Staphylococcus aureus reprograms CASP8 (caspase 8) signaling to evade cell death and Xenophagy

ABSTRACT

Regulated cell death and xenophagy constitute fundamental cellular mechanisms against invading microorganisms. Staphylococcus aureus, a notorious pathogen, can invade and persist within host cells for extended periods. Here, we describe a novel mechanism by which S. aureus subverts these host defenses through the manipulation of the CASP8 (caspase 8) signaling pathway. Upon invasion, S. aureus triggers the assembly of a RIPK3 (receptor interacting serine/threonine kinase 3) complex to induce CASP8 autoprocessing. However, the bacterium inhibits CUL3 (cullin 3)-dependent K63-linked ubiquitination, leading to an atypical activation of CASP8. This non-canonical activation does not initiate the CASP8-CASP3 cascade but instead suppresses RIPK3-dependent necroptosis, a regulated cell death pathway typically activated when apoptosis fails. The resulting non-apoptotic, cleaved CASP8 redirects its enzymatic activity toward cleaving SQSTM1/p62, a selective macroautophagy/autophagy receptor, thus enabling S. aureus to evade antimicrobial xenophagy. The results of this study suggest that S. aureus reprograms the CASP8 signaling pathway from inducing cell death to preserving cell survival and inhibiting xenophagy, a critical strategy that supports its stealthy replication and persistence within host cells.

Abbreviations: CASP3: caspase 3; CASP8: caspase 8; CFU: colony-forming units; CUL3: cullin 3; DUB: deubiquitinating enzyme; MAP1LC3B-II/LC3B-II: microtubule associated protein 1 light chain 3 beta-II; MOI: multiplicity of infection; RIPK1: receptor interacting protein kinase 1; RIPK3: receptor interacting protein kinase 3; S. aureus: Staphylococcus aureus

Introduction

The Gram-positive bacterium Staphylococcus aureus (S. aureus) is a well-recognized human and animal pathogen, implicated in a diverse array of infectious diseases [1,2]. S. aureus can enter, replicate, and persist within various types of non-phagocytic cell types, including epithelial cells, keratinocytes, and osteoblasts [3–6]. This intracellular niche serves as a privileged sanctuary for S. aureus, effectively shielding it from the host’s immune effector mechanisms and the bactericidal activity of antibiotics, ultimately contributing to the establishment of chronic and recalcitrant infections [7].

Once inside host cells, invading bacteria encounter formidable cellular defense mechanisms. Two well-studied host defense pathways crucial for eliminating intracellular pathogens are antimicrobial macroautophagy/autophagy (hereafter referred to as xenophagy) and programmed cell death [8,9]. Xenophagy, an evolutionarily conserved mechanism, targets and eliminates intracellular bacteria while preserving host cell viability [10]. During xenophagy, selective autophagy receptors like SQSTM1/p62 (sequestosome 1) recognize the bacteria, leading to their engulfment by phagophores that mature into autophagosomes leading to subsequent degradation in lysosomes [11–13].

Programmed cell death, in contrast, represents a sacrificial strategy employed by host cells to restrict bacterial replication and trigger inflammatory responses. Apoptosis, a default cell death modality, and necroptosis and pyroptosis are the three main types. Although these cell death programs are distinctly regulated, accumulating evidence suggests they are deeply interconnected, with CASP8 (caspase 8) emerging as a critical control point at the crossroads of these pathways [14]. Upon receipt of proapoptotic stimuli, CASP8 undergoes auto-cleavage and initiates the caspase cascade leading to apoptosis, while simultaneously preventing necroptosis mediated by RIPK3 (receptor interacting protein kinase 3) [15–18]. CASP8 can also drive the switch of cell death to pyroptosis when both apoptosis and necroptosis are blocked [14,19]. In addition, CASP8 has been shown to suppress autophagic induction [20].

For intracellular survival, invading bacteria have to overcome the intricate host defense mechanisms that aim to eliminate the intracellular replication niche of microorganisms. Despite eventual cell death, S. aureus infection often delays the execution of host cell death programs, creating an extended window of opportunity for bacterial replication within the host [21–24]. It has been reported that S. aureus possesses factors that target cell death pathways, modulating apoptosis to enable the survival of infected host cells [25]. Moreover, accumulating evidence suggests S. aureus can subvert the autophagic machinery in both professional and nonprofessional phagocytes [26,27]. Despite our understanding of the genetic factors that contribute to S. aureus virulence [28–30], the specific host pathways exploited by this bacterium to achieve evasion of both cell death and xenophagy remain largely unknown.

The present study investigated the early defensive response of epithelial cells to S. aureus infection. Notably, infected cells remained viable despite robust bacterial replication. We unveil a novel and highly efficient strategy employed by S. aureus to disarm cellular defense mechanisms. This strategy centers on manipulating CASP8. We demonstrate that S. aureus induces auto-cleavage of CASP8 while simultaneously inhibiting its pro-apoptotic activity, without compromising its ability to suppress necroptosis in infected cells. This non-apoptotic CASP8 also functions to cleave SQSTM1, an essential autophagic cargo receptor protein, which facilitates the ability of S. aureus to evade antimicrobial xenophagy. Our findings reveal a previously unknown pro-survival function of CASP8 mediated by its atypical activation, and add new aspects to the intricate interplay between apoptosis and autophagy pathways.

Results

S. aureus infection induces CASP8 cleavage without triggering cell death

Preserving host cell viability, even for a brief period, appears to be a fundamental strategy employed by S. aureus across multiple, if not all, strains to successfully infect nonprofessional phagocytes. This is exemplified by Krut et al. (2003) [4], who found that only a small fraction of 23 clinical isolates caused host cell death, which occurred more than 4 h after infection. Moreover, it has been reported that internalized S. aureus actively replicates before the induction of host cell death [23], further supports the hypothesis that the bacterium is capable of delaying or preventing host cell death to facilitate its intracellular survival and replication. To elucidate the underlying mechanisms, we utilized bovine mammary epithelial cells (Mac-T cells), which are well-documented for their ability to internalize staphylococci [6,21,22]. Two pathogenic strains were employed: methicillin-resistant USA300 and Newbould 305 (associated with chronic infections). Cells were infected with the respective strains for 1 h to allow bacterial internalization, followed by lysostaphin treatment to eliminate extracellular bacteria, and incubated for an additional 6 h. As anticipated, both USA300 and N305 efficiently internalized and replicated within the cells (Figure 1A). However, neither strain induced lytic cell death, as evidenced by limited LDH (lactate dehydrogenase) release (Figure 1B). Consistent with this observation, infection with either strain did not result in alterations to the levels of key markers associated with necroptosis, namely phosphorylated RIPK3 and MLKL (mixed lineage kinase domain like pseudokinase), nor markers indicative of pyroptosis, such as cleaved CASP1 (caspase 1) and mature IL1B (interleukin 1 beta) (Figure S1). Similarly, TUNEL staining revealed no significant increase in apoptosis following infection (Figure 1C). Intriguingly, immunoblot analysis demonstrated a time-dependent increase in cleaved CASP8 fragments (p43/p41 and p18 subunits) in infected cells, while levels of cleaved CASP3 (caspase 3), a critical executioner caspase downstream in the apoptotic pathway, remained unchanged (Figure 1D), suggesting a disconnect in the classical caspase cascade. To validate the effect of S. aureus on CASP8 processing, we ectopically expressed Flag-tagged CASP8 within the cells. Immunoblotting with an anti-Flag antibody confirmed the increased cleavage of CASP8, mirroring the patterns observed with endogenous CASP8 (Figure 1E).

Figure 1.

S. aureus induces CASP8 autoprocessing without inducing cell death. (A) Intracellular bacterial enumeration. Mac-T cells were infected with the indicated S. aureus strains. At the specified time points post-infection, cells were lysed, and the number of intracellular bacteria was determined by colony-forming unit (CFU) assay. Data from a single experiment carried out in technical triplicate representative of three independent experiments are shown. Data are represented as mean ± SD. (B) LDH (lactate dehydrogenase) levels in cell culture supernatants. Mac-T cells were infected with the indicated S. aureus strains and LDH levels were measured at 6 hpi. Data represent LDH release, adjusted for total cellular LDH content. Triton X-100 is used as a positive control. Data from a single experiment carried out in technical triplicate representative of three independent experiments are shown. Data are represented as mean ± SD. (C) Mac-T cells were infected with the indicated S. aureus strains and evaluated for apoptosis via TUNEL staining at 6 hpi. TUNEL labeling (green) indicates apoptosis, and DAPI (blue) marks nucleus. Bar graphs showing the percentage of apoptotic cells. Data from a single experiment carried out in technical triplicate representative of two independent experiments are shown. Data are represented as mean ± SD. (D) Cells were infected with the indicated bacterial strains and harvested at 6 hpi. Whole cell lysates were then subjected to immunoblot analysis. Bar graphs showing densitometric quantifications of CASP8 p18:ACTB/β-actin as well as CASP3 p17:ACTB. Data are presented as mean ± SD (n = 3). Representative results from three independent experiments are shown. *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA. (E) Cells were transiently transfected with Flag-CASP8 for 48 h prior to the infection with the indicated bacterial strains. At the indicated time points post-infection, whole cell lysates were analyzed by immunoblotting using an anti-Flag antibody as probe. Representative results from three independent experiments are shown. #: indicate nonspecific bands. UI: uninfected.

To assess whether our findings from Mac-T cells are applicable to other nonprofessional phagocytes, we conducted additional experiments with MCF-10A (human breast epithelial cell line), 4T1 (murine mammary carcinoma cell line), and RAW264.7 (mouse macrophage cell line) infected with S. aureus N305 for 6 h. Consistent with observations in Mac-T cells, S. aureus N305 induced an increase in the p18 CASP8 fragment in both MCF-10A and 4T1 cells, but did not trigger the processing of CASP3 (Figure S2), suggesting that the response observed in Mac-T cells is not unique to this cell line but rather reflects a common response to S. aureus infection in nonprofessional phagocytes. In contrast, and consistent with a previous study demonstrating that uptake of S. aureus results in CASP3 activation in human monocyte-derived macrophages [31], RAW264.7 cells responded to S. aureus with enhanced CASP8 autoprocessing and a concomitant increase in cleaved CASP3 (Figure S2), indicating that S. aureus is less effective at targeting the CASP8 signaling pathway in macrophages.

In context, our findings suggest that S. aureus employs a targeted mechanism to subvert or delay host cell death in nonprofessional phagocytes during the early phase of infection, specifically by disrupting CASP8 signaling downstream of its autoprocessing to maintain cell viability.

S. aureus induces a RIPK3-dependent CASP8 processing and the cleaved CASP8 is required for necroptosis inhibition

CASP8 processing typically occurs within complex IIa (also known as DISC) containing RIPK1 (receptor interacting serine/threonine kinase 1), FADD (Fas associated via death domain), and CASP8. When CASP8 activity is inhibited, RIPK3 can be recruited downstream of RIPK1 kinase activity to form complex IIb (necrosome) and induce necroptosis [32]. To understand how S. aureus influences CASP8 processing, we investigated the composition of CASP8 associated complexes during infection. Surprisingly, the isolated complex contained RIPK3, FADD, and CASP8, but lacked RIPK1 (Figure 2A), suggesting the formation of an alternative complex IIa. This finding was further substantiated by overexpression of Flag-CASP8 in Mac-T cells (Figure S3). To confirm the pathogen-specificity of this alternative complex IIa formation, we analyzed complex II components in Salmonella Typhimurium infection. In contrast to S. aureus, Salmonella induced a canonical complex IIa containing RIPK1 (Figure 2A).

Figure 2.

S. aureus induces the assembly of RIPK3-FADD-CASP8 complex to promote CASP8 autoprocessing. (A) Co-immunoprecipitation interaction between CASP8 and RIPK3, RIPK1, and FADD. Mac-T cells were infected with bacterial strains as indicated for 1 h. Whole cell lysates were prepared at 6 hpi and immunoprecipitated with anti-CASP8 beads followed by immunoblotting as indicated. Input, immunoblotting of total cell lysates. (B, C and D) Evaluation of the effect of RIPK3 and RIPK1 on CASP8 processing. Cells were transiently transfected with GFP-RIPK3 (B), RIPK3 siRNA (C) or RIPK1 siRNA (D) and subjected to S. aureus N305 infection (MOI: 25). Whole cell lysates were prepared at 6 hpi and the cleaved CASP8 (p18) fragments were detected by immunoblot. Bar graphs showing densitometric quantifications of CASP8 p18:ACTB. Data are presented as mean ± SD (n = 3). Results are representatives from at least three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA.

RIPK3 acts as a scaffolding protein for CASP8 processing independent of its kinase activity [33,34]. Because S. aureus infection upregulated RIPK3 expression without inducing its necroptosis-associated phosphorylation (Figure S1), we therefore hypothesized that RIPK3 contributes to CASP8 processing in infected cells. As anticipated, overexpression of RIPK3 significantly promoted CASP8 processing during S. aureus infection (Figure 2B), while siRNA-mediated RIPK3 knockdown yielded the opposite effect (Figure 2C). RIPK1 knockdown also had no effect on CASP8 cleavage (Figure 2D), further reinforcing the exclusion of RIPK1 in this pathway.

CASP8 activation during apoptosis normally inhibits necroptosis by inactivating key signaling molecules like RIPK1 and RIPK3 [18,35,36]. Recent studies suggest that cleaved CASP8 itself suppresses necroptosis by hindering the formation of the necrosome complex IIb [17]. To investigate whether cleaved CASP8 in our S. aureus model suppressed necroptosis, we treated the cells with z-IETD-fmk, an inhibitor that specifically targets CASP8 cleavage. As expected, z-IETD-fmk treatment resulted in a notable increase in RIPK3 and MLKL phosphorylation in infected cells (Figure 3A), along with increased LDH release (Figure 3B), indicative of necroptosis induction. Knockdown of CASP8 with siRNA mirrored the effects of z-IETD-fmk (Figure 3C,D). These findings suggest that the cleaved, nonapoptotic CASP8 retained the ability to inhibit the induction of necroptosis in infected cells.

Figure 3.

CASP8 autoprocessing functions to suppress cell death by necroptosis. (A and B) Evaluation the effect of z-ietd-fmk on necroptosis in cells infected with S. aureus. Mac-T cells were pre-treated with z-ietd-fmk at 10 μM for 1 h and proceeded to S. aureus infection for another 1 h. At 6 hpi, whole cell lysates were analyzed by immunoblot (A) and cell culture supernatants were collected for LDH analysis (B). Data are presented as mean ± SD (n = 3). *p < 0.05, Student t test. (C) Evaluation of the effect of CASP8 knockdown on necroptosis in cells infected with S. aureus. Mac-T cells were transiently transfected with CASP8 siRNA and infected with S. aureus for 1 h. The cells were subjected to immunoblot (C) and the cell culture supernatants were collected for LDH analysis (D) at the indicated time point post infection. The chart shows the relative level of LDH, with a value of ± SD (n = 3). *p < 0.05, one-way ANOVA. Results are representatives from at least three independent experiments.

Collectively, these results provide evidence that S. aureus induced CASP8 processing within a non-canonical complex IIa through the control of RIPK3.

S. aureus disrupts the ubiquitination of CASP8 by the E3 ubiquitin ligase CUL3

To elucidate the mechanism by which S. aureus evades apoptosis despite inducing CASP8 cleavage, we examined CASP8 ubiquitination, a pivotal regulatory step for its apoptotic function. While K48-linked ubiquitination targets the p18 CASP8 fragment for rapid proteasomal degradation, terminating apoptosis, K63-linked polyubiquitination mediated by CUL3 (cullin 3)-based E3 ubiquitin ligases is essential for CASP8 full activation and processing within SQSTM1-dependent foci, thereby driving commitment to apoptosis [37,38]. Cells expressing Flag-tagged CASP8 were treated with MG132 to preserve ubiquitinated proteins with or without MLN4924 to block CUL3 neddylation, an essential post-translational modification for activating CUL3-based E3 ubiquitin ligase complexes [39], and subjected to immunoprecipitation followed by immunoblot analysis. S. aureus did not affect K48-linked ubiquitination of CASP8 when compared to non-infected cells (Figure 4A), ruling out the possibility that accelerated proteasomal degradation contributes to the compromised apoptotic function of CASP8 in infected cells. Also, S. aureus did not trigger the K63-linked ubiquitination of CASP8, and MLN4924 treatment did not produce any additional impact (Figure 4A). These findings prompted us to investigate the effect of S. aureus on CUL3 neddylation. Surprisingly, we observed that Mac-T cells responded to S. aureus with a marked increase in CUL3 expression by 4 hpi, followed by enhanced neddylation of CUL3 at 6 hpi (Figure 4B). Together, the data allow us to argue that S. aureus prevents the K63-linked ubiquitination of CASP8 mediated by CUL3 following its neddylation modification, thereby impairing host cell death by apoptosis.

Figure 4.

S. aureus disrupts CUL3-mediated CASP8 ubiquitination. (A) Ubiquitination of CASP8 in response to S. aureus. Mac-T cells transfected with flag-CASP8 were pre-treated with 10 μM MG132 and 10 μM MLN4924 for 1 h prior to S. aureus infection. At 6 hpi, whole cell lysates were immunoprecipitated with anti-flag beads followed by immunoblotting as indicated. (B) Neddylation of CUL3 in response to S. aureus. Mac-T cells were infected with S. aureus for 1 h and followed by addition of Baf A1 (10 nM), MG132 (10 μM) or MLN4924 (10 μM). Cells were subjected to immunoblot at the indicated time points post infection. Bar graphs showing densitometric quantifications of NEDD-CUL3:ACTB and CUL3:ACTB. Data are presented as mean ± SD (n = 3). Results are representatives from three independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, one-way ANOVA.UI: uninfected.

Impairment of CUL3-mediated ubiquitination of CASP8 directs its substrate specificity towards

Although SQSTM1 is required for CASP8 activation and apoptosis commitment, there are also reports suggesting that SQSTM1 can be cleaved by CASP8 through unknown mechanisms [40,41]. Given the non-apoptotic function of CASP8 in S. aureus-infected cells, we were interested in determining whether it switches to SQSTM1 cleavage. Indeed, by using immunoblot analysis, we observed a distinct SQSTM1 cleavage fragment (~45 kDa) alongside the full-length protein (62 kDa) in cells infected with either N305 or USA300 S. aureus strains (Figure 5A), confirming the cleavage of SQSTM1. In contrast, neither Salmonella nor Listeria monocytogenes infection resulted in SQSTM1 cleavage (Figure 5A). To validate these findings, we ectopically expressed SQSTM1 in Mac-T cells. Only S. aureus infection, but not Salmonella or Listeria monocytogenes infection, led to the cleavage of exogenous SQSTM1 (Figure 5B). In addition, SQSTM1 cleavage was also absent in S. aureus-infected RAW264.7 cells (Figure S4).

Figure 5.

S. aureus induces cleavage of SQSTM1 through CUL3-dependent ubiquitination of CASP8. (A) Immunoblot analysis of SQSTM1 cleavage. Mac-T cells were infected with the indicated bacterial strains for 1 h and subjected to immunoblot at 6 hpi. (B) Exogenous SQSTM1 is cleaved in S. aureus-infected cells. Mac-T cells were transiently transfected with MYC-SQSTM1 before infection with the indicated bacteria strains for 1 h. Cell lysates were subjected to immunoblot at 6 hpi. (C and D) CASP8 autoprocessing is responsible for SQSTM1 cleavage. Cells were transfected with siRNA targeting CASP8 for 72 h (C) or pre-incubated with CASP8 inhibitor z-ITED-FMK at 10 μM for 1 h (D) prior to S. aureus N305 infection. At 6 hpi, cell lysates were analyzed by immunoblotting with anti-SQSTM1. (E) CASP8 promotes the cleavage of SQSTM1 at Asp329 upon S. aureus infection. Mac-T cells were co-transfected with Flag-CASP8 with or without MYC-SQSTM1, MYC-SQSTM1^D329A and MYC- SQSTM1^D347A for 48 h following endogenous CASP8 knockdown. The cells were then infected with S. aureus N305. At 6 hpi, cell lysates were subjected to immunoblot with anti-myc. (F) CUL3 is required for CASP8-induced cleavage of SQSTM1. Cells were transfected with siRNA targeting CUL3 for 72 h, followed by the transfection with Flag-CASP8 for 48 h. The cells were then infected with the indicated bacterial strains. At 6 hpi, whole cell lysates were immunoprecipitated with anti-Flag beads. MYC-SQSTM1 was prepared from MYC-SQSTM1 overexpressing cells (without S. aureus infection) by immunoprecipitation. The CASP8-containing immunoprecipitate were then incubated with SQSTM1-containing immunoprecipitate for 3 h and the mixture was subjected to immunoblot analysis. Results are representatives from at least three independent experiments. UI: uninfected.

To confirm the involvement of CASP8 in SQSTM1 cleavage in S. aureus-infected cells, we performed CASP8 knockdown. We observed that CASP8 knockdown significantly attenuated SQSTM1 cleavage (Figure 5C). Subsequently, we investigated the necessity of CASP8 autoprocessing for SQSTM1 cleavage. For this purpose, Mac-T cells were pretreated with z-IETD-fmk, which completely abrogated SQSTM1 cleavage in infected cells (Figure 5D). Collectively, these results suggest that the cleaved, non-apoptotic CASP8 accounts for SQSTM1 cleavage.

Previous studies have suggested Asp329 and Asp347 as potential CASP8 cleavage sites within SQSTM1 [40,41]. We sought to confirm CASP8 dependent cleavage and to identify the specific cleavage site of SQSTM1 at the setting of S. aureus infection. To avoid the confounding effects of endogenous CASP8, we knocked down CASP8 using siRNA. The cells were then co-transfected with Flag-tagged CASP8 and MYC-tagged wild-type SQSTM1, or SQSTM1 mutants where Asp329 or Asp347 were replaced with alanine (SQSTM1^D329A and SQSTM1^D347A, respectively). Immunoblot analysis demonstrated the cleavage of SQSTM1 in cells expressing WT SQDTM1 and SQSTM1^D347A, but not SQSTM1^D329A (Figure 5E), suggesting that CASP8 cleaves SQSTM1 at Asp329.

Based on the aforementioned findings, we proposed that the disruption of K63-linked ubiquitination of CASP8 by CUL3 redirects its proteolytic activity toward SQSTM1 cleavage. To test this hypothesis, we transfected Mac-T cells with Flag-tagged CASP8 with or without CUL3 siRNA prior to S. aureus or Salmonella infection. Flag-CASP8 was then immunoprecipitated and incubated in vitro with MYC-tagged SQSTM1 immunoprecipitated from uninfected cells. As anticipated, CASP8 derived from S. aureus-infected cells, but not from Salmonella-infected cells, cleaved SQSTM1 in vitro (Figure 5F). In addition, K63-linked polyubiquitination of CASP8 was observed in Salmonella- but not S. aureus-infected cells (Figure S5). However, the CASP8 derived from cells infected with Salmonella gained the ability to cleave SQSTM1 when CUL3 was knockdown, which led to reduced K63-linked polyubiquitination of CASP8 (Figure 5E). Collectively, these results are consistent with the notion that CUL3-dependent K63-linked ubiquitination acts as a key regulator of CASP8 substrate selectivity.

SQSTM1 cleavage disrupts autophagic targeting of S. aureus

In addition to its role in apoptosis, SQSTM1 plays a critical role in antimicrobial autophagy (also called xenophagy), tagging and directing invading pathogens for lysosomal degradation [42]. It has been well documented that S. aureus can evade selective autophagy mechanism [27], but the underlying mechanisms remain elusive. To investigate the functional consequence of CASP8-mediated SQSTM1 cleavage in the context of xenophagy against S. aureus, we first performed an MAP1LC3B-II/LC3B-II (microtubule associated protein 1 light chain 3 beta)-II turnover assay to assess whether the pathogen induces an autophagic response in our cell model. We observed that S. aureus triggered a robust autophagic response characterized by active LC3B-II turnover (Figure S6). To investigate the functional impact of SQSTM1 cleavage on intracellular bacterial clearance, we reintroduced wild-type (WT) SQSTM1, caspase-resistant SQSTM1 mutant (SQSTM1^D329A), or a mutant mimicking the CASP8-cleaved SQSTM1 fragment (SQSTM1Δ330–440) into cells following endogenous SQSTM1 knockdown. Following transfection, cells were infected with FITC-labeled S. aureus. Intracellular bacterial burden was assessed by monitoring colony-forming units (CFU) counts 4 h after internalization, and the colocalization of SQSTM1 variants with intracellular S. aureus vacuoles was examined using immunofluorescence microscopy. Our results revealed that reintroducing WT SQSTM1 significantly reduced the CFU counts compared to the empty vector control, supporting a role for SQSTM1 in S. aureus clearance. Notably, the CASP8-resistant variant SQSTM1^D329A exhibited ~ 40% greater suppression of intracellular bacterial replication than WT SQSTM1, whereas SQSTM1Δ330–440 failed to restrict S. aureus replication (Figure 6A). These results are consistent with the notion that CASP8-mediated cleavage of SQSTM1 impairs its role in autophagic control of S. aureus. Supporting this, we observed increased recruitment of SQSTM1^D329A to S. aureus vacuoles compared to WT SQSTM1, whereas SQSTM1Δ330–440 failed to form punctate structures and did not associate with intracellular bacteria (Figure 6B). Furthermore, immunoblot analysis showed that SQSTM1^D329A underwent increased degradation via the autophagy-lysosomal pathway relative to WT SQSTM1, whereas the SQSTM1Δ330–440 remained undegraded (Figure 6C). Collectively, these findings suggest that CASP8-mediated SQSTM1 cleavage disrupts bacterial targeting, thereby shielding S. aureus from autophagic degradation.

Figure 6.

Cleaved SQSTM1 loses the function in executing xenophagy. (A) Intracellular bacterial survival and replication evaluated by colony-forming unit (CFU) assay. Cells were transiently transfected with myc-tagged WT SQSTM1, SQSTM1^D329A or SQSTM1Δ330–440 for 48 h following siRNA-based endogenous SQSTM1 knockdown for 72 h. CFU assay was performed at 6 hpi. Data represent the mean ± SD from triplicate culture wells of one representative experiment. ***p < 0.001, one-way ANOVA. (B) Representative microscopic fluorescent images showing co-localization of S. aureus and SQSTM1. Mac-T cells were transfected with siRNA targeting SQSTM1 to knock down endogenous SQSTM1, followed by transiently transfection with mCherry-tagged WT SQSTM1, SQSTM1^D329A or SQSTM1Δ330–440. The cells were then infected with fitc-labelled S. aureus N305. Colocalization of intracellular bacteria and SQSTM1 were observed at 3 hpi. The percentage of cells with SQSTM1 puncta colocalizing with bacteria was enumerated. At least 50 infected cells in 10 fields (20×) were counted for each condition. Data are represented as mean ± SD from triplicate culture wells of one representative experiment. ***p < 0.001, one-way ANOVA. (C) Immunoblots showing the accumulation of full-length and cleaved SQSTM1 in response to S. aureus infection. Mac-T cells were transiently transfected with myc-tagged WT SQSTM1, SQSTM1^D329A or SQSTM1Δ330–440 and infected with N305 in the presence or absence of 10 nM Baf A1. At 6 hpi, whole cell lysates were subjected to immunoblot analysis using anti-myc. Results are representatives from at least three independent experiments.

Discussion

S. aureus has been reported to induce cell death in various host cell types [43–46]. Paradoxically, accumulating evidence suggests that this pathogen is able to survive within host cells [47–49]. This apparent contradiction may be explained by the existence of a temporal window during which S. aureus modulates host cell death pathways to preserve host cell viability, thereby ensuring its own intracellular survival. The duration of this window, however, is likely to vary depending on both the S. aureus strain and host cell type involved. For instance, S. aureus strain UAMS-1 (ATCC 49230) induces osteoblast death as early as 4 h after infection [50], whereas certain clinical S. aureus isolates have been shown to avoid inducing cell death in nonprofessional phagocytes even 24 h after infection [4].

While apoptotic cell death induced by S. aureus has been extensively documented in earlier studies [21,22,45,51], recent evidence has increasingly demonstrated that the bacterium can suppress apoptosis to facilitate its intracellular replication in both nonprofessional and professional phagocytes [25,31,52]. Our findings from S. aureus-infected epithelial cells provide further evidence that the bacterium has evolved strategies to prevent the full activation of the host cell apoptotic program. Specifically, we propose a mechanistic model wherein S. aureus induces CASP8 autoprocessing but suppresses the subsequent cleavage of downstream CASP3. This observation may help explain the findings of a previous study, where S. aureus triggered CASP8 activation as early as 45 min post-infection in osteoblasts, yet only a small fraction of these cells underwent apoptosis by 20 h [46]. In contrast, our results differ from previous data indicating that S. aureus induces a rapid apoptotic death in Mac-T cells, which is associated with the activation of both CASP8 and CASP3 [21,22]. This discrepancy may stem from the serum deprivation conditions in those studies. Serum starvation is known to induce apoptotic cell death [53] and may, therefore, amplify the apoptotic response to S. aureus. Given the ability of S. aureus to suppress caspase cascades, it is reasonable to argue that the infected cells eventually die through a caspase-independent mechanism, as reported in previous studies [54,55].

Notably, in contrast to our findings in the epithelial cell models, we found that S. aureus induced both CASP8 cleavage and CASP3 processing in RAW264.7 cells at 6 hpi. Similarly, increased CASP3 activity was observed in macrophages 5 h after phagocytosis of S. aureus [56]. The precise mechanisms underlying this difference remain unclear; however, the cleavage of SQSTM1 observed in our epithelial cell model, but not in the macrophage model, may explain these discrepancies. It is well established that the formation of SQSTM1 aggresomes is required for the cleavage and activation of effector caspases by CASP8 [37]. Therefore, the mechanisms identified in our epithelial cell model may not apply to professional phagocytes.

We occasionally observed that RIPK3 was required for S. aureus-induced CASP8 autoprocessing, although its kinase activity was not activated in this process, consistent with the notion that RIPK3 acts as a scaffold for CASP8 processing [33,57]. Early studies have shown that death receptor-induced, RIPK3-dependent CASP8 processing relies on the recruitment of RIPK1 to CASP8 cell death complex [33]. Interestingly, in our S. aureus-infected cells, RIPK1 was absent from the CASP8 complex, and its knockdown did not influence CASP8 processing. Therefore, our findings, along with previous studies, suggest that RIPK1 is not essential for RIPK3-dependent CASP8 processing and that this dependency may vary depending on the specific stimuli involved.

Findings from our S. aureus-infected cell model highlighted not only the non-kinase activity of RIPK3 regulated CASP8 autoprocessing but also that CASP8 autoprocessing serves to suppress the pro-necroptosis activation of RIPK3. This is consistent with a recent study showing that CASP8 auto-cleavage is indispensable for both apoptosis induction and necroptosis inhibition in mice, both in vitro and in vivo [17]. However, alternative evidence suggests that, during embryo development, it is the catalytic activity of CASP8, rather than its autoprocessing, that is responsible for mediating the suppression of necroptosis [14–16,58]. These findings indicate that the regulation of necroptosis by CASP8 is context-dependent, with distinct mechanisms operating in different physiological conditions. Overall, our results extend the existing knowledge of CASP8 in the cell fate determination by demonstrating that the cleaved CASP8, even in its non-apoptotic state, can restrain necroptosis.

CUL3 has been reported to interact with the DISC, where it promotes K63-linked ubiquitination of CASP8, resulting in CASP8 full activation and apoptosis induction [37]. Our subsequent investigations unveiled an absence of CUL3-dependent K63-ubiquitination of CASP8 within the infected cells, explaining the observed resistance to apoptosis despite CASP8 autoprocessing. Notably, the E3 ligase activity of CUL3 appeared to be intact, as evidenced by the increased levels of its active neddylated form in response to S. aureus infection, in line with previous studies [59]. The mechanisms behind the inability of activated CUL3 to attach K63 polyubiquitination chains to CASP8 during S. aureus infection remains elusive. One possible explanation is the interference of deubiquitinating enzymes (DUBs) secreted by the pathogen. Indeed, numerous bacterial pathogens possess virulence factors with DUB activity specially targeting K63-linked polyubiquitin chains, or to inhibit the assembly of these chains by compromising host E2 conjugating enzyme function [60]. It is also possible that certain virulence factors of S. aureus activate host DUBs, leading to the removal of K63-linked chains from CASP8. Currently, The DUBs associated with S. aureus and the host DUB responses to intracellular S. aureus remain poorly characterized and require further studies.

Conventional maturation and activation of CASP8 is mediated through the interaction between ubiquitinated CASP8 with selective autophagy receptor SQSTM1, facilitating its sequestration into autophagosomes, where it cleaves and activates downstream effector caspases [37,61]. However, SQSTM1 is reportedly prone to cleavage by CASP8 [40,62]. We have previously reported that S. aureus infection suppresses SQSTM1 phosphorylation [63], suggesting SQSTM1 as a target of this pathogen. In the present study, we further revealed that S. aureus promoted SQSTM1 cleavage, and provided compelling evidence that this process was associated with the impairment of CUL3-dependent K63-linked ubiquitination, which conferred upon CASP8 the capacity to cleave SQSTM1. Given the fact that the CASP8 lost the capacity to activate CASP3 and induce apoptosis in S. aureus-infected cells, our findings allow us to argue that the disruption of CUL3-mediated K63-linked polyubiquitination modulates the substrate selectivity of CASP8 under some conditions, redirecting its activity from effector caspases toward SQSTM1. We also observed that inhibition of CASP8 autoprocessing completely blocked SQSTM1 cleavage in S. aureus-infected cells. This observation, coupled with our findings that RIPK3 contributed to CASP8 autoprocessing in the infected cells, indicates a role of RIPK3 in CASP8-dependent SQSTM1 cleavage. It is possible that RIPK3 forms a complex with SQSTM1 and mediates this effect [62].

The intracellular survival and replication of S. aureus have been previously linked to its ability to escape autophagic clearance in nonprofessional phagocytes, primarily through the inhibition of autophagosome maturation [27,54,64]. However, in our cell model, we observed an enhanced LC3B-II turnover, suggesting that autophagosome maturation proceeds normally. This discrepancy may be attributed to the specific strain of S. aureus used in our study. The N305 strain is representative of a subset of S. aureus strains with an agrI-type regulatory system, characterized by high intracellular survival, persistence and a propensity for mild, chronic infections [65–67]. These characteristics may influence its interaction with the autophagic pathway, differentiating it from S. aureus strains typically associated with acute and severe infections. Considering the pivotal role of autophagy in cellular homeostasis, the observed modulation of this pathway likely represents an additional adaptive strategy, complementing the manipulation of the CASP8 pathway, employed by this subset of S. aureus to promote host cell survival.

The results from our cell model suggest that S. aureus employs CASP8-mediated SQSTM1 processing as a mechanism to evade autophagic clearance. As a selective autophagy receptor, SQSTM1 binds ubiquitinated substrates through its ubiquitin-binding sites, delivering them to the growing phagophore, while simultaneously interacting with LC3B via its LC3-interacting region [10]. The cleavage of SQSTM1 by CASP8 generates a stable fragment (~45 kDa) lacking both the LC3B-interacting region and ubiquitin binding sites [40,62], which may explain the observed reduced recruitment of the wild-type SQSTM1 to S. aureus vacuoles compared to the CASP8-resistant SQSTM1^D329A and the resistance of SQSTM1Δ330–440 to autophagy degradation, consistent with the notion that the cleaved SQSTM1 cannot execute autophagy-related functions in cells [40,68]. In addition, we observed that the CASP8-resistant SQSTM1^D329A mutant significantly improved xenophagic efficiency relative to wild-type SQSTM1, confirming the critical role of an intact SQSTM1 in mediating antibacterial autophagy. In context, our findings suggest that CASP8-dependent cleavage of SQSTM1 constitutes a critical mechanism utilized by S. aureus to evade autophagic elimination, which may also provide an explanation for previous observations in which a fraction of S. aureus appears to escape targeting by SQSTM1 [27].

Notably, it has been demonstrated that the ~ 45 kDa fragment resulting from CASP8-mediated SQSTM1 processing activates the MTOR (mechanistic target of rapamycin kinase) complex 1 (MTORC1) [40], a multiprotein complex well known for its function in cell growth and metabolism. Further investigation is thus warranted to determine whether the SQSTM1-MTORC1 pathway is involved in the strategies employed by invasive S. aureus to maintain host cell survival. Moreover, SQSTM1 has been implicated in regulating the transition from apoptosis to necroptosis [69]. Thus, the observed SQSTM1 cleavage may underlie the suppression of necroptosis in S. aureus-infected cells with compromised apoptotic signaling, a hypothesis that requires further exploration.

In contrast to epithelial cells, we observed that S. aureus did not induce SQSTM1 cleavage in professional phagocytes, aligns with previous studies indicating that SQSTM1 plays a crucial role in autophagic control of S. aureus in neutrophils and macrophages [70,71]. However, emerging evidence suggests a dual role for the autophagy machinery in neutrophils, wherein the LC3B-associated phagocytosis/LAP facilitates the intracellular survival of S. aureus, while the SQSTM1-mediated xenophagic pathway provides protection [71]. These findings, together with our data, underscore the intricate interplay between S. aureus and host autophagic processes and highlight the potential for cell-type-specific regulation of autophagy in immune and nonimmune contexts.

In summary, our study has uncovered a previously uncharacterized strategy employed by S. aureus to evade apoptosis, necroptosis and xenophagy by reprogramming host CASP8 signaling pathways during early infection. Continued research efforts are warranted to unravel the specific bacterial virulence factors responsible for the evasion strategies and to explore novel therapeutic approaches based on these discoveries.

Materials and methods

Reagents and antibodies

The reagents and compounds used were bafilomycin A₁ (Baf A1, Sangon Biotechnology, A601116), z-IETD-fmk (Selleck Chemicals, S7314), lysostaphin (Yuanye Biotechnology, S10039), MG132 (Selleck, S2619), MLN4924 (Selleck, S7109) and fluorescein isothiocyanate (FITC; SolarBio, F8070). The following antibodies were used for western blot analysis: CASP8 (Cell Signaling Technology, 9746), cleaved CASP8 (Asp374; Cell Signaling Technology, 9496), CASP3 (Cell Signaling Technology, 9664T), RIPK1 (HuaBio, ET1701–79), MLKL (HuaBio, ET1601–25), phospho-RIPK3 (Ser227; HuaBio, HA500330), DYKDDDK Tag (Flag; HuaBio, 0912–1), MYC (HuaBio, EM31105), phospho-RIPK1 S166 (ABclonal Biotechnology, AP1230), FADD (ABclonal Biotechnology, A18044), CUL3 (ABclonal Biotechnology, A16455), LC3B (ABclonal Biotechnology, A11280), CASP1 (ABclonal Biotechnology, A0964), RIPK3 (Affinity Biosciences, DF10141), phospho-MLKL (Ser358; Affinity Biosciences, AF7420), SQSTM1/p62 (Affinity Biosciences, AF5384), ubiquitin (Santa Cruz Biotechnology, sc-8017S), ubiquitin K48 (Zen-Bioscience 381517), ubiquitin K63 (Zen-Bioscience 381564) and IL1B (self-prepared).

Bacterial strains and growth conditions

S. aureus strains Newbould 305 (N305) and USA300 (NRS384) was grown on Brain Heart Infusion (BHI) agar plates and Salmonella typhimurium (ATCC14028) and Listeria monocytogenes (ATCC19114) on Lysogeny Broth (LB) agar plates. For infection experiment, colonies from agar plates were incubated at 37°C with rotary shaking at 220 rpm until mid-logarithmic phase of growth (A600 = 1.0 for S. aureus and A600 = 0.6 for others). Colony-forming units (CFU) were counted on LB plates incubated overnight at 37°C. Bacteria were collected by centrifugation, washed with phosphate-buffered saline (PBS) and adjusted to a concentration of 1 х 10⁹ CFU/mL prior to use. In some experiments, staphylococci were labeled with fluorescein isothiocyanate (FITC) as previously described [6].

Cell culture

The Mac-T, MCF-10A, 4T1, and RAW264.7 cell lines were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) containing high glucose supplemented with 10% fetal bovine serum (FBS). Unless otherwise stated, 100 units/mL penicillin G and 100 µg/mL streptomycin sulfate were added to the culture medium. All culture plates are incubated at 37°C in a humid atmosphere containing 5% CO₂.

Plasmids and RNA interference

Bovine CASP8 cDNA was subcloned into the p3×Flag-CMV-14 vector (Sigma-Aldrich, E7908), while SQSTM1 and its mutant cDNAs were subcloned into the pmCherry-C1 vector (Clontech 632524) and pCMV-N-MYC vector (Clontech 631064), respectively. Additionally, CUL3 and RIPK3 cDNAs were subcloned into into the pEGFP-C3 vector (Clontech, 6082–1). The SQSTM1 mutant cDNAs were generated by pCMV-N-MYC-SQSTM1 using a mutagenesis kit (Vazyme, C214). For all cDNAs, a Kozak sequence (GCCACC) was placed before the start codon to optimize the translation initiation. Transient transfection of the constructs was performed by using Lipo8000^TM (Beyotime Biotechnology, C0533). Negative control (NC) siRNA, CASP8 siRNA, RIPK1 siRNA, RIPK3 siRNA and CUL3 siRNA were synthesized by GenePharma Company. siRNAs were transfected with lipofectamine 2000 reagent (Invitrogen 11668019) in accordance with the manufacturer’s instruction. The detailed information for all of the plasmids, primers and siRNAs used in this study is provided in Table S1.

Invasion assays and intracellular bacterial survival assays

Eukaryotic cells were subcultured on culture plates for 24 h before infection. The culture medium was then aspirated and replaced with invasion medium (growth medium without FBS) followed by adding bacteria to the cell culture monolayers at a multiplicity of infection (MOI) of 25. For invasion assays, cells were incubated with bacteria for 1 h, followed by adding 10 µg/mL of lysostaphin to the medium for 15 min to kill extracellular bacteria. Cells were then washed with PBS and lysed by the addition of 0.25% Triton X-100 (Sigma-Aldrich, V900502). Serial dilutions were plated on LB agar for bacterial enumeration. For intracellular bacterial survival assays, cells were maintained in growth medium containing 10% FBS and 10 µg/mL gentamicin after lysostaphin treatment. At the indicated time points, cells were lysed, and bacterial enumeration was performed by plating serial dilutions on LB agar.

TUNEL assay

Cells were fixed in situ with 4% paraformaldehyde in PBS and treated with 0.2% Triton X-100 for permeabilization. The TUNEL assay was conducted using the BrightGreen Apoptosis Detection System (Vazyme, A112), which includes FITC-12-dUTP and recombinant terminal deoxynucleotidyl transferase (TdT) enzyme.

LDH release assay

LDH release was quantified using the LDH Release Assay Kit (Beyotime Biotechnology, C0016) in accordance with the manufacturer’s protocol. In brief, cells were seeded in 96-well plates at a density of 1 × 10⁴ cells per well and infected with S. aureus at a MOI of 25 for 1 h. At 6 hpi, 120 μL of culture medium from each well was collected for the assessment of LDH activity. Absorbance was measured at 490 nm utilizing a K3-TOUCH plate reader (Thermo Fisher Scientific).

Immunoprecipitation and immunoblotting

Total cell lysates were prepared using lysis buffer (20 mm Tris, pH7.5, 150 mm NaCl, 1% Triton X-100) containing protease inhibitor cocktail (Beyotime Biotechnology, P0013B) on ice. For immunoprecipitation, the lysates were incubated with Protein A/G beads (Bimake, B23202) conjugated with the indicated antibodies. For immunoblot analysis, immunoprecipitates or total cell lysates were separated by sodium dodecyl sulfate (SDS)-10% polyacrylamide gel electrophoresis (PAGE), transferred onto polyvinylidene fluoride (PVDF) membranes (Millipore, IPVH00010). The membrane was blocked in a QuickBlock^TM Blocking Buffer (Beyotime Biotechnology, P0252) and immunoblotted with the primary and secondary antibodies. Immunoreactive bands were visualized using electrochemiluminescent (ECL) detection system (FDbio Science, FD8020). Quantification of immunoblot signals was performed using the NIH ImageJ software.

In vitro SQSTM1 cleavage assay

Mac-T cells were transiently transfected with Flag-CASP8. After 72 h, the cells were transfected with or without CUL3 siRNA and subsequently infected with either S. aureus stain N305 or Salmonella for 1 h. At 6 hpi, the CASP8 complex was immunoprecipitated using anti-Flag beads. For the preparation of the SQSTM1 complex, Mac-T cells were transiently transfected with MYC-SQSTM1 and subjected to immunoprecipitation with anti-MYC beads. The SQSTM1 immunoprecipitates were eluted using 1× SDS-PAGE loading buffer at 95°C for 5 min. Following elution, a total of 5 μg of the SQSTM1 complex was mixed with the CASP8 beads and incubated for 3 h at 37°C before proceeding to immunoblot analysis.

Statistical analysis

Data are presented as means ± SD and analyzed by using the two-tailed unpaired Student t test or one-way ANOVA followed by a LSD post hoc test, as appropriate. A P-value less than 0.05 was considered statistically significant. SPSS software version 22.0 (IBM Corp, Armonk, NY, USA) was used for statistical analyses.

Funding Statement

This work was supported by the National Natural Science Foundation of China [Grant No. 32272949], the foundation from Hangzhou Chengxi Sci-tech Innovation Corridor Management Committee, the Key R&D Program of Zhejiang Province [Project No. 2020C02032] and the National Key R&D Program of China [Project No. 2017YFD0502200].

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the Lead contact, Dr. Xun Tan (tanxun@zju.edu.cn), upon reasonable request. Uncropped images for all blots and gels are provided in Data S1.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/15548627.2025.2483887