Elucidating Mechanisms of Tolerance to Salmonella Typhimurium across Long-Term Infections Using the Collaborative Cross

Abstract

Understanding the molecular mechanisms underlying resistance and tolerance to pathogen infection may present the opportunity to develop novel interventions. Resistance is the absence of clinical disease with a low pathogen burden, while tolerance is minimal clinical disease with a high pathogen burden. Salmonella is a worldwide health concern. We studied 18 strains of collaborative cross mice that survive acute Salmonella Typhimurium (STm) infections. We infected these strains orally and monitored them for 3 weeks. Five strains cleared STm (resistant), six strains maintained a bacterial load and survived (tolerant), while seven strains survived >7 days but succumbed to infection within the study period and were called "delayed susceptible." Tolerant strains were colonized in the Peyer's patches, mesenteric lymph node, spleen, and liver, while resistant strains had significantly reduced bacterial colonization. Tolerant strains had lower preinfection core body temperatures and had disrupted circadian patterns of body temperature postinfection sooner than other strains. Tolerant strains had higher circulating total white blood cells than resistant strains, driven by increased numbers of neutrophils. Tolerant strains had more severe tissue damage and higher circulating levels of monocyte chemoattractant protein 1 (MCP-1) and interferon gamma (IFN-γ), but lower levels of epithelial neutrophil-activating protein 78 (ENA-78) than resistant strains. Quantitative trait locus (QTL) analysis revealed one significant association and six suggestive associations. Gene expression analysis identified 22 genes that are differentially regulated in tolerant versus resistant animals that overlapped these QTLs. Fibrinogen genes (Fga, Fgb, and Fgg) were found across the QTL, RNA, and top canonical pathways, making them the best candidate genes for differentiating tolerance and resistance. IMPORTANCE To survive a bacterial infection, an infected host can display resistance or tolerance. Resistance is indicated by a decrease in pathogen load, while for tolerance a high pathogen load is accompanied by minimal disease. We infected genetically diverse mice with Salmonella Typhimurium for 21 days and discovered new phenotypes for disease outcome (delayed susceptible, tolerant, and resistant). Tolerant strains showed the lowest preinfection core body temperatures and the most rapid disruption in circadian patterns of body temperature postinfection. Tolerant strains had higher circulating neutrophils and higher circulating levels of MCP-1 and IFN-γ, but lower levels of ENA-78 than did resistant strains, in addition to more severe tissue damage. QTL analysis revealed multiple associated regions, and gene expression analysis identified 22 genes that are differentially regulated in tolerant versus resistant animals in these regions. Fibrinogen genes (Fga, Fgb, and Fgg) were found across the QTL, RNA, and the top canonical pathways, suggesting a role in tolerance.

Keywords: QTL; Salmonella; collaborative cross; fibrinogen; resistance; tolerance.

FIG 1

Delayed susceptible, tolerant, and resistant strains have distinct responses to STm infections. (A and B) Survival time (A) and percent weight change (B) after infection with STm. Strains are shown in ascending order of survival and weight change if survival was equal between strains. Means and standard deviations are shown. (C and D) Liver (C) and spleen (D) CFU counts per gram of organ at time of necropsy. Medians and interquartile ranges are shown. Circles represent individual mice (black circles represent males, and gray circles represent females). A total of 18 CC strains are represented. Slc11a1 and Ncf2 status (+, wild type; –, mutated) are shown at the bottom, as well as the response to infection (tolerant shown in blue).

FIG 2

CC strains resistant to STm infection have lower MLN and PP colonization compared to tolerant strains. Peyer’s patches (PP) (A) and MLN (B) CFU counts per gram of organ at necropsy. Each circle represents an individual mouse (black circles represent males, and gray circles represent females), and medians and interquartile range, are indicated by lines. Tolerant (T, open circles) and resistant (R, filled circles) are indicated. A Kruskal-Wallis test was performed to identify significant differences (*, P < 0.05; **, P < 0.01; ***, P < 0.001).

FIG 3

Tolerant mice have cooler body temperatures and deviate from normal body temperature patterns sooner than resistant mice. (A) Core body temperatures during resting, 24-h, and active periods for delayed susceptible (DS, filled triangles), tolerant (T, open circles), and resistant (R, filled circles) mice. (B) Times to deviation from circadian pattern (“off pattern”) for temperature and activity. Each symbol represents an individual mouse (black symbols represent males, and gray symbols represent females). Medians and interquartile ranges are indicated by blue lines. A Kruskal-Wallis test was performed to identify significant differences (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001).

FIG 4

Tolerant mice have higher WBC and higher NEU than resistant mice. Difference (infected – uninfected = Δ) between infected and uninfected for circulating. (A) Total white blood cells (WBC); (B) neutrophils (NEU); (C) monocytes (MON); (D) lymphocytes (LYM). Each circle represents an individual mouse (black circles represent males, and gray circles represent females), and lines represent medians and interquartile ranges. T, tolerant (open circles); R, resistant (filled circles). A Kruskal-Wallis test was performed to identify significant differences (*, P < 0.05; **, P < 0.01).

FIG 5

Tolerant strains have higher circulating MCP-1 and IFN-γ levels but a lower ENA-78 level at 3 weeks postinfection than resistant strains. Differences (Δ) between infected and uninfected levels of MCP-1 (A), IFN-γ (B), and ENA-78 (C) in tolerant and resistant strains. Circles represent individual mice (black circles represent males, and gray circles represent females), and lines represent medians and interquartile ranges. Outliers were removed. T, tolerant (open circles); R, resistant (closed circles). A Kruskal-Wallis test was performed to identify significant differences (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001).

FIG 6

Resistant strains have less damage in the spleen and liver at 3 weeks postinfection than do tolerant and delayed susceptible strains. The mean spleen and liver histopathology scores for tissue damage and the standard deviations are indicated. DS, delayed susceptible (red bars), T, tolerant (black bars); R, resistant (gray bars). A Kruskal-Wallis test was performed to identify significant differences (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001).

FIG 7

QTLs identified for spleen and liver colonization using 18 CC strains. (A and B) QTL of spleen CFU at 3 weeks postinfection (A) and allele effect plots for Chr 3 (B). (C and D) QTL of liver CFU at 3 weeks postinfection (C) and allele effect plots focused in on Chr 13 (D). The green dotted line indicates 85% significance, blue dotted line indicates 90% significance, and the red dotted line indicates 95% significance. Results were obtained using gQTL.

FIG 8

Binary categorization to identify QTLs linked to STm resistance using 32 CC strains. (A) QTL of resistant categorization (score of 1) versus susceptible, delayed susceptible, and tolerant categorization (score of 0) after STm infections. (B to D) Allele effect plots zoomed in for Chr 3 (B), Chr 13 (C), and Chr 17 (D). Results were obtained using gQTL.

FIG 9

Binary categorization to identify QTLs linked to STm tolerance using 27 CC strains. (A) QTL of tolerant categorization (score of 1) versus susceptible and delayed susceptible categorization (score of 0) after STm infections. (B and C) Allele effect plots zoomed in for Chr 2 (B) and Chr 6 (C). Results were obtained using gQTL.