Cytoplasmic LPS Activates Caspase-11:Implications in TLR4-IndependentEndotoxic ShockJon A. Hagar,1 Daniel A. Powell,2 Youssef Aachoui,1 Robert K. Ernst,2 Edward A. Miao1*

Inflammatory caspases, such as caspase-1 and -11, mediate innate immune detection of pathogens.Caspase-11 induces pyroptosis, a form of programmed cell death, and specifically defends againstbacterial pathogens that invade the cytosol. During endotoxemia, however, excessive caspase-11activation causes shock. We report that contamination of the cytoplasm by lipopolysaccharide (LPS) is thesignal that triggers caspase-11 activation in mice. Specifically, caspase-11 responds to penta- andhexa-acylated lipid A, whereas tetra-acylated lipid A is not detected, providing a mechanism of evasion forcytosol-invasive Francisella. Priming the caspase-11 pathway in vivo resulted in extreme sensitivity tosubsequent LPS challenge in both wild-type and Tlr4-deficient mice, whereas Casp11-deficient micewere relatively resistant. Together, our data reveal a new pathway for detecting cytoplasmic LPS.

Caspases are evolutionarily ancient pro-teases that are integral to basic cellularphysiology. Although some caspases me-

diate apoptosis, the inflammatory caspases-1 and-11 trigger pyroptosis, a distinct form of lytic pro-grammed cell death. In addition, caspase-1 pro-cesses interleukin-1b (IL-1b) and IL-18 to theirmature secreted forms. Caspase-1 is activated bythe canonical inflammasomes, which signal viathe adaptor ASC; NLRC4 and NLRP1a and -1bcan additionally activate caspase-1 directly (1, 2).In contrast to caspase-1, caspase-11 is activated in-dependently of all known canonical inflammasomepathways; the hypothetical caspase-11–activatingplatform has been termed the noncanonical in-flammasome (3). Casp1–/– mice generated from129 background stem cells are also deficient inCasp11 because of a passenger mutation back-crossed from the 129 background into C57BL/6.Caspase-11 is responsible for certain phenotypesinitially attributed to caspase-1, such as shock afterendotoxin challenge (3). The physiologic functionof caspase-11 is to discriminate cytosolic from vac-uolar bacteria (4). In the absence of caspase-11,mice become acutely susceptible to infection bybacteria that escape the phagosome and replicatein the cytosol, such as Burkholderia pseudomalleiandB. thailandensis (4). Caspase-11 also respondsto vacuolar Gram-negative bacteria, albeit with de-layed kinetics (3, 5–7), which may have relevanceto its aberrant activation during sepsis. Althoughthese studies demonstrated both detrimental andprotective roles for caspase-11, the precise nature ofthe caspase-11–activating signal remained unknown.

Because caspase-11 specifically responds tocytosolic bacteria, we hypothesized that detectionof a conserved microbial ligand within the cy-tosol triggers caspase-11. To address this hypoth-esis, we generated lysates of Gram-negative andGram-positive bacteria and transfected them into

lipopolysaccharide (LPS)–primed Nlrc4–/–Asc–/–

Casp11+/+ or Casp1–/–Casp11–/– bone marrow–derived macrophages (BMMs). By comparingthese strains, we can examine caspase-11 activa-tion in the absence of canonical inflammasomedetection of flagellin and DNA (fig. S1). Al-though boiled Gram-negative bacterial lysateswere detected through caspase-11 upon transfec-tion into BMMs, Gram-positive lysates were not(Fig. 1A). Ribonuclease (RNase), deoxyribonu-clease (DNase), lysozyme, and proteinase Kdigestion was sufficient to dispose of canonicalinflammasome agonists but failed to eliminatethe caspase-11–activating factor (or factors) (Fig.1B). We then treated boiled lysates with am-monium hydroxide, which is known to deacylatelipid species (8), and observed that the caspase-11–activating factor was degraded, whereas canon-ical inflammasome agonists persisted (Fig. 1C).

These results suggested LPS as the caspase-11 agonist. Consistent with this hypothesis, BMMsunderwent caspase-11–dependent pyroptosis aftertransfection of ultrapure Salmonella minnesotaRE595 LPS (Fig. 1D). Caspase-11 can promoteIL-1b secretion by triggering the canonical NLRP3pathway (fig. S1) (3). Consistently, IL-1b secre-tion and caspase-1 processing after transfectionof LPSwere also caspase-11 dependent (Fig. 1, Eto G). Moreover, caspase-11 alone promotedpyroptosis (Fig. 1H). In contrast to caspase-1, wewere unable to convincingly visualize caspase-11processing by means of Western blot (Fig. 1, Fand G, and fig. S2A), despite the vast majority ofcells exhibiting pyroptotic morphology as seenby phase microscopy. Although these data do notexclude the possibility that processing of a smallamount of caspase-11 is required for pyroptosis,they do indicate that processing is not a good proxymeasure for rapid caspase-11 activation. This is con-sistent with direct caspase-1 activation by NLRC4,which is not accompanied by processing (9). Theseresults suggest that the presence of LPS in thecytosol is sufficient to trigger caspase-11; however,we cannot rule out the formal possibility that thissignaling arises from a membrane-bound com-partment such as the endoplasmic reticulum or

golgi. Future identification of the noncanonicalinflammasome will permit this determination.

The caspase-11 pathway is not responsive unlessmacrophages are previously stimulated (primed)with either LPS, polyinosinic:polycytidylic acid[poly(I:C)], interferon-b (IFN-b), or IFN-g, whichlikely induces multiple components of the non-canonical inflammasome pathway, includingcaspase-11 (fig. S2B) (4–7, 10). LPS and poly(I:C)prime via Toll-like receptor 4 (TLR4) and TLR3, re-spectively, which both stimulate IFN-b production;IFN-b and IFN-g signaling overlap in their acti-vation of the signal transducers and activators oftranscription–1 (STAT1) transcription factor, whichis critical to caspase-11 activation (5, 7). In order toseparate the priming and activation stimuli ofcaspase-11, we verified that poly(I:C) and IFN-gcould substitute for LPS as priming agents (Fig. 1I).

To corroborate our LPS transfection results,we sought another means to deliver LPS to thecytoplasm. Listeria monocytogenes lyses thephagosome via the pore-forming toxin listeriolysinO (LLO) and, as aGram-positive bacterium, doesnot contain LPS. L. monocytogenes infection didnot activate caspase-11 in BMMs; however, co-phagocytosis of wild-type, but not LLO mutant(∆hly), L. monocytogenes with exogenous LPStriggered pyroptosis, IL-1b secretion, and caspase-1–processing dependent on caspase-11 (Fig. 2, Ato F). Despite this genetic evidence of caspase-11activation, we again did not observe proteolyticprocessing of caspase-11 (Fig. 2, E and F). In con-junction with our previous data indicating thatcaspase-11 discriminates cytosolic from vacuolarGram-negative bacteria (4), these results indicatethat detection of LPS in the cytoplasm triggerscaspase-11–dependent pyroptosis.

Previous studies have shown that another ago-nist, cholera toxin B (CTB), activates caspase-11.However, LPS was present with CTB for theduration of these experiments (3), and caspase-11failed to respond to CTB in the absence of LPS(Fig. 2G). The physiological function of CTB isto mediate the translocation of the enzymaticallyactive cholera toxin A (CTA) into host cells. There-fore, we hypothesized that activation of caspase-11by CTB results from delivery of co-phagocytosedLPS into the cytosol. Under this hypothesis, CTBshould likewise be able to shuttle canonical in-flammasome agonists, which are detected in thecytosol. Indeed, when LPS was replaced withPrgJ, an NLRC4 agonist (11), the pyroptotic re-sponse switched from caspase-11 dependence toNLRC4 dependence (Fig. 2G). Therefore, in theseexperiments CTB is not a caspase-11 agonist, butrather an LPS delivery agent. Whether CTB dis-rupts vacuoles during its use as an adjuvant orwhether complete cholera toxin (CTA/CTB) dis-rupts vacuoles during infection withVibrio choleraremain to be examined.

We next examined the LPS structural deter-minants required for detection through caspase-11and found that the lipid A moiety alone was suf-ficient for activation (Fig. 3A). It is well estab-lished that lipid A modifications enable TLR4

1Department of Microbiology and Immunology and LinebergerComprehensive Cancer Center, University of North Carolina atChapel Hill, Chapel Hill, NC 27599, USA. 2Department of Mi-crobial Pathogenesis, School ofDentistry, University ofMaryland,Baltimore, MD 21201, USA.

*Corresponding author. E-mail: emiao@med.unc.edu

13 SEPTEMBER 2013 VOL 341 SCIENCE www.sciencemag.org1250

REPORTS

on

Janu

ary

26, 2

017

http

://sc

ienc

e.sc

ienc

emag

.org

/D

ownl

oade

d fr

om

evasion, and we therefore hypothesized that cy-tosolic pathogens could evade caspase-11 bymeansof a similar strategy. Indeed,Francisella novicida,a Gram-negative cytosolic bacteria, was not de-tected by caspase-11 (no signal in Nlrc4–/–Asc–/–

BMMs) (Fig. 3B). F. novicida lysates containingDNA activated caspase-1; however, after DNasedigestion the remaining LPS failed to activatecaspase-11, whichwas not restored by temperature-dependent alterations in acyl chain length (Fig. 3C)(12). Aswith L. monocytogenes, co-phagocytosisof F. novicida with exogenous S. minnesota LPSresulted in caspase-11 activation (Fig. 3D). To-

gether, these results suggest that Francisella spe-cies evade caspase-11 bymodifying their lipid A.Francisella species have peculiar tetra-acylatedlipidA unlike the hexa-acylated species of entericbacteria (13). F. novicida initially synthesizes apenta-acylated lipid A structure with two phos-phates and then removes the 4′ phosphate and 3′acyl chain in reactions that do not occur in lpxFmutants (14, 15) (Fig. 3E). Conversion to thepenta-acylated structure restored caspase-11 ac-tivation, whereas other modifications that main-tained the tetra-acylated structures [ f lmK mutantor 18°C growth (12, 16)] did not (Fig. 3F). lpxF

mutant lipid A is not detected by TLR4 (14),suggesting that the TLR4 and caspase-11 path-ways have different structural requirements.

Deacylation of lipid A is a common strategyused by pathogenic bacteria. For example, Yersiniapestis removes two acyl chains from its lipid Aupon transition from growth at 25°C to 37°C (Fig.3G) (17). Consistent with our structural studiesof F. novicida lipid A, caspase-11 detected hexa-acylated lipid A from Y. pestis grown at 25°C, butnot tetra-acylated lipid A from bacteria grown at37°C (Fig. 3H). Together, these data indicate thatcaspase-11 responds to distinct lipid A structures,

Fig. 1. Cytoplasmic LPS triggers caspase-11 activation. (A to H) BMMwere LPS-primed overnight before transfection. [(A) to (C)] BMMs were trans-fected with the indicated bacterial lysates packaged in Lipofectamine 2000(Invitrogen, Carlsbad, CA). Cytotoxicity was determined by lactate dehydro-genase release 4 hours later. Where indicated, lysates were treated with RNase,DNase, lysozyme, and proteinase K (RDLP) (B) or ammonium hydroxide (C).[(D and (E)] BMMs were transfected with ultrapure LPS from S. minnesotaRE595 packaged with DOTAP (Roche, Mannheim, Germany), a liposomaltransfection reagent. Cytotoxicity (D) and IL-1b secretion by means of enzyme-linked immunosorbent assay (E) were determined 4 hours after transfection.

[(F) to (G)] BMMs were stimulated as in (D), and caspase-1 and -11 process-ing by means of Western blot were examined 2 hours after transfection.(H) Immortalized Casp1–/–Casp11–/– BMMs (iBMMs) complemented by ret-roviral transduction of Casp1 or Casp11 were transfected with LPS fromS. minnesota RE595. Cytotoxicity was determined after 4 hours. (I) Macro-phages were primed overnight with LPS (50 ng/mL), poly(I:C) (1 mg/mL),IFN-g (8 ng/mL), or left untreated. Cells were then transfected with LPSfrom S. minnesota RE595, and cytotoxicity was determined 2 hours later.Data are representative of at least three experiments. Error bars indicate SDof technical replicates.

www.sciencemag.org SCIENCE VOL 341 13 SEPTEMBER 2013 1251

REPORTS

on

Janu

ary

26, 2

017

http

://sc

ienc

e.sc

ienc

emag

.org

/D

ownl

oade

d fr

om

and pathogens appear to exploit these structuralrequirements in order to evade caspase-11.

In addition to detection of extracellular/vacuolarLPS by TLR4, our data indicate that an additionalsensor of cytoplasmic LPS activates caspase-11.These two pathways intersect because TLR4primes the caspase-11 pathway. However, Tlr4 –/–

BMMs responded to transfected or CTB-deliveredLPS after poly(I:C) priming (Fig. 4, A to C).Therefore, caspase-11 can respond to cytoplasmicLPS independently of TLR4.

In establishedmodels of endotoxic shock, bothTlr4–/– and Casp11–/– mice are resistant to lethal

challenge with 40 to 54 mg/kg LPS (3, 18, 19),whereas wild-type mice succumb in 18 to 48 hours(Fig. 4D). We hypothesized that TLR4 detectsextracellular LPS and primes the caspase-11pathway in vivo. Then, if high concentrations ofLPS persist, aberrant localization of LPS withinthe cytoplasm could trigger caspase-11, resultingin the generation of shock mediators. We soughtto separate these two events by priming and thenchallenging with otherwise sublethal doses ofLPS. C57BL/6 mice primed with LPS rapidlysuccumbed to secondary LPS challenge in 2 hours(Fig. 4D). TLR4 was required for LPS priming

because LPS-primed Tlr4–/– mice survived sec-ondary LPS challenge (Fig. 4E). To determinewhether alternate priming pathways could sub-stitute for TLR4 in vivo, we primed mice withpoly(I:C) and observed that both C57BL/6 andTlr4–/– mice succumbed to secondary LPS chal-lenge (Fig. 4E). This was concomitant withhypothermia (Fig. 4F), seizures, peritoneal fluidaccumulation, and occasionally intestinal hemor-rhage. In contrast, poly(I:C)–primedCasp11–/–micewere more resistant to secondary LPS challenge(Fig. 4G), demonstrating the consequences ofaberrant caspase-11 activation. Collectively, our

Fig. 2. Listeria and CTB mediatecaspase-11 activation by LPS. (A toF) The indicated macrophages wereprimed with poly(I:C) or LPS and then

infected by L. monocytogenes (MOI 5) in the presence or absence of LPS from S. minnesotaRE595 (1 mg/mL). Cytotoxicity [(A), (C), and (D)], IL-1b secretion (B), or caspase-1 and

caspase-11 processing [(E) and (F)] were examined 4 hours after infection. (G) Poly(I:C)– and Pam3CSK4–primed macrophages were incubated with the indicatedcombinations of CTB (20 mg/mL), LPS from E. coli O111:B4 (1 mg/mL), and PrgJ (10 mg/mL). Cytotoxicity was determined 16 hours later. Data are representativeof three [(A), (D), and (G)] or two [(B), (C), (E), and (F)] experiments. Error bars indicate SD of technical replicates.

Fig. 3. Caspase-11 responds to distinct lipid A structures. (A) Poly(I:C)–primed BMMs were transfected with LPS from S. minnesota RE595 or S. typhimuriumlipid A. Cytotoxicity was determined after 2 hours. (B) Cytotoxicity in LPS-primedBMMs was determined 4 hours after infection with F. novicida (MOI 200). (C) LPS-primed BMMs were transfected with mock or DNase-treated F. novicida lysates.Cytotoxicity was determined 4 hours later. (D) Macrophages were infected as in (B) inthe presence or absence of LPS from S. minnesota RE595. (E) Structural comparisonof lipid A from wild-type F. novicida or the lpxF mutant. Structural changes are

indicated. (F) Poly(I:C)–primed macrophages were transfected with lipid A fromF. novicida grown at 18°C or 37°C or the indicated F. novicida mutants grownat 37°C. Cytotoxicity was determined after 2 hours. (G) Structural comparisonof lipid A from Y. pestis grown at 25°C or 37°C. (H) Poly(I:C)–primed BMMswere infected with L. monocytogenes in the presence of lipid A from Y. pestisgrown at 25°C or 37°C. Cytotoxicity was determined after 4 hours. Data arerepresentative of at least three [(A), (F), and (G)] or two [(B), (C), and (D)]experiments. Error bars indicate SD of technical replicates.

13 SEPTEMBER 2013 VOL 341 SCIENCE www.sciencemag.org1252

REPORTS

on

Janu

ary

26, 2

017

http

://sc

ienc

e.sc

ienc

emag

.org

/D

ownl

oade

d fr

om

data indicate that activation of caspase-11 by LPSin vivo can result in rapid onset of endotoxic shockindependent of TLR4.

Mice challenged with the canonical NLRC4agonist flagellin coupled to the cytosolic trans-location domain of anthrax lethal toxin alsoexperience a rapid onset of shock (20). In thismodel, NLRC4-dependent caspase-1 activationtriggers lethal eicosanoid production via COX-1with similar kinetics to our prime-challengemodel,suggesting convergent lethal pathways downstreamof caspase-1 and caspase-11. Indeed, the COX-1inhibitor SC-560 rescued poly(I:C)–primed micefrom LPS lethality (Fig. 4H).

Although physiological activation of caspase-11 is beneficial in defense against cytosolic bac-terial pathogens (4), its aberrant hyperactivationbecomes detrimental during endotoxic shock. Ourdata suggest that when LPS reaches critical con-centrations during sepsis, aberrant LPS local-ization occurs, activating cytosolic surveillancepathways. Clinical sepsis is amore complex patho-physiologic state, in which multiple cytokines,eicosanoids, and other inflammatorymediators arelikely to be hyperactivated. Eicosanoid mediatorsand other consequences of pyroptotic cellular

lysis (21) should be considered in future ther-apeutic options designed to treat Gram-negativeseptic shock. This underscores the concept thatGram-negative andGram-positive sepsismay causeshock via divergent signaling pathways (22) andthat treatment options should consider these asdiscreet clinical entities.

References and Notes1. J. von Moltke, J. S. Ayres, E. M. Kofoed,

J. Chavarría-Smith, R. E. Vance, Annu. Rev.Immunol. 31, 73–106 (2013).

2. S. L. Masters et al., Immunity 37, 1009–1023 (2012).3. N. Kayagaki et al., Nature 479, 117–121 (2011).4. Y. Aachoui et al., Science 339, 975–978 (2013).5. P. Broz et al., Nature 490, 288–291 (2012).6. P. Gurung et al., J. Biol. Chem. 287, 34474–34483 (2012).7. V. A. K. Rathinam et al., Cell 150, 606–619 (2012).8. A. Silipo, R. Lanzetta, A. Amoresano, M. Parrilli,

A. Molinaro, J. Lipid Res. 43, 2188–2195 (2002).9. P. Broz, J. von Moltke, J. W. Jones, R. E. Vance,

D. M. Monack, Cell Host Microbe 8, 471–483 (2010).10. C. L. Case et al., Proc. Natl. Acad. Sci. U.S.A. 110,

1851–1856 (2013).11. E. A. Miao et al., Proc. Natl. Acad. Sci. U.S.A. 107,

3076–3080 (2010).12. Y. Li et al., Proc. Natl. Acad. Sci. U.S.A. 109, 8716–8721

(2012).13. J. S. Gunn, R. K. Ernst, Ann. N.Y. Acad. Sci. 1105,

202–218 (2007).

14. X. Wang, A. A. Ribeiro, Z. Guan, S. N. Abraham,C. R. H. Raetz, Proc. Natl. Acad. Sci. U.S.A. 104,4136–4141 (2007).

15. D. Kanistanon et al., Infect. Immun. 80, 943–951 (2012).16. D. Kanistanon et al., PLoS Pathog. 4, e24 (2008).17. R. Rebeil, R. K. Ernst, B. B. Gowen, S. I. Miller,

B. J. Hinnebusch, Mol. Microbiol. 52, 1363–1373 (2004).18. S. Wang et al., Cell 92, 501–509 (1998).19. O. Takeuchi et al., Immunity 11, 443–451 (1999).20. J. von Moltke et al., Nature 490, 107–111 (2012).21. M. Lamkanfi et al., J. Immunol. 185, 4385–4392 (2010).22. H. Gao, T. W. Evans, S. J. Finney, Crit. Care 12, 213 (2008).

Acknowledgments: The authors thank V. Dixit for sharingkey mouse strains (Casp11–/– and Nlrc4–/–Asc–/– mice wereprovided under an materials transfer agreement withGenentech). We also thank R. Flavell, M. Heise, and J. Brickey forsharing mice. We thank D. Mao, L. Zhou, and D. Trinh formanaging mouse colonies. The data presented in this manuscriptare tabulated in the main paper and in the supplementarymaterials. This work was supported by NIH grants AI007273(J.A.H.), AI097518 (E.A.M.), AI057141 (E.A.M.), and AI101685(R.K.E.).

Supplementary Materialswww.sciencemag.org/cgi/content/full/341/6151/1250/DC1Materials and MethodsFigs. S1 and S2Tables S1References (23–29)

24 May 2013; accepted 15 July 201310.1126/science.1240988

Fig. 4. LPS detection and rapid induction of shock in primed mice occurindependently of TLR4. (A) BMMs were primed overnight with poly(I:C) andthen transfected with S. minnesota RE595 LPS. Cytotoxicity was determined 2 hourslater. (B and C) Poly(I:C)– and Pam3CSK4-primed macrophages were incubated

with the indicated combinations of CTB (20 mg/mL) and LPS from E. coli O111:B4 (1 mg/mL). Cytotoxicity (B) and IL-1b secretion (C) were determined 16 hourslater. Data are representative of at least three experiments; error bars indicate SD of technical replicates [(A) to (C)]. (D) Survival of mice challenged with theindicated doses of Escherichia coli LPS, or primed with LPS and then rechallenged 7 hours later. Data are pooled from three experiments; n = 9 mice percondition. (E) Survival of mice primed with LPS (400 mg/kg) or poly(I:C) (LMW, 10 mg/kg) and then challenged 7 hours later with LPS (100 ng/kg). Data arepooled from three experiments; n = 7 mice per LPS prime group and n = 8 mice for poly(I:C) prime group. (F) Rectal temperatures of mice in panel (E) afterLPS challenge. Data are representative of three experiments; n = 4 mice per condition. (G) Survival of poly(I:C)–primed mice challenged 6 hours later with LPS(10 ng/kg). Data are pooled from three experiments, n = 11 mice (C57BL/6) or 12 (Casp11–/–). (H) Mice were primed with poly(I:C) and then challenged 6 hourslater with LPS (100 ng/kg) and monitored for survival. One hour before LPS challenge, mice were given 5 mg/kg of COX-1 inhibitor or dimethyl sulfoxide control.Data are pooled from two experiments; n = 11 mice per condition.

www.sciencemag.org SCIENCE VOL 341 13 SEPTEMBER 2013 1253

REPORTS

on

Janu

ary

26, 2

017

http

://sc

ienc

e.sc

ienc

emag

.org

/D

ownl

oade

d fr

om

(6151), 1250-1253. [doi: 10.1126/science.1240988]341Science and Edward A. Miao (September 12, 2013) Jon A. Hagar, Daniel A. Powell, Youssef Aachoui, Robert K. ErnstTLR4-Independent Endotoxic ShockCytoplasmic LPS Activates Caspase-11: Implications in

 Editor's Summary

   presence of TLR4.independently of TLR4. Mice that lack caspase-11 are resistant to LPS-induced lethality, even in thebacteria is able to access the cytoplasm and activate caspase-11 to signal immune responses

(p. 1250) report that the hexa-acyl lipid A component of LPS from Gramnegativeet al.Hagar and (p 1246, published online 25 July)et al.Kayagaki ). However, Kagan(TLR4) (see the Perspective by

The innate immune system senses bacterial lipopolysaccharide (LPS) through Toll-like receptor 4Move Over, TLR4

This copy is for your personal, non-commercial use only.

Article Tools

http://science.sciencemag.org/content/341/6151/1250article tools: Visit the online version of this article to access the personalization and

Permissionshttp://www.sciencemag.org/about/permissions.dtlObtain information about reproducing this article:

is a registered trademark of AAAS. ScienceAdvancement of Science; all rights reserved. The title Avenue NW, Washington, DC 20005. Copyright 2016 by the American Association for thein December, by the American Association for the Advancement of Science, 1200 New York

(print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last weekScience

on

Janu

ary

26, 2

017

http

://sc

ienc

e.sc

ienc

emag

.org

/D

ownl

oade

d fr

om

www.sciencemag.org/cgi/content/full/341/6151/1250/DC1

Supplementary Material for

Cytoplasmic LPS Activates Caspase-11: Implications in TLR4-Independent Endotoxic Shock

Jon A. Hagar, Daniel A. Powell, Youssef Aachoui, Robert K. Ernst, Edward A. Miao*

*Corresponding author. E-mail: emiao@med.unc.edu

Published 13 September 2013, Science 341, 1250 (2013)

DOI: 10.1126/science.1240988

This PDF file includes:

Materials and Methods

Figs. S1 and S2

Tables S1

References (23–29)

Supplementary Materials:

Materials and Methods

Figure S1-S2

Tables S1

Supplementary Materials: Materials and Methods: Mice and in vivo challenges

Wild-type C57BL/6 (Jackson Laboratory), Nlrc4–/–Asc–/– (23, 24), Casp1–/–

Casp11129mt/129mt referred to as Casp1–/–Casp11–/– (25), Casp11−/− (3), Tlr4lps-del/lps-del referred to as Tlr4–/– (Jackson # 007227) mice were used in this study. Mice were housed in a specific pathogen–free facility. All protocols were approved by the Institutional Animal Care and Use Committee at the University of North Carolina at Chapel Hill and met guidelines of the US National Institutes of Health for the humane care of animals.

For study of lethal endotoxia, mice were challenged via intraperitoneal (i.p.) injection with the indicated quantities of LPS in OptiMEM. To prepare LPS challenge doses, serial dilutions of 1mg/mL LPS were carried out, where each dilution was allowed to equilibrate at room temperature for at least 5 minutes before further dilution. In some experiments, mice were primed with LPS (400μg/kg) or poly(I:C) (LMW, 10mg/kg, Invivogen) in DBPS without divalent cations for 6-8 hours. In COX-1 inhibition experiments, mice were given SC-560 (5 mg/kg, Calbiochem) 1 hour prior to LPS challenge. In some experiments, rectal temperatures were recorded at the indicated time points using a MicroTherma 2T thermometer (Braintree Scientific) with a lubricated RET-3 probe. The presence of blood in the intestinal lumen was confirmed using the Sure-Vue Fecal Occult Blood Test (Guaic test). Mice were considered moribund when unable to right themselves. Bacterial growth conditions

All bacterial strains were grown in Luria-Bertani medium (LB) overnight at 37°C, except for L. monocytogenes and F. novicida, which were grown in BHI at 30°C or BHI + 0.1% cystine at 37°C, respectively. Bacterial lysates and lipid A preparation For generation of bacterial lysates, 1mL of overnight culture normalized to an OD600 of 1.0 was pelleted, resuspended in 50μL PBS, and boiled 10 minutes. Insoluble lysate components were pelleted at 20,000 x g for 10 minutes, and the supernatant was collected. In some experiments, supernatants were treated with DNase (10μg/mL), RNase (10μg/mL), Proteinase K (100μg/mL), and/or lysozyme (100μg/mL) for 3 hours, or adjusted to contain 30% ammonium hydroxide, incubated overnight at room temperature, air-dried, and resuspended in PBS.

LPS was extracted essentially as previously described (Westphal O, Journal of medicinal and pharmaceutical chemistry 1961). Briefly, 500ml overnight cultures were pelleted at 5000 xg for 10 minutes. Bacterial pellets were then resuspended in 40mL of sterile water and 90% aqueous phenol at a 1:1 ratio. Mixtures were incubated in a rotating oven at 65°C for 1 hour.

Samples were then cooled on ice for 5 minutes and centrifuged at 3000 xg for 30 minutes. The aqueous layer was transferred to a new tube. 20mL of sterile water was then added to the phenol layer, mixed, and the aqueous layer extracted a second time. The aqueous fractions were combined and then dialyzed for 24 hours against water in 1000 MWC tubing. After dialysis, insoluble material was removed by centrifugation at 17,000 xg for 20 minutes. The LPS sample was then frozen and lyophilized overnight. The dried sample was resuspended in 10mL of 10mM Tris, pH 8.0 containing 25μg/mL RNaseI and 100μg/mL DNaseI and incubated for 2 hours at 37°C. To digest protein, Proteinase K was added at 100μg/mL and the incubation was continued at 37°C for 2 additional hours. 5mL of water-saturated phenol was then added and the mixture was centrifuged at 3000 xg for 30 minutes. The aqueous layer was removed and dialyzed for 24 hours against water in 1000 MWC tubing. The final LPS sample was then frozen and lyophilized overnight.

To liberate lipid A from the core and O-antigen of LPS, 500μL of 10mM sodium acetate + 1% SDS was added to the dried LPS sample. This mixture was incubated at 100°C for 1 hour. This sample was then frozen and lyophilized overnight. The dried sample was resuspended in 100 μL of water along and then added to 1mL of acidified ethanol (100 μL 4N HCl in 20 mL of 95% ethanol). This sample was then centrifuged at 5000 xg for 5 minutes and the supernatant discarded. The sample was then washed with 1mL of 95% ethanol 3 times (centrifugation at 5000 x g for 5 minutes between washes, supernatant discarded) to remove residual SDS. Following the last centrifugation step, the remaining pellet was resuspended in 500μL of water and lyophilized overnight.

Contaminating lipids were removed following the protocol of Folch et al. (Folch J, JBC, 1957) Briefly, lipid A sample was resuspended 1-2% in chloroform/methanol (2:1). Samples were vortexed for 1 minute and then centrifuged at 10,000 xg for 10 minutes at 4°C. Supernatants were removed and then this extraction step was repeated. The remaining pellet was suspended in 500μL of water and lyophilized overnight.

Contaminating proteins were removed following the protocol of Hirschfeld et al. (Hirschfeld, JI, 2000). Briefly, 5mg of lipid A was resuspended in 1mL of water containing 0.2% triethylamine (TEA). Deoxycholate (DOC) was added to a final concentration of 0.5% and followed addition of 500μL of water-saturated phenol. Samples were vortexed for 5 minutes and then the phases were allowed to separate at room temperature for 5 minutes. Samples were cooled on ice and then centrifuged at 10,000 xg for 2 minutes. The aqueous layer was removed to a new tube and the phenol layer was reextracted with 500μL of water containing 0.2%TEA and 0.5% DOC. The aqueous layers were pooled and adjusted to contain a final concentration of 75% ethanol and 30mM sodium acetate. The sample was then allowed to precipitate at -20°C for 1 hour. Precipitates were centrifuged at 10,000 xg for 10 minutes at 4°C, washed in 100% ethanol and then air-dried. The precipitates were then resuspended in 500μL of water containing 0.2% TEA and lyophilized overnight. Liposome preparation and transfections Bacterial lysate transfection complexes were generated as follows: For transfection of 5x104 BMMs, 0.25μL lysate suspended in 25μL OptiMEM and 0.25μL Lipofectamine 2000 suspended in 25μL were allowed to equilibrate at room temperature for 5 minutes. Suspensions were mixed and incubated at room temperature for 20 minutes. For LPS and lipid A transfection complexes, 75ng of LPS or lipid A suspended in 2μL OptiMEM and 375ng DOTAP suspended

in 2μL OptiMEM were allowed to equilibrate at room temperature for 5 minutes. Suspensions were then mixed and incubated for 30 minutes at room temperature. Reaction volumes were then brought up to 50μL with OptiMEM. Macrophage culture, treatment, and analysis of inflammasome activation

BMMs were prepared as described (26). For experiments, macrophages were seeded into

96-well tissue culture treated plates at a density of 5x104 cells/well. When indicated, macrophages were primed with lipopolysaccharide (50 ng/ml), poly(I:C) (1μg/mL) or IFN-γ (8 ng/ml ) overnight. For infections, culture media was replaced with OptiMEM and then bacteria were added to BMMs at MOI 200 (F. novicida) or MOI 5 (L. monocytogenes), centrifuged for 5 min at 200 x g, and incubated at 37°C. After 1 hour, extracellular bacterial growth was stopped by addition of 20 μg/ml gentamicin. In some experiments, 1μg/mL LPS or lipid A was included during the infection. For transfections, culture media was replaced with OptiMEM and then LPS and lysate transfection complexes were added to BMMs and centrifuged for 5 min at 200 x g. For CTB experiments, macrophages were primed for 5h with poly(I:C) and Pam3CSK4 (100μg/mL). Media was replaced with OptiMEM containing poly(I:C), Pam3CSK4, and combinations of CTB, LPS, and PrgJ as indicated in the figure legends. Supernatant samples were collected at the indicated time points. Cytotoxicity was defined as the percentage of total lactate dehydrogenase released into the supernatant and was determined as described (27). IL-1β secretion was determined by enzyme-linked immunosorbent assay (ELISA) (R&D Systems). PrgJ was purified as described (11). S. minnesota LPS was obtained from List Biologicals. E. coli O111:B4 LPS was from Sigma. Complementation and knockdown of Casp1 and Casp11

Bone marrow derived macrophages were immortalized (iBMM) as described (28). For complementation of Casp1 and Casp11 in Casp1–/–Casp11–/– iBMMs, macrophages were transduced with pMXsIP (29) derived retrovirus carrying Casp1 or Casp11.

Western blots

Total protein from lysates and supernatants of 5 x 104 cells was analyzed by Western blot. Caspase-11 content and processing was analyzed using anti-caspase-11 antibody (17D9, Novus) diluted 1:500. Caspase-1 processing was determined using an anti-caspase-1 antibody (clone 4B4, Genentech). Blots were stripped and equivalent loading of protein was ensured by Western blot using anti-β-actin HRP antibody (Cat. # 20272, AbCam) diluted 1:20,000. Supplementary Figure Legend Fig. S1. Schematic of canonical and non-canonical inflammasome pathways. The canonical inflammasomes (shown here: NLRP3, AIM2, and NLRC4) activate caspase-1, whereas a hypothetical non-canonical inflammasome activates caspase-11. Both caspase-1 and caspase-11 initiate pyroptosis. Via its CARD domain, NLRC4 can directly bind caspase-1 to mediate its

activation. In contrast, Pyrin domain containing inflammasomes, such as NLRP3 and AIM2, cannot directly bind caspase-1; rather, they interact indirectly via the adaptor protein ASC, which polymerizes into a structure called the ASC focus. Caspase-1 processes IL-1β and IL-18 to their mature secreted forms in the ASC focus. In contrast, caspase-11 does not process IL-1β alone, rather it activates the NLRP3/ASC/caspase-1 pathway (arching arrow). In the absence of Nlrc4 and Asc, macrophages are unable to signal via any known canonical inflammasome pathways; however, they retain the ability to activate caspase-11, which functions independently of the canonical inflammasomes. Casp1–/–Casp11–/– macrophages, on the other hand, are deficient in all inflammasome pathways. Fig. S2. Upregulation of caspase-11 potentiates pyroptosis but not proteolytic processing. (A, left) Overexposure of caspase-11 western blot from Fig. 1F reveals minor processing of caspase-11 in LPS primed C57BL/6 BMMs 2 hours after transfection with LPS. (A, right) Nlrc4–

/–Asc–/– BMMs (2 x 105 cells in 5μL endotoxin free water + protease inhibitor) were subjected to three freeze thaw cycles. Insoluble components were pelleted and supernatants were then incubated for 5 minutes at 37°C. SDS-sample buffer was added to the reaction and then processed for western blotting. Significant spontaneous processing was observed in the lysates. Interpretation of (A, left): Although the minor processed band observed after LPS transfection could represent caspase-11 processing in response to cytoplasmic LPS, it could also be due to spontaneous processing that occurs in the media after pyroptotic cell lysis. (B) Upregulation of full-length caspase-11 in unprimed BMMs, or BMMs primed overnight with LPS (50ng/mL), poly(I:C) (1000ng/mL), or IFN-γ (8ng/mL), as revealed by western blot.

Supplementary Table: Name of Strain Designation Notes Reference S. typhimurium ATCC 14028s wild type www.atcc.org B. thailandensis E264 wild type www.atcc.org P. aeruginosa PAO1 wild type Joseph Mougous UPEC ATCC 19110 wild type www.atcc.org C. rodentium ATCC 51459 wild type www.atcc.org S. flexneri M90T wild type Alan Aderem L. monocytogenes 10403S Dan Portnoy L. monocytogenes Δhly

Dan Portnoy

L. innocua www.atcc.org B. subtilis Tony Richardson S. aureus ATCC25923 wild type www.atcc.org S. pyogenes MGAS-5005 wild type Tony Richardson E. faecalis V583 wild type Tony Richardson F. novicida U112 wild type www.nwrce.org F. novicida lpxF TBE264 lpxF mutant, hexa-

acyl lipid A (14)

F. novicida flmK TBE229 flmK mutant, tetracyl lipid A

(16)

Y. pestis KIM6 WT Joe Hinnebusch Table S1. Strain list.

NLRC4

NLRP3

ASC

AIM2

CASP1p10/p20

pro-IL-1!pro-IL-18

IL-1!IL-18

Pyroptosis

ASCFocus CASP11

Flagellin

T3SSRod

Multipleagonists

DNA

CASP1

CARDPyrinNODLRRCaspaseHIN200

Domain KeyCytosolicGram (-)Bacteria

Hypotheticalnon-canonicalinflammasome

LPS

Figure S1.Schematic of canonical and non-canonical inflammasome pathways. The canon-ical inflammasomes (shown here: NLRP3, AIM2, and NLRC4) activate caspase-1, whereas ahypothetical non-canonical inflammasome activates caspase-11. Both caspase-1 and caspase-11 initiate pyroptosis. Via its CARD domain, NLRC4 can directly bind caspase-1 to mediate itsactivation. In contrast, Pyrin domain containing inflammasomes, such as NLRP3 and AIM2, can-not directly bind caspase-1; rather, they interact indirectly via the adaptor protein ASC, whichpolymerizes into a structure called the ASC focus. Caspase-1 processes IL-1! and IL-18 to theirmature secreted forms in the ASC focus. In contrast, caspase-11 does not process IL-1! alone,rather it activates the NLRP3/ASC/caspase-1 pathway (arching arrow). In the absence of Nlrc4and Asc, macrophages are unable to signal via almost all known canonical inflammasome path-ways; however, they retain the ability to activate caspase-11, which functions independently ofthe canonical inflammasomes. Casp1–/–Casp11–/– macrophages, on the other hand, are deficientin all inflammasome pathways.

IFN-"poly(I:C)LPS

STAT1

IFN-!

p26 -

p43 -p38 -

casp11

actin -

C57BL/6

2h exposure5x104 cells

5 minute exposure2x105 cells

Nlrc4!/!Asc!/!

p43 -p38 -

casp11

actin -

prime:

C57BL/6

A B

Fig. S2. Upregulation of caspase-11 potentiates pyroptosis but not proteolytic processing.(A, left) Overexposure of caspase-11 western blot from Fig. 1F reveals minor processingof caspase-11 in LPS primed C57BL/6 BMMs 2 hours after transfection with LPS. (A,right) Nlrc4– /–Asc– /– BMMs (2 x 105 cells in 5µL endotoxin free water + protease inhibitor)were subjected to three freeze thaw cycles. Insoluble components were pelleted andsupernatants were then incubated for 5 minutes at 37°C. SDS-sample buffer was addedto the reaction and then processed for western blotting. Significant spontaneous process-ing was observed in the lysates. Interpretation of (A, left): Although the minor processedband observed after LPS transfection could represent caspase-11 processing in responseto cytosolic LPS, it could also be due to spontaneous processing that occurs in the mediaafter pyroptotic cell lysis. (B) Upregulation of full-length caspase-11 in unprimed BMMs, orBMMs primed overnight with LPS (50ng/mL), poly(I:C) (1000ng/mL), or IFN-" (8ng/mL), asrevealed by western blot.

References and Notes 1. J. von Moltke, J. S. Ayres, E. M. Kofoed, J. Chavarría-Smith, R. E. Vance, Recognition of

bacteria by inflammasomes. Annu. Rev. Immunol. 31, 73–106 (2013). doi:10.1146/annurev-immunol-032712-095944 Medline

2. S. L. Masters, M. Gerlic, D. Metcalf, S. Preston, M. Pellegrini, J. A. O’Donnell, K. McArthur, T. M. Baldwin, S. Chevrier, C. J. Nowell, L. H. Cengia, K. J. Henley, J. E. Collinge, D. L. Kastner, L. Feigenbaum, D. J. Hilton, W. S. Alexander, B. T. Kile, B. A. Croker, NLRP1 inflammasome activation induces pyroptosis of hematopoietic progenitor cells. Immunity 37, 1009–1023 (2012). doi:10.1016/j.immuni.2012.08.027 Medline

3. N. Kayagaki, S. Warming, M. Lamkanfi, L. Vande Walle, S. Louie, J. Dong, K. Newton, Y. Qu, J. Liu, S. Heldens, J. Zhang, W. P. Lee, M. Roose-Girma, V. M. Dixit, Non-canonical inflammasome activation targets caspase-11. Nature 479, 117–121 (2011). doi:10.1038/nature10558 Medline

4. Y. Aachoui, I. A. Leaf, J. A. Hagar, M. F. Fontana, C. G. Campos, D. E. Zak, M. H. Tan, P. A. Cotter, R. E. Vance, A. Aderem, E. A. Miao, Caspase-11 protects against bacteria that escape the vacuole. Science 339, 975–978 (2013). doi:10.1126/science.1230751 Medline

5. P. Broz, T. Ruby, K. Belhocine, D. M. Bouley, N. Kayagaki, V. M. Dixit, D. M. Monack, Caspase-11 increases susceptibility to Salmonella infection in the absence of caspase-1. Nature 490, 288–291 (2012). doi:10.1038/nature11419 Medline

6. P. Gurung, R. K. Malireddi, P. K. Anand, D. Demon, L. V. Walle, Z. Liu, P. Vogel, M. Lamkanfi, T. D. Kanneganti, Toll or interleukin-1 receptor (TIR) domain-containing adaptor inducing interferon-β (TRIF)-mediated caspase-11 protease production integrates Toll-like receptor 4 (TLR4) protein- and Nlrp3 inflammasome-mediated host defense against enteropathogens. J. Biol. Chem. 287, 34474–34483 (2012). doi:10.1074/jbc.M112.401406 Medline

7. V. A. K. Rathinam, S. K. Vanaja, L. Waggoner, A. Sokolovska, C. Becker, L. M. Stuart, J. M. Leong, K. A. Fitzgerald, TRIF licenses caspase-11-dependent NLRP3 inflammasome activation by gram-negative bacteria. Cell 150, 606–619 (2012). doi:10.1016/j.cell.2012.07.007 Medline

8. A. Silipo, R. Lanzetta, A. Amoresano, M. Parrilli, A. Molinaro, Ammonium hydroxide hydrolysis: a valuable support in the MALDI-TOF mass spectrometry analysis of Lipid A fatty acid distribution. J. Lipid Res. 43, 2188–2195 (2002). doi:10.1194/jlr.D200021-JLR200 Medline

9. P. Broz, J. von Moltke, J. W. Jones, R. E. Vance, D. M. Monack, Differential requirement for Caspase-1 autoproteolysis in pathogen-induced cell death and cytokine processing. Cell Host Microbe 8, 471–483 (2010). doi:10.1016/j.chom.2010.11.007 Medline

10. C. L. Case, L. J. Kohler, J. B. Lima, T. Strowig, M. R. de Zoete, R. A. Flavell, D. S. Zamboni, C. R. Roy, Caspase-11 stimulates rapid flagellin-independent pyroptosis in response to Legionella pneumophila. Proc. Natl. Acad. Sci. U.S.A. 110, 1851–1856 (2013). doi:10.1073/pnas.1211521110 Medline

11. E. A. Miao, D. P. Mao, N. Yudkovsky, R. Bonneau, C. G. Lorang, S. E. Warren, I. A. Leaf, A. Aderem, Innate immune detection of the type III secretion apparatus through the NLRC4 inflammasome. Proc. Natl. Acad. Sci. U.S.A. 107, 3076–3080 (2010). doi:10.1073/pnas.0913087107 Medline

12. Y. Li, D. A. Powell, S. A. Shaffer, D. A. Rasko, M. R. Pelletier, J. D. Leszyk, A. J. Scott, A. Masoudi, D. R. Goodlett, X. Wang, C. R. Raetz, R. K. Ernst, LPS remodeling is an evolved survival strategy for bacteria. Proc. Natl. Acad. Sci. U.S.A. 109, 8716–8721 (2012). doi:10.1073/pnas.1202908109 Medline

13. J. S. Gunn, R. K. Ernst, The structure and function of Francisella lipopolysaccharide. Ann. N. Y. Acad. Sci. 1105, 202–218 (2007). doi:10.1196/annals.1409.006 Medline

14. X. Wang, A. A. Ribeiro, Z. Guan, S. N. Abraham, C. R. H. Raetz, Attenuated virulence of a Francisella mutant lacking the lipid A 4′-phosphatase. Proc. Natl. Acad. Sci. U.S.A. 104, 4136–4141 (2007). doi:10.1073/pnas.0611606104 Medline

15. D. Kanistanon, D. A. Powell, A. M. Hajjar, M. R. Pelletier, I. E. Cohen, S. S. Way, S. J. Skerrett, X. Wang, C. R. Raetz, R. K. Ernst, Role of Francisella lipid A phosphate modification in virulence and long-term protective immune responses. Infect. Immun. 80, 943–951 (2012). doi:10.1128/IAI.06109-11 Medline

16. D. Kanistanon, A. M. Hajjar, M. R. Pelletier, L. A. Gallagher, T. Kalhorn, S. A. Shaffer, D. R. Goodlett, L. Rohmer, M. J. Brittnacher, S. J. Skerrett, R. K. Ernst, A Francisella mutant in lipid A carbohydrate modification elicits protective immunity. PLoS Pathog. 4, e24 (2008). doi:10.1371/journal.ppat.0040024 Medline

17. R. Rebeil, R. K. Ernst, B. B. Gowen, S. I. Miller, B. J. Hinnebusch, Variation in lipid A structure in the pathogenic yersiniae. Mol. Microbiol. 52, 1363–1373 (2004). doi:10.1111/j.1365-2958.2004.04059.x Medline

18. S. Wang, M. Miura, Y. K. Jung, H. Zhu, E. Li, J. Yuan, Murine caspase-11, an ICE-interacting protease, is essential for the activation of ICE. Cell 92, 501–509 (1998). doi:10.1016/S0092-8674(00)80943-5 Medline

19. O. Takeuchi, K. Hoshino, T. Kawai, H. Sanjo, H. Takada, T. Ogawa, K. Takeda, S. Akira, Differential roles of TLR2 and TLR4 in recognition of gram-negative and gram-positive bacterial cell wall components. Immunity 11, 443–451 (1999). doi:10.1016/S1074-7613(00)80119-3 Medline

20. J. von Moltke, N. J. Trinidad, M. Moayeri, A. F. Kintzer, S. B. Wang, N. van Rooijen, C. R. Brown, B. A. Krantz, S. H. Leppla, K. Gronert, R. E. Vance, Rapid induction of inflammatory lipid mediators by the inflammasome in vivo. Nature 490, 107–111 (2012). doi:10.1038/nature11351 Medline

21. M. Lamkanfi, A. Sarkar, L. Vande Walle, A. C. Vitari, A. O. Amer, M. D. Wewers, K. J. Tracey, T. D. Kanneganti, V. M. Dixit, Inflammasome-dependent release of the alarmin HMGB1 in endotoxemia. J. Immunol. 185, 4385–4392 (2010). doi:10.4049/jimmunol.1000803 Medline

22. H. Gao, T. W. Evans, S. J. Finney, Bench-to-bedside review: sepsis, severe sepsis and septic shock - does the nature of the infecting organism matter? Crit. Care 12, 213 (2008). doi:10.1186/cc6862 Medline

23. L. P. Shornick, P. De Togni, S. Mariathasan, J. Goellner, J. Strauss-Schoenberger, R. W. Karr, T. A. Ferguson, D. D. Chaplin, Mice deficient in IL-1beta manifest impaired contact hypersensitivity to trinitrochlorobenzone. J. Exp. Med. 183, 1427–1436 (1996). doi:10.1084/jem.183.4.1427 Medline

24. K. Takeda, H. Tsutsui, T. Yoshimoto, O. Adachi, N. Yoshida, T. Kishimoto, H. Okamura, K. Nakanishi, S. Akira, Defective NK cell activity and Th1 response in IL-18-deficient mice. Immunity 8, 383–390 (1998). doi:10.1016/S1074-7613(00)80543-9 Medline

25. K. Kuida, J. A. Lippke, G. Ku, M. W. Harding, D. J. Livingston, M. S. Su, R. A. Flavell, Altered cytokine export and apoptosis in mice deficient in interleukin-1 beta converting enzyme. Science 267, 2000–2003 (1995). doi:10.1126/science.7535475 Medline

26. E. A. Miao, C. M. Alpuche-Aranda, M. Dors, A. E. Clark, M. W. Bader, S. I. Miller, A. Aderem, Cytoplasmic flagellin activates caspase-1 and secretion of interleukin 1β via Ipaf. Nat. Immunol. 7, 569–575 (2006). doi:10.1038/ni1344 Medline

27. T. Decker, M. L. Lohmann-Matthes, A quick and simple method for the quantitation of lactate dehydrogenase release in measurements of cellular cytotoxicity and tumor necrosis factor (TNF) activity. J. Immunol. Methods 115, 61–69 (1988). doi:10.1016/0022-1759(88)90310-9 Medline

28. S. E. Warren, A. Armstrong, M. K. Hamilton, D. P. Mao, I. A. Leaf, E. A. Miao, A. Aderem, Cutting edge: Cytosolic bacterial DNA activates the inflammasome via Aim2. J. Immunol. 185, 818–821 (2010). doi:10.4049/jimmunol.1000724 Medline

29. T. Kitamura, Y. Koshino, F. Shibata, T. Oki, H. Nakajima, T. Nosaka, H. Kumagai, Retrovirus-mediated gene transfer and expression cloning: Powerful tools in functional genomics. Exp. Hematol. 31, 1007–1014 (2003). Medline