MOLECULAR AND CELLULAR BIOLOGY, Dec. 2010, p. 5726–5740
Copyright © 2010, American Society for Microbiology. All Rights Reserved.
Vol. 30, No. 24
Cdk1/Cyclin B1 Controls Fas-Mediated Apoptosis by
Regulating Caspase-8 Activity?
Yves Matthess,1† Monika Raab,1† Mourad Sanhaji,1Inna N. Lavrik,2and Klaus Strebhardt1*
Department of Obstetrics and Gynecology, School of Medicine, J. W. Goethe University, Theodor-Stern-Kai 7, 60590 Frankfurt,
Germany,1and Division of Immunogenetics, Tumorimmunology Program, German Cancer Research Center,
Im Neuenheimer Feld 280, 69120 Heidelberg, Germany2
Received 24 June 2010/Returned for modification 15 August 2010/Accepted 27 September 2010
Caspase activation is a hallmark of apoptosis. However, the molecular mechanisms underlying the regula-
tion of caspase-8 activation within the extrinsic death pathway are not well understood. In this study, we
demonstrate that procaspase-8 is phosphorylated in mitotic cells by Cdk1/cyclin B1 on Ser-387, which is
located at the N terminus of the catalytic subunit p10. This phosphorylation of procaspase-8 on Ser-387 occurs
in cancer cell lines, as well as in primary breast tissues and lymphocytes. Furthermore, RNA interference-
mediated silencing of cyclin B1 or treatment with the Cdk1 inhibitor RO-3306 enhances the Fas-mediated
activation and processing of procaspase-8 in mitotic cells. A nonphosphorylatable procaspase-8 (S387A)
facilitates Fas-induced apoptosis during mitosis. Our findings suggest that Cdk1/cyclin B1 activity shields
human cells against extrinsic death stimuli and unravel the molecular details of the cross talk between cell
cycle and extrinsic apoptotic pathways. Finally, this new mechanism may also contribute to tumorigenesis.
Intact apoptotic machinery is essential to maintain the in-
tegrity and homeostasis of multicellular organisms (8). The
evasion of apoptosis is a hallmark of cancer, as demonstrated
by the inability of cancer cells to respond appropriately to
signals that normally control unrestricted growth (11). In mam-
malian cells, the apoptotic response is mediated by either the
intrinsic or the extrinsic pathway, depending on the origin of
the death stimulus. After stimulation of the death receptor Fas
(APO-1/CD95) the death-inducing signaling complex (DISC)
assembles, which contains the oligomerized receptor, the adap-
tor molecule FADD, two isoforms of procaspase-8 (pro-
caspase-8a and -8b), procaspase-10, and c-FlipL/S/R(28, 37). In
accordance with the induced proximity model, immediately
after DISC formation, procaspase-8, which consists of two
death effector domains (DED) and a protease domain contain-
ing the p18 and p10 subunits, is proteolytically processed (28).
This autoprocessing follows a sequential order of events: while
the first cleavage step occurs at Asp-374 and results in the
formation of the subunits p43/p41 and p12, the second cleav-
age at Asp-216 and Asp-384 produces the enzymatically active
subunits p18, p10, and the prodomain p26/p24 (5, 14, 27, 40).
The mature caspase-8 heterotetramer p182-p102then translo-
cates from the DISC to the cytosol, where it cleaves several
substrates, such as Bid, and effector caspases to initiate the
apoptotic cascade (22).
Increasing evidence highlights the functional importance of
procaspase-8 for carcinogenesis; several findings suggest that
the impairment of procaspase-8 function by genetic and epi-
genetic mechanisms correlates with the malignant potential of
different types of cancer (6, 12, 43, 44). In the present study, we
investigated the functional correlation of the cell cycle with the
extrinsic death pathway. The cyclin-dependent kinase 1 (Cdk1)
in complex with cyclin B1 (Cdk1/cyclin B1) is one of the key
mitotic kinases. The kinase activity of Cdk1/cyclin B1 governs
the entry into mitosis from G2phase of the cell cycle (29, 30).
Through mediating phosphorylation of a variety of substrates,
Cdk1/cyclin B1 also plays an important role in multiple pro-
cesses during mitosis, including chromosome condensation,
nuclear envelope breakdown, centrosome separation, regula-
tion of spindle microtubule dynamics, and metaphase-to-an-
aphase transition (4, 21, 31, 36). In the present investigation,
our molecular analyses of the roles that cell cycle kinases play
in the apoptosis signaling pathway demonstrated that Cdk1/
cyclin B1 and procaspase-8 interact in vitro and in vivo. The
phosphorylation of procaspase-8 by Cdk1/cyclin B1 at Ser-387
inhibits processing of procaspase-8. In a physiological situa-
tion, such as breast cancer and proliferating T cells, where
Cdk1/cyclin B1 activity is upregulated, we observed elevated
procaspase-8 phosphorylation on Ser-387. We suggest a model
in which Cdk1/cyclin B1 helps to protect mitotic cells against
extrinsic death stimuli.
MATERIALS AND METHODS
Antibodies. Mouse monoclonal antibodies against Fas (CD95/Apo-1, clone
2R2 [Calbiochem]), ?-actin (clone AC-15), Flag tag (FlagM2 F1804 [Sigma]),
Plk1 (F-8, sc-17783), Cdk1 (17, sc-54), cyclin B1 (GNS1, sc-245), glutathione
S-transferase (GST; B-15, sc138 [Santa Cruz]), phospho-histone H3 Ser10 (clone
3H10 [Millipore]), V5 (Invitrogen), and caspase-8 (1C12 [Cell Signaling Tech-
nology] and C15 [Alexis]) and a rabbit polyclonal antibody against poly(ADP-
ribose) polymerase (PARP) (46D11 [Cell Signaling Technology]) were used
according to the manufacturers’ recommendations. To generate a polyclonal
antibody against caspase-8 phosphorylated on Ser-387, rabbits were immunized
by Eurogentec using the peptide YLEMDLSpSPQTRY. According to the im-
munization procedure, antibodies were affinity purified. Secondary antibodies
conjugated to CY2 and CY3 were obtained from Jackson Immunoresearch Labs.
Cell culture, transfection, and synchronization. Cell lines were purchased
from the American Type Culture Collection and grown in the recommended
medium supplemented with 10% heat-inactivated fetal calf serum at 37°C in a
* Corresponding author. Mailing address: Department of Obstetrics
and Gynecology, School of Medicine, J. W. Goethe University, Theodor-
Stern-Kai 7, 60590 Frankfurt, Germany. Phone: (49) 69-6301-6894. Fax:
(49) 69-6301-6364. E-mail: email@example.com.
† Y.M. and M.R. contributed equally to this study.
?Published ahead of print on 11 October 2010.
humidified environment with 5% CO2. Human lymphocytes from peripheral
blood were cultured as described previously (13).
To generate knockdown clones for cyclin B1 and caspase-8, cells were trans-
fected with plasmids in which the H1 promoter drives the expression of an
shRNA targeting the open reading frame of caspase-8 (5?-CCTGGTACATCC
AGTCACT-3?) or cyclin B1 (5?-AAGAAATGTACCCTCCAGA-3?). Stably
transfected cells were selected for 4 weeks using 1.5 mg of G418/ml (PAA
Laboratories). Transfection and synchronization were performed as described
previously (48). For the induction of apoptosis, cells were stimulated with FasL
(100 ng/ml), TRAIL (10 ng/ml), or tumor necrosis factor alpha (TNF-?; 10
ng/ml) (Enzo Life Sciences).
Generation of plasmids and mutagenesis. Sequences of the primers used in
the present study will be provided upon request. Caspase-8 was inserted into the
pGEX (GE Healthcare) and the 3?Flag-tagged pcDNA3.1-Hygro? (Invitro-
gen) vectors. The N-terminal domain was cloned into the BamHI/EcoRI sites of
the pGEX vector. The 3?Flag-tagged caspase-8 sequence was inserted into the
pcDNA3.1-V5 vector (Invitrogen). Deletion clones were generated by standard
techniques. Mutants were generated by site-directed mutagenesis using the
QuikChange protocol and Pfu Ultra II Fusion HS DNA polymerase (Strat-
agene). RNA interference (RNAi) vectors were constructed as described previ-
ously (18). Sequences for efficient silencing of caspase-8 and cyclin B1 were
obtained using our web-based shRNA design tool (www.molgyn.kgu.de
/genesilencer). All constructs were confirmed by sequencing. Information on
cloning procedures is available from the authors.
Immunoprecipitation and phospho amino acid analysis. Both methods were
performed as described previously (34).
GST pulldown. The expression of recombinant glutathione S-transferase
(GST) proteins was induced in Escherichia coli BL21 cells at 37°C for 2 h by the
addition of 1 mM IPTG (isopropyl-?-D-thiogalactopyranoside) (33). GST-fused
proteins were purified first and then incubated with lysates of HeLa cells trans-
fected with the Flag-Cdk1 expression vector in TBSN buffer (20 mM Tris [pH
8.0], 150 mM NaCl, 0.5% Nonidet P-40, 5 mM EGTA, 0.5 mM Na3VO4, 20 mM
p-nitrophenyl phosphate) supplemented with protease inhibitors (Complete
Mini EDTA-free Protease Inhibitor Cocktail; Roche Diagnostics) at 4°C for 2 h.
GST, GST–caspase-8, and GST-fused subdomains of caspase-8 were adsorbed to
glutathione-Sepharose 4B (GE Healthcare) for an additional 1 h. The bound
proteins were resolved by SDS-PAGE and transferred to an Immobilon-P mem-
brane (Millipore). The blots were then probed with anti-Flag antibodies to detect
the Flag-Cdk1 protein.
Cell cycle analysis. Cell cycle analysis was carried out as described previously
(49). Briefly, cells were washed with phosphate-buffered saline (PBS), treated
with RNase A, and stained with propidium iodide. Flow cytometry was per-
formed by using a FACScan machine and CellQuest software (Becton Dickin-
son). The determination of the mitotic indices and mitotic phase distribution was
performed as previously described (35). All experiments were performed in
Immunofluorescence assays. For immunofluorescence analysis, HeLa cells
were grown on a glass coverslip, fixed with 4% paraformaldehyde, washed with
PBS, and permeabilized with 0.2% Triton X-100 before addition of the appro-
priate primary and secondary antibodies (33). Microscopy was performed with a
Zeiss Axio Imager Z1 microscope equipped with a 40? objective lens, and the
images were captured and processed by using AxioVision Software (Zeiss).
Analysis of apoptosis. Cells were processed by using a Vybrant apoptosis assay
kit 2 (Alexa Fluor 488-annexin V/propidium iodide staining) according to the
manufacturer’s protocol (Invitrogen) and analyzed by flow cytometry using a
FACScan (Becton Dickinson).
Cdk1/cyclin B1 and caspase-8 assays. Immunoprecipitations with a cyclin
B1-specifc antibody to measure the kinase activity of Cdk1/cyclin B1 were per-
formed as described previously (48). In vitro kinase assays were performed using
10? Cdk1 buffer (New England Biolabs) supplemented with 0.05 mM ATP and
1 ?Ci of [?-32P]ATP (3,000 Ci/mmol; Amersham Pharmacia) for 30 min at 30°C
in the presence of bacterially expressed purified GST–caspase-8 fusion proteins.
For the inhibition of Cdk1 activity in kinase assays, RO-3306 was diluted 3-fold
in assay buffer in a final volume of 20 ?l, and the assay was initiated by the
addition of 40 ?l of assay buffer containing the caspase-8 substrate. Samples were
resolved by SDS-PAGE and subjected to autoradiography. Phosphate groups
were removed by incubation with ? phosphatase (New England Biolabs).
Caspase-8 activity was determined by using a Caspase Glo 8 assay kit (Promega)
according to the manufacturer’s instructions.
Patients and tumor samples. The study group consisted of 13 patients who
were treated at the Department of Obstetrics and Gynecology of J. W. Goethe
University School of Medicine (Frankfurt, Germany). All breast cancer patients
met the following major inclusion criteria: unilateral primary carcinoma of the
breast confirmed histologically by core-cut needle or incisional biopsy (fine-
needle aspiration was not considered sufficient); a tumor measurable two dimen-
sionally by mammography, breast ultrasound, or breast magnetic resonance
imaging; a primary tumor ?3 cm at its largest diameter (in patients with multi-
focal or multicentric breast cancer, the largest lesion was measured); and a
patient age between 18 and 70 years. In all cases, vital tumor biopsy specimens
from the primary site were removed during surgery, immediately frozen in liquid
nitrogen, and stored at ?80°C. All histological evaluations were carried out by
two independent investigators.
Statistical methods. Experimental in vitro data are presented as mean ? the
standard deviations from three or more independent experiments. Two-way
analysis of variance (ANOVA; GraphPad Prism; GraphPad Software, Inc., San
Diego, CA) was done to consider random effects of individual gels and different
treatments. For two-way ANOVAs, all treatment groups were compared to
Signaling via the extrinsic death pathway is attenuated in
mitotic cells. The apoptosis signaling pathway was evaluated in
mitotic and nonmitotic B lymphoblastoid SKW 6.4 cells to
determine whether the Fas-induced cellular response is regu-
lated during the cell cycle. Cells were treated once with excess
thymidine to accumulate the majority of them at G1/S (Fig. 1a,
lane 1). In addition, we performed the analyses in the presence
of the microtubule depolymerizing drug nocodazole combined
with MG132 to prevent mitotic exit, as indicated by the phos-
phorylation of histone H3 at Ser10 (pH3S10), by the high
abundance of polo-like kinase 1 (Plk1) and cyclin B1 (Fig. 1a,
lane 2), and by fluorescence-activated cell sorting analysis (Fig.
1a, lower panel). When we compared the regulation of the
extrinsic death pathway status in mitotic and nonmitotic cells,
we noted that Fas-mediated procaspase-8 processing is im-
paired in mitotic cells: procaspase-8a/b was barely processed,
and reduced levels of the previously described procaspase-8a/b
cleavage products p43/p41 and p18 were observed (Fig. 1b)
(37). To investigate this mechanism in other caspase-8-depen-
dent pathways, we used the death ligands TRAIL or TNF-?
(Fig. 1b). Again, upon stimulation with TRAIL or TNF-?
procaspase-8a/b processing was also impaired in mitotic cells
compared to nonmitotic control cells. Caspase-8 activity was
also significantly blocked following stimulation with different
death ligands (FasL, TRAIL, or TNF-?) in a mitotic cell-
enriched population (Fig. 1c). These results prompted us to
investigate the regulation of procaspase-8 in mitotic cells dur-
ing Fas-induced apoptosis.
Although it has been reported that overexpressed kinases
can protect cells against apoptosis (11), the underlying mech-
anisms, particularly those regulating the extrinsic death path-
way, remain elusive. Interestingly, the inhibition of certain cell
cycle kinases, including Cdks, polo-like kinase 1 (Plk1), and
Aurora kinases, which often exhibit elevated activity in human
cancers, triggers apoptosis in many experimental contexts (13,
20, 24, 39, 41, 42). To shed more light on the role cell cycle
kinases play for the regulation of the extrinsic death pathway,
we used at first the selective ATP-competitive inhibitor RO-
3306. This small molecule inhibits Cdk1/cyclin B1 activity with
a 10-fold selectivity compared to Cdk2/cyclin E and ?50-fold
relative to Cdk4/cyclin D (45). The treatment of mitotic SKW
6.4 cells with RO-3306 restored the sensitivity to Fas stimula-
tion as indicated by enhanced enzymatic activity and process-
ing of procaspase-8 with elevated levels of p43/p41 and p18
VOL. 30, 2010CDK1/CYCLIN B1 REGULATES CASPASE-8 ACTIVITY5727
FIG. 1. Inhibition of the extrinsic death pathway in Fas-induced mitotic cells. (a) B lymphoblastoid SKW6.4 cells were enriched in G1/S phase
by treatment with thymidine (lane 1), in pro-/metaphase with nocodazole for 16 h, followed by MG132 for 2 h (lane 2) or in a G1-like state with
nocodazole for 16 h, followed by MG132 for 2 h and RO-3306 for the inhibition of Cdk1/cyclin B1 (lane 3) (upper panel). Cell lysates were
immunoblotted for cyclin B1, histone H3 phosphorylated at Ser10 (H3 pS10), cyclin E, Plk1, and ?-actin. The cell cycle status was monitored by
flow cytometry (lower panel). (b) Cells treated as described in panel a were stimulated with 100 ng of FasL/ml, 10 ng of TRIAL/ml, or 10 ng of
TNF-?/ml for the times indicated, and total cellular lysates were analyzed by Western blotting with anti-caspase-8 MAb C15. (c) Caspase-8 activity
was determined by measuring the cleavage of the luminogenic substrate containing the IETD peptide. All experiments were performed in
triplicate. Error bars represent the standard deviation (SD). (d) Apoptosis analyses were performed by using an annexin V kit. All experiments
were performed in triplicate. Error bars represent the SD.
(Fig. 1a to c). Most interestingly, RO-3306 enhanced the sus-
ceptibility of mitotic cells to Fas-induced cell death (Fig. 1d).
These findings pointed out the possible role of Cdk1/cyclin B1
in the regulation of Fas-mediated procaspase-8 activation.
Cdk1/cyclin B1 phosphorylates procaspase-8 on Ser-387 in
vitro. To explore the role of Cdk1/cyclin B1 in the modulation
of the extrinsic death pathway in more detail, we started to
directly investigate the possible interactions between pro-
caspase-8 and Cdk1/cyclin B1. Recombinant subdomains of
human procaspase-8 expressed as GST fusion proteins were
analyzed in kinase assays to determine whether caspase-8 is an
in vitro substrate of Cdk1/cyclin B1 (Fig. 2a, upper and middle
panels). Under these conditions, only the p10 subdomain was
labeled (Fig. 2a, middle panel). To investigate whether this
phosphorylation occurs on a common residue, the labeled
band was excised and analyzed. The amino acid analysis dem-
onstrated that the site(s) phosphorylated by Cdk1/cyclin B1
was mainly serine residues (Fig. 2a, lower right panel). A
putative Cdk1 phosphorylation site within p10, Ser-387, is fol-
lowed immediately by a proline residue; this is characteristic of
sites phosphorylated by kinases such as Cdk1. Point mutants of
p10 expressed as GST fusion proteins (S387A, SS386/387AA,
FIG. 2. Cdk1/cyclin B1 phosphorylates procaspase-8 in vitro at Ser-387. (a) Recombinant GST-fused subdomains of procaspase-8 were analyzed
by SDS-PAGE (lower left panel). Subdomains were incubated with active Cdk1/cyclin B1 and [?-32P]ATP, analyzed by SDS-PAGE, and visualized
by autoradiography (upper left panel). This was followed by phospho amino acid analysis (lower right panel). (b) Wild-type (WT) and mutated
(S387A, S387E) recombinant full-length GST-caspase-8 proteins were subjected to kinase assays using Cdk1/cyclin B1, analyzed by SDS-PAGE,
and autoradiography. (c) Recombinant caspase-8 (WT) was incubated with active Cdk1/cyclin B1. Samples were immunoblotted with antibodies
specific for caspase-8, for a caspase-8 peptide [ABcasp.-8(pSer387)] containing phospho-Ser387 (YLEMDLSpSPQTRY) and for Cdk1. (d) GST,
recombinant caspase-8 (GST-WT), and a mutated form (GST-S387A) were incubated with or without active Cdk1/cyclin B1 and immunoblotted
with ABcasp.-8(pSer387) (e) Recombinant caspase-8 was incubated with or without active Cdk1/cyclin B1 in kinase buffer, followed by treatment
with 100 U of ?-phosphatase and immunoblotted with ABcasp.-8(pSer387).
VOL. 30, 2010CDK1/CYCLIN B1 REGULATES CASPASE-8 ACTIVITY5729
and T419A) were analyzed in kinase assays, and the substitu-
tion of Ser-387 led to a strong inhibition of Cdk1/cyclin B1-
mediated phosphorylation of p10 (Fig. 3a). To further deter-
mine whether procaspase-8 is a bona fide substrate of Cdk1/
cyclin B1, we examined the phosphorylation of the full-length
protein. Site-directed mutagenesis of Ser-387 followed by in
vitro kinase assays, did not block the phosphorylation of full-
length procaspase-8 completely but led to a massive downregu-
lation of the phosphorylation signal, suggesting that Ser-387 is
the major phosphorylation site for Cdk1/cyclin B1 in full-
length procaspase-8 (Fig. 2b). We generated a phospho-S387-
specific antibody to explore the significance of Ser-387 phos-
phorylation in vitro and in vivo. Phosphorylation by Cdk1/cyclin
B1 was required for the pS387-specific antibody to recognize
procaspase-8 (Fig. 2c). Furthermore, the mutation of Ser-387
to Ala or the treatment of phosphorylated wild-type pro-
caspase-8 with ?-phosphatase abolished this recognition com-
pletely, validating the specificity of our affinity-purified anti-
body for the Ser387-phosphorylated form of procaspase-8 (Fig.
2d and e).
Next, we analyzed whether procaspase-8 and Cdk1/cyclin B1
interact in vitro. Using lysates of cells transfected with Flag-
Cdk1 expression constructs, we performed pulldown assays
with procaspase-8 or its subdomains fused to GST to deter-
mine the regions involved in the association of both proteins in
vitro. Flag-Cdk1 was associated with procaspase-8, its N-termi-
nal portion (GST-NT), and all fragments of GST-NT compris-
ing the linker region between DED1 and DED2 (Fig. 3b and
c). We concluded that the region connecting both DED do-
mains is involved in binding to Cdk1/cyclin B1.
Inhibition of Cdk1/cyclin B1 activity impairs the phosphor-
ylation of procaspase-8 at Ser-387 in vivo. We evaluated
whether Cdk1/cyclin B1 and procaspase-8 also interact in vivo.
In contrast to extracts of asynchronous HeLa cells, specific
complexes containing procaspase-8, Cdk1, and cyclin B1 were
detected following the immunoprecipitation of mitotic cell ex-
tracts using caspase-8-, Cdk1-, or cyclin B1-specific antibodies
(Fig. 4a). Next, we used indirect immunofluorescence to
visualize the subcellular localization of caspase-8, caspase-8
(pS387), and cyclin B1. Whereas caspase-8 was found predom-
inantly in the cytoplasm, the Ser-387 phosphorylated form was
also found in the cytoplasm associated with intense staining of
centrosomes from prometaphase to anaphase (Fig. 4b, upper
panels). Cyclin B1 also localized to the centrosomes and mi-
crotubules, confirming previous findings (3, 16). The merged
image demonstrates a colocalization of caspase-8 (pS387) and
cyclin B1 at the centrosomes. Depletion of caspase-8 by RNAi
abolished the staining of cytoplasm and centrosomes using our
phospho-caspase-8-specific antibodies completely (Fig. 4b,
To further address the significance of S387 phosphorylation
during mitosis, cells treated with nocodazole and MG132 were
FIG. 3. Cdk1/cyclin B1 associates with and phosphorylates procaspase-8 in vitro at Ser-387. (a) The p10 subunit was mutated at different sites
(S387A, SS386/387AA, and T419A), and the corresponding GST fusion proteins were incubated with active Cdk1/cyclin B1 and [?-32P]ATP,
analyzed by SDS-PAGE and visualized by autoradiography. (b) Bacterially expressed, purified GST-fused full-length caspase-8 (GST-WT), and the
N-terminal (GST-NT) and C-terminal (GST-CT) subdomains of caspase-8 were incubated with lysates of HeLa cells that had been transfected with
a Flag epitope-tagged Cdk1 (Flag-Cdk1) expression construct for pulldown assays. The Flag-Cdk1 protein that associated with GST-caspase-8 or
its subdomains was detected by immunoblot with an anti-Flag antibody. (c) The binding of the GST-fused N-terminal subdomains (GST-2xDED,
-DED1?, and -DED1) was also analyzed in pulldown assays. The Flag-Cdk1 protein that associated with GST-fused subdomains of the N-terminal
portion of caspase-8 was detected by immunoblotting with an anti-Flag antibody. A 5% total input lane was also run.
5730 MATTHESS ET AL.MOL. CELL. BIOL.
collected by mitotic shake-off. Although procaspase-8 showed
strong S387 phosphorylation only in mitotic cells, the treat-
ment of shake-off cells with the Cdk1 inhibitor RO-3306 led to
an almost complete loss of the phosphorylation signal (Fig. 4c).
We also determined the kinetics of procaspase-8 phosphory-
lation in vivo. The quality of the synchronization of the double
thymidine-blocked HeLa cells was monitored by analyzing the
expression of cyclin B1 and Plk1, the phosphorylation of his-
tone H3 (Fig. 4d, upper left panel), and the mitotic index (Fig.
4d, lower right panel), as well as by flow cytometry (Fig. 4d,
upper right panel). In these cells, the peak levels of pro-
caspase-8 phosphorylation at Ser-387 at 8 to 10 h after release
from the double-thymidine block (Fig. 4d, upper left panel)
coincided with maximum activity of the Cdk1/cyclin B1 com-
plex (Fig. 4d, lower left panel). In semisynchronized colon
(SW480) and breast cancer (T47D, MDA-MB-231) cells, the
phosphorylation of procaspase-8 also peaked in mitosis, sup-
porting a correlation between Cdk1/cyclin B1 activity and pro-
caspase-8 Ser-387 phosphorylation in cell lines of different
origins (Fig. 5a).
To investigate this correlation in more detail, we depleted
cyclin B1 using our previously developed RNAi vectors (18, 19,
26, 49). Because the complete small interfering RNA (siRNA)-
mediated knockdown of cyclin B1 in our earlier studies in-
duced apoptosis in a number of cancer cell types (49), we
selected stable HeLa clones that exhibited only a partial knock-
down of cyclin B1 (up to 80%) (Fig. 4e and 5b). Furthermore,
consistent with the downregulation of the cyclin B1 protein,
the kinase activity of Cdk1/cyclin B1 was decreased to ca. 20%
in the cellular extract of the analyzed Cycl.B1?cell clones
compared to mock-transfected cells (Fig. 4e and Fig. 5b),
which confirms our previous results (48). Similar Cdk1/cyclin
B1-kinase activities were also measured in different Cycl.B1?
clones (data not shown). Upon Fas stimulation of mitotic
Cycl.B1?cells procaspase-8a/b processing products p43/p41
and p18 were significantly upregulated compared to mitotic
mock-transfected cells (Fig. 4e, upper panel, and Fig. 5b, left
panel). Upon Fas stimulation, cyclin B1-depleted mitotic cells
also showed an elevated level of apoptosis compared to their
mock-transfected mitotic counterparts (Fig. 4e, lower right
panel). By knocking down Cdk1 by RNAi, we could also sen-
sitize mitotic cells to Fas-mediated apoptosis (Fig. 5b, right
To further evaluate the role of Cdk1/cyclin B1 activity for
the phosphorylation of procaspase-8 at Ser-387 and its pro-
cessing, we used the Cdk1 inhibitor RO-3306 in vitro and in
vivo (45). The level of procaspase-8 (pS387) correlated in-
versely with the concentration of RO-3306 in our in vitro kinase
assays (Fig. 6a). By treating proliferating cells for 18 h with
RO-3306, we confirmed the previous finding that RO-3306
arrests cells at the G2/M border. Phospho-histone H3 staining
demonstrated that mitotic cells appeared as early as 10 min
after drug removal, coinciding with the appearance of phos-
phorylated procaspase-8 (pS387) in SW480 cells (Fig. 6b). This
suggested that the enzymatic activity of Cdk1/cyclin B1 is im-
portant for procaspase-8 phosphorylation at Ser-387. To fur-
ther study the correlation of Cdk1 activity and processing of
procaspase-8 in living cells, we analyzed protein extracts of
mitotic B lymphoblastoid SKW 6.4 cells treated with FasL in
the absence or presence of the Cdk inhibitor RO-3306. Fas
stimulation led to increased levels of p43/p41 and p18 in mi-
totic cells with downregulated Cdk1/cyclin B1 activity com-
pared to mitotic control cells (Fig. 4f, upper panel). Treatment
of mitotic SKW 6.4 cells with RO-3306 also induced enhanced
apoptosis upon Fas stimulation compared to untreated mitotic
SKW 6.4 cells (Fig. 4f, lower panel). Interestingly, we observed
that an increase in caspase-8 activation upon the addition of
RO-3306 occurred in type I, e.g., B lymphoblastoid SKW 6.4
cells. Caspase-8 activation, which triggers the apoptotic cas-
cade in type I cells, takes place directly at the DISC, indepen-
dently of mitochondria and caspase-9. Therefore, experiments
in SKW6.4 cells demonstrate that RO-3306 influences
caspase-8 activation directly at the DISC independently of
Mimicking the phosphorylation at Ser-387 modulates pro-
cessing of procaspase-8 and Fas-mediated apoptosis in mam-
malian cells. Interestingly, Ser-387 is located at the N-terminal
end of p10 close to the cleavage residue Asp-384, which, in
conjunction with Asp-374, facilitates the separation of the two
catalytic subunits. To evaluate whether the phosphorylation of
Ser-387 by Cdk1 can affect the processing and concomitantly
the stability of procaspase-8 upon Fas stimulation in nonsyn-
chronized cells, HeLa cells were transfected with Flag-tagged
procaspase-8 wild type (WT) or the procaspase-8 mutants
S387A and S387E. Extracts from Fas-stimulated transfected
HeLa cells were analyzed by immunoblotting with a Flag-
specific antibody. Whereas the S387A mutant, which mimics
the nonphosphorylated form, disappeared rapidly after Fas
stimulation, the S387E mutant, which mimics the phosphory-
lated form, was more stable (Fig. 7a).
In our previous experiments we demonstrated that Cdk1
phosphorylates caspase-8 at Ser-387 in mitotic cells. Since
caspase-8 is the main initiator of the extrinsic death pathway, it
would be interesting to monitor first the sensitivity of nonmi-
totic cells expressing either the nonphosphorylatable form
S387A or the phosphomimic form S387E to Fas-induced apop-
tosis. We generated double-tagged (N-terminal Flag, C-termi-
nal V5) mutated versions of caspase-8 and monitored the ki-
netics of p12/p10 production as an early step of procaspase-8
processing. The p12/p10 subunit was detectable in HeLa cells
transfected with caspase-8 (S387A) expression constructs
within 1 h of Fas stimulation, and its level increased after 2 to
3 h (Fig. 7b, upper panel). In contrast, V5-tagged p12/p10 was
at the limit of detection in HeLa cells transfected with the
caspase-8 (S387E) expression construct and became visible
only after prolonged exposure. PARP cleavage in these cells
indicated efficient Fas-induced signaling via endogenous
caspase-8 (Fig. 7b, lower panel). To evaluate the correlation
between procaspase-8 processing as measured by p12/p10
cleavage and apoptosis in a rescue experiment of nonsynchro-
nized cells, caspase-8-depleted cells (caspase-8?) were trans-
fected with the constructs expressing double-tagged caspase-8
(S387A or S387E) mutants (Fig. 7c). The S387A mutant ex-
pressed in caspase-8-depleted cells was processed with the
passing of time upon Fas stimulation, as indicated by decreas-
ing amounts of procaspase-8 and increasing levels of p12/p10
(Fig. 7c, upper and middle panels). First signs of procaspase-8
processing correlated with the onset of apoptosis as measured
by PARP cleavage (Fig. 7c, lower panel). In contrast, the
S387E mutant was not processed and blocked the apoptotic
VOL. 30, 2010CDK1/CYCLIN B1 REGULATES CASPASE-8 ACTIVITY 5731
FIG. 4. The phosphorylation of procaspase-8 correlates with Cdk1/cyclin B1 activity in vivo. (a) Lysates from asynchronous (?) or mitotic (?)
HeLa cells were immunoprecipitated with caspase-8-, Cdk1-, or cyclin B1-specific antibodies. Anti-mouse IgG was used as a control. The
immunoprecipitated proteins were resolved by SDS-PAGE and immunoblotted for caspase-8, cyclin B1, and Cdk1. Lanes 1 and 2 show 2.5% total
input. (b) HeLa cells were arrested in G1/S using a double-thymidine block and released into fresh medium for 10 h. Semisynchronized cells were
stained with DAPI (4?,6?-diamidino-2-phenylindole) and subsequently categorized according to their staining patterns. Cells were labeled using
5732 MATTHESS ET AL.MOL. CELL. BIOL.
response in caspase-8-depleted cells upon Fas stimulation (Fig.
7c). These data indicate that both mutants displayed a different
apoptotic response even in nonmitotic cells where the activity
of Cdk1 is nearly inexistent, showing that the replacement of
endogenous procaspase-8 by a procaspase-8 mutant mimicking
phosphorylation at S387 inhibits the extrinsic death pathway
even in an unsynchronized cell population. This corroborates
our previous findings that the mitotic phosphorylation of
caspase-8 by Cdk1 at Ser-387 blocks the Fas-induced apoptotic
Still, since our data demonstrate that the phosphorylation of
procaspase-8 at Ser-387 occurs exclusively during mitosis and is
mediated by Cdk1/cyclin B1, we focused in subsequent exper-
iments on the role of procaspase-8 phosphorylation in mitotic
cells. First, we sought to determine whether phosphorylation of
Ser-387 controls Fas-induced apoptosis in mitotic cells. We
used a strategy based on the depletion of endogenous
caspase-8 by siRNA targeting the untranslated region and its
replacement in mitotic cells by similar amounts of transfected
caspase-8, either with or without mutation of Ser387 to a non-
phosphorylatable Ala residue (Fig. 7d, upper and middle
panel). Upon Fas stimulation of mitotic cell cultures, only a
small increase in caspase-8 activity and a small increase in
apoptosis were determined (Fig. 7d, lower left and right pan-
els). In contrast to the replacement of endogenous caspase-8
with wild-type caspase-8, the replacement with the S387A mu-
tant led to elevated caspase-8 activity and to an increased
incidence of apoptosis, indicating that prevention of Ser-387
phosphorylation increases sensitivity to Fas-induced apoptosis
in mitotic cells (Fig. 7d, lower left and right panels). Moreover,
the expression of the S387E mutant in an exponentially grow-
ing, caspase-8-depleted HeLa cell population, which encom-
passes ?80% nonmitotic cells, protected against Fas-mediated
apoptosis in a prolonged observation period of 24 h compared
to S387A-expressing cells (Fig. 7e).
Furthermore, to evaluate a putative contribution of the in-
trinsic death pathway to the apoptotic response that we ob-
served upon Fas stimulation, we used HCT116 cells with a
disrupted mitochondrial death pathway due to a lack of a
functional BAX gene (50). We replaced endogenous caspase-8
with wild-type caspase-8 or the S387A mutant in mitotic BAX-
negative HCT116 cells (Fig. 8a). Expression of the nonphos-
phorylatable caspase-8/S387A-form in caspase-8-depleted,
BAX-negative HCT116 cells led to elevated caspase-8 activity
and to an increased incidence of apoptosis upon Fas stimula-
tion during mitosis. The replacement experiment in BAX-pos-
itive cells revealed an elevated level of apoptosis compared to
the corresponding experiment using BAX-negative cells, indi-
cating a weak contribution of the intrinsic pathway to overall
apoptosis (Fig. 8c). Our data suggest that prevention of Ser-
387 phosphorylation of procaspase-8 increases sensitivity to
Fas-induced apoptosis also in mitotic BAX-negative HCT116
cells with a disrupted mitochondrial pathway (Fig. 8b to d).
Phosphorylation of procaspase-8 in primary cells and tis-
sues. Because alterations in cyclin B1 and Cdk1 are a wide-
spread feature of tumorigenesis (46), we assessed whether
procaspase-8 is phosphorylated at Ser-387 in breast cancer
cells. We analyzed primary normal breast epithelial cells, pre-
malignant ductal carcinomas in situ (DCIS) and invasive breast
cancers (Fig. 9a, upper panel). The level of cyclin B1 was low
in normal epithelial cells and higher in the more proliferative
DCIS and invasive breast cancer. Interestingly, elevated cyclin
B1 expression in the breast tissue often correlated with an
increased level of procaspase-8 phosphorylation at Ser-387
(Fig. 9a, lower panel). We detected phosphorylated pro-
caspase-8 (S387) in 12 of 13 breast cancer samples, suggesting
a pathophysiological role for this event in primary breast
To investigate the physiological role of procaspase-8 phos-
phorylation in primary cells of different origin, we analyzed the
extrinsic pathway in T cells. Resting peripheral blood lympho-
cytes (PBLs) stimulated with phytohemagglutinin (PHA)
started to proliferate within 1 to 2 days depending on the
donor. The onset of proliferation was associated with increased
ABcasp.-8 (pSer387), caspase-8, or cyclin B1 antibodies for analysis by microscopy. Caspase-8-deficient HeLa cells were arrested in G1/S using a
double-thymidine block and released into fresh medium for 10 h. Semisynchronized cells were stained with DAPI or labeled using ABcasp.-8
(pSer387) or cyclin B1 antibodies for analysis by microscopy. (c) To enrich for mitotic HeLa cells, an incubation with nocodazole (Noc) for 16 h
was carried out; cells were captured by shake-off and then treated with MG132 for 2 h (lane 2). Shake-off cells were treated as described for lane
2, except that, 1 h after MG132 addition, the Cdk1 inhibitor RO-3306 was added (lane 3). The cell lysates were immunoblotted for caspase-8,
caspase-8 (pS387), cyclin B1, Cdk1, pH3S10, securin, and ?-actin. (d) HeLa cells were arrested in G1/S using a double-thymidine block (time zero)
and released for the times indicated. The cellular lysates were immunoblotted for caspase-8 (pS387), caspase-8, pH3S10, cyclin B1, Cdk1, Plk1,
CD95/Fas, and ?-actin (upper left panels). To measure the Cdk1/cyclin B1 activity in cell lysates of synchronized cells, immunoprecipitation assays
were performed with cyclin B1-specific antibodies, and immunoprecipitated proteins were then subjected to kinase assays with histone H1 as the
substrate (lower left panel). The cell cycle status was monitored using flow cytometry (upper right panel). Boxed numbers exhibit the percentage
of cells with 4 N DNA content. The mitotic index and distribution of mitotic phases of HeLa cells 10 h after release from the G1/S block are shown:
the leftmost bar graph indicates 30% mitotic index 10 h after G1/S (of these, 46% were in prophase and 25% were in metaphase). The mitotic index
was determined by counting 6 times 250 to 300 cells (lower right panel). All experiments were performed in triplicate. The error bars represent
the SD. (e) Characterization of mock-transfected HeLa cells and of a cyclin B1-depleted HeLa clone using a stably integrated RNAi cassette
(Cycl.B1?-HeLa). Mitotic shake-off cells (mock-transfected and Cycl.B1?) were stimulated with 100 ng of FasL/ml for the times indicated. Lysates
were immunoblotted to analyze procaspase-8 processing using the caspase-8 MAb C15 and antibodies for caspase-8 (pS387), cyclin B1, and ?-actin
(*, unspecific band). To measure Cdk1/cyclin B1 activity, immunoprecipitation assays were performed using cyclin B1-specific antibodies, and the
immunoprecipitated proteins were subjected to kinase assays with histone H1 as the substrate (middle panel). Apoptosis analyses were performed
by using an annexin V kit (lower panel). All experiments were performed in triplicate. Error bars represent the SD. (f) To enrich for mitotic SKW
6.4 cells, an incubation with nocodazole (Noc) for 16 h was carried out; cells were captured by shake-off and then treated with MG132 for 2 h. For
the inhibition of Cdk1 activity, RO-3306 was added 1 h after MG132 treatment. Cells were stimulated with 100 ng of FasL/ml for the times
indicated. Lysates were immunoblotted to analyze procaspase-8 processing using the anti-caspase-8 MAb C15. Apoptosis analyses were performed
by using an annexin V kit (lower panel). All experiments were performed in triplicate. Error bars represent the SD.
VOL. 30, 2010 CDK1/CYCLIN B1 REGULATES CASPASE-8 ACTIVITY5733
Plk1 and cyclin B1 expression, as well as Cdk1/cyclin B1 activity
(Fig. 9b, upper panels). The detection of phosphorylated pro-
caspase-8 (S387) correlated with the pronounced increase of
Cdk1/cyclin B1 activity (Fig. 9b, lower panel). Cycling PBLs,
trapped in mitosis with nocodazole and MG132, were treated
with or without RO-3306 and FasL. Incubation with RO-3306
decreased the level of procaspase-8 phosphorylation at Ser-387
within 2 to 3 h and led to elevated processing (Fig. 9c). Therefore,
the inhibition of procaspase-8 phosphorylation at Ser-387 by
downregulating Cdk1/cyclin B1 activity correlates with a more
robust Fas-induced response also in primary mitotic lymphocytes.
Because many types of cancer are characterized by acceler-
ated cell proliferation driven by the increased activity of cell
cycle kinases and by resistance to death stimuli, in this study we
examined the impact of these protein kinases on the apoptotic
FIG. 5. The phosphorylation of procaspase-8 correlates with Cdk1/cyclin B1 activity in cell lines of different origin. (a) T47D, SW480, and
MDA-MB-231 cells were arrested in G1/S using a double-thymidine block (time zero) and released for the times indicated. The cellular lysates were
immunoblotted for caspase-8 (pS387), caspase-8, pH3S10, cyclin B1, Cdk1, Plk1, and ?-actin. (b) Characterization of a cyclin B1-depleted clone
using a stably integrated RNAi cassette (Cycl.B1?-HeLa). Mitotic shake-off cells (wild-type and Cycl.B1?) were stimulated with 100 ng of FasL/ml
for the times indicated. Lysates were immunoblotted to analyze procaspase-8 processing using the caspase-8 MAb C15 and antibodies for caspase-8
(pS387), cyclin B1, and ?-actin (*, unspecific band) (left panel). Characterization of Cdk1-depleted cells by RNAi. Mitotic shake-off cells (wild-type
and Cdk1-depleted cells) were stimulated with 100 ng of FasL/ml for the times indicated. Lysates were immunoblotted to analyze procaspase-8
processing using the caspase-8 MAb C15 and antibodies for caspase-8 (pS387), Cdk1, and ?-actin (*, unspecific band) (right panel).
FIG. 6. Inhibition of Cdk1 activity using the small molecule RO-3306 suppresses the phosphorylation of procaspase-8 at Ser-387 in vitro and
in vivo. (a) Procaspase-8 purified from bacteria was subjected to kinase assays with Cdk1/cyclin B1 and increasing concentrations of RO-3306.
Samples were immunoblotted for Cdk1, caspase-8, and caspase-8 (pSer387). (b) Treatment of proliferating human SW480 cancer cells with
RO-3306 for 14 h led to a complete block of the cell cycle at the G2/M border. After releasing SW480 cells from RO-3306 a Western blot analysis
was performed for caspase-8 (pS387), caspase-8, pH3S10, cyclin B1, Cdk1, and ?-actin.
5734 MATTHESS ET AL.MOL. CELL. BIOL.
FIG. 7. A nonphosphorylatable S387 mutant of procaspase-8 sensitizes mitotic cells to Fas-mediated apoptosis. (a) HeLa cells were transiently
transfected with vectors encoding Flag-tagged caspase-8 (Flag-WT, Flag-S387A, and Flag-S387E). At 18 h after transfection, cells were incubated
with 100 ng of FasL/ml plus 1 ?g of cycloheximide/ml for 3 and 6 h. Lysates were immunoblotted for exogenous caspase-8 with a Flag-specific
antibody and for ?-actin (upper panels). The signal intensity of immunoblotted, Flag-tagged caspase-8 was standardized to the level of ?-actin
expression (lower panel). All experiments were performed in triplicate. The error bars represent the SD. (b) Analysis of procaspase-8 processing
in HeLa cells expressing different double-tagged (N-terminal Flag, C-terminal V5) forms of procaspase-8 (S387A, S387E). Lysates were immu-
noblotted for V5 and PARP. (c) Analysis of procaspase-8 processing in caspase-8 knockdown HeLa cell clones (caspase-8?) expressing different
double-tagged (N-terminal Flag, C-terminal V5) forms of procaspase-8 (S387A, S387E). Lysates were immunoblotted for V5 and caspase-8. (d)
On day 1, HeLa cells were transfected with siRNA targeting the untranslated region of caspase-8, followed by the transfection of Flag-tagged
caspase-8 (mock, WT, or S387A) on day 2. On day 3, the cells were treated overnight with nocodazole, and then a mitotic shake-off was performed
on day 4. Subsequently, cells were reseeded in nocodazole-containing medium and stimulated with 100 ng of FasL/ml. The cell cycle status was
monitored by flow cytometry (upper panels). Lysates were immunoblotted for exogenous caspase-8 using a Flag-specific antibody, caspase-8,
PARP, and ?-actin (middle panels). Caspase-8 activity and apoptosis (annexin staining) were determined (lower panels). All experiments were
performed in triplicate. Error bars represent the SD. (e) On day 1 HeLa cells were transfected with siRNA targeting the untranslated region of
caspase-8, followed by the transfection of Flag-tagged caspase-8 (mock, S387A, or S387E) on day 2. On day 3 cells were stimulated with 100 ng
of FasL/ml. After 24 h, the cell numbers were determined, and lysates were immunoblotted for exogenous caspase-8 using a Flag-specific antibody
and ?-actin. The experiment was performed in triplicate. Error bars represent the SD.
VOL. 30, 2010 CDK1/CYCLIN B1 REGULATES CASPASE-8 ACTIVITY5735
cascade. Although various genetic and epigenetic mechanisms,
including mutations, homo- or heterozygous genomic dele-
tions, or allelic imbalance on chromosome 2q that inactivate
caspase-8, have been described previously (9), our understand-
ing of the posttranslational regulation of procaspase-8 is lim-
ited. Here, we describe for the first time the phosphorylation of
procaspase-8 on a serine residue. We propose a model in which
death receptor signaling is inhibited by the Cdk1/cyclin B1-
mediated phosphorylation of procaspase-8 on Ser-387 in mi-
totic cells (Fig. 10).
Various investigations established a two-step cleavage
model for the processing of procaspase-8, and it was proposed
that its activation occurs in a strict sequence of events: Al-
though the cleavage between the small and the large catalytic
subunit represents the first step of autoprocessing, the second
step occurs between the prodomain and the large catalytic
subunit (5, 27). A phosphorylation of procaspase-8 at Ser-387
close to Asp-384 and Asp-374, which are used to generate
p43/p41 and the p12/p10 subunits, inhibits also the generation
of p18. From these data, we conclude that a phosphorylation at
the N-terminal end of p10 inhibits the liberation of the p12/p10
subunits by blocking the Asp-374/Asp-384 cleavage sites and
thereby also influences the two-step cleavage of procaspase-8.
Our observations suggest that both steps of caspase-8 process-
ing occur on a mutual basis. By controlling the autoprocessing
of procaspase-8, Cdk1/cyclin B1 seems to elevate the threshold
of cells against extrinsic death signals during mitosis. This
mechanism is not specific for Fas-induced cell death, it can be
generalized to other caspase-8-dependent pathways including
the stimulation of cells with TRAIL or TNF-?.
FIG. 8. A nonphosphorylatable S387 mutant of procaspase-8 sensitizes mitotic BAX-deficient HCT116 cells to Fas-mediated apoptosis. (a) On
day 1, wild-type (BAX-positive) and BAX-deficient HCT116 cells were transfected with siRNA targeting the untranslated region of caspase-8,
followed by the transfection of Flag-tagged caspase-8 (mock, WT, or S387A) on day 2. On day 3, cells were treated overnight with nocodazole, and
then a mitotic shake-off was performed on day 4. Subsequently, the cells were reseeded in nocodazole-containing medium and stimulated with 100
ng of FasL/ml. Lysates were immunoblotted for exogenous caspase-8 using a Flag-specific antibody and BAX. (b) Lysates of BAX-positive and
-negative cells that were treated as described in panel a were immunoblotted for PARP. (c and d) Apoptosis (annexin staining) and caspase-8
activity were determined. All experiments were performed in triplicate. Error bars represent the SD.
5736 MATTHESS ET AL.MOL. CELL. BIOL.
Whereas the activation of effector caspases by Fas-induced
stimulation of caspase-8 suffices for cell killing in type I cells,
caspase cascade amplification through caspase-8 mediated ac-
tivation of the proapoptotic BCL-2 family member BID lead-
ing to BAX/BAK-dependent activation of caspases-9 is essen-
tial in type 2 cells. Using type I cells such as the B
lymphoblastoid cell line SKW 6.4 or HCT116 cells with a
disrupted mitochondrial death pathway due to a lack of a
functional BAX gene, we could demonstrate that the post-
translational modification of procaspase-8 regulates Fas-medi-
ated apoptosis independent of caspases cascade amplification.
A previous study has described the src-mediated phosphor-
ylation of caspase-8 on Tyr-380, which is located in the linker
region between the p18 and p11 subunits. The overexpression
of a hyperactive mutant of c-src resulted in a delayed accumu-
lation of the large catalytic subunit p18 after Fas stimulation.
Therefore, src activity triggers endogenous caspase-8 tyrosine
phosphorylation and also protects cells from Fas-induced
FIG. 9. Phosphorylation of procaspase-8 at Ser-387 in primary cells and tissues. (a) Lysates were prepared from human breast cancer tissues,
premalignant ductal carcinoma in situ (DCIS) samples and normal breast tissues. Total protein was resolved by SDS-PAGE and immunoblotted for
caspase-8 (pSer387), caspase-8, cyclin B1, and ?-actin (top panel). Blue bars represent the level of caspase-8 (pSer387) normalized to the level of ?-actin
expression (lower panel). Red bars represent the level of cyclin B1 expression standardized to the level of ?-actin expression. (b) Peripheral blood
lymphocytes (PBLs) were treated with PHA for the times indicated. Cell lysates were immunoblotted for caspase-8 (pS387), caspase-8, histone H3
phosphorylated at Ser10 (H3 pS10), cyclin B1, Cdk1, Plk1, CD95/Fas, and ?-actin (upper panels). To measure Cdk1/cyclin B1 activity in cell lysates of
synchronized cells, immunoprecipitation assays were performed using a cyclin B1-specific antibody, and the purified proteins were subjected to kinase
assays with histone H1 as the substrate (lower panel). (c) PBLs were cultured with PHA for 16 h. Cells were then washed three times and cultured for
an additional 5 days in the presence of interleukin-2, followed by nocodazole treatment for 16 h and MG132 for 2 h. For the inhibition of Cdk1/cyclin
B1, RO-3306 was added.
VOL. 30, 2010CDK1/CYCLIN B1 REGULATES CASPASE-8 ACTIVITY5737
apoptosis. Although elevated Cdk1/cyclin B1 activity is a more
general feature of mammalian tumors, the protective effect of
elevated src activity under physiological conditions is limited to
certain types of cancers such as human colon cancer (7). The
present study also highlights the functional importance of post-
translational modifications within the region linking the small
and large catalytic subunits in the regulation of the activity of
caspase-8. Phosphorylation of Tyr-380 and/or Ser-387 re-
presses the activation of procaspase-8, which requires two
cleavage events to remove the linker region between the p18
and the p10 subunits to liberate p10 (7). In addition, Cdk1/
cyclin B1 was shown to phosphorylate caspase-9 (Thr-125) and
most recently also caspase-2 (Ser-340) and thereby blocks
apoptosis in mitotic cells (1, 2). Most interestingly, these
posttranslational events, including the phosphorylation of
caspase-8 (Tyr-380) by src (7), all modulate the processing of
caspases by targeting different linker regions connecting func-
tional domains of caspases. Taken together, the description of
caspases-2, -8, and -9 as substrates of Cdk1/cyclin B1 suggests
that Cdk1/cyclin B1 targets predominantly initiator caspases
that act upstream in the apoptotic cascade thereby blocking
apoptotic signaling in an early stage.
Furthermore, pTyr-380 of caspase-8 can function as a dock-
ing site for the p85 subunit of phosphatidylinositol 3-kinase
(PI3K), which affects Rac activation, thereby promoting cell
adhesion and motility (38). Hence, caspase-8 can be consid-
ered as a molecular switch involved in the cellular decision to
initiate apoptosis or migration. Whether phosphorylation at
Ser-387 by Cdk1/cyclin B1 regulates this balance between the
apoptotic and nonapoptotic functions of caspase-8 by influenc-
ing its interaction with PI3K will require further study.
In many cancers, the normal apoptotic program is dysregu-
lated, which can lead to uncontrolled proliferation and the
development of resistance to conventional chemotherapy and
radiotherapy. Many tumors become particularly resistant to
death receptor-induced apoptosis with increasing treatment
modalities (15). Thus, it is important to identify and develop
therapies that restore the sensitivity of cancer cells to Fas-
mediated apoptosis. Interestingly, the potentiation of Cdk1 via
overexpression of its cognate partner cyclin B1 and the inac-
tivation of Cdk inhibitors is often linked to tumor development
(23, 25, 32, 39). Our observations suggest that procaspase-8
phosphorylation on Ser-387 raises the threshold for Fas-medi-
ated apoptosis in mitotic cells of different origin. In addition,
antimitotic drugs that inhibit microtubule dynamics cause cells
to arrest in mitosis with elevated levels of cyclin B1 and acti-
vated Cdk1/cyclin B1 kinase activity. Although this treatment
elevates the threshold toward extrinsic and intrinsic apoptotic
stimuli considering the Cdk1/cyclin B1-mediated phosphoryla-
tion of capases-8, -9, and -2, the activation of the spindle
assembly checkpoint by antimitotic drugs is of central impor-
tance for the fate of cancer cells.
Because increased proliferation is one trait of cancer, neo-
plastic cells could have hijacked the protective effect of ele-
FIG. 10. Proposed model for the impact of the active Cdk1/cyclin B1 complex on the extrinsic death pathway. During the cell cycle, Cdk1/cyclin
B1 associates with procaspase-8. As cyclin B1 becomes abundant in G2, the active Cdk1/cyclin B1 complex phosphorylates Ser-387 on procaspase-8
at the centrosomes. This phosphorylation event inhibits the processing of procaspase-8, thus blocking its enzymatic activation and the downstream
initiation of apoptotic signaling. This cellular barrier seems to protect mitotic cells against Fas-mediated apoptosis. The identification of Ser-387
phosphorylation of procaspase-8 in primary lymphocytes and breast tissues suggests both a physiological and a pathophysiological role for this
process in vivo.
5738 MATTHESS ET AL.MOL. CELL. BIOL.
vated Cdk1/cyclin B1 activity to evade apoptosis by raising the
threshold for extrinsic death signals originating from the infil-
tration of immune cells into tumors. The selective inhibition of
Cdk1/cyclin B1 leading to reduced procaspase-8 phosphoryla-
tion may constitute a novel approach to fighting the resistance
of cancer to therapeutic strategies targeting the extrinsic death
pathway. Moreover, several strategies have been developed to
elevate caspase-8 expression to restore its functions in human
cancer and to resensitize cancer cells to extrinsic death stimuli:
the treatment of cancer cells with demethylating agents, inter-
ferons, and retinoic acid are promising strategies to raise the
levels of the initiator caspase-8 in tumors lacking this critical
apoptosis regulator (10, 17, 47). Taken together, combinations
of these agents with Cdk1/cyclin B1 inhibitors might further
increase the sensitivity of cancer cells to extrinsic cell death.
We thank Kaufmann for providing the tissue samples. We thank
Vogelstein for the generous gift of BAX-negative HCT116 cells.
This study was supported by grants from the Else Kro ¨ner-Fresenius/
Carls-Stiftung, the Deutsche Krebshilfe, and the LOEWE Centre
1. Allan, L. A., and P. R. Clarke. 2007. Phosphorylation of caspase-9 by CDK1/
cyclin B1 protects mitotic cells against apoptosis. Mol. Cell 26:301–310.
2. Andersen, J. L., C. E. Johnson, C. D. Freel, A. B. Parrish, J. L. Day, M. R.
Buchakjian, L. K. Nutt, J. W. Thompson, M. A. Moseley, and S. Kornbluth.
2009. Restraint of apoptosis during mitosis through interdomain phosphor-
ylation of caspase-2. EMBO J. 28:3216–3227.
3. Bentley, A. M., G. Normand, J. Hoyt, and R. W. King. 2007. Distinct se-
quence elements of cyclin B1 promote localization to chromatin, centro-
somes, and kinetochores during mitosis. Mol. Biol. Cell 18:4847–4858.
4. Blangy, A., H. A. Lane, P. d’Herin, M. Harper, M. Kress, and E. A. Nigg.
1995. Phosphorylation by p34cdc2 regulates spindle association of human
Eg5, a kinesin-related motor essential for bipolar spindle formation in vivo.
5. Chang, D. W., Z. Xing, V. L. Capacio, M. E. Peter, and X. Yang. 2003.
Interdimer processing mechanism of procaspase-8 activation. EMBO J. 22:
6. Cox, A., A. M. Dunning, M. Garcia-Closas, S. Balasubramanian, M. W.
Reed, K. A. Pooley, S. Scollen, C. Baynes, B. A. Ponder, S. Chanock, J.
Lissowska, L. Brinton, B. Peplonska, M. C. Southey, J. L. Hopper, M. R.
McCredie, G. G. Giles, O. Fletcher, N. Johnson, S. S. dos, I. L. Gibson, S. E.
Bojesen, B. G. Nordestgaard, C. K. Axelsson, D. Torres, U. Hamann, C.
Justenhoven, H. Brauch, J. Chang-Claude, S. Kropp, A. Risch, S. Wang-
Gohrke, P. Schurmann, N. Bogdanova, T. Dork, R. Fagerholm, K. Aaltonen,
C. Blomqvist, H. Nevanlinna, S. Seal, A. Renwick, M. R. Stratton, N. Rah-
man, S. Sangrajrang, D. Hughes, F. Odefrey, P. Brennan, A. B. Spurdle, G.
Chenevix-Trench, J. Beesley, A. Mannermaa, J. Hartikainen, V. Kataja,
V. M. Kosma, F. J. Couch, J. E. Olson, E. L. Goode, A. Broeks, M. K.
Schmidt, F. B. Hogervorst, L. J. Van’t Veer, D. Kang, K. Y. Yoo, D. Y. Noh,
S. H. Ahn, S. Wedren, P. Hall, Y. L. Low, J. Liu, R. L. Milne, G. Ribas, A.
Gonzalez-Neira, J. Benitez, A. J. Sigurdson, D. L. Stredrick, B. H. Alex-
ander, J. P. Struewing, P. D. Pharoah, and D. F. Easton. 2007. A common
coding variant in CASP8 is associated with breast cancer risk. Nat. Genet.
7. Cursi, S., A. Rufini, V. Stagni, I. Condo, V. Matafora, A. Bachi, A. P.
Bonifazi, L. Coppola, G. Superti-Furga, R. Testi, and D. Barila. 2006. Src
kinase phosphorylates caspase-8 on Tyr380: a novel mechanism of apoptosis
suppression. EMBO J. 25:1895–1905.
8. Degterev, A., and J. Yuan. 2008. Expansion and evolution of cell death
programmes. Nat. Rev. Mol. Cell. Biol. 9:378–390.
9. Fulda, S. 2009. Caspase-8 in cancer biology and therapy. Cancer Lett. 281:
10. Fulda, S., M. U. Kufer, E. Meyer, V. F. van, B. Dockhorn-Dworniczak, and
K. M. Debatin. 2001. Sensitization for death receptor- or drug-induced
apoptosis by re-expression of caspase-8 through demethylation or gene
transfer. Oncogene 20:5865–5877.
11. Hanahan, D., and R. A. Weinberg. 2000. The hallmarks of cancer. Cell
12. Harada, K., S. Toyooka, N. Shivapurkar, A. Maitra, J. L. Reddy, H. Matta,
K. Miyajima, C. F. Timmons, G. E. Tomlinson, D. Mastrangelo, R. J. Hay,
P. M. Chaudhary, and A. F. Gazdar. 2002. Deregulation of caspase 8 and 10
expression in pediatric tumors and cell lines. Cancer Res. 62:5897–5901.
13. Holtrich, U., G. Wolf, A. Brauninger, T. Karn, B. Bohme, H. Rubsamen-
Waigmann, and K. Strebhardt. 1994. Induction and down-regulation of
PLK, a human serine/threonine kinase expressed in proliferating cells and
tumors. Proc. Natl. Acad. Sci. U. S. A. 91:1736–1740.
14. Hughes, M. A., N. Harper, M. Butterworth, K. Cain, G. M. Cohen, and M.
MacFarlane. 2009. Reconstitution of the death-inducing signaling complex
reveals a substrate switch that determines CD95-mediated death or survival.
Mol. Cell 35:265–279.
15. Igney, F. H., and P. H. Krammer. 2002. Death and anti-death: tumour
resistance to apoptosis. Nat. Rev. Cancer 2:277–288.
16. Jackman, M., C. Lindon, E. A. Nigg, and J. Pines. 2003. Active cyclin
B1-Cdk1 first appears on centrosomes in prophase. Nat. Cell Biol. 5:143–
17. Jiang, M., K. Zhu, J. Grenet, and J. M. Lahti. 2008. Retinoic acid induces
caspase-8 transcription via phospho-CREB and increases apoptotic re-
sponses to death stimuli in neuroblastoma cells. Biochim. Biophys. Acta
18. Kappel, S., Y. Matthess, M. Kaufmann, and K. Strebhardt. 2007. Silencing
of mammalian genes by tetracycline-inducible shRNA expression. Nat. Pro-
19. Kappel, S., Y. Matthess, B. Zimmer, M. Kaufmann, and K. Strebhardt.
2006. Tumor inhibition by genomically integrated inducible RNAi-cassettes.
Nucleic Acids Res. 34:4527–4536.
20. Keen, N., and S. Taylor. 2004. Aurora-kinase inhibitors as anticancer agents.
Nat. Rev. Cancer 4:927–936.
21. Kimura, K., O. Cuvier, and T. Hirano. 2001. Chromosome condensation by
a human condensin complex in Xenopus egg extracts. J. Biol. Chem. 276:
22. Lavrik, I., A. Krueger, I. Schmitz, S. Baumann, H. Weyd, P. H. Krammer,
and S. Kirchhoff. 2003. The active caspase-8 heterotetramer is formed at the
CD95 DISC. Cell Death. Differ. 10:144–145.
23. Malumbres, M., and M. Barbacid. 2001. To cycle or not to cycle: a critical
decision in cancer. Nat. Rev. Cancer 1:222–231.
24. Malumbres, M., and M. Barbacid. 2009. Cell cycle, CDKs and cancer: a
changing paradigm. Nat. Rev. Cancer 9:153–166.
25. Mashal, R. D., S. Lester, C. Corless, J. P. Richie, R. Chandra, K. J. Propert,
and A. Dutta. 1996. Expression of cell cycle-regulated proteins in prostate
cancer. Cancer Res. 56:4159–4163.
26. Matthess, Y., S. Kappel, B. Spankuch, B. Zimmer, M. Kaufmann, and K.
Strebhardt. 2005. Conditional inhibition of cancer cell proliferation by tet-
racycline-responsive, H1 promoter-driven silencing of PLK1. Oncogene 24:
27. Medema, J. P., C. Scaffidi, F. C. Kischkel, A. Shevchenko, M. Mann, P. H.
Krammer, and M. E. Peter. 1997. FLICE is activated by association with the
CD95 death-inducing signaling complex (DISC). EMBO J. 16:2794–2804.
28. Muzio, M., B. R. Stockwell, H. R. Stennicke, G. S. Salvesen, and V. M. Dixit.
1998. An induced proximity model for caspase-8 activation. J. Biol. Chem.
29. Nigg, E. A. 2001. Mitotic kinases as regulators of cell division and its check-
points. Nat. Rev. Mol. Cell. Biol. 2:21–32.
30. Nurse, P. 1990. Universal control mechanism regulating onset of M-phase.
31. Peter, M., J. Nakagawa, M. Doree, J. C. Labbe, and E. A. Nigg. 1990. In vitro
disassembly of the nuclear lamina and M phase-specific phosphorylation of
lamins by cdc2 kinase. Cell 61:591–602.
32. Pines, J. 2006. Mitosis: a matter of getting rid of the right protein at the right
time. Trends Cell Biol. 16:55–63.
33. Raab, M., S. Pfister, and C. E. Rudd. 2001. CD28 signaling via VAV/SLP-76
adaptors: regulation of cytokine transcription independent of TCR ligation.
34. Raab, M., M. Yamamoto, and C. E. Rudd. 1994. The T-cell antigen CD5 acts
as a receptor and substrate for the protein-tyrosine kinase p56lck. Mol. Cell.
35. Reindl, W., J. Yuan, A. Kramer, K. Strebhardt, and T. Berg. 2008. Inhibition
of polo-like kinase 1 by blocking polo-box domain-dependent protein-pro-
tein interactions. Chem. Biol. 15:459–466.
36. Rudner, A. D., and A. W. Murray. 2000. Phosphorylation by Cdc28 activates
the Cdc20-dependent activity of the anaphase-promoting complex. J. Cell
37. Scaffidi, C., I. Schmitz, P. H. Krammer, and M. E. Peter. 1999. The role of
c-FLIP in modulation of CD95-induced apoptosis. J. Biol. Chem. 274:1541–
38. Senft, J., B. Helfer, and S. M. Frisch. 2007. Caspase-8 interacts with the p85
subunit of phosphatidylinositol 3-kinase to regulate cell adhesion and mo-
tility. Cancer Res. 67:11505–11509.
39. Shapiro, G. I. 2006. Cyclin-dependent kinase pathways as targets for cancer
treatment. J. Clin. Oncol. 24:1770–1783.
40. Srinivasula, S. M., M. Ahmad, T. Fernandes-Alnemri, G. Litwack, and E. S.
Alnemri. 1996. Molecular ordering of the Fas-apoptotic pathway: the Fas/
APO-1 protease Mch5 is a CrmA-inhibitable protease that activates multiple
Ced-3/ICE-like cysteine proteases. Proc. Natl. Acad. Sci. U. S. A. 93:14486–
VOL. 30, 2010 CDK1/CYCLIN B1 REGULATES CASPASE-8 ACTIVITY5739
41. Strebhardt, K. 2010. Multifaceted polo-like kinases: drug targets and anti- Download full-text
targets for cancer therapy. Nat. Rev. Drug Discov. 9:643–660.
42. Strebhardt, K., and A. Ullrich. 2006. Targeting polo-like kinase 1 for cancer
therapy. Nat. Rev. Cancer 6:321–330.
43. Stupack, D. G., T. Teitz, M. D. Potter, D. Mikolon, P. J. Houghton, V. J.
Kidd, J. M. Lahti, and D. A. Cheresh. 2006. Potentiation of neuroblastoma
metastasis by loss of caspase-8. Nature 439:95–99.
44. Teitz, T., T. Wei, M. B. Valentine, E. F. Vanin, J. Grenet, V. A. Valentine,
F. G. Behm, A. T. Look, J. M. Lahti, and V. J. Kidd. 2000. Caspase 8 is
deleted or silenced preferentially in childhood neuroblastomas with ampli-
fication of MYCN. Nat. Med. 6:529–535.
45. Vassilev, L. T., C. Tovar, S. Chen, D. Knezevic, X. Zhao, H. Sun, D. C.
Heimbrook, and L. Chen. 2006. Selective small-molecule inhibitor reveals
critical mitotic functions of human CDK1. Proc. Natl. Acad. Sci. U. S. A.
46. Winters, Z. E., N. C. Hunt, M. J. Bradburn, J. A. Royds, H. Turley, A. L.
Harris, and C. J. Norbury. 2001. Subcellular localization of cyclin B, Cdc2,
and p21WAF1/CIP1in breast cancer. association with prognosis. Eur. J. Cancer
47. Yang, X., M. S. Merchant, M. E. Romero, M. Tsokos, L. H. Wexler, U.
Kontny, C. L. Mackall, and C. J. Thiele. 2003. Induction of caspase 8 by
interferon gamma renders some neuroblastoma (NB) cells sensitive to tumor
necrosis factor-related apoptosis-inducing ligand (TRAIL) but reveals that a
lack of membrane TR1/TR2 also contributes to TRAIL resistance in NB.
Cancer Res. 63:1122–1129.
48. Yuan, J., A. Kramer, Y. Matthess, R. Yan, B. Spankuch, R. Gatje, R. Knecht,
M. Kaufmann, and K. Strebhardt. 2006. Stable gene silencing of cyclin B1 in
tumor cells increases susceptibility to taxol and leads to growth arrest in vivo.
49. Yuan, J., R. Yan, A. Kramer, F. Eckerdt, M. Roller, M. Kaufmann, and K.
Strebhardt. 2004. Cyclin B1 depletion inhibits proliferation and induces
apoptosis in human tumor cells. Oncogene 23:5843–5852.
50. Zhang, L., J. Yu, B. H. Park, K. W. Kinzler, and B. Vogelstein. 2000. Role
of BAX in the apoptotic response to anticancer agents. Science 290:989–
5740 MATTHESS ET AL.MOL. CELL. BIOL.