Cutting Edge: Caspase-8 Is a Linchpin in Caspase-3 and Gasdermin D Activation to Control Cell Death, Cytokine Release, and Host Defense during Influenza A Virus Infection

Innate immunity acts as the first line of defense against invading pathogens. A key component of the innate immune response is programmed cell death (PCD), which can eliminate infected cells and prevent pathogen replication and spread. Because aberrant activation of PCD can cause inflammation and pathology, regulation of PCD pathways is critical for organismal homeostasis. Caspases are key proteins in the regulation of PCD pathways. Caspase-8 in particular is one of the most versatile caspases and is involved in regulating all three of the most well-defined PCD pathways: pyroptosis, apoptosis, and necroptosis. Caspase-8 is essential for both canonical and noncanonical Nod-like receptor family pyrin domain-containing 3 (NLRP3) inflammasome activation (1) and can play a role in gasdermin D (GSDMD)-mediated IL-1β and IL-18 release and pyroptosis (2, 3). It also acts as an initiator caspase downstream of death receptors to activate caspase-3 and -7 for apoptosis (4). Furthermore, it inhibits necroptosis by cleaving receptor interacting serine/threonine kinase (RIPK) 1 (RIPK1) (5). Despite these known roles, study of caspase-8 in vivo has been hindered by a lack of genetic tools. Mice lacking caspase-8 undergo embryonic lethality that can be rescued by additional deletion of Ripk3 (6, 7); therefore, study of caspase-8 has been limited to studies with Casp8^−/−Ripk3^−/− mice, which prevents identification of the role of caspase-8 under circumstances where RIPK3 is also functionally involved. Due to the critical roles of RIPK3 in both necroptotic and many nonnecroptotic contexts (8), the function of caspase-8 and the PCD pathways regulated by it under such physiological conditions has not been well-defined.

Influenza A virus (IAV) infection is a global health threat that continues to cause significant outbreaks every year. PCD pathways play an important role in the host response to IAV, and this virus has been shown to trigger pyroptosis, apoptosis, and necroptosis (PANoptosis), which is controlled by the master regulator Z-DNA-binding protein 1 (ZBP1) (9). Downstream of ZBP1 sensing of IAV, the ZBP1-PANoptosome is formed, which contains key components from pyroptosis (the NLRP3 inflammasome), apoptosis, and necroptosis (10). Formation of the inflammasome triggers caspase-1 activation, release of IL-1β and IL-18, and cell death (11). Inflammasome activation is vitally important to protect the host during IAV infection, as mice deficient in Nlrp3, Pycard (gene encoding the inflammasome adaptor protein apoptosis-associated speck-like protein containing a caspase recruitment domain [ASC]) or Casp1 are all highly susceptible to IAV infection (12, 13). Additionally, mice deficient in Il1r1 and Il18 also exhibit defects in controlling viral titer (13–15), suggesting that inflammasome-mediated cytokine maturation and release from pyroptotic cells is critical in host protection. Given the central role of caspase-8 in regulating PCD pathways and NLRP3 inflammasome activation and the importance of inflammasome activation in host defense during IAV infection (1, 9, 12–15), it is important to understand how caspase-8 functions during IAV-induced cell death. However, RIPK3 is also critically involved in IAV-induced pyroptosis, apoptosis, and necroptosis (9, 16), thereby preventing the use of Casp8^−/−Ripk3^−/− mice as a model to investigate caspase-8 functions during IAV infection. Therefore, the role of caspase-8 in the modulation of cell death during IAV infection remains unclear. Gaining a mechanistic understanding of the regulation of PCD pathways in IAV infection is important for identifying new strategies to prevent the morbidity and mortality associated with this infection.

To study caspase-8 functions during IAV infection, we generated a caspase-8 mutant mouse with a deficiency in caspase-8 autoprocessing, yielding an uncleavable version of the protein (CASP8^DA) that, consistent with previous findings, reduces the activation of apoptotic proteins (17). We found that IAV-induced activation of GSDMD and release of the inflammasome-dependent cytokines IL-1β and IL-18 were enhanced in CASP8^DA bone marrow-derived macrophages (BMDMs). Mechanistically, caspase-3 was activated downstream of caspase-8 to cleave GSDMD to yield its inactive N-terminal p20 fragment, thus limiting the cytokine release from dying cells. Collectively, our results show that during IAV infection, caspase-8 coordinates the negative regulation of GSDMD-mediated pyroptosis and inhibits proinflammatory cytokine release through the caspase-8–caspase-3 axis.

Casp8^DA/DA (caspase-8^D387A/D387A) mice were generated by the St. Jude Transgenic/Gene Knockout Shared Resource using CRISPR/Cas9 technology by injecting pronuclear stage C57BL/6J zygotes with a single-guide RNA (Caspase8_D387A_Guide 01, 5′-ATGAATCCACTTCTAAAGTGNGG-3′; 125 ng/μl) designed to introduce a dsDNA break into exon 7 of the Casp8 gene, human codon-optimized Cas9 mRNA transcripts (50 ng/μl), and a 200-nucleotide ssDNA molecule containing the desired mutations (Casp8–D387A homology-directed repair, 5′-GAAAGTCTGCCTCATCCGGAATATAGTTCTTGTGAGATGAGCTCGCCACTTCTAAAGTGTGGTTCTGTTGCTCGAAGCCTGCCTCATCAGGCACTC-3′). To facilitate the identification of founder mice and genotyping, silent substitutions generating an SacI restriction site (5′-GAGCTC-3′) were also introduced. Zygotes were surgically transplanted into the oviducts of pseudopregnant CD1 females, and newborn mice carrying the Casp8 D387A allele were identified by PCR followed by Sanger sequencing using forward primer 5′-GAAGGAGGCCTCCATCTATGACC-3′ and reverse primer 5′-GCTCTAACCACAGAAATGAGTAAGGAAGC-3′.

Casp3^−/− (18), Casp7^−/− (19), Ripk3^−/− (20), Casp8^−/−Ripk3^−/− (7), Gsdmd^−/− (21), Zbp1^−/− (22), and ASC-Citrine (JAX 030743) (23) mice have been described previously. IAV infection was performed using a sublethal dose (around 50 PFU of IAV PR8 strain) as previously described (24). Animal studies were conducted under protocols approved by the St. Jude Children’s Research Hospital Committee on the Use and Care of Animals.

Primary BMDMs were cultured and infected with IAV (multiplicity of infection [MOI] = 20, unless otherwise indicated) as previously described (24). The caspase-3 inhibitor (Z-DEVD-FMK, Selleckchem S7312) was added at 25 μM 6 h postinfection (hpi) where indicated, and samples were collected at 12 hpi.

Immunoblotting for in vitro samples was performed as previously described (24, 25). For immunoblotting of in vivo samples, day 3 postinfection lung tissue was used. The Abs used were anti–caspase-1 (AdipoGen, AG-20B-0042, 1:2000), anti–caspase-3 (Cell Signaling Technology [CST], #9662, 1:1000), anti-cleaved caspase-3 (CST, #9661, 1:1000), anti–caspase-7 (CST, #9492, 1:1000), anti-cleaved caspase-7 (CST, #9491, 1:1000), anti–caspase-8 (CST, #4927, 1:1000), anti-cleaved caspase-8 (CST, #8592, 1:1000), anti-pMLKL (CST, #37333, 1:1000), anti-MLKL (Abgent, AP14272b, 1:1000), anti-GSDMD (Abcam, ab209845, 1:1000), anti-gasdermin E (GSDME) (Abcam, ab215191, 1:1000), anti–β-actin (Proteintech, 66009-1-IG, 1:5000), and HRP-conjugated secondary Abs (Jackson ImmunoResearch Laboratories, anti-rabbit [111-035-047], 1:5000; anti-mouse [315-035-047], 1:5000). Densitometry of the bands was quantified in ImageJ and normalized to the band density of β-actin. IL-18 release was measured using ELISA kits (Invitrogen, BMS618-3). IL-1β and TNF release were measured by multiplex ELISAs (Millipore MCYTOMAG-70K). Cytokines from bronchoalveolar lavage fluid were measured as previously described (26).

Real-time cell death analysis was performed as previously described (24). Caspase-3/7 activity analysis was monitored following the same protocol using the Incucyte Caspase-3/7 Red reagent (Essen BioScience, 4704).

After the indicated treatment, BMDMs were fixed with 4% paraformaldehyde, permeabilized in PBS with 0.05% Triton, blocked with 1% skim milk in PBS, and stained with the primary Abs conjugated to Alexa Fluor 488, anti-cleaved CASP3 (CST #9669), and anti-ASC (Millipore 2EI-7), followed by staining with the secondary Alexa Fluor 647 anti-mouse IgG Ab and DAPI. Images were acquired using a Marianis spinning disk confocal microscope (Intelligent Imaging Innovations).

GraphPad Prism v9.0 software was used for data analysis. Statistical significance was determined by Student t test between two groups or one-way ANOVA for three or more groups and log-rank test for survival analysis. Mean ± SEM of the data is presented.

IAV infection triggers pyroptosis, apoptosis, and necroptosis (PANoptosis) in BMDMs (9). Given the important role of caspase-8 in balancing PCD pathways during embryogenesis (27, 28), we sought to understand the function of caspase-8 during IAV infection. To first determine whether caspase-8 is involved in regulating PCD pathways during IAV infection, we infected wild-type (WT), Casp8^−/−Ripk3^−/− and Ripk3^−/− BMDMs with IAV and monitored the activation of key molecules involved in PANoptosis. Infected WT cells showed activation of caspase-1, GSDMD and GSDME (pyroptotic markers), caspase-8, caspase-3, and caspase-7 (apoptotic markers), and MLKL (necroptotic marker), whereas infected Casp8^−/−Ripk3^−/− BMDMs displayed minimal or no activation of these markers at the same time points (Supplemental Fig. 1). However, we also observed a notable reduction in the activation of each of these markers in Ripk3^−/− BMDMs infected with IAV (Supplemental Fig. 1), as previously reported (9, 16). Thus, the role of RIPK3 in IAV-induced cell death prevents interpretation of the caspase-8 function in Casp8^−/−Ripk3^−/− cells.

To overcome this obstacle and enable evaluation of the role of caspase-8 in IAV-induced cell death, we next generated Casp8^DA/DA mice, where the caspase-8 autoprocessing function was inhibited by replacing aspartic acid 387 (D387) with alanine (CASP8^DA) (Supplemental Fig. 2), yielding an uncleavable form of the protein. Upon infection with IAV, WT and Casp8^DA/DA BMDMs showed comparable numbers of dead cells and similar rates of viral protein production (Fig. 1A–C). However, further biochemical examination of PCD pathways showed key differences in the activation of pyroptotic, apoptotic, and necroptotic proteins in IAV-infected Casp8^DA/DA BMDMs (Fig. 1C). Autoprocessing of caspase-8 is essential for the downstream activation of caspase-3 and -7 in the extrinsic apoptosis pathway (17) and for the subsequent activation of GSDME (29). Casp8^DA/DA BMDMs showed impaired caspase-3, caspase-7, and GSDME cleavage (Fig. 1C). However, although inflammasome activation (caspase-1 cleavage) remained consistent between WT and Casp8^DA/DA BMDMs, an increase in the GSDMD activation (p30 fragment) was observed in Casp8^DA/DA BMDMs (Fig. 1C, 1D). In addition, phosphorylation of MLKL, a hallmark of necroptosis, was also increased (Fig. 1C). Activated GSDMD mediates the release of processed IL-1β and IL-18 downstream of caspase-1 activation (30). Indeed, the amounts of IL-1β and IL-18, but not TNF, were significantly higher in the supernatant at 12 hpi in Casp8^DA/DA BMDMs compared with WT BMDMs, consistent with the enhanced activation of the GSDMD pore-forming molecule (Fig. 1E). The equivalent cell death, increased GSDMD activation, and higher IL-1β and IL-18 release were consistently observed in Casp8^DA/DA BMDMs compared with WT across different IAV MOIs and at different time points of infection (Supplemental Fig. 3).

To understand the mechanism underlying the increased activation of GSDMD and release of IL-1β and IL-18 in Casp8^DA/DA BMDMs during IAV infection, we investigated the role of other cell death molecules in the cleavage of GSDMD under these conditions. Caspase-3 and -7 can cleave full-length GSDMD at its N-terminal D87 residue to form an inactive p45 fragment in a recombinant system (31). We also observed a p45 fragment in WT BMDMs upon IAV infection, and the formation of this fragment was decreased in Casp8^DA/DA BMDMs (Fig. 1C). Moreover, we observed a p20 GSDMD fragment that was inversely correlated with the formation of the active p30 fragment at 9 and 12 hpi. Dual cleavage of GSDMD by both caspase-1 and caspase-3/7 can generate a p20 fragment, as suggested by a previous report in cells overexpressing both caspase-1 and caspase-3 (32). We therefore hypothesized that the increased GSDMD p30 formation observed in Casp8^DA/DA BMDMs after IAV infection was due to a lack of caspase-3/7–mediated cleavage, as activation of caspase-3/7 was substantially reduced in Casp8^DA/DA BMDMs. To test this possibility, we infected BMDMs from either Casp3^−/− or Casp7^−/− mice with IAV and analyzed the activation of cell death markers and cytokine release. The robust increase in the formation of the active GSDMD p30 fragment and decrease in the formation of the p20 fragment seen in Casp8^DA/DA BMDMs (Fig. 1C) were seen in Casp3^−/− but not in Casp7^−/− BMDMs postinfection (Fig. 2A). Additionally, the release of IL-1β and IL-18 was also significantly increased in Casp3^−/− but not in Casp7^−/− BMDMs, whereas the secretion of the GSDMD-independent cytokine TNF remained comparable between WT, Casp3^−/−, and Casp7^−/− BMDMs (Fig. 2B). Despite the increase in the release of IL-1β and IL-18, phosphorylation of the necroptosis marker MLKL was not changed in Casp3^−/− (Fig. 2A), suggesting that the release of these cytokines is independent of MLKL in Casp3^−/− cells. To further confirm that enzymatic activity of caspase-3 is critical to modulate GSDMD activation, we treated WT and Gsdmd^−/− cells with the caspase-3 inhibitor Z-DEVD-FMK. We observed that inhibition of caspase-3 activity increased the formation of the active GSDMD p30 fragment and reduced the formation of the inactive form p20 (Fig. 2C). These observations suggest an inhibitory role for the caspase-8–caspase-3 axis to keep GSDMD activity in check and limit the release of proinflammatory cytokines processed by caspase-1.

Based on the known cleavage sites of the caspases, the presence of both activated caspase-1 and caspase-3 in the same cell could potentially lead to the formation of the GSDMD p20 fragment. Although concerted activation of caspase-8 and caspase-3 can be observed during AIM2, NLRP3, NLRP1b, and NLRC4 inflammasome activation, in particular under conditions where caspase-1 or GSDMD are deleted (33–37), caspase-1 and -3 are biochemically activated at the same time in a population of cells during various conditions (10). However, the simultaneous engagement of apoptotic and pyroptotic proteins in a single cell has not been visualized and characterized under physiological conditions. To determine whether apoptotic and pyroptotic processes occur in the same cell at the same time during IAV infection, we first stained the ASC speck, the hallmark of inflammasome/pyroptotic process activation, and cleaved caspase-3, to mark cells with activated apoptotic processes (Fig. 3A). At 12 hpi with IAV, nearly all WT BMDMs observed to contain ASC specks were also positive for cleaved caspase-3, whereas Casp8^DA/DA cells containing ASC specks were generally negative for cleaved caspase-3 (Fig. 3A, 3B), consistent with our biochemical observation (Fig. 1C, Supplemental Fig. 3B). To gain further insight into the kinetics of ASC speck formation and the activation of apoptotic caspases, we used BMDMs derived from ASC-citrine mice (23) and performed live-cell imaging in the presence of a fluorogenic caspase-3/7 probe. In line with the confocal observations, IAV-infected cells with ASC specks also frequently showed caspase-3/7 activity (Fig. 3C). The formation of ASC specks generally preceded the emergence of caspase-3/7 activity by less than 10 min (Fig. 3C), suggesting that the onset of pyroptotic and apoptotic pathways occurs together in ASC speck-positive cells during IAV infection. The presence of cleaved caspase-3 in nearly all ASC speck-positive WT cells after IAV infection further supports that caspase-1 and caspase-3 could cleave GSDMD to form the inactive p20 fragment. Together, these findings suggest that the activation of GSDMD is held in check by the caspase-8–caspase-3 axis.

Activation of the NLRP3 inflammasome and IL-1β and IL-18 cytokine release are known to be protective during IAV infection in mice (11–15). Based on the essential role of the caspase-8–caspase-3 axis we observed in regulating GSDMD activation and IL-1β and IL-18 release, we tested the effect of having the uncleavable caspase-8 (Casp8^DA/DA) during IAV infection in vivo. Casp8^DA/DA mice had improved survival and lost less body weight compared with WT mice during IAV infection (Supplemental Fig. 4A, 4B). Lungs from infected Casp8^DA/DA mice at day 3 postinfection also had increased GSDMD activation (Supplemental Fig. 4C, 4D) together with increased IL-1β and IL-18 in bronchoalveolar lavage fluid compared with WT mice (Supplemental Fig. 4E, 4F), consistent with our in vitro finding in BMDMs. These findings further suggest that caspase-8 activation negatively regulates pyroptotic effectors in vivo during infection, providing a potential explanation for viral resistance of Casp3^−/− mice (38).

Overall, our findings showed an anti-inflammatory function of caspase-8 through its autoprocessing during IAV infection. At the cellular level, autoprocessing of caspase-8 leads to activation of caspase-3, which subsequently results in cleavage of GSDMD to produce its inactive form, limiting IL-1β and IL-18 release. The activity of the caspase-8–caspase-3 axis observed in this study may represent a coevolved feature of the virus developed to suppress the host inflammatory response and improve viral propagation, although the specific viral proteins involved in this process require further study. Although mounting evidence shows that the concerted activation of apoptotic and pyroptotic proteins can occur under various conditions (10), our model provides an explanation for the functional consequences from such PCD cross-talk under physiological conditions during viral infection. Improved understanding of this cross-talk and the essential roles of caspase-8 to act as a linchpin in these processes will improve our ability to modulate the immune system to limit viral replication and prevent pathogenesis.

We thank all members of the Kanneganti laboratory for discussions. We also thank R. Tweedell for scientific editing and writing support.

The authors have no financial conflicts of interest.