Tài liệu Báo cáo khoa học: Upregulation of DR5 by proteasome inhibitors potently sensitizes glioma cells to TRAIL-induced apoptosis doc

Upregulation of DR5 by proteasome inhibitors potently
sensitizes glioma cells to TRAIL-induced apoptosis
Holger Hetschko
1
, Valerie Voss
1
, Volker Seifert
1
, Jochen H. M. Prehn
2
and Donat Ko
¨
gel
1
1 Department of Neurosurgery, Centre for Neurology and Neurosurgery, Johann Wolfgang Goethe University Clinics, Frankfurt ⁄ Main,
Germany
2 Department of Physiology and Medical Physics, Royal College of Surgeons in Ireland, Dublin, Ireland
Gliomas are the most common and malignant primary
brain tumors in humans. Glioblastoma multiforme is
the highest-grade as well as the most aggressive and
frequent glioma [1]. Because gliomas are characterized
by a diffuse infiltrative growth into the surrounding
brain tissue, complete surgical resection of glioblas-
toma multiforme tumors is virtually impossible [2]. In
addition, high-grade gliomas exhibit only limited sensi-
tivity to ensuing multimodal treatment with radio-
therapy and chemotherapy [2], which in large part is
Keywords
apoptosis; astrocytoma; death receptor;
proteasome; stress kinase
Correspondence

D. Ko
¨
gel, Experimental Neurosurgery,
Johann Wolfgang Goethe University Clinics,
Theodor-Stern-Kai 7, Neuroscience Centre,
D-60590 Frankfurt am Main, Germany
Fax: +49 69 6301 5575
Tel: +49 69 6301 6940
E-mail:
(Received 28 January 2008, revised 19
February 2008, accepted 21 February 2008)
doi:10.1111/j.1742-4658.2008.06351.x
This study was undertaken to explore the potential of new therapeutic
approaches designed to reactivate cell death pathways in apoptosis-refrac-
tory gliomas and to characterize the underlying molecular mechanisms of
this reactivation. Here we investigated the sensitivity of a panel of glioma
cell lines (U87, U251, U343, U373, MZ-54, and MZ-18) to apoptosis
induced by the death receptor ligand tumor necrosis factor-related apopto-
sis-inducing ligand (TRAIL), TRAIL in combination with gamma irradia-
tion, and TRAIL in combination with proteasome inhibitors (MG132 and
epoxomicin). Analysis of these six glioma cell lines revealed drastic differ-
ences in their sensitivity to these treatments, with two of the six cell lines
revealing no significant induction of cell death in response to TRAIL
alone. Interestingly, the proteasome inhibitors MG132 and epoxomicin
were capable of potentiating TRAIL-induced apoptosis in TRAIL-sensitive
U87 and U251 cells and of reactivating apoptosis in TRAIL-resistant U343
and U373 cells. In contrast, gamma irradiation had no synergistic effects
with TRAIL in the two TRAIL-resistant cell lines. RNA interference
against death receptor 5 (DR5) revealed that reactivation of TRAIL-
induced apoptosis by proteasome inhibitors depended on enhanced tran-

scription and surface expression of DR5. Transient knockdown of the
transcription factor GADD153⁄ C ⁄ EBP homologous protein and applica-
tion of the synthetic c-Jun N-terminal kinase inhibitor SP600125 indicated
that enhanced DR5 expression occurred independently of GADD153 ⁄
C ⁄ EBP homologous protein, but required activation of the c-Jun N-termi-
nal kinase ⁄ c-Jun signaling pathway. Novel therapeutic approaches using
TRAIL or agonistic TRAIL receptor antibodies in combination with pro-
teasome inhibitors may represent a promising approach to reactivate apop-
tosis in therapy-resistant high-grade gliomas.
Abbreviations
Ac-DEVD-AMC, acetyl-DEVD-7-amido-4-methylcoumarin; CHOP, C ⁄ EBP homologous protein; DR4, death receptor 4; DR5, death receptor 5;
FACS, fluorescence-activated cell sorting; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; JNK, c-Jun N-terminal kinase; NF-jB,
nuclear factor kappa B; PI, proteasome inhibitor; siRNA, small interfering RNA; TRAIL, tumor necrosis factor-related apoptosis-inducing
ligand.
FEBS Journal 275 (2008) 1925–1936 ª 2008 The Authors Journal compilation ª 2008 FEBS 1925
caused by inherent and potent apoptosis resistance [3].
Clearly, overcoming this resistance by approaches
designed to reactivate apoptosis in malignant glioma
has important implications for the development of
novel glioma therapies.
Tumor necrosis factor-related apoptosis-inducing
ligand (TRAIL), a member of the tumor necrosis factor
superfamily, has been shown to be a promising candi-
date for novel anticancer therapies [4]. Interestingly,
TRAIL is capable of inducing apoptosis in a wide
range of tumor cells, but not in normal tissue. This
tumor-selective cytotoxicity has been shown for glioma
cells in comparison to non-transformed astrocytes
in vitro [5,6]. The physiological function of TRAIL
remains unclear, although studies suggest that TRAIL

plays a key role in tumor surveillance by the immune
system [7]. TRAIL induces apoptosis by binding to
its agonistic cognate cell surface receptors death
receptor 4 (DR4) ⁄ TRAIL R1 and death receptor 5
(DR5) ⁄ TRAIL R2, and this then leads to recruitment
of the adaptor protein Fas-associated death domain
and the initiator caspases procaspase-8 and procaspase-
10, and formation of a membrane-bound multiprotein
complex called death-inducing signaling complex [4].
Death-inducing signaling complex formation then
leads to autoproteolytic cleavage of procaspase-8 and
procaspase-10, and subsequent downstream activa-
tion of the extrinsic and intrinsic pathways of apoptosis
[4].
Despite the considerable interest raised by the poten-
tial of TRAIL in novel anticancer therapies, accumu-
lating evidence suggests that TRAIL alone may not be
sufficient to efficiently activate apoptosis in many types
of cancers, inluding gliomas [8]. Hence, much effort
has been made to establish new modalities for com-
bined treatments with TRAIL and other antineoplastic
agents or apoptosis inducers to improve the potency of
TRAIL-based therapeutic approaches.
Proteasome inhibitors (PIs) represent a highly prom-
ising novel class of anticancer agents [9]. PIs are
already in clinical use, as bortezomib (PS-341 ⁄
Velcade) has been approved for the treatment of multi-
ple myeloma [10]. Recent evidence suggests that PIs
also might be efficient agents for the treatment of solid
tumors, such as lung and prostate cancer [11,12].

This study reveals that TRAIL-induced apoptosis
can be efficiently reactivated in TRAIL-resistant malig-
nant glioma cell lines by combined treatment with PIs.
Furthermore, we show that reactivation of TRAIL-
induced apoptosis by proteasome inhibition induces
the c-Jun N-terminal kinase (JKN) ⁄ c-Jun stress signal-
ing pathway and requires enhanced JNK ⁄ c-Jun-depen-
dent surface expression of DR5.
Results
Synergistic effects of combined treatment with
TRAIL and gamma irradiation
To analyze the sensitivity of high-grade gliomas to cell
death induced by the death ligand TRAIL, we
employed a panel of six grade III–IV glioma cell lines
(U87, U251, U343, U373, MZ-18, and MZ-54). In an
initial experiment, the cells were treated for 48 h at a
final concentration of 250 ng TRAILÆmL
)1
(Fig. 1A).
Surprisingly, only two of six cell lines (U87 and U251)
significantly responded to TRAIL treatment as
measured by annexin-V–FLUOS ⁄ propidium iodide
staining and flow cytometry (Fig. 1A). Prolonged incu-
bation for up to 96 h also did not induce detectable
cell death in the four TRAIL-resistant cell lines (data
not shown).
As gamma irradiation is an existing component of
current glioma therapies, and as activation of TRAIL
receptors and DNA damage were shown to have syner-
gistic death-inducing effects in other types of cancer

[13], we investigated whether similar effects could also
be observed in glioma cells. Cell lines U87, U343 and
U373 were subjected to gamma irradiation with single
doses of 10 Gy and 20 Gy respectively. Twenty-four
hours postirradiation, the cultures were treated with
250 ngÆmL
)1
TRAIL for an additional 24 h, after which
cell death was measured by fluorescence-activated cell
sorting (FACS) analysis (Fig. 1B). In comparison to the
controls, a dose of 10 Gy had no effect on cell viability,
either alone or in combination with TRAIL (data not
shown). The higher dose of 20 Gy induced significant
cell death in all three cell lines after 48 h ( 20–30%),
and was capable of potentiating TRAIL-induced cell
death in the TRAIL-sensitive cell line U87. However, in
U343 and U373 cells, there was no significant synergis-
tic effect of TRAIL and gamma irradiation, indicating
that the activation of DNA damage-induced signaling
pathways was not sufficient to reactivate TRAIL sensi-
tivity in the TRAIL-resistant cell lines (Fig. 1B). It is of
note that this lack of apoptosis reactivation was inde-
pendent of the p53 status of these cells, as U343 cells,
but not U373 cells, have been shown to express func-
tional p53 [14,15].
PIs potently reactivate TRAIL-induced apoptosis
in glioma cells
Next, we investigated the synergistic effects of TRAIL
and PIs in our glioma cell lines. For that purpose, we
employed two different PIs in combination with

TRAIL: MG132, a potent reversible inhibitor targeting
Apoptosis by TRAIL and proteasome inihibitors H. Hetschko et al.
1926 FEBS Journal 275 (2008) 1925–1936 ª 2008 The Authors Journal compilation ª 2008 FEBS
the 26S complex of the proteasome [16], and epoxomi-
cin, a highly specific and irreversible inhibitor of
several hydrolyzing activities of the proteasome [17],
at concentrations of 2.5 lm and 50 nm respectively
(Fig. 2). Whereas treatment with MG132 alone had a
moderate cytotoxic effect in three cell lines (U87,
U373, and MZ-54), treatment with epoxomicin signifi-
cantly induced cell death in only two of six cell lines
(U373 and MZ-54) (Fig. 2). Interestingly, both PIs
potently enhanced TRAIL-induced apoptosis in the
two TRAIL-sensitive cell lines (U87 and U251), and
were able to reactivate apoptosis in four TRAIL-resis-
tant cell lines (U343, U373, MZ-54, and MZ-18). It is
of note that all six investigated cell lines were previ-
ously shown to express DR5, whereas DR4 expression
was undetectable in U251, U373 and MZ-54 cells [18].
Furthermore, reactivation of TRAIL-induced cell
death did not depend on functional p53, as it was also
observed in mutant p53 expressing U373 cells.
PIs potently induce expression of DR5
in glioma cells
To elucidate the underlying molecular mechanisms of
the observed synergistic effects of PIs, we focused on
the transcriptional activation of pro-apoptotic genes
after proteasome inhibition with MG132 and epoxo-
micin. Samples of cells exposed to gamma irradiation
served as a control for p53-dependent gene activation.

Although we found markedly enhanced DR5 trans-
criptional activation after proteasome inhibition and
gamma irradiation in all four cell lines investigated
(Fig. 3A), there was no evidence for activation of
DR4 by any stimulus. However the stress-activated
pro-apoptotic transcription factors C ⁄ EBP homolo-
gous protein (CHOP) and c-Jun were transcriptionally
activated after proteasome inhibition in all four cell
lines, whereas induction of CHOP and c-Jun was less
prominent or even absent after gamma irradiation
(Fig. 3A).
Western blot analysis confirmed that DR4 protein
levels were not elevated after proteasome inhibition or
gamma irradiation (Fig. 3B). Although we found a
strong elevation of DR5 protein levels after protea-
some inhibition, we could detect only moderately
enhanced protein levels after gamma irradiation. Simi-
larly, the amount of CHOP protein was markedly
increased by proteasome inhibition in all three cell
lines, but in only one cell line (U87) after gamma
irradiation.
B
A
Fig. 1. Synergistic effects of combined treatment with TRAIL and gamma irradiation. (A) TRAIL resistance is commonly observed in glioma
cell lines. U87, U251, U343, U373, MZ-54 and MZ-18 cells were treated with 250 ngÆmL
)1
TRAIL for 48 h. Cells were stained with annexin-
V–FLUOS ⁄ propidium iodide, and cell death was measured with flow cytometry. Data are means ± SEM from four independent cultures.
*P < 0.05 as compared to untreated control. (B) TRAIL-sensitive cells (U87) and TRAIL-resistant cells (U343, U373) were subjected to
gamma irradiation (20 Gy) and subsequently were treated with 250 ngÆmL

)1
TRAIL or were left untreated for an additional 24 h. Cells were
irradiated once in an Elekta SL75 ⁄ 5 linear accelerator (6 MeV). Apoptosis was quantified with annexin-V–FLUOS ⁄ propidium iodide staining
and flow cytometry. Data are means ± SEM from four independent cultures. *P < 0.05 as compared to untreated control.
#
P < 0.05 as com-
pared to treatment of the same cell line with gamma irradiation alone. Similar results were obtained in three separate experiments.
H. Hetschko et al. Apoptosis by TRAIL and proteasome inihibitors
FEBS Journal 275 (2008) 1925–1936 ª 2008 The Authors Journal compilation ª 2008 FEBS 1927
Upregulation of DR5 by PIs contributes to the
reactivation of TRAIL-induced cell death
in glioma cells
Although the PIs prominently upregulated DR5
(Fig. 3), it was not clear whether this elevated expres-
sion was responsible for the dramatic reactivation of
TRAIL-induced apoptosis observed in malignant gli-
oma cells. To address this question, we knocked down
expression of DR5 by small interfering RNA (siRNA)
duplexes targeted against DR5 mRNA and studied the
effect on caspase activation after treatment of the cells
with TRAIL, MG132, and TRAIL in combination
with MG132. In initial experiments, U343 cells were
transfected with DR5 siRNA, treated with or without
2.5 lm MG132, and subjected to western blot analysis
(Fig. 4A). Transfection with siRNA against DR5
resulted in a robust reduction of DR5 activation after
MG132 treatment as compared to MG132-treated con-
trol cells transfected with a nontargeting scrambled
siRNA (Fig. 4A). In contrast, transfection with the
siRNAs had no major effect on expression levels of

DR4. To determine whether the changes in the amount
of DR5 protein in the protein lysates were also
reflected by the levels of DR5 surface expression, we
performed a flow cytometric analysis, which indeed
confirmed enhanced DR5 surface expression after
treatment of the cells with MG132. In analogy to the
western blot data, DR5 surface expression was not
completely abolished in DR5 siRNA-transfected cells
treated with MG132, but was reduced to basal surface
expression levels of scrambled siRNA-transfected con-
trol cells (Fig. 4B). Under the experimental conditions
chosen, TRAIL alone caused only weak induction of
caspase-3-like activity in the TRAIL-sensitive cell lines
U87 and U251 (Fig. 4C). Nevertheless, we observed a
clear trend towards an apoptosis-inhibitory effect of
the DR5 knockdown, indicating that basal expression
of DR5 is required for the TRAIL sensitivity in these
cells. Apoptosis induced by TRAIL in combination
Fig. 2. Reactivation of TRAIL-induced apop-
tosis after combined treatment with TRAIL
and PIs. TRAIL-sensitive cells (U87 and
U251) and TRAIL-resistant cells (U343,
U373, MZ-54, and MZ-18) were treated with
250 ngÆmL
)1
TRAIL and the PIs MG132
(2.5 l
M) and epoxomicin (50 nM) or vehicle
(dimethylsulfoxide) for 16 h. Apoptosis was
quantified with annexin-V–FLUOS ⁄ propidium

iodide staining and flow cytometry. Data are
means ± SEM from four independent cul-
tures. *P < 0.05 as compared to dimethyl-
sulfoxide-treated control.
#
P < 0.05 as
compared to treatment of the same cell line
with the respective PI alone. Similar results
were obtained in three separate experi-
ments.
Apoptosis by TRAIL and proteasome inihibitors H. Hetschko et al.
1928 FEBS Journal 275 (2008) 1925–1936 ª 2008 The Authors Journal compilation ª 2008 FEBS
with MG132 was potently attenuated in TRAIL-resis-
tant U343 cells, TRAIL-sensitive U251 cells and
TRAIL-sensitive U87 cells after knockdown of DR5 in
comparison to the respective controls (Fig. 4C). These
results suggest that enhanced surface expression of
DR5 plays a critical role in the reactivation of
TRAIL-induced apoptosis after proteasome inhibition
in glioma cells.
CHOP is not involved in DR5 upregulation by PIs
and cell death induced by proteasome inhibition
plus TRAIL
It was recently reported that CHOP, an endoplasmic
reticulum stress-inducible member of the CCAAT ⁄
enhancer-binding protein family, can act as an
upstream activator of DR5 in certain types of cancer
cells [19,20]. In line with our previous findings [21],
we found CHOP to be strongly activated after pro-
teasome inhibition (Fig. 3), and were therefore inter-

ested in whether CHOP was also involved in DR5
upregulation by PIs in glioma cells. To determine
this, U343 cells were transfected with siRNA duplexes
targeted against CHOP or with nontargeting scram-
bled siRNA, and then treated with MG132 (Fig. 5A).
Western blot analysis revealed that although CHOP
expression was knocked down significantly in the
CHOP siRNA-transfected cells as compared to the
scrambled siRNA-transfected control cells, the activa-
tion of DR5 after MG132 treatment was not affected
by the knockdown (Fig. 5A). Furthermore, the
knockdown of CHOP did not result in a significant
attenuation of apoptosis induced by TRAIL plus
MG132 in CHOP siRNA-transfected cells as com-
pared to scrambled siRNA-transfected control cells
(Fig. 5B).
Inhibition of the JNK/c-Jun pathway abrogates
PI-mediated DR5 upregulation and cell death
Our data so far had suggested that PIs potently
induced the expression of DR5 in glioma cells in a
CHOP-independent manner. Although the tumor sup-
pressor p53 has also been implicated in regulation of
DR5 expression [22], the p53-deficient cell line U373
showed a similiar increase in DR5 upregulation as the
wild-type p53-expressing cell lines U87 and U343 after
treatment with PIs in this study (Fig. 3A,B), indicating
that p53 was not responsible for the observed DR5
induction. As a third potential upstream regulator of
DR5, we next focused on the JNK ⁄ c-Jun stress signal-
ing pathway. It is well established that PIs such as

MG132 are capable of inducing the JNK ⁄ c-Jun path-
way [23]. To address the issue of whether activation of
B
A
Fig. 3. PIs enhance expression of DR5. (A)
DR5, CHOP and c-Jun are transcriptionally
induced after proteasome inhibition. Cells
were either treated with 2.5 l
M MG132,
50 n
M epoxomicin or vehicle (dimethylsulf-
oxide) for 16 h or subjected to gamma irradi-
ation (20 Gy) and subsequently lysed 24 h
after exposure. Expression of DR4 (28
cycles), DR5 (25 cycles), CHOP (28 cycles)
and GAPDH (25 cycles) was determined by
semiquantitative RT-PCR. For detection of
c-Jun expression, RT-PCR with GAPDH
serving as an internal control (in the same
PCR reaction) was performed (30 cycles for
c-Jun and 30 cycles for GAPDH). Similar
results were obtained in at least three sepa-
rate experiments. (B) Western blot analysis
of DR4, DR5 and CHOP expression levels.
Forty micrograms of protein were loaded
onto each lane, with a-tubulin serving as a
loading control. Similar results were
obtained in two separate experiments.
H. Hetschko et al. Apoptosis by TRAIL and proteasome inihibitors
FEBS Journal 275 (2008) 1925–1936 ª 2008 The Authors Journal compilation ª 2008 FEBS 1929

JNKs and c-Jun contributes to PI-dependent DR5
upregulation in human glioma cells, we investigated
the effect of MG132 on DR5 upregulation in the pres-
ence of the JNK-specific inhibitor SP600125 by wes-
tern blot analysis. At concentrations of 10–30 lm,
SP600125 abrogated the MG132-induced increase in
phosphorylated c-Jun levels significantly (Fig. 6A). We
also observed that in combination with MG132,
SP600125 effectively blocked MG132-induced DR5
expression in a dose-dependent manner in U251 cells,
and this was subsequently confirmed for the highest
concentration used (30 lm) in U343 cells (Fig. 6B).
There was no detectable effect of SP600125 on
MG132-triggered induction of CHOP, suggesting that
the AP-1 site in the CHOP promoter [24] is not
required for enhanced CHOP expression after protea-
some inhibition, and lending further support to the
notion that the observed DR5 upregulation occured in
a CHOP-independent manner. To analyze the rele-
vance of JNK ⁄ c-Jun-dependent DR5 upregulation in
cell death triggered by TRAIL in combination with
MG132, we again performed caspase assays. Blocking
the JNK ⁄ c-Jun signaling pathway before treatment of
the cells with MG132 led to a significant decrease of
caspase-3-like activity induced by TRAIL plus MG132
in both TRAIL-sensitive and TRAIL-resistant glioma
cells (Fig. 6C). These results strongly suggest a promi-
nent role of the JNK ⁄ c-Jun signaling pathway in
A
B

C
Fig. 4. RNA interference against DR5 effi-
ciently protects cells from TRAIL-induced
apoptosis after proteasome inhibition.
Twenty-four hours after transfection with
scrambled control siRNA or DR5 siRNA,
TRAIL-resistant U343 cells were treated
with 2.5 l
M MG132 or dimethylsulfoxide for
16 h. (A) Analysis of DR5 and DR4 expres-
sion levels by western blotting. a-Tubulin
served as a loading control. Similar results
were obtained in two separate experiments.
(B) Analysis of cell surface expression of
DR5 after RNA interference against DR5
(DR5) or treatment with control siRNA (scr),
subsequent treatment with 2.5 l
M MG132
or dimethylsulfoxide (16 h), and staining
with a specific goat anti-DR5 IgG. Unspecific
goat IgG served as isotype control. The
experiment was repeated twice with similiar
results. (C) Knockdown of DR5 inhibits
apoptosis after treatment with TRAIL and
MG132. Following transfection with DR5
siRNA (DR5) and control siRNA (scr), cells
were treated with dimethylsulfoxide (con-
trol), 2.5 l
M MG132, TRAIL (250 ngÆmL
)1

)
or TRAIL in combination with MG-132 for an
additional 16 h, after which the cells were
harvested and whole cell lysates were moni-
tored for caspase-3-like activity by measur-
ing cleavage of the fluorogenic substrate
Ac-DEVD-AMC (10 l
M). Data are
means ± SEM from four to eight indepen-
dent cultures. *P < 0.05 as compared to
dimethylsulfoxide-treated control. Similar
results were obtained in two separate
experiments.
Apoptosis by TRAIL and proteasome inihibitors H. Hetschko et al.
1930 FEBS Journal 275 (2008) 1925–1936 ª 2008 The Authors Journal compilation ª 2008 FEBS
upregulation of DR5 and resensitization to apoptosis
after proteasome inhibition.
Discussion
The high resistance of malignant gliomas to the cur-
rently used radiation therapy and chemotherapy proto-
cols is a persistent obstacle to the improvement of the
outcome in glioma patients. Hence, a much effort has
been made in recent years to obtain more data on the
molecular mechanisms underlying gliomagenesis and
therapy resistance of gliomas, with the aim of develop-
ing more efficient and target-specific therapies.
The death receptor ligand TRAIL and agonistic
TRAIL-R antibodies are being considered as novel
and promising therapeutic agents for the treatment of
hematological malignancies and solid tumors, includ-

ing gliomas. Despite the described tumor-selective pro-
apoptotic properties of TRAIL, many cancer cells are
innately resistant to TRAIL. In line with previous
studies [6,25], all 18 glioma cell lines investigated in
this study exhibited expression of DR5, but four of six
cell lines showed no significant response to TRAIL.
In order to identify synergistic treatments to modu-
late TRAIL sensitivity and to reactivate apoptosis in
glioma cells, we compared the effects of already exist-
ing treatment modalities (gamma irradiation) with PIs
(MG132 and epoxomicin), a new class of chemothera-
peutic drugs with tremendous therapeutic potential [9].
The correct functioning of the ubiquitin–proteasome
pathway is essential for the degradation of the major-
ity of intracellular proteins and is implicated in many
cellular processes, including regulation of apoptosis.
Both epoxomicin and MG132 were able to potently
enhance apoptosis in the TRAIL-sensitive cell lines.
Both PIs increased apoptosis  5–6-fold in the
TRAIL-sensitive cell lines U87 and U251, and even at
subtoxic levels they were able to reactivate apoptosis
in both TRAIL-resistant cell lines (U343 and U373).
It is widely accepted that PIs can act on mutiple cellu-
lar targets implicated in regulation of apoptosis
[20,21,26,27]. However, despite abundant evidence for
the therapeutic potential of PIs in a variety of malig-
nancies, the relevant signaling pathways leading to
apoptosis triggered by proteasome inhibition seem to
vary substantially between different types of cancer. In
some types of cancer, such as prostate cancer and leu-

kemia, upregulation of DR5 seems to play a critical
role in the potentiating effects of PIs on TRAIL-
induced cell death [20,28,29]. Our data clearly indicate
that treatment with PIs potently elevated DR5 expres-
sion at the mRNA and protein levels in glioma cells,
whereas gamma irradiation induced only a moderate
increase in DR5 protein levels, even in the wild-type
p53-expressing cell lines U87 and U343. Enhanced pro-
tein levels and surface expression of DR5 after PI
treatment might be caused by a cumulative effect of
transcriptional induction and decreased degradation of
DR5 protein, thus providing an explanation for the
high efficiency of PIs in reactivation of TRAIL-depen-
dent cell death observed in this study, and also for the
reduced potency of gamma irradiation in elevating
DR5 protein levels.
To assess the importance of DR5 in PI-triggered
reactivation of TRAIL-dependent apoptosis, we per-
formed transient RNA interference against DR5
expression. Indeed, the knockdown of PI-triggered
DR5 induction had a very potent effect on DR5
A
B
Fig. 5. DR5 protein levels are not affected by RNA interference
against CHOP. (A) Four hours after transfection with CHOP siRNA
(CHOP) and scrambled control siRNA (scrambled), TRAIL-resistant
U343 cells were treated with 2.5 l
M MG132 or dimethylsulfoxide
for 16 h. CHOP and DR5 expression levels were analyzed by wes-
tern blotting. a-Tubulin served as a loading control. (B) Knockdown

of CHOP expression has no effect on TRAIL-induced apoptosis
after proteasome inhibition. U343 cells were treated as described
in (A), and caspase-3-like activity was measured by Ac-DEVD-AMC
cleavage. Data are means ± SEM from four to eight independent
cultures. *P < 0.05 as compared to dimethylsulfoxide-treated con-
trol. Similar results were obtained in two separate experiments.
H. Hetschko et al. Apoptosis by TRAIL and proteasome inihibitors
FEBS Journal 275 (2008) 1925–1936 ª 2008 The Authors Journal compilation ª 2008 FEBS 1931
surface expression levels and apoptosis induced by
TRAIL in combination with MG132 in U343 and
U251 cells. These data suggest that, in contrast to
other cancer types, such as hepatocellular carcinoma
[30], transcriptional upregulation of DR5 plays a piv-
otal role in potentiation and reactivation of TRAIL-
induced apoptosis in glioma cells.
We then addressed the question of which upstream
signaling pathways might lead to enhanced DR5 sur-
face expression after PI treatment, and in line with
previous observations in other types of cancer cells
[21], we demonstrated that the pro-apoptotic transcrip-
tion factor CHOP was potently activated on the tran-
scriptional level by PIs in glioma cells. Although
CHOP has been previously described as a putative
upstream regulator of DR5 in different types of cancer
cells [19,20], we did not observe any significant CHOP-
dependent effects on DR5 expression and cell death,
indicating that CHOP is not required for DR5 induc-
tion triggered by PIs in glioma cells.
Two other apoptosis-regulating transcription fac-
tors whose activity is known to be modulated by PIs

are p53 and nuclear factor kappa B (NF-jB).
Although the majority of reports suggest that PIs
induce a p53-dependent form of apoptosis in most
cancer cell types, such as leukemia cells [27], as well
as melanoma, colon cancer and myeloma cells
[21,26,31], PIs have been shown to induce p53-inde-
pendent apoptosis in glioma cells [32,33]. In line
with these observations, enhanced DR5 expression
did not depend on the p53 status (wild-type or
mutated) in the three glioma cell lines investigated in
this study.
As IjBa, the endogenous inhibitor of the transcrip-
tion factor NF-jB, is continuously degraded via the
proteasome pathway, PIs act as indirect inhibitors of
the antiapoptotic transcription factor NF-jB [34].
However, we have previously shown that treatment of
neuroblastoma and colon cancer cells with epoxomicin
is not associated with noticeable transcriptional
A
C
B
Fig. 6. The JNK ⁄ c-Jun signaling pathway contributes to DR5 induction and apoptosis induced by TRAIL plus PIs in glioma cells. (A) Inhibition
of JNK ⁄ c-Jun signaling downmodulates DR5 induction in a dose-dependent manner. U251 cells were pretreated with the indicated concen-
trations of the JNK inhibitor SP600125 or vehicle (dimethylsulfoxide) for 2 h and treated with 2.5 l
M MG132 or dimethylsulfoxide for an addi-
tional 16 h. Whole cell lysates were analyzed by western blotting for levels of phosphorylated c-Jun and DR5. a-Tubulin served as a loading
control. (B) SP600125 downmodulates PI-triggered DR5 induction in U343 cells. U343 cells were pretreated with 30 l
M SP600125 or di-
methylsulfoxide for 2 h, and subsequently treated with 2.5 l
M MG132 for an additional 16 h in the presence or absence of 30 lM

SP600125. Protein levels of phosphorylated c-Jun, DR5 and CHOP were analyzed by western blotting. a-Tubulin served as a loading control.
(C) SP600125 inhibits apoptosis induced by TRAIL plus MG132 in three different glioma cell lines. U87, U251 and U343 cells were treated
as described in (B). Caspase-3-like activity was measured by Ac-DEVD-AMC cleavage. Data are means ± SEM from four to eight indepen-
dent cultures. *P < 0.05 as compared to dimethylsulfoxide-treated control. Similar results were obtained in three separate experiments.
Apoptosis by TRAIL and proteasome inihibitors H. Hetschko et al.
1932 FEBS Journal 275 (2008) 1925–1936 ª 2008 The Authors Journal compilation ª 2008 FEBS
changes of antiapoptotic NF-jB target genes such as
Bcl-2, Bcl-xL, and the inhibitors of apoptosis (IAPs)
[21]. In addition, it was recently reported that most
glioblastoma cell lines exhibit only low constitutive
NF-jB activity, and inhibition of NF-jB did not sig-
nificantly influence apoptosis induced by DNA dam-
age, TRAIL and PIs in glioma cells in this study [35].
Although inactivation of NF-jB has been suggested to
play a major role in the antitumorigenic effect of the
PI bortezomib (Velcade ⁄ PS-341) in multiple myeloma
[36] and melanoma cells [26], inhibition of NF-kB is
not required to sensitize hepatocellular carcinoma cells
and lymphoma cells to apoptosis [37,38], again empha-
sizing the cancer type-specific effects of PIs.
As PIs are also known to trigger the stress-induced
JNK ⁄ c-Jun pathway [23,33], we finally asked the
question of whether this pathway might be involved
in the observed DR5 induction triggered by PIs.
RT-PCR analysis revealed a potent upregulation of
c-Jun at the transcriptional level, suggesting upstream
activation of JNKs, phosphorylation of c-Jun and
enhanced c-Jun expression by an autoregulatory feed-
forward mechanism [39]. Indeed, application of the
synthetic JNK inhibitor SP600125 significantly

reduced levels of phosphorylated c-Jun in a dose-
dependent manner after PI treatment in glioma
cells, whereas it had no discernible effect on CHOP
induction. Inhibition of the JNK ⁄ c-Jun pathway with
SP600125 also significantly downmodulated
PI-induced DR5 induction and cell death triggered by
TRAIL in combination with MG132.
In conclusion, our data emphasize the importance of
the JNK ⁄ c-Jun signaling cascade in transcriptional
induction of DR5 and resensitization to TRAIL-
induced apoptosis in malignant glioma cells, and sug-
gest that new target-specific therapeutic approaches
employing PIs in combination with TRAIL may repre-
sent a highly promising strategy to reactivate apoptosis
in therapy-resistant high-grade astrocytomas.
Experimental procedures
Materials
The caspase substrate acetyl-DEVD-7-amido-4-methyl-
coumarin (Ac-DEVD-AMC) was purchased from Bachem
(Heidelberg, Germany). MG132 and epoxomicin were pur-
chased from Sigma-Aldrich (Deisenhofen, Germany). JNK
inhibitor II (SP600125) was from Merck Biosciences
(Darmstadt, Germany). Recombinant human TRAIL was
from PeproTech Inc. (Rocky Hill, NJ, USA). All other
chemicals were of analytic grade purity and were from
Sigma-Aldrich (Deisenhofen, Germany).
Cell lines and culture
Human glioma cell lines U87, U251, U343, and U373, as
well as the newly established glioma cell lines, were main-
tained in DMEM with 10% heat-inactivated fetal bovine

serum, 100 U ÆmL
)1
penicillin, and 100 mgÆmL
)1
strepto-
mycin. Human glioma cell lines MZ-18 and MZ-54 were
established from a primary glioblastoma and a recurrent
grade IV tumor, respectively. To isolate glioma cells, tumor
specimens were homogenized, suspended in NaCl ⁄ P
i
and
centrifuged at 400 g. Pellets were resuspended in 10 mL of
medium, and plated into Petri dishes. Cultivation was per-
formed under standard conditions at 37 °C and a humidi-
fied 5% CO
2
atmosphere. When confluent monolayers had
been obtained, tumor-derived cells were trypsinized and
replated in new Petri dishes for serial passaging. All newly
established cell lines were analyzed for glial fibrillary acidic
protein expression by immunostaining with a monoclonal
glial fibrillary acidic protein IgG (R&D Systems,
Wiesbaden, Germany). Isotypic primary antibody (Serotec,
Du
¨
sseldorf, Germany) was used as control.
RT-PCR
Extraction of total cellular RNA, reverse transcription and
PCR were performed as previously described [40]. Primer
sequences were as follows: CHOP-sense, 5¢-GGT

CCT GTC TTC AGA TGA AAA TG-3¢; CHOP-antisense,
5¢-CCT GGT GCA GAT TCA CCA TTC-3¢; c-Jun-sense,
5¢-TGA CTG CAA AGA TGG AAA CG-3¢; c-Jun-antisense,
5¢-CCT GCT CAT CTG TCA CGT TC-3¢; DR4-sense,
5¢-AGA GAG AAG TCC CTG CAC CA-3¢;DR4-antisense,
5¢-AGA GAG AAG TCC CTG CAC CA-3¢; DR5-sense,
5¢-CAG AGG GAT TGT GTC CAC CT-3¢; DR5-anti-
sense, 5¢-TAC GGC TGC AAC TGT GAC TC-3¢; glyceral-
dehyde-3-phosphate dehydrogenase (GAPDH)-sense;
5¢-CCT GAC CTG CCG TCT AGA AA-3¢; GAPDH-anti-
sense, 5¢-TTA CTC CTT GGA GGC CAT GT-3¢.
SDS/PAGE and western blotting
SDS ⁄ PAGE and western blotting were performed as
described elsewhere [40]. The resulting blots were probed
with rabbit polyclonal anti-DR5 serum (Imgenex, San Diego,
CA, USA) diluted 1 : 200, rabbit polyclonal anti-DR4 serum
(ProSci, Poway, CA, USA) diluted 1 : 500, rabbit polyclonal
anti-phospho-c-Jun serum (Cell Signaling Technology,
Danvers, MA, USA) diluted 1 : 800, mouse monoclonal
anti-CHOP IgG diluted 1 : 200, or mouse monoclonal
anti-a-tubulin IgG (clone DM 1A; Sigma), diluted 1 : 5000.
Determination of caspase-3-like protease activity
DEVDase (caspase-3-like) activity was determined as previ-
ously described [40].
H. Hetschko et al. Apoptosis by TRAIL and proteasome inihibitors
FEBS Journal 275 (2008) 1925–1936 ª 2008 The Authors Journal compilation ª 2008 FEBS 1933
Flow cytometry
For cell death analysis, cells were stained with annexin-
V–FLUOS ⁄ propidium iodide (Roche Applied Science,
Mannheim, Germany), following the manufacturer’s

instructions. For analysis of DR5 surface expression, cells
were stained with goat polyclonal anti-DR5 IgG (Axxora,
Gru
¨
nberg, Germany) and a goat IgG isotype control
(SouthernBiotech, Birmingham, AL, USA), respectively,
according to the manufacturer’s instructions. In all cases, a
minimum of 10
4
events per sample were acquired. Flow
cytometric analyses were performed on a FACScan (BD
Biosciences; Heidelberg, Germany) followed by analysis
using cellquest and winmdi software.
Gene silencing using siRNA
The following annealed double-stranded siRNAs from
Dharmacon (Chicago, IL, USA) were used: CHOP siGe-
nome duplexes D-004819-01-0005 and D-004819-02-0005;
and DR5 siGenome duplexes D-004448-01-0005 and
D-004448-03-0005. Scrambled siRNA siCONTROL from
Dharmacon was used as a negative, nonsilencing control.
Cells were transfected with 100 nm siRNAs using siM-
PORTER (Biomol, Hamburg, Germany) as described by
the manufacturer.
Statistics
Data are given as means ± SEM. For statistical compari-
son, a t-test or one-way ANOVA followed by a Tukey test
were employed using spss software (SPSS GmbH Software,
Munich, Germany). P-values smaller than 0.05 were consid-
ered to be statistically significant.
Acknowledgements

The authors would like to thank Dr Sigrid Horn and
Monika Herr, Department of Neurosurgery, Mainz
University Clinics, for cell lines MZ-18 and MZ-54, Dr
Klaus Eberlein and Jussi Moog, Department of Radia-
tion Therapy and Radiation Oncology, Frankfurt
University Clinics, for help with the gamma irradiation
experiments, and Hildegard Schweers for excellent
technical assistance. This study was supported by
the Wilhelm Sander Stiftung (grant 2005.067.1) to
D. Ko
¨
gel and by Science Foundation Ireland to
J. H. M. Prehn.
References
1 Kleihues P, Louis DN, Scheithauer BW, Rorke LB,
Reifenberger G, Burger PC & Cavenee WK (2002) The
WHO classification of tumors of the nervous system.
J Neuropathol Exp Neurol 61, 215–225; discussion
226–219.
2 Maher EA, Furnari FB, Bachoo RM, Rowitch DH,
Louis DN, Cavenee WK & DePinho RA (2001)
Malignant glioma: genetics and biology of a grave
matter. Genes Dev 15, 1311–1333.
3 Bogler O & Weller M (2002) Apoptosis in gliomas, and
its role in their current and future treatment. Front
Biosci 7, e339–e353.
4 Bouralexis S, Findlay DM & Evdokiou A (2005) Death
to the bad guys: targeting cancer via Apo2L ⁄ TRAIL.
Apoptosis 10, 35–51.
5 Fulda S, Wick W, Weller M & Debatin KM (2002)

Smac agonists sensitize for Apo2L ⁄ TRAIL- or antican-
cer drug-induced apoptosis and induce regression of
malignant glioma in vivo. Nat Med 8, 808–815.
6 Hao C, Beguinot F, Condorelli G, Trencia A, Van Meir
EG, Yong VW, Parney IF, Roa WH & Petruk KC
(2001) Induction and intracellular regulation of tumor
necrosis factor-related apoptosis-inducing ligand
(TRAIL) mediated apotosis in human malignant glioma
cells. Cancer Res 61, 1162–1170.
7 Takeda K, Smyth MJ, Cretney E, Hayakawa Y,
Kayagaki N, Yagita H & Okumura K (2002) Critical
role for tumor necrosis factor-related apoptosis-induc-
ing ligand in immune surveillance against tumor
development. J Exp Med 195, 161–169.
8 Song JH, Song DK, Pyrzynska B, Petruk KC,
Van Meir EG & Hao C (2003) TRAIL triggers
apoptosis in human malignant glioma cells through
extrinsic and intrinsic pathways. Brain Pathol 13, 539–
553.
9 Adams J (2004) The proteasome: a suitable antineoplas-
tic target. Nat Rev Cancer 4, 349–360.
10 Richardson PG, Barlogie B, Berenson J, Singhal S,
Jagannath S, Irwin D, Rajkumar SV, Hideshima T,
Xiao H, Esseltine D et al. (2005) Clinical factors pre-
dictive of outcome with bortezomib in patients with
relapsed, refractory multiple myeloma. Blood 106,
2977–2981.
11 Aghajanian C, Soignet S, Dizon DS, Pien CS, Adams J,
Elliott PJ, Sabbatini P, Miller V, Hensley ML, Pezzulli
S et al. (2002) A phase I trial of the novel proteasome

inhibitor PS341 in advanced solid tumor malignancies.
Clin Cancer Res 8, 2505–2511.
12 Papandreou CN, Daliani DD, Nix D, Yang H, Madden
T, Wang X, Pien CS, Millikan RE, Tu SM, Pagliaro L
et al. (2004) Phase I trial of the proteasome inhibitor
bortezomib in patients with advanced solid tumors with
observations in androgen-independent prostate cancer.
J Clin Oncol 22, 2108–2121.
13 Wendt J, von Haefen C, Hemmati P, Belka C, Dorken
B & Daniel PT (2005) TRAIL sensitizes for ionizing
irradiation-induced apoptosis through an entirely
Apoptosis by TRAIL and proteasome inihibitors H. Hetschko et al.
1934 FEBS Journal 275 (2008) 1925–1936 ª 2008 The Authors Journal compilation ª 2008 FEBS
Bax-dependent mitochondrial cell death pathway. Onco-
gene 24, 4052–4064.
14 Asai A, Miyagi Y, Sugiyama A, Gamanuma M, Hong
SH, Takamoto S, Nomura K, Matsutani M, Takakura
K & Kuchino Y (1994) Negative effects of wild-type
p53 and s-Myc on cellular growth and tumorigenicity of
glioma cells. Implication of the tumor suppressor genes
for gene therapy. J Neurooncol 19, 259–268.
15 Badie B, Goh CS, Klaver J, Herweijer H & Boothman
DA (1999) Combined radiation and p53 gene therapy
of malignant glioma cells. Cancer Gene Ther 6, 155–162.
16 Rock KL, Gramm C, Rothstein L, Clark K, Stein R,
Dick L, Hwang D & Goldberg AL (1994) Inhibitors of
the proteasome block the degradation of most cell pro-
teins and the generation of peptides presented on MHC
class I molecules. Cell 78, 761–771.
17 Hanada M, Sugawara K, Kaneta K, Toda S, Nishiy-

ama Y, Tomita K, Yamamoto H, Konishi M & Oki T
(1992) Epoxomicin, a new antitumor agent of microbial
origin. J Antibiot (Tokyo) 45, 1746–1752.
18 Hetschko H, Voss V, Horn S, Seifert V, Prehn JH &
Ko
¨
gel D (2008) Pharmacological inhibition of Bcl-2
family members reactivates TRAIL-induced apoptosis
in malignant glioma. J Neurooncol 86, 265–272.
19 Yamaguchi H & Wang HG (2004) CHOP is involved in
endoplasmic reticulum stress-induced apoptosis by
enhancing DR5 expression in human carcinoma cells.
J Biol Chem 279, 45495–45502.
20 Yoshida T, Shiraishi T, Nakata S, Horinaka M,
Wakada M, Mizutani Y, Miki T & Sakai T (2005)
Proteasome inhibitor MG132 induces death receptor 5
through CCAAT ⁄ enhancer-binding protein homologous
protein. Cancer Res 65, 5662–5667.
21 Concannon CG, Koehler BF, Reimertz C, Murphy
BM, Bonner C, Thurow N, Ward MW, Villunger A,
Strasser A, Kogel D et al. (2007) Apoptosis induced by
proteasome inhibition in cancer cells: predominant role
of the p53 ⁄ PUMA pathway. Oncogene 26, 1681–1692.
22 Wu GS, Burns TF, McDonald ER III, Jiang W, Meng
R, Krantz ID, Kao G, Gan DD, Zhou JY, Muschel R
et al. (1997) KILLER ⁄ DR5 is a DNA damage-inducible
p53-regulated death receptor gene. Nat Genet 17,
141–143.
23 Meriin AB, Gabai VL, Yaglom J, Shifrin VI &
Sherman MY (1998) Proteasome inhibitors activate

stress kinases and induce Hsp72. Diverse effects on
apoptosis. J Biol Chem 273, 6373–6379.
24 Guyton KZ, Xu Q & Holbrook NJ (1996) Induction of
the mammalian stress response gene GADD153 by oxi-
dative stress: role of AP-1 element. Biochem J 314 (Pt 2),
547–554.
25 Rieger J, Naumann U, Glaser T, Ashkenazi A &
Weller M (1998) APO2 ligand: a novel lethal weapon
against malignant glioma? FEBS Lett 427, 124–128.
26 Amiri KI, Horton LW, LaFleur BJ, Sosman JA &
Richmond A (2004) Augmenting chemosensitivity of
malignant melanoma tumors via proteasome inhibition:
implication for bortezomib (VELCADE, PS-341) as a
therapeutic agent for malignant melanoma. Cancer Res
64, 4912–4918.
27 Masdehors P, Merle-Beral H, Maloum K, Omura S,
Magdelenat H & Delic J (2000) Deregulation of the
ubiquitin system and p53 proteolysis modify the
apoptotic response in B-CLL lymphocytes. Blood 96,
269–274.
28 He Q, Huang Y & Sheikh MS (2004) Proteasome inhib-
itor MG132 upregulates death receptor 5 and cooper-
ates with Apo2L ⁄ TRAIL to induce apoptosis in
Bax-proficient and -deficient cells. Oncogene
23, 2554–
2558.
29 Kabore AF, Sun J, Hu X, McCrea K, Johnston JB &
Gibson SB (2006) The TRAIL apoptotic pathway
mediates proteasome inhibitor induced apoptosis in
primary chronic lymphocytic leukemia cells. Apoptosis

11, 1175–1193.
30 Ganten TM, Haas TL, Sykora J, Stahl H, Sprick MR,
Fas SC, Krueger A, Weigand MA, Grosse-Wilde A,
Stremmel W et al. (2004) Enhanced caspase-8
recruitment to and activation at the DISC is critical for
sensitisation of human hepatocellular carcinoma cells to
TRAIL-induced apoptosis by chemotherapeutic drugs.
Cell Death Differ 11(Suppl. 1), S86–S96.
31 Qin JZ, Ziffra J, Stennett L, Bodner B, Bonish BK,
Chaturvedi V, Bennett F, Pollock PM, Trent JM,
Hendrix MJ et al. (2005) Proteasome inhibitors trigger
NOXA-mediated apoptosis in melanoma and myeloma
cells. Cancer Res 65, 6282–6293.
32 Wagenknecht B, Hermisson M, Eitel K & Weller M
(1999) Proteasome inhibitors induce p53 ⁄ p21-indepen-
dent apoptosis in human glioma cells. Cell Physiol
Biochem 9, 117–125.
33 Yin D, Zhou H, Kumagai T, Liu G, Ong JM, Black
KL & Koeffler HP (2005) Proteasome inhibitor PS-341
causes cell growth arrest and apoptosis in human glio-
blastoma multiforme (GBM). Oncogene 24, 344–354.
34 Karin M & Ben-Neriah Y (2000) Phosphorylation
meets ubiquitination: the control of NF-[kappa]B activ-
ity. Annu Rev Immunol 18, 621–663.
35 La Ferla-Bruhl K, Westhoff MA, Karl S, Kasperczyk
H, Zwacka RM, Debatin KM & Fulda S (2007) NF-
kappaB-independent sensitization of glioblastoma cells
for TRAIL-induced apoptosis by proteasome inhibition.
Oncogene 26, 571–582.
36 Hideshima T, Chauhan D, Richardson P, Mitsiades C,

Mitsiades N, Hayashi T, Munshi N, Dang L, Castro
A, Palombella V et al. (2002) NF-kappa B as a
therapeutic target in multiple myeloma. J Biol Chem
277, 16639–16647.
H. Hetschko et al. Apoptosis by TRAIL and proteasome inihibitors
FEBS Journal 275 (2008) 1925–1936 ª 2008 The Authors Journal compilation ª 2008 FEBS 1935
37 Ganten TM, Koschny R, Haas TL, Sykora J, Li-Weber
M, Herzer K & Walczak H (2005) Proteasome
inhibition sensitizes hepatocellular carcinoma cells, but
not human hepatocytes, to TRAIL. Hepatology 42,
588–597.
38 Kurland JF & Meyn RE (2001) Protease inhibitors
restore radiation-induced apoptosis to Bcl-2-expressing
lymphoma cells. Int J Cancer 96, 327–333.
39 Minet E, Michel G, Mottet D, Piret JP, Barbieux A,
Raes M & Michiels C (2001) c-JUN gene induction and
AP-1 activity is regulated by a JNK-dependent pathway
in hypoxic HepG2 cells. Exp Cell Res 265, 114–124.
40 Ko
¨
gel D, Schomburg R, Copanaki E & Prehn JH
(2005) Regulation of gene expression by the amyloid
precursor protein: inhibition of the JNK ⁄ c-Jun path-
way. Cell Death Differ 12, 1–9.
Apoptosis by TRAIL and proteasome inihibitors H. Hetschko et al.
1936 FEBS Journal 275 (2008) 1925–1936 ª 2008 The Authors Journal compilation ª 2008 FEBS