Bcl-2 and Bcl-XLRegulate
Proinflammatory Caspase-1 Activation
by Interaction with NALP1
Jean-Marie Bruey,1Nathalie Bruey-Sedano,1Frederic Luciano,1Dayong Zhai,1Ruchi Balpai,1Chunyan Xu,1
Christina L. Kress,1Beatrice Bailly-Maitre,1Xiaoqing Li,1Andrei Osterman,1Shu-ichi Matsuzawa,1
Alexey V. Terskikh,1Benjamin Faustin,1and John C. Reed1,*
1Burnham Institute for Medical Research, La Jolla, CA 92037, USA
substrates involved in apoptosis or inflamma-
tion. In C. elegans, a paradigm for caspase
regulation exists in which caspase CED-3 is ac-
tivated by nucleotide-binding protein CED-4,
which is suppressed by Bcl-2-family protein
CED-9. We have identified a mammalian analog
of this caspase-regulatory system in the NLR-
family protein NALP1, a nucleotide-dependent
activator of cytokine-processing protease cas-
amyl-dipeptide (MDP). Antiapoptotic proteins
Bcl-2 and Bcl-XLbind and suppress NALP1,
reducing caspase-1 activation and interleukin-
1b (IL-1b) production. When exposed to MDP,
caspase-1 processing and IL-1b production,
whereas Bcl-2-overexpressing macrophages
demonstrate less caspase-1 processing and
tion of host defense and apoptosis machinery.
Caspases are intracellular proteases that cleave sub-
strates involved in either apoptosis or inflammation, with
different branches of the caspase family devoted to these
two functions in mammals. The zymogen forms of all in-
flammatory, and some apoptotic, caspases contains an
N-terminal CARD domain that mediates their interactions
with various adaptor proteins, thereby controlling their ac-
tivation, often through a mechanism involving oligomeri-
zation (reviewed in (Martinon and Tschopp, 2004). In Cae-
norhabditis elegans (C. elegans), a paradigm for apoptotic
caspase regulation has been established in which the
CARD-containing caspase CED-3 is activated by CED-4,
oligomerizes to create a platform for protease activation
(Riedl et al., 2005). CED-4 is directly suppressed by
Bcl-2-family member CED-9, an antiapoptotic protein
that binds CED-4 (Metzstein et al., 1998). Given the simi-
larities in apoptosis mechanisms throughout the animal
kingdom, it has been hypothesized that mammalian Bcl-
but no convincing examples have heretofore been
PILLER, or PAN) constitute a large family of caspase-acti-
vating and NF-kB-activating proteins found in vertebrates
and in marine vertebrates but not C. elegans or Drosoph-
binding fold called NACHT, plus leucine-rich repeat (LRR)
domains, typically in combination with additional protein-
interaction domains, including PYRIN and CARD domains
(reviewed in Kufer et al., 2005; Martinon and Tschopp,
2005; Stehlik and Reed, 2004; Ting et al., 2006). The
NACHT domain mediates oligomerization of mammalian
NLRs, analogous to the nucleotide-binding NB-ARC do-
to suggestthattheLRRssuppress NACHT-mediated olig-
crobial ligands (Martinon etal.,2002; Poyet et al.,2001). In
this regard, NLRs represent the intracellular complement
to the cell-surface TLR-family receptors involved in innate
lar host-defense proteins of plants (Kufer et al., 2005;
Martinon and Tschopp, 2005; Stehlik and Reed, 2004).
Here we show that the human NLR-family member
NALP1 (NAC/CARD7/DEFCAP/CLR17.1/NLRP1) is regu-
lated by interactions with antiapoptotic proteins Bcl-2
and Bcl-XL, which suppress NALP1-mediated activation
of caspase-1 and reduce production of the caspase-1
substrate interleukin-1b. NALP1 is similar to CED-4 in that
it contains CARD- and nucleotide-binding oligomerization
domains. The Bcl-2/Bcl-XL-mediated suppression of cas-
pase-1-activating NALP1 thus provides a mammalian
anism linking host defense and apoptosis.
Cell 129, 45–56, April 6, 2007 ª2007 Elsevier Inc. 45
NALP1 Binds Bcl-2 and Bcl-XL
We surveyed members of the NALP family for interactions
with antiapoptotic human Bcl-2-family proteins. NALP1
was found to associate with Bcl-2 and Bcl-XLby coimmu-
noprecipitation (coIP) experiments using lysates prepared
from transfected HEK293T cells expressing epitope-
tagged proteins. Of the six human antiapoptotic Bcl-2-
family proteins, only Bcl-2 and Bcl-XLassociated with
NALP1. In contrast, Mcl-1, Bcl-W, Bfl-1, and Bcl-B did
ious proapoptotic Bcl-2-family proteins, including Bax,
Bak, Bid, and Bcl-G (Figure 1B). Similar conclusions
were reached using in vitro protein-binding assays where
NALP1-containing cell lysates were incubated with bacte-
ria-produced GST-fusion proteins (Figure 1C).
To explore whether NALP1 is unique among NLR-family
proteins in its ability to bind Bcl-2 and Bcl-XL, we com-
pared NALP1 with NALP2, -3, and -4, which all contain
PYRIN, NACHT, and LRR domains like NALP1. We also
examined the proteins Pyrin and ASC, which contain
PYRIN domains. However, among these proteins tested,
only NALP1 associated with Bcl-XLand Bcl-2 (Figures
1D and 1E; data not shown).
NALP1 forms a multiprotein caspase-activating com-
plex called the ‘‘inflammasome,’’ which contains NALP1,
bipartite-adaptor protein ASC (containing PYRIN and
CARD domains), and caspase-1 (Martinon et al., 2002).
Both lipopolysaccharide (LPS) and the peptidoglycan
component muramyldipeptide (MDP) have been reported
to stimulate NALP1 inflammasome assembly (Faustin
et al., 2007; Martinon et al., 2002). To explore the interac-
tion of endogenous Bcl-2 and Bcl-XLwith endogenous
NALP1, we performed experiments with THP-1 mono-
phorbol esterTPAandfollowed procedures thatwerepre-
viously published in which treatment of these cells with
either LPS or MDP (along with ATP to enhance IL-1b
release) was shown to induce inflammasome assembly,
caspase-1 activation, and IL-1b secretion (Martinon et al.,
2002). Treatment of macrophages with LPS or MDP did
not significantly alter total cellular levels of NALP1, ASC,
blotting (Figures 2A–2D; data not shown), but it did stimu-
was immunoprecipitated from untreated macrophages
using anti-NALP1 antibody, endogenous Bcl-2 and Bcl-
XLwere associated with NALP1-containing immune com-
plexes, while ASC was not (Figure 2B). These macro-
phages evidently contain more Bcl-2 than Bcl-XL, which
possibly accounts for the clearer association of Bcl-2
with NALP1 immunoprecipitates when compared to Bcl-
XL. In contrast, when immunoprecipitated from MDP/
ATP-treated (Figure 2B) or LPS/ATP-treated (Figure S1)
macrophages, ASC was associated with NALP1-contain-
ing immune complexes, while Bcl-2 and Bcl-XLwere not.
These findings were confirmed by reciprocal coIP experi-
ments using anti-Bcl-2 (Figure 2C), anti-Bcl-XL (not
shown), or anti-ASC (Figure 2D) antibodies. Subcellular
fractionation studies showed that these LPS/ATP-induc-
ible differences in NALP1 binding to ASC and Bcl-2 also
correlated with changes in the relative amounts of NALP1
associated with membranous organelles where Bcl-2 is
located (Figure 2E).
Bcl-2 and Bcl-XLSuppress Caspase-1 Activation
The NALP1 inflammasome binds caspase-family prote-
ases involved in proteolytic processing of proinflam-
matory cytokine prointerleukin-1b (IL-1b), including pro-
caspase-1 and procaspase-5, but not caspase-9 or
caspase-12 (Figure S2). We therefore explored the effect
of overexpressing antiapoptotic Bcl-2-family proteins on
NALP1-induced production of IL-1b. When 293 cells
were transfected with plasmids encoding the inflamma-
some components NALP1, ASC, and procaspase-1 as
well as the inflammasome substrate pro-IL-1b, we
observed mature IL-1b secretion into culture medium (de-
tected by ELISA) and production of 17 kDa cleaved IL-1b
protein in cells (detected by immunoblotting; Figure 3A).
Cotransfection of Bcl-2 or Bcl-XLmarkedly suppressed
NALP1-dependent IL-1b secretion as well as production
of intracellular cleaved p17 IL-1b. Immunoblotting experi-
ments showed that Bcl-2 and Bcl-XLdid not alter the
levels of the various inflammasome components (not
shown). In contrast to Bcl-2 and Bcl-XL, antiapoptotic
Bcl-2-family proteins that do not bind NALP1 do not sup-
press IL-1b secretion or pro-IL-1b cleavage; these include
Bcl-W, Bcl-B, Bfl-1, and Mcl-1. Moreover, none of the
six antiapoptotic Bcl-2-family proteins modulated IL-1b
production induced by transfection of cells with procas-
pase-1 alone or in combination with an alternative NLR-
family protein (NALP2/PAN1) that does not bind Bcl-2-
family proteins (Figure S3), thus confirming the specificity
of these results. However, all six antiapoptotic human
Bcl-2-family proteins effectively suppressed apoptosis
and reduced activation of apoptotic caspases when ex-
pressed in 293 cells by the same transfection method
(Figure S3B), confirming the bioactivity of these proteins.
Similar results regarding Bcl-2 and Bcl-XLsuppression
of NALP1-induced IL-1b production were obtained using
HeLa cells (not shown) except that transfection of ASC
was not required because these cells express ASC en-
dogenously (Figure S4).
dent activation of procaspase-1 so that the effects of Bcl-
XLand Bcl-2 could be tested directly and modeled our ap-
proach after previously described cell-free systems for
studying NALP1-mediated activation of caspase-1 (Marti-
non et al., 2002). Extracts from THP-1 macrophages were
mixed with extracts from NALP1-transfected 293T cells
and then incubated at 37?C to induce caspase-1 activa-
tion in the presence or absence of recombinant Bcl-2-
family proteins. Adding Bcl-2 or Bcl-XLto extracts sup-
pressed caspase-1 activity as measured by hydrolysis of
46 Cell 129, 45–56, April 6, 2007 ª2007 Elsevier Inc.
fluorogenic substrate acetyl-Tryptophanyl-Glutamyl-His-
Figure 3B). In contrast, Bcl-W, Bfl-1, Bcl-B, or Mcl-1 did
not significantly suppress NALP1-dependent caspase-1
activation in extracts. Also, when THP-1 macrophages
were pretreated with LPS to induce activation of cas-
pase-1 prior to preparing extracts, then Bcl-2 (Figure S5)
and Bcl-XL(not shown) failed to suppress caspase-1 ac-
tivity in vitro, showing that Bcl-2 and Bcl-XLdo not sup-
press caspase-1 after it has become activated.
NALP1-containing extracts were also used for interro-
gating mechanisms by which Bcl-XLsuppresses NALP1
cause of its superior potency (Faustin et al., 2007). Note
inated with MDP-containing peptidoglycan, which may
account for their ability to activate NALP1. For these ex-
periments, the bacterial form of MDP was compared
withan inactiveenantiomer, MDP-DD. Prior to MDPexpo-
sure, the caspase-1-binding adaptor ASC is not associ-
ated with NALP1 (Figures S6 and 3C). When active
MDP-LD (but not MDP-DD) was added to extracts derived
from HEK293T cells transfected with plasmids encoding
GFP-tagged ASC and epitope-tagged NALP1, we ob-
served that GFP-ASC inducibly associated with NALP1
(Figures 3C and 3D). Addition of Bcl-XLor Bcl-2 to the ex-
tracts prevented GFP-ASC from binding to NALP1. Thus,
Bcl-XLand Bcl-2 prevent inflammasome formation in vitro
Figure 1. Bcl-2 and Bcl-XLBind NALP1
(A) and (B) show CoIP assays. HEK293T cells were transfected with plasmids encoding Flag-NALP1 and either Myc-tagged Bcl-2, Bcl-XL, Mcl-1,
Bcl-B, Bfl-1 (A), Bcl-XL, or Bcl-W proteins and GFP-tagged Bcl-Gs, Bax, Bak and Bid (B). Cell lysates normalized for total protein content were
analyzed directly (middle and bottom) or subjected to IP (top) using anti-Myc (A) or anti-Flag (B) antibody and were analyzed by SDS-PAGE/immu-
noblotting (WB) using anti-Flag (A), anti-Myc, and anti-GFP (B) antibodies.
(C) GST pull-down assays are shown. Cell extract from 293T cells overexpressing NALP1 (20 mg per sample) was incubated with 1 mg of immobilized
GST or GST-fusion proteins corresponding to various Bcl-2-family proteins. Proteins associated with glutathione-Sepharose were analyzed by
immunoblotting using anti-GST (bottom) and anti-NALP1 (top) antibodies.
(D and E) Bcl-XLand Bcl-2 uniquely bind NALP1 among the NLR-family proteins that were tested. Lysates were prepared from 293T cotransfected
with plasmids encoding Flag-Bcl-XL(D) or Bcl-2 (E) and plasmids encoding Myc-tagged NALP1, NALP3, NALP4, Pyrin, ASC, or untagged NALP2.
Lysates were analyzed directly (bottom) or subjected to IP using anti-Bcl-XLor anti-Bcl-2 antibody (top), then analyzed by immunoblotting using
anti-Flag, anti-Myc, or anti-NALP2 antibodies, as indicated. All results are representative of at least three independent experiments. Asterisks denote
nonspecific bands. Molecular weight (MW) markers are indicated in kilodaltons.
Cell 129, 45–56, April 6, 2007 ª2007 Elsevier Inc. 47
MDP stimulation. Control proteins, such as GST-Bcl-B,
which does not bind NALP1, did not have this effect
(Figure 3D; data not shown). We hypothesize, therefore,
that Bcl-2 and Bcl-XLrecognize an inactive conformation
of NALP1 and suppress conversion of NALP1 to the active
conformation that binds ASC and allows inflammasome
Binding Is Required for Suppression of NALP1
by Bcl-2 and Bcl-XL
Domain-mapping experiments were performed to explore
whether binding is required for Bcl-2 and Bcl-XLto sup-
press NALP1-induced activation of caspase-1 and pro-
duction of IL-1b. Antiapoptotic Bcl-2-family proteins
contain conserved BH1-4 domains and are homologous
throughout their amino acid sequences with the exception
of a loop of variable length between BH4 (a-helix-1) and
BH3 (a-helix-2; Strasser, 2005). To explore why Bcl-2
and Bcl-XLuniquely bind NALP1 among the six antiapop-
and Bcl-XLwith various deletion mutants. Removal of the
loop from Bcl-2 or Bcl-XL abolished interaction with
NALP1 (Figures 4A and S7A). In contrast, deleting BH3
or BH4 domains from Bcl-XLdid not impair binding to
NALP1, as determined by coIP experiments (Figure S7B).
These protein-interaction studies were performed by
coIP using cell lysates and were independently confirmed
Figure 2. Endogenous NALP1 Binds Bcl-2
(A) TPA-differentiated THP-1 cells were cultured with or without (CT) 1 mg/ml crude LPS for 20 hr and then pulsed with 5 mM ATP for 10 min. Cell
lysates were normalized for protein content and analyzed by immunoblotting using antibodies specific for NALP1, Bcl-2, Bcl-XL, and ASC. Culture
supernatants (normalized for volume) were also analyzed for IL-1b by immunoblot analysis.
(B) THP-1 macrophages were cultured with or without (CTL) MDP/ATP as above, and lysates were prepared for IP with control IgG or anti-NALP1
antibody. The resulting immune complexes were analyzed by immunoblotting using antibodies recognizing NALP1 (top), Bcl-2, ASC, or Bcl-XL(bot-
tom). Ten percent of lysates were also loaded directly as a control (left lanes).
(C and D)THP-1macrophageswerepreparedas above,and lysates weresubjectedtoIP usingcontrol (CTL)IgGand either anti-Bcl-2(C) oranti-ASC
(D) antibodies. Immune complexes were fractionated by SDS-PAGE (8% gels, C; 14% gels, D) and analyzed by immunoblotting using anti-NALP1
(top) antibody followed by anti-Bcl-2 or anti-ASC (bottom).
(E) LPS changes subcellular location of NALP1. THP-1 macrophages were treated with or without 5 mg/ml crude LPS for 4 hr and 2.5 mM ATP for
10 min, then cytosolic and membrane fractions were prepared and analyzed by immunoblotting using anti-NALP1, -Bcl-2, or -tubulin antibodies.
All results are representative of R3 independent experiments. Asterisks denote nonspecific bands.
48 Cell 129, 45–56, April 6, 2007 ª2007 Elsevier Inc.
by immunofluorescence confocal microscopy analysis of
intact cells, where full-length Bcl-2, but not Bcl-2(Dloop),
cytosolic to an organellar location (Figure S8). Correlating
with the protein interaction, mutants of Bcl-XL(Figure 4B)
or Bcl-2 (Figure 4C) that lacked the loop were also inactive
with respect to suppression of NALP1-induced IL-1b
secretion and NALP1-induced proteolytic processing of
intracellular pro-IL-1b. Because Bcl-XL(Dloop) and Bcl-
2(Dloop) mutants have enhanced antiapoptotic activity
(Chang et al., 1997), NALP1-suppressing activity can be
separated from antiapoptotic activity of Bcl-XL and
Bcl-2. Similarly, a point mutant of Bcl-2 (G145A) lacking
antiapoptotic activity (Yin et al., 1994) retained NALP1-
NALP1-induced IL-1b production, again dissociating
NALP1-suppressing activity from apoptosis-suppressing
activity (Figure 4C).
Using a series of truncation and internal deletion mu-
tantsof NALP1,weattempted tomap the regionof NALP1
required for binding Bcl-XL. These experiments demon-
Figure 3. Bcl-2 and Bcl-XLSuppress Caspase-1 Activation by NALP1
(A) Bcl-2 and Bcl-XLsuppress NALP1-induced IL-1b production. 293 cells were transfected with plasmids encoding procaspase-1, pro-IL-1b, and
ASC with (+) or without (?) NALP1 according to established procedures (Martinon et al., 2002) and with various plasmids encoding antiapoptotic Bcl-
2-family proteins as indicated, with total DNA constant maintained at 1 mg by addition of empty plasmid. Supernatantswere analyzed by ELISA for IL-
immunoblotting (WB) using anti-IL-1b antibody (bottom).
(B) Bcl-2 and Bcl-XLsuppress NALP1-mediated caspase-1 activation in vitro. THP-1 cells were differentiated into macrophages by overnight treat-
ment with 50 ng/ml TPA, and cytosolic extracts were prepared. Extracts from 293T cells transfected with NALP1-encoding plasmid were mixed 2:1
with THP-1 extracts and incubated with 5 mg GST protein or various GST-fusion proteins corresponding to antiapoptotic Bcl-2-family proteins for 45
min at 37?C. Caspase-1 activity was measured using Ac-WEHD-AFC substrate, measuring relative fluorescence units (RFU) of AFC released per min
per mg of protein (mean ± SD; n = 3). Asterisk indicates p < 0.05. All experiments were repeated at least three times.
(Cand D)Bcl-XLandBcl-2 inhibitMDP-inducedassociationofNALP1 withASCinvitro.Extracts(20mg) from293Tcells transfectedwithanormalized
MDP-LD or MDP-DD was added to stimulate inflammasome assembly. IPs were performed using anti-myc antibody followed by SDS-PAGE/immu-
noblot analysis using anti-GFP (top). Alternatively, lysates were directly analyzed by immunoblotting (WB) using anti-myc, -GFP, or -GST antibodies
(bottom). All results are representative of R3 independent experiments. Asterisks denote nonspecific bands, while arrowheads indicate specific
Cell 129, 45–56, April 6, 2007 ª2007 Elsevier Inc. 49
of NALP1 are necessary, but insufficient, for binding Bcl-
XL(Figures 4D, 4E, S7C, and S7D). These protein-interac-
tion studies were performed by coIP using cell lysates and
were independently confirmed by immunofluorescence
confocal microscopy analysis of intact cells, where full-
length NALP1 but not NALP1DLRR was shown to redis-
tribute from a diffuse cytosolic to an organellar location
when coexpressed with Bcl-2 (Figure S8). Consistent with
the protein-interaction data showing that the LRRs of
NALP1 are required for binding Bcl-XL, we observed that
IL-1b production induced by a mutant of NALP1 lacking
the LRRs was not suppressed by Bcl-XL, in contrast
to full-length NALP1 (Figure 4F). We conclude, therefore,
that Bcl-2 and Bcl-XL must bind NALP1 to suppress
NALP1-mediated IL-1b production.
Bcl-2 Regulates MDP-Induced IL-1b Production
We experimentally manipulated the levels of Bcl-2 or Bcl-
XLin human THP-1 macrophages using RNA interference
(RNAi) and gene transfer then studied effects on MDP-
induced IL-1b production. In cultured human THP-1 mac-
rophages, siRNA experiments demonstrated that IL-1b
production (measured by ELISA) in response to MDP is
Figure 4. Binding Is Required for Suppression of NALP1-Induced IL-1b Production by Bcl-2 and Bcl-XL
(A and B)Loops of Bcl-XLand Bcl-2 required for NALP1binding are shown.293T cells weretransfected withFlag-NALP1 plasmid in combination with
plasmids encoding Myc-Bcl-2,Myc-Bcl-2-DLOOP,Myc-Bcl-XL,orMyc-Bcl-XL-DLOOP.Lysateswereeitheranalyzeddirectly(10mgprotein;bottom)
or subjected to IP using anti-Flag (top) and analyzed by immunoblotting using anti-Myc antibody.
(B and C) Interaction of NALP1 with Bcl-XLis required for inhibition of NALP1-induced IL-1b production. HeLa cells were transfected with plasmids
encoding NALP1, procaspase-1, and pro-IL-1b in combination with plasmid encoding Bcl-XLversus Bcl-XLDloop (B) or Bcl-2 versus Bcl-2Dloop (C),
withIL-1bproductionmeasured at24hr.Additionally,wild-type Bcl-2 and Bcl-2-G145A were compared (C),whichfurtherdissociatedsuppression of
NALP1-induced IL-1b production from antiapoptotic activity. In (B), cell lysates were normalized for protein content and analyzed by immunoblotting
using anti-IL-1b antibody (bottom).
(D and E) LRR of NALP1 is necessary but insufficient for binding Bcl-XL. Interactions of Bcl-XLwith full-length NALP1 and various NALP1-deletion
mutants were tested by coIP assay. Representative data are shown for DLRR and DFIIND mutants. (E) shows summary of results (see Figures
S7C and S7D).
(F) NALP1DLRR mutant is resistant to suppression by Bcl-XL. HeLa cells were transfected as above, by comparing full-length NALP1 with
NALP1DLRR and measuring IL-1b production in the presence or absence of cotransfected Bcl-XL. Supernatants in panels (B), (C), and (E) were
analyzed by ELISA for IL-1b at 24 hr posttransfection (mean ± SD; n = 3). Asterisk indicates p < 0.05 compared to NALP1.
50 Cell 129, 45–56, April 6, 2007 ª2007 Elsevier Inc.
largely NALP1-dependent (Figures 5A and S9) even
though at least three NLR-family members (NALP1, Nod2,
andCryopyrin) areknownto respondtothisbacterial pep-
tide (Inohara et al., 2003; Martinon et al., 2004). Moreover,
MDP-induced IL-1b production by THP-1 macrophages
was suppressed by chemicals that inhibit caspase-1
(Ac-YVAD-fmk and zVAD-fmk) but not by compounds
that preferentially inhibit effector caspases involved in ap-
optosis (Ac-DEVD-fmk), consistent with involvement of
inflammatory caspases (Figure 5A). Immunoblot analysis
confirmed sequence-specific reduction in NALP1 protein
in siRNA-treated THP-1 cells and independently verified
that MDP-induced IL-1b production was suppressed
(Figures 5B and S9). Moreover, NALP1-targeting siRNA
(but not scrambled RNA control) greatly reduced pro-
teolytic processing of caspase-1 and of intracellular
Figure 5. Bcl-2 and Bcl-XLRegulate MDP-Inducible IL-1b Production in Macrophages
(A) Regulation of MDP-induced IL-1b production by NALP1, Bcl-2, and Bcl-XL. THP-1 monocytes were electroporated with various siRNAs (2 mg to-
tal), including NALP1, control (CT; scrambled NALP1 sequence), or a 1:1 mixture of siRNA (1 mg each) targeting Bcl-XLand Bcl-2. Cells were differ-
(B)Immunoblot analysisof siRNA-transfected cells is shown. Lysates were prepared from THP-1cells treated as above, or culture supernatants were
TCA precipitated and analyzed by immunoblotting, using antibodies specific for IL-1b (supernatant; top), NALP1 (lysate; middle), or b-actin (lysates;
(C) NALP1 regulates MDP-induced caspase-1 activation. THP-1 cells were electroporated with NALP1 or scrambled siRNAs and treated as above
with MDP-LD or MDP-DD for 2 hr. Cell lysates (L) were normalized for protein content and either incubated with biotinyl-VAD-fmk to capture active
caspase-1 using streptavidin-Sepharose (SA; top) or analyzed directly. Immunoblotting was performed using antibodies specific for p10 fragment of
active caspase-1 (top), procaspase1 (p45), NALP1, or mature IL-1b (bottom).
(D) Bcl-2 overexpression suppresses MDP-induced IL-1b production. THP-1 macrophages infected with control (white circles) or Bcl-2-encoding
(black circles) recombinant lentiviruses were treated with various concentrations of either MDP (left) or bacterial flagellin (right). IL-1b production
was measured 24 hr later (mean ± SD; n = 3).
(E) Bcl-2 suppresses MDP-induced caspase-1 activation. THP-1 cells were infected by control or Bcl-2-encoding lentivirus, differentiated into mac-
rophages with TPA, and then cultured with MDP-LD or MDP-DD. Lysates were processed as in (C).
(F) Bcl-2 inhibits inflammasome assembly in macrophages. THP-1 macrophages infected with control or Bcl-2-encoding lentiviruses were unstimu-
lated (C = control) or stimulated with MDP-DD, MDP-LD, or LPS for 6 hr. Then, lysates were used for IP (top) using anti-NALP1 antibody. Immune
complexes (top) or lysates (bottom) were analyzed by SDS-PAGE/immunoblotting using antibodies recognizing ASC, Bcl-2, or NALP1.
Cell 129, 45–56, April 6, 2007 ª2007 Elsevier Inc. 51
pro-IL-1b induced in THP-1 macrophages by MDP-LD
In THP-1 macrophages where MDP-induced IL-1b pro-
duction is mostly NALP1 dependent, siRNA-mediated
reductions in Bcl-2 and Bcl-XL(Figure S10) caused an in-
crease in MDP-stimulated IL-1b production (Figure 5A),
suggesting that endogenous Bcl-2 and Bcl-XLrestrain
NALP1-dependent IL-1b production. In contrast, siRNAs
targeting Bcl-2-family proteins that fail to bind NALP1
(i.e., Bcl-B and Bcl-W) did not significantly impact MDP-
induced IL-1b production (Figure S10). Immunoblot analy-
sisconfirmed thatsiRNA-treatments producedreductions
in the relevant proteins (Figure S10). Some siRNA re-
agents targeting other Bcl-2-family members have nucle-
otide compositions closely approximating either the Bcl-
2- or Bcl-XL-specific siRNAs, and thus serve as controls.
While siRNA-mediated knockdown of Bcl-2 and Bcl-XL
enhanced MDP-induced IL-1b production, overexpres-
sion of Bcl-2 in THP-1 macrophages had the opposite
effect (Figure 5D, left). The specificity of Bcl-2-mediated
suppression of MDP-induced IL-1b production was con-
firmed by experiments using bacterial flagellin (Figure 5D,
right), which stimulates an alternative NLR-family member
(Ipaf/CLAN; Franchi et al., 2006; Miao et al., 2006) that
does not bind Bcl-2 or Bcl-XL (shown). Time course
studies suggested that Bcl-2-mediated suppression of
MDP-induced IL-1b production is demonstrable within
4 hr and excluded differences in macrophage survival as
an explanation for the difference in IL-1b release (Fig-
ure S11). Bcl-2 overexpression in THP-1 macrophages
also inhibited MDP-stimulated proteolytic processing of
caspase-1 (Figure 5E). We also observed that Bcl-2 over-
expression inhibited inflammasome assembly in THP-1
cells whether induced by MDP or by LPS, and less endog-
enous ASC coIPed with endogenous NALP1 in Bcl-2-
overexpressing THP-1 macrophages (Figure 5F).
Similar conclusions were reached from studies using
cultured bone marrow-derived macrophages from bcl-2
knockout and bcl-2 transgenic mice (Domen et al., 2000;
Wang et al., 2005). Direct comparisons showed that
MDP induced more IL-1b production in cultures of macro-
phages from bcl-2?/?mice compared to bcl-2+/?mice,
which in turn produced more IL-1b than cells from bcl-
2+/+mice (Figure 6A). Indeed, Bcl-2-deficient macro-
phages produced 33% ± 6% more IL-1b than wild-type
macrophages. Conversely, macrophages from transgenic
mice that overexpress Bcl-2 in blood cells that are driven
by a H2K promoter (Domen et al., 2000) elaborated 37% ±
7% less IL-1b compared to control cells from nontrans-
genic littermates (Figure 6A). These findings are particu-
larly striking when recognizing that MDP is capable of trig-
gering both NALP1-dependent and NALP1-independent
pathways for IL-1b production and that Bcl-2 only sup-
presses the NALP1-dependent contribution. Also, the ab-
solute difference in IL-1b production under the conditions
of these experiments was ?500–1000 pg/mL, which is
Muckles-Wells patients (which contain hereditary muta-
tions in a NLR gene) show IL-1b differences of only 30–
900 pg/ml compared to normal cells (Agostini et al.,
2004), and levels of IL-1b in serum of septic mice report-
edly average 350 pg/mL (Saito et al., 2003). In contrast
to IL-1b, MDP-induced production of TNFa was not dif-
ferent among macrophages derived from bcl-2+/+, -2+/?,
and -2?/?mice (Figure 6A) nor among macrophages
derived from bcl-2 transgenic mice (Figure 6B), showing
specificity and implying that MDP activates other mole-
cules (e.g., Nod2 or Cryopyrin) besides NALP1 that regu-
late signaling pathways leading to TNFa production. Note
that preparations of murine macrophages varied in their
sensitivity to MDP and sometimes required priming with
a small amount of LPS as described (Sutterwala et al.,
2006; Figure S12).
The differences in MDP-induced IL-1b production ob-
served for bcl-2 knockout and bcl-2 transgenic mice cor-
teolytic processing of caspase-1 (Figures 6C and 6D).
Immunoblotting showedcomparable levelsof NALP1pro-
tein in bcl-2 transgenic and nontransgenic macrophages,
excluding a trivial explanation (not shown). We therefore
conclude that Bcl-2 restrains the MDP-induced activation
of caspase-1 and secretion of the caspase-1 substrate
IL-1b in primary cultured macrophages.
Bcl-2 Regulates MDP-Induced IL-1b Production
We compared IL-1b production in wild-type versus bcl-2
knockout mice injected with MDP. Prior to injection, no
IL-1b was detectable in serum of bcl-2?/?, bcl-2+/?, or
tion, IL-1b serum levels rose to ?100 pg/mL in wild-type
mice, declining substantially by 24 hr. In contrast, in
bcl-2 knockout mice, IL-1b serum levels were over 4-
fold higher at 3 hr and remained persistently elevated at
24 hr (Figure 6E). In contrast to IL-1b, MDP-induced pro-
duction of TNFa in vivo was not affected by bcl-2 and
was measured at 0, 3, and 24 hr postinjection with MDP
(Figure 6E; data not shown). While a clear increase in
MDP-induced IL-1b production was observed in bcl-2
knockout mice, reproducible differences in IL-1b produc-
tion were not detectable in the Bcl-2 transgenic mice (not
in myeloid-lineage cells, whereas NALP1 is expressed in
many tissues (Chu et al., 2001).
In C. elegans, it has been shown that the Bcl-2 ortholog
CED-9 binds caspase-activator CED-4 and suppresses
CED-3 protease activation (Metzstein et al., 1998). Here-
tofore, an analogous mechanism for controlling caspase
activation has been lacking in higher organisms. We
show that Bcl-2-family members Bcl-2 and Bcl-XLinteract
with NALP1, blunting NALP1-mediated activation of
52 Cell 129, 45–56, April 6, 2007 ª2007 Elsevier Inc.
family proteins (Kufer et al., 2005; Martinon and Tschopp,
2005). By analogy to structurally similar host-defense
genes in plants, presumably the reason for expansion of
this gene family is to provide diversity in recognition of
pathogen-associated molecules through diversification
Figure 6. Bcl-2 Regulates MDP-Driven Caspase-1 Activation and IL-1b Production in Mice
(A and B) Bcl-2 regulates IL-1b production induced by MDP in cultured mouse macrophages. Macrophages from age-matched bcl-2+/+, -2+/?,
and -2?/?mice (A) or from bcl-2 transgenic (Tg) mice or control littermates (B; three mice per group) were stimulated with MDP-LD or MDP-DD
for 2 hr, then pulsed with 2.5 mM ATP. After 2 hr, levels of IL-1b (left) or TNFa (right) released into culture supernatants were measured (mean ±
SD ; n = 3). Asterisks denote p < 0.05 compared to wild-type cells.
(C and D) Bcl-2 regulates MDP-inducible caspase-1 activation in mouse macrophages. Macrophages from bcl-2+/+versus bcl-2?/?littermates and
from bcl-2-Tg versus control littermates were incubated with biotinyl-VAD-fmk (30 mM) and subsequently stimulated with MDP-LD or MDP-DD for
2 hr. Lysates were analyzed as in Figure 5C.
(E) Bcl-2 regulates MDP-induced IL-1b production in vivo. Age-matched female bcl-2+/+(n = 5), -2+/?(n = 4), and -2?/?(n = 4) mice were injected i.p.
with 100 ug/kg MDP. Serum was collected before (t0) and at 3 hr and 24 hr after injection for measuring levels of IL-1b (left) and TNFa (right; mean ±
SEM). Asterisks denote p < 0.05 compared to+/+control mice.
Cell 129, 45–56, April 6, 2007 ª2007 Elsevier Inc. 53
quired for Bcl-2/XLbinding, implying that the same do-
main used by NALP1 to recognize pathogen-associated
MDP also binds Bcl-2/ XL. The binding of Bcl-2 and ASC
to NALP1, however, is unlikely to be directly competitive
because ASC has been shown to interact with the PYRIN
domain of NALP1, while the LRRs arenecessary for Bcl-2/
ognize different conformational states of NALP1. Differ-
ences in the LRRs of NALP1 relative to other members
of the NLR family may explain why Bcl-2 and Bcl-XL
bind NALP1 but not NALP2-4.
binding arethe least-conserved segments among the Bcl-
2-family proteins, presumably explaining why Bcl-2 and
Bcl-XL, but not other Bcl-2-family proteins, bind NALP1.
Since the loop region is subject to posttranslational mod-
ifications that modulate the antiapoptotic activity of Bcl-2
and Bcl-XL, it will be interesting to explore the impact on
NALP1 binding. The apparent utilization of the loop region
by Bcl-2 and Bcl-XLfor engaging NALP1 differs structur-
ally from the mechanisms used by CED-9 for binding
CED-4 (Yan et al., 2005), implying that different means
can be employed to accomplish the same goal. In this re-
gard, profound structural differences have also been
noted between orthologous human and C. elegans apo-
ptosis regulators, such as CED-4 and its mammalian
counterpart Apaf1 (Adams and Cory, 2002), which illus-
trates how basic paradigms for function are preserved de-
spite structural diversification during evolution. However,
it should be noted that the loop domains of Bcl-2 and
Bcl-XL may be necessary to generate conformational
states competent to bind NALP1 rather than serving di-
rectly as ligands for binding NALP1.
The data presented here demonstrate an apoptosis-
independent phenotype for Bcl-2 and Bcl-XL. However,
while the proinflammatory branch of the caspase family
(caspases-1,4,5 in humans) that NALP1 regulates is prin-
cipally involved in cytokine activation, these proteases
have also been implicated in apoptosis induction in a vari-
ety of pathological contexts, including infection of macro-
phages by bacteria and neuronal cell death induced by is-
Thus, the abilityof Bcl-2 andBcl-XLto suppressaninflam-
matory caspase-activating NLR-family member (NALP1)
during stress (Figure S13). Additional links between NLR-
family proteins and the core components of the apoptosis
machinery have been reported that may also be relevant.
For example, ASC has been reported to bind Bax, collab-
orating in apoptosis induction (Ohtsuka et al., 2004).
Moreover, NALP1 (NAC) can associate with Apaf-1 (Chu
et al., 2001; Figure S6), an activator of apoptotic cas-
pases. Thus, an intricate network of protein interactions
appears to exist that involves components of the innate
immunity and apoptosis machineries, presumably allow-
ing for coordination of cell death and host defense. A pre-
diction of these findings is that some viral homologs of
Bcl-2 will be found to interact with and inhibit NLR-family
members as a mechanism of blunting host-defense while
simultaneously suppressing cell death for purposes of
preserving hosts for viral replication.
Additional details for methods are provided in the Supplemental Data.
Production of GST Proteins
GST-fusion proteins containing Bcl-XL, Bcl-2, Bcl-W, Bcl-B, Bfl-1, and
Mcl-1 were expressed from pGEX 4T-1 (Zhai et al., 2006). All con-
structs were engineered with a stop codon to exclude the C-terminal
transmembrane (TM) domains. GST-fusion proteins corresponding
to Bcl-2-family proteins were confirmed to be properly folded based
on ability to bind fluorochrome-conjugated BH3 peptides with submi-
cromolar affinity as shown by fluorescence polarization assays (Zhai
et al., 2006).
Cell Lines and Transfections
HEK293T and HeLa cells were cultured in DMEM high-glucose me-
dium (Irvine Scientific), while THP-1 cells were grown in RPMI 1640
medium containing 10% fetal bovine serum (FBS). Transfection of
HEK293T and HeLa cells was performed using Lipofectamine 2000 re-
agent (Invitrogen). THP-1 cells were electroporated using an instru-
ment from AMAXA (see below). THP-1 cells were differentiated into
macrophages by culturing overnight with 50 ng/ml TPA (Calbiochem).
THP-1 Cell Transfection with siRNA
siRNAs were purchased from AMBION (Table S1). Various double-
stranded ribo-oligonucleotides (dsRNA) with overhanging 30deoxy
TT were prepared that target NALP1, Bcl-2, Bcl-XL, Bcl-W, or Bcl-B
mRNAs. Control dsRNAs constituting scrambled versions of NALP1
siRNA sequences were also tested (Table S1).
THP-1 cells were electroporated using AMAXA system (Nucleofec-
tor), suspending 2 31 06cells in 100 ml of solution V (AMAXA) contain-
ing 2 mg siRNA or dsRNA control, and applying an electrical discharge
using a fluorescence isothiocyanate (FITC)-conjugated control dsRNA
(Ambion). Cells were cultured for 56 hr then differentiated by culturing
with 50 ng/mL TPA for 24 hr. Adherent macrophages were then stim-
ulated for 4 hr with 1 mg/ml LPS or 2 mg MDP-LD, followed by 2.5 mM
ATP for 20 min. Culture supernatants were assayed for IL-1b after 2 hr.
Culture supernatants and cell extracts were also analyzed by immuno-
blotting in some cases.
described (Zufferey et al., 1998). THP-1 cells were transduced with the
virus at a MOI = 15 in suspension with 10 mg/ml polybrene. At 3 days
postinfection, infected cells were serially diluted to achieve single cells
per well in 96-well plates. Bcl-2 expression in 19 of the 40 recovered
clones was confirmed by immunofluorescence using FITC-conjugated
anti-Bcl-2 antibody. Two independent clones were further amplified
and used for subsequent experiments.
Murine Macrophage Cultures
Bone marrow cells were harvested from leg bones of sacrificed 5-
week-old bcl-2 transgenic, bcl-2 knockout, and littermate control
mice on a C57Bl/6 background (Domen et al., 2000; Wang et al.,
2005). Bone marrow cells were cultured at 37?C with 5% CO2in
RPMI 1640 medium with 10% FBS. CSF-1 (SIGMA) was added to me-
dium for 1 week to promote differentiation. Adherent macrophages in
24-well plates were stimulated for 3 hr with various amounts of LPS,
MDP-LD, or MDP-DD then cultured for a further 20 min in fresh
54 Cell 129, 45–56, April 6, 2007 ª2007 Elsevier Inc.
medium containing 2.5 mM ATP, washed, and cultured in 0.5 ml fresh
medium for 2 hr before collecting culture supernatants for IL-b assay.
Forassessingeffects ofBcl-2-family proteinson caspase-1 activation,
cytosolic extracts were prepared from differentiated THP-1 cells be-
fore or after treatment with LPS or MDP (Bruey et al., 2000) using lysis
buffer: 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 20 mM EDTA, 50 mM
NaF, 0.5% NP-40, and 0.1 mM Na3VO4. THP-1 cell extracts (10 mg)
were mixed with 30 mg of extract from NALP1-expressing HEK293T
cells in a total volume of 0.1 mL, then 1–5 mg of recombinant GST pro-
teins were added, and the mixtures were incubated at 37?C for 45 min
before adding caspase substrate, Ac-WEHD-AFC.
To recover active caspases from cells, differentiated THP-1 macro-
phages were incubated with biotinyl-VAD-fmk (30 mM; Alexis) 0.5 hr
before stimulation with 2 mg/ml MDP-LD or MDP-DD for 3 hr. Cells
were then resuspended in ‘‘lysis solution’’ (50 mM Tris-HCl, pH 7.4,
150 mM NaCl, 20 mM EDTA, 50 mM NaF, 0.3% NP-40, 0.1 mM
Na3VO4, 20 mg/ml-1leupeptin, 20 mg/ml-1aprotinin, and 1 mM
PMSF), and active caspases were recovered using Streptavidin-Se-
pharose (SIGMA), adding 30 ml of 1:1 Streptavidin-Sepharose suspen-
sion per 250 ml of lysate. Beads were washed in the lysis buffer and an-
alyzed by SDS-PAGE/immunoblotting using an antibody that detects
the p10 small subunit of processed human caspase-1 (Santa Cruz,
IPs and Immunoblotting
Immunoblotting was performed as described previously (Guo et al.,
2003). For IPs, 108THP-1 cells were used. For coIPs, 2 3 106
HEK293T cells were cultured in 50 mM benzyloxycarbonyl-Val-Ala-
Asp (O-methyl)-fluoromethyl ketone (zVAD-fmk; Enzyme Systems
Products) to prevent apoptosis. Cells were suspended in 0.5mL lysis
solution, cleared by incubation with 15 ml protein G Sepharose 4B
(Zymed), and then incubated with 15 ml of either polyclonal antibody
or rabbit IgG control serum (Zymed; San Francisco) and 40 ml protein
G at 4?C overnight. Samples were then washed four times with lysis
buffer, boiled in Laemmli buffer, and analyzed by SDS-polyacrylamide
gel electrophoresis (PAGE)/immunoblotting. Alternatively, lysates
were directly analyzed by immunoblotting after normalization for total
Stepwise extraction of cytosolic fractions and organelle/membrane
fractions was performed using ProteoExtract Subcellular Proteome
Extraction Kit (Calbiochem) and 2 3 106THP-1 macrophages.
A cohort of mice (8 to 10 weeks old) was injected intraperitoneally with
100 mg/kg-1 MDP-LD in PBS. For serum cytokine measurements,
animals were bled retro-orbitally before or at various times after MDP
Most data were presented as the mean ± SD from at least three inde-
pendent experiments. Statistical comparisons between different treat-
ments were performed by unpaired Student’s t test and p % 0.05 was
considered statistically significant.
Supplemental Data include one table, 13 figures, Experimental Proce-
dures, and References and can be found with this article online at
Supported by grants from the NIH and Lymphoma Research Founda-
tion. We thank Drs. C.M. Sorenson, N. Sheibani, H. Mori, and A. Porter
Received: May 9, 2006
Revised: August 15, 2006
Accepted: January 19, 2007
Published: April 5, 2007
Adams, J.M., and Cory, S. (2002). Apoptosomes: engines for caspase
activation. Curr. Opin. Cell Biol. 14, 715–720.
Agostini, L., Martinon, F., Burns, K., McDermott, M.F., Hawkins, P.N.,
and Tschopp, J. (2004). NALP3 forms an IL-1beta-processing inflam-
order. Immunity 20, 319–325.
Bruey, J.-M., Ducasse, C., Bonniaud, P., Ravagnan, L., Susin, S.A.,
Diaz-Latoud, C., Gurbuxani, S., Arrigo, A.-P., Kroemer, G., Solary,
E., and Garrido, C. (2000). Hsp27 negatively regulates cell death by
interacting with cytochrome c. Nat. Cell Biol. 2, 645–652.
Chang, B.S., Minn, A.J., Muchmore, S.W., Fesik, S.W., and Thomp-
son, C.B. (1997). Identification of a novel regulatory domain in Bcl-xL
and Bcl-2. EMBO J. 16, 968–977.
Chu, Z.-L., Pio, F., Xie, Z., Welsh, K., Krajewska, M., Krajewski, S.,
tosome involved in cytochrome c-dependent caspase activation and
apoptosis. J. Biol. Chem. 276, 9239–9245.
Domen, J., Cheshier, S.H., and Weissman, I.L. (2000). The role of ap-
optosis in the regulation of hematopoietic stem cells: Overexpression
of Bcl-2 increases both their number and repopulation potential.
J. Exp. Med. 191, 253–264.
Maitre, B.,Hanein,D., Rouiller, I., andReed,J.C.(2007).Reconstituted
NALP1 inflammasome reveals two-step mechanism of caspase-1
activation. Mol. Cell 25, 713–724.
Franchi, L., Amer, A., Body-Malapel, M., Kanneganti, T.D., Ozoren, N.,
Jagirdar, R., Inohara, N., Vandenabeele, P., Bertin, J., Coyle, A., et al.
(2006). Cytosolic flagellin requires Ipaf for activation of caspase-1 and
interleukin 1beta in salmonella-infected macrophages. Nat. Immunol.
Guo, B., Zhai, D., Cabezas, E., Welsh, K., Nouraini, S., Satterthwait,
A.C., and Reed, J.C. (2003). Humanin peptide suppresses apoptosis
by interfering with Bax activation. Nature 423, 456–461.
Inohara, N., Ogura, Y., Fontalba, A., Gutierrez, O., Pons, F., Crespo, J.,
Fukase, K., Inamura, S., Kusumoto, S., Hashimoto, M., et al. (2003).
Host recognition of bacterial muramyl dipeptide mediated through
NOD2. Implications for Crohn’s disease. J. Biol. Chem. 278, 5509–
Kufer, T.A., Fritz, J.H., and Philpott, D.J. (2005). NACHT-LRR proteins
(NLRs) in bacterial infection and immunity. Trends Microbiol. 13, 381–
Martinon, F., and Tschopp, J. (2004). Inflammatory caspases: linking
an intracellular innate immune system to autoinflammatory diseases.
Cell 117, 561–574.
Martinon,F.,and Tschopp, J.(2005). NLRs joinTLRs asinnate sensors
of pathogens. Trends Immunol. 26, 447–454.
Martinon, F., Burns, K., and Tschopp, J. (2002). The Inflammasome: A
molecular platform triggering activation of inflammatory caspases and
processing of proIL-b. Mol. Cell 10, 417–426.
Cell 129, 45–56, April 6, 2007 ª2007 Elsevier Inc. 55
Martinon, F., Agostini, L., Meylan, E., and Tschopp, J. (2004). Identifi- Download full-text
cation of bacterial muramyl dipeptide as activator of the NALP3/cryo-
pyrin inflammasome. Curr. Biol. 14, 1929–1934.
Metzstein, M., Stanfield, G., and Horvitz, H. (1998). Genetics of pro-
grammed cell death in C. elegans: past, present and future. Trends
Genet. 14, 410–416.
Miao, E.A., Alpuche-Aranda, C.M., Dors, M., Clark, A.E., Bader, M.W.,
Miller, S.I., and Aderem, A. (2006). Cytoplasmic flagellin activates cas-
pase-1 and secretion of interleukin 1beta via Ipaf. Nat. Immunol. 7,
Navarre, W.W., and Zychlinsky, A. (2000). Pathogen-induced apopto-
sis of macrophages: a common end for different pathogenic strate-
gies. Cell. Microbiol. 2, 265–273.
Ohtsuka, T., Ryu, H., Minamishima, Y.A., Macip, S., Sagara, J., Na-
kayama, K.I., Aaronson, S.A., and Lee, S.W. (2004). ASC is a Bax
adaptor and regulates the p53-Bax mitochondrial apoptosis pathway.
Nat. Cell Biol. 6, 121–128.
Poyet, J.-L., Srinivasula, S.M., Tnani, M., Razmara, M., Fernandes-
Alnemri, T., and Alnemri, E.S. (2001). Indentification of Ipaf, a human
caspase-1 activating protein related to Apaf-1. J. Biol. Chem. 276,
Riedl, S.J., Li, W., Chao, Y., Schwarzenbacher, R., and Shi, Y. (2005).
Structure of the apoptotic protease-activating factor 1 bound to ADP.
Nature 434, 926–933.
Saito, H., Sherwood, E.R., Varma,T.K., and Evers, B.M. (2003).Effects
of aging on mortality, hypothermia, and cytokineinduction in mice with
endotoxemia or sepsis. Mech. Ageing Dev. 124, 1047–1058.
Stehlik, C., and Reed, J.C. (2004). The PYRIN connection: novel
players in innate immunity and inflammation. J. Exp. Med. 200, 551–
Strasser, A. (2005). The role of BH3-only proteins in the immune sys-
tem. Nat. Rev. Immunol. 5, 189–200.
Sutterwala, F.S., Ogura, Y., Szczepanik, M., Lara-Tejero, M., Lichten-
berger, G.S., Grant, E.P., Bertin, J., Coyle, A.J., Galan, J.E., Askenase,
P.W., and Flavell, R.A. (2006). Critical role for NALP3/CIAS1/Cryopyrin
in innate and adaptive immunity through its regulation of caspase-1.
Immunity 24, 317–327.
Ting, J.P., Kastner, D.L., and Hoffman, H.M. (2006). CATERPILLERs,
pyrin and hereditary immunological disorders. Nat. Rev. Immunol. 6,
Wang, S., Sorenson, C.M., and Sheibani, N. (2005). Attenuation of ret-
inal vascular development and neovascularization during oxygen-
induced ischemic retinopathy in Bcl-2?/?mice. Dev. Biol. 279, 205–
Yan, N., Chai, J., Lee, E.S., Gu, L., Liu, Q., He, J., Wu, J.W., Kokel, D.,
Li, H., Hao, Q., et al. (2005). Structure of the CED-4-CED-9 complex
provides insights into programmed cell death in Caenorhabditis ele-
gans. Nature 437, 831–837.
Yin, X.M., Oltvai, Z.N., and Korsmeyer, S.J. (1994). BH1 and BH2
domains of Bcl-2 are required for inhibition of apoptosis and heterodi-
merization with bax. Nature 369, 321–333.
Zhai, D., Jin, C., Satterthwait, A.C., and Reed, J.C. (2006). Comparison
of chemical inhibitors of anti-apoptotic Bcl-2-family proteins. Cell
Death Differ., in press.
Zhang, W.H., Wang, X.,Narayanan, M.,Zhang, Y.,Huo, C.,Reed, J.C.,
and Friedlander, R.M. (2003). Fundamental role of the Rip2/caspase-1
pathway in hypoxia and ischemia-induced neuronal cell death. Proc.
Natl. Acad. Sci. USA 100, 16012–16017.
and Trono, D. (1998). Self-inactivating lentivirus vector for safe and
efficient in vivo gene delivery. J. Virol. 72, 9873–9880.
56 Cell 129, 45–56, April 6, 2007 ª2007 Elsevier Inc.