Yersinia deploys type III-secreted effectors to evade caspase-4 inflammasome activation in human cells

ABSTRACT Yersinia are Gram-negative zoonotic bacteria that use a type III secretion system (T3SS) to inject Yersinia outer proteins into the host cytosol to subvert essential components of innate immune signaling. However, Yersinia virulence activities can elicit activation of inflammasomes, which lead to inflammatory cell death and cytokine release to contain infection. Yersinia activation and evasion of inflammasomes have been characterized in murine macrophages but remain poorly defined in human cells, particularly intestinal epithelial cells (IECs), a primary site of intestinal Yersinia infection. In contrast to murine macrophages, we find that in both human IECs and macrophages, Yersinia pseudotuberculosis T3SS effectors promote evasion of the caspase-4 inflammasome, which senses cytosolic lipopolysaccharide. The anti-phagocytic YopE and YopH, as well as the translocation regulator YopK, were collectively responsible for evading inflammasome activation, in part by inhibiting Yersinia internalization mediated by YadA and β1-integrin signaling. These data provide insight into the mechanisms of Yersinia-mediated inflammasome activation and evasion in human cells and reveal species-specific differences underlying the regulation of inflammasome responses to Yersinia. IMPORTANCE Yersinia are responsible for significant disease burden in humans, ranging from recurrent disease outbreaks (yersiniosis) to pandemics (Yersinia pestis plague). Together with rising antibiotic resistance rates, there is a critical need to better understand Yersinia pathogenesis and host immune mechanisms, as this information will aid in developing improved immunomodulatory therapeutics. Inflammasome responses in human cells are less studied relative to murine models of infection, though recent studies have uncovered key differences in inflammasome responses between mice and humans. Here, we dissect human intestinal epithelial cell and macrophage inflammasome responses to Yersinia pseudotuberculosis. Our findings provide insight into species- and cell type-specific differences in inflammasome responses to Yersinia.

Inflammasome responses to Yersinia have primarily been studied in murine macro phages.However, fundamental differences exist between human and murine inflam masome responses to bacterial infection that may impact pathogenesis and disease severity in the host, including ligand recognition, virulence-driven immune suppression, and inflammasome component expression .Furthermore, beyond phagocytic cells, inflammasomes are expressed in multiple cell types, including intestinal epithe lial cells (IECs) (69), which are a primary site of infection and interact significantly with enteric pathogens during intestinal colonization.While specialized IECs known as M cells, which overlie intestinal immune compartments called Peyer's patches, are considered the primary point of bacterial invasion of the intestinal epithelial barrier (70)(71)(72)(73), Yersinia microcolonies are frequently found outside of Peyer's patches within submucosal pyogranulomas and in close proximity to non-M cell IECs (74).Importantly, many enteric bacterial pathogens use secreted virulence factors to inhibit IEC death in order to preserve their replicative niche during early infection (75,76).Y. enterocolitica uses two anti-phagocytic Yops, YopE and YopH, to evade caspase-1 and NLRP3-depend ent inflammasome activation in human IECs (77).However, the role of other Yops or other inflammasomes in IECs and human cells broadly is unknown.
Here, we find that in contrast to prior findings in murine macrophages (38,44,78,79), the Y. pseudotuberculosis (Yptb) effector YopJ does not induce death of human IECs or macrophages.In contrast, in human cells, Yptb evades caspase-4-and GSDMDdependent cell death and inflammasomedependent cytokine release.Notably, the type III-secreted effectors YopE, YopH, and YopK collectively enable Yptb to evade caspase-4 activation in human cells.Mechanistically, YopE and YopH blockade of bacterial internalization prevented accumulation of intracellular bacteria and correlated with reduced inflammasome activation.Furthermore, the Yptb adhesin YadA and host β1-integrins were required for Yptb internalization into human IECs and subsequent caspase-4 activation in the absence of YopE, YopH, and YopK.These findings demon strate a key role for disruption of actin-mediated internalization in Yersinia evasion of the non-canonical inflammasome in human IECs, and uncover important species and celltypespecific differences in inflammasome responses to Yersinia infection.Caco-2 cells (Fig. S2A).Similarly, WT Caco-2 cells pretreated with the caspase-1 inhibitor YVAD exhibited incomplete loss of IL-18 release after Δ6 Yptb infection compared to dimethylsulfoxide (DMSO) vehicle-treated cells (Fig. S2B), suggesting that caspase-1 contributes to, but is not absolutely required for, Δ6 Yptb-induced inflammasome activation.Caspase-8 is activated in murine macrophages and IECs in response to infection by multiple pathogens, including Yersinia (44,56,82), and can process cas pase-1 substrates such as IL-1β and GSDMD to mediate pyroptosis in the absence of caspase-1 (56,64,83).Notably, siRNA knockdown of CASP8 and pretreatment with the caspase-8 inhibitor IETD resulted in a partial reduction in IL-18 release following Δ6 Yptb infection (Fig. S2C through E), suggesting that like caspase-1, caspase-8 partially contributes to inflammasome activation during Δ6 Yptb infection.However, CASP8 siRNA knockdown in CASP1 −/− Caco-2 cells did not abrogate inflammasome activation during Δ6 Yptb infection (Fig. S2F and G), indicating that additional caspase activation pathways likely mediate cell death during Δ6 Yptb infection of human IECs.
Caspase-4 plays a critical role in inflammasome responses to a variety of enteric pathogens in human IECs (48,49,60,67), and its activation triggers both IEC death and IL-18 release (48).To test whether caspase-4 contributes to inflammasome responses to Δ6 Yptb, we infected two independent single-cell clones of CASP4 −/− Caco-2 cells (49) with either WT or Δ6 Yptb.As expected, WT Yptb infection did not elicit inflammasome activation in either WT or CASP4 −/− Caco-2 cells, and Δ6 Yptb infection of WT Caco-2 cells resulted in robust release of cleaved IL-18 and cell death.Notably, CASP4 deficiency abrogated cleavage and release of active IL-18 and cell death in response to Δ6 Yptb, indicating that caspase-4 is absolutely required for inflammasome responses to Δ6 Yptb infection in human IECs (Fig. 2C through E).Furthermore, CASP4 deficiency in THP-1 macrophages (84,85) resulted in a partial but significant decrease in inflammasome activation (Fig. S3A), suggesting that Yop-mediated evasion of the caspase-4 inflamma some is a conserved evasion mechanism across human cell types.Loss of inflamma some activation in THP-1 macrophages lacking caspase-4 largely mirrored our findings with pretreatment with the pan-caspase inhibitor ZVAD (Fig. S3B), further supporting caspase-4 contribution to inflammasome activation in human macrophages.Caspase-5 contributes to Salmonella-induced inflammasome activation in Caco-2 cells (49).To test whether caspase-5 also contributes to Δ6 Yptb-induced inflammasome activation in Caco-2 cells, we treated WT Caco-2 cells with either a control scramble siRNA or CASP5 siRNA.Knockdown of CASP5 resulted in a partial decrease in IL-18 release (Fig. S3C  and D), suggesting that caspase-5 contributes to, but is not absolutely required for, inflammasome activation.Collectively, these data indicate, for the first time, that Yptb deploys its Yops to evade caspase-4 inflammasome activation in human macrophages and IECs.

GSDMD is required for Δ6 Yptb-induced inflammasome-dependent cytokine release and cell death in human cells
Inflammasome activation leads to cleavage of the protein GSDMD and liberation of its active pore-forming N-terminal domain, leading to its oligomerization into a large ungated pore (19,25,86).Formation of the GSDMD pore in the plasma membrane leads to release of IL-1 family cytokines as well as cell lysis and death, collectively termed "pyroptosis" (87)(88)(89)(90).Caspase-4 cleaves and activates GSDMD via release of its N-terminal domain (25,91).Notably, Δ6 Yptb infection led to robust GSDMD cleavage in WT Caco-2 cells, which was completely absent in CASP4 −/− Caco-2 cells, indicating that caspase-4 is required for GSDMD cleavage in human IECs in response to Yersinia lacking its secreted effectors (Fig. 3A).In contrast, consistent with a lack of observed cell death and IL-18 release, WT Yptb infection did not elicit GSDMD cleavage in either WT or CASP4 −/− IECs (Fig. 3A).To test whether GSDMD is required for cell death and inflammasomedependent cytokine release during Δ6 Yptb infection, we pretreated Caco-2 cells with disulfiram, a chemical inhibitor of GSDMD pore formation (92).Critically, disulfiram treatment completely abrogated IL-18 release and cell death downstream of inflammasome activation in Δ6 Yptb-infected cells compared to infected vehicle control-treated cells (Fig. 3B and C).Consistent with this observation in human IECs, disulfiram treatment of human THP-1 macrophages resulted in an abrogation of IL-1β release (Fig. 3D).Collectively, these results indicate that Yptb Yops enable evasion of the caspase-4 inflammasome and GSDMD activation.
The NLRP3 inflammasome can be activated by a variety of stimuli during infection, including potassium efflux downstream of caspase-4-dependent GSDMD activation and pore formation (5)(6)(7)(8)(9).Previous studies of human IEC responses during Y. enterocolitica infection identified a critical role for the NLRP3 inflammasome (77).However, studies of human IECs found that NLRP3 does not play a role in inflammasome activation in response to Salmonella infection, potentially due to very low levels of NLRP3 expres sion in human IECs as compared to human macrophages (49,93,94).In agreement, WT Caco-2 cells pretreated with a chemical inhibitor of the NLRP3 inflammasome, MCC950, underwent comparable levels of inflammasome activation in response to Δ6 Yptb infection as infected vehicle control-treated Caco-2 cells (Fig. S4A).Caco-2 cells stimulated with LPS and nigericin, a known agonist of the NLRP3 inflammasome, also failed to induce IL-18 release, further suggesting a lack of NLRP3 inflammasome activity in Caco-2 cells.The NAIP/NLRC4 inflammasome, which senses and responds to flagellin and T3SS ligands (10-18) (Fig. S4B), and the inflammasome adaptor protein ASC (adaptor molecule apoptosis-associated speck-like protein containing a CARD) (Fig. S4C) were also dispensable for Δ6 Yptb-induced inflammasome activation, consistent with prior findings that expression of these proteins is very low in human IECs (49).Collectively, these results indicate that the canonical NLRP3 and NAIP/NLRC4 inflammasomes, as well as broadly ASC-dependent inflammasomes, are not activated during Δ6 Yptb infection of Caco-2 cells, and that GSDMD cleavage and activation occur downstream of caspase-4 and are required for cell death and cytokine release.

YopE, YopH, and YopK synergistically enable Yptb to evade human inflamma some responses
Our findings demonstrate that Yptb lacking its entire repertoire of injected effectors induces inflammasome activation in human IECs (Fig. 1).In contrast, Caco-2 cells infected with a panel of Yptb mutant strains each lacking one of the six Yops failed to elicit IL-18 secretion (Fig. 4A), indicating that loss of any individual secreted Yop was insuffi cient to alleviate inflammasome evasion and that several Yops likely have overlapping functions in evading inflammasome activation.Notably, single loss of YopK and YopM failed to induce inflammasome activation (Fig. 4A), despite their roles in evading the NLRP3/caspase-11 and pyrin inflammasomes, respectively, in murine macrophages (40,43,45,46).Y. enterocolitica was previously reported to regulate NLRP3 inflammasome activation in Caco-2 cells by a combination of YopE-and YopH-mediated blockade of integrin signaling (77).Despite a lack of a role for the NLRP3 inflammasome during Δ6 Yptb infection of human IECs (Fig. S4A), Yptb lacking both YopE and YopH (ΔyopEH Yptb) elicited significantly elevated IL-18 release in Caco-2 cells, indicating that combinatorial loss of both YopE and YopH was sufficient to induce inflammasome activation in human IECs (Fig. 4B).Nonetheless, IL-18 levels during ΔyopEH Yptb infection were significantly lower than IL-18 levels released during Δ6 Yptb infection (Fig. 4B), suggesting that additional Yops contribute to inflammasome evasion during Yersinia infection of human IECs.
YopK is a translocated effector that negatively regulates the translocation of other Yops and T3SS components (42,43,45).In murine macrophages, YopK, mediates inflammasome evasion, in contrast to YopE and YopH, which do not contribute to inflammasome evasion (42,43,45).Given that deletion of yopK alone failed to elicit IL-18 release in human IECs (77) (Fig. 4A), we considered that YopK could modulate inflamma some activation in human IECs in a manner that is masked by YopE and YopH, perhaps because these effectors are hyper-translocated in a ΔyopK mutant.Indeed, IL-18 release during ΔyopEHK Yptb infection was substantially elevated compared to ΔyopEH Yptb, and fully recapitulated levels of IL-18 release were observed during Δ6 Yptb infection (Fig. 4B), indicating that YopE, YopH, and YopK function together to enable Yptb evasion of inflammasome activation during infection.Critically, individual loss of YopK, dual loss of YopK and YopE (ΔyopEK), or dual loss of YopK and YopH (ΔyopHK) all failed to induce inflammasome activation.Only in a yopEH mutant background did additional deletion of YopK lead to an increase in IL-18 release (Fig. 4B).Consistent with infection of human IECs, ΔyopEHK infection also triggered higher levels of inflammasome activation than either ΔyopEH or ΔyopK infection in human THP-1 macrophages (Fig. 4C and D).Further more, as with Δ6 Yptb infection, ΔyopEHK-induced inflammasome activation in Caco-2 cells was fully dependent on caspase-4 (Fig. 4E and F).Taken together, these data suggest that YopE, YopH, and YopK act in concert to mediate non-canonical inflamma some evasion in both human IECs and macrophages.These results are distinct from murine macrophages, in which YopE instead activates the pyrin inflammasome, and YopK and YopM contribute to inflammasome evasion (42,43,45,46).

YadA and β1-integrin are necessary for bacterial internalization and inflam masome activation
Because cytochalasin D has pleiotropic inhibitory effects, we sought to evaluate the role of bacterial internalization in triggering inflammasome activation by manipulating components of β1-integrin signaling.As macrophages are naturally phagocytic and do not require β1-integrin signaling to induce bacterial uptake, we focused on human IECs.Yersinia-β1-integrin interactions initiate an intracellular signaling cascade, including phosphorylation of the serine/threonine kinase Akt, that ultimately regulates actindependent cellular changes necessary for bacterial internalization (110)(111)(112)(113)(114)(115).Consistent with previous reports, chemical inhibition of Akt isoforms 1 and 2 diminished Δ6 Yptb internalization in IECs (Fig. S6A) (116).Critically, inflammasome activation induced by Δ6 Yptb was also significantly blunted in the presence of Akt inhibitors (Fig. S6B).
Furthermore, pretreatment with an anti-β1-integrin blocking antibody, previously found to reduce Yersinia entry into cells (117), also significantly decreased bacterial internaliza tion and inflammasome activation during both ΔyopEH and Δ6 Yptb infection (Fig. 5E and  F).Moreover, siRNA knockdown of α5β1 integrin similarly reduced levels of internalized bacteria and inflammasome activation (Fig. S6C through E).Together, these data suggest that disruption of β1-integrin receptor engagement and downstream components of β1integrin signaling prevents Yersinia internalization and limits inflammasome activation in IECs.
The Yersinia adhesin, invasin, binds to β1-integrin to facilitate efficient bacterial internalization into mammalian cells, particularly during early intestinal invasion and translocation through M cells (70,118).Inflammasome activation during Y. enterocoli tica (Ye) infection of human IECs was dependent on invasin-β1-integrin signaling that upregulates IL-18 transcript levels (77).However, Yptb does not express high levels of invasin under T3SS-expressing conditions (117,119,120).Consistently, deletion of invasin had a minor impact on ΔyopEH and Δ6 Yptb-induced internalization and inflammasome activation (Fig. S6F and G) suggesting the possibility that other adhesins may mediate T3SS-dependent inflammasome activation by Yptb.YadA is encoded on the Yersinia virulence plasmid, and its expression is co-regulated with the T3SS and Yops (121).As YadA mediates efficient cellular entry under conditions where invasin expression is repressed (117), we hypothesized that YadA-mediated bacterial internal ization might drive caspase-4-dependent inflammasome activation.Indeed, deletion of YadA in both ΔyopEH and Δ6 Yptb significantly reduced bacterial internalization to levels comparable to WT infection (Fig. 5G) and completely abrogated inflammasome activation (Fig. 5H).Notably, deletion of YadA alone largely phenocopied deletion of both invasin and YadA with respect to internalization and inflammasome activation (Fig. 5G and H), implying that YadA is the primary adhesin responsible for bacterial internalization and inflammasome activation during Yptb infection of human IECs.
The NLRP3 inflammasome is reported to be activated by Y. enterocolitica in human IECs (77).However, we found that NLRP3 was dispensable for inflammasome responses to Yptb in Caco-2 cells (Fig. S4A).NLRP3 expression is very low in Caco-2 IECs and primary human epithelium (49,93,94), and other enteric pathogens or NLRP3 stimuli fail to activate the NLRP3 inflammasome in Caco-2 cells (49) (Fig. S4A).Differences in cell culture conditions or between Y. enterocolitica and Y. pseudotuberculosis may account for the differential role of NLRP3 during infection of human IECs.Moreover, the adaptor protein ASC, which is necessary for the function of the NLRP3, pyrin, and other inflammasomes, was dispensable for inflammasome responses to Δ6 Yptb (Fig. S4C), indicating that canonical inflammasomes are unlikely to be involved in this response.Notably, although YopE-mediated cytoskeletal disruption activates the pyrin inflamma some, which is inhibited by YopM during infection of murine macrophages (40,46), YopM had no impact on inflammasome activation in human IECs even in the presence of YopE (Fig. 4A), potentially due in part to low expression of pyrin in IECs (122,123).We found instead for the first time that Yptb activates the caspase-4 inflammasome in human epithelial cells (Fig. 2C through E, Fig. 4E and F Fig. 5A and B).Caspase-4 is more highly expressed in human IECs than other inflammasome components (49), and a broad range of intracellular enteric pathogens engages or inhibits caspase-4/11 in human and murine IECs (48, 49, 60, 62-67, 75, 76).We also found a partial role for caspase-1 and caspase-8 in inflammasome activation in Caco-2 cells (Fig. S2A and B, Fig. S2D and  E).Caspase-8 and caspase-1 can be recruited to the same inflammasome complexes and have compensatory or sequential roles (44,64,81).Simultaneous knockdown of caspase-1 and caspase-8 still allowed for the release of IL-18 during Δ6 Yptb infection (Fig. S2F), suggesting that they are unlikely to play overlapping roles and may act sequen tially.Whether caspase-8 and caspase-1 are activated downstream of caspase-4 or are acting in a parallel pathway during Yersinia infection of human IECs is unknown.CASP4 −/− THP-1 macrophages also had reduced levels of inflammasome activation during Δ6 Yptb infection (Fig. S3A and B), suggesting that like in human IECs, caspase-4 contributes to Δ6 Yptb-induced inflammasome activation in human macrophages.
In addition to blocking entry into epithelial cells (98, 101-104) (Fig. S5), YopE and YopH also block bacterial internalization into M cells (77), specialized follicular epithelial cells overlying Peyer's patches that are considered the primary site of entry across the intestinal epithelium for Yersinia (70)(71)(72)(73)104).As such, invasin-dependent entry into M cells may occur prior to full T3SS upregulation, thereby allowing the bacteria to cross the epithelium without triggering inflammasome activation.Intriguingly, we found that Δ6 Yptb induces inflammasome activation during both apical and basolateral infection of polarized Caco-2 cells, suggesting a role for non-M cell IECs in inflammasomemedi ated host defense.While integrins are predominantly expressed apically on M cells and basolaterally on non-M cell IECs, intestinal pathogens can interact with basolateral host receptors at cellular junctions either due to depolarization events leading to apical receptor relocation or focal discontinuities leading to luminal exposure of basolateral elements (124)(125)(126).Furthermore, Yptb microcolonies are found outside Peyer's patches within submucosal pyogranulomas in close proximity basolaterally to IECs (74).YopE and YopH may prevent apical or basolateral uptake by non-M cell IECs that would otherwise trigger inflammasomemediated responses within the intestinal epithelium.
Y. enterocolitica YopE and YopH were previously reported to inhibit inflammasome activation by blocking a priming signal downstream of integrin signaling that upregu lates IL-18 transcript levels (77).However, integrin signaling is also crucial for bacterial internalization into IECs.Using a variety of orthogonal approaches, we show here that disruption of integrin signaling reduced inflammasome activation commensurate with reduced levels of Yptb internalization by IECs (Fig. 5; Fig. S6).Additionally, YadA, rather than invasin, was required for Δ6 Yptb-induced inflammasome activation and bacterial internalization (Fig. 5G and H), consistent with Yptb downregulation of invasin and YadA co-expression with the T3SS (117,121).Our findings lead us to propose a model in which YopE/H-dependent blockade of YadA-mediated bacterial uptake by IECs limits delivery of LPS into the host cell, thereby allowing bacteria to evade the non-canonical inflammasome.Because adhesin-integrin binding and signal transduction mediate a variety of cellular outcomes, it is difficult to fully exclude the impact of additional factors on inflammasome activation and evasion facilitated by YopE and YopH.
While YopE and YopH interfere with inflammasome responses during Y. enterocolitica infection of human IECs, no role was previously found for the Y. enterocolitica homolog of YopK and YopQ (77).Consistently, while loss of YopK alone had no effect on inflam masome responses to Yptb, combinatorial loss of YopK, YopE, and YopH recapitulated the levels of inflammasome activation observed with Δ6 Yptb infection, suggesting that YopK enables Yptb to evade a component of an inflammasome pathway masked by YopE and YopH (Fig. 4).In murine macrophages, YopK prevents LPS-mediated inflammasome activation by preventing destabilization of the Yersinia-containing vacuole (42,43).In the absence of YopE, YopH, and YopK, Yersinia may be more readily taken up and exposed to the host cell cytosol, potentially due to vacuolar damage as a result of T3SS-medi ated pore formation (42,127).We found that loss of YopK augments ΔyopEH-induced inflammasome activation in both human macrophages and IECs, potentially pointing to a conserved mechanism by which YopK, YopE, and YopH enable Yptb to evade inflam masome responses across human cell types.Further studies are needed to determine precisely how the T3SS and YopK prevent cytosolic Yersinia LPS exposure and caspase-4 activation.
The synergistic evasion of the caspase-4 inflammasome by YopE, YopH, and YopK, as well as the absence of YopJ-induced cell death, in both human IECs and THP-1 macrophages, are fundamentally distinct from the response to Yptb observed in murine macrophages.Human cells may be intrinsically more resistant to pathogen-induced cell death due to altered expression of prosurvival factors such as cellular FADD-like interleukin-1 beta converting enzyme (FLICE)-inhibitory protein (cFLIP) or A20.Fur thermore, unlike macrophages, IECs are not phagocytic, potentially allowing contamina tion of the host cell cytosol by bacterial LPS to be a more specific indicator of virulence activity in IECs and providing a rationale for increased reliance on caspase-4-mediated responses.Further studies dissecting inflammasome responses between murine and human IECs and macrophages will provide additional insight into cell typespecific and speciesspecific differences in the mechanisms of inflammasome activation by bacterial pathogens.
Overall, our data demonstrate that Yops E, H, and K enable Yersinia pseudotuberculosis to evade caspase-4 inflammasome responses downstream of YadA-β1-integrin signaling in human cells, thereby revealing a major difference in interactions between Yops and inflammasomes in murine and human cells.Our study further highlights the distinct nature of inflammasome responses and bacterial effector activities in different cell types in mice and humans, which provides insight into how inflammasome responses and bacterial virulence activities shape health and disease.

Bacterial strains and growth conditions
Yersinia strains are described in Table S1 in the supplemental material.ΔyopE (102), ΔyopH (105), ΔyopJ (79), ΔyopK (45), ΔyopO (128), ΔyopEH (105) and Δ6 (79) strains were described previously.Δyop ΔyopEHK, ΔyopEK, and ΔyopHK were generated by introduc ing a frameshift mutation of the yopK open reading frame into the ΔyopEH, ΔyopE, and ΔyopH backgrounds, respectively, using a plasmid provided by Dr. James Bliska (129) and an allelic exchange method (130).ΔyopM was generated by introducing an unmarked deletion of the yopM open reading frame into IP2666 using a plasmid provided by Dr. James Bliska and the same allelic exchange method.ΔyopEHΔinv and Δ6Δinv strains were generated by introducing an unmarked deletion of the invasin open reading frame into ΔyopEH and Δ6 strains, respectively, using a plasmid provided by Dr. Joan Mecsas and the same allelic exchange method.ΔyopEHΔyadA, ΔyopEHΔinvΔyadA, Δ6ΔyadA, and Δ6ΔinvΔyadA strains were generated by introducing a kanamycin resistance cassette in place of the yadA open reading frame into ΔyopEH, ΔyopEHΔinv, Δ6, and Δ6Δinv strains, respectively, using a plasmid provided by Dr. Petra Dersch and the same allelic exchange method.Yersiniae were cultured overnight at 26°C with aeration in 2× yeast extract-tryptone (YT) broth.To induce T3SS expression, in the morning, the bacteria were diluted into fresh 2× YT containing 20 mM sodium oxalate and 20 mM MgCl 2 .Bacteria were grown with aeration for 1 hour at 26°C followed by 2 hours at 37°C prior to infection.All cultures were pelleted at 6,000 × g for 3 minutes and resuspended in PBS.Cells were infected at an MOI of 60 unless otherwise indicated, centrifuged at 290 × g for 10 minutes and incubated at 37°C.At 1-hour post-infection, epithelial cells were treated with 20 ng/mL or 100 ng/mL of gentamicin for 6-hour or 2-hour time points, respectively, and macrophages were treated with 100 ng/mL of gentamicin for all time points.Infections proceeded at 37°C for the indicated length of time for each experiment.In all experiments, control cells were mock infected with PBS.

Cell culture conditions
All cells were grown at 37°C in a humidified incubator with 5% CO 2 .

Cell culture of Caco-2 cells
Caco-2 cells (HTB-27; American Type Culture Collection) were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (vol/vol) heat-inactivated fetal bovine serum (FBS), 100 IU/mL penicillin and 100 µg/mL streptomycin.One day prior to infection, Caco-2 cells were incubated with 0.25% trypsin-EDTA (Gibco) diluted 1:1 with 1× PBS at 37°C for 15 minutes to dissociate cells.Trypsin was neutralized with serum-containing medium.Cells were replated in medium without antibiotics in a 24-well plate at a density of 3 × 10 5 cells/well and unprimed prior to infection as we have not observed differential inflammasomedependent cytokine release in IECs during infection with priming (49).All Caco-2 knockout cell lines are described previously (49).

Cell culture of polarized Caco-2 cells
Polarized Caco-2 cells were grown on polycarbonate 3 µM pore size cell culture inserts (Corning 3415) in a 24-well plate.Inserts were coated with collagen coating solution containing 30 µg/mL collagen, 10 µg/mL fibronectin, and 10 µg/mL bovine serum albumin (BSA) in DMEM and incubated for 3 hours.Caco-2 cells were then plated in a growth medium containing Corning MITO+ serum extender (Fisher Scientific CB-50006) on inverted (for basolateral infection) or noninverted (for apical infection) inserts.After 24 hours, the growth medium was replaced with Corning enterocyte differentiation medium (Fisher Scientific 355357) with MITO+ serum extender.The media were replaced daily and following 3 days of incubation in differentiation media, the transepithelial electrical resistance was measured using an Epithelial Volt/Ohm (TEER) Meter (World Precision Instruments) to ensure that resistance was above 250 Ω • cm 2 prior to infection.Infections were administered on the apical or basolateral side of cells as indicated.

Cell culture of THP-1 monocyte derived macrophages
THP-1 macrophages (TIB-202; American Type Culture Collection) were maintained in Roswell Park Memorial Institute 1640 medium (RPMI) supplemented with 10% (vol/vol) heat-inactivated FBS, 0.05 nM β-mercaptoethanol, 100 IU/mL penicillin, and 100 µg/mL streptomycin.Two days prior to infection, THP-1 cells were replated in medium without antibiotics in a 48-well plate at a density of 2 × 10 5 cells/well and incubated with phorbol 12-myristate 13-acetate for 24 hours to allow differentiation into macrophages.Macrophages were primed with 100 ng/mL Pam3CSK4 (Invivogen) for 16 hours prior to bacterial infections in order to upregulate pro-IL-1β transcript levels.

Cell culture of primary hMDMs
Purified human monocytes from deidentified healthy human donors were obtained from the University of Pennsylvania Human Immunology Core.Monocytes were differentiated into macrophages by culturing in RPMI supplemented with 10% (vol/vol) heat-inactivated FBS, 2 mM L-glutamine, 100 IU/mL penicillin, 100 µg/mL streptomycin, and 50 ng/mL recombinant human M-CSF (Gemini Bio-Products) for 6 days.Two days prior to infection, adhered hMDMs were replated in media with 25 ng/mL human M-CSF lacking antibiotics at 1 × 10 5 cells/well in a 48-well plate.hMDMs were then primed with 100 ng/mL Pam3CSK4 (Invivogen) for 16 hours prior to bacterial infection in order to upregulate pro-IL-1β transcript levels.

Cell culture of murine bone marrow derived macrophages
Bone marrow cells were grown in RPMI containing L-cell supernatant, heat-inactivated FBS, penicillin, and streptomycin for 8 days.One day prior to infection, differentiated bone marrow-derived macrophages were replated into 24-well dishes in media lacking antibiotics at a density of 2 × 10 5 cells/well.

ELISAs
Supernatants harvested from infected cells were assayed using ELISA kits for human IL-18 (R&D Systems) and IL-1β (BD Biosciences).

LDH cytotoxicity assays
Supernatants harvested from infected cells were assayed for cytotoxicity by measuring loss of cellular membrane integrity via LDH assay.LDH release was quantified using an LDH Cytotoxicity Detection Kit (Clontech) according to the manufacturer's instructions and normalized to mock-infected (min.cytotoxicity) and 2% triton-treated cells (max.cytotoxicity)

Immunoblot analysis
Cells were replated and infected on serum-free medium to collect supernatant samples.Supernatant samples were centrifuged at 200 × g to pellet any cell debris and treated with trichloroacetic acid (TCA) (25 µL TCA per 500 µL supernatant) overnight at 4°C.The following day, TCA-treated samples were centrifuged at max speed (15,871 × g) for 15 minutes at 4°C and washed with ice-cold acetone.TCA-precipitated supernatant samples and cell lysates were resuspended in 1× SDS-PAGE sample buffer and boiled for 5 minutes.Samples were separated by SDS-PAGE on a 12% (vol/vol) acrylamide gel and transferred to polyvinylidene difluoride Immobilon-P membranes (Millipore).Primary antibodies specific for human IL-18 (MLB International PM014), IL-1β (R&D Systems MAB201), β-actin (4967L; Cell Signaling), GSDMD (G7422; Sigma-Aldrich), and horse radish peroxidase (HRP)-conjugated secondary antibodies anti-rabbit IgG (7074S; Cell Signaling) and anti-mouse IgG (7076S; Cell Signaling) were used.Enhanced chemilu minescence Western blotting substrate or SuperSignal West Femto (Pierce Thermo Scientific) HRP substrate was used for detection.

Quantitative reverse transcription (RT)-PCR analysis
RNA was isolated using the RNeasy Plus Mini Kit (Qiagen) following the manufacturer's instructions.Cells were lysed in 350 µL RLT buffer with β-mercaptoethanol and centri fuged through a QIAshredder spin column (Qiagen).cDNA was synthesized from isolated RNA using SuperScript II Reverse Transcriptase (Invitrogen) following the manufacturer's instructions.Quantitative PCR was conducted with the CFX96 real-time system from Bio-Rad using the SsoFast EvaGreen Supermix with Low ROX (Bio-Rad).For analysis, mRNA levels of siRNA-treated cells were normalized to housekeeping gene HPRT and control siRNA-treated cells using the 2 −ΔΔCT (cycle threshold) method to calculate knockdown efficiency (131).The following primers were used: CASP5 forward: TTCAACACCACATAACGTGTCC CASP5 reverse: GTCAAGGTTGCTCGTTCTATGG CASP8 forward: GTTGTGTGGGGTAATGACAATCT Casp8 reverse: TCAAAGGTCGTGGTCAAAGCC ITGA5 forward: GGCTTCAACTTAGACGCGGAG ITGA5 reverse: TGGCTGGTATTAGCCTTGGGT HPRT forward: CCTGGCGTCGTGATTAGTGAT HPRT reverse: AGACGTTCAGTCCTGTCCATAA

Bacterial uptake enumeration with CFUs
Cells were infected with indicated strains of Yersinia at an MOI of 20.At 1 hpi, cells were treated with 100 µg/mL of gentamicin to kill extracellular bacteria.At 2 hpi, the supernatants were aspirated and cells were lysed with PBS containing 0.5% triton to collect intracellular bacteria.Harvested bacteria were serially diluted in PBS and plated on Luria-Bertani (LB) agar plates containing 2 µg/mL Irgasan.Plates were incubated at 28°C for 2 days and CFUs were counted.

Fluorescence microscopy of intracellular Yersinia
One day before infection, 2 × 10 5 cells/well were plated on glass coverslips in a 24-well plate.Cells were infected with indicated strains of Yersinia constitutively expressing GFP at an MOI of 20.At 2 hpi, cells were washed two times with PBS, fixed with 4% parafor maldehyde for 10 minutes at 37°C and stored overnight at 4°C in PBS.The following day, cells were blocked for 30 minutes at room temperature in blocking solution containing 1% BSA in PBS and incubated for 1 hour at room temperature in blocking solution with the polyclonal anti-Yersinia antibody SB349 diluted 1:1,000 (kindly provided by Dr. James Bliska) (102).AF594-conjugated goat anti-rabbit IgG antibody (A-11012 Thermo Fisher Scientific) was diluted 1:500 in blocking solution and was added to cells and then incubated for 45 minutes at room temperature.Cells were mounted on glass slides with 4′,6-diamidino-2-phenylindole (DAPI) mounting medium (Sigma Fluoroshield).Coverslips were imaged on an inverted fluorescence microscope (IX81; Olympus) and images were collected using a high-resolution charge-coupled-devise camera (FAST1394; QImaging) at a magnification of 60×.Images were analyzed and presented using SlideBook (version 5.0) software (Intelligent Imaging Innovations, Inc.).Intracellular bacteria were scored in unblinded fashion by counting 20 captures per coverslip for coverslips across independ ent triplicate experiments.

Statistical analysis
Prism 9.4.1 (GraphPad software) was utilized for the graphing of data and all statistical analyses.Statistical significance for experiments was determined using the appropriate test and is indicated in each figure legend.Differences were considered statistically significant if the P value was <0.05.

FIG 3
FIG 3 GSDMD is required for Δ6 Yptb-induced inflammasomedependent cytokine release and cell death in human cells.(A) WT or two independent single-cell clones of CASP4 −/− Caco-2 cells were infected with PBS (mock), WT Yptb, or Δ6 Yptb.Lysates and supernatants were collected at 6 hpi and immunoblotted for GSDMD and β-actin.One hour prior to infection (B and C), WT Caco-2 cells or (D) THP-1 macrophages were treated with 30 µM disulfiram or DMSO as a vehicle control.Cells were then infected with PBS (mock), WT Yptb, or Δ6 Yptb.Release of (B) IL-18 or (D) IL-1β into the supernatant and (C) percentage of cytotoxicity normalized to cells treated with 2% triton were measured at 6 hpi.***P < 0.001, ****P < 0.0001 by two-way ANOVA.Shown are averages and error bars representing the standard deviation from at least three pooled experiments.ANOVA, analysis of variance; PBS, phosphatebuffered saline.

8 FIG 4
FIG 4 YopE, YopH, and YopK synergistically enable Yptb to evade human inflammasome responses.(A and B) WT Caco-2 cells or (C and D) WT THP-1 macrophages were infected with PBS (mock) or indicated strain of Yptb.Release of IL-18 (A and B) or IL-1β (C) into the supernatant and percentage of cytotoxicity (D) normalized to cells treated with 2% triton were measured at 6 hpi.(E and F) WT or two independent single-cell clones of CASP4 −/− Caco-2 cells were infected with PBS (mock) or the indicated strain of Yptb.(E) Release of IL-18 into the supernatant and (F) percentage of cytotoxicity normalized to cells treated with 2% triton were measured at 6 hpi.*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 by one-way ANOVA (A-D) or two-way ANOVA (E and F).Shown are averages and error bars representing the standard deviation from at least three pooled experiments.ANOVA, analysis of variance; PBS, phosphatebuffered saline.

FIG 5 (
FIG 5 (Continued) the indicated multiplicity of infection (MOI) and strain of Yptb.(C) Cells were lysed at 2 hpi and bacteria were plated on Yersinia-selective agar to calculate CFUs.(D) Release of IL-18 into the supernatant was measured at 6 hpi.(E and F) WT Caco-2 cells were incubated for 1 hour with or without a 1:50 dilution of antibody against β-1 integrin and then infected with PBS (mock) or the indicated strain of Yptb at an MOI of 20 (E) or 60 (F).(E) Cells were lysed at 2 hpi and bacteria were plated on Yersinia-selective agar to calculate CFUs.(F) Release of IL-18 into the supernatant was measured at 6 hpi.(G and H) WT Caco-2 cells were infected with PBS (mock) or the strain of Yptb at an MOI of 20 (G) or 60 (H).(G) Cells were lysed at 2 hpi and bacteria were plated on Yersinia-selective agar to calculate CFUs.(H) Release of IL-18 into the supernatant was measured at 6 hpi.*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 by two-way ANOVA (A and B) or one-way ANOVA (C-H).Shown are averages and error bars representing the standard deviation from at least three pooled experiments.ANOVA, analysis of variance; hpi, hours post-infection; PBS, phosphatebuffered saline.