Pyroptosis in host defence against bacterial infection

Abstract

Pyroptosis, a regulated form of pro-inflammatory cell death, is characterised by cell lysis and by the release of cytokines, damage- and pathogen-associated molecular patterns. It plays an important role during bacterial infection, where it can promote an inflammatory response and eliminate the replicative niche of intracellular pathogens. Recent work, using a variety of bacterial pathogens, has illuminated the versatility of pyroptosis, revealing unexpected and important concepts underlying host defence. In this Review, we overview the molecular mechanisms underlying pyroptosis and discuss their role in host defence, from the single cell to the whole organism. We focus on recent studies using three cellular microbiology paradigms - Mycobacterium tuberculosis, Salmonella Typhimurium and Shigella flexneri - that have transformed the field of pyroptosis. We compare insights discovered in tissue culture, zebrafish and mouse models, highlighting the advantages and disadvantages of using these complementary infection models to investigate pyroptosis and for modelling human infection. Moving forward, we propose that in-depth knowledge of pyroptosis obtained from complementary infection models can better inform future studies using higher vertebrates, including humans, and help develop innovative host-directed therapies to combat bacterial infection.

Keywords: Salmonella; Shigella; Bacterial infection; Cell death; Cell-autonomous immunity; Host-pathogen interaction; Mycobacteria; Pyroptosis.

Fig. 1.

Different infection models to investigate pyroptosis induced by mycobacteria, Salmonella or Shigella. (A) Representative time-lapse microscopy images of a THP-1 macrophage expressing GFP-tagged ASC, infected with M. tuberculosis (blue). Shown is a pyroptotic cell death event of an infected cell. Left image: ASC specks, a hallmark of inflammasome formation, are shown in green. Right image: uptake of the DNA dye DRAQ7 (red), was used to detect cell death. The images show the induction of pyroptosis in human macrophages within 40 minutes after infection with M. tuberculosis and illustrate how dying cells become permeabilised, as shown by the uptake of DRAQ7. ASC speck formation can be detected when M. tuberculosis is intracellular. Figure panels adapted from Beckwith et al. (2020). (B) Confocal and bright-field (inset) microscopy image of a granuloma in a zebrafish larva 3 days post infection with M. marinum (red). ASC specks are shown in green. Active caspases were labelled using a Flica assay based on a fluorescent inhibitor probe (FAM-YVAD-FMK) with Caspa shown in blue (Flica). Figure panel adapted from Varela et al. (2019 preprint). (C) Confocal microscopy images of wild-type and caspase-1-deficient mice cecal enteroid monolayers infected with mCherry-expressing S. Typhimurium (red). Increased intracellular bacterial burden can be seen in the caspase-1-deficient (Caspase-1^−/−) cells. In the absence of caspase-1, the epithelial layer fails to control bacterial proliferation, highlighting caspase-1 as a crucial factor in cell-autonomous immunity. These results demonstrate the importance of caspase-1 activation for infection control in Salmonella-infected epithelial cells. Figure panels adapted from Crowley et al. (2020). (D) Confocal microscopy images of S. Typhimurium-infected murine intestine showing the significantly increased bacterial burden in the lamina propria of NLRC4-deficient (NLRC 4^−/−) mice at 48 hours post infection compared with that of heterozygous (NLRC 4^+/−) control mice. EpCAM staining (orange) highlights the presence of epithelial cells, iCAM-1 staining (red) indicates endothelium, arrowheads indicate infection with S. Typhimurium (green). In the absence of the NLRC4 inflammasome, mice fail to control infection with Salmonella and this image highlights the role of NLRC4 in controlling infection in vivo. The location of the lamina propria is indicated by dashed lines. EpCAM, epithelial cell adhesion molecule (EPCAM); iCAM-1, intercellular adhesion molecule 1. Figure panels adapted from Fattinger et al. (2021). (E) Time-lapse confocal microscopy images of Tg(il-1b:GFP-F) x Tg(lyz:dsRed) zebrafish larvae injected with either control or sept15-targeting morpholino oligonucleotides (CTRL MO or Sept15 MO, respectively), followed by infection of the hindbrain ventricle for 19 hours with E2-Crimson-expressing S. flexneri M90T. Notice that lack of Sept15 significantly increased the bacterial burden (see staining for S. flexneri, dark green) and inflammation (see staining for neutrophils, red), and indicated by expression of IL-1β (white). The Shigella-zebrafish infection model enables investigation of the cell biology of infection in vivo, in this case showing increased cytokine expression and immune cell recruitment during S. flexneri infection. Zebrafish larvae are optically accessible, allowing non-invasive imaging of cellular events in vivo at high resolution. Figure panels adapted from Mazon-Moya et al. (2017). (F) Confocal microscopy images of wild-type or 129.NLRC4-deficient (NLRC 4^−/−) cecum-derived mouse intestinal epithelial cells grown in a monolayer and infected with S. flexneri (red). Here, Mitchell et al. discovered that, in the absence of the NAIP/NLRC4 inflammasome, mice develop a shigellosis-like phenotype. The images show that S. flexneri can replicate and form actin tails (arrows) in NLRC 4^−/− murine epithelial cells. Figure panels adapted from Mitchell et al. (2020).

Fig. 2.

Bacterial interactions with the inflammasome. (A) Top: Bacterial interactions with the inflammasome in epithelial cells. IpaH7.8 is an E3 ubiquitin ligase expressed by Shigella (Sandstrom et al., 2019), which can activate murine NLRP1B by ubiquitylating its N-terminus and targeting it to proteasomal degradation. Middle: Following the detection of Salmonella and Shigella by NAIP, NLRC4 is recruited to form the inflammasome (Broz, 2015; Fattinger et al., 2021; Gram et al., 2021; Mitchell et al., 2020). Bottom: IpaH7.8 also targets human gasdermin D (GSDMD) for proteasome degradation and, this way, can block pyroptosis of human but not mouse cells (Luchetti et al., 2021). The Shigella effector arginine ADP-riboxanase (OspC3) blocks the activation of caspase-4/11 (Kobayashi et al., 2013; Li et al., 2021a) and uses the effector IpaH9.8 to block GBP-mediated bacterial recognition (Li et al., 2017; Wandel et al., 2017). However, Salmonella is recognised by GBPs and this initiates the recruitment and activation of caspase-4, followed by GSDMD maturation and pyroptosis (Santos et al., 2020). (B) Top: Bacterial interactions with the inflammasome in macrophages. M. tuberculosis induce ESX-1-dependent phagosomal damage, leading to K⁺ efflux and formation of the NLRP3 inflammasome (Beckwith et al., 2020). This process can also be induced by EST-12 (Qu et al., 2020). However, M. tuberculosis can also inhibit the NLRP3 inflammasome via the bacterial phosphokinase PknF (Rastogi et al., 2021). Middle: The NAIP/NLRC4 inflammasome also recognises bacterial OMVs from Salmonella (Yang et al., 2020). In mouse macrophages, Shigella-induced cell death depends on NLRC4 (Mitchell et al., 2020). Bottom: The Shigella effector OspC3 blocks the activation of caspase-4/-11 (Kobayashi et al., 2013; Li et al., 2021a). During Salmonella infection, GBP-mediated recognition of bacteria is important for induction of pyroptosis (Fisch et al., 2019). In infected murine macrophages, induction of pyroptosis by M. marinum depends on caspase-11 (Varela et al., 2019 preprint). EST12, cell pyroptosis-inducing protein in M. tuberculosis (officially known as Rv1579c); ESX-1, ESAT-6 secretion system 1; NLRP1B, NACHT, LRR and PYD domain-containing protein 1b allele 1 (Mus musculus).