The two murein lipoproteins of Salmonella enterica serovar Typhimurium contribute to the virulence of the organism

Abstract

Septic shock due to Salmonella and other gram-negative enteric pathogens is a leading cause of death worldwide. The role of lipopolysaccharide in sepsis is well studied; however, the contribution of other bacterial outer membrane components, such as Braun (murein) lipoprotein (Lpp), is not well defined. The genome of Salmonella enterica serovar Typhimurium harbors two copies of the lipoprotein (lpp) gene. We constructed a serovar Typhimurium strain with deletions in both copies of the lpp gene (lpp1 and lpp2) by marker exchange mutagenesis. The integrity of the cell membrane and the secretion of the effector proteins through the type III secretion system were not affected in the lpp double-knockout mutant. Subsequently, the virulence potential of this mutant was examined in a cell culture system using T84 intestinal epithelial and RAW264.7 macrophage cell lines and a mouse model of salmonellosis. The lpp double-knockout mutant was defective in invading and inducing cytotoxic effects in T84 and RAW264.7 cells, although binding of the mutant to the host cell was not affected when compared to the wild-type (WT) serovar Typhimurium. The motility of the mutant was impaired, despite the finding that the number of flagella was similar in the lpp double knockout mutant and the WT serovar Typhimurium. Deletion in the lpp genes did not affect the intracellular survival and replication of Salmonella in macrophages and T84 cells. Induction of the proinflammatory cytokines tumor necrosis factor alpha and interleukin-8 (IL-8) was significantly reduced in macrophages and T84 cells infected with the lpp double-knockout mutant. The levels of IL-8 remained unaffected in T84 cells when infected with either live or heat-killed WT and lpp mutant, indicating that invasion was not required for IL-8 production and that Toll-like receptor 2 signaling might be affected in the Lpp double-knockout mutant. These effects of the Lpp protein could be restored by complementation of the isogenic mutant. The lpp double-knockout mutant was avirulent in mice, and animals infected with this mutant were protected from a lethal challenge dose of WT serovar Typhimurium. The severe combined immunodeficient mice, on the other hand, were susceptible to infection by the lpp double-knockout mutant. The serovar Typhimurium mutants from which only one of the lpp (lpp1 or lpp2) genes was deleted were also avirulent in mice. Taken together, our data indicated that Lpp specifically contributed to the virulence of the organism.

FIG. 1.

Schematic diagram showing construction of lpp isogenic mutants of serovar Typhimurium via homologous recombination. In panel A, the Kn^r gene cassette-interrupted lpp1 gene replaced the original lpp1 gene on the genome of serovar Typhimurium and generated the lpp1 mutant, while for the lpp2 single-knockout mutant, the Kn^r gene cassette replaced the whole coding sequence of the lpp2 gene. To generate the lpp double-knockout mutant, both copies of the lpp gene including the sequence between them were replaced by a Kn^r gene cassette. In panel B, the λ Red system was used, and the Kn^r gene cassette replaced the corresponding lpp copy on the genome of the single lpp knockout mutants of serovar Typhimurium. Arrows indicate the primers' position used for the generation and identification of the lpp knockout mutants. The exact DNA sequences of the primers were also shown in Table 2. The open bars indicate the copies of the lpp gene on the genome, and the solid bars indicate the Kn^r gene cassette used to replace or interrupt the lpp genes on the genome of serovar Typhimurium. The solid circles represent the flank DNA sequence to the lpp genes of serovar Typhimurium for homologous recombination in the λ Red system.

FIG. 2.

Nucleotide and amino acid sequence homologies between lipoproteins of E. coli and serovar Typhimurium 14028. lpp, lipoprotein gene; Lpp, lipoprotein. Both E. coli lpp and S. enterica serovar Typhimurium lpp1 genes contained 237 nucleotides, whereas the S. enterica serovar Typhimurium lpp2 gene contained 240 nucleotides.

FIG. 3.

(A) Western blot analysis on the lpp double-knockout mutants of serovar Typhimurium. The outer membranes of serovar Typhimurium WT, lpp double-knockout mutants (mutants 66 and 67), and E. coli JE5505 (Lpp⁻) were isolated and separated by SDS-15% PAGE. The protein bands were then transferred to a nitrocellulose membrane and probed with Y. enterocolitica Lpp monoclonal antibody (1:1,000 dilution, determined empirically). The secondary antibodies (1:25,000 dilution) were goat antimouse conjugated with HRP. The blots were developed using an enhanced chemiluminescence kit. A band corresponding to the size of approximately 50 kDa in E. coli JE5505 appeared to be nonspecific and reacted with the antibodies. (B) Western blot analysis on the complemented lpp double-knockout mutants of serovar Typhimurium. The whole-cell lysates of serovar Typhimurium, lpp double-knockout mutants (66 and 67), the complemented lpp double-knockout mutants (66-C and 67-C), E. coli JE5505 (Lpp⁻), and E. coli DH5α (Lpp positive) were isolated and separated by SDS-15% PAGE. The protein bands were then transferred to a nitrocellulose membrane and probed with Y. enterocolitica Lpp monoclonal antibody (1:1,000 dilution). The secondary antibodies (1:25,000 dilution) were goat antimouse conjugated with HRP. The blots were developed by using an enhanced chemiluminescence kit. The arrow indicates the correct size of Lpp.

FIG. 4.

Invasion of T84 cells by the lpp double-knockout mutant (lpp⁻) (mutant 67), the WT, and the complemented (compl) strain of the lpp double-knockout mutant. The T84 cells were infected with Salmonella at an MOI of 10:1 for 1 h. The monolayers were washed and incubated with medium containing gentamicin for 1 h. After incubation, cells were washed, lysed with 0.1% TX-100, and plated on SS agar plates. The asterisk denotes statistical significance at P values of ≤0.05 by Student's t test, between the WT and mutant 67 (lpp⁻) and between 67 (lpp⁻) and 67-C [lpp⁻compl]. The values for invasion between the WT and 67-C were not statistically significant. Arithmetic means ± standard deviations were plotted.

FIG. 5.

Cytotoxicity and cell death induced by WT serovar Typhimurium and its various mutants in T84 (A) and RAW264.7 (B) cells. Cells were infected for 1 h with WT serovar Typhimurium, the lpp double-knockout mutant (67), or its complemented strain. Cells were washed and incubated with gentamicin-containing medium for 1 h. After incubation, cells were washed and further incubated with the antibiotic-free fresh medium for 24 h. Cells were gently washed twice in PBS and examined under a confocal microscope. Frame 2, cells infected with WT serovar Typhimurium (note dead cells); frame 3, cells infected with lpp double-knockout mutant (67) (note normal morphology of the cells); frame 4, cells infected with complemented strain (note dead cells); frame 1 represents a noninfected control.

FIG. 6.

Intracellular replication of the lpp isogenic mutant inside RAW264.7 cells (A) and T84 cells (B). Cells were infected with the lpp double-knockout mutant (67), the complemented lpp double-knockout mutant (67-C), and WT serovar Typhimurium 14028 for 1 h. Cells were washed and incubated for 1 h with gentamicin-containing (100 μg/ml) medium. After incubation, cells were washed and incubated in fresh medium containing a minimum concentration of gentamicin (5 μg/ml) for different time points (1, 6, 12, and 24 h). Finally, cells were lysed with 0.1% TX-100 and plated on SS agar plates to determine numbers of CFU. Three independent experiments were performed. Data from a representative experiment are shown here.

FIG. 7.

TNF-α induced by lpp double-knockout mutant in RAW264.7 cells (A) and IL-8 production by T84 cells infected with live and heat-killed bacteria (B). Macrophages and T84 cells were stimulated with the heat-killed lpp double-knockout mutant or WT serovar Typhimurium at an MOI of 0.1:1 and incubated for 8 h. T84 cells were infected with live WT, the lpp isogenic mutant (67) (lpp⁻), or the complemented lpp mutant [lpp⁻ (compl)] at an MOI of 10:1 and incubated for 1 h. After incubation, cells were washed and incubated in gentamicin-containing medium for 1 h, after which cells were washed and incubated in antibiotic-free medium for 12 h. TNF-α and IL-8 levels were determined using ELISA as described in Materials and Methods. An asterisk denotes statistically significant data (P ≤ 0.05) by Student's t test, between the WT and the lpp double-knockout mutant and between the lpp double-knockout mutant and its complemented strain. The values were not significant between WT and the lpp mutant complemented strain. Negative denotes no addition of the bacterial cells. The arithmetic means ± standard deviations were plotted. Compl denotes complemented strain.

FIG. 8.

(A) Mortality in mice following infection with the lpp double-knockout mutant (67) and WT serovar Typhimurium. Four groups of mice were used: two groups were infected orally with 3 × 10³ CFU of the lpp mutant (67) or WT serovar Typhimurium. The other two groups were infected with 40 CFU of the lpp mutant (67) or WT serovar Typhimurium i.p. The group infected with WT serovar Typhimurium orally or i.p. showed 100% mortality compared to 100% survival in the groups that were infected with the mutant serovar Typhimurium. (B) Mortality in mice challenged with a virulent strain of serovar Typhimurium following immunization with the lpp double-knockout mutant of serovar Typhimurium (67). Out of four groups of mice used for the challenge study, two groups were immunized with mutant strain 67, and one group was orally inoculated (3 × 10³ CFU) while the second group received i.p. (40 CFU) inoculation. The other two groups were left unimmunized as controls. After 2 months postimmunization, all groups of mice (immunized and nonimmunized) were challenged with a lethal dose of a virulent strain of serovar Typhimurium, either orally with 3 × 10³ CFU or by the i.p. route with 3 × 10³ CFU. Then, mice were observed for 4 weeks, and mortality was recorded.