Regulation Fe65 localization to the nucleus by SGK1 phosphorylation of its Ser566 residue.

Abstract

Fe65 is characterized as an adaptor precursor (APP) through its PID2 element, as well as with the other members of the APP protein family. With the serum- and glucocorticoid-induced kinase 1 (SGK1) substrate specificity information, we found that the putative site of phosphorylation in Fe65 by SGK1 is present on its Ser(566) residue in (560)CRVRFLSFLA(569)(X60469). Thus, we demonstrated that Fe65 and the fluorescein-labeled Fe65 peptide FITC-(560)CRVRFLSFLA(569) are phosphorylated in vitro by SGK1. Phosphorylation of the Ser(566) residue was also demonstrated using a Ser566 phospho-specific antibody. The phospho Fe65 was found mainly in the nucleus, while Fe65 S556A mutant was localized primarily to the cytoplasm. Therefore, these data suggest that SGK1 phosphorylates the Ser(566) residue of Fe65 and that this phosphorylation promotes the migration of Fe65 to the nucleus of the cell.

Full text

Page 1

41
http://bmbreports.org
BMB reports
*Corresponding Author. Tel: 82-43-261-3278; Fax: 82-43-271-0526;
E-mail: jin95324@cbu.ac.kr
Received 19 June 2007, Accepted 10 August 2007
Keywords: Confocal microscopy, Fe65, Localization, Protein phos-
phorylation, Protein-protein interaction, SGK1, Subcellular
Regulation Fe65 localization to the nucleus by SGK1
phosphorylation of its Ser566 residue
Eun Jeoung Lee1, Jaesun Chun2, Sunghee Hyun3, Hye Rim Ahn1, Jae Myung Jeong2, Soon-Kwang Hong4, Jin Tae Hong5, In
Kyeong Chang1, Hye Yeon Jeon1, Yeon Soo Han6, Chung-Kyoon Auh7, Jae In Park8 & Sang Sun Kang1,*
1School of Science Education, Chungbuk National University, Cheongju, 2Department of Biological Education, Korea National University of
Education, Cheongwon, 3Department of Premedicine, Eulji University School of Medicine, Daejeon, 4Division of Life Science, Myungji
University, Yongin, 5Department of Pharmacy, Chungbuk National University, Cheongju, 6Department of Agricultural Biology, Chonnam
National University, Gwangju, 7Division of Applied Biotechnology, Mokpo National University, Muan, 8Department of Forest Science,
Chungbuk National University, Cheongju, Korea
Fe65 is characterized as an adaptor precursor (APP) through its
PID2 element, as well as with the other members of the APP
protein family. With the serum- and glucocorticoid-induced
kinase 1 (SGK1) substrate specificity information, we found
that the putative site of phosphorylation in Fe65 by SGK1 is
present on its Ser566 residue in 560CRVRFLSFLA569(X60469).
Thus, we demonstrated that Fe65 and the fluorescein-labeled
Fe65 peptide FITC-560CRVRFLSFLA569 are phosphorylated in vi-
tro by SGK1. Phosphorylation of the Ser566 residue was also
demonstrated using a Ser566 phospho-specific antibody. The
phospho Fe65 was found mainly in the nucleus, while Fe65
S556A mutant was localized primarily to the cytoplasm.
Therefore, these data suggest that SGK1 phosphorylates the
Ser566 residue of Fe65 and that this phosphorylation promotes
the migration of Fe65 to the nucleus of the cell. [BMB reports
2008; 41(1): 41-47]
INTRODUCTION
Fe65 was originally identified as an expressed sequence tag
(EST) corresponding to an mRNA that is expressed in large
amounts in the rat brain. However, the cDNA of Fe65 encodes
a protein consisting of 711 amino acids (1). A comparison of
its amino acid sequence with sequences contained in data
banks reveals that a region within this protein can be aligned
with a region of the retroviral integrases, DNA binding pro-
teins that catalyze the integration of proviral DNA into the host
genome (2).
Serum- and glucocorticoid-induced kinase 1 (SGK1) is a
Ser/Thr protein kinase that is transcriptionally regulated by se-
rum and/or glucocorticoids in mammary epithelial cells and by
Rat-2 fibroblasts (3, 4). SGK1 is also transcriptionally regulated
by cortico-steroids in several types of cells (5, 6). SGK1 has
been identified as a protein that was designed to promote cell
survival, it has the same function as Akt kinase (7-9). We found
that SGK1 plays a role in the phosphorylation of MEKK3 and,
thus, inhibits its kinase activity (10). The consensus SGK1 sub-
strate site has been described by the structural formula
R-X-R-X-X-(S/T)-φ, where φ is any hydrophobic amino acid. The
arginine residues are conserved at positions -5 and -3 relative
to the positions of the Ser/Thr residues that are phosphorylated
in the presence of SGK1 (11, 12). Using SGK1 substrate con-
sensus sequence information, we noticed that the putative site
of SGK1 phosphorylation site in Fe65 is Ser566 residue of
560CRVRFLSFLA569 motif which is conserved in all Fe65 var-
iants (1, 2). (We count the amino acid number, according to
Fe65 variant X60469; NCBI Accession Number which was
used in this study). We investigated this further to determine
whether the phosphorylation of Ser566 in Fe65 would affect the
intrinsic kinase activity of Fe65 in vivo and in vitro, and found
that SGK1 actually does induce phosphorylation of Ser566 in
Fe65. It was shown that Fe65 fusion protein and fluorescein-la-
beled Fe65 peptides (FITC 560CRVRFLSFLA569) are phosphory-
lated by SGK1 in vitro. Further, the phosphorylation of Ser566
by SGK1 was demonstrated using a phospho-specific antibody.
We also observed that a mutant Fe65 containing Ala instead of
Ser at amino-acid 566 (Fe65S566A) is localized mainly in the cy-
toplasm, while the phosphorylated Fe65 is found mainly in the
nucleus, with the phospho-specific antibody. These observa-
tions suggest that SGK1 phosphorylates the Ser566 of Fe65 and
that this phosphorylation promotes the localization of Fe65 to
the nucleus of the cell.
RESULTS AND DISCUSSION
Phosphorylation of the recombinant Fe65 proteins and the
Fe65 peptides by SGK1
We observed the putative SGK1 phosphorylation site in Fe65

Page 2

Regulation of Fe65 translocation by phosphorylation
Eun Jeoung Lee, et al.
42
BMB reports
http://bmbreports.org
Fig. 2. The phosphorylation of Fe65 by SGK1. (A) 100 ng of the
purified Fe65 PTB2 fragment (39 kDa) or Fe65 PTB2 Ser566A was
incubated with 10 ng of active SGK1 (Upstate Biotech) in the
presence of 1 mCi γ-[32P]-ATP. Phosphorylation of the Fe65 frag-
ment and SGK1 appears to have taken place. Alkaline phosphatase
(AP) (+) was used to confirm that the phosphorylation took place.
(B) Phosphorylation of the FITC-Fe65 peptide (559ecRvRflSFl568) by
SGK1. To determine whether SGK1 phosphorylates the 566Ser moi-
ety of Fe65, SGK was incubated with the FITC-Fe65 peptide
(559ecRvRflSFl568). The phosphorylated product migrated to the
cathode (+) faster than the nonphosphorylated substrate did. The
increase in FITC-Fe65 peptide phosphorylation is shown to occur
along with increase in SGK1 concentration. The phosphorylated
product and nonphosphorylated substrate are marked.
Fig. 3. The immunoblotting with the phosphorylated 566Ser Fe65-specific antibody. COS-1 cells were transiently transfected
with expression vectors containing cDNA for myc epitope-tagged Fe65 wt or myc epitope-tagged Fe65S566A mutant. The
untransfected COS-1 represents the negative control of each panel. After immunoprecipitation with the myc antibody
(mouse), a western blot analysis was performed using rabbit Fe65 Antibody (A) or the phospho 566Ser Fe65 antibody (B).
The phospho 566Ser Fe65 antibody only recognized the wild type Fe65, not the Fe65S566A mutant.
Fig. 1. Functional domains of Fe65 in the rat brain. The WW do-
main is located near the N-terminal of Fe65 and is followed by
2 phosphotyrosine-binding (PTB) domains--PTB1 and PTB2 (which
is responsible for amyloid precursor protein cytoplasmic tail bind-
ing)-- and the putative nuclear localization sequence (NLS) in
Fe65. The SGK1 phosphorylation site, 566Ser (561RvRflSF567), lies in
the PTB2 domain, which is also responsible for the bond be-
tween Fe65 and APP. NLS and the SGK1 phosphorylation site are
indicated. The SGK1 phosphorylation site,
seems to overlap with the SGK1 phosphorylation site (see text for
details). In this study, we used a Fe65 variant (X60469; NCBI
Accession Number).
566Ser (561RvRflSF567),
(Fig. 1), Ser566 in the 561RvRflSF567 of the Fe65 PTB2 domain,
which is also responsible for the bond between Fe65 and APP
(1, 13-16). We also notice that the putative nuclear export se-
quences (NES; characterized as L/M/FX(2-3), L/I/VX(2-3), and
LXL/I) seem to overlap with the SGK1 phosphorylation site
(557LgecRVRfLSFLaV570; underline indicates the SGK1 phos-
phorylation site, is the Ser566 residue) (17-21).
To determine whether SGK1 induces phosphorylation of
Fe65, SGK1 was incubated in the presence of [γ-32P] ATP with
a Fe65 fragment (370-666) and Fe65 mutant fragment (370-
666, S566A) purified from recombinant E coli. The recombi-
nant wild-type Fe65 protein (352-666) was phosphorylated by
SGK1, whereas the mutant Fe65 (370-666 S566A), in which
Ser566 was replaced with Ala, was not phosphorylated by
SGK1 (Fig. 2A). Phosphorylation was verified by treating the
reaction mixture with alkaline phosphatase and observing that
the phosphorylated protein bands disappeared (Fig. 2A, mid-
dle lane). These results suggest that SGK1 induces phosphor-
ylation of the Ser566 residue of Fe65.
To further determine whether the Ser566 residue of Fe65 is
phosphorylated by SGK1, the fluorescein-labeled Fe65 peptide
FITC-560CRVRFLSFLA569 was incubated with SGK1. The pro-
portion of the Fe65 peptide that was phosphorylated increased
with the duration of the incubation or the amount of SGK1
used (Fig. 2B). We did not observe any phosphorylation of the
FITC-560CRVRFLAFLA569 peptide (in which Ser was replaced
with Ala), with or without SGK1 (data not shown).

Page 3

Regulation of Fe65 translocation by phosphorylation
Eun Jeoung Lee, et al.
43
http://bmbreports.org
BMB reports
The phospho 566Ser Fe65-specific antibody
We developed an antibody that recognizes the phosphorylated
566Ser Fe65 with a
Materials and Methods section). A western blot analysis was
performed. COS-1 cells were transiently transfected with ex-
pression vectors containing the cDNA for myc epitope-tagged
Fe65 or a myc epitope-tagged Fe65S566A mutant. The non-
transfected COS-1 cell (Fig. 3A) was used as a negative control
in each panel. After immunoprecipitation with the myc anti-
body (mouse), a western blot analysis was performed with rab-
bit Fe65 antibody (Fig. 3A) or with the phospho 566Ser Fe65
specific antibody (Fig. 3B). These results showed that the phos-
pho 566Ser Fe65 specific antibody only recognized the wild-
type Fe65 which is phosphorylated by SGK1, not the Fe65S566A
mutant.
560CRVRFLSFLA569 peptide (see the
Comparison of subcellular localization of wild-type Fe65 and
Fe65S566A mutant
To determine the significance of phosphorylation of Fe65
Ser566, we compared the subcellular localization of wild-type
Fe65 with that of the Fe65S566A mutant in COS-1 cells. The ex-
pression vectors containing cDNA for myc epitope-tagged
Fe65 (Fig. 4A) or myc epitope-tagged Fe65S566A mutant (Fig.
4B) were transiently transfected in COS-1 cells. Myc-Fe65 was
visualized by immunofluorescence in both fixed and per-
meabilized cells using monoclonal antibodies against myc and
Alexa Fluor 568-conjugated donkey anti-mouse immunoglob-
ulin G (red color). To define the cell boundary and cytoplasm,
we also used an anti-caveolin-1 rabbit antibody and goat an-
ti-mouse Alexa Flour 488 (green). The wild-type Fe65 was de-
tected both the nucleus and in the cytoplasm (Fig. 4A), while
the Fe65S566A mutant was observed in the cytoplasm only (Fig.
4B). These results suggest that the Fe65S566A mutant is pro-
hibited its nuclear localization.
Confocal microscopy with the phospho-specific antibody
against the phosphorylated 566Ser Fe65
To confirm the observations summarized in Fig. 4 (that the
phosphorylated wild-type Fe65 is transported into the nucleus
and the Fe65S566A mutant is not in the nucleus), we monitored
the phospho-specific antibody against the phosphorylated
566Ser Fe65 in COS-1 cells, which were transfected with myc-
FE65 or myc-Fe65S566A mutant, using confocal microscopy.
The confocal microscopic images show that the phosphory-
lated 566Ser Fe65, which was detected with the phospho-spe-
cific antibody (green), localizes in the nucleus (Figs. 4C and
4D). By contrast, the localization of Fe65S566A was limited in
the cytoplasm (Fig. 4D), while that of the Fe65 wild-type was
not (Fig. 4C). The confocal fluorescence micrographs in Figure
4 show the endogenous phosphorylated 566Ser Fe65 (green),
the transfected myc epitope-tagged wild-type Fe65 (red), and
the merge (yellow). Phosphorylated 566Ser Fe65 was detected
in the nucleus (green), while myc epitope-tagged Fe65S566A
was observed in the cytoplasm (red), consistent with the find-
ings represented in Fig. 4. The merger (yellow) was detected in
the nucleus when DNA for myc epitope-tagged wild-type Fe65
was transfected (Fig. 4C). However, the merger (yellow) was
not detected in the nucleus when DNA for myc epitope-tagged
Fe65S566A was transfected (Fig. 4D). These results suggest that
the nuclear Fe65 is mainly the phosphorylated 566Ser residue
and that phosphorylation of Fe65 566Ser residue is required for
translocation of Fe65 into the nucleus.
To test whether the translocation of Fe65 into the nucleus
by phosphorylation of Fe65 566Ser residue depends on the cell
line, we used HEK293 cell line instead of COS-1. With the
phospho-specific antibody against the phosphorylated 566Ser
Fe65, the confocal microscopy was performed in HEK293
cells which were transfected with myc-FE65 or myc-Fe65S566A
mutant. And the similar Fe65 subcellular localization results
were observed (Fig. 4E and F). The endogenous phosphory-
lated 566Ser Fe65 (green), the transfected myc epitope-tagged
wild-type Fe65 or Fe65S566A (red), and the merge (yellow) were
detected with their specific antibody. Phosphorylated 566Ser
Fe65 was observed in the nucleus (green), however myc epit-
ope-tagged Fe65S566A was shown in the cytoplasm (red), similar
with the findings represented in Fig. 4C and D. While the
merger (yellow) was shown in the nucleus when DNA for myc
epitope-tagged wild-type Fe65 was transfected (Fig. 4E), the
merger (yellow) was not detected in the nucleus when DNA
for myc epitope-tagged Fe65S566A was transfected (Fig. 4F) in
HEK293 cells. Together, these results suggest that phosphor-
ylation of Fe65 566Ser residue is required for translocation of
Fe65 into the nucleus, regardless of the transfected cell line
specificity.
Further phosphorylation of the Fe65 Ser566 residue by SGK1
seems to be required for its movement into the nucleus (Fig.
4).
The identification and characterization of SGK1 substrates
may provide clues to our understanding of how SGK1 contrib-
utes to cell survival and death (3, 4, 8, 10, 11, 18). We demon-
strated that Fe65 is a SGK1 substrate that contains the con-
sensus motif RxRxxS/Tφ: where φ is a hydrophobic amino acid,
which subsequently demonstrated that phosphorylation of
Fe65 by SGK1 controls its subcellular localization (Figs. 2 and
3). Because of the similarity in substrates for Akt kinase and
SGK1, we suspect that Akt kinase may also phosphorylate the
566Ser moiety of Fe65 and contribute to the control of Fe65
subcellular localization. However, it remains to be determined
whether Akt kinase also phosphorylates the Ser566 moiety of
Fe65.
It is unknown whether phosphorylation of the 566Ser moiety
of Fe56 influences its ability to bind to APP. Thus, we specu-
lated that its phosphorylation by SGK1 overtly influences
APP/Fe65 binding, because the 566Ser residue that is phos-
phorylated by SGK1 resides within the second PTB domain.
However, it remains to be seen whether the precise mecha-
nisms by which phosphorylation of Fe65 by SGK1 stimulate its

Page 4

Regulation of Fe65 translocation by phosphorylation
Eun Jeoung Lee, et al.
44
BMB reports
http://bmbreports.org
C: myc-Fe65 wt
D: myc-Fe65S566A
Anti-phospho-Ser566Fe65 Ab Anti-Myc AbMerge
Anti-phospho-Ser566Fe65 Ab
Anti-Myc Ab
Merge
Fig. 4. The subcellular localization of
wild-type Fe65 and Fe65S566A mutant.
COS-1 cells were transiently transfected
with expression vectors containing cDNA
for myc epitope-tagged Fe65 (A) or myc
epitope-tagged Fe65S566A mutant (B). Myc-
Fe65 was visualized by immunofluore-
scence in fixed and permeabilized cells
using monoclonal antibodies against myc
and Alexa Fluor 568-conjugated donkey
anti-mouse immunoglobulin G. (red). To
define the cell boundary and cytoplasm,
we also used an anti-caveolin-1 rabbit
antibody (green). The wild-type Fe65
was detected both the nucleus and cyto-
plasm (A), while the Fe65S566A mutant
was observed in the cytoplasm only (B).
COS-1 cells were transiently transfected
with expression vectors containing the
cDNA for myc epitope-tagged wild-type
Fe65 (C) and for the cDNA for myc epit-
ope-tagged S-A mutant Fe65 (D). HEK293
cells were transiently transfected with
expression vectors containing the cDNA
for myc epitope-tagged wild-type Fe65
(E) and for the cDNA for myc epit-
ope-tagged S-A mutant Fe65 (F). Confocal
fluorescence micrographs showing the
endogenous phosphorylated Fe65 in
COS-1 cells (left lane of C, D, E, or F),
the transfected myc epitope tagged
wild-type Fe65 or Fe65S566A mutant
(middle lane of C, D, E, or F), and the
merge (right lane of C, D, E, or F). The
phosphorylated 566Ser in Fe65 was de-
tected in the nuclear region (green),
while the myc epitope-tagged Fe65S566A
was observed in the cytoplasm (red).
The merger (yellow) was detected in the
nuclear region when DNA for myc epit-
ope-tagged wild-type Fe65 was trans-
fected (C and E), but not when the
Fe65S566A mutant DNA was transfected
(C and F). The transfected DNA and the
primary antibody are indicated above
the picture.
ability to migrate to the nucleus of the cell.
In conclusion, our results suggest that SGK1 phosphorylates
the Ser566 residue of Fe65 and that this phosphorylation is re-
quired for the nuclear localization of Fe65

Page 5

Regulation of Fe65 translocation by phosphorylation
Eun Jeoung Lee, et al.
45
http://bmbreports.org
BMB reports
E: myc-Fe65 wt
F: myc-Fe65S566A
Anti-phospho-Ser566Fe65 Ab
Anti-Myc AbMerge
Anti-phospho-Ser566Fe65 Ab
Anti-Myc Ab Merge
Fig. 4. Continued.
MATERIALS AND METHODS
Cell culture
COS-1 was purchased from ATCC. Media and supplements
were obtained from GIBCO-BRL. The cell line was maintained
in Dulbecco's Modified Essential Medium (DMEM) containing
10% fetal bovine serum (FBS), 10,000 U potassium penicillin/
streptomycin/ml, 2 mM glutamine, and 20 mM sodium
bicarbonate. The cells were incubated at 5% CO2 with 95%
humidity and in a 37oC chamber. The growth medium was
changed every 3 days.
Antibodies
The monoclonal antibody against the myc epitope was pur-
chased from Cell Signaling Technology. Antibodies against
caveolin-1 and Fe65 were purchased from Santa Cruz Biotech-
nology. An active SGK1 was purchased from Upstate Biotech-
nology.
DNA constructs and site-directed mutagenesis
Fe65 (X60469; NCBI Accession Number, cloned in pRK5 Myc
vector) was obtained from Dr. Margolis as a gift. To generate
the mutant Fe65S566A construct, 2 mutagenic primers (up: 5'-
GTGCGGTTCCTCGCCTTCCTGGCTG-3'; and down: 5'-AGC-
CAGGAAGGCG AGGAACCGCAC-3') and the QuikChange
XL Mutagenesis Kit (Stratagene, TX, USA) were used according
to the manufacturer's instructions. To express Fe65 fragment in
E. coli Fe65 and mutant Fe65S566A were subcloned into pro-
karyotic expression vector pGEX-5X-1 with forward primer;
5'-CTACATGATCCCGATGTCTGGA-3' and reverse primer;
5'-CTCTGGGGTCT GGGATCCTAG-3'). Each DNA construct
was confirmed by DNA sequencing.
Transfection and immunoprecipita-tion
Wild-type and mutant myc-Fe65S566A constructs were trans-
fected into the COS-1 or HEK293 cell using FuGENE6 (Roche
Molecular Biochemicals) according to the manufacturer's
instructions. Cells (2 ×107) were lysed in 1 ml RIPA lysis buf-
fer (50 mM Tris-HCl, 150 mM NaCl, 1% NP-40, and 0.5% so-
dium deoxycholate) for 10 minutes at 4oC. Cell lysates were
then centrifuged at 12,000 g for 15 minutes at 4oC, and the su-
pernatant was precleaned using protein A agarose beads
(Upstate Biochem). Anti-myc or anti-Fe65 antibodies were
used to precipitate Fe65. The antibodies were incubated with
the precleaned supernatant for 16 hours at 4oC and then pre-
cipitated using protein A agarose beads. Glutathione-Sepharose
beads were used to purify the antibodies. The beads were
washed 3 times with 1 ml of RIPA lysis buffer and then twice
with phosphate-buffered saline (PBS). The final pellet was used
for Fe65, western blot assay, as described.
Phosphorylation of recombinant Fe65
GST-Fe65 (a peptide consisting of amino acids 370 through
666 of the Fe65 protein) or GST-Fe65S566A 370-666 proteins
was expressed in Escherichia coli BL21 and purified using GST
agarose beads according to the manufacturer's instructions.
Purified proteins were used for the SGK1 kinase assay. Purified
and preheated GST-Fe65 and GST-Fe65S566A 370-666 proteins