PMC

MyD88 is not required for Yersinia pestis induction of IFN-β. Wild-type and Myd88^−/− BMDMs were infected with the Y. pestis T3SS⁻ strain, and the IFN-β (A) or TNF-α (B) level was measured at 4 hpi. Error bars show standard deviations from the means. Data shown are representative of results from two independent trials; each sample was tested in duplicate. Data were analyzed by two-way ANOVA followed by the Holm-Sidak test for multiple comparisons. *, P < 0.05; nd, not detected.

Model for atypical MyD88-independent TLR7 signaling during Yersinia pestis infection. Wild-type Y. pestis prevents the acidification of the macrophage phagosome, and the bacteria remain viable in a pathogen-containing vacuole. For this compartment, we hypothesize that Y. pestis stimulates atypical TLR7 signaling (MyD88 independent), which activates the expression of IFN-β. In contrast, the phoP mutant is unable to prevent the acidification and maturation of the phagosome. Consequently, bacteria lyse and likely stimulate canonical TLR signaling through MyD88 and TRIF, which activates the transcription of IFN-β, TNF-α, and other proinflammatory cytokines. During pneumonic plague, the TLR7 type I IFN response protects the lungs but causes pathology in the liver.

TLR7 is required for optimal IFN-β expression by macrophages infected with Yersinia pestis. (A) WT BMDMs were infected with Y. pestis T3SS⁺ or T3SS⁻ bacteria (both pgm negative) at an MOI of 20; the IFN-β level in the supernatant was measured by an ELISA at 4 hpi. (B) RAW 264.7 macrophages were pretreated with cytochalasin D (CytoD) or the vehicle (UN) and then infected with the Y. pestis T3SS⁻ strain, and the IFN-β level in the supernatant was measured at 4 hpi. (C to F) WT or Tlr7^−/− BMDMs were infected with the Y. pestis T3SS⁻ strain (C and D) or Staphylococcus aureus (E and F) at an MOI of 20 (Y. pestis) or an MOI of 5 (S. aureus), control cells were mock infected, and IFN-β (C and E) and TNF-α (D and F) levels in the supernatant were measured by an ELISA at 4 hpi. Error bars show standard deviations from the means. Data shown are representative of results from at least two independent trials; each sample was tested in duplicate. Data were analyzed by two-way ANOVA followed by the Holm-Sidak test for multiple comparisons. *, P < 0.05; ns, not significant; nd, not detected.

Tlr7^−/− mice are more resistant to plague and express reduced serum IFN-β levels. Groups of 5 to 8 WT and Tlr7^−/− mice (A to C) or WT and MyD88^−/− mice (D) were challenged by intranasal infection with 1 × 10⁶ CFU Y. pestis KIM D27 (T3SS⁺ pgm negative) and monitored for survival over 14 days (n = 15 per strain, collected in 3 independent trials) (A) or euthanized on day 5 (B and C) or day 3 (D) postinfection, blood was collected, and serum was analyzed for IFN-β (B and D) or TNF-α (C) by an ELISA (n = 10 per strain, collected in 2 independent trials). X indicates that the animal died prior to analysis, and bars indicate the medians. Data were analyzed by a Mantel-Cox log rank test (A) or by Student's t test (B to D). *, P < 0.05.

Tlr7^−/− mice induce granulomatous lesions in the liver during Y. pestis KIM D27 infection. Groups of 5 to 8 WT and Tlr7^−/− mice were challenged by intranasal infection with Y. pestis KIM D27 (T3SS⁺ pgm negative). On day 5 postinfection, animals were euthanized, and formalin-fixed livers were sectioned, stained with hematoxylin and eosin, and analyzed by histopathology. (A and B) Representative lesions for WT (n = 6) (A) and Tlr7^−/− (n = 10) (B) mice (collected in two independent trials). Boxes outline the zoomed-in areas shown in the bottom panels. Bars, 100 μm (50 μm in the zoomed-in images). (C to E) Quantification of H&E-stained necrotic and inflammatory lesions in the livers from day 5 postinfection showing the number of inflammatory (C) or necrotic (D) lesions per low-power field (lpf) (left) and the percentage of inflammatory foci that were >100 μm in diameter (E). Data were collected in two independent trials and evaluated by unpaired Student's t test. *, P < 0.05; ns, not significant.

Y. pestis KIM YopJ partially inhibits IFN-β expression. RAW 264.7 macrophages were infected with the virulent wild-type CO92 (T3SS⁺ pgm⁺) (A), KIM D27 (T3SS⁺ pgm-negative) (A, C, and D), CO92pCD1⁻ (T3SS⁻ pgm-negative) (B), or KIM6 (T3SS⁻ pgm-negative) (A and B) Y. pestis strain at an MOI of 20. Mock cells were not infected. The Ifnβ mRNA level was quantified by real-time PCR at 2 hpi (A), or the IFN-β protein level was quantified by an ELISA at 4 hpi (B to D). (C and D) RAW 264.7 macrophages were pretreated with 20 μM z-VAD-fmk or the vehicle control prior to infection with Y. pestis KIM D27 or the yopJ_C172A mutant at an MOI of 20. At 4 hpi, LDH (C) and IFN-β (D) levels in the supernatant were measured, IFN-β levels were normalized to the amount present in 4 × 10⁵ macrophages (equal to the average number of viable cells in the WT-infected samples measured in panel C), and data shown are representative of results from three independent trials, each with biological replicates that were analyzed in duplicate. Statistical significance was evaluated by one-way ANOVA followed by Sidak's (A) or Tukey's (C and D) test for multiple comparisons or by paired Student's t test (B). ns, not significant; *, P < 0.05; nd, not detected.

Tlr7^−/− mice experience reduced serum TNF-α levels and sepsis during Y. pestis CO92 infection. Groups of 10 WT and Tlr7^−/− mice were challenged by intranasal infection with 1 × 10³ CFU Y. pestis CO92 (T3SS⁺ pgm⁺). (A) Survival was monitored over 14 days. Data shown were collected in 3 independent trials (n = 26 or 30 per group). (B to F) On day 3 postinfection, animals were euthanized; blood was collected by cardiac puncture for multiplex serum cytokine analysis; and lungs, liver, and spleen were homogenized for the enumeration of the bacterial load. (B) Bacterial counts in the indicated tissues. Bars indicate medians, and the percentages of each tissue that had undetectable bacterial titers are indicated in parentheses. (C) Serum TNF-α. (D) IL-6. The right panel shows percentages of animals in each group with high IL-6 titers. (E) IL-10. (F) IL-1β. Bars indicate the means, and the percentages of mice with undetectable cytokines are indicated (n = 17 WT and 19 Tlr7^−/− mice [3 WT mice and 1 Tlr7^−/− mouse died prior to 72 hpi], collected in two independent trials). Statistical significance was evaluated by a Mantel-Cox log rank test (A) or by Student's t test (B to F). ns, not significant.

TLR7 is associated with increased liver pathology during pneumonic plague. (A to F) Groups of 5 WT (A, C, E, and F) or Tlr7^−/− (B, D, E, and F) mice were challenged by intranasal infection with 1 × 10³ CFU Y. pestis CO92 (T3SS⁺ pgm⁺). (A to D) On day 3 postinfection, animals were euthanized, and formalin-fixed lungs (A and B), liver (C and D), and spleen (see Fig. S4A and S4B in the supplemental material) were sectioned, stained with hematoxylin and eosin, and analyzed by histopathology. Representative lesions are shown; boxes outline the zoomed-in areas shown in the bottom panels, and arrows point to bacteria in the zoomed-in images. Bars, 100 μm (50 μm in the zoomed-in images). (E and F) Quantification of histopathology. (E) Severity scoring; (F) percentages of samples with bacterial microcolonies visible by histopathology. Bars indicate medians. (G and H) Serum IFN-β and TNF-α levels were measured by an ELISA. Bars indicate medians, dotted lines indicate the limit of detection, and X indicates that the animal died prior to analysis. Statistical significance was evaluated by Student's t test. *, P < 0.05. Data shown in all panels were collected in 2 independent trials (n = 10 per group).