Tài liệu Báo cáo khoa học: An alternative transcript from the death-associated protein kinase 1 locus encoding a small protein selectively mediates membrane blebbing pdf

An alternative transcript from the death-associated
protein kinase 1 locus encoding a small protein selectively
mediates membrane blebbing
Yao Lin
1
, Craig Stevens
1
, Roman Hrstka
2
, Ben Harrison
1
, Argyro Fourtouna
1
, Suresh Pathuri
1
,
Borek Vojtesek
2
and Ted Hupp
1
1 Institute of Genetics and Molecular Medicine, Cell Signalling Unit, CRUK p53 Signal Transduction Group, University of Edinburgh, UK
2 Masaryk Memorial Cancer Institute, Brno, Czech Republic
Death-associated protein kinase 1 (DAPK-1) is a
Ca
2+
⁄ calmodulin-regulated serine ⁄ threonine kinase
composed of multiple functional domains, including a
kinase domain, a calmodulin-binding domain, eight
ankyrin repeats, two P-loop motifs, a cytoskeletal
binding domain, a death domain, and a C-terminal
regulatory tail [1]. It has been shown that DAPK-1

is involved in the regulation of distinct processes,
Keywords
DAPK-1; ERK; membrane blebbing; p53;
proteolysis
Correspondence
T. Hupp, Institute of Genetics and Molecular
Medicine, Cell Signalling Unit, CRUK p53
Signal Transduction Group, University of
Edinburgh, Edinburgh EH4 2XR, UK
Fax: +44 131 7773542
Tel: +44 131 7773583
E-mail:
(Received 14 January 2008, revised 11
March 2008, accepted 14 March 2008)
doi:10.1111/j.1742-4658.2008.06404.x
Death-associated protein kinase 1 (DAPK-1) is a multidomain protein
kinase with diverse roles in autophagic, apoptotic and survival pathways.
Bioinformatic screens were used to identify a small internal mRNA from
the DAPK-1 locus (named s-DAPK-1). This encodes a 295 amino acid
polypeptide encompassing part of the ankyrin-repeat domain, the P-loop
motifs, part of the cytoskeletal binding domain of DAPK-1, and a unique
C-terminal ‘tail’ extension not present in DAPK-1. Expression of
s-DAPK-1 mRNA was detected in a panel of normal human tissues as well
as primary colorectal cancers, indicating that its expression occurs in vivo.
s-DAPK-1 gene transfection into cells produces two protein products: one
with a denatured mass of 44 kDa, and a smaller product of 40 kDa. Dou-
ble alanine mutation of the C-terminal tail extension of s-DAPK-1
(Gly296 ⁄ Arg297) prevented production of the 40 kDa fragment, suggesting
that the smaller product is generated by in vivo proteolytic processing. The
s-DAPK-1 gene cannot substitute for full-length DAPK-1 in an mitogen-

activated protein kinase kinase ⁄ extracellular signal-regulated kinase-depen-
dent apoptotic transfection assay. However, the transfection of s-DAPK-1
was able to mimic full-length DAPK-1 in the induction of membrane bleb-
bing. The 44 kDa protease-resistant mutant s-DAPK-1G296A ⁄ R297A had
very low activity in membrane blebbing, whereas the 40 kDa s-DAPK-
1Dtail protein exhibited the highest levels of membrane blebbing. Deletion
of the tail extension of s-DAPK-1 increased its half-life, shifted the equilib-
rium of the protein from cytoskeletal to soluble cytosolic pools, and altered
green fluorescent protein-tagged s-DAPK-1 protein localization as observed
by confocal microscopy. These data highlight the existence of an alternative
product of the DAPK-1 locus, and suggest that proteolytic removal of the
C-terminal tail of s-DAPK-1 is required to stimulate maximally its mem-
brane-blebbing function.
Abbreviations
GFP, green fluorescent protein; GST, glutathione S-transferase; ERK, extracellular signal-regulated kinase; MEK, mitogen-activated protein
kinase kinase; TM, tail mutant.
2574 FEBS Journal 275 (2008) 2574–2584 ª 2008 The Authors Journal compilation ª 2008 FEBS
including apoptosis, cell survival, and autophagy path-
ways, with each role depending on the cellular context
and the upstream signals [1–5]. No apparent defects
in developmental cell death were observed in DAPK-
1-knockout mice [1], thus providing no obvious
insights into its stress-regulated functions. However,
recent research has found that loss or reduced expres-
sion of DAPK-1 underlies cases of heritable predispo-
sition to chronic lymphocytic leukemia and the
majority of cases of sporadic chronic lymphocytic leu-
kemia [6], suggesting an important role of DAPK-1 in
altering the incidence of certain cancer types. This is
consistent with the ability of DAPK-1 to play a funda-

mental role in oncogene activation of the p53 tumor
suppressor pathway [7].
DAPK-1 is relatively large for a protein kinase, and
its independent functional domains are involved in var-
ious regulatory activities. The DAPK-1 kinase domain
is required to mediate cytoskeleton remodeling by
phosphorylating myosin light chain 2 [8], inhibiting cell
migration [9] and inducing membrane blebbing [10].
The latter has been characterized as a common mor-
phology correlating with apoptosis or autophagic cell
death signals, and the actin–myosin system is
considered to be the source of the contractile force
underlying the bleb formation [11]. Furthermore,
microtubule-associated protein 1B interaction with the
kinase domain of DAPK-1 stimulates membrane bleb-
bing and autophagy [12]. The roles of its other func-
tional domains in its regulatory effects are being
characterized. For example, the death domain of
DAPK-1 forms a docking site for its interaction with
extracellular signal-regulated kinase (ERK) [6], and is
thus required for DAPK-1’s proapoptotic effect in
response to the mitogen-activated protein kinase kinase
(MEK) ⁄ ERK signaling pathway [3]. Moreover, a
germline mutation in the death domain of DAPK-1
has been found to reduce intrinsic oligomerization of
the death domain, disrupt the binding of ERK, and
thus prevent MEK ⁄ ERK-induced apoptosis [13]. The
death domain of DAPK-1 also promotes its interaction
with the netrin-1 receptor UNC5H2, whose proapop-
totic effect when unbound to netrin-1 is partially

attenuated in the absence of DAPK-1 [14]. The anky-
rin-repeat region of DAPK-1 is required for its proper
localization to the actin stress fibers [8] and for stable
binding with DAPK-1’s ubiquitin E3 ligase, called
DAPK-1-interacting protein 1 [4]. Recently, it was
shown that the leukocyte common antigen-related
tyrosine phosphatase interacts with the ankyrin-repeat
region of DAPK-1 and dephosphorylates DAPK-1 at
pY491 ⁄ 492 to stimulate its proapoptotic and antimi-
gration activities [15]. There are many regions ⁄ minido-
mains on DAPK-1 without an ascribed function, and
it is likely that further biochemical characterization
will result in a greater understanding of the DAPK-1
gene product in autophagic and apoptotic cell
signaling.
Here we report on an mRNA product of the
DAPK-1 locus that encodes a small miniprotein
(named s-DAPK-1), which shares some domains with
full-length DAPK-1: from part of the ankyrin-repeat
region, through to part of the cytoskeleton binding
domain, and concluding with a unique tail extension
of 42 amino acids that is not present in full-length
DAPK-1. Unlike DAPK-1, s-DAPK-1 cannot induce
apoptosis in response to MEK ⁄ ERK signaling. How-
ever, s-DAPK-1 can mimic full-length DAPK-1’s
ability to promote membrane blebbing. The unique
C-terminal tail of s-DAPK-1 contains an internal pro-
teolytic processing site whose removal stimulates maxi-
mally the membrane-blebbing-promoting effect of
s-DAPK-1. These data together identify a novel func-

tion for the DAPK-1 locus through the expression of a
gene product with a relatively specific role in mem-
brane blebbing.
Results
DAPK-1 is composed of multiple independent minido-
mains, and in an attempt to determine whether homol-
ogous minidomains encoded by alternative genes might
exist that compete with or mimic DAPK-1 function,
we searched for evidence of the existence of alterna-
tively expressed messages in databases. In particular,
the ankyrin repeat of DAPK-1 is a potentially versatile
protein–protein interaction motif [16], and similar pro-
teins in the human genome might be found that cross-
talk to the DAPK-1 pathway. Therefore, we evaluated
the homology of the ankyrin-repeat region of DAPK-1
with other genes in the human genome using the NCBI
nucleotide blast tool. One Homo sapiens cDNA was
identified: FLJ45958 fis, clone PLACE7011559, from
the NEDO human cDNA sequencing project. The
mRNA of this expression clone starts on intron 13–14
of the DAPK-1 gene and stops on intron 20–21
(Fig. 1A). The start codon, ATG, of this expression
clone is located on the 10–12th base pairs of exon 15
within the DAPK-1 gene, which makes the translation
of this clone in-frame with that of DAPK-1 mRNA.
After the start codon, this expression clone shares the
same sequence as DAPK-1 mRNA through the end of
exon 20, as indicated, and its stop codon, TAG, is
located on the 124–126th base pairs of intron 20–21
(Fig. 1A). Thus, the first 295 amino acids of this 337

amino acid protein are identical to the region of the
Y. Lin et al. Functional transcript expressed by DAPK-1 locus
FEBS Journal 275 (2008) 2574–2584 ª 2008 The Authors Journal compilation ª 2008 FEBS 2575
Fig. 1. The identification of a small transcript from the DAPK-1 locus. (A) A schematic map of s-DAPK-1 mRNA in relation to the DAPK-1
gene structure. The mRNA of s-DAPK-1 starts in intron 13–14 of the DAPK-1 gene. Its coding region starts from the 10th base pair on
exon 15 of the DAPK-1 gene, and shares the same splicing as full-length DAPK-1 through the rest of exons 15, 16, 17, 18, 19 and 20.
s-DAPK-1’s coding region stops at the 126th base pair of intron 20–21 of the DAPK-1 gene, and the 3¢-UTR extends through the middle of
intron 20–21. (B) Comparison of the protein sequences of DAPK-1 and s-DAPK-1. The first 295 amino acids of s-DAPK-1 are identical to
amino acids 447–743 of full-length DAPK-1; however, the last 42 amino acids comprise a unique tail. (C) Identification of s-DAPK-1 mRNA.
RT-PCR was performed using the Stratagene QPCR Human Reference Total RNA, and the products were subjected to electrophoresis and
staining with ethidium bromide. (D) mRNA level test using SYBR Green real-time PCR. The relative mRNA level is depicted as a ratio of
DAPK-1 ⁄ s-DAPK-1 to actin. (E, F) s-DAPK-1, DAPK-1 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA quantification in colon
carcinoma and rectal carcinoma as compared to normal colonic tissue. Colon carcinoma cells, rectal carcinoma cells and their normal healthy
tissue counterparts were harvested (1a, carcinoma cells; 4a, normal tissues), and the relative mRNA was quantified using SYBR Green real-
time PCR as described previously for the DAPK-1 gene [2].
Functional transcript expressed by DAPK-1 locus Y. Lin et al.
2576 FEBS Journal 275 (2008) 2574–2584 ª 2008 The Authors Journal compilation ª 2008 FEBS
DAPK-1 protein from residues 447–743, whereas the
last 42 amino acids are unique for this product
(Fig. 1B). These data suggest that this product is
highly similar to and may be a splice variant of
DAPK-1. Because of its smaller size as compared to
full-length DAPK-1, we have named it s-DAPK-1. The
transcription of s-DAPK-1 was demonstrated further
by RT-PCR using the Stratagene (La Jolla, CA, USA)
QPCR Human Reference Total RNA and the primers
located on both ends of the coding region of s-DAPK-1
mRNA (Fig. 1C).
In order to determine the expression of s-DAPK-1,
we first compared its mRNA expression with that of

DAPK-1 in three widely used tumor cell lines, and we
saw a general coincidence between full-length DAPK-1
and s-DAPK-1 mRNA levels (Fig. 1D). Next, we set
out to determine whether the s-DAPK-1 expression
occurs in normal human tissue as well as primary
human cancers, rather than just cell lines and cDNA
from the NEDO human sequencing project. We
evaluated the expression of the mRNAs encompassing
full-length DAPK-1 and s-DAPK-1 in colorectal carci-
nomas (1a) and their normal tissue counterpart (4a)
using real-time PCR. As indicated, DAPK-1 and
s-DAPK-1 seem to possess similar mRNA expression
profiles throughout the samples (Fig. 1E,F). When
full-length DAPK-1 was found to be repressed in
C18_222_1a tissue, the s-DAPK-1 isoform was also
found to be repressed and undetectable (Fig. 1E).
These data indicate that s-DAPK-1 expression can
occur in primary human cancers, and the product of
this mRNA was subsequently evaluated as described
below. Furthermore, s-DAPK1 expression in normal
intestinal tissue indicates that its expression is not the
result of aberrant splicing, which is known to occur in
human cancers.
To begin functional studies of s-DAPK-1, the
s-DAPK-1 cDNA was cloned into a Flag–Myc vector
(Fig. 2A), which contains an N-terminal Flag tag and
a C-terminal Myc tag, and this was followed by
expression in HCT116 p53
+ ⁄ +
cells. Two major trans-

fected bands were observed: a 44 kDa upper band,
and a 40 kDa lower band (Fig. 2B). In order to deter-
mine which band corresponded to s-DAPK-1, the
C-terminal Myc tag was deleted (Fig. 2A). Upon trans-
fection, the same lower protein band was observed in
the Flag–s-DAPK-1- and the Flag–s-DAPK-1-Myc-
transfected cells, whereas the upper band in the Flag–
s-DAPK-1 transfection lane was slightly smaller
(Fig. 2C, lane 2 versus lane 1). This suggests that the
depletion of the Myc tag only changes the size of the
upper band, and that therefore the upper band repre-
sents the ‘full-length’ s-DAPK-1.
Two s-DAPK-1 deletion mutants, Flag-AO (Anky-
rin repeat Only) and Flag-TD (Tail Deletion;
s-DAPK-1Dtail) were created (Fig. 2A) to further
investigate why the lower molecular mass protein was
observed. Upon transfection, the Flag-TD vector pro-
duces only one major band (s-DAPK-1Dtail) of lower
molecular mass (Fig. 2D, lane 3) similar to the 40 kDa
lower band produced from the full-length s-DAPK-1
(Fig. 2D, lane 4 versus lane 3). This suggests that the
lower band might be a cleavage product of the full-
length s-DAPK-1, and that the cleavage signal is
within the C-terminal tail extension. This is further
suggested by the in vitro cleavage assay, in which the
purified glutathione S-transferase (GST)–s-DAPK-1
was incubated with HCT116 p53
+ ⁄ +
cell lysates. With
increasing amount of cell lysates, GST–s-DAPK-1 was

cleaved in vitro at a faster rate than GST alone
(Fig. 2E), supporting the existence of a protease that
cleaves s-DAPK-1 protein in vivo. The reason why the
cleavage band was not observed in this assay may be
the rapid degradation of the purified protein from the
cell lysate. When subjected to a longer exposure, the
blot showed multiple bands below GST–s-DAPK-1,
which may mask the actual cleavage band (data not
shown). The higher molecular mass protein band
( 54 kDa) might result from a covalent adduct result-
ing from ubiquitin-like modification; nevertheless, this
apparent adduct is dependent upon the integrity of the
C-terminal tail extension.
Since it had been confirmed that the cleavage is
within the tail, we next investigated sites within the tail
that are the critical targets for the cleavage. Because
the transfected Flag-TD vector (s-DAPK-1
Dtail
) is simi-
lar to the in vivo cleaved form of Flag–s-DAPK-1 in
size, five tail mutants (TMs) of s-DAPK-1 were created
to screen the first 10 amino acids on the tail for prote-
olytic susceptibility (Fig. 3A). Upon transfection, only
s-DAPK-1G296A ⁄ R297A (TM1) exhibited a reduced
proteolytic band, and s-DAPK-1
N298A ⁄ L299A
(TM2)
showed a weakened cleavage band (Fig. 3B). These
data suggest that the first two amino acids of the tail
are critical for proteolytic susceptibility, and that the

third and fourth amino acids are involved in the regu-
lation of this cleavage. This also further fine-maps the
site of cleavage, and indicates that the tail deletion
(s-DAPK-1Dtail) may be used as a mimic of the in vivo
processed form of full-length s-DAPK-1. s-DAPK-
1
H300A
(TM3) surprisingly produced a specific shift in
size under denaturing conditions, suggesting that the
modification of the fifth amino acid on the tail may
alter its secondary structure in denaturing polyacryl-
amide gels or might yield an undefined covalent adduct
(Fig. 3B, lane 4).
Y. Lin et al. Functional transcript expressed by DAPK-1 locus
FEBS Journal 275 (2008) 2574–2584 ª 2008 The Authors Journal compilation ª 2008 FEBS 2577
Fig. 2. Identification of a proteolytic cleavage within the C-terminal tail of s-DAPK-1 protein. (A) A schematic diagram of the Flag–Myc
vector with the s-DAPK-1 clone and its mutants created by site-directed mutagenesis. The vector encoding the s-DAPK-1Dtail with a 42
amino acid tail deletion is named Flag-TD. (B–D) Transfected s-DAPK-1 and its mutants identified a cleavage within its tail. HCT116
p53
+ ⁄ +
cells were transfected with the respective vectors, as indicated, for 24 h prior to harvesting. Expression of the
ectopically expressed s-DAPK-1 and its mutants was detected using an antibody to Flag (Sigma). (E) In vitro cleavage of purified
GST–s-DAPK-1. Recombinant GST–s-DAPK-1 was purified from Bl21 cells and incubated at 30 °C with increasing amounts of
HCT116 p53
+ ⁄ +
cell lysates (0, 1, 5, 10 and 20 lL) as indicated. The sample mixtures after in vitro cleavage were subjected to immu-
noblotting.
Functional transcript expressed by DAPK-1 locus Y. Lin et al.
2578 FEBS Journal 275 (2008) 2574–2584 ª 2008 The Authors Journal compilation ª 2008 FEBS
The proapoptotic effect in response to MEK ⁄ ERK

signaling and the membrane-blebbing-promoting effect
upon transfection are two well-characterized functions
of DAPK-1 [3,10,13]. Therefore, we set out to define the
role of s-DAPK-1 in these two pathways. We used, as
expressed constructs, full-length s-DAPK-1, s-DAPK-
1Dtail, and s-DAPK-1G296AR297A, which allowed us
to evaluate whether the tail contributes to the function
of the s-DAPK-1 protein. Unlike DAPK-1, s-DAPK-1
does not induce poly (ADP-ribose) polymerase (PARP)
cleavage in response to the MEK ⁄ ERK signal input
(Fig. 4A, lane 6 versus lane 5). However, transfection
of Flag–s-DAPK-1 was able to cause significant mem-
brane blebbing, to levels similar to those caused by full-
length DAPK-1, although the effect was weaker
(Fig. 4B). Given the biological activity of s-DAPK-1 in
the membrane-blebbing assay, we evaluated the activity
of the mutant with the tail deletion (Flag-TD; s-DAPK-
1Dtail) and the protease-resistant substitution (Flag-
TM1; s-DAPK-1G296AR297A). As compared to
full-length s-DAPK-1, the s-DAPK-1G296AR297A
showed a reduced membrane-blebbing effect (Fig. 4C),
whereas s-DAPK-1Dtail was almost as active as full-
length DAPK-1 (Fig. 4C). These data indicate that the
‘tail’ of s-DAPK-1 has a negative regulatory function
with regard to s-DAPK-1 activity, and that its removal
serves to enhance the membrane-blebbing effect of
s-DAPK-1.
To determine the mechanism that underlies the effect
of the tail on the membrane-blebbing-promoting ability
of s-DAPK-1, the localization and half-lives of the

full-length s-DAPK-1, s-DAPK-1Dtail and s-DAPK-
1G296AR297A were examined. As compared to
DAPK-1, s-DAPK-1 shows more specific localization
in the cytoplasm (Fig. 5A). s-DAPK-1Dtail predomi-
nantly localizes around the nucleus, and s-DAPK-
1G296AR297A spreads throughout the cytosol and
tends to form some ‘aggregating bodies’ (Fig. 5). More-
over, the half-life of s-DAPK-1Dtail is much longer than
those of s-DAPK-1 and s-DAPK-1G296AR297A
(Fig. 6A–D), suggesting that the increased membrane-
blebbing function of s-DAPK-1Dtail is due to its slower
Fig. 3. Identification of the critical sites for
proteolytic cleavage of the C-terminal tail of
s-DAPK-1. (A) A schematic diagram of the
tail mutants of s-DAPK-1 created by site-
directed mutagenesis. (B) Expression of the
tail mutants of s-DAPK-1. HCT116 p53
+ ⁄ +
cells were transfected with the respective
vectors, as indicated, for 24 h prior to har-
vesting. Expression of the s-DAPK-1 tail
mutants was detected by immunoblotting.
(C) Cleavage of the tail of s-DAPK-1 is not
inhibited by common protease inhibitors.
HCT116 p53
+ ⁄ +
cells were transfected with
the Flag–s-DAPK-1 vector for 24 h and trea-
ted with the indicated protease inhibitors
6 h prior to harvesting. The Flag–s-DAPK-1

protein was detected by immunoblotting.
Y. Lin et al. Functional transcript expressed by DAPK-1 locus
FEBS Journal 275 (2008) 2574–2584 ª 2008 The Authors Journal compilation ª 2008 FEBS 2579
degradation (Fig. 6B). Furthermore, upon chemical
subcellular fractionation based on differential protein
solubility, s-DAPK was found to localize in both the
‘insoluble’ cytoskeletal and soluble cytosolic fractions
(Fig. 5B), whereas the s-DAPK-1Dtail equilibrium was
shifted more into the soluble cytosolic fraction
(Fig. 5C).
Discussion
DAPK-1 is a stress-regulated kinase whose down-
stream functions are linked to a variety of diverse
signaling pathways, including ERK kinase activation,
autophagic signaling, and oncogene-mediated p53 tran-
scriptional responses. DAPK-1 is also regulated by
tumor necrosis factor signaling, p90 ribosomal S6
kinase (RSK), and leukocyte common antigen-related
phosphatase, which alter the specific activity of the
kinase as a prosurvival or proapoptotic factor.
Although the DAPK-1 protein is now known to be
regulated post-translationally, the gene is also subject
to methylation, which has the potential to reduce the
specific activity of DAPK-1 [6]. In this work, we have
identified another function of the DAPK-1 locus: it
can express a message whose product possesses part of
DAPK-1’s ankyrin-repeat region, P-loop, and cytoskel-
etal binding domain, and a unique tail of 42 amino
acids encoded by intron 20–21 of the DAPK-1 gene.
In our examination of DAPK-1 and s-DAPK-1 expres-

sion using real-time PCR, we found a significant corre-
lation in their expression, whether using cancer cell
lines or normal human tissues, suggesting that mRNA
from the locus is coordinately produced. Future work
will be required to understand the regulation of the
translation of these mRNAs and whether stress-regu-
lated signaling pathways regulate these two proteins
differently in cell growth control.
Despite the many functions attributed to DAPK-1,
the two standard cellular assays for defining its func-
tion include proapoptotic pathways and membrane
blebbing. Therefore, we have examined the ability of
the s-DAPK-1 protein to play a role in these two pro-
cesses. We found that although s-DAPK-1 cannot
induce apoptosis in response to the activated
MEK ⁄ ERK signal like DAPK-1, it can mimic DAPK-
1 and induce membrane blebbing. A function was also
attributed to the unique tail of s-DAPK-1: it can regu-
late the localization and half-life of the protein and
Fig. 4. The C-terminal tail of s-DAPK-1 negatively regulates its mem-
brane-blebbing function. (A) s-DAPK-1 does not induce apoptosis in
response to MEK ⁄ ERK signaling. HEK293 cells were transfected
with the respective vectors, as indicated, for 24 h prior to harvest-
ing. PARP and PARP cleavage were detected with a PARP-specific
antibody (Cell Signalling). (B) s-DAPK-1 induces membrane blebbing.
A375 cells were transfected with the respective vectors as indi-
cated, and evaluated for membrane blebbing in transfected cells as
described previously [10]. The top panel (B) shows the normal (1)
and the blebbing (2) morphology. (C) The C-terminal tail modulates
membrane blebbing by s-DAPK-1. A375 cells were transfected with

the respective vectors as indicated (wt, TM1, and TD; s-DAPK-1Dtail)
and evaluated for membrane blebbing in transfected cells as
described previously [10]. The top panel (B) shows the normal (1)
and blebbing (2) morphology. The bar graph in the bottom panels of
(B) and (C) summarizes the mean percentage of blebbing cells upon
each transfection. Each experiment was repeated four times.
Functional transcript expressed by DAPK-1 locus Y. Lin et al.
2580 FEBS Journal 275 (2008) 2574–2584 ª 2008 The Authors Journal compilation ª 2008 FEBS
Fig. 5. The C-terminal tail of s-DAPK-1 regulates its localization. (A) Confocal microscopy. A375 cells were transfected with the respective
vectors as indicated [GFP control, GFP-wt s-DAPK, GFP-TM1, GFP-TD (s-DAPK-1Dtail), and HA-DAPK-1]. The localizations of GFP and GFP-
tagged proteins were detected under the microscope. HA–DAPK-1 protein expression was detected using an antibody to HA tag. (B, C)
Subcellular protein fractionation. Chemical fraction of cell pellets after FLAG–s-DAPK transfection into a cytosolic (A) and cytoskeletal (B)
fractions for Flag–s-DAPK or Flag–s-DAPK-1Dtail (TD) as indicated in Experimental procedures.
Y. Lin et al. Functional transcript expressed by DAPK-1 locus
FEBS Journal 275 (2008) 2574–2584 ª 2008 The Authors Journal compilation ª 2008 FEBS 2581
can be subjected to proteolytic cleavage in cells.
Presumably, the tail has evolved to relocalize s-DAPK-
1 to mediate its rapid degradation, which would
greatly attenuate its membrane-blebbing function.
Signals that produce proteolytic cleavage would in turn
reduce its degradation (Fig. 6) and allow it to function
fully as a membrane-blebbing factor.
DAPK-1 has been shown to induce membrane bleb-
bing and promote the formation of actin stress fibers
and disassembly of focal adhesions [17]. These biologi-
cal events can occur in cooperation with microtubule-
associated protein 1B [12], for which the ability of
DAPK-1 to phosphorylate myosin light chain 2 [8] and
tropomyosin-1 [18] are considered to be important.
However, it was also shown that the ankyrin-repeat

region deletion mutant of DAPK-1 mislocalized to
focal contacts and lost its ability to induce morphologi-
cal changes [8], indicating a functional role of this
region in DAPK-1’s activity. This might explain the
membrane-blebbing-promoting effect of s-DAPK-1, as
it shares a portion of the ankyrin-repeat region of
DAPK-1. However, the functional significance of the
s-DAPK-1-induced membrane blebs is not clear, as
s-DAPK-1 cannot induce MEK ⁄ ERK-stimulated apop-
totic signals (Fig. 5A). A recent study has provided a
novel insight into membrane blebbing [19]; it was
shown that membrane blebbing is due to the reassembly
of the contractile cortex. Therefore, distinct from the
alternative models showing that membrane blebbing is
linked to autophagic or cell death pathways, membrane
blebbing may also be part of a normal cell division pro-
cesses such as cytokinesis. Considering that ankyrin B
plays an important role in the membrane-blebbing pro-
cess [19], DAPK-1 and s-DAPK-1 may be able to inter-
act with ankyrin B via their ankyrin repeats and thus
promote membrane blebbing. Although these data pro-
vide an explanation for the significance of the ankyrin-
repeat region of DAPK-1 in inducing morphological
changes, they do not necessarily indicate that DAPK-1-
or s-DAPK-1-induced membrane blebbing is part of a
normal cell division cycle. Considering that physiologi-
cal membrane blebbing is a transient process [19], it
also remains possible that DAPK-1 and s-DAPK-1
arrest the cells at the blebbing stage and thus halt the
cell division cycle. Therefore, the actual biological sig-

nificance of the s-DAPK-1- and DAPK-1-induced
membrane blebbing requires further investigation.
Experimental procedures
Cell culture and harvest, plasmids and
transfection, and treatment
HEK293 (human embryonic kidney cell line) and A375
(human melanoma) cells were cultured in DMEM medium,
and HCT116 (human colon carcinoma) cells were cultured
in McCoy medium. The medium was supplemented with
10% fetal bovine serum and a penicillin and streptomycin
mixture at 37 °C with 5% CO
2
in a humidified atmosphere.
In a typical experiment, 10
6
cells were seeded into a 10 cm
tissue culture plate and left for at least 24 h to attach to
the bottom of the container. Before harvesting, cells were
first washed twice with NaCl ⁄ P
i
and then scraped into
1 mL of NaCl ⁄ P
i
. The HA–DAPK-1 was a kind gift from
A. Kimchi (Weizmann Institute, Rehovot, Israel). s-DAPK-
1 was cloned into the Flag–Myc vector from Sigma (Poole,
UK), and the mutants were created using the Quickchange
site-directed mutagenesis kit from Stratagene. GST–s-
DAPK-1 was cloned into the pDEST-15 gateway GST vec-
tor from Invitrogen. The primers for cloning and mutations

are available upon request. Prior to transfection, Lipofecta-
mine ( 2 lLÆl g
)1
DNA) was added to Optimum medium
without fetal bovine serum. After a 5 min incubation, the
mixture was added to the DNA constructs, and after a
30 min incubation at room temperature, the whole solution
was added to the cells. The translation inhibitor cyclohexi-
mide from Supleco (Bellefonte, PA, USA) was used at a
concentration of 10 lgÆmL
)1
.
Protein analysis
Proteins were extracted by lysing the cells with lysis buffer
(1% NP40, 0.15 m NaCl, 50 mm Tris, pH 7.5, 1 mm
Fig. 6. The C-terminal tail of s-DAPK-1 regu-
lates its half-life. (A–D) The half-life of
s-DAPK-1 is regulated by its C-terminal tail.
HCT116 p53
+ ⁄ +
cells were transfected with
the respective vectors as indicated for 24 h,
in combination with cycloheximide treat-
ment at the indicated times, prior to har-
vesting. Expression of the Flag-tagged
proteins was evaluated by immunoblotting.
Functional transcript expressed by DAPK-1 locus Y. Lin et al.
2582 FEBS Journal 275 (2008) 2574–2584 ª 2008 The Authors Journal compilation ª 2008 FEBS
dithiothreitol and 1· protease inhibitor mixture), and the
protein concentrations were determined using the Bradford

reagent (Bio-Rad, Hercules, CA, USA). Immunoprecipita-
tion, protein separation by SDS ⁄ PAGE and detection by
immunoblotting were done as previously described [2].
The following antibodies were used: anti-HA (Covance,
Princeton, NJ, USA), anti-GST and anti-Flag (Sigma),
and anti-PARP (Cell Signal). The ProteoExtract Subcellu-
lar Proteome Extraction Kit (Calbiochem, La Jolla, CA,
USA) was used to extract proteins from mammalian cells
according to their cytosolic or cytoskeletal subcellular
localization. The kit was used in accordance with the
manufacturer’s recommendations. All extraction buffers
contained protease inhibitors, and all steps were carried
out at 4 °C unless stated. The cytosolic and cytoskeletal
fractions were stored at )70 °C and analyzed by immuno-
blotting.
RNA extraction, reverse transcription, PCR and
real-time PCR
mRNA was extracted from cells and tissue (obtained with
local ethical permission from the Masaryk Institute ethics
committee) using the Qiagen RNeasy Mini kit, following
the manufacturer’s suggested procedures. The optional step
of DNase treatment using the Qiagen RNase-free DNase
set was also included. After the extraction, RT-PCR was
performed using the Omniscript RT kit from Qiagen
(Valencia, CA, USA) and pfu polymerase from Stratagene,
following the manufacturer’s suggested protocols. Real-time
PCR was performed using the Qiagen QuantiTect SYBR
Green one-step PCR kit, following the manufacturer’s sug-
gested protocols. The actin primers were as follows: for-
ward, 5¢-CTACGTCGCCCTGGACTTCGAGC-3¢; reverse,

5¢-GATGGAGCCGCCGATCCACACGG-3¢. The DAPK-
1 primers were as follows: forward, 5¢-CGAGGTGA
TGGTGTATGGTG-3¢, reverse, 5¢-CTGTGCTTTGCTGG
TGGA-3¢. The s-DAPK-1 primers were as follows: for-
ward, 5¢-CGTCTCTCCAGCAGGTGTT-3¢; reverse, 5¢-TA
AGGCCACAGGGTCCAGTA-3¢.
Immunostaining and membrane-blebbing assay
A375 cells were analyzed by immunostaining and mem-
brane blebbing. Twenty-four hours post-transfection, cells
were fixed with 4% paraformaldehyde in NaCl ⁄ P
i
for
10 min, washed, and blocked with antibody dilution buf-
fer (3% BSA in NaCl ⁄ P
i
) for 1 h. For the non-green
fluorescent protein (GFP)-tagged proteins, the transfected
cells were then visualized using HA.11 antibody
(Covance) and antibody to Flag (Sigma). After incuba-
tion with the appropriate primary antibodies for 1 h,
cells were washed with NaCl ⁄ P
i
, stained with mouse
Alexa488-conjugated secondary antibody, and mounted
for observation by immunostaining or by examining
membrane-blebbing morphology using a Leica fluorescent
microscope. For immunostaining, the transfected cells
were incubated with Topro-3 from Invitrogen (1 : 1000 in
NaCl ⁄ P
i

) for 15 min at 37 °C before mounting. In mem-
brane-blebbing assays, 300 transfected cells were counted
upon each transfection, and each experiment was
repeated four times.
In vitro cleavage of bacterial purified protein
The GST–s-DAPK-1-transformed BL21 cells were induced
with arabinose for 3 h and lysed with 0.2% Triton in
NaCl ⁄ P
i
. The GST–s-DAPK-1 protein was then extracted
from the lysate using glutathione–Sepharose (GE Health-
care, Amersham, UK) and eluted with 50 mm glutathione.
For in vitro cleavage assays, 2 l L of the purified GST
fusion protein was incubated at 30 °C with various
amounts of lysate, as mentioned above, for 30 min. The
reaction was then stopped by adding SDS sample buffer,
and the mixture was subjected to immunoblotting.
Acknowledgements
B. Vojtesek and R. Hrstka are funded by
grants 301 ⁄ 05 ⁄ 0416 and 301 ⁄ 08 ⁄ 1468 from GACR and
grant LC06035. T. Hupp is funded by grants from
Cancer Research UK.
References
1 Bialik S & Kimchi A (2006) The death-associated pro-
tein kinases: structure, function, and beyond. Annu Rev
Biochem 75, 189–210.
2 Lin Y, Stevens C & Hupp T (2007) Identification of a
dominant negative functional domain on DAPK-1 that
degrades DAPK-1 protein and stimulates TNFR-1-med-
iated apoptosis. J Biol Chem 282, 16792–16802.

3 Chen CH, Wang WJ, Kuo JC, Tsai HC, Lin JR, Chang
ZF & Chen RH (2005) Bidirectional signals transduced
by DAPK–ERK interaction promote the apoptotic
effect of DAPK. EMBO J 24, 294–304.
4 Jin Y, Blue EK, Dixon S, Shao Z & Gallagher PJ
(2002) A death-associated protein kinase (DAPK)-inter-
acting protein, DIP-1, is an E3 ubiquitin ligase that
promotes tumor necrosis factor-induced apoptosis and
regulates the cellular levels of DAPK. J Biol Chem 277,
46980–46986.
5 Gozuacik D & Kimchi A (2006) DAPk protein family
and cancer. Autophagy 2, 74–79.
6 Raval A, Tanner SM, Byrd JC, Angerman EB, Perko
JD, Chen SS, Hackanson B, Grever MR, Lucas DM,
Matkovic JJ et al. (2007) Downregulation of death-
associated protein kinase 1 (DAPK1) in chronic
lymphocytic leukemia. Cell 129, 879–890.
Y. Lin et al. Functional transcript expressed by DAPK-1 locus
FEBS Journal 275 (2008) 2574–2584 ª 2008 The Authors Journal compilation ª 2008 FEBS 2583
7 Raveh T, Droguett G, Horwitz MS, DePinho RA &
Kimchi A (2001) DAP kinase activates a p19ARF ⁄
p53-mediated apoptotic checkpoint to suppress onco-
genic transformation. Nat Cell Biol 3, 1–7.
8 Bialik S, Bresnick AR & Kimchi A (2004) DAP-kinase-
mediated morphological changes are localization depen-
dent and involve myosin-II phosphorylation. Cell Death
Differ 11, 631–644.
9 Kuo JC, Wang WJ, Yao CC, Wu PR & Chen RH
(2006) The tumor suppressor DAPK inhibits cell motil-
ity by blocking the integrin-mediated polarity pathway.

J Cell Biol 172, 619–631.
10 Inbal B, Bialik S, Sabanay I, Shani G & Kimchi A
(2002) DAP kinase and DRP-1 mediate membrane
blebbing and the formation of autophagic vesicles
during programmed cell death. J Cell Biol 157,
455–468.
11 Torgerson RR & McNiven MA (1998) The actin–myo-
sin cytoskeleton mediates reversible agonist-induced
membrane blebbing. J Cell Sci 111, 2911–2922.
12 Harrison B, Krauss M, Burch L, Stevens C, Craig A,
Gordon-Weeks P & Hupp T (2008) DAPK-1 binding
to a linear peptide motif in MAP1B stimulates auto-
phagy and membrane blebbing. J Biol Chem (in
press).
13 Stevens C, Lin Y, Sanchez M, Amin E, Copson E,
White H, Durston V, Eccles DM & Hupp T (2007) A
germ line mutation in the death domain of DAPK-1
inactivates ERK-induced apoptosis. J Biol Chem 282,
13791–13803.
14 Llambi F, Lourenco FC, Gozuacik D, Guix C, Pays L,
Del Rio G, Kimchi A & Mehlen P (2005) The depen-
dence receptor UNC5H2 mediates apoptosis through
DAP-kinase. EMBO J 24, 1192–1201.
15 Wang WJ, Kuo JC, Ku W, Lee YR, Lin FC, Chang
YL, Lin YM, Chen CH, Huang YP, Chiang MJ et al.
(2007) The tumor suppressor DAPK is reciprocally reg-
ulated by tyrosine kinase Src and phosphatase LAR.
Mol Cell 27, 701–716.
16 Li J, Mahajan A & Tsai MD (2006) Ankyrin repeat: a
unique motif mediating protein–protein interactions.

Biochemistry 45, 15168–15178.
17 Kuo JC, Lin JR, Staddon JM, Hosoya H & Chen RH
(2003) Uncoordinated regulation of stress fibers and
focal adhesions by DAP kinase. J Cell Sci 116, 4777–
4790.
18 Houle F, Poirier A, Dumaresq J & Huot J (2007) DAP
kinase mediates the phosphorylation of tropomyosin-1
downstream of the ERK pathway, which regulates the
formation of stress fibers in response to oxidative stress.
J Cell Sci 120, 3666–3677.
19 Charras GT, Hu CK, Coughlin M & Mitchison TJ
(2006) Reassembly of contractile actin cortex in cell
blebs. J Cell Biol 175, 477–490.
Functional transcript expressed by DAPK-1 locus Y. Lin et al.
2584 FEBS Journal 275 (2008) 2574–2584 ª 2008 The Authors Journal compilation ª 2008 FEBS