Salmonella re-engineers the intestinal environment to break colonization resistance in the presence of a compositionally intact microbiota

Abstract

The gut microbiota prevents harmful microbes from entering the body, a function known as colonization resistance. The enteric pathogen Salmonella enterica serovar (S.) Typhimurium uses its virulence factors to break colonization resistance through unknown mechanisms. Using metabolite profiling and genetic analysis, we show that the initial rise in luminal pathogen abundance was powered by a combination of aerobic respiration and mixed acid fermentation of simple sugars, such as glucose, which resulted in their depletion from the metabolome. The initial rise in the abundance of the pathogen in the feces coincided with a reduction in the cecal concentrations of acetate and butyrate and an increase in epithelial oxygenation. Notably, these changes in the host environment preceded changes in the microbiota composition. We conclude that changes in the host environment can weaken colonization resistance even in the absence of overt compositional changes in the gut microbiota.

Keywords: Salmonella; colonization resistance; microbiota; mixed acid fermentation; short-chain fatty acids.

Figure 1:. An initial rise in the absolute S. Typhimurium abundance is driven by mixed acid fermentation in mice with an intact microbiota.

(A) Groups (n = 8) of antibiotic-naïve, genetically resistant (CBA/J) mice were infected with 10⁹ colony-forming units (CFU) of the S. Typhimurium wild type (WT) or an avirulent invA spiB mutant and the recovery of CFU from feces was recorded over time. To compare these data with findings from the mouse colitis model, CBA/J mice were pretreated with a single dose of streptomycin (20 mg/animal) and infected one day later with 10⁹ CFU of the S. Typhimurium wild type. Solid lines indicate the geometric mean. Dotted lines indicate the geometric standard deviation. (B and E-G) CBA/J mice were infected with a 1:1 mixture of the indicated strains. The graphs show CFU of the S. Typhimurium WT and the respective mutant recovered from feces at the indicated days after infection (Days p.i.). WT and mutant CFU recovered from the same animal are connected by dotted lines. The competitive index (CI) determined for each time point is the ratio of wild type to mutant bacteria. (B) Mice were infected with WT and a napA narZ narG cyxA mutant. (C and D) CBA/J mice were mock infected or infected with 10⁹ CFU of the S. Typhimurium WT. Cecal contents were collected three days after infection (Day 3 p.i.) or four days after infection (Day 4 p.i.) for untargeted metabolomics analysis. (C) Principal component analysis of the cecal metabolome in the indicated groups of mice. (D) Volcano plot of metabolites with carbohydrates colored in yellow. (E) Mice were infected with the S. Typhimurium WT and a frdABCD mutant. (F) Mice were infected with the S. Typhimurium WT and an ackA pta mutant. (G) Mice were infected with the S. Typhimurium WT and an adhE mutant. (B, C and E-G) Each dot represents data from one animal (n). *, P < 0.05; **, P < 0.01; *****, P < 0.001; ns, P > 0.05.

Figure 2:. S. Typhimurium depletes cecal glucose by catabolizing the sugar to increase absolute pathogen abundance in the feces.

(A) Antibiotic-naïve, genetically resistant (CBA/J) mice were infected with a 1:1 mixture of the S. Typhimurium wild type (WT) and a mutant unable to catabolize glucose (ptsG glk manXYZ mutant). The graph shows colony-forming units (CFU) of each strain recovered from feces at the indicated days after infection (Days p.i.). WT and ptsG glk manXYZ mutant CFU recovered from the same animal are connected by dotted lines. The competitive index (CI) determined for each time point is the ratio of wild type to mutant bacteria. (B-D) CBA/J mice were mock infected or infected with the S. Typhimurium WT and ileal epithelial cells were collected three days later to isolate mRNA. Fold changes in transcript levels of Lcn2 (B), Sglt1 (C), and Glut2 (D) were determined by quantitative real-time PCR. (E and F) CBA/J mice were mock infected (n = 16) or infected with the S. Typhimurium WT (n = 15) and received a bolus of glucose three days later. (E) Blood glucose levels were measured after the indicated time points after receiving the bolus of glucose (glucose tolerance test). Data is from two independent experiments. (F) The graph shows the area under the curve of the glucose tolerance test. (G) Groups (n =8) of CBA/J mice were infected with an avirulent S. Typhimurium strain (invA spiB mutant). Two days after infection, mice were mock-treated or treated with sotagliflozin, followed 30 minutes later by feeding a bolus of glucose. The graph shows recovery of the S. Typhimurium invA spiB mutant from the feces at the indicated time points after infection. (H and I) CBA/J mice were mock infected or infected with the indicated S. Typhimurium strains. Data is from two independent experiments. (H) Concentrations of glucose in cecal contents was measured three days after infection. (I) CFU in feces were determined three days after infection (B-D, F, H and I) The graphs show geometric means ± geometric standard deviation. (E and G) Solid lines indicate the geometric mean. Dotted lines indicate the geometric standard deviation. Each symbol represents data from one animal. (A-D, F, H and I) Each dot represents data from one animal (n). *, P < 0.05; **, P < 0.01; *****, P < 0.001; ****, P < 0.0001; ns, P > 0.05.

Figure 3:. Intestinal inflammation elicited during S. Typhimurium infection.

Antibiotic-naïve, genetically resistant (CBA/J) mice were mock infected or infected with the S. Typhimurium wild type (WT) and the cecum was collected at the indicated time points to assess the severity of intestinal inflammation. Each dot (A-D) or bar (E) represents data from one animal (n). (A-D) RNA was extracted from cecal tissue and transcript levels of genetic markers of inflammation, including Il17a (A), Mip2 (B), Kc (C), and Lcn2 (D), were determined by quantitative real time PCR. Transcript levels are expressed as fold-changes compared to transcript levels detected in mock-infected animals. The graphs show geometric means ± geometric standard deviation. (E and F) Hematoxylin and eosin-stained histological sections from the cecum were blinded and analyzed by a veterinary pathologist. (E) Histopathology score for each animal (bars). (F) Representative images are shown. *, P < 0.05; **, P < 0.01; ***, P < 0.005; ns, P > 0.05.

Figure 4:. Kinetics of compositional changes in the fecal microbiota during S. Typhimurium infection.

Groups of antibiotic-naïve, genetically resistant (CBA/J) mice were infected with the S. Typhimurium wild type and fecal samples were collected for DNA extraction at the indicated time points. (day 0: n = 12; day 3: n = 10; day 5: n = 9; day 7: n = 8; day 10: n = 8) (A) PCoA plot showing variation of the mouse fecal microbiota composition by treatment group. Each dot represents data from one animal. (B) Comparison of distances between samples collected at the indicated time points after infection using weighted unique fraction metric (UniFrac). (C) Relative abundance of Clostridia ASVs detected from feces collected at the indicated time points after infection. (D) Identification of predictors of variation between samples on the class level using random forest analysis. (E) Bar plot displaying the number of significantly increased and decreased ASVs by class in feces collected three days after infection (top panel) or 10 days after infection (bottom panel) compared to feces collected prior to infection (FDR corrected P value =< 0.05). (F) Relative abundance of ASVs belonging to the indicated Clostridia families in feces collected at the indicated time points after infection. (B, C, and F) The box plots represent the first to third quartiles, and the line indicates the median value. *, P < 0.05; **, P < 0.01; ns, P > 0.05.

Figure 5:. Reactive nitrogen species reduce concentrations of microbiota-derived acetate and butyrate three days after infection with S. Typhimurium.

(A, B, D-F) Antibiotic-naïve, genetically resistant (CBA/J) mice were mock infected (A and E), infected with the S. Typhimurium wild type (WT) (A, B, D-F), or infected with a S. Typhimurium fadD ydiQRSTD mutant (B) and cecal contents were collected at the indicated time points after infection. (A, B, and D) The graph shows the geometric mean concentration (bars) of the indicated short-chain fatty acids ± geometric standard deviation. (C) Growth of S. Typhimurium in fecal slurries containing short-chain fatty acid concentrations resembling conditions encountered prior to infection or 3 days after S. Typhimurium infection was monitored by measuring the optical density at 600 nm (OD₆₀₀) over time. Solid lines indicate the geometric mean of n = 4 repeats. Dotted lines indicate the geometric standard deviation. (D and F) Mice received regular drinking water (treatment: mock), or drinking water supplemented with 80 μg/mL apocynin (Apo) or 1 mg/mL aminoguanidine (Amino). (E) RNA was extracted from cecal tissue and transcript levels of Nos2 was determined by quantitative real time PCR. Transcript levels are expressed as fold-changes compared to transcript levels detected in mock-infected animals. The graphs show geometric means ± geometric standard deviation. (F) The graph shows the geometric mean of colony-forming units (CFU) recovered from feces of mice at the indicated time points after infection ± geometric standard deviation. (G) The relationship between S. Typhimurium wild type CFUs and the indicated cecal short-chain fatty acid levels at day 3 after infection were assessed by analyzing data from five independent experiments. (A, B, and D-G) Each symbol represents data from one animal and indicates n. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, P > 0.05.

Figure 6:. Loss of epithelial hypoxia drives cydAB mediated growth at 3 days after S. Typhimurium infection.

(A - C) CBA/J mice were infected with a 1:1 mixture of the indicated strains. The graphs show CFU of the S. Typhimurium WT and the respective mutant recovered from feces at the indicated days after infection (days p.i.). WT and mutant CFU recovered from the same animal are connected by dotted lines. The competitive index (CI) determined for each time point is the ratio of wild type to mutant bacteria. (D and E) CBA/J mice were mock infected or infected with the S. Typhimurium wild type (WT) and the cecum was collected three days later. One hour before euthanasia, mice were injected with pimonidazole HCl. (D) Binding of pimonidazole was detected using hypoxyprobe-1 primary antibody and a Cy-3 conjugated goat anti-mouse secondary antibody (red fluorescence) in histological sections from the cecum that were counterstained with DAPI nuclear stain (blue fluorescence). Representative images are shown with 100 μm scale bars. (E) Pimonidazole staining was quantified by measuring mean pixel intensities from the lumen to the border of the colonocytes, and into the tissue. The graph shows the peak pixel intensity for each mouse (symbols) and the mean peak intensity for each group (bars) ± standard deviation. (F) CBA/J mice were mock infected or infected with the S. Typhimurium WT and cecal epithelial cells were collected three days later to isolate mRNA. Fold changes in transcript levels of the Angptl4, Pgc1a, Ndufs1, Ndufv1, Atp5g1, cox1, and Uqcr genes were determined by quantitative real-time PCR. (A-C, and E-F) Each symbol represents data from one animal and indicates n. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, P > 0.05.