RIPK1-dependent apoptosis bypasses pathogen blockade of innate signaling to promote immune defense

Abstract

Many pathogens deliver virulence factors or effectors into host cells in order to evade host defenses and establish infection. Although such effector proteins disrupt critical cellular signaling pathways, they also trigger specific antipathogen responses, a process termed "effector-triggered immunity." The Gram-negative bacterial pathogen Yersinia inactivates critical proteins of the NF-κB and MAPK signaling cascade, thereby blocking inflammatory cytokine production but also inducing apoptosis. Yersinia-induced apoptosis requires the kinase activity of receptor-interacting protein kinase 1 (RIPK1), a key regulator of cell death, NF-κB, and MAPK signaling. Through the targeted disruption of RIPK1 kinase activity, which selectively disrupts RIPK1-dependent cell death, we now reveal that Yersinia-induced apoptosis is critical for host survival, containment of bacteria in granulomas, and control of bacterial burdens in vivo. We demonstrate that this apoptotic response provides a cell-extrinsic signal that promotes optimal innate immune cytokine production and antibacterial defense, demonstrating a novel role for RIPK1 kinase-induced apoptosis in mediating effector-triggered immunity to circumvent pathogen inhibition of immune signaling.

Figure 1.

RIPK1 kinase activity is required for Yersinia-induced apoptosis. WT and Ripk1^kd BMDMs were treated with 60 µM of the RIPK1 inhibitor Nec-1 or control media and infected with YopJ-deficient (ΔYopJ) or WT (Yp) Y. pseudotuberculosis. (A) Cell death was measured by LDH release assay at 4 h postinfection. (B and C) Apoptotic caspase cleavage was assessed by flow cytometry staining for CC3 (B) and Western blotting of lysates (C) 2 h postinfection. Molecular mass is indicated in kilodaltons. (D and E) Supernatants of BMDMs were measured for release of TNF at 4 h postinfection (D) or for TNF, IL-12p40, and IL-6 at 6 h after treatment with TLR ligands (E). Data are representative of more than three independent experiments in A and two independent experiments in B–E. Graphs show mean and SD of triplicate experiments. (F) Splenocytes were harvested 4 h after intravenous infection with Yp and stained for the indicated cell types and CC3. (G) Percentage of WT and Ripk1^kd CC3⁺ spleen cells infected with Yp relative to WT mice infected with the virulence plasmid–deficient strain (YpP⁻). Flow cytometry numbers and bar graph show mean and SD (n = 4–5 mice per group) values representative of two independent experiments. #, P < 0.05 compared with YpP⁻ infection. Statistical differences were determined by Student’s t test. **, P < 0.01; ***, P < 0.001; NS, not significant.

Figure 2.

RIPK1 kinase activity protects against in vivo Yersinia infection. (A) Survival of WT and Ripk1^kd mice infected with 3 × 10⁸ CFUs Y. pseudotuberculosis (Yp) by oral gavage (total n = 10 for WT and 17 for Ripk1^kd). (B and C) Bacterial burdens were quantified on day 5 postinfection as in A. Solid line represents the geometric mean. Dotted line represents the limit of detection in the lung. Data are representative of more than four independent experiments in A and B and three independent experiments in C. (D) Survival of mice provided with control or GSK′547 chow before being infected as in A (total n = 10–19 mice per condition). (E–G) MLNs from WT and Ripk1^kd mice were harvested 5 d postinfection, as in A. (E) Parallel sections were stained by immunohistochemistry with antibodies against Yp or CC3. Bars, 100 µm. (F) H&E-stained sections show WT MLNs with fewer bacterial colonies (arrows), often contained within well-circumscribed pyogranulomas (asterisks). Bacterial colonies in Ripk1^kd MLNs are dispersed throughout areas of less discrete necrotizing inflammation, including in the space adjacent to the subcapsular sinus (arrowheads). Bars, 500 µm. Images in E and F are representative of two independent experiments (n = 9–10 mice per group). (G) Tissue destruction and effacement of MLNs from Yp- and ΔYopJ-infected mice, quantified from samples stained as in F and outlined as in Fig. S1 B. Line represents the mean (total n = 8–10 mice per group). (H) Survival of WT and Ripk1^kd mice infected with 3 × 10⁸ CFUs of either Yp or ΔYopJ by oral gavage (total n = 18–26 mice per condition). (I) Bacterial burdens from mice infected as in H were measured on day 5 postinfection. Data pooled from two independent experiments (n = 8–10 mice per group). Statistical significance in differences between bacterial burdens in B, C, and I was determined by Mann–Whitney U test and between percent effaced area in G by Student’s t test. The interaction between mouse genotype and bacterial infection in G and I was tested by two-way analysis of variance. Survival data in A, D, and H were pooled from two independent experiments and statistical significance determined by log-rank test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, not significant.

Figure 3.

Hematopoietic RIPK1 kinase activity is required for optimal control of bacterial infection and in vivo cytokine production. (A) BM chimeras were generated by reconstituting lethally irradiated congenic hosts with BM from either WT or Ripk1^kd mice. (B) BM chimeric mice were infected with 1–2 × 10⁸ CFUs Y. pseudotuberculosis, and tissue bacterial burdens were measured on day 5 postinfection. Line represents the geometric mean. (C and D) Frequency of live cells and the number of inflammatory monocytes (CD11b^hiLy6C^hiLy6G⁻) and neutrophils (CD11b^hiLy6G⁺) in the MLNs of infected mice were measured by flow cytometry on day 5 postinfection. (E–G) Cytokine expression by innate immune cell populations isolated from MLNs was measured by flow cytometry on day 5 postinfection. Line represents the mean. Data in B–G are representative of three independent experiments (n = 7 mice per group). (H) Chimeric mice were generated by reconstituting lethally irradiated WT or Ripk1^kd hosts with WT congenic BM. (I) BM chimeric mice were infected as in A, and bacterial burdens measured on day 5 postinfection. Line represents the geometric mean. Data were pooled from two independent experiments, each with similar results (n = 18–21 mice per group). Statistical significance in B and I was determined by Mann–Whitney U test and in C–G by Student’s t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001; NS, not significant.

Figure 4.

RIPK1 is a cell-extrinsic regulator of in vivo cytokine production. (A) WT and RIPK3-deficient mice were infected with 2 × 10⁸ CFUs Y. pseudotuberculosis (Yp) by oral gavage and bacterial burdens measured on day 5 postinfection. Line represents the geometric mean. Data pooled from three independent experiments, each with similar results (n = 23–25 mice per group). (B) Mixed BM chimeras were generated by reconstituting lethally irradiated congenic hosts with a 1:1 mixture of either WT or Ripk1^kd BM (CD45.2⁺) and WT congenic B6.SJL BM (CD45.1⁺). (C) Chimeric mice were infected with 2 × 10⁸ CFUs Yp, and cytokine expression on day 5 postinfection was measured by flow cytometry. Congenic labeling of donor cells allowed for comparison within individual mice (paired t test) in addition to comparison across experimental groups (unpaired t test). Lines connect congenic cell populations within individual mice. (D) Bacterial burdens were measured on day 5 postinfection of mixed BM chimeras. Data in C and D are representative of three independent experiments (n = 7 mice per group). Statistical significance in bacterial burdens in A and D was determined by Mann–Whitney U test. *, P < 0.05; **, P < 0.01; NS, not significant. (E) Model of YopJ-mediated inhibition, RIPK1-dependent apoptosis, and the induction of host-protective antibacterial inflammatory responses.