Tumor necrosis factor: an apoptosis JuNKie?
TNF's main function is to stimulate inflammation by turning on gene transcription through the IKK/NFkappaB and JNK/AP-1 signaling cascades. TNF also can trigger apoptosis through caspase-8, but the role and underlying mechanism of this activity are not fully understood. Here, we review recent data on the role of JNK in the regulation of TNF-dependent apoptosis and discuss what is known so far about how cells decide whether to live or die in response to TNF.
Cell, Vol. 116, 491–497, February 20, 2004, Copyright 2004 by Cell Press
ReviewTumor Necrosis Factor:
An Apoptosis JuNKie?
Chen and Goeddel, 2002) (Figure 1). IKK stimulates
NF-B by catalyzing phosphorylation and degradation
of the NF-B inhibitor, I-B (reviewed in Karin and Lin,
Eugene E. Varfolomeev and Avi Ashkenazi*
1 DNA Way
South San Francisco, California 94080 2002). TRAF2 functions also as an obligatory conduit
for stimulation of JNK through its MAPK kinase MKK7,
promoting phosphorylation of c-Jun and thus increasing
AP-1 activity. Unlike TNFR1, TNFR2 binds TRAF2 di-TNF’s main function is to stimulate inflammation by
turning on gene transcription through the IKK/NFB rectly, hence activating IKK and JNK (Figure 1). TRAF2
also recruits ancillary proteins that modulate signalingand JNK/AP-1 signaling cascades. TNF also can trig-
ger apoptosis through caspase-8, but the role and through each TNFR, i.e., cIAP (cellular inhibitor of apo-
ptosis protein) 1 and 2, and TRAF1. cIAP1 supportsunderlying mechanism of this activity are not fully un-
derstood. Here, we review recent data on the role of ubiquitination and degradation of TRAF2, whereas
TRAF1 inhibits TNFR2-dependent signaling through anJNK in the regulation of TNF-dependent apoptosis and
discuss what is known so far about how cells decide unknown mechanism (Wajant et al., 2003).
whether to live or die in response to TNF.
TNF as a Conditional Death Ligand Blocked
TNF as a Key Inducer of Proinflammatory Genes
Tumor necrosis factor-␣ (TNF) plays a pivotal role in
While TNF’s cardinal role is to stimulate inflammation,
orchestrating innate inflammatory responses in verte-
it is capable also of inducing apoptosis when NF-B
brates. Upon detection of invading intracellular patho-
signaling is blocked. The precise biological role of this
gens, tissue macrophages and T cells produce either
activity is unclear. TNF may function alongside of “pro-
membrane-associated TNF (mTNF) or proteolytically
fessional” death ligands such as FasL and Apo2L/TRAIL
derived soluble TNF (sTNF). TNF triggers local expres-
to help cytotoxic leukocytes kill pathogen-infected cells.
sion of chemokines and cytokines, promoting the adhe-
Perhaps TNF’s apoptotic capability contributes to its
sion, extravasation, attraction, and activation of leuko-
established pathological role in rheumatoid arthritis and
cytes at the site of infection. Later, TNF facilitates
inflammatory bowel disease.
transition from innate to acquired immunity by enhanc-
The “intrinsic” apoptosis pathway—triggered by intra-
ing antigen presentation and T cell costimulation.
cellular injury such as DNA damage—controls caspase
TNF is the prototype of ⵑ20 related cytokines that act
activation through the Bcl-2 gene family (reviewed in
through specific members of the TNF receptor (TNFR)
Puthalakath and Strasser, 2002; Cory and Adams, 2002).
superfamily, mainly to modulate immunity (reviewed in
In this pathway, damage sensors induce transcription
Locksley et al., 2001). TNF homologs exist in insects,
of Bcl-2 homology 3 (BH3) domain proteins (e.g., Puma,
primitive chordates, amphibians, fish, birds, and mam-
Noxa, Bim, Bmf). These apical regulators activate down-
mals. Mammalian TNF signals through two distinct cell-
stream proapoptotic Bcl-2 relatives (e.g., Bax, Bak),
surface receptors: TNFR1, the primary receptor for
overcoming inhibition by antiapoptotic Bcl-2 family
sTNF, and TNFR2, the main receptor for mTNF (reviewed
members (e.g., Bcl-2, Bcl-X
). The activated Bcl-2 rela-
in Wajant et al., 2003). These receptors trigger several
tives trigger mitochondrial release of factors that pro-
intracellular signaling pathways, including the I-kB ki-
mote caspase activation in the cytosol. One factor is
nase (IKK), c-Jun N-terminal kinase (JNK), and p38 or
cytochrome c, which cooperates with Apaf-1 to activate
p42/44 mitogen-activated protein kinase (MAPK) cas-
caspase-9. This apical caspase activates the effector
cades, which control gene expression through tran-
caspases 3, 6, and 7, causing apoptotic death. Two
scription factors such as NF-B and AP-1.
other mitochondrial factors, Smac/Diablo and Omi/
Most cell types constitutively express TNFR1 while
HtrA2, prevent IAPs from inhibiting caspase activation.
TNFR2 expression is highly regulated. TNFR1 and 2 re-
The “extrinsic” pathway is triggered by extracellular
semble each other in their extracellular, cysteine-rich
death ligands such as the TNF relatives FasL and Apo2L/
domains. TNFR1 contains a cytoplasmic death domain
TRAIL, which signal respectively through Fas and DR4
(DD) that binds to the adaptor TRADD (TNFR-associated
or DR5 (reviewed in Ashkenazi, 2002). These death re-
DD). TNFR2 lacks a DD, but has a cytoplasmic motif
ceptors bind directly to the adaptor FADD (Fas-associ-
that binds TRAFs (TNFR-associated factors). The homo-
ated DD), which mediates recruitment and activation of
trimeric TNF ligand binds a pre-associated receptor ho-
caspases-8 and -10 within a death-inducing signaling
motrimer, inducing conformational changes that enable
complex (DISC). Caspases-8 and -10 activate apoptotic
the cytoplasmic motifs to bind cognate signaling adap-
death through the same effector caspases as the
tors (Locksley et al., 2001). Upon binding to ligated
intrinsic pathway. Modulation of the extrinsic pathway
TNFR1, TRADD recruits the secondary adaptors RIP1,
occurs at several levels. Decoy receptors can compete
TRAF2, or TRAF5. This causes activation of the IKK
with death receptors for ligand binding. The caspase-
complex, which consists of IKK ␣, ␤, and ␥ (also called
related molecule c-FLIP (cellular FLICE-inhibitory pro-
Nemo), through an unknown mechanism (reviewed in
tein), which lacks catalytic activity, competes with cas-
pases-8 and -10 for DISC binding (reviewed in Thome
and Tschopp, 2001). Further downstream, IAPs inhibit
liver damage that requires both TNFR1 and 2 for maximal
apoptosis. Thus, mTNF uses both receptors to trigger
hepatocyte death (Maeda et al., 2003 and references
TNFR1 shares several components of the extrinsic
death pathway with Fas, DR4, and DR5. Mouse embryo
fibroblasts (MEFs) deficient in either FADD or cas-
pase-8 resist TNF-induced apoptosis, demonstrating an
obligatory role for these molecules. Previous studies
with transfected cells suggested that TNFR1 assembles
a DISC similar to that of Fas and DR4/5, except that this
occurs indirectly through TRADD (Chen and Goeddel,
2002). A recent study examined a wild-type HT1080 hu-
man fibroblast cell line, resistant to sTNF-induced apo-
ptosis, and a mutant line with defective NF-B activa-
tion, sensitive to TNF killing (Micheau and Tschopp,
2003). In both lines, sTNF induces TRADD-mediated as-
sembly of a TNFR1-associated complex (complex I) that
contains RIP1, TRAF2, and cIAP1 and activates the IKK/
NF-B pathway (Figure 2). Subsequently, TRADD, RIP1,
and TRAF2 undergo biochemical modifications and the
complex dissociates from TNFR1, moving to the cytosol.
FADD and caspase-8 bind to this cytosolic complex
(complex II). In wild-type cells, complex II contains abun-
dant c-FLIP but little caspase-10, while in mutant cells
the converse is true. Thus, c-FLIP may be an important
Figure 1. Model for Control of Gene Transcription by TNF
NF-B-dependent factor preventing apical caspase ac-
tivation at the level of complex II, possibly by competing
for caspase-10 binding. Besides c-FLIP, cIAP1 and
effector caspase activation (reviewed in Salvesen and
TRAF1 are more abundant in complex II of the mutant
Duckett, 2002). In “type I cells,” the DISC generates
cells and might inhibit caspase activation (Micheau and
sufficient caspase activity to trigger death. In “type II
Tschopp, 2003; Wang et al., 1998).
cells,” apical caspase activation is weaker, and apopto-
Simultaneous engagement of both TNFRs amplifies
sis requires amplification through crosstalk to the intrin-
TNF-induced apoptosis (Wajant et al., 2003; Maeda et
sic pathway: caspase-8 cleaves the BH3 protein Bid,
al., 2003). This correlates with increased TNFR2-induced
which stimulates Bax and Bak to augment caspase acti-
TRAF2 degradation. Since TRAF2 recruits cIAPs to
vation (reviewed in Peter and Krammer, 2003).
TNFR1, its degradation via TNFR2 may facilitate cell
The pleiotropic nature of TNF has hindered elucidation
death. TRAF2 destruction also attenuates TNFR1-medi-
of its apoptosis signaling mechanism. TNF does not
ated NF-B activation, further promoting apoptosis.
usually trigger apoptosis in TNFR-bearing cells. How-
Thus, TRAF2 may provide an additional switch between
ever, general inhibition of transcription or translation or
inflammation and cell death downstream of TNF.
selective blockade of the IKK/NF-B pathway uncovers
TNF’s proapoptotic capacity (Chen and Goeddel, 2002;
JNK as a Regulator of Apoptosis
Karin and Lin, 2002). Although mice with TNF or TNFR
JNK1, 2, and 3 (also known as stress-activated protein
gene knockouts develop normally, mice deficient in
kinases) form a subgroup of the MAPK superfamily that
NF-B signaling die in utero from TNF-dependent apo-
is activated by cell stressors such as ultraviolet (UV)
ptosis of liver cells. By activating NF-B, TNF induces a
radiation and by proinflammatory cytokines such as TNF
number of antiapoptotic genes, including c-FLIP, cIAP1,
and interleukin-1 (reviewed in Weston and Davis, 2002;
cIAP2, A1, A20, TRAF1, and TRAF2.
Shaulian and Karin, 2002; Lin, 2003). JNK phosphory-
lates specific subunits, namely c-Jun, JunB, JunD, and
ATF-2, of the AP-1 transcription factor, turning on genesTNFR1 and 2 as Mediators
of TNF-Induced Apoptosis that control diverse cellular functions including prolifera-
tion, differentiation, and apoptosis. JNK’s precise roleChallenge of mice with bacterial lipopolysaccharide
(LPS) together with the liver-specific transcription inhibi- in apoptosis remains controversial since it appears to
have conflicting effects depending on the species, typetor D-galactosamine (GalN) stimulates systemic release
of sTNF, which induces hepatocyte apoptosis and liver of cell, or nature of death stimulus.
Drosophila melanogaster has counterparts to the ma-failure. TNFR1-deficient mice are resistant to this effect,
while TNFR2 knockouts are sensitive. Thus, sTNF sig- jor components of the mammalian JNK cascade as well
as orthologs of many mammalian cell death genes. Innals hepatocyte apoptosis mainly through TNFR1. In-
deed, TNF-deficient mice expressing a transgenic, non- Drosophila, apoptosis during embryonic patterning of
the wing, eye, and gut requires the fly’s JNK orthologcleavable mTNF mutant are largely resistant to LPS/
GalN challenge. Unlike LPS, the T cell stimulator Conca- DJNK (also called Basket) (Kockel et al., 2001; Moreno
et al., 2002; and references therein) (Figure 3). Ectopicnavalin A (ConA) induces mostly mTNF. ConA causes
Figure 2. Model for Apoptosis Control by TNFR1
DJNK activation in the eye imaginal disc causes exces- In mammals, there is evidence both for proapoptotic
and for antiapoptotic JNK activity (Lin, 2003 and refer-sive cell death and eye ablation. This phenotype involves
DJNK-dependent phosphorylation of DJun, which pro- ences therein). Apoptotic death of rat PC12 neuronal
cells deprived of nerve growth factor (NGF) requiresmotes transcription of Hid and Rpr. Hid, Rpr, and an-
other fly gene called Grim encode proteins with a related JNK activity. Knockout of JNK1 and JNK2 in the mouse
suppresses apoptosis in the hindbrain neuroepitheliumsequence motif that enables them to induce cell death
by binding to the fly IAP ortholog DIAP1. This binding at day 9.25, but causes increased apoptosis in the hind-
brain and forebrain at day 10.5. JNK1/JNK2-deficientprevents DIAP1 from blocking activation of the fly cas-
pase DRONC, much like the interaction of mammalian MEFs resist apoptosis induction by UV radiation, protea-
some inhibitors, or genotoxic drugs. Furthermore, inSmac and IAPs (Salvesen and Duckett, 2002). DIAP1
also attenuates DJNK activation by promoting degrada- JNK1 or JNK2 knockout mice, thymocytes are refractory
to death in response to T cell receptor ligation, while intion of DTRAF1, the fly ortholog of mammalian TRAF2.
DJun turns on an additional negative-regulatory feed- JNK3-deficient mice, hippocampal neurons resist apo-
ptosis induction by excitotoxic stress. JNK may promoteback loop by promoting transcription of Puckered (Puc),
a dual specificity phosphatase that inactivates DJNK. mammalian cell apoptosis by engaging the cell-intrinsic
pathway (Figure 4). Whereas wild-type MEFs die in re-Thus, in Drosophila, JNK plays a crucial role in a tightly
regulated signaling pathway that promotes apoptosis. sponse to UV or as a result of ectopic expression of a
Figure 3. Model for Apoptosis Control by Drosophila Eiger
Figure 4. Model for Mammalian Apoptosis Modulation by JNK
constitutively active MKK7-JNK fusion protein, Bax/Bak survival downstream of IL-3, e.g., Akt, remains to be de-
termined.knockout MEFs do not (Lei et al., 2002). Moreover, JNK1/
JNK2-deficient MEFs fail to show Bax activation, cyto-
chrome c release, or death upon UV exposure. JNK- JNK as a Contextual Modulator
of TNF-Induced Apoptosisdependent phosporylation of the BH3 proteins Bim and
Bmf causes their dissociation from dynein and myosin While there is experimental evidence that JNK can inhibit
apoptosis induction by TNF, other data suggest that itmotor complexes, as does UV radiation; this may free
Bim and Bmf to activate apoptosis through Bax and Bak can function as an important positive regulator of TNF-
induced apoptosis. In Drosophila, a homolog of the(Lei and Davis 2003 and references therein). JNK also
promotes Bim transcription through c-Jun (Shaulian and mammalian TNF superfamily called Eiger stimulates
apoptosis through a JNK-dependent mechanism (IgakiKarin, 2002).
In contrast to the proapoptotic activity of JNK in NGF- et al., 2002; Moreno et al., 2002; and references therein)
(Figure 3). Eiger binds to a TNFR-related fly proteindeprived PC12 cells, new data suggest that JNK medi-
ates pro-survival signals downstream of interleukin-3 called Wengen (Kanda et al., 2002). Although Wengen
is required for Eiger activity, it does not possess recog-(IL-3) in human FL5.12 pro-B cells (Yu et al., 2004). IL-3,
a crucial survival factor for FL5.12 cells, stimulates JNK, nizable binding motifs for signaling adaptors such as
TRAFs or TRADD, nor does it display detectable signal-while its withdrawal decreases JNK activity. Inhibition of
JNK with the low molecular weight compound SP600125 ing function. Hence, Wengen may be a ligand binding
subunit of a more complex signaling receptor. Regard-partially attenuates apoptosis of FL5.12 cells after IL-3
withdrawal, whereas expression of a constitutively ac- less, Eigen acts through Wengen to induce DTRAF1-
dependent stimulation of DJNK, thereby inducing apo-tive JNKK2-JNK1 fusion protein promotes this re-
sponse. Thus, JNK contributes to survival signaling in ptosis. As with DJNK, ectopic Eiger expression in the
eye results in excessive apoptosis and eye ablation, andFL5.12 cells downstream of IL-3. Earlier work showed
that IL-3 promotes phosphorylation of the BH3 protein this phenotype can be suppressed by Puc.
Several studies in mice demonstrate an inhibitory roleBad predominantly on serine residues (Ser122 and 136),
and that this phosphorylation mediates the survival sig- for JNK in TNF-induced cell death. TRAF2 knockout
MEFs, which have largely intact NF-B signaling butnal of IL-3 in FL5.12 cells; current work shows that acti-
vated JNK phosphorylates Bad on threonine 201; this reduced JNK stimulation, display increased apoptosis
sensitivity to TNF (Wajant et al., 2003). The JNK inhibitorinhibits Bad’s association with Bcl-X
, probably freeing
up more Bcl-X
to block apoptosis (Yu et al., 2004 and SP600125 enhances TNF-induced apoptosis of MEFs
deficient in the RelA subunit of NF-B. Furthermore,references therein). These findings suggest a novel
mechanism for JNK-dependent inhibition of cell death MEFs deficient in JNK1 and JNK2 show increased sensi-
tivity to TNF killing, and their transfection with JNK1 or(Figure 4). However, the prevalence of Bad phosphoryla-
tion on threonine 201 by JNK, as compared to phosphor- JunD rescues TNF resistance (Lamb et al., 2003 and
references therein). Conversely, JNK2-deficient MEFs,ylation on serines by other kinases that may promote
isolated from a different mouse genetic background N-terminal structure? (4) How does jBid selectively re-
lease mitochondrial Smac without affecting cytochromethan the latter MEFs, display a moderate resistance
rather than sensitization to TNF killing (Dietrich et c? (5) Does ectopic expression of jBid inhibit the associ-
ation of cIAP1 with TRAF2 within complex II? (6) Is jBidal., 2003). Moreover, constitutive JNK1/2 activation in
cells with deficient NF-B signaling sensitizes to TNF- involved in TNF-induced apoptosis in type I as well as
type II cells?induced apoptosis (Lin, 2003). JNK1 and JNK2 knock-
outs are much less sensitive than wild-type mice to Besides acting at the level of direct or indirect post-
transcriptional modification of death signaling mole-ConA-induced liver damage (which involves mTNF);
furthermore, SP600125 blocks TNF-induced death of cules, JNK may modulate apoptosis through AP-1-
dependent gene transcription. This is exemplified byIKK␤-deficient mouse hepatocytes (Maeda et al., 2003).
JNK’s involvement in TNF-induced apoptosis may de- JNK’s regulation of Bim mRNA levels through c-Jun
(Shaulian and Karin, 2002). The level of c-Jun in cells ispend on NF-B. TNF activates JNK transiently; however,
in cells with general inhibition of transcription or specific controlled by ubiquitination and consequent proteaso-
mal degradation, but the underlying enzymatic machin-inhibition of NF-B, TNF leads instead to prolonged JNK
activation (Lin, 2003; Maeda et al., 2003). The NF-B- ery has been elusive. New work uncovers two specific
ubiquitin ligases that support c-Jun destruction in neu-dependent genes A20, GADD45␤, and XIAP attenuate
TNF activation of JNK upon ectopic expression (Lin, ronal and nonneuronal cells, affecting AP-1 activity as
well as apoptosis (Nateri et al., 2004; Wertz et al., 2004;2003). However, A20-deficient MEFs, which display in-
creased sensitivity to TNF-induced death, have normal and references therein) (Figure 4). Numerous mamma-
lian F box proteins function as substrate adaptors forJNK activation (Lee et al., 2000). Similarly, GADD45␤-
deficient MEFs or mouse splenocytes show unaltered ubiquitin ligases of the SCF (Skp1/Cullin/F box protein)
type. A yeast two-hybrid screen of a brain cDNA library,JNK activation or apoptosis induction by TNF plus cyclo-
heximide, although this result has been challenged designed to identify specific binders of phosphorylated
c-Jun, detected the F box protein Fbw7 (Nateri et al.,(Amanullah et al., 2003). While XIAP knockout mice de-
velop normally (Harlin et al., 2001), sensitivity of XIAP- 2004). Fbw7 bound to phosphorylated c-Jun and pro-
moted its ubiquitination, but did not interact with a phos-deficient cells to TNF stimulation of JNK or apoptosis
has yet to be analyzed. Thus, the specific mechanisms phorylation-defective c-Jun mutant or with ATF2. Ec-
topic expression of Fbw7 and c-Jun in 293T humanby which NF-B limits the duration of JNK activation
remain obscure. embryonic kidney cells induced proteasomal degrada-
tion of phosphorylated c-Jun, attenuating AP-1 activity.A recent study suggests that JNK may play an essen-
tial positive role in TNF-induced apoptosis (Deng et al., Conversely, siRNA knockdown of Fbw7 in rat PC12 cells
led to accumulation of phospho-c-Jun, increasing AP-12003). In RelA-deficient MEFs or human HeLa cells ex-
pressing an I-B mutant that blocks NF-B, small in- activity. Importantly, Fbw7 depletion in PC12 cells ele-
vated basal apoptosis levels and substantially aug-terfering RNA (siRNA) knockdown of MKK7 prevented
TNF-induced caspase-8 processing and cell death. TNF mented apoptosis induced by NGF deprivation. Expres-
sion of the JNK-inhibitory scaffolding protein JIP-1stimulated an MKK7-dependent (and presumably, JNK-
dependent) cleavage of the BH3 protein Bid. MKK7/ reversed these proapoptotic effects, indicating that they
are required for JNK activity. Fbw7 depletion also aug-JNK-mediated Bid processing generated a unique prod-
uct (termed jBid), distinct from the previously identified mented apoptosis of mouse primary cerebellar neurons
in a c-Jun-dependent fashion. Thus, in neuronal cells,product of Bid cleavage by caspase-8. A Bid deletion
mutant similar to jBid in size translocated to the mito- modulation of c-Jun levels by the SCF
complex appears to provide an important mechanismchondria, where it selectively triggered the release of
Smac, but not of cytochrome c. SiRNA knockdown of for controlling JNK-dependent apoptosis (Nateri et al.,
2004).Smac prevented TNF from inducing caspase-8 pro-
cessing and apoptosis, while transfection of activated Different ubiquitin ligase complexes may control
c-Jun degradation in other tissues. Indeed, pulldownSmac augmented these TNF effects, suggesting that
Smac acts upstream of caspase-8 activation. A Smac- experiments with hDET1 (the human homolog of de-
Etiolated 1, an Arabidopsis thaliana protein that regu-based peptide, known from other studies to block IAP
interactions with caspases, inhibited association of lates plant photomorphogenesis), identified a distinct
ubiquitin ligase complex that regulates c-Jun ubiquitina-cIAP1 and TRAF2. These data suggest a model in which
TNF activates caspase-8 by relieving it from cIAP inhibi- tion in 293T epithelial cells and in U2OS osteosarcoma
cells (Wertz et al., 2004) (Figure 4). hDET1 binds totion through MKK7/JNK-dependent jBid generation and
consequent Smac release (Figure 2). However, despite hCOP1, a human homolog of Arabidopsis Constitutively
Photomorphogenic-1, together forming a ubiquitin li-evidence for inhibitory association of cIAPs with cas-
pases-3, -7, or -9, support for interaction with caspase-8 gase substrate adaptor that brings c-Jun into contact
with a ubiquitin ligase complex that consists of three(or 10) is lacking (Salvesen and Puckett, 2002). More-
over, overexpression of cIAP1 and 2 in HT1080 cells proteins: DNA damage binding protein 1 (DDB1), Cullin
4A (CUL4A), and Regulator of Cullins-1 (ROC1). Likeexpressing mutant I-B did not protect against TNF kill-
ing (Wajant et al., 2003). So it remains unclear how jBid Fbw7, hCOP1 possesses WD40 repeats; however, nei-
ther hDET1 nor hCOP1 contains an F box, suggestingpromotes caspase-8 activation through Smac. The jBid
model raises several other intriguing questions: (1) How that an unidentified “X box” mediates binding of the
hDET1/hCOP1 heterodimer to the rest of the complexdoes MKK7/JNK activation generate jBid? Is transcrip-
tion required? (2) Do inhibitors of caspases or other (Figure 4). This ligase complex, termed DCX
catalyzed ubiquitination of c-Jun in vitro and in 293Tproteases block jBid generation? (3) What is jBid’s
and U2OS cells. Ectopic coexpression of hCOP1 and doubtedly shed new light on the role of TNF-induced
apoptosis in health and disease.hDET1 led to proteasome-mediated c-Jun degradation,
and siRNA knockdown of hDET1 led to c-Jun accumula-
tion, increasing AP-1 activity and apoptosis. Thus, in
nonneuronal cells, modulation of c-Jun levels by the
We thank Chris Clark for editing this manuscript.
complex provides an important mecha-
nism for controlling JNK-dependent apoptosis, similar
to the action of SCF
in neuronal cells. Given the ex-
tensive diversity of the ubiquitin-proteasomal system, it
Amanullah, A., Azam, N., Balliet, A., Hollander, C., Hoffman, B.,
is likely that multiple ubiquitin ligases regulate JNK-
Fornace, A., and Liebermann, D. (2003). Cell signalling: cell survival
and a Gadd45-factor deficiency. Nature 424, 741.
dependent apoptosis in various cell types at the level
Ashkenazi, A. (2002). Targeting death and decoy receptors of the
of c-Jun and perhaps of other AP-1 subunits. These new
tumour-necrosis factor superfamily. Nat. Rev. Cancer 2, 420–430.
findings raise the intriguing, as yet untested hypothesis
Bouwmeester, T., Bauch, A., Ruffner, H., Angrand, P.O., Bergamini,
that ubiquitin ligase activity against c-Jun may regulate
G., Croughton, K., Cruciat, C., Eberhard, D., Gagneur, J., Ghidelli,
JNK’s contribution to TNF-induced apoptosis. If so, then
S., et al. (2004). A physical and functional map of the human TNF-a/
a logical question that follows is whether NF-B uses
NF-kB signal transduction pathway. Nat. Cell Biol. 6, 97–105. Pub-
this mechanism to exert additional inhibitory control
lished online January 25, 2004.
over TNF-induced cell death, perhaps by promoting the
Chen, G., and Goeddel, D.V. (2002). TNF-R1 signaling: a beautiful
expression of specific ubiquitin ligase components.
pathway. Science 296, 1634–1635.
Cory, S., and Adams, J.M. (2002). The Bcl2 family: regulators of the
cellular life-or-death switch. Nat. Rev. Cancer 2, 647–656.
TNF’s primary role is to stimulate inflammatory cells to
Deng, Y., Ren, X., Yang, L., Lin, Y., and Wu, X. (2003). A JNK-
dependent pathway is required for TNFalpha-induced apoptosis.
fight infection, whereas the function of its proapoptotic
Cell 115, 61–70.
capability remains mysterious. Nonetheless, it makes
Dietrich, N., Thastrup, J., Holmberg, C., Gyrd-Hansen, M., Fehren-
sense that the main conduit for TNF’s proinflammatory
bacher, N., Lademann, U., Lerdrup, M., Herdegen, T., Jaattela, M.,
action, NF-B, strongly inhibits TNF’s induction of apo-
and Kallunki, T. (2003). JNK2 mediates TNF-induced cell death in
ptosis. TNF turns on several antiapoptotic genes
mouse embryonic fibroblasts via regulation of both caspase and
through NF-B. Notable amongst those is c-FLIP, which
cathepsin protease pathways. Cell Death Differ., in press. Published
blocks apical caspase activation in a cytosolic DISC
online December 12, 2003.
that forms downstream of TNFR1. NF-B stimulation
Harlin, H., Reffey, S.B., Duckett, C.S., Lindsten, T., and Thompson,
also limits the duration of TNF-induced JNK activity;
C.B. (2001). Characterization of XIAP-deficient mice. Mol. Cell. Biol.
conversely, NF-B inhibition is associated with pro-
longed TNF-induced JNK activation and apoptosis. How
Igaki, T., Kanda, H., Yamamoto-Goto, Y., Kanuka, H., Kuranaga, E.,
Aigaki, T., and Miura, M. (2002). Eiger, a TNF superfamily ligand that
NF-B controls JNK activity is not yet clear, though
triggers the Drosophila JNK pathway. EMBO J. 21, 3009–3018.
several NF-B-induced genes have been implicated.
Kanda, H., Igaki, T., Kanuka, H., Yagi, T., and Miura, M. (2002).
New evidence supports a positive, if not essential,
Wengen, a member of the tumor necrosis factor receptor superfam-
role of JNK in TNF-induced apoptosis. In Drosophila,
ily, is required for Eigen signaling. J. Biol. Chem. 277, 28372–28375.
the TNF superfamily homolog Eiger requires DJNK to
Karin, M., and Lin, A. (2002). NF-kappaB at the crossroads of life
induce cell death. However, Eiger engages the DRONC
and death. Nat. Immunol. 3, 221–227.
caspase independently of DFADD and the Drosophila
Kockel, L., Homsy, J.G., and Bohmann, D. (2001). Drosophila AP-1:
caspase-8 homolog DREDD, which in the fly regulate
lessons from an invertebrate. Oncogene 20, 2347–2364.
inflammation rather than apoptosis. In contrast, mam-
Lamb, J.A., Ventura, J.J., Hess, P., Flavell, R.A., and Davis, R.J.
malian TNF requires TRADD, FADD, and caspase-8 for
(2003). JunD mediates survival signaling by the JNK signal transduc-
apoptosis induction, and JNK may modulate this by
tion pathway. Mol. Cell 11, 1479–1489.
promoting caspase-8 activation. In cells with blocked
Lee, E.G., Boone, D.L., Chai, S., Libby, S.L., Chien, M., Lodolce,
NF-B, JNK may trigger an unusual path to caspase-8
J.P., and Ma, A. (2000). Failure to regulate TNF-induced NF-kappaB
activation, involving a novel modification of Bid. Future
and cell death responses in A20-deficient mice. Science 289, 2350–
work should define biochemically how JNK supports
Bid processing and caspase-8 activation.
Lei, K., and Davis, R.J. (2003). JNK phosphorylation of Bim-related
members of the Bcl2 family induces Bax-dependent apoptosis.
Both in neuronal and nonneuronal cells, newly identi-
Proc. Natl. Acad. Sci. USA 100, 2432–2437.
fied ubiquitin ligase complexes that support proteaso-
Lei, K., Nimmua, A., Zong, W.X., Kennedy, N.J., Flavell, R.A., Thomp-
mal degradation of c-Jun reveal a novel mechanism for
son, C.B., Bar-Sagi, D., and Davis, R.J. (2002). The Bax subfamily
control of JNK-associated apoptosis. It will be interest-
of Bcl2-related proteins is essential for apoptotic signal transduction
ing to examine whether this mechanism also modulates
by c-Jun NH(2) terminal kinase. Mol. Cell. Biol. 22, 4929–4942.
JNK’s contribution to TNF-induced apoptosis, and if
Lin, A. (2003). Activation of the JNK signaling pathway: breaking the
so, whether NF-B affects the transcription of specific
brake on apoptosis. Bioessays 25, 17–24.
ubiquitin ligase components. The broader question of
Locksley, R.M., Killeen, N., and Lenardo, M.J. (2001). The TNF and
how cells integrate NF-B-controlled determinants with
TNF receptor superfamilies: integrating mammalian biology. Cell
other signals downstream of TNF is ripe for the emerging
field of systems biology. Indeed, the first example for
Maeda, S., Chang, L., Li, Z.W., Luo, J.L., Leffert, H., and Karin, M.
such exploration of TNF/NF-B signaling has just been
(2003). IKK␤ is required for prevention of apoptosis mediated by
cell-bound but not by circulating TNFalpha. Immunity 19, 725–737.
reported (Bouwmeester et al., 2004). Understanding how
cells choose life or death after TNF stimulation will un-
Micheau, O., and Tschopp, J. (2003). Induction of TNF receptor
I-mediated apoptosis via two sequential signaling complexes. Cell
Moreno, E., Yan, M., and Basler, K. (2002). Evolution of TNF signaling
mechanisms: JNK-dependent apoptosis triggered by Eiger, the Dro-
sophila homolog of the TNF superfamily. Curr. Biol. 12, 1263–1268.
Nateri, A.S., Riera-Sans, L., Da Costa, C., and Behrens, A. (2004).
The ubiquitin ligase SCF
antagonizes apoptotic JNK signaling.
Science, in press. Published online January 22, 2004. 10.1126/
Peter, M.E., and Krammer, P.H. (2003). The CD95(APO-1/Fas) DISC
and beyond. Cell Death Differ. 10, 26–35.
Puthalakath, H., and Strasser, A. (2002). Keeping killers on a tight
leash: transcriptional and post-transcriptional control of the pro-
apoptotic activity of BH3-only proteins. Cell Death Differ. 9, 505–512.
Salvesen, G.S., and Duckett, C.S. (2002). IAP proteins: blocking the
road to death’s door. Nat. Rev. Mol. Cell Biol. 3, 401–410.
Shaulian, E., and Karin, M. (2002). AP-1 as a regulator of cell life
and death. Nat. Cell Biol. 4, E131–E136.
Thome, M., and Tschopp, J. (2001). Regulation of lymphocyte prolif-
eration and death by FLIP. Nat. Rev. Cancer 1, 50–58.
Wajant, H., Pfizenmaier, K., and Scheurich, P. (2003). Tumor necrosis
factor signaling. Cell Death Differ. 10, 45–65.
Wang, C.Y., Mayo, M.W., Korneluk, R.G., Goeddel, D.V., and
Baldwin, A.S., Jr. (1998). NF-kB antiapoptosisi: induction of TRAF1
and TRAF2 and c-IAP1 and c-IAP2 to suppress caspase-8 activa-
tion. Science 281, 1680–1683.
Wertz, I.E., O’Rourke, K.M., Zhang, Z., Dornan, D., Arnott, D., De-
shaies, R.J., and Dixit, V.M. (2004). Human de-etiolated-1 regulates
c-Jun by assembling a CUL4A ubiquitin ligase. Science, in press.
Published online January 22, 2004. 10.1126/science.1093549.
Weston, C.R., and Davis, R.J. (2002). The JNK signal transduction
pathway. Curr. Opin. Genes Dev. 12, 14–21.
Yu, C., Minemoto, Y., Zhang, J., Liu, J., Tang, F., Bui, T.N., Xiang,
J., and Lin, A. (2004). JNK suppresses apoptosis via phosphorylation
of the proapoptotic Bcl-2 family protein Bad. Mol. Cell 13, 1–20.