The Inflammasomes: Guardians of the Body

ArticleinAnnual Review of Immunology 27(1):229-65 · February 2009with201 Reads
Impact Factor: 39.33 · DOI: 10.1146/annurev.immunol.021908.132715 · Source: PubMed

The innate immune system relies on its capacity to rapidly detect invading pathogenic microbes as foreign and to eliminate them. The discovery of Toll-like receptors (TLRs) provided a class of membrane receptors that sense extracellular microbes and trigger antipathogen signaling cascades. More recently, intracellular microbial sensors have been identified, including NOD-like receptors (NLRs). Some of the NLRs also sense nonmicrobial danger signals and form large cytoplasmic complexes called inflammasomes that link the sensing of microbial products and metabolic stress to the proteolytic activation of the proinflammatory cytokines IL-1beta and IL-18. The NALP3 inflammasome has been associated with several autoinflammatory conditions including gout. Likewise, the NALP3 inflammasome is a crucial element in the adjuvant effect of aluminum and can direct a humoral adaptive immune response. In this review, we discuss the role of NLRs, and in particular the inflammasomes, in the recognition of microbial and danger components and the role they play in health and disease.


Available from: Fabio Martinon
ANRV371-IY27-09 ARI 16 February 2009 8:57
The Inflammasomes:
Guardians of the Body
Fabio Martinon,
Annick Mayor,
and J
urg Tschopp
Department of Immunology and Infectious Diseases, Harvard School of Public Health,
Boston, Massachusetts 02115
Department of Biochemistry, University of Lausanne, 1066 Epalinges, Switzerland;
Annu. Rev. Immunol. 2009. 27:229–65
First published online as a Review in Advance on
December 8, 2008
The Annual Review of Immunology is online at
This article’s doi:
2009 by Annual Reviews.
All rights reserved
Key Words
adjuvanticity, danger signal, inflammation, innate immunity,
NOD-like receptors
The innate immune system relies on its capacity to rapidly detect in-
vading pathogenic microbes as foreign and to eliminate them. The
discovery of Toll-like receptors (TLRs) provided a class of membrane
receptors that sense extracellular microbes and trigger antipathogen
signaling cascades. More recently, intracellular microbial sensors have
been identified, including NOD-like receptors (NLRs). Some of the
NLRs also sense nonmicrobial danger signals and form large cytoplas-
mic complexes called inflammasomes that link the sensing of microbial
products and metabolic stress to the proteolytic activation of the proin-
flammatory cytokines IL-1β and IL-18. The NALP3 inflammasome
has been associated with several autoinflammatory conditions includ-
ing gout. Likewise, the NALP3 inflammasome is a crucial element in the
adjuvant effect of aluminum and can direct a humoral adaptive immune
response. In this review, we discuss the role of NLRs, and in partic-
ular the inflammasomes, in the recognition of microbial and danger
components and the role they play in health and disease.
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Click here for quick links to
Annual Reviews content online,
Other articles in this volume
Top cited articles
Top downloaded articles
• Our comprehensive search
Page 1
ANRV371-IY27-09 ARI 16 February 2009 8:57
receptor (PRR):
membrane receptor
expressed by cells of
the immune system to
identify molecules
associated with
microbial pathogens or
cellular stress
molecular pattern
(PAMP): highly
conserved microbial
structure that is
essential for microbial
survival and is detected
by host innate immune
Danger signal: signal
released by injured or
damaged tissues that
trigger an innate
immune response
Toll-like receptor
(TLR): membrane
receptor involved in
innate immune sensing
Vertebrates have evolved two complementary
systems to detect and clear pathogens: the in-
nate and the adaptive immune systems. The
innate immune system is the first one to be
activated by pathogens (1) and is usually suffi-
cient to clear the infection. However, when the
innate immune system is overwhelmed, it trig-
gers and directs the adaptive arm, thus activat-
ing specific B and T cells for pathogen clear-
ance. Receptors expressed by B and T cells are
generated through somaticgene rearrangement
and hypermutation. This process enables the
generation of a virtually infinite repertoire of
antigen receptors, allowing the adaptive im-
munity to specifically recognize any type of
In contrast, innate immunity is character-
ized by its ability to recognize a wide range of
pathogens such as viruses, bacteria, and fungi,
but through a limited number of germline-
encoded receptors called pattern-recognition
receptors (PRRs) (2, 3). PRRs are expressed by
many cell types including macrophages, mono-
cytes, dendritic cells (DCs), neutrophils, and
epithelial cells, and they allow the early de-
tection of pathogens directly at the site of in-
fection. PRRs recognize conserved microbial
signatures (4) termed pathogen-associated
molecular patterns, or PAMPs (see below).
Once activated, the innate immune system ini-
tiates the inflammatory response by secreting
cytokines and chemokines, inducing the expres-
sion of adhesion and costimulatory molecules
in order to recruit immune cells to the site of
infection and to trigger the adaptive immune
Pathogens can rapidly evolve and, in prin-
ciple, could avoid detection by the innate im-
mune system by simply altering the targeted
PAMPs. By doing so, the pathogen would not
only escape the recognition by the innate im-
mune system but also avoid the adaptive im-
mune response. However, the immune system
has evolved to recognize PAMPs that are es-
sential for the viability of microbes and are
thus less prone to modifications. PAMPs can
be of diverse origins; sugars, flagellin, and the
cell wall components peptidoglycan (PGN) and
lipopolysaccharide (LPS) are all recognized by
the innate immune system.
The model proposed by Charles Janeway
based on PAMP recognition is, however, too
simplistic. Indeed, if PAMPs activate the im-
mune pathway, how does the immune system
distinguish pathogenic microorganisms from
commensal and other non-pathogenic bacteria?
To answer this question, Matzinger (5, 6) sug-
gested that the activation of the innate immune
system is not only based on the recognition of
PAMPs but also relies on the presence of danger
signals or danger-associated molecular patterns
(DAMPs) released by injured cells. These two
models seem completely opposed, but several
reports now show activation of innate immu-
nity by host molecules. Indeed among others,
mammalian dsDNA (7) and uric acid crystals
(8) activate an inflammatory response.
The release of DAMPs is a common event,
as tissue damage and cell lysis are often associ-
ated with infections and lead to the release of
host molecules. Recognition of these DAMPs
by the immune system not only allows the sens-
ing of an ongoing infection and subsequent re-
cruitment of more immune cells, but also can
initiate the repair of the damaged tissue (9). It
seems then that the innate immune pathway not
only scans the cellular environment for signs
of invading pathogens, but also recognizes the
damage caused by them.
PAMPs are recognized by PRRs that are
either cytoplasmic, membrane-bound, or se-
creted; the most intensely studied are the
Toll-like receptors (TLRs). These receptors
are expressed by many cell types including
mononuclear, endothelial, and epithelial cells.
Once activated by PAMPs, the TLRs induce
different signaling cascades depending on the
adaptor protein, ultimately leading to the ac-
tivation of the transcription factors NF-κB,
AP-1, and interferon-regulatory factor (IRF)-
3. TLR activation results in the production
of antimicrobial peptides, inflammatory cy-
tokines and chemokines, tumor necrosis factor
(TNF)-α, and costimulatory and adhesion
230 Martinon
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 2
ANRV371-IY27-09 ARI 16 February 2009 8:57
molecules, as well as in the upregulation of ma-
jor histocompatibility complexes (MHCs). As
one given pathogen does not trigger only one
specific TLR, but rather a set of TLRs leading
to the expression of different proteins depend-
ing on the nature of the activated TLRs, the
immune system is instructed on the type of the
invading microorganism and mounts the most
appropriate response to fight it. Excellent re-
views on the biology of TLRs have been pub-
lished recently (10–12). TLRs are therefore not
the focus of this review.
Recently, two other families of PRRs were
described: the NLRs (NOD-like receptors) and
the RLHs (RIG-like helicases). Unlike TLRs,
these families consist of soluble proteins that
survey the cytoplasm for signs that advertise the
presence of intracellular invaders. Two RNA
helicases, namely RIG-I and MDA5 (13–15),
were identified as cytoplasmic, viral RNA sen-
sors. Upon viral stimulation of the two RLHs,
NF-κB and IRF3/7 are activated and, in turn,
induce the transcription of type I interferon
(IFN). Based on the CARD (caspase recruit-
ment domain) homology with RIG-I or MDA5,
the CARD adaptor (Cardif, also known as IPS-
1, MAVS, or VISA) that induces IFN-β was
identified (16–18). Both the CARD domain and
the mitochondrial localization of Cardif are re-
quired to induce NF-κB and IRF3/7 activation
(16, 18).
Studies on TLR signaling pathways and
the analysis of key TLR-deficient mice re-
vealed that TLRs could not be the only in-
nate immune receptors responsible for cytokine
production. Indeed, computational analysis of
the genome identified the NLR proteins.
NLR proteins are intracellular LRR (leucine-
rich repeat)-containing proteins that resemble
plant disease–resistance genes. The character-
ization of these NLRs has advanced greatly
in recent years, underscoring their essential
roles in innate immunity. In particular, cy-
toplasmic complexes called inflammasomes,
in which the scaffolding and sensing pro-
teins are members of the NLR family, have
been found to be central platforms of innate
NOD-like receptor
(NLR): cytosolic
protein involved in
innate immune sensing
RIG-like helicase
(RLH): cytosolic
helicase involved in
innate immune sensing
of nucleic acids
molecular complex
involved in the
activation of
inflammatory caspases
resulting in the
processing of
immature proIL-1β
and proIL-18 into
their mature forms
NOD signalosome:
complex that is
assembled upon
oligomerization of
NOD1 or NOD2 that
activates RIP2 and
triggers NF-κB
In this review, we examine the remarkably
important and emerging functions of inflam-
masomes as guardians of the body. We begin
by describing the general molecular nature of
the inflammasome complexes and the known
pathways that activate them. We then highlight
the current understanding of the function of
this pathway—its role in orchestrating host de-
fenses and in the pathogenesis of inflammatory
In recent years, the central role of the NLR
family has become increasingly appreciated (19,
20). NLRs form central molecular platforms
that organize signaling complexes such as in-
flammasomes and NOD signalosomes. Most
NLRs are expressed in the cytosol. Structurally,
NLRs are multidomain proteins with a tripar-
tite architecture containing a C-terminal region
characterized by a series of LRRs, a central
nucleotide domain termed the NACHT do-
main (also referred to as NOD domain), and an
N-terminal effector domain (Figure 1).
The LRR domain has been implicated in
ligand sensing and autoregulation of NLRs,
yet the precise mechanism of how NLR LRRs
sense their ligands is largely unknown. The
LRR is a widespread structural motif of 20–30
amino acids with a characteristic pattern rich
in the hydrophobic amino acid leucine. LRR
domains are formed by tandem repeats of a
structural unit consisting of a β strand and an
α helix, and are organized in such a way that
all the β strands and the helices are parallel
to the same axis, resulting in a nonglobular,
horseshoe-shaped molecule with the curved β
parallel sheet lining the inner circumference
and the α helices lining the outer circumfer-
ence (21). These modules are associated with
a wide range of functions including a role as
pathogen sensors in various innate immune re-
ceptors such as TLRs and NLRs. TLRs con-
tain LRRs that recognize or sense the pres-
ence of a wide range of PAMPs including LPS,
lipoproteins, flagellin, and RNA from bacteria
or viruses. They are believed to sense directly
The Inflammasomes 231
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 3
ANRV371-IY27-09 ARI 16 February 2009 8:57
NOD-like receptors
Figure 1
Domain organization of representative NOD-like receptors (NLRs). NLRs are characterized by three distinct domains: the
ligand-sensing leucine-rich repeats (LRRs); the NACHT domain, which is responsible for the capacity of NLRs to oligomerize; and
the effector domain, which can be a pyrin domain (PYD), CARD, or BIR domain. Most NLRs also contain a NACHT-associated
domain (NAD) C-terminal of the NACHT domain. NLRs are divided into two large subfamilies: the 14 members of the
PYD-containing NALP clan and the five members of the NODs and CIITA. IPAF and the BIR-containing NAIP form the remaining
NLR members. For comparison, the structural organization of a plant NLR-like gene and APAF1 are shown. (Abbreviations: CARD,
caspase recruitment domain; PYD, pyrin domain; FIIND, function to find; NACHT, domain conserved in NAIP, CIITA, HET-E and
TP1; NAD, NACHT-associated domain; BIR, baculovirus IAP repeat; AD, activation domain; NALP, NACHT, LRR, and
PYD-containing protein; CIITA, MHC class II transcription activator; IPAF, ICE-protease-activating factor.)
oligomeric structure
that is assembled when
APAF1 interacts with
cytochrome c released
from mitochondria;
triggers apoptosis by
activating caspase-9
STAND family of
NTPases: Subfamily
of AAA+ NTPases
that includes NLRs
AAA+ NTPases:
superfamily of
ATPases associated
with a variety of
cellular activities and
characterized by their
extended P-loop
ATPase domain
capable of forming
or indirectly their activating PAMPs. Double-
stranded RNA and lipopeptides have recently
been shown to bind TLR3 and TLR1/TLR2
complexes, respectively, whereas LPS-induced
TLR4 activation is presumed to be indirect and
to involve binding of LPS to MD2 (22, 23). In
contrast, no experimental data have convinc-
ingly demonstrated a direct interaction between
the LRRs of NLRs and their respective activa-
tors, suggesting that sensing of pathogens and
other signals by NLRs may be indirect.
The NACHT domain, which is central to
all NLRs, has similarity to the NB-ARC mo-
tif of the apoptotic mediator APAF1. APAF1
performs its cellular function through the for-
mation of a caspase-9 activating, heptameric
platform termed an apoptosome. The NB-
ARC domain is responsible for dATP/
ATP-dependent oligomerization of APAF1
upon cytochrome c binding, a process that initi-
ates apoptosis. Both the NACHT and NB-ARC
domains belong to the recently defined
STAND family of NTPases (24). Oligomeriza-
tion has been reported for several STAND fam-
ily proteins, as well as in other related AAA+
NTPases. Similarly, it is believed that the
crucial step in NLR activation lies in the
oligomerization of the NACHT domain,
thereby forming active, high molecular weight
complexes that characterize inflammasomes
and NOD signalosomes (25, 26).
NLR subfamilies differ in their N-terminal
effector domains, which mediate signal trans-
duction to downstream targets, leading to ac-
tivation of inflammatory caspases by inflam-
masomes or NF-κB by NOD signalosomes.
The vast majority of NLRs harbor a death-fold
domain at the N terminus, which is either a
CARD or a pyrin domain (PYD) (Figure 1,
Table 1). The death-fold domain superfamily
was originally described in proapoptotic signal-
ing pathways and, in addition to CARD and
232 Martinon
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 4
ANRV371-IY27-09 ARI 16 February 2009 8:57
Table 1 NOD-like receptors (NLRs)
NLR Commonly used nomenclature
human mouse Other names and aliases Structure
NALP3 Cias1, Pypaf1, Mmig1 PYD-NACHT-NAD-LRR
NALP12 Pypaf7, Monarch1, RNO2,
(Continued )
The Inflammasomes 233
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 5
ANRV371-IY27-09 ARI 16 February 2009 8:57
Table 1 (Continued )
NLR Commonly used nomenclature
human mouse Other names and aliases Structure
NAIPb Birc1b, Naip-rs6, NAIP2 BIR3x-NACHT-LRR
NAIPc Birc1c, Naip-rs5, NAIP3 BIR3x-NACHT-LRR
NAIPd Birc1d, Naip-rs2, NAIP4 BIR3x-NACHT-LRR
NAIPe Birc1e, Naip-rs3, NAIP5 BIR3x-NACHT-LRR
NAIPf Birc1f, Naip-rs4, NAIP6 BIR3x-NACHT-LRR
NLRs are divided into two large subfamilies: the 14 members of the PYD-containing NALP clan and the five members of the NODs and CIITA. IPAF
and the BIR-containing NAIP form the remaining NLR members.
PYD, includes the death domain (DD) and the
death effector domain (DED) (27). The death-
fold domain is characterized by six α helices
that are tightly packed in a Greek key fold and
form trimers or dimers with other members
of the same subfamily. In most known cases
a DD interacts with a DD, a CARD with a
CARD, a DED with a DED, and a PYD with a
PYD. These four domains are frequently found
in pathways that lead to the activation of cas-
pases or that activate the transcription factor
NF-κB. The observation that every death-fold-
containing family member is able to interact
with another partner harboring the same do-
main was instrumental in the identification of
major signaling pathways involved in apoptosis
and immunity. The death fold acts as a molec-
ular velcro that bridges receptors to adaptors
and effector proteins. Similarly, the N-terminal
PYD or CARD present in NLRs recruits PYD-
or CARD-containing molecules to signaling
NLR Subfamilies
Structurally and functionally, NLRs are di-
vided into subfamilies. In this review, we use
the nomenclature that is most commonly used
and highlight the various NLR subfamilies. An
overview of the family members and their his-
torical or alternative nomenclature is listed in
Table 1. The NLR family emerged almost
ten years ago with the cloning of NOD1 and
NOD2 and the subsequent identification of
234 Martinon
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 6
ANRV371-IY27-09 ARI 16 February 2009 8:57
the PYD domain (28–31). Based on phyloge-
netic distribution, we distinguish three differ-
ent NLR subfamilies, all characterized by a spe-
cific molecular structure (29, 32) (Figure 1).
NALPs represent the largest NLR subfamily
and have 14 genes identified in humans (29)
(Table 1). Some of them, such as NALP1,
NALP2, and NALP3, were shown to be the
central scaffold of caspase-1-activating com-
plexes known as inflammasomes. NALP pro-
teins harbor a NACHT and a LRR and are
characterized by an N-terminal PYD domain.
Interestingly, the LRR region within NALPs
is organized in the genome in a very con-
served and precise manner. NALP LRR re-
gions are formed by tandem repeats of ex-
ons of exactly 171 nucleotides and are defined
completely by a preserved intron-exon struc-
ture. Each exon encodes one central LRR and
two halves of the neighboring LRRs (33). The
phasing and position of the introns are consis-
tent with rapid and efficient exon amplification
during evolution. This particular modular or-
ganization possibly allows extensive alternative
splicing of the LRR region without disturbing
the three-dimensional fold of the region and,
as a consequence, maximizing variability in the
ligand-sensing unit.
Evolutionarily, IPAF and NAIP group to-
gether and are well separated from other NLRs.
IPAF contains an N-terminal CARD, whereas
NAIP has three BIR domains, which are often
found in proteins involved in apoptosis such as
the IAP family of caspase inhibitors. Both IPAF
and NAIP are involved in the formation of in-
flammasomes, either alone or in combination
with one another.
The third class of NLRs includes the
remaining CARD-containing NLRs such as
NOD1, NOD2, NOD4, and CIITA. This
clade also contains a slightly separated group
with NOD3 and NOD5/NLRX1 (19, 32).
NOD5/NLRX1 does not have a defined N-
terminal domain, whereas at least one splice
variant of CIITA has been reported to harbor
a CARD. NOD1 and NOD2 activate the tran-
scription factor NF-κB, a major regulator of
inflammatory responses. CIITA regulates the
transcriptional regulation of genes encoding
MHC II (34). NOD5/NLRX1 is recruited to
the outer membrane of mitochondria (35). The
function of NOD5/NLRX1 is still controver-
sial. One study suggested that NOD5/NLRX1
interacts and negatively regulates the antivi-
ral pathway involving the CARD-containing
adapter MAVS/CARDIF/IPS-1/VISA (36),
whereas another study proposed that NLRX1
promotes the production of reactive oxygen
species (ROS) (37). NOD3 and NOD4 have
no identified functions yet.
NLR Expression Patterns
and Gene Regulation
Expression patterns of most NLRs in vari-
ous cell populations and tissues have not yet
been studied in detail. Nevertheless, the im-
portance of NLRs in defense strategies of the
body is supported by the fact that several NLRs
are expressed in cells and tissues that have a
role in immunity such as phagocytes. Some
NLRs are also critical in epithelial cells, which
form the first barrier of defense against bac-
teria in human tissues and express NOD1,
NOD2, NALP3, and NAIP (20, 38, 39).
NALP1 is widely expressed, whereas NALP3 is
found mainly in immune cells, epithelial cells,
and osteoblasts (40). NAIP and IPAF are ex-
pressed in the brain and in macrophages and
macrophage-rich tissues such as spleen, lung,
and liver (41, 42). The expression of some
NLRs seems to be highly restricted; for ex-
ample, NALP5, NALP8, NALP4, NALP7,
NALP10, and NALP11 are mainly expressed
in germ cells and preimplantation embryos
(43). Most NLRs may be induced by other
branches of the innate immunity as part of a reg-
ulatory network. TLR stimulation, for exam-
ple, increases the expression of NLRs, such as
NOD1, NOD2, and NALP3, possibly reflect-
ing enhancement of NLR responses after TLR
stimulation (44).
NLRs: Lessons from Evolution
Ever since the discovery of NLRs in mammals,
the similarity of these genes with a family of
The Inflammasomes 235
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 7
ANRV371-IY27-09 ARI 16 February 2009 8:57
plant genes involved in immune defenses has
cross-fertilized NLR research (28). The plant
genes, known as R-genes (R for resistance), are
crucial for the immune defense of plants against
bacteria, fungi, viruses, and other pathogens.
The largest known class of R-genes structurally
resembles mammalian NLRs. They have a
C-terminal LRR, a central oligomerization
module related to the NB-ARC subtype of
STAND domains, and an N-terminal effector
domain that is generally either a coiled-coil do-
main or a TIR domain (45). The TIR domain
is a well-known recruitment domain involved
in the TLR and IL-1R family of immune me-
diators. There are more than 150 NLR-like
R-genes in Arabidopsis (46). Many of these genes
have been implicated in sensing pathogens. The
vast repertoire of NLRs in plants is believed
to be the result of a complex host-pathogen
race that promoted the evolution of specific
NLR genes that genetically interact with spe-
cific avirulence genes from distinct pathogens
(47). Although plant NLR-like proteins are
functionally and structurally similar to mam-
malian NLRs, there is no evidence for a com-
mon evolutionary origin. It is more likely that
these innate immune sensors are an exam-
ple of convergent evolution. The NLR struc-
ture (based on a C-terminal LRR sensing unit,
a STAND oligomerization module, and an
N-terminal recruitment domain) probably
originated independently through evolution,
which emphasizes key molecular constraints
required to design successful innate immune
systems (48, 49). Supporting the convergent
evolution idea, no NLR-like proteins have been
found in insects. In the animal kingdom, the
first evolutionarily conserved NLRs are ob-
served in the echinoderm sea urchin.
The observation by Mechnikov of phagocy-
tosis in echinoderms is considered to be a crit-
ically defining moment that led to the original
concept of immune defenses and, more specifi-
cally, innate immunity (50). More than one cen-
tury after these experiments, the sequencing of
the sea urchin genome provided us with another
appreciation of the complexity and general con-
servation of innate immune system mechanisms
(51, 52). The sea urchin genome contains 222
TLRs and 203 NLRs. Interestingly, most of the
NLRs are expressed in the gut, suggesting that
gut-related immunity is a likely driving force
behind the expansion of this family in echino-
derms (51). These findings highlight NLRs as
well as TLRs as evolutionarily important im-
mune genes that preceded the acquisition of the
adaptive immune system in vertebrates.
In nonmammalian vertebrates such as ze-
brafish, three distinct families of NLRs have
been identified (53, 54). The first subfam-
ily is related to the NOD subclass of mam-
malian NLRs and contains orthologs of NOD1,
NOD2, NOD3, and NOD4. The second sub-
family resembles the NALPs and contains at
least six genes. Finally, the last subfamily has the
highest similarity with the NACHT domain of
NOD3 and has expanded in teleost fish into sev-
eral hundreds of predicted genes. Most of these
genes encode a PYD domain at the N terminus
(similar to the one that characterizes NALPs).
Interestingly, some of these NLRs have a
C-terminal extension following the LRR that
contains a PRY-SPRY domain. The PRY-SPRY
domain is also found in human Pyrin, a PYD-
containing regulator of the inflammasome (see
below). The precise role of the PRY-SPRY in
Pyrin or in nonmammalian, vertebrate NLRs is
unknown; however, this finding further high-
lights the role of Pyrin and possibly other
PRY-SPRY-containing proteins in the regula-
tion of NLRs (55, 56).
Extensive diversification of the NLRs oc-
curred also within the mammalian lineage; this
is particularly true for NALPs that mainly
evolved through gene duplication events. Some
NALPs such as NALP2 and NALP7 in hu-
mans are clearly paralogs, whereas others, such
as NALP4 and NALP9, expanded only in the
mouse (Table 1). A similar evolutionary trend
was followed by NAIP in mice, where the locus
expanded to seven NAIP genes (33).
All these observations indicate that the NLR
repertoire within a species, and across verte-
brates, is large. Some of these NLRs have un-
dergone lineage-specific amplification, such as
NOD3-related NLRs in zebrafish or NALPs in
236 Martinon
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 8
ANRV371-IY27-09 ARI 16 February 2009 8:57
mammals. Genes involved in interactions with
pathogens are likely to diversify by undergoing
lineage-specific expansion (57), reflecting the
adaptive dynamics of a species to new environ-
ments with emerging pathogens.
NOD Signalosomes
NOD1 was one of the first NLRs to be de-
scribed (58, 59). Both NOD1 and NOD2,
once activated, recruit and engage the ki-
nase RIP2 through CARD-CARD interac-
tions. Oligomerization of RIP2 in the NOD
signalosome results in the activation of the
transcription factor NF-κB (60). Recently, the
CARD-containing protein CARD9 has also
been found to interact with NOD2 and RIP2
and to be involved in the activation of JNK
and p38 by NOD2 (61). Both NOD1 and
NOD2 detect muropeptides released from bac-
terial PGNs. NOD1 and NOD2 sense dis-
tinct PGN structures. Whereas NOD2 detects
muramyl dipeptides (MDP), the largest motif
common to Gram-negative and Gram-positive
bacteria, NOD1 detects meso-diaminopimelic
acid (meso-DAP), which is mainly found in
Gram-negative bacteria (62, 63). Both PGNs
are degradation products of bacterial cell wall
components released by intracellular or phago-
cytosed bacteria. Pathogens that do not reach
the intracellular compartment of the host cell
may use specialized secretion systems to in-
ject the PGN fragment into the host cytosol.
Helicobacter pylori, for example, elicits NOD1
activation by delivering PGN fragments into
the host cell through a mechanism that requires
a functional type IV secretion system (64).
NOD1 and NOD2 are crucial innate im-
mune receptors in epithelial cells, where
they are important to control infection via
the gastro-intestinal route, for example, by
H. pylori and Listeria monocytogenes (64, 65). Im-
portantly, mutations in NOD2 have been asso-
ciated with Crohn’s disease, a form of inflamma-
tory bowel disease. Most NOD2 mutations in
these patients affect the LRR region of NOD2
and are believed to disrupt the protein’s ability
to sense bacteria (66). This probably confers a
loss of tolerance toward commensal bacteria or
allows the proliferation of pathogenic bacteria
in the gut. Gain-of-function mutations in the
NACHT domain of NOD2 have been shown
to be responsible for Blau syndrome, a rare au-
toinflammatory disorder starting in childhood
and characterized by skin rashes, uveitis, and
joint inflammation (67). More recently, NOD2
has been suggested to play a role in the ac-
tivation of some types of inflammasomes (68,
69). These findings are discussed in more detail
The term inflammasome was coined to describe
a high molecular weight complex that activates
inflammatory caspases and the cytokine IL-
1β (70). Inflammasome is assembled from the
word inflammation—to reflect the function of
this complex—and the suffix “some” from the
Greek soma meaning body, which is frequently
used in cell biology to define entities or molec-
ular complexes such as proteasome, liposome,
ribosome, etc. Importantly, the term inflamma-
some was also chosen to highlight structural and
functional similarities with another well-known
caspase-activating complex, the apoptosome, a
molecular platform that triggers apoptosis (71).
Inflammasomes Lead to Activation
of Inflammatory Caspases
Caspases are proteases produced in cells as cat-
alytically inactive zymogens and usually un-
dergo proteolytic processing during activation
(72). The subset of caspases that cleave sub-
strates during apoptosis are known as exe-
cutioner caspases. These executioner caspases
(caspase-3, -6, and -7 in mammals) are gener-
ally activated by the initiator caspases such as
caspase-8, caspase-10, caspase-2, or caspase-9.
Initiator caspases harbor an N-terminal death-
fold domain (CARD or DED in mammals) that
is required for the activation of their C-terminal
catalytic region. The mechanism of activation
of initiator caspases depends on the engage-
ment and activation of platforms such as the
The Inflammasomes 237
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 9
ANRV371-IY27-09 ARI 16 February 2009 8:57
In mammals, the inflammatory caspases include human and
murine caspase-1, human and murine caspase-12, murine
caspase-11, and the two caspase-1-related human caspases,
caspase-4 and caspase-5 (33, 204). Inflammatory caspases in mam-
mals have a CARD domain followed by a domain containing the
catalytic residue cysteine. These caspases are termed inflamma-
tory because the main substrates of caspase-1 identified to date
are cytokines (such as IL-1β, IL-18, and possibly IL-33) that
are crucial mediators of the inflammatory response. Caspase-1
was the first caspase to be discovered in mammals, but only re-
cently have the pathways (inflammasomes) leading to its activa-
tion been discovered (70). Although both human caspase-5 and
mouse caspase-11 have been associated with caspase-1 activation,
no specific substrates have been described for them. Caspase-5 is
recruited by the C-terminal CARD of NALP1, suggesting that
it may be involved in the activity of inflammasomes harboring
NALP1. Caspase-12 and caspase-4 are activated by endoplas-
mic reticulum (ER) stress, an adaptive response that copes with
protein overload in the ER. However, the function of these in-
flammatory caspases upon ER stress is unclear (75). Moreover,
caspase-12 appears to be an inhibitor of the inflammasome, pos-
sibly by interfering with caspase-1 activation, a process that has
been associated with susceptibility to sepsis (209).
death-inducing signaling complex (DISC) for
caspase-8 and -10, the PIDDosome for caspase-
2, and the apoptosome for caspase-9 (73). These
platforms integrate cellular signals, recruit ini-
tiator caspases via their death-fold domain, and
promote dimerization of the caspases, which
all lead to the formation of an active enzyme
proficient enough to initiate specific signaling
cascades (74). Inflammasomes activate a class
of caspases known as inflammatory caspases
(25, 75) (see sidebar, Inflammatory Caspases).
An increasing number of studies highlight the
importance and complexity of inflammatory
caspase activation.
Prototypical Inflammasomes
Although the biochemistry and diversity of in-
flammasomes are still poorly understood, we
distinguish three prototypes of inflammasomes:
The NALP1 inflammasome, the NALP3 in-
flammasome, and the IPAF inflammasome. For
several NALPs, there is evidence for their roles
as scaffolding proteins of inflammasomes (76).
It is assumed that the PYD of NALPs inter-
acts and recruits the adaptor ASC (apoptosis-
associated speck-like protein containing a
caspase recruitment domain) via PYD-PYD
interaction (Figure 2). ASC contains an N-
terminal PYD and a C-terminal CARD and is
an essential component for inflammasome for-
mation (70, 77). The CARD domain within
ASC binds and recruits caspase-1 to the in-
flammasome. NALP1 has a C-terminal exten-
sion that harbors a CARD, which was shown
to recruit caspase-5 (70) or a second caspase-1
(26). Other NALPs do not have the NALP1
C-terminal extension; instead, CARDINAL
(a protein very similar to the NALP1 C termi-
nus) interacts with other inflammasomes such
as the NALP3 or NALP2 inflammasome (78).
Neither CARDINAL nor NALP1 is highly
conserved in mice. The NALP1 locus in mice
contains three paralogs that have no functional
PYD, whereas CARDINAL is not present in
the mouse genome at all. It is therefore pos-
sible that NALP1 genes fulfill CARDINAL
functions in mice. IPAF has an N-terminal
CARD and directly recruits caspase-1 (41)
(Figure 2).
Basic mechanisms implicated in the activa-
tion of NLRs are also involved in inflamma-
some assembly. IPAF and NALP3 selectively
bind ATP/dATP, and nucleotide binding is nec-
essary for oligomerization of the NACHT do-
main (79, 80). Both IPAF and NALP3 bind
SGT1 and HSP90 (81, 82), two proteins whose
plant orthologs were previously shown to regu-
late and interact with plant NLRs. In mammals,
as in plants, the activity of the HSP90-SGT1
complex is essential for NALP3 activation,
probably for keeping the inflammasome inac-
tive but competent for activation. HSP90 has
been suggested to act upstream of NALP1 in-
flammasome activation by anthrax lethal toxin
238 Martinon
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 10
ANRV371-IY27-09 ARI 16 February 2009 8:57
Although the IPAF, NALP3, and NALP1
inflammasomes form prototypical inflamma-
some complexes, recent genetic evidence
suggests that other NLRs, such as NAIP or
NOD2, may be involved in forming inflam-
masomes or in modulating their activity. Ex-
actly how and at what step these proteins are
connected to the formation of inflammasomes
are unknown. It is also unclear whether multi-
ple NLRs may assemble as heterocomplexes to
form competent inflammasomes (84, 85). Thus,
the complexity and diversity of inflammasomes
may turn out to be considerable (86).
Although inflammasomes are emerging more
and more as key players in inflammatory and
immune responses, a growing number of stud-
ies reveal their function in the sensing of a con-
troversial signal: danger. It is well known that
a major function of the immune system is to
differentiate self from nonself—to respond to
self with tolerance and to mount an immune
response against nonself. Innate immunity, for
example, detects PAMPs from microbes, in-
cluding pathogens. There is also evidence
that innate immunity is able to discriminate
pathogenic microbes from nonpathogenic mi-
crobes or commensals; but this raises the ques-
tion of how the immune system interprets the
microbial environment allowing the discrimi-
nation between nonpathogenic and pathogenic
microbes (6). Matzinger and colleagues (87),
to account for some unsolved questions of the
self-from-nonself model, promoted an alterna-
tive hypothesis (the danger hypothesis), sug-
gesting that it is the presentation of an anti-
gen in the context of a danger signal that
triggers an efficient immune response, not sim-
ply the foreignness of the antigen. This led to
multiple studies identifying molecules and sig-
nals that are released by damaged or stressed tis-
sues and that trigger or modulate the immune
response (88, 89). Both the self-from-nonself
model and the danger model may synergize
to determine the quality and extent of the in-
NALP3 inflammasome
IPAF inflammasome
Figure 2
Structural organization of the typical NALP3 and IPAF inflammasomes. The
core structure of the NALP3 inflammasome is formed by NALP3, the adaptor
ASC, and caspase-1. PYD-PYD and CARD-CARD homotypic interactions are
crucial for the recruitment and activation of either the adaptor ASC or the
inflammatory caspases (left panel ). IPAF recruits caspase-1 directly via
CARD-CARD interactions (right panel ). The leucine-rich repeats of NALP3
or IPAF are proposed to sense the activating signals leading to the
oligomerization of the NACHT domain and initiating the formation of the
donut-shaped inflammasome. Based on the structure of the apoptosome, the
caspases and IL-1β processing activity most likely face toward the inside of the
donut (lower panel ).
nate immune response (90). Although the dan-
ger hypothesis was first proposed in mammals,
evidence for such a type of immune response
was first documented in plants. Plant immunity
relies on an effector-triggered immunity that
mainly detects pathogen-driven modifications,
stress, or danger signals in the infected host cell
(91). Similarly, the mammalian NLR NALP3
was found to be involved in sensing danger
The Inflammasomes 239
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 11
ANRV371-IY27-09 ARI 16 February 2009 8:57
Cell Disruption Activates the
Early observations that cell lysis in a hypotonic
buffer can lead to the processing of proIL-1β
provided an important model system that
permitted the identification of caspase-1 as
the protease responsible for the processing
of proIL-1β (92). Using this cell-free system,
investigators noticed that caspase-1 activation
was restricted to a few cell types, such as the
monocytic cell line THP-1. A similar assay
subsequently allowed the first biochemical
identification and characterization of an in-
flammasome complex (70). After disruption of
cellular integrity, the inflammasome is sponta-
neously formed. Assembly can be inhibited by
complementing the cell extracts with potassium
levels that mimic normal levels found in the cy-
tosol of healthy cells (over 70 mM) (70, 93, 94).
The observation that subphysiological amounts
of potassium are required for spontaneous
inflammasome formation suggests that the in-
flammasome may sense drops in potassium lev-
els. This possibility is supported by recent find-
ings demonstrating that various danger signals
and stimuli that activate the NALP3 inflam-
masome can trigger potassium efflux, thereby
lowering the cytosolic potassium concentration
of stimulated cells (93, 95) (Figure 3).
Sensing Extracellular ATP
Extracellular ATP serves as a danger signal
that alerts the immune system by binding to
the purinoreceptor P2X7, thereby activating
NALP3 and caspase-1 (8, 96–98). Although
ATP is emerging as an important modulator
of inflammation (99), the critical role of extra-
cellular ATP as a danger signal is unclear. The
amount of extracellular ATP that is required to
activate macrophages in vitro is relatively high
(2 to 5 mM), and, moreover, in vivo most of the
extracellular ATP may be rapidly hydrolyzed by
ectonucleotidases (100). ATP is released from
cells as a consequence of cell damage and/or cel-
lular stress. In endothelial as well as in epithe-
lial cells, ATP release is triggered by nonlytic,
mechanical stimuli as diverse as compression,
Bacterial toxins
A. hydrophila (Aerolysin)
Dinoflagellates (Maitotoxin)
L. monocytogenes (LLO)
S. aureus
Danger signals
Skin irritants
Viral DNA
NALP3 inflammasome
Figure 3
Multiple NALP3 inflammasome activators trigger cellular signals, such as potassium efflux and reactive
oxygen species, that eventually activate an inflammasome dependent on caspase-1, ASC, and NALP3. Note
that muramyl dipeptide (MDP) activation of caspase-1 may also require the NOD-like receptors NOD2 and
NALP1. (Abbreviations: CPPD, calcium pyrophosphate dihydrate crystals; MSU, monosodium urate
crystals; PAMP, pathogen-associated molecular pattern; UV, ultraviolet light.)
240 Martinon
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 12
ANRV371-IY27-09 ARI 16 February 2009 8:57
hydrostatic pressure changes, and hypotonic
shock (101). ATP release can also occur through
secretory organelles that store large amounts
of ATP (102). Adrenal medullary chromaffin
granules, which may be released upon physical
or psychological stress, have concentrations of
ATP around 100 mM, whereas platelet-dense
granules contain concentrations of ATP that
can reach 500 mM (100 times more than the
cytosolic concentration). Similarly, secretion of
insulin-containing granules from pancreatic β
cells releases the ATP pool stored in the gran-
ules (103). ATP can also be released from mi-
crobial flora and pathogens. Salivary histatins,
for example, trigger ATP efflux from Candida
albicans, increasing extracellular ATP, a mech-
anism that may contribute to the antifungal
properties of these proteins (104). Exposure of
cells to extracellular ATP activates caspase-1
(105). Several studies have demonstrated that
ATP-induced caspase-1 activation and subse-
quent IL-1β maturation requires activation of
the purinoreceptor P2X7 in combination with
another type of channel, the pannexin-1 chan-
nel (106). Interestingly, pannexin-1, besides its
role as a gap junction protein, can act as a spe-
cific ATP release channel (107). This suggests
an amplifying mechanism for P2X7-mediated
inflammasome activation via pannexin-1, which
is indeed observed, at least in vitro (108).
The generation of ASC-deficient mice
demonstrated that ATP-mediated caspase-1 ac-
tivation requires ASC, and it was therefore
probably dependent on the activation of a
NALP protein (77, 109). This hypothesis was
confirmed in studies using NALP3-deficient
mice (8, 96–98), demonstrating that extracel-
lular ATP can act as a danger signal to acti-
vate the NALP3 inflammasome and promote
caspase-1 activation and IL-1β maturation.
Although extracellular ATP has been shown
to be involved in inflammatory conditions,
as in asthmatic airway inflammation (110),
the physiological significance of extracellular
ATP-mediated NALP3 inflammasome ac-
tivation still remains to be demonstrated
in vivo.
Uric Acid: A Danger Signal
Involved in Gout
In addition to ATP, cells release other danger
signals to activate the immune system. In a
seminal paper, the Rock laboratory (111) puri-
fied, from the supernatant of dying cells, a low
molecular fraction that could trigger adjuvan-
ticity in vivo and identified uric acid as the active
compound of that fraction. Uric acid is the end
product of the cellular catabolism of purines
and is present at near saturating amounts in
body fluids and at much higher concentration
in the cytosol of healthy cells. It is believed that
extracellular uric acid coming in contact with
the high levels of free sodium present in the
extracellular environment nucleates and forms
monosodium urate (MSU) crystals. MSU is
considered to be the biologically active struc-
ture that is responsible for the adjuvantic effect
of uric acid. Therefore, formation of this dan-
ger signal is the result of a multistep process that
starts with the release of uric acid. The biologi-
cal activity of MSU, including its adjuvanticity,
depends on the activation of the NALP3
inflammasome and the production of IL-1, but
not on TLRs (8, 112, 113). MSU stimulates
the NALP3 inflammasome to produce active
IL-1β (8), and macrophages from mice defi-
cient in components of the inflammasome, such
as caspase-1, NALP3, and ASC, have a highly
reduced crystal-induced IL-1β activation
The in vivo relevance of uric acid signals has
been addressed in a follow-up study by Rock
and colleagues (114), who showed that elimi-
nation of uric acid reduced the generation of
cytotoxic T cells to an antigen in transplanted
syngeneic cells and the proliferation of autore-
active T cells in a transgenic diabetes model.
It has also been suggested that erythrocytes
infected with Plasmodium parasites accumulate
high levels of the uric acid precursor hypox-
anthine, which is released and converted into
uric acid upon rupture of the erythrocytes, a
process that results in inflammation (115). Uric
acid release also occurs in DCs incubated in
the presence of alum (aluminum hydroxide),
The Inflammasomes 241
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 13
ANRV371-IY27-09 ARI 16 February 2009 8:57
Adjuvant: substance
that enhances the
capacity of an antigen
to stimulate the
immune system
which is the most widely used adjuvant in hu-
man vaccines (116). Interestingly, alum was
recently found to be a direct inflammasome
activator (113, 117–119) (see below).
High levels of circulating uric acid (hy-
peruricemia) has been associated with vari-
ous inflammatory diseases, including multiple
sclerosis, hypertension, and cardiovascular dis-
eases (120). Hyperuricemia and MSU forma-
tion are strongly linked to gout. The autoin-
flammatory disease gout is characterized by
arthropathies generated by the inflammatory
reaction to MSU in the joints and periarticu-
lar tissues (121). In a model of MSU crystal-
induced peritonitis in mice, impaired inflamma-
tion is found in inflammasome-deficient mice
or mice deficient in the IL-1 receptor (IL-1R)
(8, 112), suggesting that the inflammatory re-
sponse in gout is dependent on the inflamma-
some. The importance of IL-1 in the pathology
of gout is supported by promising preliminary
clinical trials in patients with acute gout. Pa-
tients responded positively to the injection of
the IL-1R antagonist IL-1ra (122, 123). Sim-
ilarly, patients with pseudogout, an inflamma-
tory disease caused by the deposition of calcium
pyrophosphate dihydrate crystals (CPPD), an-
other type of pathogenic microcrystal that acti-
vates the NALP3 inflammasome (8), responded
well to treatment with the IL-1ra (124).
Silica and Asbestos and Inflammation
in the Lung
Alveolar macrophages reside in the respiratory
surfaces, one of the major boundaries between
the body and the outside world. These cells are
phagocytes that play an important role in host
defenses against microorganisms and remove
particles such as dust. Silica and asbestos dust
are particularly strong inflammation inducers
in the lungs (125). Macrophages can dissolve
MSU, but are not able to efficiently eliminate
microparticles of silica and asbestos. Inhaling
finely divided crystalline silica or asbestos dust
in very small quantities over time can lead to
inflammatory conditions known as silicosis and
asbestosis, respectively (126). As the dust be-
comes lodged in the lungs, continuous irrita-
tion ensues, resulting in chronic inflammation
that favors the development of cancer. This, in
particular with asbestos, is associated with the
development of malignant mesotheliomas.
Similar to what was found for MSU, asbestos
and silica microparticles activate the NALP3
inflammasome (127–129). It is significant to
note that pulmonary inflammation is greatly re-
duced in NALP3-deficient mice after in vivo
inhalation of asbestos or silica (127, 128).
Aluminum Particles: An
Inflammasome-Dependent Adjuvant
Vaccine adjuvants are exogenic preparations
that boost the immune response to achieve pro-
tective immunity. Most adjuvants activate in-
nate immune receptors such as TLRs (130).
More recently, NLRs and inflammasomes were
found to respond to specific adjuvants. Alum has
been the most widely used adjuvant in human
vaccination for more than half a century (131).
Among the early observations on the adjuvantic
effect of alum was that immunization with this
adjuvant led to an increase in antigen-induced
T cell proliferation, apparently resulting from
the augmented production of IL-1 (132). Anti-
IL-1 antibodies are able to inhibit an antigen-
specific T cell proliferative response after im-
munization with alum adjuvant, but not with
Freund’s complete adjuvant (133). The abil-
ity of alum to activate caspase-1 and produce
active IL-1β and IL-18 was demonstrated in
vitro (134). Alum-induced caspase-1 activation
depends on the NALP3 inflammasome (113,
117–119). Mice deficient in NALP3, ASC, or
caspase-1 fail to mount a significant antibody
response to antigen immunization with alum
adjuvants (113, 117, 119), confirming that the
NALP3 inflammasome is a key player that links
the innate immune system with the adaptive
immune system.
Inflammasomes and Inflammation
in the Skin
The skin is the body’s first line of de-
fense against external threats and serves as an
242 Martinon
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 14
ANRV371-IY27-09 ARI 16 February 2009 8:57
effective barrier against ordinary environmen-
tal intrusions. As such, the skin is often
damaged by various insults and proficient in
mounting efficient immune responses. Ultra-
violet irradiation, for example, was recently
shown to activate the NALP3 inflammasome
and promote IL-1β maturation in keratinocytes
(135). A role of the inflammasome in the skin
was also found in contact hypersensitivity, an
inflammatory disease caused by irritant chem-
icals that penetrate the skin surface and that
induce a T cell–mediated immune response
(136). This response is divided into two phases.
The first is a sensitization phase in which the
sensitizing chemical acts both as an adjuvant
and as a foreign hapten. Uptake of the chem-
ical by skin-resident antigen-presenting cells
(APCs) and their migration to draining lymph
nodes ensue, promoting T cell priming. Reex-
posure to the chemical defines a second phase,
also known as the elicitation phase. Here, chal-
lenge with the corresponding antigen triggers
the activation of primed T cells. The innate
immune module of the sensitization phase de-
pends on the presence of functional caspase-1,
IL-1β, and IL-18 (137–140), suggesting a po-
tential involvement of the inflammasome. The
role of the inflammasome was confirmed in
ASC- and NALP3-deficient mice that showed
an impaired contact hypersensitivity response
to the irritants trinitrophenylchloride (TNP-
Cl) (97), trinitrochlorobenzene (TNCB), and
dinitrofluorobenzene (DNFB) (141). In these
mice, transfer of primed T cells results in a
normal contact hypersensitivity, suggesting that
only the sensitization phase requires NALP3
and ASC. Interestingly, DNFB promotes the
release of IL-1β in a caspase-1-dependent man-
ner in primary keratinocytes as well as in a DC
line, suggesting that the inflammasome may ei-
ther detect such compounds directly or, more
likely, may detect some danger signals released
or produced by these irritants (142). On the
contrary, dinitrothiocanobenzene (DNTB) is
unable to activate the inflammasome and, as
such, fails to induce a strong immune response
in vivo despite the fact that DNTB is compe-
tent in inducing the effector phase if the re-
cells (APCs): cells
that process antigens
and have special
molecules (MHC) that
bind the processed
antigens and display
them on the cell
surface for T cells to
lated antigen DNFB (that activates the inflam-
masome) was used for sensitization (143). To-
gether, these findings suggest that the inflam-
masome can bridge danger signals triggered by
the irritant effect of sensitizing chemicals with
the activation of IL-1β and IL-18, thus promot-
ing efficient activation of the adaptive immune
ROS, the Common NALP3 Activator?
Most danger signals described to activate the
NALP3 inflammasome trigger similar intracel-
lular changes that may converge on a common
mechanism of NALP3 activation (Figure 3).
Potassium efflux, the induction of frustrated
phagocytosis, and ROS production are the most
striking features associated with NALP3 acti-
vators (127). Large particles and crystals such
as MSU, alum, asbestos, and silica can induce
the so-called frustrated phagocytosis at the sur-
face of the cell, provoking the formation of
cytoskeletal filaments (144). Inhibition of cy-
toskeletal filament generation with the pharma-
cological agents cytochalasin D or colchicine
disrupts the ability of particles to trigger
IL-1β activation (8, 113, 127), suggesting that
the process of phagocytosis or frustrated phago-
cytosis is involved in NALP3 activation. In-
terestingly, colchicine (145) was one of the
first anti-inflammatory drugs identified for the
treatment of gout by the Greeks more than
fifteen centuries ago (146) and is still used in
modern medicine to treat inflammatory dis-
eases. ROS production occurs quickly upon ex-
posure of macrophages with silica or asbestos
dust (127, 128, 147, 148). Similarly, MSU and
alum produce ROS (113, 127). Both ATP and
the toxin nigericin (that do not require phago-
cytosis to activate the inflammasome) activate
ROS (149). A cellular redox imbalance also oc-
curs upon cellular stimulation with the skin
sensitizer DNCB, and ROS are also induced
by UV (150). Along this line, knockdown of
NAPDH subunits or the use of antioxidants
inhibit inflammasome activation induced by
alum, MSU, ATP, nigericin, asbestos, and silica
(93, 113, 127, 149). It is therefore reasonable to
The Inflammasomes 243
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 15
ANRV371-IY27-09 ARI 16 February 2009 8:57
Receptor for
advanced glycation
end product (RAGE):
involved in the sensing
of HMGB1, a danger
signal released by
injured cells
propose that ROS are either directly sensed
by NALP3 or indirectly sensed through cy-
toplasmic proteins that modulate inflamma-
some activity. ROS production by hydrogen
peroxide activates DCs in a similar way to
TLRs (151) and activates the inflammasome
ROS production is a well-known, highly
conserved signal involved in damage and stress
sensing. ROS are also important players in in-
nate immune responses in plants (152). Ara-
bidopsis mutants that contain disruptions of
NADPH oxidases fail to generate a full ox-
idative burst in response to infection by bac-
terial and fungal pathogens (48). Interestingly,
plant potassium efflux has been linked to ROS
production at the membrane (152). Potassium
efflux has also been implicated in NADPH
activation in granulocytes (153). It is therefore
possible that potassium efflux by NALP3 acti-
vators may be involved in ROS generation.
A key function of the innate immune sys-
tem is the recognition of invading microbes.
Responses to extracellular PAMPs and some
extracellular danger signals are mediated by
membrane receptors such as TLRs and RAGE
(receptor for advanced glycation end product).
However, NLRs are localized in the cytosol
and thus are specialized in sampling PAMPs
and danger signals that ultimately reach or af-
fect this particular cellular compartment. Al-
though microbes may reach the cytosol of cells
during their life cycles, degradation products
from phagocytosed bacteria and viruses may
also be present in the cytosol and contribute
to NLR and inflammasome activation (154).
By definition, PAMPs represent molecules vi-
tal for microbial survival and are therefore un-
likely to vary in their structures because any
major change would be detrimental. Exam-
ples of PAMPs are bacterial structural com-
ponents, such as LPS and PGNs, or viral nu-
cleic acids. PAMPs are signatures that define
classes of microbes and, as such, are critical
in alerting the immune system. In addition to
detecting PAMPs, inflammasomes also detect
toxins and signals that are restricted to certain
pathogens (155) (Figures 3 and 4). It is tempt-
ing to hypothesize that these signals may or-
chestrate specific innate immune responses as
a result of a unique host-pathogen coevolution
maximizing fitness for both the pathogen and
the host. The pathogen may benefit from viru-
lence to promote spreading replication and sur-
vival, whereas the host evolves to cope with the
infection (156).
Shigella flexneri
Legionella pneumophila
Salmonella typhimurium
Pseudomonas aeruginosa
factor Co-factors
Figure 4
Gram-negative pathogens secrete factors such as flagellin and possibly other virulence factors through type
III (T3SS) or type IV (T4SS) secretion systems to trigger an inflammasome dependent on caspase-1 and
IPAF. Genetic studies have demonstrated that caspase-1 activation in this context may also require the
adaptor ASC or the NOD-like receptor protein NAIP. Whether NAIP and ASC contribute to the formation
of IPAF inflammasomes directly or indirectly is unknown.
244 Martinon
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 16
ANRV371-IY27-09 ARI 16 February 2009 8:57
Inflammasome Activation by PAMPs
Inflammasomes respond to immunomodula-
tory PAMPs, mainly bacterial PGNs and nu-
cleic acids. PGNs are structural units of cell
walls common to all bacteria (157). Degrada-
tion of PGNs leads to the release of several
structural units including MDP. MDP is sensed
in the cytosol by the NLR NOD2, which ac-
tivates NF-κB. MDP also activates caspase-1
and IL-1β (158) via NALP3 in human mono-
cytes, suggesting that NALP3 is an additional
MDP sensor (69, 159, 160). The strength of
the immune response to MDP varies greatly
and depends on the animal species and genetic
background. Mice are much less sensitive than
humans, guinea pigs, or rats to these PGN-
derived peptides; moreover, C57BL/6 mice are
less sensitive than BALB/c mice (161, 162). In-
terestingly, both NF-κB and IL-1β activation
are greatly enhanced by MDP in the presence
of the protein synthesis inhibitor cycloheximide
(CHX), demonstrating that a CHX-sensitive
pathway may affect MDP internalization or its
presentation to NLRs (69). Genetic studies in
mice have shown that IL-1β activation by MDP
requires both NOD2 and NALP3, suggesting
that both these NLRs may cooperate either in-
directly or directly as part of the same molecular
complex (69). This finding is consistent with
the observation that monocytes from Crohn’s
disease patients who have functional mutations
in the NOD2 gene fail to activate IL-1β upon
MDP stimulation (163). Moreover, Muckle-
Wells patients that harbor a gain-of-function
mutation in the NALP3 gene overproduce
IL-1β upon stimulation with MDP (159). Sim-
ilarly, a probable gain-of-function mutation in
NOD2 in the mouse leads to increased IL-1β
production upon stimulation of macrophages
with MDP (164). NOD2 has also been sug-
gested to play a role, together with NALP1,
in MDP-induced caspase-1 activation, further
suggesting that multiple NLRs may cooperate
and synergize to mount host defenses (68, 165).
TLR9 and the intracellular protein DAI
have been identified as sensors for DNA result-
ing in the triggering of a type I IFN response
(166, 167). Similar to DAI, the inflammasome
is capable of sensing cytosolic DNA (168), al-
though this is unlikely to be direct. Infection
of a monocytic cell line or mouse macrophages
with adenoviruses and herpesviruses leads to ac-
tivation of caspase-1 and IL-1β. NALP3- and
ASC-deficient mice display reduced innate in-
flammatory responses to infection with aden-
ovirus. Inflammasome activation also occurs as
a result of transfected cytosolic bacterial, viral,
and mammalian (host) DNA; however, in this
case sensing is dependent on ASC only and not
on NALP3. It is also independent of TLRs and
IRFs (168). Studies have also identified RNA as
an activator of NALP3 (98, 169), although this
study could not be confirmed by other groups.
Pore-Forming Bacterial Toxins
Activate the NALP3 Inflammasome
Many bacterial pathogens produce toxins that
contribute to virulence by modifying host re-
sponses. Bacterial toxins are generally either
enzymes or pore-forming proteins (170). Most
of the bacterial toxins that activate NALP3
are pore-forming toxins. These toxins are re-
leased by bacteria in a soluble form and sub-
sequently polymerize into a ring-like structure
forming a pore in the membrane of the host
cell. Pore formation triggers ionic imbalance;
in particular, potassium efflux and calcium in-
fluxes are observed. These two processes are
common danger signals and are frequently as-
sociated with NALP3 inflammasome activation
(155). Among pore-forming toxins, α-toxin
from Staphylococcus aureus and aerolysin from
Aeromonas hydrophila are potent activators of the
NALP3 inflammasome (96, 171). Similarly, lis-
teriolysin O (LLO), a toxin released by Listeria
monocytogenesis, activates caspase-1 in an ASC-
and NALP3-dependent manner (96, 172).
Interestingly, ivanolysin O, a LLO-related cy-
tolysin that exhibits quite similar function re-
garding the contribution to the escape of
Listeria from the phagosome into the cytosol, is
unable to restore caspase-1 activation in LLO-
deficient strains, suggesting that these toxins
may engage specific signals beyond their ability
The Inflammasomes 245
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 17
ANRV371-IY27-09 ARI 16 February 2009 8:57
Virulence factor:
molecules produced by
pathogens that are
involved in host/
pathogen interaction
and aimed at
increasing the rate of
to form pores (173). The release into the cytosol
of flagellin during Listeria infection activates the
IPAF inflammasome (but not the NALP3 in-
flammasome, see below), highlighting the abil-
ity of pathogens to engage specific and multiple
inflammasomes (174).
Anthrax Lethal Toxin
Activates NALP1
Bacillus anthracis is the causative agent of an-
thrax and depends for its virulence on the se-
cretion of factors that form functional toxins.
Anthrax lethal toxin (LeTx) is one of the major
toxins produced by B. anthracis and is believed
to be responsible for causing death in systemic
anthrax infections.
Macrophages from inbred mice are either
susceptible or resistant to cell death in response
to LeTx. This trait difference has been mapped
to a locus on chromosome 11 and is associ-
ated with a polymorphism in the nalp1b gene
that impinges on caspase-1 activation (175).
How LeTx activates NALP1 and the role of
the inflammasome in anthrax pathology are
still unknown. Murine NALP1b does not con-
tain a PYD; hence, it is not clear whether
it requires ASC or dimerization with another
NALP for caspase-1 recruitment. On the other
hand, NALP1b possesses a CARD and a region
related to CARDINAL. It is therefore possi-
ble that this CARD-containing region is able
to activate caspase-1 in an ASC-independent
manner, as was suggested for human NALP1
in vitro (26). Activation of caspase-1 by LeTx
requires binding, uptake, and endosome acidi-
fication to mediate translocation of lethal fac-
tor (a functional subunit of LeTx) into the host
cell cytosol. Interestingly, catalytically active
lethal factor activates caspase-1 by a mechanism
involving proteasome activity and potassium
efflux (176–178).
IPAF Inflammasome Activation
by Injected Virulence Factors
IPAF and NAIP5 have been involved in the
detection of virulence factors from Gram-
negative bacteria. Recognition of Salmonella ty-
phimurium and Shigella flexneri activates the
IPAF inflammasome that requires the ASC
adaptor (77, 179). The role of ASC in the IPAF
inflammasome is still unclear, but ASC may
stabilize or facilitate caspase-1 recruitment to
IPAF. Alternatively, IPAF may cooperate with a
yet to be defined NALP to activate caspase-1.
In contrast to S. typhimurium and S. flexneri,
Legionella pneumophila requires two murine
NLRs, NAIP5 and IPAF, for inflammasome
formation, but does not require ASC (180–
182). The role of NAIP5 in IPAF inflamma-
some activation is unknown. It has been sug-
gested that defective NAIP5 signaling ren-
ders macrophages permissive to L. pneumophila
despite caspase-1 activation, suggesting that
NAIP5 may have additional functions be-
yond its role in IPAF and caspase-1 acti-
vation (183). Unlike the seven NAIP genes
found in the murine genome, humans only
harbor one copy. Consistent with findings in
the mouse, knockdown of NAIP or IPAF in
human cell lines leads to enhanced suscepti-
bility to L. pneumophila (38). IL-18 produc-
tion and protection against Anaplasma phago-
cytophilum, a neutrophilic obligate bacteria that
causes human anaplasmosis, relies on ASC and
caspase-1 and partially on IPAF, but not on
NALP3 (184). Similarly, Pseudonomas aeruginosa
specifically activates the IPAF inflammasome
Most Gram-negative pathogens that acti-
vate IPAF require the type III secretion system
(T3SS) or the type IV secretion system (T4SS)
to inject into the host cell IPAF-activating vir-
ulence factors, mainly flagellin. These bacte-
rial injection machines span the two membranes
from the bacteria and the host cell membrane
(84). Polymers of flagellin form the flagella,
a structure anchored to the bacterial cell wall
that enables bacterial motility. Flagellin is a
well-known activator of host innate immunity
through its capacity to trigger TLR5 activation.
In this context, flagellin is considered a PAMP
as it is a vital, evolutionarily conserved ele-
ment of mobile bacteria. In the context of IPAF
activation, flagellin could be interpreted as a
246 Martinon
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 18
ANRV371-IY27-09 ARI 16 February 2009 8:57
virulence factor, as flagella formation is not re-
quired and flagellin release in the cytosol is
apparently the result of an active process de-
pendent on T3SS or T4SS injection systems
(84). Interestingly, S. flexneri, a pathogen that
does not have a flagellum and apparently does
not express flagellin, still requires the T3SS
for IPAF-induced caspase-1 activation, indicat-
ing that flagellin may not be the only T3SS
virulence factor used by pathogens to activate
IPAF inflammasomes (188). How flagellin or
other virulence factors engage IPAF inflamma-
somes as well as the precise function of ASC or
NAIP in detecting these signals and activating
caspase-1 are fascinating questions that need to
be solved in future studies.
Little is known about the mechanisms that reg-
ulate inflammasome activity. Inflammasome,
caspase-1, and IL-1β activation are best per-
formed in cells that express all the components
at high concentrations and in active forms. Most
cells do not express all the components re-
quired for inflammasome activation, necessitat-
ing prior stimulation or sensitization. THP-1
cells, for example, require differentiation of this
monocytic cell line into a macrophage-like cell
(70). Similarly, mouse macrophages are gen-
erally primed with LPS, and the response to
MDP may require incubation with the protein
synthesis inhibitor CHX (69). The pathogen
Francisella tularensis first triggers type I IFN ac-
tivation to enable ASC-dependent inflamma-
some activation (189). These experimental find-
ings hint of synergisms, feedback loops, and
checkpoints that ultimately control inflamma-
some activation and orchestrate the physiolog-
ical inflammatory response. Of particular inter-
est are negative feedback loops that are crucial
for the resolution phase of inflammation. NF-
κB for example, a well-known proinflamma-
tory transcription factor, is also a crucial player
in the down-modulation of the inflammatory
response including inflammasome activation
(190). Although we know little at the physio-
logical level, investigators have identified vari-
ous proteins that may interfere with inflamma-
some assembly and inflammatory caspase acti-
vation. Based on the modular structure of these
proteins, we can distinguish two major types
of inflammasome regulators: those containing
a CARD domain and those with a PYD domain
(Figure 5).
Pyrin Domain–Containing
Inflammasome Regulators
PYD-containing regulators are believed to in-
terfere with PYD-PYD interaction between
NALPs and the adaptor ASC. These PYD reg-
ulators include Pyrin, POP1, POP2, and viral
PYDs (vPYDs). POPs and the poxviral gene
product M13L-PYD (also known as vPYD)
are short proteins that contain mainly a PYD
(191). Poxviruses deficient in vPYD produce
an enhanced activation of caspase-1 and secre-
tion of IL-1β, further strengthening the idea
that inflammasomes (that sense viral DNA) are
important players in immunity against viruses
(192–194). Similarly, POP1 and POP2 mod-
ulate inflammasome activity probably by dis-
rupting ASC-NALP interactions (191, 195).
The absence in the mouse genome of both
POP1 and POP2 makes the evaluation of their
physiological importance in vivo challenging.
Pyrin was initially identified as the product
of the MEFV gene, which is mutated in pa-
tients with familial Mediterranean fever (FMF)
(196), a hereditary autoinflammatory syndrome
characterized by episodic fever and serosal or
synovial inflammation. Targeted disruption of
Pyrin in mice causes increased endotoxin sensi-
tivity and enhanced caspase-1 activation (197).
Most of the mutations in Pyrin in FMF pa-
tients affect the C terminus, which harbors a
PRY-SPRY domain. The function of this do-
main, which is partially absent in the mouse,
is not clear, but a role in the regulation of
the inflammasome and caspase-1 was proposed
(55, 56, 197). The PYD of Pyrin interacts
with the PYD of ASC (197), suggesting that it
may be involved in blocking the recruitment of
ASC and inflammasome formation. However,
artificial overexpression of Pyrin can also be
The Inflammasomes 247
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 19
ANRV371-IY27-09 ARI 16 February 2009 8:57
Pi9, vCrmA,
Figure 5
Inflammasome activity is inhibited by PYD-containing proteins that interfere with ASC and NALP
interaction and CARD-containing proteins that disrupt caspase-1 interaction with IPAF or ASC. Bcl-2 and
Bcl-XL proteins have been suggested to inhibit NALP oligomerization (165). Caspase-1 activity can be
directly blocked by Flightless-I (259), and the serpin protease inhibitor 9 (Pi9) (260) or cowpox
virus-encoded inhibitor of caspase-1, vCrmA (261).
proinflammatory. This led Fernandes-Alnemri
et al. (198) to propose an alternative hypoth-
esis, namely that Pyrin, like NALP3, assem-
bles an inflammasome with ASC and caspase-1.
On the other hand, FMF is mainly considered
to be an autosomal recessive autoinflamma-
tory disorder, an observation more consistent
with the notion that Pyrin loss-of-function may
cause the aberrant inflammation in these pa-
tients by allowing inflammasome hyperactiva-
tion (199). PSTPIP1, a Pyrin-interacting pro-
tein, is mutated in PAPA syndrome (
arthritis, pyoderma gangrenosum, and acne), an
autoinflammatory disease associated with over-
production of IL-1β (200). Moreover, muta-
tions in a mouse-related protein, PSTPIP2,
cause a macrophage-dependent autoinflamma-
tory syndrome, further delineating the im-
portance of Pyrin and inflammasome regula-
tion in autoinflammatory disorders (201, 202)
(see below).
CARD Domain-Containing
A small family of inflammasome regulators
harbor a CARD that is highly similar to the
CARD of caspase-1. These proteins most likely
emerged from successive gene duplications and
include in humans iceberg, INCA, COP, and
caspase-12 (33, 203, 204). Through CARD-
CARD interactions these proteins presumably
inhibit processing of proIL-1β by preventing
recruitment and/or activation of the caspase by
the adaptor ASC or IPAF. Except for caspase-
12, most of these proteins are not present in
the mouse or rat genomes, again highlighting
considerable differences in the regulation of
the inflammasome in diverse species. Caspase-
12 is present in two main polymorphic vari-
ants in human, resulting in the production of
either a truncated protein containing the N-
terminal CARD domain (CARD-only) or a full-
length variant molecule (caspase-12L) (205).
248 Martinon
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 20
ANRV371-IY27-09 ARI 16 February 2009 8:57
The full-length variant of caspase-12, which is
the less frequent allele, confined to the popu-
lation of African descent, is linked to hypore-
sponsiveness to LPS-induced production of
cytokines (205, 206). Genetic studies have pre-
dicted that the polymorphism generating the
caspase-12 short variant was driven by posi-
tive selection to complete fixation in the human
genome approximately 60,000–100,000 years
ago (207, 208). Loss of the caspase-12 C ter-
minus may have conferred a selective bene-
fit, possibly by increasing sepsis resistance in
human populations that experienced emergent
infectious diseases as geographic expansion oc-
curred in association with increases in human
population size and density. In line with this
hypothesis, caspase-12-deficient mice clear
bacterial infection more efficiently than do
wild-type littermates and have an enhanced
production of proinflammatory cytokines, in-
cluding IL-1β and IL-18 (209). Caspase-12 was
proposed to be a decoy caspase that blocks
caspase-1 activation resulting in enhanced vul-
nerability to bacterial infection and septic mor-
tality, similar to cFLIP (a decoy caspase-8-like
protein), which regulates caspase-8-mediated
apoptosis. However, contrary to FLIP, the full-
length variant of caspase-12 has autoproteolytic
activities, a mechanism that may regulate its
anti-inflammasome properties or may under-
line the possibility that the full-length variant
of caspase-12 cleaves specific substrates yet to
be identified (210).
Although inflammasomes are involved in both
pathways of pathogen and danger signal sens-
ing, their function converges in the activation
of inflammatory caspases (mainly caspase-1),
which have few known substrates, primarily IL-
1β, IL-18, and possibly IL-33. The complex-
ity of inflammasome assembly contrasts with
the reductionist vision of its main role as a
trigger of IL-1β and IL-18 maturation. More-
over, IL-1β, IL-18, and IL-33 are related cy-
tokines that initiate a MyD88-dependent sig-
naling pathway, very similar to the pathway
engaged by TLRs. Therefore, pathogen sens-
ing by inflammasomes can be interpreted as
an indirect activation of a TLR-like receptor
(IL-1R or IL-18R), a scenario that is compa-
rable to the activation mechanism of the Toll
receptor in Drosophila (19). Both mammalian
and Drosophila signaling pathways involve the
activation of, in mammals, cytokines IL-1β
and IL-18 and, in Drosophila,Sp
atzle through
proteolytic processing, which is initiated by
microbial sensors that engage and activate spe-
cific proteases. Although IL-1-processing in-
flammasomes in mammals sense pathogens in
the cytosol, the proteases that activate Sp
are localized and activated in the hemolymph
of the fruit fly. With this in mind, the complex-
ity and diversity of inflammasomes may reflect
the multitude of pathogens and danger signals
that are detected in specific locations, whereas
the conservation of TLRs and IL-1β/IL-18 re-
ceptor signaling cascades emphasizes the im-
portance of MyD88 signaling in innate immu-
nity (19). The discovery of specific activators
that signal through an inflammasome-IL-1β
pathway brought new interest to the biology of
IL-1β. As new caspase-1 substrates are being
uncovered, new inflammasome functions will
emerge beyond its role in the maturation of
IL-1β, IL-18, and possibly IL-33 (171, 211–
Inflammasomes, Inflammation,
and Inflammatory Diseases
Originally identified as the endogenous py-
rogen, exogenous IL-1β triggers fever in ex-
perimental animals (214). In addition to fever,
IL-1β has multiple other effects on the cen-
tral nervous system. These include induction
of slow-wave sleep, anorexia, and inflammatory
pain hypersensitivity, typically associated with
infections or injury (214, 215). The important
role of IL-1β and the inflammasome in inflam-
mation and fever is strongly supported by ge-
netic evidence that links the inflammasome to
The Inflammasomes 249
Annu. Rev. Immunol. 2009.27:229-265. Downloaded from
by UNIVERSITY OF AUCKLAND on 02/10/10. For personal use only.
Page 21
ANRV371-IY27-09 ARI 16 February 2009 8:57
Table 2 Diseases associated with inflammasome activity
Disease Clinical features
Familial cold
syndrome (FCAS)
Fever, arthralgia,
cold-induced urticaria
NALP3 overactive yes
Muckle-Wells syndrome
Fever, arthralgia, urticaria,
sensorineural deafness,
NALP3 overactive yes
Chronic infantile
neurological cutaneous and
articular syndrome
Fever, severe arthralgia,
urticaria, neurological
problems, severe
NALP3 overactive yes
Familial Mediterranean
fever (FMF)
Fever, peritonitis, pleuritis,
Pyrin overactive partial
Pyogenic arthritis, pyoderma
gangrenosum, and acne
syndrome (PAPA)
Pyogenic sterile arthritis PSTPIP1 overactive yes
Hyperimmunoglobulin D
syndrome (HIDS)
Arthralgia, abdominal pain,
to be demonstrated yes
Tumor necrosis factor
syndrome (TRAPS)
Fever, abdominal pain, skin
TNF-R1 to be demonstrated yes
Systemic juvenile idiopathic
arthritis (SOJIA)
Chronic joint inflammation unknown to be demonstrated yes
Adult-onset Still’s disease
Arthralgia, fever unknown to be demonstrated yes
Behcet’s disease Arthralgia, uveitis, ulcers unknown to be demonstrated yes
Schnitzler’s syndrome Urticaria, fever arthralgia unknown to be demonstrated yes
Gout Metabolic arthritis, pain uric acid
activated yes
Pseudogout Arthritis CPPD activated yes
Contact dermatitis Urticaria irritants activated unknown
Fever syndrome Fever NALP12 unknown unknown
Hydatidiform mole Hydatid mole NALP7 unknown unknown
Vitiligo Skin depigmentation,
NALP1 unknown unknown
a family of hereditary periodic fevers (HPFs)
(216) (Table 2). HPFs are heritable disor-
ders characterized by unexplained and recur-
rent episodes of fever and severe inflamma-
tion. These patients suffer from rashes and
serosal and synovial inflammation with vary-
ing degree of neurological involvement. HPFs
are part of the expanding family of so-called
autoinflammatory diseases that differ from au-
toimmune disorders in that evidence for adap-
tive immunity components such as autoreac-
tive T cells or immunoglobulins to self-antigens
is lacking (217). Familial