005. Microextraction of bacterial lipid A: easy and rapid method for mass spectrometric characterization. J. Lipid Res. 46:1773–1778. 10.1194/jlr.D500014-JLR200. [DOI] [PubMed] [Google Scholar]
  • 27.Tirsoaga A, Novikov A, Adib-Conquy M, Werts C, Fitting C, Cavaillon JM, Caroff M. 2007. Simple method for repurification of endotoxins for biological use. Appl. Environ. Microbiol. 73:1803–1808. 10.1128/AEM.02452-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Chessa D, Dorsey CW, Winter M, Baümler AJ. 2008. Binding specificity of Salmonella plasmid-encoded fimbriae assessed by glycomics. J. Biol. Chem. 283:8118–8124. 10.1074/jbc.M710095200. [DOI] [PubMed] [Google Scholar]
  • 29.Murray GL, Attridge SR, Morona R. 2003. Regulation of Salmonella Typhimurium lipopolysaccharide O antigen chain length is required for virulence; identification of FepE as a second Wzz. Mol. Microbiol. 47:1395–1406. 10.1046/j.1365-2958.2003.03383.x. [DOI] [PubMed] [Google Scholar]
  • 30.Nevola JJ, Stocker BA, Laux DC, Cohen PS. 1985. Colonization of the mouse intestine by an avirulent Salmonella typhimurium strain and its lipopolysaccharide-defective mutants. Infect. Immun. 50:152–159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Skurnik M, Venho R, Bengoechea JA, Moriyón I. 1999. The lipopolysaccharide outer core of Yersinia enterocolitica serotype O:3 is required for virulence and plays a role in outer membrane integrity. Mol. Microbiol. 31:1443–1462. 10.1046/j.1365-2958.1999.01285.x. [DOI] [PubMed] [Google Scholar]
  • 32.Deligios M, Bacciu D, Deriu E, Corti G, Bordoni R, De Bellis G, Leori GS, Rubino S, Uzzau S. 2014. Genome sequence of the host-restricted Salmonella enterica serovar Abortusovis strain SS44. Genome Announc. 2:e00261–14. 10.1128/genomeA.00261-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Uzzau S, Leori GS, Petruzzi V, Watson PR, Schianchi G, Bacciu D, Mazzarello V, Wallis TS, Rubino S. 2001. Salmonella enterica serovar-host specificity does not correlate with the magnitude of intestinal invasion in sheep. Infect. Immun. 69:3092–3099. 10.1128/IAI.69.5.3092-3099.2001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Uzzau S, Marogna G, Leori GS, Curtiss R, III, Schianchi G, Stocker BA, Rubino S. 2005. Virulence attenuation and live vaccine potential of aroA, crp cdt cya, and plasmid-cured mutants of Salmonella enterica serovar Abortusovis in mice and sheep. Infect. Immun. 73:4302–4308l. 10.1128/IAI.73.7.4302-4308.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Bravo D, Silva C, Carter JA, Hoare A, Alvarez SA, Blondel CJ, Zaldívar M, Valvano MA, Contreras I. 2008. Growth-phase regulation of lipopolysaccharide O-antigen chain length influences serum resistance in serovars of Salmonella. J. Med. Microbiol. 57(Pt 8):938–946. 10.1099/jmm.0.47848-0. [DOI] [PubMed] [Google Scholar]
  • 36.Vogel SN, Awomoyi AA, Rallabhandi P, Medvedev AE. 2005. Mutations in TLR4 signaling that lead to increased susceptibility to infection in humans: an overview. J. Endotoxin Res. 11:333–339. 10.1177/09680519050110060801. [DOI] [PubMed] [Google Scholar]
  • 37.Akira S, Uematsu S, Takeuchi O. 2006. Pathogen recognition and innate immunity. Cell 124:783–801. 10.1016/j.cell.2006.02.015. [DOI] [PubMed] [Google Scholar]
  • 38.Dziarski R. 2006. Deadly plague versus mild-mannered TLR4. Nat. Immunol. 7:1017–1019. 10.1038/ni1006-1017. [DOI] [PubMed] [Google Scholar]
  • 39.Montminy SW, Khan N, McGrath S, Walkowicz MJ, Sharp F, Conlon JE, Fukase K, Kusumoto S, Sweet C, Miyake K, Akira S, Cotter RJ, Goguen JD, Lien E. 2006. Virulence factors of Yersinia pestis are overcome by a strong lipopolysaccharide response. Nat. Immunol. 7:1066–1073. 10.1038/ni1386. [DOI] [PubMed] [Google Scholar]

Articles from Infection and Immunity are provided here courtesy of American Society for Microbiology (ASM)

RESOURCES