Contribution of the stg Fimbrial Operon of Salmonella enterica Serovar Typhi during Interaction with Human Cells

Bacterial strains, plasmids, media, and growth conditions.Bacterial strains and plasmids used in this study are listed in Table 1. Bacteria were routinely grown in Luria-Bertani (LB) broth at 37°C, unless indicated otherwise. When required, antibiotics, amino acids, or supplements were added at the following concentrations: kanamycin, ampicillin, and diaminopimelic acid (DAP), 50 μg/ml; chloramphenicol, 34 μg/ml; and tryptophan, cysteine, and arginine, 22 μg/ml. Transformation of bacterial strains was routinely done by using the calcium/manganese-based or electroporation method as described previously (27).

Cloning of the stg fimbrial operon.The stg operon was amplified from genomic DNA of strain ISP1820 using the Elongase enzyme mixture (Invitrogen) with primer StgA-F (5′CGGGATCCGAGATGAGAATAACGGAATA-3′) containing a BamHI restriction site (underlined) and primer StgD-R (5′GCTCTAGACATTGATATGACTTATTTTG-3′) containing an XbaI restriction site (underlined). The 5-kb PCR product was purified and cloned into vector pCR2.1 using a TOPOXL PCR cloning kit (Invitrogen), resulting in plasmid pSIF018. The XbaI-HindIII fragment was subcloned into low-copy-number vector pWSK29 at the same restriction sites, resulting in plasmid pSIF026. The different constructs were transformed into the nonfimbriate E. coli K-12 Δfim mutant strain ORN172 (42) or into S. enterica serovar Typhimurium and Typhi strains.

Adherence to human epithelial cells.The ability of E. coli strain ORN172 containing the stg operon (pSIF018) or only the vector (pCR2.1) to adhere to human epithelial cells (INT-407) was assessed. A total of 2.5 × 10⁵ cells grown in minimal essential medium (Wysent) supplemented with 10% heat-inactivated fetal calf serum (Wysent) and 25 mM HEPES (Wysent) were seeded in 24-well tissue culture plates 24 h before the adherence assays. One hour before infection, cells were washed three times with prewarmed phosphate-buffered saline (PBS) (pH 7.4), and fresh complete medium was added to each well. Bacteria were grown overnight on LB medium plates and were resuspended in PBS to an optical density at 600 nm (OD₆₀₀) of 1.5 (∼1.5 × 10⁹ CFU/ml). Approximately 2.5 × 10⁷ CFU was added to each well (multiplicity of infection [MOI], 100). The 24-well plates were then centrifuged at 1,000 × g for 5 min to synchronize infection, incubated at 37°C in 5% CO₂ for 90 min, and rinsed three times with PBS. PBS-0.1% deoxycholic acid sodium salt was added to each well, and samples were diluted and spread on LB medium plates for enumeration by viable colony counting. The results were expressed as the percentage of the initial inoculum. Statistical differences were assessed using Student's t test.

A similar protocol was used to test adherence of Salmonella and/or the isogenic stg mutant strains to INT-407 cells, except that bacteria were grown overnight without shaking in LB medium containing 0.3 M NaCl and an MOI of 20 was used. When indicated below, an additional 90-min incubation with 100 μg/ml gentamicin to kill extracellular bacteria was performed in order to assess the invasion level.

Generation of a single-copy stgA-lacZ transcriptional fusion and β-galactosidase assay.The stgA promoter region was amplified using the Elongase enzyme mixture (Invitrogen) and the following primers: StgA-F and StgA-R (5′AACTGCAGCCAGCAAATGCCGTTTTGTT3′). The PCR product was cloned into vector pCR2.1 using a TOPOXL PCR cloning kit (Invitrogen), resulting in plasmid pSIF016. A 530-bp fragment digested with XbaI and SpeI was purified and ligated to pFUSE digested with XbaI (2), resulting in plasmid pSIF020. Plasmid pSIF020 was confirmed to contain the stgA promoter in the correct orientation for lacZ fusion. To generate a single copy of the PstgA-lacZ fusion in serovar Typhi, pSIF020 was transferred by conjugation and integrated into the genome by homologous recombination as described previously (2, 3). A strain carrying a single integrated copy of PstgA-lacZ in ISP1820 was designated DEF068. The expression of stg was evaluated by β-galactosidase assays of the reporter strain DEF068 grown in different conditions. β-Galactosidase activity was measured using o-nitrophenyl-β-d-galactopyranoside as described previously (23).

Construction of a serovar Typhi strain with an stg deletion.A suicide vector for deletion of the stg fimbrial operon (STY3918 to STY3922) was constructed as follows. A 530-bp fragment of the 5′ end of stgA was generated by PCR using primers StgA-F and StgA-R, and a 482-bp fragment of the 3′ end of stgD was generated by PCR using primers StgD-F (5′AACTGCAGGCCGCAGAGCTGTGAAAATG3′) and StgD-R. These two fragments were ligated and cloned into the XbaI and BamHI sites of pMEG-375 (15). A resulting suicide vector containing the stgA′-stgD′ fragment (pSIF004) was used for allelic replacement of the stg region. The pSIF004 suicide vector was conjugated from E. coli MGN-617 to serovar Typhi strain ISP1820 by overnight plate mating on LB medium with DAP. Transconjugants were selected by growth on LB medium plates containing chloramphenicol without DAP. Selection for double-crossover allele replacement was performed by sacB counterselection on LB agar plates without NaCl containing 5% sucrose (16). Isogenic strain DEF004 has a deletion of the stg region resulting from a double crossover, as determined by the absence of resistance to ampicillin and chloramphenicol encoded on the suicide vector, and the expected stg deletion, as confirmed by PCR (data not shown).

Bacterial survival in human macrophages.The human monocyte cell lines THP-1 (= ATCC TIB-202) and U937 (= ATCC CRL 1593) were maintained in RPMI 1640 (Invitrogen) containing 10% fetal calf serum, 25 mM HEPES, 2 mM l-glutamine, 1% minimal essential medium nonessential amino acids (Wisent), and 1 mM sodium pyruvate (Sigma). Stock cultures of these cells were maintained as monocyte-like, nonadherent cells at 37°C in an atmosphere containing 5% CO₂. Before infection, cells were differentiated by addition of 10⁻⁷ M phorbol 12-myristate 13-acetate (Sigma) for 24 to 72 h. For macrophage infection assays, cells were seeded at a concentration of 5 × 10⁵ cells per well in 24-well tissue culture dishes. Bacteria grown overnight at 37°C in static conditions were added to a cell monolayer at an MOI of 10 and centrifuged for 5 min at 1,000 × g to synchronize phagocytosis. After incubation for 20 min at 37°C (zero time), the infected cells were washed three times with prewarmed PBS and incubated with supplemented medium as described above containing 100 μg/ml of gentamicin to kill extracellular bacteria. The infected monolayers were either lysed from the tissue culture dishes by addition of 0.1% deoxycholic acid sodium salt in PBS or incubated further. After lysis the number of surviving bacteria was determined by bacterial plate counting (CFU). The level of phagocytosis was expressed as a percentage of the initial inoculum. The survival rate was expressed as a percentage determined by comparing the number of intracellular bacteria with the number at the previous time.

Statistical differences were assessed using Student's t test. Where indicated, the macrophages were incubated 1 h prior to infection with 1 μg/ml of cytochalasin D (Sigma) to inhibit bacterial uptake as described previously (32). The level of cytochalasin D was maintained throughout the infection.

stg fimbrial operon.The stg fimbrial cluster has a G+C content of 49% and is a member of a distinct group of related fimbrial genes that are located in the glmS-pstS intergenic region (21, 39). In the sequenced genomes of S. enterica (including unfinished genomes) this fimbrial gene cluster has been identified only in serovar Typhi. Moreover, stg sequences were not detected by comparative genomic hybridization in the genomes of 140 strains belonging to many serovars of subspecies I (30; M. McClelland, personal communication). The previously described distribution of stg determined by Southern blotting may therefore represent cross-hybridization with other less homologous fimbrial genes (39). However, a putative fimbrial gene inserted in the glmS-pstS region in S. bongori belongs to the Stg group, and its product exhibits the highest level of identity to the predicted stg fimbrial gene products of serovar Typhi (Table 2). The genes encoding a number of fimbrial systems in pathogenic E. coli are also inserted in the glmS-pstS region and belong to the Stg group; these systems include the Stg (21), Lpf_O113 (5), and Lpf2 (O-island 154) (38) systems. In addition, Lpf and related fimbriae encoded in the yhjX-yhjW region in Salmonella and E. coli (36, 37) exhibit some identity to the predicted stg gene products of serovar Typhi, but less identity than other fimbriae belonging to the Stg group (Table 2). The serovar Typhi stg fimbrial cluster contains five ORFs designated stgABCC′D as stgC is a predicted pseudogene and contains a premature stop codon. The stgC ORF may code for a 170-amino-acid (aa) protein, and a second ORF designated stgC′ may code for a 605-aa protein. The stgC stop codon is present in the stgC sequence of serovar Typhi strain ISP1820 (data not shown), as well as in the sequenced genomes of serovar Typhi strains TY2 and CT18 (4, 29).

Adhesion of E. coli containing the stg operon.To examine the capacity of the stg fimbrial cluster to mediate adherence to INT-407 cells, the stg operon was cloned in different vectors and transformed into E. coli strain ORN172. ORN172 is an E. coli K-12 noninvasive strain with a deletion in the fim operon that does not express type 1 fimbriae and is commonly used to study adherence conferred by recombinant fimbrial systems (42). E. coli ORN172 cells containing the vector alone (pCR2.1) adhered poorly between the cells or without pattern on the cell surface and were often isolated (Fig. 1A). However, ORN172 cells containing stg (pSIF018) adhered in aggregates or clusters on the cell surface (Fig. 1B). The introduction of stg into E. coli conferred a significantly higher level of adhesion to epithelial cells, which was threefold higher than that of the strain harboring the vector alone (Fig. 1C). A higher level of adhesion was also observed when a low-copy-number vector (pSIF026) was used (data not shown).

• Open in new tab
• Download powerpoint

FIG. 1.

Adherence and expression of the stg fimbrial operon by E. coli and S. enterica serovar Typhimurium. (A and B) Adherence of E. coli strain ORN172 to human epithelial cells (INT-407) containing the vector (pCR2.1) (DEF045) (A) or the stg genes (pSIF018) (DEF049) (B). Slides were stained with 5% Giemsa stain. Bacteria are indicated by arrows. (C) Percentage of the initial inoculum associated with epithelial cells after 90 min of incubation for E. coli and serovar Typhimurium carrying the stg operon (DEF047) or the control vector (DEF048). All assays were conducted in duplicate and repeated independently at least three times. The results are expressed as the means ± standard errors of the replicate experiments. An asterisk indicates that there is a significant difference between the strain containing the control vector and the strain containing the stg operon (P < 0.005).

Adhesion of serovar Typhimurium containing the stg operon.As the stg fimbrial operon is absent in serovar Typhimurium, we used this serovar to establish whether stg could contribute to adherence to INT-407 cells by a heterogeneous Salmonella serovar. Serovar Typhimurium strain χ3339 harboring stg (pSIF018) exhibited a significantly higher level of adhesion to INT-407 cells, which was 30-fold higher than that of the strain harboring the vector alone (pCR2.1) (Fig. 1C). As salmonellae are able to invade epithelial cells, the level of invasion was also determined by a gentamicin protection assay. An invasion level similar to that exhibited by the wild-type parent harboring only the vector was observed (data not shown).

stg expression in serovar Typhi.To study the expression of the stg fimbrial operon in the native serovar Typhi strain, an stgA::lacZ fusion was inserted into the chromosome of strain ISP1820, generating strain DEF068. Strain DEF068 was used to determine the influence of a number of in vitro growth conditions on stg expression. The expression of the promoter fusion was determined for bacteria grown in LB medium from early log phase to stationary phase. β-Galactosidase expression increased from early to stationary phase, following overnight growth in LB medium (Fig. 2). The β-galactosidase expression following growth on LB agar was nearly twofold higher (54 U) than the expression following overnight growth in LB broth (29 U) (Fig. 2). The highest levels of β-galactosidase expression were observed following overnight growth in minimal medium (M9-glucose) (76 U) (Fig. 2). Expression in conditions that mimic those encountered during invasion and infection of host cells was also studied. The effect of the sodium chloride concentration in the medium was evaluated, as this concentration represents a condition that can influence cell invasion by Salmonella (1, 8). The effect of iron availability and pH on stg expression was also evaluated. Changes in these conditions did not result in any significant changes in β-galactosidase expression (data not shown).

• Open in new tab
• Download powerpoint

FIG. 2.

stg expression in serovar Typhi: β-galactosidase activity expressed from the PstgA::lacZ fusion in serovar Typhi (DEF068) in different growth conditions. Bacteria were grown in LB medium with agitation to early log phase (OD₆₀₀, 0.3), mid-log phase (OD₆₀₀, 0.6), late log phase (OD₆₀₀, 0.9), and stationary phase (overnight), on LB agar, and in M9-glucose broth (M9 min broth) (overnight). The error bars indicate standard deviations.

stg contributes to adherence of serovar Typhi to epithelial cells.We assessed whether stg contributes to adherence of serovar Typhi to INT-407 cells by constructing an isogenic ΔstgABCC′D mutant by allelic exchange. The mutated strain, DEF004, exhibited a significantly lower level of adherence (80% of the wild-type strain adherence) (Fig. 3A). A level of adherence significantly higher than that of the wild-type strain was observed when the stg mutant was complemented with the stg genes on a low-copy-number vector (pSIF026) (Fig. 3A). In spite of the lower level of adherence of the mutant, its level of invasion was higher than that of the wild-type parent, but not significantly higher (Fig. 3A).

• Open in new tab
• Download powerpoint

FIG. 3.

Role of stg in the interaction of serovar Typhi with human cells: capacity of the wild-type strain, the stg mutant (DEF004), and the complemented strain (DEF066) to adhere to and invade INT-407 cells (A) or to survive within THP-1 macrophage-like cells (B). All assays were conducted in duplicate and repeated independently at least three times. The results are expressed as the means ± standard errors of the replicate experiments. Significant differences (P < 0.005) in adherence or phagocytosis between the mutant and the wild-type strain of serovar Typhi are indicated by asterisks. The values for percent recovery were normalized to the wild-type control value, which was defined as 100% at each time point.

Loss of stg results in increased phagocytosis of serovar Typhi by macrophages.As survival in macrophages plays an essential role in systemic infection by Salmonella, we characterized the interaction of the isogenic stg mutant with human macrophages. The wild-type strain and the mutant were used to infect human macrophage-like cells, and the numbers of bacteria present after phagocytosis at 2 and 24 h postinfection were determined. The mutant showed a significantly higher level of phagocytosis than the wild-type strain (Fig. 3B). The levels of bacterial survival at 2 or 24 h postinfection were similar for both the stg mutant and the wild-type strain (Fig. 3B). Complementation of the stg mutant with stg on a low-copy-number vector (pSIF026) restored the wild-type phagocytosis phenotype (Fig. 3B).

Role of stg in macrophage interactions.As bacterial uptake of the stg mutant by macrophages was altered, we wanted to evaluate the effect of stg overexpression on phagocytosis. The uptake of both serovar Typhi strain ISP1820 and serovar Typhimurium strain χ3339 harboring stg (pSIF018) on a multicopy vector was significantly lower than the uptake of the bacterial strain harboring the vector alone (pCR2.1) (Fig. 4). This lower level of phagocytosis was also observed using macrophage-like U937 cells (data not shown). Then, in order to differentiate between the initial levels of bacteria associated with or internalized by macrophages, we used an inhibitor of cytoskeletal function, cytochalasin D, to block bacterial uptake. In the presence of cytochalasin D, less then 2% of the initial inoculum was associated with macrophages. The percentages of serovar Typhi that were associated with macrophages were similar when stg was present at a high copy number and when the wild-type harboring the vector alone was used (Fig. 5). In addition, the stg mutant also showed a level of association with macrophages similar to that of the wild-type strain when bacterial uptake was inhibited by cytochalasin D (Fig. 5). Since the levels of association with macrophages were similar in cytochalasin D-treated cells regardless of the presence of stg, these results indicate that the stg fimbrial system contributes to a reduction in internalization of serovar Typhi by macrophages.

• Open in new tab
• Download powerpoint

FIG. 4.

Effect of overexpression of stg on phagocytosis. Serovar Typhimurium carrying the stg cluster (DEF047) or the control vector (DEF048) and serovar Typhi carrying the stg cluster (DEF033) or the control vector (DEF064) were incubated with THP-1 macrophage-like cells. The percentage of the initial inoculum associated with cells after 120 min of incubation is indicated. All assays were conducted in duplicate and repeated independently at least three times. The results are expressed as the means ± standard errors of the replicate experiments. An asterisk indicates that there is a significant difference in phagocytosis between the wild-type strain containing the vector alone and the strain with the stg operon (P < 0.05).

• Open in new tab
• Download powerpoint

FIG. 5.

Role of stg fimbrial operon in bacterial association with macrophages. Bacterial uptake was inhibited with cytochalasin D, and the numbers of bacteria with stg (DEF033) and without stg (DEF004) associated with macrophages were compared. All assays were conducted in duplicate and repeated independently at least three times. The values for percent recovery were normalized to the wild-type control value, which was defined as 100% at each time point. The results are expressed as the means ± standard errors of the replicate experiments.