Content uploaded by Hiroshi Shima
Author content
All content in this area was uploaded by Hiroshi Shima on Mar 28, 2015
Content may be subject to copyright.
Phosphorylation of Ser-446 Determines Stability of MKP-7*
Received for publication, January 6, 2005
Published, JBC Papers in Press, February 2, 2005, DOI 10.1074/jbc.M500200200
Chiaki Katagiri‡, Kouhei Masuda‡, Takeshi Urano§, Katsumi Yamashita¶, Yoshio Araki**,
Kunimi Kikuchi‡, and Hiroshi Shima‡储
From the ‡Division of Biochemical Oncology and Immunology, Institute for Genetic Medicine, Hokkaido University,
Kita-15, Nishi-7, Kita-ku, Sapporo 060-0815, Japan, the §Department of Biochemistry II, Nagoya University Graduate
School of Medicine, 65 Tsurumai, Showa-ku, Nagoya 466-0065, Japan, the ¶Division of Life Science, Graduate School of
Natural Science and Technology, Kanazawa University, Kakuma-machi, Kanazawa, 920-1192, Japan, and the **Division
of Bioscience, Graduate School of Environmental Earth Science, Hokkaido University, Kita-10, Nishi-5, Kita-ku,
Sapporo 060-0810, Japan
MAPK cascades can be negatively regulated by mem-
bers of the MAPK phosphatase (MKP) family. However,
how MKP activity is regulated is not well characterized.
MKP-7, a JNK-specific phosphatase, possesses a unique
COOH-terminal stretch (CTS) in addition to domains
conserved among MKP family members. The CTS con-
tains several motifs such as a nuclear localization sig-
nal, a nuclear export signal, PEST sequences, and a ser-
ine residue (Ser-446) that can be phosphorylated by
activated ERK, suggesting an important regulatory
role(s).
35
S-pulse labeling experiments indicate that the
half-life of MKP-7 is 1.5 h, a period significantly elon-
gated by deleting the CTS. We also show that overex-
pressed MKP-7 is polyubiquitinated when co-expressed
with ubiquitin and that proteasome inhibitors markedly
inhibit MKP-7 degradation. We also determined that
MKP-7 phosphorylated at Ser-446 has a longer half-life
than unphosphorylated form of the wild type protein, as
does a phospho-mimic mutant of MKP-7. These results
indicate that activation of the ERK pathway strongly
blocks JNK activation through stabilization of MKP-7
mediated by phosphorylation.
In all eukaryotic organisms mitogen-activated protein
(MAP)
1
kinase modules are involved in signal transduction of
numerous cellular responses including proliferation, differen-
tiation, and apoptosis (1, 2). Three subfamilies of MAP kinases
(MAPKs) have been well characterized: ERKs (extracellular
signal-regulated protein kinases), JNKs (c-Jun NH
2
-terminal
kinases), and the p38 MAPK kinases. It is well established that
ERK1/2 are typically stimulated by growth-related stimuli,
while JNK and p38 are primarily activated by stress-related
signals such as heat and osmotic shock, UV irradiation, and
inflammatory cytokines. MAPK pathways are regulated at
multiple levels to ensure the specificity, timing, and strength of
their activity. One critical aspect of this regulation is reversible
phosphorylation of MAPKs.
Negative regulation of MAPKs is achieved by dual dephos-
phorylation of the TXY motif by phosphatases. As in vivo can-
didates for negative regulators, the MAPK phosphatases
(MKPs), a family of dual specificity protein phosphatases, have
been identified (3). MKPs are primarily composed of two do-
mains, a rhodanese-like domain and a dual specificity phospha-
tase catalytic domain (4). In mammals 10 genes encoding
MKPs differing in substrate specificity and subcellular local-
ization have been reported. According to phylogenetic analysis
and gene structure, MKPs can be classified into four groups
(5–7). Group I contains the nuclear MKPs: MKP-1/DUSP1,
PAC1/DUSP2, MKP-2/DUSP4, and hVH-3/DUSP5, all of which
target the three primary MAPKs, ERK, JNK, and p38. Group II
includes cytoplasmic MKPs that mainly target ERK, namely,
MKP-3/DUSP6, PYST2/DUSP7, and MKP-4/DUSP9. Group III
contains MKP-5/DUSP10, which exhibits a unique NH
2
-termi-
nal domain in addition to the MKP common structure. MKP-5,
which is both nuclear and cytoplasmic, dephosphorylates JNK
and p38. Group IV consists of the nuclear and cytoplasmic
MKPs, hVH5/DUSP8, and MKP-7/DUSP16. These proteins ex-
hibit a unique COOH-terminal sequence of about 300 amino
acid residues in addition to the common MKP structure (8, 9).
We previously showed that MKP-7 shuttles between the
nucleus and the cytoplasm and suppresses activation of MAP
kinases in COS-7 cells in the order of selectivity JNK ⬎⬎ p38 ⬎
ERK. Using several mutant proteins, we found that a long
COOH-terminal stretch (CTS) contains a functional nuclear
export signal and a nuclear localization signal, both of which
enable MKP-7 to shuttle between the nucleus and the cyto-
plasm, and that the CTS determines JNK preference for
MKP-7 by masking MKP-7 activity toward p38 (8). Recently we
found that the CTS domain is bound by ERK and that Ser-446
in the CTS is phosphorylated by ERK depending on several
external stimuli (10). These data strongly suggest that the CTS
plays important regulatory role(s) in cells.
Although very few studies address their post-translational
regulation, it is known that some MKP/dual specificity phos-
phatases can be phosphorylated and/or polyubiquitinated.
MKP-1 is phosphorylated by ERK on two COOH-terminal ser-
ine residues, Ser-359 and Ser-364, which does not directly
affect phosphatase activity but results in stabilization of the
protein due to reduced degradation by the ubiquitin-directed
proteasome complex (11). VH-1-related phosphatase (VHR) has
also been reported to be phosphorylated on Tyr-138 by ZAP-70,
resulting in translocation of VHR to the immune synapse (12).
So far some MKPs have been reported to be phosphorylated.
* This work was supported in part by Grants-in-aid for Scientific
Research (B) provided by Japan Society for the Promotion on Science,
and a Grant-in aid for Scientific Research on Priority Area (A) provided
by the Ministry of Education, Culture, Sports, Science, and Technology
of Japan. The costs of publication of this article were defrayed in part by
the payment of page charges. This article must therefore be hereby
marked “advertisement” in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
储To whom correspondence should be addressed. Tel.: 81-11-706-5536;
Fax: 81-11-706-7541; E-mail: hshima@igm.hokudai.ac.jp.
1
The abbreviations used are: MAP, mitogen-activated protein;
MAPK, MAP kinase; ERK, extracellular signal-regulated kinase; JNK,
c-Jun NH
2
-terminal kinase; MEK, MAP kinase/ERK kinase; MKP,
MAP kinase phosphatase; HA, hemagglutinin; PMA, 12-O-tetradecano-
ylphorbol-13-acetate; CTS, COOH-terminal stretch, DAPI, 4⬘,6-dia-
midino-2-phenylindole; WT, wild type.
THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 280, No. 15, Issue of April 15, pp. 14716–14722, 2005
© 2005 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.
This paper is available on line at http://www.jbc.org14716
at TOHOKU UNIVERSITY on March 27, 2015http://www.jbc.org/Downloaded from
MKP-1 is phosphorylated by ERK1/2 in vivo as well as in vitro,
which inhibits protein degradation. hVH-5 is phosphorylated in
response to PMA treatment, but the physiological conse-
quences of such modification are not known (13). Xenopus
CL100 (XCL100), a homologue of human MKP-1, is phospho-
rylated by ERK in a cell cycle-dependent manner (14). In the
case of XCL100, serine residue(s) are phosphorylated during
the G
2
phase, and serine and threonine residues are phospho-
rylated during M phase. Experiments with proteasome inhibi-
tors demonstrate that not only MKP-1 but MKP-2 degradation
is mediated via the ubiquitin-proteasome pathway (15).
Recently we found that the CTS of MKP-7 is bound by ERK
and that Ser-446 in the CTS is phosphorylated by ERK in
response to several external stimuli (10). Since MKP-7 is a JNK
phosphatase, it is important to analyze physiological signifi-
cance of phosphorylation of MKP-7 by ERK. Such analysis
could explain how MKP-7 links ERK activity to stress kinase
activations. The CTS in MKP-7 contains two PEST sequences
(regions abundant in proline, glutamate, serine, and threonine
residues) (8), which are found in many rapidly degraded pro-
teins and have been suggested to signal proteolytic degradation
(16, 17). The presence of such sequences predicts that MKP-7
could be a relatively unstable protein (8). We and others (10,
18) previously showed that expression levels in cells of COOH-
terminally truncated MKP-7 and mouse MKP-M, a homologous
to the human MKP-7, are higher than that of the wild type
protein, supporting such an idea. In the present study, we
determine the stability of MKP-7 and identify sequences me-
diating protein degradation. We also analyze the effect of Ser-
446 phosphorylation on MKP-7 stability.
EXPERIMENTAL PROCEDURES
Plasmid Construction—Mammalian expression vectors, pFLAG-
MKP-7, pFLAG-MKP-7-(1–370), -(1–568), and -(1– 604), and -S446A
have been described previously (8, 10). Point mutations of pFLAG-
MKP-7S446A and S446D were generated by site-directed mutagenesis
using the QuikChange site-directed mutagenesis system (Stratagene).
pFLAG-MKP-7-(1– 435), -(1–511), -(162– 665), -(371– 665), -(371– 665)-
S446A, -371– 665)S446D, -(390 – 665), -(436 – 665), -436 – 665)S446A,
-⌬406 –540, -⌬P1 (deletion from 332–353), and -⌬P2 (deletion from
441– 462) were constructed by PCR. To construct pFLAG-MKP-2,
pFLAG-MKP-3, and pFLAG-MKP-5, the coding region of human
MKP-2, MKP-3, and MKP-5 were amplified by reverse transcription-
PCR and subcloned into the pFLAG-CMV2 vector (Sigma). pCI-neo-T7-
ubiquitin and pCI-neo-HA-ubiquitin were kind gifts of Dr. Yokosawa
(19).
Cell Culture and DNA Transfection—COS-7 and HeLa cells were main-
tained in Dulbecco’s modified Eagle’s medium containing 10% fetal bovine
serum at 37 °C under 5% CO
2
. Cells were transfected with various ex-
pression vectors using the FuGene-6 transfection reagent (Roche Diag-
nostics Inc.) according to the manufacturer’s recommendation.
35
S Pulse-Chase Analysis—At 32 h after DNA transfection, COS-7
cells were cultured in the presence of a Expre
35
S
35
S protein labeling
mix (PerkinElmer Life Sciences) for 1 h. The radioactive medium was
removed, and cells were chased with non-radioactive medium. At the
indicated times, cells were lysed with immunoprecipitation buffer con-
taining 50 mMTris-HCl (pH 7.5), 150 mMNaCl, 2 mMEDTA, 10%
glycerol, 1% Triton X-100, 1 mMphenylmethylsulfonyl fluoride, 10
g/ml leupeptin, and 10
g/ml aprotinin. Cell lysates were centrifuged
at 20,000 ⫻gfor 10 min, and the resulting supernatants were used as
cell extracts. Each sample was immunoprecipitated with 2
g of mouse
anti-FLAG M2 antibody (Sigma) and 15
l of protein G-Sepharose 4 fast
flow (Amersham Biosciences). Immunoprecipitated proteins were sep-
arated by SDS-PAGE on a 10% gel. The intensity of bands of FLAG-
MKP-7 and mutant proteins was quantified using an FLA-3000 imag-
ing system (Fujifilm, Tokyo, Japan).
Immunoblot Analysis of Cells Treated with Cycloheximide and Pro-
teasome Inhibitors—Eighteen hours after DNA transfection, COS-7
cells (2 ⫻10
5
/35-mm-diameter plate) were starved for 14 h and treated
with 10
g/ml cycloheximide for the indicated times. To determine the
effect of proteasome-mediated proteolysis, 20
MMG115 (Peptide In-
stitute, Osaka, Japan), 20
MMG132 (Peptide Institute), or Me
2
SO as
a vehicle was added 2 h before cycloheximide treatment. For immuno-
blot analysis, cells were lysed in MAPK lysis buffer as described previ-
ously (10). Cell lysates were subjected to immunoblot with anti-FLAG
M2 or anti-actin (Sigma) antibody.
Ubiquitination Assay—COS-7 cells (4 ⫻10
5
/60-mm-diameter plate)
were transfected with 2.4
g of pFLAG-MKP-7 together with 0.4
g
each of pCI-neo-T7-ubiquitin and pCI-neo-HA-ubiquitin. Eighteen
hours after transfection, cells were starved for 12 h and then treated
with 20
MMG132 for 6 h. The cells were lysed in MAPK buffer and
subjected to immunoprecipitation using 3
g of anti-FLAG M2 mono-
clonal antibody and 15
l of protein G-Sepharose 4 fast flow. Immuno-
precipitated proteins were separated on a 10% gel and subjected to
immunoblot using anti-T7 monoclonal antibody (Novagen) or anti-
FLAG polyclonal antibody (8).
Cell Staining—Immunohistochemical analyses were performed as
described (8). To detect FLAG-tagged proteins immunofluorescence was
performed using anti-FLAG rabbit antibody with AlexaFluor 488-con-
jugataed anti-rabbit IgG secondary antibody (Invitrogen) or AlexaFluor
546-conjugated anti-rabbit IgG secondary antibody (Invitrogen). To de-
tect phospho-Ser-446 of MKP-7 immunofluorescence was undertaken
using anti-phospho-Ser-446 antibody with AlexaFluor 546-conjugated
goat anti-mouse IgG secondary antibody (Invitrogen) or AlexaFluor
488-conjugataed anti-mouse IgG secondary antibody (Invitrogen). Nu-
clei were detected by staining with 1
g/ml 4 – 6-diamidino-2-phenyl-
indole (DAPI). Fluorescence was visualized under a fluorescence confo-
cal microscope (Olympus). Anti-phospho-Ser-446 monoclonal antibody
was developed against a phosphopeptide NKLCQFpSPVQEC as de-
scribed previously (20).
RESULTS
MKP-7 Is an Unstable Protein in COS-7 Cells—To deter-
mine whether MKP-7 is rapidly degraded, pulse-chase exper-
iments were performed. COS-7 cells transiently transfected
with pFLAG-MKP-7 were pulse-labeled with [
35
S]methionine
for 1 h and chased with non-radioactive medium up to 4 h.
Radiolabeled FLAG-MKPs were precipitated with an anti-
FLAG antibody at 0-, 1-, 2-, and 4-h chase time points and
subjected to autoradiography (Fig. 1, insets). The half-life of
MKP-7 was shown to be 1.5 h (Fig. 1a). Under the same
experimental conditions, half-lives of MKP-2, MKP-3, and
MKP-5, which are representative of groups I, II, and III,
respectively, were analyzed. The half-life of MKP-2, known to
be a short-lived protein due to proteasome-dependent degra-
dation (15), was shown to be 1.2 h (Fig. 1b). The half-lives of
MKP-3 and MKP-5 were both 4 h (Fig. 1cand d). These data
indicate that like MKP-2, but unlike MKP-3 and MKP-5,
MKP-7 is a highly unstable protein.
MKP-7 Is Degraded by a Ubiquitin-dependent Pathway—To
determine whether the rapid turnover of MKP-7 is due to
proteasomal activity, we analyzed the effect of two proteasome
inhibitors, MG115 and MG132, on degradation of MKP-7 (Fig.
2A). Immunoblot analysis showed that in the absence of inhib-
itors, levels of FLAG-MKP-7 were rapidly decreased in the
presence of cycloheximide, an inhibitor of de novo protein syn-
thesis. This reduction was completely inhibited by treatment of
COS-7 cells with MG115 or MG132. Under these conditions,
the amounts of actin were constant. These results suggest that
MKP-7 is degraded by the proteasome. An important compo-
nent of proteasome-mediated degradation is the proper target-
ing of the protein to be degraded by the ubiquitin conjugation
complex (21). This process results in the attachment of multiple
ubiquitin chains to the target protein. To determine whether
MKP-7 was ubiquitinated, we transiently overexpressed
FLAG-MKP-7 in COS-7 cells with or without T7- and HA-
ubiquitin. Proteins were similarly ubiquitinated in lysates
from cells with or without co-expressed FLAG-MKP-7 (Fig. 2B,
right). FLAG-MKP-7 was purified from the FLAG-MKP-7 con-
taining cell extracts using anti-FLAG-Sepharose (Fig. 2B,left).
A high molecular weight smear, characteristic of polyubiquiti-
nation, was observed in lysates from cells in which MKP-7 and
ubiquitin were co-expressed, demonstrating that MKP-7 is de-
graded by the ubiquitin/proteosomal pathway (Fig. 2B,left).
Ubiquitination of MKP-7 14717
at TOHOKU UNIVERSITY on March 27, 2015http://www.jbc.org/Downloaded from
Deletion of the COOH-terminal Region Affects Stability of
MKP-7 but Deletion of the PEST Sequences Does Not—To de-
termine sequences required for ubiquitin-proteasome depend-
ent degradation of MKP-7, we prepared constructs of two de-
letion mutants, MKP-7-(162– 665) and MKP-7-(1–370). Pulse-
chase experiments showed that the degradation rate of MKP-
7-(162– 665) was similar to that of the wild type protein, while
MKP-7-(1–370) was stable until 4 h (Fig. 3A), suggesting that
the stability of MKP-7 is determined primarily by amino acid
residues between 371 and 665. Previously we reported the
presence of two PEST sequences, PEST1 (332–353, PEST
score ⫽7.16) and PEST2 (441– 462, PEST score ⫽5.16). To
clarify whether the PEST sequences of MKP-7 are required for
its degradation, we prepared constructs encoding mutant pro-
teins, MKP-7⌬P1 and -⌬P2, which lack PEST1 and PEST2,
respectively. Interestingly, the half-lives of MKP-7⌬P1 and
-⌬P2 were similar to that of the wild type protein (Fig. 3B),
indicating that instability of MKP-7 is not conferred by the
PEST regions.
Identification of Sequences in the CTS Required for Degra-
dation of MKP-7—To determine which sequences in the CTS
play a role in stability, expression vectors encoding COOH-
terminal deletion mutant proteins, MKP-7-(1– 604), -(1–568),
-(1–511), -(1– 435), and -⌬406 –540 were constructed (Fig. 4A).
Using pulse-chase analysis, the half-lives of wild type and
mutant proteins were calculated (Fig. 4B). The half-life of
MKP-7-(1– 604), which was 1.7 h, was similar to that of the
wild type protein. However, the half-lives of MKP-7-(1–568)
and MKP-7-(1–511) were 5.0 and 4.8 h, respectively, indicating
that amino acid residues 569 – 604 are involved in protein deg-
radation. By contrast, degradation of MKP-7-(1– 435) was not
observed until 4 h. The observation that PEST2 is equivalent to
amino acid residues 441– 462 indicates that it is not involved in
degradation (Fig. 3B) and suggests rather that amino acid
residues 463–511 are involved in protein degradation. We des-
ignated residues 463–511 and 569 – 604 as regions 1 and 2,
respectively. The half-life of MKP-7⌬406 –540, which contains
region 2 but lacks region 1, was shown to be 3.5 h. These data
indicated that the CTS, in particular regions 1 and 2, are
involved in degradation of MKP-7.
Detection of Upward Mobility Shift of the CTS—To examine
the relationship between phosphorylation at Ser-446 of MKP-7
and its degradation through the CTS, we prepared expression
vectors of CTS fragments including Ser-446. COS-7 cells, trans-
fected with pFLAG-MKP-7-(371– 665), -(390 – 665), or -(436 –
665) (Fig. 5), were labeled with [
35
S]methionine and chased. We
found that
35
S-labeled MKP-7-(371– 665) and MKP-7-(390 –
665) were detected as doublets, while
35
S-labeled MKP-7-(436 –
665) was seen as a single band by autoradiography (Fig. 5,
insets). The half-lives of the lower bands of pFLAG-MKP-7-
(371– 665) and MKP-7-(390 – 665) were 1.2 and 1.5 h, respec-
tively, which are equivalent to that of the wild type MKP-7
FIG.1.Comparison of stability of MKP-7 and other MKPs in-
cluding MKP-2, MKP-3, and MKP-5. COS-7 cells (2 ⫻10
5
/35-mm-
diameter plate) were transfected with 1.2
g of pFLAG-MKP-7 (a),
pFLAG-MKP-2 (b), pFLAG-MKP-3 (c), or pFLAG-MKP-5 (d). After 32 h
in culture, cells were pulsed with [
35
S]methionine for 1 h and chased for
the indicated times. Cells were lysed, and samples were subjected to
immunoprecipitation with anti-FLAG antibody followed by SDS-PAGE.
The levels of [
35
S]methionine-labeled pFLAG-MKPs were monitored by
autoradiography as shown in the insets. The graph shows the relative
intensity of [
35
S]methionine labeled MKPs. Intensities relative to that
seen in cells without chase are presented. Data shown are the means
from three independent experiments.
FIG.2. MKP-7 is degraded by an ubiquitin-dependent path-
way. A, COS-7 cells (2 ⫻10
5
/35-mm-diameter plate) were transfected
with 1.2
g of pFLAG-MKP-7. After 30 h of culture, cells were cultured
without or with either 20
MMG115 or MG132 for 2 h. Cells were then
treated with 10
g/ml cycloheximide (CHX) for the indicated times, and
levels of FLAG-MKP-7 and actin were monitored by immunoblot (IB)
with anti-FLAG M2 antibody (upper panel) or anti-actin antibody
(lower panel). DMSO, dimethyl sulfoxide. B, COS-7 cells (4 ⫻10
5
/60-
mm-diameter plate) were transfected with 2.4
g of pFLAG-MKP-7
with or without 0.4
g each of pCI-neo-T7-ubiquitin and pCI-neo-HA-
ubiquitin. Eighteen hours after transfection, cells were starved for 12 h
and treated with 20
MMG132 for 6 h. An immunoprecipitation (IP)/
immunoblot (IB) of the cell lysate was done using anti-FLAG M2 anti-
body for the immunoprecipitation and either anti-T7 antibody (left,
upper panel) or anti-FLAG polyclonal antibody (left,lower panel) for the
immunoblot. Levels of ubiquitinated proteins and FLAG-MKP-7 in cell
extracts were monitored by immunoblot with anti-T7 antibody (right,
upper panel) or anti-FLAG polyclonal antibody (right,lower panel). Vec,
vector; Ub, ubiquitin.
Ubiquitination of MKP-714718
at TOHOKU UNIVERSITY on March 27, 2015http://www.jbc.org/Downloaded from
protein. In contrast, the half-lives of the upper bands of MKP-
7-(371– 665) and MKP-7-(390 – 665) and band of MKP-7-(436 –
665) were 2.9, 3.2, and 2.8 h, respectively. We demonstrated
previously that upon PMA stimulation, Ser-446 of FLAG-
MKP-7 is phosphorylated by co-expressed HA-ERK2, resulting
in an upward mobility shift of FLAG-MKP-7 seen on SDS-
PAGE (10). By analogy, it is possible that the upwardly shifted
bands of MKP-7-(371– 665) and MKP-7-(390 – 665) correspond
to phosphorylated forms.
Detection of Phosphorylation at Ser-446 in Cells—To detect
phosphorylation of Ser-446 in cells, we developed an antibody
raised against a peptide containing phosphoserine-446. Cells
transfected with FLAG-MKP-7 and HA-ERK2 were stained
with this antibody after treatment with or without PMA. No
protein was detected by the antibody in cells without PMA-
treatment (Fig. 6A,upper panels). FLAG-MKP-7 was phospho-
rylated in PMA-treated cells (Fig. 6A,lower panels), but MKP-
7S446A under the same conditions was not (data not shown).
These results indicate that the antibody is mono-specific for
phospho-Ser-446. Since this antibody did not detect any protein
in an immunoblot analysis (data not shown), we conclude that
the antibody is suitable only for immunohistological analyses.
To determine whether FLAG-MKP-7-(371– 665) and FLAG-
MKP-7-(390 – 665) are phosphorylated in cells even in the ab-
sence of HA-ERK2 and without PMA-treatment, COS-7 cells
were transfected with pFLAG-MKP-7-(371– 665), -(390 – 665),
-(436 – 665), or -(371– 665)S446A and stained with anti-phos-
pho-Ser-446 antibody (Fig. 6B) without any stimuli. As shown
in Fig. 6B, specific interaction with phospho-Ser-446 was con-
firmed since the antibody does not recognize MKP-7-(371–
665)446A (compare lanes m and n). As shown in Fig. 6B,
MKP-7-(371– 665) was phosphorylated in some cells indicated
by arrows but not in others (compare lanes a and b). Likewise,
MKP-7-(390 – 665) was phosphorylated in some cells as indi-
cated by arrows but not in others (compare lanes e and f). In
FIG.3. COOH-terminal residues but not PEST sequences af-
fect stability of MKP-7. COS-7 cells (2 ⫻10
5
/35-mm-diameter plate)
were transfected with 1.2
g of pFLAG-MKP-7WT, pFLAG-MKP-7-
(162– 665), or pFLAG-MKP-7-(1–370) (A) and pFLAG-MKP-7WT, -⌬P1,
or -⌬P2 (B). At 32 h after transfection, cells were pulsed with [
35
S]me-
thionine for 1 h and chased for the indicated times. The graph shows the
relative intensity of [
35
S]methionine-labeled MKP-7 mutant proteins.
The intensity in labeled cells without chase was defined as 100%. Data
shown are the means from three independent experiments. A t1⁄2, ob-
tained from the graph, is presented.
FIG.4. Stability of COOH-terminal truncated mutants of
MKP-7. A, schematic representation of MKP-7 mutant proteins. B,
COS-7 cells (2 ⫻10
5
/35-mm-diameter plate) were transfected with 1.2
g each of pFLAG-MKP-7WT, pFLAG-MKP-7-(1– 604), -(1–568), -(1–
511), -(1– 435), and -⌬406 –540. Thirty-two hours after transfection,
cells were pulsed with [
35
S]methionine for 1 h and chased for the
indicated times. The graph shows the relative intensity of [
35
S]methi-
onine-labeled MKP-7 mutant proteins. The intensity in labeled cells
without chase was defined as 100%. Data shown are the means from
three independent experiments. A t1⁄2, obtained from the graph,
is presented.
Ubiquitination of MKP-7 14719
at TOHOKU UNIVERSITY on March 27, 2015http://www.jbc.org/Downloaded from
contrast, MKP-7-(436 – 665) was phosphorylated in almost all
transfected cells (compare lanes i and j). These strongly sug-
gested that the upper bands of MKP-7-(371– 665) and MKP-7-
(390 – 665) and the single band of MKP-7-(436 – 665) seen in
Fig. 5 are phosphorylated forms. Since the half-lives of these
bands are longer than those of the lower bands of MKP-7-(371–
665) and MKP-7-(390 – 665), it is likely that phosphorylation at
Ser-446 stabilizes the CTS fragment. It is significant that ex-
pressed COOH-terminal fragments can be phosphorylated
even in the absence of HA-ERK2 without PMA stimuli. We
showed previously that HA-JNK1 can phosphorylate Ser-446
in a MKP-7CS mutant that lacks JNK phosphatase activity
(10). Phosphorylation of the COOH-terminal fragment may be
catalyzed not only by endogenous ERK but by endogenous JNK
in cells.
Phosphorylation on Ser-446 Stabilizes MKP-7—To confirm
that phosphorylation at Ser-446 of the CTS confers stability,
the half-lives of MKP-7-(371– 665)S446A and -(371– 665)-
S446D, which are nonphosphorylatable and phosphorylation-
mimicking mutants, respectively, were analyzed (Fig. 7, left
FIG.5. Stability of COOH-terminal
fragments of MKP-7. COS-7 cells (2 ⫻
10
5
/35-mm-diameter plate) were trans-
fected with 1.2
g of pFLAG-MKP-7-
(371– 665), -(390 – 665), or -(436 – 665).
Thirty-two hours after transfection, cells
were pulsed with [
35
S]methionine for 1 h
and chased for the indicated times. The
levels of [
35
S]methionine-labeled MKP-7
mutant proteins were monitored by auto-
radiography as shown in insets. The
graph shows the relative intensity of
[
35
S]methionine-labeled MKP-7 mutant
proteins. The intensity in labeled cells
without chase was defined as 100%. Data
shown are the means from three inde-
pendent experiments.
FIG.6. Detection of phospho-Ser-
446 in the COOH-terminal fragments.
A, 36 h after transfection with pFLAG-
MKP-7, HeLa cells were cultured without
or with 5 ng/ml PMA for 15 min. FLAG-
MKP-7 was detected by immunofluores-
cence using an anti-FLAG rabbit antibody
with AlexaFlour-546-conjugated goat an-
ti-rabbit secondary antibody (red). Phos-
pho-Ser-446 was detected by immunoflu-
orescence using an anti-phospho-Ser-446
mouse antibody with AlexaFluor-488 con-
jugated goat anti-mouse secondary anti-
body (green). Cell structure was examined
by differential interference contrast (DIC).
B, COS-7 cells transfected with MKP-7-
(371– 665) (panels a– d), -(390 – 665) (panels
e– h), -(436 – 665) (panels i–l), or -(371–
665)S446A (panels m–p) were cultured for
36 h. FLAG-MKP-7 was detected by immu-
nofluorescence using an anti-FLAG rabbit
antibody with AlexaFluor-488-conjugated
goat anti-rabbit secondary antibody (pan-
els a,e,i, and m). Phospho-Ser-446 was
detected by immunofluorescence using
an anti-phospho-Ser-446 mouse antibody
with AlexaFluor-546-conjugated goat an-
ti-mouse secondary antibody (panels b,f,
j, and n). Nuclei were stained with DAPI
(panels c,g,k, and o). Cells in which
FLAG-MKP-7 is phosphorylated are indi-
cated by arrows.
Ubiquitination of MKP-714720
at TOHOKU UNIVERSITY on March 27, 2015http://www.jbc.org/Downloaded from
panel). The half-lives of MKP-7-(371– 665)S446A and -S446D
were 1.3 and 3.2 h, respectively, which is equivalent to the
half-lives of the lower and upper bands of MKP-7-(371– 665)
(Fig. 5, left panel). Furthermore, the half-life of MKP-7-(436 –
665)S446A was shown to be 1.3 h (Fig. 7, right panel), which
was shorter than that of MKP-7-(436 – 665) (Fig. 5, right panel).
These data suggest that phosphorylation at Ser-446 stabilizes
the CTS fragment. We then asked whether this finding applies
to wild type MKP-7. The time course of degradation of MKP-
7S446D, a phosphorylation mimicking mutant, and MKP-
7S446A, a nonphosphorylatable mutant in cells with no stim-
uli, was analyzed (Fig. 8). The half-life of MKP-7S446D was
4.0 h, and that of MKP-7S446A and the wild type protein was
1.5 h, indicating that phosphorylation of Ser-446 stabilizes
MKP-7 protein. These data strongly suggest that phosphoryl-
ation of Ser-446 increases the half-life of MKP-7 by regulating
the stability of the COOH-terminal region.
DISCUSSION
In this study, we found that MKP-7 is a short-lived protein
that is degraded via ubiquitin-mediated proteolysis. Two re-
gions, amino acid residues 463–511 and 569 – 604, in the car-
boxyl terminus of MKP-7 were shown to target the protein for
degradation. We then asked whether protein degradation is
regulated by phosphorylation on Ser-446. Using several deletion
mutants, we found that upwardly shifted bands in SDS-PAGE
corresponding to phosphorylated forms have longer half-lives
than the non-shifted ones. Furthermore, a phosphorylation-
mimicking mutant of MKP-7 showed a longer half-life than a
phosphorylation-deficient one. These results strongly sug-
gested that phosphorylation of Ser-446 of MKP-7 blocks its
rapid degradation. MKP-7 suppresses MAPK activation in the
order of selectivity, JNK ⬎⬎ p38 ⬎ERK. We also determined
whether phosphorylation at Ser-446 affects substrate specific-
ity. Replacement of Ser-446 by aspartate or glutamate did not
mediate any difference in substrate specificity (data not
shown). Therefore, we propose that the physiological impor-
tance of phosphorylation at Ser-446 is as follows (Fig. 9). In
quiescent cells, MKP-7 is maintained at low levels due to rapid
turnover by the ubiquitin-mediated protein degradation path-
way. Upon activation of cells by growth factors, ERK is acti-
vated through the MAPKKK/MEK/ERK pathway, and acti-
vated ERK can phosphorylate Ser-446 of MKP-7. This
phosphorylation does not modify the substrate specificity of
MKP-7 but leads to stabilization of the protein. Accumulation-
phosphorylated MKP-7 strongly suppress JNK activation.
So far several motifs such as the PEST sequences (16, 17),
the cyclin destruction box (22, 23), the KEN destruction box
(24), degradation signals of a hydrophobic nature (25, 26),
phosphorylation-dependent degradation signals (27), and the
myc degron (28) are reported to be required for rapid proteol-
ysis by the ubiquitin-proteasome pathway. Since the identifi-
FIG.7.Effect of SA and SD mutations on stability of the COOH-
terminal fragments. COS-7 cells (2 ⫻10
5
/35-mm-diameter plate)
were transfected with 1.2
g of pFLAG-MKP-7-(371– 665)S446A, -(371–
665)S446D, or -(436 – 665)S446A. 32 h after transfection, cells were
pulsed with [
35
S]methionine for 1 h and chased for the indicated times.
The levels of [
35
S]methionine-labeled MKP-7 were monitored on auto-
radiography as shown in the insets. The graph shows the relative
intensity of [
35
S]methionine-labeled MKP-7. The intensity in labeled
cells without chase was defined as 100%. Data shown are the means
from three independent experiments.
FIG.8. A phosphomimicking mutant of MKP-7 has a longer
half-life than phosphodefficient MKP-7. COS-7 cells (2 ⫻10
5
/35-
mm-diameter plate) were transfected with 1.2
g of pFLAG-MKP-7WT,
-S446A, or -S446D. 32 h after transfection, cells were pulsed with
[
35
S]methionine for 1 h and chased for the indicated times. The levels of
[
35
S]methionine-labeled proteins were monitored by autoradiography.
The graph shows the relative intensity of [
35
S]methionine-labeled
MKP-7 mutant proteins. The intensity in labeled cells without chase
was defined as 100%. Data shown are the means from three independ-
ent experiments.
FIG.9.Regulation of MKP-7 by phosphorylation on Ser-446.
Ubiquitination of MKP-7 14721
at TOHOKU UNIVERSITY on March 27, 2015http://www.jbc.org/Downloaded from
cation of MKP-7, it has been speculated that MKP-7 is unsta-
ble, since it contains two PEST sequences (8). To determine
whether these motifs play a role in protein stability, we mu-
tated them (Fig. 3B); however, analyses of the mutants showed
that their disruption does not increase MKP-7 stability. In-
stead, regions 1 (463–511) and 2 (569 – 604) were shown to
mediate MKP-7 instability. Regions 1 and 2 may provide bind-
ing sites for components of the ubiquitin system, for example by
interacting with an E3 ubiquitin ligase. Since deletion of region
1 or 2 independently increases stability, it is possible that such
a ligase interacts with both regions 1 and 2. Phosphorylation of
Ser-446 may block an E3 ligase from interacting with these
regions. On the other hand, our site-directed mutagenesis ex-
periments indicate that the 12 lysine residues present in re-
gions 1 and 2 are not sites for ubiquitination (data not shown).
By doing a comparison of the amino acid sequences of regions 1
and 2 with other candidate motifs mentioned above, we found
aRXXL box, core sequences in the cyclin destruction box (29),
at positions 487– 490 in region 1. The RXXL box is a targeting
signal of the anaphase-promoting complex/cyclosome (APC/C),
a 1500-kDa complex comprised of many different subunits that
serves as the ubiquitin ligase (E3) (30). Identification of targets
recognized by regions 1 and 2 is now in progress.
Several reports demonstrated that MKP-1 and MKP-2,
which are nuclear MKPs, are short-lived due to ubiquitin-
mediated proteolysis (11, 15, 31, 32). Here, we show that
MKP-7 is also highly unstable, as is MKP-2 (Fig. 1). We re-
ported previously that MKP-7 is localized exclusively in the
cytoplasm, but this localization becomes nuclear following lep-
tomycin B treatment or replacement of leucine by alanine in
the nuclear export signal (8). To determine whether localiza-
tion of MKP-7 affects its degradation, we estimated the half-life
of the Leu-to-Ala mutant, which affects nuclear localization.
The half-life of the Leu-to-Ala mutant was similar to that of the
wild type (data not shown), indicating that the efficiency of
MKP-7 degradation is similar in the cytoplasm and in
the nucleus.
Previously we mapped MKP-7 to 12p12, an area prone to
deletions in acute lymphoblastic leukemia, acute and chronic
myeloid leukemia, and myelodysplastic syndrome (8). It has
been reported that JNK is constitutively activated in several
tumor cell lines and that the transforming action of some
oncogene is JNK-dependent (33, 34). Since MKP-7 was identi-
fied as a phosphatase specific for JNK, it could also function as
a tumor suppressor in cancers through negative regulation of
the JNK pathway. Whether the MKP-7 gene is deleted or
mutated in tumors from patients is crucially important. Re-
cently it was reported that among 22 leukemia patients ana-
lyzed, 17 were hemizygous for MKP-7/DUSP16, but no inacti-
vating mutations could be detected (35), and investigators
hypothesized that MKP-7/DUSP16 could be haploinsufficient
for tumor suppression. Another group also reported that ex-
pression of MKP-7/DUSP16 gene is significantly down-regu-
lated in both clinical tumors and cultured prostate cancer cell
lines (36). These data suggest that down-regulation of the
MKP-7/DUSP16 gene may be critical for initiation or progres-
sion of several tumors. Here we demonstrate that MKP-7 pro-
tein levels are regulated by a phosphorylation-dependent,
ubiquin-mediated protein degradation pathway. Whether this
pathway is impaired in certain cancers remains to be
determined.
Recent reports indicate that ERK1/2 activity functions in
MKP-1 degradation via the ubiquitin-proteasome pathway. Ac-
tivation of ERK1/2 induces phosphorylation and reduced deg-
radation of MKP-1 in CCL39 hamster fibroblasts (11) and in
Xenopus oocytes (14). These observations indicate a control
mechanism designed to limit undesirable long term activation
of ERK1/2. In contrast, ERK1/2 signaling can trigger degrada-
tion of MKP-1 via the ubiquitin-proteasome pathway in CL3
human lung cancer cells treated with Pb(II), a carcinogenic
metal (31). In this case, ERK1/2 activation is sustained by
stimulating MKP-1 degradation. Data presented here are the
first to report that the stability of an MKP other than MKP-1 is
regulated by phosphorylation. It is important to note that phos-
phorylation of MKP-7 by ERK does not affect the ERK pathway
positively or negatively but rather affects the JNK pathway.
Our results demonstrate that MKP-7 is involved in negative
cross-talk between the JNK and ERK pathways and strongly
suggest that sustained activation of ERK can result in attenu-
ation of JNK activation through phosphorylation of MKP-7.
Acknowledgements—We thank Dr. Yokosawa for pCI-neo-T7-ubiq-
uitin and pCI-neo-HA-ubiquitin. We also thank E. Yoshida for secre-
tarial assistance.
REFERENCES
1. Chang, L., and Karin, M. (2001) Nature 410, 37– 40
2. Wada, T., and Penninger, J. M. (2004) Oncogene 23, 2838 –2849
3. Camps, M., Nichols, A., and Arkinstall, S. (2000) FASEB J. 14, 6 –16
4. Keyse, S. M., and Ginsburg, M. (1993) Trends Biochem. Sci. 18, 377–378
5. Theodosiou, A., and Ashworth, A. (2002) Oncogene 21, 2387–2397
6. Farooq, A., and Zhou, M.-M. (2004) Cell. Signal. 16, 769 –779
7. Alonso, A., Rojas, A., Godzik, A., and Mustelin, T. (2004) Topics Curr. Genet. 5,
333–358
8. Masuda, K., Shima, H., Watanabe, M., and Kikuchi, K. (2001) J. Biol. Chem.
276, 39002–39011
9. Tanoue, T., Yamamoto, T., Maeda, R., and Nishida, E. (2001) J. Biol. Chem.
276, 26629 –26639
10. Masuda, K., Shima, H., Katagiri, C., and Kikuchi, K. (2003) J. Biol. Chem. 278,
32448 –32456
11. Brondello, J. M., Pouyssegur, J., and McKenzie, F. R. (1999) Science 286,
2514 –2517
12. Alonso, A., Rahmouni, S., Williams, S., van Stipdonk, M., Jaroszewski, L.,
Godzik, A., Abraham, R. T., Schoenberger, S. P., and Mustelin, T. (2003)
Nat. Immunol. 4, 44–48
13. Johnson, T. R., Biggs, J. R., Winbourn, S. E., and Kraft, A. S. (2000) J. Biol.
Chem. 275, 31755–31762
14. Sohaskey, M. L., and Ferrell, J. E. Jr. (2002) Mol. Biol. Cell 13, 454 – 468
15. Torres, C., Francis, M. K., Lorenzini, A., Tresini, M., and Cristofalo, V. J.
(2003) Exp. Cell Res. 290, 195–206
16. Rechsteiner, M. C., and Rogers, S. W. (1996) Trends Biochem. Sci. 21, 267–271
17. Rechsteiner, M. C. (2004) Adv. Exp. Med. Biol. 547, 49 –59
18. Matsuguchi, T., Musikacharoen, T., Johnson, T. R., Kraft, A. S., and Yoshikai,
Y. (2001) Mol. Cell. Biol. 21, 6999 –7009
19. Sasajima, H., Nakagawa, K., and Yokosawa, H. (2002) Eur. J. Biochem. 269,
3596 –3604
20. Yasui, Y., Urano, T., Kawajiri, A., Nagata, K., Tatsuka, M., Saya, H., Fu-
rukawa, K., Takahashi, T., Izawa, I., and Inagaki, M. (2004) J. Biol. Chem.
279, 12997–13003
21. Ciechanover, A. (1998) EMBO J. 17, 7151–7160
22. Glotzer, M., Murray, A. W., and Kirschner, M. W. (1991) Nature 349, 132–138
23. King, R. W., Glotzer, M., and Kirschner, M. W. (1996) Mol. Biol. Cell 7,
1343–1357
24. Pfleger, C. M., and Kirschner, M. W. (2000) Genes Dev. 14, 655– 665
25. Brodbeck, D., Hill, M. M., and Hemmings, B. A. (2001) J. Biol. Chem. 276,
29550 –29558
26. Gilon, T., Chomsky, O., and Kulka, R. G. (2000) Mol. Cell. Biol. 20, 7214 –7219
27. Laney, J. D., and Hochstrasser, M. (1999) Cell 97, 427– 430
28. Salghetti, S. E., Kim, S. Y., and Tansey, W. P. (1999) EMBO J. 18, 717–726
29. Zur, A., and Brandeis, M. (2002) EMBO J. 21, 4500 – 4510
30. Peters, J. M., King, R. W., Hoog, C., and Kirschner, M. W. (1996) Science 274,
1199 –1201
31. Lin, Y. W., Chuang, S. M., and Yang, J. L. (2003) J. Biol. Chem. 278,
21534 –21541
32. Kassel, O., Sancono, A., Kratzschmar, J., Kreft, B., Stassen, M., and Cato,
A. C. (2001) EMBO J. 20, 7108 –7116
33. Bost, F., McKay, R., Dean, N., and Mercola, D. (1997) J. Biol. Chem. 272,
33422–33429
34. Antonyak, M. A., Moscatello, D. K., and Wong, A. J. (1998) J. Biol. Chem. 273,
2817–2822
35. Hoornaert, I., Marynen, P., Goris, J., Sciot, R., and Baens, M. (2003) Oncogene
22, 7728 –7736
36. Kibel, A. S., Huagen, J., Guo, C., Isaacs, W. B., Yan, Y., Pienta, K. J., and
Goodfellow, P. J. (2004) Int. J. Cancer 109, 668 – 672
Ubiquitination of MKP-714722
at TOHOKU UNIVERSITY on March 27, 2015http://www.jbc.org/Downloaded from
Kunimi Kikuchi and Hiroshi Shima
Urano, Katsumi Yamashita, Yoshio Araki,
Chiaki Katagiri, Kouhei Masuda, Takeshi
Stability of MKP-7
Phosphorylation of Ser-446 Determines
Mechanisms of Signal Transduction:
doi: 10.1074/jbc.M500200200 originally published online February 2, 2005
2005, 280:14716-14722.J. Biol. Chem.
10.1074/jbc.M500200200Access the most updated version of this article at doi:
.JBC Affinity SitesFind articles, minireviews, Reflections and Classics on similar topics on the
Alerts:
When a correction for this article is posted• When this article is cited•
to choose from all of JBC's e-mail alertsClick here
http://www.jbc.org/content/280/15/14716.full.html#ref-list-1
This article cites 36 references, 21 of which can be accessed free at
at TOHOKU UNIVERSITY on March 27, 2015http://www.jbc.org/Downloaded from