You are on page 1of 13

REVIEWS

Expansion and evolution of cell death


programmes
Alexei Degterev* and Junying Yuan‡
Abstract | Cell death has historically been subdivided into regulated and unregulated
mechanisms. Apoptosis, a form of regulated cell death, reflects a cell’s decision to die in
response to cues and is executed by intrinsic cellular machinery. Unregulated cell death
(often called necrosis) is caused by overwhelming stress that is incompatible with cell survival.
Emerging evidence, however, suggests that these two processes do not adequately explain
the various cell death mechanisms. Recent data point to the existence of multiple non-
apoptotic, regulated cell death mechanisms, some of which overlap or are mutually exclusive
with apoptosis. Here we examine how and why these different cell death programmes have
evolved, with an eye towards new cytoprotective therapeutic opportunities.

Cell death is a fundamental cellular response that has X protein (BAX) and BCL2-antagonist/killer-1 (BAK)
a crucial role in shaping our bodies during develop- double mutant mice) results in significant abnormalities4–6.
ment and in regulating tissue homeostasis by elimi- Apoptosis also functions to maintain homeostasis,
nating unwanted cells. The first form of regulated or especially in the immune system, because it eliminates
programmed cell death (PCD) to be characterized was unwanted cells. Dysregulation of apoptosis leads to vari-
apoptosis, which was described in Caenorhabitis elegans ous human diseases, such as cancer and autoimmunity.
in the early 1990s1. Subsequent genetic analysis of mam- Inappropriate activation of cell death is also the leading
malian apoptosis presented a more complex picture, in cause of tissue injury and functional decline in a large
which individual apoptosis genes from C. elegans have number of acute diseases (such as stroke, myocardial
expanded into large multi-protein families (FIG. 1). These infarction and brain trauma) and chronic diseases (such
findings suggest that redundancy, functional special­ as diabetes and neurodegeneration). However, effective
ization and compensatory regulation of mammalian cytoprotective therapies for these diseases remain a major
apoptotic signalling and execution might be important unmet medical need.
features of mammalian apoptosis. To a significant extent, the limited success of cyto­
Because of its conserved and uniform nature, apop­ protective drug development can be traced to the simpli-
tosis is frequently defined mechanistically as a pathway fied view that cell death is either intrinsically regulated
of regulated cell death that involves the sequential activa- by apoptosis or that it is unregulated, caused by over-
tion of caspases, a family of Cys proteases, and that is whelming stress (so-called necrosis). Necrosis possesses
controlled both positively and negatively by B-cell lym- characteristic features, such as organelle swelling, mito-
phoma protein-2 (BCL2) family members. Assays have chondrial dysfunction, massive oxidative stress and rapid
*Tufts University,
now been developed for multiple steps of the pathway, plasma-membrane permeabilization, that are thought to
School of Medicine, allowing the characterization of apoptotic death in vitro be indicative of the catastrophic nature of cell death, rather
Department of Biochemistry, and in vivo. Apoptotic cell death is characterized by than a result of cellular regulation. The general view of
136 Harrison Ave., Boston, distinctive morphological features, including nuclear the relationship between apoptosis and necrosis is that
Massachusetts 02111, USA.
fragmentation, membrane blebbing and the formation milder insults to the cell cause apoptosis, whereas more

Harvard Medical School,
Department of Cell Biology, of apoptotic bodies (see Ref. 2 for further details), that intense insults induce uncontrollable necrosis7. It is thought
200 Longwood Ave., Boston, can be used to identify apoptotic cell death events. that such an apparently unregulated — and hence untarg-
Massachusetts 02115, USA. Genetic studies have shown that apoptosis has a sig- etable — process accounts for the bulk of cell death events in
Correspondence to J.Y. nificant role during normal mammalian development3, acute pathologies. However, in the past few years, evidence
e‑mail:
jyuan@hms.harvard.edu
especially in the central nervous system where genetic has emerged for a number of regulated non-apoptotic
doi:10.1038/nrm2393  deficiency of apoptotic genes (such as caspase‑9, apoptotic cell death pathways, including some with morphological
Published online 16 April 2008 protease-activating factor-1 (APAF1), or BCL2-associated features that were previously attributed to necrosis8.

378 | May 2008 | volume 9 www.nature.com/reviews/molcellbio


© 2008 Nature Publishing Group
REVIEWS

It has become clear that the simple apoptosis–necrosis C. elegans Mammalian


classification does not adequately represent the complex- Upstream signal Upstream signal
(transcription) (processing, transcription)
ity of endogenous cell death regulation. We are now just
beginning to appreciate the important functions of reg­u­ Activated
lated non-apoptotic cell death in homeostatic regulation BH3-only
and development of human disease. EGL-1 proteins
In this review, we first provide a brief overview of Mitochondria
apoptosis and then examine how mammals might have
acquired more apoptotic genes and more complex CED-9 BCL2
apoptotic regulation through gene duplication. We then
describe examples of regulated non-apoptotic cell death BAX/BAK
pathways and consider the possible relationships and
evolutionary origins of the diverse cell death processes. Cytochrome c
release
Apoptosis — the fundamental mechanism of PCD
The mechanism of PCD elucidated by genetic studies in APAF1
the nematode C. elegans defines a primordial paradigm of
apoptosis9. In C. elegans, the activation of PCD is controlled CED-4
by an elegant and simple pathway (FIG. 1). The initiation Apoptosome
of apoptosis is regulated by transcriptional upregulation of
egl‑1, a pro-apoptotic BCL2 homology-3 (BH3)-only member Caspase-9
of the BCL2 protein family10. Binding of EGL‑1 to anti-
apoptotic CED‑9 relieves the inhibition that CED-9 exerts Caspase-3
CED-3
on the adaptor protein CED‑4, allowing CED‑4 to bind
and activate the Cys protease CED‑3, which in turn cleaves
Cell death Cell death
multiple specific cellular substrates to execute cell death11.
Analysis of apoptosis in mammalian cells has led to Figure 1 | Evolutionary expansion of C. elegans
the identification of multiple mammalian homologues apoptotic machinery in mammalian
Nature cells. Side-by-side
Reviews | Molecular Cell Biology
for each class of the C. elegans CED proteins. Extensive comparison of the Caenorhabditis elegans CED protein
studies have revealed that the mechanism of mam­malian pathway and the core apoptotic machinery in mammalian
cells shows the conservation of the general outline of the
apoptosis is similar to that of C. elegans, although it
pathway. Extension of the apoptotic machinery can also be
has become much more complex (FIG. 1). For example, observed at every step of the pathway, including multiple
although transcriptional upregulation of pro-apoptotic B-cell lymphoma protein-2 (BCL2) homology-3 (BH3)-only-
members of the BCL2 family can still play a part in the protein activating signals, complex regulation of the BCL2
initiation of apoptosis in mammalian cells12, many BH3- family and the addition of mitochondrial cytochrome c
only, pro-apoptotic BCL2 family members can be activ­ release, which drives the formation of an apoptosome and
ated in a number of different ways (including cleavage, activation of the upstream caspases (first caspase‑9 and
phosphoryl­ation, myristoylation and ubiquitylation13–16) then the executioner caspases, such as caspase‑3 and
to regulate the initiation of apoptosis17. caspase‑7). Added complexity is provided by the existence
In mammalian cells, the activated EGL‑1-like BH3- of multiple family members in each class of the apoptotic
regulators, with both redundant and non-redundant
only members of the BCL2 family also inhibit CED‑9-like
functions. These regulators provide ‘fail-safe’ apoptosis
anti-apoptotic members of the BCL2 family (such as BCL2, machinery that can generate specialized responses to
BCL-XL and MCL1) by direct interaction17 (BOX 1; FIG. 1). various upstream stimuli. Possible direct activation of
However, a major mechanism by which anti-apoptotic BAX and BAK by BH3-only proteins is indicated by a dotted
members of the BCL2 family protect cells is not through line. APAF1, apoptotic protease-activating factor-1; BAK,
the direct inhibition of caspase activation at the level of BCL2-antagonist/killer-1; BAX, BCL2-associated X protein.
adaptor molecules (for example, the mammalian CED‑4-
like APAF1 protein), but by protecting the integrity of
the mitochondria (BOX 1; FIG. 1). BH3-only factors bind damage and cytochrome c release. In some cases, BAX and
and antagonize anti-apoptotic BCL2 family members BAK can act through interactions with the components
that reside in the outer mitochondrial membrane. Anti- of the mito­chondrial permeability pore, namely with the
apoptotic BCL2 proteins exert their activity by preventing voltage-dependent anionic channel (VDAC)20. Mitochon­
the pro-apoptotic multidomain BCL2 family members drial damage might also be caused by BAX- and BAK-
BAX and BAK from causing mitochondrial damage18, independent mechanisms, such as those induced by intra­
and binding of BH3-only proteins relieves this inhibi- mitochondrial K+ influx, or caused by the direct action
tion. Additionally, a subclass of BH3-only factors might of caspase‑2 on mitochondria (which occurs, curiously,
directly induce the formation of a BAX–BAK channel18, independently of the protease activity of caspase-2)20.
BH3-only member although how much this direct mechanism contributes Mitochondrial damage and the release of mitochon-
Pro-apoptotic BCL2 family
member possessing only a
to apoptosis remains a subject of debate19. BAX and drial proteins amplifies apoptotic signalling in mam­
BCL2 homology-3 (BH3) BAK have been proposed to form an oligomeric (at least malian cells, a step that is considered to be less important
domain. tetrameric or larger6) channel that leads to mitochondrial for apoptosis in lower organisms, including flies and

nature reviews | molecular cell biology volume 9 | May 2008 | 379


© 2008 Nature Publishing Group
REVIEWS

Box 1 | Major classes of apoptosis mediators


Caspases Caspases
Caspases are a family of Cys proteases (humans have Prodomain p20 p10
11 caspases) that cleave their substrates after Asp residues. Executioner caspases
Caspases contain three main domains: a prodomain and
large (p20) and small (p10) catalytic subunits. The large Activator caspases
domain contains the active site Cys residue. Activation of DED, CARD
caspases involves the proteolytic cleavage of zymogens, the Inflammatory caspases
removal of the prodomain and separation of the p20 and DED, CARD
p10 subunits, or allosteric conformational changes. The
prodomains of activator and inflammatory caspases contain BCL2 protein family
protein–protein-interaction domains (such as the caspase- BH4 BH3 BH1 BH2 TM
recruitment domain (CARD) and the death-effector domain Anti-apoptotic
(DED)) that link them to apoptosis signalling molecules21.
Multidomain pro-apoptotic
BCL2 family
The B-cell lymphoma protein-2 (BCL2) family is subdivided BH3-only
into three sub-classes: anti-apoptotic (such as BCL2,
BCL-XL and MCL1), multidomain pro-apoptotic (BAX and
BAK) and BH3-only (such as BID, BIM, BAD, NOXA and APAF1, NLR and PIDD adaptors
PUMA)17. The BH1, BH2 and BH3 domains of multidomain PYD, CARD NACHT NAD LRR
family members form a hydrophobic cleft that serves as a NLRs
heterodimerization interface for the BH3 domains of
BH3-only proteins. The BH4 domain of anti-apoptotic CARD NB-ARC WD-40
BCL2 family members directly interacts with a voltage- APAF1
dependent anion channel and inhibits apoptotic LRR ZU-5 ZU-5 DD
mitochondrial changes113. Anti-apoptotic BCL2 family PIDD
members can be cleaved by caspases, which results in the
loss of the BH4 domain and in the proteins exhibiting pro-apoptotic, rather than anti-apoptotic, activity114.
BH1/BH2/BH3, BCL2 homology‑1/2/3 domain; TM, transmembrane domain. Nature Reviews | Molecular Cell Biology

APAF1, NLR and PIDD adaptors


NLR (nucleotide-binding and oligomerization domain (NOD)-like receptor), APAF1 (apoptotic protease-activating factor-1)
and PIDD (p53-induced protein with a death domain) activate caspases following pathogenic infection (caspase‑1),
cytochrome c release (caspase‑9) and genotoxic stress (caspase‑2), respectively. The NLR family (22 proteins in humans) can
be subdivided into NOD factors (which sense bacterial peptidoglycans and signal through RIP2 to activate NF‑κB),
inflammasome-forming NALPs, IPAF, NAIP and CIITA (which regulates major histocompatibility (MHC) class II transcription).
Activation of APAF1, NLRs and PIDD proteins leads to the formation of large multimeric caspase-activating complexes115,116.
Caspase recruitment can be direct (through CARD domains) or can involve additional adaptors (such as RAIDD (also known
as CRADD) interacting with the death-domain (DD) of PIDD116). The nucleotide-binding and oligomerization domain (the
NACHT domain) and NB‑ARC domains are required for oligomerization. The NB‑ARC domain of APAF1 also binds to dATP,
which is required for apoptosome formation. The WD‑40 repeats of APAF1 bind cytochrome c. The Leu-rich repeats (LRRs)
of NLRs probably have a key role in sensing pathogen-associated molecular patterns83. CIITA, major histocompatibility
complex (MHC) class II trans-activator; IPAF, ICE-protease activating factor; NAIP, neuronal apoptosis-inhibitory protein;
NALP, NACHT‑, LRR- and pyrin domain (PYD)-containing proteins; NF-κB, nuclear factor-κB.

worms. Cytochrome c, which is released from damaged in turn cleave downstream caspases, such as caspase‑3 and
mitochondria, promotes the formation of a heptameric caspase‑7, to execute cell death; alternatively, caspase‑8
‘apoptosome’ megacomplex of APAF1 and caspase‑9 (a can cleave the BH3-only pro-apoptotic protein BID,
member of the CED‑3-like Cys protease family). This which in turn amplifies the cell death signal by causing
leads to the conformational change and activation of mitochondrial damage and cell death24. The development
caspase‑9 (FIG. 1). Activated caspase‑9 in turn cleaves of cytokine-mediated apoptosis programmes in higher
and activates downstream caspases, including caspase‑3, multicellular organisms provides a crucial way to coordin­
caspase‑6 and caspase‑7, that carry out the execution ate the regulation of cell numbers at the organismal
phase of apoptosis21. level in response to the environmental stimuli.
In addition to the intrinsic apoptosis pathway, which
resembles PCD in C. elegans, mammalian cells also pos- Evolutionary expansion of apoptosis
sess an extrinsic apoptosis pathway, which is induced by Although the core regulators of PCD in C. elegans consist
pro-apoptotic and pro-inflammatory cytokines (such as of only 4 genes — egl‑1, ced‑3, ced‑4 and ced‑9 — each
FAS ligand (FASL) and tumour-necrosis factor‑α (TNFα), has multiple mammalian homologues. On the basis of
which are ligands for the death-receptor family)22. By bind- close homology in the key regions, the mammalian ced-
ing to death-domain receptors, FASL and TNFα induce like genes probably arose through gene duplication and
the formation of specific intracellular death-induced might have evolved and been selected during evolution
signalling complexes (DISCs23), which activate upstream to meet the challenges that highly complex multicellular
caspases such as caspase‑8. The activation of caspase‑8 can organisms face.

380 | May 2008 | volume 9 www.nature.com/reviews/molcellbio


© 2008 Nature Publishing Group
REVIEWS

Multiple genes enable functional specialization. One cells is shown by the compensatory upregulation
direct consequence of this gene duplication is specifica- of caspases in different caspase mutant mice — the loss of
tion — different apoptosis regulators respond to differ­ent one caspase can be compensated by the upregulation
pro-apoptotic signals. The mammalian caspase family of another caspase37,38. Although the specification of
provides an excellent example of specification. Different different apoptotic regulators in a signal and/or subcellu-
caspase family members possess distinct functions in lar compartmentalization manner provides a mechanism
mammalian cells that are predominantly based on their to fine-tune cellular responses, danger might arise if a
subcellular localization and protein–protein interactions specified response is lost because of a genetic mutation.
rather than on their substrate specificities25. Mammalian In this case, it appears that upregulation of an alternative
caspases were initially assigned to three major classes: caspase, or of its regulators, can compensate for the loss
apical or activator caspases, such as caspase-2, -4, ‑8, -9, of a specific caspase.
‑10 and ‑12, which initiate the caspase cascade in apop- Although different caspases exhibit specificity towards
tosis; executioner caspases, such as caspase‑3, ‑6 and ‑7, certain cleavage sites39, such specificity is relative. When
which act in the downstream execution steps of the proc- present in a sufficient concentration and given sufficient
ess; and inflammatory caspases, such as caspase‑1, -5 incubation time, most caspases can cleave most, if not
and ‑11, which mediate cell death and inflammatory all, of the caspase substrates that have been identified to
responses. Further analyses suggest that there might be date. The preferential cleavage of a selective subgroup
additional important distinctions between family mem- of caspase substrates in the early phase of apoptosis
bers in the same class. Activator caspases have distinct may only reflect the proximity of the substrates to the
roles that depend on the activating complexes that they activ­ated caspases at that specific time or they may serve
are recruited to. For example, caspase‑8 specifically to modulate the kinetics of the process. The ability of
contributes to death-receptor signalling26 and normal multiple caspases to cleave common substrates might
pro­liferation of lymphocytes27; caspase‑2 mediates serve to insure the execution of apoptosis even when
genotoxic stress-mediated death28; mouse caspase‑12 one caspase is lost. Furthermore, although upstream
and human caspase‑4 mediate endoplasmic reticulum caspases can be activated preferentially during the early
(ER)‑stress-mediated death29,30; and caspase‑9 is activ­ stages of apoptosis (such as caspase-8 and ‑10, which are
ated by the apoptosome downstream of cytochrome c preferentially activated in response to a FASL or TNFα
release31. Such division of labour might provide a sensi- signal26), all caspases are cleaved and activated during
tive mechanism to allow complex multicellular organ- the later stages of apoptosis. Therefore, all of the caspases
isms to detect and differentially respond to distinct may contribute to the execution of apoptosis21. Apoptosis
environmental stimuli. in mammalian cells is genetically programmed to maxi-
A similar division of labour has been observed for mize the ability to kill the cells once an appropriate order
the members of the BCL2 family. BCL2 family members is received.
can be subdivided into three major classes on the basis of
structural and functional differences17 (BOX 1). However, Expansion or reduction. Although the findings dis-
recent studies suggest that subtle but important differ- cussed above lead us to suggest that apoptotic pathways
ences exist in each sub-class. In particular, although all expanded during evolution, another view is also poss­
BH3-only factors possess a similar EGL‑1-like mode ible. C. elegans, along with its evolutionary relatives,
of action, they can be activated non-redundantly by might have experienced a reduction in the complexity
particular types of apoptotic signals. For example, BID of the apoptotic machinery from primordial ancestors
mediates signalling by the TNF-family and by genotoxic with apoptotic machinery that might have been closer
stress16,32; BIM plays a key part in inducing death in lym- to that of humans. However, we find this scenario less
phoid and myeloid cells following cytokine withdrawal33; likely, because no such developed apoptotic machinery,
NOXA and PUMA signal downstream of p53 (Ref. 34); resembling that of mammals, has been identified in more
and BMF has a key role in anoikis35. The specificities of primitive organisms, such as plants, fungi and bacteria.
BH3-only factors not only result from their interactions
with upstream regulators, they also result from their Non-apoptotic mechanisms of cell death
distinct binding preferences for different multidomain Emerging evidence suggests that apoptosis is not the
Anoikis BCL2 family members. For example, the BH3 domain of only mechanism of cellular suicide, but rather that cells
Cell death caused by cell
detachment from the matrix.
BAD binds BCL2 and BCL-XL, whereas the BH3 domain might choose one of many mechanisms to die when
of NOXA displays selectivity towards MCL1 and the they are ready, with apoptosis often representing the
Linker cell BCL2-related protein A1 (Ref. 36). Such selectivity in top choice. In C. elegans, most developmental cell death
A cell in a C.elegans embryo the inter­actions of different members of the BCL2 family events occur through the apoptotic mechanism and only
that is positioned between the
provides an important mechanistic basis for activating rare events, such as the developmental death of a linker
gonad and the cloacal tube.
distinct apoptosis signalling pathways in response to cell, are non-apoptotic40. Type I cell death, which displays
Type I cell death different pro-apoptotic stimuli. the characteristic morphology of apoptosis as discussed
A term introduced as part of a above, is also the most frequently observed form of
morphological classification of The benefits of redundancy. An additional and non- death during normal mouse development2. Although
developmental cell death.
Type I death shows the
conflicting significance of the multiplication of apoptosis induction of morphologically non-apoptotic develop-
morphological features of regulators is the benefit that is provided by redundancy. mental cell death becomes prominent in mice in vivo
apoptosis. The redundancy of apoptosis programmes in mammalian when apoptotic machinery is genetically disrupted

nature reviews | molecular cell biology volume 9 | May 2008 | 381


© 2008 Nature Publishing Group
REVIEWS

Box 2 | Three pathways of non-apoptotic cell death


a Autophagic cell death b Necroptosis c PARP1-mediated cell death
TNFα DNA DNA
damage degradation Cell
death
TNFR
Cell PARP1 Nucleus
LC3-II membrane
Death RIP1 NAD+ PAR
domain Specific loss polymer
ATG5–ATG12, Death
initiation
ATG16L BCL2, domain
BCL-XL TRAF2
Beclin-1, UVRAG Kinase RIP1
VPS34, Energy
VPS15 Lipoxy- Acid collapse
genases SMase Ceramide JNK
WIPI ULK mTOR
Autophagy Rac1/NADPH JNK
ATG9 PAS oxidase AIF
Mitochondrial Mitochondrial Mitochondrial
dysfunction ROS dysfunction
Autophagosome formation Regulated necrosis Cell death Mitochondria
Autophagic cell death. Eighteen yeast autophagy-related (ATG) genes, which are required for autophagosome (AP)
formation, have been identified (AP–Atg proteins)117. Several mammalian homologues have been identified (shown in
figure panel a). The activity of the class III phosphatidylinositol 3-kinase complex is subject to inhibition by B-cell
Nature Reviews | Molecular Cell Biology
lymphoma protein-2 (BCL2) and to activation by the tumour suppressor UVRAG (UV radiation resistance-associated
gene protein). Nutrient-deprivation signalling induces autophagy through the suppression of mammalian target of
rapamycin (mTOR) activity. Functional analysis has implicated the same core ATG proteins in the formation of different
types of autophagosomes, including those that have pro-death roles118. The molecular machinery of mammalian
autophagosomes is still only partially characterized. Intrinsic differences in the composition of autophagosome protein
machinery might determine its function as a pro-survival mechanism, rather than a pro-death mechanism. PAS,
phagophore assembly site. See REF.117 for further details. This panel is modified, with permission, from ref. 117
(2007) Cold Spring Harbor Laboratory Press.
Necroptosis. The activation of RIP1 kinase increases reactive oxygen species (ROS) production (from the mitochondrial
respiratory chain and the RIP1–Rac1–NADPH oxidase complex) and activates c-Jun N-terminal kinase (JNK) kinase
(which might be crucial for the execution of necroptotic cell death in some cell types)119,120. Autophagy can be prominently
activated during necroptosis, but it only appears to contribute to cell death in some cell types63. Other execution steps,
including the activation of phospholipase A2, lipoxygenases121 and acid sphingomyelinase122, have been described.
The exact roles of these steps remain to be elucidated. Given the similarity of many of the downstream execution steps
in necroptosis to those attributed to ‘classic’ unregulated necrosis, the main difference between these two mechanisms
might be in the method of activation (regulated by internal signalling mechanisms rather than caused by overwhelming
stress) of similar relatively nonspecific execution events. SMase, sphingomyelinase; TNFα, tumour-necrosis factor-α;
TNFR, TNFα receptor.
PARP1-mediated cell death. Two pathways are shown: energy collapse and apoptosis-inducing factor (AIF) translocation.
In addition to the cell death role, poly(ADP–ribose) polymerase‑1 (PARP1) is involved in initiating DNA repair. Genetic
ROS deletion of PARP1 leads to sensitivity to DNA-damaging agents123. Oxidoreductase activity of AIF also has a homeostatic
(Reactive oxygen species). role, as it is crucial for mitochondrial complex I function124 and a mutation in mice that is associated with severe
Highly reactive intermediates progressive neurodegeneration has been mapped to AIF124. TRAF2, TNF receptor-associated factor-2.
in the reduction of oxygen to
water. These are produced, for
example, as a byproduct of
oxidative phosporylation in the (such as in caspase-deficient motor neurons41), one Type II cell death. Type II cell death is characterized
mitochondria. might question whether non-apoptotic cell death can by the accumulation of double-membrane-enclosed
also occur under normal conditions when apoptosis is vesicles. These vesicles are characteristic of auto­phagy
Type II cell death
A term introduced as part of a
possible. One could argue that non-apoptotic mecha- and so type II cell death is often called autophagic cell
morphological classification of nisms represent rudimentary back-up forms of cell death death, although the role of autophagy in this type of
developmental cell death. that are only relevant in rare circumstances when the cell death is still under debate2. Autophagy is an evolu-
Type II death shows the apoptotic machinery is genetically unavailable. However, tionarily-conserved intracellular catabolic mechanism
morphological features of
an understanding of the molecular mechanisms under- that operates at low levels under normal conditions to
autophagy.
lying non-apoptotic cell death in vitro has recently begun mediate the degradation of cytoplasmic components,
Necroptosis to emerge, and we are now beginning to appreciate the protein aggregates and expired intracellular organelles
A process of regulated non- importance of these processes. Here, we summarize (for example, the ER and mitochondria) by forming
apoptotic cell death displaying the data that describes three emerging regulated non- double-membrane-enclosed vesicles called autophago-
necrotic morphology, which
can be induced by death
apoptotic cell death pathways: type II cell death, necroptosis somes. The contents of autophagosomes are degraded
domain receptors through RIP1 and poly(ADP–ribose) polymerase‑1 (PARP1)-mediated by lysosomal enzymes after autophagosomes fuse with
kinase activity. necrotic death. lysosomes.

382 | May 2008 | volume 9 www.nature.com/reviews/molcellbio


© 2008 Nature Publishing Group
REVIEWS

Under normal conditions, autophagy has an important as death-domain-receptor engagement by correspond-


role in maintaining intracellular homeostasis. In this role, ing ligands, can lead to non-apoptotic cell death (as
autophagy removes damaged or dysfunctional organelles assessed using morphological criteria) when apoptosis
and misfolded proteins that can be detrimental to sur- is inhibited by caspase inhibitors58–61 or through muta-
vival (for example, of neurons)42. Under conditions of tions in caspase‑8 or FAS-associated death-domain
nutrient deprivation, autophagy promotes cell survival protein (FADD)62–65. Although necroptosis is activated
by degrading disposable intracellular contents, thereby by the same stimuli that initiate apoptosis, the morpho­
generating energy and building blocks for protein syn- logical features of necroptosis — organelle swelling,
thesis. Autophagy is regulated by a large group of ATG rapid mitochondrial dysfunction, plasma membrane
(autophagy-related) genes that are conserved from yeast permeabilization and lack of nuclear fragmentation
to humans (BOX 2). — are characteristic of pathological necrosis, which
Activation of autophagy during Drosophila melano­ is presumed to be unregulated death that is caused by
gaster metamorphosis has been well established43. Using overwhelming stress8,66.
genetic elimination of Atg genes, it was recently shown Emerging evidence suggests that the initiation of the
that a lack of autophagy attenuated the degradation of necroptotic programme by TNFα occurs at the receptor
D. melanogaster salivary glands by blocking PCD44. Thus, level through the recruitment and activation of an intra­
this study provides evidence for the role of autophagy in cellular signalling complex that involves the adaptor mole­
developmental cell death. However, the lack of cell death cule RIP1 (but not the TRADD–RIP1 complex, which
defects in autophagy-deficient mutant mice (such as ATG5 mediates the activation of nuclear factor-κB (NF‑κB)
or ATG7 mouse mutants), as well as in ATG7-deficient and apoptosis)67. Activation of necroptosis requires the
Drosophila, raises the question of whether autophagy kinase activity of RIP1, which is not required for NF‑κB
per se or only part of the autophagy pathway is involved and apoptosis signalling64. The distinct nature of apop-
in type II cell death during development2,45–48. tosis and necroptosis was further emphasized by the
Autophagy might contribute to cell death that is discovery, in a random cell-based screen, of a series of
induced by viruses. Human immunodeficiency virus‑1 structurally distinct small molecule necrostatins, all
(HIV1) induces the accumulation of Beclin-1 protein and of which specifically and efficiently inhibit necroptosis,
cell death in uninfected bystander CD4+ T cells when the but not TNFα-induced apoptosis63,68–70. The mechanism
HIV1 envelope protein (Env) interacts with the chemo­ that leads to the execution of necroptosis downstream
kine receptor CXCR4 (Refs 49,50). Although the mecha- of RIP1 kinase activation remains unclear. RIP1 is
nism by which autophagy is activated by Env is not yet translocated into the mitochondria, which leads to the
clear, this result suggests that autophagy might contribute disruption of the association of ADP–ATP translocase
to cell death in a non-cell-autonomous manner, providing (ANT) with cyclophilin D (Ref. 71) and this might explain
a mechanism by which a virus might induce cell death the rapid mitochondrial dysfunction that is associated
independently of viral replication. with necroptosis63. Interestingly, the protective effect of
Autophagy can assume the killer role when apoptosis is the necroptosis inhibitor necrostatin‑1 in an in vivo heart
unavailable51. For example, autophagy mediates cell death ischaemia–reperfusion injury was dependent on the expres-
in apoptosis-deficient BAX–/– BAK–/– cells in response sion of cyclophilin D (Ref. 72). However, the mechanistic
to genotoxic or ER stress stimuli52,53. It is possible that details of this step remain to be determined.
autophagy is induced, albeit at low levels, under normal
conditions, but that it becomes exacerbated in response to PARP1-mediated necrotic death. PARP1 is a nuclear
cellular stress during apoptosis-deficient conditions to pro- enzyme that has a key role in maintaining genome
mote cell death. In this regard, it is interesting to note that stability. PARP1 is rapidly activated by DNA-strand
Beclin-1 — the mammalian homologue of yeast Atg6 and breaks and recruits DNA-repair factors by attaching
a regulator of the type III phosphatidylinositol 3-kinase ADP–ribose units to chromatin-associated proteins.
VPS34 — has a BH3 domain54 and interacts with BCL2, Loss of PARP1 leads to an increased sensitivity to DNA
which reflects the convergent regulation of apoptotic and damage73, which prompted the development of PARP1
autophagic cell death55. Furthermore, a subset of BH3-only inhibitors as chemopotentiators of DNA-damaging anti-
pro-apoptotic BCL2 family members, including BNIP3 cancer agents74. However, over-activation of PARP1 can
CD4+ T cell and BIK, can induce the activation of autophagic cell lead to caspase-independent cell death75.
One of a subpopulation of
mature T cells that express the
death56,57. Although the mechanism by which BNIP3 and PARP1 can mediate cell death in a number of differ-
CD4 receptor. BIK induce autophagy is not clear, and although it remains ent scenarios (BOX 2). In one scenario, alkylating DNA
to be seen whether autophagy is induced as a secondary damage promotes rapid PARP1-mediated depletion
NF-κB consequence of mitochondrial damage, such studies of cytosolic NAD+, which leads to necrotic death by
(Nuclear factor-κB). A family of
raise the possibility that autophagic cell death might be ‘energy collapse’ in glycolytic cells76 (these cells depend
transcription factors that are
activated by many induced in a manner similar to that of apoptosis. on cytosolic NAD+ for glycolysis and energy genera-
inflammatory signals. tion). This mechanism can be viewed as an extension
Necroptosis as a form of regulated necrosis. Necroptosis of the genome-surveillance function of PARP1, as
Cyclophilin D (BOX 2), which represents a type of programmed necro- it provides an elegant way to differentially regulate
A mitochondrial matrix protein
that is associated with the
sis, is another intriguing example of a regulated non- DNA-damage responses in rapidly proliferating glyco­
mitochondria permeability apoptotic cell death mechanism. Its discovery was prompted lytic cells and in cells in a vegetative state, relying on
pore. by observations that classic apoptotic stimuli, such mitochondrial respiration for maintaining ATP levels76.

nature reviews | molecular cell biology volume 9 | May 2008 | 383


© 2008 Nature Publishing Group
REVIEWS

Box 3 | The contribution of non-apoptotic cell death to pathological injury Although the fine details of PARP1-dependent cell death
remain to be sorted, it is clear that PARP1-mediated cell
Recent studies have shown that non-apoptotic cell death is not just a nuisance of death is an important subject for investigation because
in vitro experimentation — it makes important contributions to pathological of its role in various human pathologies (BOX 3).
regulation in vivo.
Autophagy has been prominently observed in various disease models and its
inhibition can provide therapeutic benefit. For example, recent studies suggest that
The roots of regulated cell death
Beclin-1-dependent autophagy promotes injury in mouse models of heart ischaemia– Curiously, although the phenomenon of apoptosis was
reperfusion injury and heart failure125,126. In addition, autophagy suppresses apoptosis first described in the context of developmental regulation,
in MYC-dependent lymphomas, promoting tumour growth127. Conversely, genetic the function of apoptosis in development is not clear-
analysis has clearly established that Beclin-1 functions as a tumour suppressor128. cut; C. elegans ced‑3 (loss-of-function) or ced‑4 (loss-of-
Autophagy promotes the survival of cancer cells that express apoptosis-inhibiting function) mutants are developmentally normal81. If PCD
BCL2 family members under the conditions of hypoxia129. However, inhibiting does not provide a developmental advantage for nema-
autophagy under such conditions unleashes necrosis, which might promote todes that have it compared with ancient variants that
inflammation and, ultimately, tumorigenesis. These data suggest a complex role of lacked it, what factors might have led to the selection and
autophagy in tumour formation.
evo­lution of PCD mechanisms? Interestingly, ced‑3 and
As the morphology of necroptotic cell death is similar to that of necrosis, it has been
ced‑4 mutants were found to be significantly more sensi-
investigated whether pathological necrosis might actually represent, if only in part,
regulated necroptosis. Indeed, administration of necroptosis inhibitor necrostastin‑1 tive to death caused by Salmonella typhimurium infection
(Nec‑1) provides significant tissue protection and functional improvements in a range compared with wild-type worms82, raising the possibility
of acute tissue injuries in vivo in mouse models (brain and heart ischaemia–reperfusion) that the host defence response, rather than develop­
by mechanisms that are clearly distinct from the inhibition of pathological mental cell death, might be the primordial function
apoptosis63,72,130. In particular, in a mouse model of stroke, the effect of Nec‑1 was both of apoptosis.
temporally and mechanistically (measured through caspase‑3 activation) distinct from The host defence function of apoptosis has been
that of apoptosis inhibitors. Furthermore, the protective effect of Nec‑1 did not require expanded in mammalian cells. A large family of mamma-
co-administration of caspase inhibitors, although the pan-caspase inhibitor zVAD.fmk lian NLR (NOD-like receptor) proteins, homologous to 
and Nec‑1 did have an additive effect (FIG. 3).
C. elegans CED‑4 and mammalian APAF1, function
Poly(ADP–ribose) polymerase‑1 (PARP1) inhibitors provide significant protective
to regulate the activation of caspases in response to
effects in mouse brain and heart ischaemia–reperfusion injury, mouse models of
colitis and other inflammatory diseases, neurodegeneration and diabetes mellitus intracellular pathogens83. NLRs contain three distinct
(reviewed in Refs 131–133). PARP1 inhibitors have also emerged as promising domains: an N‑terminal caspase-recruitment domain
anti-cancer agents, increasing the sensitivity of resistant cancer cells to various (CARD) or pyrin effector domain (with the exception
DNA-damaging agents and also selectively killing some tumour cells134. of neuronal apoptosis-inhibitory protein (NAIP) and,
possibly, NOD5 (also known as NLRX1)); a nucleotide-
binding and oligomerization domain (the so-called
Rapidly proliferating glycolytic cells might pose a sig- NACHT domain); and a variable number of C‑terminal
nificant danger for the organism if they accumulate Leu-rich repeats (LRRs) (BOX 1). The 22 identified NLRs
DNA damage and, hence, have to be efficiently elimi- can be divided into two large sub-classes — the NODs
nated; by contrast, vegetative cells can be allowed (NOD1–5), which activate the RIP2–NF-κB pathway,
more time to complete DNA repair. Along these lines, and the NLRs, consisting of NALP1–14 (NALP is named
PARP1 activation also leads to the specific release of the after NACHT‑, LRR- and pyrin domain (PYD)-containing
inflammatory cytokine high mobility group protein‑B1 proteins), IPAF (also known as NLRC4) and NAIP,
(HMGB1), which can alert immune cells to the presence which promote caspase‑1 activation84. In addition, the
of dangerous cells with damaged DNA77. PARP1 also NLR protein CIITA (major histocompatibility complex
mediates cell death that is induced by secondary DNA (MHC) class II trans-activator) serves the function of
damage associated with acute neuronal injury. In this master regulator of MHC class II transcription85. At least
case, excitatory neuronal cell death leads to the trans- some of the NALPs function by recruiting the adaptor
location of the poly(ADP–ribose)-polymer into the protein ASC through a homotypic PYD–PYD inter-
cytosol, triggering translocation of protein apoptosis- action, and ASC in turn recruits caspase‑1 through a
inducing factor (AIF) from the mitochondria to the CARD–CARD interaction86. Oligomerization of NALPs
nuclei, where it mediates cell death78,79. Interestingly, brings the inflammatory caspases, such as caspase‑1 and
DNA-damage-induced PARP1-mediated cell death caspase‑5, into close proximity to promote their activa-
involves TNF receptor-associated factor-2 (TRAF2)– tion. This NALP protein complex is called the inflam-
RIP1-dependent activation of c-Jun N‑terminal masome87. The activation of caspase‑1 in turn leads to
kinase‑1 (JNK1, also known as MAPK8), which con- the processing and release of interleukin‑1β (IL-1β) and
tributes to mitochondrial dysfunction and necrotic IL‑18, which serve important pro-inflammatory func-
death80. However, the relationship between this process tions. The NALP1-mediated caspase‑1 activation is also
and necroptosis remains unclear. regulated by anti-apoptotic proteins of the BCL2 family88,
Currently, it is difficult to make definitive conclu- providing a possible role for BCL2 family proteins in
sions regarding the relationship of the two pathways regulating the host defence response.
NLR of PARP1-induced cell death. However, it appears pos- Interestingly, the origins of the NLR family can be
A mammalian protein that is
characterized by the presence
sible that NAD+ depletion and AIF activity act in the traced to non-apoptotic regulators in simple organisms,
of NACHT and Leu-rich repeat same pathway and that their relative contributions to potentially providing an exciting insight into the evo­
(LRR) domains. cell demise might depend on the specific cell milieu. lutionary origins of mammalian apoptosis. Mammalian

384 | May 2008 | volume 9 www.nature.com/reviews/molcellbio


© 2008 Nature Publishing Group
REVIEWS

NLRs functionally and structurally resemble NLR-like a NLR PAMPs


proteins, such as plant R-gene-encoded proteins89, that
are present in many primitive organisms (FIG. 2). The
DAMPs
R-gene-encoded proteins contain LRRs, which sense SGT1 HSP90 (ATP)
intracellular pathogens and pathogen-induced danger sig-
nals and activate inflammatory signalling90. In a manner LRR NACHT
Cell death
that is highly analogous to the plant innate immune sys-
Pyrin Caspase-1
tem, mammalian inflammasomes recognize signals that CARD Cytokine
are presented by intracellular pathogens and mediate the RIP2
processing
activation of death in infected cells (such as macrophages) Gene expression
through caspase‑1-dependent apoptosis (pyroptosis) and b NB-LRR PAMPs (NODs)
DAMPs
induction of inflammatory signalling through caspase‑1-
dependent processing of IL-1β, IL-18 (Ref. 5) and probably
IL-33 (Ref. 91). SGT1 HSP90
Similar to hundreds of plant R-gene-encoded proteins, LRR NB
Cell death
distinct mammalian inflammasome complexes sense (hypersensitive response)
particular conserved pathogen-associated molecular pat- TIR
terns (PAMPs; these include components of the bacterial CC
Gene expression
cell wall, bacterial flagellin, and bacterial and viral RNA)
Figure 2 | Mammalian NLR proteins and NB‑LRR proteins
and pathogen-induced danger-associated molecular from simpler organisms Nature
haveReviews
similar| Molecular
functions.Cell Biology
Although
patterns (DAMPs; these include changes in ATP levels and NLR (panel a) and NB‑LRR (panel b) proteins (such as those
intracellular potassium concentrations)89. For example, encoded by plant R-genes) have variable domain
the NALP1 inflammasome senses bacterial cell wall pepti­ architecture83,89,90, all NLR and NB‑LRR proteins are
doglycans92, whereas the IPAF inflammasome is activated characterized by the presence of: Leu-rich repeats (LRR),
by cytosolic flagellin93,94. In addition, recent studies which have significant variability (because of alternative
showed that plant R-gene-encoded proteins can sense splicing136) and are involved in pathogen-associated
PAMPs in the nuclei to induce transcriptional changes molecular pattern (PAMP) sensing; nucleotide-binding
that promote an immune response95. It is important to ATPase oligomerization domains (NACHT or NB; these
make-up the ‘oligomerization domain’); and effector
determine whether this can also represent an additional
domains (CARD and pyrin in NLRs; TIR and coiled coil (CC)
mode of action of mammalian NLRs. Curiously, whereas in NB‑LRRs), which are involved in the recruitment of
the plant innate immune response can lead to cell death downstream factors, such as caspase‑1 or RIP2 in the case of
at the infection site90, involving measurable induction of IPAF and NOD proteins, or additional adaptors, such as ASC,
caspase-like activity96, it is carried out by non-caspase in the case of NALP3 protein83. The mammalian NLRs
molecules, such as legumains. Plant proteins displaying (panel a) also display multiple functional similarities with
sequence homology and structure similarity to caspases, NB‑LRRs (panel b) in that they function as part of the innate
termed metacaspases, probably represent a different immune response. Other similarities include: NLRs and
branch of the evolutionary tree97. NB‑LRRs both require SGT1–HSP90 binding to maintain the
Based on these observations, the mammalian proteins in an inactive, but signal-competent state; they are
both activated by PAMPs and danger-associated molecular
inflammasome might be a convergence point between
patterns (DAMPs) in vivo and in vitro; and their cell death and
a simple ced-gene-mediated apoptotic pathway and inflammatory responses are activated through effector
the NLR-mediated innate immune pathway, providing domains. These similarities suggest that mammalian NLR
capabilities for PAMP or DAMP sensing by NLRs, as proteins (and other APAF1-like molecules, such as APAF1
well as caspase activation. Therefore, the mammalian and PIDD) probably evolved from primitive NB‑LRR proteins.
apoptotic machinery might have evolved by converging ASC, apoptotic speck protein; IPAF, ICE-protease activating
the simple apoptotic machinery, such as that found in factor; NB-LRR, nucleotide-binding Leu-rich-repeat; NLR,
C. elegans, with a separate and even more ancient and nucleotide-binding and oligomerization domain (NOD)-like
PAMPs
Components of pathogens that highly conserved innate immune system. Furthermore, receptor; NALP3, NACHT‑, LRR- and pyrin domain (PYD)-
are recognized by the innate because mammalian CED‑4-like factors (such as NLRs, containing protein-3; TIR, translocated intimin receptor.
immune system, including APAF1 and p53-induced protein with a death domain
lipopolysaccharide,
peptidoglycans and viral RNA.
(PIDD)) function similarly in caspase activation,
develop­ment of the mammalian apoptosis machinery patho­gens) and divergence of the developmental, geno-
DAMPS might have been primarily driven by the requirements toxic and inflammatory pathways, which are regulated by
Changes in the cellular for a more efficient host defence system, rather than the separate sets of activating complexes (the apoptosome,
environment that are
more complex homeostatic and developmental regula- PIDDosome and inflammasome, respectively) (BOX 1).
associated with pathogenic
infection, such as an increase tion in higher eukaryotes. Although retaining the basic
in ATP concentration. layout of the PCD mechanism in C. elegans, the conver­ Evolution of non-apoptotic cell death mechanisms
gence of this pathway with NLR regulation led to the Whereas PARP1-mediated non-apoptotic cell death
PIDDosome acquisition of a number of unique features that are probably evolved as part of the cellular response to
A caspase‑2-activating
complex that is formed by the
characteristic of mammalian apoptosis, including DNA damage, other non-apoptotic cell death mecha-
multimerization of PIDD and direct regulation of the mammalian CED‑4 orthologues nisms might have evolved as a part of host defence
RAIDD proteins. by specific activators (such as cytochrome c, ATP and responses.

nature reviews | molecular cell biology volume 9 | May 2008 | 385


© 2008 Nature Publishing Group
REVIEWS

a Predominantly necroptosis b Predominantly apoptosis One could speculate that autophagic cell death might have
arisen from the need to ensure the survival of the whole
organism through sacrificing infected cells. Both type I and
II interferons (IFNs) modulate macroautophagy as a part
of the antiviral response98. In this regard, it is interesting
that treatment of human HeLa cells with IFNγ leads to
the induction of autophagy through the activation of
death-associated protein (DAP) kinases104, and overactiv­
ation of DAP kinases can lead to autophagic cell death.
Thus, autophagic cell death might be part of the antiviral
No cytoprotection Maximal Maximal No cytoprotection response that is activated by IFNs.
cytoprotection cytoprotection The functional significance of necroptosis in the host
defence response is less clear. However, RIP1, a kinase
c Mixed model d Synergistic model
that is crucially involved in the activation of necroptosis,
has an important role mediating the activation of NF‑κB
and IFN genes in the context of the innate immune
system105, although apparently in a kinase-independent
manner. RIP1 signalling can be triggered by extracellular
pathogen-sensing Toll-like receptors106,107, components
of invading pathogens108 and inflammatory cytokines,
such as TNFα. The closest D. melanogaster homologue
of RIP1, IMD, is also a key mediator of the innate immune
response to Gram-negative bacteria, acting upstream of
Partial cytoprotection Partial cytoprotection No cytoprotection No cytoprotection a number of factors that are also linked to RIP1 signal-
ling in mammalian cells, including the D. melanogaster
homologues of FADD, caspase‑8, IKK (inhibitor of
nuclear factor (NF)-κB (IκB) kinase) complex and
NF‑κB109. Furthermore, IMD can also mediate caspase-
dependent apoptosis109, similar to the recently discovered
RIP1-dependent, caspase‑8-mediated apoptosis cascade
Maximal Maximal
cytoprotection cytoprotection in mammalian cells110.
Curiously, IMD lacks a kinase domain, so kinase-
zVAD.fmk Apoptotic cell Healthy cell dependent induction of necroptosis might be an evo-
Nec-1 Necroptotic cell lutionarily novel addition to the repertoire of RIP1
functions. Therefore, RIP1 might promote both the
Figure 3 | Plasticity of cell death activation in vivo. The existence of multiple cell
death mechanisms suggests that careful consideration should be given to determine induction of specific inflammatory signalling (by NF-κB
Nature Reviews | Molecular Cell Biology
which mechanism (or mechanisms) is primarily activated in any particular injury and IFN upregulation) and the elimination of infected
paradigm if therapeutic approaches are to prove useful. As an example, multiple cells through a pro-inflammatory process (necrosis). The
outcomes of anti-apoptotic or anti-necroptotic therapies can be anticipated depending promotion of necrosis can by itself serve to potentiate
on the specific injury paradigm. a | Predominant necrotic death might occur when the the antibacterial response by causing a leakage of cellular
endogenous conditions prohibit apoptosis. Use of necrostatin‑1 (Nec-1), which inhibits contents and the specific release of pro-inflammatory
RIP1-dependent necroptosis, might provide maximal cytoprotective benefit as a single mediators, such as IL-6 (Ref. 111). Recently discovered
agent under these circumstances. b | Alternatively, apoptotic cell death might be cell-death-independent activation of autophagy by
predominant in cell populations that are not subjected to excessive external stress, or TLR4–RIP1 as a potential mechanism for bacterial clear-
that are intrinsically deficient in necroptosis activation58,63. zVAD.fmk prevents apoptosis
ance also fits well with this notion103. These data suggest
by inhibiting caspases. c | A mixture of apoptotic and necroptotic cell death might occur,
leading to a significant, but partial, cytoprotective effect of each treatment and an that the evolution of the innate immune response might
additive effect of combination therapy. d | Apoptosis might be the predominant ‘primary’ have led to the acquisition of the RIP1-kinase-mediated
form of cell death. However, inhibition of apoptosis might result in the activation of necroptotic response by the IMD pathway.
necroptosis. In this scenario, neither apoptosis nor necroptosis inhibitors might work as
single agents, and combined treatment could provide maximal cytoprotective benefit. Opportunities for cytoprotective therapy
Catastrophic cell death is the main underlying cause of
death and lifelong disabilities in a broad range of human
Autophagy has a well-established role in defending diseases, from acute disorders (such as stroke, myo­cardial
against viral and bacterial invasion (reviewed in Ref. 98). infarction, brain and spinal cord trauma and septic shock)
Sindbis virus, a single-stranded RNA virus of the Togavirus to chronic neurodegenerative conditions. Cell death is
Macroautophagy family, causes encephalitis, which can be amelio­rated by also an important compounding factor as a side effect
The sequestration of cytosolic overexpression of Beclin-1 in transgenic mice99. Bacteria of chemotherapy, many inflammatory diseases, diabetes
components in that escape into the cytosol from the endosomes can be and other conditions. Therefore, the development of effi-
autophagosomes and their
subsequent degradation when
engulfed by macroautophagy100,101. Furthermore, autophagic cient strategies to inhibit pathological cell death remains
autophagosomes fuse with machinery promotes the clearance of extracellular a key challenge of cell death research and a crucial unmet
lysosomes. bacteria that are recognized by the TLR4 receptor102,103. medical need.

386 | May 2008 | volume 9 www.nature.com/reviews/molcellbio


© 2008 Nature Publishing Group
REVIEWS

a Survival mode b Apoptosis mode c Necroptosis mode

TNFα TNFα TNFα

TNFR TNFR TNFR


Cell
TRAF2

TRAF2

TRAF2
membrane

Necroptosis complex
RIP1

RIP1

RIP1

RIP1

RIP1
RIP1
TRADD

TRADD

TRADD
Ub ? Ub ? Ub ?
Complex I

Ub Ub Ub
Ub Ub Ub
Rac1

Rac1

Rac1
Ub ? Ub ? Ub ?
DISC
NIK

NIK
IKK

IKK

IKK
NIK
TRADD

NADPH NADPH NADPH


RIP1

oxidase oxidase oxidase


FADD DISC DISC
FLIP
NF-κB Caspase-8 Caspase-8 NF-κB
Cycloheximide,
Low level ZFRA, FAK–/–
zVAD.fmk,
ischaemic insult,
Survival Apoptosis caspase-8–/– Necroptosis
Inflammation
Proliferation Figure 4 | Activation of alternative cell fates following TNFα stimulation. In many cell types, apoptosis is not the
default response to tumour-necrosis factor‑α (TNFα) stimulation — multiple cell fates can be independently adopted.
a | In many cases, activation of nuclear factor-κB (NF‑κB) signalling, resulting from NF‑κB-activating complex I formation
and RIP1 polyubiquitylation136, is a primary response to TNFα. In this scenario, RIP1 ubiquitylation limits the formation of
pro-apoptotic signalling complexes137. In addition, NF-κB transcriptionally upregulates the expression of pro-survival
genes137. The low level of caspase‑8 cleaves RIP1, which inhibits necroptosis. b | Activation of apoptosis as a primary
response to TNFα requires specific apoptosis-promoting conditions, such as the presence Nature Reviews
of protein| Molecular
synthesisCell Biology
inhibitors
(cycloheximide), the overexpression of zinc finger-like protein (ZFRA) polypeptide138 or a deficiency in focal adhesion
kinase (FAK) kinase signalling139. This leads to efficient pro-apoptotic death-inducing-signalling complex (DISC)
formation136 and caspase‑8 activation. c | Activation of necroptosis as a primary response to TNFα requires suppression
of apoptotic signalling or, at least, caspase activity (by caspase inhibitors). Rapid loss of ATP140–143, conditions of
excessive reactive oxygen or nitrogen species production144 or ischaemic conditions63,130 can provide environments
that are non-permissive to apoptosis. Notably, all three of these pathways represent independent cell fates that are
selected on the basis of the specifics of the cellular regulation. IKK, inhibitor of NF-κB (IκB) kinase; NIK, NF-κB-inducing
kinase; RIP1, receptor-interacting protein-1; TRADD, TNF receptor type 1-associated death domain protein;
TRAF2, TNF receptor-associated factor-2; Ub, ubiquitin.

The discovery of apoptosis and the development of It is important to keep in mind that although apopto-
specific genetic and small molecule methods to inhibit sis may be the preferred type of physiological cell death,
pathological apoptosis in vivo have shown that pathological the option to die by apoptosis might not always be avail-
cell death can, indeed, be targeted for therapeutic benefit. able under in vivo conditions. Situations that involve an
However, the success of anti-apoptotic therapies has been imbalance of ROS-generation and ROS-detoxification,
limited, perhaps because of our lack of understanding of limited energy metabolism or a lack of proper protein
the complexity of cell death regulation in mammalian synthesis might restrict the ability of cells to activate
cells. The appreciation of such complexity leads us to sug- apoptotic cell death. Under such circumstances, cells
gest that when considering the possibility of inhibiting a might choose to die through one of the alternative
specific pathological mechanism of cell death, we must cell death pathways. The existence of other cell death
consider several issues. Does one particular form of cell options suggests that there might be some plasticity in
death have a major role in the injury (FIG. 3)? Alternatively, the choice of cell death programmes, with apoptosis
are several cell death mechanisms operational in the being only one of the spectrum of available regulated cell
injured tissue? If several mechanisms are operational, death options. Furthermore, rather than considering a
then combination therapy might provide maximal benefit. hierarchal regulation of multiple cell death mechanisms,
Whether a combination therapy will be effective depends in which cells first activate apoptosis and only undergo
on the contribution of each form of death, not only to the non-apoptotic cell death if apoptosis is inhibited, we
tissue injury, but also to the functional decline of the tissue propose that commitment to apoptosis might not even
(FIG. 3c). We must also consider whether a possible backup be necessary for the activation of non-apoptotic cell
cell death mechanism that might be activated in the event demise. In other words, non-apoptotic signalling can
of the primary mechanism is being inhibited. Given these be initiated independently through alternative mecha-
Interferon
An inflammatory cytokine that
possibilities, cytoprotective treatment might not only be nisms to carry out the order of cellular execution under
is produced by cells as part of improved by combination treatment, it might require conditions that are ill-suited for apoptosis, but ideal for
the innate immune response. combination treatment (FIG. 3d). non-apoptotic cell death (FIG. 4).

nature reviews | molecular cell biology volume 9 | May 2008 | 387


© 2008 Nature Publishing Group
REVIEWS

The discovery of the apoptotic programme opened the these features allowed specialized activation of cell death
door for the development of specific cell-death-targeting responses by various upstream stimuli, improved inte-
‘smart’ therapies. Pathological cell death pro­cesses might gration of apoptosis with other cellular signalling and
represent a ‘conscious’ decision of the cell in response metabolic pathways, and increased fidelity of apoptosis
to specific pathological signals, which would allow the execution, due to the redundancy in the functions of
development of specific approaches to influence this individual members of apoptosis regulatory families.
‘choice’ by small molecule or protein-based agents that Recent characterization of the inflammasome pathway
could target the apoptotic signalling machinery. This of caspase‑1 activation revealed that the evolution of
has already led to the development of multiple classes mammalian apoptosis probably involved convergence
of agent, such as BCL2 and inhibitor of apoptosis (IAP) of the primitive apoptosis machinery with the innate
proteins, that specifically trigger apoptosis in cancer immune system.
cells112. Although these agents are still in the early stages In the past few years we have also begun to appreciate
of clinical development, preliminary evidence is promis- that apoptosis is not the only form of regulated cell death.
ing. At the same time, therapies to eliminate catastrophic Three of the best-understood examples of non-apoptotic
tissue damage and functional decline, which are asso- cell death are type II cell death, necroptosis and PARP1-
ciated with pathological cell death in various human mediated necrotic death (see above). Although these
pathologies (from stroke to myocardial infarction), are processes can serve functions that are complimentary
still limited. The recent discovery of regulated non- or reinforce apoptosis, it is likely that they have evolved
apoptotic cell death might offer a new hope for treating to serve specific non-redundant functions in responses to
these diseases. Although the study of these forms of cell pathogen infection, nutrient and energy deprivation,
death is still in its infancy, a number of promising results and DNA damage. The changing perception of regu-
have already been generated (BOX 3). lated cell death as an array of diverse responses, rather
than a single apoptotic pathway, implicates complexity
Conclusion and provides novel opportunities for cytoprotective
Less than 2 decades ago, cell death was categorically con- therapies. In particular, the discovery of the specific
sidered to be passive and uninteresting. The discovery regulated, morphologically necrotic, non-apoptotic cell
of mammalian homologues of C. elegans cell death genes death mechanisms suggests that at least a subset of
led to the understanding that cell death, in the form of necrotic pathological cell death might also be regulated
apoptosis, can be a highly regulated cellular mechanism. by cellular mechanisms and, therefore, could be amend-
In-depth characterization of mammalian apoptosis able to therapeutic drug development. Understanding
uncovered both conservation of the basic layout of apop- how cell death operates under the specific conditions of
totic signalling in C. elegans and evolutionary expansion particular human diseases might bring in a new era
of the protein families of apoptotic regulators. Moreover, of cytoprotective drug development.

1. Yuan, J. & Horvitz, H. R. A first insight into the 12. Hofmann, E. R. et al. Caenorhabditis elegans HUS‑1 21. Degterev, A., Boyce, M. & Yuan, J. A decade of
molecular mechanisms of apoptosis. Cell 116, is a DNA damage checkpoint protein required for caspases. Oncogene 22, 8543–8567 (2003).
S53–S56 (2004). genome stability and EGL‑1‑mediated apoptosis. 22. Schulze-Osthoff, K., Ferrari, D., Los, M., Wesselborg, S.
The authors provide in-depth insight into the Curr. Biol. 12, 1908–1918 (2002). & Peter, M. E. Apoptosis signaling by death receptors.
discovery of the apoptotic machinery in C. elegans. 13. Tait, S. W. et al. Apoptosis induction by Bid requires Eur. J. Biochem. 254, 439–459 (1998).
2. Clarke, P. G. Developmental cell death: morphological unconventional ubiquitination and degradation of its 23. Micheau, O. & Tschopp, J. Induction of TNF receptor
diversity and multiple mechanisms. Anat. Embryol. N‑terminal fragment. J. Cell Biol. 179, 1453–1466 I‑mediated apoptosis via two sequential signaling
181, 195–213 (1990). (2007). complexes. Cell 114, 181–190 (2003).
This paper provides an early morphological 14. Konishi, Y., Lehtinen, M., Donovan, N. & Bonni, A. In this paper, upstream signalling events that lead
analysis of developmental cell death and describes Cdc2 phosphorylation of BAD links the cell cycle to the to the activation of apoptosis are described.
the existence of multiple forms of PCD. cell death machinery. Mol. Cell 9, 1005–1016 (2002). 24. Barnhart, B. C., Alappat, E. C. & Peter, M. E. The
3. Twomey, C. & McCarthy, J. V. Pathways of apoptosis 15. Zha, J., Weiler, S., Oh, K. J., Wei, M. C. & Korsmeyer, CD95 type I/type II model. Semin. Immunol. 15,
and importance in development. J. Cell. Mol. Med. 9, S. J. Posttranslational N‑myristoylation of BID as a 185–193 (2003).
345–359 (2005). molecular switch for targeting mitochondria and In this paper, distinct pathways (direct or
4. Yoshida, H. et al. Apaf1 is required for mitochondrial apoptosis. Science 290, 1761–1765 (2000). mitochondria-mediated) of executioner caspase
pathways of apoptosis and brain development. Cell 16. Li, H., Zhu, H., Xu, C. J. & Yuan, J. Cleavage of BID by activation by death receptors are described in detail.
94, 739–750 (1998). caspase 8 mediates the mitochondrial damage in the 25. McStay, G. P., Salvesen, G. S. & Green, D. R.
5. Gu, Y. et al. Activation of interferon-γ inducing factor Fas pathway of apoptosis. Cell 94, 491–501 (1998). Overlapping cleavage motif selectivity of caspases:
mediated by interleukin-1β converting enzyme. 17. Youle, R. J. & Strasser, A. The BCL‑2 protein family: implications for analysis of apoptotic pathways. Cell
Science 275, 206–209 (1997). opposing activities that mediate cell death. Nature Death Differ. 15, 322–331 (2007).
6. Wei, M. C. et al. tBID, a membrane-targeted death Rev. Mol. Cell Biol. 9, 47–59 (2008). 26. Sprick, M. R. et al. Caspase‑10 is recruited to and
ligand, oligomerizes BAK to release cytochrome c. 18. Scorrano, L. & Korsmeyer, S. J. Mechanisms of activated at the native TRAIL and CD95 death-
Genes Dev. 14, 2060–2071 (2000). cytochrome c release by proapoptotic BCL‑2 family inducing signalling complexes in a FADD-dependent
7. McConkey, D. J. Biochemical determinants of members. Biochem. Biophys. Res. Commun. 304, manner but can not functionally substitute caspase‑8.
apoptosis and necrosis. Toxicol. Lett. 99, 157–168 437–444 (2003). EMBO J. 21, 4520–4530 (2002).
(1998). This paper describes in detail the molecular 27. Su, H. et al. Requirement for caspase‑8 in NF‑kB
8. Zong, W. X. & Thompson, C. B. Necrotic death as a cell mechanism of the key step in apoptosis where activation by antigen receptor. Science 307,
fate. Genes Dev. 20, 1–15 (2006). mitochondrial cytochrome c is released. 1465–1468 (2005).
9. Horvitz, H. R., Shaham, S. & Hengartner, M. O. The 19. Willis, S. N. et al. Apoptosis initiated when BH3 28. Lassus, P., Opitz-Araya, X. & Lazebnik, Y. Requirement
genetics of programmed cell death in the nematode ligands engage multiple Bcl‑2 homologs, not Bax or for caspase‑2 in stress-induced apoptosis before
Caenorhabditis elegans. Cold Spring Harb. Symp. Bak. Science 315, 856–859 (2007). mitochondrial permeabilization. Science 297,
Quant. Biol. 59, 377–385 (1994). This paper presents evidence that suggests that 1352–1354 (2002).
10. Conradt, B. & Xue, D. Programmed cell death. anti-apoptotic BCL2 family members, rather than In this paper, a specific role for caspase‑2 in
WormBook, 1–13 (2005). BAX and BAK, are the targets of BH3-only proteins. genotoxic stress-induced apoptosis is shown.
11. Metzstein, M. M., Stanfield, G. M. & Horvitz, H. R. 20. Gogvadze, V., Orrenius, S. & Zhivotovsky, B. Multiple 29. Hitomi, J. et al. Involvement of caspase‑4 in
Genetics of programmed cell death in C. elegans: past, pathways of cytochrome c release from mitochondria endoplasmic reticulum stress-induced apoptosis and
present and future. Trends Genet. 14, 410–416 in apoptosis. Biochim. Biophys. Acta 1757, 639–647 Aβ-induced cell death. J. Cell Biol. 165, 347–356
(1998). (2006). (2004).

388 | May 2008 | volume 9 www.nature.com/reviews/molcellbio


© 2008 Nature Publishing Group
REVIEWS

30. Nakagawa, T. et al. Caspase‑12 mediates 51. Maiuri, M. C., Zalckvar, E., Kimchi, A. & Kroemer, G. 71. Temkin, V., Huang, Q., Liu, H., Osada, H. & Pope,
endoplasmic‑reticulum‑specific apoptosis and Self-eating and self-killing: crosstalk between R. M. Inhibition of ADP/ATP exchange in receptor-
cytotoxicity by amyloid-β. Nature 403, 98–103 (2000). autophagy and apoptosis. Nature Rev. Mol. Cell Biol. interacting protein-mediated necrosis. Mol. Cell Biol.
This paper provides the first demonstration of the 8, 741–752 (2007). 26, 2215–2225 (2006).
specific role of caspase‑12 in ER stress response. 52. Ullman, E. et al. Autophagy promotes necrosis in 72. Lim, S. Y., Davidson, S. M., Mocanu, M. M.,
31. Zou, H., Li, Y., Liu, X. & Wang, X. An apoptosis-deficient cells in response to ER stress. Cell Yellon, D. M. & Smith, C. C. The cardioprotective effect
APAF‑1cytochrome c multimeric complex is a Death Differ. 15, 422–425 (2008). of necrostatin requires the cyclophilin‑D component of
functional apoptosome that activates procaspase‑9. 53. Shimizu, S. et al. Role of Bcl‑2 family proteins in a non- the mitochondrial permeability transition pore.
J. Biol. Chem. 274, 11549–11556 (1999). apoptotic programmed cell death dependent on Cardiovasc. Drugs Ther. 21, 467–469 (2007).
The authors establish the existence of the autophagy genes. Nature Cell Biol. 6, 1221–1228 73. Suzuki, S. et al. Nur77 as a survival factor in tumor
caspase‑9‑activating apoptotosome complex. (2004). necrosis factor signaling. Proc. Natl Acad. Sci. USA
32. Gao, Z., Shao, Y. & Jiang, X. Essential roles of the 54. Oberstein, A., Jeffrey, P. D. & Shi, Y. Crystal structure 100, 8276–8280 (2003).
Bcl‑2 family of proteins in caspase‑2‑induced of the Bcl‑XL‑Beclin 1 peptide complex: Beclin 1 is a 74. Jagtap, P. & Szabo, C. Poly(ADP-ribose) polymerase
apoptosis. J. Biol. Chem. 280, 38271–38275 (2005). novel BH3-only protein. J. Biol. Chem. 282, and the therapeutic effects of its inhibitors. Nature
33. Bouillet, P. et al. Proapoptotic Bcl‑2 relative Bim 13123–13132 (2007). Rev. Drug Discov. 4, 421–440 (2005).
required for certain apoptotic responses, leukocyte 55. Pattingre, S. et al. Bcl‑2 antiapoptotic proteins inhibit 75. Oei, S. L., Keil, C. & Ziegler, M. Poly(ADP-ribosylation)
homeostasis, and to preclude autoimmunity. Science Beclin 1‑dependent autophagy. Cell 122, 927–939 and genomic stability. Biochem. Cell Biol. 83,
286, 1735–1738 (1999). (2005). 263–269 (2005).
This is one of the earliest papers establishing This paper presents evidence for the regulation of 76. Zong, W. X., Ditsworth, D., Bauer, D. E., Wang, Z. Q. &
specific roles of a particular BH3-only factor, BIM, autophagy by the anti-apoptotic BCL2 family Thompson, C. B. Alkylating DNA damage stimulates a
in response to specific upstream signals. members, establishing convergent regulation of regulated form of necrotic cell death. Genes Dev. 18,
34. Yu, J. & Zhang, L. The transcriptional targets of p53 in apoptosis and autophagy. 1272–1282 (2004).
apoptosis control. Biochem. Biophys. Res. Commun. 56. Tracy, K. & Macleod, K. F. Regulation of mitochondrial This paper establishes the existence of the ‘death
331, 851–588 (2005). integrity, autophagy and cell survival by BNIP3. by energy collapse’ mechanism of
35. Schmelzle, T. et al. Functional role and oncogene- Autophagy 3, 616–619 (2007). PARP‑1‑mediated cell death.
regulated expression of the BH3-only factor Bmf in 57. Rashmi, R., Pillai, S. G., Vijayalingam, S., Ryerse, J. & 77. Ditsworth, D., Zong, W. X. & Thompson, C. B.
mammary epithelial anoikis and morphogenesis. Chinnadurai, G. BH3-only protein BIK induces Activation of poly(ADP)-ribose polymerase (PARP‑1)
Proc. Natl Acad. Sci. USA 104, 3787–3792 (2007). caspase-independent cell death with autophagic induces release of the pro-inflammatory mediator
36. Chen, L. et al. Differential targeting of prosurvival features in Bcl‑2 null cells. Oncogene 27, 1366–1375 HMGB1 from the nucleus. J. Biol. Chem. 282,
Bcl‑2 proteins by their BH3-only ligands allows (2007). 17845–17854 (2007).
complementary apoptotic function. Mol. Cell 17, 58. Khwaja, A. & Tatton, L. Resistance to the cytotoxic 78. Andrabi, S. A. et al. Poly(ADP-ribose) (PAR) polymer is
393–403 (2005). effects of tumor necrosis factor a can be overcome a death signal. Proc. Natl Acad. Sci. USA 103,
This paper demonstrates that selective interactions by inhibition of a FADD/caspase-dependent signaling 18308–18313 (2006).
of pro- and anti-apoptotic BCL2 family members pathway. J. Biol. Chem. 274, 36817–36823 79. Yu, S. W. et al. Apoptosis-inducing factor mediates
contribute to the regulation of the mitochondrial (1999). poly(ADP-ribose) (PAR) polymer-induced cell death.
step in apoptosis. 59. Matsumura, H. et al. Necrotic death pathway in Fas Proc. Natl Acad. Sci. USA 103, 18314–18319
37. Takai, Y. et al. Caspase‑12 compensates for lack of receptor. J. Cell Biol. 151, 1247–1256 (2000). (2006).
caspase‑2 and caspase‑3 in female germ cells. 60. Vercammen, D. et al. Inhibition of caspases increases This paper describes the mechanism of the
Apoptosis 12, 791–800 (2007). the sensitivity of L929 cells to necrosis mediated by PARP1–AIF cell death.
38. Troy, C. M. et al. Death in the balance: alternative tumor necrosis factor. J. Exp. Med. 187, 1477–1485 80. Xu, Y., Huang, S., Liu, Z. G. & Han, J. Poly(ADP-ribose)
participation of the caspase‑2 and ‑9 pathways in (1998). polymerase‑1 signaling to mitochondria in necrotic cell
neuronal death induced by nerve growth factor 61. Vercammen, D. et al. Dual signaling of the Fas death requires RIP1/TRAF2-mediated JNK1
deprivation. J. Neurosci. 21, 5007–5016 (2001). receptor: initiation of both apoptotic and necrotic cell activation. J. Biol. Chem. 281, 8788–8795
39. Garcia-Calvo, M. et al. Inhibition of human caspases death pathways. J. Exp. Med. 188, 919–930 (1998). (2006).
by peptide-based and macromolecular inhibitors. 62. Chan, F. K. et al. A role for tumor necrosis factor 81. Ellis, H. M. & Horvitz, H. R. Genetic control of
J. Biol. Chem. 273, 32608–32613 (1998). receptor‑2 and receptor-interacting protein in programmed cell death in the nematode C. elegans.
40. Abraham, M. C., Lu, Y. & Shaham, S. programmed necrosis and antiviral responses. J. Biol. Cell 44, 817–829 (1986).
A morphologically conserved nonapoptotic program Chem. 278, 51613–51621 (2003). 82. Aballay, A. & Ausubel, F. M. Programmed cell death
promotes linker cell death in Caenorhabditis elegans. 63. Degterev, A. et al. Chemical inhibitor of nonapoptotic mediated by ced‑3 and ced‑4 protects Caenorhabditis
Dev. Cell 12, 73–86 (2007). cell death with therapeutic potential for ischemic brain elegans from Salmonella typhimurium-mediated
41. Oppenheim, R. W. et al. Programmed cell death of injury. Nature Chem. Biol. 1, 112–119 (2005). killing. Proc. Natl Acad. Sci. USA 98, 2735–2739
developing mammalian neurons after genetic deletion This paper describes a first‑in‑class potent and (2001).
of caspases. J. Neurosci. 21, 4752–4760 (2001). selective inhibitor of necroptosis. 83. Martinon, F., Gaide, O., Petrilli, V., Mayor, A. &
42. Yuan, J. Inducing autophagy harmlessly. Autophagy 4, 64. Holler, N. et al. Fas triggers an alternative, Tschopp, J. NALP inflammasomes: a central role in
249–250 (2007). caspase‑8‑independent cell death pathway using the innate immunity. Semin. Immunopathol. 29, 213–229
43. Lee, C. Y. & Baehrecke, E. H. Steroid regulation of kinase RIP as effector molecule. Nature Immunol. 1, (2007).
autophagic programmed cell death during 489–495 (2000). This paper provides an extensive review of the
development. Development 128, 1443–1455 In this paper, the crucial role of RIP1 kinase activity mechanism of action and regulation of mammalian
(2001). in the activation of necroptosis is first inflammasomes.
44. Berry, D. L. & Baehrecke, E. H. Growth arrest and demonstrated. 84. Martinon, F. & Tschopp, J. NLRs join TLRs as innate
autophagy are required for salivary gland cell 65. Kawahara, A., Ohsawa, Y., Matsumura, H., Uchiyama, Y. sensors of pathogens. Trends Immunol. 26, 447–454
degradation in Drosophila. Cell 131, 1137–1148 & Nagata, S. Caspase-independent cell killing by Fas- (2005).
(2007). associated protein with death domain. J. Cell Biol. 85. Ting, J. P., Kastner, D. L. & Hoffman, H. M.
This paper demonstrates that autophagic cell 143, 1353–1360 (1998). CATERPILLERs, pyrin and hereditary immunological
death is specifically activated during development This paper is one of the initial reports showing disorders. Nature Rev. Immunol. 6, 183–195
under apoptosis-competent conditions. induction of necrotic death by death-domain (2006).
45. Marino, G. et al. Tissue-specific autophagy alterations receptor signals. 86. Srinivasula, S. M. et al. The PYRIN-CARD protein ASC
and increased tumorigenesis in mice deficient in 66. Festjens, N., Vanden Berghe, T. & Vandenabeele, P. is an activating adaptor for caspase‑1. J. Biol. Chem.
Atg4C/autophagin‑3. J. Biol. Chem. 282, Necrosis, a well-orchestrated form of cell demise: 277, 21119–21122 (2002).
18573–18583 (2007). signalling cascades, important mediators and 87. Martinon, F., Burns, K. & Tschopp, J. The
46. Komatsu, M. et al. Impairment of starvation-induced concomitant immune response. Biochim. Biophys. inflammasome: a molecular platform triggering
and constitutive autophagy in Atg7-deficient mice. Acta 1757, 1371–1387 (2006). activation of inflammatory caspases and processing of
J. Cell Biol. 169, 425–434 (2005). This paper provides an in-depth analysis of the proIL-β. Mol. Cell 10, 417–426 (2002).
47. Kuma, A. et al. The role of autophagy during the early signalling and execution pathways of regulated This paper provides an initial report demonstrating
neonatal starvation period. Nature 432, 1032–1036 necrosis. the existance of the caspase‑1 activating
(2004). 67. Zheng, L. et al. Competitive control of independent inflammasome complex.
This paper demonstrates the crucial role of programs of tumor necrosis factor receptor-induced 88. Bruey, J. M. et al. Bcl‑2 and Bcl-XL regulate
autophagy in vivo in maintaining survival under cell death by TRADD and RIP1. Mol. Cell Biol. 26, proinflammatory caspase‑1 activation by interaction
nutrient-deprivation conditions. 3505–3513 (2006). with NALP1. Cell 129, 45–56 (2007).
48. Juhasz, G., Erdi, B., Sass, M. & Neufeld, T. P. This paper shows that apoptotic and necroptotic This paper presents evidence that inflammasome
Atg7-dependent autophagy promotes neuronal signalling pathways diverge at the level of the activity is inhibited by BCL2 family members.
health, stress tolerance, and longevity but is death-domain receptor. 89. Petrilli, V., Dostert, C., Muruve, D. A. & Tschopp, J.
dispensable for metamorphosis in Drosophila. Genes 68. Wang, K. et al. Structure-activity relationship analysis The inflammasome: a danger sensing complex
Dev. 21, 3061–3066 (2007). of a novel necroptosis inhibitor, Necrostatin‑5. triggering innate immunity. Curr. Opin. Immunol. 19,
49. Espert, L. et al. Autophagy is involved in T cell death Bioorg. Med. Chem. Lett. 17, 1455–1465 (2007). 615–622 (2007).
after binding of HIV‑1 envelope proteins to CXCR4. 69. Jagtap, P. G. et al. Structure-activity relationship study 90. Jones, J. D. & Dangl, J. L. The plant immune system.
J. Clin. Invest. 116, 2161–2172 (2006). of tricyclic necroptosis inhibitors. J. Med. Chem. 50, Nature 444, 323–329 (2006).
50. Pattingre, S., Espert, L., Biard-Piechaczyk, M. & 1886–1895 (2007). 91. Schmitz, J. et al. IL‑33, an interleukin‑1‑like cytokine
Codogno, P. Regulation of macroautophagy by mTOR 70. Teng, X. et al. Structure-activity relationship study of that signals via the IL‑1 receptor-related protein ST2
and Beclin 1 complexes. Biochimie 90, novel necroptosis inhibitors. Bioorg. Med. Chem. Lett. and induces T helper type 2-associated cytokines.
313–323(2007). 15, 5039–5044 (2005). Immunity 23, 479–490 (2005).

nature reviews | molecular cell biology volume 9 | May 2008 | 389


© 2008 Nature Publishing Group
REVIEWS

92. Faustin, B. et al. Reconstituted NALP1 inflammasome 113. Shimizu, S., Konishi, A., Kodama, T. & Tsujimoto, Y. 132. Horvath, E. M. & Szabo, C. Poly(ADP-ribose)
reveals two-step mechanism of caspase‑1 activation. BH4 domain of antiapoptotic Bcl‑2 family members polymerase as a drug target for cardiovascular disease
Mol. Cell 25, 713–724 (2007). closes voltage-dependent anion channel and inhibits and cancer: an update. Drug News Perspect. 20,
93. Franchi, L. et al. Cytosolic flagellin requires Ipaf for apoptotic mitochondrial changes and cell death. Proc. 171–181 (2007).
activation of caspase‑1 and interleukin 1β in Natl Acad. Sci. USA 97, 3100–3105 (2000). 133. Kauppinen, T. M. & Swanson, R. A. The role of
salmonella-infected macrophages. Nature Immunol. 7, 114. Cheng, E. H. et al. Conversion of Bcl‑2 to a Bax-like poly(ADP-ribose) polymerase‑1 in CNS disease.
576–582 (2006). death effector by caspases. Science 278, 1966–1968 Neuroscience 145, 1267–1272 (2007).
94. Miao, E. A. et al. Cytoplasmic flagellin activates (1997). 134. Zaremba, T. & Curtin, N. J. PARP inhibitor
caspase‑1 and secretion of interleukin 1β via Ipaf. 115. Acehan, D. et al. Three-dimensional structure of the development for systemic cancer targeting. Anticancer
Nature Immunol. 7, 569–575 (2006). apoptosome: implications for assembly, procaspase‑9 Agents Med. Chem. 7, 515–523 (2007).
References 93 and 94 establish that IPAF binding, and activation. Mol. Cell 9, 423–432 (2002). 135. Martinon, F. & Tschopp, J. Inflammatory caspases and
inflammasomes can sense intracellular PAMP This is the first paper to provide structural insight inflammasomes: master switches of inflammation. Cell
signals. into the functioning of the apoptosome. Death Differ. 14, 10–22 (2007).
95. Shen, Q. H. et al. Nuclear activity of MLA immune 116. Park, H. H. et al. Death domain assembly mechanism 136. Muppidi, J. R., Tschopp, J. & Siegel, R. M. Life and
receptors links isolate-specific and basal disease- revealed by crystal structure of the oligomeric death decisions: secondary complexes and lipid rafts
resistance responses. Science 315, 1098–1103 PIDDosome core complex. Cell 128, 533–546 (2007). in TNF receptor family signal transduction. Immunity
(2007). 117. Mizushima, N. Autophagy: process and function. 21, 461–465 (2004).
96. Woltering, E. J. Death proteases come alive. Trends Genes Dev. 21, 2861–2873 (2007). 137. O’Donnell, M. A., Legarda-Addison, D., Skountzos, P.,
Plant Sci. 9, 469–472 (2004). This is an in-depth review that describes recent Yeh, W. C. & Ting, A. T. Ubiquitination of RIP1
97. Vercammen, D., Declercq, W., Vandenabeele, P. & progress in understanding autophagic regulation. regulates an NF‑κB‑independent cell-death switch in
Van Breusegem, F. Are metacaspases caspases? 118. Yu, L. et al. Regulation of an ATG7–beclin 1 program TNF signaling. Curr. Biol. 17, 418–424 (2007).
J. Cell Biol. 179, 375–380 (2007). of autophagic cell death by caspase‑8. Science 304, 138. Hong, Q. et al. Zfra affects TNF-mediated cell death by
98. Schmid, D. & Munz, C. Innate and adaptive immunity 1500–1502 (2004). interacting with death domain protein TRADD and
through autophagy. Immunity 27, 11–21 (2007). This paper demonstrates activation of autophagic negatively regulates the activation of NF‑κB, JNK1,
Reference 98 provides a detailed overview of the cell death resulting from caspase inhibition. p53 and WOX1 during stress response. BMC Mol.
role of autophagy in the immune regulation. 119. Kim, Y. S., Morgan, M. J., Choksi, S. & Liu, Z. G. Biol. 8, 50 (2007).
99. Liang, X. H. et al. Protection against fatal Sindbis virus TNF-induced activation of the Nox1 NADPH oxidase 139. Takahashi, R. et al. Focal adhesion kinase determines
encephalitis by beclin, a novel Bcl‑2‑interacting and its role in the induction of necrotic cell death. Mol. the fate of death or survival of cells in response to
protein. J. Virol. 72, 8586–8596 (1998). Cell 26, 675–687 (2007). TNFα in the presence of actinomycin D. Biochim.
100. Nakagawa, I. et al. Autophagy defends cells against 120. Schulze-Osthoff, K. et al. Cytotoxic activity of tumor Biophys. Acta 1770, 518–526 (2007).
invading group A Streptococcus. Science 306, necrosis factor is mediated by early damage of 140. Leist, M. et al. Inhibition of mitochondrial ATP
1037–1040 (2004). mitochondrial functions. Evidence for the involvement generation by nitric oxide switches apoptosis to
This paper presents evidence that the activation of of mitochondrial radical generation. J. Biol. Chem. necrosis. Exp. Cell Res. 249, 396–403 (1999).
autophagy can serve a host defence function. 267, 5317–5323 (1992). 141. Leist, M., Single, B., Castoldi, A. F., Kuhnle, S. &
101. Py, B. F., Lipinski, M. M. & Yuan, J. Autophagy limits 121. Festjens, N. et al. Butylated hydroxyanisole is more Nicotera, P. Intracellular adenosine triphosphate (ATP)
Listeria monocytogenes intracellular growth in the than a reactive oxygen species scavenger. Cell Death concentration: a switch in the decision between
early phase of primary infection. Autophagy 3, Differ. 13, 166–169 (2006). apoptosis and necrosis. J. Exp. Med. 185,
117–125 (2007). 122. Thon, L. et al. Ceramide mediates caspase- 1481–1486 (1997).
102. Sanjuan, M. A. et al. Toll-like receptor signalling in independent programmed cell death. FASEB. J. 19, This paper introduces the idea that intrinsic
macrophages links the autophagy pathway to 1945–1956 (2005). differences in cellular energy levels might have a
phagocytosis. Nature 450, 1253–1257 (2007). 123. Ame, J. C., Spenlehauer, C. & de Murcia, G. The PARP key role in the selection of a cell’s form of death:
103. Xu, Y. et al. Toll-like receptor 4 is a sensor for autophagy superfamily. Bioessays 26, 882–893 (2004). either apoptotic or necrotic.
associated with innate immunity. Immunity 27, This is a detailed review of the structure and 142. Volbracht, C., Leist, M. & Nicotera, P. ATP controls
135–144 (2007). function of PARP family members. neuronal apoptosis triggered by microtubule
104. Inbal, B., Bialik, S., Sabanay, I., Shani, G. & Kimchi, A. 124. Vahsen, N. et al. AIF deficiency compromises oxidative breakdown or potassium deprivation. Mol. Med. 5,
DAP kinase and DRP‑1 mediate membrane blebbing phosphorylation. EMBO J. 23, 4679–4689 (2004). 477–489 (1999).
and the formation of autophagic vesicles during In this report, the metabolic function of AIF in the 143. Nicotera, P., Leist, M., Fava, E., Berliocchi, L. &
programmed cell death. J. Cell Biol. 157, 455–468 regulating activity of mitochondrial complex I is Volbracht, C. Energy requirement for caspase
(2002). established. activation and neuronal cell death. Brain Pathol. 10,
105. Meylan, E. & Tschopp, J. The RIP kinases: crucial 125. Zhu, H. et al. Cardiac autophagy is a maladaptive 276–282 (2000).
integrators of cellular stress. Trends Biochem. Sci. 30, response to hemodynamic stress. J. Clin. Invest. 117, 144. Dimmeler, S., Haendeler, J., Sause, A. & Zeiher, A. M.
151–159 (2005). 1782–1793 (2007). Nitric oxide inhibits APO‑1/Fas-mediated cell death.
A detailed review that discusses the structure and 126. Matsui, Y. et al. Distinct roles of autophagy in the Cell Growth Differ. 9, 415–422 (1998).
function of the RIP kinase family. heart during ischemia and reperfusion: roles of AMP-
106. Cusson-Hermance, N., Khurana, S., Lee, T. H., activated protein kinase and Beclin 1 in mediating Acknowledgements
Fitzgerald, K. A. & Kelliher, M. A. Rip1 mediates the autophagy. Circ. Res. 100, 914–922 (2007). J.Y. is supported in part by grants from the National Institute
Trif-dependent toll-like receptor 3‑ and 4‑induced 127. Amaravadi, R. K. et al. Autophagy inhibition enhances on Aging (NIA) Merit Award (R37 AG12859) and the National
NF‑κB activation but does not contribute to interferon therapy-induced apoptosis in a Myc-induced model of Institutes of Health (NIH) Director’s Pioneer Award. A.D. is
regulatory factor 3 activation. J. Biol. Chem. 280, lymphoma. J. Clin. Invest. 117, 326–336 (2007). supported in part by NIA Mentored Research Scientist Career
36560–36566 (2005). This paper demonstrates that the induction of Development Award and Massachusetts Medical Foundation
107. Meylan, E. et al. RIP1 is an essential mediator of Toll- autophagy in cancer cells in vivo enhances tumour Smith Family New Investigator Award.
like receptor 3‑induced NF‑κB activation. Nature growth through the inhibition of apoptosis.
Immunol. 5, 503–507 (2004). 128. Edinger, A. L. & Thompson, C. B. Defective autophagy
108. Balachandran, S., Thomas, E. & Barber, G. N. A FADD- leads to cancer. Cancer Cell 4, 422–424 (2003). DATABASES
dependent innate immune mechanism in mammalian 129. Degenhardt, K. et al. Autophagy promotes tumor cell Interpro: http://www.ebi.ac.uk/interpro
cells. Nature 432, 401–405 (2004). survival and restricts necrosis, inflammation, and LRRs | NACHT | Pyrin
109. Georgel, P. et al. Drosophila immune deficiency (IMD) tumorigenesis. Cancer Cell 10, 51–64 (2006). UniProtKB: http://beta.uniprot.org/
is a death domain protein that activates antibacterial This paper demonstrates the plasticity of cancer AIF | ANT | APAF1 | ASC | ATG5 | ATG7 | BAK | BAX | BCL2 |
defense and can promote apoptosis. Dev. Cell 1, cell death activation under hypoxic conditions by BCL-XL | Beclin-1 | BID | BIM | Caspase‑9 | CED‑3 | CED‑4 |
503–514 (2001). demonstrating the activation of apoptosis, CED‑9 | CIITA | CXCR4 | DAP | EGL‑1 | HMGB1 | MAPK8 |
110. Petersen, S. L. et al. Autocrine TNFα signaling autophagy and necrosis that is dependent on the MCL1 | NAIP | NALP1 | NLRC4 | NLRX1 | PARP1 | PIDD | RIP1 |
renders human cancer cells susceptible to expression of specific protein factors in the cell. TNFα | TRAF2 |
Smac‑mimetic‑induced apoptosis. Cancer Cell 12, 130. Smith, C. C. et al. Necrostatin: a potentially novel
445–456 (2007). cardioprotective agent? Cardiovasc. Drugs Ther. 21, FURTHER INFORMATION
111. Vanden Berghe, T. et al. Necrosis is associated with 227–233 (2007). Junying Yuan’s homepage:
IL‑6 production but apoptosis is not. Cell Signal 18, 131. de la Lastra, C. A., Villegas, I. & Sanchez-Fidalgo, S. http://www.hms.harvard.edu/dms/bbs/fac/yuan.html
328–335 (2006). Poly(ADP-ribose) polymerase inhibitors: new Alexei Degterev’s homepage: http://www.tufts.edu/med/
112. Korkina, O. & Degterev, A. in Wiley Encyclopedia of pharmacological functions and potential clinical biochemistry/faculty/degterev/degterev.html
Chemical Biology (ed. Begley, T. P.) (John Wiley and implications. Curr. Pharm. Des. 13, 933–962 All links are active in the online pdf
Sons, Ltd, 2008). (2007).

390 | May 2008 | volume 9 www.nature.com/reviews/molcellbio


© 2008 Nature Publishing Group

You might also like