Flexible Usage and Interconnectivity of Diverse Cell Death Pathways Protect against Intracellular Infection

Abstract

Programmed cell death contributes to host defense against pathogens. To investigate the relative importance of pyroptosis, necroptosis, and apoptosis during Salmonella infection, we infected mice and macrophages deficient for diverse combinations of caspases-1, -11, -12, and -8 and receptor interacting serine/threonine kinase 3 (RIPK3). Loss of pyroptosis, caspase-8-driven apoptosis, or necroptosis had minor impact on Salmonella control. However, combined deficiency of these cell death pathways caused loss of bacterial control in mice and their macrophages, demonstrating that host defense can employ varying components of several cell death pathways to limit intracellular infections. This flexible use of distinct cell death pathways involved extensive cross-talk between initiators and effectors of pyroptosis and apoptosis, where initiator caspases-1 and -8 also functioned as executioners when all known effectors of cell death were absent. These findings uncover a highly coordinated and flexible cell death system with in-built fail-safe processes that protect the host from intracellular infections.

Keywords: Salmonella; apoptosis; caspase-1; caspase-11; caspase-8; cell death; effector caspases; gasdermin D; necroptosis; pyroptosis.

Graphical abstract

Figure 1

Combined Loss of Caspases-1, -11, -12, and -8 (plus RIPK3) Leads to Lack of Bacterial Control upon Salmonella Infection (A) Bacterial replication over time in WT and Casp1^–/–;Casp11^–/– mice infected with Salmonella ΔAroA (200 CFU). n = 10−22 mice per group per time point. Mean and SEM are shown. ^∗∗p < 0.005, ^∗p < 0.05, ^nsp > 0.05 = not significant. (B) Bacterial loads in spleen and liver of mice of the indicated genotypes 3 weeks post-infection with Salmonella ΔAroA (200 CFU). n = 7−48 mice per genotype. Mean and SEM are shown. ^∗∗p < 0.005, ^∗p < 0.05, ^nsp > 0.05 = not significant. (C) Bacterial loads in spleen and liver from mice of the indicated genotypes 1 to 3 weeks post-infection with Salmonella ΔAroA (200 CFU). n = 3−4 mice per genotype and time point. Mean and SEM are shown. ^∗∗p < 0.005, ^∗p < 0.05, ^nsp > 0.05 = not significant. (D) Mouse survival curves and corresponding bacterial loads in the spleen and liver at time of sacrifice in WT and Casp1^–/–;Casp11^–/–;Casp12^–/–;Casp8^–/–;Ripk3^−/− mice infected with Salmonella ΔAroA (200 CFU). n = 7−8 mice per genotype. Mean and SEM are shown. ^∗∗p < 0.005. (E) Bone marrow chimeras of the indicated genotypes were infected with Salmonella ΔAroA (200 CFU) and culled for analysis of bacterial loads in spleen and liver 3 weeks post-infection. n = 10 mice per group. Mean and SEM are shown. ^∗∗p < 0.005. Please also see Figure S1.

Figure 2

Combined Loss of Caspases-1, -11, -12, and -8 Abrogates the Death of BMDMs upon Salmonella Infection (A) LDH release cell death assay of primary BMDMs of the indicated genotypes after infection with Salmonella SL1344 (MOI = 50). Data pooled from two or more experiments. Mean and SEM are shown. ^∗∗p < 0.005, ^nsp > 0.05 = not significant. (B) BMDMs of the indicated genotypes were infected with Salmonella SL1344 (MOI = 50) and cleavage associated with activation of the indicated cell death regulators was analyzed by immunoblotting at the indicated time points. Probing for β-actin served as a loading control. (C) Confocal or lattice light-sheet imaging of BMDMs of the indicated genotypes after infection with GFP-expressing Salmonella (MOI = 50) at the indicated time points. Yellow, membrane; Magenta, Salmonella; Cyan, PI. Scale bars: 10 μm. Please see also Figure S2.

Figure 3

Caspase-8-Mediated Apoptosis Is the Default Backup Mechanism when Caspases-1- and 11-Mediated Pyroptosis Is Disabled in Salmonella-Infected iBMDMs (A) LDH release cell death assay of iBMDMs of the indicated genotypes after infection with Salmonella SL1344 (MOI = 50). Data pooled from two or more experiments. Mean and SEM are shown. ^∗∗p < 0.005, ^nsp > 0.05 = not significant. (B) Immunoblot analysis of the indicated proteins in iBMDMs of the indicated genotypes after infection with Salmonella SL1344 (MOI = 50). Probing for β-actin served as a loading control. (C) LDH release cell death assay of Salmonella SL1344 (MOI = 50) -infected WT and Casp1^–/–;Casp11^–/–;Casp12^–/– iBMDMs that had been left untreated or treated with the RIPK1 inhibitor, Nec1s (30 μM). Data pooled from two or more experiments. Mean and SEM are shown. ^∗∗p < 0.005, ^nsp > 0.05 = not significant. (D) Immunoblot analysis of the indicated proteins in Salmonella SL1344 (MOI = 50) -infected WT and Casp1^–/–;Casp11^–/–;Casp12^–/– iBMDMs that had been left untreated or treated with the RIPK1 inhibitor, Nec1s (30 μM). Probing for β-actin served as a loading control. Please see also Figures S3 and S4.

Figure 4

Caspase-11 Can Compensate for the Loss of Caspases-1 and -8 to Ensure GSDMD-Mediated Killing of Salmonella-Infected Cells (A) LDH release cell death assay of Salmonella SL1344 (MOI = 50) -infected iBMDMs of the indicated genotypes that had been left untreated or treated with VX-765 or Emricasan. Data pooled from two or more experiments. Mean and SEM are shown. ^∗∗p < 0.005, ^nsp > 0.05 = not significant. (B) LDH release cell death assays of iBMDMs of the indicated genotypes that had been infected with Salmonella SL1344 (MOI = 50). Data pooled from two or more experiments. Mean and SEM are shown. ^∗∗p < 0.005, ^nsp > 0.05 = not significant. (C) LDH release cell death assays of Salmonella-infected iBMDMs of the indicated genotypes or Casp1^–/–;Casp8^–/–;Ripk3^–/– iBMDMs that had been left untreated or treated with Emricasan and infected with Salmonella SL1344 (MOI = 50). Data pooled from two or more experiments. Mean and SEM are shown. ^∗∗p < 0.005, ^nsp > 0.05 = not significant. (D) Immunoblot analysis of caspase-11, GSDMD, BID, and PARP in Casp1^–/–;Casp8^–/–;Ripk3^–/– iBMDMs that had been left untreated or treated with Emricasan and infected with Salmonella SL1344 (MOI = 50). WT iBMDMs that had been left untreated or treated with LPS for 4 h were used as a control for the induction of caspase-11. Probing for β-actin served as a loading control. Please see also Figure S4.

Figure 5

Caspase-1 Can Activate Caspases-3, -7, and -9 Independently of Caspase-8 and BID (A) LDH release cell death assays of iBMDMs of the indicated genotypes that had been infected with Salmonella SL1344 (MOI = 50). Data pooled from two or more experiments. Mean and SEM are shown. ^∗∗p < 0.005, ^nsp > 0.05 = not significant. (B) iBMDMs of the indicated genotypes were infected with Salmonella SL1344 (MOI = 50) and cleavage associated with activation of the indicated cell death proteins was analyzed by immunoblotting at the indicated time points. Probing for β-actin served as a loading control. (C) LDH release cell death assays of iBMDMs of the indicated genotypes that had been infected with Salmonella SL1344 (MOI = 50). Data pooled from two or more experiments. Mean and SEM are shown. ^∗∗p < 0.005, ^nsp > 0.05 = not significant. (D) GsdmD^–/–;Casp8^–/–;Ripk3^–/–;Mlkl^–/–;Casp3^–/–;Casp7^–/– iBMDMs were infected with Salmonella SL1344 (MOI = 50) and expression and cleavage associated with activation of the indicated cell death proteins was analyzed by immunoblotting at the indicated time points. Probing for β-actin served as a loading control. (E) LDH release cell death assays of iBMDMs of the indicated genotypes that had been infected with Salmonella SL1344 (MOI = 50). Data pooled from two or more experiments. Mean and SEM are shown. ^nsp > 0.05 = not significant. (F) LDH release cell death assays of iBMDMs of the indicated genotypes that had been infected with Salmonella SL1344 (MOI = 50). Data pooled from two or more experiments. Mean and SEM are shown. ^nsp > 0.05 = not significant. Please see also Figure S5.

Figure 6

CRISPR Screen Reveals a Central Role for Caspase-1 in Mediating Salmonella-Infection-Induced Cell Death Independent of All Known Downstream Effectors of Cell Killing (A) LDH release cell death assays of iBMDMs of the indicated genotypes that had been infected with Salmonella SL1344 (MOI = 50). Data pooled from two or more experiments. Mean and SEM are shown. ^∗p < 0.05, ^nsp > 0.05 = not significant. (B) GsdmD^–/–;Bid^–/–;Mlkl^–/–;Casp3^–/–;Casp7^–/–;Casp9^–/– iBMDM whole genome CRISPR-Cas9 screen mean-difference (MD) plot showing log-fold change versus average log counts per million (CPM) after three rounds of infection with Salmonella SL1344 (MOI = 50) (please also see Figures S5A and S5B). (C) GsdmD^–/–;Bid^–/–;Mlkl^–/–;Casp3^–/–;Casp7^–/–;Casp9^–/– iBMDMs were infected with Salmonella SL1344 (MOI = 50) and cleavage associated with activation of caspases-1 and -8 was analyzed by immunoblotting at the indicated time points. Probing for β-actin served as a loading control. (D) LDH release cell death assays of WT, Casp1^–/–;Casp11^–/–;Casp12^–/–;Casp8^–/–;Ripk3^–/–, and two independent clones (#1 and #2) of GsdmD^–/–;Bid^–/–;Mlkl^–/–;Casp3^–/–;Casp7^–/–;Casp9^–/–;Casp1^–/– iBMDMs that had been infected with Salmonella SL1344 (MOI = 50). Data pooled from two or more experiments. Mean and SEM are shown. ^nsp > 0.05 = not significant. Please see also Figure S6.

Figure 7

Caspase-1 Can Act Upstream of and Requires Caspase-8 to Induce Cell Death in the Absence of All Known Downstream Effectors of Pyroptosis and Apoptosis (A) iBMDMs of the indicated genotypes were infected with Samonella SL1344 (MOI = 50) and cleavage associated with activation of caspases-1 and -8 was analyzed by immunoblotting at the indicated time points. Probing for HSP70 served as loading control. (B) LDH release death assays of Salmonella SL1344 (MOI = 50) -infected Casp1^–/–;Casp11^–/–;Casp12^–/–;Casp8^–/–;Ripk3^–/– and GsdmD^–/–;Bid^–/–;Mlkl^–/–;Casp3^–/–;Casp7^–/–;Casp9^–/– iBMDMs that had been left untreated or treated with VX-765 or Emricasan. Data pooled from two experiments. Mean and SEM are shown. ^∗∗p < 0.005, ^nsp > 0.05 = not significant. (C) Immunoblot analysis of caspases-1 and -8 activation at the indicated time points in Salmonella SL1344 (MOI = 50) -infected GsdmD^–/–;Bid^–/–;Mlkl^–/–;Casp3^–/–;7^–/–;9^–/– iBMDMs that had been left untreated or treated with VX-765. Probing for HSP70 served as a loading control. Please see also Figure S6.