FNR is a global regulator of virulence and anaerobic metabolism in Salmonella enterica serovar Typhimurium (ATCC 14028s)

Abstract

Salmonella enterica serovar Typhimurium must successfully transition the broad fluctuations in oxygen concentrations encountered in the host. In Escherichia coli, FNR is one of the main regulatory proteins involved in O2 sensing. To assess the role of FNR in serovar Typhimurium, we constructed an isogenic fnr mutant in the virulent wild-type strain (ATCC 14028s) and compared their transcriptional profiles and pathogenicities in mice. Here, we report that, under anaerobic conditions, 311 genes (6.80% of the genome) are regulated directly or indirectly by FNR; of these, 87 genes (28%) are poorly characterized. Regulation by FNR in serovar Typhimurium is similar to, but distinct from, that in E. coli. Thus, genes/operons involved in aerobic metabolism, NO. detoxification, flagellar biosynthesis, motility, chemotaxis, and anaerobic carbon utilization are regulated by FNR in a fashion similar to that in E. coli. However, genes/operons existing in E. coli but regulated by FNR only in serovar Typhimurium include those coding for ethanolamine utilization, a universal stress protein, a ferritin-like protein, and a phosphotransacetylase. Interestingly, Salmonella-specific genes/operons regulated by FNR include numerous virulence genes within Salmonella pathogenicity island 1 (SPI-1), newly identified flagellar genes (mcpAC, cheV), and the virulence operon (srfABC). Furthermore, the role of FNR as a positive regulator of motility, flagellar biosynthesis, and pathogenesis was confirmed by showing that the mutant is nonmotile, lacks flagella, is attenuated in mice, and does not survive inside macrophages. The inability of the mutant to survive inside macrophages is likely due to its sensitivity to the reactive oxygen species generated by NADPH phagocyte oxidase.

FIG. 1.

Location of the tnpA insertion (between bp 106 and 107) in the fnr gene. WT fnr sequences are in bold, and the sequences of the beginning and ending junctions of the tnpA insert are in italics. Arrows indicate the direction of transcription. IGS, intergenic spacer region. (Complete DNA sequences [i.e., ogt, tnpA/fnr junctions, and ydaA] are available at GenBank accession number AH015911.)

FIG. 2.

Logo graph of the information matrix obtained from the consensus alignment of FNR motif sequences for serovar Typhimurium (derived from the corresponding FNR-regulated genes in E. coli). The total height of each column of characters represents the amount of information for that specific position, and the height of each character represents the frequency of each nucleotide.

FIG. 3.

Correlation between the microarray and qRT-PCR data for 19 selected genes. The ratios of changes in gene expression, from the microarray and qRT-PCR experiments, for the FNR mutant relative to the WT were log₂ transformed and linearly correlated. The genes selected and the primers used in qRT-PCR are listed in Table 2.

FIG. 4.

Scheme representing the structural organization of the major genes involved in virulence/SPI-1 (A), ethanolamine utilization (B), and flagellar biosynthesis and motility/swarming (C to E). The names of genes are listed to the right of the arrows, an asterisk next to the gene indicates the presence of at least one FNR motif in the 5′ region, and the numbers to the left of the arrows indicate the ratio of gene expression in the fnr mutant relative to that in the WT.

FIG. 5.

Comparison of the WT, the fnr mutant, and the mutant strain harboring pfnr for motility (left) and comparison of the WT and the mutant for the presence of surface appendages by SEM (center) and for the presence of flagella by negative staining and TEM (right). Cells were grown anaerobically in LB-MOPS-X media, and samples were prepared as described in Materials and Methods.

FIG. 6.

Comparison of the fnr mutant and the WT strain for virulence in 6- to 8-week-old C57BL/6 mice. (A) Groups of 10 mice were inoculated p.o. with 5 × 10⁶ and 5 × 10⁷ CFU/mouse. (B) Groups of five mice were challenged i.p. with 250 CFU/mouse, as described in Materials and Methods. Percent survival is the number of mice surviving relative to the number of mice challenged at time zero.

FIG. 7.

Comparison of the WT, the fnr mutant, and the mutant strain harboring pfnr for survival inside peritoneal macrophages from C57BL/6 mice. The macrophages were harvested and treated as described in Materials and Methods. (A) Comparison between the fnr mutant and the WT strain. The number of viable cells found inside the macrophages, at time zero, following the removal of extracellular bacteria by washing/gentamicin treatment is defined as 100% survival. (B) Comparison between the WT, the fnr mutant, and the pfnr-complemented mutant. The number of viable cells found inside macrophages at 20 h is expressed as percent survival relative to that found inside macrophages at 2 h.

FIG. 8.

Virulence of the WT and the fnr mutant in C57BL/6 mice and congenic gp91^phox−/− mice and survival of the bacteria inside peritoneal macrophages. The mice were challenged i.p. with 250 CFU/mouse, as described in Materials and Methods. (A) C57BL/6 and gp91^phox−/− mice treated with the WT strain. (B) C57BL/6 and gp91^phox−/− mice treated with the fnr mutant. (C) Survival of the WT and the fnr mutant inside macrophages from C57BL/6 and gp91^phox−/− mice. The number of viable cells at 20 h is expressed as percent survival relative to that found inside the macrophages at time zero.