Macrophages mediate flagellin induced inflammasome activation and host defense in zebrafish

Abstract

The inflammasome is an innate immune complex whose rapid inflammatory outputs play a critical role in controlling infection; however, the host cells that mediate inflammasome responses in vivo are not well defined. Using zebrafish larvae, we examined the cellular immune responses to inflammasome activation during infection. We compared the host responses with two Listeria monocytogenes strains: wild type and Lm-pyro, a strain engineered to activate the inflammasome via ectopic expression of flagellin. Infection with Lm-pyro led to activation of the inflammasome, macrophage pyroptosis and ultimately attenuation of virulence. Depletion of caspase A, the zebrafish caspase-1 homolog, restored Lm-pyro virulence. Inflammasome activation specifically recruited macrophages to infection sites, whereas neutrophils were equally recruited to wild type and Lm-pyro infections. Similar to caspase A depletion, macrophage deficiency rescued Lm-pyro virulence to wild-type levels, while defective neutrophils had no specific effect. Neutrophils were, however, important for general clearance of L. monocytogenes, as both wild type and Lm-pyro were more virulent in larvae with defective neutrophils. This study characterizes a novel model for inflammasome studies in an intact host, establishes the importance of macrophages during inflammasome responses and adds importance to the role of neutrophils in controlling L. monocytogenes infections.

Figure 1. L. monocytogenes virulence and dissemination in zebrafish larvae

(A, B, C) Survival of larvae infected with WT, Δhly, and ΔactA, respectively at 10, 100, 1000, and 10000 CFU. (D, E, F) Dissemination of GFP expressing WT, Δhly, and ΔactA infection, respectively. (A, B, C) L. monocytogenes is virulent in zebrafish larvae, dependent upon the classic virulence factors listeriolysin O (hly) and ActA (actA). 48 hpf zebrafish larvae were infected with WT, Δhly, and ΔactA strains. Survival following inoculation was checked every 12 hours for 5 days. Each graph is the pooled data from three independent survival experiments. P-values and hazard ratios (HR) are relative to WT infection at the same dose of bacteria. (D, E, F) Representative images of dissemination over the course of 100 CFU infection. Larvae infected with WT, Δhly, and ΔactA strains constitutively expressing GFP were imaged daily. Δhly and ΔactA strains did not disseminate. Scale bars are 500 μm.

Figure 2. Lm-pyro is attenuated in zebrafish larvae

(A) Survival of larvae infected with WT and Lm-pyro at 10, 100, 1000, and 10000 CFU. (B) Dissemination of GFP expressing WT and Lm-pyro. (A) Larvae were infected as in Figure 1. WT L. monocytogenes repeated with similar virulence. Lm-pyro displayed virulence defects at 10 and 100 CFU. At 1000 CFU virulence was equivalent, and slightly increased at 10000 CFU. Each graph is the pooled data from three independent survival experiments. P-values and hazard ratios (HR) are relative to WT infection at the same dose of bacteria. (B) Representative images of dissemination over the course of 100 CFU infection with WT and Lm-pyro. WT disseminated similar to Figure 1. Lm-pyro did not disseminate. Scale bars are 500 μm.

Figure 3. Innate immune responses to Lm-pyro include ASC specks and pyroptosis

(A) Quantification of ASC specks at the site of infection and (B) representative images of larvae expressing GFP-ASC mRNA following 100 CFU infection. (C) Image stills (see also Supplementary Movie) showing macrophage pyroptosis during 100 CFU Lm-pyro infection. (A, B) Larvae were fixed at 12 hpi and imaged. ASC specks were quantified and normalized to the number of specks in PBS mock inoculated larvae. The graph is representative of three independent experiments. Scale bars are 100 μm. (C) Macrophages expressing the fluorophores dendra2 and mCherry in the cytosol and fused to histone-2B, respectively, were imaged. In macrophages undergoing pyroptosis, cytoplasmic signal was rapidly lost, coinciding with nuclear condensation. Scale bars are 10 μm.

Figure 4. Lm-pyro virulence is rescued in the absence of caspase A

Survival of morphant larvae infected with (A) WT and (B) Lm-pyro at 100 CFU. (A) WT virulence was unchanged in control or caspase A morphants. (B) Lm-pyro virulence was rescued during the depletion of caspase A. Each graph is the pooled data from three independent survival experiments.

Figure 5. Phagocyte recruitment to WT and Lm-pyro infection

(A, B) Quantification of macrophages [Tg(mpeg1:dendra)] recruited to the HBV and representative images. (C, D) Quantification of neutrophils [Tg(lyz:eGFP)] recruited to the HBV and representative images. Larvae were inoculated with 100 CFU of bacteria, fixed at 6 hpi and imaged, and cells were quantified. Graphs are a single representative experiment from three independent experiments.

Figure 6. Macrophage depletion removes the virulence defect of Lm-pyro

(A, B) Survival of Irf8 morphant larvae lacking macrophages following infection with WT and Lm-pyro, (A) and (B) respectively, at 100 CFU. (C, D) Survival of Tg(mpx-Rac2 D57N) larvae with functionally defective neutrophils following infection with WT and Lm-pyro, (C) and (D) respectively. (E, F) Survival of Pu.1 morphant larvae lacking both macrophage and neutrophil development following infection with WT and Lm-pyro, (E) and (F) respectively. Each graph is the pooled data from three independent survival experiments. Statistical analyses compare phagocyte impaired to non-impaired condition. (G) Comparison table displaying Fisher’s exact test comparisons, assessing the significant difference of final survival outcomes.