NEMO and RIP1 Control Cell Fate in
Response to Extensive DNA Damage
via TNF-a Feedforward Signaling
Sharon Biton1and Avi Ashkenazi1,*
1Department of Molecular Oncology, Genentech, 1 DNA Way, South San Francisco, CA 94080 USA
Upon DNA damage, ataxia telangiectasia mutated
(ATM) kinase triggers multiple events to promote
cell survival and facilitate repair. If damage is exces-
sive, ATM stimulates cytokine secretion to alert
neighboring cells and apoptosis to eliminate the
afflicted cell. ATM augments cell survival by acti-
vating nuclear factor (NF)-kB; however, how ATM
induces cytokine production and apoptosis remains
elusive. Here we uncover a p53-independent mech-
anism that transmits ATM-driven cytokine and
caspase signals upon strong genotoxic damage.
Extensive DNA lesions stimulated two sequential
NF-kB activation phases,
NEMO/IKK-g: The first phase induced TNF-a-
TNFR1 feedforward signaling, promoting the second
kinase triggered JNK3/MAPK10-dependent inter-
leukin-8 secretion and FADD-mediated proapoptotic
caspase-8 activation. Thus, in the context of exces-
sive DNA damage, ATM employs NEMO and RIP1
kinase through autocrine TNF-a signaling to switch
on cytokine production and caspase activation.
These results shed light on cell-fate regulation by
Metazoan cells are capable of maintaining genomic integrity
despite frequent exposure to exogenous or endogenous
genotoxic insults. This capacity is mediated by a specialized
DNA-damage response (DDR), which is essential for tissue
homeostasis and avoidance of malignancy. The DDR involves
a network of sensing and signaling components that detect
DNA lesions and orchestrate their repair (Harper and Elledge,
2007; Jackson and Bartek, 2009; Kastan and Bartek, 2004;
O’Driscoll and Jeggo, 2006; Shiloh, 2003, 2006). The DDR also
engages signaling pathways to arrest cell growth and promote
survival during repair or, alternatively, if lesions are excessive,
to alarm neighboring cells and eliminate the damaged cell by
Cancer treatment often relies on genotoxic strategies such as
ionizing radiation and certain types of chemotherapy. Many
treatments induce a specific type of DNA lesion, characterized
by double-strand breaks (DSBs). The primary transducer of the
response to DSBs is the nuclear kinase ataxia telangiectasia
mutated (ATM) (Kastan and Bartek, 2004; O’Driscoll and Jeggo,
2006; Shiloh, 2003, 2006). Upon detection of DSBs, ATM is
rapidly activated to phosphorylate multiple target proteins on
serine or threonine residues followed by glutamine (SQ/TQ
motif). ATM thereby mobilizes an array of events, facilitating
points, and activation of prosurvival and stress-response
pathways. Depending on the extent and duration of DNA
damage, ATM can also activate cytokine secretion and pro-
grammed cell death (Fumagalli and d’Adda di Fagagna, 2009;
Harper and Elledge, 2007; Kastan and Bartek, 2004; O’Driscoll
and Jeggo, 2006; Rashi-Elkeles et al., 2006; Shiloh, 2003, 2006).
pathway, which regulates genes involved in diverse biological
functions, including inhibition of apoptosis and promotion of
cell survival (Criswell et al., 2003; Huang et al., 2003; Li et al.,
2001; Piret et al., 1999). ATM-induced NF-kB signaling is impli-
cated not only in permitting normal cells to repair DNA but also
in enabling malignant cells to resist genotoxic therapy (Janssens
and Tschopp, 2006). The mechanism underlying NF-kB activa-
tion by ATM shares some features with the mode of NF-kB
stimulation by the proinflammatory ligand tumor necrosis factor
a (TNF-a). Upon binding to its type 1 cell-surface receptor
(TNFR1), TNF-a assembles a proximal signaling complex, con-
sisting of several components (Chen and Goeddel, 2002;
Hayden and Ghosh, 2008; Karin, 2006; Karin and Lin, 2002; Wil-
son et al., 2009). These include the adaptor, TNFR-associated
death domain (TRADD); the DD-containing serine/threonine
(S/T) kinase, receptor-interacting protein 1 (RIP1); and the E3
ubiquitin ligases, TNF receptor-associated factor 2 (TRAF2)
and cellular inhibitor of apoptosis protein (cIAP) 1 and 2.
TNFR1 triggers cIAP-mediated, K63-linked polyubiquitylation
of RIP1, promoting RIP1 interaction with TAK1 kinase and
TAK1-binding proteins 1 and 2. TAK1 stimulates inhibitor of kB
(IkB) kinase (IKK), which consists of the catalytic subunits IKK-a
and b and the regulatory subunit IKK-g (NEMO). A universal
requirement of RIP1 for TNF-a-induced NF-kB activation has
92 Cell 145, 92–103, April 1, 2011 ª2011 Elsevier Inc.
been challenged (Wong et al., 2010). Upon TNFR1 activation,
IKK mediates IkB phosphorylation, leading to IkB ubiquitylation
and proteasomal degradation. This allows NF-kB to enter the
nucleus and activate gene transcription.
ATM is essential for NF-kB activation in response to DNA
damage (Li et al., 2001; Piret et al., 1999; Rashi-Elkeles et al.,
2006). Initial studies suggested that DNA lesions induce
ATM-dependent NEMO phosphorylation and subsequent ubiq-
uitylation to activate the cytoplasmic IKK complex (Bartek and
Lukas, 2006; Huang et al., 2003; Janssens and Tschopp, 2006;
Wu et al., 2006). More recent evidence indicates that nuclear
exit of ATM also drives NEMO ubiquitylation, leading to TAK1
phosphorylation and IKK activation (Hinz et al., 2010; Wu et al.,
2010). In contrast to the TNF-a pathway, DDR-driven IkB degra-
dation in mouse embryo fibroblasts did not require TNFR1,
TRAF2, TRAF5, or FADD (Hur et al., 2003). NEMO also forms
a tripartite complex with RIP1 and p53-inducible DD-containing
protein (PIDD), augmenting NEMO ubiquitylation and NF-kB
activation (Janssens et al., 2005). Alternatively, PIDD binds
caspase-2 via the adaptor RIP1-associated protein with a DD,
forming a ‘‘PIDDosome’’ that induces apoptosis (Janssens and
Tschopp, 2006; Tinel and Tschopp, 2004).
NEMO associates with RIP1, which is reported to be required
for DDR-induced NF-kB activation in mouse embryo fibroblasts
but not human HEPG2 cells (Hinz et al., 2010; Hur et al., 2003).
RIP1 is implicated not only in controlling NF-kB but also in regu-
lating cell death in response to TNF-a, viral infection, or cIAP
antagonists (Bertrand et al., 2008; Cho et al., 2009; He et al.,
2009; Petersen et al., 2007; Varfolomeev et al., 2007; Wang
et al., 2008). Kinase activity of RIP1 is dispensable for NF-kB
activation (Kelliher et al., 1998; Ting et al., 1996) but required
for induction of necroptotic cell death (Cho et al., 2009; Degterev
et al., 2008; He et al., 2009; Hur et al., 2003). RIP1 ubiquitylation
favors downstream prosurvival signaling via NF-kB over cell
death induction through the adaptor Fas-associated death
domain (FADD) and the apoptosis-initiating protease caspase-8
(C8) (Declercq et al., 2009; Wang et al., 2008).
Despite the critical relevance of the DDR to physiology and
disease, little is known about how cellular life or death decisions
are made upon ATM activation. In the present work we identify
a p53-independent mechanism that selectively triggers ATM-
of extensive DNA damage. Our results provide insight into the
control of cell fate downstream of ATM.
Extensive DNA Damage Induces ATM-Dependent
Interaction of NEMO, RIP1, FADD, and Caspase-8,
Leading to Caspase Activation
Toinvestigate howATM controls cellularoutcome inresponse to
different levels of DNA damage, we treated HeLa cells with the
topoisomerase II inhibitor etoposide and analyzed interactions
between potentially relevant intracellular signaling components.
Etoposide caused significant formation of DSBs, indicated by
phosphorylation of H2AX, BRCA1, SMC-1, and the ATM
substrate KAP1 (Figures S1A and S1B available online and
Figures 1A and 1B) (Shiloh, 2006). Etoposide also induced
NF-kB activation, detectable within 2 hr by immunoblot (IB) anal-
ysis of IkB degradation; furthermore, at high doses and upon
longer incubation, etoposide stimulated phosphorylation of
stress-activated c-Jun N-terminal kinase (JNK) (Figures 1A and
1B). Consistent with previous data (Hur et al., 2003), etoposide
triggered association of NEMO with RIP1 (Figures 1C and 1E).
with apparent ubiquitylation of NEMO (Figure S1C)—a modifica-
tion known to be important for NF-kB activation in response to
DSBs (Huang et al., 2003; Hinz et al., 2010). Surprisingly, at
higher doses and on longer incubation, etoposide also stimu-
lated recruitmentof FADD
(Figures 1C and 1E) and association of RIP1 with FADD and C8
(Figures 1D and 1F). FADD binds RIP1 and C8 via respective
DD and death effector domain (DED) interactions (Wilson et al.,
2009). C8 activation involves self-processing between the p18
and p10 subunits to form a p43 fragment, whereas cleavage
between p18 and the DEDs releases C8 from FADD (Boatright
et al., 2003). The pan caspase inhibitor zVAD-fmk (zVAD)
efficiently blocks the latter but not the former event (unpublished
data). Consistent with this, zVAD augmented association of
FADD and RIP1 with NEMO and with C8 (Figures 1G and 1H);
reciprocal immunoprecipitation (IP) of FADD or RIP1 further
confirmed their interaction (Figure 1I, Figure S1D).
The recruitment of FADD and C8 appeared to be functionally
significant because IP of C8 revealed C8 processing into p43
and a marked elevation in C8 enzymatic activity, specifically in
response to high-dose etoposide (Figure 1H and Figure 2A).
zVAD did not affect H2AX and KAP1 phosphorylation or p43
generation (Figure S1B and Figure 2B), nor did it change the
cellular levels of NEMO, RIP1, FADD, or C8 (Figure S1E).
However, zVAD blocked further C8 processing and consequent
activation of caspase-3 (C3), as indicated by attenuated
cleavage of C3 and its substrate, PARP; in contrast, zVAD
slightly enhanced stimulation of NF-kB and JNK (Figure 2B).
Consistent with its lack of effect on p43 generation, zVAD did
not decrease C8 activity in FADD IPs (Figure S1F). Due to
antibody (Ab) cross-reactivity, we were unable to visualize C8
by IB in NEMO IPs. Nevertheless, C8 activity was readily detect-
able by NEMO IP, and this was blocked by siRNA silencing of
NEMO (Figure 2C); moreover, C8 activity was detected by
RIP1 IP (Figure S1G). C8 depletion by siRNA further verified
specific C8 association with FADD and RIP1 and C8 activation
(Figures S1H and S1I). Together, these results suggest that
high-dose etoposide induces a functional interaction of NEMO,
RIP1, FADD, and C8.
Ionizing radiation, neocarcinostatin, doxorubicin, or campto-
thecin also induced recruitment of FADD and C8 to RIP1, as
well as NF-kB and JNK activation (Figures S2A and S2B),
demonstrating that other DSB inducers stimulate similar events.
Moreover, high-dose etoposide triggered association between
RIP1 and C8 and C8 activation also in MDA231 and MDA435
cells (Figures S2C and S2D), extending these effects to other
cell types. p53 is an important mediator of ATM signals in
response to DSBs (Miyashita et al., 1994; Nakano and Vousden,
2001; Oda et al., 2000). Given that HeLa, MDA231 and MDA435
cells lack functional p53, we turned to RKO cells, which
express wild-type p53, to assess p53’s involvement in these
Cell 145, 92–103, April 1, 2011 ª2011 Elsevier Inc. 93
latter events. Silencing of p53 by siRNA did not significantly
affect JNK activation in response to high-dose etoposide, or
recruitment of FADD and C8 to RIP1, or C8 activation (Figures
S2E–S2G). In contrast, the ATM-specific inhibitor Ku-55933
(Hickson et al., 2004), which blocked KAP1 phosphorylation
and apparent NEMO ubiquitylation (Figure S1C), attenuated
IkB degradation and JNK phosphorylation (Figure 2D). ATM inhi-
bition also suppressed the recruitment of FADD and RIP1 to
NEMO and C8 IPs as well as C8 activation (Figures 2E–2G).
We obtained similar results upon ATM depletion by siRNA
(Figures 2H–2J). Thus, stimulation of NF-kB, JNK, and C8 in
response to high-dose etoposide requires ATM but not p53.
NEMO and RIP1 Operate Upstream of Cytokine
Secretion and C8 Activation
Next, we investigated whether NEMO and RIP1 are required for
ATM-mediated C8activation.Consistentwiththeprevious impli-
cation of NEMO in ATM-mediated NF-kB activation, knockdown
of NEMO by siRNA abolished etoposide-stimulated IkB degra-
dation; moreover, NEMO depletion also inhibited JNK activation
(Figure 3A). Persistent DNA lesions promote cellular secretion of
Figure 1. Extensive DNA Damage Induces Interac-
tion of NEMO, RIP1, FADD, and Caspase-8
HeLa cells were pretreated with zVAD (20 mM) for 4 hr,
followed by treatment with various doses of etoposide
(Eto) for 20 hr (A) or 100 mM etoposide for the indicated
period of time (B). Cell lysates were analyzed by immu-
noblot (IB) with specific antibodies to phosphorylated
KAP1 or JNK, or to IkB.
(C–F) Cells were treated as above and analyzed by
immunoprecipitation (IP) followed by IB as indicated.
(G–I) Cellswere pretreated withDMSO or zVAD(20mM)for
4 hr, followed by 100 mM etoposide for 20 hr, and analyzed
by IP and IB as indicated. Iso: isotype-matched control for
the IP Ab in the etoposide-treated cells. Asterisk: back-
See also Figure S1.
proinflammatory cytokines such as interleukin
(IL)-6 and IL-8 via NF-kB (Fumagalli and
d’Adda di Fagagna, 2009; Linard et al., 2004;
Rodier et al., 2009; Zhou et al., 2001). We
confirmed ATM-dependent secretion of IL-6
and IL-8 in response to high-dose etoposide
(Figures S3A–S3F). NEMO knockdown inhibited
generation of both cytokines (Figures 3B and
3C); furthermore, it blocked recruitment of
FADD and RIP1 to NEMO and C8 IPs as well
as C8 activation (Figures 3D–3F). RIP1 siRNA
also inhibited these latter events (Figures 3G–
3J), with comparable effects in MDA231 cells
(Figures S3G–S3I). By contrast, p53 silencing
in RKO cells did not significantly inhibit IL-8
secretion (Figure S2H), consistent with other
data (Rodier et al., 2009). Hence, NEMO and
secretion and C8 activation in response to
extensive DNA damage. Knockdown of RIP1
or C8 by siRNA similarly attenuated etoposide-induced cell
death (Figures S3J and S3K), confirming that C8 activation is
mediated primarily by RIP1 and contributes significantly to this
NF-kB Stimulation by ATM Controls Both JNK-Mediated
Cytokine Secretion and JNK-Independent C8 Activation
The IKK inhibitor BMS-345541 blocked IkB degradation, as well
as JNK activation and IL-8 secretion, in response to high-dose
etoposide (Figures 4A and 4B). IKK inhibition also attenuated
the recruitment of FADD and RIP1 to NEMO and C8 IPs as well
as C8 activation (Figures 4C–4E). We obtained similar results
by siRNA silencing of the NF-kB subunit p65/RelA (Figures
S4A–S4D). Consistent with these findings, IKK inhibition sup-
pressed cell-death induction by etoposide (Figure S4E). The
JNK inhibitor SP600125 did not prevent etoposide-induced IkB
(Figures 4F and 4G). In contrast, JNK inhibition did not substan-
tially affect association of FADD with NEMO and RIP1, interac-
tion of RIP1 and FADD with C8, or C8 activation (Figures
4H–4J). DepletionofJNK1or JNK2bysiRNAdidnotsignificantly
94 Cell 145, 92–103, April 1, 2011 ª2011 Elsevier Inc.
inhibit IL-8secretion or C8activation (data not shown).However,
knockdown of JNK3/MAPK10 attenuated IL-8 secretion without
significantly affecting IkB degradation, association of RIP1 and
FADD with C8, or C8 activation (Figures S4F–S4I). These results
suggest that NF-kB activation occurs upstream of both JNK
stimulation and FADD recruitment: JNK3 supports IL-8 secre-
tion, whereas FADD independently mediates C8 activation.
RIP1 Kinase Activity Is Required for IL-8 Secretion
as well as C8 Stimulation
To assess whether any of the latter events depends on RIP1
kinase activity, we used the RIP1-specific inhibitor Necrostatin1
(Nec1) (Degterev et al., 2008, and see below). Nec1 did not alter
H2AX or KAP1 phosphorylation in response to high-dose etopo-
side; however, it partially blocked IkB degradation and JNK
phosphorylation (Figure 5A, Figure S5A). Unexpectedly, we
detected two phases of NF-kB activation, evident by an initial
wave of RelA/p65 phosphorylation and IkB degradation at
2–6 hr, followed by a second wave at 12–20 hr (Figures 5B
and 5C). Although the level of activation was comparable during
both phases, pretreatment with Nec1 selectively inhibited the
second but not the first phase (Figures 5B and 5C). Nec1
strongly attenuated etoposide-induced mRNA transcription of
TNF-a and IL-8 at 27 hr but not at 7 hr (Figures S5B and S5C),
consistent with its selective inhibition of the second but not the
first NF-kB phase. These data confirm earlier work in
Figure 2. High-Dose
ATM-Dependent C8 Activation
HeLa cells were treated with various etoposide
doses for 20 hr and subjected to C8 IP followed by
either IB analysis (A, top) or C8 enzymatic activity
assay (A, bottom).
(B) Cells were treated with zVAD (20 mM) or DMSO
for 4 hr, followed by 100 mM etoposide for 20 hr.
Cell lysates were analyzed by IB as indicated.
(C) HeLa cells were transfected with control or
NEMO siRNA for 48 hr, followed by treatment with
100 mM etoposide for 20 hr. Cell lysates were
(D–G) HeLa cells were pretreated with the ATM-
specific inhibitor Ku-55933 (ATMi) or vehicle plus
zVAD for 4 hr, followed by treatment with 100 mM
and/or IB as indicated (D–F)or by C8 IP and activity
(H–J) Cells were transfected with ATM or control
siRNA for 48 hr, followed by treatment with 100 mM
(H) or by C8 IP, followed by IB (I) or by C8 activity
assay (J). Pretreatment with zVAD was included
before etoposide addition.
See also Figure S2.
which IkB was monitored up to 9 hr after
genotoxic exposure (Hur et al., 2003);
however, they reveal a second phase of
NF-kB induction in response to high-
dose etoposide, which uniquely depends
on RIP1 kinase activity.
Nec1 also inhibited etoposide-driven IL-8 secretion (Fig-
ure 5D). In contrast, Nec1 did not block RIP1 association with
NEMO (Figure 5E), consistent with its lack of effect on early
NF-kB activation. However, Nec1 did prevent association of
FADD with NEMO and RIP1, interaction of FADD and RIP1
with C8, or C8 activation (Figures 5E–5G). We observed similar
Nec1 effects in MDA231 and MDA435 cells (Figures S5D–S5F
and data not shown). Hence, RIP1 kinase activity is important
for both IL-8 secretion and C8 stimulation.
Nec1 inhibits RIP1 autophosphorylation (Degterev et al.,
panel of 98 diverse kinases, using the KINOMEscan technology
(Table S1A). As compared to its complete inhibition of RIP1,
Nec1 inhibited two other kinases, namely, PAK1 and PKAca,
by more than 60%. Depletion of PAK1 or PKAca by siRNA
did not inhibit IL-8 secretion or interaction of RIP1 and FADD
with C8 in response to high-dose etoposide (Figures S5G–
S5J). Furthermore, Nec1 bound tightly to RIP1 but not to RIP2
or RIP4 (Table S1B), consistent with other evidence of its
selectivity for RIP1 over RIP2, 3, or 4 (Degterev et al., 2008).
To examine whether RIP1 undergoes phosphorylation in
response to high-dose etoposide, we took advantage of
RIP1’s character as an S/T-kinase by screening a panel of
available anti-phospho-S/T antibodies for ability to detect RIP1
upon IP and IB (see Experimental Procedures). One of the
antibodies (clone 100G7E) specifically revealed robust RIP1
Cell 145, 92–103, April 1, 2011 ª2011 Elsevier Inc. 95
phosphorylation by 22 hr in response to 100 mM but not 10 mM
etoposide (Figures 5H and 5I). Silencing of ATM or NEMO by
siRNA attenuated RIP1 phosphorylation, as did IKK inhibition
(Figures 5J–5L). The anti-phospho-SQ/TQ antibody didnot react
with RIP1 (data not shown), suggesting that RIP1 is not a direct
substrate for ATM. Nec1 suppressed RIP1 phosphorylation
(Figure 5M), whereas knockdown of PAK1 or PKAca did not
(Figures S5K and S5L). Furthermore, ectopic expression of
a kinase-domain-deficient RIP1 mutant (D300), albeit less
abundant than endogenous RIP1 expression, significantly atten-
uated RIP1 phosphorylation (Figure 5N). Consistent with the
Nec1 data, D300 inhibited IL-8 secretion, recruitment of FADD
to RIP1 and NEMO, and C8 activation in response to high-dose
etoposide (Figures S5M–S5P). Importantly, phosphorylated
RIP1 was associated with NEMO and FADD, and vice versa,
FADD was detectable upon RIP1 IP with anti-phospho-S/T
antibody (Figures 5O–5Q). These results suggest that RIP1
etoposide—a modification that depends upon ATM, NEMO,
and IKK and drives both IL-8 secretion and C8 activation.
in responseto high-dose
Upstream of Cytokine Secretion and C8
(A–C) HeLa cells were transfected with control or
NEMO siRNA for 48 hr, followed by treatment with
100 mM etoposide for 20 hr. Cell lysates were
analyzed by IB as indicated (A), and supernatants
were analyzed for IL-6 and IL-8 secretion by ELISA
(B and C).
(D–F) Cells were transfected with NEMO or control
siRNA for 48 hr, followed by etoposide as in
(A)–(C). Cell lysates were analyzed by IP and IB
as indicated (D and E), or by C8 IP and activity
(G–J) Cells were transfected with RIP1 or control
siRNA for 48 hr, followed by etoposide as in (A)–
(C). Cell supernatants were analyzed for IL-8
secretion by ELISA (G), and cell lysates were
analyzed by IP and IB (H and I), or by C8 IP and
activity assay (J). Pretreatment with zVAD was
included before etoposide addition.
Error bars in (B), (C), and (G) represent mean ±
standard deviation (SD) of duplicates. See also
3. NEMOandRIP1 Operate
We were intrigued by the high-molec-
ular-mass RIP1 species induced at high
doses of etoposide (Figures 1C–1H);
these species disappeared upon siRNA
knockdown of RIP1 (Figure 3H), suggest-
ing RIP1 polyubiquitylation. RIP1 IP under
denaturing conditions and IB against
ubiquitin confirmed this possibility (Fig-
ure S6A). Further analysis revealed K63-
rather than K48-linked polyubiquitylation
of RIP1 (Figure S6B). Genetic ablation of
cIAPs or induction of their proteasomal
destruction by Smac-mimetics prevents
TNF-a-induced RIP1 polyubiquitylation,
disrupting NF-kB activation and promoting cell-death signaling
via FADD and C8 (Bertrand et al., 2008; Cho et al., 2009; De-
clercq et al., 2009; He et al., 2009; Petersen et al., 2007; Varfo-
lomeev et al., 2007). Hence, to probe the involvement of RIP1
ubiquitylation in the response to DNA damage, we used the
bivalent IAP antagonist BV6 (Varfolomeev et al., 2007, 2008).
BV6 blocked etoposide-induced RIP1 K63 polyubiquitylation
as well as IkB degradation, but not KAP1 phosphorylation
(Figures S6C–S6E). In the absence of etoposide, BV6 induced
a modest, yet detectable, association of RIP1 and FADD with
C8 and a measurable increase in C8 activation (Figures S6F
and S6G). Upon etoposide treatment, BV6 further enhanced
association of RIP1 and FADD with C8, processing of C8 into
p43, and C8 activation, as well as cell death (Figures S6F–
S6H). Consistent with these data, siRNA knockdown of cIAP2
augmented etoposide-induced association of RIP1, FADD,
and C8, as well as C8 activation (Figures S6I and S6J). Thus,
although cIAP-mediated RIP1 polyubiquitylation supports NF-
kB activation, it opposes the recruitment of FADD and C8 to
RIP1. Nec1 abolished the appearance of ubiquitylated RIP1 in
96 Cell 145, 92–103, April 1, 2011 ª2011 Elsevier Inc.
NEMO IPs (Figure 5E), indicating that RIP1 kinase activation
occurs upstream of RIP1 ubiquitylation.
TNF-a-TNFR1 Feedforward Signaling Mediates
the Induction of IL-8 Secretion and C8 Activation
We reasoned that initial NF-kB activation downstream of ATM
might turn on positive feedforward signals that could re-engage
NEMO and RIP1 kinase to drive IL-8 secretion and C8 activation
in the context of persistent DNA damage. TNF-a was a plausible
candidate for mediating such signals because it can be induced
via NF-kB and is also capable of activating NF-kB and of
engaging RIP1 via TNFR1. Indeed, exposure to high doses of
etoposide or prolonged incubation with the drug induced signif-
icant accumulation over time of secreted TNF-a in cell superna-
tants (Figures 6A and 6B). We observed initial upregulation of
Figure 4. Role of the IKK and JNK Pathways
in the Response to High-Dose Etoposide
(A–E) HeLa cells were pretreated with the IKK-
specific inhibitor BMS-345541 (IKKi, 5 mM) or
vehicle for 4 hr, followed by treatment with 100 mM
etoposide for 20 hr. Cell lysates were analyzed by
IP and IBas indicated (A,C,and D),orby C8IP and
activity assay (E); cell supernatants were analyzed
for IL-8 secretion by ELISA (B).
(F–J) Cells were prereated with the JNK-specific
followed by treatment with 100 mM etoposide for
20 hr. Cell lysates were analyzed by IP and IB as
indicated (F, H, and I), or by C8 IP and activity
assay (J); cell supernatants were analyzed for IL-8
secretion by ELISA (G). Pretreatment with zVAD
was included before etoposide addition.
Error bars in (B) and (G) represent mean ± SD of
duplicates. See also Figure S4.
TNF-a mRNA (62-fold) by 6 hr, followed
by massive further induction (6900-fold)
at 26 hr(Figure 6C). IKK inhibition blocked
TNF-a secretion (Figure 6D), indicating
that this event is driven primarily by
NF-kB. JNK inhibition partially attenuated
TNFa production (data not shown), sug-
gesting that JNK signaling may coamplify
To assess the functional significance of
TNF-a secretion, we tested the effect of
a soluble TNFR1-Fc decoy receptor.
TNFR1-Fc inhibited the late (20 hr) but
not the early (2 hr) phase of etoposide-
induced NF-kB activation, as well as
TNFR1-Fc also suppressed RIP1 phos-
phorylation, RIP1 polyubiquitylation, and
IL-8 secretion (Figures 6F–6H). Although
not preventing RIP1 binding to NEMO at
2 hr, TNFR1-Fc partially attenuated this
interaction at 20 hr, blocking FADD
tion of C8 (Figures 6I–6K). We obtained similar results by siRNA
depletion of TNFR1 (Figures 6L–6P) or with anti-TNF-a neutral-
izing antibody (Figures S6K and S6L). Consistent with these
findings, TNFR1-Fc significantly inhibited induction of cell death
by high-dose etoposide (Figure S6M). Hence, an autocrine
TNF-a-TNFR1 feedforward loop supports the second wave of
NF-kB stimulation and re-engages RIP1 kinase in conjunction
with NEMO to trigger IL-8 secretion and C8 activation.
Metazoan homeostasis requires aneffective ability to respondto
the frequent DNA damage inflicted by environmental or endoge-
nous insults. Whereas successful resolution of DNA lesions
enables cells to continue to perform their normal functions and
Cell 145, 92–103, April 1, 2011 ª2011 Elsevier Inc. 97
Figure 5. RIP1 Kinase Activity Is Required for IL-8 Secretion and C8 Activation
(A) HeLa cells were pretreated with the RIP1 kinase-specific inhibitor Nec1 (30 mM) or vehicle for 4 hr, followed by treatment with 100 mM etoposide for 20 hr. Cell
lysates were analyzed by IB as indicated.
(B and C) Cells were pretreated with Nec1 (30 mM) or vehicle for 4 hr, followed by treatment with 100 mM etoposide for the indicated time, and cell lysates were
analyzed by IB with Ab against phosphorylated p65 or IkB.
(D–G) Cells were treated as in (A) and cell supernatants were analyzed for IL-8 secretion by ELISA (D), or cell lysates were analyzed by IP and IB as indicated
(E and F) or by C8 IP and activity assay (G).
(H) Cells were treated with 100 mM etoposide for 20 hr, and cell lysates were analyzed by RIP1 IP followed by IB using phospho-S/T Ab (clone 100G7E, see
(I) Cells were treated with indicated etoposide doses for 2 hr or 22 hr, and cell lysates were analyzed as in (H).
(J and K) Cells were transfected with control, ATM (J), or NEMO siRNA (K) for 48 hr, followed by treatment with 100 mM etoposide for 20 hr, and analyzed as in (H).
(L and M) Cells were pretreated with IKKi (5 mM) versus vehicle (L) or Nec1 (30 mM) versus vehicle (M) for 4 hr, followed by treatment with 100 mM etoposide for
20 hr, and analyzed as in (H).
(N)CellsweretransfectedwiththeD300RIP1deletion constructoremptyvector for24hr,followed bytreatmentwith100mMetoposide for20hr,andanalyzedas
(O–Q) Cells were treated with 100 mM etoposide for 20 hr and analyzed by NEMO (O) or FADD (P) IP followed by IB using phospho-S/T Ab. Vice versa, cell lysates
were analyzed by IP using phospho-S/T Ab followed by IB with RIP1 and FADD Ab (Q). Pretreatment with zVAD was included before etoposide addition.
Error bars in (D) represent mean ± SD of duplicates. See also Figure S5 and Table S1.
98 Cell 145, 92–103, April 1, 2011 ª2011 Elsevier Inc.
Figure 6. Autocrine TNF-a-TNFR1 Feedforward Signaling Mediates Induction of IL-8 Secretion and C8 Activation
(A and B) HeLa cells were treated with various doses of etoposide for 20 hr or with 100 mM etoposide for different periods of time. Cell supernatants were
collected, concentrated 5-fold, and analyzed by TNF-a ELISA.
(C) HeLa cells were treated for 6 hr or 26 hr with 100 mM etoposide and the mRNA induction of TNF-a was measured by TaqMan Human NF-kB Pathway Array
(see Experimental Procedures).
(D) Cells were pretreated with IKKi (5 mM) or vehicle for 4 hr, followed by treatment and analysis as in (A).
(E,G, and I)Cells weretreated with100mMetoposide forthe indicated timeinthe presence orabsenceof TNFR1-Fc (100mg/ml).Celllysates wereanalyzed by IB
as indicated (E), or by RIP1 IP under denaturing conditions (G), or by NEMO IP (I), followed by IB as indicated.
S/T Ab (clone 100G7E).
ELISA (H). Cell lysates were analyzed by IP and IB as indicated (J) or by C8 IP and activity assay (K).
(L–P)HeLa cellsweretransfected withcontrol orTNFR1siRNAfor48hrandtreatedwith100mMetoposide for20hrorasindicated. Celllysateswereanalyzed by
IB for KAP1 phosphorylation, IkB degradation, and JNK phosphorylation (L). Cell supernatants were analyzed by IL-8 ELISA (M). Cell lysates were analyzed by IP
and IB as indicated (N and O) or by C8 IP and activity assay (P). Pretreatment with zVAD was included before etoposide addition.
Error bars in (C), (D), (H), and (M) represent mean ± SD of duplicates. See also Figure S6.
Cell 145, 92–103, April 1, 2011 ª2011 Elsevier Inc. 99
hence the organism to thrive, excessive damage can lead to
genomic instability syndromes, malignant cell transformation,
and cancer (Harper and Elledge, 2007; Jackson and Bartek,
2009; Kastan and Bartek, 2004; O’Driscoll and Jeggo, 2006;
Shiloh, 2003, 2006). Mechanisms to eliminate cells harboring
extensive DNA lesions have evolved alongside DNA repair func-
tions. Thus, defining the molecular events that orchestrate an
effective DDR is important not only for understanding homeo-
stasis but also for developing effective approaches to treat
diseases associated with DNA damage.
to arrest cell-cycle progression and promote cell survival,
enhancing the cell’s ability to repair the damage. Alternatively,
ATM may direct the cell toward an apoptotic fate if the damage
to DNA is excessive or persistent. ATM activates p53, which
controls both cell-cycle checkpoints and apoptosis via tran-
scriptional as well as nontranscriptional mechanisms (Miyashita
et al., 1994; Nakano and Vousden, 2001; Oda et al., 2000).
However, little is known about how ATM controls cell fate in
coordination with the extent of genotoxic damage. Key among
the signaling cascades engaged by ATM is the NF-kB pathway,
activated through nuclear as well as cytoplasmic events (Bartek
and Lukas, 2006; Hinz et al., 2010; Huang et al., 2003; Hur et al.,
2003; Janssens and Tschopp, 2006; Wu et al., 2006, 2010).
NF-kBisgenerallyviewed asaprosurvivalfactor; however,there
certain circumstances, through repression of prosurvival genes
or upregulation of proapoptotic factors (Baud and Karin, 2009;
Hayden and Ghosh, 2008; Karin, 2006; Papa et al., 2006).
Indeed, activation of NF-kB and AP-1 in response to genotoxic
stress induces apoptosis of T lymphocytes via upregulation of
Fas/CD95 ligand (Kasibhatla et al., 1998). NF-kB also mediates
proinflammatory cytokine secretion in the context of persistent
DNA lesions—a process dubbed senescence-associated secre-
tory phenotype (Fumagalli and d’Adda di Fagagna, 2009; Rodier
et al., 2009). However, the underlying molecular interactions
have not been well defined. Our study reveals that NF-kB activa-
tion by ATM sets up a p53-independent signaling mechanism
that enables the cell to switch on a new set of events in the
face of strong genotoxic stress.
To model extensive DNA damage, we used concentrations of
the DSB-inducing agent etoposide that mimic peak plasma
levels in cancer patients receiving high-dose chemotherapy
(Hande et al., 1984). Phosphorylation of H2AX, KAP1, and other
DDR markers confirmed extended induction of DSBs and ATM
activation. We discovered that a NEMO-RIP1 complex, which
is induced also in response to low levels of DNA damage,
continues to assemble in the presence of extensive DSBs.
Under the latter circumstance, NEMO-RIP1 interaction triggers
two important new functions: (1) stimulation of JNK phosphory-
lation, leading to IL-8 production; and (2) recruitment of FADD
and C8, leading to C8 activation. Knockdown of NEMO or
RIP1 blocked both of these functions, as did pharmacologic
inhibition of IKK or siRNA silencing of RelA/p65, indicating
that not only NEMO and RIP1 but also NF-kB-dependent
signals are required to drive these events. Further studies
demonstrated that JNK3 is critical for IL-8 induction but not
for C8 activation. These results provide direct evidence sup-
porting the previous notion that NF-kB mediates cytokine
secretion in response to persistent DSBs (Fumagalli and
d’Adda di Fagagna, 2009) and reveal that this function depends
nonredundantly on JNK3.
interaction in response to strong DNA damage, we examined the
high-dose etoposide induces two consecutive phases of NF-kB
activation. Consistently with other work (Hur et al., 2003), the
selective RIP1 kinase inhibitor Nec1 did not block the initial
phase, indicating that RIP1 kinase activity is dispensable for
this function. However, Nec1 inhibited the second phase,
demonstrating that RIP1 kinase activity supports this additional
wave of NF-kB stimulation. Nec1 also inhibited JNK phosphory-
lation and IL-8 secretion, as well as recruitment of FADD and C8
to RIP1 and C8 activation, placing RIP1 kinase activity upstream
of these events. Importantly, RIP1 phosphorylation, which
occurred in an ATM- and NEMO-dependent manner, was
detectable only in response to strong DNA damage, in conjunc-
tion with the second NF-kB activation phase. Moreover, phos-
phorylated RIP1 associated with NEMO and FADD, supporting
the involvement of RIP1 kinase in C8 activation. RIP1 likely
undergoes autophosphorylation, given that its modification
was inhibited by Nec1 and by an ectopically expressed RIP1
mutant lacking the kinase domain. We have been unable to
map the RIP1 phosphorylation site(s) by mass spectrometry,
due to difficulty with purifying the protein from overexpression
with NEMO at late time points after induction of strong DNA
damage. Inhibition of this RIP1 modification with BV6 or by
siRNA knockdown of cIAP2 augmented the recruitment of
FADD and C8 to RIP1 and C8 activation, suggesting that RIP1
ubiquitylation does not favor the latter events. This finding is
consistent with other evidence that RIP1 ubiquitylation interferes
with recruitment of FADD and C8 in response to TNF-a (Wang
et al., 2008). On the other hand, BV6 also inhibited NF-kB activa-
tion, in keeping with the role of K63-linked RIP1 ubiquitylation in
NF-kB activation by TNF-a (Wertz and Dixit, 2008).
RIP1 phosphorylation was blocked by NF-kB inhibition and
occurred during the second wave of NF-kB activity. This promp-
ted us to investigate whether the initial NEMO-RIP1 interaction
induces a positive reinforcement loop to re-engage RIP1 kinase.
Two lines of evidence indicated that such a mechanism is medi-
ated by autocrine TNF-a-TNFR1 feedforward signaling: First,
high-dose etoposide induced NF-kB-dependent TNF-a produc-
tion, which was detectable by 6 hr yet greatly accelerated over
time. JNK inhibition partially attenuated TNF-a induction, sug-
gesting that JNK signaling coamplifies TNF-a upregulation.
Second, whereas intervention with TNF-a signaling spared initial
NEMO-RIP1 association and early NF-kB activation, it disrupted
further NEMO-RIP1 binding, RIP1 phosphorylation, RIP1 ubiqui-
tylation, and subsequent NF-kB stimulation. TNF-a blockade
also inhibited the downstream events linked to IL-8 secretion,
C8 activation, and cell death. A requirement for sufficient
TNF-a accumulation to drive the TNFR1-mediated feedforward
loop likely explains the observed lag in RIP1 phosphorylation.
Moreover, the first NF-kB activation phase induced by DNA
lesions is typically slower than NF-kB stimulation by TNF-a
100 Cell 145, 92–103, April 1, 2011 ª2011 Elsevier Inc.
(Janssens and Tschopp, 2006), which may further contribute to
the delay in RIP1 modification.
Taken together with other findings, our studies suggest the
following model for cell-fate regulation in response to DNA
damage (Figure 7). As previously established, in response to
low levels of DSBs, ATM stimulates an early phase of NF-kB
activity in a NEMO-dependent yet TNFR1-independent manner,
promoting transcription of several well-established prosurvival
genes (Karin, 2006; Karin and Lin, 2002; Luo et al., 2005;
Rashi-Elkeles et al., 2006). Indeed, we observed mRNA induc-
tion of the prosurvival genes cIAP1 and cFLIP within 6 hr of eto-
poside addition, and depletion of either protein significantly
augmented C8 activation (data not shown). Critically, NF-kB
stimulation also sets up an autocrine feedforward signaling
loop that involves TNF-a and TNFR1: In the presence of exten-
sive DSBs, this loop is sustained and promotes a second wave
of NF-kB signaling, accelerating TNF-a production and driving
TNFR1-mediated RIP1 autophosphorylation. In conjunction
with NEMO, RIP1 kinase promotes JNK3-mediated induction
of IL-8 secretion to alert other cells and recruits FADD to drive
C8 activation and cell death. K63-linked polyubiquitylation of
RIP1, which occurs following RIP1 phosphorylation, supports
the second phase of NF-kB activation but also can counteract
FADD-C8 recruitment, raising the possibility that RIP1 ubiquity-
lation provides a further regulatory node.
C8 is a key mediator of proapoptotic signaling by death recep-
tors (Ashkenazi and Dixit, 1998; Peter and Krammer, 2003;
Wilson et al., 2009). Several studies indicate that C8 can be acti-
vated in response to genotoxic stress (Engels et al., 2000; Jones
et al., 2001; Sohn et al., 2005; von Haefen et al., 2003;
Wesselborg et al., 1999). Our present work identifies a specific
path leading to C8 activation in response to extensive DSBs
via NF-kB and TNFR1. The finding that IL-8 secretion occurs in
concert with C8 activation raises the possibility that cells
destined to die in the context of strong DNA lesions send signals
to mobilize other cells in the tissue microenvironment to respond
to the damage, for example by helping to remove apoptotic cell
debris or to reinforce growth arrest (Acosta et al., 2008;
Varfolomeev et al., 2005; Waugh and Wilson, 2008). The
molecular interplay we defined bears intriguing similarity to the
mechanism of apoptosis activation by Smac-mimetic drugs,
which also engage C8 via NF-kB-mediated induction of TNF-a
(Bertrand et al., 2008; Petersen et al., 2007; Varfolomeev et al.,
2007). Our findings provide support for the conclusion that the
NF-kB cascade is not merely a prosurvival pathway; rather,
NF-kB also has the capacity to set up conditional pro-death
Figure 7. A Model Depicting ATM Regulation of Cell Fate in Response to Low versus High Levels of DNA Damage
In response to a low level of DSBs, ATM activation induces ubiquitylation of NEMO/IKK-g and its interaction with RIP1 and with IKK-a and -b, driving NF-kB
activation. This early NF-kB activation phase is independent of RIP1 kinase activity and TNFR1 and promotes cell survival by inducing transcription of
antiapoptotic genes such as cIAPs and cFLIP. Another key effect of NF-kB activation downstream of ATM is to set up an autocrine TNF-a-TNFR1 signaling loop.
If the damage is resolved, this loop subsides. In contrast, if the level of DSBs is extensive, ATM activity is sustained and the feedforward signaling loop continues
to operate. This accelerates further TNF-a production, supporting a second wave of NF-kB activity and driving TNFR1-mediated RIP1 phosphorylation. In turn,
RIP1 kinase promotes JNK3-mediated induction of IL-8 production to alert other cells of the damage and recruits FADD to activate C8 and trigger programmed
cell death. K63-linked polyubiquitylation of RIP1 by cIAPs, which occurs downstream of RIP1 kinase activation, promotes the second NF-kB activation phase;
however, it can also counteract the recruitment of FADD and C8 to RIP1, providing an additional level of regulation.
Cell 145, 92–103, April 1, 2011 ª2011 Elsevier Inc. 101
signals, which can be decoded in the context of cellular param-
eters such as the extent of DNA damage to dictate whether the
cell should live or die.
In conclusion, we have identified a NEMO- and RIP1-based
signaling that mediates ATM-induced cytokine secretion and
caspase activation selectively in the context of severe DNA
damage. These results provide important insight into the molec-
ular events that control cell fate downstream of ATM.
HeLa-M cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM),
and MDA231, MDA435, RKO cells were cultured in RPMI 1640. Culture media
were supplemented with 10% fetal bovine serum (FBS), 2 mM L-glutamine.
Reagents, siRNA, and cDNA Transfection
Please see Extended Experimental Procedures.
Immunoprecipitation and Immunoblot Analysis
Cellsweregrownon15cmplates, treatedasindicated, washedtwotimeswith
PBS, and harvested with 2 ml of the following lysis buffer: 20 mM Tris HCL
(pH 7.5), 150 mM NaCl, 10% glycerol, 1% Triton X-100, 2 mM EDTA, with
complete, EDTA-free protease Inhibitor (Roche) and phosphatae inhibitors
Cells were left on ice for 20 min and centrifuged at 14,000 rpm for 15 min.
Three milligrams of protein lysate was used for immunoprecipitation using
2 mg of the following antibodies: FADD (sc-5559), Caspase-8 (sc-6136),
NEMO (FL-419) for overnight at 4?C. The following day, 25 ml of protein A/G
ultralink resin (Thermo Scientific) was added for 2 hr at 4?C, and the IPs
were washed four times with lysis buffer, sample buffer was added, and beads
were boiled for 5 min at 95?C. Samples were then analyzed by SDS-PAGE
followed by IB. For further information on analysis of C8 activity and of RIP1
polyubiquitylation and phosphorylation, please see Extended Experimental
Supplemental Information includes Extended Experimental Procedures, six
figures, and one table and can be found with this article online at doi:10.
We thank Scot Marsters for help with constructs, Vishva Dixit and Kim Newton
for the RIP1 kinase domain deletion mutant and ubiquitin chain-specific anti-
bodies, John Moffat for help with KINOMEscan coordination, Domagoj Vucic
for BV6, members of Genentech’s FACS and DNA synthesis labs for technical
help, and members of the Ashkenazi lab for helpful discussions.
Received: July 2, 2010
Revised: January 21, 2011
Accepted: February 4, 2011
Published: March 31, 2011
Acosta, J.C., O’Loghlen, A., Banito, A., Guijarro, M.V., Augert, A., Raguz, S.,
Fumagalli, M., Da Costa, M., Brown, C., Popov, N., et al. (2008). Chemokine
Ashkenazi, A., and Dixit, V.M. (1998). Death receptors: signaling and modula-
tion. Science 281, 1305–1308.
Bartek, J., and Lukas, J. (2006). Cell biology: The stress of finding NEMO.
Science 311, 1110–1111.
Hopes and pitfalls. Nat. Rev. Drug Discov. 8, 33–40.
Bertrand, M.J., Milutinovic, S., Dickson, K.M., Ho, W.C., Boudreault, A.,
Durkin, J., Gillard, J.W., Jaquith, J.B., Morris, S.J., and Barker, P.A. (2008).
cIAP1 and cIAP2 facilitate cancer cell survival by functioning as E3 ligases
that promote RIP1 ubiquitination. Mol. Cell 30, 689–700.
Boatright, K.M., Renatus, M., Scott, F.L., Sperandio, S., Shin, H., Pedersen,
I.M., Ricci, J.E.,Edris, W.A., Sutherlin, D.P., Green,D.R.,et al. (2003). A unified
model for apical caspase activation. Mol. Cell 11, 529–541.
Chen, G., and Goeddel, D.V. (2002). TNF-R1 signaling: a beautiful pathway.
Science 296, 1634–1635.
Cho, Y.S., Challa, S., Moquin, D., Genga, R., Ray, T.D., Guildford, M., and
Chan, F.K. (2009). Phosphorylation-driven assembly of the RIP1-RIP3
complex regulates programmed necrosis and virus-induced inflammation.
Cell 137, 1112–1123.
Criswell, T., Leskov, K., Miyamoto, S., Luo, G., and Boothman, D.A. (2003).
Transcription factors activated in mammalian cells after clinically relevant
doses of ionizing radiation. Oncogene 22, 5813–5827.
Declercq, W., Vanden Berghe, T., and Vandenabeele, P. (2009). RIP kinases at
the crossroads of cell death and survival. Cell 138, 229–232.
Degterev, A., Hitomi, J., Germscheid, M., Ch’en, I.L., Korkina, O., Teng, X.,
Abbott, D., Cuny, G.D., Yuan, C., Wagner, G., et al. (2008). Identification of
RIP1 kinase as a specific cellular target of necrostatins. Nat. Chem. Biol. 4,
Engels, I.H., Stepczynska, A., Stroh, C., Lauber, K., Berg, C., Schwenzer, R.,
Wajant, H., Janicke, R.U., Porter, A.G., Belka, C., et al. (2000). Caspase-8/
FLICE functions as an executioner caspase in anticancer drug-induced
apoptosis. Oncogene 19, 4563–4573.
Fumagalli, M., and d’Adda di Fagagna, F. (2009). SASPense and DDRama in
cancer and ageing. Nat. Cell Biol. 11, 921–923.
Hande, K.R., Wedlund, P.J., Noone, R.M., Wilkinson, G.R., Greco, F.A., and
Wolff, S.N. (1984). Pharmacokinetics of high-dose etoposide (VP-16-213)
administered to cancer patients. Cancer Res. 44, 379–382.
Harper, J.W., and Elledge, S.J. (2007). The DNA damage response: Ten years
after. Mol. Cell 28, 739–745.
Hayden, M.S., and Ghosh, S. (2008). Shared principles in NF-kappaB
signaling. Cell 132, 344–362.
He, S., Wang, L., Miao, L., Wang, T., Du, F., Zhao, L., and Wang, X. (2009).
Receptor interacting protein kinase-3 determines cellular necrotic response
to TNF-alpha. Cell 137, 1100–1111.
Hickson, I., Zhao, Y., Richardson, C.J., Green, S.J., Martin, N.M., Orr, A.I.,
Reaper, P.M., Jackson, S.P., Curtin, N.J., and Smith, G.C. (2004). Identifica-
tion and characterization of a novel and specific inhibitor of the ataxia-telangi-
ectasia mutated kinase ATM. Cancer Res. 64, 9152–9159.
Hinz, M., Stilmann, M., Arslan, S.C., Khanna, K.K., Dittmar, G., and Scheider-
eit, C. (2010). A cytoplasmic ATM-TRAF6-cIAP1 module links nuclear DNA
damage signaling to ubiquitin-mediated NF-kappaB activation. Mol. Cell 40,
Huang, T.T., Wuerzberger-Davis, S.M., Wu, Z.H., and Miyamoto, S. (2003).
Sequential modification of NEMO/IKKgamma by SUMO-1 and ubiquitin medi-
ates NF-kappaB activation by genotoxic stress. Cell 115, 565–576.
Hur, G.M., Lewis, J., Yang, Q., Lin, Y., Nakano, H., Nedospasov, S., and Liu,
Z.G. (2003). The death domain kinase RIP has an essential role in DNA
damage-induced NF-kappa B activation. Genes Dev. 17, 873–882.
Jackson, S.P., and Bartek, J. (2009). The DNA-damage response in human
biology and disease. Nature 461, 1071–1078.
Janssens, S., and Tschopp, J. (2006). Signals from within: the DNA-damage-
induced NF-kappaB response. Cell Death Differ. 13, 773–784.
Janssens, S., Tinel, A., Lippens, S., and Tschopp, J. (2005). PIDD mediates
NF-kappaB activation in response to DNA damage. Cell 123, 1079–1092.
Jones, D.T., Ganeshaguru, K., Virchis, A.E., Folarin, N.I., Lowdell, M.W.,
Mehta, A.B., Prentice, H.G., Hoffbrand, A.V., and Wickremasinghe, R.G.
102 Cell 145, 92–103, April 1, 2011 ª2011 Elsevier Inc.
(2001). Caspase 8 activation independent of Fas (CD95/APO-1) signaling may
mediate killing of B-chronic lymphocytic leukemia cells by cytotoxic drugs or
gamma radiation. Blood 98, 2800–2807.
Karin, M. (2006). Nuclear factor-kappaB in cancer development and progres-
sion. Nature 441, 431–436.
Karin, M., and Lin, A. (2002). NF-kappaB at the crossroads of life and death.
Nat. Immunol. 3, 221–227.
Kasibhatla, S., Brunner, T., Genestier, L., Echeverri, F., Mahboubi, A., and
Green, D.R. (1998). DNA damaging agents induce expression of Fas ligand
and subsequent apoptosis in T lymphocytes via the activation of NF-kappa
B and AP-1. Mol. Cell 1, 543–551.
Kelliher, M.A., Grimm, S., Ishida, Y., Kuo, F., Stanger, B.Z., and Leder, P.
(1998). The death domain kinase RIP mediates the TNF-induced NF-kappaB
signal. Immunity 8, 297–303.
Li, N., Banin, S., Ouyang, H., Li, G.C., Courtois, G., Shiloh, Y., Karin, M., and
Rotman, G. (2001). ATM is required for IkappaB kinase (IKKk) activation in
response to DNA double strand breaks. J. Biol. Chem. 276, 8898–8903.
Linard, C., Marquette, C., Mathieu, J., Pennequin, A., Clarencon, D., and
Mathe, D. (2004). Acute induction of inflammatory cytokine expression after
gamma-irradiation in the rat: effect of an NF-kappaB inhibitor. Int. J. Radiat.
Oncol. Biol. Phys. 58, 427–434.
Luo, J.L., Kamata, H., and Karin, M. (2005). IKK/NF-kappaB signaling:
balancing life and death–a new approach to cancer therapy. J. Clin. Invest.
D.A., Hoffman, B., and Reed, J.C. (1994). Tumor suppressor p53 is a regulator
of bcl-2 and bax gene expression in vitro and in vivo. Oncogene 9, 1799–1805.
Nakano, K., and Vousden, K.H. (2001). PUMA, a novel proapoptotic gene, is
induced by p53. Mol. Cell 7, 683–694.
O’Driscoll, M., and Jeggo, P.A. (2006). The role of double-strand break repair -
insights from human genetics. Nat. Rev. Cancer 7, 45–54.
family and candidate mediator of p53-induced apoptosis. Science 288,
Papa, S., Bubici, C., Zazzeroni, F., Pham, C.G., Kuntzen, C., Knabb, J.R.,
Dean, K., and Franzoso, G. (2006). The NF-kappaB-mediated control of the
JNK cascade in the antagonism of programmed cell death in health and
disease. Cell Death Differ. 13, 712–729.
Peter, M.E., and Krammer, P.H. (2003). The CD95(APO-1/Fas) Dros. Inf.
Serv.C and beyond. Cell Death Differ. 10, 26–35.
Petersen, S.L., Wang, L., Yalcin-Chin, A., Li, L., Peyton, M., Minna, J., Harran,
P., and Wang, X. (2007). Autocrine TNFalpha signaling renders human cancer
cells susceptible to Smac-mimetic-induced apoptosis. Cancer Cell 12,
Piret, B., Schoonbroodt, S., and Piette, J. (1999). The ATM protein is required
for sustained activation of NF-kappaB following DNA damage. Oncogene 18,
Rashi-Elkeles, S.,Elkon,R.,Weizman, N., Linhart,C., Amariglio,N., Sternberg,
of ATM-dependent pro- and antiapoptotic signals in response to ionizing radi-
ation in murine lymphoid tissue. Oncogene 25, 1584–1592.
Rodier, F., Coppe, J.P., Patil, C.K., Hoeijmakers, W.A., Munoz, D.P., Raza,
S.R., Freund, A., Campeau, E., Davalos, A.R., and Campisi, J. (2009). Persis-
tent DNA damage signalling triggers senescence-associated inflammatory
cytokine secretion. Nat. Cell Biol. 11, 973–979.
Shiloh, Y. (2003). ATM and related protein kinases: safeguarding genome
integrity. Nat. Rev. Cancer 3, 155–168.
Shiloh, Y. (2006). The ATM-mediated DNA-damage response: taking shape.
Trends Biochem. Sci. 31, 402–410.
Sohn, D., Schulze-Osthoff, K., and Janicke, R.U. (2005). Caspase-8 can be
activated by interchain proteolysis without receptor-triggered dimerization
during drug-induced apoptosis. J. Biol. Chem. 280, 5267–5273.
Tinel, A., and Tschopp, J. (2004). The PIDDosome, a protein complex impli-
cated in activation of caspase-2 in response to genotoxic stress. Science
Ting, A.T., Pimentel-Muinos, F.X., and Seed, B. (1996). RIP mediates tumor
necrosis factor receptor 1 activation of NF-kappaB but not Fas/APO-1-initi-
ated apoptosis. EMBO J. 15, 6189–6196.
Varfolomeev, E., Maecker, H., Sharp, D., Lawrence, D., Renz, M., Vucic, D.,
by Apo2 ligand/tumor necrosis factor-related apoptosis-inducing ligand.
J. Biol. Chem. 280, 40599–40608.
Varfolomeev, E., Blankenship, J.W., Wayson, S.M., Fedorova, A.V., Kayagaki,
N., Garg, P., Zobel, K., Dynek, J.N., Elliott, L.O., Wallweber, H.J., et al. (2007).
IAP antagonists induce autoubiquitination of c-IAPs, NF-kappaB activation,
and TNFalpha-dependent apoptosis. Cell 131, 669–681.
Varfolomeev, E., Goncharov, T., Fedorova, A.V., Dynek, J.N., Zobel, K., De-
shayes, K., Fairbrother, W.J., and Vucic, D. (2008). c-IAP1 and c-IAP2 are crit-
ical mediators of tumor necrosis factor alpha (TNFalpha)-induced NF-kappaB
activation. J. Biol. Chem. 283, 24295–24299.
von Haefen, C., Wieder, T., Essmann, F., Schulze-Osthoff, K., Dorken, B., and
Daniel, P.T. (2003). Paclitaxel-induced apoptosis in BJAB cells proceeds via
a death receptor-independent,caspases-3/-8-driven mitochondrial amplifica-
tion loop. Oncogene 22, 2236–2247.
Wang, L., Du, F., and Wang, X. (2008). TNF-alpha induces two distinct
caspase-8 activation pathways. Cell 133, 693–703.
Waugh, D.J., and Wilson, C. (2008). The interleukin-8 pathway in cancer. Clin.
Cancer Res. 14, 6735–6741.
ubiquitin system. Cell Death Differ. 17, 14–24.
Wesselborg, S., Engels, I.H., Rossmann, E., Los, M., and Schulze-Osthoff, K.
(1999). Anticancer drugs induce caspase-8/FLICE activation and apoptosis in
the absence of CD95 receptor/ligand interaction. Blood 93, 3053–3063.
Wilson, N.S., Dixit, V., and Ashkenazi, A. (2009). Death receptor signal trans-
ducers: nodes of coordination in immune signaling networks. Nat. Immunol.
Wong, W.W., Gentle, I.E., Nachbur, U., Anderton, H., Vaux, D.L., and Silke, J.
(2010). RIPK1 is not essential for TNFR1-induced activation of NF-kappaB.
Cell Death Differ. 17, 482–487.
Wu, Z.H., Shi, Y., Tibbetts, R.S., and Miyamoto, S. (2006). Molecular linkage
between the kinase ATM and NF-kappaB signaling in response to genotoxic
stimuli. Science 311, 1141–1146.
Wu, Z.H., Wong, E.T., Shi, Y., Niu, J., Chen, Z., Miyamoto, S., and Tergaonkar,
V. (2010). ATM- and NEMO-dependent ELKS ubiquitination coordinates
TAK1-mediated IKK activation in response to genotoxic stress. Mol. Cell 40,
Zhou, D., Yu, T., Chen, G., Brown, S.A., Yu, Z., Mattson, M.P., and Thompson,
J.S. (2001). Effects of NF-kappaB1 (p50) targeted gene disruption on ionizing
radiation-induced NF-kappaB activation and TNFalpha, IL-1alpha, IL-1beta
and IL-6 mRNA expression in vivo. Int. J. Radiat. Biol. 77, 763–772.
Cell 145, 92–103, April 1, 2011 ª2011 Elsevier Inc. 103